Incidence and Predictors of Cancer Following Kidney Transplantation in Childhood

Transcription

1 American Journal of Transplantation 2017; 17: Wiley Periodicals Inc The American Society of Transplantation and the American Society of Transplant Surgeons doi: /ajt Incidence and Predictors of Cancer Following Kidney Transplantation in Childhood A. Francis 1,2, *, D. W. Johnson 3,4,5, J. C. Craig 1,2 and G. Wong 1,2,6,7 1 Sydney School of Public Health, University of Sydney, Camperdown, NSW, Australia 2 Centre for Kidney Research, Kids Research Institute at The Children s Hospital at Westmead, Westmead, NSW, Australia 3 Department of Nephrology, Princess Alexandra Hospital, Brisbane, Australia 4 Australasian Kidney Trials Network, Diamantina Institute, University of Queensland, Brisbane, Australia 5 Translational Research Institute, Brisbane, Australia 6 Centre for Transplant and Renal Research, Westmead Hospital, Westmead, NSW, Australia 7 Australian and New Zealand Dialysis and Transplant Registry, Adelaide, SA, Australia *Corresponding author: Anna Francis, anna.francis@health.qld.gov.au Cancer risk is increased substantially in adult kidney transplant recipients, but the long-term risk of cancer in childhood recipients is unclear. Using the Australian and New Zealand Dialysis and Transplant Registry, the authors compared overall and site-specific incidences of cancer after transplantation in childhood recipients with population-based data by using standardized incidence ratios (SIRs). Among 1734 childhood recipients (median age 14 years, 57% male, 85% white), 289 (16.7%) developed cancer (196 nonmelanoma skin cancers, 143 nonskin cancers) over a median follow-up of 13.4 years. The 25-year cumulative incidences of any cancer were 27% (95% confidence intervals 24 30%), 20% (17 23%) for nonmelanoma skin cancer, and 14% (12 17%) for nonskin cancer (including melanoma). The SIR for nonskin cancer was 8.23 (95% CI ), with the highest risk for posttransplant lymphoproliferative disease (SIR 45.80, 95% CI ) and cervical cancer (29.4, 95% CI ). Increasing age at transplantation (adjusted hazard ratio [ahr] per year 1.10, 95% CI ), white race (ahr 3.36, 95% CI ), and having a functioning transplant (ahr 2.27, 95% CI ) were risk factors for cancer. Cancer risk, particularly for virus-related cancers, is increased substantially after kidney transplantation during childhood. Abbreviations: AIHW, Australian Institute of Health and Welfare; ahr, adjusted hazard ratio; CI, confidence interval; HR, hazard ratio; ANZDATA, Australian and New Zealand Dialysis and Transplant Registry; BCC, basal cell carcinoma; CMV, cytomegalovirus; EBV, Epstein Barr virus; HPV, human papillomavirus; ICD-10, International Classification of Diseases, Tenth Revision; IQR, interquartile range; NHL, non-hodgkin lymphoma; PTLD, posttransplant lymphoproliferative disease; SCC, squamous cell carcinoma; SIR, standardized incidence ratio Received 25 November 2016, revised 10 March 2017 and accepted for publication 18 March 2017 Introduction Cancer is an important and well-recognized complication after kidney transplantation in adults (1,2). The cumulative absolute risk of any cancer 20 years after transplantation is approximately 35%, corresponding to a relative increased risk of fourfold compared with the general population (3 5). The excess risk is site specific, being higher for virus-related cancers such as posttransplant lymphoproliferative disorder (PTLD) and cervical cancer and lower for non virus- and non immune-related cancers, such as breast and bowel cancers (6). Causal pathways are yet to be determined in detail but may be attributed to cumulative exposure to chronic immunosuppression, leading to loss of antitumor and antiviral immune surveillance (2,6 8). Cancer risk in recipients who received a kidney transplant during childhood is less well defined. A single study of 231 children reported a 20-year cumulative cancer incidence of approximately 11% after transplantation, with nonmelanoma skin cancer as the most common (9). Others have suggested PTLD is the most common cancer after kidney transplantation in children (10 13). Most previous studies are limited by their small sample size, relatively short follow-up periods, and sparse data on the risk of specific cancer types. Risk factors for cancer after kidney transplantation in childhood have not been explored for any cancer except PTLD. The aims of this study were to determine the absolute and relative risks of cancer and the risk factors for the development of cancer in patients who received their first kidney transplants as children or adolescents. 2650

2 Cancer After Childhood Kidney Transplantation Methods Study population and setting This long-term cohort study included all children and adolescents (younger than 20 years) who underwent kidney transplantation in Australian or New Zealand between January 1963 and December Patients with combined organ transplants (e.g. liver kidney) and those who had a cancer before their first transplantation were excluded. Data were collected from the Australian and New Zealand Dialysis and Transplant (ANZDATA) Registry, which prospectively collects data on all patients who commence renal replacement therapy in Australia and New Zealand (14). Covariates of interest: Covariate data included recipient characteristics (age, sex, self-reported race, cause of end-stage renal disease [ESRD], weight, height, cytomegalovirus [CMV] serological status, Epstein Barr virus [EBV] serological status, PRA), donor characteristics (source, age, CMV and EBV serological status), transplant characteristics (date, HLA mismatch, induction agent, baseline immunosuppressant), and transplant outcomes (graft loss, death). Outcomes: The ANZDATA registry records all incident cancers (nonmelanoma skin and nonskin) of all renal transplant recipients. Incident cancers were defined as the first cancer diagnosed after transplantation. The treating nephrology team reports all nonskin cancers and the first skin squamous cell carcinoma (SCC) or basal cell carcinoma (BCC) after transplantation. Prior work has validated the cancer reporting in ANZDATA against the statutory reporting Central Cancer Registry and found good concordance (15). Diagnoses of incident cancers (C0 C97) were coded according to the International Classification of Diseases, Tenth Revision (ICD-10) for cancers (16). Premalignant and in situ lesions were excluded. Non-Hodgkin s lymphoma (NHL) and PTLD were classified together as NHL/PTLD as there was insufficient information to distinguish between the two groups. Only malignant cases of PTLD are reported to ANZDATA. Statistical analysis Results were expressed as frequencies (percentages) for categorical data, mean (SD) for continuous normally distributed data, or median (interquartile range [IQR]) for continuous non-normally distributed data. Nonskin cancer incidence was compared with the Australian Institute of Health and Welfare (AIHW) general population data from 1982 to 2010 by using standardized incidence ratios (SIRs). Cancer incidences (except nonmelanoma skin cancer) are collected by the AIHW from the Australian Cancer Database and reported in age categories of 5 years, stratified by sex, for each calendar year. In addition, the AIHW reports the age- and sex-stratified population for each year. The general population at risk for cancer, stratified by sex, was established by subtracting the number of people who had developed an incident cancer in any given year from the person-time for all subsequent years, divided into 5-year age categories. The expected nonskin cancer event rate was calculated based on the crude incidence in each age group. The ratios of the observed to the expected incidences of nonskin cancers were calculated by using the indirect age standardization method, and their 95% CIs were calculated by assuming a Poisson distribution. The excess risks of site-specific incident cancers with >10 cases were calculated by using the same method. The expected numbers of incident cases were calculated by multiplying the overall person-years at risk by the corresponding population cancer incidence. Person-years for the observed were counted from the time of transplantation to the time of death, loss to follow-up, end of study (December 31, 2013), or diagnosis, whichever came first. For survival analyses, the follow-up period was defined from the time of transplantation to the time of first cancer diagnoses after transplantation. Those who did not develop cancer were censored at the time of death or end of study. A nonparametric estimation of the adjusted cumulative incidence function of cancer after transplantation was performed by using competing risk analysis. This considered the event of interest (i.e. cancer incidence) and other competing events (death). Anyone who did not experience the event (i.e. event free) was censored. The proportions free from incident cancers were calculated by using the Fine Gray method, allowing for the competing event of death. Univariable and multivariable Cox proportional hazard models assessed the risk factors for cancer development after transplantation. Variables that had an association with incident cancer at a p-value of <0.25 in the unadjusted analyses were included in the initial multivariable-adjusted analyses. In addition, sex, immunosuppression, and age at transplantation were included a priori in the final adjusted model because these variables have previously been shown to influence the risk of cancer after transplantation. Potential effect modification was tested between age, race, and sex by using two-way interaction terms. Transplant status (functioning transplant or receiving dialysis) was also assessed as a time-varying covariate. Given a time-varying covariate was included in the multivariable analyses, we did not account the other causes of death as competing events. Instead, participants were censored at the time of death. The proportional hazards assumptions of all Cox models were assessed by fitting time-dependent covariates in the multivariable models and plotting the Schoenfeld residuals. Model suitability was tested by using a stepwise backward elimination method to determine the optimal model. Data were structured by using Python 2.7, with the modules Pandas and NumPy. Data analysis were performed with SAS Studio (SAS Institute, Cary, NC), with p values <0.05 considered statistically significant. Results From 1963 to December 2013, 1748 children and adolescents received their first renal transplant. Fourteen children with cancer before their first transplantation were excluded, leaving 1734 children, of whom 576 (33%) received two or more subsequent kidney transplants. The median follow-up time was 13.4 years (IQR years) with a total follow-up time of patient-years. Baseline characteristics: The baseline characteristics of the cohort are presented in Table 1. The majority of recipients were transplanted in their teenage years and had congenital anomalies of the kidney and urinary tract. Incidence of skin and nonskin cancers: There were 289 transplant recipients (16.7%, 95% CI %) who developed at least one cancer, with 196 incident nonmelanoma skin cancers and 143 incident nonskin cancers (including melanoma). There were 50 people who developed both an incident skin and nonskin cancer. Of the 289 first cancers of any cause, the majority (161 [56%]) occurred within the first graft period, 96 (33%) occurred during a subsequent graft, and only 32 developed with the patient receiving dialysis after a failed American Journal of Transplantation 2017; 17:

3 Francis et al Table 1: Baseline characteristics (N = 1734) Variable n (%) Recipient characteristics Age at transplantation (years) (11.4) (16.0) (27.1) (45.5) Sex Female 741 (42.7) Race White 1470 (84.8) Indigenous/Pacific Islander 130 (7.5) Asian 70 (4.0) Other 64 (3.7) Cause of end-stage renal disease CAKUT 734 (42.3) Glomerulonephritis 575 (33.1) Cystic diseases 136 (7.8) Other/unknown 289 (16.7) Preemptive transplantation 265 (16.9) Donor characteristics Donor status Living donor 849 (49.1) Donor age 1 >37 years 835 (50.0) Transplant characteristics Transplant era (24.9) (24.1) (23.7) (27.4) HLA mismatch (11.7) (50.3) (19.5) Induction agent Basiliximab 440 (25.4) T cell depleting agent 155 (8.9) Other 17 (0.9) None 1122 (64.6) Baseline immunosuppression 3 Cyclosporine/azathioprine/prednisone 449 (25.9) Tacrolimus/mycophenolate/prednisone 378 (21.8) Azathioprine/prednisone 253 (14.6) Cyclosporine/mycophenolate/prednisone 244 (14.1) Other regimen 187 (10.8) Missing data 223 (12.9) CAKUT, congenital anomalies of the kidney and urinary tract. 1 Missing data for Missing data for Immunosuppression regimen in the first month of transplantation. graft. After adjustment for the competing risk of death, the cumulative incidences of cancers at 25 years after first renal transplant were 27% (95% CI 24 30%) for all cancers, 20% (95% CI 17 23%) for nonmelanoma skin cancer, and 14% (95% CI 12 17%) for nonskin cancer (Figure 1). Incidence of site-specific cancer (Table 2): The most common incident skin cancer type was SCC (n = 129, 7.4%, 95% CI %), followed by BCC (n = 67, 3.9%, 95% CI %). The most common nonskin cancer was NHL/PTLD (n = 40, 2.3%, 95% CI %), followed by cervical cancer (n = 18, %, 95% CI %). Time to cancer development: The median time to cancer development was 14.7 years (IQR years), with 12.4 years (IQR years) to first nonskin cancer and 17.2 years (IQR years) to first nonmelanoma skin cancer. The median time to cancer for nonskin cancers had two peaks, with PTLD occurring earlier (6.6 years, IQR 12.7 years) and all others occurring later (14.8 years, IQR years). The median time to first nonmelanoma skin cancer was similar for SCCs (17.3 years, IQR years) and BCCs (16.9 years, IQR years). The median age of first cancer varied between 18 and 32 years, depending on the type. NHL/PTLD occurred at an earlier age than other cancers (median age 18, IQR years), compared with 31 years (IQR 24-31) for other nonskin cancers and 32 years (IQR years) for nonmelanoma skin cancers (Table 2) American Journal of Transplantation 2017; 17:

4 Cancer After Childhood Kidney Transplantation Years since transplantation Figure 1: Time to first cancer, stratified by type (adjusted for competing risk of death). [Color figure can be viewed at wiley onlinelibrary.com] Table 2: Characteristics of site-specific cancers after childhood kidney transplantation Cancer site ICD-10 codes n (%) 1 transplantation, years 2 (IQR) Median time from Median age of first cancer, years 2 (IQR) Nonmelanoma skin cancer (12 23) 32 (27 39) Basal cell carcinoma C44 67 (33) 17 (13 23) 32 (27 39) Squamous cell carcinoma C (66) 17 (12 23) 32 (27 39) Nonskin cancer (7 21) 27 (21 37) Melanoma C43 12 (8) 14 (10 18) 29 (22 34) Lip/oral cavity C00 14, C (3) Digestive C (7) 26 (18 31) 41 (32 48) Respiratory C (1) Connective tissue C (1) Reproductive and breast C (26) 16 (9 23) 31 (26 37) Cervix C53 18 (13) 12 (9 19) 27 (24 35) Urinary C (11) 18 (9 25) 32 (19 41) Neurological C (2.8) Endocrine C (4.2) Hematological C81 85, C91 95, 48 (34) D47.Z1 Hodgkin disease C81 4 (3) Leukemia C (3) NHL/PTLD C82-85 and D47.Z1 40 (28) 7 (1 13) 18 (13 25) Unknown/other C76 80, C97 2 (1) ICD-10, International Classification of Disease-10; IQR, interquartile range; NHL, non-hodgkin s lymphoma; PTLD, posttransplant lymphoproliferative disease. 1 Proportion of total cases of nonmelanoma skin cancer or solid organ cancer for that site-specific cancer. 2 Reported for groups 10. Risk factors associated with the development of cancer: The results of univariate analysis are shown in Table S1. When adjusted for cause of ESRD, sex, donor source, and baseline immunosuppression, factors associated with cancer development included increasing age at transplant (hazard ratio [HR] 1.10/year, 95% CI , p < 0.001), white race (HR 3.36, 95% CI , p < 0.001), and functioning transplant American Journal of Transplantation 2017; 17:

5 Francis et al Variable Adjusted Hazard Ratio, 95%CI 1 Adjusted Hazard Ratio, 95%CI 1 P value Patient level Age at transplant (6-1.14) <0.001 Cause of ESKD CAKUT 3 Cystic 0.73 ( ) 0.28 GN ( ) 0.48 Other 1.38 ( ) 0.11 Non-Caucasian race <0.001 Caucasian race 3.36 ( ) Male sex 0.89 Female sex 3 ( ) Donor level Living donor Deceased donor 1.10 ( ) 0.57 Transplant level Dialysis after failed graft Functioning transplant Immunosuppression regimen Tac 5 /MMF 6 /Pred 7 Cyc 8 /MMF/Pred Cyc/Aza 9 /Pred Aza/Pred Other 2.27 ( ) 1.62 ( ) 1.27 ( ) 1.44 ( ) 1.39 ( ) < % lower confidence interval, 2= age variable increment is one year 3= congenital anomalies of the kidney and urinary tract, 4= Glomerulonephritis, 5= Tacrolimus, 6 = Mycophenolate, 7= Prednisone, 8=Cyclosporin, 9= Azathioprine Figure 2: Risk factors for cancer in childhood recipients of a kidney transplant (n = 1511). (HR 2.27, 95% CI , p < 0.001) (Figure 2) Functioning transplant status was assessed as a timevarying covariate. The overall risk of cancer doubled for each time point the recipient had a functioning transplant compared with times when recipients returned to dialysis after a failed graft. Transplant era was not associated with risk of cancer (p = 0.20) (Table S1 and Figure S1). Risk factors for nonmelanoma skin cancer and nonskin cancers were examined by using Cox regression (Tables S2 and S3). When adjusted for cause of ESRD, sex, donor source, and baseline immunosuppression, factors associated with nonmelanoma skin cancer development included increasing age at transplantation (adjusted HR [ahr] 1.17/year, 95% CI , p < 0.001), white race (HR 4.98, 95% CI , p = 0.026), latitude (ahr 5/degree closer to equator, 95% CI , p = 0.008), and functioning transplant (ahr 2.84, 95% CI , p = 0.003). Nonskin cancer was associated only with increasing age at transplant (ahr 6/year, 95% CI , p = 0.020) and white race (ahr 2.81, 95% CI , p = 0.026), when adjusted for sex, donor source, underlying disease, baseline immunosuppression, and functioning transplant. Incidence compared with the general population: After a median follow-up time of 13.4 years, 143 incident nonskin cancers were observed in the transplanted population compared with an expected number of 17.4 (SIR = 8.23, 95% CI ). The risk was elevated equally for males and females (Table 3) The risk of cancer for the transplant cohort was similar to that in the general population who were years older. For example, the cancer risk of a 20-year-old kidney transplant recipient was similar to that of people aged 50 without kidney disease (297/100,000, compared with /100,000, respectively). The excess risk compared with that of the general population was not constant but varied with age, with the risk declining with increasing age (Figure 3). Table 3: Age-standardized incidence rates and ratios Cancer site Transplant population: events, person-years at risk General population: events, person-years at risk Expected events Standardized incidence ratio (95% CI) Nonskin cancers 143, , ( ) Nonskin cancers (female) 71, , ( ) Nonskin cancers (male) 72, , ( ) NHL/PTLD 40, , ( ) Cervical cancer 18, , ( ) Melanoma 12, , ( ) NHL, non-hodgkin lymphoma; PTLD, posttransplant lymphoproliferative disease American Journal of Transplantation 2017; 17:

6 Cancer After Childhood Kidney Transplantation Figure 3: Standardized incidence ratios of nonskin cancer by age at cancer diagnosis. [Color figure can be viewed at wileyonlinelibrary.com] Some site-specific cancers also occurred at higher rates than the overall baseline increase of risk of around eightfold. There were 40 cases of NHL/PTLD in the transplant population compared with 0.9 expected case (SIR 45.8, 95% CI 33 62) and 18 observed cases of cervical cancer compared with 0.6 expected case (SIR 29.4, 95% CI ). Discussion Cancer is a common complication after kidney transplantation during childhood. Accounting for the competing risks of death, more than one-fourth of childhood recipients are expected to develop cancer by 25 years after transplantation. Nonmelanoma skin cancers are the most common cancer type, and cancers related to viruses, such as NHL, PTLD, and cervical cancer are heavily overrepresented. Compared with the general population, childhood kidney transplant recipients experienced around an eightfold increased risk of nonskin cancer. The median time to cancer development varied by cancer type. For instance, NHL and PTLD presented early, within 7 years after transplantation, while all other cancer types occurred around years after the first transplantation. With the exception of NHL and PTLD, most other cancers occurred after the child had transitioned to adult care. Increasing age at first transplantation, having a functioning transplant (a proxy for taking immunosuppression), and white race were independent risk factors for cancer development after transplantation. The cumulative incidence of cancer observed in this Australian/New Zealand cohort was higher than in most other pediatric reports (10,11,17 20). In Europe, the reported cumulative incidence of cancers varied between 7% at 10 years and 18% at 25 years after kidney transplantation in childhood (9,12). The higher rates of cancer observed in Australia and New Zealand may have been attributable to risk factors, such as the increased exposure to ultraviolet light. Latitude of residence closer to the equator (and, hence, increased ultraviolet light exposure) was associated with increased risk of nonmelanoma skin cancer in this cohort. Alternately, the larger cohort size and longer follow-up may have provided improved precision of the estimates of cancer risk. Transplant era did not have any significant effects on cancer risk after transplantation. While the incidence of cancer in this cohort was higher than that in other pediatric studies, it remains markedly lower than that in adult kidney transplant recipients (4,21). Previous studies reported a higher incidence of cancer, with a 20-year cumulative incidence of cancer (including nonmelanoma skin cancer) approaching as high as 60% in adult transplant recipients (3,17,18,22). The lower cumulative incidence of cancer in those who received a kidney transplant during childhood was expected, as increasing age at transplantation has been associated with increased cancer risk (3,21 23). This study further confirms findings from previous observational studies. For every 1-year increase in age at transplantation, the overall risk of cancer increased by at least 10%. The consistent increase in the relative and absolute risk of cancer over time and across all age groups suggest that age alone is not the only risk American Journal of Transplantation 2017; 17:

7 Francis et al factor for cancer but may be directly associated with the long-term exposure to immunosuppression. This cohort had an excess risk of cancer by eight times that of the general population, which is higher than reports from previous adult studies (1,5,6). Previous findings from the ANZDATA registry show a threefold increased risk for all solid organ cancers in the adult kidney transplant population compared with the age- and sex-matched population during a similar era (3). The SIR of cancer for childhood kidney transplant recipients is higher than that for adult kidney transplant recipients, as in childhood and early adulthood the general population is at a very low risk of most types of cancer (24).The cumulative effects of longterm immunosuppression and decreased antiviral activity may have accounted for the sizable differences in cancer risk between the transplant population and their healthy peers. This study extends prior knowledge by showing that for childhood recipients of a kidney transplant, a functioning transplant is a significant risk factor for nonmelanoma skin cancer development, with an adjusted increased risk of at least twofold compared with receiving dialysis after a failed graft. White race and living closer to the equator were associated with increased risk of nonmelanoma skin cancer. The higher cancer incidence observed in the transplanted population may also be partially explained by early detection bias. Transplant recipients may undergo frequent medical assessments and cancer screening, thus resulting in an overestimation of cancer incidence compared with the general population. Previous studies in the adult transplant population have shown that virus-related cancers are common after transplantation (2,5,6). Viruses, such as human papillomavirus (HPV) and EBV, are well-established risk factors for multiple tumor types. HPV may be associated with skin SCC and is causal for tongue, mouth, anal, cervical, and penile cancer (25,26). EBV also plays a causative role in Hodgkin lymphoma and NHL/PTLD (26). This study confirms and extends the previous observations, showing that the most common site-specific cancers in those transplanted in childhood were cancers caused or possibly caused by viruses, such as nonmelanoma skin cancer (predominantly SCC), NHL, PTLD, and cervical cancer. Consistent with other studies, virus-associated cancers such NHL and PTLD occurred at higher rates than other non virusrelated cancer types (11,27). Given that HPV infection is potentially preventable, the HPV vaccine may provide a unique opportunity for disease prevention in high-risk groups, such as those transplanted during early childhood. The magnitude of the increased risk for cervical cancers in women who had undergone transplantation as a child, with reference to the general population, was substantially higher than that of adult female transplant recipients (6). This implies that the cumulative risk of immunosuppression may have had an impact on immune system cancer surveillance and antiviral efficacy. In addition, women who are transplanted in childhood are much more likely to be immunosuppressed when they first encounter HPV and thus less likely to clear the virus and more likely to develop cancer (28,29). With the availability of the HPV vaccine, protection of HPV-related cancer through herd immunization of all at-risk preteens before exposure to the virus is crucial to prevent cancer development after transplantation. The time to first cancer in this study was similar to other times reported in pediatric studies with similar follow-up times (10 12,19). NHL and PTLD developed early (within a median of 7 years) after transplantation and at a much younger age (median age 18 years) compared with other cancer types. The relatively short time to cancer suggested that childhood kidney transplant recipients experienced cancer as young adults or, particularly for PTLD, even while still in pediatric care. This study was strengthened by its long follow-up time, large sample size, and completeness of cancer records within the ANZDATA registry. The reliability and comprehensive nature of cancer data collection in ANZDATA meant that selection and ascertainment biases were minimized (15). Incidence was also calculated accounting for the competing event of death. This study may have several weaknesses. The nature of a registry-based analysis suggested potential residual confounding and selection, detection, and coding biases may have occurred, particularly for nonmelanocytic skin cancers. Inability to inspect patient records meant that potentially confounding data such as in-depth information on comorbidities, medications, and environmental exposures was unable to be elicited. Drug dosing information was not routinely collected within the ANZDATA registry. As such, the impact of the cumulative dose of immunosuppressive agents and cancer risk after transplantation were unable to be assessed. In addition, data on the treatment of rejection have been reliably collected only since 1997, so the effect of rejection and its treatment on cancer risk were unable to be explored. Other than for nonmelanoma skin cancer, risk factors for specific cancer types could not be examined due to the small numbers of events in each group. EBVpositive donor to EBV-negative recipient confers an increased risk of PTLD, but donor EBV data were mostly incomplete and so the association between EBV and PTLD was unable to be examined. The risk of PTLD and PTLD-related mortality are the subjects of our future research. In conclusion, cancer occurs commonly after kidney transplantation in childhood and adolescence. Similar to adult kidney transplant recipients, nonmelanoma skin cancers and virus-related cancers predominated and occurred much more commonly compared with the general population. NHL/PTLD occurred earlier than all other cancer types. Increasing age at transplantation, white race, and having a functioning transplant were associated 2656 American Journal of Transplantation 2017; 17:

8 Cancer After Childhood Kidney Transplantation with increased risk of cancer. Given the high rates of virus-related cancer, such as cervical cancers, interventions such as HPV vaccine may potentially reduce the future burden of vaccine-preventable cancers. Acknowledgments The authors gratefully acknowledge the assistance of Dr Omri Bahat- Treidel in data processing and the substantial contributions of the entire Australian and New Zealand nephrology community (physicians, surgeons, database managers, nurses, renal operators, and patients) that provide information to and maintain the ANZDATA Registry database. Disclaimer The data reported here have been supplied by the Australian and New Zealand Dialysis and Transplant Registry. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the Australia and New Zealand Dialysis and Transplant Registry. Disclosure The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation. References 1. Birkeland SA, Lokkegaard H, Storm HH. Cancer risk in patients on dialysis and after renal transplantation. Lancet 2000; 355: Vajdic CM, McDonald SP, McCredie MR, et al. Cancer incidence before and after kidney transplantation. JAMA 2006; 296: Webster AC, Craig JC, Simpson JM, Jones MP, Chapman JR. Identifying high risk groups and quantifying absolute risk of cancer after kidney transplantation: A cohort study of 15,183 recipients. Am J Transplant 2007; 7: Lim WH, Chadban SJ. Cancer in ESRD: Clear on the epidemiology, hazy on the mechanisms. J Am Soc Nephrol 2016; 27: Engels EA, Pfeiffer RM, Fraumeni Jr JF, et al. Spectrum of cancer risk among US solid organ transplant recipients. JAMA 2011; 306: Yanik EL, Clarke CA, Snyder JJ, Pfeiffer RM, Engels EA. Variation in cancer incidence among patients with ESRD during kidney function and nonfunction intervals. J Am Soc Nephrol 2016; 27: Gutierrez-Dalmau A, Campistol JM. Immunosuppressive therapy and malignancy in organ transplant recipients: A systematic review. Drugs (Abingdon Engl) 2007; 67: Tessari G, Naldi L, Boschiero L, et al. Incidence of primary and second cancers in renal transplant recipients: A multicenter cohort study. Am J Transplant 2013; 13: van Amstel SP, Vogelzang JL, Starink MV, Jager KJ, Groothoff JW. Long-term risk of cancer in survivors of pediatric ESRD. Clin J Am Soc Nephrol 2015; 10: Debray D, Baudouin V, Lacaille F, et al. De novo malignancy after solid organ transplantation in children. Transplant Proc 2009; 41: Simard JF, Baecklund E, Kinch A, et al. Pediatric organ transplantation and risk of premalignant and malignant tumors in Sweden. Am J Transplant 2011; 11: Koukourgianni F, Harambat J, Ranchin B, et al. Malignancy incidence after renal transplantation in children: A 20-year singlecentre experience. Nephrol Dial Transplant 2010; 25: Berardinelli L, Raiteri M, Ghio L, Messa PG, Montagnino G. The role of immunosuppression in malignancies among 351 pediatric renal transplant patients. Transplant Proc 2010; 42: ANZDATA Registry. 38th Report, Australia and New Zealand Dialysis and Transplant Registry, Adelaide, Australia Webster AC, Supramaniam R, O Connell DL, Chapman JR, Craig JC. Validity of registry data: Agreement between cancer records in an end-stage kidney disease registry (voluntary reporting) and a cancer register (statutory reporting). Nephrology 2010; 15: World Health Organisation. [cited 2016 Jan 11]. Available from: Wisgerhof HC, van der Geest LG, de Fijter JW, et al. Incidence of cancer in kidney-transplant recipients: A long-term cohort study in a single center. Cancer Epidemiol 2011; 35: Kasiske BL, Snyder JJ, Gilbertson DT, Wang C. Cancer after kidney transplantation in the United States. Am J Transplant 2004; 4: Dharnidharka VR, Tejani AH, Ho PL, Harmon WE. Posttransplant lymphoproliferative disorder in the United States: Young Caucasian males are at highest risk. Am J Transplant 2002; 2: Dharnidharka VR, Kulsum-Mecci N. Non-post-transplant lymphoproliferative disorder cancers in children after organ transplantation. Pediatr Transplant 2013; 17: Ma MK, Lim WH, Turner RM, Chapman JR, Craig JC, Wong G. The risk of cancer in recipients of living-donor, standard and expanded criteria deceased donor kidney transplants: A registry analysis. Transplantation 2014; 98: Bichari W, Bartiromo M, Mohey H, et al. Significant risk factors for occurrence of cancer after renal transplantation: A single center cohort study of 1265 cases. Transplant Proc 2009; 41: Aguiar B, Santos Amorim T, Romaozinho C, et al. Malignancy in kidney transplantation: A 25-year single-center experience in Portugal. Transplant Proc 2015; 47: Australian Institute of Health and Welfare. [cited 2016 Jan 10]. Available from: Cogliano V, Baan R, Straif K, et al. Carcinogenicity of human papillomaviruses. Lancet Oncol 2005; 6: Wong M, Pagano JS, Schiller JT, Tevethia SS, Raab-Traub N, Gruber J. New associations of human papillomavirus, Simian virus 40, and Epstein-Barr virus with human cancer. J Natl Cancer Inst 2002; 94: Smith JM, Martz K, McDonald RA, Harmon WE. Solid tumors following kidney transplantation in children. Pediatr Transplant 2013; 17: Chin-Hong PV, Palefsky JM. Natural history and clinical management of anal human papillomavirus disease in men and women infected with human immunodeficiency virus. Clin Infect Dis 2002; 35: American Journal of Transplantation 2017; 17:

9 Francis et al 29. Massad LS, Ahdieh L, Benning L, et al. Evolution of cervical abnormalities among women with HIV-1: Evidence from surveillance cytology in the women s interagency HIV study. J Acquir Immune Defic Syndr 2001; 27: Supporting Information Additional Supporting Information may be found in the online version of this article. Table S1: Univariate analysis of associations with cancer. Table S2: Risk factors for nonmelanoma skin cancer in childhood recipients of a kidney transplant (n = 1089). Table S3: Risk factors for nonskin cancer in childhood recipients of a kidney transplant (n = 1358). Figure S1: Time to cancer stratified by transplant era American Journal of Transplantation 2017; 17: