https://orcid.org/0000-0003-4668-9176
Abstract
Early efforts at risk-adapted therapy for neuroblastoma are predicted to result in differential late effects; the magnitude of these differences has not been well described.
Late mortality, subsequent malignant neoplasms (SMNs), and severe/life-threatening chronic health conditions (CHCs), graded according to CTCAE v4.03, were assessed among 5-year Childhood Cancer Survivor Study (CCSS) survivors of neuroblastoma diagnosed 1987-1999. Using age, stage at diagnosis, and treatment, survivors were classified into risk groups (low [n = 425]; intermediate [n = 252]; high [n = 245]). Standardized mortality ratios (SMRs) and standardized incidence ratios (SIRs) of SMNs were compared with matched population controls. Cox regression models estimated hazard ratios (HRs) and 95% confidence intervals for CHC compared with 1029 CCSS siblings.
Among survivors (49.8% male; median age = 21 years, range = 7-42; median follow-up = 19.3 years, range = 5-29.9), 80% with low-risk disease were treated with surgery alone, whereas 79.1% with high-risk disease received surgery, radiation, chemotherapy ± autologous stem cell transplant (ASCT). All-cause mortality was elevated across risk groups (SMRhigh = 27.7 [21.4-35.8]; SMRintermediate = 3.3 [1.7-6.5]; SMRlow = 2.8 [1.7-4.8]). SMN risk was increased among high- and intermediate-risk survivors (SIRhigh = 28.0 [18.5-42.3]; SIRintermediate = 3.7 [1.2-11.3]) but did not differ from the US population for survivors of low-risk disease. Compared with siblings, survivors had an increased risk of grade 3-5 CHCs, particularly among those with high-risk disease (HRhigh = 16.1 [11.2-23.2]; HRintermediate = 6.3 [3.8-10.5]; HRlow = 1.8 [1.1-3.1]).
Survivors of high-risk disease treated in the early days of risk stratification carry a markedly elevated burden of late recurrence, SMN, and organ-related multimorbidity, whereas survivors of low/intermediate-risk disease have a modest risk of late adverse outcomes.
Neuroblastoma, the most common extracranial solid tumor of childhood, is marked by its clinical heterogeneity (1,2). Treatment is tailored according to risk-group classification on the basis of a combination of markers that are predictive of risk of relapse and death (1,3-,6). During the past 4 decades, therapy has been sequentially reduced for patients with low- and intermediate-risk disease, whereas increasingly intensive multimodal therapy has been used to treat high-risk patients (7). This risk-based approach has led to excellent survival among patients with low- and intermediate-risk disease and improved survival for high-risk patients (1,8). However, outcomes for high-risk patients remain suboptimal, with approximately 40% developing recurrent disease with current standard of care treatment (9). Survivors of high-risk disease were among the first pediatric patients to receive novel therapies, such as isotretinoin and immunotherapy, either alone or in combination (7,10). Yet granular data are lacking on how therapeutic shifts over the past several decades have impacted neuroblastoma survivors’ risk for late mortality and other long-term chronic health conditions (CHCs).
The current project leverages the Childhood Cancer Survivor Study (CCSS), which includes longitudinal follow-up of 5-year survivors of childhood cancer. Prior analyses of late health outcomes among neuroblastoma survivors in CCSS have been limited by the lack of available data on clinical stage, which precluded any analyses of late morbidity and mortality in this cohort by disease risk group. Here, we used the more contemporary expansion cohort, which includes data on disease stage, to analyze late mortality, risk of subsequent malignant neoplasm (SMN), and CHCs among survivors of neuroblastoma treated in the early era of risk stratification. Late outcomes among survivors both overall and by risk group were assessed, and the burden of late mortality and morbidity was compared with matched population-based controls (mortality and SMN) and with an unaffected sibling cohort (CHCs).
Methods
Study design and participants
The CCSS is a retrospective cohort of 25 665 5-year survivors of childhood cancer diagnosed at younger than 21 years of age at 1 of 31 North American institutions between 1970 and 1999 with longitudinal prospective follow-up. Details of the cohort methodology and study design have been previously reported (11,12). Participating sites abstracted data on cancer diagnosis and treatment exposures from the medical records of all participants who provided authorization (12,13). Select variables, such as disease stage, were collected among participants diagnosed between 1987 and 1999 (“expansion cohort”) but not among those diagnosed between 1970 and 1986 (“original cohort”). This analysis, which focuses on late adverse outcomes by disease risk group, was limited to neuroblastoma survivors diagnosed between 1987 and 1999.
Using age at diagnosis, clinical stage at diagnosis, and treatment exposures (including cooperative group protocol numbers, when available [Supplementary Table 1, available online]), participants were assigned to a low-, intermediate-, or high-risk disease group (5,6). Data were not available on histopathology or tumor biology. Seventy-seven participants were not assigned a risk group because of lack of treatment information or missing stage data. Of note, we previously reported on late adverse outcomes among 1397 neuroblastoma survivors diagnosed during infancy between 1970 and 1999 (14); a subset of those survivors (N = 469) is included in this analysis.
Overall, 1452 neuroblastoma survivors diagnosed between 1987 and 1999 were evaluable for late mortality; 999 survivors completed a baseline survey and up to 2 follow-up questionnaires, which included questions on late recurrence, SMN, and CHCs (Supplementary Figure 1, available online). A sample of 1029 siblings of CCSS participants completed nearly identical surveys and served as a comparison group. Informed consent was obtained from participants or guardians. Each site obtained institutional review board approval.
Outcomes of interest: Late mortality, SMN, and CHCs
Late mortality was defined as death occurring 5 years or longer from primary neuroblastoma diagnosis as ascertained from the National Death Index through December 31, 2017 (15). Cause of death was classified into 1 of 3 mutually exclusive categories: late recurrence/progression of neuroblastoma; other health-related causes, including SMN, cardiac, pulmonary, infectious, or other medical causes (and excludes deaths from neuroblastoma); or external causes, including accident, injury, and suicide.
SMNs, defined as new primary malignancies occurring 5 years or longer from the date of primary neuroblastoma diagnosis, were ascertained through questionnaires and confirmed by pathology report, medical records, and/or death certificate for all CCSS participants. Recurrences of neuroblastoma were excluded from SMN categorization and considered separately.
CHCs were ascertained from detailed questionnaires, which are publicly available at https://ccss.stjude.org/tools-and-documents/questionnaires.html, and graded for severity using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE, v4.03) as mild (grade 1), moderate (grade 2), severe or disabling (grade 3), life-threatening (grade 4), or fatal (grade 5) (16), using previously described methods (17). Grade 1 or mild CHCs, which do not require medication or other intervention, were combined with “no condition” as a reference group.
Statistical analysis
Survivor and sibling characteristics were summarized using frequencies and median values for categorical and continuous measures, respectively. Cumulative mortality was calculated from cohort entry using Kaplan-Meier estimation for overall mortality and cumulative incidence for cause-specific mortality, treating death from other causes as competing risks. Follow-up extended to the date of the last National Death Index search or death, whichever came first. Comparisons of cumulative mortality curves were completed using Gray’s test (18). Standardized mortality ratios (SMRs) of observed to expected deaths and excess absolute risk (EAR) of death (observed minus expected deaths per 10 000 person-years) were computed for the population of US-based eligible neuroblastoma survivors using age-, sex-, and calendar-year-matched rates to calculate expected numbers of deaths. US population rates were obtained from the Centers for Disease Control and Prevention (15).
Among participants, standardized incidence ratios (SIRs) were calculated as ratios of observed numbers of SMNs to number of expected events using age-, sex-, and calendar-year-matched US population rates from Surveillance, Epidemiology, and End Results (19). EAR of SMN was calculated as observed minus expected SMNs per 10 000 person-years.
Cumulative incidence of SMN and CHCs were calculated starting from cohort entry at 5 years after diagnosis with follow-up to event of interest or last contact, treating death from other causes as a competing risk. Cox proportional hazards models were fit for grade 3-5 CHCs occurring after cohort entry, to compare survivors with siblings and, among survivors, to assess associations of treatment and cancer characteristics with hazards of CHCs. Cox models used age as the time scale, with participants entering the risk set at the age they entered the cohort, and were adjusted for sex and race/ethnicity. SAS 9.4 and R 3.5.3 were used for these analyses. All reported P values are 2-sided and considered significant if less than .05.
Results
Table 1 describes characteristics of the 999 participants (median age = 21 years, range = 7-42; median follow-up = 19 years, range = 5-30), which included 425 (42.5%) survivors with low-risk disease, 252 (25.2%) with intermediate-risk disease, and 245 (24.5%) with high-risk disease, and a sibling comparison group (N = 1029). Whereas 80% of individuals with low-risk disease were treated with surgery alone, 79% of individuals with high-risk disease received multimodality therapy with chemotherapy, radiation, surgery, and at least 1 autologous stem cell transplant (ASCT). Seventy-seven of the 195 individuals (39.5%) treated with ASCT received total body irradiation (TBI)-based conditioning.
Demographic and treatment characteristics of 5-year survivors of neuroblastoma, overall and by risk group and enrolled in the Childhood Cancer Survivor Study (CCSS) expansion cohort, when compared to a sibling comparison group
| Characteristic | All neuroblastoma survivors (N = 999) | Survivors with low-risk disease (n = 425) | Survivors with intermediate-risk disease (n = 252) | Survivors with high-risk disease (n = 245) | P valuea | Siblings (N = 1029) |
|---|---|---|---|---|---|---|
| Sex, No. (%) | ||||||
| Male | 498 (49.8) | 199 (46.8) | 127 (50.4) | 134 (54.7) | .14 | 471 (45.8) |
| Female | 501 (50.2) | 226 (53.2) | 125 (49.6) | 111 (45.3) | 558 (54.2) | |
| Race/ethnicity, No. (%) | ||||||
| Hispanic | 104 (10.4) | 53 (12.5) | 26 (10.3) | 17 (6.9) | 66 (6.6) | |
| Non-Hispanic, Black | 68 (6.8) | 20 (4.7) | 14 (5.6) | 22 (9.0) | 37 (3.7) | |
| Non-Hispanic, White | 785 (78.6) | 335 (78.8) | 205 (81.3) | 190 (77.6) | .03 | 864 (86.9) |
| Other | 42 (4.2) | 17 (4.0) | 7 (2.8) | 16 (6.5) | 27 (2.7) | |
| Age at diagnosis, years, No. (%) | ||||||
| <1 | 469 (46.9) | 229 (53.9) | 177 (70.2) | 16 (6.5) | <.001 | – |
| 1-4 | 422 (42.2) | 154 (36.2) | 69 (27.4) | 175 (71.4) | – | |
| 5-9 | 82 (8.2) | 29 (6.8) | 6 (2.4) | 44 (18.0) | – | |
| ≥10 | 26 (2.6) | 13 (3.1) | 0 (0.0) | 10 (4.1) | – | |
| Treatment category, No. (%) | ||||||
| Surgery only | 351 (36.6) | 337 (80.1) | 8 (3.2) | 0 (0.0) | <.001 | – |
| Chemotherapy ± surgery | 343 (35.8) | 67 (15.9) | 200 (79.4) | 50 (20.9) | – | |
| Radiation ± chemotherapy ± surgery | 265 (27.6) | 17 (4.0) | 44 (17.5) | 189 (79.1) | – | |
| Missing | 40 | 4 | 0 | 6 | – | |
| Autologous stem cell transplant, No. (%) | ||||||
| Yes | 195 (20.2) | 4 (0.9) | 5 (2.0) | 186 (77.5) | <.001 | – |
| Total body irradiation (TBI), No. (%) | ||||||
| Yes | 77 (8.2) | 0 (0.0) | 0 (0.0) | 77 (34.1) | <.001 | – |
| TBI dose (Gy), median, range | 12 (2-13.8) | 0 (0.0) | 0 (0.0) | 12 (2.0-13.8) | <.001 | – |
| Abdominal radiation, N (%) | ||||||
| Yes | 207 (22.0) | 10 (2.4) | 31 (12.5) | 156 (69.0) | <.001 | – |
| Abdomen radiation dose (Gy), median, range | 20.0 (2.0-51.0) | 22.5 (4.5-46.0) | 12.0 (3.0-51.0) | 21.0 (2.0-51.0) | .02 | – |
| Cranial radiation, No. (%) | ||||||
| Yes | 111 (11.8) | 1 (0.2) | 4 (1.6) | 100 (44.2) | <.001 | – |
| Cranial radiation dose (Gy), median, range | 12.0 (2.0-66.0) | 21 (21.0-21.0) | 13.5 (6.0-15.0) | 12 (2.0-48.0) | .54 | - |
| Chest radiation, No. (%) | ||||||
| Yes | 126 (13.4) | 6 (1.4) | 4 (1.6) | 111 (49.1) | <.001 | – |
| Chest radiation dose, (Gy), median, range | 12 (2.0-45.0) | 22.5 (3.0-27.0) | 13.2 (4.5-22.0) | 12 (2.0-45.0) | .32 | – |
| Alkylating agents, No. (%) | ||||||
| Yes | 598 (61.8) | 80 (18.9) | 239 (94.8) | 238 (98.8) | <.001 | – |
| Cyclophosphamide equivalent dose | 6467.6 (2.2-68 944.4) | 5311.5 (178.3-43 742.1) | 5131.2 (2.2-59 939.1) | 13 146.2 (1050.7-68 944.4) | <.001 | – |
| Anthracyclines, No. (%) | ||||||
| Yes | 536 (55.3) | 69 (16.3) | 209 (82.9) | 229 (94.6) | <.001 | – |
| Anthracycline dose, mg/m2, median, range | 147.7 (0.0-750.8) | 150.5 (0.0-515.7) | 141.2 (0.3-507.1) | 151.5 (0.1-685.1) | .01 | – |
| Platinum agents, No. (%) | ||||||
| Yes | 501 (51.6) | 44 (10.4) | 187 (74.2) | 241 (99.6) | <.001 | – |
| Platinum dose, mg/m2, median, range | 497.6 (0.6-3,203.4) | 439.4 (35.2-1251.9) | 429.9 (0.6-3203.4) | 621.4 (90.9-2994.4) | <.001 | – |
| Epipodophyllotoxins, No. (%) | ||||||
| Yes | 490 (50.7) | 42 (9.9) | 186 (73.8) | 236 (98.7) | <.001 | – |
| Epipodophyllotoxin dose, mg/m2, median, range | 1721.4 (0.5-14 305.2) | 1223.3 (0.5-5567.1) | 1421.5 (43.6-8334.1) | 2050.8 (1.9-14 305.2) | <.001 | – |
| Isotretinoin, No. (%) | ||||||
| Yes | 76 (7.9) | 5 (1.2) | 7 (2.8) | 63 (26.7) | <.001 | – |
| Years of follow-up from diagnosis, median, range | 19.3 (5.0-29.9) | 19.4 (5.0-29.7) | 20.2 (6.5-29.2) | 17.9 (5.1-29.2) | <.001 | |
| Age at last contact, y, median, range | 21 (7-42) | 21 (7-42) | 21 (9-30) | 20 (7-38) | .31 | 28 (7-54) |
| Vital status, No. (%) | 1018 (98.9) | |||||
| Alive | 906 (90.7) | 409 (96.2) | 244 (96.8) | 184 (75.1) | <.001 | 1018 (98.9) |
| Dead | 93 (9.3) | 16 (3.8) | 8 (3.2) | 61 (24.9) | 11 (1.1) |
| Characteristic | All neuroblastoma survivors (N = 999) | Survivors with low-risk disease (n = 425) | Survivors with intermediate-risk disease (n = 252) | Survivors with high-risk disease (n = 245) | P valuea | Siblings (N = 1029) |
|---|---|---|---|---|---|---|
| Sex, No. (%) | ||||||
| Male | 498 (49.8) | 199 (46.8) | 127 (50.4) | 134 (54.7) | .14 | 471 (45.8) |
| Female | 501 (50.2) | 226 (53.2) | 125 (49.6) | 111 (45.3) | 558 (54.2) | |
| Race/ethnicity, No. (%) | ||||||
| Hispanic | 104 (10.4) | 53 (12.5) | 26 (10.3) | 17 (6.9) | 66 (6.6) | |
| Non-Hispanic, Black | 68 (6.8) | 20 (4.7) | 14 (5.6) | 22 (9.0) | 37 (3.7) | |
| Non-Hispanic, White | 785 (78.6) | 335 (78.8) | 205 (81.3) | 190 (77.6) | .03 | 864 (86.9) |
| Other | 42 (4.2) | 17 (4.0) | 7 (2.8) | 16 (6.5) | 27 (2.7) | |
| Age at diagnosis, years, No. (%) | ||||||
| <1 | 469 (46.9) | 229 (53.9) | 177 (70.2) | 16 (6.5) | <.001 | – |
| 1-4 | 422 (42.2) | 154 (36.2) | 69 (27.4) | 175 (71.4) | – | |
| 5-9 | 82 (8.2) | 29 (6.8) | 6 (2.4) | 44 (18.0) | – | |
| ≥10 | 26 (2.6) | 13 (3.1) | 0 (0.0) | 10 (4.1) | – | |
| Treatment category, No. (%) | ||||||
| Surgery only | 351 (36.6) | 337 (80.1) | 8 (3.2) | 0 (0.0) | <.001 | – |
| Chemotherapy ± surgery | 343 (35.8) | 67 (15.9) | 200 (79.4) | 50 (20.9) | – | |
| Radiation ± chemotherapy ± surgery | 265 (27.6) | 17 (4.0) | 44 (17.5) | 189 (79.1) | – | |
| Missing | 40 | 4 | 0 | 6 | – | |
| Autologous stem cell transplant, No. (%) | ||||||
| Yes | 195 (20.2) | 4 (0.9) | 5 (2.0) | 186 (77.5) | <.001 | – |
| Total body irradiation (TBI), No. (%) | ||||||
| Yes | 77 (8.2) | 0 (0.0) | 0 (0.0) | 77 (34.1) | <.001 | – |
| TBI dose (Gy), median, range | 12 (2-13.8) | 0 (0.0) | 0 (0.0) | 12 (2.0-13.8) | <.001 | – |
| Abdominal radiation, N (%) | ||||||
| Yes | 207 (22.0) | 10 (2.4) | 31 (12.5) | 156 (69.0) | <.001 | – |
| Abdomen radiation dose (Gy), median, range | 20.0 (2.0-51.0) | 22.5 (4.5-46.0) | 12.0 (3.0-51.0) | 21.0 (2.0-51.0) | .02 | – |
| Cranial radiation, No. (%) | ||||||
| Yes | 111 (11.8) | 1 (0.2) | 4 (1.6) | 100 (44.2) | <.001 | – |
| Cranial radiation dose (Gy), median, range | 12.0 (2.0-66.0) | 21 (21.0-21.0) | 13.5 (6.0-15.0) | 12 (2.0-48.0) | .54 | - |
| Chest radiation, No. (%) | ||||||
| Yes | 126 (13.4) | 6 (1.4) | 4 (1.6) | 111 (49.1) | <.001 | – |
| Chest radiation dose, (Gy), median, range | 12 (2.0-45.0) | 22.5 (3.0-27.0) | 13.2 (4.5-22.0) | 12 (2.0-45.0) | .32 | – |
| Alkylating agents, No. (%) | ||||||
| Yes | 598 (61.8) | 80 (18.9) | 239 (94.8) | 238 (98.8) | <.001 | – |
| Cyclophosphamide equivalent dose | 6467.6 (2.2-68 944.4) | 5311.5 (178.3-43 742.1) | 5131.2 (2.2-59 939.1) | 13 146.2 (1050.7-68 944.4) | <.001 | – |
| Anthracyclines, No. (%) | ||||||
| Yes | 536 (55.3) | 69 (16.3) | 209 (82.9) | 229 (94.6) | <.001 | – |
| Anthracycline dose, mg/m2, median, range | 147.7 (0.0-750.8) | 150.5 (0.0-515.7) | 141.2 (0.3-507.1) | 151.5 (0.1-685.1) | .01 | – |
| Platinum agents, No. (%) | ||||||
| Yes | 501 (51.6) | 44 (10.4) | 187 (74.2) | 241 (99.6) | <.001 | – |
| Platinum dose, mg/m2, median, range | 497.6 (0.6-3,203.4) | 439.4 (35.2-1251.9) | 429.9 (0.6-3203.4) | 621.4 (90.9-2994.4) | <.001 | – |
| Epipodophyllotoxins, No. (%) | ||||||
| Yes | 490 (50.7) | 42 (9.9) | 186 (73.8) | 236 (98.7) | <.001 | – |
| Epipodophyllotoxin dose, mg/m2, median, range | 1721.4 (0.5-14 305.2) | 1223.3 (0.5-5567.1) | 1421.5 (43.6-8334.1) | 2050.8 (1.9-14 305.2) | <.001 | – |
| Isotretinoin, No. (%) | ||||||
| Yes | 76 (7.9) | 5 (1.2) | 7 (2.8) | 63 (26.7) | <.001 | – |
| Years of follow-up from diagnosis, median, range | 19.3 (5.0-29.9) | 19.4 (5.0-29.7) | 20.2 (6.5-29.2) | 17.9 (5.1-29.2) | <.001 | |
| Age at last contact, y, median, range | 21 (7-42) | 21 (7-42) | 21 (9-30) | 20 (7-38) | .31 | 28 (7-54) |
| Vital status, No. (%) | 1018 (98.9) | |||||
| Alive | 906 (90.7) | 409 (96.2) | 244 (96.8) | 184 (75.1) | <.001 | 1018 (98.9) |
| Dead | 93 (9.3) | 16 (3.8) | 8 (3.2) | 61 (24.9) | 11 (1.1) |
Treatment data are shown among those with medical record abstraction; 77 had missing treatment data and were unable to be classified into a disease risk group.
χ2, Fisher exact, or Kruskal-Wallis test for test of association with risk.
Demographic and treatment characteristics of 5-year survivors of neuroblastoma, overall and by risk group and enrolled in the Childhood Cancer Survivor Study (CCSS) expansion cohort, when compared to a sibling comparison group
| Characteristic | All neuroblastoma survivors (N = 999) | Survivors with low-risk disease (n = 425) | Survivors with intermediate-risk disease (n = 252) | Survivors with high-risk disease (n = 245) | P valuea | Siblings (N = 1029) |
|---|---|---|---|---|---|---|
| Sex, No. (%) | ||||||
| Male | 498 (49.8) | 199 (46.8) | 127 (50.4) | 134 (54.7) | .14 | 471 (45.8) |
| Female | 501 (50.2) | 226 (53.2) | 125 (49.6) | 111 (45.3) | 558 (54.2) | |
| Race/ethnicity, No. (%) | ||||||
| Hispanic | 104 (10.4) | 53 (12.5) | 26 (10.3) | 17 (6.9) | 66 (6.6) | |
| Non-Hispanic, Black | 68 (6.8) | 20 (4.7) | 14 (5.6) | 22 (9.0) | 37 (3.7) | |
| Non-Hispanic, White | 785 (78.6) | 335 (78.8) | 205 (81.3) | 190 (77.6) | .03 | 864 (86.9) |
| Other | 42 (4.2) | 17 (4.0) | 7 (2.8) | 16 (6.5) | 27 (2.7) | |
| Age at diagnosis, years, No. (%) | ||||||
| <1 | 469 (46.9) | 229 (53.9) | 177 (70.2) | 16 (6.5) | <.001 | – |
| 1-4 | 422 (42.2) | 154 (36.2) | 69 (27.4) | 175 (71.4) | – | |
| 5-9 | 82 (8.2) | 29 (6.8) | 6 (2.4) | 44 (18.0) | – | |
| ≥10 | 26 (2.6) | 13 (3.1) | 0 (0.0) | 10 (4.1) | – | |
| Treatment category, No. (%) | ||||||
| Surgery only | 351 (36.6) | 337 (80.1) | 8 (3.2) | 0 (0.0) | <.001 | – |
| Chemotherapy ± surgery | 343 (35.8) | 67 (15.9) | 200 (79.4) | 50 (20.9) | – | |
| Radiation ± chemotherapy ± surgery | 265 (27.6) | 17 (4.0) | 44 (17.5) | 189 (79.1) | – | |
| Missing | 40 | 4 | 0 | 6 | – | |
| Autologous stem cell transplant, No. (%) | ||||||
| Yes | 195 (20.2) | 4 (0.9) | 5 (2.0) | 186 (77.5) | <.001 | – |
| Total body irradiation (TBI), No. (%) | ||||||
| Yes | 77 (8.2) | 0 (0.0) | 0 (0.0) | 77 (34.1) | <.001 | – |
| TBI dose (Gy), median, range | 12 (2-13.8) | 0 (0.0) | 0 (0.0) | 12 (2.0-13.8) | <.001 | – |
| Abdominal radiation, N (%) | ||||||
| Yes | 207 (22.0) | 10 (2.4) | 31 (12.5) | 156 (69.0) | <.001 | – |
| Abdomen radiation dose (Gy), median, range | 20.0 (2.0-51.0) | 22.5 (4.5-46.0) | 12.0 (3.0-51.0) | 21.0 (2.0-51.0) | .02 | – |
| Cranial radiation, No. (%) | ||||||
| Yes | 111 (11.8) | 1 (0.2) | 4 (1.6) | 100 (44.2) | <.001 | – |
| Cranial radiation dose (Gy), median, range | 12.0 (2.0-66.0) | 21 (21.0-21.0) | 13.5 (6.0-15.0) | 12 (2.0-48.0) | .54 | - |
| Chest radiation, No. (%) | ||||||
| Yes | 126 (13.4) | 6 (1.4) | 4 (1.6) | 111 (49.1) | <.001 | – |
| Chest radiation dose, (Gy), median, range | 12 (2.0-45.0) | 22.5 (3.0-27.0) | 13.2 (4.5-22.0) | 12 (2.0-45.0) | .32 | – |
| Alkylating agents, No. (%) | ||||||
| Yes | 598 (61.8) | 80 (18.9) | 239 (94.8) | 238 (98.8) | <.001 | – |
| Cyclophosphamide equivalent dose | 6467.6 (2.2-68 944.4) | 5311.5 (178.3-43 742.1) | 5131.2 (2.2-59 939.1) | 13 146.2 (1050.7-68 944.4) | <.001 | – |
| Anthracyclines, No. (%) | ||||||
| Yes | 536 (55.3) | 69 (16.3) | 209 (82.9) | 229 (94.6) | <.001 | – |
| Anthracycline dose, mg/m2, median, range | 147.7 (0.0-750.8) | 150.5 (0.0-515.7) | 141.2 (0.3-507.1) | 151.5 (0.1-685.1) | .01 | – |
| Platinum agents, No. (%) | ||||||
| Yes | 501 (51.6) | 44 (10.4) | 187 (74.2) | 241 (99.6) | <.001 | – |
| Platinum dose, mg/m2, median, range | 497.6 (0.6-3,203.4) | 439.4 (35.2-1251.9) | 429.9 (0.6-3203.4) | 621.4 (90.9-2994.4) | <.001 | – |
| Epipodophyllotoxins, No. (%) | ||||||
| Yes | 490 (50.7) | 42 (9.9) | 186 (73.8) | 236 (98.7) | <.001 | – |
| Epipodophyllotoxin dose, mg/m2, median, range | 1721.4 (0.5-14 305.2) | 1223.3 (0.5-5567.1) | 1421.5 (43.6-8334.1) | 2050.8 (1.9-14 305.2) | <.001 | – |
| Isotretinoin, No. (%) | ||||||
| Yes | 76 (7.9) | 5 (1.2) | 7 (2.8) | 63 (26.7) | <.001 | – |
| Years of follow-up from diagnosis, median, range | 19.3 (5.0-29.9) | 19.4 (5.0-29.7) | 20.2 (6.5-29.2) | 17.9 (5.1-29.2) | <.001 | |
| Age at last contact, y, median, range | 21 (7-42) | 21 (7-42) | 21 (9-30) | 20 (7-38) | .31 | 28 (7-54) |
| Vital status, No. (%) | 1018 (98.9) | |||||
| Alive | 906 (90.7) | 409 (96.2) | 244 (96.8) | 184 (75.1) | <.001 | 1018 (98.9) |
| Dead | 93 (9.3) | 16 (3.8) | 8 (3.2) | 61 (24.9) | 11 (1.1) |
| Characteristic | All neuroblastoma survivors (N = 999) | Survivors with low-risk disease (n = 425) | Survivors with intermediate-risk disease (n = 252) | Survivors with high-risk disease (n = 245) | P valuea | Siblings (N = 1029) |
|---|---|---|---|---|---|---|
| Sex, No. (%) | ||||||
| Male | 498 (49.8) | 199 (46.8) | 127 (50.4) | 134 (54.7) | .14 | 471 (45.8) |
| Female | 501 (50.2) | 226 (53.2) | 125 (49.6) | 111 (45.3) | 558 (54.2) | |
| Race/ethnicity, No. (%) | ||||||
| Hispanic | 104 (10.4) | 53 (12.5) | 26 (10.3) | 17 (6.9) | 66 (6.6) | |
| Non-Hispanic, Black | 68 (6.8) | 20 (4.7) | 14 (5.6) | 22 (9.0) | 37 (3.7) | |
| Non-Hispanic, White | 785 (78.6) | 335 (78.8) | 205 (81.3) | 190 (77.6) | .03 | 864 (86.9) |
| Other | 42 (4.2) | 17 (4.0) | 7 (2.8) | 16 (6.5) | 27 (2.7) | |
| Age at diagnosis, years, No. (%) | ||||||
| <1 | 469 (46.9) | 229 (53.9) | 177 (70.2) | 16 (6.5) | <.001 | – |
| 1-4 | 422 (42.2) | 154 (36.2) | 69 (27.4) | 175 (71.4) | – | |
| 5-9 | 82 (8.2) | 29 (6.8) | 6 (2.4) | 44 (18.0) | – | |
| ≥10 | 26 (2.6) | 13 (3.1) | 0 (0.0) | 10 (4.1) | – | |
| Treatment category, No. (%) | ||||||
| Surgery only | 351 (36.6) | 337 (80.1) | 8 (3.2) | 0 (0.0) | <.001 | – |
| Chemotherapy ± surgery | 343 (35.8) | 67 (15.9) | 200 (79.4) | 50 (20.9) | – | |
| Radiation ± chemotherapy ± surgery | 265 (27.6) | 17 (4.0) | 44 (17.5) | 189 (79.1) | – | |
| Missing | 40 | 4 | 0 | 6 | – | |
| Autologous stem cell transplant, No. (%) | ||||||
| Yes | 195 (20.2) | 4 (0.9) | 5 (2.0) | 186 (77.5) | <.001 | – |
| Total body irradiation (TBI), No. (%) | ||||||
| Yes | 77 (8.2) | 0 (0.0) | 0 (0.0) | 77 (34.1) | <.001 | – |
| TBI dose (Gy), median, range | 12 (2-13.8) | 0 (0.0) | 0 (0.0) | 12 (2.0-13.8) | <.001 | – |
| Abdominal radiation, N (%) | ||||||
| Yes | 207 (22.0) | 10 (2.4) | 31 (12.5) | 156 (69.0) | <.001 | – |
| Abdomen radiation dose (Gy), median, range | 20.0 (2.0-51.0) | 22.5 (4.5-46.0) | 12.0 (3.0-51.0) | 21.0 (2.0-51.0) | .02 | – |
| Cranial radiation, No. (%) | ||||||
| Yes | 111 (11.8) | 1 (0.2) | 4 (1.6) | 100 (44.2) | <.001 | – |
| Cranial radiation dose (Gy), median, range | 12.0 (2.0-66.0) | 21 (21.0-21.0) | 13.5 (6.0-15.0) | 12 (2.0-48.0) | .54 | - |
| Chest radiation, No. (%) | ||||||
| Yes | 126 (13.4) | 6 (1.4) | 4 (1.6) | 111 (49.1) | <.001 | – |
| Chest radiation dose, (Gy), median, range | 12 (2.0-45.0) | 22.5 (3.0-27.0) | 13.2 (4.5-22.0) | 12 (2.0-45.0) | .32 | – |
| Alkylating agents, No. (%) | ||||||
| Yes | 598 (61.8) | 80 (18.9) | 239 (94.8) | 238 (98.8) | <.001 | – |
| Cyclophosphamide equivalent dose | 6467.6 (2.2-68 944.4) | 5311.5 (178.3-43 742.1) | 5131.2 (2.2-59 939.1) | 13 146.2 (1050.7-68 944.4) | <.001 | – |
| Anthracyclines, No. (%) | ||||||
| Yes | 536 (55.3) | 69 (16.3) | 209 (82.9) | 229 (94.6) | <.001 | – |
| Anthracycline dose, mg/m2, median, range | 147.7 (0.0-750.8) | 150.5 (0.0-515.7) | 141.2 (0.3-507.1) | 151.5 (0.1-685.1) | .01 | – |
| Platinum agents, No. (%) | ||||||
| Yes | 501 (51.6) | 44 (10.4) | 187 (74.2) | 241 (99.6) | <.001 | – |
| Platinum dose, mg/m2, median, range | 497.6 (0.6-3,203.4) | 439.4 (35.2-1251.9) | 429.9 (0.6-3203.4) | 621.4 (90.9-2994.4) | <.001 | – |
| Epipodophyllotoxins, No. (%) | ||||||
| Yes | 490 (50.7) | 42 (9.9) | 186 (73.8) | 236 (98.7) | <.001 | – |
| Epipodophyllotoxin dose, mg/m2, median, range | 1721.4 (0.5-14 305.2) | 1223.3 (0.5-5567.1) | 1421.5 (43.6-8334.1) | 2050.8 (1.9-14 305.2) | <.001 | – |
| Isotretinoin, No. (%) | ||||||
| Yes | 76 (7.9) | 5 (1.2) | 7 (2.8) | 63 (26.7) | <.001 | – |
| Years of follow-up from diagnosis, median, range | 19.3 (5.0-29.9) | 19.4 (5.0-29.7) | 20.2 (6.5-29.2) | 17.9 (5.1-29.2) | <.001 | |
| Age at last contact, y, median, range | 21 (7-42) | 21 (7-42) | 21 (9-30) | 20 (7-38) | .31 | 28 (7-54) |
| Vital status, No. (%) | 1018 (98.9) | |||||
| Alive | 906 (90.7) | 409 (96.2) | 244 (96.8) | 184 (75.1) | <.001 | 1018 (98.9) |
| Dead | 93 (9.3) | 16 (3.8) | 8 (3.2) | 61 (24.9) | 11 (1.1) |
Treatment data are shown among those with medical record abstraction; 77 had missing treatment data and were unable to be classified into a disease risk group.
χ2, Fisher exact, or Kruskal-Wallis test for test of association with risk.
Late mortality
Among 1452 survivors included in the mortality analysis (Supplementary Table 2, available online), the cumulative incidence of death at 25 years after neuroblastoma diagnosis was 8.3% (95% CI = 6.8 to 9.8). The 25-year cumulative incidence of mortality differed significantly among individuals with a history of high-risk disease (22.6%, 95% CI = 17.7 to 27.5) when compared with individuals with low (2.7%, 95% CI = 1.3 to 4.1) or intermediate-risk (3.0%, 95% CI = 1.0 to 5.1) disease (Figure 1, A).
Cumulative incidence of A) mortality, B) subsequent malignant neoplasm, C) recurrence, D) grade 3-5 chronic health conditions among survivors of neuroblastoma by disease risk group at 25 years after neuroblastoma diagnosis.
The SMR and EAR of death at 25 years after diagnosis were 9.5 (95% CI = 7.8 to 11.5) and 40.5 (95% CI = 32.6 to 50.0), respectively. SMR and EAR of death were greatest for individuals with high-risk disease (SMR = 27.7, 95% CI = 21.4 to 35.8; EAR = 137.8, 95% CI = 105.5 to 179.7), although they remained significantly increased for individuals with low- and intermediate-risk disease as well. SMRs and EARs were increased for all ages of diagnosis and for both males and females (Table 2).
Overall and cause-specific standardized mortality ratios (SMR) and excess absolute risk (EAR) among survivors of neuroblastoma enrolled in the Childhood Cancer Survivor Study at 25 years after diagnosis
| Observed | Expected | SMR (95% CI) | EAR (95% CI)/10 000 person-years | |
|---|---|---|---|---|
| All deathsa | 111 | 11.691 | 9.5 (7.8 to 11.5) | 40.5 (32.6 to 50.0) |
| Sex | ||||
| Male | 62 | 8.168 | 7.6 (5.9 to 9.8) | 42.2 (31.3 to 56.3) |
| Female | 49 | 3.523 | 13.9 (10.4 to 18.5) | 38.7 (28.3 to 52.5) |
| Risk | ||||
| Low risk | 14 | 4.952 | 2.8 (1.7 to 4.8) | 8.8 (3.2 to 18.2) |
| Intermediate risk | 9 | 2.694 | 3.3 (1.7 to 6.5) | 10.4 (3.2 to 24.3) |
| High risk | 69 | 2.493 | 27.7 (21.4 to 35.8) | 137.8 (105.5 to 179.7) |
| Missing risk | 19 | |||
| Age at diagnosis (years) | ||||
| <1 | 12 | 5.071 | 2.4 (1.3 to 4.2) | 5.6 (1.4 to 13.1) |
| 1-4 | 54 | 4.945 | 10.9 (8.3 to 14.4) | 49.1 (36.0 to 66.4) |
| 5-9 | 33 | 1.139 | 29.0 (19.9 to 42.3) | 199.5 (134.5 to 294.3) |
| ≥10 | 12 | 0.536 | 22.4 (11.9 to 42.0) | 194.1 (99.2 to 372.3) |
| Health-related cause of deathb | 35 | 3.596 | 9.7 (7.0 to 13.6) | 1.3 (0.9 to 1.8) |
| Subsequent malignancy | 11 | 0.832 | 13.2 (7.3 to 23.9) | 0.4 (0.2 to 0.8) |
| Cardiac | 2 | 0.450 | 4.4 (1.1 to 17.8) | 0.1 (0 to 0.3) |
| Pulmonary | 7 | 0.272 | 25.7 (12.3 to 53.9) | 0.3 (0.1 to 0.6) |
| Other heath conditions | 15 | 2.042 | 7.3 (4.4 to 12.2) | 0.5 (0.3 to 0.9) |
| External causes | 6 | 8.095 | 0.7 (0.3 to 1.6) | −0.1 (−0.2 to 0.2) |
| Unknown causes | 5 | – | – | – |
| Observed | Expected | SMR (95% CI) | EAR (95% CI)/10 000 person-years | |
|---|---|---|---|---|
| All deathsa | 111 | 11.691 | 9.5 (7.8 to 11.5) | 40.5 (32.6 to 50.0) |
| Sex | ||||
| Male | 62 | 8.168 | 7.6 (5.9 to 9.8) | 42.2 (31.3 to 56.3) |
| Female | 49 | 3.523 | 13.9 (10.4 to 18.5) | 38.7 (28.3 to 52.5) |
| Risk | ||||
| Low risk | 14 | 4.952 | 2.8 (1.7 to 4.8) | 8.8 (3.2 to 18.2) |
| Intermediate risk | 9 | 2.694 | 3.3 (1.7 to 6.5) | 10.4 (3.2 to 24.3) |
| High risk | 69 | 2.493 | 27.7 (21.4 to 35.8) | 137.8 (105.5 to 179.7) |
| Missing risk | 19 | |||
| Age at diagnosis (years) | ||||
| <1 | 12 | 5.071 | 2.4 (1.3 to 4.2) | 5.6 (1.4 to 13.1) |
| 1-4 | 54 | 4.945 | 10.9 (8.3 to 14.4) | 49.1 (36.0 to 66.4) |
| 5-9 | 33 | 1.139 | 29.0 (19.9 to 42.3) | 199.5 (134.5 to 294.3) |
| ≥10 | 12 | 0.536 | 22.4 (11.9 to 42.0) | 194.1 (99.2 to 372.3) |
| Health-related cause of deathb | 35 | 3.596 | 9.7 (7.0 to 13.6) | 1.3 (0.9 to 1.8) |
| Subsequent malignancy | 11 | 0.832 | 13.2 (7.3 to 23.9) | 0.4 (0.2 to 0.8) |
| Cardiac | 2 | 0.450 | 4.4 (1.1 to 17.8) | 0.1 (0 to 0.3) |
| Pulmonary | 7 | 0.272 | 25.7 (12.3 to 53.9) | 0.3 (0.1 to 0.6) |
| Other heath conditions | 15 | 2.042 | 7.3 (4.4 to 12.2) | 0.5 (0.3 to 0.9) |
| External causes | 6 | 8.095 | 0.7 (0.3 to 1.6) | −0.1 (−0.2 to 0.2) |
| Unknown causes | 5 | – | – | – |
All deaths include 65 deaths because of recurrence (n = 3 low risk, n = 5 intermediate risk, n = 50 high risk, n = 7 missing risk). CI = confidence interval.
Health-related causes of death do not include deaths because of recurrence.
Overall and cause-specific standardized mortality ratios (SMR) and excess absolute risk (EAR) among survivors of neuroblastoma enrolled in the Childhood Cancer Survivor Study at 25 years after diagnosis
| Observed | Expected | SMR (95% CI) | EAR (95% CI)/10 000 person-years | |
|---|---|---|---|---|
| All deathsa | 111 | 11.691 | 9.5 (7.8 to 11.5) | 40.5 (32.6 to 50.0) |
| Sex | ||||
| Male | 62 | 8.168 | 7.6 (5.9 to 9.8) | 42.2 (31.3 to 56.3) |
| Female | 49 | 3.523 | 13.9 (10.4 to 18.5) | 38.7 (28.3 to 52.5) |
| Risk | ||||
| Low risk | 14 | 4.952 | 2.8 (1.7 to 4.8) | 8.8 (3.2 to 18.2) |
| Intermediate risk | 9 | 2.694 | 3.3 (1.7 to 6.5) | 10.4 (3.2 to 24.3) |
| High risk | 69 | 2.493 | 27.7 (21.4 to 35.8) | 137.8 (105.5 to 179.7) |
| Missing risk | 19 | |||
| Age at diagnosis (years) | ||||
| <1 | 12 | 5.071 | 2.4 (1.3 to 4.2) | 5.6 (1.4 to 13.1) |
| 1-4 | 54 | 4.945 | 10.9 (8.3 to 14.4) | 49.1 (36.0 to 66.4) |
| 5-9 | 33 | 1.139 | 29.0 (19.9 to 42.3) | 199.5 (134.5 to 294.3) |
| ≥10 | 12 | 0.536 | 22.4 (11.9 to 42.0) | 194.1 (99.2 to 372.3) |
| Health-related cause of deathb | 35 | 3.596 | 9.7 (7.0 to 13.6) | 1.3 (0.9 to 1.8) |
| Subsequent malignancy | 11 | 0.832 | 13.2 (7.3 to 23.9) | 0.4 (0.2 to 0.8) |
| Cardiac | 2 | 0.450 | 4.4 (1.1 to 17.8) | 0.1 (0 to 0.3) |
| Pulmonary | 7 | 0.272 | 25.7 (12.3 to 53.9) | 0.3 (0.1 to 0.6) |
| Other heath conditions | 15 | 2.042 | 7.3 (4.4 to 12.2) | 0.5 (0.3 to 0.9) |
| External causes | 6 | 8.095 | 0.7 (0.3 to 1.6) | −0.1 (−0.2 to 0.2) |
| Unknown causes | 5 | – | – | – |
| Observed | Expected | SMR (95% CI) | EAR (95% CI)/10 000 person-years | |
|---|---|---|---|---|
| All deathsa | 111 | 11.691 | 9.5 (7.8 to 11.5) | 40.5 (32.6 to 50.0) |
| Sex | ||||
| Male | 62 | 8.168 | 7.6 (5.9 to 9.8) | 42.2 (31.3 to 56.3) |
| Female | 49 | 3.523 | 13.9 (10.4 to 18.5) | 38.7 (28.3 to 52.5) |
| Risk | ||||
| Low risk | 14 | 4.952 | 2.8 (1.7 to 4.8) | 8.8 (3.2 to 18.2) |
| Intermediate risk | 9 | 2.694 | 3.3 (1.7 to 6.5) | 10.4 (3.2 to 24.3) |
| High risk | 69 | 2.493 | 27.7 (21.4 to 35.8) | 137.8 (105.5 to 179.7) |
| Missing risk | 19 | |||
| Age at diagnosis (years) | ||||
| <1 | 12 | 5.071 | 2.4 (1.3 to 4.2) | 5.6 (1.4 to 13.1) |
| 1-4 | 54 | 4.945 | 10.9 (8.3 to 14.4) | 49.1 (36.0 to 66.4) |
| 5-9 | 33 | 1.139 | 29.0 (19.9 to 42.3) | 199.5 (134.5 to 294.3) |
| ≥10 | 12 | 0.536 | 22.4 (11.9 to 42.0) | 194.1 (99.2 to 372.3) |
| Health-related cause of deathb | 35 | 3.596 | 9.7 (7.0 to 13.6) | 1.3 (0.9 to 1.8) |
| Subsequent malignancy | 11 | 0.832 | 13.2 (7.3 to 23.9) | 0.4 (0.2 to 0.8) |
| Cardiac | 2 | 0.450 | 4.4 (1.1 to 17.8) | 0.1 (0 to 0.3) |
| Pulmonary | 7 | 0.272 | 25.7 (12.3 to 53.9) | 0.3 (0.1 to 0.6) |
| Other heath conditions | 15 | 2.042 | 7.3 (4.4 to 12.2) | 0.5 (0.3 to 0.9) |
| External causes | 6 | 8.095 | 0.7 (0.3 to 1.6) | −0.1 (−0.2 to 0.2) |
| Unknown causes | 5 | – | – | – |
All deaths include 65 deaths because of recurrence (n = 3 low risk, n = 5 intermediate risk, n = 50 high risk, n = 7 missing risk). CI = confidence interval.
Health-related causes of death do not include deaths because of recurrence.
More than half of late deaths (65/120) in the cohort were due to late recurrence, with most recurrence-related deaths occurring in individuals with high-risk disease (n = 50). Other nonrelapse/nonrecurrence health-related causes of death included SMN (SMR = 13.2, 95% CI = 7.3 to 23.9; EAR = 0.4, 95% CI = 0.2 to 0.8); pulmonary causes (SMR = 25.7, 95% CI = 12.3 to 53.9; EAR = 0.3, 95% CI = 0.1 to 0.6); and other health conditions (SMR = 7.3, 95% CI = 4.4 to 12.2; EAR = 0.5, 95% CI = 0.3 to 0.9). SMR because of cardiac causes was elevated, but the EAR of death because of cardiac etiologies was not (Table 2).
Late recurrence and SMN
The 25-year cumulative incidence of developing a late recurrence was 7.2% (95% CI = 5.6 to 8.9), whereas the cumulative incidence of developing an SMN was 3.7% (95% CI = 2.2 to 5.1), accounting for competing risk of death. When analyzed by risk group, the cumulative incidence of SMN was highest in individuals with high-risk disease (11.4%, 95% CI = 6.5 to 16.2) and much lower in those with intermediate- or low-risk disease (Figure 1, B). The 25-year cumulative incidence of late recurrence in survivors of high-risk disease was 18.5% (95% CI = 13.3 to 23.8), 3.1% (95% CI = 1.4 to 4.7) for survivors of intermediate-risk disease, and not significantly elevated for survivors of low-risk disease (Figure 1, C). Only 1 survivor had a recurrence prior to SMN and was alive at the time of last data analysis. Another survivor had an SMN and then experienced a subsequent relapse and died. Otherwise, the other survivors who had SMNs only had SMNs without relapse/recurrence.
Compared with age-, sex-, and calendar-year-matched population-based controls, survivors had a significantly increased risk of developing an SMN (SIR = 8.1, 95% CI = 5.4 to 12.1), corresponding to 16.1 excess cases per 10 000 person-years (95% CI = 10.0 to 25.0), which was largely driven by individuals with high-risk disease (Table 3). In fact, when analyzed by risk group, EAR for SMN was markedly elevated for those with high-risk disease (64.7, 95% CI = 42.0 to 99.0), moderately elevated for individuals with intermediate-risk disease (5.4, 95% CI = 0.4 to 20.9), and did not differ from the general population among those with low-risk disease. SIRs and EARs were elevated for all age groups except those diagnosed at younger than 1 year of age. SIRs and EARs for SMN subtypes are shown in Table 3.
Standardized incidence ratio (SIR) (95% CI) and excess absolute risk (EAR) (95% CI) for subsequent malignant neoplasms (SMN) at 25 years after primary cancer diagnosis
| Observed | Expected | SIR (95% CI) | EAR (95% CI) per 10 000 person-years | |
|---|---|---|---|---|
| All SMNs | 28 | 3.451 | 8.1 (5.4 to 12.1) | 16.1 (10.0 to 25.0) |
| Sex | ||||
| Male | 13 | 1.559 | 8.3 (4.9 to 14.3) | 15.3 (8.1 to 27.7) |
| Female | 15 | 1.892 | 7.9 (4.4 to 14.2) | 16.8 (8.3 to 31.8) |
| Risk | ||||
| Low risk | 2 | 1.552 | 1.3 (0.2 to 9.1) | 0.7 (−1.9 to 18.7) |
| Intermediate risk | 3 | 0.815 | 3.7 (1.2 to 11.3) | 5.4 (0.4 to 20.9) |
| High risk | 22 | 0.787 | 28.0 (18.5 to 42.3) | 64.7 (42 to 99.0) |
| Age at diagnosis | ||||
| <1 | 3 | 1.430 | 2.1 (0.7 to 6.5) | 2.1 (−0.6 to 10.4) |
| 1-4 | 15 | 1.431 | 10.5 (6.2 to 17.8) | 21.3 (11.6 to 37.6) |
| 5-9 | 7 | 0.366 | 19.1 (8.5 to 43.2) | 62.1 (25.6 to 144.6) |
| ≥10 | 3 | 0.224 | 13.4 (3.6 to 50.3) | 87.7 (18.2 to 347.9) |
| SMN subtype | ||||
| Thyroid | 9 | 0.437 | 20.6 (9.5 to 44.5) | 5.6 (2.4 to 12.4) |
| Sarcoma | 7 | 0.367 | 19.1 (8.3 to 44.1) | 4.3 (1.7 to 10.4) |
| Acute myeloid leukemia | 2 | 0.090 | 22.3 (5.6 to 89.1) | 1.3 (0.3 to 5.2) |
| Other glial | 1 | 0.088 | 11.3 (1.6 to 80.3) | 0.6 (0 to 4.6) |
| Breast | 1 | 0.109 | 9.1 (1.3 to 65.7) | 0.6 (0 to 4.6) |
| Non-Hodgkin lymphoma | 1 | 0.232 | 4.3 (0.6 to 30.6) | 0.5 (−0.1 to 4.5) |
| Small intestine and colorectal | 1 | 0.094 | 10.6 (1.5 to 75.3) | 0.6 (0.0 to 4.6) |
| Other cancersa | 6 | 0.559 | 10.7 (4.8 to 23.9) | 3.6 (1.4 to 8.4) |
| Observed | Expected | SIR (95% CI) | EAR (95% CI) per 10 000 person-years | |
|---|---|---|---|---|
| All SMNs | 28 | 3.451 | 8.1 (5.4 to 12.1) | 16.1 (10.0 to 25.0) |
| Sex | ||||
| Male | 13 | 1.559 | 8.3 (4.9 to 14.3) | 15.3 (8.1 to 27.7) |
| Female | 15 | 1.892 | 7.9 (4.4 to 14.2) | 16.8 (8.3 to 31.8) |
| Risk | ||||
| Low risk | 2 | 1.552 | 1.3 (0.2 to 9.1) | 0.7 (−1.9 to 18.7) |
| Intermediate risk | 3 | 0.815 | 3.7 (1.2 to 11.3) | 5.4 (0.4 to 20.9) |
| High risk | 22 | 0.787 | 28.0 (18.5 to 42.3) | 64.7 (42 to 99.0) |
| Age at diagnosis | ||||
| <1 | 3 | 1.430 | 2.1 (0.7 to 6.5) | 2.1 (−0.6 to 10.4) |
| 1-4 | 15 | 1.431 | 10.5 (6.2 to 17.8) | 21.3 (11.6 to 37.6) |
| 5-9 | 7 | 0.366 | 19.1 (8.5 to 43.2) | 62.1 (25.6 to 144.6) |
| ≥10 | 3 | 0.224 | 13.4 (3.6 to 50.3) | 87.7 (18.2 to 347.9) |
| SMN subtype | ||||
| Thyroid | 9 | 0.437 | 20.6 (9.5 to 44.5) | 5.6 (2.4 to 12.4) |
| Sarcoma | 7 | 0.367 | 19.1 (8.3 to 44.1) | 4.3 (1.7 to 10.4) |
| Acute myeloid leukemia | 2 | 0.090 | 22.3 (5.6 to 89.1) | 1.3 (0.3 to 5.2) |
| Other glial | 1 | 0.088 | 11.3 (1.6 to 80.3) | 0.6 (0 to 4.6) |
| Breast | 1 | 0.109 | 9.1 (1.3 to 65.7) | 0.6 (0 to 4.6) |
| Non-Hodgkin lymphoma | 1 | 0.232 | 4.3 (0.6 to 30.6) | 0.5 (−0.1 to 4.5) |
| Small intestine and colorectal | 1 | 0.094 | 10.6 (1.5 to 75.3) | 0.6 (0.0 to 4.6) |
| Other cancersa | 6 | 0.559 | 10.7 (4.8 to 23.9) | 3.6 (1.4 to 8.4) |
Other cancers include Pseudosarcomatous carcinoma; sebaceous adenocarcinoma of the skin; eccrine adenocarcinoma; precursor T-cell lymphoblastic lymphoma. CI = confidence interval.
Standardized incidence ratio (SIR) (95% CI) and excess absolute risk (EAR) (95% CI) for subsequent malignant neoplasms (SMN) at 25 years after primary cancer diagnosis
| Observed | Expected | SIR (95% CI) | EAR (95% CI) per 10 000 person-years | |
|---|---|---|---|---|
| All SMNs | 28 | 3.451 | 8.1 (5.4 to 12.1) | 16.1 (10.0 to 25.0) |
| Sex | ||||
| Male | 13 | 1.559 | 8.3 (4.9 to 14.3) | 15.3 (8.1 to 27.7) |
| Female | 15 | 1.892 | 7.9 (4.4 to 14.2) | 16.8 (8.3 to 31.8) |
| Risk | ||||
| Low risk | 2 | 1.552 | 1.3 (0.2 to 9.1) | 0.7 (−1.9 to 18.7) |
| Intermediate risk | 3 | 0.815 | 3.7 (1.2 to 11.3) | 5.4 (0.4 to 20.9) |
| High risk | 22 | 0.787 | 28.0 (18.5 to 42.3) | 64.7 (42 to 99.0) |
| Age at diagnosis | ||||
| <1 | 3 | 1.430 | 2.1 (0.7 to 6.5) | 2.1 (−0.6 to 10.4) |
| 1-4 | 15 | 1.431 | 10.5 (6.2 to 17.8) | 21.3 (11.6 to 37.6) |
| 5-9 | 7 | 0.366 | 19.1 (8.5 to 43.2) | 62.1 (25.6 to 144.6) |
| ≥10 | 3 | 0.224 | 13.4 (3.6 to 50.3) | 87.7 (18.2 to 347.9) |
| SMN subtype | ||||
| Thyroid | 9 | 0.437 | 20.6 (9.5 to 44.5) | 5.6 (2.4 to 12.4) |
| Sarcoma | 7 | 0.367 | 19.1 (8.3 to 44.1) | 4.3 (1.7 to 10.4) |
| Acute myeloid leukemia | 2 | 0.090 | 22.3 (5.6 to 89.1) | 1.3 (0.3 to 5.2) |
| Other glial | 1 | 0.088 | 11.3 (1.6 to 80.3) | 0.6 (0 to 4.6) |
| Breast | 1 | 0.109 | 9.1 (1.3 to 65.7) | 0.6 (0 to 4.6) |
| Non-Hodgkin lymphoma | 1 | 0.232 | 4.3 (0.6 to 30.6) | 0.5 (−0.1 to 4.5) |
| Small intestine and colorectal | 1 | 0.094 | 10.6 (1.5 to 75.3) | 0.6 (0.0 to 4.6) |
| Other cancersa | 6 | 0.559 | 10.7 (4.8 to 23.9) | 3.6 (1.4 to 8.4) |
| Observed | Expected | SIR (95% CI) | EAR (95% CI) per 10 000 person-years | |
|---|---|---|---|---|
| All SMNs | 28 | 3.451 | 8.1 (5.4 to 12.1) | 16.1 (10.0 to 25.0) |
| Sex | ||||
| Male | 13 | 1.559 | 8.3 (4.9 to 14.3) | 15.3 (8.1 to 27.7) |
| Female | 15 | 1.892 | 7.9 (4.4 to 14.2) | 16.8 (8.3 to 31.8) |
| Risk | ||||
| Low risk | 2 | 1.552 | 1.3 (0.2 to 9.1) | 0.7 (−1.9 to 18.7) |
| Intermediate risk | 3 | 0.815 | 3.7 (1.2 to 11.3) | 5.4 (0.4 to 20.9) |
| High risk | 22 | 0.787 | 28.0 (18.5 to 42.3) | 64.7 (42 to 99.0) |
| Age at diagnosis | ||||
| <1 | 3 | 1.430 | 2.1 (0.7 to 6.5) | 2.1 (−0.6 to 10.4) |
| 1-4 | 15 | 1.431 | 10.5 (6.2 to 17.8) | 21.3 (11.6 to 37.6) |
| 5-9 | 7 | 0.366 | 19.1 (8.5 to 43.2) | 62.1 (25.6 to 144.6) |
| ≥10 | 3 | 0.224 | 13.4 (3.6 to 50.3) | 87.7 (18.2 to 347.9) |
| SMN subtype | ||||
| Thyroid | 9 | 0.437 | 20.6 (9.5 to 44.5) | 5.6 (2.4 to 12.4) |
| Sarcoma | 7 | 0.367 | 19.1 (8.3 to 44.1) | 4.3 (1.7 to 10.4) |
| Acute myeloid leukemia | 2 | 0.090 | 22.3 (5.6 to 89.1) | 1.3 (0.3 to 5.2) |
| Other glial | 1 | 0.088 | 11.3 (1.6 to 80.3) | 0.6 (0 to 4.6) |
| Breast | 1 | 0.109 | 9.1 (1.3 to 65.7) | 0.6 (0 to 4.6) |
| Non-Hodgkin lymphoma | 1 | 0.232 | 4.3 (0.6 to 30.6) | 0.5 (−0.1 to 4.5) |
| Small intestine and colorectal | 1 | 0.094 | 10.6 (1.5 to 75.3) | 0.6 (0.0 to 4.6) |
| Other cancersa | 6 | 0.559 | 10.7 (4.8 to 23.9) | 3.6 (1.4 to 8.4) |
Other cancers include Pseudosarcomatous carcinoma; sebaceous adenocarcinoma of the skin; eccrine adenocarcinoma; precursor T-cell lymphoblastic lymphoma. CI = confidence interval.
CHCs
The 25-year cumulative incidence of grade 3-5 CHCs among survivors by risk group is shown in Figure 1, D. When compared with siblings, survivors had a 6.4-fold increased risk of developing a grade 3-5 CHC (95% CI = 4.5 to 9.1, P < .001), after adjusting for sex and race/ethnicity. Consistent with other outcomes, the highest risks of grade 3-5 CHCs were noted among high-risk patients (Grade 3-5 CHCs: HRhigh = 16.1, 95% CI = 11.2 to 23.2; P < .001; HRintermediate = 6.3, 95% CI = 3.8 to 10.5; P < .001; HRlow = 1.8, 95% CI = 1.1 to 3.1; P = .02) (Figure 2). Similar patterns of risk were noted for developing grade 3-5 endocrinopathies and cardiac conditions among survivors relative to siblings (Figure 2). Interestingly, no increased risk was noted for grade 3-5 respiratory or gastrointestinal CHCs, either overall or by risk group, relative to siblings. Curiously, risk of grade 3-5 neurologic CHCs was increased among survivors of intermediate-risk disease but not among those with low- or high-risk disease.
Risk of grade (gr) 3-5 chronic health conditions among neuroblastoma survivors, overall and by disease risk group, compared with a sibling comparison group. CI = confidence interval; HR = hazard ratio.
Multivariable analyses of treatment exposures associated with specific CHCs in the overall cohort are shown in Supplementary Table 3 (available online).
Late morbidity and mortality in survivors of high-risk disease
Given the relative dearth of long-term outcomes data among survivors of high-risk disease, we assessed the risk of late outcomes in this subgroup by treatment category. In Cox models adjusted for sex and race/ethnicity, there was no difference in risk of all-cause or cause-specific mortality among high-risk patients treated with ASCT, whether TBI-based or chemotherapy-based, when compared with individuals treated without ASCT. However, those treated with TBI-based ASCT were 2.2-fold more likely to develop a grade 3-5 CHC (95% CI = 1.3 to 3.8). Individuals treated with chemotherapy-based ASCT were at 1.8-fold risk for a grade 3-5 CHC (95% CI = 1.0 to 3.5, P = .04), when compared with those treated without ASCT, after adjusting for sex and race/ethnicity (Supplementary Table 4, available online).
Discussion
In this comprehensive assessment of late adverse outcomes in neuroblastoma survivors treated in the early days of risk stratification, we found strikingly different risk profiles for late mortality, late recurrence, SMN, and grade 3-5 CHCs by primary disease risk group. Specifically, we showed that survivors of high-risk disease carry a markedly elevated burden of late recurrence, SMN, and organ-related multimorbidity, whereas survivors of low- or intermediate-risk disease have a modest risk of late adverse outcomes. In subgroup analyses limited to high-risk survivors, individuals treated with TBI-based or chemotherapy-based ASCT were not at increased risk of all-cause or cause-specific mortality but were at two-fold risk of developing grade 3-5 CHCs, relative to high-risk survivors treated without ASCT. Our findings suggest that intensification of therapy, which has been critical in improving cure rates for individuals with high-risk disease, has come with significant cost, which must be considered in future trial design. At the same time, reduction in therapy decreased the burden of late effects in survivors of low- and intermediate-risk disease.
Notably, efforts toward risk stratification in the treatment of neuroblastoma began in the 1980s when it first became evident that localized, low-risk neuroblastoma was treatable with surgery alone and therapy was accordingly deintensified (20,21). At the same time, treatment for high-risk disease was intensified with myeloablative therapy followed by autologous or allogeneic transplantation (22-25). Monoclonal antibody–based immunotherapy that targeted the cell-surface ganglioside GD2, with or without GM-CSF, was introduced during this decade as well (26-28). In the 1990s, treatment was increasingly tailored on the basis of tumor biology with further reduction in therapy for individuals with both low- and intermediate-risk disease (29,30) and intensification of multimodality therapy for high-risk patients. Improved outcomes were noted after trials of myeloablative chemotherapy and ASCT, total body irradiation, and isotretinoin in the mid-1990s (31).
Although survival outcomes have been reported after these significant therapeutic shifts (24,25,31-41), comprehensive data are lacking on the long-term burden of morbidity and mortality associated with these changes. Prior reports within the CCSS have explored late effects among neuroblastoma survivors diagnosed between 1970 and 1986 (42), before risk stratification efforts, or among those diagnosed with neuroblastoma during infancy (<12 months) (14), the majority of whom likely had low- or intermediate-risk disease. Recent studies from other cohorts have focused on SMN (43,44) or late effects among survivors of high-risk neuroblastoma, although the latter analyses have been limited by small cohort size and/or single-center design (45-49).
The current analysis fills this knowledge gap by elucidating the risk of late morbidity and mortality in a large, multicenter cohort of 5-year neuroblastoma survivors treated during the early era of risk stratification. Although late mortality in this cohort remained substantial (SMR = 9.5, EAR = 40.5), risk was particularly pronounced among high-risk patients (SMR = 27.7, EAR = 135.7). It is important to note that more than half of late deaths in the overall cohort occurred because of late recurrence (65/120, with 50 of these deaths occurring in those with high-risk disease), and not due to SMN or late toxicity, suggesting that disease biology was an important contributor to suboptimal late outcomes in this cohort (7,50). Novel therapies and strategies are still needed to ensure enduring and complete remission of the primary cancer (51), particularly among individuals with high-risk disease.
Still, our findings indicate that tailoring of treatment by disease risk group has resulted in significant improvement in late health outcomes for subgroups of neuroblastoma survivors. The 25-year cumulative incidence of SMN was low for survivors of low- and intermediate-risk disease; encouragingly, the SIR of SMN was comparable to the general population for those with low-risk disease. Similarly, a stepwise progression in risk of developing a grade 3-5 CHC was noted by risk group, with the lowest risks demonstrated among survivors of low-risk disease.
When considering SMN risk in this cohort, thyroid carcinoma was the most frequently identified SMN, which is consistent with other reports (52,53). It is not clear whether this risk was due to radiation exposure to the thyroid gland at a very young age (53), metaiodobenzylguanidine (MIBG) therapy (54,55), or a combination thereof, although MIBG utilization (for imaging or therapeutically) was not assessed in this cohort. Other frequent SMNs in this cohort included therapy-related myeloid neoplasms (t-MN) and sarcomas, which have also been reported in other cohorts of neuroblastoma survivors (56,57). It is likely that the true prevalence of t-MN, which are typically of short latency (58), was higher than reported here since events occurring before 5 years after diagnosis were not assessed.
With respect to CHCs, we identified an exceedingly high burden of endocrinopathies across risk groups with particularly elevated risks among those with a history of high-risk disease. This risk has also been demonstrated in other cohorts of high-risk neuroblastoma survivors (47,59,60). As expected in multivariable analysis, TBI exposure was associated with risk of developing grade 3-5 endocrinopathies. It is also noteworthy that even those in the intermediate-risk group, mostly treated with surgery and minimal chemotherapy, had an increase in risk of CHCs compared with siblings, although there is an overlap of the confidence intervals in several of the CHCs between risk groups.
We were surprised to find no significantly increased risk of respiratory CHCs across risk groups given existing evidence of pulmonary adverse outcomes from other cohorts of neuroblastoma survivors (61-63). Similarly, we expected to find an increased risk of gastrointestinal CHCs given high rates of thoracoabdominal resections and/or abdominal radiation, but we detected an increased risk only in high-risk survivors. It is likely that other reports of CHCs included conditions of mild-moderate severity (16), which were not captured in the current report.
Several limitations must be considered when interpreting these results. Importantly, tumor biology data and disease risk groups, which are critical for neuroblastoma risk stratification, such as the International Neuroblastoma Risk Group Staging System (6,64,65), were not available in the CCSS cohort; risk group was thus assigned using available data on age at diagnosis, disease stage, and treatment. We acknowledge that this approach may have led to misclassification of risk group for select patients. We used the participant’s first treatment protocol and disease stage at diagnosis for risk classification purposes, which occasionally led to a mismatch of treatment intensity and assigned risk group (eg, four patients with low-risk designation had an ASCT). In general, however, treatment intensity mapped to disease risk group, with high-risk group serving as a proxy for intensive, multimodality therapy. Importantly, risk stratification approaches have continued to evolve significantly since the year 2000, and the results presented herein may not be directly applicable to contemporarily treated patients. It is also important to note that although SMNs were validated with pathology review, CHCs were otherwise self-reported by participants.
Nonetheless, these data can help inform the design of future treatment protocols for children with neuroblastoma and guide the care of long-term neuroblastoma survivors who were treated in this era and continue to be followed longitudinally. Importantly, these data demonstrate that survivors with low- or intermediate-risk disease have benefited from risk stratification efforts with resultant lower risk of late adverse outcomes. Although these improvements remain elusive for survivors of high-risk disease, future therapeutic efforts, focused on the use of immunotherapy and other novel therapies (8), will hopefully optimize cure rates and sustained remission for high-risk survivors while minimizing risk for late effects. As survival rates continue to increase with these therapeutic modifications (51), close monitoring of survivors of all risk groups will be required for emerging late effects and CHCs.
Data availability
The CCSS is a publicly available data resource. Investigators can apply for specific analyses through a proposal process available on the website. The dataset specific to these analyses is not publicly available.
Author contributions
Danielle Novetsky Friedman, MD (Conceptualization; Investigation; Methodology; Writing—original draft; Writing—review & editing), Gregory Armstrong, MD (Funding acquisition; Investigation; Methodology; Project administration; Writing—review & editing), Kevin Oeffinger, MD (Investigation; Writing—review & editing), Leslie Robison, PhD (Investigation; Methodology; Project administration; Writing—review & editing), Brent Weil, MD (Investigation; Writing—review & editing), Lucie Turcotte, MD (Investigation; Writing—review & editing), Paul Nathan, MD, MSc (Investigation; Writing—review & editing), Todd Gibson, PhD (Investigation; Writing—review & editing), Kirsten Ness, PhD (Investigation; Writing—review & editing), Joseph Neglia, MD, MPH (Investigation; Writing—review & editing), Suzanne L. Wolden, MD (Investigation; Writing—review & editing), Emily Tonorezos, MD, MPH (Investigation; Writing—review & editing), Susan Smith, MA (Investigation; Writing—review & editing), Rebecca Howell, PhD (Investigation; Writing—review & editing), Sue Cohn, MD (Investigation; Writing—review & editing), Lisa Diller, MD (Conceptualization; Methodology; Supervision; Writing—review & editing), Wendy M. Leisenring, ScD (Data curation; Formal analysis; Investigation; Methodology; Supervision; Writing—review & editing), Pamela J. Goodman, MS (Data curation; Formal analysis; Methodology; Writing—original draft; Writing—review & editing), Charles Sklar, MD (Investigation; Writing—review & editing), Tara Henderson, MD (Conceptualization; Investigation; Methodology; Supervision; Writing—review & editing).
Funding
This study was supported by a grant (U24 CA-55727, G.T. Armstrong, Principal Investigator) from the US the National Cancer Institute, National Institute of Health, Bethesda, MD, USA. Support to St Jude Children’s Research Hospital, Memphis, TN, USA, was also provided by the Cancer Center Support (CORE) grant (CA-21765, C. Roberts, Principal Investigator) and the American Lebanese Syrian Associated Charities (ALSAC).
Conflicts of interest
GA and TH, who are JNCI Associate Editors and co-authors on this paper, were not involved in the editorial review or decision to publish the paper. The other authors have no conflict of interest to disclose.
Acknowledgments
The funders had no role in study design, data collection, data interpretation, or writing of the manuscript. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Preliminary findings from this analysis were presented at the American Society of Clinical Oncology Meeting in 2021 in Chicago, IL.
