University of Southern California Dissertations and Theses

Does young adult Hodgkin's disease cluster by school, residence and age?

(USC Thesis Other)

Does young adult Hodgkin's disease cluster by school, residence and age?

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zed) Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DOES YOUNG ADULT HODGKIN’S DISEASE CLUSTER BY SCHOOL, RESIDENCE AND AGE? by Khandaker Talat Shamsul Islam A Thesis Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment O f The Requirements For The Degree MASTER OF SCIENCE (APPLIED BIOMETRY AND EPIDEMIOLOGY) December 1998 Copyright 1998 K.Talat S.Islam Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI N u m b er: 1 3 9 4 7 6 7 UMI Microform 1394767 Copyright 1999, by UMI Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Arbor, MI 48103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. U NIV ERSITY O F S O U T H E R N CALIFO RNIA T H E G R A D U A T E S C H O O L U N IV E R S IT Y P A R K L O S A N G E L E S . C A L IF O R N IA 8 0 0 0 7 This thesis, written by /C H A rbM ££ r G / f # **£ C /£ . X & A _________ under the direction of ki~*. Thesis Committee, and approved by all its members, has been pre sented to and accepted by the Dean of The Graduate School, in partial fulfillment of the requirements fo r the degree of Jtta.s.tL£r-„3£~S.c.ienc_e__. Dram / ' Date 8 THESIS COMMITTEE / Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For my family Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents: I. Introduction Page 1 n. Methods Page 6 in. Results Page 15 IV. Discussion Page 19 V. Bibliography Page 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Table 1. Table! Demographics Tablela. Distribution of Cases on the Basis of : 10 Ethnicity Tablelb. Gender Distribution of Sub-types o f : 11 Hodgkin’s Disease Table lc. Age Distribution of Cases : 11 Table Id. Distribution of Cases on the Basis : 12 of Religion. Cell Counts And Relative Risk Of HD Among Schoolmates Table2a. For all zip codes :16 Table2b. For zip codes with single school : 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List o f Figures: Figure 1. Figure 2. Figure 3. Histogram showing the age distribution of Cases of HD in Los Angeles County age variability of HD from 1968-L997. Variability of Age in Both Sexes is Compared by Box-Plot Box plots showing age variability in HD sub-types. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT The hypothesis of common childhood exposure for Hodgkin’s Disease (HD) as suggested by previous studies was tested in this study by detecting school clusters. The cases consisting of Los Angeles County residents, diagnosed with HD within 1968-1997, ascertained from the University of Southern California Cancer Surveillance Program population-based cancer registry. Data were obtained from the USC-CSP database Each case was paired with all other cases of similar subtype and then paired data were tested by chi-square test .We found that cases of HD of similar age were more likely to be at same school than expected by chance among subtypes Lymphocytic Predominance (OR=1.6, p=0.00005) and HD-not otherwise specified (1.28,0.0023). A weaker association was detected for Lymphocytic Depletion and Mixed Cellularity types. The relative risks for nodular sclerosis, the most common type in young adults, were equivocal. We conclude that some etiological factor is important in the development of some subtypes of HD. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. INTRODUCTION Background The enigma regarding the nature of etiology of Hodgkin’s Disease (HD) started with its discovery by Thomas Hodgkin in 1832 and is yet to be resolved.1 It has been speculated to be of infectious origin since 19th century and was considered to be an infectious disease as late as in the 1940s. Since then numerous epidemiological, immunological and molecular studies have shown the possible infectious component of HD in association with viral exposure but no specific infectious agent has yet been identified. The binucleated or multinucleated Reed-Stemberg mirror cells, Hodgkin cells and lacunar cell formation is pathognomonic of HD. The Hodgkin tissue shows the purported tumor cells scattered in an array of lymphocytes, granulocytes, plasma cells, histocytes and fibrous tissue. This is consistent with the giant cell pattern seen in chronic infectious conditions such as tuberculosis, infectious mononucleosis or cat’s scratch disease. The clinical pattern characterized by insidious onset, progressive weakness, loss o f weight, slowly progressive lymphadenopathy starting at a single group and drenching night sweat with classical Pel-Ebstein fever is also suggestive of infectious disease.2 These features are almost similar to that of tuberculosis and explain the putative roll of Bacillus hodgkini in the etiology of HD as stated by Bunting and Yates in earlier part of this century.3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Epidemiological Features The epidemiological characteristics of Hodgkin’s disease are also suggestive of an infectious nature of the disease. Different surveys and studies6 ’1 3 revealed familial association o f HD. The number o f sib-sib pairs was found to be greater than other possible association like parent-sib pairs or spouse pairs. The few incidence o f spouse pairs is a potential proof against HD as a communicable disease. The higher incidence of sib-sib pairs than parent-sib pairs and greater similarity in time of onset among sibship pairs are suggestive of similar environmental exposure as a potential risk of Hodgkin’s disease.6 The higher incidence of HD then expected in twins of cases, but relatively low percentage of concordant twins of HD are suggestive of the interaction between genetic predisposition and environmental exposure. The cases of NS are often from high-socioeconomic status with small sibship size, reflecting delayed exposure to some common childhood disease and thus mimicking the polio pattern. The descriptive epidemiology o f HD has a characteristic bimodal age-incidence pattern and association with socioeconomic status. Cozen et al. showed that the biomodality o f age-incidence curve of HD in the Los Angeles County population is due to the high incidence o f Nodular Sclerosis (NS) among the younger age (<25) group and Mixed Cellularity (MC) among the older age (>45) group.4 The variation of risk pattern of HD in association with socio-economic status is also dependent on the geographic region.5 These features of HD have augmented the postulation o f c two-disease/multiple etiology’ hypothesis by MacMahon6 and 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ‘delayed exposure’ hypothesis by Gutensohn and Cole.7 In 1992, Cozen et al. showed the similarity between the risk pattern of NS and poliomyelitis and that of MC and tuberculosis. The elusive infectious pattern of HD have attracted intense attention of researchers all along and have resulted in numerous studies in the clustering and aggregation of HD in different settings such as school clusters,8 ’ 9 ’1 0 ’1 1 occupational settings,1 2 ’1 3 ’1 4 familial aggregation6 ,1 5 and community clusters.1 6 ’1 7 ’1 8 Vianna et al.8 showed significant case-case and case-controi linkage o f 31 Cases of HD in a high school at Albany, New York over a period of 23 years in 1971. They also reported aggregation of Cases of HD in the high schools at Nassau and Suffolk Counties of Long Island.9 A study by Zack MM. et al. of Connecticut Tumor registry, in 1977, revealed higher proportion of school contacts between the cases of H D .1 1 Milham detected a higher proportional mortality for HD among the teachers in Washington state.1 2 Higher mortality rate from HD than normal was also detected by Vianna et al. among the physicians in the upper state New York.1 3 But none of these finding have been conclusive and above criticism. The • O Q study of Vienna et al. is highly confounded by the socioeconomic status. ’ They failed to account for the variation of HD by socioeconomic status. The study of infectious component of HD can be undertaken by assessing either of the two patterns: (1) Spread of disease as in a communicable disease, by one to one contact or (2) Exposure o f cases to a common source. Vienna et al. and others often has used the first pattern. Other researchers using this pattern for HD 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. transmission failed to show any significant case-case link of HD. Smith and Pike failed to show any evidence of more contact among Cases of HD than control in a study done in Oxford.1 9 A study by Scherr et al. of cases of HD from July 1,1973 through December 1977 among the residents o f Boston and Worcester Standard Metropolitan Statistical areas failed to show any statistically significant school contact among Cases of H D .1 0 Smith et al. in contradiction to Vianna. et al. detected lower mortality rate from HD among the physicians then expected.2 0 Objective o f the Study Our study has been undertaken to seek a way out o f this etiological maze of Hodgkin’s disease. In this study, the hypothesis being tested is that the cases of HD will cluster by place of residence, age and high school, implying that they were exposed to the same infectious agent at the same time. If we do find clustering, it will support the idea that an infectious agent is involved in the etiology o f Hodgkin’s disease. Considering these assumptions, all cases of Hodgkin’s Disease, aged less than 50years, from 1968-1997 of Los Angeles County have been analyzed in this study to detect whether the proportion o f pairs studying in the same school and within one year of age difference was greater than the number of pairs expected by chance alone. The objective of this study is not to detect personal contact for HD but rather to explore if some exposure shared by people at childhood leads to the development of one of the sub-types of HD in the later years. Instead of traditional analyses of clustering, the analyses 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. employed in this study are chi-square tests or exact tests depending on cell 7 1 counts. Chen et al pointed out in a simulation study that the three techniques commonly used for the cluster analyses, namely Ederer-Myers-Mantel’s method, Knox’s method and Mantel’s method, are not adequately sensitive and thus the inconclusive results in many studies may be due to lack o f statistical power. All cases of HD have been stratified into the subtypes according to Rye’s classification. The cases within each group have been paired with all other cases, chi-square test was performed to detect whether the number o f pairs in same school of similar age was any different from the numbers expected from chance alone given the distribution of the cases. The school for each independent case was ascertained through the zip codes. The relative risk in each test reflected the risk of the cases being schoolmates. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. METHODS Data Collection The data presented in this study regarding the cases HD in Los Angeles County from. 1968-1997.Data of cases from 1972-1997 have been obtained from the University of Southern California Cancer Surveillance Program (CSP), the population-based cancer registry for Los Angeles County. The completeness of the data is assured by state mandated passive surveillance and CSP pursued active surveillance. The conscientious quality control standard o f the CSP checks against any misclassification of disease status and incompleteness o f the data. Information regarding each case’s gender, date of birth, ethnicity, date of diagnosis, socioeconomic status, histopathology, involved site and zip code was recorded. A survey was performed to characterize each high school in the Los Angeles County by zip codes. Study Subjects The subjects in this study consist o f all the cases o f HD in Los Angeles County from 1968-1997, ascertained by the University o f Southern California Cancer Surveillance Program, the population-based cancer registry o f Los Angeles County. A total of 4147 cases were grouped by Rye classification comprising of 2,565 cases of Nodular Sclerosis (NS), 812 cases of Mixed Cellularity (MC), 88 cases of Lymphocytic Depletion (LD) and 270 Lymphocytic Predominance (LP). A total of 398 (9.6%) cases could not be grouped in any of the four subclasses so they were grouped as ‘Not Otherwise Specified (NOS)’. 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Definitions of Study Variables Similar Age: Cases bom within one year of each other is defined as being of similar age. These cases are likely to be in the same class due to their age proximity. Same School: A survey was performed to characterize each high school in Los Angeles County by zip codes, and zip codes were subsequently replaced by the corresponding high school. The schools that the cases had attended were assumed to be the schools in the zip codes in which the cases lived at the time of diagnosis. In some of the zip codes there were multiple schools. In those instances a single school was designated for each zip code and all patients who resided in those zip codes were coded to that one school. In 138 zip codes there were one school per zip code. Pairs of Interest: Cases of similar age who have studied in same school are the pairs of interest, as these pairs represent the cases who had similar exposure in school. These pairs represent schoolmates who are more likely to be together and often be classmates. Relative Risk :The ‘Odds Ratio (OR)’ or ‘Relative Risk (RR)’ mentioned in this study should not be considered as individual relative risk but rather an event relative risk. It detects whether the proportion of pairs of cases of similar age and in same school is greater than that is expected by chance alone. This can be interpreted as the probability of cases being classmates or schoolmates. It should not be considered as proof o f any case-case links. 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Strict Definition: There was potential risk of misclassification of the exposure variable, ‘same school’, at zip codes with more than one school. To address this issue and quantify the extent o f misclassification the data was also analyzed using only the zip codes (138 )that have single school. Statistical Analysis The data were classified initially into five histopathological groups as NS, MC, LP, LD and NOS. Then the NS and MC groups had been further subdivided into six subgroups, each consisting of cases diagnosed within a five year interval starting from 1968 uptol997 (Table2). These subdivisions were done for the purpose of both computational feasibility and to investigate the possibility in the nature of case pairing across different time periods. Each case within the smallest unit (subgroups for NS and MC and groups for others) was then paired with rest of the cases within that unit. The possible number of pairs for ‘n’ cases is expressed by nC2=n(n-l)/2............................................................ (I) As only possible pairs are considered in this analyses and not detection of triplets or higher combination exact test as well as Chi-square test can be used for analyses. The ‘A’ cell of the 2x2 contingency table (figurel) represents the number of pairs o f cases of HD of similar age at same school. Other cells correspond to their definition. Chi-square method was applied instead of McNemar’s method, as each cell count represents pairs of cases with specific exposure characteristic. 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figurel: 2x2Contigency table of the pairs of cases. Cases o f Similar Age Cases of Different Age Studied in Same School A B Studied in C D Different School None o f the analyses was adjusted for any other factor. The study design and pairing adjusted for possible confounders, sex and socio-economic status. The pairing of cases in each unit was achieved by S-plus. The statistical analyses were attained by using SAS. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS Demographic Features The demographic features o f all the cases have been provided in TablelA- D. The male-female ratio of cases aged between 16-35 is almost 1 but the ratio is about 1.5:1 in cases aged below 16 years and above 35 years (Table 1A). The age of the cases appeared to be normally distributed as presented in figurel. No variability of age was noticed between the sexes from figure2. Except for the Jewish population the number of male cases were greater than the total number of female cases (TablelB). But male-to-female ratio for HD was almost one among Jewish patients. TablelC shows the gender distribution by sub-types o f HD. It is noticeable that the male-to-female ratio is almost 1 for NS whereas it is about 2:1- 3:1 in all other subtypes. About 73% of all female cases of HD suffer from NS. The variability of age in different sub-types is presented in figure3. The box-plots depicted that the mean age of Cases of NS was less than the mean age o f any other subtype. The total number of male cases was greater than that of female cases in all ethnicities (TablelD). The male-female ratio for HD cases was 1.3. Table l:Dem ographic Characteristics of the HD cases of Los Angeles County (1968-1997) Table 1A: Age Distribution 0-15 242 (5.8) 161 (3.9) 403(9.7) 16-25 783 (18.9) 682 (16.4) 1465(353) 26-35 838 (20.2) 624 (15.1) 1462(35.3) 36-45 518 (12.5) 299 (7.2) 817(19.7) All 2381 57.4 1766 42.6 4147(100) 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TablelB: Distribution of Religion BUBi MlMMM M iM M Christian Sects 1358 (32.8) 1062 (25.6) 2420 (58.4) Jewish 166 (4.0) 129 (3.0) 295 (7) Eastern Religion 14 (0.3) 10 (0.2) 24 (0.5) Atheist 178 (4.3) 111 (2.7) 289 (7) Unspecified 665 (16.1) 453 (11.0) 1118(27.1) All 2381 (57.5) 1765 (42.5) 4147(100) TablelC: Gender by Sub-types of Hodgkin’s Disease1 i i i i l l l l l l l l i l i a |B M B Nodular Sclerosis 1292 (31.2) 1286 (31.0) 2579 (62.2) Mixed Cellularity 573 (13.8) 239 (5.8) 812(19.6) Lymphocyte Predominance 203 (4.9) 67 (1.6) 270(6.5) Lymphocyte Depletion 61 (1.5) 27 (0.7) 88 (2.1) Not Otherwise Specified 252 (6.1) 146 (3.5) 398 (9.6) AH 2381 57.4 1766 42.6 4147(100) 1 According to Rye's classification of Hodgian's disease. 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TablelD: Ethnic Distribution. ^W flrn frr White 1640 (40.1) 1239 (30.3) 2880 (70.5) Hispanic 437 (10.7) 280 (6.9) 717 (17.5) African 212 (5.2) 177 (4.3) 389 (9-5) American Asian 31 (0.8) 30 (0.7) 61(1.5) Other 23 (0.6) 18 (0-4) 41(1) All 2343 (57.3) 1744 (42.7) 40882(100) “ Ethnicity o f 59 cases is unknown. Figurel :Histogram showing the age distribution of HD cases in Los Angeles County age variability of HD from 1968-1997 j ______ [______ I ______ [ ______ [ ______ ! ______ [ ______ 1 ______ [ ______ L 0 5 10 15 20 25 30 35 40 45 50 AGE 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. AGE Figure2:Variability of Age in Both Sexes is Compared by Box-Plot 30 U J 20 1 0 0 Figure 3: Box-Plots Showing Age Variability in HD Subtypes. 40 30 20 10 o Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Relative Risk fo r Cases o f HD Sharing Common exposure The result of the Chi-square analyses for clustering of Cases of HD in schools have been presented in Table2A and Table2B.The odds ratio (OR) or relative risk of cases being schoolmates or classmates have been presented in both the tables with 95% confidence interval (Cl) and the cell counts of the two-by-two table. Table2A presents the result from the analyses using all cases of HD. The finding for NS was equivocal. Except for the 1988-1992 period (during this period cases of NS within one year age difference were 1.2 times more likely to have been schoolmates or classmates than expected, the 95% Cl being 1-1.4.) no significant finding was noted for the cases of NS. The probability for cases being schoolmates or classmates appeared to be more then expected among cases of MC during 1988-1992 (RR=1.7 95% CI=l-3). The relative risk of 1.2 (with 95%CI ranging from 0.7 to 2.4) was detected for time periods 1983-1987 and 1993-1997. Separate analyses for each five-year period were not performed for LD, LP and NOS due to the small sample size. The relative risk of 2.1, with 95%CI being 1- 4.5, was noted for the cases of LD. The analyses revealed a relative risk of 1.59 among cases of LP with statistical significance. The 95% Cl ranged from 1.2 to as high as 2.2. Similar statistically significant association was also found for HD not otherwise specified (NOS). The relative risk being 1.3 with 95%CI being 1-1.6. Further analyses, using strict definition for ‘same school exposure’, was performed to evaluate the effect of possible misclassification of schools by zip codes. According to the ‘strict definition’ cases diagnosed in zip codes with a 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. single school were only considered. Zip codes that had more than one school were left out of the analyses. Table2B presents the relative risk with 95% Cl for the cases of the different histological types following a strict definition. No significant change was noticed for cases of NS. But remarkable changes were noticed among the cases of MC.RR greater than 1 with statistical significance was noted during the period 1993-1997 (RR=2.8 with 95% Cl =1.2-6.5). Relative risk during 1988-1992 and 1983-1988 was respectively 1.9 (with 95% Cl being 0.9- 4.3) and 1.8 (with 95% CI=0.8-3.8). No relative risk for time period!968-1972 could be calculated as no ‘pair of interest’ was detected in this time period. A significant increase o f RR was also noted among cases of LD. The relative risk for LD was 3.8 with the 95% Cl being 1.2-11.8. The relative risk for LP also increased to 1.7 from 1.6 using the strict definition. Though the 95% Cl (1-3) included 1 yet the p-value was almost 0.05. The relative risk of NOS remained almost unchanged but the 95%CI included 1 in this analyses. It should be noted that the relative risks in most of the analyses were above 1 except for cases of NS. Applying the strict definition the relative risks appeared to be higher among cases of MC, LD and LP. Due to reduction of sample size the 95% Cl often became wider and included 1. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TabIe2A: Cell counts and relative risk of HD among schoolmates: (All Zip Codes) Group No. of pairs Same Age ,Same School(A) No. of pairs Same School Different Age (B) No. of Pairs: Same Age- Different School (C) Total No. of Pairs (without any missing values) Relative Risk (95% Cl) Nodular sclerosis (including Hodgkin’s granuloma) 1968 - 1972 10 125 459 4425 0.7 (0.4-1.4) 1973 - 1977 84 631 3648 32640 1.02 (0.8-1.3) 1978 - 1982 125 1183 4884 49770 0.95 (0.8-1.14) 1983 - 1987 162 1640 7759 78606 0.88 (0.75-1.04) 1988 - 1992 207 1701 8176 89253 1.2 (1.0-1.4) 1993 - 1997 145 1532 6099 67528 0.93 (0.8-1.1) Mixed cellular (including Hodgkin’s granuloma) 1968 - 1972 3 38 135 1540 0.8 (0.2-2.6) 1973 - 1977 24 229 893 11175 1.2 (0.8-1.8) 1978 - 1982 23 203 753 7875 1.04 (0.7-1.6) 1983 - 1987 15 118 592 6328 1.2 (0.7-2.1) 1 988- 1992 14 96 488 6326 1.7 (1.0-3) 1993 - 1997 8 80 326 4098 12 (0.6-2.4) Lymphocytic depletion 8 61 155 2701 2.1 (1.0-4.5) Lymphocytic predominance 43 419 1352 22791 1.6 (1.2-2.2) NOS 97 1096 2476 39621 1.3 (1.0-1.6) 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TabIe2B: Cell counts and relative risk of HD among schoolmates: (Zip Codes With Single School) Group No. o f pairs Same Age .Same SchooI(A) No. o f pairs Same School Different Age (B) No. of Pairs: Same Age- Different School (C) Total No. o f Pairs (without any missing values) Relative Risk (95% Cl) Nodular sclerosis (including Hodgkin’s granuloma) 1968 -1972 4 56 125 1326 0.7 (0.2-1.8) 1973 -1977 27 182 844 7140 l.l (0.7-1.6) 1978-1982 38 387 1484 13530 0.8 (0.5-1.1) 1983 -1987 74 655 2477 23653 0.9 (0.7-1.2) 1988-1992 104 868 2992 30628 1.1 (0.9-1.3) 1993 -1997 75 613 2108 21738 1.1 (0.9-1.4) Mixed cellular (including Hodgkin’s granuloma) 1968 -1972 0 15 15 465 — 1973 -1977 13 102 332 3916 1.3 (0.7-2) 1978 -1982 9 78 206 2278 1.1 (0.6-2.3) 1983 - 1987 8 42 162 1711 1.8 (0.8-3.8) 1988-1992 8 45 200 2485 1.9 (0.9-4.3) 1993-1997 7 30 78 1035 2.8 (1.2-6.5) Lymphocytic depletion 4 14 46 666 3.7 (1.2-11.8) Lymphocytic predominance 14 115 316 4851 1-7 (1-3) NOS 44 480 734 11325 1.26 (0.9-1.7) 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION Study Outcome and Previous Studies This study is unique both in its approach and finding. Most of the prior cluster studies of HD have dealt with HD as a single entity and did not delve in the individual histological subtype of HD. To evaluate individual infectious component, each subtype o f HD was analyzed separately in this study. Previous studies revealed that the incidence of NS is higher among younger population below 45 of the high socioeconomic status. This unique frisk pattern of NS is similar to that of the poliomyelitis risk model. In this analysis we failed to show any excess in common childhood exposure among cases of NS. The relative risk in all periods was around 1. This result can be explained in the light of sample size and strong association o f NS with socioeconomic status and age. The breaking of the data in six five year interval periods have greatly reduced the number of effective pairs. A case diagnosed at one of the endpoints of any time period could not be paired with the cases of adjacent period. As for example a case diagnosed in the year 1983 could not be paired with the case diagnosed in the year 1982. The age dependent incidence of NS is most noted during the age25-35 but in this study the age of the cases ranged from 2 to 45 years.4 But to study the ‘delayed exposure’ hypothesis this broad age spectrum might not be suitable. The matching of cases by zip codes have reduced the bias due to socioeconomic status but by considering cases of all socio-economic status in a single stratum might have caused dilution effect. The cases o f NS diagnosed in 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the lower socioeconomic status with large family size do not correspond to the risk model of poliomyelitis. The evidence of clustering was more prominent for MC. The relative risk for MC in all time periods but one were greater than 1. Considering the strict definition for ‘same school’ the relative risk in all time periods was above 1. In some periods the relative risk were aro und two. This should be interpreted as strong findings regarding possible clustering of cases of MC in school. The high relative risk of LD was very interesting from both epidemiological and pathological consideration. From histological point LD is an outcast of HD, as it is characterized by Reed-Stemberg cells with sparse lymphocyte and greater bulk of fibrous tissue. The sparse distribution of lymphocyte is not explainable. Though LD is rather rare in comparison to MC yet their likelihood in risk pattern is well appreciated. Both conditions are rare in the younger population with increase of incidence with age.4 The relative risk for NS, MC, NOS and LP detected in this study well corresponds with previous studies regarding aggregation of Cases of HD in schools. Scherr et al. studied cases of HD diagnosed from July 1,1973 through December 31,1977, among residents of Boston and Worcester Metropolitan area. 245 cases were matched to population based controls on age and sex.1 0 Each subject was interviewed and schooling history was gathered. They observed that in high schools Cases of HD were schoolmates (RR=1.2) more than expected and so were classmates (1.9). Cases were found to be schoolmates (RR=1.4) more 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. often in elementary school. But in junior schools the cases appeared less likely to be schoolmates (RR=0.8). Grufferman et al. studied 448 cases of HD, diagnosed in Boston from 1959 through 1973.2 2 They applied an extension o f Knox's model for space-time clustering, to detect links of HD in schools. The relative risk of links for schoolmates was 1.2 and classmates 0.9. Zack et al. studied the ‘patterns of high school’ contact for HD among cases o f Leukemia and Lymphoma diagnosed among Connecticut residents aged 15-29 from 1960 through 1971.1 1 The list of cases was supplied by Connecticut Tumor Registry. A total of 229 cases of HD were included in the analyses of the potential 606 cases on the basis of attending 10-12 grades in Connecticut schools. The data were analyzed by case-control, index-case cohort and two-interval cohort approach. Cases attending same school simultaneously was compared to the number of cases in simulated controls (on the basis of grade levels and years of attendance of the patients were randomly selected from Connecticut high schools with similar 10th grade enrolment). In this extensive analyses the researchers detected relative risk ranging from 1.24 to 1.7 for grade specific links for HD. The index-case cohort analyses showed that the diagnosis of HD among the exposed was twice that of the unexposed (RR=2.05). Role o f Novel Approach to Clustering Analysis It should be noted that in most of the previous studies the 95% Cl often included 1. In the traditional time cluster methods one often has to declare a time 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and space boundary for the method to work. But in diseases like Hodgkin’s disease this limits are often arbitrary. As Chen, et al. pointed out that in real life situation the sensitivity o f Knox’s method and other common time-space clustering methods are not sufficient to detect clustering o f diseases like HD.2 1 On the more those methods are applicable to detect case-case linkage rather than exposure to common source. In this Chi-square approach the time and space limit imposed are derived directly from the study concept and it measured the odds ratio of the proportion exposed to a common source rather than case-case linkage. Another limitation o f the traditional methods of cluster analysis is that one has to define an arbitrary space-time for contact. The difficulty in defining the distance required for transmission of a disease like HD is easily perceivable. This constraint was circumvented in this study by addressing common exposure in schools. Due to use o f pairing of cases the analyses were stratified on the basis of possible confounders such as age at diagnosis, sex and socio-economic conditions. Though the cases are detected at different age yet they have been analyzed on the basis o f birth cohort. The age difference among cases has been the variable of interest, which is not affected by the age at diagnosis. The lack of variance in age among different subtypes(Figure3) and between male and female(Figure2) also negates the necessity of treating age as a confounder in all of the analysis. Each case has been paired with all possible cases of both sexes. Thus pairs of classmates and schoolmates have been established between both sexes. The socio-economic status of the population is defined by the census tract 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and zip code. Considering schools by zip codes by definition accounts for socioeconomic status of the cases. Pairing also improves the sample size and thus increases the power of the study. A total of 100 cases provides a sample size of 9,900 cell counts. This is greatly helpful in the study o f rare diseases. Implication o f the result The findings of this study provide further evidence in support to the hypothesis that cases of HD shares similar exposure during childhood. It is impossible to provide conclusive proof for infectious nature of HD by epidemiological studies alone. The strongest argument against the infectious hypothesis o f HD is the lack of numerous case-case link and all contacts of cases not attaining the disease. But this condition should be viewed from the perspective o f the possible involved microbe. Low-grade infection among the majority with fatal infection among few is a characteristic nature of many viral infections. It should be noted that high fatality is counterproductive for the survival of the involved microbe. The lower incidence of HD among Asians, higher incidence among Jews4 and greater association with certain HLA types and familial susceptibility25 implies the possible genetic predisposition for HD. The disease can be postulated as an abnormal response to certain common infection in genetically susceptible population. Research has demonstrated that HD develops in stepwise pattern as documented by Okuyama “...beginning with a reversible, biological stage during which Hodgkin and Reed-Stemberg cells are formed, followed by constitutive but reversible Hodgkinogenic medical stage that leads to 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. an irreversible, systemic and fatal proto-oncogenic stage.” 2 6 Association o f HD with Epstein-Barr virus (EBV) and HTV-infection has been well documented in numerous studies. Cases of HD among HTV infected person often shows the presence of EBV genome in the RS, Hodgkin and other tumor cells. The association of HD with EBV ranges from 26-50% of cases. The rate of association also varies in different studies but a similar pattern is documented by all researchers. The association of EBV have been noted to be highest, almost 100% for LD. MC shows association in about 50-70% cases and 10-40% cases o f NS. In a study in Kenya showed that 100% of pediatric cases of HD ( all subtypes) and • • 9 7 66% cases of adult HD (all subtypes) were associated with EBV infection. The high relative risk for LD and relative risk above 1 for MC can be explained by the association of the subtypes with EBV and. Limitations o f the Study and Future Approach In this study the relative risk attained following the ‘strict’ definition for ‘same school’ was often higher than the one detected by ‘loose’ definition, underlies the limitation of the study. The detection of schools of the cases by the zip codes at the time of diagnosis has some inherent probability of mistake. The chance o f misclassification arises from two sources: 1) Cases might not have lived at the same zip code during his childhood as he did at the time of diagnosis. 2) At zip codes with more than two schools the cases might have been at a different school than that is considered in this analyses. But there is no reason to think that this misclassification will affect those o f ‘similar age’ more than those of different age and vice versa. So, from epidemiological point this is introducing non- 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. differential misclassification, which always results in changing the odds ratio towards the null. This condition can be appreciated from the increase in the relative risk after employing ‘strict’ definition. The non-significant result for cases of Nodular Sclerosis and significant result for other subtypes is perplexing in the sense that previous studies have shown that NS commonly follows the polio risk pattern. In the light of the limitations o f this study the data is required to be further analyzed by pairing all cases o f nodular sclerosis from 1968 to 1997. To increase the precision o f the study cases of NS aged between 20 to 40 should be analyzed as subset analysis as this age group is found to follow the polio risk pattern. To protect the study from residual effect of socioeconomic status, pairs should be formed on the basis of socioeconomic status and then analyzed by stratified analysis. Conclusion The findings of this study is not a conclusive proof on behalf of the infectious nature of HD but certainly aid in strengthening the hypothesis by providing further evidence in the same direction. The novel approach used in this study is an easily applicable tool for analyzing already existing data of population based cancer registries. If data can be collected from each case regarding his/her schooling with specific time period than it will significantly save the study from possible misclassification. Addition of this single information will significantly increase the power of detecting school clusters in Hodgkin’s disease. Further studies regarding the aggregation of HD by subtypes should be undertaken to unearth the variation in the risk factors of different sub-types. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bibliography: Encyclopaedia Britannica,: "Hodgkin, Thomas" Britannica Online. <www.eb.com: 180/cgi-bin/g?DocF=micro/273/39.ht3nl>, 1994-1998, Encyclopaedia Britannica Inc. 2 Allan N.C. and Ludlam C.A.: Diseases of the blood, Davidson’s Principal’s and Practice of Medicine, 16th Edition,1991,ELBS. 3 Yates JL., and Bunting C.H.: The Rational Treatment of Hodgkin’s Disease. J. Am. Med. Assoc.64:1953-1961, 1915. 4 Cozen W., Katz J. and Mack T.M: Risk Patterns of Hodgkin’s Disease in Los Angeles Vary by Cell Type, Cancer Epidemiology, Biomarkers & Prevention vol. 1,261-268, May/June 1992. 5 Glaser S.L: Regional Variation in Hodgkin’s Disease. Cancer Research 26:1189-1200,1966. 6 MacMahon B.: Epidemiology of Hodgkin’s disease. Cancer Research 26:1189-1200,1966. 7 Gutensohn N.M. and Cole P.: Epidemiology of Hodgkin’s disease. Seminars n Oncology7:92-102,1980. 8 Vianna N.J., Greenwald P. et al: Hodgkin’s Disease: Cases with Features of a Community Outbreak. Ann Intern Med. 1972;77:169-180. 9 Vianna N J. and Polan A.K.: Epidemiologic Evidence for Transmission of Hodgkin’s Disease. N.Engl. J. Med. 1973; 289:499-502. 1 0 Scherr P.A., Gutensohn N. and Cole P.: School Contract among Persons with Hodgkin’s Disease, Am. J. of Epidemiology:vol. 120,No. 1,29-38,1984. 1 1 Zack M.M et al.: High School Contact among Persons with Leukemia and Lymphoma. JNatl. Cancer Inst.:vol.59, No.5, 1343-1349, Nov 1977. 1 2 Milham S.: Hodgkin’s Disease as an Occupational Disease of School Teachers. N Engl. J. Med.:290-1329, 1974. 1 3 Vianna N.J. et al: Disease Mortality among Physicians. Lancet,2:131-133, 1974. 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 4 Matanoski G.M. et al.: Hodgkin’s Disease Mortality among Physician. Lancet:926-927, 1975. 1 5 Razis D.V. et al: Familial Hodgkin’s Disease :Its Significance and Implications. A m Intern Med,51:933-171, 1959. 1 6 Dworsky R.L. and Henderson BE.: Hodgkin’s Disease Clustering Families and Communities. Cancer Res.,34:1161-1163,1974. 1 7 Klinger R.J. and Minton J.P.: Case Clustering of Hodgkin’s Disease in a Small Rural Community with Association among Cases.Lancet, 1:168- 170,1973. 1 8 Schimpff S.C. et al.: Leukemia and Lymphoma Patients Interlinked by Prior Social ContactXancet, 1:124-129,1975. 1 9 Smith P.G and Pike M.C.: Case Cluster in Hodgkin’s Disease: A Brief Review of Present Position and Report of Current Work in Oxford. Cancer Res 34:1156-1160,1974. 2 0 Smtih P.G. et al.: Hodgkin’s Disease Mortality among Physicians. Lancet,ii:525,1974. 2 1 Chen R. et al: A Study of three Techniques for Time-Space Clustering in Hodgkin’s Disease. Statistics in medicine VoI:.3,173-184,1984 2 2 Grufferman S., Cole P. and Levitan T.: Evidence of Transmission of Hodgkin’s Disease in High School. N.Engl. J. Med. 300:1006-1011,1979. “Parkin D.M. et al.: Cancer Incidence in Five Continents, volVI IARC Scientific Publications no.l20,Lyon:Intemational Agency for Research for Cancer. 2 4 MacFarlane G J. et al: International patterns in the Occurrence of Hodgkin’s Disease in Children and Young Adult Males. International Journal of Cancer 61:165-169. “Mack T.M., Cozen W. et al.: Concordance of Hodgkin’s Disease in Identical Twins Suggesting Genetic Susceptibility to the Young-Adult form of the Disease. N. Engl. J. of Med.332:413-418. 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 6 Okuyama S.: Evolution of Hodgkin’s Disease. Medical Hypotheses. 47(3):199-213,1996, September. 2 7 Weinreb M. et al.: The Consistent Association between Epstein-Barr Virus (EBV) And Hodgkin’s Disease in Children in Kenya. Blood.87(9):3828-3836, 1996 May Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ■ > > IMAGE EVALUATION TEST TARGET (Q A -3 ) / ✓ 4 b * A j * : * ? / , 1 . 0 l.l ys £ H Sii IM IjJJ^ 1 2.2 2.0 1 . 8 1.25 1.4 1 . 6 150mm A P P L I E D ^ IIW IG E . In c 1653 East Main Street - ^ = ~ - Rochester, NY 14609 USA — Phone: 716/482-0300 - = ~ - = Fax: 716/288-5989 O 1993. Applied Image. Inc.. All Rights R eserved Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Interaction of dietary fiber and serum cholesterol on early atherosclerosis

PDF

Immune recovery vitritis in AIDS: Incidence, clinical predictors, sequellae, and treatment outcomes

PDF

Extent, prevalence and progression of coronary calcium in four ethnic groups

PDF

Evaluation of the efficacy of two lipid-lowering treatments using serial quantitative coronary angiography: Two- and four-year treatment results

PDF

Application of a two-stage case-control sampling design based on a surrogate measure of exposure

PDF

A pilot survey of medical abortion knowledge and practices among obstetrician/gynecologists and family practitioners in Los Angeles County

PDF

Comparisons of metabolic factors among gestational diabetes mellitus probands, siblings and cousins

PDF

Imputation methods for missing items in the Vitality scale of the MOS SF-36 Quality of Life (QOL) Questionnaire

PDF

Descriptive epidemiology of thyroid cancer in Los Angeles County, 1972-1995

PDF

Familiality and environmental risk factors of peptic ulcer: A twin study

PDF

The Magnitude And Rate Of Decline In Alpha-Fetoprotein Levels In Children With Unresectable Or Metastatic Hepatoblastoma, Predict Outcome

PDF

Endometrial cancer following breast cancer treatment: Tumor characteristics and predictors of survival

PDF

Cluster analysis of p53 mutational spectra

PDF

Molecular And Clinical Prognosticators For Survival In Epithelial Ovarian Carcinoma

PDF

Do reproductive factors have the same effect on breast cancer risk in African-American and White women?

PDF

Association between latchkey status and smoking behavior in middle school children

PDF

The fit of the genetic relative risk models of the human leucocyte antigen (HLA) in the risk of insulin-dependent diabetes mellitus in the Finnish population

$High-dose chemotherapy followed by autologous hematiopoietic progenitor cell transplantation for relapsed or refractory Hodgkin's disease$

PDF

High-dose chemotherapy followed by autologous hematiopoietic progenitor cell transplantation for relapsed or refractory Hodgkin's disease

PDF

Rates of cognitive decline using logitudinal neuropsychological measures in Alzheimer's disease

PDF

Associations of weight, weight change and body mass with breast cancer risk in Hispanic and non-Hispanic white women

Asset Metadata

Creator Islam, Khandaker Talat Shamsul (author)

Core Title Does young adult Hodgkin's disease cluster by school, residence and age?

School Graduate School

Degree Master of Science

Degree Program Applied Biometry

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

Tag biology, biostatistics,health sciences, medicine and surgery,health sciences, oncology,health sciences, public health,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor [illegible] (committee chair), [illegible] (committee member), Azen, Stanley (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-26975

Unique identifier UC11336981

Identifier 1394767.pdf (filename),usctheses-c16-26975 (legacy record id)

Legacy Identifier 1394767.pdf

Dmrecord 26975

Document Type Thesis

Rights Islam, Khandaker Talat Shamsul

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