Abstract
Background:
Curative therapy places childhood cancer survivors at increased risk of second primary malignancies (SPMs). However, there have been few population-based attempts to characterize differences in outcomes between SPMs in childhood cancer survivors compared with outcomes from first primary malignancies (FPMs).
Methods:
We extracted clinical and demographic information about childhood cancer survivors who developed SPMs and from individuals with comparable FPMs using data from the Surveillance, Epidemiology, and End Results (SEER) program. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated with Cox proportional hazards models comparing overall survival (OS) between individuals with and without a history of childhood cancer. OS was evaluated both overall and within specific cancers diagnosed in ≥50 childhood cancer survivors. Models accounted for potential confounders, including sex, race, age, treatment decade, histology, and disease stage.
Results:
Compared with individuals with FPMs (n=1,332,203), childhood cancer survivors (n=1,409) with a SPM experienced poorer OS (HR=1.86, 95% CI: 1.72-2.02) after accounting for cancer type, age, sex, race, and decade of diagnosis. A history of childhood cancer remained a poor prognostic factor for all specific cancers evaluated, including: breast (HR=2.07, 95% CI: 1.63-2.62), thyroid (HR=3.59, 95% CI: 2.08-6.19), acute myeloid leukemia (HR=2.38, 95% CI:1.87-3.05), brain (HR=2.09, 95% CI:1.72-2.55), melanoma (HR=2.57, 95% CI: 1.55-4.27), bone (HR=1.88, 95% CI:1.37-2.57), and soft tissue sarcoma (HR=2.44, 95% CI: 1.78-3.33).
Conclusion(s):
Compared to individuals without a prior cancer diagnosis, survivors of childhood cancer with an SPM experience inferior outcomes. Survival disparities were observed for the most frequent SPMs diagnosed in childhood cancer survivors.
Keywords: Second primary malignancy, childhood cancer survivor, survival
Precis:
Compared to individuals without a prior cancer diagnosis, survivors of childhood cancer with a second primary malignancy experience inferior outcomes. These disparities in overall survival are observed across cancer types, therapeutic exposures, and clinical factors.
Background
Progress in the treatment of childhood malignancies over the past several decades has contributed to five-year survival rates exceeding 80% (1). As a result, the estimated number of childhood cancer survivors living in the United States is expected to surpass 500,000 by 2020 (2). However, curative therapy for childhood malignancies places many survivors at increased risk for late effects of treatment, including second primary malignancies (SPMs). An elevated risk of SPMs is observed for all major childhood cancer diagnostic groups, including survivors of childhood leukemia, central nervous system tumors, Hodgkin’s disease, soft-tissue sarcoma, neuroblastoma, bone cancer, non-Hodgkin’s lymphoma, and kidney tumors (3). In particular, patients treated with radiotherapy, anthracyclines, platinum agents, or alkylating agents are vulnerable to subsequent neoplasms (4–9). Although fewer than 3% of childhood cancer survivors develop an SPM within 15 years of their initial diagnosis (4), the cumulative incidence approaches 10% by 30 years post-diagnosis in high risk populations (10, 11). In particular, survivors of childhood cancers experience an excess risk of second primary female breast cancer, thyroid cancer, hematologic cancers, bone and soft tissues sarcomas (3, 7, 12–14).
SPMs are the second leading cause of late mortality among survivors of childhood cancer, behind only relapse of the primary malignancy (15, 16). While an elevated risk of SPM among survivors of childhood cancer is well-established, SPM outcomes are not as well understood. Specifically, there have been few population-based studies comparing survival rates after SPM with survival rates after a first primary malignancy (FPM) of the same diagnostic category. To address this knowledge gap, we used data from the National Cancer Institute (NCI) Surveillance, Epidemiology, and End Results (SEER) program to evaluate differences in survival outcomes for individuals diagnosed with FPMs and childhood cancer survivors diagnosed with comparable SPMs.
Materials and Methods
The NCI’s SEER database provides de-identified cancer incidence and survival data on nearly a third of the U.S. population (17). Data included in this study were extracted from nine SEER registries from five states (Connecticut, Hawaii, Iowa, New Mexico, and Utah) and four metropolitan areas (Atlanta, Detroit, San Francisco-Oakland, and Seattle-Puget Sound). All variables were extracted from the November 2017 release of SEER 9 using SEER*Stat version 8.3.5 software, capturing cancers diagnosed between 1973 and 2015. Use of the SEER publically available resources are not subject to Institutional Review Board review.
Identification of SPMs in Childhood Cancer Survivors
Incident histologically distinct SPMs were identified among individuals with an index malignancy diagnosed between 0 and 19 years of age and recorded in one of the SEER 9 registries. A total of 1,573 individuals with SPMs were identified, of which 95 were excluded because the second malignancy was diagnosed within six months of the index cancer, making it difficult to distinguish an SPM from an incidence of multiple primary tumors (18). An additional 69 events with the same histology as the index cancer were excluded due to an inability to distinguish the SPM from a recurrence of the index tumor. The analysis was restricted to the first eligible SPM diagnosed in each survivor of childhood cancer; however, consistent with the FPM comparison group, individuals who developed multiple subsequent malignancies were not excluded from the analysis. The remaining 1,409 eligible SPMs were categorized based on the International Classification of Diseases for Oncology, Third Edition (ICD-O-3) coding system.
Extraction of First Primary Malignancies
We restricted our analysis to FPMs that were comparable to the SPMs observed among survivors of childhood cancer. For example, OS from an SPM of breast cancer in a survivor of childhood cancer was compared with OS from a breast cancer FPM. For each SEER cancer site code observed among the 1,409 eligible SPMs diagnosed in survivors of childhood cancer, corresponding FPMs were identified from individuals without a documented history of prior cancer. Comparisons were restricted to individuals with an FPM diagnosis in the age (0-60 years) and date (1973-2015) range of SPMs diagnosed in childhood cancer survivors.
Demographic and Clinical Variables
Demographic data, including age at diagnosis, race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, and other), sex, clinical data including ICD-O-3 codes, tumor site, year of diagnosis, surgery (yes, no/unknown), radiotherapy (yes, no/unknown), and chemotherapy (yes, no/unknown), were extracted for each eligible individual. SEER-generated Historic and Summary stage variables (localized, regional, distant disease) were extracted for available cancer sites and years (1973-2009). The primary outcome of overall survival (OS) was calculated in months from diagnosis of an SPM in survivors of childhood cancer or from diagnosis of an FPM for individuals without a history of cancer.
Statistical Analysis
Statistical analyses were conducted in Stata version 15 (StataCorp LLC, College Station, Texas) at a 5% significance level. Descriptive statistics were compared between childhood cancer survivors with SPMs and individuals with FPMs. Patients were censored at date of last contact or December 31, 2015 if still alive at the end of follow-up. Kaplan-Meier plots and log-rank test were used to compare OS between SPMs diagnosed in childhood cancer survivors and a comparable set of FPMs, identified using one-to-one propensity score matching (PS-matching) with a caliper of ±0.05 to account for the potential for residual confounding. Propensity scores reflected the conditional probability that an individual has a history of childhood cancer, given a set of observed baseline characteristics. Scores were estimated from sex, race/ethnicity, age at diagnosis, decade of diagnosis, and cancer diagnosis (i.e., SEER-generated recoded groups of ICD-O-3 site codes for SPMs observed among survivors of childhood cancer) (19).
Hazard ratios (HR) and accompanying 95% confidence intervals (CI) were estimated using Cox proportional hazards models to evaluate the association between a history of childhood cancer and survival in the overall cohort as well as among cancer groups with at least 50 SPMs identified in survivors of childhood cancer: breast cancer, thyroid cancer, acute myeloid leukemia, cancers of the central nervous system, soft-tissue sarcoma, bone cancer, and melanoma. Adjusted HRs were estimated in multivariable regression models, adjusting for sex, race/ethnicity, age at diagnosis, and decade of diagnosis. Because cancer stage was not available for every individual for every year, we generated separate Cox proportional hazards models to adjust for SEER-generated Historic and Summary stage variables, restricting our analysis to individuals with complete data. In order to evaluate whether the association between OS and a history of childhood cancer was modified by other characteristics, we assessed interaction terms between childhood cancer history and decade of diagnosis, disease stage, and demographic factors in separate regression models. A likelihood ratio test p-value <0.05 was considered statistical evidence of effect modification. Separate Cox proportional hazards models, stratifying on age at childhood cancer diagnosis, years since childhood cancer diagnosis, age at SPM diagnosis, and exposure to radiation therapy during childhood, were also generated to evaluate potential difference in outcomes across features of the childhood cancer. We evaluated potential departures from the proportional hazards assumption in each Cox model by visual inspection of the scaled Schoenfeld residuals and inclusion of time-dependent covariates in the model.
Results
Overall, 1,409 eligible SPMs were identified among survivors of childhood cancer and compared to 1,332,203 similar FPMs diagnosed in individuals <60 years of age (Table 1). Most SPMs were diagnosed in survivors of childhood leukemia (15.2%), cancers of the central nervous system (15.6%), or Hodgkin lymphoma (23.6%). On average, survivors of childhood cancer developed SPMs at a younger age (mean=28.04) than the comparable FPMs (mean=47.55) and were diagnosed during more recent decades (73.5% SPMs diagnosed 2000-2015 vs 49.0% FPMs diagnosed 2000-2015). After PS-matching, radiation therapy was documented for 26.0% of SPMs in childhood cancer survivors and 36.4% of FPMs (p<0.001), while chemotherapy was documented in 42.6% of SPMs in childhood cancer survivors and 46.9% of FPMs (p=0.021).Notably, while treatment information included in the SEER database has a high positive predictive value, the negative predictive value is low (20).
Table 1.
Demographic and clinical characteristics of study population
| SPMs in Childhood Cancer Survivors (n=1,409) | First Primary Malignancies | ||||
|---|---|---|---|---|---|
| Overall (n=1,332,203) | P-value | PS-match (n=1,409) | P-value | ||
| Sex, n(%) | |||||
| Male | 609 (43.22) | 593,216 (44.53) | 0.324 | 602 (42.73) | 0.790 |
| Female | 800 (56.78) | 738,987 (55.47) | 807 (57.27) | ||
| Race/Ethnicity, n(%) | |||||
| Non-Hispanic White | 1,052 (74.66) | 1,004,691 (75.42) | < 0.001 | 1,063 (75.44) | 0.868 |
| Non-Hispanic Black | 131 (9.30) | 121 (8.59) | |||
| Hispanic | 122 (8.66) | 148,588 (11.15) | 116 (8.23) | ||
| Other | 104 (7.38) | 81,714 (6.13) | 109 (7.74) | ||
| 97,210 (7.30) | |||||
| Cancer Diagnosis, n(%) | |||||
| Female Breast | 185 (13.13) | 280,351 (21.04) | < 0.001 | 185 (13.13) | 0.998 |
| Thyroid | 173 (12.28) | 60,088 (4.51) | 175 (12.42) | ||
| Brain | 146 (10.36) | 31,793 (2.39) | 152 (10.79) | ||
| AML | 103 (7.31) | 9,974 (0.75) | 96 (6.81) | ||
| Melanoma | 82 (5.82) | 83,179 (6.24) | 80 (5.68) | ||
| Soft-tissue Sarcoma | 71 (5.04) | 14,033 (1.05) | 76 (5.39) | ||
| Bone Cancer | 70 (4.97) | 6,363 (0.48) | 73 (5.18) | ||
| Other | 579 (41.09) | 846,422 (63.54) | 572 (40.60) | ||
| Year of Diagnosis, n(%) | |||||
| 1973-1979 | 9 (0.64) | 132,708 (9.96) | < 0.001 | 6 (0.43) | 0.864 |
| 1980-1989 | 94 (6.67) | 242,329 (18.19) | 85 (6.03) | ||
| 1990-1999 | 270 (19.16) | 304,692 (22.87) | 268 (19.02) | ||
| 2000-2009 | 542 (38.47) | 404,409 (30.36) | 541 (38.40) | ||
| 2010-2015 | 494 (35.06) | 248,065 (18.62) | 509 (36.12) | ||
| Age at Diagnosis, mean(SD) | 28.04 (12.77) | 47.55 (11.70) | < 0.001 | 28.05 (13.12) | 0.984 |
| SEER Historic Stage, n(%) | |||||
| Localized | 519 (36.83) | 516,729 (38.79) | < 0.001 | 487 (34.56) | 0.626 |
| Regional | 298 (21.15) | 283,792 (21.30) | 309 (21.93) | ||
| Distant | 276 (19.59) | 203,638 (15.29) | 293 (20.79) | ||
| Unknown/NA | 316 (22.43) | 328,044 (24.62) | 320 (22.71) | ||
| Radiation Therapy, n(%) | |||||
| Yes | 366 (25.98) | 463,912 (34.82) | < 0.001 | 513 (36.41) | < 0.001 |
| No/Unknown | 1,043 (74.02) | 868,291 (65.18) | 896 (63.59) | ||
| Chemotherapy, n(%) | |||||
| Yes | 600 (42.58) | 470,991 (35.35) | < 0.001 | 661 (46.91) | 0.021 |
| No/Unknown | 809 (57.42) | 861,212 (64.65) | 748 (53.09) | ||
| Childhood Cancer, n(%) | -- | -- | |||
| Leukemia | 214 (15.19) | ||||
| CNS Tumor | 220 (15.61) | ||||
| Hodgkin’s Disease | 332 (23.56) | ||||
| NHL | 84 (5.96) | ||||
| Neuroblastoma | 42 (2.98) | ||||
| Soft-tissue Sarcoma | 90 (6.39) | ||||
| Bone Cancer | 84 (5.96) | ||||
| Other | 343 (24.34) | ||||
| Childhood Radiation Therapy, n(%) | -- | -- | |||
| Yes | 700 (49.7) | ||||
| No/Unknown | 709 (50.3) | ||||
| Childhood Chemotherapy, n(%) | -- | -- | |||
| Yes | 793 (56.3) | ||||
| No/Unknown | 616 (43.7) | ||||
| Median time to SPM, year (range) | 16.0 (0.5-41.0) | -- | -- | ||
Abbreviation: second primary malignancies, SPMs; propensity score matched, PS-match; acute myeloid leukemia, AML; standard deviation, SD; central nervous system, CNS; non-Hodgkin lymphoma, NHL
Survivors of childhood cancer who developed an SPM experienced significantly poorer OS than FPMs diagnosed in individuals without a history of childhood cancer (HR=1.88, 95% CI: 1.66-2.14), after adjusting for sex, age at diagnosis, decade of diagnosis, race/ethnicity, and cancer diagnosis (Table 2). Similar to the association observed with all-cause mortality, a history of childhood cancer was associated with an increased risk of death due to cancer (Supplemental Table S1; HR=1.79, 95% CI: 1.64-1.95). The estimated five-year OS for PS-matched FPMs was 77.4% (95% CI: 75.0-79.5), compared to 61.7% (95% CI: 58.9-64.4) for SPMs in survivors of childhood cancer (Figure 1, log-rank p<0.001). Separate regression models restricted to the PS-matched set of individuals and adjusting for treatment exposures did not meaningfully impact the observed association between a history of childhood cancer and OS. Among individuals with available SEER Historic and Summary Stage variables (n=1,093 SPMs in childhood cancer survivors, n=1,004,159 FPMs) there was statistical evidence of interaction (p=0.017) between stage and a history of childhood cancer, indicating that differences in OS were larger for individuals with localized disease (HR=3.43, 95% CI: 2.89-4.11) than for those with regional (HR=2.19, 95% CI: 1.82-2.64) or distant (HR=1.75, 95% CI: 1.52-2.01) disease. In stratified models (Table 3), poorer OS was observed among childhood cancer survivors diagnosed before 10 years of age (HR=2.83, 95% CI: 2.46-3.25), survivors diagnosed within 10 years of the childhood cancer (HR=2.61, 95% CI: 2.29-2.99), and survivors treated with radiation therapy for the childhood cancer (HR=2.13, 95% CI: 1.92-2.37). However, disparities in OS were observed for all subgroups, including childhood cancer survivors diagnosed after 10 years of age (HR=1.61, 95% CI: 1.46-1.77), survivors diagnosed ≥10 years after the childhood cancer (HR=1.61, 95% CI: 1.46-1.78), and survivors without a known exposure to radiotherapy during childhood (HR=1.62, 95% CI: 1.44-1.83). Stratifying on decade of diagnosis, a history of childhood cancer was consistently associated with poorer overall survival across each decade of diagnosis evaluated (Supplemental Table S2).
Table 2.
Association between a history of childhood cancer and overall survival, SEER 1973-2015
| SPMs in Childhood Cancer Survivors (N) | First Primary Cancers (N) | Overall Survival HR (95% CI) | |
|---|---|---|---|
| Unadjusted Model | 1,409 | 1,332,203 | 1.25(1.15-1.35) |
| PS-matched Model | 1,409 | 1,409 | 1.88(1.66-2.14) |
| Adjusted Model1 | 1,409 | 1,332,203 | 1.86(1.72-2.02) |
| Adjusted Model + Treatment2 | 1,409 | 1,332,203 | 1.97(1.81-2.13) |
| Adjusted Model + SEER Stage3 | 1,093 | 1,004,159 | 2.07(1.88-2.28) |
| Stratified Adjusted Models1 | |||
| SEER Localized Stage | 519 | 516,729 | 3.43(2.89-4.11) |
| SEER Regional Stage | 298 | 283,792 | 2.19(1.82-2.64) |
| SEER Distant Stage | 276 | 203,638 | 1.75(1.52-2.01) |
Abbreviations: Surveillance, Epidemiology, and End Results, SEER; second primary malignancies, SPMs; propensity score, PS; confidence interval, CI
Adjusted for sex, age at diagnosis, decade of diagnosis, race/ethnicity, and cancer diagnosis
Adjusted for sex, age at diagnosis, decade of diagnosis, race/ethnicity, cancer diagnosis, chemotherapy and radiation therapy
Adjusted for sex, age at diagnosis, decade of diagnosis, race/ethnicity, cancer diagnosis, and SEER historic stag
Figure 1.
Kaplan-Meier survival curves for SPMs diagnosed among survivors of childhood cancer (n=1,409) and PS-matched first primary malignancies (n=1,409)
Table 3:
Stratified comparison of overall survival between SPMs diagnosed among survivors of childhood cancer and FPM
| N | Overall Survival HR (95% CI) | |
|---|---|---|
| Childhood Cancer Age at Diagnosis | ||
| FPMs | 1,332,203 | Ref. |
| SPMs - childhood cancer diagnosed 0-9 years | 486 | 2.83 (2.46 - 3.25) |
| SPMs - childhood cancer diagnosed 10-19 years | 923 | 1.61 (1.46 - 1.77) |
| Years Since Childhood Cancer Diagnosis | ||
| FPMs | 1,332,203 | Ref. |
| SPMs <10 years since childhood cancer | 462 | 2.61 (2.29 - 2.99) |
| SPMs ≥10 years since childhood cancer | 947 | 1.61 (1.46 - 1.78) |
| Exposure to Radiation Therapy in Childhood | ||
| FPMs | 1,332,203 | Ref. |
| SPMs - exposed to radiation in childhood | 700 | 2.13 (1.92 - 2.37) |
| SPMs - no/unknown exposure to radiation in childhood | 709 | 1.62 (1.44 - 1.83) |
Abbreviations: second primary malignancies, SPMs; first primary malignancies, FPMs; hazard ratio, HR; confidence interval, CI
Adjusted for sex, age at diagnosis of index cancer, decade of diagnosis of index cancer, race/ethnicity, and index cancer diagnosis
Group specific associations were also conducted for SPM categories with at least 50 childhood cancer survivors, including breast (n=185), thyroid (n=173), brain (n=146), acute myeloid leukemia (n=103), melanoma (n=82), soft-tissue sarcoma (n=71), and bone cancer (n=70). For each specific cancer evaluated, a history of childhood cancer remained a prognostic factor for poor outcomes (Figure 2): breast (HR=2.07, 95% CI: 1.63-2.62), thyroid (HR=3.59, 95% CI: 2.08-6.19), acute myeloid leukemia (HR=2.38, 95% CI: 1.87-3.05), brain (HR=2.09, 95% CI: 1.72-2.55), melanoma (HR=2.57, 95% CI: 1.55-4.27), bone (HR=1.88, 95% CI: 1.37-2.57), and soft tissue sarcoma (HR=2.44, 95% CI: 1.78-3.33). A history of childhood cancer remained significantly associated with poorer OS in models restricted to ICD-O-3 histology codes with ≥5 childhood cancer survivors and adjusting for histology.
Figure 2.
Forest plot of hazard ratios and 95% confidence intervals comparing overall survival for specific SPMs diagnosed among survivors of childhood cancer and comparable first primary malignancies: (A) adjusted for age, diagnosis decade, race, and sex; (B) adjusted for age, diagnosis decade, race, sex, histology and restricted to histology codes with ≥5 SPMs
Discussion
Compared to the general population, survivors of childhood cancer face increased rates of cancer-specific mortality (21, 22). The results of this study indicate that the increase in cancer-specific mortality is likely attributed to a combination of both an increase in the incidence of SPMs (3) and poorer relative survival following a SPM diagnosis. Specifically, in this population-based evaluation, we observed that survivors of childhood cancer diagnosed with an SPM experience inferior outcomes compared with their peers without a history of cancer. Disparities in OS were observed for survivors of childhood cancer independent of demographic factors, age at childhood cancer diagnosis, time from childhood cancer to SPM diagnosis, and exposure to radiotherapy during childhood. Notably, a history of childhood cancer was consistently associated with a two- to three-fold increased risk of death for the most commonly diagnosed SPMs.
This is the first study to specifically evaluate survival outcomes for SPMs and comparable FPMs between individuals with and without a history of childhood cancer. Few studies have evaluated the impact of prior cancer history on subsequent cancer outcomes among adults with prostate, lung, or gastrointestinal cancers (23–25). Zhou et al. recently explored whether a history of any prior cancer adversely influenced survival across 20 adult cancer sites (26). Although the HRs were attenuated relative to those observed in the current study, a history of prior cancer was associated with inferior outcomes for secondary breast cancer (HR=1.27, 95% CI: 1.24-1.30), melanoma (HR=1.23, 95% CI: 1.19-1.27), bone and soft tissue sarcoma (HR=1.17, 95% CI:1.09-1.25), and thyroid cancer (HR=1.56, 95% CI: 1.43-1.69). Notably, our results are consistent with previous reports that a prior history of cancer was associated with larger relative difference in survival for cancers with generally favorable outcomes, such as thyroid, breast and melanoma (26, 27). Keegan et al. evaluated differences in survival by age at SPM and observed that the five-year OS was 33% lower for individuals diagnosed with an SPM <15 years of age, 20% lower for individuals diagnosed with an SPM 15-39 years of age, and 8% lower for individuals diagnosed with an SPM ≥40 years of age (27). We did not observe survival differences across age at SPM; however, we did observe elevated HRs for individuals diagnosed with a childhood cancer <10 years of age or diagnosed with a SPM within 10 years of their initial childhood cancer. Although not yet evaluated in the scientific literature, these findings may be partly explained by an increased risk for late effects among children treated at younger ages or the potential that SPMs with a shorter latency between FPM and SPM are more aggressive, occur more frequently in individuals with a genetic predisposition, or are associated with particular childhood therapeutic exposures that might impact treatment efficacy.
A number of complicating factors may contribute to poor outcomes for childhood cancer survivors diagnosed with a SPM, including presence of comorbidities, prior cancer therapy interference with SPM treatment options, or unique etiologies of SPMs. Survivors of childhood cancer are at increased risk for developing chronic health conditions (28), including obesity (29, 30), frailty (31), and cardiovascular injury (32). These conditions have been linked to increases in mortality among adult cancer patients and may diminish the ability of survivors to tolerate intensive therapy for SPMs (33–36). In order to explore the potential impact of comorbid conditions on survival, we conducted a secondary analysis of cause of death. Compared to FPMs, the risk of death due to cancer among survivors of childhood cancer diagnosed with a SPM (HR=1.79, 95% CI: 1.64-1.95) was similar to all cause mortality (HR=1.86, 95% CI: 1.72-2.02). In the current analysis, greater than 75% of the deaths observed in both the childhood cancer survivors with SPMs and FPMs groups were attributed to cancer; however, assigning cause of death in the presence of co-morbidities may not be straightforward or reliably reported. Therefore, given the similarities in these associations, we only presented associations with overall survival. In addition to the impact of comorbidities, treatment exposures during childhood may further interfere with or limit treatment options for the SPM due to cumulative exposures to specific therapeutic modalities or agents. Although our ability to evaluate therapeutic differences is limited in SEER, we did observe lower frequencies of exposure to both radiation therapy (26.0% vs 36.4%) and chemotherapy (42.6% vs 46.9%) in SPMs diagnosed in survivors of childhood cancer than comparable FPMs after matching on potential confounders. In addition to treatment differences, it is possible that standard treatment options are less efficacious in treatment-related SPMs. Treatment exposures during childhood cancer therapy, including ionizing radiation, may result in the development of SPMs with distinct mutational profiles (37), which may influence response to therapy. For example, RB1 and TP53 inactivation is common in radiation-related sarcomas (38, 39), and somatic mutations in these genes have been linked or variability in treatment efficacy (40–42).
Strengths of this assessment include a relatively large sample size of SPMs in childhood cancer survivors with up to 42 years of follow-up. However, the number of specific SPMs (i.e., cancers with <50 SPMs diagnosed in childhood cancer survivors) limited our ability to evaluate the impact of a prior childhood cancer on outcomes for approximately 40% of specific SPMs. Still, this study provides valuable insight into the detrimental association between prior childhood cancer and survival for the most common SPMs, including breast cancer, thyroid cancer, cancers of the central nervous system, acute myeloid leukemia, melanoma, soft-tissue sarcoma, and bone cancer. Exposure to radiation therapy and certain chemotherapy agents during childhood are well-established risk factors for the development of SPMs (3–6, 43). However, detailed treatment information is incomplete or unavailable for individuals included in the SEER database. Additionally, given the growing recognition of SPM risk, survivors of childhood cancer may be more likely than the general population to receive regular cancer screenings. Participation in screening programs may result in earlier diagnosis of SPMs, which may bias survival rates. However, cancer screening is more likely to artificially inflate survival time in childhood cancer survivors, thereby biasing our effect estimates towards the null. Interestingly, we observed the largest relative differences in survival outcomes for localized disease, although a prior childhood cancer diagnosis remained a prognostic factor for poor outcome in more advanced cancers. Finally, the SEER program only captures incident cancers at participating registries. Therefore, any SPMs diagnosed among survivors of childhood cancer that relocate outside of SEER catchment areas were not be included in this analysis.
Changes in childhood cancer therapy, including less intensive therapy, have led to a reduction, but not elimination, in the risk of SPM (4, 11, 44). This study describes a novel association between a history of childhood cancer and poor outcomes after SPMs. Notably, we did not observe evidence that the relative difference in OS decreased with more modern treatment eras, suggesting that despite overall improvements in cancer treatment, a history of prior childhood cancer confers a survival disadvantage even with contemporary treatment approaches. Given the growing number of childhood cancer patients surviving well beyond their initial diagnosis, the long-term health of childhood cancer survivors requires effective approaches to reduce both risk for SPMs and adverse outcomes following diagnosis of an SPM. A combination of host and tumor factors likely influence an individual’s response to treatment for an SPM. Additional work is needed to fully evaluate predictors of suboptimal survival in childhood cancer survivors diagnosed with specific secondary cancers.
Supplementary Material
Abbreviations:
- SPMs
second primary malignancies
- FPMs
first primary malignancies
- SEER
Surveillance, Epidemiology, and End Results
- HR
hazard ratio
- CI
confidence interval
- OS
overall survival
- NCI
National Cancer Institute
- ICD-O-3
International Classification of Diseases for Oncology, Third Edition
- PS
propensity score
References
- 1.Smith MA, Altekruse SF, Adamson PC, Reaman GH, Seibel NL. Declining childhood and adolescent cancer mortality. Cancer. 2014;120(16):2497–506. Epub 2014/05/24. doi: 10.1002/cncr.28748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Robison LL, Hudson MM. Survivors of childhood and adolescent cancer: life-long risks and responsibilities. Nature Reviews Cancer. 2014;14(1):61–70. Epub 2013/12/07. doi: 10.1038/nrc3634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Neglia JP, Friedman DL, Yasui Y, Mertens AC, Hammond S, Stovall M, Donaldson SS, Meadows AT, Robison LL. Second malignant neoplasms in five-year survivors of childhood cancer: childhood cancer survivor study. Journal of the National Cancer Institute. 2001;93(8):618–29. Epub 2001/04/20. [DOI] [PubMed] [Google Scholar]
- 4.Turcotte LM, Liu Q, Yasui Y, Arnold MA, Hammond S, Howell RM, Smith SA, Weathers RE, Henderson TO, Gibson TM, Leisenring W, Armstrong GT, Robison LL, Neglia JP. Temporal Trends in Treatment and Subsequent Neoplasm Risk Among 5-Year Survivors of Childhood Cancer, 1970-2015. JAMA. 2017;317(8):814–24. Epub 2017/03/01. doi: 10.1001/jama.2017.0693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Henderson TO, Rajaraman P, Stovall M, Constine LS, Olive A, Smith SA, Mertens A, Meadows A, Neglia JP, Hammond S, Whitton J, Inskip PD, Robison LL, Diller L. Risk factors associated with secondary sarcomas in childhood cancer survivors: a report from the childhood cancer survivor study. International Journal of Radiation Oncology, Biology, Physics. 2012;84(1):224–30. Epub 2012/07/17. doi: 10.1016/j.ijrobp.2011.11.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Borgmann A, Zinn C, Hartmann R, Herold R, Kaatsch P, Escherich G, Moricke A, Henze G, von Stackelberg A. Secondary malignant neoplasms after intensive treatment of relapsed acute lymphoblastic leukaemia in childhood. European Journal of Cancer (Oxford, England : 1990). 2008;44(2):257–68. Epub 2007/11/06. doi: 10.1016/j.ejca.2007.09.019. [DOI] [PubMed] [Google Scholar]
- 7.Henderson TO, Whitton J, Stovall M, Mertens AC, Mitby P, Friedman D, Strong LC, Hammond S, Neglia JP, Meadows AT, Robison L, Diller L. Secondary sarcomas in childhood cancer survivors: a report from the Childhood Cancer Survivor Study. Journal of the National Cancer Institute. 2007;99(4):300–8. Epub 2007/02/22. doi: 10.1093/jnci/djk052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kim SH, Shin KH, Seok SO, Cho YJ, Noh JK, Suh JS, Yang WI. Secondary malignant neoplasms after osteosarcoma: early onset and cumulative alkylating agent dose dependency. Annals of Surgical Oncology. 2015;22(3):859–65. Epub 2014/09/07. doi: 10.1245/s10434-014-4070-2. [DOI] [PubMed] [Google Scholar]
- 9.Bassal M, Mertens AC, Taylor L, Neglia JP, Greffe BS, Hammond S, Ronckers CM, Friedman DL, Stovall M, Yasui YY, Robison LL, Meadows AT, Kadan-Lottick NS. Risk of selected subsequent carcinomas in survivors of childhood cancer: a report from the Childhood Cancer Survivor Study. Journal of Clinical Oncology. 2006;24(3):476–83. Epub 2006/01/20. doi: 10.1200/jco.2005.02.7235. [DOI] [PubMed] [Google Scholar]
- 10.Meadows AT, Friedman DL, Neglia JP, Mertens AC, Donaldson SS, Stovall M, Hammond S, Yasui Y, Inskip PD. Second neoplasms in survivors of childhood cancer: findings from the Childhood Cancer Survivor Study cohort. Journal of Clinical Oncology. 2009;27(14):2356–62. Epub 2009/03/04. doi: 10.1200/jco.2008.21.1920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Friedman DL, Whitton J, Leisenring W, Mertens AC, Hammond S, Stovall M, Donaldson SS, Meadows AT, Robison LL, Neglia JP. Subsequent neoplasms in 5-year survivors of childhood cancer: the Childhood Cancer Survivor Study. Journal of the National Cancer Institute. 2010;102(14):1083–95. Epub 2010/07/17. doi: 10.1093/jnci/djq238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Rihani R, Bazzeh F, Faqih N, Sultan I. Secondary hematopoietic malignancies in survivors of childhood cancer: an analysis of 111 cases from the Surveillance, Epidemiology, and End Result-9 registry. Cancer. 2010;116(18):4385–94. Epub 2010/06/16. doi: 10.1002/cncr.25313. [DOI] [PubMed] [Google Scholar]
- 13.Reulen RC, Frobisher C, Winter DL, Kelly J, Lancashire ER, Stiller CA, Pritchard-Jones K, Jenkinson HC, Hawkins MM. Long-term risks of subsequent primary neoplasms among survivors of childhood cancer. JAMA. 2011;305(22):2311–9. Epub 2011/06/07. doi: 10.1001/jama.2011.747. [DOI] [PubMed] [Google Scholar]
- 14.Bluhm EC, Ronckers C, Hayashi RJ, Neglia JP, Mertens AC, Stovall M, Meadows AT, Mitby PA, Whitton JA, Hammond S, Barker JD, Donaldson SS, Robison LL, Inskip PD. Cause-specific mortality and second cancer incidence after non-Hodgkin lymphoma: a report from the Childhood Cancer Survivor Study. Blood. 2008;111(8):4014–21. Epub 2008/02/09. doi: 10.1182/blood-2007-08-106021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mertens AC, Liu Q, Neglia JP, Wasilewski K, Leisenring W, Armstrong GT, Robison LL, Yasui Y. Cause-specific late mortality among 5-year survivors of childhood cancer: the Childhood Cancer Survivor Study. Journal of the National Cancer Institute. 2008;100(19):1368–79. Epub 2008/09/25. doi: 10.1093/jnci/djn310. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Armstrong GT, Chen Y, Yasui Y, Leisenring W, Gibson TM, Mertens AC, Stovall M, Oeffinger KC, Bhatia S, Krull KR, Nathan PC, Neglia JP, Green DM, Hudson MM, Robison LL. Reduction in Late Mortality among 5-Year Survivors of Childhood Cancer. The New England Journal of Medicine. 2016;374(9):833–42. Epub 2016/01/14. doi: 10.1056/NEJMoa1510795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.SEER*Stat Database: Incidence SEER18 Research Database, November 2016 submission (Katrina/Rita Population Adjustment) [Internet]. National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch; November 2016. Available from: www.seer.cancer.gov. [Google Scholar]
- 18.Lupo PJ, Brown AL, Hettmer S. Second malignancy risk among pediatric, adolescent, and young adult survivors of fusion-positive and fusion-negative sarcomas: Results from the SEER database, 1992 through 2012. Cancer. 2016. Epub 2016/08/03. doi: 10.1002/cncr.30222. [DOI] [PubMed] [Google Scholar]
- 19.National Cancer Institute. Surveillance E, and End Results (SEER) Program. Site Recode ICD-O-3/WHO 2008 Definition: National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch; July 2018. Available from: https://seer.cancer.gov/siterecode/icdo3_dwhoheme/. [Google Scholar]
- 20.Noone AM, Lund JL, Mariotto A, Cronin K, McNeel T, Deapen D, Warren JL. Comparison of SEER Treatment Data With Medicare Claims. Medical Care. 2016;54(9):e55–64. Epub 2014/03/19. doi: 10.1097/mlr.0000000000000073. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tukenova M, Diallo I, Hawkins M, Guibout C, Quiniou E, Pacquement H, Dhermain F, Shamsaldin A, Oberlin O, de Vathaire F. Long-term mortality from second malignant neoplasms in 5-year survivors of solid childhood tumors: temporal pattern of risk according to type of treatment. Cancer Epidemiology, Biomarkers & Prevention. 2010;19(3):707–15. Epub 2010/03/05. doi: 10.1158/1055-9965.epi-09-1156. [DOI] [PubMed] [Google Scholar]
- 22.MacArthur AC, Spinelli JJ, Rogers PC, Goddard KJ, Phillips N, McBride ML. Risk of a second malignant neoplasm among 5-year survivors of cancer in childhood and adolescence in British Columbia, Canada. Pediatric Blood & Cancer. 2007;48(4):453–9. Epub 2006/06/13. doi: 10.1002/pbc.20921. [DOI] [PubMed] [Google Scholar]
- 23.Dinh KT, Mahal BA, Ziehr DR, Muralidhar V, Chen YW, Viswanathan VB, Nezolosky MD, Beard CJ, Choueiri TK, Martin NE, Orio PF, Sweeney CJ, Trinh QD, Nguyen PL. Risk of prostate cancer mortality in men with a history of prior cancer. BJU International. 2016;117(6b):E20–8. Epub 2015/04/08. doi: 10.1111/bju.13144. [DOI] [PubMed] [Google Scholar]
- 24.Laccetti AL, Pruitt SL, Xuan L, Halm EA, Gerber DE. Effect of prior cancer on outcomes in advanced lung cancer: implications for clinical trial eligibility and accrual. Journal of the National Cancer Institute. 2015;107(4). Epub 2015/02/11. doi: 10.1093/jnci/djv002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Smyth EC, Tarazona N, Peckitt C, Armstrong E, Mansukhani S, Cunningham D, Chau I. Exclusion of Gastrointestinal Cancer Patients With Prior Cancer From Clinical Trials: Is This Justified? Clinical Colorectal Cancer. 2016;15(2):e53–9. Epub 2016/01/10. doi: 10.1016/j.clcc.2015.11.003. [DOI] [PubMed] [Google Scholar]
- 26.Zhou H, Huang Y, Qiu Z, Zhao H, Fang W, Yang Y, Zhao Y, Hou X, Ma Y, Hong S, Zhou T, Zhang Y, Zhang L. Impact of prior cancer history on the overall survival of patients newly diagnosed with cancer: A pan-cancer analysis of the SEER database. International Journal of Cancer. 2018. Epub 2018/04/19. doi: 10.1002/ijc.31543. [DOI] [PubMed] [Google Scholar]
- 27.Keegan THM, Bleyer A, Rosenberg AS, Li Q, Goldfarb M. Second Primary Malignant Neoplasms and Survival in Adolescent and Young Adult Cancer Survivors. JAMA Oncology. 2017;3(11):1554–7. Epub 2017/04/21. doi: 10.1001/jamaoncol.2017.0465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bowers DC, Liu Y, Leisenring W, McNeil E, Stovall M, Gurney JG, Robison LL, Packer RJ, Oeffinger KC. Late-occurring stroke among long-term survivors of childhood leukemia and brain tumors: a report from the Childhood Cancer Survivor Study. Journal of Clinical Oncology. 2006;24(33):5277–82. Epub 2006/11/08. doi: 10.1200/jco.2006.07.2884. [DOI] [PubMed] [Google Scholar]
- 29.Wilson CL, Liu W, Yang JJ, Kang G, Ojha RP, Neale GA, Srivastava DK, Gurney JG, Hudson MM, Robison LL, Ness KK. Genetic and clinical factors associated with obesity among adult survivors of childhood cancer: A report from the St. Jude Lifetime Cohort. Cancer. 2015;121(13):2262–70. Epub 2015/05/13. doi: 10.1002/cncr.29153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Meacham LR, Gurney JG, Mertens AC, Ness KK, Sklar CA, Robison LL, Oeffinger KC. Body mass index in long-term adult survivors of childhood cancer: a report of the Childhood Cancer Survivor Study. Cancer. 2005;103(8):1730–9. Epub 2005/03/12. doi: 10.1002/cncr.20960. [DOI] [PubMed] [Google Scholar]
- 31.Ness KK, Armstrong GT, Kundu M, Wilson CL, Tchkonia T, Kirkland JL. Frailty in childhood cancer survivors. Cancer. 2015;121(10):1540–7. Epub 2014/12/23. doi: 10.1002/cncr.29211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Armenian SH, Armstrong GT, Aune G, Chow EJ, Ehrhardt MJ, Ky B, Moslehi J, Mulrooney DA, Nathan PC, Ryan TD, van der Pal HJ, van Dalen EC, Kremer LCM. Cardiovascular Disease in Survivors of Childhood Cancer: Insights Into Epidemiology, Pathophysiology, and Prevention. Journal of Clinical Oncology. 2018;36(21):2135–44. Epub 2018/06/07. doi: 10.1200/jco.2017.76.3920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Protani M, Coory M, Martin JH. Effect of obesity on survival of women with breast cancer: systematic review and meta-analysis. Breast Cancer Research and Treatment. 2010;123(3):627–35. Epub 2010/06/24. doi: 10.1007/s10549-010-0990-0. [DOI] [PubMed] [Google Scholar]
- 34.Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. The New England Journal of Medicine. 2003;348(17):1625–38. Epub 2003/04/25. doi: 10.1056/NEJMoa021423. [DOI] [PubMed] [Google Scholar]
- 35.Schmid D, Leitzmann MF. Cardiorespiratory fitness as predictor of cancer mortality: a systematic review and meta-analysis. Annals of Oncology. 2015;26(2):272–8. Epub 2014/07/11. doi: 10.1093/annonc/mdu250. [DOI] [PubMed] [Google Scholar]
- 36.Handforth C, Clegg A, Young C, Simpkins S, Seymour MT, Selby PJ, Young J. The prevalence and outcomes of frailty in older cancer patients: a systematic review. Annals of Oncology. 2015;26(6):1091–101. Epub 2014/11/19. doi: 10.1093/annonc/mdu540. [DOI] [PubMed] [Google Scholar]
- 37.Behjati S, Gundem G, Wedge DC, Roberts ND, Tarpey PS, Cooke SL, Van Loo P, Alexandrov LB, Ramakrishna M, Davies H, Nik-Zainal S, Hardy C, Latimer C, Raine KM, Stebbings L, Menzies A, Jones D, Shepherd R, Butler AP, Teague JW, Jorgensen M, Khatri B, Pillay N, Shlien A, Futreal PA, Badie C, McDermott U, Bova GS, Richardson AL, Flanagan AM, Stratton MR, Campbell PJ. Mutational signatures of ionizing radiation in second malignancies. Nature Communications. 2016;7:12605 Epub 2016/09/13. doi: 10.1038/ncomms12605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gonin-Laurent N, Gibaud A, Huygue M, Lefevre SH, Le Bras M, Chauveinc L, Sastre-Garau X, Doz F, Lumbroso L, Chevillard S, Malfoy B. Specific TP53 mutation pattern in radiation-induced sarcomas. Carcinogenesis. 2006;27(6):1266–72. Epub 2006/02/24. doi: 10.1093/carcin/bgi356. [DOI] [PubMed] [Google Scholar]
- 39.Gonin-Laurent N, Hadj-Hamou NS, Vogt N, Houdayer C, Gauthiers-Villars M, Dehainault C, Sastre-Garau X, Chevillard S, Malfoy B. RB1 and TP53 pathways in radiation-induced sarcomas. Oncogene. 2007;26(41):6106–12. Epub 2007/03/21. doi: 10.1038/sj.onc.1210404. [DOI] [PubMed] [Google Scholar]
- 40.Olivier M, Langerod A, Carrieri P, Bergh J, Klaar S, Eyfjord J, Theillet C, Rodriguez C, Lidereau R, Bieche I, Varley J, Bignon Y, Uhrhammer N, Winqvist R, Jukkola-Vuorinen A, Niederacher D, Kato S, Ishioka C, Hainaut P, Borresen-Dale AL. The clinical value of somatic TP53 gene mutations in 1,794 patients with breast cancer. Clinical Cancer Research. 2006;12(4):1157–67. Epub 2006/02/21. doi: 10.1158/1078-0432.ccr-05-1029. [DOI] [PubMed] [Google Scholar]
- 41.Yin M, Grivas P, Emamekhoo H, Mendiratta P, Ali S, Hsu J, Vasekar M, Drabick JJ, Pal S, Joshi M. ATM/RB1 mutations predict shorter overall survival in urothelial cancer. Oncotarget. 2018;9(24):16891–8. Epub 2018/04/24. doi: 10.18632/oncotarget.24738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Russo A, Bazan V, Iacopetta B, Kerr D, Soussi T, Gebbia N. The TP53 colorectal cancer international collaborative study on the prognostic and predictive significance of p53 mutation: influence of tumor site, type of mutation, and adjuvant treatment. Journal of Clinical Oncology. 2005;23(30):7518–28. Epub 2005/09/21. doi: 10.1200/jco.2005.00.471. [DOI] [PubMed] [Google Scholar]
- 43.Veiga LH, Bhatti P, Ronckers CM, Sigurdson AJ, Stovall M, Smith SA, Weathers R, Leisenring W, Mertens AC, Hammond S, Neglia JP, Meadows AT, Donaldson SS, Sklar CA, Friedman DL, Robison LL, Inskip PD. Chemotherapy and thyroid cancer risk: a report from the childhood cancer survivor study. Cancer Epidemiology, Biomarkers & Prevention. 2012;21(1):92–101. Epub 2011/10/27. doi: 10.1158/1055-9965.epi-11-0576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Peterson KM, Shao C, McCarter R, MacDonald TJ, Byrne J. An analysis of SEER data of increasing risk of secondary malignant neoplasms among long-term survivors of childhood brain tumors. Pediatric Blood & Cancer. 2006;47(1):83–8. Epub 2005/12/01. doi: 10.1002/pbc.20690. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.