Effect of genetic factors in the development of childhood lymphocytic leukemia (ALL)

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content EFFECT OF GENETIC FACTORS IN THE DEVELOPMENT OF CHILDHOOD LYMPHOCYTIC LEUKEMIA (ALL) by Wen Liu-Mares A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA in Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (EPIDEMIOLOGY) May 2001 Copyright 2001 Wen Liu-Mares Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UM I Number: 3027741 Copyright 2001 by Liu-Mares, Wen All rights reserved. ___ ® UMI UMI Microform 3027741 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90007 This dissertation, written by Mn..MH.7.Mares ........................................... under the direction of h.§z...... Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Dean of Graduate Studies DISSERTATION COMMITTEE Chairperson .J £ ^ L b J L & .............. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. To my parents, my husband and my children. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgements I would first like to thank my academic advisers and committee members- Dr s. Stan Azen, Duncan Thomas, Mark Krailo, Smita Bhatia, Juergen Reichardt, and Jonathan Buckley - for their guidance throughout my doctoral training. Drs. Stan Azen and Duncan Thomas provided invaluable mentorship on both the science and art of academic life. My many year professors Drs. Harland Sather, Malcolm Pike provided critical encouragement and helpful comments. Special thanks and gratitude go to Drs. Richard Gatti, Robert Haile, Xiao Ou Shu, Julie Ross and Leslie Robison for sharing with me their invaluable expertise knowledge in their respective fields. I thank my friends and classmates Marleen Voortmans and Bill Howells for their assistance with data collection at the early stage of the study. Finally, I would like to thank my parents, my parents-in-law, my husband, and children for their unwavering support and encouragement all the time. 111 with permission of the copyright owner. Further reproduction prohibited without permission. T able o f Contents Dedication.............................................................................................. ii Acknowledgements.............. iii List of Tables............................................................................................................... vi Abstract......................................................................................................................viii Chapter 1: Overview............. 1 Chapter 2: Childhood Acute Lymphocytic Leukemia - An Epidemiologic Review................................ 4 Introduction.................................. 4 Genetic Factors for Acute Lymphocytic Leukemia (ALL)........................ ,.......... 10 International Variation of Childhood Leukemia................................................... 10 Leukemia in Syndromic Disease........................................................................... 11 Familial Leukemia not Associated with a Recognized Syndrome.......................12 Leukemia in twins......................................... 34 Environmental Factors for Childhood Acute Lymphocytic Leukemia (ALL) 43 Ionizing Radiation ................................................................................ 43 Other Environmental Factors..................................................................................53 Ataxia Telangiectasia (A-T).................................. 60 ATM Gene and Leukemia..........................................................................................64 Conclusion........................................................................................ 76 Chapter 3: Incidence of Cancer Among Family Members of Childhood Acute Lymphocytic Leukemia with a Family History of Cancer................... 78 Introduction............................... 78 Background.................................................. 79 Methods ........................................................ 80 Study Design ....................... 80 Identification of Study Population......................................................................... 81 Selection Criteria............................................................... .81 Data Collection............................................................... 83 Database Establishment and Management ......... 84 Statistical Analysis...................................................................................................84 Results... .................................................... 87 Characteristics of the Study Population.................................................................87 Excluded Cases........................................................................................................90 Included Cases......................................................................................................... 91 Family History of Cancer............................................ 91 Overview of Results............................................................................ 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Discussion.................................................................................................................108 Limitations and Potential Biases of the Study....................................................108 Conclusion................................................................................................................ 139 Chapter 4: ATM Gene Mutation and Childhood Acute Lymphocytic Leukemia Study - A Grant Proposal..................................................141 Introduction........................ .................................................................. .................141 Specific Aims...................................................... 143 Background................................................................................ 144 ALL Incidence....................................................................................................144 T-cell Leukemia Among A-T Patients..............................................................145 Ionizing Radiation.................................................................. 146 Ataxia Telangiectasia (A-T)....................................................................... 150 ATM Gene..........................................................................................................152 ATM Gene Status in Leukemia.......................... 154 Significance............................................................................................................... 156 Preliminary Study...................... 157 Research Design and Methods.................................................................................160 Overview................................................................................................................ 160 Case Selection........................................................................................................161 Control Selection............................................................................... 162 Selection of Controls.................................. 163 Data Collection........................................ 171 Blood Collection and Processing......................... 174 Single Strand Conformational Polymorphism (SSCP) for ATM Gene Mutations..................................................................................175 Protein Truncation Test (PTT) for ATM Gene Mutations................................. 177 Data Analysis......................................................................................................... 179 Sample Size and Power Considerations...............................................................180 Timeline................................................................................................................. 184 Alternative Design of the Case-Control Study.................................................... 184 Limitations and Potential Problems..................................................................... 189 Human Subjects........................................................................................................ 193 Chapter 5: Conclusion..............................................................................................197 References..................................................... 205 Appendix A. Questionnaire for the Proposed Study: ATM Gene Mutation and Childhood Acute Lymphocytic Leukemia..............................223 Appendix B. Detailed Budget for Initial Budget Period of the Proposed Study...................................................................................................337 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Table 1. Summary of Major findings, Unanswered Questions of Studies Reviewed in Chapter 2 ............... 6 Table 2. Summary of Leukemia as a Component of Inherited Disease............... 13 Table 3. Summary of Familial Leukemia Case-Control Studies and Case Reports........................................................................... 15 Table 4. Summary the results of association between radiation and adult/ childhood leukemia in studies before 1990 (BEIR V )......................... 55 Table 5. Summary of Recent Studies on Association Between Radiation and Childhood Leukemia............................ 57 Table 6. Summary of ATM Polymorphisms Detected Among Six T-ALL Patients by SSCP......................................................................................67 Table 7. Summary of ATM Mutations Detected in Tumor Samples of Eight T-PLL Patients.............................................................................. 68 Table 8. Summary of ATM Mutations by PTT in Four T-PLL Cases.................69 Table 9. Summary of ATM Mutations Among Seventeen T-PLL Patients.........70 Table 10. Summary of ATM Mutations Detected in One Allele Among Six T-PLL Patients Who Had a Deletion in the Other Allele..............71 Table 11. Eligibility Criteria of CCG-B903 Study .................... 82 Table 12. Demographic Characteristics of the Childhood ALL Patients with Family History of Cancer..................................... 88 Table 13. Weights of Cancers among Family Members of Childhood ALL Patients...................................................................... 89 Table 14. The Observed Cancer among ALL Patient Relatives with Family History of Cancer (after Excluding the Second Proband) in CCG-B903 Study Compared to the Expected Number Based on the Age, Gender, Race-Adjusted Cancer Incidence in Los Angeles County 1983-1987....................................... 93 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 15. The Observed Cancer among Male and Female ALL Patient Relatives with Family History of Cancer (after Excluding the Second Proband) in CCG-B903 Study Compared to the Expected Number Based on the Age, Gender, Race-Adjusted Cancer Incidence in Los Angeles County 1983-1987 ......................... 98 Table 16. The Observed Cancer among First-Degree ALL Patient Relatives with Family History of Cancer (after Excluding the Second Proband) in CCG-B903 Study Compared to the Expected Number Based on the Age, Gender, Race-Adjusted Cancer Incidence in Los Angeles County 1983-1987 ......................... 99 Table 17. The Observed Cancer among Second-Degree ALL Patient Relatives with Family History of Cancer (after Excluding the Second Proband) in CCG-B903 Study Compared to the Expected Number Based on the Age, Gender, Race-Adjusted Cancer Incidence in Los Angeles County 1983-1987 ................... 100 Table 18. Comparison between Using Unrelated and Parental Controls 172 Table 19. Sample Size and Power (1-p) Used for Design not Matching on Family History of Cancer................................... 182 Table 20. Sample Size and Power (l-(3) Used for Design Restricting on Family History of Cancer...................................................................... 183 Table 21. Number of Cases Required to Detect a Gene-Environment (G- E) Interaction for Different Levels of G and E Exposures with 80% Power............................................................................................. 185 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Background Childhood acute lymphocytic leukemia (ALL) is the most common childhood malignancy. Despite decades of research, the etiology of ALL remains largely unknown. Because childhood ALL occurs at a very young age (median age: 4-5 years) with a lack of long term exposure to possible environmental risk factors, genetic factors may play a greater role in causing childhood ALL than with adult cancers. Review of the literature A proportion of familial leukemia has been found associated with Li- Fraumeni syndrome, ataxia telangiectasia (A-T) and other inherited syndromes. Multiple leukemia cases among offspring of consanguineous marriages suggest inherited factors may play a role in the etiology. A-T patients who are carriers of two copies of mutated ATM gene and the presumed carriers of one copy of the mutated gene have been found to be more sensitive to radiation than those who do not carry the mutated gene. Previous studies identified 32%-62.5% somatic ATM mutations among T- PLL patients. ATM germline mutations were found among 5 of 19 T-ALL patients (26%) in one study. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Data analysis The B-903 study is one of the largest similar type studies in the U.S., indicating there is an increased risk of leukemia and breast cancer among family members of childhood acute lymphocytic leukemia (ALL). The results of the study help to formulate the hypothesis that there are genetic factors playing roles in childhood ALL and in the increased risk of cancers among family members of childhood ALL. Proposal The proposed case-control study will be the first attempt in the U.S. to examine the association of a germline mutation and childhood ALL. The proposed study is an important step toward testing the etiologic role of germline mutations in the recently cloned ATM gene in the development of T-ALL. The findings of this study will indicate whether future, larger scale studies are indicated to determine whether the majority of ALL patients are predisposed to this disease because of carrying a germline mutation in ATM, and whether low dose radiation interacts with germline mutation of ATM gene on the risk of developing childhood ALL. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1 - Overview Childhood acute lymphocytic leukemia (ALL) is the most common childhood malignancy. Despite decades of research, the etiology of ALL remains largely unknown. Although high dose radiation has been found to be one cause of leukemia, studies of demographic and other environmental factors have produced inconclusive findings. Because childhood ALL occurs at a very young age (median age: 4-5 years), before long term exposure to environmental risk factors is possible, genetic factors may play a greater role in causing childhood ALL than with adult cancers. In this dissertation, the genetic etiology of childhood ALL is explored to search for answers. Chapter 2 of this dissertation reviews: 1) studies investigating the genetic component of leukemia, which includes familial leukemia and leukemia in twins; 2) studies of environmental factors, focusing on radiation; and 3) studies of mutations in the ATM gene among leukemia patients. Available familial leukemia case reports, and twin studies from 1960 to the present are included in Chapter 2 to show how genetic factors of leukemia were studied and reported in years past when no modem biological techniques were available. All recent studies related to ataxia telangiectasia (A-T) and the AT mutated (ATM) gene, which was cloned in 1995, and their associations with leukemia are also discussed in Chapter 2. The findings of these studies inform the hypotheses to be tested in the proposed study concerning the association of the ATM gene and childhood leukemia. 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In Chapter 2, the major findings, limitations, and clues regarding genetic etiology of each study are discussed. At the end of each section, the results, limitations and implications of these studies discussed in that section are summarized. Studies included in Chapter 2 are also summarized in several tables to help to review results across studies. The major findings, unanswered questions and future directions of the entire chapter are given in the three-page table at the beginning of Chapter 2 to help provide the “big picture” of these studies. The contents of the review chapter are referenced or mentioned only briefly in subsequent Chapters 3 (data analysis) and 4 (grant proposal), to avoid redundancy with Chapter 2 (review of the literature). Chapter 3 of the dissertation presents an analysis of the Childhood Cancer Group (CCG) B-903 study (Jonathan Buckley, P.I.). In this study, the frequency of 25 types of cancer was compared among family members of children having ALL with the number of cases expected in the general population. This retrospective cohort study covered the period 1990-96, and involved 387 ALL patients (probands) and 5,103 proband family members. Family history information on the incidence of cancer was collected for first and second degree relatives of newly diagnosed ALL probands under 21 years of age treated at a CCG-affiliated hospital. Standardized incidence ratios (SIRs) were calculated for each type of cancer by dividing the number of observed by the number of expected incident cases of cancer among ALL family members, controlling for age. 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Data analysis results are reported using the format of a potentially publishable article in a scientific journal. The results and discussion are summarized for each cancer site. The discussion focuses on major genetic findings of each cancer in literature, and what genetic or environmental path these cancers possibly shared with leukemia. These cancers including leukemia and breast cancer are reviewed in the discussion of the analysis chapter. The leukemia literature, reviewed in Chapter 2, is not repeated in the discussion section. Chapter 4 proposes a case-control study to examine the association of germline mutations in ATM gene and development of childhood ALL. The ATM germline mutation rate among ALL cases will be compared with population controls matched on age, gender and race. It is hypothesized that the ATM mutation rate will be higher among ALL cases than controls. This study will recruit 125 pairs of cases (T-ALL childhood patients) and controls. Blood samples will be collected to assess the germline mutations in the entire ATM gene. The proposal chapter follows the format required by NIH, except that additional information has been provided to discuss alternative study designs, and the two main methods of screening ATM gene mutations are compared and discussed in greater detail than would ordinarily be done in an NIH grant proposal. A draft budget of the proposal is included at the end of proposal as an appendix, along with the modified study questionnaire proposed to collect data from the parents of patients and controls on family history of diseases and environmental exposures. 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2 - Childhood Acute Lymphocytic Leukemia - An Epidemiologic Review Introduction Cancer is the most common natural cause of death among children, exceeded only by accidents. Although the incidence of childhood cancer is low, the potential years of life lost associated with childhood cancer exceeds any type of adult cancer. Acute lymphocytic leukemia (ALL) is the most common childhood malignancy. Although early prophylactic treatment of the central nervous system (CNS) and combination chemotherapy greatly improved disease-free survival over past decades, treatment remains eventually unsuccessful in about thirty percent of such patients. ALL is costly to the child patient, the child’s family, and to the society. During treatment, patients must cope with adverse side effects and feelings of anxiety and depression. Behavioral problems among patients often arise as normal coping strategies are exceeded. Even among those ‘cured’ from the disease, patients often struggle with educational challenges, social, physiological, and psychological problems in their attempt to re-construct normal lives (Bleyer, 1990; Birch, 1983). With the possible exception of high dose radiation which has been indicated as a cause of leukemia, studies for other environmental factors have resulted in inconsistent findings (Pendergrass et al, 1985; Ross, 1999). Researchers have further pursued the association of genetic factors in the development of childhood ALL. This chapter will review studies over the last 30-40 years that investigate the 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. etiology of childhood leukemia. Findings for these studies have been used to formulate new hypotheses concerning the association of the ATM gene and childhood leukemia (see Chapter 4). The aim of this present chapter is to review three types of studies related to the etiology of leukemia. First, studies relating to the genetics of leukemia include reports of familial leukemia and leukemia in twins. Next, studies of environmental risk factors are examined. The importance of radiation is emphasized because i) high dose radiation is one of the few proven risk factors for leukemia; ii) carriers of both copies of mutated ATM gene (homozygotes) and carriers of one copy of mutated ATM gene (heterozygotes) have been found to be more sensitive to radiation than normal individuals; and iii) finally, studies of mutations in the ATM gene among leukemia patients are discussed. The findings of these previous studies provide a rationale for examining the possible association of the ATM gene mutation, radiation and the development of childhood ALL . The following table (Table 1) summarizes the major findings, unanswered questions of the studies reviewed in Chapter 2. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1. Summary of major findings, unanswered questions of studies reviewed in Chapter 2 Review Sections (Summary of Literatures) Major findings, unanswered questions and future directions: Genetic Factors: Familial ALL (see Table 2, 3) Familial Leukemia: 1) Leukemia is found among patients with Li-Fraumeni syndrome (germline p53 gene mutations) and with ataxia telangiectasia (germline ATM gene mutations) and other inherited syndromes. 2) Familial leukemia has been reported in the literature, but the underlying causal mechanisms behind the familial aggregation are not clear. Very few such studies had performed thorough research for all possible shared environmental factors, and molecular genetic techniques such as mutation detection were limited in those days. On the other hand, multiple leukemia cases among offspring of consanguineous marriage suggest inherited factors may play a role in the etiology. 3) No specific mode of genetic transmission has been identified so far based on a limited number of segregation analyses. No studies have been done to search for an association between germline mutations and leukemia. 4) There is a large international variation of ALL and acute myeloid leukemia (AML) incidence. The difference between the incidence among the migrants and incidence in their homeland would provide clues related to the contribution of genetic or environmental factors in the etiology of childhood leukemia. 5) Future directions: a) to establish a systematic childhood leukemia registry and family pedigree to perform segregation analysis; b) to perform a well designed matched case-control study to study both environmental and genetic factors; c) to collect blood samples of certain high risk groups such as infant leukemia, childhood leukemia with FF1, or one histological group such as T or B-cell ALL to study the association of these homogeneous leukemia with germline mutations of genes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1 (continued). Summary of major findings, unanswered questions and future directions of studies reviewed in Chapter 2 Genetic Factors: Leukemia in Twins (see Table 2, 3) Leukemia in Twins: 1) The reported concordance rate of ALL ranges from 5% to 25% in the literature. Because of the subjective judgment by physical features, blood groups, placenta, membrane of twins and the ascertainment of identical twins vary from study to study, those traditional methods are less accurate than the molecular techniques used nowadays to ascertain MZ twins. 2) The most likely explanation so far of the increased concordance rate of leukemia is that leukemia cells are transferred from one twin to the other in utero. 3) There is an overall deficit of leukemia in twins compared to the general population. 4) Although the above 2) and 3) findings do not provide direct supportive evidence that genetic factors play a role in leukemia, they raise the following questions: a) if a twin could be affected by getting leukemia cells from the other twin, why is the concordance rate not 100%? Are there any genetic or environmental factors associated with whether the other twin accepts the leukemia cell transplantation? b) Most twins had leukemia onset during infancy if the leukemia cells are transferred from one twin to the other. However, there is one case report of a pair of twins who had ALL at 9 and 11 years suggesting that the leukemias originated from a single leukemic cell, developed in one twin, and transferred to the other twin (Ford et al, 1997). How and in what way do genetic and environmental factors influence the twin leukemia concordance in infants or in teenagers? 5) Future direction: to establish a complete twin registry to follow all twins from birth (including twin pairs with 0,1 or 2 twins affected) and to collect information on their genetic and environmental exposures. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1 (continued). Summary of major findings, unanswered questions and future directions of studies reviewed in Chapter 2 Environmental Factor: Radiation and ALL (see Table 4, 5) 1) High level radiation exposure is associated with increased risk of leukemia. The excess of leukemia appears to increase with increasing mean dose of radiation to the bone marrow. Risk is higher for those exposed at 20 years of age or younger. High dose radiation has become very rare nowadays. 2) The association of low dose radiation with leukemia is inconsistent. Studies with positive findings have usually found a 1.5 to 4-fold increased risk among those exposed to low dose radiation. These findings may also reflect effects of chance, or possible confounding factors, bias and exposure measurements. 3) Testing the association of low dose radiation and leukemia has become more challenging since low dose radiation studies have usually focused on medical diagnostic radiation - the radiation doses of which have been greatly reduced with the improvement of diagnostic equipment. Moreover, the proportion of medical radiation is only about 11% of the total radiation exposure. The other main sources of radiation come from radon (55%), outer space (8%), rock and soil (8%) and from inside the human body (11%). The total radiation exposure and their differences could not be reflected solely by medical radiation exposure. 4) Future studies of association of genes that predispose humans to be more sensitive to radiation exposure, and to find more accurate methods of measuring total low dose radiation could help to determine the risk of low dose radiation for developing leukemia. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 1 (continued). Summary of major findings, unanswered questions and future directions of studies reviewed in Chapter 2 ATM gene and Leukemia (see Table 6-10) Results of a few initial studies of ATM gene mutations among leukemia patients indicated that: a) these studies have focused on T-lineage ALL or PLL patients with 32% -62.5% somatic/germline mutations identified; b) missense mutations are dominant in T-PLL and T-ALL patients, in contrast to mostly truncating mutations found among A-T patients; c) almost all mutations identified so far have been unique regarding mutation types and locations in the ATM; d) most studies, except one, revealed somatic mutations instead of constitutional mutations. These studies could not estimate the risk of ATM gene mutations and leukemia because: a) small samples of convenience (8-37 cases); b) non-random selection of “unmatched” controls; and c) the sensitivity of screening methods of ATM mutations among cases and controls could be improved to be higher. These studies raise several questions: a) whether these somatic mutations among T-PLL could also be identified in the germline; b) whether ALL patients have the same or different type and frequency of ATM gene mutations as T-PLL patients; and c) the generality of the majority of germline mutations identified in a small sample of T- ALL patients (Luo et al, 1998) in one previous study to the majority of childhood ALL is uncertain. Questions for future studies include: a) whether these nucleotide substitutions cause loss of function of ATM gene or they are “neutral” polymorphisms; b) whether the number of missense mutations is significantly different from expected; and c) what is the difference between the missense or truncating mutations among leukemia patients compared to those found among A-T and general population? 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Genetic Factors for Acute Lymphocytic Leukemia (ALL) Genetic factors are discussed in four sections. First, studies of international variation in the incidence of childhood leukemia will be briefly reviewed. Secondly, the discussion of familial leukemia includes leukemia in syndromic disease and familial leukemia not reported as a component of syndromic disease, which includes: a) multiple leukemia in some families; b) clustering of leukemia among children of consanguineous marriage; c) excess leukemia among families with leukemia, or multiple childhood cancer patients. Finally, the review of leukemia in twins discuss the increased concordance for leukemia in twins and the overall deficit of leukemia among twins. International Variation o f Childhood Leukemia Leukemia is the most common type of childhood cancer, accounting for 30%-40% of childhood cancer. There is marked international variation in the incidence of childhood leukemia. For example, ALL is the most common childhood malignancy in western countries (except in the African American population in the US) with an annual incidence between 24.6 to 31.7 per million children compared to 12.5, 12.7 per million children in Japanese and US black children. On the other hand, the most common childhood neoplasm in Japan and China is AML instead of ALL with an annual incidence of 13.3 per million children compared to the annual incidence of 2.7 to 6.6 per million children in western countries (including the 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. African American population in the US) (Birch, 1983). It would be interesting to study the incidence of leukemia among migrant Chinese and Japanese in the US. The difference between the incidence among these migrants and incidence in their homeland would provide clues related to the contribution of genetic or environmental factors in the etiology of childhood ALL. It is not known that any of such studies has been carried out. Leukemia in Syndromic Disease Reports of familial leukemia are infrequent in the literature. When aggregation in families is reported, it is commonly a component of other diseases or medical syndromes. Germline mutations of genes are believed to be responsible for these medical syndromes and patients with these syndromes are predisposed to develop leukemia. Some of the genes characterizing these syndromic diseases are summarized in Table 2. Germline mutations in p53, ATM, NF1 have been linked to the Li-Fraumeni syndrome, ataxia telangiectasia (A-T), and neurofibromatosis, respectively. Swift et al (1976) found five individuals who died of leukemia or lymphoma before age 45 among relatives of patients with A-T in 27 families compared to one expected death (P<0.01). The relative risk of dying of leukemia and lymphoma before the age of 45 among ATM heterozygotes was estimated to be 7 times (P<0.03) the risk in the US white population using a maximum likelihood method (Swift et al, 1976). Neurofibromatosis was found in excess among children with all kinds of leukemia, including ALL and CML (0/E= 3.8 and 2.7 and 71.4 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. respectively) (Narod et al, 1991). Leukemias are also clustered among patients with Bloom syndrome, Fanconi syndrome, immunodeficiency syndrome, Wiskott-Aldrich syndrome, Blackfan-Diamond syndrome, etc. (Narod et al, 1991). Constitutional mutations and the excess of leukemia found in individuals with these syndromes provide evidence for a genetic origin among a subset of patients. Further studies are required to test the association of germline mutations of these genes with childhood leukemia to determine the relative risk in both homozygote and heterozygote carriers of the mutation and identify biological function of these genes and their pathological path to initiate leukemia. Familial Leukemia not Associated with a Recognized Syndrome Multiple Leukemia in Some Families Multiple cases of leukemia occurring in a single family have been reported in the literature. While case reports of familial aggregation of leukemia are individually unable to establish the association between heredity and leukemia, they collectively suggest a genetic component to ALL etiology. Table 3 summarizes case reports and case-control studies of familial leukemia. A brief summary of these major findings and my critique of these findings for each of these studies follow. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2. Summary of leukemia as a component of inherited disease Relation to leukemia Gene (Locus) Function Leukemia as a component of syndromic illness Constitutional trisomy Down syndrome ALL predominates > 3yrs AML1 CBFA to DNA AML predominant <3 yrs (21q22.3) IFNAR, CRF2-4, GART, SON sequence modification common to many genes expression AD platelet granule defect predisposition to AML Chromosome 8 trisomy mosaicism DNA repair deficiency AML/CML (8) uninvestigated function expressed during hematopoiesis; coding sequence of MOZ- a novel zinc finger containing gene with putative acetyltransferase activity Bloom syndrome ALL/AML BLM (15q26.1) increased sister chromatid exchange (AR) Ataxia telangiectasia(A-T) T-cell ALL ATM (1lq22-23) ATM, a putative phosphatidylinositol kinase involving meiotic recombination and cell cycle control i.e. failure to inhibit DNA synthesis following radiation and disobedience to the Gl-S transition checkpoint of the cell cycle (AR) Nijmegen/Berlin breakage syndrome (8q21) two variants of A-T (AR) Fanconi anemia AML (9q22.3) (16q24.3) FACC Encoding a cytoplastic protein of unknown function (group C) containing a leucine zipper and nuclear localization signals (group A) (AR) Tumor suppressor gene syndromes Neurofibromatosis 1 ALL/CML NF1 protein encoded is GAP (17qll .2) family and down regulate the p21-rais proto-oncogene; associated with skin findings of xanthogranuloma, loss of heterozygosity with deletion of normal allele in myeloid-derived clones; more common in boys and associated with maternal inheritance (AD) Li Fraumeni syndrome leukemia p53 germline mutation (AD) (17ql3.1) 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2 (continued). Summary of leukemia as a component of inherited disease Relation to leukemia Gene (Locus) Function Leukemia as a component of syndromic illness Immunodeficiency syndromes Wiscott-Adldrich syndrome ALL/AML WASP (Xp 11.23) encoding a proline-rich protein (XLR) Bruton agammaglobulinemia BTK (Xq21.3) tyrosine kinase i.e. failure to produce B cell lymphocytes and Ig heavy chain rearrangement (XLR) Schwachman-Bodian pancreatic lipomatosis hematologic malignancy ?t(6;12) i.e. congenital pancreatic lipomatosis and early onset of pancytopenia and malignancies (AR) Kostmann's infantile genetic agranulocytosis AML (Ip35-p34.3) mutations in GCSF receptor (AR) Blackfan-Diamond syndrome AML AD,AR AD: autosomal dominant; AR: autosomal recessive; XLR: x-linked recessive Bridges et al (1961) reported two cases eaeh of leukemia among three families in a series of 182 leukemia patients diagnosed at one hospital from 1948 to 1959. One patient and his first cousin had acute leukemia at ages 26 and 22, respectively. In another family, a male was diagnosed with sub-acute lymphocytic leukemia at age 52. Five years later, his son was diagnosed with acute lymphocytic leukemia at age 21. The third family had two female cases of acute leukemia: a four- year-old diagnosed with acute monocytic leukemia, and her first cousin diagnosed with leukemia and Down syndrome at 15 months. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3. Summary of familial leukemia case-control studies and case reports Authors number of cases Type of leukemia mode of inheritance environmental factors /cytogenetic analysis Case-control study Videbaek 17 cases had family history (1947) of leukemia (>=2 cases in a family) among 209 families compared to 1 case among 200 control families Examples: 4 cases in three generations Gunz (1975) Li (1976) 72 out 909 had >=2 cases among first degree relatives three times the control (matched by age, sex) 38 families had 2 or more cases of childhood cancers among 5000 families, two age and diagnosis matched controls were selected O/E=3/0.2=T5 not stated 2 CLL, 1 Chronic granulocytic leukemia, 1 eosinophilic leukemia CLL more than AL noCM L aggregation not stated dominant with low penetrance not monogenic but more likely polygenic Two to 5 fold increased risk among parents, aunts and uncles, grandparents compared to the controls and general population Linet 23 siblings of 342 leukemia (1986) pts had leukemia /lymphoma compared to 4 in matched cancer controls; and 10 in noncancer controls CLL Pottern Siblings of adult leukemia (1991) patients (>=30) had 2.3 and 3 fold increased risk of leukemia and CLL respectively compared to the controls Shpilberg (1994) , Cohort Study Draper (1977) Relatives of patients with hematological neoplasms had 3.6 fold increased risk of hematological neoplasms compared with the relatives of patients with non- malignant hematological neoplasms or type II diabetes in 12 families both sibs were affected by leukemia; in 7 families, one sib had leukemia, the other sib had lymphoma; in 18 families, one sib had leukemia, the other sib had a solid tumor among 20,000 families with a childhood cancer; The relative risk of leukemia among siblings of leukemia patients compared to the general population was estimated as 2.3. important demographic characteristics (such as age) were not matched 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3 (continued). Summary of familial leukemia case-control studies and case reports Authors number of cases Type of leukemia mode of inheritance environmental factors /comments Cohort Study (continued) Narod (1991) Relative risk of ALL between ag( 0-14 in Down syndrome patients was estimated as 10. The heritable fraction of leukemia was estimated As 2.6% Draper (1996) Siblings of AL patients had a 2-4 fold increased risk than children from the general population Familial Cases R eport Decastello 4 cases in two generations /Weiss CLL Anderson 5 out of 8 siblings Age 5-8 years childhood acute Leukemia Johnson /Peters 3 adult and 1 lymphosarcoma Among 12 siblings acute leukemia Peters 4 childhood leukemia among 12 siblings parents were first cousins Steinberg 3 cases and 1 lymphosarcoma Among 7 children 4 among 12 children acute leukemia acute leukemia recessive consanguinity parents were second cousins Gordon 4 adult in two generations acute leukemia Leverger three cousins of same generation (at ages of 2,3,5 years) childhood ALL had same HLA haplotype Bridges (1961) 2 cases each in three families First cousins at 26,22 yrs Father and son at 52, 21 yrs First cousins at 4, 1.25 yrs acute leukemia sub-acute LL and ALL AML and AL Health (1964) 5 cases in three generations Three sisters at 26,53,65 (case 1-3) case3’s daughter and her son at 39 and 7 (case 5 and 4) acute leukemia AML no environmental factors investigated no virologic exams 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3 (continued). Summary of familial leukemia case-control studies and case reports Authors number of cases Type of leukemia mode of inheritance environmental factors /comments Familial Cases Report (continued) Gunz (1966) ALL in four siblings at 3 yrs, 10 months, 10 months, and 6 yrs among 8 full siblings and 1 half-sibling; a fifth sibling suspected having chromosome break ALL dominant with full penetrance normal karyotypes and a few cells were hypodiploid this patient had leukemia at 11 months no virus disease, radiation, insecticides sprays, no contact industry poisons Rigby (1968) 2 or more cases in 39 families leukemia In 32 families, the cases are siblings or parent-child pairs Mcphedran (1969) three cases each in two families at 59,58,51,45 59 and 44 yrs parents of these 6 patients were siblings 4 cases in mother, daughter, two distant cousins in another family CLL 1 A LL, 1 AML 2 AGL genetic factors suggested Kurita (1974) 2 or more siblings in 20 families 30% (6/20) parents of these siblings were first cousins compared to 4.5% (9 out of200) among non-familial case families single recessive or polygenic two patients exposed to benzene Gunz (1978) 13 cases in a family of three generations at ages 31,66,50,21,17,5, 14,4, (AML), 41 (CML), 2,9,9,7 (AL) 8 AML, 1 CML 4 AL single gene dominance with varying degree of penetrance contact with grey hound, dogs, cats, birds, fish chromosome 11 deletion in one patient (acquired) Kende(1994) multiple ALL in two Arab ancestry with intermarriage and consanguinity ALL autosomal recessive Although no environmental factors were studied, the results suggest that: 1) a small percentage of leukemia cases (3% in this study) had familial aggregation and a proportion of these familial cases might provide clues of genetic etiology; 2) the 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. youngest female leukemia case among the three instances, who was diagnosed at 15 months (with Down syndrome) might support a hypothesis that genetic factors were more likely to play a role, since she had onset at a young age and now it has been demonstrated that patients with Down syndrome or other chromosome abnormalities have higher incidence of leukemia than the general population. Heath et al (1964) reported five AML cases from 1904-1963 among three generations of a single family. The cytogenetic report of case 5 and her two surviving normal sons did not reveal abnormalities of chromosome numbers. No environmental factors were investigated, nor were any virologic examinations done. The authors carefully validated all reported cancers and diseases in this family through death certificate or hospital records. The five cases of acute leukemia (3 AML and 2 possible AML) with similar clinical features, the two cases of breast cancer, and a family member with Down syndrome whose mother had chromosome 12-15 translocation seems to suggest a genetic origin. However, it was not possible to determine at that time the genotype for genes like BRCA1, BRCA2, p53, ATM. The presence of multiple cases of breast cancer and leukemia in this family suggested a possible role of constitutional mutations in p53, ATM (Li-Fraumeni syndrome, A-T) or other unidentified genes. Gunz et al (1966) observed four ALL cases (including two fraternal twins) at age 3 years, 10 months, 10 months and 6 years among 8 full siblings and 1 half sibling. A fifth sibling developed a disease suggestive of leukemia at age 11 months. No distinctive blood and serum groups, or chromosome abnormalities were found 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. among these sibling patients and their parents. One ALL sibling and one sibling suspected of having ALL had some degree of chromosome breakage. Possible environmental factors were studied. Background radiation and radioactivity in milk were measured by the National Radiation Laboratory and showed no unusual levels compared to other parts of the country. No viral diseases and other childhood leukemia were reported in this small Maori community during 1953-1963. No unusual level of insecticides/agriculture sprays was detected in this area compared to other parts of the country. The family had no history of exposure to industrial or other poisons. The large number of cases in one nuclear family, and the young age at onset (3 infant leukemia, and the other 2 under 6 years) in this family suggested a genetic basis. Although it is not clear without the knowledge of other cancers in this family since the extended pedigree of the family was not collected, these multiple childhood leukemias could be part of Li-Fraumeni syndrome. The unusual observation of 4 independent acute leukemia cases in a sibship of 9 would furnish a strong argument for the presence of a germline mutation. Rigby et al (1968) reported 39 families with multiple leukemia/lymphoma cases involving 91 relatives from a total of 855 family members. Of the 39 families, 32 involved a close genetic relationship between affected individuals (19 siblings, 1 twin pair and 12 parent-child pairs). Of these 32 families with multiple cases, four pairs and three triplets had acute leukemia. The presence of multiple cases of leukemia/lymphoma in these families suggests that genetic factors might play a role. 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. More deaths occurred under the age of 30 among multiple family cases (N=86) than single-family cases (N -l 12). Families with multiple cases ("cases") and single cases ("controls") were selected through surveys and tumor registries in hospitals in a central six-county area in Nebraska during 1958 to 1963. The authors carefully validated the diagnosis through medical records and death certificate. No comparison of incidence and death due to leukemia and lymphoma was made between the "cases" and "controls". All lymphoma (including Hodgkin's disease and Non-Hodgkin’s lymphoma) and leukemia (including acute leukemia, CLL) were lumped up together, making it impossible to study acute leukemia separately. McPhedran et al (1969) reported three sibling CLL leukemia cases each in two families. The three CLL in one family and the three CLL in the other family were also first cousins. The ages of onset of these six CLL cases were 59, 58, 51, 45, 59 and 44 respectively. In these families with CLL, four additional cases of acute leukemia occurred in a mother-daughter pair, and in two distant cousins; the acute leukemia developed at ages of 47, 17, 63, and 13, respectively. The two families were related through a second generation marriage. The four acute leukemia types included acute granulocytic, acute monocytic, acute lymphocytic and acute granulocytic. Among the 111 death records of the relatives of the two families, ten were attributed to malignancies other than leukemia, including two cases of lung cancer, two cases of liver cancer, one case each of throat, neck, cervix, testicular cancer and one unspecified cancer. Two other relatives developed lung and skin 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cancer at the time of study. None of these occurred among first degree relatives of the leukemia patients and the incidence of each type of cancer did not appear excessive. Although CLL and acute leukemia occur at different ages (old vs. young) and have different clinical features, these two related families provide some clues for genetic etiology and may support the theory that an abnormal stem cell is differentiated into different leukemia cells. The pattern of cancers in these families does not seem compatible with typical Li-Fraumeni syndrome. The genetic abnormalities responsible for these multiple leukemia cases were not known. Gunz et al (1978) reported thirteen cases of leukemia in a single family. These thirteen cases included twelve acute myeloid leukemia and acute leukemia, and one case of chronic myeloid leukemia in a family of three generations among a total of 293 family members. Neither chromosome abnormalities nor genetic markers were linked to the occurrence of familial aggregation, except that a chromosome 11 deletion was found in one patient. These findings suggested a genetic predisposition to develop leukemia. Of the twelve acute leukemia cases, eight were AML which included five children (diagnosed at ages of 21, 17, 5, 14 and 4, respectively). The three adult AML cases were diagnosed at the ages of 31, 66 and 50. The other four acute leukemia types were not specified and occurred during childhood, at ages of 2, 9, 9, and 7 separately. An attempt to link the haplotypes with the occurrence of leukemia was unsuccessful. An immunological test on one patient of this family failed to find any constitutional immune deficiency. 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Preliminary results of virologic studies indicated the presence of viruses in at least one individual of the family. All 13 cases of leukemia were verified from death certificates and, when available, doctor's/hospital records. These findings suggested a strong genetic predisposition to leukemia within this family with little evidence of known environmental risk factors such ionizing radiation. There were no consanguineous marriages. The significance of contact with pets among the family members was unknown. Exogenous factors such as virus infection or their interaction with genetic factors could not be excluded. The remarkable similarity of the morphologic type of leukemia among the family (most were AML or acute leukemia) suggested a genetic origin. An attempt was made to strengthen the hypothesis of a genetic origin by examining the linkage between the leukemia aggregation and genetic markers, but was unsuccessful. It was found that the chromosome 11 deletion in one patient was acquired rather than constitutional. In this family, the precise nature of leukemogenic etiologic factor(s) remained unknown. Horwitz (1997) suggested that leukemia may be inherited with 'anticipation', a pattern of declining age of onset from one generation to the next. The models of trinucleotide repeat expansion as in Huntington disease and neurodegenerative illness inherited of multiple downstream mutations may result from a defect in a DNA repair gene. These hypotheses were based on that the interesting findings of CAG/CTG repeats in Huntington disease and spinal cerebellar ataxia. Meiotic and mitotic instability resulting from abnormalities in DNA repair were also observed in 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A-T patients. Subsets of leukemia demonstrated both microsatellite and trinucleotide repeat instability. Horwitz (1997) further suggested that the risk of a childhood leukemia may result from 1) inheritance of not only a single defect, but also multiple de novo germline mutations or 2) inheritance of a single major ancestral gene (to account for anticipation) and multiple minor genes (secondary mutations to account for clinically variations of leukemia subtypes) rather than an autosomal dominant germline mutation. These two theories may help explain the observation that siblings of leukemia patients have an increased incidence of birth defects, Down syndrome, solid tumors and increased noncancer death rates. The incidence of Klinefelter XXY syndrome, Turner XO syndrome, congenital anomalies (a de novo aneuploidy, chromosomal micro-deletion), spontaneous abortions among mothers of leukemia patients, increased growth and developmental delay of the patients (mentally and physically, such as underdevelopment of upper limb, abnormalities of the thumb) in leukemia families reflect de novo mutational events. This implies that the first significant ‘hit’ in the multi-step evolution of leukemia is a germline, rather than somatic event in a subgroup of familial leukemia. However, based on the Horwitz’s own data, it was concluded that the absence of increased risk of malignancy among offspring of ALL survivors (N=382), suggested that the heritable component may be small or non-existent for ALL (Horwitz, 1997). The absolute excess risk in each generation was not established. Even so, the results could not rule out a genetic component in ALL since first, ALL survivors could be different from the overall ALL population, and in particular, 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different from those who had onset during infancy. It is reasonable to assume that ALL with an early age onset is more likely to have a genetic origin. About 60-80% of infant leukemia have a chromosome 1 lq23 (somatic) rearrangement (Maloney et al, 1997, cited by Ross). Secondly, 5 year survival of infants with ALL is about 25% compared to a 70% survival for those diagnosed with ALL between age 2 and 5 (Pui et al, 1994). Finally, a substantial proportion of ALL survivors went through radiation therapy, which might improve the survival of these ALL patients. On the contrary, for ALL patients who were carriers of ATM mutations, radiation therapy would have increased the chromosome breakage and reduced survival rates. Therefore, the subset of infant ALL with a strong genetic background and poor survival were not likely to be selected as survivors in Horwitz study. The hypotheses Horwitz suggested for familial leukemia needs more studies to provide further evidence. In a summary, these familial leukemia reports suggest, first, that there may be a role for genetic factors. The evidence for a genetic etiology appeared in those families with multiple cases among several generations, and within these families there was a remarkable similarity of histologic type, morphologic leukemia cells, and age of onset. Secondly, constitutional mutations responsible for familial aggregation of leukemia are largely unknown. Only p53 germline mutations in Li-Fraumeni syndrome with leukemia as a component have been identified. The association of other germline mutations with leukemia as part of the syndrome have not been 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. studied. Finally, the mode of inheritance of leukemia is not clear. Segregation analysis would provide further evidence. In the next section, reports of clustering of leukemia in an unusual situation - among children of a consanguineous marriage- will be discussed. It may offer specific clues to the genetic basis of leukemia familiarity. Clustering of Leukemia among Children of Consanguineous Marriages Kurita et al (1974) studied twenty families with two or more leukemia cases occurring among siblings in Japan. Thirty percent (6 of 20) of the parents of these sibling patients were first cousins, compared to 4.5% (9 out of 200) among non- familial leukemia families. Moreover, there were an additional four families with multiple siblings affected whose parents were second cousins or first cousins once removed. The age at onset among familial cases whose parents were first cousins was younger (median 6 years, range 13 days to 41 years) than for those cases whose parents were not related (median 27 years, range 5 months to 59 years). Control families were selected from 383 non-familial leukemia families diagnosed in a 13- year period at one cancer center and one university. Case and control selections were not clearly described in this article. Cases were identified through author's own investigation (N=l), literature (N=16) and questionnaires (N=T2) sent to hospitals in Japan. How these reports and hospitals were selected, what percentage of hospitals responded to the questionnaires, and how these case diagnoses were confirmed was not specified. Similarly, the method of 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. selection for 383 non-familial controls was not described. The study indicated that parental consanguinity might be associated with more numbers of siblings affected with leukemia and at younger age than that of a single sibling affected. Kende et al (1994) described multiple ALL cases in two Arab ancestry families with intermarriage and consanguinity. The similarity of early age at onset (4-6 years), mode of presentation, similar tumor biology, similar characteristics of leukemia cells- ALL, similar poor response to treatment suggested a genetic etiology. In one family, the parents of two common ALL were second-cousins, and two siblings of these ALL cases died in infancy from unknown causes. The two ALL cases were diagnosed at 5.5 and 6 years respectively. The second family was closely interrelated with three consanguineous marriages. The parents of one T- ALL, one lymphoma and one unclassified leukemia were first cousins, the parents of the other two T-ALL and two deaths in infancy with gastrointestinal symptoms were first cousins. Two out of these five leukemia/lymphoma patients had cafe au lait spots but no other stigmata characteristic of neurofibromatosis. These multiple ALL cases reported by Kende suggested an association between leukemia and consanguinity. These familial leukemias might be associated with either a single recessive gene (Mendelian mechanism) inheritance or associated with increased risk of syndromic diseases such as neurofibromatosis, immunologic deficiencies(A-T, Wiscot-Aldrich syndrome, etc.), or abnormality of metabolizing enzymes, enzymatic repair mechanism (DNA repair enzymes) among consanguineous marriages, which predisposed family members to leukemia. 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Steinberg (1960) also found that in two of the three families with multiple childhood leukemia, consanguineous marriage was involved (cited by Gunz et al, 1966). Li et al (1976) found that among 11 sibships with multiple cases of acute leukemia or lymphoma, two families (18%) had consanguineous marriage. In a summary, these early case reports conducted 20-30 years ago have been limited by the lack of advanced molecular technology for detecting abnormalities of chromosome structure, gene mutations, and polymorphisms. Therefore, the nature of genetic basis of leukemia in these families was largely unknown. Studies of familial leukemia in offspring of consanguineous marriage offer some clues regarding the etiology of familial leukemia. First, the similarity of histologic type, morphologic leukemia cells, age of onset among multiple leukemia cases with intermarriage and consanguinity indicated an impact of shared genetic material in the family. Secondly, in at least one study (Kende et al, 1994), two out of five leukemia/lymphoma patients had atypical signs of neurofibromatosis, a hereditary syndrome predisposing to leukemia. This may suggest that the increased incidence of hereditary syndromes and chromosome abnormalities among offspring of related parents were possibly responsible for the clusters of leukemia in these families. The familial aggregation of leukemia in consanguinity would indicate a recessive mode of inheritance of leukemia. If there is a single low frequency recessive gene shared by family members, one would expect an increased chance of two copies of the abnormal gene being inherited by the offspring of these related 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. parents than that among offspring of unrelated parents. These previous studies could not directly prove these hypotheses. Future studies searching for constitutional abnormalities among family members of intermarriage will provide further evidence. Although both studies of familial leukemia in offspring of consanguineous marriage and studies in familial aggregation in the previous section provided some clues for the genetic basis of leukemia, these two type of studies are descriptive, and could not establish an increased susceptibility or elevated risk. Case-control and cohort studies are more useful in this regard. The limited numbers of case-control or cohort studies are reviewed below. Excess Leukemia among Families with Leukemia, or Multiple Childhood Cancer Patients The incidence of leukemia among first-degree relatives of leukemia patients was first reported to be 16 times higher (8.1% vs. 0.5%) than in the control families by Videbaek et al (1947 cited by Ponz de Leon et al, 1994) (Table 3). However, the selection of cases and controls was questionable. It was not described how these cases and controls were selected, and whether the controls were randomly selected and whether confounding factors such as age, gender and race were matched in this study. Moreover, all different kinds of leukemia were mixed together in this study. First degree relatives of leukemia patients (all types) were three times more likely to develop leukemia than first degree relatives of controls matched by age and sex (Gunz et al, 1975 cited by Ponz de Leon et al, 1994). Since it is known that there 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. are significant differences in the incidence of leukemia among different races, not matching on race between cases and controls made the results of this study questionable unless the race distribution between cases and controls were similar. Again, different leukemia subtypes were lumped together. More CLL was found in case families. In this study, the question whether or not there is increased risk of ALL among relatives of childhood ALL was not addressed. More cancers were observed than expected among families with multiple childhood cancer patients. Thirty-eight families with two or more childhood cancers were ascertained among 5,000 cases during 1947-1974 and 1940-1974 respectively (Li et al, 1976). The 38 case families included one family with AML among identical twins, six with three or four affected siblings and 31 families with two childhood cancers. The “index case” was the last diagnosed childhood cancer from the 38 families with multiple cases. Among those families diagnosed with one childhood cancer in the same center, two “controls” matched on age and tumor type were selected for each “index case”. Three cases of childhood cancers were observed among siblings of “index cases”, compared to 0.2 expected. Parents had 5- fold (6.6% vs. 1.4%) and 3-fold (6.6% vs. 1.9%) risks compared to control parents and the general population by using Connecticut cancer incidence data, respectively. Aunts and uncles of multiple sibships with cancer had 3 times (6.8% vs. 2.1%) increased risk compared to their counterpart of the controls and the general population. Grandparents had 2-fold increased risk compared to the control 2 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. grandparents. Carcinogenic agents, chromosome analysis and immunologic studies were recommended for future studies. The authors of this study carefully reviewed childhood cancers diagnosed from 1947-74 at ages 0 to 19. Among 5000 patients, 38 families had two or more childhood cancers, whose diagnosis and the reported cancers among first degree relatives were completely documented in medical records. Therefore, the 3-5 fold increased risk among first-degree relatives seems reliable. In this study, aggregates of tumors among three families may a result of inherited disorders. One family of two-sib pairs with brain tumors had a family history of neurofibromatosis. The other two families both had one child with brain tumor and the other with sarcoma. These two families also had soft tissue sarcomas among first and second-degree relatives. The brain tumors and soft-tissue sarcomas in these two families may present Li- Fraumeni syndrome. However, the familial aggregation of childhood cancer compared to the general population seems inconclusive due to small numbers of observed and expected cases (3 observed vs. 0.2 expected). Malignancy history among second- degree relatives and other extended family members was obtained through interview rather than confirmed by medical records. The increased risk among aunts, uncles and grandparents may be subject to recall bias and misclassification. A larger sample size would be required to address whether there is an increased risk of childhood leukemia, and particularly childhood ALL, among family members of these patients. 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shpilberg et al (1994) found that family members (N-4,061) of patients with hematological neoplasms (N=189) showed a higher risk for developing hematological neoplasms (i.e. Hodgkin's disease, multiple myeloma, non-Hodgkin's lymphoma, acute leukemia, chronic lymphocytic leukemia, and myeloproliferative disorders) compared to relatives (N=1463) of patients with non-malignant hematological disorders (i.e. anemia, immune thrombocytopenia, neutropenia, etc.) or type II diabetes (OR=3.62, 95% Cl: 1.44-9.07) (N=69). Unfortunately, disease categories used for the selection of cases and the outcomes were not very specific. Different types of neoplasm such as lymphoma, leukemia, anemia were lumped together, both in selecting the cases and in observing events among family members. Also, cases and controls were not matched on important demographic characteristics such as age. Draper et al (1977) estimated the risks for cancer among siblings of children who had had cancer. Among 20,000 cases (parents of 15,000 cases were interviewed) of childhood cancers diagnosed between 1953 and 1974,102 families were found to have two or more affected children. There were 12 families in which * both sibs were affected by leukemia, 7 families in which one sib had leukemia and another sib had lymphoma, and 18 families with one sib affected by leukemia and another sib affected by other solid tumors (e.g., CNS tumor, neuroblastoma, Wilms’ tumor, bone tumors). There were 22 families in which 2, 3, or 4 sibs were affected by retinoblastoma. In one family, one sib was affected by retinoblastoma and the other sib had osteoclastoma (but not retinoblastoma). Aggregation within sibship of 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the childhood cancers was observed more frequently than expected by chance. The actual risk for such siblings was estimated as 1 in 300, compared to 1 in 600 in the general population. The siblings of a leukemia patient had 2.3 and 2.9 relative risks of developing leukemia and lymphoma respectively compared to a child from the general population. This study was based on a large survey of childhood cancers and all diagnoses had been confirmed by death certificates or, more often, by hospital records. The large sample size (20,000 cases) of childhood cancer patients made it possible to estimate the increased risk of leukemia among siblings of childhood leukemia patients. Families with multiple childhood cancers such as leukemia, retinoblastoma may be attributable to genetically determined conditions. The heritable fraction of leukemia was estimated to be 2.6% (142 out of 5564) by Narod et al (1991) in the UK childhood cancer registry of children under age 15 during the period of 1971 to 1983. Ethnic variation, unknown environmental factors and incomplete documentation of constitutional abnormalities and other underlying conditions or inherited mutations suggest the heritable fraction may be higher than this estimate. In this study, cases considered be hereditary were: 1) all bilateral retinoblastomas and Wilm's tumors; 2) all children with established, recorded underlying genetic conditions (i.e. a constitutional genetic mutation, either chromosomal or involving a single gene); and, 3) tumors occurring in children with aniridia (absence of iris) or hemihypertrophy (over growth one half of the body, an organ or part) which signals the presence of an underlying mutation. Based on these 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. criteria, the estimated 2.6% heritable fraction of leukemia seems indicate a proportion of leukemia attributable to recognized syndrome. More acute non- lymphocytic leukemia (ANLL) was attributable to Down syndrome (5.3%) than ALL (1.7%). The relative risk of ALL between age 0-14 in Down patients was comparatively constant at 10. Less than 1% of all leukemia presents as multiple cases within a single family. Siblings of AL patients have been found to have a 2-4 fold greater risk of developing the disease than children from the general population (Draper et al, 1996). In summary, these studies of familial leukemia among children of unrelated parents or consanguineous marriage suggest that among small groups of high risk families, genetic factors play an important role in the onset of one or more leukemia cases among these families. The inherited factors play a role in families predisposed to developing leukemia and/or other cancers in the form of multiple cases, early onset of the disease, or similarity of the histologic subtype of the disease and clinical prognosis, and among those multiple case families with consanguineous marriage. One segregation study of leukemia suggested a multifactorial origin for this neoplasm (Ponz de Leon, 1994). Although autosomal dominant, autosomal recessive, and autosomal dominant with reduced penetration were also suggested in studies of familial leukemia, no specific mode of genetic transmission has been identified so far. Limitations were identified among some case-control and cohort studies of leukemia as the following: 1) sample sizes were generally small (after excluding the 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. multiple probands), which resulted in reduced power; 2) controls were not randomly selected and were not matched on confounding factors such as age, gender and race; 3) malignancy history among family members, especially second-degree relatives and other extended family members, was obtained through interview and not confirmed by medical records; 4) different neoplasm types such as lymphoma, leukemia, etc., were lumped together in both selecting the cases and observed endpoints among family members; 5) very few case-control studies have examined childhood leukemia; and, 6) environmental factors such as social economic status, rural or suburb communities, etc. have usually either not been collected or have been reported incompletely. Larger controlled and carefully designed studies are required to address the excess risk of childhood leukemia. Acquired translocation breakpoints, screening candidate tumor suppressor genes, apoptotic pathways, cell cycle regulators, cytokines, transcription factors, oncogenes as well as testing for association of germline mutations (such as ATM gene in the hereditary syndrome ataxia telangiectasia) with leukemia are some potential future directions for searching the new answers for the etiology. Leukemia in Twins Increased Concordance for Leukemia in Twins About twenty percent of identical twins developed leukemia if one twin developed AL before the age of 6. Shared genetic composition and intrauterine 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. insults or blood-bome metastases provide possible explanations for the high concordance rates of leukemia pairs among twins. Despite attempts to identify a leukemogen that might have been shared by the same family, few have been found (Milvihill et al cited by Kende et al, 1994). Danis et al (1982) summarized five studies conducted in the 1960s and early 1970s, and estimated that the concordance rate for ALL among monozygotic twins (MZ) ranged from 17% to 25%. The most concordant leukemia subtype was ALL. The concordance rate decreased with increased age of the twins. Buckley et al (1996) studied 76 MZ twin pairs with leukemia. Three pairs were concordant, resulting in a 5% concordance rate among children from birth up to 20 years of age. Two pairs were ALL and diagnosed at 13 months and 3 years respectively. The other pair was diagnosed with AML by age 2 years. The low concordance rate compared to previous studies (17% to 25%) suggests a smaller overall heritable genetic contribution to childhood leukemia or difference in methodology. In this study, three additional twin pairs known to be concordant for leukemia were excluded from analyses since the families could not be contacted and interviewed. After including these three pairs, and assuming that the proportion of these twins that were MZ and had leukemia was the same as for those who returned the questionnaires, the case-wise concordance for leukemia would have been 10%. Pearson et al (1963) reported that one pair of male twins developed ALL at age 3. The mother had a single x-ray exposure at 7 months of pregnancy. Based on 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. their common placenta, strong physical resemblance and 21 blood antigens being the same, the twins were more likely identical than fraternal. The authors summarized ten pairs of concordant leukemia (including the present study) among identical twins from 1938 to 1963. Seven of the ten pairs (6 female pairs and 1 male pair) had onset under 11 months, ranging from 4 to 10 months. Six out of seven infant pairs had acute leukemia — , 3 ALL pairs, 1 AML pairs, 1 pair with acute leukemia unspecified, and the sixth pair with AML in one twin and acute leukemia unspecified in the other. The seventh pair had subacute lymphocytic leukemia. The remaining 3 pairs (2 male pairs and 1 female pair) developed leukemia from 2 to 4 years of age. All three of these pairs were diagnosed with ALL. Three mothers of the twin leukemia patients had single X-ray exposure at 7, 7 and 8 months of pregnancy. The authors of this study carefully examined the twin status of the two ALL patients and concluded that they had 94.4% probability of being identical twins. In the 1960’s ascertainment of twin status were based on the fetal membranes, physical features, blood groups, and finger prints. Since the subjective judgment of physical features, and variation of other characteristics (e.g., MZ or DZ twins could have either one placenta, membrane, or chorion, or have two), the reported concordant pairs of leukemia could over-estimate the concordance rates. Of the eight pairs with documented fetal membranes, at least six had one placenta, and 5 of these six were infant leukemia pairs. These data provide some indirect evidence that leukemia cells 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. may have been transferred from one twin to the other. Even so, the mechanism of initiation of leukemia cells in first twin was not known. A Transfer of a Malignant Cell from One Twin to Another Ford et al (1997) studied identical twin boys who were diagnosed with T-cell nomHodgkin’s lymphoma (T-NHL) and acute T-cell lymphocytic leukemia (ALL) at ages 9 and 11 years. Both twins had bi-allelic rearrangements of T-cell receptor p, y (TCRp, TCRy) with identical 11 bpN regions suggesting identical clonal rearrangements. The results suggested that 1) some pediatric T malignancies are initiated in utero and latency can be 10 years or more; 2) there can be a common biological origin in T-cell precursors (from the same single cell) for different types of malignancies (i.e. NHL and ALL in this study); 3) leukemia developed prenatally in one twin and then transferred to the other twin, similar to transplant of cells or organs between genetically identical subjects (i.e. twins, highly inbred mice, etc.) or cancer cells to immunosuppressed recipients. This study could not prove or reject the hypothesis that a single T cell in one twin transformed or mutated with TCRp, TCRy rearrangements and then prenatally metastasized to another twin via intraplacental anastomoses through the sharing of a monochorionic placenta. Candidate genes for T-NHL/ALL such as p53 mutations (exons 5-8), TAL deletion, TCL-1 rearrangement were not detected in these twin patients. Campbell et al (1996) studied identical twin females diagnosed with pre-B ALL at 9 months who demonstrated different clinical responses to therapy and had 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. different clinical courses. One twin (twin 1) relapsed 21 months after treatment and died 6 months after the first relapse. The other twin remained in complete remission after finishing 2 years of therapy and during a 28 months post treatment follow-up. The twins shared acquired clonal rearrangement of the MLL gene which indicated that the concordant leukemia originated from a single cell in utero. However, karyotypic analysis of twin 1 revealed 47 XX inv(3)(q21;q26), 1(4:1 l)(q21:23), +8 which was not observed in twin 2. Unless these abnormalities were acquired, the poor outcome for twin 1 could not be totally explained by the different rate of evolution of leukemia cell between the twins. This study supported the idea that these identical twins shared a common leukemia clone and MLL gene rearrangements and also suggests that despite a common clonal origin in utero, post-natally these leukemia cells can evolve independently. Super et al (1994) studied one pair of female identical twins diagnosed with ALL at birth and 3 months. Twin A was diagnosed at birth and died several hours later. Twin B was diagnosed at 3 months, was treated with aggressive chemotherapy and remained in remission for 47 months after diagnosis. Identical, somatic rearrangements of MLL gene located at the 1 lq23 breakpoint and a t( 11; 19) (q23;pl3.3) were detected from peripheral blood cells of both twins. Although the twins had separate placentas and were diamniotic and dichorionic, they were diagnosed as identical twins by having identical RFLPs (restrictive fragment length polymorphism) for eight markers for which the parents were heterozygous. Together 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with the same father’s X chromosome and the same ABO blood group, the probability that these twins were monozygotic was 0.998. It was concluded that a B-cell precursor or hematopoietic stem cell was transferred from twin A to twin B or vice versa in utero even though they did not share placental circulation. The leukemia cells from twin A were transferred either to the maternal circulation and then back to twin B, or through rare placental anatomosis. Mahmoud et al (1995) described a pair of female monozygotic twins who both developed pre-B ALL at two months of age. Rearrangements in the HRX gene at 1 lq23 were absent during one twin’s remission of the disease and also absent in another twin’s postmortem liver sample. Rearrangements in the HRX gene, which were present at diagnosis of ALL in both twins were absent in twins' parents, sister and in the germline of a control. The results suggested an acquired chromosomal abnormality or polymorphism, and supported the theory of intrauterine single cell origin of leukemia cell and transmission from one twin to the other. The etiological mechanisms responsible for inducing 1 Iq23 translocations during pregnancy were not identified. Occurrence of Leukemia in Twins Compared to the General Population Rodvall et al (1992) reported that no increased childhood cancer was found among 35,582 Swedish twins under age 16 in the period of 1952 and 1967. All cause mortality among twins was substantially increased (0/E=3.7, 95% Cl: 3.6-3.8) 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and cancer incidence was significantly reduced among males twins under age of 5 (O/E=0.3, 95% Cl: 0.1-0.7). Seventeen leukemia patients (9 males and 8 females) were found among the twins compared to 18.1 expected from the general population (O/E=0.9, 95% Cl: 0.5-1.5). Similarly, 19 deaths were observed among twins compared to 21.6 expected (O/E=0.88, 95% 0=0.5-1.4). No significant difference was detected after the sex, age, calendar year of diagnosis were adjusted. A pair in which one twin had leukemia and the other twin had a brain tumor was reported. No concordant leukemia pairs were found. This study was a large population-based study. All diagnoses were confirmed either by Swedish Cause-Death Register or by Cancer Register. The reason for the observed deficit of childhood cancer among twins is not known. A higher mortality rate than expected was found among twins, attributable to a high mortality during the first year of life. A lower than expected cancer mortality and incidence among twins was found. Factors that may be related to the deficit include low birth weight, competing causes of death or fetal mortality, or sensitivity to environmental carcinogens. Inskip et al (1991) reported that leukemia occurred in thirteen twins, including two female concordant pairs among 30,925 twins bom in Connecticut between 1930 and 1969. The twin concordance rate for female leukemia was 18% (2 of 11 pairs) overall and 29% (2 of 7 pairs) for like-sex pairs. Among the thirteen cases among twins, eight were ALL, three AML, one acute leukemia, one lymphocytic leukemia. Both leukemia among twin pairs were ALL. As the Swedish 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. twin study reported, both leukemia and all other childhood cancer combined were observed to be lower than expected from the general population (SIR=0.8, 95% CI=0.4-1.4, and SIR=0.6, 95% 0=0.3-0.9 respectively). The deficit was greater among males, especially males under 5 years of age (SIR=0.5, 95% 0=0.2-0.8, and SIR=0.2, 95% 0=0.0-0.7, respectively). Single leukemic or pre-leukemia stem cells transferred from one twin to the other via a shared placental circulation explained the reported similarities in age of onset and cytogenetic evidence of monoclonality among MZ leukemia patients. Since data between 1930 and 1965 was used in this study, suggesting status was not determined for the two concordant pairs out of 7 like-sex ALL pairs, which made it difficult to compare this concordance rate (29%) with previous studies. This Connecticut study again found a deficit of childhood cancers among twins. The possible explanations included low birth weight among twins, and selective early mortality of twin fetuses or neonates who could otherwise have developed a childhood cancer. Swerdlow et al (1996) found that among the 1,063 twins with cancer in England and Wales cancer registry during 1971-1984, 114 childhood leukemia patients and young adult leukemia were reported. There was a borderline significant excess risk of leukemia among first bom twins compared to the second bom twins (OR=1.44, 95%: 0.98-2.14). This non-significant tendency remained when the analysis was stratified by gender (same or opposite gender of the twins), or by 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. leukemia type (lymphoid leukemia or myeloid leukemia). These results supported some of the previous studies. Possible explanations for these results included either 1) the first bom twins had a higher risk of ascending infection and infections acquired through birth canal, which resulted in increasing the risk of leukemia; or 2) the second bom twins had higher perinatal death because of poor placental implantation, severe anoxia, and low birth weight, which may reduce the number of surviving second bom twins, who might have developed leukemia at a later date. In summary, the higher than expected concordance rate for leukemia among twins, especially the nearly 100% rate for twin pairs less than 2 years old could be explained by a single intrauterine event via co-joined placental circulation, in which leukemia cells are transferred from one twin to the other. The second explanation for the concordant leukemia is in utero radiation exposure. Previous studies on the association of in utero radiation exposure and childhood leukemia are inconclusive and contradictory. High X-ray exposures in utero among twins (especially in early years prior to equipment improvements) appear to not increase leukemia risk among twins, relative to the general population (Swedish and Connecticut twin studies, Inskip et al, 1991). The third theory is that the leukemia cells from one twin were transferred to the maternal circulation and then transplanted to another twin. This theory was supported by a recent study which the identical twins did not share the placental circulation (Super et al, 1994). The fourth possible explanation involved environmental factors such as viral etiology. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The hypothesis of viral infection in the etiology of leukemia is based on the following observations: 1) MZ twins have different chromosomal abnormalities which are acquired; 2) MZ twin discordance of leukemia represents differential environmental effects on identical genome and acquired genetic changes associated with leukemia; and 3) leukocytes of leukemia patients do not have HLA antigens that can stimulate the lymphocytes of their healthy identical twins, indicating the postnatal leukemogenic influences. Environmental Factors for Childhood Acute Lymphocytic Leukemia (ALL) Ionizing Radiation High Dose in Utero/Postnatal Radiation The most clearly etiologic agent associated with acute leukemia in man is ionizing radiation. High level radiation exposure is clearly associated with an increased risk of developing leukemia. Japanese atomic bomb survivors had a probability of approximately 1 in 60 of developing leukemia within 12 years of that exposure. ALL occurred most frequently among those less than 15 years of age at the time of atomic bomb exposure, with the peak occurring within 8 years. Further evidence was found among patients with polycythemia vera or ankylosing spondylitis who were treated with ionizing radiation. One in six within 12 years in the former group and one in 720 within 25 years in the latter group developed leukemia. Chromosome fragility following the radiation treatment was observed in many patients. Together with the high degree chromosome fragility observed in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hereditary conditions such as ataxia telangiectasia, the importance of chromosome damage becomes more evident (Pendergrass et al, 1985). Based on the reports of committee on Biological Effects of Ionizing Radiation (BEIR V), the risk of acute leukemia is increased by irradiation to hemopoietic cells, the magnitude of the increase depending on the dose of radiation, its distribution in time and space, and the age of exposed individuals. The mean latent period preceding the clinical onset of the leukemia also varies, depending on the hematologic type of the disease as well as age at the time of irradiation. The dose-mortality (dose-incidence data were not available) curve for the total excess cases of leukemia appears to increase in slope with increasing mean dose to the marrow. Life Span Study among A-bomb survivors (LSS) demonstrated that risks are initially higher for those exposed at under 20 years of age but decrease somewhat more rapidly with time after exposure than for those exposed at older ages. There was no clear indication that the risks for those under 10 years of age were significantly greater than for persons 10-20 years old at the time of exposure since there were few cases. Yoshimoto et al (1990) summarized epidemiological studies of cancer risk among the children of atomic bomb survivors conducted at the Radiation Effects Research Foundation. These children included two groups: (1) the in utero-exposed children; and (2) the FI population, which was conceived after the atomic-bombings and bom to parents of whom one or both were atomic bomb survivors. Two leukemia patients were identified among the 1,630 children exposed in utero (920 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. out of 1,630 received uterine dose >=0.01 Gy), compared to 31 leukemia among 15,895 children exposed at age of 0 to 9 years. Among the 41,066 in the FI population who did not receive radiation, there were 17 leukemia cases compared to 16 leukemia cases identified among 31,150 who received 0.01 Sv (Sv: sievert is defined as that producing the same biologic effect in a special tissue as 1 gray of high energy x-rays;) or more maternal or paternal radiation before conception. The small number of children exposed in utero and the small number of leukemia cases in this cohort precluded analyses by cancer site, different gestational stages at exposure or by time of paternal exposure before conception. Since no significant cancer risk under the age of 20 was observed with increasing dose, the effect of germline mutations induced by radiation exposure was extremely low. The absence of any remarkable increase in childhood cancer (including leukemia) mortality among Japanese A-bomb survivors may be the result of variable effects, and types of radiation, dose and confounding factors. Delongchamp et al (1997) analyzed cancer mortality from October 1950 through May 1992 among atomic bomb survivors exposed in utero. Risk estimates for this group were also compared to those survivors who were less than 6 years old at the time of exposure. The cohorts studied include 807 in utero survivors and 5,545 persons exposed during childhood, with members of both groups having estimated radiation doses of at least 0.01 Sv. The comparison group included 10,453 persons with little (<0.01 Sv) or no exposure. Analyses were limited mainly to cancer deaths occurring between the ages of 17 and 46. There were significantly increased 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. numbers of leukemia cases among those exposed to radiation at age 6 or younger compared to controls (P<0.0001). Although there were only two leukemia deaths among those exposed in utero, the leukemia death rate for this group was five times higher than that in the comparison group (P = 0.054). One male died of lymphoid leukemia at age 30 with an estimated 0.04 Gy gamma rays exposure. A female died of myeloid leukemia at age 18 with 0.023 Gy gamma rays exposure. However, the data did not provide evidence of a dose response among those exposed in utero because no high-dose leukemia deaths were observed, a result that differs considerably from those exposed as children. There is a need for caution in the interpretation of these data. First, the number of cancer deaths is small; second, there is an unexplained significant difference in the mortality from solid cancer between the sexes; and third, the excess of leukemia in those exposed in utero is not reflected in an increasing dose response. Low Dose in /Utero/Preconception/Postnatal Radiation The association of low-level radiation exposure with ALL is less clear. The classic study by Stewart et al (1958) showed a 2-fold increased leukemia mortality among children of mothers who received diagnostic radiographs during pregnancy based on confirmed diagnosis and radiation exposure in hospital records. Considerable debate continues on subsequent reports of associations of preconception, intrauterine, and occupational exposure to low level radiation and the development of childhood leukemia. 4 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shu et al (1988) conducted a population-based case-control study of 172 childhood ALL cases and 344 healthy population control children in urban Shanghai, China. Excess risk for ALL was associated with intrauterine (OR=1.6, 95% CL 0.9- 2.8) and paternal preconception diagnostic x-ray exposure (OR=2.6, 95% CI:1.5- 4.6). Elevated risk of ALL was also found among children receiving six or more postnatal x-rays (OR=3.3 95% 0=0.7-15.9). However, this finding was based on a small number of cases. Shu et al (1994b) conducted a population-based case-control study of 166 childhood acute leukemia cases and the same number of matched controls in Shanghai, China. Prenatal X-ray exposure was found to be associated with a 2.4 fold (95% Cl: 0.5-10.6) increased risk of developing childhood acute leukemia. Post natal X-ray exposure was also linked with a 1.6 fold (95% Cl: 1.0-2.6) elevated risk of developing childhood acute leukemia. The increased risk associated with more frequent post-natal x-ray examinations was marginally statistically significant (P=0.07). Little evidence, however, was found relating either maternal or paternal preconception X-ray exposure with the subsequent cancer risk in offspring. The possible reasons for the conflicting findings on the effect of paternal preconception X-ray exposure between the present and the previous study (Shu et al, 1988) included methodology (source of exposure information): in 1994, as high as 92% of case parents and 96% of control parents were interviewed, compared to 83.8% of case mothers, 11% case fathers, 91.4% control mothers, 7% control fathers were interviewed in 1988. Therefore, the information on paternal preconception X-ray 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exposure was more accurate in the 1994 study because the exposure information was obtained from both fathers and mothers. The other explanations for the differences between the two studies included: effects of chance or higher exposure dose in 1950s/60s (previous study) vs. lower dose of X-ray examinations after 1970s in the present study. Shu et al (1994a) found a positive or marginally positive association between infant ALL and number of paternal X-rays of the lower gastro-intestinal (GI) and lower abdomen (trend test, P < 0.01), upper GI (trend, P = 0.04), limb (trend, P = 0.08) and chest (trend, P=0.08). The relative risks associated with 2 or more of lower or upper GI exams, or with 10 or more times of limb or chest radiation x-rays were estimated as 3.78 (95% Cl, 1.49-9.64), 2.71 (95% Cl, 0.99-7.44), 1.41 (95% Cl, 0.65-3.05) and 2.21 (95% Cl, 1.0-4.9) respectively compared to those with no exposure to one of the above four sites. Gardner et al (1990) studied leukemia and Non-Hodgkin’s lymphoma (NHL) incidence in young people near the Sellafield nuclear facility in West Cumbria in a case-control study. The association of leukemia among children and their fathers’ occupational exposure to ionizing radiation was studied. It was reported that total paternal occupational exposure to radiation before conception greater or equal to 100 mSv was associated with 6.3 (95% Cl: 1.52-26.03) and 8.38 (95% Cl: 1.35-51.99) increased risk for leukemia under age 25 (4 out of 46) compared to their area (unmatched) (5 out of 288) and local controls (matched by sex and date of birth) (3 out of 276) respectively. Paternal exposure to radiation 6 month before conception Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. greater or equal to 10 mSv were associated 7.38 (95% Cl: 1.74-31.31) and 6.82 (95% Cl: 1.46-31.86) increased risk for leukemia (N=4) compared to their area (N=5) and local controls (N=4) respectively. Among children of fathers employed at Sellafield at their conception, the relative risks of developing leukemia compared with area and local controls were 2.79 (95% Cl: 1.04 to 7.52) and 2.07 (95% Cl: 0.69 to 6.14) respectively. Among the five Seascale leukemia cases (living within a 5 km-radius from Sellafield nuclear plant), four leukemia cases received the highest paternal radiation doses (97, 102, 162, and 188 mSv). The paternal exposure of the fifth leukemia case was not documented. Three of the four leukemia cases with paternal exposure before conception were ALL. Kinlen et al (1993) disagreed that paternal preconception radiation was solely responsible for the increased leukemia in Seascale. A significant excess of leukemia at ages 0-24 was found in Seascale in the years 1951-91 in those who were bom there (ratio of observed to expected cases 8.6 ; p<0.01). This also applied to those not bom there (7.2; p<0.01). It was concluded that paternal preconception radiation cannot be the sole cause of the excess in Seascale since it would not explain the excess among those bom outside Seascale. His alternative hypothesis was that the true cause was a factor (such as infection) other than paternal preconception radiation. On this basis, the association found among those bom there, if not partly due to chance, may reflect an indirect relation with the tme cause. Stevens et al (1990) reported a weak association between bone marrow dose and all types of leukemia, all ages, and all time periods after exposure among 1177 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. individuals who died of leukemia (cases) and 5330 other deaths (controls) between 1952 to 1981 in Utah radioactive fallout from NTS study. The greatest excess risk was found in those individuals in the high-dose group with acute leukemia who were younger than 20 years at exposure and who died before 1964. There was a 5.8- fold of increased risk of dying leukemia between age of 0-19 years if exposed at higher doses (6-30 mGy) compared to be exposed at lower doses (0-2.9 mGy). There was a 5.3- fold increased risk of dying ALL if exposed at higher dose (6-30 mGy) compared to those exposed at lower dose (0-2.9 mGy). Using mortality data may be subject to accuracies related to cause of death coding and introduce uncertainty in estimates of risk, especially after year 1968 when leukemia was not uniformly fatal within 3 year of diagnosis. Nevertheless, this study indicated a dose related trend for childhood acute leukemia in the earliest time period and was consistent with previous studies that individuals exposed at young age, and high dose were at higher risk of leukemia, especially acute leukemia, such as ALL. Petridou et al (Greece, 1996) reported that infants throughout Greece under 12 months exposed in utero to ionizing radiation from the Chernobyl accident had 2.6 times the incidence of leukemia, compared to unexposed children (95% Cl, 1.4 to 5.1; P =0.003). It was also reported that children bom to mothers residing in regions with high radioactive fallout were at higher risk of developing infant leukemia. No significant difference in leukemia incidence was found among children 12 to 47 months of age. Preconception radiation had no demonstrable effect on leukemia risk in any of the studied age groups. Exposure of the Greek population to Chernobyl 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. radiation started soon after the accident and was notable for about a year; the average exposure has been estimated at about 2mSv. However, the dose that each child received in utero based on his/her residence location was not estimated. The trend of increased risk of infant leukemia with increased dose of radiation in utero could not be examined. The histological types of these infant leukemias were not reported in this study. The small sample size of the twelve infant leukemia cases precluded a stratified analysis by leukemia type, such as ALL or AML. A similar study carried out in Finland concluded that the incidence of childhood leukemia did not increase over the period 1976-92 and that the excess risk during the period 1989-1992 was not significantly different from zero. (Auvinen et al, 1994). Tables 4 and 5 summarize the studies of association between low dose radiation and leukemia. In conclusion, radiation may alter genetic material at chromosomal and molecular levels. DNA may be altered, lost, or added during the process of repair or recovery from the damage. These changes may result in the development of leukemia, along with the genetic predisposition or altered susceptibility to other environmental agents. There is no consistent evidence suggesting that children in utero at the time of the atomic bomb in Japan had increased incidence of ALL during childhood. Also, no consistent association has emerged between the use of diagnostic radiation postnatally and ALL (Neglia et al, 1988). Studies included in this text of literature review are almost all positive except for preconception radiation which is controversial. Other negative studies are included in Table 4, 5. The inconsistent findings of the low dose studies reviewed 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. above suggest that the association of leukemia and diagnostic radiography remains uncertain. Studies with positive findings have usually found a 1.5 to 4-fold increased risk among those exposed to low dose radiation. The findings suggested a positive, but weak association between radiation and leukemia. And these findings may reflect effects of chance, or possible confounding factors, bias and exposure measurements which could be summarized as: 1) the excess radiography may have been for preleukemic conditions; 2) the association between leukemia and radiography may have been confounded by the conditions under which x-rays were taken; 3) possible selection and recall bias; 4) different populations have different genetic predisposition to radiation exposure; and 5) uncertainty of exposure estimation and wide range of doses administered for each type o f radiography examination. In summary, over the past decades, epidemiologic studies of childhood leukemia have provided little consistent etiologic information, including the association between low dose radiation and childhood leukemia. In most studies, relative risks were not evaluated for specific histological types of leukemia (such as ALL), or immunophenotypes (i.e. T or B cell ALL), or age-specific groups (such as infant leukemia), much less focus on the genetic susceptibility in the etiology. However, important etiologic clues might just exist in these homogeneous subgroups (Ross, 1999). Although no studies have been carried out to provide evidence that the frequency of mutations in ataxia telangiectasia mutated (ATM) gene might be 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. substantially different in the different population studied, it could be one of these potential genetic factors to explain part of the inconsistent findings. Ataxia telangiectasia (A-T) patients who are carriers of two copies of mutated ATM gene and the presumed carriers of one copy of the mutated gene (parents of the A-T patients) have been found to be more sensitive to radiation than those who do not carry the mutated gene (Paterson, et al, 1979). The proposed study will explore the association of carrying a mutation in the ATM gene and the development childhood ALL. The proposal for the study is discussed in Chapter 4. Other Environmental Factors Studies examining the effects of electromagnetic fields (EMF), electricity transmission and distribution equipment on childhood leukemia have reported inconsistent findings. One of recent case-control studies found no difference in EMF exposure levels and duration (Linet et al, 1997). Paternal occupation, preconception use of alcohol, paternal exposure to plastic resin fumes, solvents, maternal smoking, consumption of alcohol, nicotine and medicaments, ultrasound examinations during pregnancy, hormone-assisted conception or in vitro fertilization, maternal reproductive history have also been explored in previously epidemiologic studies without conclusive findings (Linet et al, 1985, 1997; Ross, 1999). Infectious agents, particularly viruses, have long been suspected of causing leukemia. Despite extensive efforts, a viral origin for leukemia has not been 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unequivocally demonstrated. However, human retroviruses have been identified and characterized in certain rare T-cell lymphoproliferative malignancies. There is no evidence that childhood ALL is associated with either prenatal or postnatal infectious disease exposures (Linet et al 1985, Neglia et al, 1988). Immune deficiency (immunologic abnormalities) has been associated with susceptibility to neoplasia of lymphoid tissue (Kurita et al, 1974). The evidence includes: 1) immunoglobin abnormalities reported in an AL father and his son; 2) deficiency in levels of IgG and IgM among leukemia patients and nonleukemia siblings; 3) increased incidence of leukemia among patients with primary immunodeficiencies; 4) various drugs, including alkylating agents and other types of chemotherapy associated with leukemia (particularly ANLL), believed to be caused by the immunosuppressive activity of these agents (Linet et al, 1985); 5) Bacillus Calmette-Guerin (BCG) vaccination, widely used for human tuberculosis and shown to be an immunologic adjuvant and antitumor agent, may significantly reduce incidence or mortality from leukemia (Linet et al, 1985). The only chemical agent that is likely to produce excess leukemia risk is occupational exposure to benzene (Pendergrass, 1985). However, this exposure can only account for small proportion of adult leukemia. The following factors have also been studied but revealed no consistent findings (Pendergrass, 1985, Neglia et al, 1988, Peters et al, 1994): birth order, birth weight, advanced maternal age, paternal age, social class, maternal reproductive 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 4. Summary the results of association between radiation and adult/childhood leukemia in studies before 1990 (BEIR V) Study Radiation exposure Risk estimate Comments Diagnostic radiography on adult chronic myeloid leukemia (CML) Preston-Martin et al 1988 x-ray examinations of the back, gastrointestinal tract (GI) and kidneys 3-20 years prior to diagnosis O R =1.0,1.4, 1.7 and 2.4 for 0-0.99,1-0.99, 10-19.99 and >20 Gy (P<0.05 for the highest exposure category) Support the association between CML and x-ray examinations, and with the frequency of exams found in two previous studies. Preleukemic conditions and other medical conditions were not found to confound the observed association. Low dose radiation and ML among children and adults, Linos et al, 1980 x-ray examinations No positive findings Controls were nonleukemia patients and visited the same clinic as cases at two times. Small sample size and inappropriate selections of controls who may have received more x-ray examination than the general populations might explain the negative findings. Cancer (-including childhood leukemia under 15) among residents down wind from Nevada nuclear weapon test site Lyon et al, 1979 Fallout from nuclear weapon testing to nearby residents About 2.4 times excess death of childhood leukemia were found in 'exposed cohort1 compared to 'unexposed cohort' in 'high exposed' counties Lower death rate among the 'unexposed cohort' than that of U.S. population and lack of excess death from other childhood cancers suggest that the finding might have been an artifact of diagnostic error. Further more, the ’ high exposed' counties were not, in fact higher than the 'low exposed' counties. Leukemia among residents down wind from Nevada nuclear weapon test site among Mormon families Johnson et al, 1979 Fallout from nuclear weapon testing to nearby residents The observed and expected of leukemia incidence was 3 1/7=4.4 Not medically confirmed, self-reported cases and using volunteers gathering data are the weak points of the study. Chance variation may explain the findings. Leukemia from the three counties southwestern Utah Machado et al, 1987 Fallout from nuclear weapon testing to nearby residents The observed and expected of leukemia mortality for all ages 62/42.8=1.4 and 9/3.2=3 for 0-14 years old Although there appeared to be a small excess in these three southwestern Utah counties, the causality of fallout exposure could not be assessed from the study. Leukemia among participants of nuclear test explosion Smoky Caldwell et al, 1980 Fallout from nuclear weapon testing to participants Significant increase in leukemia incidence with Observed and expected incidence 10/4=2.5 Following study by Robinette et al confirmed the findings but the death from leukemia was not significantly increased compared to the expected. The excess cases of leukemia could be explained by chance variations. However, the association may be real and reflect an underestimation either the doses or of the risk per unit since leukemia occurred the most frequently in two most heavily exposed groups. U \ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 4 (continued). Summary the results of association between radiation and adult/childhood leukemia in studies before 1990 (BEIR V) Study Radiation exposure Risk estimate Comments Leukemia among participants of nuclear test -British Study Darby et al, 1988 Fallout from nuclear weapon testing to participants Increase in leukemia incidence with observed and expected incidence 22/6=3.67 Death from leukemia was not significantly increased compared to the expected based on the national rate Leukemia from global fallout Archer et al, 1987 Global fallout across the U.S. Fallout peaked in 1957 and 1962. Death rate for acute and myeloid leukemia in children aged 5-9 years rose to an initial peak in 1962 and secondary peak in 1968. No such peak was observed for other types of leukemia. Leukemia death rate for all ages and all cell types peaked during 1960 to 1969 and consistently highest in states with high “ Sr levels. Other three studies in Europe reviewed by Darby et al suggested that the increased incidence of childhood leukemia may be attributable to improvements in diagnosis and no convincing evidence of association with fallout. Leukemia among individuals near British (near Sellafield) nuclear reprocessing plants Gardner etal, 1984, 1987 Radiation from nearby nuclear installations Six (6) leukemia deaths in children ages - 24 in 1968-1974 compared to 1.4 expected with O/E ratio=4.3. Five (5) death from leukemia compared to 0.53 expected among children bom to women resident near the plant in 1950 to 1983. The viral infection among large number of new workers to this area proposed as a causative factor for excess childhood leukemia. (Kinlen et al, Watson et al) The viral infection may also explain the five cases of leukemia in children aged 0-24 compared to 0.5 expected near Dounreay nuclear reprocessing plant. Leukemia incidence and mortality near nuclear installations during 1959-1980 in UK Cook-Mozaffari et al, Forman et al, 1987 Significant overall excess of cancer mortality due to lymphoid leukemia and brain cancer in children. Unusually high mortality ALL near Sellafield was identified by using a Geographic Analysis machine to free cancer clusters from bias due to selection (Openshaw et al). However, the mortality rate in the control area were lower than expected, there has not been a general increase in cancer rates in the area. And no significant trend in cancer rates with distance from nuclear installations. Beral et al also reported significantly elevated leukemia incidence and all cancer in children in all exposed area combined titan those in control areas whereas the mortality was not. Cook-Mozaffari confirmed the observation of Beral et al, but suggest the difference may be due to a variation in case registration, possibly owing to social class differences. V i os Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 5. Summary of recent studies on association between radiation and childhood leukemia Author/ Disease Preconception-paternal OR (95%CI) Preconception-maternal OR (95%CI) Prenatal (Intrauterine) OR (95% Cl) Postnatal OR (95%CI) Shu, 1988 /ALL, age 0-14 (N=172 cases, 344 controls) 2.6(1.5-4.6) (>=11 vs. none) l.l(0.3-2.8) (>=11 vs. none) 1.6(0.9-2.8) (ever vs. never) 1.3 (0.7-15.9) (6+ vs, none) Shu, 1994b/ Acute Leukemia Age 0-14 (N=166 pairs) 0.7 (0.3-1.4) (>=10 vs. <5) (Lifetime) 0.9 (0.5-1.7)(>=2 vs. none) (2 year prior) 1.5 (0.6-3.4) (>=10 vs. <5) (Lifetime) 0.9 (0.5-1.8) (>=2 vs. none) (2 year prior) 2.4(0.5-10.6) (ever vs. never) 1.6 (1.0-2.6) (ever vs. never) 2.0 (0.8-5.0) (3+ vs. none) Shu, 1994a/ Infant ALL 0-18 months (N=191 for paternal exposure) (N=203 for paternal exposure) Total x-ray: 1.1 (0.4-2.8) Chest: 2.2(1.0-4.9) (10+ vs. none) Limb: 1.4(0.7-3.1) (10+ vs. none) Upper GI: 2.7 (1.0-7.4) (2+ vs. none) Lower GI; 3.8 (1-5-9.6) (2+ vs. none) NS Gardner, 1990/ Childhood leukemia Age: 0-24 (N=52) Employed at conception: 2.8 (1.0-7.5) (area); 2.1 (0.7-6.1) (local) Dose recorded at conception: 3.7 (1.3-10.8); 2.9 (0.9-9.1) >=100 msv before conception: 6.3 (1.5-26); 8.4 (1.4-52) >=10 msv 6 month before conception: 7.4 (1.7-31.3); 6.8 (1.5-31.9) NS 1.2 (0.3-4.3) (Area controls) 1.2 (0.3-4.7) (Local controls) Kinlen, 1993/ Childhood leukemia Age:0-24 0/E=8.6 (P<0.01) For those bom in Seascale 0/E=7.2 (P<0.01) For those not bom in Seascale < 1 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 5 (continued). Summary of recent studies on association between radiation and childhood leukemia Author/ Disease Preconception-paternal OR (95%CI) Preconception-maternal OR (95%CI) Prenatal (Intrauterine) OR (95% Cl) Postnatal OR (95%CI) Petridou, 1996/ Infant leukemia Age: 0-12 months (N=12) NS 2.6 (1.4-5.1) Yoshimoto,1990/ Age: 0-40 (N=1630 exposed in utero) (N=31150 exposed to parental preconception exposure) 1950-1984 16 cases of leukemia were observed among children whose parents received 0.01 Sv or more exposure compared to 17 leukemia observed among those whose parents received 0 Sv ATB 2 cases were observed among those exposed in utero, compared 31 cases exposed at 0-9 years of age ATB) Relative risk at 1 Gy: 3.77 for all cancer Relative risk (compared at 1 Gy: 17.25 for leukemia 2.23 for other cancers Delongchamp, 1997/ Age: 17-46 (N=807 exposed in utero) (N=5545 exposed at age 5 or younger ATB) (N=10453 persons with <0.01 Sv) 1950-1992 2 leukemia (one LL, one ML) death were observed among those exposed in utero compared to 4 case in controls 0.49 death per 104 PY/0.10 deaths per 104 PY=5 (P=0.054) (95% upper Cl: 18.7) (no dose response) 24 leukemia cases among those exposed at 0-5 years of age ATB) (show a dose response) 0 0 history, maternal oral contraceptive use, occupational exposure of parents, hydrocarbon and petroleum products, pesticide, herbicide exposure and N-nitroso compounds (NOC). In summary, over the past three decades, epidemiologic studies of environmental etiology of childhood leukemia revealed inconclusive findings. Environmental factors are thought to be important but no causal relationships with leukemia have been established except for radiation. It is believed that genetic factors do affect the occurrence of leukemia but these genetic factors are not strong enough to be a single component. The interaction of environmental and genetic factors could be more important and stronger than either one of them individually, but the extent and mode of inheritance are not clear. The evidence suggesting an inherited component in leukemia includes: 1) Patients with inherited syndrome are susceptible to leukemia; 2) Familial occurrence of leukemia, although its genetic basis is not certain. The immune deficiency cancer registry reported that 17% of patients with immune deficiency - component of inherited syndromes such as A-T developed leukemia. Whether this is associated with immune response, radiosensitivity, or some other mechanism is unknown. Familial aggregation of leukemia associated with chronic exposures such as benzene could be found only in a few cases. In the majority of familial occurrences of leukemia, a genetic association is suspected, but not either proven or disproven (Hamden, 1985). Environmental factors could not totally explain familial aggregation since other families could also have been exposed to similar environmental factors but not 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. developed multiple cases of leukemia. Conservatively, it could be concluded that genetic factors at least play partial role in the etiology of childhood leukemia. Studies focusing on specific, more homogeneous rare groups, and searching for interaction of both environmental and distinct hereditary abnormalities involving gene mutations, polymorphism might help to formulate new hypotheses and provide new answers. Ataxia Telangiectasia (A-T) Ataxia telangiectasia is an autosomal recessive syndrome in children characterized by progressive cerebellar ataxia and culocutaneous telangiectasia. Patients with this syndrome also have an unusual hypersensitivity of fibroblasts and lymphocytes to ionizing radiation. There is a 61-fold and 184-fold increased cancer incidence among White and African American patients, respectively. Non-random chromosomal rearrangements in lymphocytes, thymic hypoplasia with IgA and IgG2 immunodeficiencies have also been observed among A-T patients. Dysfunctional DNA processing or repair protein is suspected among A-T patients (Paterson et al, 1964, Taylor et al, 1975). Lymphomas and lymphocytic leukemia are the most common malignancies among A-T patients, comprising 60% and 27%, respectively (Morrell et al, 1986). Other primary carcinomas of the ovary, oral cavity, breast, pancreas, liver and stomach have also been reported. The etiology of high cancer risk among the A-T patients remains unknown. Candidate mechanisms include: 1) immune defects, 2) 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. random chromosome breaks and stable translocations, 3) post irradiation abnormalities, and 4) sensitivity to gamma rays, X-rays and cytotoxic drugs. Some of these mechanisms have been demonstrated among heterozygotes, such as sensitivity to irradiation, although not as striking as among homozygotes (Gatti et al, 1985, 1988, Hecht et al, 1990). Several studies have estimated the risk of cancers among A-T relatives. One study found that five relatives died of leukemia or lymphoma before age 45 among 27 A-T families compared to one expected death (P<0.01). The relative risk of dying of leukemia and lymphoma before the age of 45 among ATM heterozygotes was estimated to be 7 times (P<0.03) the risk in the US white population. It was estimated that ATM heterozygotes might comprise over 5% of all persons dying from a cancer before the age of 45 (Swift et al, 1976). Based on the estimated 1.4 to 2.8% frequency of ATM heterozygotes in the general population and a relative risk of breast cancer among these ATM gene carriers of 6.8-7.6 (Swift et al, 1986, 1987), it is estimated that 8.8-18% of all patients with breast cancer are ATM gene heterozygotes. Lymphoid malignancy occurred at a higher rate among White non-Amish relatives than spouse controls (13 vs. 1). However, no firm conclusion could be made (Swift et al, 1986,1987). Cancer incidence among obligate heterozygotes of 110 White non-Amish A- T patients was compared to spouse controls in a prospective study. An estimated 2.3 and 3-fold relative risk among men and women of developing cancers (all types 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. combined) was found among ATM heterozygotes relative to noncarrier spouses (Swift etal, 1990). Cancer incidence among 1,599 relatives and 821 spbuse controls of 161 White A-T patients were compared. Among 294 ATM heterozygotes, the estimated rate ratio was 3.8 among men and 3.5 among women. The odds ratio of exposure to diagnostic x-ray, therapeutic, or occupational radiation to the development of breast cancer among A-T relatives was 5.8, compared to controls matched on year of birth (Swift et al, 1991). Cancer incidence among 574 relatives and 213 spouse controls of forty-four (44) A-T families was compared. The relative risk of developing cancers among ATM heterozygotes was 6.1 relative to spouse controls. There were five (5) cases of leukemia and lymphoma (two lymphomas, two CLL, one ALL) observed when only three (3) were expected. However, the excess of leukemia and lymphoma was not significant (Morrell et al, 1990). Easton (1994) reviewed two U.S. studies by Swift et al (1987 and 1991), a U.K. study by Pippard et al (1988), and a Norwegian study by Borresen et al (1990). A meta-analysis of the data from these four studies yield an estimate of the allele frequency to be 0.2-1%, with 0.5% being the best point estimate. The pooled estimated relative risk of breast cancer was 3.9 (95% Cl: 2.1-7.2) and 1.9 (95% Cl 1.5-2.5) for any other cancers. In summary, the association between ATM heterozygotes and leukemia remains unclear. In studies by Swift et al (1976, and 1987), leukemia and lymphoma 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were aggregated among A-T relatives (Swift et al, 1976, 1987, Morrell et al, 1990). In the other studies reviewed above, the risk of leukemia was not estimated. Since the ATM gene was not cloned until 1995, it was not possible to determine ATM gene status among A-T relatives in these previous studies. Furthermore, the estimation of leukemia and lymphoma was based on those relatives older than 20 years of age. The exclusion of childhood leukemia and lymphoma may have resulted in an underestimation of the risk since children are at greater risk of developing acute leukemia compared to an adult aged 25 to 65. Finally, the use of spouse controls was questionable, given that married individuals are known to have lower incidence for certain cancers than the general population. However, Easton (1994) demonstrated that this bias can not explain all of the excess risk observed. In conclusion, more cancers occurred among A-T relatives than was observed among the general population. Leukemia is one of the most common malignancies among A-T patients who are carriers of two ATM mutations. Leukemia was found more among A-T relatives and among presumed carriers of one copy of the mutated gene (parents of the A-T patients) although this elevated risk was based on small numbers of leukemia cases. The risk for childhood leukemia among A-T patient siblings and cousins has not been estimated. It is reasonable to hypothesize that more ATM gene carriers may be found among ALL patients with a family history of cancers for any first or second degree relatives under age 45. Moreover, children carrying one copy of the mutated ATM gene may be at higher risk of developing ALL than randomly selected population controls. 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ATM Gene and Leukemia Since the ATM gene was cloned in 1995, six studies have analyzed ATM mutations among individuals having leukemia. Most of these studies have had small samples of T-cell pro-lymphocytic leukemia (T-PLL). Only one study has examined ATM mutations among ALL patients. Table 6 to Table 10 summarizes the findings of these ATM mutations among leukemia patients. The ATM gene was found inactivated in a rare sporadic malignancy, T-cell prolymphocytic leukemia (T-PLL), which is often associated with cytogenetic aberrations of chromosome 14. T-PLL is a rare form of mature leukemia that occurs both in adults as a sporadic disease and in younger patients suffering a hereditary condition, such as A-T. Both T-PLL and T-ALL developed among A-T patients. A high incidence of T-cell leukemia/lymphomas was observed in ATM-deficient mice. The ATM gene was shown to sustain frequent loss-of-function mutations in T-PLL tumor cells, consistent with its functioning as a tumor suppressor gene in this type of leukemia (Luo et al, 1998). Luo et al (1998) studied 19 T-ALL patients for ATM mutations. The 19 T- ALL patients were selected from a hospital in London. The average age at onset of the T-ALL cases was 21.9 years, ranging from 2 to 44 years old. These T-ALL patients exhibited rare nucleotide substitutions not previously found in ATM. Rare polymorphisms were found among six patients (32%). Five of the six missense mutations (26%) were identified in the germ-line, indicating constitutional 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. polymorphism. The racial background of these six patients included three Arabs, one Indian, one African and one European Caucasian. All six nucleotide changes included a single base substitution and none were found in the kinase homology domain in contrast with mutations in this domain dominant among T-PLL patients. Among 17 of 19 patients analyzed by a cytogeneticist, three had both chromosome 14 translocations and the rare ATM polymorphism. The method for collecting constitutional cells to study germline mutations was not described. The absence of somatic nucleotide changes (loss of function) in ATM gene among T-ALL patients as compared with T-PLL patients suggests different genetic events in the development of these two types of leukemia. The authors indicated these five germline mutations were also found in unaffected individuals, but did not indicate their racial identity. ATM gene mutations on exon 13 were examined among 95 unaffected Iranians and failed to demonstrate similar nucleotide substitutions in the germline. Thus, whether or not these 'rare polymorphisms' causally link to the development of T-ALL needs further investigation. The authors concluded that the existence of frequent point mutations or small deletions/insertions is uncommon among T-ALL patients. However, this conclusion needs additional empirical data to be verified. As the authors indicated, no small deletions/insertions were observed among these 19 patients but a point mutation rate as high as 26% was observed in the germline cells. The findings relating to germline cells are more meaningful to genetic predisposition than the somatic mutations found in the leukemic cells. 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Because results were based on a small sample of T-ALL patients, further studies are needed to confirm these preliminary findings. Yuille et al (1998) analyzed eight T-PLL samples. Structural lesions were found in all eight patients. Both alleles were affected in four of the eight patients. This provides strong evidence that ATM acts as a tumor suppressor during T-PLL tumorigenesis. Nucleotide changes were also present in addition to structural lesions. Two patients had 1 lq23 abnormalities. Nucleotide changes were detected in five of eight patients (62.5%) by SSCP. Three missense mutations were observed. Two of the three missense mutations were reported in the conserved PI-3K domain (exon 61 and 65 respectively). Only one of these point mutations was present in one T-PLL patient in a previous study. None of the other changes have been reported either among A-T patients or normal Caucasians. Two blood samples of patients in remission were collected. The missense mutations found in leukemic cells were not observed in the germline cells. An association of ATM gene mutation with T-PLL tumorigenesis is suggested since nucleotide changes present in addition to structural lesions disrupting both alleles were found. The mechanism of inactivation appeared to be unusual because multiple structural lesions on one allele were often observed. Stoppa-Lyonnet et al (1998) found ten of fifteen sporadic T-PLL had loss of heterozygosity (LOH) of the 1 lq22-23 region (67%). Frozen leukemic cells and Epstein-Barr virus (EBV)-transformed cell lines from blood samples, allowed the 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 6. Summary of ATM polymorphism detected among six T-ALL patients by SSCP (N=19) (Luo et al, 1998) * Patient /ethnic origin Cytogenetics (leukemic cells ?) Segment Wild type allele in tumor sample ** Nucleotide sequence change Predicted Protein change Presence of alternation in germline + Comment 11/Hindu NA exon 7 50 410A ^T Y137F Not analyzed Not informative for assessing LOH 18/Arab 46,XY ;45,XY, -5,10, d el(ll), +marker exon 7 50 378T-»A D126E + 4/Arab t (11;14) exon 13 50 1744T->C F582L + 10/Eur.C auc. t (1;14) Intron 16 50 1VS16+ 22A-»C ? + 15/Africa n 46,XY,-10, +marker exon 23 50 3118A^G M l 040V + This is a rare polymorphism in unaffected Gambians and Iranians. This substitution was also reported in a B- cell NHL 12/Arab Del (9)/ t(10;14) exon 41 0 5882A-»G Y1961C + * Six somatic mutations identified in tumor DNA (32%) were found in the germline among five T-ALL (26%) patients. ** Percentage of normal ATM copy in tumor DNA. 50% represents normal amount. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 7. Summary of ATM mutations detected in tumor samples of eight T-PLL patients (N=8) (Yuille et al, 1998) * Patient Segment Nucleotide sequence change Predicted Protein change Comment 1 Not detected 2 Not detected 3 exon 61 8668C->G L2890V in the conserved PI-3K domain reported before in a T-PLL case somatic origin (absent in remission sample ) 4 exon 65 9022C->T R3008C in the conserved PI-3K domain abnormality was also shown in exon 3 in PCR-SSCP analysis the tumor sample showed del( 11 )(q23-24) 5 intron40 exon 41 5763-22del31 Site splicing site loss The deletions removed intron 40 splice site and nine bases of exon 41 The mutation is predicted to lead exon-skipping. Resulted in a missense mutation? 6 exon 44 6116A->G E2139G (E2039G?) Somatic origin (absent in remission sample ) The tumor sample showed add(l l)(q23) 7 Not detected 8 intron 36 5177 + 44T->C ? The tumor sample had a trans location on chromosome 11 and the nucleotide of unknown significance * Mutation analysis by PCR-SSCP. PCR products from exon 3-65 in tumor samples were analyzed by SSCP to identify somatic mutations (62.5%) which were then sequenced. All eight patients had structural lesions. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 8. Summary o f ATM mutations by PTT in four T-PLL cases (N=15) (Stoppa-Lyonnet et al, 1998) * Patient /ethnic origin Segment Nucleotide sequence change Predicted Protein change Presence of alternation in germline Comment Big Exon 7 2891insT Truncated protein The frameshift created a stop codon 51 bases downstream Dia exon 7 6536dell05 (cDNA) IVS46+1G->A (leukemic gDNA) Truncated protein (Splicing aberrations) The donor site G~> A substitution caused exon 46 was skipped and created no frameshift and stop codon Close to PI-3K domain Bat Intron 28 4182ins29 Truncated protein ? (Splicing aberrations) The cause of 29 nt insertion remains to be determined Bui intron 16 7645C->G R2486G - PI-3K domain * Ten of the fifteen T-PLL (67%) were found loss o f heterozygosity (LOH). Somatic mutations were identified in four patients (44%) with LOH among seven (7) with LOH and two without T-PLL patients tested. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 9. Summary of ATM mutations among seventeen T-PLL patients (N=37) (Vorechovsky et al, 1997) * Sample Exon Nucleotide change Presence of the wild type allele Protein change Comment 5b3 15 2119del TGTGA Frameshift Truncated lc8 30 4174insC Frameshift Premature termination of the translation product 5b3 30 4220T->C 20% I1407T somatic origin (not present in remission samples) lb8 36 5044G-*C D 168211 lb2 40 5729T^A 15% L1910H Partial allelic loss at the ATM locus ld4 47 6490G-»A E2164K lc8 51 7187C-»G T2396S lclO 51 7271T-»G V2424G Reported previously in a A-T, putative A-T carrier la9 52 7325A->C Q2442P ld5 54 7636del TCTAGAATT S2546del R2547del,12548 del Reported previously in a A-T, a breast cancer patient with family history of multiple malignancies. Putative A-T carrier (represent the most frequent A-T allele). Truncated. 6c4 55 7880insT 25% Frameshift Partial allelic loss at the ATM locus la8 57 8084 G->C G2695A lb4 58 8165 T->G L2722R lb7 58 8174 A->T D2725V 5a6 58 8194 T ^ C F2732L Tla5 60 8430delAAA 50% K2810del Truncated, somatic origin (not present in remission samples) la l 61 8613delACA 50% R2871 S,I12872del Truncated, no allelic loss at the ATM locus 6b 1 61 8668 C-»G 40% L2890V no allelic loss at the ATM locus BJ01 65 9139 C~>T R3047X Nonsense, stop codon. Wild type was lost in tumor DNA but presence in a granulocyte-enriched fraction * Somatic mutations were identified in seventeen (17) T-PLL patients (46%) among 37 T-PLL tested. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 10. Summary of ATM mutations detected in one allele among six T-PLL patients who had a deletion in the other allele (N=24) (Stilgenbauer et al,1997) * Patient Segment Nucleotide sequence change Predicted Protein change Comment 1 exon la- 3 Deletions of 5' exons No ATM transcript Absence of ATM mRNA 21 exon 28 3873del 120 Del Vall292-Glnl331 loss 40 amino acids and the alternations close vicinity of its c-Abl binding domain 11 exon 37 5309C->G Serl770ter Nonsense mutation Premature truncation resulting in lacking the functionally important PI-3K domain 13 exon 58 8174A->G Asp2725Gly Missense mutation in the PI-3K domain 15 exon 65 9016G-»C Ala3006Pro Missense mutation in the PI-3K domain 3 exon 65 9022C^T Arg3008Cys Missense mutation in the PI-3K domain * Using direct sequencing of the corresponding PCR products. Somatic mutations were identified. Fifteen of the 24 T-PLL patients (62.5%) had small deleted segments at 1 lq22.3-23.1. Six o f these fifteen (15) patients (40%) with a deletion in one allele had mutations on the second allele. analysis of paired normal and leukemic DNAs. The protein truncation test (PTT) was then used to search for mutations in the ATM gene. Four mutations were detected (1 nonsense, 2 aberrant splicings, and 1 missense) among patients with LOH (N=7); no ATM mutations were found among patients without LOH of the region (N=2). It was suggested that ATM acts as a tumor suppressor gene whose inactivation is a key event in the development of T-cell prolymphocytic leukemia. Loss of heterozygosity (LOH) was defined as germinal heterozygosity for a given marker, complete or nearly complete (>=90% decreased intensity) absence of an allele in the tumor samples. The risk of T-PLL among individuals with germline heterozygosity of ATM mutations remains to be established. The two splicing errors resulting in truncated protein were detected by PTT. However, PTT infrequently detects missense mutations which are predominant in T lineage leukemia. How the missense mutation was found without generating a stop codon (as the nonsense mutation found in this study) was not explained. The single strand confirmation polymorphism (SSCP) approach might allow for the demonstration of a higher number of ATM mutations than PTT approach. Vorechovsky et al (1997) demonstrated a high frequency of ATM mutations among sporadic T-PLL patients. In contrast to the ATM mutation pattern in A-T, the most frequent nucleotide changes in this leukemia were missense mutations. Tumor DNA from 37 patients with sporadic T-cell prolymphocytic leukemia (T-PLL) were analyzed by exon-scanning SSCP. Seventeen patients (46%) demonstrated nineteen types of ATM mutations. In contrast to the ATM mutation pattern in A-T most often 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. characterized by small deletions and insertions, the most frequent nucleotide changes in this leukemia were missense mutations (30%). These mutations clustered in the region corresponding to the kinase domain, which is believed either to interfere with the catalytic function of ATM or to directly abrogate ATP binding or substrate recognition. The authors suggested that it is possible that patients with these missense mutations in the germline will not develop clinically recognizable A-T, but rather may predispose to lymphocytic neoplasia. This hypothesis warrants further study to examine whether a proportion of leukemia patients carry an ATM mutation. Two of seventeen mutated T-PLL samples had a previously reported A-T allele. The other fifteen mutations were not observed in a laboratory study of 308 Caucasians. Biallelic polymorphism in an amplified segments corresponding exon 26 were analyzed. Two alleles were present in 19% (7/37) of T-PLL samples compared to 49% (189/384) heterozygosity in controls of northern European ancestry. This suggests a loss of one of the ATM alleles in a subset of tumor DNAs. The evidence of a significant proportion of loss-of-function mutations and complete absence of the normal copy of ATM in the majority of mutated tumors establishes somatic inactivation of this gene in the pathogenesis of sporadic T-PLL. These findings support the hypothesis that ATM acts as a tumor suppressor. As constitutional DNA was not available, a putative hereditary predisposition to T-PLL will require further investigation. Stilgenbauer et al (1997) demonstrated that a small commonly deleted segment at 1 lq22.3-23.1 was identified in 15 of 24 T-PLLs patients (62.5%). The 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. remaining copies of the gene were further analyzed among six cases since this critical region contained ATM. Deletions affecting one ATM allele were found among all six patients. Mutations of the second ATM allele were also identified among all six cases examined, leading to the absence, premature truncation or alteration of the ATM gene product. Thus, this study demonstrated disruption of both ATM alleles by deletion or point mutation in T-PLL, providing further evidence that the ATM gene functions as a tumor-suppressor among non-A-T individuals. Results from these initial studies of mutations in ATM among leukemia patients may be summarized as follows: 1) the studies have focused on T-lineage ALL or PLL patients with 32% -62.5% somatic/germline mutations identified; 2) missense mutations are dominant in T-PLL and T-ALL patients, in contrast to mostly truncating mutations found among A-T patients; 3) almost all mutations identified so far have been unique regarding mutation types and locations in the ATM (i.e., no single so called ‘hot spot’ has been identified among leukemia patients); 4) most studies, except one, revealed somatic mutations instead of constitutional mutations. Several limitations of these studies precluded the estimation of the relative risk for developing leukemia associated with carrying a mutation in ATM. First, the representativeness of the source population was questionable given a lack of detailed description regarding how these cases (sample sizes ranging from 8 to 37 patients) were ascertained. Second, in at least one study, the protein truncation test (PTT) was used to detect missense mutations for which the PTT is known to be not sensitive. Even when single strand conformational polymorphism (SSCP) was used 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to identify missense mutations, its sensitivity is unclear. Third, the selection of ‘unmatched controls’ (‘unaffected individuals’) was not adequately described. Finally, the analyses of unaffected individuals were generally confined to 1 or 2 exons rather than screening the entire gene for mutations. These studies raise several questions: 1) whether these somatic mutations among T-PLL could also be identified in the germline; 2) whether ALL patients have the same or different type and frequency of ATM gene mutations as T-PLL patients; and 3) the generalization of the majority of germline mutations identified in a small sample of T-ALL patients (Luo et al, 1998) in one previous study to the majority of childhood ALL is uncertain. Further more detailed questions include: 1) whether these nucleotide substitutions cause loss of function of ATM gene or they are “neutral” polymorphisms; 2) whether the number of missense mutations is significant different from expected; 3) what is the difference between the missense mutation or truncated mutation among leukemia patients compared to those found among the A-T and general populations? Much research remains to be done to estimate the risk of leukemia among ATM heterozygotes. The proposed study will exam the mutations in the ATM gene of both cases and matched controls to examine the associated risk of carrying a mutation on the development of childhood ALL. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Conclusion A summary of the major findings regarding childhood ALL providing background information for the proposed study follows: 1) The etiology of childhood leukemia is largely unknown. 2) The lower incidence and ratio of childhood ALL to ANLL among Asians and Africans compared with that in Western nations suggest potential differences in genetic characteristics, lifestyle factors, and exposure to infectious agents and environmental factors. Additional studies are indicated. 3) The role of genetic factors in the development of childhood ALL is supported by familial leukemia case reports and by twin studies, although recent studies suggest that high concordance rates among identical twins might be caused by leukemic cells transferring from one twin to the other. 4) Increased incidence of ALL among persons with congenital and hereditary disorders characterized by inherited or induced chromosome abnormalities (such as Ataxia-telangiectasia, Down's, Klinefelter's, Bloom's syndrome or Fanconi's anemia) provides additional evidence for a genetic etiology. 5) Further genetic evidence is provided by multiple leukemia in consanguineous marriage. Among those families with two or more leukemia, there were more parents who were first cousins than those families with one single leukemia. Whether the excess of leukemia was associated with more hereditary disease in consanguinity or with recessive mode of inheritance for leukemia needs further study. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6) High dose radiation increases the risk for leukemia in childhood. In contrast, the association of low dose radiation and the risk of leukemia in childhood is not clear. 7) About 13% of ataxia telangiectasia patients developed leukemia by age 15. Recently, among small samples of adult and childhood leukemia patients, somatic and germline mutations were found in ATM gene. Whether ATM germline mutations is one of genetic basis for childhood leukemia requires further study. 8) The two peaks of ALL incidence suggest different risk factors may be operative at the different ages. Very little is known why people between age 25 to 60 have a relatively low frequency of ALL. It is possible that immunologic characteristics of the young and the elderly may be very important in the etiology of ALL. Additional studies of immunologic factors within the context of case-control studies for both the young and the elderly with ALL are warranted. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3 - Incidence of Cancer Among Family Members of Childhood Acute Lymphocytic Leukemia with a Family History of Cancer Introduction Studies of the association of genetic factors with leukemia will help to identify the genes that are critical to both normal and abnormal cellular growth, differentiation and proliferation, and locating gene mutations contributing to the development of childhood cancer. The study of childhood leukemia can provide important insights into genetic or familial factors, possibly helping to identify high risk individuals and families. Such knowledge may contribute to cancer prevention among children and their family members. Leukemia is the most common type of childhood cancer. Acute lymphocytic leukemia (ALL) accounts for 60-80% of all childhood leukemia. The aim of the present study was to investigate the genetic background of children with ALL, more specifically, to see whether the incidence of 25 types of cancer among adult family members of children with ALL is higher than expected based on gender, race and age specific rates from the general population. Specifically, the following hypotheses are to be tested in the analyses: 1) There are more leukemias among family members of childhood ALL patients; 2) There are more cancers other than leukemia among family members of childhood ALL; 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It was originally hypothesized that four familial or genetic factors observed among adult family members would be positively associated with increased ALL incidence observed among childhood patients (probands): (a) family history of cancer, (b) a major congenital abnormality in the proband, (c) a history of more than three spontaneous abortions in the proband’s mother, and (d) second malignancy in the lifetime of the proband. In this data analysis chapter, the incidence of cancers among family members of ALL is compared with that of the general population. Background Extensive epidemiological, clinical, and laboratory studies on many types of adult cancers provide information on risk factors, high risk groups, and adult cancer etiology. But much less is known of childhood carcinogenesis. Epidemiologists face formidable challenges identifying extrinsic (environmental) and intrinsic (individual and genetic) risk factors for childhood cancers and the interactive effect of the two in understanding the etiology of childhood cancer. Much of the ALL research that has been done over the past two decades has focused on extrinsic risk factors. Some of the most commonly studied extrinsic risk factors for ALL include: electromagnetic fields (EMF), ionizing radiation, maternal and paternal consumption of alcohol, medications during pregnancy, diet, smoking habits, paternal occupational exposures, and viral related infections in utero. Overall, study findings on extrinsic risk factors for ALL are inconsistent and account for a small proportion of the observed variation in disease incidence. 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Genetic risk factors may be of greater importance in explaining childhood than adult cancers because children who develop cancer are usually not old enough to have had chronic exposure to extrinsic risk factors. If inheritance of a gene (or genes) results in susceptibility to environmental carcinogens and the development of the malignancy, the genetic proportion can be much higher. Several lines of evidence suggest that genetic factors play a role in the etiology of leukemia. First, familial aggregation of the disease and multiple cases of ALL in offspring of consanguineous marriages have been reported (Kende et al, 1994). The remarkable similarities of age, mode of presentation, characteristics of the leukemia cells, response to treatment and prognosis among these family members suggests that malignancy may originate from the same mechanism. Second, there is an increased likelihood of leukemia developing in monozygotic twins, with a concordance rate as high as 25% (Danis et al, 1982). Third, there is a significant increase in the incidence of leukemia in children with chromosomal abnormalities (e.g., Down syndrome). Finally, the Philadelphia chromosome - the reciprocal translocation between chromosome 9 and 22 - might be responsible for the development of chronic myelogenous (CML) and some other leukemias. Methods Study design A retrospective cohort study was conducted. Family history information was collected for first and second degree relatives of children (under 21 years of age) 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. newly diagnosed with acute lymphoblastic leukemia (ALL) enrolled into Childhood Cancer Group (CCG) B-903 study during the period 1990-96. Identification o f Study Population All newly diagnosed childhood patients who were treated in one of CCG affiliated hospitals were eligible for screening for entry into the B-903 study. Screening was achieved through an eligibility questionnaire. In addition, prevalent cases (children seen at the hospital irrespective of diagnosis date) were also eligible if they met any of the selection criteria. Completed eligibility questionnaires were sent to CCG Operation Office to be checked and filed. Once the patient’s family was determined to meet the eligibility criteria, the on-study form and family history form were completed by a research associate or a nurse, and parents of the patient. These two forms were returned to CCG center within four weeks of registration of the patient for data screening and processing. The protocol remained open for 1990- 1996. Selection Criteria Six criteria were used to select subjects (see Table 11). Only those families whose proband had a family history of cancer among any first degree relatives or among second degree relatives under age 36 were included in the following analyses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 11. The Eligibility Criteria of CCG-B903 Study 1. Patients had a family history of cancer defined as: A first degree relative with cancer or A second degree relative or a cousin with cancer that were diagnosed before age 36; 2. Patients had a major congenital abnormality, a hereditary condition, a constitutional chromosomal abnormality, significant dismorphology or developmental retardation not secondary to birth injury; Abnormalities that were specially excluded are: Minor congenital abnormalities such as transient cardiac problems and small birthmarks; Cleft lip (with or without cleft palate), since this conditions is known to be polygenic; Cleft palate (with or without cleft lip); Abnormalities attributed to environmental factors, such alcohol abuse, maternal drug abuse, rubella and other infections; Cerebral palsy and other defects caused by in utero hypoxia; and Abnormalities caused by pregnancy complications, such as diabetes, premature delivery and toxemia. 3. Children of mothers who had a history of multiple spontaneous abortion (three or more); 4. Patients with a second malignancy ; 5. Patients with multiple primary tumors, including bilateral tumors in paired organs; 6. Being twins. * Patients were those under 21 years old at registration or diagnosis and patients met at least one of the above criteria received the on study and family history forms. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Data collection The protocol CCG-B903 was open to patient entry on December 1, 1990. A structured questionnaire was designed to obtain pedigree information on the families of registered patients. This information included the occurrence of cancer and other major illnesses among first and second degree relatives. More specifically, information was collected from parents on: i) family history of cancer; ii) hereditary conditions, congenital and constitutional chromosomal abnormalities (including skin syndromes, neurocutaneous syndromes, skeletal disorders, endocrine system abnormalities, intestinal syndromes, chromosome fragility syndromes, immunodeficiency syndromes, and other hereditary conditions); iii) history of spontaneous abortion by the patient’s mother; iv) occurrence of a second malignancy in the patient; v) occurrence of multiple primary tumors in the patient; and, vi) the patient’s twin status. Family pedigrees for each family were developed, including information on patients’ parents, siblings, maternal and paternal grandparents, and maternal and paternal aunts and uncles. For those family members identified as having cancer or congenital conditions, the age of diagnosis was reported. Family members’ ages at the time of interview (or death if deceased) were also collected. If the family members had more than one malignancy or congenital abnormality, up to three types of cancer or conditions and age of onset were collected. The proband’s age, gender, date of birth, and selection eligibility criterion were also collected. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Database Establishment and Management Completed eligibility and family history questionnaires were checked for completeness and consistency by the data manager and database administrator. If necessary, respondents were re-contacted by the interviewer to clarify inconsistent and missing information. Data on patients (probands) were entered and maintained in the CCG registration data base. The probands’ and their family members’ disease diagnoses, ages, and relationships information were stored in the Genetic Analysis Package (GAP) software (Buckley, 1997). Family pedigrees were drawn using GAP. Statistical Analysis Standard Incidence Ratio (SIR) A cohort analysis was used to compare the incidence of different types of cancers among the study families with the incidence from the general population. Person-years of follow-up were calculated from the date of birth to the date of cancer diagnosis, death, or date of study entry, whichever occurred first. The standardized incidence ratio (SIR) was calculated by dividing the observed by the expected number of incident cancer cases. The age-specific cancer incidence rates in Los Angeles County between 1983-87 were used to calculate the expected number of incident cancer cases. SIRs were calculated separately for White, Hispanic and African American families (as classified by the race recorded for the proband), and for cancers among Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. family members occurring before and after age 45. Male and female ALL relatives were pooled. For male cancers (i.e., prostate, testicular cancers), only male family members contributed person-years to the observed and expected numbers. Similarly only female relatives were used for female specific cancers (female breast cancer, uterus, cervix, and ovary cancers). All leukemia subtypes were combined to calculate the observed and expected values for leukemia. This pooled leukemia group included lymphoid leukemia, myeloid leukemia, monocytic leukemia, other leukemia, and leukemia unspecified. Cancer of the uterus (unspecified) and corpus uteri were combined. SIR by Race, Age, Gender, and Degree of Relatives The SIRs were stratified by race and age (Table 14.) since cancer incidence is different among races and age groups. Cancers that occurred at younger ages may be more likely to be related to genetic factors than those that occurred at older ages. Therefore, cancer incidence among ALL relatives younger than 45 years of age or 45 years or older was analyzed separately. For cancer sites with an overall significant confidence interval, this excess observed cancer incidence was examined to see if it was stronger in a race or age group. Since the majority of ALL patients in this study were White (nearly 90%) and there were far fewer Hispanic and African American relatives at risk, the 95% Cl of overall SIRs combining three race groups stratified by age groups was used to discuss the significance of the findings. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. An attempt was made to stratify the results by gender and degree of relationship to the proband. Since the data were sparse within these subgroups, only SIRs of all three races combined were presented (Table 15-17). Weights and Adjustments If the family was entered into the study only because a first degree relative had cancer or because a second degree relative had cancer before age 36 (hereafter called “second proband”), there will be selectively more cancers in the family members because of the selection criteria than if those case families had been selected randomly. Therefore, an adjustment was made to the observed and expected values. If only one family member had cancer and was responsible for the family being eligible for the current study, this second proband was dropped from the numerator and denominator (weighted as “zero”). If more than one relative had cancer that qualified the family for the present study, one case was excluded using a weighting system. Specially, if n such “secondary probands” were present, each secondary proband was weighted for both the cancer event and person-years by the factor If there was no on-study form to provide eligibility data, these data were taken from the initial eligibility form. When any family member had more than one type of cancer, only the first diagnosed type of cancer was used in the analysis. Each family member contributed to the person-years from birth to age of cancer onset, or 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. until the age at time of interview if the person had not experienced any malignancy in his life. Results Characteristics o f the Study Population There were total 406 newly diagnosed cases of childhood ALL (probands). The racial/ethnic mix of the 406 probands included 345 White (85.0%), 21 Hispanic (5.2%), 23 African American (5.7%), and 15 other races (3.7%). Two patients did not have race/ethnicity reported (0.4%). There were 246 males (60.6%). The median age at diagnosis was 5 years. Ninety-seven (23.9%) had leukemia at age 2 years or younger. Of the total 406 newly diagnosed cases of childhood ALL (probands), 294 (72.4%) were included in the analysis because these patients had a family history of cancer. The age, gender and race distribution is summarized in Table 12, types of cancer reported among family members and weights of these cancers in the analysis are summarized in Table 13. There were 4300 individuals from these probands and their family members. After excluding the 294 probands, there were 4006 family members. Since 166 family members were weighted as zero (see the following section for details), there were 3,840 family members, or an average of 13.1 family members per proband included in the analysis. Among the 3,840 proband family members, 1,909 (49.7%) were males. The racial/ethnic composition included 3,408 White (88.8%), 263 Hispanic (6.8%), and 169 African American (4.4%). 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 12. Demographic characteristics of the childhood ALL patients with family history of cancer Frequency (Percentage) Gender Male 177 (60.2%) Female 117 (39.8%) Race White 263 (89.5%) Hispanic 17 ( 5.7%) Black 14 ( 4.8%) Age at diagnosis (years) 0 6 (2.0%) 1 22 (7.5%) 2 42 (14.3%) 3 45 (15.3%) 4 25 (8.5%) 5 21 (7.2%) 6 10 ( 3.4%) 7 14 (4.8%) 8 20 (6.8%) 9 7 (2.4%) 10 6 (2.0%) 11 15 (5.1%) 12 11 (3.7%) 13 10 (3.4%) 14 14 (4.8%) 15 14 (4.8%) 16 8 (2.7%) 17 1 ( 0.3%) 18 1 ( 0.3%) 19 2 (0.7%) 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 13. Weights of cancers among family members of childhood ALL patients Weight Cancer Frequency 0.5 Colon 1 0.5 Pancreas 1 0.5 GI, Tract, NOS 1 0.5 Bronchus/Lung 1 0.5 Bones and Cartilage 1 0.5 Melanoma 2 0.5 Skin 7 0.5 Breast 8 0.5 Uterus 1 0.5 Cervix 7 0.5 Ovary, Fallopian Tube 3 0.5 Kidney 1 0.5 Brain/Neuroma 2 0.5 Thyroid 1 0.5 Lymph Nodes 3 0.5 Non-Hodgkin's Lymphoma 1 0.5 Hodgkin's Disease 1 0.5 Leukemia 3 0.5 Malignant Tumors, NOS 5 0.67 Skin 2 0.67 Leukemia 1 0.75 Breast 2 0.75 Kidney 1 0.75 Hodgkin's Disease 1 1 No cancer 3492 1 Tongue 1 1 Mouth, NOS 1 1 Oropharynx 1 1 Esophugus 3. 1 Stomach 5 1 Small Intestine 1 1 Colon 22 1 Liver 9 1 Gallbladder 1 Pancreas 7 1 GI, Tract, NOS 2 1 Nasal Gavities 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 13 (continued). Weights of cancers among family members of childhood ALL patients Weight Cancer Frequency 1 Larynx, Throat 9 1 Bronchus/Lung 31 1 Respiratory Tract, NOS 1 1 Blood/Bone 9 1 Bones and Cartilage 1 1 Melanoma 15 1 Skin 41 1 Breast 46 1 Uterus 14 1 Cervix 1 1 Ovary, Fallopian Tube 4 1 Prostate 10 1 Testis 2 1 Penis & Other Male Organs 1 1 Bladder 5 1 Kidney 2 1 Brain/Neuroma 10 1 Thyroid 3 1 Other Site, NOS 1 1 Lynph Nodes 4 1 Non-Hodgkin’s Lymphoma 1 1 Hodgkin’s Disease 3 1 Leukemia 12 1 Malignant Tumors, NOS 11 Excluded cases Of the 4,006 original family members, 166 (4.1%) were excluded from the analysis. These included 54 first degree relatives who had cancer and this cancer led to ascertainment of the family, 92 second degree relatives had cancer under age 36 as the only criterion that qualified the family for the study, and 20 family 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. members of adopted children. Thus, the analyses which follow are based on a total of 3,840 eligible family members of childhood ALL probands. If the patient was adopted, the entire family was excluded from the analysis. Since there were only 15 other race patients, the data was too sparse to perform SIR analyses of these race groups. Included cases If the family was registered into the study by any criteria other than family history (FH) of cancer (even if FH was also present), the whole family was included in the analysis (excluding the proband). There were fifty-seven (57) so called ‘second probands’ who were “partially” included in the analysis (weighted less than 1). Four had weights of 0.75, three had weights of 0.67, and fifty (50) had weights of 0.5. Family History o f Cancer The SIR outcomes for the 3,840 family members are reported below for 25 types of cancer for which family history information was collected. The CCG disease code was converted to the ICDO code (2n d Edition, 1990). SIRs are stratified by age, race/ethnicity, gender and degree of relationship to the proband. SIR age group categories include less than 45 years (referred to as “younger” relatives) and 45 years and above (called “older” relatives). Racial/ethnic categories included White, Hispanic, and African American. In discussing the results, the term “overall” 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. will be used often. Overall age SIR results include family members of all ages. Overall racial/ethnic group SIR results include results for all three racial/ethnic groups. If there was an overall significant excess observed, the difference by age groups was discussed. Difference by racial groups was not emphasized since the data for Hispanic and Black relatives was sparse. SIRs for males and females have been combined (Table 14, 16, 17). SIRs stratified by gender are demonstrated in Table 15. Overview o f the Results The results confirmed the hypotheses of the study. There are more leukemias among family members younger than 45 years of age. Moreover, this study indicates that colon, liver, larynx, melanoma, pancreatic and uterus cancers were significantly more common among older family members. More brain/CNS tumors and leukemia were found among younger family members than in the general population, but not among older family members. Breast cancer was found to be higher in both young and old age groups. Risk of Leukemia Proband family members were 2.4 times as likely to develop leukemia than age and gender matched people in the general population (SIR=2.4, 95% CI=1.3-3.8) (Table 14). A significant 3.8-fold increased risk of developing leukemia was found 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 14. The observed cancer among ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) Disease White <=44 yr >=45 yr SIR Hispanic <=44 yr >=45 yr SIR Black <=44 yr >=45 yr SIR Total across all ethnic groups Observed/Expected SIR 95% Cl of total SIR Tongue 0.0 ( 0/ 0.28) (0/0.006 ) (0/0.004 ) 0.0 0/ 0.29) - 1.0 ( 1/ 1.03) (0/0.03 ) (0/0.03 ) 0.9 1/ 1.08) 0.0004-3.6 0.8 ( 1/ 1.31) (0/0.03 ) (0/0.03 ) 0.7 1/ 1.37) 0.0003-2.9 Esophagus 0.0 ( 0/ 0.06) (0/0.004 ) 0.0 (0/0.01 ) 0.0 0/ 0.07) - 1.7 ( 2/ 1.18) (0/0.07 ) 10. 9 (1/0.09 ) 2.2 3/ 1.35) 0.4-5.4 1.6 ( 2/ 1.24) (0/0.08 ) 9.9 (1/0.10 ) 2.1 3/ 1.42) 0.4-5.2 Stomach 2.8 ( 1/ 0.36) (0/0.05 ) (0/0.03 ) 2.3 1/ 0.44) 0.0009-8.9 1.4 ( 3/ 2.18) (0/0.30 ) (0/0.10 ) 1.2 3/ 2.57) 0.2-2. 9 1.6 ( 4/ 2.54) (0/0.35 ) (0/0.13 ) 1.3 4/ 3.01) 0.3-3.0 Colon 1.8 ( 2.0/ 1.12) (0/0.05 ) (0/0.06 ) 1.6 2.0/ 1.23) 0.2-4.7 1.9 (17.5/ 9.26) (0/0.43 ) 3.5 (1/0.28 ) 1.9 18.5/ 9.97) 1.1-2.8 1.9 (19.5/10.38) (0/0.48 ) 3.0 (1/0.34 ) 1.8 20.5/11.20) 1.1-2. 7 Liver 0.0 ( 0/ 0.21) (0/0.02 ) (0/0.02 ) 0.0 0/ 0.24) _ 10. 6 ( 7/ 0.66) (0/0.09 ) (0/0.04 ) 8.8 7/ 0.80) 3.5-16.4 8.1 ( 7/ 0.87) (0/0.11 ) (0/0.06 ) 6.7 7/ 1.04) 2.7-12.6 Pancreas 1.7 (0.5/0.29) 0.0 (0/0.02 ) 0.0 (0/0.01 ) 1.6 0.5/0.32) 0.2-8. 9 1.6 (4.0/2.52) 12.1 (2/0.16 ) 10. 8 (1/0.09 ) 2.5 7.0/2.78) 1.0-4.7 1.6 (4.5/2.81) 11.0 (2/0.18 ) 9.6 (1/0.10 ) 2.4 7.5/3.10) 1.0-4.5 vo u > Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 14 (continued). The observed cancer among ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) Disease White <=44 yr >=4 5 yr SIR Hispanic <=44 yr >=45 yr SIR Black <=44 yr >=45 yr SIR Total across all ethnic groups Observed/Expected SIR 95% Cl of total SIR Larynx 9.2 ( 2/ 0.22) (0/0.007 ) 0.0 (0/0.01 ) 8.5 2/0.24) 0.8-23.9 3.2 ( 6/ 1.89) (0/0.08 ) 12.5 (1/0.08 ) 3.4 7/ 2.05) 1.4-6.4 3.8 ( 8/ 2.11) (0/0.09 ) 11.0 (1/0.09 ) 3.9 9/ 2.29) 1.8-6.9 Lung/Bronchus 0.3 ( 0.5/ 1.68) 27.2 (1/0.04 ) 0.0 (0/0.08 ) 0.8 1.5/ 1.80) 0.03-2.7 1.3 (27.0/20.61) 0.0 (0/0.71 ) 2.6 (2/0.75 ) 1.3 29.0/22.10) 0.9-1.8 1.2 (27.5/22.29) 1.3 (1/0.75 ) 2.4 (2/0.83 ) 1.3 30.5/23.90) 0.9-1.8 Bone 0.5 (0.5/1.04) (0/0.07 ) (0/0.02 ) 0.4 0.5/1.13) 0.07-2.5 6.2 (1.0/0.16) (0/0.01 ) (0/0.004 ) 5.6 1.0/0.18) 0.002-21.8 1.2 (1.5/1.20) (0/0.08 ) (0/0.02 ) 1.2 1.5/1.31) 0.05-3.7 Connective 0.0 ( 0/ 1.21) (0/0.08 ) (0/0.05 ) 0.0 0/ 1.35) _ tissue 0.0 ( 0/ 0.65) (0/0.03 ) (0/0.01 ) 0.0 0/ 0.69) - 0.0 ( 0/ 1.86) (0/0.11 ) (0/0.06 ) 0.0 0/2.04) - Melanoma 0.7 ( 5.0/ 7.18) (0/0.07 ) (0/0.03 ) 0.7 , 5.0/ 7.28) 0.2-1.4 2.5 (11.0/ 4.38) (0/0.06 ) (0/0.02 ) 2.5 11.0/ 4.46) 1.2-4.1 1.4 (16.0/11.56) (0/0.13 ) (0/0.05 ) 1.4 16.0/11.74) 0.8-2.1 Urinary Bladder 0.0 ( 0/ 0.62) (0/0.01 ) (0/0.01 ) 0.0 0/ 0.65) _ 1.1 ( 5/ 4.71) (0/0.17 ) (0/0.07 ) 1.0 5/ 4.94) 0.3-2.1 0.9 ( 5/ 5.33) (0/0.18 ) (0/0.08 ) 0.9 5/ 5.60) 0.3-1.8 4 ^ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 14 (continued). The observed cancer among ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) White Hispanic Black Total across all 95% Cl <=44 yr <=44 yr <=44 yr ethnic groups of total SIR Disease >=45 yr SIR > SIR =45 yr > SIR =45 yr Observed/Expected SIR Kidney 0.8 (0.75/0.93) (0/0.04 (0/0.05 ) 0.7 (0.75/1.02) 0.01-3.3 0.4 (1.00/2.58) (0/0.19 (0/0.07 ) 0.4 (1.00/2.84) 0.0001-1.4 0.5 (1.75/3.51) (0/0.23 (0/0.12 ) 0.5 (1.75/3.86) 0.03-1.4 Brain/Nervous 2.3 ( 7/ 3.06) 6.9 (1/0.14 (0/0.06 ) 2.5 ( 8/ 3.26) 1.0-4.4 System 1.9 ( 3/ 1.59) 0.0 (0/0.09 (0/0.03 ) 1.8 ( 3/ 1.71) 0.3-4.3 2.2 (10/ 4.65) 4.3 (1/0.23 (0/0.09 ) 2.2 (11/ 4.97) 1.1-3.7 Thyroid 0.5 (1.5/3.01) (0/0.18 (0/0.04 ) 0.5 (1.5/3.22) 0.02-1.5 1.8 (2.0/1.12) (0/0.07 (0/0.02 ) 1.7 (2.0/1.21) 0.2-4.7 0.8 (3.5/4.13) (0/0.25 (0/0.06 ) 0.8 (3.5/4.43) 0.2-1.8 Hodgkin's 1.4 (4.25/3.06) (0/0.10 (0/0.07 ) 1.3 (4.25/3.22) 0.4-2.9 disease 0.0 (0.00/0.43) (0/0.03 (0/0.01 ) 0.0 (0.00/0.47) - 1.2 (4.25/3.49) (0/0.13 (0/0.08 ) 1.2 (4.25/3.69) 0.3-2.5 Non-Hodgkin's 0.2 (0.5/3.02) (0/0.14 (0/0.07 ) 0.2 (0.50/3.24) 0.02-0.9 Lymphoma 0.3 (1.0/3.89) (0/0.21 (0/0.07 ) 0.2 (1.00/4.17) 0.0001-0. 9 0.2 (1.5/6.91) (0/0.35 (0/0.14 ) 0.2 (1.50/7.41) 0.01-0.7 Leukemia 3.5 (10.67/3.09) 4.0 (1/0.25 17.8 (1.5/0.08) 3.8 (13.17/3.43) 2.0-6.2 0.4 ( 1.00/2.27) 0.0 (0/0.13 0.0 (0.0/0.05) 0.4 ( 1.00/2.45) 0.0002-1.6 2.2 (11.67/5.36) 2.6 (1/0.38 11.5 (1.5/0.13) 2.4 (14.17/5.88) 1.3-3.8 vo on Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 14 (continued). The observed cancer among ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) White Hispanic Black Total across all 95% Cl <=44 yr <=44 yr <=44 yr ethnic groups of total SIR Disease >=45 yr SIR > SIR =45 yr > SIR = 45 yr Observed/Expected SIR Primary site 3.0 (2.5/0.84) 25.1 (1/0.04 ) (0/0.03 ) 3.9 ( 3.5/0.90) 0.9-9.0 uncertain 1.4 (5.0/3.45) 10.2 (2/0.20 ) (0/0.12 ) 1.9 ( 7.0/3.77) 0.7-3.5 1.7 (7.5/4.29) 12.7 (3/0.24 ) (0/0.15 ) 2.3 (10.5/4.67) 1.1-3.8 Prostate 0.0 ( 0/ 0.06) (0/0.002 ) (0/0.003 ) 0.0 ( 0/ 0.06) - 1.1 (10/ 9.46) (0/0.63 ) (0/0.39 ) 1.0 (10/10.48) 0.5-1. 6 1.1 (10/ 9.52) (0/0.63 ) (0/0.39 ) 1.0 (10/10.54) 0.5-1. 6 Testis 0.6 (2.0/3.45) (0/0.12 ) (0/0.01 ) 0.6 ( 2/3.60 ) 0.05-1.6 0.0 (0.0/0.27) (0/0.01 ) (0/0.0004) 0.0 ( 0/0.28 ) - 0.5 (2.0/3.72) (0/0.13 ) (0/0.01 ) 0.5 ( 2/3.86 ) 0.05-1.5 Breast 1.9 (16.5/ 8.74 2.6 (1/0.39 ) 0.0 (0/0.21 ) 1.9 (17.5/ 9.34) 1.1-2.9 Female 1.5 (28.0/18.99 2.7 (2/0.73 ) 3.1 (1/0.32 ) 1.6 (31.0/20.04) 1.1-2 .1 1.6 (44.5/27.73 2.7 (3/1.12 ) 1.9 (1/0.53 ) 1.7 (48.5/29.38) 1.2-2.1 Uterus 1.8 ( 1.5/0.85) 32. 6 (2/0.06 ) (0/0.009 ) 3.8 ( 3.5/0.92) 0.9-8.8 2.4 (10.0/4.26) 0.0 (0/0.18 ) (0/0.037 ) 2.2 (10.0/4.48) 1.1-3.8 2.3 (11.5/5.11) 8.3 (2/0.24 ) (0/0.046 ) 2.5 (13.5/5.40) 1.3-4.0 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 14 (continued). The observed cancer among ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) Disease White <=4 4 yr >=45 yr SIR Hispanic <=44 yr >=45 yr SIR Black <=4 4 yr >=45 yr SIR Total across all ethnic groups Observed/Expected SIR 95% Cl of total ; Cervix 1.9 (4.5/2.33) (0/0.31 ) (0/0.07 ) 1.7 (4.5/2.71) 0.5-3.5 0.0 (0.0/1.05) (0/0.21 ) (0/0.05 ) 0.0 (0.0/1.31) - 1.3 (4.5/3.38) (0/0.52 ) (0/0.12 ) 1.1 (4.5/4.02) 0.3-2.4 Ovary 1.1 (1.5/1.42) (0/0.10 ) (0/0.03 ) 1.0 (1.50/1.55) 0.04-3.1 1.6 (4.0/2.48) (0/0.14 ) (0/0.03 ) 1.5 (4.00/2.65) 0.4-3.4 1.4 (5.5/3.90) (0/0.24 ) (0/0.06 ) 1.3 (5.50/4.20) 0.4-2.6 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 15. The observed cancer among male and female ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) Total across all ethnic groups Male Female Disease SIR (95% Cl) SIR (95% Cl) Colon 2.6 (17.5/6.7) 1.1-2.7 0.7 (3/4.5) 0.1-1.9 Liver 6.6 ( 5.0/0.8) 2.1-15.4 7 .1 (2/0.3) 0.9-25.5 Pancreas 1.9 ( 3.5/1.9) 0.4-4.3 3.3 (4/1.2) 0.9-8.5 larynx 3.2 ( 6.0/1.9) 1.2-7.1 6.8 (3/0.4) 1.4-19.9 Lung/Bronchus 1.6 (24.5/15.6) 1.0-2.3 0.7 (6/8.3) 0.3-1.6 Brain/Nervous system 1.1 ( 3.0/2.8) 0.2-3.1 3.7 (8/2.2) 1.6-6.6 Leukemia 2.3 (8.17/3.5) 1.0-4 .2 2.5 (6/2.4) 0.9-4.9 vo oo Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 16. The observed cancer among first-degree ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) Total across all ethnic groups Disease SIR (95% Cl) Colon 1.9 (0.5/0.26) 0.3-10.9 Liver 0.0 (0.0/0.05) Pancreas 0.0 (0.0/0.06) Larynx 0.0 (0.0/0.06) Melanoma 0.0 (0.0/1.15) Breast 0.0 (0.0/1.18) Uterus 0.0 (0.0/0.13) Cervix 2.7 (1.0/0.37) 0.001-10. 6 Kidney 3.4 (0.75/0.22) 0.06-15.5 Brain/Nervous System 0.8 (0.5/0.64) 0.1-4.4 Non-Hodgkin's Lymphoma 0.8 (0.5/0.59) 0.1-4. 8 Leukemia 1.7 (1.17/0.70) 0.01-6.1 Primary Site Uncertain 2.7 (0.5/0.19) 0.4-15.0 v o V O Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 17. The observed cancer among second-degree ALL patient relatives with family history of cancer (after excluding the second proband) in CCG-B903 study compared to the expected number based on the age, gender, race-adjusted cancer incidence in Los Angeles County 1983-1987 Standardized Incidence Ratio (Observed /Expected Number) Total across all ethnic groups Disease SIR (95% Cl) Colon 1.,7 (19.0/10.9) 1. , 0--2.7 Liver 7 . . 1 ( 7.0/1.0) 2.,8--14. 5 Pancreas 2..5 ( 7.5/3.0) 1..0--4 .6 Larynx 4..0 ( 9.0/2 .2) 1..8--7 .7 Melanoma 1..5 (16.,0/10.6) 0.. 9--2.3 Breast 1..7 (48..5/28.2) 1.. 3--2.2 Uterus 2.5 (13..5/5.3) 1,. 4--4 .1 Cervix 1.0 ( 3.,5/3.6) 0.2--2.3 Brain/Nervous System 2, . 0 ( 8..5/4.3) 0,, 9--3.5 Non-Hodgkin's Lymphoma 0.2 ( 1. . 0/6.8) 0,.4--0.8 Leukemia 1.2 ( 6..0/5.2) . 0,. 4--2.3 Primary Site Uncertain 2.2 (10..0/4.5) . 1. 1--3.8 o o among young relatives (SIR=3.8, 95% CI=2.0-6.2) when all racial/ethnic groups were combined. Fewer cases of leukemia were found among older relatives compared to the general population (SIR=0.4, 95% CI=0-1.6). The excess of leukemia cases among young family members was consistent across all three racial/ethnic groups with 3.5, 4.0, and 17.8-fold increased risk among White, Hispanic and African Americans relatives, respectively. There were fewer reported cases of leukemia among older relatives in all racial/ethnic groups than expected in the general population (SIR=0.4, 0.0, and 0.0, respectively). Overall, leukemia cases occurred more often for all three racial/ethnic groups relative to the general population (SIR=2.2, 2.6 and 11.5, respectively). When stratified by gender, a 2.3-fold excess of leukemia was found among male relatives (95% CI=1.0-4.2), and a borderline significant 2.5 fold increased risk among female relatives (95% CI=0.9-4.9) (Table 15). There were 1.7 (95% 0=0.01-6.1) (Table 16) and 1.2-fold (95% CI=0.4-2.3) (Table 17) increased risks of leukemia among first and second degree relatives of ALL patients although the elevated risk was not significant. Risk of Brain/Central Nervous System Tumors Two times more cases of brain tumors were observed among all relatives of the childhood leukemia patients (SIR=2.2, 95% 0=1.1-3.7). More young relatives developed brain tumors (SIR=2.5, 95% 0=1.0-4.4) than the general population. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This was also true among older relatives, but this finding was not statistically significant (SIR-1.8, 95% CI=0.3-4.3). Whites and Hispanics were the main contributors to this elevated risk. In Whites, both younger and older relatives developed more brain tumors than the general population. In Hispanics, on the other hand, only younger relatives showed excess brain tumors. The increased risk among the young, old and all relatives of the two race groups are: 2.3, 1.9, 2.2 and 6.9, 0.0 and 4.3, respectively. No brain tumors were reported among African American relatives. When stratified by gender, there was a non-significant 1.1 -fold excess of CNS tumors found among male relatives (95% CI=0.2-3.1), and a significant 3.7- fold increased risk among female relatives (95% CI=1.6-6.6). There was a borderline 2.0-fold (95% CI=0.9-3.5) increased risk of CNS tumor among second degree relatives of ALL patients. The increased risk was not observed among first degree relatives. Risk of Female Breast, Uterine, and Cervical cancers The incidence of female breast cancer was consistently higher overall for all racial/ethnic and age groups. 1.7 times more cases of breast cancer were reported among relatives than expected based on the general population (SIR=1.7, 95% CI=1.2-2.1). Among relatives of White, Hispanic and African American patients younger than 45 years old, there was a significant 1.9-fold increased risk of breast cancer compared to a woman from the general population of same age and race 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (SIR=T .9, 95% CI=1.1-2.9). Similarly, older relatives were 1.6 times more likely to develop breast cancer than were adults of the same age and gender from the general population (SIR=1.6, 95% CRT.1-2.1). The excess of cases reported by both younger and older relatives was consistent across all three racial/ethnic groups: 1.9, 1.5, and 1.6-fold increased risks were found among Whites, 2.6, 2.7, and 2.7-fold increased risks were found among Hispanics, and 0, 3.1, and 1.9-fold greater risks were found among African Americans. Increased risk of uterine cancers was found among younger, older and all relatives of the pooled racial/ethnic groups (SIR=3.8, 2.2 and 2.5, 95% CI=0.9-8.8, 1.1-3.8 and 1.3-4.0, respectively). The overall significance was mainly influenced by younger White and Hispanic relatives and older White relatives. The observed elevated risks were 1.8, 2.4, and 2.3 among young, old and all White relatives and 32.6, 0, and 8.3 among Hispanic relatives. No cases of uterine cancers were reported by African American relatives. No overall excess of cervical cancer was found in this study. Nevertheless, it occurred more frequently among younger relatives, and all relatives of three races combined (SIR=1.7, 1.1, 95% CR0.5-3.5, 0.3-2.4, respectively), and no cervical cancers were reported in older relatives. The elevation was mainly present among White relatives (SIR=1.9). On the other hand, none was reported among older relatives, which resulted in overall incidence ratio of 1.3 for all relatives in this ethnic group. No cases of cervical cancers were observed among Hispanic and African American relatives. 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. There was no increased risk of cervical cancer among first and second degree relatives. There were 1.7 (95% CI=1.3-2.2) and 2.5-fold (95% CI=1.4-4.1) increased risks of breast and uterine cancer among second degree relatives. Excesses of breast and uterine cancer were not seen among first degree relatives. Risk of Colon, Liver and Pancreatic Cancers There were 1.8 to 6.7-fold increased risks among the three digestive system cancers among proband relatives. 1.8 times more cases of colon cancer was reported among family members (SIR=1.8, 95% CI=1.1-2.7). Among younger relatives, a 1.6 times more frequent of colon cancers were reported than expected from general population (SIR=1.6, 95% 0=0.2-4.7). Among older relatives, the chance of developing colon cancer was 1.9 times that of the general population (SIR=1.9, 95% 0=1.1-2.8). The excess of cases of colon cancer among both younger and older relatives were found among both White and African American family members. There was a 1.8, 1.9, and 1.9-fold increased risk of developing colon cancer for Whites, and a 0.0, 3.5, and 3-fold increased risk among African Americans. No cases of colon cancers were reported among Hispanic family members. Among White family members, older relatives showed an increased risk for liver cancer as high as 10.6 times that of their counterparts in the general population. Overall, White relatives had a 8.1-fold increased risk of developing liver cancer. When both Hispanic and African American relatives were included in the pooled cohort of person-years, the additional expected number of cancers did not change the 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. significantly increased risks for older relatives or for all relatives (SIR=8.8 and 6.7 and 95% Cl =3.5-16.4, 2.7-12.6, respectively). No liver cancers were reported among either Hispanic or African American relatives. The incidence of pancreas cancer among proband relatives was consistently higher for all three racial/ethnic groups among older relatives and among relatives overall. A significant, nearly three times greater number of pancreas cancer cases was reported among older and all proband relatives (SIR=2.5, 2.4, 95% CI=1.0-4.7, 1.0-4.5, respectively). The finding among older relatives and among family members of all ages was mirrored in each racial/ethnic group, with 1.6, 1.6 increased risk among Whites, 12.1, 11.0 among Hispanics, and 10.8 and 9.6 among African Americans. No cases of pancreas cancer were reported among either younger Hispanic or African American relatives. The elevated pancreas cancer rate among younger White relatives (SIR=1.7) did not result in a significant excess number of pancreas cases among younger relatives of the three racial/ethnic groups than for the general population (SIR=1.6, 95% CI=0.2-8.9). The overall increased colon cancer was observed among male relatives (SIR=2.6, 95% CI=1.1-2.7) and among second degree relatives (SIR=1.7, 95% CI=1.0-2.7). Liver cancer was reported more among male (SIR=6.6, 95% 0=2.1- 15.4) and female relatives (SIR=7.1, 95% 0=0.9-25.5) and among second degree relatives (SIR=7.1, 95% 0=2.8-14.5). Pancreas cancer was increased among female relatives (SIR=3.3, 95% 0=0.9-8.5) and among second degree relatives 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (SIR=2.5, 95% CI=1.0-4.6). The increased risk of these three type of cancer among first degree relatives was not indicated by the data. Risk of Larynx Cancer Cases of larynx cancer occurred more frequently among older relatives and among all relatives of the three races (SIR=3.4, 3.9, 95% CT=1.4-6.4, 1.8-6.9, respectively). Increased risk of 8.5 times the general population among young relatives was not significant due to the small number of reported cases (SIR=8.5, 95% CI= 0.8-23.9). The elevation was mainly present among older White and African American family members (SIR=3.2, 12.5) which led to a 3.8 and 11-fold increased risk among all White and African American relatives, respectively. No cases of larynx cancer were reported among Hispanic relatives. Increased risk of larynx cancer was found among both male (SIR=3.2, 95% CI=1.2-7.1) and female (SIR=6.8, 95% 0=1.4-19.9) relatives, and also among second degree relatives (SIR=4.0, 95% 0=1.8-7.7), but not among first degree relatives. Risk of Melanoma A 2.5 times significantly increased risk was found among older relatives of ALL patients (SIR=2.5, 95% 0=1.2-4.1). In contrast, younger relatives showed no significant difference in melanoma incidence than the general population (SIR=0.7, 95% 0=0.2-1.4). For the combined relatives of the two age groups, the chance of 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. developing melanoma was 1.4 times greater than for the general population (SIR=1.4, 95% CTO.8-2.1). Only Whites contributed to the observed melanoma cases, which paralleled the overall results. The excess of cases among older and all relatives and fewer cases among younger relatives showed the exact same incidence ratios as those pooled from all three race categories (SIR=2.5, 1.4 and 0.7, respectively). There was a borderline increased risk of melanoma among second degree relatives (SIR=1.5, 95% CI=0.9-2.3). There was no observed risk of melanoma among first degree relatives or among male, female relatives. Risk of Tumors for Unknown Primary Site About two times more tumor primary site unknown was observed among all relatives of the childhood leukemia patients (SIR=2.3, 95% 01=1.1-3.8). A 3.9-fold increased risk of developing tumors with unknown primary site was observed in young relatives of the probands (SIR=3.9, 95% 01=0.9-9.0) when the three ethnic groups were considered together. More unknown site tumors were also elevated among older relatives, although it was not significant (SIR=1.9, 95% 01=0.7-3.5). The young and old cancer family members had 25.1, 10.2 increased risk in Hispanics and 3.0, 1.4 in Whites. Among Hispanic and White relatives, there was a 12.7 and 1.7-fold increased risk than for the general population. No tumors with primary site uncertain were reported among African American relatives, which diluted the overall finding. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. There was an increased risk of melanoma among second degree relatives (SIR=T .5, 95% CI=0.9-2.3). There was no observed risk of melanoma among first degree relatives or among male or female relatives. Discussion Limitations and Potential Biases o f the Study Selection Bias The study was designed to include all childhood ALL patients who were treated at one of CCG institutions and met one of the eligibility criteria. Each parent was asked ten questions on a one page eligibility questionnaire and the questionnaire was sent to the CCG Operation Center if there was at least one positive answer for ten questions. Since the selection of newly diagnosed patients was institution- dependent, there was a potential for ‘more interesting’ families to be enrolled into the study, which could skew the results towards positive associations. These ‘more interesting’ families usually mean the child met more than one of the eligibility criteria, but does not mean there are more cancers in the family. Since the parents only answered “yes” or “no” to these ten questions, and there were only two questions related to family history of cancer among first and second degree relatives respectively, it is not expected that families with more cancers had a higher chance to be enrolled than families with fewer cancers. After considering that at least one cancer was eliminated from each family for the analysis, the impact of this selection bias is minimal. 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Although the eligibility was based on self-reported family history (FH), the accuracy of self-reported cancers among first degree relatives is very high- 83%- 100% (see the following section: Validity of Self-Reported Cancers). It is reasonably expected that there was not a potential for serious selection bias based on reported FH. The results must be reviewed with some caution. Excesses seen within one subgroup, but not generally, might well be genuine unless it is assumed that any biased ascertainment was specific for these families. Validity of Self-Reported Cancers The study is subjected to disease misclassification since the self-reported cancers were not validated through medical records, cancer registries or death certificates. There was no evidence from the data that cancers were over-reported at all sites. For example, lung cancer which is known mainly caused by smoking, was not significantly increased among family members of childhood ALL compared to that expected from the general population. In the literature, the accuracy of self-reported cancer was dependent on the site of cancer and the degree of relationship. For breast cancer, the sensitivity was as high as 94% and for cancers of bowel, ovary, endometrium, pancreas, and prostate, the sensitivity was more than 81% (Douglas et al, 1999). And the sensitivity of reporting was also influenced by closeness of relationship. If cancers were among first degree relatives, the sensitivity was from 86% to 94%, and if cancers were 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. among second degree relatives, the sensitivity was from 78% to 82%, respectively in two cities of confirmation (Douglas et al, 1999). The majority of reported cases were accurate, and especially, the self-reported cancer among first-degree relatives achieved high sensitivity. A correlation between the accuracy of reported diagnoses and the closeness of the relationship was also confirmed by two previous studies (Bondy et al, 1994, Love et al, 1985). One of the main reasons for difficulty in validating self-reported cancer was the lack of crucial information known by the proband about affected relatives. Bondy et al (1994) reported a difference in accuracy between first and second degree relatives. Of all 22 cancers in first-degree relatives reported by the parents of childhood sarcoma patients, all (100%) were confirmed as invasive cancer. Of 86 cancers in second-degree relatives reported by the parents of childhood sarcoma patients, 50 (58%) were confirmed as invasive cancer. This study indicated that there was over-reporting of cancers in the second degree relatives, and the self- reported cancers among first-degree relatives were accurate. In a comparison of self-reported family history with a cancer information database, high sensitivities were observed for subjects' reports of breast (83%), colorectal (73%), and prostate (70%) cancers, while ovarian (60%) and uterine (30%) cancers were not reported as well. Younger subjects tended to report family history more accurately than older subjects, with women reporting slightly better than men. The authors concluded that the results indicate subjects in a case-control 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. study are able to report accurately family histories of several common kinds of cancer and that they can do so without observable recall bias (Kerber et al, 1997). Aitken et al. (1995) compared self-report of colorectal cancer family history in patients undergoing colonoscopy with relatives' medical records. The records confirmed 77% of positive family histories (reporting colorectal cancer) and negative family histories (reporting no colorectal cancers). A study of patients with breast cancer found that their reports of cancer sites in first-degree relatives (FDR) were 85% to 100% accurate, whereas reports of cancer sites in second-degree relatives (SDR) dropped to an accuracy of 72% to 85% (Theis et al, 1994). However, these percentages were higher than those in one of the earliest studies on self-report accuracy, in which Love et al. (1985) documented 83% accuracy in FDR reports, 67% in SDR reports, and 60% in the reports of third-degree relatives (TDR). In a study cohort consisted of 10,523 participants from the First National Health and Nutrition Examination Survey in 1971-1975 who were aged 25-74 years, self-reports of hospitalization for breast cancer were confirmed as accurate for 100% of cases where a hospital record was available. The accuracy was moderate for lung cancer (78%), prostate cancer (75%), colon cancer (71%) (Bergmann et al, 1998). In the other study cohort including 65,582 men and women aged 39-96 years who were participants in the Cancer Prevention Study II Nutrition Survey, estimates of sensitivity (the proportion of study participants with a registry-documented cancer who self-reported the cancer) ranged from 0.79 for an exact match of cancer site and 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. year of diagnosis (+/-1 year) to 0.93 for a match of any reported cancer. The sensitivity of exact matches varied considerably by cancer site and was highest for breast, prostate, and lung cancers (0.91, 0.90, and 0.90, respectively) and lowest for rectal cancer and melanoma (0.16 and 0.53, respectively) (Bergmann et al, 1998). In a summary, these studies indicated that there is a high accuracy of self- reported cancers among first-degree relatives ranging from 83% to 100%, and less accuracy among second-degree relatives ranging from 67% to 85% except a 58% positive rate of self-reported cancers among second degree relatives in one childhood sarcoma study. Moreover, correlation between self-reported cancer and medical records varies for different cancers. Breast cancer has the highest sensitivity ranging from 83% to 100%, followed by colon cancer and prostate cancer with sensitivities from 71% to 77%, 70% to 90% respectively. In these studies, ovarian, uterus and melanoma had sensitivity of 60%, 30% and 53% respectively, comparatively lower than that of the above common cancers. The validity for reporting of other malignancies such as leukemia was limited in the literature. Therefore, the increased risk of breast cancer and colon cancer in the present study seems reliable. One needs to be cautious when interpreting the excess of uterine cancer and melanoma because of lower sensitivity of these two malignancies. Since the parents of the childhood ALL patients reported the cancers among their children, siblings and parents, all these were first degree relatives in terms of their relationship to the parents of the proband. Therefore, the sensitivity of the self-reported cancers by parents of the childhood ALL is expected to be high according to the literature (83%-100%). The 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. highest accuracy of self-reported cancer among first degree relatives in the literature makes the results of observed increased risk of cancers among second degree relatives in this study important and worthy of further study. In spite of the highest sensitivities of self-reported cancer among first-degree relatives, there was no observed increased risk of cancer among first degree relatives o f ALL patients in this study. It is probably due to the smaller number of first degree relatives of ALL patients compared to the number of their second degree relatives, and also probably due to the younger age range of the siblings and parents of these young ALL patients (median age at diagnosis: 5 years), who were not at high risk of developing cancer yet. Further larger scale studies are needed to explore these findings. The possibility of using these literature estimates of sensitivity in a quantitative way to adjust our own SIRs for misclassification might be considered. However, this is complicated by the different approaches in these studies since most sensitivity analyses were stratified either by cancer type without considering the degree of relationship, or stratified by relationship without considering the type of cancer. In the present study, all cancers were reported among first-degree relatives, therefore, the sensitivity should be higher than those cancer-specific sensitivities pooling self-reported cancers among all degrees of relatives. Use of Cancer Incidence Rates in Los Angeles (LA) County The age, gender and race specific cancer incidence in LA was used to calculate the number of cancers expected among the family members of a child with 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ALL. The incidence rates of LA do not differ significantly from those of other parts of U.S., and the rates between years 1983-87 do not differ significantly from the rates of those years when cancers among family members occurred. The minor difference in the LA incidence rates versus the national rates should have a negligible effect on the calculation of expected number of cancers. Adjustments (Weights) Adjustments for number of cancers among family members to be included in the analysis was necessary since the ascertainment made families with at least one additional cancer other than childhood ALL were enrolled into the study (also see Weights and Adjustments in “Method” section). However, whether the standardized incidence ratio was over or under-adjusted or appropriately adjusted needs further study. Interpretation of Results It seems that the observed number of cancers appears reasonable after the adjustments. Moreover, some of the families in the study could be ascertained by other criteria than family history of cancer, (i.e., congenital abnormality of the proband, mother’s multiple spontaneous abortions, etc.), and the observed cancer incidence among these families was also adjusted. Although the self-reported cancers were not validated, the sensitivity seems to be high in the literature. The expected number of cancers is reasonably accurate although LA cancer incidence Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. instead of national cancer incidence was used. As indicated above, it is expected that LA age, gender, and race-specific cancer incidence is not significantly different from the national cancer incidence, and therefore, did not have an important impact on the calculation of the expected numbers. Leukemia The reason for the familial aggregation of leukemia among young family members of childhood ALL patients is not known. The evidence supporting a genetic explanation of the elevated risk of leukemia among young family members include: 1) an excess of leukemia was found among patients with certain medical syndromes which are attributable to dysfunction of gene(s); 2) familial aggregation of leukemia has been reported in the literature; 3) leukemia was found higher among relatives of leukemia patients, among offspring of intermarriage and among families that had two or more childhood cancer patients (see literature review, Chapter 2). Our study found a 2.4-fold increased risk of all types of leukemia among all relatives, and a 3.8-fold excess of leukemia among younger relatives (i.e., proband family members less than 45 years of age). A previous study had shown a three-fold excess of first-degree relatives with leukemia compared to a control group matched for age and sex (Ponz de Leon et al, 1994). The increased risk of leukemia was found among both male and female relatives with similar magnitude of SIR and 95% confidence interval. The present study did not observe any excess of leukemia among first degree relatives. One explanation for the difference of study findings is 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that not only first and second degree relatives contribute to the significant increased risk of leukemia, but also third and fourth degree relatives. Excess of leukemia among relatives younger than 45 years may suggest roles of genetic factors. It was not known whether the observed elevated risk, together with increased risk of breast cancer among younger and older relatives, was part of medical syndrome (i.e. Li-Fraumeni syndrome) or interactions of other genetic and environmental factors. Carcinogenic agents, chromosome analysis, immunologic studies, and association of gene(s) and leukemia are recommended for future studies. Colon Cancer Colon and breast cancers present the best examples of the interaction of genetic and environmental factors in carcinogenesis. Pedigree analysis has shown familial aggregation of both hereditary nonpolyposis and adenomatosis coli as well as apparently sporadic forms of colorectal cancer. It has been suggested that the hereditary nonpolyposis is transmitted in an autosomal dominant pattern of inheritance and adenomatosis coli is most likely multifactorial, implying a close interaction between exogenous agents and several genes (Ponz de Leon, 1994). The excess of colon cancers among older relatives of leukemia patients may reflect such genetic and environmental agent interaction. Significantly elevated colon cancer incidence among all relatives and older relatives (i.e., relatives 45 years of age and above) were observed in this study (1.8 and 1.9-fold higher risks, respectively). Several studies found a positive family 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. history of colorectal neoplasms in 10%-20% of probands, yielding a 1.8-2.1 fold increased risk among family members (Ponz de Leon, 1994). A study found a 5.2- fold relative risk of colon cancer in younger first degree relatives of patients diagnosed with colorectal cancer (Hall et al, 1996). Risk of colon cancer was 1.77 times among those with a family history of colorectal cancer among first degree relatives compared to those without a family history. Risk of colon cancer was increased about the same amount among those with a family history of colorectal polyps. Gender, age and subset had little modifying effect on the increased risk (Kerber et al, 1998). In a case-control study of cancer of the colon it was found that 96 out of 332 (29%) cases had a positive family history of cancer of the colon (2 cases and more) as compared with 19 out of 473 (4%) controls (Ghadirian et al, 1993). A recently study indicated that environmental factors such as physical activity, energy intake, and vegetable intake showed no relationship with colon cancer in those with family history of colorectal cancer. The findings may suggest that with the presence of genetic predisposition, the influence of environmental factors on colon cancer risk was not significant (La Vecchia et al, 1999). One way to explain the elevated risk is that colon cancer is inherited as a component of medical syndrome such as Lynch syndrome II (hereditary nonpolyposis colorectal cancer, HNPCC) which is characterized by an excess of extra-colonic colorectal cancer (CRC) or carcinoma of the endometrium, ovary, small bowel, stomach, pancreas and transitional cell carcinoma of the ureter and renal pelvis (Lynch et al, 1995a; 1995b). Two mismatch repair genes hMLHl and 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hMSH2 have been related to HNPCC and to a lesser degree connected to neoplasms of skin, breast, and pancreas (Hutter et al, 1996). Recently it was confirmed that germ line mutations explained the excess of colon cancer risk among those diagnosed at a young age. It was found that fourteen out of 50 colon cancer patients (28%) diagnosed under 30 years of age had pathogenic mutations, and a number of other possible pathogenic variants of either the hMSH2 or hMLHl gene (Farrington et al, 1998). In earlier studies, different germ line mutations of either hMSH2 or hMLHl were identified in eight of 17 (47%) English HNPCC kindreds (Froggatt et al, 1996), in eight of 39 (21%) Swedish HNPCC families with mutations of hMLHl (Tannergard et al, 1995), and in nineteen of 30 (63%) Finnish HNPCC families with mutations of hMLHl (Nystrom-Lahti et al, 1995). In the other study, germline protein truncating mutations of hMLHl or hMSH2 were found in 50% of families with HNPCC (6 of 12) (Luce et al, 1995). Patients with HNPCC and their relatives could be at higher risk of other tumors. Female relatives of HNPCC patients are at higher risk for developing endometrial cancer (i.e. ovary & uterine cancer) (Watson et al, 1994). Breast cancer may occur as an integral tumor in the HNPCC syndrome (Risinger et al, 1996). It was observed that colon cancer patients are at greater risk of developing additional primary colon, rectal, and pancreatic cancer. The standardized incidence ratio (SIR) was 2.77 (95 percent confidence interval (Cl), 2.07-3.70) for additional primary colon cancer, 2.26 (95 percent Cl, 1.34-3.81) for rectal cancer, and 2.38 (95 percent Cl, 1.32-4.30) for pancreatic cancer, respectively (Slattery et al, 1995). 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. On the other hand, people with other cancers could be at increased risk of colon cancer. Evidence of a significant risk for colon cancer in BRCA1-linked families has been reported. Loss of heterozygosity in the 17q21 region was present in tumor cells of 39 (49%) patients among 79 colon cancer cases, which suggested that BRCA1 and other genes in this region may play an important role in the development of both breast and colon cancers (Garcia-Patino et al, 1998). More colon cancers were observed among first degree relatives of childhood brain tumor patients (O/E-5/1 .6). The present study suggests an association of leukemia and colorectal cancer among family members. The current study found an excess of cases of colon (SIR-1.8, 95% Cl: 1.1- 2.7), pancreas (SIR-2.4, 95% Cl: 1.0-4.5), uterus (SIR-2.5, 95% Cl: 1.3-4.0), and breast cancer (SIR-1.7, 95% Cl: 1.2-2.1) among relatives of ALL than those of in the general population. Although increased risk of colon cancer was only observed among relatives 45 years or older, the role of genetic factors can not be ruled out. Reliable results could be produced based on sensitivity of self-reported colon cancer ranging from 71% to 77% in the literature. It is not known whether BRCA1, hMSH2, hMLHl or other yet to be identified gene(s) play a role in the elevated risk of these cancers. The results might suggest complex gene(s) are associated with developing multiple primary tumors, or both genetic and environmental factors play roles in the elevated risk. How these factors interact and mediate steps in the progression from normal to malignant cells and influence cellular function needs further investigation. 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Liver Cancer Hepatocellular carcinoma (HCC) varies among different geographic area, socioeconomic status, gender, and age-specific incidence, suggesting genetic differences in susceptibility (Olsson et al, 1996). Hepatitis B virus and dietary chemical carcinogen aflatoxin Bl, alcohol consumption, occupation (farming), and family history of HCC are associated with a higher risk of liver cancer. A molecular signature of chemical carcinogen is found in the mutated p53 gene (Harris, 1996). The carriers of EPHX mutant alleles and hepatitis B infection had 135 times higher risk of HCC then those who did not have the mutants and the infection (Olsson et al, 1996). Aflatoxin Bl has been postulated to be a hepatocarcinogen by causing p53 mutation (McGlynn et al, 1995). Significant associations were observed between family history of hepatocellular carcinoma and primary liver cancer [relative risk (RR) = 2.4; 95% confidence interval (Cl), 1.3 to 4.4] among 320 liver cancer patients and 1408 hospital controls. The elevated risk of liver cancer associated with family history was not modified by adjustment for tobacco, alcohol, and personal history of cirrhosis and hepatitis (RR = 2.9; 95% Cl, 1.5 to 5.3) (Fernandez et al, 1994). Further studies are needed to confirm the existence of a genetic component in the familial aggregation of liver cancer. Hepatoblastoma of children, another common histological type of liver cancer has been found to be associated with familial adenomatous polyposis (FAP), which is believed to be a germ line mutation of adenomatous polyposis coli (Giardiello et al, 1996). 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A human T-lymphotropic virus type I infected cell line MT-2 was fairly sensitive to hepatitis C virus (HCV) infection and capable of supporting HCV replication (Kato et al, 1995). This finding may explain partially the association of liver cancer among family members and children’s leukemia if a virus infection was the mechanism for leukemia (see Table 14). The present study showed an overall 6.7-fold increased risk for all relatives of ALL patients. The increased risk was observed among both male and female relatives, and among second degree relatives. It is probably the only study that found an association between childhood leukemia and liver cancer among family members. The mechanisms behind the elevated risk were not known. One of the possible explanations is that the self reported liver cancer included both primary and secondary (metastatic) liver tumors. The interaction of environmental factors (such as viral infection) and genetic predisposition might be another possible reason. Pancreas Cancer The etiology of pancreatic cancer is largely unknown. Exogenous risk factors commonly cited in the literature include smoking, fat and meat intake, and lack of fruits or vegetables in the diet (Ponz de Leon et al, 1994). There might be a genetic component in the etiology of pancreatic cancer. Previous studies suggest that 5% of pancreas cancer patients show positive family histories. These cancers are common in the neoplastic spectrum of Lynch syndrome II (Ponz de Leon et al, 1994). The relatives of four brothers all of whom developed 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pancreas cancers included cases of colorectal tumors, prostate cancer, and a carcinoma ‘in situ’ of the cervix uteri. Many family members of pancreas cancer patients reported other cancers of breast, colorectal, lung, and melanoma, as well as pancreatic malignancies. Individuals with family history of pancreatic or other cancers are at greater risk of pancreatic cancer. The risk of pancreas cancer and cancer of the stomach and kidneys in females is raised in relatives of breast cancer patients, but it needs to be confirmed further (Tulinius et al, 1994). Significant associations were observed between family history of pancreatic cancer and pancreatic cancer (RR = 3.0; 95% Cl, 1.4 to 6.6) among 362 pancreatic cancer and 1408 hospital controls. The risk for pancreatic cancer did not appreciably change after allowance for tobacco, alcohol, dietary factors, and medical history of diabetes and pancreatitis (RR = 2.8; 95% Cl, 1.3 to 6.3) (Fernandez et al, 1994). Significantly increased risk of pancreatic cancer was found for subjects whose first-degree relative had cancers of the pancreas (OR = 3.2), colon (OR = 1.7) or ovary (OR = 5.3). The findings are consistent with the familial predisposition reported for pancreatic cancer and its association with hereditary non-polyposis colon cancer (Silverman et al, 1999). In an earlier large case-control study (179 case-control pairs), it was found that 7.8% of the pancreatic cancer patients reported a positive family history of the same disease, as compared with 0.6% among controls, a 13-fold difference between cases and controls. If this unusual aggregation of familial pancreatic cancer was not attributable to reporting 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bias, it may suggest the potential importance of genetic component in the development of pancreatic cancer (Ghadirian etal, 1991). The gene(s) responsible for familial pancreatic cancer has not been identified so far. Incidence of p53 expression in pancreatic tumors was found to be significantly lower in patients with a family history of cancer compared to those without. The findings suggest that p53 germ line mutations might be responsible for a small proportion of individuals at an elevated risk of pancreatic cancer as a component of the Li-Fraumeni syndrome. The other as yet unidentified or inherited or familial risk factors might be related to a majority of pancreatic tumor patients with normal p53 protein (Dergham et al, 1997). It was reported that pancreatic cancer as well as breast cancer, ovary cancer and leukemia was likely associated with ataxia-telangiectasia heterozygosity (Swift et al, 1990). Recently, germline mutations in BRCA2, p i6 have been shown to predispose to pancreatic cancer as well as breast cancer and melanoma (Hruban et al, 1999). The current study demonstrates a 2.4-fold increased risk of pancreatic cancer among all relatives of childhood ALL patients and it is consistent with previous 2-5 fold increased risk associated with family history of pancreatic, colon and ovary cancer. Together with the elevated risk of leukemia, laryngeal cancer, breast, colon cancer and melanoma among family members of ALL patients, germline p53, BRCA2, hMSH2 or hMLHl, and pi 6 mutations might further explain of elevated pancreatic cancer risk in a subset of the population. 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Breast Cancer Previous studies have estimated the hereditary fraction to be approximately 8% of all breast tumors. A family history of breast cancer is the single strongest risk factor for developing breast cancer (Ponz de Leon et al, 1994). The highest breast cancer risk is believed to be in daughters or sisters of breast cancer patients. It was found that a history of breast cancer in relatives was strongly associated with breast cancer risk (OR: 2.95 [1.63-5.34]) in a large matched case-control study (Ghadirian et al, 1998). Relative risk (RR) for breast cancer associated with a history of first-degree relative having a breast cancer was 1.70 [95% confidence interval (Cl): 1.55-1.87] and 2.34 (95% Cl, 1.80-3.02) if the mother or the sister had breast cancer at age 45 years or younger. The risk associated with family history of breast cancer seems to be largely independent of other known risk factors (Egan et al, 1998). Similar results were produced in three large case-control studies in New Zealand, US and Sweden. The relative risks for breast cancer among first-degree and second-degree relatives were 2.6 and 1.7 respectively in the New Zealand study (McCredie et al, 1997); 2.45 (95% Cl: 1.84 - 3.06) and 1.82 (95% Cl: 1.39 -2.24) respectively in the US study (Slattery et al, 1993) and 1.96 (95% Cl: 1.67-2.3) among first-degree relatives in the Swedish study (Magnusson et al, 1998). In two earlier studies, the risk of breast cancer was 2.1 (95% Cl: 1.6- 2.8) among women whose mother had breast cancer diagnosed before the age of 40 years and was 2.5 (95% Cl: 1.5 -4.2) among those who had a sister with breast cancer (Colditz et al, 1993), and a 1.9 fold (95% Cl: 1.6-2.3) risk was associated with history of first- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. degree relatives with breast cancer (Byrne et al, 1991). The pooled estimate of relative risk (RR) from a meta-analysis of 74 studies indicated that the risk associated with various family histories was 1.9 (95% Cl, 1.7-2.0) among relatives, 2.1 (Cl - 2.0, 2.2) among first-degree relatives and 1.5 (Cl = 1.4, 1.6) among second-degree relatives. Relative risks were increased in subjects under age 50 when a relative had been diagnosed before age 50 (Pharoah et al, 1997). A couple of genes have been identified as being associated with breast cancer risk. It was found that mutations in BRCA1 and BRCA2 are present in 59% (10 out of 17) Ashkenazi Jewish families with four or more breast or ovarian cancers (Schubert et al, 1997). Similarly, 47% of Welsh families (8 out of 17) with two or more cases of breast cancer under age 50 and/or ovarian cancer had demonstrated mutations in BRCA1 or BRCA2 (Lancaster et al, 1998). Similar results were found in a recent large population-based study among young women with breast cancer in Britain. Overall, there were 4.9% of women diagnosed before age 45 who were carriers of either BRCA1 or BRCA2 ( 30 of 617). This percentage of carriers increased to 45% (5 of 11) if patients had two or more affected first- or second- degree relatives with breast or ovarian cancer by age 60 years (Peto et al, 1999). In another study, mutations of BRCA 1 were found in 14 (61%) of the 23 families with hereditary breast cancer and no BRCA1 or BRCA2 mutations were detected in five (22%) families (Rebbeck et al, 1996). Among those who were predicted noncarriers of BRCA 1 or BRCA2, case subjects were 2.06 times (95% Cl =1.69-2.50) and 1.24 times (95% Cl = 1.17-1.32) 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. more likely to report a first-degree or second-degree family history of breast cancer, respectively, than were control subjects (Claus et al, 1998). It was concluded that family history remains a factor in predicting breast cancer risk even among predicted noncarriers of BRCA 1 and/or BRCA2 mutations. These results suggested that although BRCA1 and BRCA2 are the most important genes which are present in about 50% of breast cancer patients who had strong family history of breast and/or ovarian cancer, one or more additional genes may yet be found that explain the remaining proportion of hereditary breast cancer. Heterogeneity within and across families in the pattern of cancer susceptibility has suggested that different susceptibility alleles may exist. Li-Fraumeni syndrome is characterized by early onset breast cancer, and by the occurrence of several other neoplasms including soft tissue sarcoma, osteosarcoma, brain tumors, leukemia and lung, laryngeal and adrenocortical cancer. p53 mutations in germ line cells were discovered among patients with this syndrome. Germline p5 3 mutations have also been found among patients with familial breast cancer, but to a much more limited extent than BRCA1. Other genes that have been associated with breast cancer risk are the androgen receptor (AR) gene, the ataxia telangiectasia (AT) gene and the Cowden disease gene PTEN/MMAC1. It was suspected that other genes such as AT may confer a less increased risk of the disease, but may account for a larger proportion of cases, but this hypothesis has yet to be confirmed. It was found that 3.4% (3 of 88 ) and 1% (1 of 100) of breast cancer patients with a family history of breast cancer carried a germ 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. line mutations of ataxia telangiectasia mutated (ATM) gene in two previous studies (Chen et al, 1998). On the other hand, other tumor types were found in BRCA1 mutation/haplotype carriers. A study in Sweden found germ line mutations in the BRCA1 breast and ovarian cancer susceptibility gene identified in 15 of 47 kindreds. Other tumors associated with BRCA1 mutation/haplotype carries include prostate, pancreas, skin, lung cancer and malignant melanoma (Johannsson et al, 1996). It was found that family history of prostate cancer, endometrial and ovarian cancers in some families could be associated with breast cancer risk (Anderson et al, 1993). It was reported in the literature that breast cancer and kidney tumors occurred more often among leukemia case relatives than control relatives. In our study, the same excess breast cancers were found but failed to support previous findings of more frequent kidney tumors among leukemia patient relatives (SIR=0.3). Despite the remarkable findings of molecular changes occurring in breast cancer, none seems to be unique to breast cancer, and none has been found in all breast cancers (Ponz de Leon et al, 1994). In a summary, previous studies have found a 2-3 fold higher risk of developing breast cancer among first degree relatives affected by breast cancer, compared to a 1.5-2 fold increase among second degree relatives. The present study is generally consistent with the previous research finding, although not completely. The present study showed a significant increase of 1.7 times more breast cancers among relatives of childhood ALL patients and significant 1.9, 1.6-fold increased 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. risks among younger (<45 years) and older relatives respectively. Our study did not find excess ovary cancer among all relatives which were described together with breast cancer in previous studies. The present study also demonstrated a 1.7 times (95% Cl: 1.3-2.2) increased risk of breast cancer among second degree relatives, which fell in the range of increased risk in previous studies (1.5-2 fold). However, the current study failed to indicate a significant excess of breast cancer among first degree relatives. It may result in sisters of ALL patients who were too young (likely in their childhood) to develop breast cancer, and young mothers of childhood ALL, who were likely under age 40 with low incidence of breast cancer even when genetic factors are involved. Larger sample size and long follow-up of these families are required for further evidence. According to the literature, the increased risk of breast cancer seems reliable because of high sensitivity of self-reported breast cancer. Genetic roles of increased risk in this study, especially among relatives younger than 45 years are indicated. The susceptibility gene BRCA1 was reported responsible for some families aggregating with female breast and ovarian cancers. Current findings in our study may support the Li-Fraumeni syndrome which is associated with germ line mutations of p53, a tumor suppressor gene mapped to chromosome 17p. This mutation is extremely rare and is unlikely to play a major role in etiology of hereditary breast cancer other than those in the Li-Fraumeni syndrome. The BRCA2 gene on the long arm of chromosome 13 not only confers risk to breast cancer but also contributes to the development of tumors of cervix, melanoma, leukemia and lymphoma (Grimmond et al, 1996). The findings in present study could also not 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exclude the possibility that a BRCA2 gene inherited in leukemia proband families contributed to the excess leukemia and breast cancers observed. In our data set, the contribution of BRCA1, BRCA2, ATM, P53 or other genes to this elevated risk of breast cancer and leukemia is not known. The standardized incidence ratio of the present study is similar to those relative risk estimates of a positive family history of cancer, suggesting to some extent having a child with ALL is an equivalently strong indicator as family history of breast cancer on increasing the risk of this disease. The possibility that genetic factors play a role is indicated and further studies are required. Cervical cancer Cervical cancer is characterized more as a sexually transmitted disease than as a hormonal related malignancy like ovarian cancer, endometrial cancer and breast cancer (Hulka, 1997). Infections of specific types of human papillomaviruses (HPV) constitute a major risk factor in development of 90% of cervical cancers (Munger, 1995; Sastre-Garau et al, 1996; zur Hausen et al, 1996). Detailed molecular analysis of cervical cancer cells revealed that the DNA of the cancer- associated papilloma viruses is usually integrated into the cellular genome of the carcinoma cells. Only two viral genes (E6 and E7) are always preserved in a functionally active state. However, it was found that deregulated expression of the papilloma virus E6-E7 genes appears not to be sufficient to evoke cervical cancer (von Rnebel Doeberitz, 1990). Host environment such as genetic predisposition 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. might play a role. It was suggested that certain human leukocyte antigen (HLA) haplotypes confer an increased risk for severe cervical dysplasia and invasive cervical carcinoma following human papillomavirus type 16, 18 (HPV 16,18) and several other infections. It implied that the dysregulation of immune responses to foreign antigens and loss of recognition of self and non-self antigens increased the susceptibility to HPV associated tumors (Apple et al, 1995; Odunsi et al, 1995; Allen et al, 1996;). P53 tumor suppressor protein and the product of the retinoblastoma susceptibility gene, pRB was expressed consistently in cervical carcinoma (Munger, 1995). Familial relative risks (FRRs) were calculated separately for mothers and daughters, and were between 1.8 and 2.3 for in situ and invasive cancers in a large Sweden population-based study (Hemminki et al, 1999). A comparison of cancers in mothers and daughters showed an association between cervical cancer and many cancer types observed in immunosuppressed patients, suggesting a role for a mild form of immunosuppression, in addition to sexual behavior leading to human papilloma-virus infections in familial cervical cancer. Whether or not there is a hereditary component in the etiology of cervical cancer was not certain (Hemminki et al, 1999). Recently, HPV types have been reported to be linked with a high percentage of non-melanoma skin cancers (basal and squamous cell carcinoma), and presence of HPV infections in cancers of oral cavity, larynx and esophagus (zur Hausen et al, 1996). This could suggest interactions of HPV causing mutagenic and aneuploidizing activities and environmental factors such as sunlight, smoking and alcohol exposures. 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The present study indicated a 1.7 times higher risk of cervical cancer and a significant 3.8 times higher risk of leukemia among relatives of ALL younger than 45 years of age. No such previous studies have been done. One possible explanation of the excess of cervical cancer could be due to excess of cervical cancer in situ, rather than invasive cervical cancers reported among family members since these two types of tumor were clearly distinguished at time of diagnosis. On the other hand, whether the infections in uterine cervix during pregnancy result in the excess of maternal cervical cancer and children’s leukemia through the pathway of viral protein related genomic instability, accumulation of chromosomal abnormalities, colonial expansion of malignant cells and its interaction with the host factors remains to be studied further. Uterus and other Endometrial Cancer There was an increased risk of endometrial cancer among the first-degree female relatives of endometrial cancer patients by nearly 3-fold [odds ratio (OR), 2.8; 95% Cl: 1.9-4.2]. Cases also reported significantly more colorectal cancer in family members than did controls (OR, 1.9; 95% Cl, 1.1-3.3). The observed family history of endometrial cancer as an independent risk factor for cancer of the endometrium, the association of family history of colorectal cancer and endometrial cancer suggests that genetic factors play an important role in the development of familial endometrial cancer and colorectal cancer (Gruber et al, 1996). Among the first and second degree relatives of families with more than three ovarian cancers 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reported, cancers of breast, uterus and prostate cancer were 2.5, 5 and 4.5 times, respectively, that of the general population (Jishi et al, 1995). It was reported that non-polyposis colorectal cancer (HNPCC) and colorectal carcinoma (CRC) have been related to endometrial cancer (EC) (Sumoi et al, 1995). A potential tumor suppressor gene PTEN/MMAC1 most often is found mutated in endometrial cancers (Risinger et al, 1997). Mutation of p53 tumor suppressor gene with its mutant p53 protein products occurred in 20% of endometrial adenocarcinomas (Berchuck et al, 1995). Microsatellite instability has been identified in 30% of the endometrial cancers and 6% of the cervical cancers (Berchuck et al, 1995). A recent study indicated that family history of cancer or carrying the codon 31 Arg allele of the WAF1 gene may be associated with development of endometrial carcinomas (odds ratio, 2.81, 4.33 respectively) (Hachiya et al, 1999). It was found that the first degree relatives of women with double primary cancers and not carrying a mutation of either hMSH2 or hMLHl are at greater relative risk (RR) of endometrial cancer (RR=5.4 ; Cl 2.0-11.7) as well as colorectal cancer (RR= 2.8; Cl 1.7-4.5). If the women with double primary cancers cany a mutation on either hMSH2 or hMLHl, their first degree relatives had a very high risk of colorectal cancer (RR = 8.1, Cl 3. 5-15.9) and endometrial cancer (RR = 23.8, Cl 6.4-61.0) (Millar et al, 1999). Although the molecular pathogenesis of endometrial cancer remains obscure, more recent evidence supports genetic factors might play a role in a subset of 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. endometrial cancer patients. Our study revealed a 3.8 fold and 2.2 fold increased risk of uterine cancer among both younger (<=44 years) and older (>44 years) relatives of ALL patients. There is also an elevated colon cancer risk among older family members of ALL patients. One needs to be cautious when interpreting the elevated risk since the sensitivity of self-reported uterine cancer is low (30%, Kerber et al, 1997). Whether or not there is a connection between uterine and colon cancer by carrying a mutation on either hMSH2 or hMLHl gene is not known. Further studies are required to address the hypothesis. Melanoma It was estimated that about 8-12% of all malignant melanoma (MM) have a genetic component. Ultraviolet sunlight exposure (especially among white, fair skinned individuals) is one of the major environmental factors for the development of MM. Carcinomas of other organs such as lung, breast, prostate, pancreas, and colorectum have been described in hereditary dysplastic nevus syndrome (HDNS) (Goldstein et al, 1995; Ponz de Leon et al, 1994). Significant excess of skin cancers including BCC (basal cell carcinoma), SCC (squamous cell carcinoma) have been found in chronic lymphocytic leukemia (CLL) patients. Cutaneous malignant melanoma (CMM) are reported more frequently in NHL (Non-Hodgkin’s lymphoma) but not CLL patients (Levi et al, 1996). Hematopoietic cytokine family, which includes leukemia inhibitory factor (LIF) and interleukin-11 (IL-11) has been associated with loss of growth inhibition 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in advanced melanomas. Deletions of DNA on chromosome 9p21-22 were frequently observed in cells derived from melanoma and ALL. The observation of more nervous system tumors as second malignancy to melanoma suggested differentiation toward neural crest, neuroepithelium and/or mesenchymal derivation and supported a putative association with a hereditary cancer susceptibility trait (Azizi et al, 1995). Germline mutations in INK4A that codes for the target protein of pl6 (9p21), CDK4, underlies susceptibility in very rare melanoma families. The INK4A mutations were found in 35% ( 8 of 22) families with three or more cases of melanoma (Newton Bishop et al, 1999). In a case-control study, the risk of multiple primary melanoma (MPM) was greatest in those with a family history of melanoma, with large numbers of benign nevi, and the presence of clinically or histologically atypical nevi. Germline mutations of CDKN2A (pi 6) were present in 26% (6 of 23) patients with MPM (Burden et al, 1999). In two earlier studies, germline mutations of CDKN2A (p i6) were present in 18% (5 of 28) melanoma patients who had at least one affected first or second degree relative and in 7.8% of the 64 Swedish melanoma kindreds that each included at least two first-degree relatives with melanoma and dysplastic nevus syndrome (FitzGerald et al, 1996; Platz et al, 1997). The risk of melanoma increased 1.1-1.2 times if one parent had a cancer other than melanoma. When both parents had melanoma, the respective risk in the offspring was 9.3. This genetic evidence for a causal role was significantly strengthened by the observation that p i6 was frequently inactivated in familial melanoma kindreds. 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Since then, a high frequency of pl6 gene alterations were observed in many primary tumors. Pancreatic carcinoma and digestive tract cancers were observed among families with familial atypical multiple mole melanoma (FAMMM) syndrome (Lynch et al, 1991; Bergman et al, 1990). The current study found a 2.5 fold increased risk of melanoma (among older relatives), 2.4 to 3.8 fold elevated risk for leukemia (among all and younger relatives), and 2.2 to 2.5 times more frequent brain /nervous system tumors (among all and younger relatives). These findings may provide indirect evidence of loss of genetic material (9p21-22), dysfunction of hematopoietic cytokine family and aberration of cell differentiation as reported in previous studies. Because of low sensitivity of self-reported melanoma, the excess of melanoma among leukemia relatives in the current study needs to be confirmed further. Brain and other Central Nervous System .(CNS) Cancers The relationship between tumors of nervous system and environmental factors is poorly understood. Brain tumors (i.e. medulloblastoma, glioblastoma) together with breast cancer, and sarcoma, define the clinical spectrum of Li- Fraumeni syndrome. The p53 germ line mutations were found in astrocytomas and choroid plexus tumors in two families, each having four members with familial brain tumor syndrome (Ch'ene et al, 1999; Vital et al, 1998). Neoplasms of central nervous occur with increased frequency in neurofibromatosis (NF-1). A 4-fold relative risk was reported among probands 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. compared to the expected numbers. Other types of cancers including gastric, lung, breast, malignant melanoma, thyroid cancer, sarcomas, and leukemia have been reported in excess among patients with neurofibromatosis. A significantly increased risk [standardized incidence ratio, SIR = 2.12, 95% confidence interval (Cl) = 1.18- 3.49] was shown for developing primary brain tumors among first degree relatives of astrocytoma patients. Interestingly, the increased risk among relatives was for astrocytoma only (SIR = 3.12, 95% Cl 1.42-5.92), and not for other brain tumors (Maimer et al, 1999). Among 163 cases of astrocytoma and their matched controls, six brain tumors were found among relatives of astrocytoma patients diagnosed between 0-4 years of age compared to none was found among control relatives. The significantly increased risks for brain tumors might suggest a genetic etiology (Kuijten et al, 1990). In the current study, a significantly increased brain tumor risk was found among young and all relatives of ALL patients (SIR=2.5, 2.2; 95% Cl: 1.0-4.4, 1.1- 3.7, respectively). The sensitivity of self-reported brain tumor is not known. However, usually there is high sensitivity of self-reported cancer among first degree relatives and among young relatives of family members. The excess of brain tumors among relatives younger than 45 years of childhood ALL indicated a possible genetic basis for an increased risk of leukemia, brain tumors and breast cancers among these younger family members. It could be part of the Li-Fraumeni syndrome, in which germline p53 mutations are responsible for the familial aggregation of these malignancies. However, how our findings of excess breast 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cancer, melanoma, leukemia, pancreatic, liver cancer relate to brain tumor and through which genetic mechanism is not certain. Further studies need to be carried out to explore these issues. Laryngeal Cancer The main etiologic factors for laryngeal cancer are alcohol and tobacco use together with low consumption of vegetables and fruits (Riboli et al, 1996). There was evidence that familial oral and pharyngeal cancers are attributable to smoking and drinking rather than a genetic predisposition. The excess smoking related cancers among male relatives supported this finding. However, the role of genetic factors could not be totally ruled out. Glutathione S-transferase M l (GSTM1) gene deletion was found more in smokers than non-smokers. In laryngeal cancer, GSTM1 null genotype was found more frequent in patients less than 60 years old than those 60 years or older (78.6% vs. 36.8%). The finding suggested the GSTM1 gene polymorphism potentially modify the risk for larynx cancer depending on smoking history and age (Kihara et al, 1997). Also, in larynx cancer the DNA adducts were observed to be derived predominantly from polycyclic aromatic hydrocarbons (PAHs) and were evident only in tissue from smokers. Since the variation in expression of cytochrome p450 (CYP) 2C9/10 was greater than 10-fold in different individuals, further investigation is needed to confirm this finding (Badawi et al, 1996). Most previous studies 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. concluded that environmental factors play the main role in the etiology of laryngeal cancer. Overall, the p53 mutational spectrum explained the differences among cancers of colon, lung, esophagus, breast, liver, brain, reticuloendothelial tissues and hemopoietic tissues. The induction of skin carcinoma by ultraviolet light is also indicated to be associated with the occurrence of p53 mutations (Harris, 1996). This most common genetic mutation may be contributed by both exogenous and endogenous factors that cause the predisposition to several human malignancies among leukemia relatives. In our present study, there was significantly more laryngeal cancer observed among older and all relatives of childhood ALL patients (SIR=3.4, 3.9; 95% Cl: 1.4- 6.4,1.8-6.9 respectively). This elevated risk was observed among both male (SIR=3.2, 95% Cl: 1.2-7.1) and female relatives (SIR=6.8, 95% Cl: 1.4-19.9). The assumption that there are more smokers and drinkers among males than females could not totally explain the increased risk among both male and female and higher risk among female relatives of ALL probands. Whether this excess of laryngeal cancer was part of Li-Fraumeni syndrome or due to carrying germ line mutations of yet to be identified gene(s), or by shared environmental factors requires further study. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Conclusion The results of this study indicate that colon, liver, larynx, melanoma and pancreas cancers are significantly more common among older family members. More brain/CNS tumors and leukemia were found among younger family members than in the general population, but not among older family members. Breast cancer was found to be higher in both young and old age groups. These findings of increased risk of cancer among relatives, especially among those younger than 45 year relatives of ALL patients, suggest a genetic susceptibility contributing to the development of childhood ALL. The findings of this study also leave us many questions. For example, if genetic factors instead of shared environmental factors are mainly responsible for the increased risk of leukemia and other cancers among family members of ALL, which specific gene or genes are associated with the increased risk? Carriers of the gene(s) are predisposed to only one type of cancer or predisposed to different types of cancer based on the age, gender, and other environmental exposures? That is to say, what genetic path(s) these family members of ALL patients shared are associated with increased risk of leukemia, brain tumor and breast cancer among younger relatives and colon, liver, larynx, melanoma, pancreas and uterus cancers among older relatives? The broad findings of this study will help us to formulate hypotheses for future studies. The B-903 study indicated there is an increased risk of leukemia, breast cancer and other common cancers among family members of childhood acute lymphocytic leukemia (ALL). The analysis of this study is complicated by the design of the 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. study, since family history of cancer is both an eligibility criterion and the end point to assess if there are excess cancers among family members. Development of new analytical methods are required to handle such data. Development of efficient methods to validate family history of cancer, radiation exposure, and to apply these data in the analysis is another area needs to be studied further. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4 - ATM Gene Mutation and Childhood Acute Lymphocytic Leukemia - A Grant Proposal Introduction Cancer is the most common natural cause of death among children, exceeded only by accidents. Although the incidence of childhood cancer is low, the potential years of life lost associated with childhood cancer exceeds any type of adult cancer. Acute lymphocytic leukemia (ALL) is the most common childhood malignancy. Although early prophylactic treatment of the central nervous system (CNS) and combination chemotherapy greatly improved disease-free survival over past decades, treatment remains eventually unsuccessful in about thirty percent of such patients. ALL is costly to the patient, the child’s family, and to the society. During treatment, patients must cope with adverse side effects and feelings of anxiety and depression. Behavioral problems among patients often arise as normal coping strategies are exceeded. Even among those ‘cured’ from the disease, patients often struggle with educational challenges, social, physiological, and psychological problems in their attempt to re-construct normal lives (Maguire, 1983). Despite decades of research, the etiology of ALL remains largely unknown. High dose radiation is one likely cause of leukemia. Leukemia risk was found to be 10-20 times higher among Japanese A-bomb survivors. However, direct evidence of an association of low dose radiation and childhood leukemia has been controversial, 1 4 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. so estimates of risk remain somewhat uncertain. Studies of demographic and other environmental factors have produced inconclusive findings. Because childhood ALL occurs at a very young age (median age: 4-5 years) with a lack of long term exposure to possible environmental risk factors, genetic factors may play a greater role in causing childhood ALL than with adult cancers. There is growing evidence in the literature suggesting the importance of genetic factors. Multiple cases of leukemia reported in the literature suggest a genetic etiology. Concordance rates of leukemia among monozygotic (MZ) twins have been found to be 5-25%, which is generally higher than would have occurred by chance. Increased susceptibility to childhood ALL appears to be associated with inheritance of a syndrome such as ataxia telangiectasia (A-T) and associated with susceptibility to environmental carcinogens such as xeroderma pigmentosum (Strong et al, 1977, as cited by Birch 1983). Researchers have further pursued the association of inheritance of a gene (or genes) in the development of childhood ALL. Earlier studies focused on somatic mutations in leukemic cells. Since the ataxia telangiectasia mutated (ATM) gene was cloned in 1995, several studies have demonstrated 30% of T-cell prolymphocytic leukemia (T-PLL) or T-ALL patients with a wide age range were carriers of a somatic or germline mutation of ATM gene. However, these studies were based on non randomly selected, small samples of cases and no control subjects were included in the studies 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to compare the rates of carrying an ATM mutation with the cases. The contribution of the ATM gene in the development of childhood ALL remains to be established. Specific Aims The overall objective of the proposed study is to determine whether subjects who are heterozygotes for a mutation in the ataxia telangiectasia mutated (ATM) gene have a higher than expected risk of developing T-cell childhood acute lymphocytic leukemia (T-ALL) than children without the ATM gene mutation. The specific aims of the study follow: 1. The primary aim of the study is to examine the association of germline mutations in the ATM gene and development of childhood T-ALL. The ATM germline mutation rate among T-ALL cases will be compared with population controls matched on age, gender and race. It is hypothesized that the ATM mutation rate will be higher among T-ALL cases than controls. 2. The secondary aim of the study is to explore the interaction effect of radiation and ATM gene mutations in the development of childhood T-ALL. It is hypothesized that relative risk from an ATM mutation will be higher for those receiving radiation exposure than those without. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Background ALL Incidence Leukemia is the most comm on type of childhood cancer, accounting for about 30-40% of childhood cancer. Acute lymphocytic leukemia (ALL) is the most common type of leukemia in Western countries, with an annual incidence rate of 40 cases per million children under 15 years of age. ALL comprises 60-80% of leukemia in children under 5 years of age in England, Wales, Australia and the United States. Approximately 20-30% more males than females develop ALL. This male excess is consistent through all ages except among infants among whom the male to female ratio is about 1. The incidence is higher in Whites than other races. The racial difference is most pronounced for children 2-5 years of age. Based on immunologic phenotyping, four subtypes of leukemia have been identified: T-cell ALL (thymic), B-cell ALL, common ALL (also called pre-B ALL), and null or unclassified ALL. Pre-B cell ALL is most common (73%), followed by T-ALL (11-20%), null or unclassified ALL (12%), and B-ALL (1-2%) (Pendergrass, 1985, Greaves, 1984; Hardisty, 1983). The young age peak (4-5 years) is dominated by Pre-B cell ALL, with a small contribution from T-ALL. The T-ALL age distribution is broader, peaking between 8 and 12 years of age. Unclassified ALL has a sharp peak in infancy with a persistent low level from 3 to 50 years of age. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. T-cell Leukemia among A -T Patients There is a large increase in lymphoid malignancy among A-T patients and a total absence of myeloid tumors. Penetrance of the tumor phenotype is about 10- 15% by early adulthood. The increase in lymphoid malignancy includes both B-and T-cell tumors. T-cell tumors may occur at any age and may be T-ALL, T-cell lymphoma, or T-prolymphocytic leukemia (T-PLL). There also appears to be a 4 to 5-fold increased frequency of T-cell tumors, relative to B-cell tumors among A-T patients. Thus, it is possible that a significant proportion of T-ALL/T-cell lymphoma children may be associated with undiagnosed A-T (Taylor et al, 1996). A-T siblings have been shown to have clonal chromosome rearrangements of both B and T cells, simultaneously, but in these siblings the T-cell clones occupied the entire T-cell compartment and the B-cell clones were small. An important inference from these facts is that the A-T defect preferentially affects immune system gene recombination in T cells rather than B cells. A-T children predispose to develop T-ALL, the A-T adult patients have particular susceptibility to develop T cell prolymphocytic leukemia (T-PLL). T-PLL is very aggressive and the median survival is only several months. It was concluded in the limited series of studies that T-cell leukemia is predominant in A-T patients. The more common pre-B cell leukemia is less commonly observed. Myeloid leukemia has not been reported in A- T patients (Taylor et al, 1996). 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It was speculated that the same genetic events are associated with T-PLL in A-T patients and T-PLL in non-A-T patients because of similarities in both clinical and laboratory manifestations ( Stilgenbauer et al, 1997). Given the marked predisposition of A-T patients to develop neoplasms of the T-cell lineage, a series of T-cell ALL in children will be studied to search for an association with ATM mutations in this proposal. Ionizing Radiation The most clearly etiologic agent associated with acute leukemia is ionizing irradiation. High level exposure to ionizing radiation is strongly associated with increased ALL risk. Japanese A-bomb survivors had a probability of approximately 1 in 60 of developing leukemia within 12 years of exposure. ALL occurred most frequently in those less than 15 years old at the time of exposure, with the peak occurring within 8 years of exposure (Neglia et al, 1988). The rate of mortality of leukemia is significantly elevated at 0.4 Gy and above but not significantly elevated at lower doses. At bone marrow does of 3-4 Gy, the estimated dose-response curve peaks and turns downward (BEIRV, 1990). The saturation of the leukemia response at high doses has been attributed to the reduced survival of potentially transformed myeloblasts in the range above 3-4 Gy. A recent study of Japanese A-bomb survivors with doses less than 0.5 Sv found that there is a statistically significant risk of solid tumors in the range 0-0.1 Sv, and an upper confidence limit on any possible threshold is computed as 0.06 Sv (Pierce et al, 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2000). The study relates only to solid cancer, not leukemia. Further study of low dose radiation on the risk of leukemia will provide further evidence of low dose radiation in the etiology of leukemia. Further evidence of the strong association of radiation and ALL was provided by patients with polycythemia vera or ankylosing spondylitis treated with ionizing irradiation. One in six polycythemia vera patients developed leukemia within 12 years of exposure through treatment. One in 720 ankylosing spondylitis patients developed leukemia within 25 years of exposure (Pendergrass et al, 1985). One in 1073 women with invasive cervical cancer who received radiation therapy developed leukemia. The excess of leukemia in the irradiated women with cervical cancer was confined to acute and myeloid leukemia with 58 observed and 41 expected (RR=1.4, 95% 01=1.1-1.8) (Day and Boice, 1983). Chromosome fragility following the irradiation treatment was observed among many of these patients. These findings, together with the high degree chromosome fragility observed in hereditary conditions such as ataxia telangiectasia, suggest that chromosome damage may play an important role in the development of ALL (Pendergrass et al, 1985). Direct observational evidence of the association of low-level radiation exposure with ALL remains elusive. The classic study by Stewart et al (1956, 1958) showed a 2-fold increased leukemia mortality among children of mothers who received diagnostic radiographs during pregnancy. It may indicate an usually high susceptibility of the embryo and fetus to radiation. Considerable debate continues about subsequent reports of associations of preconceptional, intrauterine, 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. occupational exposures with increased childhood leukemia. There is no consistent evidence to suggest that children in utero at the time of A-bomb in Japan developed ALL at a higher rate than expected. No consistent association has been found between the use of diagnostic irradiation postnatally and ALL (Neglia et al, 1988). It has been estimated that about 1% of all leukemia cases in the general population may be attributable to diagnostic radiography (BEIR V, 1990). However, Thomas and Preston-Martin estimated that approximately 12% instead of 1% leukemias may be attributable to diagnostic x-rays using BEIR V risk estimates and age-specific population dose estimates (Thomas et al, 1992). Increased incidence of acute leukemia and chronic granulocytic leukemia were shown among radiologists in the U.S., U.K. and P.R.China. Because of uncertain doses to the bone marrow in occupationally and internally irradiated populations, it is not clear how their risks per unit dose compare with those in the more acutely irradiated populations such as A- bomb survivors. For high dose radiation, more than a 5-fold increased risk of leukemia was reported. One of these is the Stevens (1990) Utah fallout study which found there was a 5.3- fold increased risk of dying from ALL if exposed at high dose (6-30 mGy) compared to being exposed at low dose (0-2.9 mGy). For low dose radiation such as diagnostic radiography, its association with leukemia remains uncertain. Studies with positive findings have usually found a 1.5 to 4-fold increased risk among those exposed to low dose radiation. 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In summary, over the past decade, epidemiologic studies of childhood leukemia have revealed little consistent information regarding etiology, including the association between low dose radiation and childhood leukemia. In most cases, the relative risks were not evaluated for specific histological types of leukemia (such as ALL), or immunophenotypes (i.e. T or B cell ALL), or age-specific groups (such as infant leukemia), much less focus on the genetic susceptibility. However, important etiologic clues might exist in these homogeneous subgroups (Ross, 1999). One possible explanation for inconsistent radiation findings is that children may have different genetic susceptibilities to ALL. Ataxia telangiectasia (A-T) patients who are carriers of two copies of mutated ATM gene and the presumed carriers of one copy of the mutated gene (parents of the A-T patients) have been found to be more sensitive to radiation than those who do not carry the mutated gene (Paterson, et al, 1979). Recently, it was found that the association of childhood ALL and diagnostic radiation was modified in the presence of variants of DNA repair genes such as hMSH3, XRCCI, and probably hMLHl (Infante-Rivard et al, 2000). Although the results were not statistically significant and must be interpreted with caution, the effect of polymorphisms in DNA repair genes in the development of childhood ALL deserves further study. This study will explore the association of carrying a mutation in the ATM gene and the development of childhood ALL and also explore whether there is a significant association of radiation and ALL among those carrying an ATM gene mutation. 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ataxia Telangiectasia (A-T) There is a 61-fold and 184-fold increased cancer incidence among White and African American A-T patients, respectively (Paterson et al, 1964, Taylor et al, 1975). Lymphomas and lymphocytic leukemia are the most common malignancies among A-T patients, comprising 60% and 27%, respectively (Morrell et al, 1986). There is an increased risk of cancers among A-T blood relatives. An estimated 2.3 and 3-fold relative risk among men and women of developing cancers (all types combined) was found among ATM heterozygotes relative to noncarrier spouses (Swift et al, 1990). The estimated cancer incidence rate ratio was 3.8 among men and 3.5 among women among ATM heterozygotes compared to spouse controls. The odds ratio of exposure to diagnostic x-ray, therapeutic, or occupational radiation to the development of breast cancer among A-T relatives was 5.8, compared to controls matched on year of birth (Swift et al, 1991). The relative risk of developing cancers among ATM heterozygotes was 6.1 relative to spouse controls (Morrell et al, 1990). There is an increased risk of breast cancers among A-T blood relatives. A relative risk of breast cancer among these ATM gene carriers of was estimated as 6.8-7.6 and an estimated 8.8-18% of all patients with breast cancer are ATM gene heterozygotes (Swift et al, 1986, 1987). The increased risk of leukemia and lymphoma among A-T relatives remains inconclusive. One study found that five blood relatives died of leukemia or 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. lymphoma before age 45 among 27 A-T families compared to one expected death (P<0.01). The relative risk of dying of leukemia and lymphoma before the age of 45 among ATM heterozygotes was estimated to be 7 times (P<0.03) the risk in the US white population using a maximum likelihood method (Swift et al, 1976). In another study, there were five (5) cases of leukemia and lymphoma (two lymphomas, two CLL, one ALL) observed when only three (3) were expected. However, the excess of leukemia and lymphoma was not significant (Morrell et al, 1990). Lymphoid malignancy occurred at a higher rate among White non-Amish blood relatives than spouse controls (13 vs. 1). However, no firm conclusion could be made (Swift et al, 1986, 1987). A meta-analysis of the data from four previous studies yield an estimate of the allele frequency to be 0.2-1%, with 0,5% being the best point estimate. The pooled estimated relative risk of breast cancer was 3.9 (95% Cl: 2.1-7.2) and 1.9 (95% Cl 1.5-2.5) for any other cancers (Easton, 1994). In conclusion, more cancers occurred among A-T relatives than was observed among the general population. Leukemia is one of the most common malignancies among A-T patients who are carriers of two ATM mutations. Leukemia was found more among A-T relatives and among presumed carriers of one copy of the mutated gene (parents of the A-T patients) although this elevated risk was based on small numbers of leukemia cases. The risk for childhood leukemia among A-T patient siblings and cousins has not been estimated. 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The association between ATM heterozygotes and leukemia has not been thoroughly studied. Since the ATM gene was not cloned until 1995, it was not possible to determine ATM gene status among A-T relatives in these previous studies. Furthermore, the estimation of leukemia and lymphoma was based on those relatives older than 20 years of age. The exclusion of childhood leukemia and lymphoma may have resulted in an underestimation of the risk since children are at greater risk of developing acute leukemia compared to an adult aging 25 to 65. It is reasonable to hypothesize that more ATM gene carriers may be found among T cell childhood ALL patients. Moreover, children carrying one copy of the mutated ATM gene may be at higher risk of developing ALL than randomly selected population controls. ATM gene The ATM gene is located on chromosome band 1 lq22-23 containing 13,000 base pairs and 66 exons that cover 150 kb of genomic DNA. The largest exon is exon 12 which contains 372 nucleotides. The gene product is also large, containing 3056 amino acids. ATM is expressed in all types of tissues tested, which include tissues of the pancreas, kidney, skeletal muscle, liver, lung, placenta, brain, thymus, prostate, testis, ovary, small intestine, colon and leukocytes (Savitsky et al, 1995, Gatti, 1998). Alternative splicing occurs at the 5' end. The 3' portion of the gene has strong homology to phosphatidylinositol 3-kinase (PI-3K). Although the function of 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the ATM gene is unknown, it is believed to function as an intracellular signal transduction, warning the cell to repair DNA damages before next cell division. There are two types of ATM gene mutations identified at the level of genomic DNA: deletion or insertion of one or a few nucleotides, and point mutations. The first type of mutation (also called frameshift mutations) usually results in changes of the triple codes for all the codons that follow, thus altering the sequence of the amino acids. In the ATM gene, these mutations frequently lead to premature termination of the protein (truncated protein) by creating a stop codon. Until recently, 70% of detected ATM mutations have resulted in shortened (truncated) proteins. The second type of ATM gene mutation is point mutation, which may (missense mutation) or may not (silent mutation) affect codons by a single nucleotide substitution. The influence of missense mutations in ATM is more complicated than frameshift mutations. A missense mutation can alter the codon for a particular amino acid, and therefore change the function or structure of proteins. However, sometimes missense mutations are difficult to distinguish from polymorphisms. Moreover, a point mutation that fails to change the coding of an amino acid can be deleterious one (Gatti, 1998). Over 250 mutations in the ATM gene have been identified. About 70% of ATM mutations result in truncated proteins, most of which affect splice sites. Most A-T patients are compound heterozygotes. That is, most non-consanguineous A-T patients have two distinct mutations. One potential ‘hot spot’ is exon 54. Fifteen percent (15%) of ATM mutations have been identified at exon 54. 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ATM Gene Status in Leukemia Since the ATM gene was cloned in 1995, six studies have analyzed ATM mutations in leukemia patients. Most of these studies have had small samples Of T- cell pro-lymphocytic leukemia (T-PLL). Only one study has examined ATM mutations among ALL patients. The ATM gene was found inactivated in a rare sporadic malignancy, T-cell prolymphocytic leukemia (T-PLL), which is often associated with cytogenetic aberrations of chromosome 14. T-PLL is a rare form of mature leukemia that occurs both in adults as a sporadic disease and in younger patients suffering a hereditary condition, such as A-T. A-T patients are predisposed to develop T-PLL. High incidence of T-cell leukemia/lymphomas was observed in ATM-deficient mice. The ATM gene was shown to sustain frequent loss-of-function mutations in T-PLL tumor cells, consistent with functioning as a tumor suppressor gene in this type of leukemia (Luo etal, 1998). Luo et al (1998) studied 19 T-ALL patients for ATM mutations. These T- ALL patients exhibited rare nucleotide substitutions not previously found in the ATM gene. Rare polymorphisms were found among six patients (32%). Five of these six point mutations (26%) were identified in the germ-line, indicating constitutional polymorphisms. The lack of small deletions/insertions observed among these 19 patients and point germline mutation rate as high as 26% need to be further verified. Because 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. results were based on a small sample of T-ALL patients, further studies are needed to confirm these preliminary findings. Results from these initial studies of mutations on ATM gene among leukemia patients may be summarized as follows: 1) the studies have focused on T-lineage ALL or PLL with 32% -62.5% somatic mutations identified; 2) missense mutations are dominant in T-PLL and T-ALL, in contrast to mostly truncated mutations found among A-T patients; 3) almost all mutations identified so far have been unique regarding mutation types and locations on the ATM gene (i.e., no single so called ‘hot spot’ has been identified among leukemia patients); 4) most studies, except one, revealed somatic mutations instead of constitutional mutations. Several limitations of these studies precluded the estimation of the relative risk for developing leukemia associated with carrying a mutation on ATM gene. First, the representativeness of the source population was questionable given a lack of detailed description regarding how these cases (sample sizes ranging from 8 to 37 patients) were ascertained. Second, in at least one study, the protein truncation test (PTT) was used to detect missense mutations which were known not sensitive to PTT. Even when single strand conformational polymorphism (SSCP) was used to identify missense mutations, the sensitivity was unclear. Third, the selection of ‘unmatched controls’ (‘unaffected individuals’) was not adequately described. Finally, the analyses among unaffected individuals were generally confined to 1 or 2 exons rather than screening the entire gene for mutations. 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. These studies raise several questions: 1) whether these somatic mutations among T-PLL could also been identified in the germline; 2) whether ALL patients have the same or different type and frequency of ATM gene mutations as T-PLL patients; and, 3) the generalization of majority germline mutations identified in a small sample of T-ALL patients in one previous study to majority childhood ALL is uncertain. Further more detailed questions include: 1) whether these nucleotide substitutions cause loss of function of ATM gene or they are “neutral” polymorphisms; 2) whether the number of missense mutation is significant different from expected; 3) what is the difference between the missense mutation or truncated mutation among leukemia patients compared to those found among AT patients and the general population? Much research remains to be done to estimate the risk of leukemia among ATM heterozygotes. The proposed study will examine the mutations of the ATM gene of both cases and matched controls to examine the associated risk of carrying a mutation on ATM gene on the development of childhood ALL. Significance The question of whether ALL incidence is higher among ATM gene carriers compared to matched controls is relatively new, made possible by recent advances in genetic analytic techniques. For a long time, ATM heterozygote status could only be determined among the relatives of A-T patients by their relationship and disease status. Now, our ability to identify mutations in ATM may allow association of 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ATM heterozygotes with leukemia to be established relative to matched controls. Thus, the discovery of an association of ATM carriers and childhood T-ALL would extend our knowledge of cancer risk to the molecular level. The answers to these questions may provide valuable information for identifying high-risk families of childhood ALL and learning more about the etiology of childhood ALL. The discovery of reliable genetic markers for identifying ATM gene carriers may also be useful in genetic counseling of families and investigation of whether ATM carriers worldwide comprise a significant proportion of cancers such as breast cancer and leukemia, and patients showing excessive sensitivity to conventional doses of radiation therapy. The discovery may also help us understand the complex interactions of apparently unrelated physiological systems, such as immunodeficiency and radiosensitivity. Preliminary Study Increased Risk of Leukemia and Breast Cancer among Family Members of Childhood ALL The Childhood Cancer Group (CCG) B-903 study (Jonathan Buckley, P.I.) was a retrospective cohort study covering the period 1990-96, involving 294 ALL patients (probands) with a family history of cancer and their 3860 family members in this analysis. Family history information was collected for first degree and second degree relatives of newly diagnosed ALL probands under 22 years of age treated at a CCG affiliated hospital. Standardized incidence ratios (SIRs) were calculated for 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. each type of cancer by dividing the number of observed by the number of expected incident cases of cancer among ALL family members, controlling for age. Family members of ALL patients were 2.6 times as likely to develop leukemia than expected calculated from age and gender-specific cancer rates in the general population (SIR=2.6, 95% 0=1.4-4.0). A significant four-fold increased risk of developing leukemia was reported among young relatives (SIR=4.1, 95% CI=2.2-6.5) when all racial/ethnic groups were combined. The excess of leukemia cases among young family members was consistent across all three racial/ethnic groups with 3.8, 3.9, and 17.6-fold increased risk among White, Hispanic and African Americans relatives, respectively. The incidence of female breast cancer was consistently higher overall for all racial/ethnic and age groups. 1.8 times more cases of breast cancer were reported among relatives than expected based on the general population (SIR=1.8, 95% Cl— 1.4-2.3). Among relatives of White, Hispanic and African American younger than 45 years old, there was a significant 2.4-fold increased risk of having breast cancer (SIR=2.4, 95% 0=1.5-3.5). Similarly, older relatives were 1.6 times more likely to develop breast cancer than were adults of the same age and gender from the general population (SIR=1.6, 95% 0=1.1-2.1). The excess of cases reported by both younger and older relatives was consistent across all three racial/ethnic groups: 2.3, 1.5, and 1.7-fold increased risks were found among Whites, 2.6, 2.7, and 2.7- fold increased risks were found among Hispanics, and 4.6, 3.1, and 3.7-fold greater risks were found among African Americans was reported. 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The association between ATM heterozygotes and leukemia remains unclear. In previous studies, risk estimates for leukemia and lymphoma among A-T relatives were reported only combined (Swift et al, 1976, 1987, Morrell et al, 1990). The pooled estimated relative risk of breast cancer among heterozygotes was 3.9 (95% Cl: 2.1-7.2) (Easton, 1994). The increased leukemia and breast cancer risk among family members of childhood ALL, especially relatives under 45 years old in B-903 study suggests a genetic etiology. Carriers of the ATM gene will have increased susceptibility to leukemia and breast cancer. It is reasonable to investigate if there is an association between the ATM gene and childhood ALL. The B-903 study provides a data screening and collection instrument and data collection network to collect information on family history of cancer and other hereditary diseases among all childhood cancer patients treated at CCG institutions. It can be used to screen newly diagnosed ALL patients and identify those having cancers among first and second degree of relatives. These ALL patients are thought to have a higher risk of carrying a mutation in the ATM gene than those without a family history of cancer among any first-degree relatives or among second-degree relatives under age 45. The B-903 instrument will be adapted and modified for use in the proposed study. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Research Design and Methods Overview The proposed postdoctoral fellowship population-based case-control study will examine the association between ATM gene mutations and the development of childhood T-ALL, and its interaction with low dose radiation exposure in the development of childhood ALL. ATM gene mutation rates and radiation exposures among newly diagnosed ALL with a family history of cancers will be compared with their age, gender and race matched controls. To address aim 1 of the study, blood samples for both cases and controls will be drawn, stored, and shipped to the network laboratory for analysis. Constitutional DNA will be extracted and germline mutations in the entire ATM gene will be screened for each case and control by SSCP which can identify point mutations. The association between ATM mutations and childhood T-ALL will be analyzed using conditional logistic regression. Assuming that the carrier rate of a mutation in the ATM gene is 1% among controls, the selection of 125 case-control pairs will result in more than 80% power to detect a relative risk of 7. To address the aim 2 of the study, self reported radiation exposure information on paternal, maternal, and study subject diagnostic X-rays and other radiation will be collected through telephone interviews with parents or guardians of the participants. The association between ATM gene status and childhood T-ALL and their interaction effect will be tested using conditional logistic regression. 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This study will advance the current state of knowledge by examining the association of ATM and T-ALL. Newly diagnosed T-ALL patients will be selected from a national network of children’s hospitals during the one-year period 1/1/2001- 12/31/2001. Controls will be matched on age, gender and race using RDD. An estimated 125 case-control pairs will be obtained over a one-year period. The findings of the study may justify a need for a larger study investigating the possible etiologic role of ATM gene mutations in the development of childhood ALL. Case Selection Patients registered at CCG institutions and diagnosed with T-ALL under 22 years of age in 2001 will be identified through the CCG registration database. All ethnic groups will be included. Based on the registration of approximate 5500 newly diagnosed childhood patients in current CCG studies annually (Bleyer, 1990), it is expected that 1155 ALL patients will be screened to enter the study. Based on estimated 12% of T-ALL , 139 T-ALL among ALL are expected to participate in the study. After considering patient or physician refusals to participate (8%), and lost to follow-up (2%), 90% (N=125) cases will be enrolled in this study. For each eligible, participating case, one matched control will be selected. Each newly diagnosed T-ALL must meet the following eligibility criteria: 1) The diagnosis of T-ALL must be made between January 1, 2001 and December 31, 2001; 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2) The patient’s physician must give permission to contact the parents of the case; 3) The family of the case must have had a telephone in the household in which they resided at the time the child was diagnosed with leukemia; 4) The biological mother of the case must be available, speak English or Spanish, and consent to be interviewed; 5) Patients must reside in the U.S.; This approach will allow cases with a higher ATM allele frequency to be identified, while still allowing estimation of population parameters. This approach will only take one year to accrue the target sample size and the proportion of cases carrying a mutation on the ATM gene (pi) might be higher than that of pooled ALL, although it is not known for certain at this stage. Control Selection Once the case is interviewed and has blood drawn, a control matched to the case by year of birth (+ 1 year), gender, and race/ethnicity will be randomly selected using random digital dialing (RDD). Since childhood ALL is comparatively rare, this specific histological type of childhood leukemia will be ascertained through national wide cooperative children hospitals affiliated to CCG. RDD provides an economical, and efficient method to select controls for cases residing throughout the US. 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Once a potential control is identified, a brief description of the study will be presented to eligible controls contacted after verbal consent to participate is obtained from parents or guardians. Names and mailing address will be obtained and a 45 minutes to 1-hour telephone interview will be scheduled. Arrangements will then be made for blood samples to be drawn at the pediatric clinic or at the nearest CCG affiliated hospital and sent to the collaborative lab. The advantages and disadvantages of RDD and other alternative methods of selection of controls are discussed in details in the following section. Selection o f Controls Methods for selecting controls considered include RDD, clinic-based controls, parents, siblings or friends of the cases or children treated for non- malignant disease at the same hospital treating the case. The strengths and weaknesses in term of bias and efficiency of these alternatives are discussed below. RDD Using RDD, controls will be selected from households in the same general geographic location as the case through the generation of a series of telephone numbers based upon the telephone number of the case. The area code, exchange, and first two digits of the number will be used to form a set of unique telephone numbers by randomly generating the last two digits. Once the first potential control number is obtained, a trained survey interviewer will call until one of the following 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. outcomes is achieved: non-working number, nonresidential, no answer after nine attempts, complete refusal, partial refusal, eligible (match available), or not eligible. The number is called three times per day (morning, afternoon, evening) for three days (one of which falls on the weekend) until a contact is made. If no contact is achieved after the nine attempts, the interviewer proceeds to the next potential control telephone number. A prepared text is provided to determine if a potential matched control is available. This method offers the advantage of approaching all households with phones in a given geographic area and comes close to sampling randomly from the general population (Rothman et al, 1998). Random-digit dialing is frequently used for control selection in epidemiological studies of childhood cancer. Although random-digit dialing may be limited by several difficulties including undercoverage, nonresponse, and the possibility of oversampling households with more children, it is popular due to its cost-effeetiveness and ability to efficiently select a suitable control population from a wide geographic region. One major problem of this approach is that it now takes a longer time than before to identify a potential control because of a recent increase in commercial calls to households resulting in unwillingness to talk in person over the phone. Previously, it is estimated that an average of 32-42 (median 28, range 2-339) residential calls are required to find a matched control (Robison et al, 1984, Ward et al, 1984). 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Until recently only limited data were available regarding the use of random digit dialing for control selection in analytic epidemiologic studies within pediatric populations. Subsequently, as pediatric clinical cooperative groups have become more commonly used for case ascertainment, it has been demonstrated that random digit dialing provides an acceptable and cost-effective method for control selection (Robison et al, 1995). In our most recently completed CCG study to date (E l5), which used similar matching criteria to obtain infant controls for the infant cases, an average of 94 calls (this includes businesses, disconnected lines, etc.) were required to obtain a matched control for cases <1 year of age. The maximum number of calls to find an infant control could be up to 200 calls (Dr. J Ross, personal communication). In this proposed study, controls will not be limited to infants, but with a broader range of age: from infants to young adults under 22 years. Therefore, it is expected that it will probably take a shorter time than an average of 94 calls to identify a matched control. However, it is not known if the refusal rate will increase once parents learn that a blood sample is required from each eligible control, which might increase the number of calls to find a suitable control. Clinic-Based Controls Controls could also be selected from the same pediatric clinic where the case visits his/her primary pediatrician for routine physical exam and vaccination. In addition to requesting parental contact of the ALL cases, the CCG physicians would also be asked to provide us with the name of the primary physician who initially 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. referred the child to the CCG institution for treatment. This primary (non-oncology) physician, typically a pediatrician, would be sent a letter describing the study, and requesting assistance in the identification of a normal control child with the same age (+ 1 year), gender and race. This would be accomplished by providing us with a list of all eligible normal children that are in his or her pediatric practice. This roster would include no patient identifiers; only date of birth, sex, and race of the children. Upon receipt of this roster, one control that fits the matching criteria (age, gender, and race) could be randomly selected. Upon selection of the control child, the parents would be contacted by the primary physician for consent to release their name and telephone number for further contact. The advantages of using clinic controls include saving time on phone calls to identify eligible controls plus the ease of blood draw. Once the control is found and is willing to participate to the study, the blood sample could be drawn at the clinic and shipped to the collaborative lab. The request for permission to enroll both case and control and releasing the information of these children’s medical history, treatment of the non-malignancy disease and family history of cancer could be obtained through the same clinic if this information is documented. The clinic staff might be more motivated and interested in finding the cause of childhood ALL and therefore, more willing to help to find a control. One main disadvantage of clinic controls is that the list of potential eligible controls depends on the pediatrician. It is critical that we obtain the complete list in order to select a control randomly to avoid selection bias. Moreover, the number of 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. children of the same age, gender and race might be limited in one clinic. If no control is available from within the clinic, rosters from a neighboring clinic or RDD might have to be utilized. Although using the same clinic adjusts for the social economic status (SES) to an extent, the other indicator of SES such as living in the same geographic area might not be as well matched between cases and controls as using RDD. The other unsolvable selection bias is that there will be much less opportunity to be selected as a control for children without insurance and primary pediatrician. However, this similar problem also exists among those households without a phone when using RDD method to find a control. The cost of finding a clinical control is not certain. Based on a recent CCG study, $400 is estimated as the time and cost for identifying a control with Down syndrome (Dr. J Ross, personal communication). It is possible that the time and cost could be reduced when finding a normal child without disease as a control. Therefore, we probably could conservatively estimate the cost as $400 for a clinic control. Parent or Sibling Controls The use of family member controls is both cost-effective and convenient. They are perfectly matched on ethnicity, common environmental factors, and other potential confounding variables. Lower refusal rates may be expected relative to unrelated controls. Blood samples of the case and his parents or siblings could be obtained during a single visit. 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Parents of these ALL patients are likely to be alive at the time of study since the disease occurs at a young age. Blood samples and genotypes of both parents could be obtained to avoid the potential bias when the genotype of one parent is missing. Using parent or sibling controls could also adjust for a form of confounding bias known as “population stratification”, a situation arises when both the gene frequency and penetrance differ across ethnic groups. Using parent controls will automatically match on FH, thus not posing the difficult of screening many potential controls to match on FH. However, the cost of ATM gene screening (estimated as $700-1000 /person) is higher than using unrelated controls since two individuals (both parents) instead of one are to be genotyped. The convenience and low cost of parent controls does not seem to offset its high cost of mutation detection. This approach is also limited by its ability to detect the main effect of the environmental exposure. Siblings are better matched controls than parents in term of age and being at risk of developing childhood ALL. Older siblings could be better candidates than younger siblings for controls unless there is a time trend since these older siblings are disease free at the age that the case had the disease. The potential biases of younger siblings having not yet reached the age of the cases requires statistical adjustment. Also, eligible siblings might not be available due to impaired parental reproduction associated with the studied genotype, or by choosing not to have another child after a child has been diagnosed with ALL. Even if the siblings are unaffected at the time of study, there is potential overmatching on environmental and 168 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. other factors including the genotype. This would likely reduce the efficiency of the study. Gauderman and Thomas (1999) and Witte et al (1999) compared the relative efficiency of a sibling control or a pseudosib (parent) control to a population control for estimating a major effect of a recessive gene. Sib controls are the least efficient, requiring about twice the sample size compared to using unrelated controls (58% efficiency). The pseudosib controls are more efficient for recessive genes. However, three genotypes instead of two were required per case. After considering the high cost of screening mutations on ATM gene, this design is less efficient. In summary, the greatest strength of family-based case control design is the avoidance of potential bias caused by variation of gene frequency and penetrance in different populations that can not be controlled by matching, stratification or statistical adjustment. Although the relevant but unknown ethnic effects could not be controlled by using RDD controls, RDD controls will be considered in this study since first, the extent of the bias causing by ethnic effects (also called “population stratification”) remains unknown. Secondly, as discussed above, family controls are less efficient than population-based controls. Finally, the high cost of screening mutations in the ATM gene for three individuals in the family-control design compared to screening mutations for two individuals when using RDD controls makes it more difficult practically to use family controls. The comparisons between using unrelated controls versus using parental controls are summarized in Table 18. 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Friend Controls Using friends of the cases as controls offers the advantage that the age and neighborhood are naturally matched, and it would likely be more convenient and easier to identify a more willing control subject to participate in the study. But this method poses two problems. First, being a friend of a case might be related to sharing certain exposures such as radiation from soil, building materials and radon. Second, the investigator who uses friend controls will depend on the cases to provide a list of friends and choose randomly from the list. The dependence on cases to provide a list of friends adds potential selection bias (Rothman et al, 1998). As a result, certain individuals who have many friends, will have more chance to be selected as the friend control, and will be over-represented in the control series. Therefore, any exposure associated with such individuals will be over-represented in the data set. The efficiency of this method over random digit dialing is not certain. Hospital-Based Controls Hospital-based controls are not considered in this proposed study because the cases of this study will be selected through a national network of children’s hospital rather than a small number of hospitals. Thus, referral patterns to the hospitals does not seem so important to be matched between cases and controls. Second, this nonrandom sampling of controls will possibly not be selected independently of exposure in the source population. Patients hospitalized with other diseases may be unrepresentative of the exposure distribution in the source population either because 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exposed individuals are more or less likely than nonexposed individuals to be hospitalized for the control disease if they develop them, or because the exposure may cause or prevent these control diseases in the first place. This second potential bias may impact this study in a less extent since the exposure of our interest is radiation and the common reason for hospitalization among children is accidents and other acute diseases. The association between these childhood acute diseases and radiation is less apparent than the relationship between those hospitalized adult diseases and exposures such as smoking, alcohol, obesity, etc. Finally, to recruit a sick child as a control during his a few days of hospitalization, the close collaboration with the parents and staff members of different departments is required. Data Collection Some information on ALL patients will be obtained using the CCG registration data base, including name, initials, residence zip code, method of payment, birth date, diagnosis date, ICDO morphology code, immunophenotype (T or B cell malignancy), race, gender and cancer therapy (chemotherapy, radiation therapy, etc.) if treated at CCG institutions. with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 18. Comparisons of Advantage and Disadvantage between Using Unrelated and Parental Controls Advantage Disadvantage Unrelated controls 1) Main effects of both G, E and GxE interactive effect can be estimated. 1) It may be used to detect GxE interaction only when G and E are common and the common alleles confer a modest or small effect on disease risk; 2) Population stratification or genetic admixture might adversely influence this design. Parental controls 1) Eliminate the systematic difference of genetic background by using relati ves of the case as the control subject; 2) Relatives are easily identified and are more willing to participate; 3) There may be a gain in efficiency for detecting GxE interaction involving a rare gene because of over-sampling on the genetic factor 1) It does not allow to estimate the main effect of the environmental exposure; 2) Difference in estimates across strata are not restricted to resulting from only gene-environment interaction. Differentiating between the various reasons for the differences may be difficult. 3) It is less efficient for detecting the genetic factor main effect and the gain in efficiency decreases as the frequency of the genetic factor increases; 4) There is a potential for overmatching, particularly on E exposure, leading to loss in efficiency.; 5) It could be expensive to genotype both parents; 6) There will be additional ethical issues associated with using related control. f O The following information will be obtained from the guardians of both patients and controls using a modified version of the B-903 instrument: 1) subject’s diagnostic X-ray exposure, location in the body and medical condition to have x-ray exam; 2) parents and siblings’ age, race, education, most recent occupation if applicable, smoking, alcohol history, parental contact to paints, solvents, petroleum and pesticides use; 3) family history of cancer, and major illness, genetic syndromes and abnormalities among first and second degree relatives, age at diagnosis, age at the time of interview if disease free, and age, cause of death, if applicable; 4) maternal and paternal radiation exposures including pre- and post-conception x-ray exams and occupational exposures, and radiation therapies etc.; 5) maternal reproductive history and infectious illness, use of medications during the relevant pregnancy; and 6) other medical conditions requiring x-ray exams to be performed among ALL patients and controls. One-hour telephone interview will be administered to obtain the information. A life event calendar focusing on medical events such as illness and medical care will be filled out before the interview begins. The life calendar has proven to be a useful technique in helping parents/guardians remember disorders and events that may have led to receive diagnostic radiation. Interviewers will be trained to follow interviewer instructions and script as worded and record responses in coding boxes. Interviewers will be trained to administer questionnaires consistently to all respondents, to prepare to answer questions, to code any responses that are not 173 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. precoded and update the coding manual. Collected data will be reviewed for consistency before input into a SAS database. Logical error checks will be further performed by a SAS program to ensure data quality. If necessary, the respondents will be recontacted to clarify or correct responses. Blood Collection and Processing Blood samples providing constitutional DNA to determine the presence of ATM gene mutations will be collected at the CCG treating institutions or at a pediatric clinic for both cases and controls. A kit for each study subject will be prepared and sent to the clinic or hospital where the blood sample is drawn. For most cases (98-99%), remission is achieved during the initial induction phase of chemotherapy. A blood sample will be collected once remission is achieved. The remaining 1% of patients who do not successfully achieve remission at the end of four-week induction of therapy will be excluded from the analysis. For patients, a 5- cc blood sample at the time of remission will be collected. For controls, a 5-cc of peripheral blood will be collected by a trained nurse at the time of a routine visit to the pediatric clinic or at the time of a scheduled visit to the nearby CCG institution. These blood samples should be drawn in a sodium heparin tube with anticoagulant to keep blood from clotting. The blood samples should not be put directly on ice or in a freezer to allow the white cells to be viable. The tubes should be labeled with patient initials, CCG registration number, gender and birth date and then stored and shipped 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to a lab for further processing. DNA will be extracted and analyzed for ATM gene mutations. Single Strand Conformational Polymorphism (SSCP) fo r ATM Gene Mutations SSCP is a simple method for detecting mutations. A segment containing a mutation is amplified by PCR from genomic DNA in the presence of a radiolabeled T 9 nucleotide, usually P. Then the PCR products are denatured to generate single stranded molecules and loaded on a non-denaturing gel. Conformational change of the single stranded mutant DNA is believed to be the reason for the mobility shift, which can be detected by electrophoresis. Then the suspect region is sequenced to confirm suspected mutations (Cotton et al, 1998). The conditions of the gel can be varied by alternating the running temperature, the degree of cross-linking in the gel matrix, or by the inclusion of glycerol or sucrose. These variations can change the type of confirmation seen and can increase the sensitivity of detection. Different types of gels, buffers and temperatures are usually tested prior to screening large number of gDNAs using the known mutations for each exon. It is believed that the detection efficiency of optimized SSCP will approach 60-70% (Gatti, personal communication). 1) SSCP is widely used and practical in mutation detection and SSCP is a better choice for screening genomic DNA. The advantages of this method of mutation detection include: 175 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2) Unlike PTT, SSCP can detect most point mutations within DNA segments smaller than 250 bp; 3) SSCP is rapid, not requiring any special equipment, and highly sensitive if properly optimized. Many different mutations within a DNA fragment can often be distinguished on the same gel; 4) Screening smaller segments increases both the accuracy and amount of work. Screening segments larger than 250 bp decreases the sensitivity. Since each of the 66 exons will be PCR amplified and screened by 'optimized' SSCP, only those exons containing more than 250bp are needed to be cleaved by restriction enzymes to obtain smaller segments. The only exon longer than 250bp in length is exon 12 which is composed of 372 nucleotides. The potential pitfalls and limitations associated with this method include: 1) Resolution is very important since differences between two confirmations can be very subtle and difficult to detect on polyacrylamide gels; 2) This approach misses splice mutations at the intron/exon borders (please see the section: The advantages of PTT); 3) Whether a specific mutation can be detected in a given condition is not predictable; 4) The major problem of SSCP is that the sensitivity depends on the sequence of context in an unpredictable way; 5) Missense mutations are often difficult to distinguish from rare polymorphism; 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6) The interpretation of the results of an SSCP gel requires some experience. Controls with known mutations can assist in the analysis. PCR-SSCP was invented in 1989 and can be applied to RNA or to reverse transcribed RNA (cDNA), either to increase sensitivity or to assay multiple exons simultaneously. Highly sensitive SSCP analysis by capillary electrophoresis (CE-SSCP) has been reported. The system is fully automated and skill-demanding manual gel casting and sample loading are no longer necessary. Most importantly, mobilities are normalized and digitized by the computer and the presence and absence of mutations are judged by statistical evaluations. Many of the advantages except the gel casting and sample loading are shared with SSCP using gel-based, automated, multicolor DNA sequencers. The disadvantage of this system is the considerable initial investment for the machine and limited number of samples analyzed per day per machine -60 to 96 samples. Protein Truncation Test (PTT) for ATM Gene Mutations When dealing with a large gene like ATM, mRNA-based mutation detection such as PTT is desirable to analyze larger fragments of the coding region. PTT is based on in vitro-coupled transcription and translation of PCR-amplified coding sequences and detects only the disease-causing mutations. PTT uses cDNA template and a forward primer containing a T7 promoter and an ATG translation initiation site. The PCR product is then translated in a rabbit reticulocyte system. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Because this method detects truncated proteins, every mutation is real, i.e. not a polymorphism nor an alternative splicing product. However, in order to detect all truncated proteins, several PTT screenings are required for each patient. Currently, the efficiency of PTT is about 70%. The advantages of PTT include: 1) Large coding segments (2-3 kb or even longer) can be screened in a single assay; 2) PTT detects mutations resulting in termination of the translation process. The detected mutations are real. Thus, sequence analysis is not strictly necessary to confirm the mutation; 3) Exon splicing mutations are easily detected. Mutations occurring deep inside introns, such as the English IVS40 + 1126 A-> G mutation (codon 1921 insl37nt) can only be detected using PTT rather than SSCP. The limitations of PTT include: 1) Point mutations are missed unless they result in protein truncation; 2) PTT is limited by the availability of RNA samples, and also by their susceptibility to degradation by nucleases; 3) In compound heterozygotes, a more stable RNA transcript might mask the presence of a second mutation in the same PTT segment; Together, PTT (using cDNA) and SSCP (using gDNA) probably detect more than 98% of ATM mutations. Since the majority of mutations found among T-PLL and T-ALL patients are missense mutations, SSCP is the first choice for this study. 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, it is preferred that mRNA obtained from fresh blood samples are processed and stored for further analysis by PTT, when additional study resources become available. Data Analysis For the main effect of ATM mutations on risk of childhood ALL, the McNemar odds ratio will be used: OR=U/V where U represents the number of pairs for which the case carries an ATM mutation but the control does not, V represents the number of pairs for which the control carriers an ATM mutation but the case does not. The standard error (SE) will be estimated as (1/U+1/V)m for matched-pair data. Therefore the 95% confidence interval will be: Exp[ln(0J?)±1.96(&D)] McNemar test will be used to test the null hypothesis H o'. O R -1: X 2 =(U-V)/(U+V)1 /2 For the interaction of radiation (R) and ATM mutations (A), the likelihood ratio test will be used to compare the following two models: Logit P 1 (X)=a+(31(R)+ f}2(A) (1) Logit P2(X)=a+p}(R)+ fa(A) + fa(RA ) (2) 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The null hypothesis being tested here is H o'. @3=0, and the likelihood ratio statistic is of the form: X 2M L R --2 1 n (X 1/L 2) where Li is the maximum likelihood value for model (1) and L2 is the maximum likelihood value for model (2). This likelihood ratio statistic has a chi-square distribution with one degree of freedom. For the effect of other environmental factors (xj, X 2, ... xn ), conditional logistic regression will be used to estimate the relative risk of the main effect on childhood ALL (X): Logit P(X)=a+ A+ yR+piXj+ j8 2X 2+ ■ ■ ■ -PrXn Each of the environmental factors will be tested for association with ALL. If a suggestive association exists (P<0.1), then this factor along with radiation exposure will be included in the model (being controlled) to assess the effect of ATM gene status on childhood ALL. Sample Size and Power Considerations For the matched case-control study, the number of discordant pairs (numbers of cases that are carriers of the ATM gene but not the controls, plus the number of controls who are carriers but not the ALL cases) required to detect a relative risk (R) is given by the formula (Schlesselman, 1982): m=[Za/2+ Zp VP(l-P)]2 / (P-0.5)2, where P=R/(1+R). 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Assuming cases and controls being carriers (“exposure”) are independent, the probability of an “exposure”-discordant pair is given by: Pe ~ (Poqi+Piqo), where pi,po are the estimated proportion of carriers in cases, and controls respectively. The quantities are defined by qo=l-po and q i= l-p \, respectively. Once the R and po are known, pi = R/>o /(I - /> o +R po ). Thus the total number of pairs M required on average to yield m discordant pairs is: M ~ ml (poqi+piqo)- After considering the sensitivity of ATM gene mutations by SSCP, the apparent relative risk (R*) will be reduced as a function of sensitivity (U),/>o, and the true relative risk R (Table 19). The following formula shows the R* (Flegal et al, 1986): R*=[URpo* ((1 -U)po+1 -po)]/[U*/?0 * ((1 -U) * R*i?o+1 -po)] • Assuming that the carrier rate of mutations in the ATM gene is 1% among controls, the selection of 115 case-control pairs will result in 80% power to detect a true relative risk of 7, an apparent relative risk of 6.9. Table 19 demonstrates that 70% sensitivity almost does not yield any difference between the apparent relative risk and the true relative risk, therefore, does not significantly increase the sample size. For example, 119 pairs are required to have 80% power to detect a relative risk of 6.9 instead of 115 pairs are required to detect a relative risk of 7. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 19. Sample size and power (1-P) used for design not matching on family history of cancer P R R* 0.05 0.1 0.15 0.2 2.0 2.0 3126 2492 2104 1820 3.0 3.0 971 782 666 581 4.0 4.0 507 412 354 311 5.0 4.9 325 266 230 203 6.0 5.9 232 192 167 148 7.0 6.9 178 148 129 115 8.0 7.8 142 119 105 94 9.0 8.8 118 99 88 79 10.0 9.7 100 85 75 68 13.0 12.5 68 58 52 48 15.0 14.4 56 48 43 40 20.0 18.9 38 34 30 28 (po=0.01, a=0.05, one sided, sensitivity-0.7) R: the true relative risk R*: the apparent relative risk, as a function of sensitivity, po, and R. If cases and controls are selected among those with family history among first degree relatives, po would be higher than the estimated 0.01 among general population. Here po is assumed to be 0.02 for the calculation. In this design, the selection of 125 case-control pairs will result in better than 85% power to detect a relative risk of 5 (Table 20). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 20. Sample size and power (1-P) used for design restricting on family history of cancer P R R* 0.05 0.1 0.15 0.2 2.0 2.0 1594 1271 1073 928 3.0 3.0 500 403 343 298 4.0 3.9 263 214 184 161 5.0 4.9 170 140 121 107 6.0 5.8 123 102 88 78 7.0 6.8 95 79 69 62 8.0 7.7 77 64 56 50 9.0 8.6 64 54 48 43 10.0 9.5 55 46 41 37 13.0 12.1 38 33 29 27 15.0 13.8 32 27 25 22 20.0 17.9 23 20 18 16 (po=0.02, a=0.05, one sided, sensitivity=0.7) R: the true relative risk R*: the apparent relative risk, as a function of sensitivity, p0 j and R. The main aim of the study is to examine the association of ATM gene (G) mutations and childhood T-ALL. There will be enough power to detect this effect if the relative risk is 7 or larger. This study will not have a large enough sample size to detect a small GxE interaction effect. The interaction is one of our interests, but this study could not be able to fully address this issue and obtain a definite answer. However, this study will be able to collect information such as radiation exposure among cases and controls and determine if the relative risk is consistent with 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. previous studies. If the data of this study provide clues of a GxE interaction, a larger scale study will be proposed to further examine the GxE effect. Timeline The accrual of newly diagnosed ALL patients since January 2001 will continue through the end of 2001. Interviews, data entry, blood sample collection should be completed for all subjects by January 2003. It is expected that 1000-2000 newly diagnosed ALL cases will visit CCG institutions each year. The matched cases and controls will be enrolled in the study during year 2001. The SSCP genotyping for the mutations in the ATM gene will be accomplished in year 2002 through 2004 with 80 subjects per year. Data analysis for at least one manuscript submission should be accomplished by the end of 2005. Alternative Design o f the Case-Control Study Stratified Sampling of Cases and Controls on Family History An alternative design to address the hypotheses of the study is a smaller scale study restricted to family history among cases and controls. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 21. Number o f cases required t o detect a gene-environment (G-E) interaction fo r different levels o f G a n d E exposures w ith 8 0 % power O I I 7 3 T3 o .2 h i T3 7 3 o O b .2 V ) -o 73 o CN I f 73 7 3 O CN I etj 1? O 1 1 O I I o 2 ? c o O I I Pi o O f O P i m O I ! Pi v o 1 1 0 7 P i <N O I I ei.c Vb vb © © O N C N Vb © Vb ^H r~- r- vo C N tJ- C N © i— i 0 0 C O C O I N - 0 0 tI- © C O V O oo vo f — 1 7t- V O O N 00 © C N O N © Os V O C O r- vb i— i 00 Vb oo vo IN C O Vb 1 — 1 © I I C O 1 2 V O Tf C O C N r - v o C O C N , “ H © co C N i — 1 cn | r- 2 ! ^C N o V ! 7 3 7 3 . . , O I O Pi o w — G C>. (N G O ( N o ^ rt o 00 o ^ m C N l- H GO i n •D 'O h (N r o Ch ^ v o r o t ' ' t m M n C N b- n vo O O v o o n o o T f fhON^-'ONO *-< c o c n ’ -h ^ i—i c o cn CN © v b o o 00 (N © © v© © © C O Os 00 o o c o * —< — * © O n - h 0 0 0 0 <N O O O 0 v o © 0 0 VO T I r b M N fH Tf- r i m VO -H - <o <n O v © vo n - t n n — in - m co o ■ ! } ■ o o vo oo m ’— I T f 0 4 i-H i— t i— I oo » -( > * t vo r~ - vo 0 0 CO <N t — 1-H t - * " S t © o n t> ^H - H © 7 f © ON f T | O l C v G T | VO M ^ to © ro t i P m G ^ 0 0 n ^ VO X E o G T ) t Tf —I -X vb ,— < o © co o o s ^ vo ov C n rj- oo vo T t m cn © *7t- v b © r - On © OS r - V O © © © vo 7 t co co r H M in o v n i— i Os 00 Ob CO t - C-- vb co Vb vo i— h oo r , r t M O N <N v b on o o o o t - ( N ’ t h - O V O VO I > O n O n T f CN rb 't N ^ Tt m in vo vb vo V O C N r f 7 f cN I> ro O n 00 C N ©\ O O O N - C N 7 t Tf O N Vb C N On 1 — 1 C O 00 V b 00 r- vo C O V O 7f © C N C N © 1 — H © C O T j* 1 — 1 i— i Tf V O O N 00 © vo C O r- vo O N < N O N V O O N vo t- c- C O V b 1-H © C N 1 — 1 t*- C O C N N- V O C O C N 1 — 1 1-H © co C N 1 -H 1-H C O V O © 7 t C N 00 00 rj* V O vb C O r~- 00 © © Vb C N Vb C O vb 00 C N C O 7f Vb 1 — 1 vO vo C N Vb C O 00 C N O O V b r - r - C N 00 00 © © vb C N C O O N 00 C N © vo 00 Vb C O C N C N C O Th C O C N 1 — 1 r - i 00 C N 1 — 1 T-H C N * — ro © co © t-~ vo o r-~ oo r- oo Vb n VO O VO O n T i N G ^ c b G (7b « -< vo oo vb r f vo vb vo C- O V O vo ro V O Vb © On 00 C O r-> O N Vb C O C N C N C N C O Vb 00 C N V O © © © © *-< 1 -c © © © © © © IN' Vb Vb 00 C N © 00 C O © C N Vb V O C N 00 © V O O N V O C N C N T t H . H H Ip C O 00 O N © O N 00 O N C N C N C O H Vb C N Vb Vb C O (N —1 I — • —( CO Vb 0 0 CN v o © © © © ^ © o © © © © © © CO C N O O V O * — C N *-( vo © C N Vb C N © !-< vb r - cn © N N- n n r - 0 0 V O r f C N © On Vb co Vb oo — co CN © co © i s O N CO C N ’ -H 00 - cb V) 00 C N vo © © © © T -H r-H © © © © © © © C N © 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Table 21 (continued). Number of cases required to detect a gene-environment (G-E) interaction for different levels of G and E exposures with 80% power Prevalence Odds ratios (q/=2) O dds ratios (tp=2) O dds ratios (t|t=2) O dds ratios (t|/=3) O dds ratios (vp=7) E G ORio ORpi O R n =2 =2 =8 ORio ORoi =2 =4 O R „ =16 ORjo ORoi =3 =4 O R ,, =24 OR,o ORo, =3 =4 O R ,, =36 O R ,0 ORo, =3 =4 O R ,, =84 0.30 0.01 7801 9942 6839 8724 6950 8867 2643 3374 780 992 0.03 2706 3449 2399 3060 2442 3115 937 1196 286 364 0.05 1692 2156 1518 1936 1548 1974 600 766 189 241 0.08 1127 1435 1031 1315 1054 1344 415 529 137 175 0.12 819 1043 772 983 791 1008 318 406 112 143 0.16 672 856 653 831 671 855 276 351 103 131 0.50 0.01 7700 9766 6574 8328 6905 8751 2747 3469 892 1111 0.03 2679 3399 2318 2938 2450 3107 985 1245 332 415 0.05 1680 2132 1475 1870 1570 1991 638 808 223 281 0.08 1124 1427 1011 1282 1086 1378 450 570 167 211 0.12 823 1045 767 973 835 1059 355 450 142 179 0.16 680 863 658 835 725 920 316 401 135 170 ORio denotes the odds ratio for nonsusceptible subjects exposed to the environmental factor (radiation); OR>, denotes the odds ratio for susceptible subjects not exposed to the environmental factor (radiation); ORn denotes the odds ratio for susceptible subjects exposed to the environmental factor (radiation); interaction parameter rp=ORn/ORio*ORoi *: the left column represents the number cases required to detect a one-sided test (a=0.05). # : the right column represents the number cases required to detect a two-sided test (a=0.05). o o o s Disadvantage: 1) It will take longer time and more interviews to find controls with a family history of cancer; 2) High cost of finding an appropriate control challenges the feasibility o f this approach; Advantages: 1) Although there is a limited literature addressing the frequency of ATM mutations among those with a family history of cancer, it is expected that the base rate of ATM mutations will be higher among both cases and controls with a family history of cancer, which would result in a reduced sample size for a given RR (see Table 19 and 20); 2) Fewer cases and controls will make it easier and at lower cost to validate the family history of cases and controls, and later even to validate the medical x-ray exposure; On the other hand, this study has several disadvantages and advantages compared to the above alternative design: The advantages of this alternative design compared to the proposed design are: 1) Not having to restrict to family history between cases and controls will reduce the time and cost to find a matched control for each case; 2) The results of the study could be applied to T-ALL patients; 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3) This study design will allow us to estimate the radiation exposure, frequency of ATM gene mutations, and proportion of controls with a family history of cancer. The extended knowledge will help to design the next study mentioned below. Moreover, we could propose a following study using parent controls to further test if the association between ATM and ALL found in the case-control study is real or due to an effect of "population stratification"; 4) If an association is found, we could propose a more general study of ALL, or other type of leukemia such as B-cell ALL, AML, or we could propose more restrictive studies such as studies of ALL with a family history of cancer, or T- cell ALL diagnosed under 15 years old, etc.; 5) The proposed study will be focused on one homogeneous group - T-ALL, which appears to have a 4 to 5-fold increased frequency relative to B-cell malignancy among A-T patients. A-T children predispose to develop T-ALL, the A-T adult patients have particular susceptibility to develop T cell prolymphocytic leukemia (T-PLL) (Taylor et al, 1996). Other types of childhood leukemia would not be considered in this study since in that case, the focus of the study will be scattered and the sample size would have to be increased to have enough power to find an increased relative risk in that particular leukemia type. The funding would have to be increased with the increased sample size if other type of leukemia is included, making the study even harder to get funded and carried out. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The disadvantages of the design include: 1) The frequency of ATM mutations would be lower in both cases and controls. To detect a same relative risk as in the design with a family history (FH) would require a larger sample size; 2) The increased number of case-control pairs would result in extra cost for ATM mutation detection among study subjects. However, the increased cost of ATM mutations would be less than the extra cost of finding a control with FH in the alternative design. In summary, the effect on the difference of frequency of ATM mutations between cases and controls, and its effect on the RR in the above two designs are not known. After weighing the convenience and lower cost for finding a control in unscreened design, versus more effort to find a control, and lower cost for detecting ATM mutations, the design using RDD is chosen for the proposed study. Hopefully, this study will contribute preliminary data for the future studies. Limitations and Potential Problems Recall Bias Recall bias might be expected to be less serious than for studies of adult cancers since the children patients and their matched controls are young at diagnosis, so activities and exposures need only be recalled over a shorter period of time. Also, children have fewer diagnostic x-rays than in adults, making it easier for parents to 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recall each x-ray exam and confirm each exposure dose from their physicians and dentists. Validation of Self-Reported X-Ray Exposure and Cancers among Family Members Independent confirmation of reported familial cases, their diagnosis, age onset would be desirable. The proposed study will confirm self-reported cancers among a 20% sample of randomly selected case and control families through their medical records, vital statistical registries, death certificate, surgical or pathological reports or other documents. In the proposed study, parents of the child proband will report cancer cases only on their first degree relatives (their children, siblings and parents). Previous studies indicated that there is a high accuracy of self-reported cancers among first-degree relatives ranging from 83% to 100% (Douglas et al, 1999, Bondy et al, 1994, Love et al, 1985). Validating of cancer cases among a 20% sample of families will provide an estimate of the magnitude and direction of misclassification, and the high cost, the time and effort for validating all reported cases would be reduced. The high sensitivity of self-reported cancers among family members in this study will provide the rationale for not validating FH in the larger scale study in the future. This validation data could be used in analysis of future studies in which FH is not validated by medical records. The issue of validating radiation exposure from medical X-rays seems more difficult and subject to much misclassification. Since exposure dose to medical 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sources of radiation is much less than it used to be, current data indicates that with the present level of exposure, medical radiation is not detectable as a risk factor for ALL (Dr. Buckley, personal communication). Although doses for specific exams are going down, newer high dose exams are being introduced and the frequency of use is increasing. There is some evidence showing that the overall population dose from X-rays is going up among adults. However, it is not clear whether this new trend applies to children (Dr. Thomas, personal communication). The other major challenge is accurately measuring the levels of x-ray exposure, since even if the type of exam, body locations receiving x-rays and the facility are given, the medical records may not contain the details on radiographic procedure and dose level. One of examples of validating x-ray exposure was provided by Dr. Preston-Martin’s adult AML study. Tremendous time and effort was made to collect all medical x-ray exposure for ten years before diagnosis and all these x-ray exposures were validated through medical records by contacting each of the treating physicians in LA county. The effect of the current comparatively low dose medical x-ray exposure, its effects on ATM mutations, and effect on the risk of the disease is not certain at this stage (Dr. Thomas, personal communication). Moreover, the difficulty of determining the dose of radiation exposure, even the type of radiation, and body location irradiated are known, and the high cost to collect medical records from a wide geographic spread of cases and controls and their treating clinics and hospitals in this study, made it not feasible to validate the 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. radiation exposure through medical records, at least at the initial stage of the study. Finally, the proportion of medical radiation is only about 11% of the total radiation exposure. The other main sources of radiation come from radon (55%), outer space (8%), rock and soil (8%) and from inside human body (11%). The total radiation exposure and their differences could not be reflected solely by medical radiation exposure. On the other hand, it makes excellent sense to consider that any effect of a mutated ATM could be due to radiation sensitivity. Therefore, the genetic component ATM gene status will be emphasized at the initial stage of the study. If a genetic effect is seen, the self-reported radiation exposure will be examined and the possibility of an even stronger association of the ATM gene with ALL in those who have radiation will be further examined. If so, the radiation exposure will be validated to confirm the findings. Sensitivity of SSCP and PTT Detection of mutations of the ATM gene is complicated by its large size and by the majority of mutations being missense rather than truncating. The limitations of mutation detection methods such as SSCP and PTT have been discussed in the previous sections. The efficiency of SSCP is about 60-70%. PTT alone is not appropriate to detect ATM gene mutations in this study since it is not sensitive to point mutations. PTT could detect a point mutation only when this mutation produces a stop codon and results in a truncated protein product. 192 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mutations in ATM are comparatively rare, so a large sample size is required. Since childhood ALL is less prevalent than most adult cancers, the cooperation of multiple children’s hospitals throughout the US will be needed to identify and register enough patients into the study. The Children’s Cancer Group (CCG) has established a nationwide network which includes more than 115 children’s hospitals in the U.S., and Canada. Each year, half of childhood cancer patients in the U.S. are treated in CCG institutions. Data quality control and large coverage of patients and resident area in this country is expected to yield a large enough sample size to achieve the desired statistical power. The proposed study is an important step toward testing the etiologic role of germline mutations in the recently cloned ATM gene in the development of childhood ALL. The findings of this study will provide important evidence if further larger scale studies are indicated to investigate whether the majority of ALL patients are predisposed to this disease because of carrying a germline mutation in ATM. Human Subjects Source o f Subjects Cases of newly diagnosed ALL since 2001 who meet the eligibility criteria. The parents or guardians of cases and controls should be able to understand the informed consent statement and be free to decline to participate. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sources o f Research Materials The following research materials will be obtained 1) exposure histories of eases and controls and a history of cancer among family members by interview; 2) venous blood samples of controls drawn at the time of interview; 3) tissues of bone marrow samples from the diagnosing hospitals for cases and venous blood samples or bone marrow samples of cases once remission has been achieved. Recruitment and Consent After each case is identified, a letter is sent to his physician asking for permission to contact the patient about the study. When the physician has given his consent, a letter is sent to the patient to describe the study. This is followed by a telephone call to further explain the study, to request participation, and to schedule the interview. At the time of interview, the respondent is given a copy of the informed consent to read, and the interviewer is available to answer any questions. Before the interview and the blood draw, the respondent is asked to sign the informed consent and given a copy of this form to keep. After an eligible control is identified by random digital dialing, the interviewer will explain the study and schedule an interview. The interview and blood drawn are conducted only after the control has read and signed the informed consent form. 194 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Potential Risks The psychological risks for a one hour-interview and venous blood drawn for cases and controls are minimal. For cases, the bone marrow samples at diagnosis and bone marrow or peripheral blood samples after complete remission has been achieved will be obtained at the institution of treatment as part of protocol to verify diagnosis and response. For controls, the potential physical risks from obtaining blood sample, including discomfort, hematoma, and fainting, are minimal and rare. The interviewers are trained and certified in correct techniques for obtaining blood samples. Confidentiality Information about ATM heterozygote and disease status of family members of childhood ALL cases will be saved in numerical codes. Identifiers linking to specific named individuals will be kept in a secured file. All analyses will use these codes and only aggregate data will be reported. Potential Benefits Although there will be no direct benefits to study subjects, the study will provide a better understanding of the cancer risk associated with being an ATM heterozygote. This information is critical for establishing meaningful testing and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. counseling guidelines and for clinical management of families and individuals who are ATM heterozygotes. Statement Regarding Inclusion o f Women and Minorities All cases and controls who are eligible will be included in the study, regardless of their gender or ethnic identity. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 5 - Conclusion The familial leukemia reports, twin studies of leukemia, studies examining the association of radiation and childhood leukemia conducted over the past four decades, and studies of ATM gene mutations among leukemia patients since the ATM gene was cloned in 1995 were reviewed. The early reports of familial aggregation of leukemia suggest a genetic etiology. The hypothesis is further supported by studies of Li-Fraumeni, ataxia telangiectasia, and other inherited syndromes. Family members of patients with these syndromes may carry germline mutations of genes such as p53, ATM, BRCA1, BRCA2, and appear to be predisposed to acquiring leukemia, breast cancer and other diseases. The findings of studies of leukemia among twins have been somewhat are less clear. On one hand, high concordance rates of leukemia, ranging from 5% to 25%, have been reported in the literature. However, recent evidence obtained using molecular techniques suggests that the increased concordance rate of leukemia may be attributable to leukemia cells being transferred from one twin to the other in utero. Furthermore, an overall deficit of leukemia among twins compared to the general population has been found. Although recent studies among twins do not provide direct supportive evidence that genetic factors play an etiologic role in the development of leukemia, they raise several questions to be investigated further. For example, a) if a twin could be affected by getting leukemia cells from the other twin, 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. why is the concordance rate not 100%? Are there genetic or environmental factors associated with whether the other twin accepts the leukemia cell transplantation? b) Most twins had leukemia onset during infants if the leukemia cells are transferred from one twin to the other. However, there is one report that both twins had ALL at 9 and 11 years respectively with the evidence suggesting that the leukemias originated from a single leukemia cell, developed in one twin, and transferred to the other (Ford et al, 1997). How and in what way do genetic and environmental factors influence the twin leukemia concordance in infants or in teenagers? There is a long way to go to establish which specific genetic factors are associated with leukemia, and to what extent these genetic factors interact with other genetic and environmental factors to thereby influence the risk of developing leukemia. The other interesting finding of this review of the literature was that multiple leukemia cases among offspring of consanguineous marriage were reported, suggesting an association between leukemia and consanguinity. This increased risk of leukemia could be Influenced by single recessive gene (Mendelian mechanism) inheritance, associated with increased risk of syndromic diseases such as neurofibromatosis, immunologic deficiencies (A-T, Wiscot-Aldrich syndrome, etc.), and/or abnormality of metabolizing enzymes, enzymatic repair mechanism (DNA repair enzymes) among consanguineous marriages, which may predispose family members to developing leukemia. 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. There is marked international variation of ALL and AML incidence. The difference between the incidence among the migrants living in the United States and those living in the nation in which they were bom could provide insight regarding the contribution of genetic and environmental factors in the etiology of childhood leukemia. It is not known that any of such studies has been carried out. Future directions of familial leukemia include: a) to establish a systematic childhood leukemia registry and family pedigree to perform segregation analysis; b) to perform well designed matched case-control studies to investigate both environmental and genetic factors; c) to collect blood samples of certain high risk groups such as infant leukemia, childhood leukemia with FH, or one histological group such as T or B-cell ALL to study the association of these more homogeneous leukemia subtypes with germline mutations of genes. In the literature, high level radiation exposure has been associated with increased risk of leukemia. But high dose radiation has become much less common nowadays. The association of low dose radiation and leukemia is inconsistent. The review of studies examining the association of radiation and childhood leukemia with positive findings have usually found 1.5 to 4.0-fold increased risk. However, these findings may reflect effects of chance, possible confounding or bias. Testing the association of low dose radiation and leukemia has become more challenging since low dose radiation studies have usually focused on medical diagnostic radiation - the radiation doses of which have been greatly reduced with 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the improvement of diagnostic equipment. Moreover, the proportion of medical radiation is only about 11% of the total radiation exposure. Future studies of the association of genes that predispose humans to be more sensitive to radiation exposure, and to find more accurate methods of measuring total low dose radiation could help to determine the risk of low dose radiation for developing leukemia. One such gene that has been found to be associated with radiation sensitivity and leukemia susceptibility is the ataxia telangiectasia mutated (ATM) gene. Leukemia is one of the most common malignancies among ataxia telangiectasia (AT) patients who are carriers of two ATM mutations (Morrell et al, 1986). Leukemia was also found more frequently than expected among relatives of A-T cases and among presumed carriers of one copy of the mutated gene (parents of the A-T patients) although this elevated risk was based on small numbers of leukemia cases (Swift et al, 1976, 1986, 1987; Morrell et al, 1990). The risk for childhood leukemia among siblings and cousins of A-T patients has not been estimated. AT patients and the presumed carriers of ATM gene mutations have been found to be more sensitive to radiation than those who do not carry the mutated gene (Paterson, et al, 1979; Gatti et al, 1985, 1988; Hecht et al, 1990). Nowadays, the dose of radiation is too low to result in large relative risk of leukemia, making studies examining the effect of mutations in the ATM gene mutation and its interaction with environmental factors more appealing and necessary to better understand the etiology of childhood ALL. 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results of a few initial studies of somatic ATM gene mutations among T- lineage PLL leukemia patients indicated that 32-63% carrying mutations. Missense mutations are believed to be dominant, in contrast to mostly truncating mutations found among A-T patients. One study of germline mutations among T-ALL patients found 5 of 19 (26%) patients had rare ATM polymorphisms. Almost all mutations identified so far have been unique regarding mutation types and locations on the ATM gene. Questions for future studies include: 1) whether somatic mutations among T-PLL patients could also be identified in the germline; 2) whether ALL patients have the same or different type and frequency of ATM gene mutations as T- PLL patients; 3) whether the germline mutations identified in a small sample of T- ALL patients in one previous study (Luo et al, 1998) could be found among the majority of childhood ALL patients; 4) whether nucleotide substitutions cause the loss of function of ATM gene or they are “neutral” polymorphisms; 5) whether the number of missense mutations observed is significantly different from the number expected; and, 6) whether there is a difference between missense mutations or truncating mutations among leukemia patients compared to those found among AT patients and the general population? The analyses of CCG-B903 study revealed increased risk among younger family members (those 45 years and under) for the following types of cancer: brain/CNS tumors, leukemia, cancers of primary site uncertain, female uterus, and cervix (SIRs = 2.4, 2.6, 2.5, 2.7, and 1.6, respectively). An increased risk of breast 201 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cancer was found among both younger and older ALL family members (SIRs - 2.4 and 1.6, respectively). Probands’ family members were 2.6 times as likely to develop leukemia than those people in the general population (SIR=2.6, 95% CI=1.4-4.0). A four-fold increased risk of developing leukemia was reported among younger relatives (SIR=4.1, 95% CI=2.2-6.5) when all racial/ethnic groups were combined. The excess of leukemia cases among younger family members was consistent across all three racial/ethnic groups with 3.8, 3.9, and 17.6-fold increased risks among White, Hispanic and African Americans relatives, respectively. Fewer cases of leukemia were found among older relatives compared to the overall population (SIR=0.4, 95% CI=0-1.6). Overall, leukemia cases occurred more often for all three racial/ethnic groups relative to the general population (SIR=2.3, 2.6 and 11.1, respectively). These findings of increased risk of cancer among relatives, especially among those younger than 45 years of age, suggest a genetic susceptibility contributing to the development of childhood ALL. Future research examining the etiologic role of specific genetic factors on childhood ALL is therefore indicated. The B-903 study is one of the largest similar type studies in the U.S., indicating there is an increased risk of leukemia and breast cancer among family members of childhood acute lymphocytic leukemia (ALL). The results of the study provide clues that there are possible genetic factors playing roles in childhood ALL and in the increased risk of cancers among family members of childhood ALL. An innovative 202 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. method is used in the analysis to adjust for the selection bias associated with the proband selection process. The analysis of this study was complicated by the nature of the familial cancer study, since family history of cancer was an eligibility criterion and it is also the end point to assess if there are excess cancers among family members. This new adjustment method needs to be further tested to see if it properly or over/under adjusts the excess cancers among family members. The probands’ and their family members’ disease diagnoses, ages, and relationships information were stored in the Genetic Analysis Package (GAP) software (Buckley, 1997). Family pedigrees were drawn using GAP. This method may be useful in future large pedigree collection and preparation for segregation analysis once data collection is completed. This method of data management will also help to follow- up the family members who are cancer free during the study, but may be at the risk of developing cancer in the future. The proposed case-control study will be the first attempt in the U.S. to examine the association of a germline mutation and childhood ALL. This study will overcome several limitations of earlier studies. First, the selection of both cases and controls will represent the source population. All T-cell ALL registered and diagnosed at one of CCG institutions during year 2001 will be ascertained. Since half of childhood cancers of U.S. are treated at CCG institutions each year, all T- ALL cases diagnosed in one-year period will be included in the study. Therefore, these T-ALL cases by complete ascertainment are representative of the target 203 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. population. Using RDD controls will come close to sampling randomly from the general population (Rothman et al, 1998). Second, this study is designed to have a large enough sample size and power to test a relative risk of 7 or larger of the ATM gene germline mutations in the development of childhood ALL. To my knowledge, a study such as the one proposed has not yet been done in the U.S. Third, single strand conformational polymorphism (SSCP) will be mainly used to detect probably dominant missense mutations among childhood ALL. If sufficient funding is available, PTT will also be used to identify missense mutations, which will increase the sensitivity. The type and frequency of ATM gene mutations detected among T- PLL patients will be further analyzed and compared among these childhood ALL patients. Finally, the analyses of unaffected individuals will cover the entire ATM gene (66 exons) instead of only confined to 1 or 2 exons for screening mutations. These results will contribute our knowledge of ATM gene mutations among controls, for which only estimated numbers are available now. The proposed study is an important step toward testing the etiologic role of germline mutations in the recently cloned ATM gene in the development of childhood ALL. The findings of this study will indicate whether future, larger scale studies are indicated to determine whether the majority of ALL patients are predisposed to this disease because of carrying a germline mutation in ATM, and whether low dose radiation interacts with germline mutation of ATM gene on the risk of developing childhood ALL. 204 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Agudo A, Gonzalez CA. Secondary matching: a method for selecting controls in case-control studies on environmental risk factors. Int JEpidemiol, 1999 Dec, 28(6):1130-3. Aitkin J, Bain C, Ward M, Siskind V, MacLennan R. How accurate is self-reported family history of colorectal cancer? Am JEpidem ol, 1995;141:863-71. Allen M; Kalantari M; Ylitalo N; Pettersson B; Hagmar B; Scheibenpflug L; Johansson B; Petterson U; Gyllensten U. HLA DQ-DR haplotype and susceptibility to cervical carcinoma: indications of increased risk for development of cervical carcinoma in individuals infected with HPV 18. Tissue Antigens, 1996 Jul, 48(1):32- 7. Anderson RC, Hermann HW. Leukemia in twin children. JAMA, 1955, 158:652-4. Anderson DE ; Badzioch MD. Familial effects of prostate and other cancers on lifetime breast cancer risk. Breast Cancer Res Treat, 1993 Nov, 28(2): 107-13. Andrieu N; Glavel F; Gairard B; Piana L; Br'emond A; Lansac J; Flamant R; Renaud R. Familial risk of breast cancer in a French case-control study. Cancer Detect Prey, 1994, 18(3):163-9. Apple RJ; Becker TM; Wheeler CM; Erlich HA. Comparison of human leukocyte antigen DR-DQ disease associations found with cervical dysplasia and invasive cervical carcinoma. Journal o f the National Cancer Institute, 1995 Mar 15, 87(6):427-36. Auvinen A ; Hakama M ; Arvela H ; Hakulinen T ; Rahola T ; Suomela M ; S'oderman B ; Ryfomaa T. Fallout from Chernobyl and incidence of childhood leukaemia in Finland, 1976-92 [see comments] BMJ, 1994 Jul, 16309(6948): 151-4. Azizi E; Friedman J; Pavlotsky F; Iscovich J; Bomstein A; Shafir R; Trau H; Brenner H; Nass D. Familial cutaneous malignant melanoma and tumors of the nervous system. A hereditary cancer syndrome. Cancer, 1995 Nov 1, 76(9): 1571-8. Badawi AF; Stem SJ; Lang NP; Kadlubar FF. Cytochrome P-450 and acetyltransferase expression as biomarkers of carcinogen-DNA adduct levels and human cancer susceptibility. Progress in Clinical and Biological Research, 1996, 395:109-40. 205 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BEIR V. Health effects of exposure to low levels of ionizing radiation. Committee on the Biological Effects of Ionizing Radiations, Board on Radiation Effects Research, Commission on Life Sciences, National Research Council. 1990. Berehuck A; Boyd J. Molecular basis of endometrial cancer. Cancer, 1995 Nov 15, 76(10 Suppl):2034-40. Bergman W; Watson P; de Jong J; Lynch HT; Fusaro RM. Systemic cancer and the FAMMM syndrome. Br J Cancer, 1990 Jun, 61(6):932-6. Bergmann MM. Byers T. Freedman DS. Mokdad A. Validity of self-reported diagnoses leading to hospitalization: a comparison of self-reports with hospital records in a prospective study of American adults. American Journal o f Epidemiology, \99% May, 15147(10):969-77. Bergmann MM. Calle EE. Mervis CA. Miracle-McMahill HL. Thun MJ. Heath CW. Validity of self-reported cancers in a prospective cohort study in comparison with data from state cancer registries. American Journal o f Epidemiology, 1998 Mar, 147(6):556-62. Birch JM. Epidemiology of Paediatric Cancer. Recent Results in Cancer Research. Vol 88. Springer-Verlag Berlin Heidelberg 1983. Bleyer WA. The impact of childhood cancer on the United States and the world. CA Cancer J Clin, 1990 Nov-Dee, 40(6):355-67. Bondy ML; Lustbader ED; Buffler PA; Schull WJ; Hardy RJ; Strong LC. Genetic epidemiology of childhood brain tumors. Genet Epidemiol, 1991, 8(4):253-67. Bondy ML; Strom SS; Colopy MW; Brown BW; Strong LC. Accuracy of family history of cancer obtained through interviews with relatives of patients with childhood sarcoma. Journal o f Clinical Epidemiology, 1994 Jan, 47(l):89-96. Bridges JM, Nelson MG. Familial leukemia. Acta Haematology, 1961, 26:246-51. Buckley JD; Buckley CM; Breslow NE; Draper GJ; Roberson PK; Mack TM. Concordance for childhood cancer in twins. Medical and Pediatric Oncology, 1996 Apr, 26(4):223-9. Burden AD; Newell J; Andrew N; Kavanagh G; Connor JM; MacKie RM. Genetic and environmental influences in the development of multiple primary melanoma [see comments]. Arch Dermatol, 1999 Mar, 135(3):261-5. 206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Byrne C ; Brinton LA ; Haile RW ; Schairer C. Heterogeneity of the effect of family history on breast cancer risk. Epidemiology, 1991 Jul, 2(4):276-84. Campbell M; Cabrera ME; Legues ME; Ridge S; Greaves M. Discordant clinical presentation and outcome in infant twins sharing a common clonal leukaemia. British Journal o f Haematology, 1996 Apr, 93(1): 166-9. Chen J ; Birkholtz GG ; Lindblom P ; Rubio C ; Lindblom A. The role of ataxia- telangiectasia heterozygotes in familial breast cancer. Cancer Res, 1998 Apr 1, 58(7): 1376-9. Ch'ene P; Ory K; R 'uedi D; Soussi T; Hegi ME. Functional analyses of a unique p53 germline mutant (Y236delta) associated with a familial brain tumor syndrome. In tJ Cancer, 1999 Jul 2, 82(l):17-22. Claus EB ; Schildkraut J ; Iversen ES Jr ; Berry D ; Parmigiani G. Effect of BRCA1 and BRCA2 on the association between breast cancer risk and family history. J Natl Cancer Inst, 1998 Dec 2, 90(23): 1824-9. Colditz GA ; Willett WC ; Hunter DJ ; Stampfer MJ ; Manson JE ; Hennekens CH ; Rosner BA. Family history, age, and risk of breast cancer. Prospective data from the Nurses' Health Study. JAMA, 1993 Jul 21, 270(3):338-43. Cooke JV. Acute leukemia in twins. JAMA, 1953, 152:1023-9. Cotton, Richard G. H. Mutation Detection. New York: Oxford University Press. 1997. 198 p. Danis, RP; Keith, LG. Some observations concerning leukemia in twins. Acta Geneticae Medicae et Gemellologiae, 1982, 31(3-4):173-7. Delongchamp RR; Mabuchi K; Yoshimoto Y; Preston DL. Cancer mortality among atomic bomb survivors exposed in utero or as young children, October 1950-May 1992. Radiation Research, 1997 Mar, 147(3):385-95. Dergham ST; Dugan MC; Arlauskas P; Du W; Vaitkevicius VK; Crissman JD; Sarkar FH. Relationship of family cancer history to the expression of p53, p21WAF- 1, HER-2/neu, and K-ras mutation in pancreatic adenocarcinoma. Int J Pancreatol, 1997 Jun, 21(3):225-34. Douglas, Fiona S; O'Dair, Lindsay C; Robinson, Marion; Evans, D Gareth R; Lynch, Sally A. The accuracy of diagnoses as reported in families with cancer: a retrospective study. Journal of Medical Genetics, 1999 Apr, 36(4): 309-312. 207 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Draper GJ; Sanders BM; Lennox EL; Brownbill PA. Patterns of childhood cancer among siblings. British Journal o f Cancer, 1996 Jul, 74(1): 152-8. Easton DF. Cancer risks in A-T heterozygotes. International Journal o f Radiation Biology, 1994 Dec, 66(6 Suppl):S177-82. Egan KM ; Stampfer MJ ; Rosner BA ; Trichopoulos D ; Newcomb P A ; Trentham- Dietz A ; Longnecker MP ; Mittendorf R ; Greenberg ER ; Willett WC. Risk factors for breast cancer in women with a breast cancer family history. Cancer Epidemiol Biomarkers Prev, 1998 May, 7(5):359-64. Farrington SM ; Lin-Goerke J ; Ling J ; Wang Y ; Burczak JD ; Robbins DJ ; Dunlop MG. Systematic analysis of hMSH2 and hMLHl in young colon cancer patients and controls. Am J Hum Genet, 1998 Sep, 63(3):749-59. Fernandez E ; La Vecchia C ; D'Avanzo B ; Negri E ; Franceschi S. Family history and the risk of liver, gallbladder, and pancreatic cancer. Cancer Epidemiol Biomarkers Prev, 1994 Apr-May, 3(3):209-12. FitzGerald MG; Harkin DP; Silva-Arrieta S; MacDonald DJ; Lucchina LC; Unsal H; O'Neill E; Koh J; Finkelstein DM; Isselbacher KJ; Sober AJ; Haber DA. Prevalence of germ-line mutations in pi 6, pl9ARF, and CDK4 in familial melanoma: analysis of a clinic-based population. Proc N atl A cad Sci USA, 1996 Aug 6, 93(16):8541-5. Flegal KM; Brownie C; Haas JD. The effects of exposure misclassification on estimates of relative risk. Am J Epidemiol, 1986 Apr, 123(4):736-51. Foppa I. Spiegelman D. Power and sample size calculations for case-control studies of gene-environment interactions with a polytomous exposure variable. American Journal o f Epidemiology, 1997 Oct 1, 146(7):596-604. Ford AM; Pombo-de-Oliveira MS; McCarthy KP; MacLean JM; Carrico KC; Vincent RF; Greaves M. Monoclonal origin of concordant T-cell malignancy in identical twins. Blood, 1997 Jan 1, 89(l):281-5. Froggatt NJ ; Brassett C ; Koch DJ ; Evans DG ; Hodgson SV ; Ponder BA ; Maher ER. Mutation screening of MSH2 and MLH1 mRNA in hereditary non-polyposis colon cancer syndrome. J M ed Genet, 1996 Sep, 33(9):726-30. Garcia-Closas M. Lubin JH. Power and sample size calculations in case-control studies of gene-environment interactions: comments on different approaches. American Journal o f Epidemiology, 1999 Apr 15, 149(8):689-92. 208 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Garcia-Closas M. Rothman N. Lubin J. Misclassification in case-control studies of gene-environment interactions: assessment of bias and sample size. Cancer Epidemiology, Biomarkers & Prevention, 1999 Dec, 8(12): 1043-50. Garcia-Pati-no E ; Gomendio B ; Lleonart M ; Silva JM ; Garcia JM ; Provencio M ; Cubedo R ; Espa~na P ; Ram'ony Cajal S ; Bonilla F. Loss of heterozygosity in the region including the BRCA1 gene on 17q in colon cancer. Cancer Genet Cytogenet, 1998 Jul 15, 104(2):119-23. Gauderman WJ, Witte JS, Thomas DC. Family-based association studies. Journal o f the National Cancer Institute. Monographs, 1999, (26):31 -7. Gardner, MJ; Hall, AJ; Snee, MP; Downes, S; Powell, CA; Terrell, JD. Methods and basic data of case-control study of leukaemia and lymphoma among young people near Sellafield nuclear plant in West Cumbria. Bmj (Clinical Research Ed.), 1990 Feb 17, 300(6722):429-34. Gardner, MJ; Snee, MP; Hall, AJ; Powell, CA; Downes, S; Terrell, JD. Results of case-control study of leukaemia and lymphoma among young people near Sellafield nuclear plant in West Cumbria. Bmj (Clinical Research Ed.), 1990 Feb 17, 300(6722):423-9. Gatti RA, Swift M (Eds). Ataxia-telangiectasia: genetics, neuropathology, and immunology of a degenerative disease of childhood: proceedings of a conference held in Solvang, California, January 16-20, 1984. 407 p. Gatti, RA; Berkel, I; Boder, E; Braedt, G; Charmley, P; Concannon, P; Ersoy, F; Foroud, T; Jaspers, NG; Lange, K; et al. Localization of an ataxia-telangiectasia gene to chromosome 1 lq22-23. Nature, 1988 Dec 8, 336(6199):577-80. Gatti RA: The genetic basis of human cancer. In: Genetic Basis of Human Cancer. B. Vogelstein and K.W. Kinzler (Eds.). McGram-Hill: New York. 1998. 731 p. Ghadirian P; Boyle P; Simard A; Baillargeon J; Maisonneuve P; Perret C. Reported family aggregation of pancreatic cancer within a population-based case-control study in the Francophone community in Montreal, Canada. Int J Pancreatol, 1991 Nov- Dee, 10(3-4): 183-96. Ghadirian P ; Cadotte M ; Lacroix A ; Baillargeon J ; Perret C. Colon cancer in seven siblings. Eur J Cancer, 1993, 29A(ll):1553-6. 209 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ghadirian P ; Lacroix A ; Perret C ; Maisonneuve P ; Boyle P. Sociodemographic characteristics, smoking, medical and family history, and breast cancer. Cancer Detect Prev, 1998, 22(6):485-94. Giardiello FM; Petersen GM; Brensinger JD; Luce MC; Cayouette MC; Bacon J; Booker SV; Hamilton SR. Hepatoblastoma and APC gene mutation in familial adenomatous polyposis. Gut, 1996 Dec, 39(96):867-9. Goldstein AM; Tucker MA. Genetic epidemiology of familial melanoma. Dermatologic Clinics, 1995 Jul, 13(3):605-12. Goldstein AM. Andrieu N. Detection of interaction involving identified genes; available study designs. Journal o f the National Cancer Institute. Monographs, 1999, (26):49-54. Greaves MF. Subtypes of acute lymphoblastic leukemia: Implications for the pathogenesis and epidemiology of leukemia. In: Pathogenesis of leukemias and lymphomas: Environmental influences. Magrath IT, O’Conor GT, Ramot B (Eds.). New York. Raven Press, 1984, ppl29-139. Greenberg ER. Random digit dialing for control selection. A review and a caution on its use in studies of childhood cancer. Am J Epidemiol, 1990 Jan, 131(1): 1-5. Grimmond SM; Palmer JM; Walters MK; Scott C; Nancarrow DJ; Teh BT; Elmes C; Pyke C; Khoo SK; Bennett I; et al. Confirmation of susceptibility locus on chromosome 13 in Australian breast cancer families. Human Genetics, 1996 Jul, 98(l):80-5. Gruber SB; Thompson WD. A population-based study of endometrial cancer and familial risk in younger women. Cancer and Steroid Hormone Study Group. Cancer Epidemiology, Biomarkers and Prevention, 1996 Jun, 5(6):411-7. Gunz, FW; Fitzgerald, PH; Crossen, PE; Mackenzie, IS; Powles, CP; Jensen, GR. Multiple cases of leukemia in a sibship. Blood, 1966 Apr, 27(4):482-9. Gunz, FW; Gunz, JP; Vincent, PC; Bergin, M; Johnson, FL; Bashir, H; Kirk, RL. Thirteen cases of leukemia in a family. Journal o f the National Cancer Institute, 1978 Jun, 60(6): 1243-50. Hachiya T ; Kuriaki Y ; Ueoka Y ; Nishida J ; Kato K ; Wake N. WAF1 genotype and endometrial cancer susceptibility. Gynecol Oncol, 1999 Feb, 72(2): 187-92. 210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hall NR; Bishop DT; Stephenson BM; Finan PJ. Hereditary susceptibility to colorectal cancer. Relatives of early onset cases are particularly at risk. Diseases o f the Colon and Rectum, 1996 Jul, 39(7):739-43. Hamden, DG. Inherited factors in leukaemia and lymphoma. Leukemia Research, 1985, 9(6):705-7. Harris CC. p53 tumor suppressor gene: at the crossroads of molecular carcinogenesis, molecular epidemiology, and cancer risk assessment. Environmental Health Perspectives, 1996 May, 104 Suppl 3:435-9. Heath CW. Five cases of acute leukemia in three generations. The New England Journal o f Medicine, 1964, 272:882-7. Hecht, F; Hecht, BK. Cancer in ataxia-telangiectasia patients. Cancer Genetics and Cytogenetics, 1990 May, 46(1):9-19. Hemminki K ; Dong C ; Vaittinen P. Familial risks in cervical cancer: is there a hereditary component? Int J Cancer, 1999 Sep 9, 82(6):775-81. Hemminki K ; Vaittinen P ; Kyyr'onen P. Modification of cancer risk in offspring by parental cancer (Sweden). Cancer Causes Control, 1999 Apr, 10(2): 125-9. Horwitz M; Goode EL; Jarvik GP. Anticipation in familial leukemia. American Journal o f Human Genetics, 1996 Nov, 59(5):990-8. Hopper JL. Chenevix-Trench G. Jolley DJ. Dite GS. Jenkins MA. Venter DJ. McCredie MR. Giles GG. Design and analysis issues in a population-based, case- control-family study of the genetic epidemiology of breast cancer and the Co operative Family Registry for Breast Cancer Studies (CFRBCS). Journal o f the National Cancer Institute Monographs, 1999, (26):95-100. Horwitz M. The genetics of familial leukemia. Leukemia, 1997 Aug, 11(8): 1347- 59. Hruban RH ; Petersen GM ; Goggins M ; Tersmette AC ; Offerhaus GJ ; Falatko F ; Yeo CJ ; Kem SE. Familial pancreatic cancer. Ann Oncol, 1999, 10 Suppl 4 (-HD-):69-73. Hulka BS. Epidemiologic analysis of breast and gynecologic cancers. Progress in Clinical and Biological Research, 1997, 396:17-29. 211 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hutter P; Couturier A; Scott RJ; Alday P; Delozier-Blanchet C; Cachat F; Antonarakis SE; Joris F; Gaudin M; D'Amato L; et al. Complex genetic predisposition to cancer in an extended HNPCC family with an ancestral hMLHl mutation. Journal o f M edical Genetics, 1996 Aug, 33(8):636-40. Hwang SJ. Beaty TH. Liang KY. Coresh J. Khoury MJ. Minimum sample size estimation to detect gene-environment interaction in case-control designs. American Journal o f Epidemiology, 1994 Dec 1, 140(11): 1029-37. Infante-Rivard C, Mathonnet G, Sinnett D. Risk of Childhood Leukemia Associated with Diagnostic Irradiation and Polymorphisms in DNA Repair Genes. Environ Health Perspect, 2000 Jun, 108(6):495-498. Inskip, PD; Harvey, EB; Boice, JD Jr; Stone, BJ; Matanoski, G; Flannery, JT; Fraumeni, JF Jr. Incidence of childhood cancer in twins. Cancer Causes and Control, 1991 Sep, 2(5):315-24. Izraeli S; Henn T; Strobl H; Ludwig WD; Kovar H; Haas OA; Harbott J; Bartram CR; Gadner H; Lion T. Expression of identical E2A/PBX1 fusion transcripts occurs in both pre-B and early pre-B immunological subtypes of childhood acute lymphoblastic leukemia. Leukemia, 1993, 7(12):2054-6. Jishi MF; Itnyre JH; Oakley-Girvan IA; Piver MS; Whittemore AS. Risks of cancer among members of families in the Gilda Radner Familial Ovarian Cancer Registry. Cancer, 1995 Oct 15, 76(8): 1416-21. Johannsson O; Ostermeyer EA; Hakansson S; Friedman LS; Johansson U; Sellberg G; Brondum-Nielsen K; Sele V; Olsson H; King MC; et al. Founding BRCA1 mutations in hereditary breast and ovarian cancer in southern Sweden. American Journal o f Human Genetics, 1996 Mar, 58(3):441-50. Kato N; Nakazawa T; Mizutani T; Shimotohno K. Susceptibility of human T- lymphotropic virus type I infected cell line MT-2 to hepatitis C virus infection. Biochemical and Biophysical Research Communications, 1995 Jan 26,206(3):863-9. Kende, G; Toren, A; Mandel, M; Neumann, Y; Kenet, G; Brok-Simoni, F; Ramot, B; Ben-Bassat, I; Rechavi, G. Familial leukemia: description of two kindreds and a review of the genetic aspects of the disease. Acta Haematologica, 1994, 92(4):208- 11. Kerber RA, Slattery ML. Comparison of self-reported and data-base-linked family history of cancer data in a case-control study. Am J Epidemiol, 1997;146:244-8. 212 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kerber RA ; Slattery ML ; Potter JD ; Caan BJ ; Edwards SL. Risk of colon cancer associated with a family history of cancer or colorectal polyps: the diet, activity, and reproduction in colon cancer study. Int J Cancer, 1998 Oct 5, 78(2): 157-60. Kihara M; Kihara M; Kubota A; Furukawa M; Kimura H. GSTM1 gene polymorphism as a possible marker for susceptibility to head and neck cancers among Japanese smokers. Cancer Letters, 1997 Jan 30, 112(2):257-62. Kinlen, LJ. Can paternal preconceptional radiation account for the increase of leukaemia and non-Hodgkin's lymphoma in Seascale? Bmj (Clinical Research Ed.), 1993 Jun 26, 306(6894): 1718-21. Knox EG; Marshall T; Barling R. Leukaemia and childhood cancer in twins. Journal o f Eidemiology and Community Health, 1984 Mar, 38(1): 12-6. Kuijten RR; Bunin GR; Nass CC; Meadows AT. TI- Gestational and familial risk factors for childhood astrocytoma: results of a case-control study. Cancer Res, 1990 May 1, 50(9):2608-12. Kurita S; Kamei Y; Ota K. Genetic studies on familial leukemia. Cancer, 1974 Oct, 34(4): 1098-101. La Vecchia C ; Gallus S ; Talamini R ; Decarli A ; Negri E ; Franceschi S. Interaction between selected environmental factors and familial propensity for colon cancer. Eur J Cancer Prev, 1999 Apr, 8(2):147-50. Lancaster JM ; Carney ME ; Gray J ; Myring J ; Gumbs C ; Sampson J ; Wheeler D ; France E ; Wiseman R ; Harper P ; Futreal PA. BRCA1 and BRCA2 in breast cancer families from Wales: moderate mutation frequency and two recurrent mutations in BRCA1. Br J Cancer, 1998 Dec, 78(ll):1417-20. Levi F; Randimbison L; Te VC; La Vecchia C. Non-Hodgkin's lymphomas, chronic lymphocytic leukaemias and skin cancers. British Journal o f Cancer, 1996 Dec, 74(11): 1847-50. Li FP; Tucker MA; Fraumeni JF Jr. Childhood cancer in sibs. Journal o f Pediatrics, 1976 Mar, 88(3):419-23. Linet MS ; Hatch EE ; Kleinerman RA ; Robison LL ; Kaune WT ; Friedman DR ; Severson RK ; Haines CM ; HartsOek CT ; Niwa S ; Wacholder S ; Tarone RE. Residential exposure to magnetic fields and acute lymphoblastic leukemia in children. N Engl J Med, 1997 Jul 3, 337(1 ):l-7 213 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Longnecker MP; Enger SM. Epidemiologic data on alcoholic beverage consumption and risk of cancer. Clinica Chimica Acta, 1996 Mar 15, 246(1 -2): 121 -41. Love RR, Evans AM, Josten DM. The accuracy of patient reports of a family history of cancer. J Chron Dis, 1985, 38:289-93. Lubbe J; von Ammon K; Watanabe K; Hegi ME; Kleihues P. Familial brain tumour syndrome associated with a p53 germline deletion of codon 236. Brain Pathology, 1995 Jan, 5(l):15-23. Lubin JH, Gail MH. On power and sample size for studying features of the relative odds of disease. Am J Epidemiol., 1990 Mar, 31(3):552-66. Luce MC ; Marra G ; Chauhan DP ; Laghi L ; Carethers JM ; Cherian SP ; Hawn M ; Binnie CG ; Kam-Morgan LN ; Cayouette MC ; et al. In vitro transcription/translation assay for the screening of hMLHl and hMSH2 mutations in familial colon cancer. Gastroenterology, 1995 Oct, 109(4): 1368-74. Luo L; Lu FM; Hart S; Foroni L; Rabbani H; Hammarstrom L; Yuille MR; Catovsky D; Webster AD; Vorechovsky I. Ataxia-telangiectasia and T-cell leukemias: no evidence for somatic ATM mutation in sporadic T-ALL or for hypermethylation of the ATM-NPAT/E14 bidirectional promoter in T-PLL. Cancer Research, 1998 Jun 1, 58(11):2293-7. Lynch HT; Fusaro RM. Pancreatic cancer and the familial atypical multiple mole melanoma (FAMMM) syndrome. Pancreas, 1991 Mar, 6(2): 127-31. Lynch HT; Lynch J. Genetics, natural history, surveillance, management, and gene mapping in the Lynch syndrome. Pathologie Biologie, 1995a Mar, 43(3):151-8. Lynch HT; Lynch JF. Clinical implications of advances in the molecular genetics of colorectal cancer. Tumori, 1995b May-Jun, 81(3 Suppl): 19-29. Magnusson C ; Colditz G ; Rosner B ; Bergstrom R ; Persson I. Association of family history and other risk factors with breast cancer risk (Sweden). Cancer Causes Control, 1998 May, 9(3):259-67. Maguire GP. The psychological sequelae of childhood leukaemia. Recent Results Cancer Res, 1983, 88:47-56. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mahmoud HH; Ridge SA; Behm FG; Pui CH; Ford AM; Raimondi SC; Greaves MF. Intrauterine monoclonal origin of neonatal concordant acute lymphoblastic leukemia in monozygotic twins. M edical and Pediatric Oncology, 1995 Feb, 24(2):77-81. Maimer B; Gr'onbergH; Bergenheim AT; Lenner P; HenrikssonR. Familial aggregation of astrocytoma in northern Sweden: an epidemiological cohort study. In tJ Cancer, 1999 May 5, 81(3):366-70. Maloney et al, unpublished data, 1997. McCredie M ; Paul C ; Skegg DC ; Williams S. Family history and risk of breast cancer in New Zealand. Int J Cancer, 1997 Nov 14, 73(4):503-7. McGlynn KA; Rosvold EA; Lustbader ED; Hu Y; Clapper ML; Zhou T; Wild CP; Xia XL; Baffoe-Bonnie A; Ofori-Adjei D; et al. Susceptibility to hepatocellular carcinoma is associated with genetic variation in the enzymatic detoxification of aflatoxin Bl. McPhedran, P; Heath, CW Jr; Lee, J. Patterns of familial leukemia. Ten cases of leukemia in two interrelated families. Cancer, 1969 Aug, 24(2):403-7. Millar AL ; Pal T ; Madlensky L ; Sherman C ; Temple L ; Mitri A ; Cheng H ; Marcus V ; Gallinger S ; Redston M ; Bapat B ; Narod S. Mismatch repair gene defects contribute to the genetic basis of double primary cancers of the colorectum and endometrium. Hum M ol Genet, 1999 May, 8(5):823-9. Morrell, D; Cromartie, E; Swift, M. Mortality and cancer incidence in 263 patients with ataxia-telangiectasia. Journal o f the National Cancer Institute, 1986 Jul, 77(l):89-92. Morrell, D; Chase, CL; Swift, M. Cancers in 44 families with ataxia-telangiectasia. Cancer Genetics and Cytogenetics, 1990 Nov 1, 50(1): 119-23. Morton CT. Is research in normal and ill children involving radiation exposure ethical? Arch Gen Psychiatry, 1996 Nov, 53(11): 1059-61. Munger K. The molecular biology of cervical cancer. Journal o f Cellular Biochemistry. Supplement, 1995, 23:55-60. Neglia, JP; Robison, LL. Epidemiology of the childhood acute leukemias. Pediatric Clinics o f North America, 1988 Aug, 35(4):675-92. 215 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Narod SA ; Stiller C ; Lenoir GM. An estimate of the heritable fraction of childhood cancer. Br J Cancer, 1991, 63(6):993-9. Narod SA. Genetic epidemiology of childhood cancer. Biochim Biophys Acta, 1996, 1288(3):F141-50. Nelson RL; Davis FG; Persky V; Becker E. Risk of neoplastic and other diseases among people with heterozygosity for hereditary hemochromatosis. Cancer, 1995 Sep 1, 76(5):875-9. Nystr'om-Lahti M ; Kristo P ; Nicolaides NC ; Chang SY ; Aaltonen LA ; Moisio AL ; JTarvinen HJ ; Mecklin IP ; Kinzler KW ; Vogelstein B ; et al. Founding mutations and Alu-mediated recombination in hereditary colon cancer. Nat Med, 1995 Nov, 1(11): 1203-6. Newton Bishop JA; Harland M; Bennett DC; Bataille V; Goldstein AM; Tucker MA; Ponder BA; Cuzick J; Selby P; Bishop DT. Mutation testing in melanoma families: INK4A, CDK4 and INK4D. B rJC ancer, 1999 Apr, 80(l-2):295-300. Odunsi K; Terry G; Ho L; Bell J; Cuzick J; Ganesan TS. Association between HLA DQB1 * 03 and cervical intra-epithelial neoplasia. Molecular Medicine, 1995 Jan, 1(2):161-71. Olsson R. Hepatitis and cancer: genetic aspects. Scandinavian Journal o f Gastroenterology. Supplement, 1996, 220:115-20. Paterson RD, Kelly WD, Good RA. Ataxia-telangiectasia-its association with a defective thymus, immunological deficiency disease and malignancy. Lancet, 1964 May, 1189-93. Paterson, MC; Anderson, AK; Smith, BP; Smith, PJ. Enhanced radiosensitivity of cultured fibroblasts from ataxia telangiectasia heterozygotes manifested by defective colony-forming ability and reduced DNA repair replication after hypoxic gamma- irradiation. Cancer Research, 1979 Sep, 39(9):3725-34. Pearson HA, Cone CTE., Jr. Leukemia in identical twins. The New England Journal o f Medicine, 1963, 268:1151-6. Pendergrass TW. Epidemiology of acute lymphoblastic leukemia. Seminar in Oncology, 1985 Jun, 12(2):80-91. 216 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Peters JM ; Preston-Martin S ; London SJ ; Bowman JD ; Buckley JD ; Thomas DC. Processed meats and risk of childhood leukemia. Cancer Causes Control, 1994 Mar, 5(2): 195-202. Peto J ; Collins N ; Barfoot R ; Seal S ; Warren W ; Rahman N ; Easton DF ; Evans C ; Deacon J ; Stratton MR. Prevalence of BRCA1 and BRCA2 gene mutations in patients with early-onset breast cancer [see comments]. J Natl Cancer Inst, 1999 Jun 2, 91(11):943-9. Petridou E; Trichopoulos D; Dessypris N; Flytzani V; Haidas S; Kalmanti M; Koliouskas D; Kosmidis H; Piperopoulou F; Tzortzatou F. Infant leukaemia after in utero exposure to radiation from Chernobyl. Nature, 1996 Jul 25, 382(6589):352-3. Pharoah PD ; DayNE ; Duffy S ; Easton DF ; Ponder BA. Family history and the risk of breast cancer: a systematic review and meta-analysis. Int J Cancer, 1997 May 29, 71(5):800-9. Pierce DA ; Preston DL. Radiation-related cancer risks at low doses among atomic bomb survivors. Radiat Res, 2000 Aug, 154(2): 178-86. Platz A; Hansson J; Mansson-Brahme E; Lagerlof B; Linder S; Lundqvist E; Sevigny P; Ingan'as M; Ringborg U. Screening of germline mutations in the CDKN2A and CDKN2B genes in Swedish families with hereditary cutaneous melanoma. JN a tl Cancer Inst, 1997 May 21, 89(10):697-702. Ponz de Leon, M. Familial and hereditary tumors. New York: Springer-Verlag. 1994. 347 p. Pottem LM; Linet M; Blair A; Dick F; Burmeister LF; Gibson R; Schuman LM; Fraumeni JF Jr. Familial cancers associated with subtypes of leukemia and non- Hodgkin's lymphoma. Leukemia Research, 1991, 15(5):305-14. Pui CH. Behm FG. Downing JR. Hancock ML. Shurtleff SA. Ribeiro RC. Head DR. Mahmoud HH. Sandlund JT. Furman WL. et al. 1 Iq23/MLL rearrangement confers a poor prognosis in infants with acute lymphoblastic leukemia. Journal o f Clinical Oncology, 1994 May, 12(5):909-15. Quesnel S; Malkin D. Genetic predisposition to cancer and familial cancer syndromes. Pediatric Clinics o f North America, 1997 Aug, 44(4):791-808. Rebbeck TR ; Couch FJ ; Kant J ; Calzone K ; DeShano M ; Peng Y ; Chen K ; Garber JE ; Weber BL. Genetic heterogeneity in hereditary breast cancer: role of BRCA1 and BRCA2. Am JHum Genet, 1996 Sep, 59(3):547-53. 217 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Riboli E; Kaaks R; Esteve J. Nutrition and laryngeal cancer. Cancer Causes and Control, 1996 Jan, 7(l):147-56. Rigby PG; Pratt PT; Rosenlof RC; Lemon HM. Genetic relationships in familial leukemia and lymphoma. Archives o f Internal Medicine, 1968 Jan 121(l):67-70. Risinger JI; Barrett JC; Watson P; Lynch HT; Boyd J. Molecular genetic evidence of the occurrence of breast cancer as an integral tumor in patients with the hereditary nonpolyposis colorectal carcinoma syndrome. Cancer, 1996 May 1, 77(9): 1836-43. Risinger JI; Hayes AK; Berchuck A; Barrett JC. PTEN/MMAC1 mutations in endometrial cancers. Cancer Research, 1997 Nov 1, 57(21):4736-8. Robison LL. Daigle A. Control selection using random digit dialing for cases of childhood cancer. American Journal o f Epidemiology, 1984 Jul, 120(1): 164-6. Robison LL; Buckley JD; Bunin G. Assessment of environmental and genetic factors in the etiology of childhood cancers: the Childrens Cancer Group epidemiology program. Environmental Health Perspectives, 1995 Sep, 103 Suppl 6 :111-6. Rodvall, Y; Hrubec, Z; Pershagen, G; Ahlbom, A; Bjurman, A; Boice, JD Jr. Childhood cancer among Swedish twins. Cancer Causes and Control, 1992 Nov, 3(6):527-32. Ross JA. Epidemiologic studies of childhood leukemia: where do we go from here? Medical and Pediatric Oncology, 1999 Jan, 32(l):65-7. Rothman KJ; Greenland S. Modem Epidemiology. 2n d Ed. Lippincott -Raven Publishers. 1998. 737p. Sastre-Garau X; Loste MN; Vincent-Salomon A; Favre M; Mouret E; de la Rochefordiere A; Durand JC; Tartour E; Lepage V; Charron D. Decreased frequency of HLA-DRB1 13 alleles in Frenchwomen with HPV-positive carcinoma of the cervix. International Journal o f Cancer, 1996 Jun 21, 69(3): 159-64. Savitsky K; Bar-Shira A; Gilad S; Rotman G; Ziv Y; Vanagaite L; Tagle DA; Smith S; Uziel T; Sfez S; et al. A single ataxia telangiectasia gene with a product similar to PI-3 kinase. Science, 1995 Jun 23, 268(5218):1749-53. 218 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Savitsky K; Sfez S; Tagle DA; Ziv Y; Sartiel A; Collins FS; Shiloh Y; Rotman G. The complete sequence of the coding region of the ATM gene reveals similarity to cell cycle regulators in different species. Human Molecular Genetics, 1995 Nov, 4(ll):2025-32. Schlesselman JJ. Case-control studies: Design, conduct, analysis. New York. Oxford University Press. 1982. Schubert EL ; Mefford HC ; Dann JL ; Argonza RH ; Hull J ; King MC. BRCA1 and BRCA2 mutations in Ashkenazi Jewish families with breast and ovarian cancer. Genet Test, 1997, l(l):41-6. Shpilberg, O; Modan, M; Modan, B; Chetrit, A; Fuchs, Z; Ramot, B. Familial aggregation of haematological neoplasms: a controlled study. British Journal o f Haematology, 1994 May, 87(l):75-80. Shu, XO; Gao, YT; Brinton, LA; Linet, MS; Tu, JT; Zheng, W; Fraumeni, JF Jr. A population-based case-control study of childhood leukemia in Shanghai. Cancer, 1988 Aug 1, 62(3):635-44. Shu, XO; Jin, F; Linet, MS; Zheng, W; Clemens, J; Mills, J; Gao, YT. Diagnostic X-ray and ultrasound exposure and risk of childhood cancer. British Journal o f Cancer, 1994 Sep, 70(3):531-6. Shu, XO; Reaman, GH; Lampkin, B; Sather, HN; Pendergrass, TW; Robison, LL. Association of paternal diagnostic X-ray exposure with risk of infant leukemia. Investigators of the Childrens Cancer Group. Cancer Epidemiology, Biomarkers and Prevention, 1994 Dec, 3(8):645-53. Silverman DT ; Schiffman M ; Everhart J ; Goldstein A ; Lillemoe KD ; Swanson GM ; Schwartz AG ; Brown LM ; Greenberg RS ; Schoenberg JB ; Pottem LM ; Hoover RN ; Fraumeni JF Jr. Diabetes mellitus, other medical conditions and familial history of cancer as risk factors for pancreatic cancer. Br J Cancer, 1999 Aug, 80(11): 1830-7. Slattery ML ; Kerber RA. A comprehensive evaluation of family history and breast cancer risk. The Utah Population Database [see comments]. JAMA, 1993 Oct 6, 270(13):1563-8. Slattery ML ; Mori M ; Gao R ; Kerber RA. Impact of family history of colon cancer on development of multiple primaries after diagnosis of colon cancer. Dis Colon Rectum, 1995 Oct, 38(10): 1053-8. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Stewart, AM. Leukemia and other neoplasms in childhood following radiation exposure in utero— a general survey of present knowledge. British Journal o f Radiology, 1968 Sep, 41(489):718-9. Stilgenbauer S; Liebisch P; James MR; Schroder M; Schlegelberger B; Fischer K; Bentz M; Lichter P; Dohner H. Molecular cytogenetic delineation of a novel critical genomic region in chromosome bands llq22.3-923.1 in lymphoproliferative disorders. Proceedings of the National Academy of Sciences of the United States of America, 1996 Oct 15, 93(21):11837-41. Stoppa-Lyonnet D; Soulier J; Lauge A; Dastot H; Gar and R; Sigaux F; Stem MH. Inactivation of the ATM gene in T-cell prolymphocytic leukemias. Blood, 1998 May 15, 91(10):3920-6. Gill Super HJ; Rothberg PG; Kobayashi H; Freeman Al; Diaz MO; Rowley JD. Clonal, nonconstitutional rearrangements of the MLL gene in infant twins with acute lymphoblastic leukemia: in utero chromosome rearrangement of 1 lq23. Blood, 1994 Feb 1, 83(3):641-4. Sakkinen PA, Severson RK, Ross JA, Robison LL. Random-digit dialing for control selection in childhood cancer studies: the geographic proximity and demographics within matched sets. Am J Public Health, 1995 Apr, 85(4):555-7. Siegmund KD. Gauderman WJ. Thomas DC. Association tests using unaffected- sibling versus pseudo-sibling controls. Genetic Epidemiology, 1999, 17 Suppl LS731-6. Stevens W ; Thomas DC ; Lyon JL ; Till JE ; Kerber RA ; Simon SL ; Lloyd RD ; Elghany NA ; Preston-Martin S. Leukemia in Utah and radioactive fallout from the Nevada test site. A case-control study. JAMA, 1990 Aug 1, 264(5):585-91. Sumoi R; Hakala-Ala-Pietila T; Lemihen A; Mecklin JP; Lehtovirta P. Hereditary aspects of endometrial adenocarcinoma. International Journal o f Cancer, 1995 Jul 17, 62(2): 132-7. Swerdlow AJ; De Stavola B; Maconochie N; Siskind V. A population-based study of cancer risk in twins: relationships to birth order and sexes of the twin pair. International Journal o f Cancer, 1996 Aug 7, 67(4):472-8. Swift M; Sholman L; Perry M; Chase C. Malignant neoplasms in the families of patients with ataxia-telangiectasia. Cancer Research, 1976 Jan, 36(1):209-15. 220 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Swift, M; Morrell, D; Cromartie, E; Chamberlin, AR; Skolnick, MH; Bishop, DT. The incidence and gene frequency of ataxia-telangiectasia in the United States. American Journal o f Human Genetics, 1986 Nov, 39(5):573-83. Swift, M; Reitnauer, PJ; Morrell, D; Chase, CL. Breast and other cancers in families with ataxia-telangiectasia. New England Journal o f Medicine, 1987 May 21, 316(21): 1289-94. Swift, M; Chase, CL; Morrell, D. Cancer predisposition of ataxia-telangiectasia heterozygotes. Cancer Genetics and Cytogenetics, 1990 May, 46(l):21-7. Tannerg°ard P ; Lipford JR ; Kolodner R ; Fr'odin JE ; Nordenskj'old M ; Lindblom A. Mutation screening in the hMLHl gene in Swedish hereditary nonpolyposis colon cancer families. Cancer Res, 1995 Dec 15, 55(24):6092-6. Taylor, AM; Hamden, DG; Arlett, CF; Ilarcourt, SA; Lehmann, AR; Stevens, S; Bridges, BA. Ataxia telangiectasia: a human mutation with abnormal radiation sensitivity. Nature, 1975 Dec 4, 258(5534):427-9. Taylor AM; Metcalfe JA; Thick J; Mak YF. Leukemia and lymphoma in ataxia telangiectasia. Blood, 1996 Jan 15, 87(2):423-38. Theis B, Boyd N, Lockwood G, Tritchier D. Accuracy of family cancer history in breast cancer patients. Eur J Cancer Prev, 1994, 3:321 -7. Thomas DC ; Preston-Martin S. Risk of leukemia caused by diagnostic x-rays. Health Phys, 1992 Nov, 63(5):576-8. Thomas DC. Design of gene characterization studies: an overview. Journal o f the National Cancer Institute. Monographs, 1999, (26): 17-23. Tulinius H ; Olafsdottir GH ; Sigvaldason H ; Tryggvadottir L ; Bjamadottir K. Neoplastic diseases in families of breast cancer patients. J M ed Genet, 1994 Aug, 31(8):618-21. Vital A ; Bringuier PP ; Huang H ; San Galli F ; Rivel J ; Ansoborlo S ; Cazauran JM ; Taillandier L ; Kleihues P ; Ohgaki H. Astrocytomas and choroid plexus tumors in two families with identical p53 germline mutations. JNeuropathol Exp Neurol, 1998 Nov, 57(11): 1061-9. von Knebel Doeberitz M. The role of papillomaviruses in the etiology of cervical cancer. Geburtshilfe Frauenheilkd, 1990 Jul, 50(7):511-7. 221 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Vorechovsky I; Luo L; Dyer MJ; Catovsky D; Amlot PL; Yaxley JC; Foroni L; Hammarstrom L; Webster AD; Yuille MA. Clustering of missense mutations in the ataxia-telangiectasia gene in a sporadic T-cell leukaemia. Nature Genetics, 1997 Sep, 17(l):96-9. Ward EM, Kramer S, Meadows AT. The efficacy of random digit dialing in selecting matched controls for a case-control study of pediatric cancer. Am J Epidemiol, 1984 Oct, 120(4):582-91. Watson P; Vasen HF; Mecklin JP; J”arvinen H; Lynch HT. The risk of endometrial cancer in hereditary nonpolyposis colorectal cancer. Am J Med, 1994, 96(6):516-20. Weber W, Gencik A, Muller H. Familial cancer: consequences for the oncological practice. Cancer Detect Prev, 1986;9:455-8. Witte JS. Gauderman WJ. Thomas DC. Asymptotic bias and efficiency in case- control studies of candidate genes and gene-environment interactions: basic family designs. American Journal o f Epidemiology, 1999 Apr 15, 149(8):693-705. Yoshimoto, Y. Cancer risk among children of atomic bomb survivors. A review of RERF epidemiologic studies. Radiation Effects Research Foundation JAMA, 1990 Aug 1, 264(5):596-600. Yuille MA; Coignet LJ; Abraham SM; Yaqub F; Luo L; Matutes E; Brito-Babapulle V; Vorechovsky I; Dyer MJ; Catovsky D. ATM is usually rearranged in T-cell prolymphocytic leukaemia. Oncogene, 1998 Feb 12, 16(6):789-96. zur Hausen H. Papillomavirus infections— a major cause of human cancers. Biochimica et Biophysica Acta, 1996 Oct 9, 1288(2):F55-78. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. A ppendix A. Q uestionnaire for the Proposed Study : A T M G ene M utation and Childhood A cute L ym phocytic Leukem ia Study I D # ______________ N am e o f C h ild :_________________________________________ _ N am e o f M o th e r:________________________________________ Nam e o f Interv iew er:___________________________________ I D # _____ D ate o f Interview : / / Start Tim e o f In te rv ie w :_____ :______central tim e E nd Tim e o f In te rv ie w :_____ :______central tim e D uration o f Interview : m inutes Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section A K ey Dates A l. W hat was ...'s date o f birth: I / M O D A Y Y R A2. H ow m any w eeks did the pregnancy last? (C alculate) W eeks N ine m onths equals 40 weeks E stim ate last norm al m enstrual period / / M O D A Y Y R A3. W hen did you learn from a physician you w ere pregnant? / / M O D A Y Y R A4. H ow m any w eeks pregnant w ere you at this tim e? ________ W eeks A5. D id you ever breast feed . . . ? 1 ...Y es (GO TO A6) 5... N o (GO TO A8) 8... D K 9 ...REF A6. H ow long did you breast feed? A7. D id you bottle f e e d ...? 1...Yes (G O T O A8) 5 ...No (GO TO B l) 8...D K 9 ...REF M onths W eeks A8. H ow old w as . . . w hen you began bottle feeding? m onths to to -p * Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. A9. H ow long did you bottle feed . . .? ________ /________ M onths W eeks A10. W hat was the brand nam e o f the form ula you used m ost o fte n ? ________ _ A l l . To the best o f your know ledge, w as this a soy-based or m ilk-based form ula? 1.. .. Soy 5 .... M ilk 8 .... DK 9 .,.. REF Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission C odes 2 years before L N M P............................ 2 Y rs before Y ear before L N M P 1 Y r before M onth before L N M P ............................ M o before Last norm al m enstrual period............ LNM P E nd T rim ester 1................................ Tri 1 / E nd T rim ester 2 ................................Tri 2 / E nd Trim ester 3 ................................T ri 3 / (D elivery date) B reast feeding ended..................... N urs / to to Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section B M O T H E R 'S P R E G N A N C Y H IST O R Y B l. A re you currently pregnant? 1 Y es 5.... N o 8 D K 9 R e f ________ B 2.H ow m any tim es have you been pregnant, including live births, stillbirths, m iscarriages, and abortions? tim es B3. W hich pregnancy was (index c h ild )? ** N ote that this includes m iscarriages/abortions 1st Pregnancy 2nd Pregnancy 3rd Pregnancy B4. W as the outcom e o f your (lst/2nd/etc. pregnancy a live birth, stillbirth, a m iscarriage, or induced abortion? Live b irth ........................... 1 S tillb irth ............................2 M iscarriage........................3 Induced a b o rtio n ............ 4 O th e r...................................5 Live b irth........................... 1 S tillb irth .............................2 M iscarriage........................3 Induced a b o rtio n............ 4 O th e r...................................5 Live b irth........................... 1 S tillb irth............................ 2 M iscarriage...................... 3 Induced ab o rtio n ............ 4 O th e r...................................5 B5. If a m iscarriage or induced abortion, was there any know n reason or abnorm ality? B6. H ow old w ere you the (lst/2 n d /etc.) tim e you becam e pregnant? Age Age Age B7. H ow long did this pregnancy last? W eeks............................ 1 M o n th s..........................2 W eeks...,........................1 M o n th s.......................... 2 W eeks..............................1 M o n th s........................... 2 B8. IF LIV E BIR TH : On w hat date did you deliver? IF O TH ER : In w hat m onth and year did the pregnancy end? / / M O D A Y YR / / M O D A Y Y R / / M O D A Y Y R IF M ISC A R R IA G E O R A B O R T IO N , R E T U R N T O B4 FO R N E X T PR E G N A N C Y . Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 1 st Pregnancy 2nd Pregnancy 3rd Pregnancy B9. W as the child foil term or prem ature? Full te r m ........................1 Prem ature...................... 2 Full te r m ........................1 Prem ature...................... 2 Full te rm ........................1 Prem ature...................... 2 BIO. W as this a single birth or did you have tw ins or triplets? Single b ir th ..................... 1 T w in s................................ 2 Triplets or m o re ............ 3 Single b ir th ..................... 1 T w in s................................ 2 T riplets or m o re ............ 3 Single b irth .....................1 T w in s................................ 2 T riplets or m o re ............ 3 IF ST IL L B IR T H , R E T U R N TO B4 F O R N EX T PR E G N A N C Y . IF L IV E B IR T H , GO TO B l l o o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 4T H Pregnancy 5TH Pregnancy 6T H Pregnancy 7TH Pregnancy B4. Live b irth ...........................1 S tillb irth ............................ 2 M iscarriage........................3 Induced a b o rtio n ............ 4 O th e r...................................5 Live b irth ........................... 1 S tillb irth .............................2 M iscarriage....................... 3 Induced abortion..............4 O ther....................................5 Live b irth ........................... 1 S tillb irth ............................ 2 M iscarriage........................3 Induced a b o rtio n ............ 4 O ther....................................5 L ive b irth ........................... 1 Stillbirth.............................2 M iscarriage........................3 Induced abortion..............4 O ther...................................5 B5. B6. Age Age Age Age B7. W eeks.............................1 M o n th s .......................... 2 W eek s............................ 1 M o n th s ..........................2 W eek s...................... . 1 M o n th s..........................2 W ee k s............................ 1 M o n th s...........................2 B8 / / / / / / / / M O D A Y Y R M O D A Y Y R M O D A Y Y R M O D A Y Y R IF M ISC A R R IA G E O R A B O R T IO N , R E T U R N TO B 4 F O R N EX T PR E G N A N C Y . Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 4TH Pregnancy 5TH Pregnancy 6TH Pregnancy 7T H Pregnancy B9. Full te rm ....... ................1 P rem atu re ..................... 2 Full te r m ........................1 Prem ature...................... 2 Full te rm ........................1 Prem ature...................... 2 Full te rm ........................1 Prem ature...................... 2 BIO Single b irth ...................................1 T w in s..............................................2 Triplets or m o re ......................... 3 Single b ir th ........................................... 1 T w in s.......................................................2 Triplets or m o re ...................................3 Single b irth ........................................1 T w in s...................................................2 Triplets or m o re ...............................3 Single b irth ........................................1 T w in s...................................................2 T riplets or m o re ...............................3 IF ST IL L B IR T H , R E T U R N TO B4 FO R N E X T P R E G N A N C Y . IF L IV E BIR T H , G O TO B 9. o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Pregnancy # 1st CHILD Pregnancy # ______ 2nd CH ILD Pregnancy # 3rd CHILD B 1 1. W hat w as the sex o f (the/each) child? M ale .............1 Fem ale .............2 M ale ,. 1 Fem ale ........2 M ale 1 Fem ale ........2 B12. W hat is the first nam e o f (the/each) child? B13. Is (child) now living? Y es 1 (N EX T L IV E B IR T H or Section C) N o ..............2 Y es 1 (N E X T L IV E B IR TH or Section C) N o ..............2 Y es..............1 (N E X T L IV E B IRTH or Section C) N o 2 .............. B 14. W hat w as the year o f (child’s) death? B15. W hat w as the cause o f death? R e tu rn to B 4 fo r n ex t p reg n an cy . Do n o t co n tin u e to Section C u n til Q 's B 4-B 15 ________ hav e been asked fo r all p regnancies._______________________ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Pregnancy # Pregnancy # Pregnancy # ______ 6th CHILD Pregnancy # 4th CHILD 5th CHILD 7th CHILD B i t . M a le ................. 1 F em ale..............2 M ale ................. 1 Fem ale ............................2 M ale ...............1 Fem ale ........................ 2 M ale ..............1 Fem ale .......................2 B12. B13. Y es 1 (N E X T LIV E BIRTH or Section C) N o ........... 2 Yes 1 (N E X T LIV E B IR TH or Section C) N o ...............2 Y es 1 (N E X T LIV E B IR TH or Section C) N o ..............2 Y es..............1 (N E X T L IV E B IR T H or Section C) N o 2 .............. B14. B15. R eturn to B4 for next pregnancy. Do not continue to B elow until Q 's B 4-B 14 __________________have been asked for all pregnancies._____________________ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section C FE R T IL IT Y Q U EST IO N S C 1. H ow old w ere you w hen you had your first m enstrual period? years 99 REF C2. H ow old were you w hen your m enstrual periods becam e regular, that is, w hen you could usually predict w ithin about a w eek w hen your periods w ould come, and they tended to last about the sam e num ber o f days each tim e? ____________ Y ears 96_______ N ever becam e regular 9 8 ______ D K 99______ REF______ C3. B efore you cam e pregnant w ith flNDF.X CHI I T ) i, did you ever use oral contraceptives for birth control or for any other reason, such as to regulate your periods or to control acne? 1 YES 5.. NO 8 DK 9 R E F ______ C4. W as that for birth control, to regulate periods, to treat acne, or for som e other reason? (C H E C K A LL TH A T A PPLY ) [____ ] B irth control [_____] R egulate periods [_____] T reat acne [_____] (O ther reason) (SPEC IFY )____________________________________ : _____________ ’ to u > u > Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. C5. W as that a low dose pill, a progesterone only pill, or a standard birth control pill? (IF M O R E TH A N O N E TYPE, CH O O SE LA ST PILL B E FO R E IN D EX C H IL D ’S BIRTH). R efer to the interview guide PA G E 3 for pill description. 1 Low dose 3 .... Progesterone Only 5 Standard 8 D K 9 R EF C6. W hat was the nam e o f the pill? 8 D K 9 R EF Cl. H ow old w ere you w hen you first started using oral co n traceptives? YEARS 9 8 ........ D K 9 9 ........ R E F ______ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. C8. D id you stop using oral contraceptives prior to your pregnancy w ith (IN D EX CH ILD - ), o r w ere you still using them w hen you becam e pregnant? 1 P R IO R > C 9. H ow long before your pregnancy did you stop using them ? M O N TH S 9 9 8 D K 99 9 R E F ______ 5 U SED D U R IN G -------> C IO . H ow far along in the pregnancy w ere you w hen you stopped PR E G N A N C Y using them ? W EEK S 9 9 8 D K 99 9 R E F ______ 8....D K 9....REF C l 1. T hinking about your oral contraceptive use before the birth o f (IN D EX CH ILD ), about how m any tim es did you go o ff the pill from the tim e that you first began taking the pill? TIM ES 99 8 D K 99 9 REF to u > e n Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. C12. A ltogether, how m any m onths or years did you use oral contraceptives prior to your pregnancy w ith IN D E X C HILD? 1...........M O N T H S :______ 3 ...........Y E A R S :_________ 9 8...........DK 9 9 R E F ______ C13. B efore you cam e pregnant w ith (IN D EX CH ILD ! , did you ever use contraceptive horm one im plants o r injections for birth control or for any other reason, such as to regulate your periods or to control acne? 1 YES 5 NO 8. D K 9 R E F ______ C14. W as that for birth control, to regulate periods, to treat acne, or for som e other reason? (C H EC K A LL TH A T A PPL Y ) | B irth control [____] R egulate periods [____] Treat acne [____] (O ther reason) (SPECIFY)_ C15. H ow old w ere you w hen you first started using contraceptive im plants or injections? YEARS 9 8 ........... D K 9 9 ............REF t o u > O S Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. C l 6. D id you stop using contraceptive im plants or injections prior to your pregnancy w ith fIN D EX CHILD'), or w ere you still using them w hen you becam e pregnant? 1 ........... PR IO R ------------- > C 17. H ow long before your pregnancy did you stop using them ? M O N TH S 99 8..... DK 99 9..... R E F ______ 5 ........... U SED D U R IN G > C 18. H ow far along in the pregnancy w ere you w hen you stopped PR EG N A N C Y using them ? ; W EEKS 99 8...... D K 99 9 ....... R E F ______ 8........... D K 9 ............R E F ______ C19. A ltogether, how m any m onths or years did you use contraceptive im plants or injections prior to your pregnancy w ith IN D EX CHILD? 1............M O N TH S:_______ 3............ Y E A R S :________ 9 8 ..........D K 9 9 ..........R E F __ C20. N ow thinking about the tw o year period before you becam e pregnant w ith (IN D EX CHILD), did you use a diaphragm , condom , or som ething sim ilar? 1 YES 5 NO 8 D K 9 R E F ______ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. C 2 1. A nd during the tw o year period before you becam e pregnant w ith fIN D EX C H IL D ), did you use jelly, cream, foam, o r suppository contraceptives? 1 Yes ------> C22. W hat brand did you use? _______________________________________ 5 NO 8 D K 9 R E F ___ C23. D id you stop using that prior to your pregnancy w ith (IN D EX CHILD'), or were you still using it w hen you becam e pregnant? 1 PR IO R ----------------> C24. H ow long before your pregnancy did you stop using it? M ON TH S 9 9 8.......D K 99 9.......R E F ______ 5 U SED D U R IN G PREG N A N C Y 8 D K 9 R EF ■ > C 25. H ow far along in the pregnancy w ere you w hen you stopped using it? W EEKS 9 9 8 DK 99 9 R E F ______ to u > 0 0 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. C26. D id you use an I.U .D .? 1 Y es --------> C 27. P rio r to your pregnancy w ith IND EX C H ILD , did you have it rem oved, or was it still in 5 ......N o place w hen you becam e pregnant? 8 D K 9 R EF_____ 1 PR IO R -----------------> C 28. H ow long before your pregnancy did you have it rem oved? (M O N T H S )_________ 99 8.........D K 99 9 R E F ____ 5...... U SED D U R IN G — > C 29. H ow m any w eeks pregnant w ere you w hen you had it rem oved? (W EEK S) _______ 99 8 D K 99 9 R EF 8 D K 9 R E F _ _ C30. D id you use any other form o f b irth control? 1 Y E S > C 31. W hat w as that? 5 N O 8 D K 9 R E F C32. A t the tim e you becam e pregnant w ith (IN D EX CHILD), w ere you trying to get pregnant? 1 YES ---------- > C 33. H ow long had you been trying?_________ M O N TH S (IF 12 O R M O R E M O N T H S, G O TO C 35) 5 N O (C4.) 999 R E F ______ 8 D K 9 R EF ts> u > < 3 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. C34. D id you ever try for one straight year or m ore to becom e pregnant and during that tim e n o t becom e pregnant? 1 YES 5 NO 8 D K 9 R E F ______ C35. D id you and (IN D EX C H ILD - ) .‘s father ever visit a doctor or clinic because it w as difficult to get pregnant? 1 YES 5 N O (G O TO SE C T IO N D) 8 DK 9 R E F ______ C36. D id the doctor tell you w hat the reason was for your difficulty in becom ing pregnant? 1 ..YES ---------- > C 37. D escribe 9 9 9....... REF 4 N O 8 .DK 9 ,.REF C38. D id you have a surgical procedure to treat the problem ? 1 ..YES ---------- > C 39. D escribe 999....... REF 5 J .N O 8 ..DK 9 ..REF Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. C40. D id you receive tablets or injections for the problem , such as Clom id? 1 YES 5 NO 8.......D K 9 R E F ______ C 4 1. D id you have som e other treatm ent for the problem ? 1 Y ES ---------- > C 42. D e sc rib e _____________________________________ 999... 5 N O 8 DK 9. . R E F ______ .REF Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section D A B N O R M A L IT IE S A N D H E R E D IT A R Y C O N D IT IO N S Were any of your children born with: B IR T H D EFEC T P re g # P re g # Preg # ______ Preg # nam e nam e name nam e D l. C left lip or palate 1....... Yes 1........Yes 1........Yes 1....... Yes 5....... N o 5........N o 5 ........No 5........No 8....... D K 8 ........D K 8 ........D K 8........D K 9....... REF 9 ........R EF 9 ....... R EF 9 ....... R EF D2. Spina bifida or other spinal defect 1....... Yes 1........Yes 1....... Yes 1........Yes 5....... N o 5........N o 5.......N o 5 ....... N o 8....... D K 8....... D K 8........D K 8........D K 9....... R EF 9........REF 9....... R EF 9 ....... R EF D3. Large or m ultiple birthm arks (A ny one larger than a 1........Yes 1....... Yes 1...... .Yes 1....... Yes quarter; or 6 or m ore about the size o f a dime) 5 ....... No 5....... N o 5....... N o 5........N o 8....... D K 8 ........D K 8....... D K 8....... DK 9....... R EF 9....... R EF 9.......REF 9 ....... REF D4. D eafness or hearing im pairm ent 1....... Yes 1....... Yes 1....... Y es 1....... Yes 5 ........N o 5....... N o 5....... N o 5....... No 8....... DK 8....... D K 8....... D K 8....... D K 9....... R EF 9 ........R EF 9 ........REF 9 ....... REF = t f c 6 0 o £ fa £ o !2 > < 2 Q fa £ o >-ZQ 8 o ^ > * 2 Q fa = 8 = 6 1 ) o j- fa 8 0 ^ > - 2 Q u o >- 2 fa £ o ^ >"2; Q fa S o « > - Z Q fa fa £ o !2 k * 2 Q £ o ><2 0 £ o >-'2 0 fa SSoZ ><2 0 fa = t t 6 0 u £ o > 4 > 2 Q fa t r t 0 O C O £ q i - I < 2 1C O< 3 \ S o !2 > Z Q £ o !2 >-2 0 fa = B = 6 0 H 0 fa fa fa 0 1 e 4 3 £ o ><2 0 fa £ o ^ ><2 0 fa £ o W ><2 0 fa — • in 00 o s cl o U c o O c-i Q d o U £ o W ><2 0 d o U -'t Q Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. B IR T H D EFEC T Preg # Preg # Preg # P re g # nam e nam e name nam e D5. B lindness or vision problem s 1........Yes 1....... Yes 1....... Yes 1....... Yes 5....... N o 5....... N o 5....... N o 5....... N o 8....... D K 8....... D K 8 ....... D K 8....... D K 9 ........REF 9 ....... REF 9....... R EF 9 ....... REF D6. D ow n syndrom e 1....... Yes 1....... Y es 1....... Yes 1........Yes 5........N o 5....... N o 5.......No 5 ........No 8........D K 8....... D K 8....... D K 8....... D K 9 ........R EF 9 ....... R EF 9 ....... REF 9 ........REF D7. M ental retardation or a severe learning disability 1........Yes 1........Yes 1.......Yes 1....... Y es 5....... N o 5....... N o 5....... No 5....... N o 8....... D K 8....... D K 8....... D K 8 ........D K 9 ........R EF 9 ....... R E F 9....... REF 9....... REF D8. U nusually sm all head or m icrocephaly 1........Y es 1....... Yes 1.......Yes 1........Yes 5........N o 5....... N o 5 ....... N o 5........N o 8....... D K 8 ....... D K 8....... DK 8 ........D K 9 ........R EF 9....... R EF 9........R EF 9 ........R EF to a 03 C 8 o Oh = t t M O h a 00 a > O^ M > ? £ q 2 r -3 <n 0 0 = f c ao O h 8 O > Z Q = 1 % ao 8 o O d > Z Q O h = t t h O o 8 o o d > « 2: Q O h H U M O h w Q f f i u C Q a o U Q Oh 8 O i4 ><z;o cn O 0 q q tn u 0 q > *. e q 2 £ q q i n C C O s q i n 0 0 8 o i4 > Z D V i C D © q q V i C D 0 q q q 2 q Q q i n 0 0 cK q »n 0 0 a o O V O Q S o « > Z O 8 o t * d ><ZQ F < Oh 8 O « > z a Oh a o o Q 8 o > Z Q O h S ofcd > < £ Q Oh a o u Q 245 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. B IR T H D EFEC T Preg # ______ P re g # Preg # Preg # nam e name name nam e D9. U nequal sized lim bs or hem ihypertrophy 1....... Yes 1........Yes 1........Yes 1........Y es 5....... N o 5........N o 5....... N o 5....... N o 8....... DK 8....... D K 8 ........D K 8........D K 9 ........REF 9 ....... REF 9....... R E F 9 ........R EF DIO. B irth defect o f the h eart (list) 1....... Yes 1....... Yes 1....... Yes 1....... Y es 5 ........N o 5 ....... No 5 ........N o 5....... N o (SPEC IFY if possible) 8........D K 8....... DK 8........D K 8 ........D K 9 ....... R EF 9 ....... R EF 9 .......REF 9 ........R EF Specify Specify Specify Specify D l l . B irth defect o f the pancreas or digestive tract (list) 1....... Yes 1...... Yes 1........Yes 1.......Yes 5....... N o 5........N o 5....... N o 5....... N o (SPEC IFY if possible) 8....... D K 8....... D K 8 ........D K 8....... D K 9 ........REF 9 ....... REF 9....... R EF 9 ....... R EF Specify Specify Specify Specify to O n = * f c t d ) ft % b O ft =tt b o b O bO ft u w ft w Q I 3 o Z Q p H p % S o ^ > > Z Q p H 8 o ^ > Z D P . c o O 0 \ Q SoW > « Z. Q Uh « E ? p . CO 8 O ^ >- Z Q P h t & * e u C O « & > o , c o 8 O ^ > < Z Q H j c & O o ft c o > ft £ ? o 1 ) ft C O c o u © 5 So^! >-ZQ ft ft 8 o & ■ > £ a ft c n < D O ^ ft] >? £ q 2 tri 06 cK S3 O U Q < & ft 0 0 ’ 5 ft c o ft C O £ 8 ft CO *? 3 < l > ft CO 247 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. B IR T H D EFEC T Preg # P reg # Preg # ______ P re g # name nam e name nam e D I2 . A ny kidney, bladder or sex organ abnorm alities 1........Yes 1........Y es 1........Yes 1........Yes 5 ........N o 5....... N o 5....... N o 5....... N o (SPEC IFY ) 8 ........D K 8....... D K 8........DK 8........D K 9 ........R EF 9....... R E F 9....... R EF 9....... REF Specify Specify Specify Specify D 13. O ther chrom osom al abnorm ality such as T urner’s 1........Y es 1........Y es 1.......Yes 1....... Yes syndrom e, K linefelter’s syndrom e, trisom y 18, o r trisom y 5....... N o 5 ....... N o 5....... N o 5....... N o 13. 8 ........D K 8 ....... D K 8....... D K 8....... D K 9 REF 9 R EF 9 R EF 9 R EF describe describe describe describe 0 0 fa & fa 00 = t f c o o o U h fa a > I s. 8 o o fa C / D = f c 0 0 O h I e fa e g * o u fa 00 = t f c 0 0 1 ) s- l fa 8 o ^ !* £ Q fa e g * 'o o fa CO I d u o « ><ZQ 8 u a. c o H O W d , W Q X § 3 S C o O ( N 3 8 o >^ZQ Z Q u o i * j > < X Q S o fc Z i * !z Q fe Hh c o U C * S 5 249 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. = t f c W > o s- Oh 5 f t 5 0 O P * 8 o > - £ Q tu =fc 5 0 c n u o >? £ q 2 >n O O O N 5 f t 5 0 O h 8 o ^ > Z * Q H U w p H w Q I 3 • * ? } ■ 5 * '§ p . C O S o « t* £ Q ft GO 00 a> o ^ q £ q 2 q i n 0 0 O n ft C O o * 4 £ Q <§• 3 o ft m ■o | !S o £ ? o u a. C/2 250 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. = t f c 8 O M > * Z Q fa O h CO = t f c O J D ( Z 1 * . q 2 - > o 0 0 O s O - co = » t b o 8 0 ^ > * Z O 8 0 ^ > * Z G fa O h O O = t f e b o < D fa 8 0 ^ fa 8 o ^ > ■ Z Q fa 'E ? '3 4 > O h C O 8 0 ^ fa O £ Q fa O . CO fa U fa fa fa Q I e s m c o O ■ * ? 5 a o C J Q Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. H ER ED ITA R Y C O N D ITIO N S P re g # Preg # Preg # Preg # name name name name D 16. Skin syndrom es such as: N evoid basal-cell carcinom a syndrom e, Fam ilial tricoepitheliom a, X eroderm a pigm entosum (X P), Fam ilial atypical m ole- m elanom a (FA M M M ) syndrom e, W erner’s syndrom e or progeria 1 Yes 5 N o 8 D K 9 .....REF 1 Yes 5 No 8 D K 9 .....R EF 1 Yes 5 N o 8 D K 9 .....R EF 1 Yes 5 N o 8 D K 9 .....REF Specify Specify Specify Specify D17. N eurocutaneous syndrom es such as: 1....... Yes 1........Yes 1........Y es 1....... Yes N eurofibrom atosis (NF), 5....... No 5....... N o 5........N o 5....... No • 8....... DK 8....... D K 8........D K 8....... DK T uberous sclerosis, 9 ....... R EF 9.......R EF 9 ....... R EF 9....... REF V on H ippel-L indau disease, Sturge-W eber syndrom e Specify Specify Specify Specify N > to = t f c b O o > s- Ph 3 4 > a o o £ 5 5 O h C / 3 = * f c W > S O ' ^ ><ZQ & 5 5 O h 00 « & C u 0 3 = t f c b o Ph o £ Q Uh o « D C u 00 5 3 o ^ > < £ Q '§ O h 0 3 = H = b Q < D t / 5 < Uo Z q in 0 0 < & O h 03 5 3 o ^ > < £ Q o - 0 3 % C U O O h >< 0 3 ‘ z C r t 1 Z , q q v S 0 0 te n < & O h 03 a o a v d s G O < DO q q i / S 0 0 o o U ■ 5 £ o < D O h 0 3 253 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. H ER ED ITA R Y C ON DITIONS Preg # Preg # ______ Preg # Preg # name nam e nam e name D18. Skeletal disorders such as: M ultiple exostoses, Endochondrom atosis, Cherubism , Fibro-osseous dysplasia, C artilage-hair-hypoplasia syndrom e, A chondroplasia (dwarfism ), O steogenesis im perfecta 1 Yes 5 No 8 D K 9 R EF Specify 1 Yes 5 No 8 D K 9 R EF Specify 1 Yes 5 N o 8 D K 9 REF Specify 1 Yes 5 No 8 D K 9 REF Specify D19. Endocrine system disorders such as: M ultiple endocrine neoplasia (M EN), syndrom e, Fam ilial pheochrom ocytom a 1 Yes 5 N o 8 D K 9 .....REF 1 Yes 5 N o 8 D K 9 .....REF 1 Yes 5 N o 8 D K 9.....R EF 1 Yes 5 No 8 D K 9 .....REF Specify Specify Specify Specify K ) L r i 4 ^ = * f c 0/J c u 'o 0 ) c p 0 3 D O . . > * z a U s c p 0 3 P h i G O t a d £ P U h » - < m 00 O n £ o < 1 > O - 0 5 O t4 Z p P P < & » a, 0 3 = t f c O D S o ^ > < £, Q Uh a 0 3 o U d P Q £ Q P d £ o D a. 03 % b o Oh § 3 O > Z Q P - £ * C P 0 3 S O ta d > H £ Q CP 03 = f f c O D P h D O t * 4 < & * '§ C P 0 3 G O u 00 5 o t ^ d Z Q C P 0 3 G O O On 5 255 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. H ER ED ITA R Y C O N D ITIO N S Preg # Preg # P re g # Preg # nam e name nam e name D20. Intestinal syndrom es such as: C ystic fibrosis (CF), Polyposis coli, juvenile polyposis or G ardner’s syndrom e, Peutz-Jeghers syndrom e 1 Yes 5 No 8 DK 9 .....R EF 1 Yes 5 N o 8 DK 9 R EF Specify 1 Yes 5 No 8 D K 9 .....REF 1 Yes 5 N o 8 D K 9 R EF Specify Specify Specify D 21. Chrom osom al fragility syndrom es such as: X eroderm a pigm entosum (XP), B loom ’s syndrom e, F anconi’s anem ia A taxia telangiectasia (AT), D yskeratosis congenita 1 Yes 5 No 8 D K 9 R EF Specify 1 Yes 5 No 8 D K 9 R EF Specify 1, ,.Y e s 5 N o 8 DK 9 REF Specify 1 Y es 5 N o 8 D K 9 REF Specify t o U\ o\ Cj Q C u a , o o O h G O C X o 4 Z Q '§ O h C O P H O h C O = t f c 00 O h c f r O h C O 8 o \ 4 > - z o E x * o . C O £ X o 4 E x P Q Z q v > 00 c K 5 & 8 C L . CO ■Oh C O = * f c a t ) o U h O h 8 o 4 ■ > - Z Q 4 ? 5 0 3 O h C O & o - c o c o o © CM Q o o u CM p 257 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. H ER ED ITA R Y C ON DITIONS Preg # Preg # ______ Preg # Preg # nam e name nam e name D 22. Im m unodeficiency syndrom es such as: 1....... Yes 1....... Yes 1....... Y es 1....... Yes W iscott-A ldrich syndrom e, 5 ....... No 5........N o 5....... N o 5....... N o A gam m aglobulinem ia, 8....... DK 8....... D K 8 ........D K 8........D K Com m on variable Im m unodeficiency, 9 ....... REF 9 ........REF 9 ....... R EF 9....... REF A taxia telangiectasia (AT), IgA deficiency, Specify Specify Specify Specify X -linked lym phoproliferative syndrom e D23. O ther hereditary conditions such as: 1....... Y es 1........Yes 1........Yes 1....... Yes A niridia, 5....... N o 5....... N o 5....... N o 5....... N o 8....... D K 8....... D K 8....... D K 8....... D K B eckw ith’s or B eckw ith-W iedem ann syndrom e, 9....... REF 9 ....... R EF 9 ........R EF 9 ........R E F Polycythem ia vera Specify Specify Specify Specify t o L h 00 = f f c O S i c 8 c u 0 3 o f e d £ Q fa Cu 0 3 =8= 00 < l > J - M fa o f c a d 2 Q fa Cu 0 3 £ o < D C L . 03 = t f e 00 fa w g fa H e o < D fa 0 3 > < fa & O. 03 = 8= 00 8 o ^ i* £ Q '£ * o < u c u 03 fa C u 03 = 8 = 00 fa fa '§ c u 0 3 fa c u 0 3 G O U < N < N Q " c o U to C N Q 259 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section E C H IL D ’S M E D IC A L H IST O R Y N O W W E W O U LD LIK E SO M E IN FO R M A TIO N O N E A C H OF Y O U R PR EG N A N C IES TH A T RESU LTED IN A STILLB IR TH O R L IV E BIRTH. Have any of your children ever had: D ISEA SES Preg # Preg # Preg # P re g # nam e name nam e nam e E l. D iabetes 1 Yes 5 No 8 D K 9 R EF if yes, age at dx 1 , Yes 5 No 8 D K 9 R EF if yes, age at dx 1 Yes 5 No 8 DK 9 .....R E F ___ if yes, age at dx 1 Yes 5 No 8 D K 9 .....R E F ___ if yes, age at dx E2. A sthm a 1 Yes 5 N o 8 DK 9 REF if yes, age at dx 1 Y es 5 N o 8 D K 9 REF if yes, age at dx 1 Yes 5 No 8 DK 9 .....R E F ___ if yes, age at dx 1....... Yes 5 N o 8 D K 9 R EF if yes, age at dx E3. Epilepsy 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 DK 9 R EF if yes, age at dx 1 Yes 5 No 8 D K 9 R EF if yes, age at dx Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D ISEA SE Preg # P re g # Preg # Preg # Preg # nam e nam e nam e nam e nam e E l. Cont. 1 Yes 5 No 8 D K 9 R EF if yes, age at dx 1....... Yes 5 N o 8 DK 9 R EF if yes, age at dx 1....... Yes 5 N o 8 D K 9 REF if yes, age at dx 1 Yes 5 N o 8 D K 9 R E F if yes, age at dx 1 Yes 5 No 8 DK 9 .....R E F ___ if yes, age at dx E2. Cont. 1 Yes 5 No 8 D K 9 .....R E F ___ if yes, age at dx I.,;,...Y es 5 No 8 D K 9 R E F if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 R E F if yes, age at dx 1 Yes 5 No 8 D K 9 R E F if yes, age at dx E3. Cont. 1 Yes 5 No 8 DK 9 R EF if yes, age at dx 1....... Yes 5 No 8 D K 9 REF if yes, age at dx 1 Yes 5 N o 8 D K 9 .....R E F ___ if yes, age at dx 1 Yes 5 N o 8 D K 9 .....R E F ____ if yes, age at dx 1 Yes 5 No 8 D K 9 R EF if yes, age at dx to On Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D ISEA SE P re g # Preg # P reg # Preg # nam e name nam e nam e E4. B enign tum ors or abnorm al grow ths 1....... Yes 1........Yes 1........Yes 1....... Y es 5....... N o 5........N o 5 ........N o 5....... N o 8 ....... DK 8........D K 8 ........D K 8....... DIC 9 ....... REF 9....... REF 9 ........REF 9 ........R EF describe describe describe describe If yes, age at dx I f yes, age at dx If yes, age at dx ______ _ I f yes, age at dx E5. C ancer or leukem ia 1....... Yes 1....... Yes 1........Yes 1....... Yes 5....... N o 5....... No 5 ........N o 5....... N o 8....... D K 8........DK 8 ....... D K 8....... D K 9 ....... R EF 9....... REF 9 ........R EF 9....... R E F describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx tO 0\ to Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D ISEA SE Preg # P re g # P re g # P re g # Preg # name nam e nam e nam e name E4. Cont. 1....... Yes 1........Yes 1....... Yes 1....... Y es 1....... Yes 5....... N o 5........N o 5....... N o 5 ....... No 5........N o 8 ........D K 8........D K 8.......DK 8 ....... D K 8........DK 9 ....... R EF 9 ....... REF 9....... REF 9 ........R EF 9 ....... R EF describe describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx E5. Cont. 1........Yes 1........Yes 1....... Yes 1....... Yes 1........Yes 5........No 5........N o 5....... N o 5........N o 5....... N o 8 ........D K 8....... D K 8....... D K 8....... D K 8....... D K 9 R EF 9 REF 9 R EF 9 R EF 9 R EF describe describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx t o C \ u > Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D ISEA SE Preg # Preg # Preg # Preg # nam e nam e nam e name E6. Thyroid disease (over active thyroid, under 1........Yes 1....... Yes 1........Yes 1........Yes active thyroid, G raves’ disease, etc) 5........No 5....... N o 5....... No 5 ....... N o 8....... D K 8 ........D K 8 ....... DK 8....... D K 9 ....... REF 9 ........R E F 9 ....... R EF 9....... R EF describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx E7. A ny hem ophilia, or sickle cell anem ia ? 1....... Yes 1....... Y es 1........Y es 1....... Y es 5 ....... N o 5 ........N o 5........No 5....... N o 8 ....... D K 8........D K 8 ........D K 8....... D K 9 ........REF 9 ....... REF 9 ........R E F 9....... REF describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx to Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D ISEA SE Preg # Preg # Preg # Preg # Preg # nam e name name nam e nam e E6. Cont. 1 . . Yes I ....... Yes 1....... Yes 1........Y es 1....... Y es 5...... .No 5....... N o 5........N o 5........N o 5....... N o 8 ....... D K 8.......D K 8........DK 8 ....... D K 8....... D K 9 R EF 9 REF 9 REF 9 R EF 9 R E F describe describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx E7. Cont. 1....... Yes 1....... Yes 1....... Y es 1........Y es 1....... Y es 5 ....... N o 5....... No 5........N o 5 ........N o 5....... N o 8 ....... D K 8....... DK 8 ......D K 8 ........DK 8 ....... D K 9 R EF 9 R EF 9 REF 9 R EF 9 R EF describe describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx to O n Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. DISEA SE Preg # Preg # Preg # Preg # nam e nam e nam e name E8. M um ps I Y es 5 N o 8 D K 9 .....REF____ if yes, age at dx 1 Yes 5 No 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 .......REF___ if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx E9. Eczem a 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 REF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx E10. C hicken pox 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 , Yes 5 No 8 D K 9 R E F if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx t o ON ON Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D ISEA SE Preg # Preg # Preg # Preg # Preg # name nam e nam e name nam e E8. Cont. 1 Yes 5 No 8 DK 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 JDK 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 REF if yes, age at dx 1 Yes 5 No 8 D K 9.....R EF_ __ if yes, age at dx E9. Cont. 1.. Yes 5 No 8 DK 9 REF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 N o 8 DK 9 .....R EF_ __ if yes, age at dx E10. Cont. 1 Yes 5 No 8 D K 9 .....R EF___ if yes, age at dx 1 Yes 5 N o 8 D K 9 REF if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx 1 Yes 5 No 8 D K 9 .....R E F___ if yes, age at dx 1 Yes 5 N o 8 D K 9 R EF if yes, age at dx o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. DISEA SES P re g # Preg # ______ Preg # Preg # name nam e nam e nam e E ll. M easles 1........Yes 1........Yes 1........Yes 1........Yes 5 ....... N o 5........N o 5........N o 5........N o 8....... D K 8........DK 8........D K 8........D K 9 ....... REF 9 ....... REF 9 ....... R EF 9........R EF if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx E l 2. A ny other diseases? 1....... Yes 1....... Y es 1....... Y es 1.......Y es 5........No 5....... N o 5 ........No 5....... No Specify 8 ........D K 8 ........D K 8....... D K 8.......D K 9 ........R EF 9 ....... R EF 9 ........R E F 9 ........R E F describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx to O s 00 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D ISEA SE P re g # P re g # P re g # Preg # Preg # ______ nam e nam e name nam e nam e E ll . Cont. 1........Yes 1....... Yes 1....... Yes 1....... Yes 1....... Y es 5....... N o 5....... No 5....... No 5....... N o 5 ........N o 8....... DK 8....... D K 8....... D K 8 ........D K 8....... D K 9 ........R EF 9....... REF 9....... REF 9 ........REF 9 ....... R EF if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx E l 2. Cont. 1....... Yes 1....... Y es 1....... Yes 1... ...Yes 1........Yes 5 ........N o 5....... N o 5 ....... N o 5....... No 5........N o 8 ........D K 8....... D K 8 ....... D K 8 ....... D K 8 ....... D K 9 ......R EF 9 ........REF 9 ....... REF 9 ........R EF 9 ....... R EF describe describe describe describe describe if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx if yes, age at dx Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section F M E D IC A L H IS T O R Y (M O T H E R : IN F E C T IO U S D IS E A S E ) In your lifetime have you ever had any o f the following diseases: IN FEC TIO U S D ISEA SE H ad disease? W as this diagnosed by a physician? age o f diagnosis or onset F I. H epatitis A (infectious hepatitis, usually from 1....... Yes 1....... Yes som ething one ate) 5 ........No 5....... N o 8....... DK 8....... D K 9 ........REF 9 ....... R EF F2. H epatitis B (serum hepatitis, from blood products, or 1.. ....Yes 1....... Yes sexual transm ission) 5........N o 5....... No 8........D K 8........D K 9....... R EF 9 ....... REF F3. H epatitis C (non-A or B ) sim ilar in transm ission as B 1........Yes 1....... Yes 5........N o 5....... No 8........D K 8........D K 9...... REF 9 ........REF F4. R ubella or G erm an M easles 1........Y es 1....... Yes 5....... N o 5 ........No 8....... D K 8........D K 9 ........REF 9 ....... REF t o o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. IN FEC TIO U S D ISEA SE H ad disease? W as this diagnosed by a physician? age o f diagnosis or onset F5. M um ps 1....... Yes 1....... Yes 5....... No 5 ....... No 8....... D K 8....... DK 9....... R EF 9....... REF F6. C hicken Pox 1....... Yes 1....... Yes 5....... No 5.......No 8....... D K 8 ....... DK 9....... REF 9 ....... R EF F7. R egular M easles 1....... Yes 1....... Yes 5....... No 5....... N o 8........D K 8....... DK 9 ....... REF 9 ....... R EF F8. Shingles 1....... Yes 1....... Yes 5....... N o 5........No 8....... D K 8........DK 9... ...R E F 9 ....... REF • K > --4 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. IN FEC TIO U S D ISEA SE H ad disease? W as this diagnosed by a physician? age o f diagnosis or onset F9. M ononucleosis (m ono) or glandular fever 1....... Yes 1....... Yes 5....... N o 5....... No 8....... D K 8....... D K 9....... REF 9 ....... REF F10. O ther infectious or contagious disease (besides 1........Yes 1....... Yes influenza) 5....... N o 5....... N o 8....... D K 8....... DK SPEC IFY 9 ....... R EF 9....... REF F I 1. O ther infectious or contagious disease (besides 1....... Yes 1....... Yes influenza) 5....... N o 5....... No 8....... D K 8 ....... DK SPEC IFY 9....... REF 9 ....... R EF to ■ O t o SECTION G FAMILY MEDICAL H IST O R Y o o o c G g OS -X C J 3 t/j "2 ‘ o 1 “ T 3 a 3 o 3 H O J? c 8 + - » < D *3 * + - » r. y c o 0 3 < D 3 o a o T 3 & O Q * ^ s 5 S « s » I 1 & § S S ' I s « 3 S Q > 3 fa O i X! C O ( D .j O t i 3 . X > < z O c * T 3 I S O SP r - I ir i 00 O n 8 O ^ ><ZQ fa ■ a 4 ) 00 a X X X 3 O > - o SP 273 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o ^ ZQoi O 8 o tJ x T3 §P 4 > . Q S Q J X ) J- O £ ' 5 O J * < \< X . *. S Sfil s s s x x x * X X X X x x x x x x x x x x x x x X X X X X X X X X X x X X X X X X X X X X X X X X X X X , X X X X X X x X X X X X X . X X X X X X X . X X S S S S x X , X X X X X X X X X X , X X X X X X X X X X X e 1) X! T 3 ts Q rS X T3 p SP < u £} a . c s ” 2 O o u 274 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. IN D IVIDUA L D ow n syndrom e over active thyroid or G raves’ disease under active thyroid leukem ia or blood disease G2. Y our B iological M other 1....... Yes 1........Yes 1........Y es 1....... Yes 5....... N o 5....... N o 5....... No 5........N o alive? 1........Yes 8....... D K 8...... DK 8 ........D K 8 ....... D K 5....... No 9....... REF 9 ....... REF 9....... REF 9 ....... REF 8....... D K 9 ........REF age at dx age at dx age at dx current age or age at death ___ D escribe: cause o f death G3. Y our B iological Father 1........Yes 1....... Y es 1........Y es 1....... Yes 5....... N o 5....... No ....... N o 5........No alive? 1....... Yes 8....... DK 8 ........D K 8....... D K 8....... D K 5....... N o 9 ....... REF 9....... R EF 9 ....... REF 9 ....... REF 8....... D K 9 .......REF age at dx age at dx age at dx current age or age at death ___ D escribe: cause o f death Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. other cancer benign tum or A lzheim er’s disease senility or dem entia epilepsy other condition G2. C ont 1........Yes 1....... Yes 1....... Y es 1....... Yes 1....... Yes 1....... Yes 5....... N o 5........No 5....... N o 5....... N o 5....... No 5 ........N o 8....... D K 8....... D K 8 ........D K 8........D K 8....... D K 8 ........D K 9 R EF 9 R EF 9 R EF 9 R EF 9 REF 9 R EF age at d x ___ age at d x ___ age at dx age at dx___ age at d x ___ age at d x ___ Describe: Describe: Describe: Describe: G3. Cont 1....... Yes 1....... Yes 1........Yes 1........Yes 1....... Yes 1........Yes 5........N o 5....... No 5........N o 5....... N o 5....... N o 5........No 8....... D K 8....... D K 8....... D K 8....... D K 8........DK 8........D K 9 REF 9 R EF 9 R EF 9 R EF 9 REF 9 R EF age at d x ___ age at d x ___ age at dx___ age at dx___ age at d x ___ age at d x ___ Describe: Describe: Describe: D escribe: t o ■ O Os Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. IN D IV ID U A L D ow n syndrom e over active thyroid or G raves’ disease under active thyroid leukem ia or blood disease G4. biological sibling 1........Yes 1........Yes 1....... Y es 1....... Yes 5........N o 5........N o 5....... N o 5 ....... N o nam e 8 ........D K 8........D K 8....... D K 8........D K 9 ....... REF 9 ........R EF 9.......R EF 9 ....... REF m ale or fem ale? age at dx age at dx age at dx full or h a lf sib? D escribe: alive? 1........Yes 5 ....... N o 8........D K 9 ....... R EF current age or age at death ___ cause o f death to S3 O T 3 a o € o P H p p x o & o P ^ £ Q X T3 O P 4 Z Q P h X T 3 C S 00 S O ^ ><Z G X •o d > S’ a a -S T 3 t> d o a S 278 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. IN D IV ID U A L D ow n syndrom e over active thyroid or G raves’ disease under active thyroid leukem ia or blood disease G5. biological sibling 1....... Yes 1....... Yes 1....... Yes 1........Yes 5....... N o 5....... No 5 ....... No 5 ........N o nam e 8....... D K 8....... D K 8.......D K 8 ....... D K 9....... REF 9 ........R EF 9 ........R EF 9 ........R E F m ale or female? full or h a lf sib? age at dx age at dx age at dx alive? 1....... Yes D escribe: 5....... N o 8....... DK 9....... R EF current age or age at death ___ cause o f death b O 8 O > < 5 5 Q . f i x o o 5 5 Q x -o ■ s 3 I* Z Q - K T3 C 6 0 o 55 Q f c £ * « Z Q • s o SP o Q c o U i r i o 280 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. IN D IV ID U A L D ow n syndrom e over active thyroid or G raves’ disease under active thyroid leukem ia or blood disease G6. biological sibling 1........Y es 1....... Y es 1....... Yes 1....... Yes 5........No 5.......No 5....... N o 5....... No nam e 8....... DK 8........D K 8.......D K 8....... DK 9 .......REF 9 ....... R E F 9....... R EF 9....... R EF m ale or female? age at dx age at dx age at dx full or h a lf sib? Describe: alive? 1....... Y es 5....... N o 8....... D K 9 R EF current age or age at death ___ cause o f death b o 00 73 c o U h 4 c £ 4 ) SP ' » r > o o X •o < D S ' o £ '3 < D X : * a " 3 < u § p o § 5 •§ 3 fe £ I 3 a 6 0 -a 6 0 03 O - fi -C O & -S C 3 O U 's d O 282 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. IN D IV ID U A L D ow n syndrom e over active thyroid or G raves’ disease under active thyroid leukem ia or blood disease G7. biological sibling 1....... Yes 1........Yes 1....... Y es 1........Yes 5........N o 5....... No 5....... N o 5........N o nam e 8....... D K 8........D K 8 ....... D K 8....... D K 9....... R EF 9 ........REF 9 ....... R EF 9........REF m ale or female? age at d x ____ age at dx age at dx full or h a lf sib? Describe: alive? 1....... Yes 5........N o 8....... D K 9 ....... R EF current age or age at death ___ cause o f death C O N TIN U E W ITH AS M A N Y B IO LO G IC A L SIBLING S U N TIL C O M PLETE 8 O M > < £ Q £ ft > - 2 Q x T3 C 2 < & B < 2 Q x T 3 a > SP 4) ft C S o a k a 284 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section H E V E N T S A R O U N D TH E B IR T H O F T H E IN D E X CH ILD H I . A t the tim e you becam e pregnant w ith (IN D EX C H ILD ), approxim ately how m uch did you w eigh? (PO U N D S) 998 DK 999 REF H2. D uring the first th re e m o n th s o f your pregnancy, did you gain weight, lose weight, or stay about the same? 1........... G A IN E D --------------- > H 3. A bout how m uch w eight did you gain in those first three m onths? (PO U N D S) 998 D K 999 R EF____ 3 ........... L O S T --------------------> H 4. A bout how m uch w eight did you lose in those first three m onths? (PO U N D S) 998 D K 999 R EF____ 5 ........... SAM E 8 ........... D K 9 ........... R EF H5. H ow m uch total w eight did you gain during the entire pregnancy? ___ (PO U N D S) 998 D K 999 REF to 00 U \ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. H6. H ow tall are you? (R EC O R D IN FE E T AN D IN C H ES) ______FE E T AN D _______ INCH ES 9 8 DK 9 9 ........R E F ________ H7. D uring your pregnancy w ith (IN D EX C H ILD ), did you have m orning sickness (nausea or vom iting)? 1 Y E S -------------- > H 8. In about what w eek o f pregnancy did it start? W EEK 98 . 5 N O (G O T O H 14) 99 8 ............ D K (G O T O H 14) 9 ............ R E F (G O T O H I 4 ) _______ H 9. D id you ju st have nausea or did you vom it? 1.................. N A U SE A O N LY (G O T O H l l ) 5 .................. V O M IT IN G (G O T O H 10) 8 .................. D K 9 .................. R E F ______ H 10. H ow m any w eeks did the vom iting last W EEKS 9 8 ...........D K 9 9 ............ R EF to 00 O s D K R E F Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. H i t . D id you take any m edication or receive any treatm ent for the m orning sickness? 1 Y E S ------------------------------- -— > H 12. W hat w as i t? ___________________________ (GO TO H 13) 5 .............. NO 8 D K H 13. W as it a prescription w ritten by your doctor and only 9 ..R E F available from a pharm acy? 1............. Yes 5............. No 8 ............. D K 9 ............. R E F _______ t o 00 -o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D uring your pregnancy w ith . . . did you have: IF Y ES: W hat treatm ent did you receive (describe) H14. Sw elling o f the hands or feet 1 Yes 5 N o 8 D K 9 ........R EF H I 5. Protein or album in in the urine 1 Yes 5 No 8 D K 9 ........REF H I 6. H igh blood pressure 1 Yes 5 N o 8 D K 9 ........REF D id you have this condition before your pregnancy? 1 Yes 5 N o 8 D K 9 ..........R E F H I 7. Pre-eclam psia or toxem ia 1 Y es 5 No 8 D K 9 ........REF 00 00 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D uring y o u r pregnancy w ith . . . did you have: IF YES: W hat treatm ent did you receive (describe) H I 8. H eart disease 1 ...........Yes D id you have this condition before your 5 ........... No pregnancy? 8 ...........D K 1............ Yes 9 ........... REF 5 ............ No 8 ............ D K 9 ............ REF H 19. B ladder or urinary tract infection 1 ........... Yes D id you have this condition before your (UTI) 5 ........... N o pregnancy? 8 ........... D K 1............ Yes 9 ........... R EF 5............ N o 8 ............ D K 9 ............ R EF H20. K idney disease or infection 1 ............Yes 5 ...........N o 8 ........... D K 9 ........... R EF 00 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D uring your pregnancy w ith . . . did you have: IF Y E S: W hat treatm ent did you receive (describe) H 21. A nem ia/iron deficiency 1 Yes 5 No 8 DK 9 ..........REF H22. Threatened m iscarriage 1 Yes 5 N o 8 D K 9 ..........R EF H23. V aginal bleeding or spotting 1 Yes 5 No 8 D K 9 ..........REF H24. Genital warts 1 Yes 5 No 8 D K 9 ..........REF D id you have this condition before your pregnancy? 1 Yes 5 N o 8 D K 9 ........REF Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. D uring your pregnancy w ith . . . did you have: IF Y E S: W hat treatm ent did you receive (describe) H25. Diabetes 1............ Yes D id you have this condition before your 5 ............ N o pregnancy? 8 ............ D K 1 ........... Yes 9 ............ REF 5 ........... N o 8 ...........D K 9 ........... R EF H26. A ny other illness 1............Yes 5 ............ N o 8 ............ D K 9 ............ R EF K > v > © Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section I M E D IC A T IO N S To the best o f your know ledge, did you take or receive any o f the follow ing m edications in the form o f an injection, ointm ent, or by m outh for in th e y e a r u p u n til y o u r p reg n an cy , d u rin g th e m o n th s b efo re you knew y o u w e re p re g n an t, d u rin g y o u r p reg n an cy w ith . . . ? (o r w hile b re a st feeding . . . ? ) T his w ould cover the tim e period b e tw ee n________ (m o/yr) a n d _________ (m o/yr). TY PE OF M ED IC A TIO N a. D uring that period did you take (M ED IC A TIO N )? b. W hat w as the nam e o f the drug? c. For w hat reason or condition w ere you taking the drug? d. W as this drug prescribed by a doctor? 11. Oral contraceptives yes....................... 1 n o .........................5 D K ...................... 8 R E F ....................9 yes........................1 n o .........................5 D K ...................... 8 R E F .....................9 12. V itam ins yes........................1 n o .........................5 D K ...................... 8 R E F ....................9 y es....................... 1 n o .........................5 D K ...................... 8 R E F ....................9 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. T Y PE O F M ED IC A TIO N a. D uring that period did you take (M ED IC A TIO N )? b. W hat was the nam e o f the drug? c. F or w hat reason or condition were you taking the drug? d. W as this drug prescribed by a doctor? 13. Iron supplem ents yes.......................1 n o .........................5 D K ......................8 R E F ....................9 y es........................1 n o .........................5 D K ...................... 8 R E F ..................... 9 14. R ecreational drugs such as m arijuana, heroin, cocaine or crack yes........................1 n o .........................5 D K ...................... 8 R E F ....................9 W W W \ X v \ \ X X A X X XXX X \ X XXX X XX XX X X X \ \ X X X \ X X \ \ XX \ x x w w x x x x x x X X X X X X X X X X \ X X X X X XXX \X X .X w x . x x x x w X X X \ X X X X X X X X X X X X X X X X X X X X X X x X X X . X X . X 15. Fem ale horm ones such as fertility drugs yes........................1 n o .........................5 D K ...................... 8 R E F ....................9 yes........................1 n o .........................5 D K ...................... 8 R E F .....................9 K > x D U > Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. c. W hen did you take the drug? 1 ear before ............................................... / Before rou knew vou were pregnant.......2 touring your prejfHuncv .............................3 Durinti breastfeeding....................................4 Don't know. ...............................................K Refuse.................................................................. f. H ow often did you take the drug and for how long? l)=duy M— month M =wrck } -year i t l i l i l l t t l l i l l l i l f l l l l i l l l 11. Yr. B efore................1 D uring breast feed..................... 4 B efore k n e w ...........2 D on’t k n o w ..................................8 D uring p reg................3 R efu se ........................................... 9 (D )(D) (W ) (W ) Tim es (M ) D uration (M ) (Y) (Y) 12. Yr. B efore................1 D uring breast feed......................4 B efore k n e w ...........2 D o n ’t k n o w ..................................8 D uring p reg ................3 R e fu se ........................................... 9 (D) (D) (W ) (W ) Tim es (M ) D uration (M ) (Y) (Y) K > VO Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. e. W hen did you lake the drug? t etir before. ............................................... / Before von knew you were pregnant...... 2 During vnur pregnancy .............................i D unn" hreastfivdinf!....................................4 Don V know. ............................................... 8 Refuse..................................................................V f. H ow often did you take the drug and for how long? S)--day M— monlh }V=week Y~veur 13. Yr. B efore................ 1 D uring breast feed......................4 B efore k n e w ........... 2 D o n ’t k n o w ..................................8 D uring p re g ................3 R e fu se ........................................... 9 (D) (D) (W ) (W ) Tim es (M ) D uration (M ) (Y ) (Y) 14. Yr. B efore................1 D uring breast feed......................4 B efore k n e w ........... 2 D o n ’t k n o w ..................................8 D uring p reg................3 R efu se............................................9 (D ) (D) (W ) (W ) Tim es (M ) D uration (M ) (Y) (Y) 15. Yr. B efore................1 D uring breast feed......................4 B efore k n e w ........... 2 D o n ’t k n o w ................................. 8 D uring p reg................ 3 R e fu se........................................... 9 (D ) (D) (W ) (W ) Tim es (M ) D uration (M ) (Y) (Y) t o VO Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. T Y PE O F M E D IC A TIO N a. D uring th at period did you take (M ED IC A TIO N )? b. W hat w as the nam e o f the drug? c. F or w hat reason or condition w ere you taking the drug? d. W as this drug prescribed by a doctor? 16. Laxatives or anti-diarrheal agents y es........................1 n o .........................5 D K ...................... 8 R E F ....................9 y es....................... 1 n o .........................5 D K ...................... 8 R E F .....................9 17. Im m unosuppressants or steroids such as cortisone, prednisone, 6- M P, im uran, cytoxan or any others? y es........................1 n o .........................5 D K ...................... 8 R E F ....................9 y es........................1 n o .........................5 D K ...................... 8 R E F .....................9 18. Insulin y es....................... 1 n o .........................5 D K ...................... 8 R E F ....................9 y es........................1 n o .........................5 D K ...................... 8 R E F ..................... 9 19. T hyroid m edications yes....................... 1 n o .........................5 D K ...................... 8 R E F ....................9 y es........................1 n o .........................5 D K ...................... 8 R E F ..................... 9 t o O N Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. T Y PE OF M ED IC A TIO N a. D uring that period did you take (M ED ICA TIO N )? b. W hat w as the nam e o f the drug? c. For w hat reason or condition were you taking the drug? d. W as this drug prescribed by a doctor? 110. Epilepsy m edications y es...................... 1 n o .........................5 D K ...................... 8 R E F ................... 9 y es....................... 1 n o .........................5 D K ...................... 8 R E F ..................... 9 111. A ny alternative, holistic or herbal m edications or treatm ents? y es....................... 1 n o .........................5 D K .....................8 R E F ....................9 yes........................1 n o ........................5 D K ...................... 8 R E F ..................... 9 112. A ny other m edications? SPEC IFY yes....................... 1 n o .........................5 D K ...................... 8 R E F ....................9 yes........................1 n o .........................5 D K ...................... 8 R E F..................... 9 t o < 1 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. e. W hen did you take the drug? f. H ow often did you take the drug and for how long? Year before ..........................................1 R ef ore you knew van were preunant.... 2 Durian you r p reg n a n cy...............................? During breastfeeding.................................... 4 Don '1 know. ............................................... 8 Refuse..................................................................9 D=day M=numth U'-week Y=yeur 16. Yr. B efore........... .... 1 D uring breast feed .......... ........4 (D) (D) B efore k n e w ..... ....2 D o n ’t k n o w ............................8 (W ) (W ) D uring p reg.......... ....3 R efu se........................... ........9 Tim es (M ) D uration (M ) (Y) (Y) 17. Yr. B efore......... ....1 D uring breast feed......... .........4 (D) (D) B efore knew...... ....2 D o n ’t k n o w ....... ............. .........8 (W ) (W ) D uring p re g .......... ....3 R efu se................................ .........9 Tim es (M ) D uration (M ) (Y) (Y) 18. Yr. B efore.......... .... 1 D uring breast feed ......... .........4 (D) (D) B efore k n e w ..... ....2 D o n ’t k n o w ...................... .........8 (W ) (W ) D uring preg........... ....3 R e fu se ............................... .........9 Tim es (M ) D uration (M ) (Y) (Y) 19. Yr. B efore.......... .... 1 D uring breast feed......... ........4 (D) (D) B efore k n e w .... ....2 D o n ’t k n o w .................. ........8 (W ) . (W ) D uring p reg......... ....3 R e fu se........................... ........9 Tim es (M ) D uration (M ) (Y) (Y) 00 w > 3 £ £ o -c •a •4 - » 3 -a < t s o £ o E b f l 3 j-« ' T3 t o rg 3 O !2 * o 1 1 j r * § 1 !i ^ V > : > ll‘ l.i’ 5. a i£ . s < • s « £ j ? « a a J P S * £ > < 3 .2 te 3 Q q £ t P B H 3 > £ 2 o 4 3 J tUD-* 3 2 . . 3 3 O _ Q O P 2 "b S . S . £ a £ ? £ •* s 0 ) 1 ) ^ m t W ) . = s s < to * _ « r q p Q U h > * I Q £ S P 3 O 3 Q a f — 1 T 3 t o f t o < H H 3 S t S 5 •° 5 M > * a ^ 'S e ^ C Q Q £ S P 3 O Q £ S P H •3 ^ g S ^ . 3 O t o Q Q c d to to (3 £ , fcM " tp t o t o ® < 2 . tf — I t o o - W ) ^ c — to 'C >< C Q 3. Q ( N 299 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Mother's Diet During Pregnancy During your pregnancy with__________ , how often did you eat or drink the following foods or beverages? A=never, B=< 1 month, C=1-3X/month, D=1-3X/week, E=4-6X/week, F=daily or more FOOD Specify: A B C D E F 113. Ice Cream 114. Margarine 115. Butter 116. Yogurt 117. Cheese 118. Eggs 119. Dried legumes (beans, lentils etc) I20. Fish 1 2 1 .Fresh Vegetables I22. Canned or Frozen Vegetables I23. Fresh Fruit I24. Milk I25. Poultry (chicken, turkey, etc) I26. Red Meats: A) LIST B) LIST C) OTHER Specify U > o o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. FOOD Specify: A B C D E F 127. Soy Products (SEE LIST ON CARD) 128. Soft Drinks 129. Coffee: A) Regular B) Decaf 130. Tea A) Black B) Green C) Herbal 131. Cocoa as a beverage I32. Beer I33. Wine I34. Liquor u > o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. X -R A Y S/R A D IA T IO N N E X T I W ILL B E A SK IN G Y O U A B O U T SPEC IA L TESTS A N D X -R A Y S Y O U M A Y H A V E HA D FR O M TW O Y EA R S B E FO R E Y O U R PR E G N A N C Y W ITH ... U P TO TH E TIM E OF H IS/H ER BIRTH. 135. First, did you have any ultrasound or sonogram ......... Y es...l exam inations during this pregnancy (these are usually ....... N o....5 (138) done w ith a hand-held probe m oved up and dow n the abdom en, D K ....8 (1 3 8 )___ and they show a picture o f the baby on th e screen)? 136. H ow m any tim es did you have an ultrasound during this Tim es pregnancy? 137. D uring w hich m onths w ere the ultrasounds done?................... ...... T1 ...3___ ___ T 2...4_ _ ___ T3...5______ 138. D id you have an am niocentesis during this pregnancy Y es...l (this is usually done w ith a needle inserted through N o 5 your abdom en into your wom b to get som e fluid)? DK 8____ u > o K ) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. DID Y O U H A V E A N Y X -R A Y S D U R IN G Y O U R PR E G N A N C Y W ITH (IN D EX C H ILD )? These w ould include any x-rays except dental. 139. 1........ Y es--------------------------------> 140. W hat w as that? (D escribe reasons and part o f body receiv ed ) (if m ore than one type use 143) 5 N o (go to 152) _____________________________________ 8 D K _______________________________________ 9 R EF 141. W hat type o f x-rays did you receive? 1 ........................... x-ray film s (141) 2...........................fluoroscopy 3...........................scan 4...........................other 142. H ow m any tim es did you have these x-rays? (From 140) (142) tim es U ) o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 143. W hat w as that (D escribe reasons and part o f body received) (if m ore than this, use 146) 144. W hat type o f x-rays did you receive? 1 2 3 4 145. H ow m any tim es did you have these x-rays? (From 143) ________________ (145) tim es .x-ray film s .fluoroscopy .scan .other (144) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 146. W hat w as that (D escribe reasons and part o f body received) (if m ore than this, use 149) 147. W hat type o f x-rays did you receive? 1 ........................... x-ray film s 2...........................fluoroscopy 3...........................scan 4...........................other 148. H ow m any tim es did you have these x-rays? (From 146) (148) tim es u > o (147) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 149. W hat w as that (D escribe reasons and part o f body received) 150. W hat type o f x-rays did you receive? 1 ........................... x-ray film s 2...........................fluoroscopy 3.................. ........scan 4...........................other 151. H ow m any tim es did you have these x-rays? (From 149) (151) tim es U J o (150) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. DID Y O U H A V E A N Y X -R A Y S FIV E Y EA R S B E FO R E Y O U R PR E G N A N C Y U P UN TIL Y O U R PR E G N A N C Y W IT H .............. ? These w ould include any x-rays except dental. 152. 1........ Y es------------------------------- > 153. W hat w as that? (D escribe reasons for x-ray, w hat part o f body received x-ray) (if m ore than one type Use 157) 5 N o (go to 169) _____________________________________ 8 D K • _______________________ 9 R E F ______ _______________________________________ I54-. W hen w ere the x-rays taken? 1 .......................... (1-5 Y ears before) _______ (154) 2.......................... (<1 year before) 3.......................... (m onth before) 155. W hat type o f x-rays did you receive? 1 ...........................x-ray film s (155) 2...........................fluoroscopy 3...........................scan 4...........................other 156. H ow m any tim es did you have these x-rays? (From 153) (156) tim es o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 157. W hat w as th at (D escribe reasons for X-ray, w hat part o f body received) (if m ore than this, use 161) 158. W hen w ere the x-rays taken? 1 ............... (1-5 years before) _______ (158) 2 ............... (<1 year before) 3............... (m onth before) 159. W hat type o f x-rays did you receive? 1 2 3 4 160. H ow m any tim es did you have these x-rays? (From 157) __________________ (160) tim es .x-ray films .fluoroscopy .scan other (159) u > o o o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 161. W hat w as th at (D escribe reasons and body o f part received) (if m ore than this, use 165) 162. W hen w ere the x-rays taken? 1 ............ (1-5 years before) _______ (162) 2 ............ (<1 year before) 3 ............ (m onth before) 163. W hat type o f x-rays did you receive? 1 ........................... x-ray films 2 ...........................fluoroscopy 3 ...........................scan 4 ...........................other 164. H ow m any tim es did you have these x-rays? (From 161) (164) tim es (163) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 165. W hat w as that (D escribe reasons and part o f body received) 166. W hen w ere the x-rays taken? 1 .............(1-5 years before) _______(166) 2 ............ (<1 year before) 3 .............(m onth before) 167. W hat type o f x-rays did you receive? 1 ........................... x-ray films 2 ........................... fluoroscopy 3 ........................... scan 4 ........................... other 168. H ow m any tim es did you have these x-rays? (From 165) (168) tim es U J I —* o (167) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. TH E FO LLO W IN G Q U ESTIO N S A R E A B O U T C ER TA IN FO R M S O F R A D IA TIO N SU C H A S SCANS, FL U O R O SC O PY AN D RA D IA TIO N TREA TM EN T. D U R IN G TH E TIM E PER IO D FR O M 5 Y EA R S B E FO R E YO UR PR EG N A N C Y TO T H E T IM E ... WAS BO RN , D ID Y O U H A V E ANY OF TH E FO LLO W IN G ? 169. Thyroid, liver, brain, lung, or pancreatic Y es ---------------- 1----------------------------- > 170. Describe scan or fluoroscopy, or any other N o 5 (go to 173)_________________________ radiation treatm ent involving any other D K ............................ 8 ______________________ part o f you body? R E F ............................ 9____ 171. W hen w ere the treatm ents done? 1...... (1-5 years before) ______ 2 ....(< 1 year before) 3 ... .(m onth before) 1... .(before you knew you were pregnant) 8 ....(w hile you w ere pregnant) 172. H ow m any tim es did you have these treatm ents? ___________________ tim es Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. DID Y O U R C H ILD (IN D EX CHILD) H A V E A N Y X -R A Y S ? These w ould include any x-rays. 173. 1........Y es------------ > 174. W hat was that? (D escribe reasons and part o f body re ce iv e d ) (if m ore than one type use 178) 5 N o (GO TO SECTIO N J)_____________ ______________________________________ 8 D K ________________________________________________________________________ 9 R EF 175. W hat type o f x-rays did you receive? 1 ........................... x-ray films 2.......................... fluoroscopy 3.......................... scan 4...........................other 176. H ow m any tim es did you have these x-rays? (From 174) (176) tim es 177. W hat the age did the child have these x-rays? (From 174) (177) tim es w to (175) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 178. W hat was that? (D escribe reasons and part o f body received ) 179. W hat type o f x-rays did you receive? 1 2 3 4 180. H ow m any tim es did you have these x-rays? (From 178) ______________ (180) tim es 181. W hat the age did the child have these x-rays? (From 178) __________________ (181) tim es .x-ray film s (179) .fluoroscopy .scan .other u > U ) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. Section J O CCU PA TIO N: J -l) W hat was your jo b title and describe your duties? J-2) O n average, how m any hours per w eek did you work? J-3) W hat w ere the dates (yrs) w orked? (Inclusive) #a From hours per w eek Thru #b From hours per w eek Thru #c From hours per w eek Thru #d From hours per w eek Thru #e From hours per w eek Thru Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. J -l) W hat was your jo b title and describe your duties? J-2) O n average, how m any hours per w eek did you w ork? J-3) W hat were the dates (yrs) w orked? (Inclusive) # f From hours per week Thru #e From hours per w eek Thru #h From hours per week Thru #i From hours per week Thru U > L f\ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. J-4) In this jo b were you exposed to any chem icals, dusts, fum es, oils or radiation? J-5) D id you w ear any protective clothing or a m ask? J-6) D escribe w hat protective clothing, etc. was w orn. JOBS # a Y e s................. 1 N o ...................5 (next jo b ) D K ................. 8 (next job) Y e s ..................... 1 N o ...................... 5(Go to J7) D K ...................... 8 (Go to J7) #b Y e s................. 1 N o .................. 5 (next jo b ) D K ................. 8 (next job) Y e s .....................1 N o ...................... 5(Go to J7) D K ...................... 8 (Go to J7) #c Y e s................. 1 N o ...................5 (next jo b ) D K ................. 8 (next job) Y e s .....................1 N o ...................... 5 (Go to J7) D K ...................... 8 (Go to J7) #d Y e s................. 1 N o .................. 5 (next jo b ) D K ............ ....8 (next job) Y e s .....................1 N o ...................... 5 (Go to J7) D K ...................... 8 (Go to J7) #e Y e s................. 1 N o ...................5 (next jo b ) D K ................. 8 (next jo b ) Y e s ..................... 1 N o ...................... 5 (Go to J7) D K .,....................8 (Go to J7) U ) 0 > Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. J-4) In this jo b were you exposed to any chem icals, dusts, fumes, oils or radiation? J-5) D id you w ear any protective clothing or a mask? J-6) D escribe w hat protective clothing, etc. was worn. # f Y e s................. 1 N o ...................5 (next jo b ) D K ................. 8 (next jo b ) Y e s ..........:......... 1 N o ...................... 5 (Go to J7) D K ...................... 8 (G o to J7) # g Y es................. 1 N o ...................5 (next jo b ) D K ................. 8 (next jo b ) Y e s ..................... 1 N o ...................... 5 (Go to J7) D K ...................... 8 (Go to J7) #h Y e s................. 1 N o ...................5 (next jo b ) D K ................. 8 (next job) Y e s ..................... 1 N o ...................... 5 (Go to J7) D K ...................... 8 (G o to J7) #i Y es................. 1 N o ...................5 (next jo b ) D K ................. 8 (next jo b ) Y e s ..................... 1 N o ......................5 (Go to J7) D K ......................8 (Go to J7) U ) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. J-7 W ere radiation or radiaoactive m aterials used by you or near where you w orked? J-8 W hat type o f radiation was it? J-9 D id you w ear a radiation badge? JOBS # a Y e s............ 1 Y e s ...................1 N o ..............5 (next jo b ) N o ........................5 D K ............ 8 (next jo b ) D K ...................... 8 #b Y es............ 1 N o ..............5 (next jo b ) Y e s .....................1 D K ............ 8 (next job) N o ........................5 D K ...................... 8 #c Y es............ 1 N o ..............5 (next jo b ) Y e s .....................1 D K ............ 8 (next jo b ) N o ........................5 D K ...................... 8 # d Y e s............ 1 N o ..............5 (next jo b ) Y e s .....................1 D K ............ 8 (next jo b ) N o ........................5 D K ...................... 8 #e Y e s............ 1 N o ..............5 (next jo b ) Y e s .....................1 D K .............8 (next jo b ) N o ........................5 D K ...................... 8 0 0 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. J-7 W ere radiation or radiaoactive m aterials used by you or near where you w orked? J-8 W hat type o f radiation was it? J-9 D id you w ear a radiation badge? # f Y e s............ 1 N o ..............5 (next jo b ) Y e s ..................... 1 D K ...........8 (next job) N o ........................5 D K ...................... 8 #g Y es............1 N o ..............5 (next jo b ) Y e s .....................1 D K ............ 8 (next job) N o ....................... 5 D K ...................... 8 # h Y es............ 1 N o ..............5 (next jo b ) Y e s .....................1 D K ............ 8 (next job) N o ........................5 D K ...................... 8 #i Y e s............ 1 N o ..............5 (next jo b ) Y e s .....................1 D K ............ 8 (next job) N o ........................5 D K ...................... 8 U > 'sO Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. U SE E X T R A SH E E T S FO R A D D IT IO N A L JO B S. A F T E R L A ST JO B G O TO SE C T IO N K. Section K H A B IT S A N D H O U SE H O L D E X PO SU R E S K l. D o you currently drink wine, beer, liquor, or any other form o f alcohol? Y e s............... 1 N o ...............5 D K ...............8 R E F 9______ K2. D id you ever drink at least 2 drinks per m onth for a period o f one year or longer? Y e s ............. 1 N o ...............5 (go to K 19) D K ...............8 R E F ............ 9 u > O Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. K3. D uring the m onth prior to your pregnancy w ith __________ did you drink any alcohol? Y e s .............1 N o .............5 (go to K 7) D K .............8 R E F .......... 9 _ _____ K4. D id you drink w ine? Y e s ..........1 --------------> describe ty p e _________________ (red or w hite) N o ............5 (go to K 5) D K ...........8 R E F 9______ H ow m uch did you drink per day, week, or m o n th ? p e r ________(4 oz glasses) K5. D id you drink beer? Y e s ...............1 N o ..............5 (go to K 6) D K ..............8 R E F ........... 9______ H ow m uch did you drink p er day, week, or m o n th ? p e r (12 oz can) K6. D id you drink liquor? Y e s ..........1 --------------> describe ty p e _____________ N o .5 (go to K 7) D K ...........8 R E F .. 9______ H ow m uch did you drink per day, week, or m o n th ? p e r (1 s h o t/1 'A oz) CO K ) Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. K7. W hile you w ere pregnant b u t before you knew you w ere pregnant w ith ___ did you drink any alcohol? Y e s ............ 1 (go to K 8) N o ............. 5 (go to K l l ) D K .............. 8 R E F ........... 9 K8. D id you drink w ine? Y e s ........... 1 ------------- -> describe type N o .............. 5 (go to K 9) D K .............. 8 R E F...............9 H ow m uch did you drink per day, week, or m onth? per K 9 . D id you drink beer? Y e s ......... 1 N o .... .........5 (go to K 10) DK.... .........8 R E F.. .........9 H ow m uch did you drink per day, week, or m onth? per K10. D id you drink liquor? Y e s ......... 1 ----------— > describe type N o .5 (go to K l l ) D K .............8 R E F .......... 9 ____ H ow m ueh did you drink per day, week, or m o n th ? p e r ____ (red or w hite) (4 oz glass) (12 oz can) (1 shot/1 ‘ /2 oz) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. K 1 1. O nce you knew you w ere pregnant, any alcohol? Y e s ... ......... 1 N o ..... ............ 5 (go to K 15) D K .... ............ 8 REF ............ 9 K12. D id you drink wine? Y e s ... ............ 1 ---------- > describe type N o ..... ............ 5 (go to K 13) D K .... ............ 8 R E F .. ...................9 H ow m uch did you drink per day, week, or m onth? per (4 oz glass) K 13. D id you drink beer? Y e s ... ...................1 N o ........ ............ 5 (go to K 14) D K .... ............ 8 R E F .. ............ 9 H ow m uch did you drink per day, week, or m onth? per (12 oz can) K 14. D id you drink liquor? Y e s ............ N o .............. D K ............ R E F .......... H ow m uch did you drink per day, week, or m onth? 1— --------- > describe ty p e _____ 5 (go to K 15) 8 9 p e r (1 shot/1 Vi oz) (red or w hite) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. K 15. D uring the tim e you breastfed Y e s ..... ...........1 (go to K 16) did you drink any alcohol? N o ..................5 (go to K 19) D K ..... ...........8 R E F ... ...........9 K 16. D id you drink wine? Y e s............. ....1 --------> describe type N o ...................5 (go to K 17) D K .................8 R E F ............ ....9 H ow m uch did you drink per day, week, or m onth? per (4 oz glass) K 17. D id you drink beer? Y e s............. ....1 N o .............. ....5 (go to K 18) D K ............ ....8 R E F ........... ....9 H ow m uch did you drink per day, week, or m onth? per (1 2 o z c a n ) K18. D id you drink liquor? Y e s............ .... 1-------------- > describe type N o .............. ....5 (go to K 19) D K ............ ....8 R EF 9 H ow m uch did you drink per day, week, or m onth? per (1 shot/1 V i U ) to 4 ^ (red or white) Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. S M O K IN G H IS T O R Y : K19. H ave you ever sm oked even one cigarette? 1 ....................... YES 5 ....................... N O (go to K 32) 8....................... DK 9....................... REF K20. A t any tim e in your life, did you sm oke at least one cigarette per day for a period o f three m onths or longer? 1 YES — > K 21. H ow old w ere you when you first began to sm oke regularly? years 5 .............N O (go to K 32) 8.............DK 9.............REF K22. Do you sm oke at present? 1 ............... YES — >H ow m any cigarettes a day do you currently smoke? 5 ................N O -------> H ow old were you w hen you sto p p ed ? Y ears 8................D K 9................R EF K 23. O ver your entire lifetim e, how m any years have you sm oked? Y ears 8...................DK 9...................R E F _ _____ K 24. D id you sm oke at all in the year before you becam e pregnant w ith (IN D EX CHILD)? 1 ................. Y E S _______(go to below , K 25) 5 ...................N O (go to K 26) 8...................D K 9................ ...REF K25. H ow m any cigarettes a day did you smoke? cigarettes per day cigarettes per day u > to C /l Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. K26. W hile you w ere pregnant but before you knew you w ere pregnant w ith , did you sm oke 1 .........YES (go to below , K 27) 5 .........N O (go to K 28) 8 .........D K 9 .........REF K 27. H ow m any cigarettes a day did you sm oke? cigarettes per day K28. D id you sm oke during your pregnancy w ith (IN D EX C H ILD )? 1 ................Y E S _______ (go to below , K 29) 5 ................N O (go to K 30) 8 ................D K 9 . ...............REF_______ K29. H ow m any cigarettes a day did you sm oke? cigarettes per day K30. D id you sm oke w hile b reastfeed in g ___________(IN D EX CHILD)? 1 .............. ...Y E S ------------> K 31. H ow m any cigarettes a day did you sm oke? cigarettes p er dav 5 ..................NO 8 ..D K 9..................R EF u > t s > o\ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. a. In the m onth before the pregnancy, during the pregnancy, or w hile breastfeeding, did you use any products used to control: b. H ow frequently did you use ? c. W hen did you use ? RflAl) C liO H IV K32. H O U SEH O LD IN SECTS such as R aid, B lack Flag bug spray, Ortho H ornet and W asp K iller, no-pest strips, ant traps, or roach baits? Y es 1 N o 5 (K32) D K 8 (K32) tim es per D a y ................ D W eek ...............W M o n th ............ M Y e a r...............Y M o. B e fo re ................. 2 D u rin g ..........................3 B rstfeed in g................. 4 D K ................................8 K 33. M oths such as m othballs? Y es 1 N o 5 (K 33) D K 8 (K 33) tim es per D a y ................. D W eek...............W M o n th ............ M Y e a r................ Y M o. b e fo re .................. 2 D u rin g ..........................3 B rstfeed in g ................. 4 D K ................................. 8 U J < 1 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. a. In the m onth before the pregnancy, during the pregnancy, or w hile breastfeeding, did you use any products used to control: b. H ow frequently did you use ? c. W hen did you use ? READ CHOICES K34. M ice, rats, gophers or m oles such as D -C on or W arfarin? Y e s...,......1 N o 5 (K 34) D K 8 (K 34) tim es per D a y ................. D W eek...............W M o n th ............ M Y e a r................ Y M o. b e fo re ...................2 D u rin g ..........................3 B rstfeed in g................. 4 D K .................................8 K35. Fleas or ticks on pets such as R aid foggers, H artz flea spray or flea collars? Yes 1 N o 5 (K35) D K 8 (K 35) tim es per D a y ................. D W eek ...............W M o n th ............ M Y e a r................ Y M o. b e fo re ...................2 D u rin g ..........................3 B rstfeed in g ................. 4 D K ................................. 8 U > 0 0 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. a. In the m onth before the pregnancy, during the pregnancy, or w hile breastfeeding, did you use any products used to control: b. H ow frequently did you use ? c. W hen did you use ? READ CHOICES K36. O utdoor hom e garden products such as dandelion killers or crabgrass killers, plant/tree insects or diseases, slug and snail bait? Y es 1 N o 5 (K36) D K 8 (K36) tim es per D a y ..................D W eek...............W M o n th ............ M Y e a r................ Y M o. b e fo re ...................2 D u rin g .......................... 3 B rstfeed in g ................. 4 D K ..................................8 K37. Pesticides used to spray or treat farm crops, or pesticides applied by a law n service? Y es 1 N o 5 (K37) D K 8 (K37) tim es per D a y ................. D W eek ...............W M o n th ............ M Y e a r................ Y M o. b e fo re....... ...........2 D u rin g .......................... 3 B rstfeed in g ................. 4 D K ..................................8 Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. a. In the m onth before the pregnancy, during the pregnancy, or w hile breastfeeding, did you use any products used to control: b. H ow frequently did you use ? c. W hen did you use ? READ CHOICES K 38. Insect repellents: such as OFF, B ackw oods Cutter, N atrapel or B ugout? Y es 1 N o 5 (K 38) D K 8 (K 38) tim es per D a y ................. D W eek ...............W M o n th .............M Y e a r................ Y M o. b e fo re ...................2 D u rin g .......................... 3 B rstfeed in g ................. 4 D K ..................................8 K39. W as you house treated by exterm inators? Y es 1 N o 5 (K 39) D K 8 (K 39) tim es per D a y ................. D W eek...............W M o n th .............M Y e a r................ Y Mo. b e fo re ...................2 D u rin g ..........................3 B rstfeed in g ................. 4 D K ..................................8 O J U > o Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. a. In the m onth before the pregnancy, during the pregnancy, or w hile breastfeeding, did you use any products used to control: b. H ow frequently did you use ? c. W hen did you use ? READ CHOICES K40. D id you com e in contact w ith any paints, stains or lacquers? Y es 1 N o 5 (K40) D K 8 (K40) tim es per D a y ................. D W eek...............W M o n th ............ M Y e a r................Y M o. b e fo re ...................2 D u rin g .......................... 3 B rstfeed in g ................. 4 D K ................................. 8 K 4 1 . D id you come in contact w ith petroleum products such as gasoline, kerosene, lubricating oils, or spot rem overs? DO N O T IN C LU D E PU M PIN G GAS Yes 1 N o 5 (K41) D K 8 (K 41) tim es per D a y ................. D W eek...............W M o n th ............ M Y e a r................Y M o. b e fo re...... ............2 D u rin g .......................... 3 B rstfeed in g ................. 4 D K ....... :........................8 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. a. In the m onth before the pregnancy, during the pregnancy, or w hile breastfeeding, did you use any products used to control: b. H ow frequently did you use ? c. W hen did you use ? READ CHOICES K 42. D id you use any hair dyes or tints, or perm anents or relaxers? Y es 1 N o 5 (K 42) D K 8 (K 42) tim es per D a y ................. D W eek ...............W M o n th ............ M Y e a r................ Y M o. b e fo re ...................2 D u rin g ..........................3 B rstfeed in g ..................4 D K .................................8 D etails (list by question num ber): U > t o Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Section O B A C K G R O U N D IN FO R M A T IO N N O W I W O U LD LIK E TO A SK Y O U SO M E Q U ESTIO N S A B O U T Y O U R B A CK GROU ND. O l. T o w hich o f the follow ing groups does your mother 0 2 . To w hich o f the follow ing groups does your father belong? belong? W hite, (not o f H ispanic origin)....,.................................. 1 W hite, (not o fH isp an ic o rig in )........................................ 1 Black, (not o fH isp an ic o rig in )........................................2 Black, (not ofH isp an ic o rig in )........................................2 H ispanic (include M exican A m erican )......................... 3 H ispanic (include M exican A m e ric an )......................... 3 A m erican Indian or A laska n a tiv e ................................. 4 A m erican Indian or A laska n ative...................................4 A sian or Pacific Islander.................................................... 5 A sian or Pacific Islander.................................................... 5 O ther - specify_____________________ 6 ______ O ther - specify_____________________ 6 ______ GO TO 0 2 G O TO 0 3 0 3 . W hat is your date o f birth?______________ ______/ /______ M O DA Y Y R 0 4 . W hat is the religion in w hich you w ere raised? (DO N O T R EA D C A TEG O R IES TO R ESPO N D EN T) N o n e .............................. ...............00 P rotestant (no preferen ce)................................07 C a th o lic ....................... ...............01 Jew ish...................... .................................... ..........08 L u th e ra n ...................... ...............02 G reek O rthodox......................................... ..........09 B a p tist.......................... ...............03 C ongregationalist..................................... ..........10 P resbyterian............... ..............04 M orm on............ ........................................... ..........11 E piscopalian............... ............... 05 Seventh D ay A dventist........................... ..........12 M ethodist..................... ...............06 O ther (specify) 13 U i u > Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. 0 5 . W hat is the highest grade or level o f schooling that you have com pleted? Less than 8 y e a rs .............................................................. 1 8 through 11 y e a r s ............................................................................................................ 2 12 years or com pleted high sc h o o l.............................................................................. 3 P ost high school training other th an college...................................................... 4 Som e c o lleg e ................................. 5 C ollege graduate............................................................ 6 Post graduate le v e l............................................ 7 0 6 . A re you currently: READ LIST M arried D ivorced Separated W idow ed Single, never m arried ....................5 L iving as m arried .....................................6 O th e r............................................................ 7 D K ................................................................. 8 0 7 . A re you and . . . 's father related by blood (such as second cousins) Y es.............. 1 N o ...................5 IF Y ES, S p ecify _________________________ _ _ _______________________ D K .............. 8 0 8 . W ould you say the area you lived in last year was primarily: U rban........................................1 S u b u rb a n ............................... 2 R u ra l........................................3 F a rm ........................................4 0 9 . H ow m any different residential telephone num bers are presently in your hom e? W e are interested in separate telephone num bers, not several phones w ith the sam e telephone num ber. u > - t- Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. 0 1 0 . Since ... was bom , how m any residences have you lived in for 1 continuous year or longer? _____________________________ num ber o f residences a. A t the tim e o f your pregnancy with . . . B. D uring (REF Y EA R ) Oil. Single fam ily residence................................ ....1 Single fam ily residence.......................... ....... 1 W hat type o f housing M obile h o m e ................................................... .....2 M obile h o m e.............................................. ........... 2 did you live in? D uplex-Q uadruplex (2-4 u n its )................ .........3 D uplex-Q uadruplex (2-4 u n its) ................... ........... 3 A partm ent w ith 5-10 u n its ............................................ .........4 A partm ent w ith 5-10 u n its .................................. ...................4 A partm ent w ith 11+ u n its .............................................. .....5 A partm ent w ith 11+ u n its .................................... ...................5 O ther (D ESC R IB E) .............. .....6 O ther (D ESC R IB E) ........... 6 D K ............................................................. .....8 D K ........................................................ ........... 8 R E F ............................................................. 9 R E F ........................................................ 9 u > U\ Reproduced w ith perm ission o f th e copyright owner. Further reproduction prohibited without permission. a.W as your total household income: b. H ow m any people w ere supported by this incom e 0 1 2 . In the year . . . w as bora? Less than $10,000 0 $10,000 to $19,999 1 $20,000 to $29,999 2 $30,000 to $39,999 3 $40,000 to $49,999 4 $50,000 to 75,000 5 M ore than $75,000 6 num ber supported 0 1 3 . Last year? Less than $10,000 0 $10,000 to $19,999 1 $20,000 to $29,999 2 $30,000 to $39,999 3 $40,000 to $49,999 4 $50,000 to $75,000 5 M ore than $75,000 6 num ber supported U > u > O n A ppendix B. Detailed B udget for Initial B udget Period o f the Proposed Study PER SO N N N EL N A M E R O LE ON PR O JE C T T Y P E A P P T . % E FFO R T SA LA RY FRIN GE* TOTALS (m onth) O N PRO J. R EQ U E ST E D Liu-M ares Principal Investigator 1 2 100% (40 hr / wk) 50,000 13,800 63,800 TBN Interview er 12 100% (40 h r / wk) Subtotals Fringe is based on 27.6% o f requested salary 20,800 5741 70,800 19,541 26,541 90,341 SU PPLIES Storage o f blood sam ples/shipm ent to lab/equipm ent and m aterials C om puter labels, photo copies Participant files folders Subtotals 300x250= 75,000 200 200 75,400 O TH ER E X PE N SES Long D istance Call 7,500 (3*125=375 hours for controls and 1*125=125 hours for case, 1*125=125 hours for scheduling blood drawn, contacting physicians at $0.2/m inutes) Postage 0.33*250*2=165 Subtotals SU BTOTAL D IR E C T C O ST FO R IN ITIA L B U D G ET PER IO D 7,665 173,406 TOTAL D IR E C T C O ST FO R IN ITIA L B U D G ET PER IO D 173,406 337 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. B U D G ET JU STIFIC A TIO N P erso n n el: W en Liu-M ares. M .D .. Ph.D .. Principal Investigator (100% Effort) w ill have the prim ary responsibility o f overseeing the overall organization and conduct o f th e study. Som e o f her specific responsibilities will include design, conduct, screen m utations, analysis and reporting o f the study, m aintaining an effective level o f com m unication betw een study participants, supervising and involving data collection, subm ission, data screening, entry and preparation o f progress reports. Interview er fTBNJ (100% Effort! will identify eligible controls by random digit dialing and schedule, conduct a telephone interview for both cases and controls. She/H e w ill also schedule to visit controls to get blood drawn or to arrange controls to visit his/her pediatric clinic or the nearest C C G institutions to have blood drawn. Institutional R esource and Environm ental T he study w ill be carried out at CC G institutions and CC G Operation C enter. Drs. Jonathan Buckley, Sm ita B hatia and M ark K railo are very interested in the study and will be available for consultation. In addition, the follow ing faculty m em bers have given m e the feedback and suggestions in the past and are likely to be available for consultation on issues involving m ethodology, data collection, analysis, m utation detection: Leslie R obison, Ph.D. : Chairm an, CCG E pidem iology Steering Com m ittee. Professor. D ivision o f Pediatric Epidem iology and C linical R esearch, U niversity o f M innesota. X iao Ou Shu, M .D ., Ph.D .: A ssociate Professor. School o f M edicine. U niversity o f V anderbilt Julie A Ross, Ph.D . : V ice Chairw om an, CCG Epidem iology Steering Com m ittee. A ssociate Professor. D ivision o f Pediatric E pidem iology and Clinical R esearch, U niversity o f M innesota. H arland Sather, Ph.D .: A ssociate Professor. D epartm ent o f Preventive M edicine. U niversity o f Southern California. D uncan T hom as, Ph.D .: Professor and D irector, B iostatistics D ivision. D epartm ent o f Preventive M edicine. U niversity o f Southern C alifornia R obert H aile, D r.P.H .: Professor, G enetic Epidem iology. D epartm ent o f Preventive M edicine. U niversity o f Southern C alifornia R ichard G atti, M .D .: Professor, D epartm ent o f Pathology, School o f M edicine, U niversity o f California, Los A ngeles O H T H E R E X PE N SE N SE S Supplies: O ne-year funds ($400) are requested for project letterhead stationary, envelopes, and labels needed for m ailing o f introductory letters, file folders, transcripts, and needed for m aintenance o f participant files. O ther Expenses L ong distance charges for identifying an eligible m atched control by random digit dialing (estim ated 100 calls w ith an average 1 m inute per call=1.6-2 hours per control) and 1 hour each for interview ing parents/guardians o f cases (N =125) and controls (N =125) and 1 hour to schedule having control blood draw n are requested based on $0.2/m inutes rate. Postage includes questionnaire, consent form s send and return, introductory letters, letters for clarifying collected data. 338 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Asset Metadata

Creator Liu-Mares, Wen (author)

Core Title Effect of genetic factors in the development of childhood lymphocytic leukemia (ALL)

Contributor Digitized by ProQuest (provenance)

School Graduate School

Degree Doctor of Philosophy

Degree Program Epidemiology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biology, molecular,health sciences, public health,OAI-PMH Harvest

Language English

Advisor Buckley, Jonathan D. (committee chair), Bhatia, Smita (committee member), Krailo, Mark (committee member), Reichardt, Juergen (committee member), Thomas, Duncan C. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-101664

Unique identifier UC11337794

Identifier 3027741.pdf (filename),usctheses-c16-101664 (legacy record id)

Legacy Identifier 3027741.pdf

Dmrecord 101664

Document Type Dissertation

Rights Liu-Mares, Wen

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA