Melanoma in children, adolescents, and young adults

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content MELANOMA IN CHILDREN, ADOLESCENTS, AND YOUNG ADULTS by Katherine Y. Wojcik A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (EPIDEMIOLOGY) AUGUST 2017 Copyright 2017 Katherine Y. Wojcik ii DEDICATION With gratitude to my Dad, whose unwavering support and endless curiosity inspired me in childhood to become a scientist. A day long in the making, but well worth it, indeed! iii ACKNOWLEDGEMENTS To my mentor and dissertation chair, Dr. Myles Cockburn, I am eternally grateful for your insightful guidance, sensible and timely advice, and steadfast encouragement throughout my doctoral training. As a result of your efforts and foresight, I have found an exciting path forward in cancer research, including melanoma and non-melanoma skin cancers, registry-based studies, and AYA-focused efforts. You’ve been an exemplary mentor, colleague, and friend. What luck to have joined your team, where there have been countless opportunities to pursue new questions, collaborate across disciplines, and enjoy the occasional group outings to the brewpub! Many thanks to my esteemed committee members: Dr. Joel Milam, Dr. Ann Hamilton, and Dr. Ashley Wysong. I am so grateful for your enduring support and critical feedback, and for providing opportunities to do hands-on work in both registry-based and clinic- based settings. Dr. Milam, thank you for providing the crucial opportunity to collect my own pilot data within the Project Forward framework. Dr. Hamilton, I am grateful to you for sharing your registry wisdom and cohort study expertise, which allowed me to better appreciate the intricacies of registry data and cohort recruitment efforts. Dr. Wysong, I am so appreciative of your infectious enthusiasm, motivation, and keen clinical perspective, which ensured each study’s results would have translatable utility. iv To Marlene Caldera, your cheerful, warm disposition and tireless support has kept the “lab” running smoothly and is immensely appreciated. I am indebted to The SunSmart team, The Project Forward team, and The Cancer Surveillance Program of Los Angeles County, especially Dr. Dennis Deapen and Dr. Lihua Liu, for their support and contributions to the collection and interpretation of the registry data involved herein. Rosa Castro, I am deeply grateful for your boundless support and friendship, advice through thick and thin, readiness to embark on road trips, and countless meals that sustained me while in the depths of data analysis! You are my sister (sestra). To friends old and new: Jennifer Kulzer, Loraine Agustin, Derek Dangerfield, Tina Patel, Jeffery Williams, Chris & Alyssa, Jeff Ruben, Sina Vakili, Nilam Dave, Charlotte Deng, Ugonna Ihenacho, Jessi Tobin, Kim Miller, Stef Thomas, Jessica Trimis, Amanda Goodrich, Carrie Tayour, Stephanie Dyal, Dardney, Christian Cerrada and many others – thank you for all the happy hours, brainstorms, commiserating, and advice along the way. To my dear mentors, Dr. David Rechtman, Dr. Martin Lee, Dr. Douglas Lewis, Dr. Richard Watanabe, and Dr. Roberta McKean-Cowdin, your support and encouragement over the years has been absolutely essential to this accomplishment. Thank you! TABLE OF CONTENTS DEDICATION ii ACKNOWLEDGEMENTS iii LIST OF TABLES vii LIST OF FIGURES ix ABBREVIATIONS x ABSTRACT xii CHAPTER 1: INTRODUCTION 1 Adolescents and Young Adults (AYAs) are an Understudied Group 2 The Need for Melanoma-Specific Study Among Young People 3 Overview of Studies 7 CHAPTER 1 REFERENCES 10 CHAPTER 2: MELANOMA INCIDENCE AND SURVIVAL PRIOR TO AGE 65 13 ABSTRACT 14 BACKGROUND 15 Study Purpose & Aims 18 MATERIALS & METHODS 20 Population & Data Sources 20 Statistical Analysis 26 RESULTS 27 Age-Adjusted Incidence Rates (AAIR) 28 5-Year Observed Survival Probability (Kaplan Meier Estimates) 29 Hazard Ratios for Risk of Death (HR; Cox Hazard Regression Estimates) 32 DISCUSSION 37 Strengths & Limitations 44 Generalizability 46 Conclusion 48 CHAPTER 2 REFERENCES 60 CHAPTER 3: HIGH BIRTH WEIGHT, EARLY UV EXPOSURE, AND MELANOMA RISK IN CHILDREN, ADOLESCENTS AND YOUNG ADULTS 68 ABSTRACT 69 BACKGROUND 70 MATERIALS & METHODS 72 Population & Data Sources 72 Statistical Analysis 73 RESULTS 75 DISCUSSION 77 Strengths and Limitations 80 vi Conclusion 81 CHAPTER 3 REFERENCES 86 CHAPTER 4: BARRIERS TO CANCER-RELATED FOLLOW-UP CARE AMONG YOUNG ADULT SURVIVORS OF MELNAOMA 96 ABSTRACT 97 BACKGROUND 98 Gaps in Knowledge for AYA Melanoma Prevention 98 Survival 100 Importance of Cancer-Related Follow-up Care (CRFC) During Survivorship 101 Role of Barriers in Low Adherence to Medical Advice 102 Consequences of Lacking Cancer-Related Follow-up During Survivorship 106 Study Aims 107 MATERIALS & METHODS 108 Statistical Analysis. 112 RESULTS 113 DISCUSSION 116 Survivorship & Denial of Cancer 119 Strengths & Limitations 121 Generalizability 123 Conclusion 125 CHAPTER 4 REFERENCES 130 CHAPTER 5: CONCLUSION 135 Implication s & Implementation 136 Future Directions 139 CHAPTER 5 REFERENCES 143 BIBLIOGRAPHY 145 APPENDIX A: Kaplan Meier (KM) Survival Plots 163 APPENDIX B: Melanoma Specific Survey 201 APPENDIX C: Skin-care behavior questions 203 APPENDIX D: Residential History Form 204 APPENDIX E: Introduction Letter/Informed Consent 205 APPENDIX F: California Cancer Registry Brochure 207 APPENDIX G: Study Protocol 208 LIST OF TABLES TABLE 2.0.1. DEMOGRAPHIC CHARACTERISTICS OF PATIENTS DIAGNOSED WITH MELANOMA, BY AGE GROUP, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* ...................................................................................................... 50 TABLE 2.0.2. TUMOR CHARACTERISTICS OF PATIENTS DIAGNOSED WITH MELANOMA, BY AGE GROUP, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014 .................................................................................................................... 51 TABLE 2.0.3. AGE-ADJUSTED INCIDENCE RATES (AAIR) FOR ALL CUTANEOUS MELANOMAS DIAGNOSED AT AGES 0-64 YEARS, BY AGE GROUP, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* ............................................................... 52 TABLE 2.0.4. FIVE YEAR OBSERVED SURVIVAL (OS) PROBABILITY AND CHANGE OVER TIME IN PATIENTS DIAGNOSED WITH MELANOMA, BY AGE GROUP AND DEMOGRAPHICS, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* ............... 53 TABLE 2.0.5. FIVE YEAR OBSERVED SURVIVAL (OS) IN PATIENTS WITH MELANOMA DIAGNOSED AT AGES 0-64 YEARS, BY AGE GROUP AND TUMOR CHARACTERISTICS, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* ............................... 54 TABLE 2.0.6. HAZARD RATIOS (HR) FOR RISK OF DEATH BY DEMOGRAPHICS, STRATIFIED BY AGE GROUP, IN PATIENTS WITH MELANOMA DIAGNOSED AT AGES 0-64 YEARS, CALIFORNIA (1988-2014)* ...................................................... 55 TABLE 2.0.7. HAZARD RATIOS FOR RISK OF DEATH BY TUMOR CHARACTERISTICS, STRATIFIED BY AGE GROUP, IN PATIENTS WITH MELANOMA DIAGNOSED AT AGES 0-64 YEARS, CALIFORNIA (1988-2014) ................................................ 56 TABLE 2.0.8. AGE GROUP COMPARISONS OF HAZARD RATIOS (HR) FOR RISK OF DEATH IN MALES VS. FEMALES IN PATIENTS DIAGNOSED WITH MELANOMA AT AGES 0-64 YEARS (1988-2014) .................................................................. 57 TABLE 2.0.9. HAZARD RATIOS (HR) FOR RISK OF DEATH IN MALES VS. FEMALES AMONG ADOLESCENTS AND YOUNG ADULTS (AYA) DIAGNOSED WITH MELANOMA AT AGES 15-30 YEARS, CALIFORNIA (1988-2014), STRATIFIED BY SELECTED CHARACTERISTICS ..................................................................................................................................... 57 TABLE 2.0.10. DEMOGRAPHIC CHARACTERISTICS OF HISPANIC ADOLESCENTS AND YOUNG ADULTS (AYAS) DIAGNOSED WITH MELANOMA AT AGES 15-39 YEARS, CALIFORNIA (1988-2014)* ..................................................................... 58 TABLE 2.0.11. TUMOR CHARACTERISTICS OF HISPANIC ADOLESCENTS AND YOUNG ADULTS (AYAS) DIAGNOSED WITH MELANOMA AT AGES 15-39 YEARS, CALIFORNIA (1988-2014)* .................................................................... 59 TABLE 3.0.1. CHARACTERISTICS OF A CASE-CONTROL STUDY OF MELANOMA DIAGNOSED AT AGES 0-29 YEARS IN CALIFORNIA (1988-2013)* ....................................................................................................................................... 82 TABLE 3.0.2. EARLY LIFE RISK FACTORS FOR MELANOMA DIAGNOSED AT AGES 0-29 IN CALIFORNIA (1988-2013) ........... 83 TABLE 3.0.3. AGE-STRATIFIED ASSOCIATION OF EARLY LIFE AMBIENT ULTRAVIOLET (UV) RADIATION AND MELANOMA DIAGNOSED AT AGES 0-29 YEARS IN CALIFORNIA (1988-2013) ....................................................................... 84 TABLE 3.0.4. EVALUATION OF DOSE-RESPONSE ACROSS BIRTH WEIGHT FOR MELANOMA DIAGNOSED AT AGES 0-29 YEARS IN CALIFORNIA (1988-2013) ........................................................................................................................ 85 viii TABLE 4.0.1. YOUNG ADULT SURVIVORS OF MELANOMA DIAGNOSED AT AGES 0-24 YEARS IN LOS ANGELES COUNTY (1996- 2010), BY CANCER-RELATED FOLLOW-UP CARE STATUS (CRFC)* ................................................................... 126 TABLE 4.0.2. BARRIERS TO CARE AMONG YOUNG ADULT SURVIVORS OF MELANOMA IN LOS ANGELES COUNTY (1996- 2010)* ................................................................................................................................................ 127 TABLE 4.0.3. SKIN PROTECTIVE BEHAVIORS AMONG YOUNG ADULT MELANOMA SURVIVORS IN LOS ANGELES COUNTY (1996- 2010)* ................................................................................................................................................ 128 TABLE 4.0.4. BARRIERS ASSOCIATED WITH LACK OF RECENT FOLLOW-UP CARE (CRFC) AMONG YOUNG ADULT MELANOMA SURVIVORS IN LOS ANGELES COUNTY (1996-2010) ..................................................................................... 129 ix LIST OF FIGURES FIGURE 1.0.1. MAJOR MILESTONES FOR AYA ONCOLOGY IN NORTH AMERICA (14, P. 2). ............................................... 9 FIGURE 1.0.2. THE GAP IN RELATIVE 5-YEAR SURVIVAL FOR AYAS WITH INVASIVE CANCER (2, P. 5). ................................ 9 FIGURE 2.0.1. MELANOMA-SPECIFIC SURVIVAL BY AGE AT DIAGNOSIS (26, P. S78.E6). ............................................... 21 FIGURE 2.0.2. CHANGE IN 5-YEAR SURVIVAL (ALL STAGES COMBINED) AMONG PATIENTS WITH MELANOMA IN CALIFORNIA, 1988-2014 ............................................................................................................................................ 30 FIGURE 2.0.3. CHANGE IN 5-YEAR SURVIVAL (REMOTE STAGE ONLY) AMONG PATIENTS WITH MELANOMA IN CALIFORNIA, 1988-2014 ............................................................................................................................................ 30 FIGURE 2.0.4. PROPOSED EXPLANATIONS FOR SEX DIFFERENCES IN MELANOMA SURVIVAL (48, P.43) ............................ 43 FIGURE 4.0.1. SEX DIFFERENCES FOR AGE-ADJUSTED INCIDENCE OF MELANOMA AMONG WHITES IN THE UNITED STATES (1975-2013), BY AGE GROUP (AGES <50 VS. 50+ YEARS) ............................................................................. 98 FIGURE 4.0.2. LITERATURE ON ADOLESCENTS AND YOUNG ADULTS HAS GROWN SINCE 2008 BUT STILL LAGS BEHIND STUDIES IN CHILDREN (9, P. VI) .............................................................................................................................. 99 FIGURE 4.0.3. RELATIVE RISK (OBSERVED TO EXPECTED RATIO [O:E]) OF SUBSEQUENT PRIMARY CANCER AFTER CUTANEOUS MALIGNANT MELANOMA (CMM) BY AGE GROUP, 1973-2006 (13, P.268) ................................................... 101 FIGURE 4.0.4. A FRAMEWORK FOR UNDERSTANDING THE BARRIERS TO ADHERENCE AMONG ADOLESCENTS AND YOUNG ADULTS (AYAS) (15 P.567) ..................................................................................................................... 102 FIGURE 4.0.5. DIMENSIONS OF AYA BARRIERS TO TREATMENT ADHERENCE (15, P. 571-573) ..................................... 103 FIGURE 4.0.6. FACTORS IMPACTING FINANCIAL DISTRESS AND MORTALITY (25, P. 2) .................................................. 105 FIGURE 4.0.7. IMPACT OF CANCER DIAGNOSIS ON WORK OR SCHOOL STATUS AMONG ADOLESCENT AND YOUNG ADULT PATIENTS (26, P. 2) ................................................................................................................................ 106 x ABBREVIATIONS AAIR Age-Adjusted Incidence Rate ALM Acral Lentiginous Melanoma AYA(s) Adolescent(s) and Young Adult(s) AYAO Adolescent and Young Adult Oncology AYAO-PRG Adolescent and Young Adult Oncology Progress Review Group BRAF B-Raf proto-oncogene, serine/threonine kinase CCR California Cancer Registry CDC Centers for Disease Control CSP Cancer Surveillance Program CI Confidence Interval CRFC Cancer-Related Follow-up Care FTO Obesity and fat-mass associated gene; produces alpha-ketoglutarate dependent dioxygenase g Grams GPM Giant Pigmented Nevus Melanoma HBW High Birth Weight HR Hazard Ratio ICD-O International Classification of Disease for Oncology LBW Low Birth Weight LM Lentigo Maligna xi LA Los Angeles KM Kaplan-Meier mm Millimeters NAACCR North American Association of Central Cancer Registries NCI National Cancer Institute NHW Non-Hispanic White NM Nodular Melanoma NOS Not Otherwise Specified NPCR National Program of Cancer Registries NRAS proto-oncogene, GTPase OR Odds Ratio SD Standard Deviation SEER Surveillance, Epidemiology, and End Results SES Socioeconomic Status SSM Superficial Spreading Melanoma US United States USPS United States Postal Service UV Ultraviolet Radiation YA Young Adult xii ABSTRACT Melanoma, the most deadly form of skin cancer, is among the top 5 most common cancers among adolescents and young adults (AYAs), who have unique underlying biology and face a distinct set of life course challenges that may affect incidence and survival, distinguishing them from younger or older patients. Existing literature predominantly represents patients well over age 50, and very little detail regarding melanoma in young patients exists. To address this deficit, we conducted a set of population-based cancer registry studies, including a detailed descriptive study of incidence and survival characteristics among AYA melanoma patients with comparisons to older and younger patients in California, a large, case-control study to identify early life risk factors for developing melanoma before age 30, and a pilot survey-based study among young adult survivors of melanoma in Los Angeles County to identify important barriers to getting cancer-related follow-up care during survivorship. We found key opportunities to improve melanoma prevention and control among AYA populations. AYA programs, as joint efforts between pediatric and adult oncologists, are gaining momentum to become the standard of care, offering a promising, dynamic approach to improving care of AYAs. Intensifying AYA-specific, site-specific research will only serve to strengthen those efforts by providing the evidence base necessary to justify resource allocation for supporting program development and maintenance to ensure that optimal care for all AYAs is being delivered. 1 CHAPTER 1: INTRODUCTION 2 Melanoma is the most deadly of the three main types of skin cancer, which also include basal and squamous cell carcinomas. Although preventable, it is responsible for >75% all of deaths from skin cancer. More than 70,000 new melanomas are diagnosed each year in the United States (US) (1). ADOLESCENTS AND YOUNG ADULTS (AYAS) ARE AN UNDERSTUDIED GROUP Historically, AYAs have been an understudied age group, existing in the grey area between pediatric and adult oncology practice. In 2005, an AYA oncology progress review group (AYAO-PRG) was convened as a joint effort between the National Cancer Institute (NCI) and the Lance Armstrong Foundation to define leadership, responsibilities, and goals, and develop a substantive rationale for distinguishing and championing the “research and cancer care needs” of AYAs diagnosed with cancer. This marked an important milestone in the development of AYA oncology (Figure 1 from the AYAO-PRG report) (2). The AYAO-PRG’s resulting report lists the following five priority recommendations (2): (1) Identify the characteristics that distinguish the unique cancer burden in the AYAO patient. (2) Provide education, training, and communication to improve awareness, prevention, access, and quality care to AYAs. (3) Create the tools to study the AYA cancer problem. (4) Ensure excellence in service delivery across the cancer control continuum (i.e. prevention, screening, diagnosis, treatment, survivorship, and end of life). 3 (5) Strengthen and promote advocacy and support of the AYA cancer patient. The report also highlighted the existence of a dramatic gap for AYAs in 5-year survival improvements over the period of 1975-1997, when compared to older or younger counterparts (Figure 2) (2). Since then, the AYA focused literature has been growing to reflect an emerging specialty of cancer practice (2-8), as AYA patterns of incidence, survival, and adherence to recommended cancer-related follow-up care (CRFC) increasingly distinguish them from older adults and children. THE NEED FOR MELANOMA-SPECIFIC STUDY AMONG YOUNG PEOPLE Despite the overall growth in AYA literature, a lack of melanoma-specific study has persisted. Instead, AYA literature on specific cancer types has consisted of comparatively more common AYA cancers, such as lymphomas and breast cancer. Yet the growing burden of AYA melanoma and the recent call to action on skin cancer by the US surgeon general (9), demands the gap in knowledge around AYA melanomas be prioritized to advance treatment and improve melanoma outcomes to reduce the impact of this potentially lethal disease. Melanoma rates increased among AYAs for three decades (10-13) with little intervention or greater study. There is some suggestion that these rates have leveled off, or perhaps may have decreased slightly (14). Continued evaluation of melanoma rates among AYAs in the future is needed to provide a more robust picture of AYA trends in melanoma. In spite of the potential recent decreases, melanoma remains one 4 of the top 5 most common cancers among AYAs and at ages 20-29, melanoma is the second most common cancer (15). Existing literature predominantly represents patients well over age 50, as the mean age of melanoma is 64 years at diagnosis. Unfortunately, evidence from studies of older adults is assumed to be relevant for much younger patients. As far as risk factors are concerned, we know that light hair, skin or eye color, having skin that burns easily and/or has a large number of nevi (moles), as well as family history of melanoma, can all predispose a person to increased risk of melanoma (16-19), whether young or old. Sunburns in childhood or adolescence have only been well studied for increased risk of melanoma in older adults. In older adults, ultraviolet (UV) exposure is thought to increase the risk of melanoma via an accumulation of DNA-related damage to the skin that has occurred over many decades of life. UV is also thought to work via a dual pathway (20, 21), meaning both chronic and intermittent exposures have been implicated in the development of melanoma. For example, chronic, high levels of UV accumulated over a lifetime of occupational-related ambient exposure may lead to a melanoma for some people, while intermittent, high-intensity exposure from recreational leisure activities like skiing, boating, and sunbathing may lead to a melanoma for others. While intermittent exposure has been linked more often to melanomas in adults (22, 23), the relationship between either ambient or intermittent exposure (particularly during early life) in relation to childhood or AYA melanoma has been less well studied. 5 Melanoma survival trends and patient characteristics are currently limited to demographics (age groups, sex, broad race groups) and some tumor details, including stage, thickness, or anatomic location. In the few reports on melanoma survival among AYAs, the focus was overwhelmingly on non-Hispanic whites or national level data. Given the differences in distributions of race/ethnicity and sun exposure behaviors across the nation, regional studies including diverse populations may allow greater insight for subgroups and aid in identifying underlying disparities. During survivorship from melanoma, the risk of a second primary melanoma is much higher for persons diagnosed under age 30 than at older ages. Cancer-related follow-up care offers an important opportunity to conduct professional skin exams, review sun- protective behaviors, and detect any subsequent melanoma as early as possible, since late stage detection is likely to have a poor outcome. Foundational AYA literature from Bleyer et al highlights the challenge of getting young people to adhere to treatment and medical advice. Among young survivors of melanoma, we know little about who gets appropriate cancer-related follow-up care, or what types of barriers may have the most detrimental impact on getting follow-up care. The need for more attention to melanoma in AYAs was explicitly identified in the AYAO- PRG report as one of three top priority areas for cancer control and prevention, noting efforts were especially needed in the development of school curricula to reduce skin cancer risk, behavioral research aimed at reducing sun exposure, and studies enabling identification of AYAs at potentially high risk of skin cancers (2). That report also noted 6 concern regarding inconsistencies in delivery of care to AYAs (i.e. pediatric or adult clinical settings) and lack of clarity around who should provide information and education regarding preventive (skin-protective) behaviors. Both pose additional challenges to optimal cancer control activities. 7 OVERVIEW OF STUDIES Creating dedicated AYA programs will be critical if AYA needs are to be adequately addressed, but a larger evidence base of information specific to cancer types affecting AYAs is needed to optimize implementing cancer-specific approaches within a dedicated AYA clinical program. Thus, the overarching goal was to conduct melanoma-specific studies to inform the refinement of strategies for prevention, treatment, and optimal care during survivorship from melanoma. Study 1 - AYA Melanoma Survival: A population-based study comparing children, AYAs, and older adults Using data from the California Cancer Registry (CCR) in 1988-2014, cases of melanoma diagnosed at ages 0-64 years were identified. Broad age groups were defined as children (0-14 years), AYAs (15-39 years), and older adults (40-64 years) to facilitate age-specific estimates and comparisons between the groups. Kaplan-Meier (KM) analysis was used to generate 5-year survival probability and Cox regressions were used to generate hazard ratios. Analyses were further stratified by gender, when sample size allowed. Study 2 - Identifying Early Life Risk Factors: A case-control study of melanoma diagnosed before age 30 Data from the CCR in 1988-2013 and California state birth registry records in1983-2011 were used to identify cases of melanoma and controls, respectively. The data from CCR were linked previously to the birth records and UV values were assigned based on the mother’s residential address on the birth record. The present study evaluated the 8 previously linked data by using logistic regression to determine if birth weight or higher birthplace UV was associated with increased odds of developing melanoma before 30 years of age. Study 3 - Melanoma Survivorship in Young Adults: A survey study to identify barriers to cancer-related follow-up care. Cancer Surveillance Program (CSP) data in 1996-2010 was used to identify melanomas diagnosed at ages 0-24 years in Los Angeles County. A melanoma-specific survey was incorporated into the parent Project Forward survey, which captured information about the last follow-up visit, any barriers to seeing a doctor in the past year, and skin care behaviors. Logistic regression was used to determine the association between lack of recent cancer-related follow-up care and number or type of barriers. 9 FIGURE 1.0.1. MAJOR MILESTONES FOR AYA ONCOLOGY IN NORTH AMERICA (14, P. 2). FIGURE 1.0.2. THE GAP IN RELATIVE 5-YEAR SURVIVAL FOR AYAS WITH INVASIVE CANCER (2, P. 5). 10 CHAPTER 1 REFERENCES 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA: a cancer journal for clinicians. 2016;66(1):7-30. 2. Adolescent and Young Adult Oncology Progress Review Group (AYAO PRG). Closing the Gap: Research and Care Imperatives for Adolescents and Young Adults with Cancer (NIH Publication No. 06-6067). Department of Health and Human Services, National Institutes of Health, National Cancer Institute, and the LIVESTRONG Young Adult Alliance Bethesda, MD; 2006. 3. Bleyer A, O'leary M, Barr R, et al. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. 2006. 4. Bleyer A, Barr R, Hayes-Lattin B, et al. The distinctive biology of cancer in adolescents and young adults. Nature Reviews Cancer. 2008;8(4):288-98. 5. Sender L, Zabokrtsky KB. Adolescent and young adult patients with cancer: a milieu of unique features. Nature Reviews Clinical Oncology. 2015. 6. Thomas DM, Albritton KH, Ferrari A. Adolescent and young adult oncology: an emerging field. J Clin Oncol. 2010;28(32):4781-2. (doi: 10.1200/JCO.2010.30.5128 [doi]). 7. Shaw PH, Reed DR, Yeager N, et al. Adolescent and Young Adult (AYA) Oncology in the United States: A Specialty in Its Late Adolescence. J Pediatr Hematol Oncol. 2015;37(3):161-9. (doi: 10.1097/MPH.0000000000000318 [doi]). 11 8. U.S. Department of Health and Human Services. The Surgeon General's Call to Action to Prevent Skin Cancer. Washington, DC: U.S. Dept of Health and Human Services, Office of the Surgeon General; 2014. 9. Wong JR, Harris JK, Rodriguez-Galindo C, et al. Incidence of childhood and adolescent melanoma in the United States: 1973-2009. Pediatrics. 2013;131(5):846-54. (doi: 10.1542/peds.2012-2520 [doi]). 10. Austin MT, Xing Y, Hayes-Jordan AA, et al. Melanoma incidence rises for children and adolescents: an epidemiologic review of pediatric melanoma in the United States. J Pediatr Surg. 2013;48(11):2207-13. 11. Purdue MP, Freeman LE, Anderson WF, et al. Recent trends in incidence of cutaneous melanoma among US Caucasian young adults. J Invest Dermatol. 2008;128(12):2905-8. (doi: 10.1038/jid.2008.159 [doi]). 12. Linos E, Swetter SM, Cockburn MG, et al. Increasing burden of melanoma in the United States. J Invest Dermatol. 2009;129(7):1666-74. 13. Barr RD, Ries LA, Lewis DR, et al. Incidence and incidence trends of the most frequent cancers in adolescent and young adult Americans, including “nonmalignant/noninvasive” tumors. Cancer. 2016. 14. Bleyer A, Barr R, Ries L, et al, eds. Cancer in Adolescents and Young Adults. Second ed. Cham, Switzerland: Springer International Publishing, 2017. 15. Cockburn M, Hamilton A, Mack T. The simultaneous assessment of constitutional, behavioral, and environmental factors in the development of large nevi. Cancer Epidemiol Biomarkers Prev. 2007;16(2):200-6. (doi: 1055-9965.EPI-06-0273 [pii]). 12 16. Cockburn M, Black W, McKelvey W, et al. Determinants of melanoma in a case– control study of twins (United States). Cancer Causes & Control. 2001;12(7):615-25. 17. Rhodes AR, Weinstock MA, Fitzpatrick TB, et al. Risk factors for cutaneous melanoma: a practical method of recognizing predisposed individuals. JAMA. 1987;258(21):3146-54. 18. Thomas NE, Edmiston SN, Alexander A, et al. Number of nevi and early-life ambient UV exposure are associated with BRAF-mutant melanoma. Cancer Epidemiol Biomarkers Prev. 2007;16(5):991-7. (doi: 16/5/991 [pii]). 19. Whiteman DC, Watt P, Purdie DM, et al. Melanocytic nevi, solar keratoses, and divergent pathways to cutaneous melanoma. J Natl Cancer Inst. 2003;95(11):806-12. 20. Lotti T, Bruscino N, Hercogova J, et al. Controversial issues on melanoma. Dermatologic therapy. 2012;25(5):458-62. 21. Bataille V. Sun exposure, sunbeds and sunscreens and melanoma. What are the controversies? Curr Oncol Rep. 2013;15(6):526-32. 22. Østerlind A, Tucker M, Stone B, et al. The Danish case-control study of cutaneous malignant melanoma. II. Importance of UV-light exposure. International Journal of Cancer. 1988;42(3):319-24. 13 CHAPTER 2: MELANOMA INCIDENCE AND SURVIVAL PRIOR TO AGE 65 14 ABSTRACT Melanoma is among the top 5 most common cancers among adolescents and young adults (AYAs), who have unique underlying biology and face a distinct set of life course challenges that may affect incidence and survival, distinguishing them from younger or older patients. Using a population-based registry study of melanomas diagnosed <age 65 in California, we identified differences in incidence and survival among AYAs versus younger/older patients. Although female AYAs had higher incidence, their male counterparts had poorer survival. Probability of 5-year survival generally improved over time for both AYAs and older adults, except with remote stage melanoma, where AYAs did not experience the same improvements as older adults. This may be related, in part, to older adults having greater access to immunotherapies and clinical trials, which have tremendously advanced the treatment of late stage melanoma. Access to clinical trials for AYAs as an age group, regardless of cancer type, remains an area of concern, complicated by AYA exclusion from eligible age ranges in these trials, their tendency to be seen in community settings rather than at comprehensive cancer centers where trials are conducted, and by relatively low participation of AYAs in clinical trials, even when available to them. 15 BACKGROUND For more than three decades, melanoma, the deadliest form of skin cancer, has been one of the most common cancers diagnosed among adolescents and young adults(1-3) (AYAs; defined as 15-39 years of age at diagnosis)(4, 5) Globally, melanoma is the 12 th most common AYA cancer, but in the US, melanoma is the 3 rd most common AYA cancer and females are at greater risk than males in this age group (6) (2 nd for males, rate 5.43 per 100K and 3 rd for females, rate 9.02 per 100K; 2000-2011) (7). While there is some suggestion that melanoma rates among AYAs in the US declined by about 1% annually in 2000-2011 (3), it is unclear if these trends have persisted in recent years, or if similar trends exist in California, where some of the highest rates of melanoma in the world occur (8) among its more than 40 million residents. More than 1000 new AYA melanomas occur annually in this state. AYAs reflect a highly heterogeneous, transitional stage of life involving aspects of cognitive, emotional, and financial development that may impact decision-making related to seeking or adhering to treatment. They may also be particularly sensitive to the cost and financial implications of treatment, because they may not yet have a stable wage, health insurance, or employment that allows them to take time off for treatment and recovery. Development of social independence from parent care may also influence access to and decision making around obtaining health care coverage. Because AYA oncology (AYAO) is an emerging specialty (9), melanoma-specific studies for this age group are lacking. Instead, the most common AYA cancers are often combined together, then described by broad demographic and clinical characteristics, and also highlight 16 much lower trial participation and treatment adherence than seen in younger or older patients. While this approach has served quite well for the purpose of distinguishing the unique needs of AYAs from other age groups (10), and thoroughly justifies AYAO as an important clinical sub-specialty, there is now an opportunity to further expand our understanding of AYA cancers by employing site-specific study strategies. Melanoma is a common AYA cancer (top 5 for AYA, top 10 for older adults), yet studies of older adults continue to dominate the literature. For older adults, where the mean age at diagnosis is 64 years, the key risk factors are sunburn and lifetime accumulation of intermittent and/or chronically high ultraviolet (UV) exposure (11-13), along with family history, genetics, and constitutional factors, such as having large numbers of nevi or light hair, eye, or skin color (14-16). In younger patients, family history and constitutional factors are also known to play a role, but aside from tanning bed use (17- 22), little else is known about risk factors, including the role of UV exposure, which among adults, involves DNA damage in the skin that was accumulated over many decades. Among children, pre-existing nevi, family history, and genetic factors may play a more distinctive role than at older ages (23, 24), owing to the short time period available for damage from UV exposure, thought this too remains understudied. Survival studies among AYAs with melanoma are also rare and inadequate, lacking the details provided for older patients. From survival studies in older adults, we know that melanoma-specific characteristics, such as Breslow thickness (25), ulceration, and histological subtype (25), as well as stage at diagnosis (26), influence survival, but the 17 impact of these factors among AYAs has not yet been examined with this level of detail, limiting our understanding of potentially important subgroups, clinical characteristics, or age-specific windows associated with higher risk of death among AYAs with melanoma. Sex differences have also become apparent in older adult studies, with males experiencing poorer survival) (27). This is less clear studied among AYAs; given the difference in AYA melanoma occurrence by sex (AYA females get more than AYA males), a closer examination of sex differences in survival is warranted. Limited information is available describing changes in survival for AYAs, and conflicting trends are reported among the handful of existing studies based on data from the Surveillance, Epidemiology, and End Results (SEER) program. One study focused only on ages 15-19 years versus children <15 years, also finding no lag in survival gains over time (1975-1995, SEER 9 data; observed/absolute survival) (28), which was supported in a different study which evaluated the entire AYA age range together, also finding no difference in the changes in melanoma survival for AYA versus older patients (1992- 2005, SEER 13 data; cause-specific survival)(26), however, a third study suggested an opposite trend - that AYAs experienced less improvement in survival from melanoma over time (1975-2002, SEER 9 data; conditional survival based on relative survival rates; employs the concept of conditional probability to account for changing hazard rates over time) (29). Including data after 2005 may help clarify this picture. 18 STUDY PURPOSE & AIMS To improve understanding of survival among AYAs with melanoma, data from the California Cancer Registry (CCR), which has a high level of diversity in demographic backgrounds, was used to describe survival among AYAs with a level of detail for melanoma not previously reported in the literature. This study aimed to evaluate survival among AYAs with melanoma (15-39 years), comparing them to younger (<15 years) and older (40-64 years) patients, with the following specific goals: 1. Describe demographic, tumor, and incidence characteristics for AYA melanomas 2. Determine the 5-year survival probabilities for AYA melanomas and characterize factors most strongly associated with poorer survival outcomes, hypothesizing that: a. Higher stage and/or greater tumor thickness will be associated with lowest AYA survival b. Minority race/ethnicity and lower SES will be associated with poorer AYA survival rates than non-minority race and high SES, respectively c. Males will have poorer survival than females across all age groups, but this effect will be more pronounced among AYA patients 3. Estimate the hazard ratio (HR) for risk of dying among AYAs with melanoma and identify factors most strongly associated with risk of death, hypothesizing that: 19 a. Higher stage at diagnosis and tumor thickness will be associated with larger HRs for AYAs than observed in younger or older patients b. Lower SES will be associated with increased risk of death in a dose- dependent manner c. Males will have higher HRs than females across demographic and clinical characteristics. 4. Estimate changes over time in 5-year survival probability for AYAs, hypothesizing that: a. Lower gains in survival will occur for AYAs compared to younger/older patients b. Lower gains will be more pronounced after 2000 (introduction of immunotherapies) 20 MATERIALS & METHODS POPULATION & DATA SOURCES Cases of cutaneous melanoma (herein “melanoma”) (total n=261,735) occurring in 1988-2014 were identified from the CCR (release date October 2016), which is the State of California’s population-based cancer registry. CCR data are ≥95% complete in accordance with standards set forth by the Centers for Disease Control’s (CDC) National Program of Cancer Registries (NPCR) and the SEER program (30). Melanomas were selected using site codes C440-449 (skin) and histology codes 8720- 8780 from The International Classification of Disease for Oncology, Third edition (ICD-O- 3). All demographic (age, sex, socioeconomic status (SES), race/ethnicity) and clinical/tumor characteristics (date of diagnosis, last follow-up date, vital status, stage at diagnosis, Breslow’s depth of invasion (tumor thickness), ulceration, anatomic location, histologic subtype) were obtained from registry variables. To improve statistical power and interpretability of the results, covariates were categorized as noted below. Age at Diagnosis Inclusion/Exclusion This study was restricted to melanomas diagnosed at <65 years of age (n=141,072 cases). For analytic purposes, three main age groups (children ages 0-14; AYAs ages 15- 39, older adults ages 40-64) were examined. Individuals aged ≥65 years of age at melanoma diagnosis were not included in the current study, because their survival differs from persons <65 years (Figure 1) and may be more strongly influenced by a set 21 of age-related factors that are distinct from factors involved for individuals aged <65 years at diagnosis (31). FIGURE 2.0.1. MELANOMA-SPECIFIC SURVIVAL BY AGE AT DIAGNOSIS (26, P. S78.E6). For example, existing and/or accumulating comorbidities, which may contribute to (or exacerbate) lower immune response or greater frailty; restricted income due to retirement status, which can influence healthcare access and ability to pay for clinical visits or treatments; deteriorating vision, which can impair safe access to transportation to/from clinical care; availability of social support and/or loss of a spouse, which can influence aspects of mental health (i.e. depression) and/or impair self-efficacy for seeking and adhering to treatment and medical advice. Further, all of these factors may impact recovery from intensive medical treatment or surgical procedures related to 22 either melanoma or comorbidities that exist with greater prevalence over the age of 65 (i.e. heart disease, hypertension, stroke, diabetes). Because the factors noted above are not routinely captured in registry data, analysis adequately accounting for such factors would be hampered. A study including linkage to Medicare/Medicaid data may be more appropriate for evaluation of survival patterns among individuals ≥65 years of age, who were excluded from the present study. Survival Time & Vital Status Observed survival was the primary outcome, which reflects survival from all causes of death after (melanoma) diagnosis and was used to describe differences between younger and older patients. Another method available is relative survival, which reflects a ratio of observed (all cause) survival among in a given cohort of melanoma patients versus expected (all cause) survival in a comparable cancer-free population. However, obtaining a comparable cancer-free population is challenging, thus the calculation for relative survival is instead based upon expected life tables for determination of expected survival, which assumes the number of cancer deaths is very small. Thus, observed survival was chosen to present observed survival statistics for their ease of interpretability and because the level of detail provided herein for observed survival has not yet been well described previously in the literature. All survival time available for cases of melanoma diagnosed over the study period (1988- 2014) was used. Survival time in months was obtained from the registry variables, which had a cut-off date of December 31, 2014 for calculation of survival time. Methods for 23 calculation have been described; briefly, because not all registries are able to actively follow-up all cases, it is necessary to use a survival time variable that presumes a subject is still alive if they are not known to be deceased at the last complete death clearance. There were no cases with entirely missing dates of diagnosis or follow-up (at a minimum, the year was available). When either the diagnosis date or follow-up date were incomplete, the following procedure was applied to “impute” missing data for both diagnosis and follow-up dates: (1) If year and month, but not day, were available, then the midpoint of that month was assigned or (2) If year, but not month or day, was available, then a midpoint of that year was assigned. This method was followed for continuity in the use of registry data to estimate survival in accordance with guidelines set forth by SEER and the North American Association of Central Cancer Registries (NAACCR). Demographic Characteristics Both males and females were included in this study. RACE AND ETHNICITY. A combined race/ethnicity variable was used, which classified individuals as non-Hispanic white (NHW), Hispanic White, non-Hispanic black, Asian/Pacific Islander, non-Hispanic Native American, Middle Eastern, or other/unknown. Additionally, a dichotomous variable defined as Hispanic/Latino or non-Hispanic/Latino (race not inclusive) was based on the NHIA registry variable. A large proportion of California’s population is of Hispanic/Latino heritage, making it important to identify the potential disparities that may exist; this may enhance clinical and public health approaches aimed at reducing disparities in the future. SES. Quintiles of SES were generated by the CCR based on census tract data;(32) 24 for individuals, geocoded data based on the address at diagnosis was obtained, and then an appropriate census block was identified and matched with the associated SES quintile (Q1-Q5) value for that census block (Q1=lowest, Q2=low-middle Q3=middle, Q4=middle- high, Q5=highest). YEAR OF DIAGNOSIS. To evaluate trends over time in this study, we categorized year of diagnosis into 5-year intervals, beginning with the most recent years; there were 27 years of data included in this study, so only two years of data were available for inclusion in the oldest time period. Tumor Characteristics Tumor characteristics were based on information from the registry, which is confirmed by histopathological report. STAGING. Stage at diagnosis was collapsed into the following categories for analysis: In situ (Stage 0), Localized (Stages I-II), Regional (Stage III), and Remote (Stage IV; also known as distant or metastatic) and unknown/not abstracted. TUMOR THICKNESS. Common clinical cutpoints for tumor thickness were used and defined as ≤1.00mm, 1.01-2.00mm, 2.01-4.00mm, >4+mm, and unknown (33). Thickness represents Breslow’s depth, which was determined by vertical measurement of the lesion (in millimeters, mm) “from the top of the granular layer to the deepest point of involvement,” in accordance with the Surveillance, Epidemiology and End Results (SEER) Program staging guidelines. Due to variations over time in reporting requirements and collection procedures, only a subset of cases had information available (51%; n=71,935) and tumor thickness in millimeters (mm) was obtained by combining variables that captured this information during different years. Unknown thickness data was included in all models, as there is evidence that these tumors tend to 25 represent advanced stages of disease, thus exclusion may lead to bias.(34) ULCERATION. Ulceration status was defined for analysis as present, not present, unknown, or missing. Ulceration was determined to be present when an intact epidermis covering the tumor was indicated to be lacking via histopathology report (33). Due to variations over time in reporting requirements and collection procedures, only a subset of cases had information available (73%; n=102,842) and ulceration status was obtained by combining variables that captured this information during different years. Unknown and missing histology categories were included in all models to assess their impact on survival outcomes. HISTOLOGICAL SUBTYPE. Histological subtypes were categorized as superficial spreading melanoma (SSM), nodular melanoma (NM), Lentigo maligna (LM), acral lentiginous melanoma (ALM), giant pigmented nevus melanoma (GPM), other (rarer subtypes), and unknown or not otherwise specified (NOS). Due to variations over time in reporting requirements and collection procedures, only a subset of cases had information available (42%; n=59,849). Unknown histology data was included as its own category in all models to assess its impact on survival outcomes. ANATOMIC LOCATION. Anatomic location was categorized as face, lip/external ear, scalp/neck, arms/shoulders/hands, trunk, legs/hips/feet, or unknown. Unknown location represented only 3% of the data. 26 STATISTICAL ANALYSIS Kaplan-Meier (KM) survival analysis was used to estimate the 5-year cumulative probability of observed survival in broad (pediatric, AYA, and adult) and to generate survival plots as visual aids to demonstrate important differences in the survival probabilities. Log-rank tests were used to detect statistically significant differences in the 5-year survival probability (alpha level=0.05). Cox Proportional Hazards regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for risk of death by age group and allowed for confounding adjustment. Criteria for presence of confounding was >10% change in the beta estimates; confounders were selected for inclusion in the final model based on these criteria and a priori knowledge. Potential effect modification by age group, sex, and time period was considered with stratified models, when sample size allowed. All survival analyses were conducted with SAS version 9.4 (SAS Institute, Inc., Cary, North Carolina). Age-adjusted incidence rates were generated using SEER*Stat software (Version 8.3.4). This was a records-based study conducted by the staff of the Cancer Surveillance Program (CSP), which is the population-based cancer registry of Los Angeles County, and reflects ongoing efforts of the CSP to monitor melanoma survival trends. 27 RESULTS In Tables 1 and 2, the distribution of demographic and tumor characteristics is provided by age group (children, AYAs, and older adults). There were a total of 141,072 cases of melanoma diagnosed in California in 1988-2014. Patients in in the AYA group accounted for 21% of all cases; less than 1% occurred among children. Females accounted for more of the cases at younger ages (52% of children were female, 61% AYAs, and 44% adults). Across all ages, melanomas occurred most frequently among NHWs and higher levels of SES. However, among children, Hispanics accounted for a larger proportion of cases (22%) than was observed in either AYAs (8%) or older adults (5%). This was also true for Asian/Pacific Islanders (5% children, 1% in either AYAs or older adults). For children and AYAs, the proportion of cases peaked during the periods 1995-1999 and 2000-2004, and may have plateaued (especially for children) thereafter. For older adults, the number of cases appeared to increase across all time periods. While the majority of melanomas were detected at the localized stage in all age groups (56% children, 64% AYAs, 54% older adults), children had greater proportions of late stage disease (regional/remote) (18%) than AYAs (7%) or older adults (7%). The same was true when tumor thickness was ≥2mm (19% children, 6% AYAs, 7% older adults). Superficial spreading melanoma (SSM) was the most common subtype (16% children, 32% AYAs, 24% older adults) after unknown/NOS subtypes, most of which may include SS; the NOS histology reflects many of the tumors that occurred prior to the discovery and use of a variety of subtypes in histology codes. Nodular melanoma (NM) was the 3 rd most common subtype in children (6%) and AYAs (4%), while lentigo maligna melanoma 28 (LMM) was the 3 rd most common subtype in older adults (10%). Across all ages, the trunk was the most common anatomic site (26% children, 38% AYAs, 33% older adults), followed by the limbs (arms/hands, legs/feet). AGE-ADJUSTED INCIDENCE RATES (AAIR) Table 3 provides melanoma rates by age group (expressed per 100,000 for AYAs and adults; expressed per million for children). Overall, rates were 2.1 per million in children, 8.4 per 100,000 in AYAs, and 42.2 per 100,000 in older adults. Sex differences were not apparent until ages 15+ years; female AYAs have higher rates than males (10.4 vs. 6.5, respectively), while older adult males have higher rates than their female counterparts (48.3 vs. 36.5, respectively). By race and ethnicity, among AYAs, females had consistently higher rates across race or Hispanic ethnicity. Among older adults, males had higher rates across most race/ethnic groups, with the notable exception of Hispanics (Hispanic whites: 1.6 vs. 2.3 for males vs. females; All Hispanics: 6.7 vs. 9.7 for males vs. females). Rates in older adult Hispanic males did not exceed the rates of their female counterparts until after age 60, nearly a decade later than was observed in their NHW counterparts (data not shown). Slightly higher rates were also observed among female versus male Asian/Pacific Islanders in the older adult group. By SES, rates increased with higher levels for all age groups, but this appeared to be stronger in AYA females and in older adult males. Over the time period observed, AYA rates were highest in the late 90’s (9.3 overall, 7.3 males, 11.4 females) and the early 2000’s (9.0 overall, 6.8 males, 11.4 females), with 29 slightly lower rates thereafter. Among older adults, rates steadily increased over the entire time period, perhaps plateauing for males after 2000, but continuing to increase among females. 5-YEAR OBSERVED SURVIVAL PROBABILITY (KAPLAN MEIER ESTIMATES) Table 4 provides melanoma survival probabilities by age group for demographics; corresponding Kaplan-Meier survival curves are found in Appendix A. Overall 5-year survival was 96% for children, 94% for AYAs, and 90% for older adults. By sex, there appears to be a survival disadvantage for males that persists across demographic and tumor characteristics, particularly in AYAs and older adults. By race/ethnicity, minorities (except for those of Middle Eastern descent) tended to have worse survival than NHWs. Among AYAs, the survival disparity was notable for black and Asian patients (81% in blacks, 88% in Asians, 93% in NHWs); this was more pronounced among older adults (73% in blacks, 79% in Asians, 90% in NHWs). By SES, survival was better at higher levels of SES within each age group (low/high SES: children 96%/99%, AYAs 89%/96%, older adults 83%/94%) and for both males and females alike (although less consistent among children). Over the time period observed (Figure 2), survival generally improved, with 5-year survival >90% for all age groups and both sexes in the most recent time period of 2010- 2014. When restricting to remote stage cases, survival improved over the time period observed for older adults, but was not apparent for AYAs (Figure 3). 30 FIGURE 2.0.2. CHANGE IN 5-YEAR SURVIVAL (ALL STAGES COMBINED) AMONG PATIENTS WITH MELANOMA IN CALIFORNIA, 1988-2014 FIGURE 2.0.3. CHANGE IN 5-YEAR SURVIVAL (REMOTE STAGE ONLY) AMONG PATIENTS WITH MELANOMA IN CALIFORNIA, 1988-2014 88% 97% 95% 100% 92% 95% 90% 91% 94% 95% 95% 95% 86% 87% 90% 91% 92% 93% 0% 20% 40% 60% 80% 100% 1988-89 1990-94 1995-99 2000-04 2005-09 2010-14 5-Year Survival Probability (%) Time Period (Year of Diagnosis) Child (n=420) AYA (n=29,220) Adult (n=111,432) 11% 20% 19% 15% 22% 14% 10% 15% 13% 15% 19% 25% 0% 5% 10% 15% 20% 25% 1988-89 1990-94 1995-99 2000-04 2005-09 2010-14 5-Year Survival Probability (%) Time Period (Year of Diagnosis) AYA (n=29,220) Adult (n=111,432) 31 Table 5 provides melanoma survival probabilities by age group for tumor characteristics; corresponding Kaplan-Meier survival curves are found in Appendix A. By stage, the male survival disadvantage was more pronounced at higher stages. Among AYAs, male 5-year survival from regional disease was 64% compared to 75% for AYA females; remote disease survival for AYA males was 15% compared to 19% in AYA females. Among older adults, male 5-year survival from regional disease was 55% compared to 63% in older adult females; remote disease survival for older adult males was 15% versus 19% in older adult females. Among children, the 5-year survival disadvantage for males was most apparent for remote disease (32% males, 43% females). By thickness category, survival decreased as thickness increased, falling below 65% for tumors ≥4mm thick. At all thickness levels, there was a male survival disadvantage observed, particularly for AYAs with tumors 2.01-4.00mm (11% lower; 86 vs. 75%, males and females, respectively) and ≥4.01mm (12% lower; 56 vs. 68%, males and females, respectively) and for older adults with tumors≥4.01mm (11% lower; 51 vs. 62%, males and females, respectively). Among children, the 5-year survival disadvantage for males was most apparent with tumors 2.01-4.00mm (8% difference; 92 vs. 100% males and females, respectively. By ulceration status, survival was worse across all age groups, but was most detrimental to AYAs when present versus not (71% vs. 96%, respectively) and adults (63% vs. 93%, respectively) and the male disadvantage persisted here as well. By histology, SSM had the best prognosis (>95% for all ages), while NM had the poorest prognosis in AYAs (77%) or older adults (68%) and GPM had the poorest prognosis in children (92%, although this lowers to 88% for male children). By anatomic site, tumors 32 of unknown/NOS origin had the poorest survival across all ages (children 51%, AYAs 51%, adults 41%; for males, these drop to 29% in children, 45% AYAs, 37% older adults), which is likely to reflect advanced disease, when it may be more difficult to determine the site of origin; otherwise, melanomas occurring on the scalp or neck had the lowest survival across all ages (88% in children, 89% in AYA, 86% in older adults), with some variation by sex. HAZARD RATIOS FOR RISK OF DEATH (HR; COX HAZARD REGRESSION ESTIMATES) Unadjusted and adjusted HRs by age group are presented in Table 6 (for demographics) and Table 7 (for tumor characteristics). Among children in the unadjusted models, race/ethnicity, stage, and ulceration were associated with statistically significant increases in the risk of death. Compared to NHW children, Hispanic children had 2.51 times the risk of death (HR: 2.51; 95%CI: 1.07, 5.89), while Asian children had 7.90 times the risk of death (HR: 7.90; 95%CI: 2.98, 20.96). Only remote stage had a statistically significantly effect on the outcome (HR: 47.76; 95%CI: 10.52, 216.78), when compared to in situ. Ulceration presence was associated with elevated risk of death, but the confidence interval was wide (HR: 2.61; 95%CI: 0.57, 11.91); only “unknown” ulceration status was significant (HR: 10.86; 95%CI: 3.90, 30.24), which may include some advanced stage disease. Increased risk was also observed for nodular and GPN melanomas, but the CIs were wide and included the null. No statistically significant associations for other characteristics were observed. 33 Unadjusted models in AYAs and older adults shared similar patterns for increased risk of death; all characteristics evaluated were statistically significant, but magnitudes tended to be larger in AYAs than older adults, especially in tumor characteristics, such as stage. Males had 2.33 times the risk of death of females (AYA HR: 2.33; 95%CI: 2.16, 2.51; Older Adult HR: 1.77; 95%CI: 1.71, 1.82), and minorities tended to have higher HRs compared to NHW patients, notably for black patients (AYA HR: 2.31; 95%CI: 1.50, 3.55; Older adult HR: 2.29; 95%CI: 1.91, 2.75), but also for Hispanics (AYA HR: 1.17; 95%CI: 1.03, 1.33; Older adult HR: 1.27; 95%CI: 1.19, 1.35) and Asians (AYA HR: 1.39; 95%CI: 1.02, 1.88; Older adult HR: 1.64; 95%CI: 1.42, 1.90). All levels of SES below the highest quintile had higher risk of death, with the greatest disparity in the lowest SES quintile (AYA HR: 2.23; 95%CI: 1.95, 2.54; Older adult HR: 2.25; 95%CI: 2.14, 2.38). Compared to the late 80s, risk of death in 2010-2014 was lower (AYA HR: 0.77; 95%CI: 0.64, 0.93; Older adult HR: 0.82; 95%CI: 0.76, 0.88). Compared to an in situ tumor, localized stage had 3.29 times the risk of death for AYAs (HR: 3.29; 95%CI: 2.82, 3.83) and 1.65 times the risk of death among older adults (HR: 1.65; 95%CI: 1.59, 1.71), while remote stage had nearly 130 times the risk among AYAs (HR: 129.29; 95%CI: 108.90, 153.50) and 36 times the risk among older adults (HR: 35.7; 95%CI: 33.87, 37.63). Having tumor thickness of 4+mm had more than 11 times the risk of a thin (<1mm) tumor among AYAs (HR: 11.06; 95%CI: 9.63, 12.70) and had 6.68 times the risk among older adults (HR: 6.68; 95%CI: 6.32, 7.06). Presence of ulceration had more than 6 times the risk of non- ulcerated tumors among AYAs (HR: 6.02; 95%CI: 5.28, 6.87) and had 4.63 times the risk among older adults (HR: 4.63; 95%CI: 4.40, 4.87). 34 Among children after adjustment (Tables 6 and 7), race/ethnic disparities persisted; Hispanic children were at 4.88 times the risk of death (HR: 4.88; 95%CI: 1.19, 19.94) and Asian children were at 7.50 times the risk of death (HR: 7.50; 95%CI: 1.70, 33.10), compared to NHW children. Remote stage was the only stage with statistically significant increased risk of death (HR: 17.52; 95%CI: 2.20, 139.18). Risk of death remained increased but not statistically significant for thicker, ulcerated, and nodular tumors. Among AYAs after adjustment (Tables 6 and 7), males remained at a survival disadvantage, with 53% higher risk of death compared to females (HR: 1.53; 95%CI: 1.42, 1.65), Hispanics were at 20% reduced risk compared to NHW (HR: 0.80; 95%CI: 0.71, 0.90), and lowest SES were at 50% higher risk than those in the highest SES (HR: 1.50; 95%CI: 1.30, 1.72). Over time, risk of death decreased by as much as 50% when compared to the earliest time period (late 1980’s), however some of this improvement may have been lost in the most recent time period (2010-2014 HR: 0.67; 95%CI: 0.54, 0.83). All stages higher than in situ remained associated with increased risk of death; localized had 3.82 times the risk of in situ tumors (HR: 3.82; 95%CI: 3.20, 3.83), regional had more than 11 times the risk (HR: 11.22; 95%CI: 9.16, 13.73), and remote was associated with 59 times the risk of death (HR: 59.07; 95%CI: 48.26, 72.31). Tumors thicker than 1mm had 2-3 times the risk of thin (<1mm) tumors, with those >4mm having the highest risk (HR: 3.33; 95%CI: 2.81, 3.96). The effect of ulceration was attenuated, but remained statistically significantly higher by 89% compared to no ulceration (HR: 1.89; 95%CI: 1.62, 2.20). Similarly, nodular (HR: 1.59; 95%CI: 1.39, 1.83) 35 and acral (HR: 2.07; 95%CI: 1.43, 2.99) effects were attenuated, but remained higher for risk of death when compared to SSM. Tumors located on the legs, hips, or feet had 20% lower risk of death (HR: 0.80; 95%CI: 0.66, 0.96), when compared to tumors located on the face. Among older adults after adjustment (Tables 6 and 7), males retained a 38% higher risk of death compared to females (HR: 1.38; 95%CI: 1.34, 1.43), and older adults of Black heritage remained at 52% increased risk of death compared to NHW counterparts (HR: 1.52; 95%CI: 1.27, 1.83). Those in the lowest SES had 79% higher risk of death compared to those in the highest SES (HR: 1.79; 95%CI: 1.70, 1.89), but all levels below the highest quintile had increased risk. Localized tumors had 98% higher risk of death compared to in situ (HR: 1.98; 95%CI: 1.86, 2.10), which increased in regional stage to 5.20 times the risk (HR: 5.20; 95%CI: 4.85, 5.59), and had 19.65 times the risk for remote stage (HR: 19.65; 95%CI: 18.30, 21.10). Tumor thickness of 4mm or greater was associated with 2.39 times risk of death (HR: 2.39; 95%CI: 2.24, 2.56) and ulceration vs. no ulceration carries a 68% greater risk of death (HR: 1.68; 95%CI: 1.59, 1.78). Nodular (HR: 1.46; 95%CI: 1.38, 1.55) and acral (HR: 1.35; 95%CI: 1.17, 1.55) tumors had the highest risk of death compared to SSM, and lentigo maligna’s effect became statistically significant in the adjusted model (HR: 1.19; 95%CI: 1.12, 1.27). A male survival disadvantage was observed among AYAs and older adults, but not children (Table 8). This disadvantage appeared to be stronger among AYAs than in older adults, and stratified models among AYAs were further explored (Table 9). Male AYAs 36 had a persistent survival disadvantage of 50% for Hispanics and 54% in non-Hispanics, 37-63% across SES levels. The disparity was highest in the most recent time period of 2010-2014, where males had more than 2 times the risk of death than their female counterparts (HR: 2.09; 95%CI: 1.48, 2.94). By stage, the disparity was highest at lower stages (up to 71% higher), remaining elevated at all stages, but no longer statistically significant at the remote stage (HR: 1.13; 95%CI: 0.92, 1.38) 37 DISCUSSION Melanoma is an aggressive and common AYA cancer, yet studies of survival in AYAs with melanoma are relatively rare. Of the little information that exists, the focus has been either entirely on NHW populations, or has come from broad, national trend reports aimed at championing the development of an AYA oncology specialty, owing to unique underlying biology and distinct distributions of cancer types occurring in this age group. As such, these reports acknowledge the limitations of existing literature, calling for an increase in site-specific study of cancers commonly affecting AYAs. Here, a detailed description of the characteristics related to survival among AYAs with melanoma is provided within the diverse population of California, where some of the highest melanoma rates in the world have been observed. The impact of melanoma-specific tumor characteristics was included and a narrative comparison of differences between AYAs and their older or younger counterparts is provided. For context, incidence in California was generated for all three age groups. AYA females had higher rates of melanoma than males (10.4 vs. 6.5 per 100,000 for females and males, respectively; 8.4 per 100,000 overall; 1988-2014). In the most recent five years (2010-2014), AYA females continued to have higher rates than males (9.2 per 100,000 and 5.5 per 100,000 for females and males, respectively; 7.3 per 100,000 overall; 2010- 2014). Compared to children and older adults, AYA rates were consistent with the literature (higher than children and lower than older adults). Risk of melanoma increased with age, which is established in the literature and is true of most cancer types. 38 Time Trend. AYA rates appear to have peaked during 1996-2000, with an AAIR of 9.3 per 100,000 (AAIR was 11.4 and 7.3 per 100,000 for females and males, respectively, all race/ethnicities) with little change through 2009 and a slight decrease in 2010-2014, with AAIR of 7.3 per 100,000 overall (9.2 and 5.5 per 100,000 for females and males, respectively). These rates of AYA melanoma in California were higher than national rates, which were 7.2 per 100,000 overall, 5.43 for males, and 9.02 in females in 2000- 2011. In Queensland, Australia, where the highest rates of melanoma in the world are observed (8), rates in AYAs were only available for ages 15-24,(7) and were 10.1 per 100,000 for invasive melanoma in 2006-2010 (11.5 and 8.8 per 100,000 for females and males, respectively) (35). While survival appears to have improved over time for both AYAs and older adults (Figure 2, all stages combined), the improvements among AYAs appear to have plateaued from 2000 onward, while older adults continued to experience incremental improvements. The effect of continued improvements in survival in older adults may be related, at least in part, to access to and benefits from immunotherapy clinical trials, which have resulted in tremendous improvements for late stage melanomas (36). If access to immunotherapy among older adults were an explanatory factor, we would expect to see greater survival from late stage melanoma among older adults, but not AYAs. Indeed, there appears to be some indication of this in Figure 3, which displays trends in survival over time for remote stage melanomas. Access to clinical trials for AYAs as an age group, regardless of cancer type, persists as an area of concern, complicated by their exclusion from eligible age ranges, their tendency to be seen at community rather than 39 comprehensive cancer centers where trials are conducted, and by their relatively low participation in clinical trials, even when available (37, 38). Overall, 5-year survival was generally high at ≥90% across the age groups (children: 96%, AYAs: 94%, older adults: 90%). Stage at diagnosis is a well-established prognostic factor and advanced melanomas have poor outcomes, with 5-year survival from remote tumors as low as 16-17% for AYAs and older adults, and just 37% in children. From adjusted Cox hazard regressions, the impact of higher stage appeared stronger among AYAs than in younger or older patients, evident even for localized tumors, which had HRs of 3.82, 1.98, and 1.13 for AYAs, adults, and children, respectively, compared to in situ melanoma. A similar pattern existed for more advanced stages, with AYAs having larger magnitudes than observed in younger or older patients. It is unclear if the reason for the higher HR in AYAs than observed in the older adults is related to more aggressive underlying disease, reflects delays in getting treatment or poor adherence to treatment, or a combination of these factors. Early detection of melanoma leads to higher 5-year survival, generally exceeding 90% for in situ and localized tumors, regardless of age at diagnosis. Thus, diligent screening, either by self-exam or clinician, is critical to early detection. Unfortunately for AYAs, lower screening for unusual skin growths, and lower suspicion of melanoma when such growths are detected, may occur because rates have historically been higher in older adults than AYAs (about 35/100k vs. 10/100k). As a result, there may be greater risk of having an AYA melanoma overlooked and detected at a later stage, when prognosis is poor. Identifying important subgroups or clinical characteristics of AYAs at greatest risk of dying after a melanoma diagnosis is needed to 40 improve our ability to identify high-risk patients to facilitate optimal clinical care and follow-up strategies. The untimely death of an AYA patient from a potentially preventable, late stage melanoma can be particularly tragic, as these young patients should otherwise be looking forward to an entire lifetime of hopes and achievements ahead. Sex Differences. A disparity in melanoma survival (observed and melanoma-specific) (39) by sex has primarily been described previously in older NHW adults (39-42), and to a lesser degree among AYAs. We looked at whether or not this would be true among AYAs, since females have a higher incidence of melanomas in this age group (10.4 vs. 6.5 per 100,000 for females and males, respectively; 8.4 per 100,000 overall; 1988-2014), whereas in older adults, males get more melanomas. Among AYA males with melanoma in California, a survival disadvantage was detected and found to persist across demographic and clinical characteristics. The disparity appeared to be slightly larger for AYAs (53% higher risk of death in males) than was observed for older adults (38% higher risk of death in males). Of the few studies found to examine AYA disparities in survival by sex for melanoma (27, 43-45), a similar survival disadvantage for males was reported, and one study included Hispanic whites, who were also found to experience worse survival than females (27, 43-45). In the current study, we observed a persistent sex disparity in hazard regression models stratified by race/ethnicity, stage, SES, and years of diagnosis. When stratified by stage, males remained at a disadvantage even for in situ and localized tumors. Although 41 this effect was diminished for remote stage melanomas, the literature suggests males are at greater risk of developing metastatic disease or experiencing relapse/progression after diagnosis (39-41) - intriguing considering this was reported among persons enrolled in clinical trials, where presumably there is higher engagement in healthcare and oversight that might otherwise at least partially account for sex differences in terms of outcome. When considering potential sex differences among persons of Hispanic ethnicity, Hispanic AYA males were at an 11% disadvantage in terms of 5-year survival, compared to their female counterparts (Table 3, 83% in males, 94% in females), which reflects a slightly stronger male-female disparity than observed among their NHW AYA counterparts (7% difference in favor of female survival; 89% in males, 96% in females). In corresponding multivariate adjusted models, the AYA male disparity was persistent for both Hispanic and NHWs (Table 5b, HR: 1.50; 95%CI: 1.16, 1.94 in Hispanic AYAs, HR: 1.54; 95%CI: 1.41, 1.67 in NHW AYAs). It should be noted that overall risk of death in the univariate model among Hispanic AYAs was 17% higher than observed in their NHW counterparts (Table 4, HR: 1.17; 95%CI: 1.03, 1.33). The Hispanic AYA HR remained elevated but not statistically significant after adjustment for demographics (sex, year of diagnosis, and age) and histology (HR: 1.09; 95%CI: 0.96, 1.24, data not shown), but was then paradoxically reversed in the fully adjusted model, with Hispanic AYAs now at 20% reduced risk of death compared to their NHW counterparts (Table 4, HR: 0.80; 95%CI: 0.70, 0.91). 42 This reversal from increased to decreased risk of death for Hispanic vs. NHW AYAs was explained solely by the addition of stage at diagnosis to a model already accounting for demographics and histology (HR: 0.82; 95%CI: 0.72, 0.93, data not shown). However, the protective effect was not reproduced when tumor thickness was used instead of stage (HR: 1.00; 95%CI: 0.87, 1.13, data not shown). Given the Hispanic paradox was explained in our data by adjustment for stage, Hispanic AYAs might be experiencing a true survival disparity compared to their NHW counterparts. A male-female disadvantage may further worsen prognosis for young Hispanic males, since recent trends suggest increased risk of delayed diagnosis and thicker tumors among Hispanics, particularly for males (46, 47). As the overall Hispanic population grows from 38% in California in 2010 (16% nationally) to a projected 47% in California by 2050, it is possible that an AYA male-female disparity could strengthen, because representation of Hispanic persons would be expected to expand more quickly within younger age groups during the overall projected population growth. This should emphasize the continued importance of monitoring AYA melanoma incidence and survival among Hispanic persons. Simultaneously, greater patient and clinician awareness of melanoma in Hispanic persons will be needed to facilitate earlier detection to reduce the morbidity and mortality associated with late stage melanoma, and public health efforts among Hispanic populations to encourage adoption of appropriate skin-protective behaviors may assist with primary prevention. 43 A number of biological explanations for a male survival disadvantage have been put forth and tested to varying degrees, as described in a recent review (details in Figure 4) (48). Not only was a male disadvantage in survival observed across many cancer types, but for melanoma specifically, this appears to occur in an age-related manner across different countries, reflecting inherently higher immune function and tumor suppression mechanisms known to exist for women, which may interact with sun exposure and dietary behaviors to influence vitamin D levels and oxidative response. FIGURE 2.0.4. PROPOSED EXPLANATIONS FOR SEX DIFFERENCES IN MELANOMA SURVIVAL (48, P.43) 44 Socioeconomic Status. SES continues to be strongly associated with melanoma. In the literature, higher SES is strongly associated with higher incidence, as was observed in the present study. Despite this, the association with risk of death is opposite and persons of lower SES are at greater risk of dying from their disease, even though they are less likely to get a melanoma in the first place. This holds true for AYAs and adults alike, there was 50-79% increased risk for the lowest SES compared to the highest SES (AYA HR: 1.50; 95%CI: 1.30, 1.72; Older Adult HR: 1.79; 95%CI: 1.70, 1.89). There are a number of other AYA-specific issues likely impacting survival after a melanoma diagnosis, including: (1) lack of trial availability for and participation by AYAs, (2) limited availability of trials for AYAs, related to policy and regulatory issues, and (3) age-specific developmental and behavioral barriers,(49) including adherence-related issues, which are understudied, but may have large impact on outcomes when adherence is lacking. Future work should explore the impact of these factors, as they have not yet been studied in AYA melanoma. STRENGTHS & LIMITATIONS This was a large, population-based study of persons diagnosed with melanoma <65 years of age in California over a long period of time (1988-2014). The size and diversity of this population allowed closer examination of demographic and tumor characteristics, which have not been well-described for AYAs in the literature, and also facilitated comparisons of AYAs to their younger and older adult counterparts. As a records-based study, no participant contact was necessary, reducing or eliminating 45 selection bias. Estimation of survival using Kaplan-Meier cumulative probabilities and Cox hazard regressions accounts for varying amounts of follow-up time across individuals, because these methods incorporate all of the time contributed by an individual along with information on vital status or loss to follow up; additionally, the Cox model allows for adjustment of confounding factors. This study is limited in that we had no information on family history of melanoma, prior personal history of melanoma if diagnosed outside California, or prior history of non- melanoma skin cancer, which is not captured in registry data. Information was also lacking on underlying genetic factors (BRAF, NRAS, etc.), constitutional factors (hair, eye, and skin color, number of moles, etc.), treatment type, treating center (community or comprehensive), enrollment in clinical trials. Survival was based on any cause of death and survival models assumed proportional hazards over time. Importantly, we had no information on behavioral factors, such as ultraviolet radiation exposure (solar or artificial/tanning beds), skin protection, exercise, and dietary habits, which may contribute not only to susceptibility, but also to the relative aggressiveness of a melanoma once it has occurred, which could influence survival outcomes. Importantly, tumor thickness was not always a required variable for collection/reporting for registry data. Further, the way tumor thickness has been measured and defined has changed over time (i.e. formerly Clark’s level, more recently Breslow’s Depth). Thus, our inferences around the impact of tumor thickness were limited (44% had unknown tumor thickness). If these missing data tended to represent thicker tumors, we would have underestimated the influence of greater tumor thickness on survival outcomes. Lastly, 46 although relative survival analysis could have been used instead for comparing survival in younger and older melanoma patients, the authors chose to present observed survival statistics for their ease of interpretability and because the level of detail provided herein for observed survival has not been well-described previously in the literature. Observed survival simply reflects who survives from all causes of death after their melanoma diagnosis, while relative survival reflects a ratio of observed (all cause) survival among in a given cohort of melanoma patients versus expected (all cause) survival in a comparable cancer-free population. However, since obtaining a comparable cancer-free population is challenging, the relative survival calculation is instead based on expected life tables to determine expected survival, operating under the assumption that that the number cancer deaths is very small. GENERALIZABILITY The results should be generalizable to persons 0-64 years of age at diagnosis with melanoma in California, but may also be applicable to persons with melanoma diagnosed in areas of similar geographical and race/ethnic distribution. The results would not apply to persons with prior history of melanoma or any other cancer due to an inherently increased risk of developing a subsequent primary, possibly due to genetic susceptibility or as a late effect from the first cancer treatment. These results would also not apply to persons 65 or older, who may have differences in underlying tumor biology, immune function, prevalence of comorbidities, or cumulative exposures related to melanoma etiology, which may confer a different profile for risk of death than that seen among persons <65 years at diagnosis. Lastly, the utility of patterns among AYAs may 47 not be limited to melanoma; it is possible that similar patterns may be observed to impact survival in other AYA cancers. Indeed, the literature has suggested that underlying themes of emerging adulthood impact engagement with healthcare among AYAs across cancer types. 48 CONCLUSION Although survival among AYAs with melanoma increased over the period studied, important subgroups with poorer outcomes were observed. Among AYA with melanoma, males were at increased risk of death despite having lower incidence than AYA females, and this disadvantage was slightly stronger among AYAs of Hispanic heritage. SES continues to play an important role in melanoma, where those in the highest SES get melanoma more frequently, but those in the lowest SES have worse survival. AYAs with melanoma had substantially higher risk of death at every stage of diagnosis than their adult counterparts, even at localized stage. Nodular and acral melanomas had poorer prognosis among AYAs than at younger or older ages. AYAs are in transitional stage of life involving aspects of cognitive, emotional, and financial development, and as such, they may perceive themselves as somewhat invincible. When an unusual growth has been detected by themselves or by a clinician, they may be less likely to seek treatment; they may be less knowledgeable or capable of processing and absorbing the seriousness of their melanoma diagnosis and reacting accordingly; they may think they are “too busy” and desire instead to focus on living a normal life (i.e. attending school, working, dating, etc.); and they may be particularly sensitive to the cost of treatment and/or experience gaps in health insurance, because they may not yet have a stable career, health insurance, or a career that allows them to take days off for treatment and recovery. These are unique “life-phase” factors that may affect an AYA’s ability to then adhere to treatment and follow-up care, in turn escalating 49 their risk of poor survival outcomes in ways that are not true of younger or older patients. Identifying important subgroups or clinical characteristics of AYAs at greatest risk of dying after a melanoma diagnosis increases our ability to identify the patients in greatest need of screening and follow-up care. This knowledge may be used to target follow-up care efforts, especially for AYAs with melanoma who are at increased risk for poorer outcomes. Efforts to reduce the gap in survival among young persons should be prioritized for clinicians, researchers, and policy-makers alike. For example, clinicians could increase screening and follow-up efforts for AYAs, especially 15-24 year olds to identify melanoma at earlier stages, reducing the risk of mortality. Researchers should also investigate ways the underlying biology may be different, which could provide insight into whether AYAs get more aggressive tumors and could potentially lead to new, age-specific drug targets or dosing protocols. Clinical trials should regularly include AYAs to determine if the type treatments and dosing may be optimized to improve survival. Recent advances in targeted therapies for older adults have led to dramatically improved prognosis during clinical trials, but these improvements have yet to be realized for AYA with melanoma. Policy related to clinical trials may also be urged to move forward by adopting revisions to allow inclusion of AYAs in clinical trials. In turn, adherence to recommended treatment may be increased, the impact of late effects may be reduced, and subsequent primaries may be caught earlier, which may all contribute to improved survival and reduce a potential excess burden of mortality among AYAs with melanoma. 50 TABLE 2.0.1. DEMOGRAPHIC CHARACTERISTICS OF PATIENTS DIAGNOSED WITH MELANOMA, BY AGE GROUP, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* 51 TABLE 2.0.2. TUMOR CHARACTERISTICS OF PATIENTS DIAGNOSED WITH MELANOMA, BY AGE GROUP, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014 52 TABLE 2.0.3. AGE-ADJUSTED INCIDENCE RATES (AAIR) FOR ALL CUTANEOUS MELANOMAS DIAGNOSED AT AGES 0-64 YEARS, BY AGE GROUP, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* 53 TABLE 2.0.4. FIVE YEAR OBSERVED SURVIVAL (OS) PROBABILITY AND CHANGE OVER TIME IN PATIENTS DIAGNOSED WITH MELANOMA, BY AGE GROUP AND DEMOGRAPHICS, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* 54 TABLE 2.0.5. FIVE YEAR OBSERVED SURVIVAL (OS) IN PATIENTS WITH MELANOMA DIAGNOSED AT AGES 0-64 YEARS, BY AGE GROUP AND TUMOR CHARACTERISTICS, CALIFORNIA CANCER REGISTRY (CCR), 1988-2014* 55 TABLE 2.0.6. HAZARD RATIOS (HR) FOR RISK OF DEATH BY DEMOGRAPHICS, STRATIFIED BY AGE GROUP, IN PATIENTS WITH MELANOMA DIAGNOSED AT AGES 0-64 YEARS, CALIFORNIA (1988-2014)* 56 TABLE 2.0.7. HAZARD RATIOS FOR RISK OF DEATH BY TUMOR CHARACTERISTICS, STRATIFIED BY AGE GROUP, IN PATIENTS WITH MELANOMA DIAGNOSED AT AGES 0-64 YEARS, CALIFORNIA (1988-2014) 57 TABLE 2.0.8. AGE GROUP COMPARISONS OF HAZARD RATIOS (HR) FOR RISK OF DEATH IN MALES VS. FEMALES IN PATIENTS DIAGNOSED WITH MELANOMA AT AGES 0-64 YEARS (1988-2014) TABLE 2.0.9. HAZARD RATIOS (HR) FOR RISK OF DEATH IN MALES VS. FEMALES AMONG ADOLESCENTS AND YOUNG ADULTS (AYA) DIAGNOSED WITH MELANOMA AT AGES 15-30 YEARS, CALIFORNIA (1988- 2014), STRATIFIED BY SELECTED CHARACTERISTICS 58 TABLE 2.0.10. DEMOGRAPHIC CHARACTERISTICS OF HISPANIC ADOLESCENTS AND YOUNG ADULTS (AYAS) DIAGNOSED WITH MELANOMA AT AGES 15-39 YEARS, CALIFORNIA (1988-2014)* 59 TABLE 2.0.11. TUMOR CHARACTERISTICS OF HISPANIC ADOLESCENTS AND YOUNG ADULTS (AYAS) DIAGNOSED WITH MELANOMA AT AGES 15-39 YEARS, CALIFORNIA (1988-2014)* 60 CHAPTER 2 REFERENCES 1. Bleyer A, O’Leary M, Barr R, Ries L, (eds), eds. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000 . Bethesda, MD: National Cancer Institute; 2006. Bleyer A., O’Leary M., Barr R. and Ries LAG, eds. NIH Pub. no. 06-5767. 2. Bleyer A, Barr R. Cancer in young adults 20 to 39 years of age: Overview. . 2009;36(3):194-206. 3. Barr RD, Ries LA, Lewis DR, et al. Incidence and incidence trends of the most frequent cancers in adolescent and young adult americans, including “nonmalignant/noninvasive” tumors. Cancer. 2016. 4. Adolescent and Young Adult Oncology Progress Review Group (AYAO PRG), ed. Closing the gap: Research and care imperatives for adolescents and young adults with cancer (NIH publication no. 06-6067). Department of Health and Human Services, National Institutes of Health, National Cancer Institute, and the LIVESTRONG Young Adult Alliance Bethesda, MD; 2006. 5. Smith AW, Seibel NL, Lewis DR, et al. Next steps for adolescent and young adult oncology workshop: An update on progress and recommendations for the future. Cancer. 2016. 61 6. Guy Jr GP, Thomas CC, Thompson T, Watson M, Massetti GM, Richardson LC. Vital signs: Melanoma incidence and mortality trends and projections-united states, 1982- 2030. MMWR Morb Mortal Wkly Rep. 2015;64(21):591-596. 7. Ries L, Trama A, Nakata K, Gatta G, Botta L, Bleyer A. Cancer incidence, survival, and mortality among adolescents and young adults. In: Cancer in adolescents and young adults. Springer; 2017:7-42. 8. Forman D, Bray F, Brewster DH, Gombe Mbalawa C, Kohler B, Piñeros M, Steliarova- Foucher E, Swaminathan R and Ferlay J, editors, ed. Cancer incidence in five continents, vol. X (electronic version). lyon: International agency for research on cancer. available from: Http://Ci5.iarc.fr, accessed [february 1, 2017]. 2013th ed. Lyon, France.: International Agency for Research on Cancer.; 2013IARC Scientific Publication no. 164. 9. Thomas DM, Albritton KH, Ferrari A. Adolescent and young adult oncology: An emerging field. J Clin Oncol. 2010;28(32):4781-4782. 10. Bleyer A, Budd T, Montello M. Adolescents and young adults with cancer. Cancer. 2006;107(S7):1645-1655. 11. De Gruijl F. Skin cancer and solar UV radiation. Eur J Cancer. 1999;35(14):2003-2009. 12. Armstrong BK, Kricker A. The epidemiology of UV induced skin cancer. Journal of Photochemistry and Photobiology B: Biology. 2001;63(1):8-18. 62 13. Narayanan DL, Saladi RN, Fox JL. Review: Ultraviolet radiation and skin cancer. Int J Dermatol. 2010;49(9):978-986. 14. Cockburn M, Hamilton A, Mack T. The simultaneous assessment of constitutional, behavioral, and environmental factors in the development of large nevi. Cancer Epidemiol Biomarkers Prev. 2007;16(2):200-206. 15. Holman CD, Armstrong BK. Pigmentary traits, ethnic origin, benign nevi, and family history as risk factors for cutaneous malignant melanoma. J Natl Cancer Inst. 1984;72(2):257-266. 16. Miller KA, Langholz BM, Zadnick J, et al. Prevalence and predictors of recent skin examination in a population-based twin cohort. Cancer Epidemiol Biomarkers Prev. 2015;24(8):1190-1198. 17. Cockburn M, Hamilton A, Mack T. Recall bias in self-reported melanoma risk factors. Am J Epidemiol. 2001;153(10):1021-1026. 18. Fu JM, Dusza SW, Halpern AC. Sunless tanning. J Am Acad Dermatol. 2004;50(5):706- 713. 19. Boniol M, Autier P, Boyle P, Gandini S. Cutaneous melanoma attributable to sunbed use: Systematic review and meta-analysis. BMJ. 2012;345:e4757. 63 20. Colantonio S, Bracken MB, Beecker J. The association of indoor tanning and melanoma in adults: Systematic review and meta-analysis. J Am Acad Dermatol. 2014;70(5):847-857.e18. 21. O’Leary RE, Diehl J, Levins PC. Update on tanning: More risks, fewer benefits. J Am Acad Dermatol. 2014;70(3):562-568. 22. Nikolaou V, Stratigos AJ. Emerging trends in the epidemiology of melanoma. Br J Dermatol. 2014;170(1):11-19. 23. Mills O, Messina JL. Pediatric melanoma: A review. Cancer control: journal of the Moffitt Cancer Center. 2009;16(3):225. 24. Whiteman DC, Valery P, McWhirter W, Green AC. Risk factors for childhood melanoma in queensland, australia. International Journal of Cancer. 1997;70(1):26-31. 25. Shaikh WR, Dusza SW, Weinstock MA, Oliveria SA, Geller AC, Halpern AC. Melanoma thickness and survival trends in the united states, 1989 to 2009. J Natl Cancer Inst. 2015;108(1):10.1093/jnci/djv294. Print 2016 Jan. 26. Pollack LA, Li J, Berkowitz Z, et al. Melanoma survival in the united states, 1992 to 2005. J Am Acad Dermatol. 2011;65(5):S78. e1-S78. e10. 27. Gamba CS, Clarke CA, Keegan TH, Tao L, Swetter SM. Melanoma survival disadvantage in young, non-hispanic white males compared with females. JAMA dermatology. 2013;149(8):912-920. 64 28. Bleyer WA. Cancer in older adolescents and young adults: Epidemiology, diagnosis, treatment, survival, and importance of clinical trials. Med Pediatr Oncol. 2002;38(1):1- 10. 29. Bleyer A, Choi M, Fuller CD, Thomas CR, Wang SJ. Relative lack of conditional survival improvement in young adults with cancer. . 2009;36(5):460-467. 30. Division of Cancer Prevention and Control, Centers for Disease Control and Prevention. National program of cancer registries (NPCR) National data quality standard . CDC’s National Program of Cancer Registries (NPCR) Web site. https://www.cdc.gov/cancer/npcr/pdf/npcr_standards.pdf. Updated August 9, 2013. Accessed February 1, 2017. 31. Lasithiotakis K, Leiter U, Meier F, et al. Age and gender are significant independent predictors of survival in primary cutaneous melanoma. Cancer. 2008;112(8):1795-1804. 32. Yost K, Perkins C, Cohen R, Morris C, Wright W. Socioeconomic status and breast cancer incidence in california for different race/ethnic groups. Cancer Causes & Control. 2001;12(8):703-711. 33. Balch CM, Gershenwald JE, Soong S, et al. Final version of 2009 AJCC melanoma staging and classification. Journal of clinical oncology. 2009;27(36):6199-6206. 34. Shaikh WR, Weinstock MA, Halpern AC, Oliveria SA, Geller AC, Dusza SW. The characterization and potential impact of melanoma cases with unknown thickness in the 65 united states’ surveillance, epidemiology, and end results program, 1989–2008. Cancer epidemiology. 2013;37(1):64-70. 35. Iannacone MR, Youlden DR, Baade PD, Aitken JF, Green AC. Melanoma incidence trends and survival in adolescents and young adults in queensland, australia. International Journal of Cancer. 2015;136(3):603-609. 36. Redman JM, Gibney GT, Atkins MB. Advances in immunotherapy for melanoma. BMC medicine. 2016;14(1):20. 37. Ferrari A, Montello M, Budd T, Bleyer A. The challenges of clinical trials for adolescents and young adults with cancer. Pediatric blood & cancer. 2008;50(S5):1101- 1104. 38. Shaw PH, Hayes-Lattin B, Johnson R, Bleyer A. Improving enrollment in clinical trials for adolescents with cancer. Pediatrics. 2014;133(Supplement 3):S109-S113. 39. Sondak VK, Swetter SM, Berwick MA. Gender disparities in patients with melanoma: Breaking the glass ceiling. J Clin Oncol. 2012;30(18):2177-2178. 40. Joosse A, De Vries E, Eckel R, et al. Gender differences in melanoma survival: Female patients have a decreased risk of metastasis. J Invest Dermatol. 2011;131(3):719-726. 41. Joosse A, Collette S, Suciu S, et al. Sex is an independent prognostic indicator for survival and relapse/progression-free survival in metastasized stage III to IV melanoma: 66 A pooled analysis of five european organisation for research and treatment of cancer randomized controlled trials. Journal of Clinical Oncology. 2013;31(18):2337-2346. 42. de Vries E, Nijsten TE, Visser O, et al. Superior survival of females among 10,538 dutch melanoma patients is independent of breslow thickness, histologic type and tumor site. Ann Oncol. 2008;19(3):583-589. 43. Keegan TH, Swetter SM, Tao L, Sunwoo JB, Clarke CA. Tumor ulceration does not fully explain sex disparities in melanoma survival among adolescents and young adults. J Invest Dermatol. 2015. 44. Haggar FA, Pereira G, Preen DD, Holman CJ, Einarsdottir K. Cancer survival and excess mortality estimates among adolescents and young adults in western australia, 1982–2004: A population-based study. PloS one. 2013;8(2):e55630. 45. McNally RJ, Basta NO, Errington S, James PW, Norman PD, Craft AW. Socioeconomic patterning in the incidence and survival of children and young people diagnosed with malignant melanoma in northern england. J Invest Dermatol. 2014;134(11):2703-2708. 46. Cockburn MG, Zadnick J, Deapen D. Developing epidemic of melanoma in the hispanic population of california. Cancer. 2006;106(5):1162-1168. 47. Clarke CA, Moy LM, Swetter SM, Zadnick J, Cockburn MG. Interaction of area-level socioeconomic status and UV radiation on melanoma occurrence in california. Cancer Epidemiol Biomarkers Prev. 2010;19(11):2727-2733. 67 48. Nosrati A, Wei ML. Sex disparities in melanoma outcomes: The role of biology. Arch Biochem Biophys. 2014;563:42-50. 49. Tai E, Beaupin L, Bleyer A. Clinical trial enrollment among adolescents with cancer: Supplement overview. Pediatrics. 2014;133(Supplement 3):S85-S90. 68 CHAPTER 3: HIGH BIRTH WEIGHT, EARLY UV EXPOSURE, AND MELANOMA RISK IN CHILDREN, ADOLESCENTS AND YOUNG ADULTS 69 ABSTRACT Melanoma, the deadliest form of skin cancer, is one of the most common cancers diagnosed before age 30. Little is known about potentially modifiable risk factors for melanoma in young patients. Determine if birth weight or higher early life ultraviolet radiation (UV) exposure is associated with increased risk of melanoma in young patients. Population-based, case-control study of 1,396 melanomas diagnosed before age 30 in 1988-2013 and 27,920 age-matched controls in California. High birth weight was associated with 19% higher risk of melanoma (adjusted OR: 1.19; 95%CI: 1.02, 1.39), low birth weight was associated with 41% lower risk (adjusted OR: 0.59; 95%CI: 0.43, 0.82); Birthplace UV levels also increased melanoma risk and the strongest effect for UV was for diagnoses at ages 15-19. Information on individual sun exposure behaviors could not be assessed; no information on genetic variants or family history. High birth weight and high early life UV increased the risk of early onset melanoma, underscoring the importance of promoting skin protective behaviors early in life for primary prevention of melanoma in young people. Increased clinician awareness may help identify high-risk, young patients to prioritize them for screening. 70 BACKGROUND Melanoma, the deadliest form of skin cancer, is responsible for about 75% of all skin cancer deaths (1). Although largely preventable (2-4), incidence of melanoma has been rising for several decades in the United States (US) (4-8), where more than 76,000 new melanomas are diagnosed annually (9). Melanoma has become one of the most common cancers among people <30 years of age in the US (10). Although more than 1,000 melanomas diagnosed in California each year will affect adolescent and young adult patients (AYA) aged 15-39 years old at diagnosis (11), our knowledge of melanoma risk factors is derived from adults older than 50 years of age. Incidence trends from the Surveillance, Epidemiology, and End Results (SEER) Program show that young females have much higher incidence of melanoma than their male counterparts, while the reverse is true for adults age 50 and older (12, 13), suggesting that risk factors identified among older persons may not explain risk in younger persons. Rates of melanoma in California are among the highest in the world (34.7 per 100,000 for non-Hispanic whites (NHW) in 2012, nearly doubling from 18.1 per 100,000 since 1990) (14). Historically, rates among Hispanic persons have been relatively low (4.2 per 100,000 in 2012) (14), but an increasing burden of thick and later stage tumors (15, 16) has been observed among California’s Hispanic population, making this state home to a diverse, high-risk population, where nearly 40% of the population is of Hispanic or 71 Latino heritage (17). Nearly 9,000 new melanomas occur annually (18) and heightened screening efforts do not explain the trends in this state (8, 19). Experimental studies suggest that early life ultraviolet (UV) exposure may affect melanoma occurrence at a young age (20-28), particularly reports that neonatal tissue may be more susceptible to invasive melanoma after a UV insult (20, 29). Underlying biology of melanoma in younger patients is also quite different (30), with childhood melanomas commonly presenting as amelanotic lesions (31). Obesity is a major risk factor for adult cancer, (32, 33) and higher birth weight is associated with pediatric, adolescent and adult obesity (34-36). Increased risk of certain childhood cancers (primarily leukemias/lymphomas) has been observed with higher birth weight (37-41), which has also been linked to adult melanoma in Danish (42) and Swedish cohorts (43), but remains less well examined at younger ages. Potentiating mechanisms for a link between obesity and melanoma include greater body surface area (44) available for sun exposure or melanoma development, increased leptin levels to support tumor growth via angiogenesis (45), and activation of BRAF mutations (46, 47). Studies of risk factors for pediatric and AYA melanomas remain sparse. We conducted a large, population-based, case-control study among California-born Hispanic and non-Hispanic persons aged <30 years to investigate the association between birth weight, early life ambient UV and a diagnosis of melanoma. 72 MATERIALS & METHODS POPULATION & DATA SOURCES Cases of cutaneous malignant melanoma (ICD-O-3 sites C440-449 and histology 8720- 8780) were identified from the California Cancer Registry. Cases (n=1,396) had a primary melanoma diagnosis at 0-29 years of age in 1988-2013 and controls (n=28,460) were identified from California birth records. Birth record linkage and selection of control subjects Cases were matched to the birth record using date of birth, sex, first and last name, mother’s residential address and/or zip, and mother’s surname at birth, as described previously (39). Controls (20 randomly selected for each case) were also identified from California Birth Registry records, with frequency matching by date of birth (±6 months from the case’s date of birth). Ultraviolet Radiation (UV) occurring at Place of Birth Birthplace UV (as a proxy for early life UV exposure) was measured in average daily watt-hours/m 2 and assigned based on mother’s home address at the time of the subject’s birth (38, 48) as described previously (48, 49). Quartiles of UV exposure were determined from the distribution observed among controls. Birth Weight Birth weight, in grams (g), was obtained from the birth record (n=79 total missing). Birth weights of <1000g (n=136) or >5250g (n=15) were considered extreme and excluded from analysis, as survival is limited with birth weight <1000g (50) and values >5250g 73 reflected <1% of our data (no appreciable impact was observed from their inclusion during sensitivity analysis). We defined low birth weight (LBW) for ≥1000 to <2500g, normal birth weight for 2500 to 4000g, and high birth weight (HBW) for ≥4000g to 5250g (50, 51). Birth weight in 1000g intervals (i.e. 1000-1999g, 2000-2999g, 3000- 3999g, etc.) was treated as a continuous variable for trend analysis. Covariates Quintiles of census tract socioeconomic status (SES) were determined from mother’s residential address on the birth record (52). Age at diagnosis in the cases was used to represent age for cases and their controls. Maternal age, race (white, black, Asian, or other), and ethnicity (Hispanic or non-Hispanic) were also obtained from the birth record, along with information on sex and gestational age (38) (weeks) for cases and controls. STATISTICAL ANALYSIS Means and frequencies were used to describe sample characteristics. Because controls were frequency matched to cases only by date of birth (±6 months), unconditional logistic regression was used to estimate odds ratios (OR) and 95% confidence intervals (CI) for melanoma. The main effect of UV was examined by comparing each quartile to the lowest quartile. Confounders in the final models were identified by a priori criteria of >10% change in the unadjusted beta estimate. Effect measure modification by age at diagnosis, sex, and ethnicity were considered with stratified analysis when permitted by 74 sample size. All analyses were conducted with SAS version 9.4 (SAS Institute, Inc., Cary, North Carolina). Approval for this study was obtained from the University of Southern California Health Science Institutional Review Board (USC-HSIRB) and the California State Committee for the Protection of Human Subjects (CPHS). 75 RESULTS Cases were less likely than controls to be of Hispanic ethnicity, but tended to have higher SES, mothers with higher maternal age, higher birth weight, lower mean UV exposure, and were more likely to be female (Table 1). Final models for birth weight and birthplace UV were mutually adjusted for variables identified as confounders, including SES, sex, age, mother’s race/ethnicity, and mother’s age. Inclusion of gestational age in our models made no difference (data not shown). In the overall adjusted model (Table 2), HBW was associated with 19% higher odds of melanoma (OR: 1.19; 95%CI: 1.02, 1.39) compared with normal birth weight, while LBW was associated with 41% lower odds of melanoma (OR: 0.59; 95%CI: 0.43, 0.82). Melanoma odds increased by 24% per each 1000g increase in birth weight (OR: 1.24; 95%CI: 1.13, 1.36). In addition, we found a dose-response with increasing birth weight (Table 4; OR: 1.12; 95%CI: 1.08, 1.17). Adjusted results for birthplace UV showed all categories of UV exposure greater than Q1 were associated with increased melanoma odds (P< 0.05; Q2-OR: 1.44, 95%CI: 1.23, 1.69; Q3-OR: 1.37, 95%CI: 1.18, 1.61; Q4-OR: 1.29, 95%CI: 1.09, 1.53). However, a dose- response relationship was not evident. We observed an interaction between birth weight and UV: among HBW infants with “high” UV (Q2-4), melanoma odds was 43% higher (OR: 1.43; 95%CI: 1.17, 1.74) than normal birth weight infants with “low” UV (Q1), yet a similar effect was not observed for HBW infants with low UV (OR: 0.92; 76 95%CI: 0.66, 1.27); there were no differences in the adjusted LBW protective effect between high and low UV (data not shown). In age-stratified analysis (Table 3), the association of birthplace UV was statistically significant only for melanomas diagnosed at 15-19 years of age (Q2-OR: 1.81, 95%CI: 1.32, 2.47; Q3-OR: 1.85, 95%CI: 1.37, 2.50; Q4-OR: 1.65, 95%CI: 1.19, 2.30). Birth weight effects across age groups were consistent with those observed among all subjects, but CIs included the null (data not shown). For both males and females, there were associations between increased risk of melanoma and higher birth weight, advanced maternal age, higher birthplace UV, and higher SES, however the strength of the associations differed by gender (Table 2). Among females the association with higher birthplace UV was not statistically significant, whereas it was among males. Among Hispanic persons (Table 2), although there were similar trends in the adjusted OR’s to the non-Hispanics, SES was the only variable for which there was a marginally significantly elevated association for Hispanics (Adjusted OR: 1.18 per SES quintile increase, 95% CI: 1.00, 1.38). 77 DISCUSSION By investigating nearly 1,400 melanomas diagnosed among California-born persons <30 years of age compared to over 27,000 matched controls, this study has identified high birth weight and birthplace residence in areas of high UV as important, independent risk factors for the development of early onset melanoma. Melanoma is an aggressive tumor with potentially devastating impact in terms of years of life lost for young people, especially with late stage disease. As incidence of melanoma in young people has risen globally in recent decades, identifying risk factors specific to this age group can facilitate primary prevention, while also highlighting aspects of early life development that warrant more exploration. After adjustment for confounders, high birth weight increased the odds of melanoma by 19%. This finding is consistent with some (37, 43, 53), but not all, (37) of the existing studies of birth weight and other childhood cancers, which rarely included melanomas. Low birth weight appeared protective (up to 40% lower odds), which has not been previously described. In an Irish study, higher birth weight was associated with melanoma at 21-30 years of age (53), however, examination at younger ages was not possible because no cases in the cohort occurred before age 21. An international, pooled case-control study of childhood cancers including melanomas up to 14 years of age observed a positive association between birth weight and childhood melanomas in UK data (n=126; OR: 1.35; 95%CI: 1.12, 1.64), but the association was null in US data (n=114; OR: 1.02; 95%CI: 0.84, 1.22) (37). In a Danish study of mostly older adults, melanoma risk increased by 7% per 1000g increase in birth weight (42); similarly, risk 78 increased by 18% per 1000g among young people in Sweden (43). We observed an even stronger dose-response effect, i.e. a 24% increase in melanoma risk per 1000g birth weight. Why high birth weight might be associated with childhood melanoma occurrence remains unclear, but might involve differences in skin surface area, propensity for obesity and tumorigenesis, and behaviors impacting infant sun exposure. Increased skin surface area is a likely feature of children born with higher birth weight and may simply reflect a greater opportunity for sun exposure or greater chance of nevi presence (number of nevi is one of the strongest predictors of melanoma (54-57). Obesity later in life has been associated with higher birth weight (58), and obesity itself is an important risk factor for cancers (32), including melanoma (44), through a variety of mechanisms (32), including an underlying propensity for tumor growth via angiogenesis, which represents the body’s ability to support rapid expansion of capillary networks that are crucial for tumor development (32, 33). Obesity remains prevalent in the US population, which may accelerate melanoma growth via increased leptin levels (45), and presence of the “obesogenic” FTO gene also appears to increase risk of melanoma (59). Maternal overweight or obesity status before pregnancy, and excessive weight gain during pregnancy, are modifiable (60) factors linked to larger offspring size, higher birth weight (61-63), and poorer cardiovascular characteristics in childhood, including elevated leptin (64). Intervention upon excess maternal weight before and during pregnancy may reduce the potential for high offspring birth weight, perhaps modifying the risk of early onset melanoma. Our results suggest the link between birth weight and 79 melanoma should be more closely examined for melanomas occurring in childhood, adolescence, and young adulthood. We observed that residence in areas with high UV exposure at birth to be an important risk factor for melanoma in younger patients, particularly for AYAs 15-19 years of age. UV exposure occurring early in life may be sufficient to increase risk for early-onset melanoma, possibly because of increased susceptibility to UV damage in infant skin, which is still developing through the first two years of life (65). Characterizing the frequency and intensity of UV exposure occurring around infancy could help explain how early life UV is linked to development of melanoma before age 30. Sun exposure behavior among infants has not been investigated to date. If larger (normal weight or HBW) infants are taken outdoors more often very early in infancy than their smaller (LBW) counterparts, then larger infants may have a greater chance of being exposed to the sun, which may at least partially explain the increased risk for early onset melanoma observed with higher birth weight (and the reduced risk among those with low birth weight). Improving our understanding of sun exposure occurring among infants could help inform the development of evidence-based guidelines for effective infant skin protection and primary prevention efforts. SES was an important confounder of the effect of early life UV, which is consistent with existing adult literature, where the highest SES levels are associated with the greatest melanoma risk. High SES may reflect a variety of behaviors that influence risk of diagnosis, including different social norms related to sunbathing, more knowledge about 80 melanoma prevention, greater access to clinical skin screenings, as well as more frequent participation in leisure time, recreational activities, and vacation-related sun exposure that increases UV exposure (8, 52, 66). High SES was also the most important risk factor among Hispanic persons, who may be “acquiring” increased melanoma risk by adopting behavioral risk factors from their non-Hispanic counterparts alongside greater acculturation, similar to phenomena described for obesity risk acquisition upon moving to the US (67), a concerning trend given one study has described low skin- protective behaviors in Hispanic children (68) and infant sun exposure patterns remain unexplored. STRENGTHS AND LIMITATIONS This population-based case-control study was conducted in a large, diverse population of California with a wide range of UV exposures (66), over a long time span (1988-2013). Linkage to birth records facilitated the study of birth weight, as well as the identification of birth residence to link objectively to early UV exposure which is free from bias associated with self-reported measures of UV (69) and no participant contact was necessary, reducing or eliminating selection bias. However, sun exposure behavior information is not available, and misclassification of a subject’s true sun exposure during infancy and early childhood may be present. We also could not assess how long a person had lived at their birthplace address and had no information on genetic variants or family history of melanoma. 81 CONCLUSION The epidemic of melanoma among young people in the US remains an important public health concern. High birth weight and high early life UV exposures may be important risk factors for early onset melanoma and underscore the importance of promoting skin protective behaviors, as early as infancy or among expectant mothers, for primary prevention of melanoma in younger people. Knowledge of these risk factors may also help clinicians identify high-risk persons and prioritize them for screening. Although practicing good skin protection is recommended across the lifespan, it may be particularly important to implement such protection for infants. While literature on adults suggests that intense, intermittent UV or lifelong accumulation of chronic exposure to UV is related to developing a melanoma after age 50, the present study suggests that early exposure to high levels of ambient UV may be enough to increase the risk of a melanoma before age 30. Learning more about sun exposure behaviors at earlier ages, especially during infancy, will be critical to further our understanding of how these early life risk factors may lead to early onset melanoma and inform the development of evidence-based guidelines for skin protection among these vulnerable populations. 82 TABLE 3.0.1. CHARACTERISTICS OF A CASE-CONTROL STUDY OF MELANOMA DIAGNOSED AT AGES 0-29 YEARS IN CALIFORNIA (1988-2013)* 83 TABLE 3.0.2. EARLY LIFE RISK FACTORS FOR MELANOMA DIAGNOSED AT AGES 0-29 IN CALIFORNIA (1988-2013) 84 TABLE 3.0.3. AGE-STRATIFIED ASSOCIATION OF EARLY LIFE AMBIENT ULTRAVIOLET (UV) RADIATION AND MELANOMA DIAGNOSED AT AGES 0-29 YEARS IN CALIFORNIA (1988-2013) 85 TABLE 3.0.4. EVALUATION OF DOSE-RESPONSE ACROSS BIRTH WEIGHT FOR MELANOMA DIAGNOSED AT AGES 0-29 YEARS IN CALIFORNIA (1988-2013) 86 CHAPTER 3 REFERENCES 1. EPA Office of Air and Radiation (6205J). Facts about: Skin Cancer. May 2010 ed. Washington, D.C.: United States Environmental Protection Agency; 2010. 2. U.S. Department of Health and Human Services. The Surgeon General's Call to Action to Prevent Skin Cancer. Washington, DC: U.S. Dept of Health and Human Services, Office of the Surgeon General; 2014. 3. Armstrong BK, Kricker A. The epidemiology of UV induced skin cancer. Journal of Photochemistry and Photobiology B: Biology. 2001;63(1):8-18. 4. Marks R. Epidemiology of melanoma. Clin Exp Dermatol. 2000;25(6):459-63. 5. Austin MT, Xing Y, Hayes-Jordan AA, et al. Melanoma incidence rises for children and adolescents: an epidemiologic review of pediatric melanoma in the United States. J Pediatr Surg. 2013;48(11):2207-13. 6. Wong JR, Harris JK, Rodriguez-Galindo C, et al. Incidence of childhood and adolescent melanoma in the United States: 1973-2009. Pediatrics. 2013;131(5):846-54. (doi: 10.1542/peds.2012-2520 [doi]). 7. Purdue MP, Freeman LE, Anderson WF, et al. Recent trends in incidence of cutaneous melanoma among US Caucasian young adults. J Invest Dermatol. 2008;128(12):2905-8. (doi: 10.1038/jid.2008.159 [doi]). 87 8. Linos E, Swetter SM, Cockburn MG, et al. Increasing burden of melanoma in the United States. J Invest Dermatol. 2009;129(7):1666-74. 9. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA: a cancer journal for clinicians. 2016;66(1):7-30. 10. Bleyer A, O’Leary M, Barr R, Ries L, (eds). Cancer Epidemiology in Older Adolescents and Young Adults 15 to 29 Years of Age, Including SEER Incidence and Survival: 1975- 2000 . Bethesda, MD: National Cancer Institute; 2006. 11. Adolescent and Young Adult Oncology Progress Review Group (AYAO PRG). Closing the Gap: Research and Care Imperatives for Adolescents and Young Adults with Cancer (NIH Publication No. 06-6067). Department of Health and Human Services, National Institutes of Health, National Cancer Institute, and the LIVESTRONG Young Adult Alliance Bethesda, MD; 2006. 12. Bleyer A, O'leary M, Barr R, et al. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. 2006. 13. Guy Jr GP, Thomas CC, Thompson T, et al. Vital signs: melanoma incidence and mortality trends and projections-United States, 1982-2030. MMWR Morb Mortal Wkly Rep. 2015;64(21):591-6. 88 14. California Cancer Registry. Incidence and Mortality Counts and Age-Adjusted (2000 US Population) Rates per 100,000 Persons by Year, Race/Ethnicity, and Sex, California, 1988-2013, Invasive Melanoma of the Skin. , 2016. 15. Cockburn MG, Zadnick J, Deapen D. Developing epidemic of melanoma in the Hispanic population of California. Cancer. 2006;106(5):1162-8. 16. Pollitt RA, Clarke CA, Swetter SM, et al. The expanding melanoma burden in California hispanics. Cancer. 2011;117(1):152-61. 17. Quick Facts - Population of United States, California, and Los Angeles County , V2015. 18. Age-Adjusted Invasive Cancer Incidence Rates by County in California, 2013. Based on December 2015 extract. , December 2015. 19. Watson M, Johnson CJ, Chen VW, et al. Melanoma surveillance in the United States: overview of methods. J Am Acad Dermatol. 2011;65(5):S6. e1,S6. e12. 20. Benjamin CL, Melnikova VO, Ananthaswamy HN. Models and mechanisms in malignant melanoma. Mol Carcinog. 2007;46(8):671-8. 21. Noonan FP, Recio JA, Takayama H, et al. Neonatal sunburn and melanoma in mice. Nature. 2001;413(6853):271-2. 89 22. Noonan FP, Otsuka T, Bang S, et al. Accelerated ultraviolet radiation-induced carcinogenesis in hepatocyte growth factor/scatter factor transgenic mice. Cancer Res. 2000;60(14):3738-43. 23. Lu C, Zhang J, Nagahawatte P, et al. The genomic landscape of childhood and adolescent melanoma. J Invest Dermatol. 2014. 24. Autier P, for Epimel D, Eortc Melanoma Cooperative Group, Jean-François. Influence of sun exposures during childhood and during adulthood on melanoma risk. International Journal of Cancer. 1998;77(4):533-7. 25. Hodis E, Watson IR, Kryukov GV, et al. A landscape of driver mutations in melanoma. Cell. 2012;150(2):251-63. 26. Berger MF, Hodis E, Heffernan TP, et al. Melanoma genome sequencing reveals frequent PREX2 mutations. Nature. 2012;485(7399):502-6. 27. Whiteman DC, Whiteman CA, Green AC. Childhood sun exposure as a risk factor for melanoma: a systematic review of epidemiologic studies. Cancer Causes & Control. 2001;12(1):69-82. 28. Narayanan DL, Saladi RN, Fox JL. Review: Ultraviolet radiation and skin cancer. Int J Dermatol. 2010;49(9):978-86. 29. Berking C, Takemoto R, Satyamoorthy K, et al. Induction of melanoma phenotypes in human skin by growth factors and ultraviolet B. Cancer Res. 2004;64(3):807-11. 90 30. Ferrari A, Bono A, Baldi M, et al. Does melanoma behave differently in younger children than in adults? A retrospective study of 33 cases of childhood melanoma from a single institution. Pediatrics. 2005;115(3):649-54. (doi: 115/3/649 [pii]). 31. Cordoro KM, Gupta D, Frieden IJ, et al. Pediatric melanoma: results of a large cohort study and proposal for modified ABCD detection criteria for children. J Am Acad Dermatol. 2013;68(6):913-25. 32. De Pergola G, Silvestris F. Obesity as a major risk factor for cancer. Journal of obesity. 2013;2013. 33. Park J, Morley TS, Kim M, et al. Obesity and cancer - mechanisms underlying tumour progression and recurrence. Nature Reviews Endocrinology. 2014;10(8):455-65. 34. Dietz WH. Critical periods in childhood for the development of obesity. Am J Clin Nutr. 1994;59(5):955-9. 35. Desai M, Beall M, Ross MG. Developmental origins of obesity: programmed adipogenesis. Current diabetes reports. 2013;13(1):27-33. 36. Eriksson J, Forsen T, Tuomilehto J, et al. Size at birth, childhood growth and obesity in adult life. International Journal of Obesity & Related Metabolic Disorders. 2001;25(5). 37. O'Neill KA, Murphy MF, Bunch KJ, et al. Infant birthweight and risk of childhood cancer: international population-based case control studies of 40 000 cases. Int J Epidemiol. 2015;44(1):153-68. (doi: 10.1093/ije/dyu265 [doi]). 91 38. Heck JE, Lombardi CA, Meyers TJ, et al. Perinatal characteristics and retinoblastoma. Cancer Causes & Control. 2012;23(9):1567-75. 39. Heck JE, Lombardi CA, Cockburn M, et al. Epidemiology of rhabdoid tumors of early childhood. Pediatric blood & cancer. 2013;60(1):77-81. 40. Johnson K, Carozza S, Chow E, et al. Birth characteristics and childhood carcinomas. Br J Cancer. 2011;105(9):1396-401. 41. Bataille V. Melanoma. Shall we move away from the sun and focus more on embryogenesis, body weight and longevity? Med Hypotheses. 2013;81(5):846-50. 42. Ahlgren M, Wohlfahrt J, Olsen LW, et al. Birth weight and risk of cancer. Cancer. 2007;110(2):412-9. 43. Crump C, Sundquist K, Sieh W, et al. Season of birth and other perinatal risk factors for melanoma. Int J Epidemiol. 2014;43(3):793-801. (doi: 10.1093/ije/dyt277 [doi]). 44. Sergentanis TN, Antoniadis AG, Gogas HJ, et al. Obesity and risk of malignant melanoma: a meta-analysis of cohort and case–control studies. Eur J Cancer. 2013;49(3):642-57. 45. Brandon EL, Gu J, Cantwell L, et al. Obesity promotes melanoma tumor growth: role of leptin. Cancer biology & therapy. 2009;8(19):1871-9. 92 46. Chen J, Chi M, Chen C, et al. Obesity and melanoma: Exploring molecular links. J Cell Biochem. 2013;114(9):1955-61. 47. Chen J. Multiple signal pathways in obesity-associated cancer. Obesity Reviews. 2011;12(12):1063-70. 48. Tatalovich Z, Wilson JP, Mack T, et al. The objective assessment of lifetime cumulative ultraviolet exposure for determining melanoma risk. Journal of Photochemistry and Photobiology B: Biology. 2006;85(3):198-204. 49. Tatalovich Z, Wilson JP, Cockburn M. A comparison of thiessen polygon, kriging, and spline models of potential UV exposure. Cartography and Geographic Information Science. 2006;33(3):217-31. 50. Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. The Lancet. 2008;371(9608):261-9. 51. Kleinman RE. Pediatric Nutrition Handbook. 6th ed. Elk Grove Village, Ill.: American Academy of Pediatrics, 2009. 52. Yost K, Perkins C, Cohen R, et al. Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes & Control. 2001;12(8):703- 11. 93 53. O’Rorke M, Black C, Murray L, et al. Do perinatal and early life exposures influence the risk of malignant melanoma? A Northern Ireland birth cohort analysis. Eur J Cancer. 2013;49(5):1109-16. 54. Pappo A. Melanoma in children and adolescents. Eur J Cancer. 2003;39(18):2651-61. 55. Strouse JJ, Fears TR, Tucker MA, et al. Pediatric melanoma: risk factor and survival analysis of the surveillance, epidemiology and end results database. J Clin Oncol. 2005;23(21):4735-41. (doi: 23/21/4735 [pii]). 56. Ribero S, Glass D, Aviv A, et al. Height and bone mineral density are associated with naevus count supporting the importance of growth in melanoma susceptibility. PLoS One. 2015;10(1):e0116863. (doi: 10.1371/journal.pone.0116863 [doi]). 57. Cockburn M, Hamilton A, Mack T. The simultaneous assessment of constitutional, behavioral, and environmental factors in the development of large nevi. Cancer Epidemiol Biomarkers Prev. 2007;16(2):200-6. (doi: 1055-9965.EPI-06-0273 [pii]). 58. Schellong K, Schulz S, Harder T, et al. Birth weight and long-term overweight risk: systematic review and a meta-analysis including 643,902 persons from 66 studies and 26 countries globally. PloS one. 2012;7(10):e47776. 59. GenoMEL Consortium. A variant in FTO shows association with melanoma risk not due to BMI. Nat Genet. 2013;45(4):428-32. 94 60. National Research Council, Rasmussen KM and Yaktine AL (Eds). Weight gain during pregnancy: reexamining the guidelines. National Academies Press, 2010. 61. STAMNES KOEPP UM, FROST ANDERSEN L, DAHL-JOERGENSEN K, et al. Maternal pre-pregnant body mass index, maternal weight change and offspring birthweight. Acta Obstet Gynecol Scand. 2012;91(2):243-9. 62. Deierlein AL, Siega-Riz AM, Adair LS, et al. Effects of pre-pregnancy body mass index and gestational weight gain on infant anthropometric outcomes. J Pediatr. 2011;158(2):221-6. 63. Rode L, Hegaard HK, Kjaergaard H, et al. Association between maternal weight gain and birth weight. Obstet Gynecol. 2007;109(6):1309-15. (doi: 109/6/1309 [pii]). 64. Fraser A, Tilling K, Macdonald-Wallis C, et al. Association of maternal weight gain in pregnancy with offspring obesity and metabolic and vascular traits in childhood. Circulation. 2010;121(23):2557-64. (doi: 10.1161/CIRCULATIONAHA.109.906081 [doi]). 65. Paller AS, Hawk JL, Honig P, et al. New insights about infant and toddler skin: implications for sun protection. Pediatrics. 2011;128(1):92-102. (doi: 10.1542/peds.2010-1079 [doi]). 66. Hausauer AK, Swetter SM, Cockburn MG, et al. Increases in melanoma among adolescent girls and young women in California: trends by socioeconomic status and UV radiation exposure. Arch Dermatol. 2011;147(7):783-9. 95 67. Kaplan MS, Huguet N, Newsom JT, et al. The association between length of residence and obesity among Hispanic immigrants. Am J Prev Med. 2004;27(4):323-6. 68. Miller KA, Huh J, Unger JB, et al. Patterns of sun protective behaviors among Hispanic children in a skin cancer prevention intervention. Prev Med. 2015;81:303-8. 69. Cockburn M, Hamilton A, Mack T. Recall bias in self-reported melanoma risk factors. Am J Epidemiol. 2001;153(10):1021-6. 96 CHAPTER 4: BARRIERS TO CANCER-RELATED FOLLOW-UP CARE AMONG YOUNG ADULT SURVIVORS OF MELNAOMA 97 ABSTRACT Melanoma is a preventable, but deadly type of skin cancer. Subsequent melanoma risk is 13 times higher in young melanoma survivors and prognosis is poor at later stages. Young Adult (YA) survivors encounter a variety of barriers interfering with access to follow-up care, placing them at increased risk of delayed cancer-related follow-up care (CRFC). Using a population-based survey study of melanoma survivors (diagnosed at ages ≤24) recruited via Los Angeles County’s cancer registry, we found that, after adjustment, financial barriers (i.e. cost, no health insurance, etc.) were associated with >12 times the risk of lacking a recent CRFC visit compared to persons with no financial barriers (OR: 12.17; 95%CI: 2.69, 55.10). Health system/organizational barriers (i.e. not having a doctor to call, not knowing where to go, etc.) were associated with >41 times the risk, compared to those without health system/organizational barriers (OR: 41.77; 95%CI: 3.97, 439.58). Programs designed to address financial concerns are needed to optimize YA access to recommended medical follow-up. Because such programs may take time to establish, or may not be available for a given setting, enhanced clinical encouragement of lower-cost strategies for early melanoma detection, such as regular total body skin self-exams, may be appropriate. 98 BACKGROUND Melanoma, a preventable but potentially deadly form of skin cancer, has been occurring at increasing rates among adolescents and young adults (AYAs; ages 15-39 years at diagnosis) for more than three decades. The incidence pattern of melanoma in AYAs immediately distinguishes this age group from older adults (see FIGURE 1, below) (1). In the United States, it is one of the top 5 most common cancers in AYAs. In California, where some of the highest melanoma rates in the world occur, there are about 1,000 new AYA melanomas annually. FIGURE 4.0.1. SEX DIFFERENCES FOR AGE-ADJUSTED INCIDENCE OF MELANOMA AMONG WHITES IN THE UNITED STATES (1975-2013), BY AGE GROUP (AGES <50 VS. 50+ YEARS) GAPS IN KNOWLEDGE FOR AYA MELANOMA PREVENTION AYA-focused literature has been growing to reflect an emerging specialty of oncology (2- 8), as patterns of incidence, survival, and adherence to medical treatment or cancer- related follow-up care (CRFC) increasingly distinguish AYAs from older adults and 99 children. Although a boost in AYA focused efforts has become apparent in the literature since 2008 (Figure 2, below) (9), a lack of melanoma-specific study at these ages has persisted, despite the growing burden of AYA melanoma and a call to action on skin cancer by the US surgeon general (10), which has recognized a dire need for AYA- specific melanoma study, if we are to improve our understanding of melanoma risk and reduce the impact of this potentially devastating disease. Risk factors in adults over age 50 involve an accumulation of ultraviolet (UV) exposure over many decades of life can lead to melanoma, and childhood UV exposure and sunburns have also been linked to melanomas occurring later in life. However, the relationship between early life UV exposure and melanoma occurrence in children and AYAs has not been well studied. FIGURE 4.0.2. LITERATURE ON ADOLESCENTS AND YOUNG ADULTS HAS GROWN SINCE 2008 BUT STILL LAGS BEHIND STUDIES IN CHILDREN (9, P. VI) 100 Constitutional factors (i.e. light skin, hair, eyes, etc.), genetics or family history of melanoma, and tanning bed use are understood to be risk factors for melanoma in children and AYAs, but potentially modifiable risk factors to inform primary prevention of melanoma in these young patients have not been well-studied, nor have patterns of sun exposure behaviors after their diagnosis, which may provide further insight into risk factors for occurrence of subsequent primary melanomas in younger populations. Development of a subsequent primary melanoma carries implications for both prognosis and quality of life during survivorship. SURVIVAL Improvements to survival among AYAs (irrespective of cancer type) have not matched the gains observed among younger or older patients (11). Beyond broad examinations at the national level among largely white populations, very little detailed information on AYA melanoma survival exists. Unpublished estimates from the Cancer Surveillance Program of Los Angeles County, which will soon release a monograph examining survival among AYAs from 1988-2013, reported sharply contrasting 5-year survival rates for early versus late stage melanomas; survival exceeded 95% for localized stage, dropped to 68% for regional stage, and then fell to a low of just 9% for distant tumors (12). Analysis at the state level using the California Cancer Registry data (see Chapter 2) provides a deeper look at melanoma-specific characteristics, including Breslow depth, found the impact of thicker tumors and later stages among AYAs to be stronger than was observed among older adults. This only underscores how essential it is for AYA survivors of melanoma to get appropriate follow-up care after diagnosis. 101 IMPORTANCE OF CANCER-RELATED FOLLOW-UP CARE (CRFC) DURING SURVIVORSHIP Among AYA survivors of melanoma, CRFC is a lifelong need. One of the greatest concerns is the risk of developing a second primary. If a patient’s first diagnosis of melanoma occurs before age 30, the risk of a subsequent melanoma is 13 times higher than expected (Figure 3) (13). FIGURE 4.0.3. RELATIVE RISK (OBSERVED TO EXPECTED RATIO [O:E]) OF SUBSEQUENT PRIMARY CANCER AFTER CUTANEOUS MALIGNANT MELANOMA (CMM) BY AGE GROUP, 1973-2006 (13, P.268) Because a late diagnosis has a meaningful impact on years of life lost for AYAs with melanoma, early detection of any subsequent primary is absolutely critical, and may be facilitated by adhering to recommended CRFC guidelines established for melanoma, which generally include having at least one full-body skin exam conducted by a health professional annually (although guidelines tend to vary with stage of disease and 102 country (13). Skipping or delaying CRFC may represent a missed opportunity to detect a subsequent melanoma at a potentially curable stage. FIGURE 4.0.4. A FRAMEWORK FOR UNDERSTANDING THE BARRIERS TO ADHERENCE AMONG ADOLESCENTS AND YOUNG ADULTS (AYAS) (15 P.567) ROLE OF BARRIERS IN LOW ADHERENCE TO MEDICAL ADVICE A significant challenge exists with getting young people to adhere to treatment and follow-up care. This age group is well-established in the literature to encounter significant challenges in their engagement with healthcare, in part attributable to underlying developmental features, which are “biopsychosocial” in nature (14), interfering with an AYA’s ability to be compliant with recommended medical care, adversely impacting adherence to treatment, participation in clinical trials, and 103 attendance at follow-up visits. Most of the available literature has examined barriers in the context of treatment adherence, with far less information available on barriers to CRFC during survivorship (limited to studies of adult survivors of childhood cancers (15) or focused only on adolescents (14), with either very few or no melanoma cases included(16-19) despite the lifelong need and critical role that CRFC may play for AYAs, particularly in early identification of a subsequent melanoma among young melanoma survivors. FIGURE 4.0.5. DIMENSIONS OF AYA BARRIERS TO TREATMENT ADHERENCE (15, P. 571-573) In a review of adherence to treatment among AYAs, Vandermorris et al. described a conceptual framework that includes patients, treatment, and the health system as three important domains underpinning barriers to adherence (Figure 4), then proposed 104 several dimensions of AYA barriers impacting adherence (14), which were adapted and refined to include sociodemographic features (Figure 5 adapted to reflect dimensions described by Vandermorris, et al). Bearing in mind there may be shared barriers to treatment adherence and clinic visits (2, 5, 11, 17, 21, 22), these dimensions may be plausibly be extended across the cancer experience into survivorship, ultimately influencing an AYA melanoma survivor’s decision to get appropriate CRFC. Briefly, these dimensions involve relationship dynamics (i.e. presence of family conflict, social support from friends, engagement with healthcare team, lack of provider continuity), knowledge, beliefs, and attitudes (i.e. inadequate understanding of their disease, how cultural backgrounds and family values may shape patient interpretation of information received), patient emotional and cognitive function (i.e. drive for normalcy, feelings of anxiety, forgetfulness, self-esteem), treatment features and engagement (i.e. “relationship with the healthcare system, rather than the treatment itself” may influence adherence), sociodemographic features (i.e. age, ethnicity, language, educational background, income level, occupation, and geospatial features such as proximity to clinicians, healthy food options, and safe places to exercise). Reported AYA barriers to care have included loss of health insurance (16), contributing to financial burdens already in place for young people working to establish a career and financial independence. Lack of peer networks and social support groups (23) during and after treatment, along with difficult physical changes associated with treatment (24) and poor access to mental health services can unnecessarily increase the unmet needs 105 for young survivors, and may heighten feelings of social isolation. These unmet psychosocial needs may become intensified under financial duress, contributing importantly to an AYA’s ability to get appropriate CRFC, which may ultimately impact survival (Figure 6) (25). FIGURE 4.0.6. FACTORS IMPACTING FINANCIAL DISTRESS AND MORTALITY (25, P. 2) There are also detrimental effects on an AYA’s plans for work or school. In one study, a greater proportion of patients working or attending school full-time prior to diagnosis were able to resume their activities compared to part-time workers/students (72% vs. 34%, for full time and part-time, respectively), but up to 50% of full time workers/students reported experiencing problems at work or school after their diagnosis (FIGURE 7) (25). This may further heighten an AYA’s sensitivity to both financial and psychological distress, as well as their ability to recover from it. 106 FIGURE 4.0.7. IMPACT OF CANCER DIAGNOSIS ON WORK OR SCHOOL STATUS AMONG ADOLESCENT AND YOUNG ADULT PATIENTS (26, P. 2) CONSEQUENCES OF LACKING CANCER-RELATED FOLLOW-UP DURING SURVIVORSHIP Low adherence to CRFC interferes with a clinician’s ability to detect recurrences, second primaries, and other late effects (26, 27). Prevalence of adherence to CRFC among AYA melanoma survivors is unknown, but assumed to be suboptimal, given the generally low adherence and low engagement that has been reported in AYAs irrespective of cancer type. The impact of financial, psychosocial, health system, or other elements on the ability to get CRFC for young melanoma survivors remains unexamined. If adherence to melanoma-specific CRFC is to be improved, it must be determined which barriers are most relevant and potentially actionable for AYAs with melanoma. 107 STUDY AIMS To identify important barriers associated with not having recent CRFC among young melanoma survivors, a population-based cross-sectional study was conducted, employing the use of a novel survey to capture self-reported CRFC, along with risk and protective factors that may be related to not having recent CRFC. Specifically, we hypothesized: 1. Prevalence of recent CRFC would be sub-optimal (>one quarter would report no recent CRFC). 2. Persons lacking recent CRFC would report higher numbers of barriers and encountering more than one type of barriers. 3. Lack of recent CRFC will be associated with lower prevalence of skin-protective behaviors. 4. Financial barriers will be most strongly related to lack of recent CRFC. 108 MATERIALS & METHODS In this cross-sectional study of AYA survivors of melanoma occurring in childhood, adolescence, or young adulthood, data was collected as a sub-study of the parent Project Forward study (ongoing), which is a large, population-based survey of young adults, who were survivors of pediatric cancer (up to age 18 at diagnosis, but expanded to age 24 for the melanoma sub-study). All cases in the parent study and sub-study had a cancer diagnosis during the period of January 1, 1996 through December 31, 2010, as identified from the Cancer Surveillance Program (CSP), which is the cancer registry of Los Angeles (LA) County. Briefly, Project Forward’s aims were to (1) identify risk and protective factors related to CRFC in Hispanic and non-Hispanic childhood cancer survivors, (2) determine if follow-up care received is consistent with published guidelines for long-term follow-up care, (3) examine milestones of emerging adulthood and receipt of CRFC, and (4) identify risk and protective factors for CRFC among persons diagnosed with melanoma up to 24 years of age. For this study, cases of melanoma (ICD-O-3 site C440-449 & histology 8720-8790; all stages except in situ) were included if diagnosed during childhood (ages <15 years) or as an AYA (ages 15-24 years) in 1996-2010 and were 18-39 years of age at study start in 2015 (birth years 1976-1997; n=292 total, n=287 unique eligible cases; target enrollment n=100). For recruitment and tracking purposes, cases were sorted by age and grouped into batches (Batch 1 included 150 eligible cases with oldest diagnosis ages; Batch 2 included the remaining 137 eligible cases). Exclusion criteria were: primary language 109 other than English or Spanish, incapable of completing survey with caregiver assistance due to illness/significant cognitive impairment, age older than 39 or younger than 18 at enrollment, or identified as deceased (either through recruitment activities or passive updates from the registry). Cases of melanoma found with multiple primaries (n=4; 3 people had n=2 primaries, 1 person had n=3 primaries) were not excluded from this sub- study, because these cases reflected multiple primaries, which were all melanoma and were diagnosed on the same date within a given individual; thus, any concerns regarding survey response in the context of multiple primaries and/or multiple sites would not apply. Study Survey. A melanoma-specific survey (APPENDIX B), along with skin-care related questions (APPENDIX B) and a residential history form (APPENDIX D) were appended to the main Project Forward survey, as part of this study. Participants were queried about their last CRFC (“Within the past year,” “1-2 years ago,” “More than 2 years ago,” “Never”), recent barriers to care (categorized as financial, health system, psychosocial, or work/school related), skin care behaviors, and residential history. All other variables, including demographics (sex, race/ethnicity, age at diagnosis, socioeconomic status) and tumor related characteristics (year of diagnosis, site, histology, stage at diagnosis, etc.) were obtained through existing registry data. Recruitment Procedures. After eligible cases were identified from the registry, contact was initiated through mailings as noted in the protocol (see APPENDIX G). After mailing the initial survey packet, follow-up included mailing reminder letters and postcards, as 110 well as telephone calls based on registry contact information. Follow-up activity commenced in April 2015 and was considered completed for an individual at the earliest occurrence of any of the following: (1) doctor refusal for patient contact, (2) patient or gatekeeper refusal (written or verbal), (3) discovery that the patient had become deceased or was incompetent, (4) patient denied cancer, or (5) patient was unreachable and/or lost to all efforts for recruitment (i.e. never heard from patient, passive refusal, promised to send, out of the country, not traceable) at melanoma study end date (December 2016). Outcome of Recent CRFC. Recent CRFC (yes/no) was captured by self-report with a question asking participants when their most recent cancer-related follow up visit occurred (see APPENDIX G for details). Responses of “within the past year” or “1-2 years ago” were classified as having recent CRFC (yes), and responses of “more than 2 years ago” or “never” were classified as not having recent CRFC (no). Barriers to Care. Barriers were captured by self-report with a two-part question asking if a participant was unable to see a doctor at any time in the past year; if yes, they were asked to let us know why (see APPENDIX G for details). There were 17 available responses, one of which allowed for free text to capture any reasons that may not have been listed. Total numbers of barriers were determined by combining their yes/no response in the first part of the question (if any barriers were encountered), with the number of specific reasons (barriers) endorsed in the second part of the question. Barriers were also classified into 4 main types: 111 • Financial (i.e. “I had no transportation to the doctor,” “It costs too much,” “I had no health insurance,” “My insurance company didn't approve a test or visit,” “Other-I anticipated limits and financial burden”) • Health system (i.e. “I didn’t have a doctor that I could call,” “I didn’t know where to go,” “I didn’t have a doctor that spoke my language,” “I don’t have a doctor that understands my background and values,” “I don’t know how to make an appointment”) • Work/school related (i.e. “I was too busy,” “I didn’t have childcare,” “I couldn’t leave work or school”) • Psychosocial (i.e. “I didn’t want to think about it,” “I was worried about finding a problem,” “I don’t trust doctors or don’t think they can do anything to help me at this time,” “Other-Anxiety”) 112 STATISTICAL ANALYSIS. Frequencies and means were used to describe the sample, along with chi-squared and t- tests to examine differences in distribution by CRFC status; alpha level 0.05 was used to identify statistically significant differences. Binary logistic regression was used to estimate the association between barriers and odds of not having recent CRFC. Criteria for presence of confounding was >10% change in the beta estimates; confounders were included in the final model based on these criteria and a priori knowledge. In the sensitivity analysis, persons originally classified as lacking recent CRFC, who also reported separately that a health professional had examined their skin within 2 years, were reclassified as having recent CRFC and all models examining barriers (number or type) were re-analyzed with the revised CRFC classification to determine if any impact on the results would be observed. All analyses were conducted in SAS v9.4 (SAS Institute, Inc., Cary, North Carolina). Approval for this study was obtained from the University of Southern California Health Science Institutional Review Board (USC-HSIRB). Because CSP’s registry is a member of the California Cancer Registry (CCR), approval was also required and obtained from both the CCR and the California State Committee for the Protection of Human Subjects (CPHS). 113 RESULTS Out of 287 young melanoma survivors initially selected from the registry, a total of 8 were known to be deceased. Of the remaining 279, there were 111 persons who participated in this study (response rate 40%). Participants consisted of 5% ages <15 years at diagnosis, 27% ages 15-19 at diagnosis, and 68% ages 20-24 at diagnosis (mean (SD) age was 20.1(3.6)). Most were female (61%) and non-Hispanic white (82%). There were 10% Hispanic whites and 34% of surveys were submitted online. Demographics and tumor characteristics did not differ by CRFC status (Table 1), except by sex, with females more likely to report recent CRFC. Barriers were reported more often, and in higher numbers, among those without recent CRFC (Table 2), which was also true across all types of barriers. Of those reporting barriers (n=27), the most commonly barriers were “It costs too much” (n=12), “I had no health insurance” (n=11), “I was too busy” (n=9), “I couldn’t leave work or school” (n=6), “My insurance company didn't approve a test or visit” (n=6), and ”I didn’t know where to go” (n=5). Overall, most participants (73%) reported high levels of perceived risk of a subsequent melanoma (Table 3), but slightly more of those with recent CRFC reported lower levels of perceived risk (21% vs. 16% for those with and without recent CRFC, respectively). Persons lacking recent CRFC reported less frequent sunscreen use (36% always used sunscreen) compared to those with recent CRFC (63% always used sunscreen) and less frequent annual skin exams (74%) compared to those in the recent CRFC group (97%). Although the differences in tanning bed use were not statistically significant, usage was higher in the group lacking recent CRFC (3% tanned 1-10 times in the past year, while 114 only one 1% reported tanning bed use one time in the past year in the recent CRFC group; P=0.30). Univariate and adjusted model results are shown in Table 4. All barriers (number and type) were associated with lacking recent CRFC in the univariate models. Although no presence of confounding by demographic or tumor characteristics was observed except with sex of the patient, all of these factors were included in the full model due to their theoretical confounding influence. In adjusted Model 1, reporting any barrier was associated with 5.67 times the odds of lacking recent CRFC, when compared to reporting no barriers (OR: 5.67; 95%CI: 1.95, 16.50). In adjusted Model 2, having 1≤3 barriers was associated with 3.59 times the odds of lacking recent CRFC (OR: 3.59; 95%CI: 1.13, 11.38) compared to those with no barriers; financial barriers were most prevalent for this group, with “it costs too much” (n=5) and “I had no health insurance” (n=6) being the most commonly endorsed in this group, but these did not appear from visual inspection to have any tendencies to co-occur with other categories of barriers. On the other hand, having >3 barriers was associated with 43.96 times the odds of lacking recent CRFC (OR: 43.96; 95%CI: 3.57, 541.75), and two types of barriers were most prevalent in this group: health system barriers (“I didn’t know where to go” (n=4)) and Financial barriers (i.e. “it costs too much” (n=3), “I had no health insurance” (n=3)). From visual inspection, it appeared among people with >3 barriers and no recent CRFC, health system barriers tended to co-occur with 115 psychosocial barriers, and financial barriers tended to co-occur with work/school barriers. When examining barriers by type (adjusted Models 3-6), financial barriers were associated with more than 12 times the odds of no recent CRFC (OR: 12.17; 95%CI: 2.69, 55.10) compared to those without financial barriers. Health system barriers were associated with more than 41 times the odds of no recent CRFC (OR: 41.77; 95%CI: 3.97, 439.58) compared to those without health system barriers. Work/school barriers were associated with more than 6 times the odds of no recent CRFC (OR: 6.70; 95%CI: 1.56, 28.84) compared to those without work/school barriers. Psychosocial barriers carried 8 times the odds of no recent CRFC (OR: 8.01; 95%CI: 1.05, 61.03), when compared to the group with no psychosocial barriers. From sensitivity analysis (Table 5), magnitudes were generally attenuated but effect sizes remained large and statistically significant; only the effect of work/school barriers did not remain statistically significant after sensitivity analysis. 116 DISCUSSION In this population-based study of young adult (YA) survivors of melanoma diagnosed in childhood, adolescence, or young adulthood, greater numbers of barriers and financial or health system barriers were most strongly associated with not having a recent cancer-related follow-up care (CRFC) visit. Since young melanoma survivors are at substantially increased risk of another melanoma, falling out of alignment with recommended CRFC may place these survivors at increased risk of a late diagnosis of any subsequent melanoma, which has poor prognosis. The impact on YA melanoma survivors has not, to our knowledge, been previously explored or quantified. Presence of any barrier was associated with more than 5 times the risk of not having a recent CRFC visit, while having greater than three barriers increased this risk by more than 40 times, compared to those with no barriers. CRFC. While the results showed higher levels of barriers were associated with higher likelihood of not having a recent CRFC visit, the results from this pilot study require replication in a setting with a larger sample. Young adults are more sensitive to financial pressures compared to older adults who may have more life experience, established careers, salaries, and have more experience managing finances while dealing with health care needs (28). To paraphrase Freyer et al, this includes experiencing greater financial impact as evidenced by larger losses in productivity than are experienced for older adults and having greater difficulty staying on health insurance (28). 117 The results of this study add evidence of an important role for work or school barriers and psychosocial barriers for young survivors of melanoma, showing there was 6-8 fold increased risk of not having recent CRFC with these types of barriers. Detrimental effects on an AYA’s plans for work or school have been demonstrated in literature among AYAs as an age group (but not for melanoma specifically), and have showed that a greater proportion of patients with full time work or school responsibilities prior to diagnosis were able to resume their activities compared to part-time workers/students (72% vs. 34%, for full time and part-time, respectively). Despite this, up to 50% of full time workers/students reported experiencing problems at work or school after their diagnosis (25). Although that study of work/school barriers was limited to evaluating the impact on AYAs as an age group (all cancer types combined) and did not evaluate melanoma-specific effects, it is possible that young survivors of melanoma could become primed toward further decreasing adherence to future CRFC under similar circumstances. If one such survivor has already experienced adverse impacts on work or school during diagnosis, they may worry about further compromising their position at work or school by taking additional time to pursue CRFC, despite the fact that it could be a life-saving measure. Poor access to mental health services may unnecessarily exacerbate distress among young survivors and heighten feelings of social isolation, resulting in unmet psychosocial needs related to lack of peer networks, social support groups (23) during and after treatment, and dealing with difficult physical changes associated with treatment (24), which may converge to negatively impact an AYA’s ability to get appropriate CRFC. Developing validated measures for capturing these 118 types of unmet needs is an ongoing area of research that also requires more attention (29). Bradford, et al demonstrated that when a patient’s first diagnosis of melanoma occurs before age 30, the risk of a subsequent melanoma is 13 times higher than expected, making it essential for AYA survivors of melanoma to get appropriate follow-up care (30). In light of this elevated risk, we sought to also understand patterns of skin- protective behaviors and perceived risk of a subsequent melanoma according to recent CRFC status. By doing this, we hoped to reveal opportunities to bolster uptake of these behaviors and inform more cost-effective strategies for survivors lacking the resources to see a physician for CRFC. The presence of high levels of perceived risk among persons without recent CRFC (79%) may suggest this group has the potential to become more motivated to increase skin protective behaviors, given proper guidance and/or reminders. Persons lacking recent CRFC also tended to report using sunscreen less often, less frequent skin exams, and had greater tanning bed usage, which may impose additional risk given their history of melanoma being established at a young age; these patterns suggest room for improvement in preventative behaviors and an opportunity to also improve early detection of a subsequent melanoma. At a minimum, it may be helpful to increase “at home” skin-screening behaviors, which may increase the likelihood of early detection of a suspicious growth. However, it remains unclear if this would lead to earlier diagnosis, as they would need to get a clinical diagnosis, which may be a 119 challenge in itself if they still face insurmountable barriers to care. The question remains as to what is the tipping point to getting a patient into a clinician’s office for the final diagnosis and treatment. Although both frequency and timing of professional skin exams were generally in agreement with CRFC status, there was some discrepancy among those lacking recent CRFC, given that only 34% of the patients reported that a physician checked their skin within the past 2 years compared to 92% of those in the recent CRFC group (P <.0001). This prompted a sensitivity analyses, which showed the impact of barriers (number or type) persisted; the magnitudes were generally attenuated but remained large, and only the effect of work/school barriers lost statistical significance, possibly due to limitations in sample size for this category. SURVIVORSHIP & DENIAL OF CANCER Most newly diagnosed melanomas in AYAs are detected at earlier stages and treated by surgical removal of the tumor, which typically occurs in the primary care or dermatologist’s office and involves short recovery time; it can be speculated from this scenario that a patient may not perceive this as a “real” cancer experience or may dismiss the gravity of their diagnosis, and it is unclear to what degree information surrounding their diagnosis has been relayed to the patient by the clinician. To our knowledge, these perceptions around a melanoma diagnosis have yet to be studied and would be informative to determining areas to improve patient education in the primary stages of this diagnosis. In contrast, a patient with late stage melanoma may require a 120 greater extent of tissue removal and/or radiation, chemotherapy, or immunotherapy treatment occurring at a hospital setting with longer recovery time and possibly greater degree of scarring from tumor removal, depending on a number of factors, including need for skin grafts and the surgical method employed. The late-stage patient may be more likely than the early-stage patient to view this as a “real” cancer, owing to a more “intense” treatment and recovery experience, and may exhibit different adherence to follow-up care and skin protection. The present study was not powered to examine differences in denial of cancer by stage, which would be challenging given poor prognosis at later stages. Anecdotally, it seemed that denial of cancer may have occurred more frequently among the melanoma cases than for other types of cancer in the parent Project Forward study, but data collection is ongoing and it remains to be seen if this observation will be upheld within the final set of participants. However, given the plausibility that those with early stage melanoma may be potentially more likely to deny history of cancer, it may be informative to compare early stage melanoma in AYAs with other early stage cancers among AYAs to understand if there is greater frequency of non-confirmation of cancer for melanoma. Furthermore, it may be beneficial to identify if there are any similarities with other “low- impact” early stage cancers, such as cervical or colorectal cancer, in which early stage disease can also involve outpatient procedures for removal of a suspicious lesion and relatively low-impact recovery time. In a multi-center study of recently diagnosed colon cancer patients, characteristics associated with non-confirmation of cancer included localized stage, lower income or education, older age, and non-white background 121 (Personal communication with HR Newman regarding unpublished data presented at ASPO conference March 2017). It remains unclear how much the denial/non- confirmation is attributable to having a low-impact procedure, beyond other factors such as educational level, cultural background, or level of patient engagement with the provider. Denial or non-confirmation of melanoma as a cancer diagnosis may also contribute to less frequent CRFC and sub-optimal skin-protective behaviors (i.e. inconsistent use of sunscreen) among AYA survivors of melanoma. As with many other areas of AYA melanoma research, this too remains understudied. STRENGTHS & LIMITATIONS This pilot study provides novel, population-based, highly detailed information on barriers to care that are specific to young adults who are survivors of childhood or AYA melanoma in the diverse population of LA County. This information has not been previously available in the literature, but there are important limitations. The present study was cross-sectional in nature and may not entirely reflect the true temporal relationship between barriers and follow-up care, and was also relatively small in sample size, limiting our ability to observe effects stratified by barrier type, sex, or age group. Barriers were self-reported, which may be affected by an individual’s ability to recall correct information from the intended time period or may be influenced by perceptions of what the “correct” answer might be. Participation in the study survey necessitated recruitment of cases not known to be deceased at the start of the study, and inclusion of late-stage cases via rapid case 122 ascertainment or proxy responses (i.e. family, friend, or caretaker) was not within the scope of the present study. While survivor bias may be present, it would be unlikely to attenuate the effects observed, but the present study assumes relevance to non- participants, who may have died (worst disease) or chosen not to participate due to lack of perceived importance of their disease experience. However, barriers encountered by late stage cases would likely produce larger effects than observed for inability to obtain follow-up care, or may, perhaps more importantly for these patients, affect quality of life and ability to get appropriate hospice care. Even so, the number of persons known to be deceased was small (n=8) and late stage disease generally represents a small proportion of all melanoma diagnoses. A number of eligible cases denied having cancer and did not participate (n=33). If all of these cases had early stage disease and/or were diagnosed at the younger end of eligible cases (i.e. diagnosed before age 18), it is possible that they would have had greater opportunity to encounter barriers and may also be more likely to not have received recent CRFC, due their denial. The reverse may be true if all deniers were late stage and/or were diagnosed at the older end of the eligible ages (i.e. 20-24 years), who may have had greater opportunity to learn how to overcome barriers as a function of time, having already entered adulthood at the time of diagnosis and having more time as an adult to gain financial or career stability, independence, and greater experience with navigating the health system. 123 GENERALIZABILITY The results should be generalizable to survivors of melanoma diagnosed at ≤24 years of age in LA County, but may also be applicable to persons with melanoma diagnosed in areas of similar geographical, ethnic, and UV exposure characteristics. Response rates were good (40%) and in alignment with expected recruitment levels among cancer survivors. Additionally, a comparison of participants to non-participants in terms of demographics or tumor characteristics, if shown to not differ, would provide further evidence to support this study was representative of the population of young melanoma survivors diagnosed in LA County who were diagnosed in childhood, adolescence, and young adulthood. If, for example, study participants differed in a statistically significant manner from non-participants by having a larger proportion of females, then results may be more reflective of females, rather than representative of both sexes, and efforts to boost male participation in future studies would become important to improving generalizability of similar studies. Importantly, it is unknown if non-participants were influenced by changes in SES status as a direct result of their melanoma diagnosis (i.e. loss of job, reduction in net income due to cost of treatment and/or time off work, etc.). Young adults may be more sensitive to financial pressures during survivorship, because they are balancing those demands with striving to attain life’s milestones (i.e. education, career, or family goals). Changes in SES could interfere with both ability to participate and with ability to adhere to recommended follow-up care, and missing these persons may lead to an underestimation of the impact of financial and/or health system barriers on a young adult’s ability to get appropriate follow-up care. 124 The results would not apply to persons with prior history of cancer due to inherently increased risk of developing another cancer (such as a second primary) after the first cancer diagnosis, possibly due to genetic susceptibility or first cancer treatment, and changes in knowledge of and ability to navigate the cancer treatment process as a result of their experience with a prior diagnosis; the results would also not apply to persons over the age 24 at diagnosis, who may have differences in underlying biology, cumulative exposures related to cancer etiology, access to healthcare, or life experience, which may confer a different risk profile for getting CRFC than that seen among younger persons. 125 CONCLUSION Cancer-related follow-up care (CRFC) among young survivors of melanoma is a lifelong need. Lack of recent CRFC may represent a missed opportunity to detect a subsequent, potentially curable, melanoma, since one of the greatest concerns for melanoma survivors is their increased risk of developing a second primary. From a prevention standpoint, it is concerning that skin protective behaviors, such as sunscreen or tanning bed use, were suboptimal for those without recent CRFC. Targeted counseling of young melanoma patients about their increased risk of a subsequent melanoma may encourage greater adoption of skin-protective behaviors during survivorship, and such counseling may be especially beneficial for those known to have limited financial resources or other barriers at the time of diagnosis. YA are uniquely sensitive to financial pressures, as they strive for educational, career, and family goals. Even early stage diagnosis can have financially destabilizing effects, because modification of work or school attendance may be necessary to accommodate CRFC, or may not be supported by the employer or school. Programs designed to address financial concerns are needed to optimize access to recommended follow-up, but may take time to establish, or may not be available for a given setting, so enhanced encouragement of lower-cost strategies for early melanoma detection, such as regular total body skin self-exams, may facilitate early detection among patients at greatest risk of not having optimal follow-up care. 126 TABLE 4.0.1. YOUNG ADULT SURVIVORS OF MELANOMA DIAGNOSED AT AGES 0-24 YEARS IN LOS ANGELES COUNTY (1996-2010), BY CANCER-RELATED FOLLOW-UP CARE STATUS (CRFC)* 127 TABLE 4.0.2. BARRIERS TO CARE AMONG YOUNG ADULT SURVIVORS OF MELANOMA IN LOS ANGELES COUNTY (1996-2010)* 128 TABLE 4.0.3. SKIN PROTECTIVE BEHAVIORS AMONG YOUNG ADULT MELANOMA SURVIVORS IN LOS ANGELES COUNTY (1996-2010)* 129 TABLE 4.0.4. BARRIERS ASSOCIATED WITH LACK OF RECENT FOLLOW-UP CARE (CRFC) AMONG YOUNG ADULT MELANOMA SURVIVORS IN LOS ANGELES COUNTY (1996-2010) 130 CHAPTER 4 REFERENCES 1. National Cancer Institute (NCI). SEER fast stats. http://seer.cancer.gov/faststats/selections.php?#Output. Accessed June, 2015. 2. Adolescent and Young Adult Oncology Progress Review Group (AYAO PRG), ed. Closing the gap: Research and care imperatives for adolescents and young adults with cancer (NIH publication no. 06-6067). Department of Health and Human Services, National Institutes of Health, National Cancer Institute, and the LIVESTRONG Young Adult Alliance Bethesda, MD; 2006. 3. Bleyer A, O'leary M, Barr R, Ries L. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. 2006. 4. Bleyer A, Barr R, Hayes-Lattin B, Thomas D, Ellis C, Anderson B. The distinctive biology of cancer in adolescents and young adults. Nature Reviews Cancer. 2008;8(4):288-298. 5. National Cancer Institute (NCI). Adolescents and young adults with cancer. National Cancer Institute Web site. http://www.cancer.gov/types/aya. Published May 15, 2015. Updated May 15, 2015. Accessed 07/12, 2015. 6. Sender L, Zabokrtsky KB. Adolescent and young adult patients with cancer: A milieu of unique features. Nature Reviews Clinical Oncology. 2015. 131 7. Thomas DM, Albritton KH, Ferrari A. Adolescent and young adult oncology: An emerging field. J Clin Oncol. 2010;28(32):4781-4782. 8. Shaw PH, Reed DR, Yeager N, Zebrack B, Castellino SM, Bleyer A. Adolescent and young adult (AYA) oncology in the united states: A specialty in its late adolescence. J Pediatr Hematol Oncol. 2015;37(3):161-169. 9. Bleyer A, Barr R, Ries L, Whelan J, Ferrari A, eds. Cancer in adolescents and young adults. Second ed. Cham, Switzerland: Springer International Publishing; 2017. Bleyer Achie, Barr Ronald, Ries Lynn, Whelan Jeremy and Ferrari Andrea, eds. Pediatric Oncology. Accessed January 2017. 10.1007/978-3-319-33679-4. 10. U.S. Department of Health and Human Services, ed. The surgeon general's call to action to prevent skin cancer. Washington, DC: U.S. Dept of Health and Human Services, Office of the Surgeon General; 2014. 11. Bleyer WA. Cancer in older adolescents and young adults: Epidemiology, diagnosis, treatment, survival, and importance of clinical trials. Med Pediatr Oncol. 2002;38(1):1- 10. 12. Wojcik K, In G, Navid F, eds. Melanoma survival in adolescents and young adults in los angeles county. Los Angeles, California: Los Angeles Cancer Surveillance Program, University of Southern California; Forthcoming (in-press) 2017. Liu L, Hwang A, Moke D, et al, eds. Cancer in Los Angeles County: Survival among Adolescents and Young Adults 1988-2013. 132 13. Bradford PT, Freedman DM, Goldstein AM, Tucker MA. Increased risk of second primary cancers after a diagnosis of melanoma. Arch Dermatol. 2010;146(3):265-272. 14. Trotter SC, Sroa N, Winkelmann RR, Olencki T, Bechtel M. A global review of melanoma follow-up guidelines. Journal of Clinical & Aesthetic Dermatology. 2013;6(9). 15. Vandermorris A, Parson KW, Greenberg ML. Adherence to treatment regimens in adolescent and young adult cancer patients. In: Bleyer A, Barr R, Ries L, Whelan J, Ferrari A, eds. Cancer in adolescents and young adults. Second ed. Springer; 2017:565. 16. Oeffinger KC, Mertens AC, Hudson MM, et al. Health care of young adult survivors of childhood cancer: A report from the childhood cancer survivor study. Ann Fam Med. 2004;2(1):61-70. 17. Keegan TH, Tao L, DeRouen MC, et al. Medical care in adolescents and young adult cancer survivors: What are the biggest access-related barriers? Journal of Cancer Survivorship. 2014;8(2):282-292. 18. Butow P, Palmer S, Pai A, Goodenough B, Luckett T, King M. Review of adherence- related issues in adolescents and young adults with cancer. J Clin Oncol. 2010;28(32):4800-4809. 19. Robison LL, Mertens AC, Boice JD, et al. Study design and cohort characteristics of the childhood cancer survivor study: A multi-institutional collaborative project. Med Pediatr Oncol. 2002;38(4):229-239. 133 20. Potosky AL, Harlan LC, Albritton K, et al. Use of appropriate initial treatment among adolescents and young adults with cancer. J Natl Cancer Inst. 2014;106(11):10.1093/jnci/dju300. Print 2014 Nov. 21. Bleyer A. The adolescent and young adult gap in cancer care and outcome. Current problems in pediatric and adolescent health care. 2005;35(5):182-217. 22. Ferrari A, Montello M, Budd T, Bleyer A. The challenges of clinical trials for adolescents and young adults with cancer. Pediatric blood & cancer. 2008;50(S5):1101- 1104. 23. Kent EE, Smith AW, Keegan TH, et al. Talking about cancer and meeting peer survivors: Social information needs of adolescents and young adults diagnosed with cancer. Journal of adolescent and young adult oncology. 2013;2(2):44-52. 24. Barakat LP, Galtieri LR, Szalda D, Schwartz LA. Assessing the psychosocial needs and program preferences of adolescents and young adults with cancer. Support Care Cancer. 2015. 25. Zafar SY. Financial toxicity of cancer care: It's time to intervene. J Natl Cancer Inst. 2015;108(5):10.1093/jnci/djv370. Print 2016 May. 26. Parsons HM, Harlan LC, Lynch CF, et al. Impact of cancer on work and education among adolescent and young adult cancer survivors. J Clin Oncol. 2012;30(19):2393- 2400. 134 27. Landier W, Bhatia S, Eshelman DA, et al. Development of risk-based guidelines for pediatric cancer survivors: The children's oncology group long-term follow-up guidelines from the children's oncology group late effects committee and nursing discipline. J Clin Oncol. 2004;22(24):4979-4990. 28. Kremer L, Mulder RL, Oeffinger KC, et al. A worldwide collaboration to harmonize guidelines for the long-term follow-up of childhood and young adult cancer survivors: A report from the international late effects of childhood cancer guideline harmonization group. Pediatric blood & cancer. 2013;60(4):543-549. 29. Freyer DR, Smith AW, Wolfson JA, Barr RD. Making ends meet: Financial issues from the perspectives of patients and their health-care providers. In: Bleyer A, Barr R, Ries L, Whelan J, Ferrari A, eds. Cancer in adolescents and young adults. Second ed. Springer; 2017:667. 30. Clinton-McHarg T, Carey M, Sanson-Fisher R, Shakeshaft A, Rainbird K. Measuring the psychosocial health of adolescent and young adult (AYA) cancer survivors: A critical review. Health Qual Life Outcomes. 2010;8(1):25. 135 CHAPTER 5: CONCLUSION 136 IMPLICATION S & IMPLEMENTATION Melanoma-specific study in young populations remains scarce in the literature, despite the fact that it is among the top 5 cancers occurring in AYAs. Over the past decade, AYAs have been increasingly recognized as a group distinct from children and older adults (1). Although awareness of the need to prioritize this age for research has grown, site- specific studies on melanoma are still lacking. Of the existing studies including any AYA melanomas, most have been limited by inconsistent or narrow definitions of AYA age ranges and often exclude minority cases, hampering our ability to detect effects in AYAs or to identify any important, underlying disparities. These limitations persist across investigations of melanoma survival, risk factors, and survivorship among AYAs. We sought to add melanoma-specific insight across these areas to provide information with potentially immediate utility for AYA patient care and to spur additional study that may hasten improvements in melanoma cancer control for young populations. In the first study (Chapter 2), we undertook a population-based registry study to describe melanoma survival among children, AYAs, and older adults in the diverse population of California, aiming to identify important differences between the age groups, and any underlying disparities. We found that AYAs had larger magnitudes for risk of death than observed among their older adult counterparts, an effect that persisted after adjustment for demographic and tumor characteristics. There is some evidence in the literature that melanomas occurring at younger ages may be more aggressive (2-7), implying potential for poorer prognosis that underscores the importance of identifying risk factors amenable to intervention to reduce incidence and 137 increase awareness to enhance early detection for AYA melanoma. It has been previously reported that males experience a disadvantage in survival, however in our study, this disadvantage appeared to be worse for AYA males than was observed for older adult males. Because this was a records based-study, we had no information on sun exposure behaviors, family history of melanoma or recurrence, as these are not variables routinely collected by the registry and the reason for the disparity remains unknown, but greater clinical follow-up of young males may be beneficial. In the second study (Chapter 3), we conducted a population-based case-control study of melanoma to identify potentially modifiable or intervenable risk factors for development of melanoma before age 30. Birth weight and early life UV were the risk factors of interest. An objective, validated measure of UV was used to assign ambient solar UV exposure from the mother’s address on the birth record to represent early life exposure to UV. This method overcomes existing limitations of survey-based measures of UV in the literature, which include recall bias and question temporality. Birth weight information was obtained from the birth record. We found evidence to support roles for both high birth weight and higher early life UV exposure. These results suggest that UV exposure may not need decades to exert its influence on melanoma risk. Skin protective behaviors could be recommended as early as infancy to aid in the reduction of melanoma at young ages, while closer monitoring of AYAs known to have had high birth weight may enhance early detection. 138 In our third study (Chapter 4), we conducted a population-based survey study aimed at identifying barriers related to lack of follow-up care among AYAs with melanoma. To date, there is almost no melanoma-specific study of survivorship for this age group. We found that greater numbers of barriers and financial or health system barriers were most strongly implicated for lacking recent cancer-related follow-up care. Since young melanoma survivors are at substantially increased risk of another melanoma, follow-up care may play an important role in early detection of any subsequent melanoma, which has poor prognosis. Poor adherence to recommended skin-protective behaviors suggests that enhanced clinical encouragement of the importance of skin protection may aid in both reducing risk of future melanoma and enhancing potential for early detection via skin-self exams. Increased implementation of sun protection and self- exams may be particularly beneficial to patients lacking resources and health insurance to cover professional skin exams. Greater awareness of these patient needs may help clinicians identify patients at high risk for low adherence to clinical follow-up, so that targeted counseling on the seriousness of their diagnosis and importance of self-care and monitoring may be delivered. 139 FUTURE DIRECTIONS Knowledge of survival characteristics for melanoma among AYAs may be further advanced by conducting registry-based studies of melanoma-specific factors, such as mitotic rate and sentinel lymph node involvement, which could add evidence that helps to better characterize the underlying tumor biology manifesting as more aggressive disease for AYA melanoma; comparisons by stage to adults and children may also be informative. Prospective clinic-based studies involving tissue collection and sentinel lymph node imaging or exams provide insight around disease progression in AYAs. In turn, this may inform research into underlying mechanisms, and may also inform the use of targeted therapies to reduce the impact of potentially more aggressive disease. Incorporating comparisons to younger and older patients can help inform approaches for treatment or protocol design in clinical trials, which may aid in the establishment of standardized treatment practices for AYA melanoma. There have been tremendous recent improvements in survival due to immunotherapy use for late stage melanomas in older adults, but important policy-level barriers to their inclusion exists and lack of awareness of the need for study in this age group persists. It is unclear if a lag in survival improvement for AYAs with melanoma could result. This presents an important opportunity for advocacy and spreading awareness of the burden of melanoma among AYAs and the importance of their inclusion in clinical trials, so that existing policy, known to interfere with AYA inclusion in clinical trials, may be rectified in the near future, ensuring that AYAs with advanced melanoma may enjoy the same improvements to survival that are currently being experienced among older adults. 140 In terms of risk reduction, we next need to understand if any risk reduction for subsequent melanomas could be achieved with better skin protective habits among a population of young melanoma survivors. Although our results suggest sun protective behaviors should be implemented as early as infancy, there is currently no information regarding existing norms for sun exposure behaviors among infants, particularly by birth weight category, which would be important to understand given the association of increasing birth weight and early-onset of melanoma. A prospective study of expectant mothers could explore sun exposure behaviors in the mother-child dyad, measure individual UV, capture skin types, and assess awareness of skin cancer risk to identify opportunities for encouraging skin protective behaviors in both the mother and infant. A diverse population should be included to understand if potential differences in early sun exposure or skin protective behaviors may further explain some of the differences in early-onset melanoma risk. Higher AYA melanoma rates over the last 30 years have led to larger numbers of AYAs entering survivorship status, which increases the number of persons at risk for subsequent melanoma. Thus, greater study of AYA melanoma survivorship, specifically to prevent subsequent melanoma, or to detect it early, should be prioritized. Our study of barriers was limited due to the cross-sectional nature of the study, which makes it difficult to establish temporality. Future studies may involve a clinic-based cohort to facilitate rapid ascertainment of newly diagnosed cases of melanoma, so that they may be queried about barriers existing at the time of diagnosis, as well as post diagnosis (1 year) and into survivorship (5 or more years post-diagnosis). This may help clarify how 141 the experience of navigating the health system, as well as payment and coverage for treatment may shape AYA capabilities to adhere to follow-up for melanoma. An intervention coupled with follow-up may be useful to determine if targeted intensive counseling and encouragement of self-exams and skin protection among resource- restricted AYAs could lead to improvements in skin-protection behaviors, including self- exams. Denial of cancer occurred among melanoma participants in the parent Project Forward study and it is unknown if their denial could be related to low adherence to follow-up care or poorer skin protection. Future studies based in registry data or in hospital or healthcare system settings may shed light on this matter, as these settings allow verification of the diagnosis, allowing us to compare it to a patient’s response (affirmation or denial). Comparison of frequency of denial among cases of early-stage melanoma to other “low-impact” early stage cancers, such as cervical or colorectal cancer, may give us insight into whether denial among patients with early stage melanoma is a distinguishing feature of this diagnosis. Denial or non-confirmation of melanoma may also contribute to less frequent follow-up care and sub-optimal skin- protective behaviors (i.e. inconsistent use of sunscreen) among AYA survivors of melanoma. The number of AYA-dedicated resources continues growing to help AYAs navigate all the challenges that come with diagnosis, treatment, and survivorship from their cancer, while still pursuing “normal” life activities. In survivorship, they are busy getting an 142 education, establishing careers, and starting families, making it easy to forget about their needs for melanoma-specific follow-up care. Organizations like Critical Mass offer a service on their website to connect AYAs with local healthcare provider to help them get cancer care, recognizing this age group tends to move around a lot. It is possible that a simple service to deliver periodic text reminders could nudge them to schedule their follow-up visits, or at least do their skin self-exam. It could be incorporated alongside this existing service and also used connect them to (or remind them of) telemedicine resources. An ideal approach to care for AYAs with melanoma may fall somewhere in between pediatric and adult practice norms, striking a balance between greater levels of clinician involvement and support for autonomy that is sensitive to developmental needs. AYA programs as joint efforts between pediatric and adult oncologists are gaining momentum to become the standard of care, offering a promising, dynamic approach to improving care of AYAs. Intensifying AYA-specific, site-specific research will only serve to strengthen those efforts by providing the evidence base necessary to justify resource allocation for supporting program development and maintenance to ensure that optimal care for all AYAs is being delivered. 143 CHAPTER 5 REFERENCES 1. Adolescent and Young Adult Oncology Progress Review Group (AYAO PRG). Closing the Gap: Research and Care Imperatives for Adolescents and Young Adults with Cancer (NIH Publication No. 06-6067). Department of Health and Human Services, National Institutes of Health, National Cancer Institute, and the LIVESTRONG Young Adult Alliance Bethesda, MD; 2006. 2. Ferrari A, Bono A, Baldi M, et al. Does melanoma behave differently in younger children than in adults? A retrospective study of 33 cases of childhood melanoma from a single institution. Pediatrics. 2005;115(3):649-54. (doi: 115/3/649 [pii]). 3. Benjamin CL, Melnikova VO, Ananthaswamy HN. Models and mechanisms in malignant melanoma. Mol Carcinog. 2007;46(8):671-8. 4. Schoellhammer L, Mikkelsen TS, Schmidt H, et al. Atypical Presentation of Malignant Melanoma in a 16 Months Old Child. American Journal of Medical Case Reports. 2015;3(3):79-81. 5. Berking C, Takemoto R, Satyamoorthy K, et al. Induction of melanoma phenotypes in human skin by growth factors and ultraviolet B. Cancer Res. 2004;64(3):807-11. 6. Livestro DP, Kaine EM, Michaelson JS, et al. Melanoma in the young: differences and similarities with adult melanoma. Cancer. 2007;110(3):614-24. 144 7. Han D, Zager JS, Han G, et al. The unique clinical characteristics of melanoma diagnosed in children. Annals of surgical oncology. 2012;19(12):3888-95. 145 BIBLIOGRAPHY Adolescent and Young Adult Oncology Progress Review Group (AYAO PRG). Closing the Gap: Research and Care Imperatives for Adolescents and Young Adults with Cancer (NIH Publication No. 06-6067). Department of Health and Human Services, National Institutes of Health, National Cancer Institute, and the LIVESTRONG Young Adult Alliance Bethesda, MD; 2006. Ahlgren M, Wohlfahrt J, Olsen LW, et al. Birth weight and risk of cancer. Cancer. 2007;110(2):412-9. Armstrong BK, Kricker A. The epidemiology of UV induced skin cancer. Journal of Photochemistry and Photobiology B: Biology. 2001;63(1):8-18. Austin MT, Xing Y, Hayes-Jordan AA, et al. Melanoma incidence rises for children and adolescents: an epidemiologic review of pediatric melanoma in the United States. J Pediatr Surg. 2013;48(11):2207-13. Autier P, for Epimel D, Eortc Melanoma Cooperative Group, Jean-François. Influence of sun exposures during childhood and during adulthood on melanoma risk. International Journal of Cancer. 1998;77(4):533-7. Balch CM, Gershenwald JE, Soong S, et al. Final version of 2009 AJCC melanoma staging and classification. Journal of clinical oncology. 2009;27(36):6199-206. Barakat LP, Galtieri LR, Szalda D, et al. Assessing the psychosocial needs and program preferences of adolescents and young adults with cancer. Support Care Cancer. 2015. (doi: 10.1007/s00520-015-2849-8 [doi]). 146 Barr RD, Ries LA, Lewis DR, et al. Incidence and incidence trends of the most frequent cancers in adolescent and young adult Americans, including “nonmalignant/noninvasive” tumors. Cancer. 2016. Bataille V. Melanoma. Shall we move away from the sun and focus more on embryogenesis, body weight and longevity? Med Hypotheses. 2013;81(5):846-50. Bataille V. Sun exposure, sunbeds and sunscreens and melanoma. What are the controversies? Curr Oncol Rep. 2013;15(6):526-32. Benjamin CL, Melnikova VO, Ananthaswamy HN. Models and mechanisms in malignant melanoma. Mol Carcinog. 2007;46(8):671-8. Berger MF, Hodis E, Heffernan TP, et al. Melanoma genome sequencing reveals frequent PREX2 mutations. Nature. 2012;485(7399):502-6. Berking C, Takemoto R, Satyamoorthy K, et al. Induction of melanoma phenotypes in human skin by growth factors and ultraviolet B. Cancer Res. 2004;64(3):807-11. Bleyer A, O’Leary M, Barr R, Ries L, (eds). Cancer Epidemiology in Older Adolescents and Young Adults 15 to 29 Years of Age, Including SEER Incidence and Survival: 1975-2000 . Bethesda, MD: National Cancer Institute; 2006. Bleyer A, O'leary M, Barr R, et al. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. Cancer epidemiology in older adolescents and young adults 15 to 29 years of age, including SEER incidence and survival: 1975-2000. 2006. 147 Bleyer A. The adolescent and young adult gap in cancer care and outcome. Current problems in pediatric and adolescent health care. 2005;35(5):182-217. Bleyer A, Barr R. Cancer in young adults 20 to 39 years of age: overview. : Elsevier, 2009. Bleyer A, Barr R, Hayes-Lattin B, et al. The distinctive biology of cancer in adolescents and young adults. Nature Reviews Cancer. 2008;8(4):288-98. Bleyer A, Barr R, Ries L, et al, eds. Cancer in Adolescents and Young Adults. Second ed. Cham, Switzerland: Springer International Publishing, 2017. Bleyer A, Budd T, Montello M. Adolescents and young adults with cancer. Cancer. 2006;107(S7):1645-55. Bleyer A, Choi M, Fuller CD, et al. Relative lack of conditional survival improvement in young adults with cancer. : Elsevier, 2009. Bleyer WA. Cancer in older adolescents and young adults: epidemiology, diagnosis, treatment, survival, and importance of clinical trials. Med Pediatr Oncol. 2002;38(1):1-10. Boniol M, Autier P, Boyle P, et al. Cutaneous melanoma attributable to sunbed use: systematic review and meta-analysis. BMJ. 2012;345:e4757. (doi: 10.1136/bmj.e4757 [doi]). Bradford PT, Freedman DM, Goldstein AM, et al. Increased risk of second primary cancers after a diagnosis of melanoma. Arch Dermatol. 2010;146(3):265-72. Brandon EL, Gu J, Cantwell L, et al. Obesity promotes melanoma tumor growth: role of leptin. Cancer biology & therapy. 2009;8(19):1871-9. 148 Butow P, Palmer S, Pai A, et al. Review of adherence-related issues in adolescents and young adults with cancer. J Clin Oncol. 2010;28(32):4800-9. (doi: 10.1200/JCO.2009.22.2802 [doi]). Age-Adjusted Invasive Cancer Incidence Rates by County in California, 2013. Based on December 2015 extract. , December 2015. California Cancer Registry. Incidence and Mortality Counts and Age-Adjusted (2000 US Population) Rates per 100,000 Persons by Year, Race/Ethnicity, and Sex, California, 1988-2013, Invasive Melanoma of the Skin. , 2016. Chen J. Multiple signal pathways in obesity-associated cancer. Obesity Reviews. 2011;12(12):1063-70. Chen J, Chi M, Chen C, et al. Obesity and melanoma: Exploring molecular links. J Cell Biochem. 2013;114(9):1955-61. Clarke CA, Moy LM, Swetter SM, et al. Interaction of area-level socioeconomic status and UV radiation on melanoma occurrence in California. Cancer Epidemiol Biomarkers Prev. 2010;19(11):2727-33. (doi: 10.1158/1055-9965.EPI-10-0692 [doi]). Clinton-McHarg T, Carey M, Sanson-Fisher R, et al. Measuring the psychosocial health of adolescent and young adult (AYA) cancer survivors: a critical review. Health Qual Life Outcomes. 2010;8(1):25. Cockburn MG, Zadnick J, Deapen D. Developing epidemic of melanoma in the Hispanic population of California. Cancer. 2006;106(5):1162-8. 149 Cockburn M, Black W, McKelvey W, et al. Determinants of melanoma in a case–control study of twins (United States). Cancer Causes & Control. 2001;12(7):615-25. Cockburn M, Hamilton A, Mack T. The simultaneous assessment of constitutional, behavioral, and environmental factors in the development of large nevi. Cancer Epidemiol Biomarkers Prev. 2007;16(2):200-6. (doi: 1055-9965.EPI-06-0273 [pii]). Cockburn M, Hamilton A, Mack T. Recall bias in self-reported melanoma risk factors. Am J Epidemiol. 2001;153(10):1021-6. Colantonio S, Bracken MB, Beecker J. The association of indoor tanning and melanoma in adults: Systematic review and meta-analysis. J Am Acad Dermatol. 2014;70(5):847,857.e18. (doi: http://dx.doi.org/10.1016/j.jaad.2013.11.050). Cordoro KM, Gupta D, Frieden IJ, et al. Pediatric melanoma: results of a large cohort study and proposal for modified ABCD detection criteria for children. J Am Acad Dermatol. 2013;68(6):913- 25. Crump C, Sundquist K, Sieh W, et al. Season of birth and other perinatal risk factors for melanoma. Int J Epidemiol. 2014;43(3):793-801. (doi: 10.1093/ije/dyt277 [doi]). De Gruijl F. Skin cancer and solar UV radiation. Eur J Cancer. 1999;35(14):2003-9. De Pergola G, Silvestris F. Obesity as a major risk factor for cancer. Journal of obesity. 2013;2013. 150 de Vries E, Nijsten TE, Visser O, et al. Superior survival of females among 10,538 Dutch melanoma patients is independent of Breslow thickness, histologic type and tumor site. Ann Oncol. 2008;19(3):583-9. (doi: mdm498 [pii]). Deierlein AL, Siega-Riz AM, Adair LS, et al. Effects of pre-pregnancy body mass index and gestational weight gain on infant anthropometric outcomes. J Pediatr. 2011;158(2):221-6. Desai M, Beall M, Ross MG. Developmental origins of obesity: programmed adipogenesis. Current diabetes reports. 2013;13(1):27-33. Dietz WH. Critical periods in childhood for the development of obesity. Am J Clin Nutr. 1994;59(5):955-9. National Program of Cancer Registries (NPCR) National Data Quality Standard , August 9, 2013. EPA Office of Air and Radiation (6205J). Facts about: Skin Cancer. May 2010 ed. Washington, D.C.: United States Environmental Protection Agency; 2010. Eriksson J, Forsen T, Tuomilehto J, et al. Size at birth, childhood growth and obesity in adult life. International Journal of Obesity & Related Metabolic Disorders. 2001;25(5). Ferrari A, Montello M, Budd T, et al. The challenges of clinical trials for adolescents and young adults with cancer. Pediatric blood & cancer. 2008;50(S5):1101-4. Ferrari A, Bono A, Baldi M, et al. Does melanoma behave differently in younger children than in adults? A retrospective study of 33 cases of childhood melanoma from a single institution. Pediatrics. 2005;115(3):649-54. (doi: 115/3/649 [pii]). 151 Forman D, Bray F, Brewster DH, Gombe Mbalawa C, Kohler B, Piñeros M, Steliarova-Foucher E, Swaminathan R and Ferlay J, editors. Cancer Incidence in Five Continents, Vol. X (electronic version). Lyon: International Agency for Research on Cancer. Available from: http://ci5.iarc.fr, accessed [February 1, 2017]. 2013th ed. Lyon, France.: International Agency for Research on Cancer.; 2013. Fraser A, Tilling K, Macdonald-Wallis C, et al. Association of maternal weight gain in pregnancy with offspring obesity and metabolic and vascular traits in childhood. Circulation. 2010;121(23):2557-64. (doi: 10.1161/CIRCULATIONAHA.109.906081 [doi]). Freyer DR, Smith AW, Wolfson JA, et al. Making Ends Meet: Financial Issues from the Perspectives of Patients and Their Health-Care Providers. In: Bleyer A, Barr R, Ries L, Whelan J, Ferrari A, eds. Cancer in Adolescents and Young Adults. Second ed. Springer, 2017:667. Fu JM, Dusza SW, Halpern AC. Sunless tanning. J Am Acad Dermatol. 2004;50(5):706-13. (doi: http://dx.doi.org.libproxy.usc.edu/10.1016/j.jaad.2003.12.006). Gamba CS, Clarke CA, Keegan TH, et al. Melanoma survival disadvantage in young, non-Hispanic white males compared with females. JAMA dermatology. 2013;149(8):912-20. GenoMEL Consortium. A variant in FTO shows association with melanoma risk not due to BMI. Nat Genet. 2013;45(4):428-32. Guy Jr GP, Thomas CC, Thompson T, et al. Vital signs: melanoma incidence and mortality trends and projections-United States, 1982-2030. MMWR Morb Mortal Wkly Rep. 2015;64(21):591-6. 152 Haggar FA, Pereira G, Preen DD, et al. Cancer survival and excess mortality estimates among adolescents and young adults in Western Australia, 1982–2004: a population-based study. PloS one. 2013;8(2):e55630. Han D, Zager JS, Han G, et al. The unique clinical characteristics of melanoma diagnosed in children. Annals of surgical oncology. 2012;19(12):3888-95. Hausauer AK, Swetter SM, Cockburn MG, et al. Increases in melanoma among adolescent girls and young women in California: trends by socioeconomic status and UV radiation exposure. Arch Dermatol. 2011;147(7):783-9. Heck JE, Lombardi CA, Cockburn M, et al. Epidemiology of rhabdoid tumors of early childhood. Pediatric blood & cancer. 2013;60(1):77-81. Heck JE, Lombardi CA, Meyers TJ, et al. Perinatal characteristics and retinoblastoma. Cancer Causes & Control. 2012;23(9):1567-75. Hodis E, Watson IR, Kryukov GV, et al. A landscape of driver mutations in melanoma. Cell. 2012;150(2):251-63. Holman CD, Armstrong BK. Pigmentary traits, ethnic origin, benign nevi, and family history as risk factors for cutaneous malignant melanoma. J Natl Cancer Inst. 1984;72(2):257-66. Iannacone MR, Youlden DR, Baade PD, et al. Melanoma incidence trends and survival in adolescents and young adults in Queensland, Australia. International Journal of Cancer. 2015;136(3):603-9. 153 Johnson K, Carozza S, Chow E, et al. Birth characteristics and childhood carcinomas. Br J Cancer. 2011;105(9):1396-401. Joosse A, Collette S, Suciu S, et al. Sex is an independent prognostic indicator for survival and relapse/progression-free survival in metastasized stage III to IV melanoma: a pooled analysis of five European organisation for research and treatment of cancer randomized controlled trials. Journal of Clinical Oncology. 2013;31(18):2337-46. Joosse A, De Vries E, Eckel R, et al. Gender differences in melanoma survival: female patients have a decreased risk of metastasis. J Invest Dermatol. 2011;131(3):719-26. Kaplan MS, Huguet N, Newsom JT, et al. The association between length of residence and obesity among Hispanic immigrants. Am J Prev Med. 2004;27(4):323-6. Keegan TH, Swetter SM, Tao L, et al. Tumor Ulceration Does Not Fully Explain Sex Disparities in Melanoma Survival among Adolescents and Young Adults. J Invest Dermatol. 2015. Keegan TH, Tao L, DeRouen MC, et al. Medical care in adolescents and young adult cancer survivors: what are the biggest access-related barriers? Journal of Cancer Survivorship. 2014;8(2):282-92. Kent EE, Smith AW, Keegan TH, et al. Talking about cancer and meeting peer survivors: social information needs of adolescents and young adults diagnosed with cancer. Journal of adolescent and young adult oncology. 2013;2(2):44-52. Kleinman RE. Pediatric Nutrition Handbook. 6th ed. Elk Grove Village, Ill.: American Academy of Pediatrics, 2009. 154 Kremer L, Mulder RL, Oeffinger KC, et al. A worldwide collaboration to harmonize guidelines for the long-term follow-up of childhood and young adult cancer survivors: A report from the international late effects of Childhood Cancer Guideline Harmonization Group. Pediatric blood & cancer. 2013;60(4):543-9. Landier W, Bhatia S, Eshelman DA, et al. Development of risk-based guidelines for pediatric cancer survivors: the Children's Oncology Group Long-Term Follow-Up Guidelines from the Children's Oncology Group Late Effects Committee and Nursing Discipline. J Clin Oncol. 2004;22(24):4979-90. (doi: JCO.2004.11.032 [pii]). Lasithiotakis K, Leiter U, Meier F, et al. Age and gender are significant independent predictors of survival in primary cutaneous melanoma. Cancer. 2008;112(8):1795-804. Linos E, Swetter SM, Cockburn MG, et al. Increasing burden of melanoma in the United States. J Invest Dermatol. 2009;129(7):1666-74. Livestro DP, Kaine EM, Michaelson JS, et al. Melanoma in the young: differences and similarities with adult melanoma. Cancer. 2007;110(3):614-24. Lotti T, Bruscino N, Hercogova J, et al. Controversial issues on melanoma. Dermatologic therapy. 2012;25(5):458-62. Lu C, Zhang J, Nagahawatte P, et al. The genomic landscape of childhood and adolescent melanoma. J Invest Dermatol. 2014. Marks R. Epidemiology of melanoma. Clin Exp Dermatol. 2000;25(6):459-63. 155 McCue DA, Lohr LK, Pick AM. Improving adherence to oral cancer therapy in clinical practice. Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy. 2014;34(5):481-94. McNally RJ, Basta NO, Errington S, et al. Socioeconomic patterning in the incidence and survival of children and young people diagnosed with malignant melanoma in Northern England. J Invest Dermatol. 2014;134(11):2703-8. Miller KA, Huh J, Unger JB, et al. Patterns of sun protective behaviors among Hispanic children in a skin cancer prevention intervention. Prev Med. 2015;81:303-8. Miller KA, Langholz BM, Zadnick J, et al. Prevalence and predictors of recent skin examination in a population-based twin cohort. Cancer Epidemiol Biomarkers Prev. 2015;24(8):1190-8. (doi: 10.1158/1055-9965.EPI-14-1389 [doi]). Mills O, Messina JL. Pediatric melanoma: a review. Cancer control: journal of the Moffitt Cancer Center. 2009;16(3):225. Narayanan DL, Saladi RN, Fox JL. Review: Ultraviolet radiation and skin cancer. Int J Dermatol. 2010;49(9):978-86. SEER Fast Stats . National Research Council, Rasmussen KM and Yaktine AL (Eds). Weight gain during pregnancy: reexamining the guidelines. National Academies Press, 2010. Adolescents and Young Adults with Cancer , May 15, 2015. 156 Nikolaou V, Stratigos AJ. Emerging trends in the epidemiology of melanoma. Br J Dermatol. 2014;170(1):11-9. (doi: 10.1111/bjd.12492). Noonan FP, Recio JA, Takayama H, et al. Neonatal sunburn and melanoma in mice. Nature. 2001;413(6853):271-2. Noonan FP, Otsuka T, Bang S, et al. Accelerated ultraviolet radiation-induced carcinogenesis in hepatocyte growth factor/scatter factor transgenic mice. Cancer Res. 2000;60(14):3738-43. Nosrati A, Wei ML. Sex disparities in melanoma outcomes: the role of biology. Arch Biochem Biophys. 2014;563:42-50. O’Leary RE, Diehl J, Levins PC. Update on tanning: More risks, fewer benefits. J Am Acad Dermatol. 2014;70(3):562-8. (doi: http://dx.doi.org/10.1016/j.jaad.2013.11.004). O’Rorke M, Black C, Murray L, et al. Do perinatal and early life exposures influence the risk of malignant melanoma? A Northern Ireland birth cohort analysis. Eur J Cancer. 2013;49(5):1109- 16. Oeffinger KC, Mertens AC, Hudson MM, et al. Health care of young adult survivors of childhood cancer: a report from the Childhood Cancer Survivor Study. Ann Fam Med. 2004;2(1):61-70. O'Neill KA, Murphy MF, Bunch KJ, et al. Infant birthweight and risk of childhood cancer: international population-based case control studies of 40 000 cases. Int J Epidemiol. 2015;44(1):153-68. (doi: 10.1093/ije/dyu265 [doi]). 157 Østerlind A, Tucker M, Stone B, et al. The Danish case-control study of cutaneous malignant melanoma. II. Importance of UV-light exposure. International Journal of Cancer. 1988;42(3):319- 24. Paller AS, Hawk JL, Honig P, et al. New insights about infant and toddler skin: implications for sun protection. Pediatrics. 2011;128(1):92-102. (doi: 10.1542/peds.2010-1079 [doi]). Pappo A. Melanoma in children and adolescents. Eur J Cancer. 2003;39(18):2651-61. Park J, Morley TS, Kim M, et al. Obesity and cancer - mechanisms underlying tumour progression and recurrence. Nature Reviews Endocrinology. 2014;10(8):455-65. Parsons HM, Harlan LC, Lynch CF, et al. Impact of cancer on work and education among adolescent and young adult cancer survivors. J Clin Oncol. 2012;30(19):2393-400. (doi: 10.1200/JCO.2011.39.6333 [doi]). Pollack LA, Li J, Berkowitz Z, et al. Melanoma survival in the United States, 1992 to 2005. J Am Acad Dermatol. 2011;65(5):S78. e1,S78. e10. Pollitt RA, Clarke CA, Swetter SM, et al. The expanding melanoma burden in California hispanics. Cancer. 2011;117(1):152-61. Potosky AL, Harlan LC, Albritton K, et al. Use of appropriate initial treatment among adolescents and young adults with cancer. J Natl Cancer Inst. 2014;106(11):10.1093/jnci/dju300. Print 2014 Nov. (doi: 10.1093/jnci/dju300 [doi]). 158 Purdue MP, Freeman LE, Anderson WF, et al. Recent trends in incidence of cutaneous melanoma among US Caucasian young adults. J Invest Dermatol. 2008;128(12):2905-8. (doi: 10.1038/jid.2008.159 [doi]). Redman JM, Gibney GT, Atkins MB. Advances in immunotherapy for melanoma. BMC medicine. 2016;14(1):20. Rhodes AR, Weinstock MA, Fitzpatrick TB, et al. Risk factors for cutaneous melanoma: a practical method of recognizing predisposed individuals. JAMA. 1987;258(21):3146-54. Ribero S, Glass D, Aviv A, et al. Height and bone mineral density are associated with naevus count supporting the importance of growth in melanoma susceptibility. PLoS One. 2015;10(1):e0116863. (doi: 10.1371/journal.pone.0116863 [doi]). Ries L, Trama A, Nakata K, et al. Cancer Incidence, Survival, and Mortality Among Adolescents and Young Adults. In: Cancer in Adolescents and Young Adults. Springer, 2017:7-42. Robison LL, Mertens AC, Boice JD, et al. Study design and cohort characteristics of the childhood cancer survivor study: A multi-institutional collaborative project. Med Pediatr Oncol. 2002;38(4):229-39. Rode L, Hegaard HK, Kjaergaard H, et al. Association between maternal weight gain and birth weight. Obstet Gynecol. 2007;109(6):1309-15. (doi: 109/6/1309 [pii]). Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. The Lancet. 2008;371(9608):261-9. 159 Schellong K, Schulz S, Harder T, et al. Birth weight and long-term overweight risk: systematic review and a meta-analysis including 643,902 persons from 66 studies and 26 countries globally. PloS one. 2012;7(10):e47776. Schoellhammer L, Mikkelsen TS, Schmidt H, et al. Atypical Presentation of Malignant Melanoma in a 16 Months Old Child. American Journal of Medical Case Reports. 2015;3(3):79-81. Sender L, Zabokrtsky KB. Adolescent and young adult patients with cancer: a milieu of unique features. Nature Reviews Clinical Oncology. 2015. Sergentanis TN, Antoniadis AG, Gogas HJ, et al. Obesity and risk of malignant melanoma: a meta- analysis of cohort and case–control studies. Eur J Cancer. 2013;49(3):642-57. Shaikh WR, Weinstock MA, Halpern AC, et al. The characterization and potential impact of melanoma cases with unknown thickness in the United States’ Surveillance, Epidemiology, and End Results Program, 1989–2008. Cancer epidemiology. 2013;37(1):64-70. Shaikh WR, Dusza SW, Weinstock MA, et al. Melanoma Thickness and Survival Trends in the United States, 1989 to 2009. J Natl Cancer Inst. 2015;108(1):10.1093/jnci/djv294. Print 2016 Jan. (doi: 10.1093/jnci/djv294 [doi]). Shaw PH, Hayes-Lattin B, Johnson R, et al. Improving Enrollment in Clinical Trials for Adolescents With Cancer. Pediatrics. 2014;133(Supplement 3):S109-13. Shaw PH, Reed DR, Yeager N, et al. Adolescent and Young Adult (AYA) Oncology in the United States: A Specialty in Its Late Adolescence. J Pediatr Hematol Oncol. 2015;37(3):161-9. (doi: 10.1097/MPH.0000000000000318 [doi]). 160 Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA: a cancer journal for clinicians. 2016;66(1):7-30. Smith AW, Seibel NL, Lewis DR, et al. Next steps for adolescent and young adult oncology workshop: an update on progress and recommendations for the future. Cancer. 2016. Sondak VK, Swetter SM, Berwick MA. Gender disparities in patients with melanoma: breaking the glass ceiling. J Clin Oncol. 2012;30(18):2177-8. (doi: 10.1200/JCO.2011.41.3849 [doi]). STAMNES KOEPP UM, FROST ANDERSEN L, DAHL-JOERGENSEN K, et al. Maternal pre-pregnant body mass index, maternal weight change and offspring birthweight. Acta Obstet Gynecol Scand. 2012;91(2):243-9. Strouse JJ, Fears TR, Tucker MA, et al. Pediatric melanoma: risk factor and survival analysis of the surveillance, epidemiology and end results database. J Clin Oncol. 2005;23(21):4735-41. (doi: 23/21/4735 [pii]). Tai E, Beaupin L, Bleyer A. Clinical Trial Enrollment Among Adolescents With Cancer: Supplement Overview. Pediatrics. 2014;133(Supplement 3):S85-90. Tatalovich Z, Wilson JP, Cockburn M. A comparison of thiessen polygon, kriging, and spline models of potential UV exposure. Cartography and Geographic Information Science. 2006;33(3):217-31. Tatalovich Z, Wilson JP, Mack T, et al. The objective assessment of lifetime cumulative ultraviolet exposure for determining melanoma risk. Journal of Photochemistry and Photobiology B: Biology. 2006;85(3):198-204. 161 Thomas DM, Albritton KH, Ferrari A. Adolescent and young adult oncology: an emerging field. J Clin Oncol. 2010;28(32):4781-2. (doi: 10.1200/JCO.2010.30.5128 [doi]). Thomas NE, Edmiston SN, Alexander A, et al. Number of nevi and early-life ambient UV exposure are associated with BRAF-mutant melanoma. Cancer Epidemiol Biomarkers Prev. 2007;16(5):991-7. (doi: 16/5/991 [pii]). Trotter SC, Sroa N, Winkelmann RR, et al. A global review of melanoma follow-up guidelines. Journal of Clinical & Aesthetic Dermatology. 2013;6(9). U.S. Department of Health and Human Services. The Surgeon General's Call to Action to Prevent Skin Cancer. Washington, DC: U.S. Dept of Health and Human Services, Office of the Surgeon General; 2014. Quick Facts - Population of United States, California, and Los Angeles County , V2015. Vandermorris A, Parson KW, Greenberg ML. Adherence to Treatment Regimens in Adolescent and Young Adult Cancer Patients. In: Bleyer A, Barr R, Ries L, Whelan J, Ferrari A, eds. Cancer in Adolescents and Young Adults. Second ed. Springer, 2017:565. Watson M, Johnson CJ, Chen VW, et al. Melanoma surveillance in the United States: overview of methods. J Am Acad Dermatol. 2011;65(5):S6. e1,S6. e12. Whiteman DC, Valery P, McWhirter W, et al. Risk factors for childhood melanoma in Queensland, Australia. International Journal of Cancer. 1997;70(1):26-31. Whiteman DC, Whiteman CA, Green AC. Childhood sun exposure as a risk factor for melanoma: a systematic review of epidemiologic studies. Cancer Causes & Control. 2001;12(1):69-82. 162 Whiteman DC, Watt P, Purdie DM, et al. Melanocytic nevi, solar keratoses, and divergent pathways to cutaneous melanoma. J Natl Cancer Inst. 2003;95(11):806-12. Wojcik K, In G, Navid F. Melanoma Survival in Adolescents and Young Adults in Los Angeles County. Los Angeles, California: Los Angeles Cancer Surveillance Program, University of Southern California; Forthcoming (in-press) 2017. Wong JR, Harris JK, Rodriguez-Galindo C, et al. Incidence of childhood and adolescent melanoma in the United States: 1973-2009. Pediatrics. 2013;131(5):846-54. (doi: 10.1542/peds.2012-2520 [doi]). Yost K, Perkins C, Cohen R, et al. Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes & Control. 2001;12(8):703-11. 163 APPENDIX A: KAPLAN MEIER (KM) SURVIVAL PLOTS KM Plot: Overall Age Group Plot (children, AYAs, adults) 164 FEMALE MALE 165 KM Plots: Survival by STAGE AYA 166 FEMALE MALE 167 KM Plots: Survival by STAGE ADULTS 168 FEMALE MALE 169 KM Plots: Survival by STAGE CHILDREN 170 FEMALE MALE 171 KM Plots: Survival by Race/Ethnicity AYA 172 FEMALE MALE 173 KM Plots: Survival by Race/Ethnicity ADULTS 174 FEMALE MALE 175 KM Plots: Survival by Race/Ethnicity CHILDREN 176 FEMALE MALE 177 KM Plots: Survival by SES Level – AYA 178 FEMALE MALE 179 KM Plots: Survival by SES Level – ADULTS 180 FEMALE MALE 181 KM Plots: Survival by SES Level – CHILDREN 182 FEMALE MALE 183 KM Plots: Survival by Time Period AYA 184 FEMALE MALE 185 KM Plots: Survival by Time Period ADULTS 186 FEMALE MALE 187 KM Plots: Survival by Time Period CHILDREN 188 FEMALE MALE 189 KM Plots: Survival by ULCERATION AYA 190 FEMALE MALE 191 KM Plots: Survival by ULCERATION ADULT 192 FEMALE MALE 193 KM Plots: Survival by ULCERATION KIDS 194 FEMALE MALE 195 KM Plots: Survival by THICKNESS AYA 196 FEMALE MALE 197 KM Plots: Survival by THICKNESS ADULTS 198 FEMALE MALE 199 KM Plots: Survival by THICKNESS KIDS 200 FEMALE MALE 201 APPENDIX B: MELANOMA SPECIFIC SURVEY 202 203 APPENDIX C: SKIN-CARE BEHAVIOR QUESTIONS 204 APPENDIX D: RESIDENTIAL HISTORY FORM 205 APPENDIX E: INTRODUCTION LETTER/INFORMED CONSENT 206 207 APPENDIX F: CALIFORNIA CANCER REGISTRY BROCHURE 208 APPENDIX G: STUDY PROTOCOL Batching. Cases of melanoma were organized into two (2) batches for tracking throughout the recruitment process (Batch 1 n=150, Batch 2 n=137). Address cleaning. After eligible cases were identified, the most current address on file was obtained from the registry and checked through USC’s address service (details here), which identifies invalid and/or incomplete address listings and tracing was used to obtain updated or corrected information. Postcards. For the addresses deemed valid, postcards were mailed to indicate a full packet would be arriving in the coming weeks. We waited at least 2 weeks post-mailing to allow time for delivery and for return/forwarding notifications from the United States Postal Service (USPS). Initial Survey Packet. If a PO return was not received, the initial recruitment packet was mailed, consisting of a 2- page introductory letter that incorporated informed consent information and individual log-in information to complete the survey online, if desired, a paper copy of the 26-page survey (pages 24-25 melanoma add-on sent only to melanoma cases), a brief Project Forward study brochure, and a California Cancer Registry (CCR) brochure explaining the purpose and function of the state cancer registry. The survey included the following questions to capture the exposure and outcome of interest for the present study: 209 A two-part question was used to capture self-reported barriers to care: “At any time in the past year, were you unable to visit a doctor or health care provider when you needed to? (Yes or No) IF YES, What are the reasons why you didn’t go?” (check all that apply) To capture the self-reported outcome of interest (CRFC), the following question was used: “When did you last see a doctor for cancer-related follow-up care? (This is where a doctor examined you and did tests to see if you had any health problems from cancer or the cancer treatment you received.)” 210 Follow-up Activities. If there was no PO return from the initial packet, after a minimum of 2 weeks, follow-up phone calls began, followed by reminder mailings as needed and/or at participant request. All callers were trained by study staff and used IRB approved scripts to conduct calls. Participation modes. Patients were able to participate by returning the completed survey in the mail, by submitting their responses online through a dedicated, secured study website, or by interview over the phone during follow-up activities. Recording of Response. Completed paper surveys received in the mail were date stamped on receipt and entered into the tracking database as received. Telephone surveys were recorded either on paper or online, as noted by the study caller in the tracking database. Online submissions were checked weekly and recorded in the tracking database. Thank you letters. After receiving a completed survey, the participant was mailed a thank you letter along with a $20 cash gift. Among participants who had already indicated interest in participating, additional effort was made to increase responses during the last 3 months of the sub-study’s recruitment period. This entailed sending a new reminder letter with the log-in information, an extra copy of the paper survey in case the prior copy had been lost, and enclosing the incentive up front.

Asset Metadata

Creator Wojcik, Katherine Y. (author)

Core Title Melanoma in children, adolescents, and young adults

Contributor Electronically uploaded by the author (provenance)

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Publication Date 07/14/2017

Defense Date 05/01/2017

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

Tag AYA,barriers to care,early life risk factors,Epidemiology,Etiology,OAI-PMH Harvest,skin cancer,Survival,survivorship

Language English

Advisor Cockburn, Myles G. (committee chair), Hamilton, Ann S. (committee member), Milam, Joel E. (committee member), Wysong, Ashley (committee member)

Creator Email katherine.y.wojcik@gmail.com,kwojcik@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-399124

Unique identifier UC11265809

Identifier etd-WojcikKath-5509.pdf (filename),usctheses-c40-399124 (legacy record id)

Legacy Identifier etd-WojcikKath-5509.pdf

Dmrecord 399124

Document Type Dissertation

Rights Wojcik, Katherine Y.

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 a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract Melanoma, the most deadly form of skin cancer, is among the top 5 most common cancers among adolescents and young adults (AYAs), who have unique underlying biology and face a distinct set of life course challenges that may affect incidence and survival, distinguishing them from younger or older patients. Existing literature predominantly represents patients well over age 50, and very little detail regarding melanoma in young patients exists. To address this deficit, we conducted a set of population-based cancer registry studies, including a detailed descriptive study of incidence and survival characteristics among AYA melanoma patients with comparisons to older and younger patients in California, a large, case-control study to identify early life risk factors for developing melanoma before age 30, and a pilot survey-based study among young adult survivors of melanoma in Los Angeles County to identify important barriers to getting cancer-related follow-up care during survivorship. We found key opportunities to improve melanoma prevention and control among AYA populations. AYA programs, as joint efforts between pediatric and adult oncologists, are gaining momentum to become the standard of care, offering a promising, dynamic approach to improving care of AYAs. Intensifying AYA-specific, site-specific research will only serve to strengthen those efforts by providing the evidence base necessary to justify resource allocation for supporting program development and maintenance to ensure that optimal care for all AYAs is being delivered.