Childhood Cancer and the Risk of ESKD

Significance Statement

The number of pediatric cancer survivors has steadily increased in recent decades because of higher incidence rates of cancer and improved treatments. Although children with cancer are exposed to various factors that might compromise kidney function, whether childhood cancer is associated with an increased risk of ESKD at adulthood is unknown. In this population-based, retrospective cohort study in Israel comprising 1,468,600 participants, the authors found that individuals with a history of childhood cancer had a risk of ESKD during a follow-up period of 30 years that was substantially elevated compared with that of individuals who were not childhood cancer survivors. These findings demonstrate a link between experiencing cancer during childhood and adverse renal outcome during adulthood, suggesting the need for closer and longer nephrological follow-up of survivors of childhood cancer.

Abstract

Background

Increasing cancer incidence among children alongside improved treatments has resulted in a growing number of pediatric cancer survivors. Despite childhood cancer survivors’ exposure to various factors that compromise kidney function, few studies have investigated the association between childhood cancer and future kidney disease.

Methods

To assess the risk of ESKD among childhood cancer survivors, we conducted a nationwide, population-based, retrospective cohort study that encompassed all Israeli adolescents evaluated for mandatory military service from 1967 to 1997. After obtaining detailed histories, we divided the cohort into three groups: participants without a history of tumors, those with a history of a benign tumor (nonmalignant tumor with functional impairment), and those with a history of malignancy (excluding kidney cancer). This database was linked to the Israeli ESKD registry to identify incident ESKD cases. We used Cox proportional hazards models to estimate the hazard ratio (HR) of ESKD.

Results

Of the 1,468,600 participants in the cohort, 1,444,345 had no history of tumors, 23,282 had a history of a benign tumor, and 973 had a history of malignancy. During a mean follow-up of 30.3 years, 2416 (0.2%) participants without a history of tumors developed ESKD. Although a history of benign tumors was not associated with an increased ESKD risk, participants with a history of malignancy exhibited a substantially elevated risk for ESKD compared with participants lacking a history of tumors, after controlling for age, sex, enrollment period, and paternal origin (adjusted HR, 3.2; 95% confidence interval, 1.3 to 7.7).

Conclusions

Childhood cancer is associated with an increased risk for ESKD, suggesting the need for tighter and longer nephrological follow-up.

Recent decades have seen significant improvements in the treatment of cancer among children and adolescents, leading to higher survival rates.^1,2 However, this is accompanied by long-term therapy-related complications among treated children who reach adulthood, including CKD,^3,4 as a result of several factors. First, chemo- and radiotherapy involve long-term deleterious effects on kidney function. The kidney is one of the organs most commonly affected by anticancer drugs and is very vulnerable to the development of drug toxicity in light of its role in metabolizing and excreting toxic agents. Specifically, cisplatin, carboplatin, and ifosfamide have been shown to result in a high prevalence of late kidney abnormalities.⁵ Second, children with any form of cancer are at increased risk for AKI from various causes, including volume depletion, sepsis, and tumor lysis syndrome,⁶ all of which confer increased risk for future CKD. Nevertheless, it is currently unknown to what extent these factors increase the future risk of CKD and ESKD among the pediatric population.⁷ Herein, we carried out a nationwide, population-based historical cohort study among 1,509,902 adolescents who were followed for 30 years and investigated the risk of developing ESKD after history of malignancy in childhood.

METHODS

Study Participants

We conducted a historical cohort study of Israeli adolescents, which included potential army recruits. One year before their conscription into military service, all eligible Israeli adolescents undergo mandatory medical board examination by a committee of two trained Israel Defense Force (IDF) medical corps physicians that includes reviewing the medical file obtained from the primary care physician, taking a medical history, and conducting a physical examination (including routine urinalysis). In case a nephrological problem is suspected, the conscript is referred to a board-certified nephrologist for confirmation or exclusion of the diagnosis.⁸ Inclusion criteria for this study were age 16 through 20 years at the time of medical board examination, which took place between 1967 and 1997. Because military service is not mandatory for Israeli non-Jews, the study included only Jewish recruits. Exclusion criteria were the presence of any of the following diagnoses, which confer an increased risk of ESKD: diabetes mellitus, any rheumatic disease, hypertension, or any past or current kidney disease at the time of enrollment, including congenital or acquired anomalies of the kidneys or urinary tract (including a congenital solitary kidney), GN, nephrolithiasis, or cystic renal disease. We also excluded participants with acute kidney disease, CKD, or proteinuria as well as participants with history of kidney cancer of any type including Wilms’ tumor.

Clinical Assessment and Diagnosis

All future conscripts provide copies of all available medical records, including health summaries from their primary physicians, and undergo a physical examination and a dipstick urinalysis test. If a kidney-related diagnosis cannot be ruled out, the future conscript is sent for additional tests and referred to a board-certified nephrologist. The accuracy and completeness of the medical information and diagnoses are additionally verified by a committee of two trained military physicians. Each diagnosis is assigned a numerical code and recorded in a central database.

Diagnosis of History of Cancer

A history of a benign or a malignant tumor was on the basis of a form submitted by the participants’ family physician, and was confirmed by a committee of two trained army physicians. When needed, additional specialists were consulted. Participants were identified according to a positive history of a tumor as per the relevant assigned numerical code in the army medical database. Malignancies included both hematologic and solid cancers. Benign tumors were defined as nonmalignant tumors resulting in any functional impairment, including gynecologic, endocrine, central nervous system, and skin tumors. Accordingly, participants were stratified into three groups: (1) no history of tumors, (2) history of a benign tumor, and (3) history of malignancy.

The Israeli Treated ESKD Registry

The Israeli treated ESKD registry is a national administrative registry maintained by the Ministry of Health.⁹ It contains information on patients receiving any form of RRT, i.e., hemodialysis, peritoneal dialysis, or kidney transplantation. All dialysis units in Israel, public and private, report to the Ministry of Health on new patients receiving RRT and changes in treatment modality. The database includes demographic data, a primary diagnosis, and initial RRT modality, as well as dates of initiating dialysis, changes of dialysis treatment modalities, renal transplantation, and death. Validation of the database includes periodic linkage with the Israeli population registry to update demographic and mortality data. Reports of cadaver donor transplants in Israel are crosschecked with the National Laboratory for Tissue Matching, and reports on living donor kidney transplants are crosschecked with the National Transplant Center. This study cohort was linked to the Israeli treated ESKD registry using the identification numbers given to all Israeli citizens at the time of birth or immigration. The institutional review board of the Israel Defense Forces approved the study and waived the requirement for informed consent on the basis of preserving participants’ anonymity.

Outcome Variables and Follow-Up

Onset of treated ESKD was defined as the date of initiation of dialysis or renal transplantation, whichever came first, and all treated patients with ESKD from January 1, 1980, to December 31, 2014, were included. Follow-up periods were measured from the initial medical board assessment until RRT initiation (incidence of ESKD), death, or December 31, 2014, whichever came first.

Statistical Analyses

Summary statistics for the study group were expressed as mean (SD) or percentage. Cox proportional hazards models¹⁰ were used to estimate the hazard ratios (HRs) and 95% confidence intervals (95% CIs) for comparing the incidence of treated ESKD among participant subgroups. The proportional hazards assumption was tested graphically using log–log graphs. Cox’s proportional hazards models were constructed controlling for age, sex, paternal origin (Europe/Americas, West Asia, North Africa, and Israel), and period of baseline examination by decade. The Mann–Whitney U test was used to compare median ages of ESKD diagnosis. Two-sided P values <0.05 were considered to indicate statistical significance. All other statistical analyses were conducted using SPSS version 24 (IBM).

RESULTS

Study Population

Figure 1 shows the study design, with stratification according to history of no tumors, benign tumor, or malignancy. The target population comprised 1,487,617 adolescents and young adults. Of these, 19,017 participants were excluded from the study due to the presence of CKD at baseline or other conditions predisposing to future ESKD. Of the 1,468,600 participants who met entry criteria, 1,444,345 (98.3%) had no history of tumors, 23,282 (1.6%) had a history of a benign tumor, and 973 (0.1%) had a history of malignancy (Table 1). The subgroup with a history of malignant tumors included 610 participants with a history of a solid tumor, 247 with a history of lymphoma, and 116 with a history of leukemia. The participants’ mean age at recruitment was 17.5±0.5, with a relatively higher proportion of young women among those with a history of benign tumors. Participants with a history of a benign tumor were also more likely to have a European/American origin versus a West Asian origin compared with participants with no history of tumors. Interestingly, in both the benign tumor and malignancy groups, we detected a gradual increase in their relative proportion in subsequent enrollment periods (Table 1).

Figure 1.

Participant assessment, designation, and outcome. “Benign” tumors were defined as nonmalignant tumors resulting in functional impairment, including gynecologic, endocrine, central nervous system, and skin tumors. The five that developed ESRD included four solid tumors and one leukemia.

Table 1.

Baseline characteristics of 1,468,600 participants examined between 1967 and 1997 according to history of benign tumor or malignancy

Characteristic	No History of Cancer (n=1,444,345)	Benign (n=23,282)	Malignant (n=973)	P Value
Age at assessment, mean (SD), yr	17.5 (0.5)	17.5 (0.5)	17.9 (0.7)	<0.001
Male sex, n (%)	878,086 (60.8)	12,233 (52.5)	600 (61.7)	<0.001
Father’s place of birth, n (%)
Europe/Americas	606,079 (42.0)	12,690 (54.5)	350 (36.0)	<0.001
West Asia	373,977 (25.9)	3848 (16.5)	192 (19.7)	<0.001
North Africa	356,250 (24.7)	5188 (22.3)	182 (18.7)	<0.001
Israel	66,958 (4.6)	1334 (5.7)	43 (4.4)	<0.001
Unknown	41,081 (2.8)	222 (1.0)	206 (21.2)	<0.001
Period of enrollment, n (%)
1967–1979	485,199 (33.6)	1809 (7.8)	146 (15.0)	<0.001
1980–1989	473,599 (32.8)	5316 (22.8)	297 (30.5)	<0.001
1990–1997	485,547 (33.6)	16,157 (69.4)	530 (54.5)	<0.001

History of Tumors and ESKD

Our study encompassed 44,534,163 person-years of follow-up (mean: 30.3), which included 43,952,983, 556,217, and 24,963 years of follow-up for participants with no history of tumors, history of benign tumors, and history of a malignancy, respectively (Table 2). During this period, 2430 participants developed ESKD, for an overall incidence rate of 5.46/100,000 person-years. ESKD developed in 2416 individuals (0.2%) without history of tumors, nine individuals (0.04%) with history of benign tumors, and five individuals (0.5%) with history of malignancy, of whom four had a solid tumor and one had leukemia. Table 2 shows the associations between history of tumors and ESKD during the follow-up period. ESKD incidence rates among participants without a history of tumors, with a history of benign tumors, and a history of malignancy were 5.5, 1.6, and 20.0/100,000 person years, respectively. Accordingly, the unadjusted HRs for ESKD among persons with history of benign tumors and malignancy were 0.5 [95% confidence interval (95% CI), 0.3 to 1.0] and 5.1 (95% CI, 2.1 to 12.2), respectively, as compared with patients with no history of tumors. After controlling for age, sex, paternal origin, and enrollment period, the subgroup of participants with history of benign tumors did not exhibit a statistically significant difference in the risk for ESKD, compared with participants without history of tumors (HR=0.6; 95% CI, 0.3 to 1.1). In contrast, the association between history of malignancy and ESKD remained statistically significant after adjustment for these confounders (HR=3.2; 95% CI, 1.3 to 7.7) (Table 2). In addition, participants with a history of tumors demonstrated a younger median age of ESKD diagnosis compared with participants without a history of tumors (39.2 versus 47.5 years, P=0.22), although this did not reach statistical significance, likely due to the small number of patients. The cumulative incidence of ESKD among the three subgroups of participants is presented in Figure 2.

Table 2.

Association between history of benign tumor or malignancy and ESKD according to the Cox proportional hazards model

Variable	No History of Cancer (n=1,444,345)	Benign (n=23,282)	Malignant (n=973)
Number of patients (%)	2416 (0.2%)	9 (0.04%)	5 (0.5%)
Incidence rate of ESKD, number of patients per 100,000 person-years	5.5	1.6	20.0
Total follow-up, yr	43,952,983	556,217	24,963
Follow-up (mean)	30.4	23.9	25.7
Age at end of follow-up, yr (SD)	47.9 (9.2)	41.4 (6.4)	43.5 (8.2)
Hazard ratio (95% CI) for ESKD (any malignancy)
Unadjusted	Reference	0.5 (0.3 to 1.0)	5.1 (2.1 to 12.2)
Adjusteda	Reference	0.6 (0.3 to 1.1)	3.2 (1.3 to 7.7)
Hazard ratio (95% CI) for ESKD– Solid tumors (n=610)	Reference
Unadjusted	Reference	—	6.3 (2.4 to 16.7)
Adjusteda	Reference	—	3.9 (1.5 to 10.4)

^aAdjusted for age, sex, paternal origin, and year of enrollment.

Figure 2.

Cumulative incidence of ESKD in the different participant subgroups.

We further analyzed the risk for future ESKD among the subgroup of 610 participants with a history of solid tumors. The latter demonstrated a crude HR of 6.3 (95% CI, 2.4 to 16.7) compared with participants without a history of tumors. Adjusting for age, sex, paternal origin, and enrollment period did not materially change the results (HR=3.9; 95%, CI 1.5 to 10.4).

DISCUSSION

Cancer patients are exposed to a myriad of factors that potentially affect kidney function, resulting both from the disease itself (e.g., paraneoplastic effects) and, to an even a greater extent, from iatrogenic factors leading to kidney injury, such as nephrotoxic effects of treatment, sepsis, and volume depletion. Nonetheless, data are severely lacking for the long-term renal function of survivors of childhood cancer in comparison with children without a history of cancer. Herein, in a retrospective cohort study using large databases, we found that adolescents with a history of cancer (both solid and hematologic) are at approximately four times the risk for future ESKD as compared with controls without history of cancer.

To the best of our knowledge, this study is the first to examine the risk of ESKD among survivors of childhood cancer, including all types of cancer (aside from kidney cancer). Several small studies have investigated the effects of cancer during childhood on kidney function, using various end points, such as reduced GFR, proteinuria, and hypertension.^11–14 Although most studies concluded that childhood cancer survivors are at increased risk for adverse renal outcomes, the reported risk varies greatly between studies. For instance, the prevalence of CKD/renal insufficiency ranged between 2.4% and 32%. This may be attributed to different GFR cutoffs used to define CKD, different formulas used to calculate GFR, varying follow-up periods, and inclusion of different subgroups with respect to both cancer type and treatment. In addition, nephrectomized patients were included in most studies. In contrast, we relied on a hard end point and included any type of cancer. Furthermore, patients who underwent nephrectomy or partial nephrectomy were excluded in order to distill the effects of cancer and cancer treatments on kidney function, since the presence of a single kidney, because of either congenital anomalies of the kidney and urinary tract or nephrectomy at an early age, is a known risk factor for renal impairment.^15,16

This study provides further support to the growing suggestion that kidney disease in adulthood often stems from injurious insults during childhood, a notion chosen as the World Kidney Day 2016 theme.¹⁷ Indeed, a substantial body of data indicate that hypertension, proteinuria, and CKD in adults represent the long-term sequelae of kidney injury during childhood, and even during fetal development.¹⁷

Notably, the increased risk of chronic organ damage due to childhood cancer is not specific to the kidney and has been shown to encompass multiple organs. These adverse outcomes are primarily pulmonary, auditory, endocrine, reproductive, cardiac, and neurocognitive.¹⁸ Although the exact mechanisms underlying tissue injury in each organ are difficult to determine due to the multiple exposures of cancer patients, this work provides further evidence of widespread organ damage after cancer in childhood. The observed association between history of cancer during childhood and ESKD likely stems mostly from both the direct (i.e., nephrotoxic), and indirect, long-term, and cumulative effects of antineoplastic treatments. The latter include volume depletion, sepsis (due to enhanced susceptibility to infection), and tumor lysis syndrome, with episodes of AKI.

The kidneys are highly susceptible to injury from toxic drugs, such as chemotherapy. This results from the fact that they receive approximately 20% of cardiac output, whereas at the same time carrying out continuous tubular uptake of drugs via endocytosis and transporter proteins. This combination of a high delivery rate and uptake leads to intracellular accumulation of drugs in the kidney, with subsequent formation of toxic metabolites and reactive oxygen species, leading to cellular injury.¹⁹ Among the various anticancer drugs, platinum-based agents and the alkylating agent ifosfamide are most commonly associated with both AKI and CKD due to their significant toxic effects on renal tubules.⁵ Additional agents that can cause tubular damage include the antifolate agent methotrexate²⁰ and immune checkpoint blockers, such as pembrolizumab and ipilimumab.²¹ In contrast, other agents, such as gemcitabine²² and antiangiogenic drugs,²³ lead primarily to glomerular damage, mainly by inducing thrombotic microangiopathy. Thus, both conventional chemotherapies and the newer biologic treatments (which have little to no relevance to this study because it refers to the period between 1967 and 1997) have the potential to cause kidney damage.

Other potential causative factors include contrast-induced nephropathy²⁴ and paraneoplastic effects. Notably, survivors of childhood cancers have been previously suggested to be at increased risk for several predisposing factors for ESKD, including diabetes and hypertension.^24,25 Hence, by excluding participants with these ESKD risk factors at baseline, we may have underestimated the true risk of ESKD among participants with a history of childhood cancer.

Interestingly, we detected an increase in the prevalence of adolescent participants with a history of childhood cancer along the study period. This likely results from the significant increase in survival rates among these patients,²⁶ as well as the increasing incidence of cancer among children in recent decades.²⁷

Several important limitations should be considered when interpreting the results of this study. First, we did not have full information about the type of cancer beyond its classification as either solid or hematologic, or on the type of treatment used. Similarly, ages at cancer diagnosis were not available to us. Hence, we were unable to explore potential causality principles that are likely relevant to ESKD development, such as chronology, biologic plausibility, and dose-response relationship, which would be interesting to explore in future studies. Nonetheless, all participants had normal renal function at the start of follow-up (i.e., during the prerecruitment assessment). Second, we did not have at hand the details of the treatment used or of any potential side effects or complications (renal or otherwise) during treatment. Third, our results are likely to be affected by survival bias because cancer survivors are more likely to have been treated by less aggressive treatments and suffer from less morbidity. Hence, although our conclusion would probably not have changed, the associated risk for ESKD demonstrated herein is likely an underestimation of the true risk. Fourth, our chosen end point (i.e., ESKD) is relatively rare, particularly at the ages presented in this study. This, in turn, probably resulted in underestimation of the lifetime risk of ESKD. Lastly, we cannot exclude the existence of participants who developed ESKD after leaving Israel and were thus not included in the ESKD registry. However, it is likely that a similar proportion of both controls and participants with a history of tumors left Israel, and thus this issue is unlikely to have affected our results.

The main strengths of this study are its reliance on a very large cohort having population-based screening for history of cancer with detailed clinical assessment parameters, a long follow-up period, and an objective end point on the basis of comprehensive documentation of ESKD.

Taken together, the findings of this study indicate that childhood cancer has to be considered as a lifelong risk factor for kidney disease secondary to decreased renal reserve. Although the absolute risk for ESKD remains low, it is noteworthy that ESKD patients represent merely a fraction of patients with CKD, and that our follow-up period ended at the fifth decade, underscoring the clinical importance of these results. Specifically, addressing this issue could entail tighter nephrological follow-up in childhood cancer survivors (e.g., via dedicated clinics) and more emphasis on minimizing nephrotoxic exposures among these patients.

Disclosures

K. Skorecki reports a Consultancy Agreement with Rambam Health Care Corporation; honoraria from the University of Montreal from a one-time residency teaching lecture and the University of Miami as a one-time grand rounds honorarium; Scientific Advisor or Membership as a Coeditor for the 11th edition of Brenner and Rector: The Kidney (published by Elsevier); and Member of Council for the Israel Science Foundation and the Scientific Advisory Committee for PSIFAS (Israel Genetics Diversity Project). All remaining authors have nothing to disclose.

Funding

None.