BACKGROUND: The effects of pediatric solid organ transplantation on cancer risk may differ from those observed in adult recipients. We described cancers in pediatric recipients and compared incidence to the general population.
METHODS: The US transplant registry was linked to 16 cancer registries to identify cancer diagnoses among recipients <18 years old at transplant. Standardized incidence ratios (SIRs) were estimated by dividing observed cancer counts among recipients by expected counts based on the general population rates. Cox regression was used to estimate the associations between recipient characteristics and non-Hodgkin’s lymphoma (NHL) risk.
RESULTS: Among 17 958 pediatric recipients, 392 cancers were diagnosed, of which 279 (71%) were NHL. Compared with the general population, incidence was significantly increased for NHL (SIR = 212, 95% confidence interval [CI] = 188–238), Hodgkin’s lymphoma (SIR = 19, 95% CI = 13–26), leukemia (SIR = 4, 95% CI = 2–7), myeloma (SIR = 229, 95% CI = 47–671), and cancers of the liver, soft tissue, ovary, vulva, testis, bladder, kidney, and thyroid. NHL risk was highest during the first year after transplantation among recipients <5 years old at transplant (SIR = 313), among recipients seronegative for Epstein-Barr virus (EBV) at transplant (SIR = 446), and among intestine transplant recipients (SIR = 1280). In multivariable analyses, seronegative EBV status, the first year after transplantation, intestine transplantation, and induction immunosuppression were independently associated with higher NHL incidence.
CONCLUSIONS: Pediatric recipients have a markedly increased risk for many cancers. NHL constitutes the majority of diagnosed cancers, with the highest risk occurring in the first year after transplantation. NHL risk was high in recipients susceptible to primary EBV infection after transplant and in intestine transplant recipients, perhaps due to EBV transmission in the donor organ.
- CI —
- confidence interval
- DLBCL —
- diffuse large B-cell lymphoma
- EBV —
- Epstein-Barr virus
- HR —
- hazard ratio
- IQR —
- interquartile range
- NHL —
- non-Hodgkin’s lymphoma
- PTLD —
- posttransplant lymphoproliferative disorder
- SIR —
- standardized incidence ratio
- SRTR —
- Scientific Registry of Transplant Recipients
- TCM —
- Transplant Cancer Match
What’s Known on This Subject:
Solid organ transplant recipients have a higher risk of cancer than the general population, attributable to use of immunosuppressant medications. Pediatric transplant recipients have a particularly elevated risk of non-Hodgkin’s lymphoma.
What This Study Adds:
This study of pediatric solid organ transplant recipients identifies an elevated risk of numerous cancer types in addition to non-Hodgkin’s lymphoma and provides evidence for a key role of Epstein-Barr virus in the etiology of most cancers in this population.
Currently in the United States, >1700 children and adolescents receive solid organ transplants per year.1 Because pediatric transplant recipients have the potential for longer life expectancies after transplantation than their adult counterparts, it is of particular importance to understand the long-term consequences of transplantation in this group.2 Transplant recipients have an elevated risk for many cancer types, in large part due to the requirement for chronic immunosuppressive therapy after transplantation.3 The effects of transplantation and immunosuppression may be different in pediatric recipients, especially the young, because they do not have fully developed immune systems before transplantation2 and because the typical distribution of cancers differs in pediatric and adult populations.4,5
In the US general population, the malignancies most frequently diagnosed in children include leukemia, non-Hodgkin’s lymphoma (NHL), and brain tumors.4,6 Because transplantation strongly increases NHL risk, lymphomas make up more than half of cancers in pediatric transplant recipients.7–9 In transplant recipients, NHL comprises part of a spectrum of related malignant and nonmalignant conditions termed posttransplant lymphoproliferative disorder (PTLD).10 PTLD risk is highest in the first year posttransplant and is associated with younger age and acquisition of Epstein-Barr virus (EBV) infection after transplantation.10–14
Because only a small number of pediatric transplants are performed annually at any single transplant center, it is difficult to comprehensively assess cancer risk in this group. The largest previous study, conducted among pediatric transplant recipients in Sweden, found that the risks of NHL, kidney cancer, multiple myeloma, and cancer of the vulva/vagina were >100 times higher than in the general pediatric population.7 However, this study included <600 recipients, and the expected number of cases for most types of cancers was <0.1. In the current study, we evaluated cancer risk and potential risk factors for cancer among pediatric recipients in the US Transplant Cancer Match (TCM) Study, which includes a pediatric population >20 times larger than the Swedish study.
The TCM Study links the Scientific Registry of Transplant Recipients (SRTR) to 16 US state or regional central cancer registries (http://transplantmatch.cancer.gov/).3 The SRTR captures all solid organ transplant recipients in the United States and collects information on recipient demographic characteristics, as well as relevant clinical characteristics, such as the indication for transplant, transplanted organ, and recipient EBV serostatus before transplant. Central cancer registries reliably ascertain all cancer diagnoses (other than basal cell and squamous cell carcinomas of the skin) within their geographic areas, including information on cancer site and histology. The TCM Study was approved by human subjects research review committees or exempted from human subjects research approval at the National Cancer Institute, the Health Resources and Services Administration, and each participating cancer registry.
For the current study, we included all recipients in the TCM Study who received solid organ transplants during 1987 to 2011 (calendar years covered by the cancer registries) when they were <18 years of age. Race/ethnicity was classified as white, African American, Hispanic, Asian, and other/unknown. Recipients of other/unknown race were excluded because this category was too heterogeneous to compare with general population cancer incidence rates. The unit of analysis was the transplant, and when recipients had multiple organ transplants before the age of 18, each transplant was included as an independent observation. At-risk time for cancer started at the later of: transplant date or start of cancer registry coverage. At-risk time ended at the first of: death, graft failure, retransplantation, loss-to-follow-up by SRTR, or end of cancer registry coverage. Follow-up for cancer risk extended beyond 18 years of age, because a transplant during childhood/adolescence could influence cancer risk into adulthood. For the evaluation of liver cancer, the first 6 months after liver transplantation were excluded to omit cancers that may have been the indication for transplant.
Cancers in transplant recipients were identified by using the linked cancer registry data and classified using International Classification of Diseases for Oncology, Third Edition topography and morphology codes as described previously.3,13 For individuals with multiple cancers of different types, each cancer type was included. Because our study assessed the risk of malignant cancer, 2 diagnoses of lymphoproliferative disease (morphology code of 9970) were excluded. The observed cancer counts were divided by the expected counts based on general population rates within strata of sex, 2-year age groups, race/ethnicity, and calendar year periods (1989–1993, 1994–1998, 1999–2003, 2004–2008, and 2009–2011), yielding standardized incidence ratios (SIRs). We calculated an overall SIR for total cancer, SIRs for specific cancer types with at least 2 observed cancer cases, and a combined SIR for cancer types with <2 observed cancer cases. We also calculated SIRs for common NHL subtypes, and within diffuse large B-cell lymphoma (DLBCL; the most common NHL subtype), we evaluated tumors with immunoblastic (morphology code 9684) and nonimmunoblastic morphology.
Because NHL made up the majority of observed cases, we assessed the risk factors for NHL and combined non-NHL cancers. We calculated incidence and SIRs according to age at transplantation, organ type, and time since transplantation. For NHL, we additionally calculated incidence and SIRs by EBV serostatus before transplantation, considering that seronegative status indicates a risk of developing primary EBV infection after transplantation. We additionally examined associations with potential risk factors by using multivariable Cox regression with calendar time as the time scale.
During 1987 to 2011, there were 40 570 solid organ transplants performed in recipients <18 years of age in the United States. Of these, 18 150 (45%) were in geographic regions covered by linked cancer registries in the TCM Study. One hundred ninety-two transplants (1%) were excluded because the recipients did not belong to 1 of the 4 main race/ethnicity groups. For our final study population, we evaluated 17 958 transplants performed in 16 732 individuals.
Among the included pediatric recipients, 54% were males, 54% were white, 24% were Hispanic, 18% were African American, and 5% were Asian/Pacific Islander (Table 1). Most transplants were for recipients <10 years of age at the time of transplantation, with 39% in recipients <5 years of age. The median year of transplantation was 2001 (interquartile range [IQR]: 1995–2005). With regard to the transplanted organ, 44% were kidney, 32% were liver, 21% were heart and/or lung, and 1% were intestine transplants. Ninety percent of transplants were first transplants. EBV serostatus before transplant was known for 8601 transplants (48% of included transplants), 46% of whom were EBV seronegative (N = 3984). The median length of follow-up was 4 years (IQR: 1–7 years), with a maximum follow-up of 22 years. A quarter of recipients were 18 years of age or older by the end of follow-up, with the oldest age of a recipient during follow-up being 38 years.
After transplantation, 392 cancers were diagnosed. Overall, the median age at cancer diagnosis was 12 years. The age at diagnosis ranged from <1 year to 35 years, with 18% of cancers occurring at ≥18 years of age (N = 72). The median time since transplantation at diagnosis was 2.5 years (IQR = 0.6–6.2 years). Of the cancers, 71% were NHL (N = 279), with a median age at diagnosis of 11 years (IQR = 5–17), and occurred a median of 1.6 years posttransplant (IQR = 0.6–5.5 years). Among NHL diagnoses, the most common subtype was DLBCL (N = 178; Supplemental Table 4), including 18 immunoblastic DLBCLs and 160 nonimmunoblastic DLBCLs. Thirty-nine percent of DLBCLs were extranodal tumors (N = 69), and among these, 16 occurred at sites corresponding to the transplanted organ. Of 5 extranodal DLBCLs identified in the kidney, 4 occurred in kidney recipients; of 9 identified in the liver, 7 occurred in liver recipients; and of 4 identified in the lung, 2 occurred in lung recipients. The second most frequently diagnosed NHL subtype was Burkitt lymphoma (N = 25).
Other common cancer diagnoses in pediatric transplant recipients were Hodgkin’s lymphoma (N = 30) and leukemia (N = 16), making up 8% and 4% of all cancers, respectively. For Hodgkin’s lymphoma, the median age at diagnosis was 12 years (IQR = 8–17 years), and mixed cellularity Hodgkin’s lymphoma was the most common subtype (N = 7; Supplemental Table 4). For leukemia, the median age at diagnosis was 14 years (IQR = 7–16 years), and the most common subtype was acute lymphocytic leukemia (N = 9). Among the less common cancers, we observed a number of subtypes that are typically disproportionately observed in children or adolescents, including 3 nephroblastomas, 2 hepatoblastomas, 2 Ewing sarcomas, and 2 dysgerminomas of the ovary (Supplemental Table 4).
Overall, cancer incidence among pediatric transplant recipients was >19 times higher than in the general population (SIR = 19.1, 95% confidence interval [CI] = 17.3–21.1) (Fig 1, Supplemental Table 5). NHL incidence was 212 times higher (SIR = 212, 95% CI = 188–238), Hodgkin’s lymphoma incidence was 19 times higher (SIR = 18.5, 95% CI = 12.5–26.4), and leukemia incidence was 4 times higher (SIR = 4.31, 95% CI = 2.46–6.99) in pediatric transplant recipients (Fig 1, Supplemental Table 4). Across all cancer types, the largest elevation in incidence was for myeloma (SIR = 229, 95% CI = 47.4–671), based on 3 observed cases (2 plasmacytomas and 1 multiple myeloma). Pediatric recipients also had elevated incidence of cancers of the kidney (SIR = 15.5), thyroid (SIR = 6.00), liver (SIR = 30.8), testis (SIR = 3.98), soft tissue (SIR = 4.45), ovary (SIR = 7.40), bladder (SIR = 28.1), and vulva (SIR = 17.4), again based on small numbers of cases (Fig 1, Supplemental Table 5).
Among NHLs, DLBCL incidence was 451 times higher than in the general population (SIR = 451, 95% CI = 397–511). For immunoblastic DLBCL specifically, incidence was 1027 times higher than in the general population (SIR = 1027, 95% CI = 609–1623), whereas nonimmunoblastic DLBCL incidence was 424 times higher (SIR = 424, 95% CI = 361–495). There was also elevated Burkitt lymphoma risk (SIR = 99, 95% CI = 64–146).
NHL incidence in pediatric transplant recipients was higher than in the general population regardless of age, organ type, time since transplantation, or EBV serostatus (Table 2). However, the magnitude of the elevation differed by subgroup. Specifically, elevations in NHL incidence were highest among recipients <5 years of age at transplantation (SIR = 313, 95% CI = 258–376) and decreased with older age (P value for heterogeneity <.001). Among the different organ types, the largest elevation was observed in intestine recipients (SIR = 1280, 95% CI = 516–2640) followed by heart/lung recipients (SIR = 318, 95% CI = 254–393), liver recipients (SIR = 197, 95% CI = 155–247), and kidney recipients (SIR = 159, 95% CI = 130–193; P value for heterogeneity <.001). The elevation in NHL incidence was highest in the first year after transplantation, whereas a smaller elevation in incidence was observed consistently across time thereafter (P value for heterogeneity <.001; Table 2). The elevation was also greater among EBV-seronegative recipients than EBV-seropositive recipients (P value for heterogeneity <.001).
In a multivariable regression model (Table 3), NHL incidence was higher in recipients of heart, lung, or intestine transplants compared with kidney recipients, with the highest incidence among intestine recipients. NHL incidence was substantially higher during the first year after transplant compared with later years (hazard ratio [HR] = 4.04, 95% CI = 3.12–5.23), in EBV seronegative recipients (HR = 2.71, 95% CI = 1.82–4.05), and in recipients who received induction immunosuppression (HR = 1.31, 95% CI = 1.01–1.72). Age at transplant, sex, and race were not associated with NHL incidence after accounting for other recipient characteristics (Table 3). The type of maintenance immunosuppressant regimen also was not associated with NHL incidence (data not shown).
For all non-NHL cancers combined, incidence was elevated compared with the general population (SIR = 5.89, 95% CI = 4.86–7.08), but elevations did not differ by age, transplanted organ, or time since transplantation (Table 2).
Overall, we found substantially elevated cancer risk among US pediatric transplant recipients. There was a particularly high elevation for NHL, which made up the majority of all cancer cases. For NHL, cancer risk differed across subgroups of recipients, with distinctly high elevations in intestine recipients and EBV-seronegative recipients. There were also significant elevations for non-NHL cancer overall and a number of specific cancer types.
Our finding that the majority of the cancer burden was attributable to NHL corresponds with results of previous studies in pediatric transplant populations.7,15 The predominance of NHL is due to the high frequency of NHL in the general pediatric population relative to other cancer types and the particularly strong elevation of NHL risk in pediatric transplant recipients. For transplant recipients overall, NHL risk is ∼8 times higher than in the general population,3 but this elevation is greatly magnified among pediatric recipients, with NHL risk >200 times higher than in the general population, as seen in our study. A similar pattern has been observed in children with AIDS, among whom NHL is the most frequently diagnosed cancer and elevations in risk are higher than those observed for NHL in adults with AIDS.16 NHL risk was particularly high within the first year after transplant, resembling the pattern seen for PTLD. This finding likely reflects primary EBV infection17 and early intensive immunosuppression, because use of induction immunosuppressant medications was also associated with NHL risk.
Within the pediatric transplant population, the elevation in NHL risk was especially increased for the youngest recipients, which is likely due to a number of factors. In the general population, NHL risk is typically lowest in the youngest age groups, but we did not observe this pattern in our pediatric transplant population. EBV seronegativity before transplantation was associated with a higher risk of NHL in our study, consistent with associations seen in adult transplant populations.18 However, most adults are already infected with EBV at the time of transplantation,18 whereas previous infection is much less common among children. In the setting of transplantation, EBV-seronegative recipients can experience a primary EBV infection after transplantation, when they are immunosuppressed and thus poorly able to control the infection, which puts them at increased risk for NHL.17,19 In our study, we saw high elevations in risk for 2 EBV-associated NHL subtypes, Burkitt lymphoma and DLBCL, with especially high elevations for immunoblastic DLBCL, which is more tightly linked to EBV than other DLBCL variants.20
The increase in NHL risk differed by type of organ transplanted, with an especially high risk for intestine recipients. Intestine transplantation may confer high risk because of the large amount of lymphatic tissue that accompanies the intestine graft,21 which could carry EBV. The relatively high frequency of EBV seronegative status and intestine transplants among children contributes to the particularly high elevations in NHL risk. Consistent with transmission of EBV through the transplanted organ, or perhaps due to chronic immune stimulation by the allograft, we found that the site of extranodal NHL strongly correlated with the transplanted organ.
Of importance, there were also significant elevations in the risk for Hodgkin’s lymphoma and myeloma, which along with NHL are considered part of PTLD. In our study, Hodgkin’s lymphoma was the second most frequent cancer diagnosis. EBV can be detected in tumor cells for the majority of Hodgkin’s lymphoma tumors arising in transplant recipients.22 Consistent with this, mixed cellularity Hodgkin’s lymphoma, which is closely tied to EBV infection,23 was the most common subtype in our population. There were only 3 cases of myeloma, but this cancer is extremely rare in children, so these cases translated to a substantial elevation in risk. Myeloma is less clearly linked to EBV infection among immunocompetent adults, but in transplant populations, a considerable fraction of tumors is EBV positive.24–28 EBV seronegative status at transplantation is associated with risk of both cancers,22,26 indicating an etiologic role of primary EBV infection during immunosuppression similar to that of NHL.
There is currently no approved EBV vaccine, and there are no established strategies for preventing EBV-related lymphomas in transplant recipients. Some transplant centers monitor high-risk recipients for circulating EBV DNA during the first posttransplant year and, for those individuals with persistently detectable EBV, subsequently reduce the intensity of immunosuppression or administer preemptive treatment of PTLD.29 One observational study found that transplant recipients who received pooled immunoglobulin as prophylaxis for cytomegalovirus disease had a lower incidence of lymphoma,30 consistent with some activity in neutralizing EBV.
We also observed elevated risks for a number of solid cancers and leukemia. Kidney, thyroid, and bladder cancer risks are elevated in people with end-stage renal disease, and in transplant populations, the risk is particularly high in kidney recipients,3,31,32 suggesting that kidney dysfunction is an important risk factor. Risks for liver cancer, soft tissue cancer, vulvar cancer, and leukemia are heightened in transplant recipients as well as people with AIDS, suggesting a relationship with immunosuppression3,33–35; liver cancer is often caused by hepatitis C or B viruses and vulvar cancer by human papillomavirus. Although our results for pediatric transplant recipients appear to mirror those seen in the adult transplant population, cancer subtypes that are rarely seen in adults contributed to some of these elevated risks. For instance, the most frequent subtype of liver cancer was hepatoblastoma, which is not caused by hepatitis, and nephroblastoma (Wilms tumor) contributed to the kidney cancer cases. Also, elevations in ovarian and testicular cancer risk have not been found in previous studies of transplant recipients. Thus, unique mechanisms may underlie some of these associations in the pediatric population. Notably, there was not a significant elevation in incidence for brain or bone cancer, 2 common cancer types in pediatric populations. No factors were clearly predictive of the magnitude of elevation in non-NHL cancer risk overall, which was perhaps foreseeable, because this group includes diverse cancer types.
Our study has several strengths. Our study is the largest study of cancer incidence in pediatric recipients conducted to date, and cancer diagnoses were based on complete, standardized information available from population-based cancer registries. The larger sample size allowed us to document risk for a wider variety of cancer types than previous studies. However, because cancer is rare among children and young adults, the expected counts for many cancers were <1 (Supplemental Table 4), and we only had statistical power to detect large SIRs. Although we followed pediatric transplant recipients into adulthood, the oldest recipients were only in their thirties by the end of follow-up. As a result, we cannot make inferences about the effects of pediatric transplantation on cancer risk in later life, a time during which most cancers develop in the general population. As this population ages, additional studies with longer follow-up will be useful to expand on our current findings.
When compared with the general population, we observed higher cancer risk in pediatric transplant recipients for several cancer types. The majority of cancer diagnoses were NHL. Risk for NHL (and likely Hodgkin’s lymphoma and myeloma) is strongly related to the combination of immunosuppression and EBV infection, which in some instances might be directly transmitted from the transplanted organ. The epidemiologic evidence from this study underscores potential cancer prevention opportunities if EBV infection can be effectively prevented or controlled in pediatric transplant recipients.
We thank individuals at the Health Resources and Services Administration (Monica Lin), the SRTR (Ajay Israni, Bertram Kasiske, Paul Newkirk, Jon Snyder), and the following cancer registries for their support and assistance: the states of California, Colorado (Jack Finch), Connecticut (Lou Gonsalves), Georgia (Rana Bayakly), Hawaii (Brenda Hernandez), Iowa, Illinois, Kentucky (Jaclyn Nee), Michigan (Glenn Copeland), New Jersey (Xiaoling Niu), New York, North Carolina (Chandrika Rao), Texas (Leticia Nogueria), Utah (Janna Harrell), and the Seattle-Puget Sound area of Washington (Margaret Madeleine). We also thank analysts at Information Management Services for programming support (David Castenson, Matthew Chaloux, Michael Curry, and Ruth Parsons).
The Scientific Registry of Transplant Recipients (SRTR) is currently operated under contract HHSH250201500009C (Health Resources and Services Administration) by the Minneapolis Medical Research Foundation, Minneapolis, MN. Previously the SRTR was managed under contracts HHSH250201000018C and HHSH234200537009C. The following cancer registries were supported by the SEER Program of the National Cancer Institute: California (contracts HHSN261201000036C, HHSN261201000035C, and HHSN261201000034C), Connecticut (HHSN261201000024C), Hawaii (HHSN261201000037C, N01-PC-35137, and N01-PC-35139), Iowa (HSN261201000032C and N01-PC-35143), New Jersey (HHSN261201300021I, N01-PC-2013-00021), Seattle-Puget Sound (N01-PC-35142), and Utah (HHSN2612013000171). The following cancer registries were supported by the National Program of Cancer Registries of the Centers for Disease Control and Prevention: California (agreement 1U58 DP000807-01), Colorado (U58 DP000848-04), Georgia (5U58DP003875-01), Illinois (5U58DP003883-03), Maryland (U58DP12-1205 3919-03), Michigan (5U58DP003921-03), New Jersey (5U58/DP003931-02), New York (U58DP003879), North Carolina (U58DP000832) and Texas (5U58DP000824-04). Additional support was provided by the states of California, Colorado, Connecticut, Illinois, Iowa, Massachusetts (Massachusetts Cancer Prevention and Control Cooperative Agreement 5458DP003920), New Jersey, New York (including the Cancer Surveillance Improvement Initiative), Texas, Utah, and Washington, as well as the University of Utah and Fred Hutchinson Cancer Research Center in Seattle, WA.
- Accepted February 2, 2017.
- Address correspondence to Elizabeth L. Yanik, PhD, ScM, Washington University in St Louis, 425 S. Euclid Ave, Campus Box 8233, St Louis, MO 63110. E-mail:
This work was presented in part at the Epidemiology Congress of the Americas; June 21–24, 2016; Miami, FL and at the American Transplant Congress; May 2–6, 2015; Philadelphia, PA.
The views expressed in this paper are those of the authors and should not be interpreted to reflect the views or policies of the National Cancer Institute, Health Resources and Services Administration, SRTR, cancer registries, or their contractors.
FINANCIAL DISCLOSURE: Dr Clarke is employed by GRAIL, Inc and has received grant funding from Genentech; the other authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Supported by the Intramural Research Program of the National Cancer Institute. Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: Dr Clarke is employed by GRAIL, Inc and has received grant funding from Genentech; the other authors have indicated they have no potential conflicts of interest to disclose.
COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2017-0542.
- US Department of Health and Human Services
- American Cancer Society
- Organ Procurement and Transplantation Network and Scientific Registry of Transplant Recipients
- Simard EP,
- Shiels MS,
- Bhatia K,
- Engels EA
- Mynarek M,
- Schober T,
- Behrends U,
- Maecker-Kolhoff B
- van Leeuwen MT,
- Grulich AE,
- Webster AC, et al
- Grogg KL,
- Miller RF,
- Dogan A
- Trappe R,
- Zimmermann H,
- Fink S, et al
- Lorsbach RB,
- Hsi ED,
- Dogan A,
- Fend F
- Copyright © 2017 by the American Academy of Pediatrics