Abstract
Children treated for cancer are at risk for neuromuscular dysfunction, but data are limited regarding prevalence, longitudinal patterns, and long-term impact.
Longitudinal surveys from 25,583 childhood cancer survivors ≥5 years from diagnosis and 5,044 siblings from the Childhood Cancer Survivor Study were used to estimate the prevalence and cumulative incidence of neuromuscular dysfunction. Multivariable models adjusted for age, sex, race, and ethnicity estimated prevalence ratios (PR) of neuromuscular dysfunction in survivors compared with siblings, and associations with treatments and late health/socioeconomic outcomes.
Prevalence of neuromuscular dysfunction was 14.7% in survivors 5 years postdiagnosis versus 1.5% in siblings [PR, 9.9; 95% confidence interval (CI), 7.9–12.4], and highest in survivors of central nervous system (CNS) tumors (PR, 27.6; 95% CI, 22.1–34.6) and sarcomas (PR, 11.5; 95% CI, 9.1–14.5). Cumulative incidence rose to 24.3% in survivors 20 years postdiagnosis (95% CI, 23.8–24.8). Spinal radiotherapy and increasing cranial radiotherapy dose were associated with increased prevalence of neuromuscular dysfunction. Platinum exposure (vs. none) was associated with neuromuscular dysfunction (PR, 1.8; 95% CI, 1.5–2.1), even after excluding survivors with CNS tumors, cranial/spinal radiotherapy, or amputation. Neuromuscular dysfunction was associated with concurrent or later obesity (PR, 1.1; 95% CI, 1.1–1.2), anxiety (PR, 2.5; 95% CI, 2.2–2.9), depression (PR, 2.1; 95% CI, 1.9–2.3), and lower likelihood of graduating college (PR, 0.92; 95% CI, 0.90–0.94) and employment (PR, 0.8; 95% CI, 0.8–0.9).
Neuromuscular dysfunction is prevalent in childhood cancer survivors, continues to increase posttherapy, and is associated with adverse health and socioeconomic outcomes.
Interventions are needed to prevent and treat neuromuscular dysfunction, especially in survivors with platinum and radiation exposure.
This article is featured in Highlights of This Issue, p. 1453
Introduction
Children with cancer in the United States have an estimated 5-year survival approaching 85% (1). However, survivors of childhood cancer often experience long-term treatment-related morbidity (2). Survivors are at risk for neuromuscular dysfunction, including weakness, impaired balance, and impaired sensation, due to the neurotoxic treatment they receive (3, 4). There is emerging evidence that neuromuscular dysfunction can persist for years posttherapy in childhood cancer survivors and impact their quality of life (3–16).
More research is needed to describe the longitudinal patterns and long-term impact of neuromuscular dysfunction that has been increasingly reported in survivors of all types of childhood cancer. Studies of neuromuscular dysfunction initially largely focused on survivors of acute lymphoblastic leukemia (ALL) due to their risk of vincristine-induced peripheral neuropathy, and found prevalence in this group to range from 25% to 100% in cross-sectional analyses (4, 9–12, 14). More recent studies have focused on survivors of childhood solid tumors given their exposures to neurotoxic chemotherapy, and found as many as 60% of these survivors experience neuromuscular dysfunction (3, 13). Few studies have evaluated neuromuscular dysfunction longitudinally, and these have been limited by small sample size and short duration of follow-up (3, 17, 18). Neuromuscular dysfunction results in impairment in critical skills necessary for school and work and can also interfere with social interactions in survivors of childhood cancer (8, 19, 20). The contribution of neuromuscular dysfunction to the increased risk for anxiety, depression, unemployment, and lower educational attainment in childhood cancer survivors should therefore be further explored (13, 21–24).
Longitudinal assessment of neuromuscular dysfunction along with other health and socioeconomic outcomes is needed to quantify the burden of neuromuscular dysfunction and guide surveillance and interventions to mitigate symptoms. The Childhood Cancer Survivor Study (CCSS) is an informative sample to study these conditions as it includes survivors of various childhood cancers who have completed detailed surveys across multiple time-points. In this cohort, we aimed to: (i) estimate prevalence and cumulative incidence of neuromuscular dysfunction by diagnosis in comparison with siblings; (ii) examine associations of treatment exposures with neuromuscular dysfunction, and (iii) determine whether neuromuscular dysfunction is associated with other adverse health and socioeconomic outcomes.
Methods
Study design and participants
The CCSS is a multicenter retrospective cohort study with longitudinal follow-up (25). It includes childhood cancer survivors diagnosed with leukemia, central nervous system (CNS) tumors, Hodgkin lymphoma, non-Hodgkin lymphoma, neuroblastoma, Wilms tumor, or bone/soft tissue sarcoma at <21 years of age at one of 31 participating CCSS institutions in the United States and Canada between January, 1, 1970 and December 31, 1999, who survived at least five years postdiagnosis. A randomly selected subset of siblings closest in age to survivors was included as a comparison group. The study was approved by an institutional review board at each participating institution, and written informed consent was obtained from each participant or a parent/guardian in accordance with the Declaration of Helsinki. Details of the CCSS have been previously reported (25, 26).
The analysis included surveys completed between November 4, 1992 and August 17, 2017. Survivors and siblings were eligible if they completed a baseline survey and had no history of congenital neuromuscular disease (e.g., cerebral palsy). Survivors with subsequent malignant neoplasms (SMNs) or late recurrence were censored at time of SMN/recurrence for time-to-event outcomes to avoid confounding from additional therapies that could cause neuromuscular dysfunction.
Risk factors
Cancer diagnosis and treatment was obtained through medical record abstraction by the treating institution and included chemotherapy exposure, radiotherapy, and surgical procedures (25, 26). Neurotoxic chemotherapeutic agents previously described to be associated with neuromuscular dysfunction were selected a-priori for analysis and included vinca alkaloids, platinum, etoposide, cytarabine and intrathecal methotrexate (4, 8, 27). Distribution curves of cumulative treatment dosage were examined in survivors with and without neuromuscular dysfunction when available. Cumulative dose of carboplatin was divided by four to convert to cisplatin equivalents (28). Distribution of platinum dosage did not differ in those with and without neuromuscular dysfunction, and therefore was analyzed as a dichotomous variable. Exposure to cranial radiotherapy, spinal radiotherapy, or other radiotherapy (defined as non-cranial/spinal radiotherapy) was included, and maximum radiation dose for each site was grouped into tertiles.
Demographic variables included age at diagnosis, sex, race, ethnicity, and years since diagnosis. Self-report of risky alcohol use based on National Institute for Alcohol Abuse and Alcoholism criteria (women; >3 drinks/day or 7 drinks/week, men; >4 drinks/day or 14 drinks/week) (29) and smoking status at baseline were also included as covariates given their known associations with peripheral neuropathy (30).
Definition of primary outcome
Neuromuscular dysfunction was defined as motor or sensory dysfunction diagnosed by a health-care provider, and was identified by a self-report survey administered to the cohort at baseline and regular follow-up intervals (2). Participants reported whether they were ever diagnosed with a neuromuscular condition and age at first occurrence which was used for time-to-event analyses. The survey closest to the event was used since recall quality decreases with time. Motor dysfunction included diagnosis of impaired balance, tremor, or extremity weakness. Sensory dysfunction included diagnosis of impaired touch sensation or prolonged pain or abnormal sensation. These conditions were selected based on previous characterization of this chronic condition within the CCSS, and were only counted as events if they occurred after the childhood cancer diagnosis (2, 31).
Definition of secondary outcomes
All secondary outcomes were based on the last available follow-up survey and considered present if neuromuscular dysfunction was reported simultaneously or prior to the secondary outcome.
Obesity was defined as body mass index (BMI) ≥30 kilograms per square meter (kg/m2) from self-reported height and weight for survivors ≥20 years at the time of evaluation, and as BMI ≥95th percentile based on age and sex matched normative data for survivors <20 years at the time of evaluation (32). Activity limitation was defined as limitation in moderate activities (e.g., walking one block, carrying groceries, walking uphill) for more than three months (33). Frailty was defined as ≥3 modified Fried frailty criteria (including low lean mass, exhaustion, low energy expenditure, walking limitation, and weakness; refs. 34, 35).
Emotional distress was evaluated in participants ≥18 years using the Brief Symptoms Inventory-18 (BSI-18; refs. 22, 36). Subscale scores for depression and anxiety were compared with population normal values, with t-scores ≥63 indicating impairment (22, 36).
Markers of socioeconomic status (SES) included employment status, ability to work, and highest level of education for participants ≥25 years at survey completion. Educational status was dichotomized as less than a college degree versus college or higher degree.
Statistical analysis
Clinical characteristics of survivors including treatment exposures were tabulated. Demographic characteristics were summarized and compared between survivors and siblings using chi-squared test statistic with bootstrapping of families instead of individuals to account for within family correlation (37).
Overall prevalence of neuromuscular dysfunction at time of study eligibility, five years postdiagnosis, and cumulative incidence after five years postdiagnosis was estimated in survivors and siblings: for comparison, a pseudo-diagnosis date was created for each sibling by adding age at diagnosis of his/her sibling survivor to the sibling's date of birth. Prevalence ratios (PR) and 95% confidence intervals (CI) of neuromuscular dysfunction were calculated in survivors compared with siblings using log-binomial regression models adjusted for age at diagnosis, sex, race, and ethnicity. Generalized estimating equations with exchangeable correlation structure were used to account for potential within-family correlation of survivors and siblings.
Among survivors, univariate analyses identified treatment exposures and sociodemographic factors associated with any neuromuscular dysfunction. Variables with P values <0.2 were included in two multivariable models. A log-binomial regression model estimated PRs and CIs of neuromuscular dysfunction at cohort entry (five years postdiagnosis), and a piecewise exponential model estimated rate ratios (RR) and CIs for developing neuromuscular dysfunction after cohort entry, until earliest date of developing neuromuscular dysfunction, SMN/late recurrence, death, or last contact. Two analyses were performed, one including all survivors and a second excluding survivors of CNS tumors, a history of cranial or spinal radiotherapy, or a history of lower limb amputation (that may have other causes of neuromuscular dysfunction unrelated to chemotherapy exposure). A priori hypothesized interactions of age at diagnosis (≥10 years vs. <10 years) were tested with sex, vinca alkaloid exposure, platinum exposure, and history of cranial radiotherapy, based on previous reports of variation in neuromuscular dysfunction by age at these exposures (3, 7). Significant interactions were reported as stratified results.
For analysis of secondary outcomes, log-binomial regression models estimated PRs and CIs for other late health and socioeconomic outcomes (reported concurrent or after neuromuscular dysfunction) in survivors with neuromuscular dysfunction compared with those without neuromuscular dysfunction, as reported in the last available survey. All models were adjusted for age at assessment, sex, race, ethnicity, and severe/disabling chronic conditions (grade 3 or 4) based on Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 (31). All tests were two-sided and P values of <0.05 were considered statistically significant. Sample weights were applied to ALL survivors in all analyses to account for under-sampling of ALL survivors diagnosed from 1987 to 1999 (38). All data analyses were performed using SAS version 9.4.
Results
Study population
Analysis included 25,583 survivors (median age 31.5 years, range 5.6–65.9, median time since diagnosis 23.0 years, range 5.0–46.7) and 5,044 siblings (median age 35.9 years, range 3.1–68.9) who completed a baseline survey and did not have a congenital neuromuscular disease (Fig. 1). Only 360 survivors (1.4%) reported neuromuscular dysfunction prior to cancer diagnosis (1.2% motor dysfunction. 0.4% sensory dysfunction). Differences between survivors and siblings are described in Table 1.
. | Survivors . | Siblings . | . |
---|---|---|---|
. | (N = 25,583) . | (N = 5,044) . | P . |
Sex, n (%) | |||
Male | 13,678 (53.5) | 2,405 (47.7) | <0.001 |
Female | 11,905 (46.5) | 2,639 (52.3) | |
Race, n (%) | |||
White | 21,637 (83.8) | 4,460 (88.4) | <0.001 |
Black | 1,685 (6.7) | 153 (3.0) | |
Other | 2,261 (9.5) | 431 (8.5) | |
Ethnicity, n (%) | |||
Hispanic | 2,028 (8.7) | 184 (3.6) | <0.001 |
Non-Hispanic | 23,479 (91.0) | 4,661 (92.4) | |
Unknown | 76 (0.3) | 199 (3.9) | |
Age at last survey (years) | |||
Mean (SD) | 32.0 (10.4) | 36.2 (11.3) | <0.001 |
Median (range) | 31.5 (5.6–65.9) | 35.9 (3.1–68.9) | |
High-risk alcohol use, n (%) | |||
Yes | 2,820 (12.1) | 675 (13.4) | <0.001 |
No | 22,695 (87.9) | 4,363 (86.6) | |
History of smoking, n (%) | |||
Yes | 5,776 (23.3) | 1,608 (32.2) | <0.001 |
No | 19,100 (76.7) | 3,387 (67.8) | |
Years since cancer diagnosis | |||
Mean (SD) | 23.8 (8.7) | ||
Median (range) | 23.0 (5.0–46.7) | ||
Diagnosis, n (%) | |||
ALL/NHL | 8,708 (43.0) | ||
Acute myeloid leukemia | 922 (3.1) | ||
Other leukemia | 328 (1.1) | ||
CNS tumor | 4,467 (15.1) | ||
Bone/soft tissue sarcoma | 3,846 (13.0) | ||
Hodgkin lymphoma | 3,104 (11.4) | ||
Kidney tumors | 2,269 (7.7) | ||
Neuroblastoma | 1,939 (6.6) | ||
Radiotherapy, n (%) | |||
Yes | 13,093 (51.3) | ||
No | 10,393 (48.7) | ||
Cranial radiotherapy, n (%) | |||
Yes | 6,694 (28.2) | ||
No | 16,190 (71.8) | ||
Spinal radiotherapy, n (%) | |||
Yes | 2,164 (9.1) | ||
No | 20,709 (90.9) | ||
Other radiotherapy, n (%) | |||
Yes | 8,562 (33.1) | ||
No | 14,321 (66.9) | ||
Chemotherapy, n (%) | |||
Yes | 19,090 (84.1) | ||
No | 4,344 (15.9) | ||
Vinca alkaloid, n (%) | |||
Yes | 15,188 (73.4) | ||
No | 6,869 (26.6) | ||
Platinum, n (%) | |||
Yes | 2,736 (10.0) | ||
No | 20,684 (90.0) | ||
Etoposide, n (%) | |||
Yes | 3,685 (17.7) | ||
No | 19,622 (82.3) | ||
Cytarabine, n (%) | |||
Yes | 6,246 (35.8) | ||
No | 16,966 (64.2) | ||
Intrathecal methotrexate, n (%) | |||
Yes | 1,748 (10.0) | ||
No | 15,624 (90.0) |
. | Survivors . | Siblings . | . |
---|---|---|---|
. | (N = 25,583) . | (N = 5,044) . | P . |
Sex, n (%) | |||
Male | 13,678 (53.5) | 2,405 (47.7) | <0.001 |
Female | 11,905 (46.5) | 2,639 (52.3) | |
Race, n (%) | |||
White | 21,637 (83.8) | 4,460 (88.4) | <0.001 |
Black | 1,685 (6.7) | 153 (3.0) | |
Other | 2,261 (9.5) | 431 (8.5) | |
Ethnicity, n (%) | |||
Hispanic | 2,028 (8.7) | 184 (3.6) | <0.001 |
Non-Hispanic | 23,479 (91.0) | 4,661 (92.4) | |
Unknown | 76 (0.3) | 199 (3.9) | |
Age at last survey (years) | |||
Mean (SD) | 32.0 (10.4) | 36.2 (11.3) | <0.001 |
Median (range) | 31.5 (5.6–65.9) | 35.9 (3.1–68.9) | |
High-risk alcohol use, n (%) | |||
Yes | 2,820 (12.1) | 675 (13.4) | <0.001 |
No | 22,695 (87.9) | 4,363 (86.6) | |
History of smoking, n (%) | |||
Yes | 5,776 (23.3) | 1,608 (32.2) | <0.001 |
No | 19,100 (76.7) | 3,387 (67.8) | |
Years since cancer diagnosis | |||
Mean (SD) | 23.8 (8.7) | ||
Median (range) | 23.0 (5.0–46.7) | ||
Diagnosis, n (%) | |||
ALL/NHL | 8,708 (43.0) | ||
Acute myeloid leukemia | 922 (3.1) | ||
Other leukemia | 328 (1.1) | ||
CNS tumor | 4,467 (15.1) | ||
Bone/soft tissue sarcoma | 3,846 (13.0) | ||
Hodgkin lymphoma | 3,104 (11.4) | ||
Kidney tumors | 2,269 (7.7) | ||
Neuroblastoma | 1,939 (6.6) | ||
Radiotherapy, n (%) | |||
Yes | 13,093 (51.3) | ||
No | 10,393 (48.7) | ||
Cranial radiotherapy, n (%) | |||
Yes | 6,694 (28.2) | ||
No | 16,190 (71.8) | ||
Spinal radiotherapy, n (%) | |||
Yes | 2,164 (9.1) | ||
No | 20,709 (90.9) | ||
Other radiotherapy, n (%) | |||
Yes | 8,562 (33.1) | ||
No | 14,321 (66.9) | ||
Chemotherapy, n (%) | |||
Yes | 19,090 (84.1) | ||
No | 4,344 (15.9) | ||
Vinca alkaloid, n (%) | |||
Yes | 15,188 (73.4) | ||
No | 6,869 (26.6) | ||
Platinum, n (%) | |||
Yes | 2,736 (10.0) | ||
No | 20,684 (90.0) | ||
Etoposide, n (%) | |||
Yes | 3,685 (17.7) | ||
No | 19,622 (82.3) | ||
Cytarabine, n (%) | |||
Yes | 6,246 (35.8) | ||
No | 16,966 (64.2) | ||
Intrathecal methotrexate, n (%) | |||
Yes | 1,748 (10.0) | ||
No | 15,624 (90.0) |
Abbreviations: ALL/NHL, acute lymphoblastic leukemia/non-Hodgkin lymphoma, CNS, central nervous system.
aPercentage is based on the total number of participants with available information. Weighting of ALL survivors due to differences in sampling in survivors diagnosed between 1975 and 1999 were accounted for in the percentage calculation.
Prevalence and cumulative incidence of neuromuscular dysfunction
Prevalence of any neuromuscular dysfunction at five years postdiagnosis was higher in survivors than in age-matched siblings (14.7% vs.1.5%, P < 0.001). After adjustment for age, sex, and race/ethnicity, this corresponded to a 9.9-fold increased prevalence of neuromuscular dysfunction in survivors compared with siblings (95% CI, 7.9–12.4). Prevalence of neuromuscular dysfunction in survivors compared with siblings varied by diagnosis (Table 2) and was most elevated in survivors of CNS tumors (PR, 27.6; 95% CI, 22.1–34.6) and sarcomas (PR, 11.5; 95% CI, 9.1–14.5). Cumulative incidence of neuromuscular dysfunction (Fig. 2) remained higher in survivors 20 years postdiagnosis (24.3%; 95% CI, 23.8–24.8); motor (18.2%; 95% CI, 17.7–18.6), sensory (13.5%; 95% CI, 13.1–13.9) than siblings (8.9%; 95% CI, 8.1–9.7); motor (5.3%; 95% CI, 4.7–6.0], sensory (5.5%; 95% CI, 4.8–6.2). When individual conditions were examined separately, survivors were more likely than siblings to report a diagnosis of weakness, tremor, impaired balance, and impaired sensation (Supplementary Fig. 1).
. | . | Survivors, % . | Siblings, % . | PR (95% CI) . |
---|---|---|---|---|
Primary diagnosis . | Type of condition . | (N = 25,506)b . | (N = 5,039)b . | (Reference siblings) . |
Overall | Any | 14.7 | 1.5 | 9.87 (7.88–12.35) |
Motor | 11.9 | 1.2 | 10.57 (8.16–13.71) | |
Sensory | 6.5 | 0.6 | 10.97 (7.60–15.31) | |
Acute lymphoblastic leukemia/non-Hodgkin lymphoma | Any | 8.4 | 5.83 (4.61–7.37) | |
Motor | 6.2 | 5.63 (4.28–7.40) | ||
Sensory | 4.2 | 7.93 (5.53–11.37) | ||
Central nervous system tumors | Any | 42.3 | 27.63 (22.05–34.64) | |
Motor | 40.5 | 35.49 (27.35–46.05) | ||
Sensory | 11.8 | 19.32 (13.52–27.61) | ||
Bone/soft tissue sarcomas | Any | 18.6 | 11.46 (9.07–14.47) | |
Motor | 12.1 | 10.16 (7.74–13.35) | ||
Sensory | 12.6 | 16.52 (11.53–23.67) | ||
Acute myeloid leukemia | Any | 9.7 | 6.41 (4.77–8.62) | |
Motor | 7.4 | 6.51 (4.63–9.17) | ||
Sensory | 4.5 | 7.12 (4.50–11.27) | ||
Other leukemia | Any | 9.1 | 5.89 (3.92–8.84) | |
Motor | 7.3 | 6.35 (4.00–10.09) | ||
Sensory | 3.7 | 5.37 (2.78–10.35) | ||
Hodgkin lymphoma | Any | 8.0 | 4.51 (3.49–5.82) | |
Motor | 4.1 | 3.34 (2.45–4.56) | ||
Sensory | 5.2 | 5.62 (3.83–8.25) | ||
Kidney tumors | Any | 3.8 | 2.86 (2.10–3.88) | |
Motor | 2.6 | 2.46 (1.71–3.53) | ||
Sensory | 1.7 | 4.01 (2.51–6.42) | ||
Neuroblastoma | Any | 12.2 | 9.20 (7.14–11.86) | |
Motor | 9.8 | 9.29 (6.94–12.44) | ||
Sensory | 6.2 | 15.08 (10.21–22.30) |
. | . | Survivors, % . | Siblings, % . | PR (95% CI) . |
---|---|---|---|---|
Primary diagnosis . | Type of condition . | (N = 25,506)b . | (N = 5,039)b . | (Reference siblings) . |
Overall | Any | 14.7 | 1.5 | 9.87 (7.88–12.35) |
Motor | 11.9 | 1.2 | 10.57 (8.16–13.71) | |
Sensory | 6.5 | 0.6 | 10.97 (7.60–15.31) | |
Acute lymphoblastic leukemia/non-Hodgkin lymphoma | Any | 8.4 | 5.83 (4.61–7.37) | |
Motor | 6.2 | 5.63 (4.28–7.40) | ||
Sensory | 4.2 | 7.93 (5.53–11.37) | ||
Central nervous system tumors | Any | 42.3 | 27.63 (22.05–34.64) | |
Motor | 40.5 | 35.49 (27.35–46.05) | ||
Sensory | 11.8 | 19.32 (13.52–27.61) | ||
Bone/soft tissue sarcomas | Any | 18.6 | 11.46 (9.07–14.47) | |
Motor | 12.1 | 10.16 (7.74–13.35) | ||
Sensory | 12.6 | 16.52 (11.53–23.67) | ||
Acute myeloid leukemia | Any | 9.7 | 6.41 (4.77–8.62) | |
Motor | 7.4 | 6.51 (4.63–9.17) | ||
Sensory | 4.5 | 7.12 (4.50–11.27) | ||
Other leukemia | Any | 9.1 | 5.89 (3.92–8.84) | |
Motor | 7.3 | 6.35 (4.00–10.09) | ||
Sensory | 3.7 | 5.37 (2.78–10.35) | ||
Hodgkin lymphoma | Any | 8.0 | 4.51 (3.49–5.82) | |
Motor | 4.1 | 3.34 (2.45–4.56) | ||
Sensory | 5.2 | 5.62 (3.83–8.25) | ||
Kidney tumors | Any | 3.8 | 2.86 (2.10–3.88) | |
Motor | 2.6 | 2.46 (1.71–3.53) | ||
Sensory | 1.7 | 4.01 (2.51–6.42) | ||
Neuroblastoma | Any | 12.2 | 9.20 (7.14–11.86) | |
Motor | 9.8 | 9.29 (6.94–12.44) | ||
Sensory | 6.2 | 15.08 (10.21–22.30) |
aTwo logistic regression models were used to estimate PRs in the entire cohort and by diagnosis. Both models were adjusted for sex, race, ethnicity, and age at diagnosis.
bSurvivors who developed a subsequent malignant neoplasm <5 years from their cancer diagnosis and siblings who developed cancer <5 years from their pseudo-diagnosis date (years from date of birth of matched survivor's age in years at diagnosis) were excluded from this analysis.
Risk factors associated with neuromuscular dysfunction
Among all survivors, treatment exposures associated with an increased prevalence of neuromuscular dysfunction included exposure to platinum, spinal radiotherapy, and ≥50 Gy of cranial radiotherapy (Table 3). We detected significant interactions of age at diagnosis with platinum (P = 0.001), vinca alkaloid (P < 0.001), and cranial radiotherapy exposure (P < 0.001) on the prevalence of neuromuscular dysfunction. Survivors <10 years of age at diagnosis with vinca alkaloid exposure had a lower prevalence of neuromuscular dysfunction than survivors without vinca alkaloid exposure. However, when excluding survivors with a history of CNS tumors, cranial or spinal radiotherapy, or lower limb amputation from the analysis, vinca alkaloid exposure (versus none) was no longer associated with a decreased prevalence of neuromuscular dysfunction in survivors <10 years of age at diagnosis. In this sample, vinca alkaloid exposure was associated with an increased prevalence of neuromuscular dysfunction in survivors ≥10 years of age at diagnosis and platinum exposure (versus none) remained associated with an increased prevalence of neuromuscular dysfunction. In a post-hoc analysis the prevalence of neuromuscular dysfunction in survivors compared with siblings remained elevated (PR, 8.2, CI 6.5–10.3) after adjusting for associated treatment exposures.
. | All survivors . | Excluding CNS tumors, cranial or spinal radiotherapy, or amputation . | ||
---|---|---|---|---|
. | (N = 25,506)b . | (N = 16,035)b . | ||
. | PR (95% CI) . | P . | PR (95% CI) . | P . |
Sex by age at diagnosis (years) | ||||
Male <10 (ref) | 1.0 | 1.0 | ||
Male ≥10 | 1.14 (1.00–1.30) | 0.049 | 1.02 (0.81–1.28) | 0.87 |
Female <10 | 0.97 (0.89–1.06) | 0.50 | 0.94 (0.78–1.14) | 0.54 |
Female ≥10 | 1.40 (1.24–1.60) | <0.001 | 1.37 (1.09–1.73) | 0.006 |
Vinca alkaloid exposure by age at diagnosis (years) | ||||
No (ref) | 1.0 | 1.0 | ||
Yes and age <10 | 0.70 (0.64–0.77) | <0.001 | 0.74 (0.61–0.88) | <0.001 |
Yes and age ≥10 | 1.02 (0.93–1.12) | 0.65 | 1.36 (1.16–1.59) | <0.001 |
Platinum exposure by age at diagnosis (years) | ||||
No (ref) | 1.0 | |||
Yes and age <10 | 1.41 (1.25–1.59) | <0.001 | 1.0 | |
Yes and age ≥10 | 1.26 (1.12–1.41) | <0.001 | 1.77 (1.51–2.07)c | <0.001 |
Cytarabine | ||||
No (ref) | 1.0 | 1.0 | ||
Yes | 0.88 (0.79–0.98) | 0.02 | 0.87 (0.72–1.06) | 0.17 |
Etoposide | ||||
No (ref) | 1.0 | 1.0 | ||
Yes | 0.91 (0.82–1.01) | 0.08 | 1.03 (0.87–1.23) | 0.70 |
Intrathecal methotrexate | ||||
No (ref) | 1.0 | 1.0 | ||
Yes | 0.65 (0.57–0.74) | <0.001 | 1.03 (0.84–1.22) | 0.91 |
Cranial radiotherapy by age at diagnosis (years) | ||||
None (ref) | 1.0 | d | d | |
>0 to <20 Gy and age <10 | 1.00 (0.80–1.26) | <0.99 | ||
>0 to <20 Gy and age ≥10 | 1.27 (0.99–1.63) | 0.05 | ||
>20 to <50 Gy and age <10 | 1.72 (1.47–2.03) | <0.001 | ||
>20 to <50 Gy and age ≥10 | 1.21 (0.97–1.50) | 0.09 | ||
≥50 Gy and age <10 | 2.81 (2.53–3.12) | <0.001 | ||
≥50 Gy and age ≥10 | 1.97 (1.75–2.21) | <0.001 | ||
Spinal radiotherapy | ||||
None (ref) | 1.0 | d | d | |
>0 to <15 Gy | 1.84 (1.32–2.46) | <0.001 | ||
≥15 Gy to <30 Gy | 2.28 (1.95–2.66) | <0.001 | ||
≥30 Gy | 1.93 (1.67–2.22) | <0.001 |
. | All survivors . | Excluding CNS tumors, cranial or spinal radiotherapy, or amputation . | ||
---|---|---|---|---|
. | (N = 25,506)b . | (N = 16,035)b . | ||
. | PR (95% CI) . | P . | PR (95% CI) . | P . |
Sex by age at diagnosis (years) | ||||
Male <10 (ref) | 1.0 | 1.0 | ||
Male ≥10 | 1.14 (1.00–1.30) | 0.049 | 1.02 (0.81–1.28) | 0.87 |
Female <10 | 0.97 (0.89–1.06) | 0.50 | 0.94 (0.78–1.14) | 0.54 |
Female ≥10 | 1.40 (1.24–1.60) | <0.001 | 1.37 (1.09–1.73) | 0.006 |
Vinca alkaloid exposure by age at diagnosis (years) | ||||
No (ref) | 1.0 | 1.0 | ||
Yes and age <10 | 0.70 (0.64–0.77) | <0.001 | 0.74 (0.61–0.88) | <0.001 |
Yes and age ≥10 | 1.02 (0.93–1.12) | 0.65 | 1.36 (1.16–1.59) | <0.001 |
Platinum exposure by age at diagnosis (years) | ||||
No (ref) | 1.0 | |||
Yes and age <10 | 1.41 (1.25–1.59) | <0.001 | 1.0 | |
Yes and age ≥10 | 1.26 (1.12–1.41) | <0.001 | 1.77 (1.51–2.07)c | <0.001 |
Cytarabine | ||||
No (ref) | 1.0 | 1.0 | ||
Yes | 0.88 (0.79–0.98) | 0.02 | 0.87 (0.72–1.06) | 0.17 |
Etoposide | ||||
No (ref) | 1.0 | 1.0 | ||
Yes | 0.91 (0.82–1.01) | 0.08 | 1.03 (0.87–1.23) | 0.70 |
Intrathecal methotrexate | ||||
No (ref) | 1.0 | 1.0 | ||
Yes | 0.65 (0.57–0.74) | <0.001 | 1.03 (0.84–1.22) | 0.91 |
Cranial radiotherapy by age at diagnosis (years) | ||||
None (ref) | 1.0 | d | d | |
>0 to <20 Gy and age <10 | 1.00 (0.80–1.26) | <0.99 | ||
>0 to <20 Gy and age ≥10 | 1.27 (0.99–1.63) | 0.05 | ||
>20 to <50 Gy and age <10 | 1.72 (1.47–2.03) | <0.001 | ||
>20 to <50 Gy and age ≥10 | 1.21 (0.97–1.50) | 0.09 | ||
≥50 Gy and age <10 | 2.81 (2.53–3.12) | <0.001 | ||
≥50 Gy and age ≥10 | 1.97 (1.75–2.21) | <0.001 | ||
Spinal radiotherapy | ||||
None (ref) | 1.0 | d | d | |
>0 to <15 Gy | 1.84 (1.32–2.46) | <0.001 | ||
≥15 Gy to <30 Gy | 2.28 (1.95–2.66) | <0.001 | ||
≥30 Gy | 1.93 (1.67–2.22) | <0.001 |
Abbreviation: CNS, central nervous system.
aTwo logistic regression models were used to estimate PRs in all survivors and excluding survivors with CNS tumors, cranial/spinal radiation, or lower limb amputation. Both models were adjusted for significant treatment exposures with P <0.2 and interactions with P <0.05 by univariate analysis. Significant interactions are reported as stratified results. Both models adjusted for race, ethnicity, other radiotherapy, alcohol use, and smoking status when significant by univariate analysis, in addition to the variables displayed.
bSurvivors who developed a subsequent malignant neoplasm <5 years from their cancer diagnosis were excluded from this analysis.
cThere was no significant interaction of platinum with age at diagnosis in this model and therefore platinum exposure is not stratified by age.
dSurvivors with cranial radiotherapy and spinal radiotherapy were excluded from this model.
In incidence analyses, treatment exposures most associated with developing neuromuscular dysfunction after cohort entry (>5 years postdiagnosis) included platinum exposure and cranial radiotherapy (Supplementary Table S1). A post-hoc analysis found motor dysfunction at five years post-diagnosis was associated with developing sensory dysfunction >5 years post-diagnosis (RR, 1.8; CI 1.6–2.0, reference no motor dysfunction at five years postdiagnosis), and sensory dysfunction at five years post-diagnosis was associated with developing motor dysfunction >5 years post-diagnosis (RR, 1.9; CI 1.6–2.1, reference no sensory dysfunction at five years postdiagnosis).
Association of neuromuscular dysfunction with health and socioeconomic outcomes
In multivariable models, survivors with any neuromuscular dysfunction were more likely to report adverse health and socioeconomic outcomes at or after their diagnosis of neuromuscular dysfunction compared to those without dysfunction (Table 4). Specifically, survivors with neuromuscular dysfunction (versus none) were more likely to experience anxiety (PR, 2.5, CI 2.2–2.9), depression (PR, 2.1; CI 1.9–2.3), obesity (PR, 1.1; CI 1.1–1.2), frailty (PR, 1.5; CI 1.4–1.5) or limitation in activity (PR, 2.7; CI 2.5–2.9). They were also less likely to have attained a college or higher degree (PR, 0.92; CI 0.90–0.94), or be employed (PR, 0.8; CI 0.8–0.9). Associations did not change when excluding survivors who had CNS tumors, who received cranial or spinal radiotherapy, or who had a lower limb amputation (Supplementary Table S2). When examined by type of dysfunction, survivors with motor dysfunction (versus none) and sensory dysfunction (versus none) also had increased prevalence of these outcomes (Supplementary Table S3).
. | Survivors with neuromuscular dysfunction . | Survivors without neuromuscular dysfunction . | . | . |
---|---|---|---|---|
. | n (%)b . | n (%)b . | PRc (95% CI) . | P . |
Emotional distress | ||||
Anxietyd | 570 (11.9) | 628 (4.7) | 2.54 (2.23–2.89) | <0.001 |
Depressiond | 789 (16.3) | 1,061 (8.0) | 2.05 (1.85–2.27) | <0.001 |
Obesity | 1,722 (26.8) | 2,712 (22.9) | 1.11 (1.05–1.17) | <0.001 |
Activity limitation | 2,122 (31.9) | 1,643 (9.2) | 2.69 (2.52–2.87) | <0.001 |
Frailty | 3,125 (61.4) | 4,467 (37.2) | 1.46 (1.41–1.51) | <0.001 |
College or higher degree | 3,666 (69.7) | 9,173 (75.8) | 0.92 (0.90–0.94) | <0.001 |
Employmente | ||||
Currently employed | 3,506 (70.0) | 10,269 (89.0) | 0.83 (0.82–0.85) | <0.001 |
Unable to work due to disability | 1,497 (29.9) | 901 (7.9) | 3.00 (2.76–3.27) | <0.001 |
. | Survivors with neuromuscular dysfunction . | Survivors without neuromuscular dysfunction . | . | . |
---|---|---|---|---|
. | n (%)b . | n (%)b . | PRc (95% CI) . | P . |
Emotional distress | ||||
Anxietyd | 570 (11.9) | 628 (4.7) | 2.54 (2.23–2.89) | <0.001 |
Depressiond | 789 (16.3) | 1,061 (8.0) | 2.05 (1.85–2.27) | <0.001 |
Obesity | 1,722 (26.8) | 2,712 (22.9) | 1.11 (1.05–1.17) | <0.001 |
Activity limitation | 2,122 (31.9) | 1,643 (9.2) | 2.69 (2.52–2.87) | <0.001 |
Frailty | 3,125 (61.4) | 4,467 (37.2) | 1.46 (1.41–1.51) | <0.001 |
College or higher degree | 3,666 (69.7) | 9,173 (75.8) | 0.92 (0.90–0.94) | <0.001 |
Employmente | ||||
Currently employed | 3,506 (70.0) | 10,269 (89.0) | 0.83 (0.82–0.85) | <0.001 |
Unable to work due to disability | 1,497 (29.9) | 901 (7.9) | 3.00 (2.76–3.27) | <0.001 |
aAnalysis was repeated excluding central nervous system tumors, cranial/spinal radiotherapy, and lower limb amputation, and associations did not change (Supplementary Table S2).
bPercentages considered the sample weight.
cReference is survivors without neuromuscular dysfunction. Each outcome was analyzed in a separate log-binomial regression model adjusted for sex, race, ethnicity, age at the assessment and presence of severe, disabling, or life-threatening chronic conditions.
dIncludes survivors 18 years of age or older.
eIncludes survivors 25 years of age or older.
Discussion
In our longitudinal cohort study of 25,583 childhood cancer survivors, we found survivors five years post-diagnosis had a nearly 10-fold increased prevalence of neuromuscular dysfunction compared with siblings, and cumulative incidence of neuromuscular dysfunction reached 24.3% by 20 years postdiagnosis. Prevalence of neuromuscular dysfunction varied by diagnosis and was highest in survivors of central nervous system (CNS) tumors, followed by sarcomas. Treatment exposures associated with neuromuscular dysfunction included platinum and cranial and spinal radiotherapy. Neuromuscular dysfunction was associated with late health/socioeconomic outcomes including obesity, activity limitation, frailty, emotional distress, lower educational attainment, and unemployment.
A key finding of this study is that while the five-year prevalence and 20-year cumulative incidence of neuromuscular dysfunction was increased in survivors compared with siblings, cumulative incidence curves demonstrated a similar slope in survivors and siblings from five to 20 years postdiagnosis. This is the first study to our knowledge to longitudinally evaluate neuromuscular dysfunction in childhood cancer survivors more than five years postdiagnosis and suggests the increased risk of developing neuromuscular dysfunction in survivors compared with siblings is highest within five years of diagnosis. Interestingly, we also found platinum and radiotherapy were associated with developing late dysfunction in survivors more than five years postdiagnosis, indicating certain treatment exposures may put survivors at risk for neuromuscular dysfunction beyond five years. In addition, a diagnosis of only motor or sensory dysfunction at five years postdiagnosis was also associated with progressing to both types of dysfunction during follow-up. One potential explanation is worsening peripheral neuropathy that has been described after cessation of treatment, also referred to as “coasting” (18, 39). Alternatively, survivors may have subclinical nerve damage from treatment that is not identified until they develop age-related changes that impair mobility and amplify symptoms. Another possibility is that the neuromuscular dysfunction is mediated by other chronic diseases such as diabetes or cerebrovascular disease that develop as childhood cancer survivors age (40–42). Regardless of the etiology of dysfunction, these findings have important implications for surveillance. Nearly one quarter of childhood cancer survivors are diagnosed with neuromuscular dysfunction by 20 years postdiagnosis, and some survivors continue to develop neuromuscular dysfunction more than five years postdiagnosis. Current guidelines recommend screening for neuromuscular dysfunction for two to three years posttherapy in certain survivors (43), however survivors exposed to platinum or radiotherapy, or with existing motor or sensory dysfunction, may benefit from longer monitoring.
Another important finding of our study is the association of neuromuscular dysfunction with adverse health and socioeconomic outcomes. While previous studies have demonstrated neuromuscular dysfunction is associated with impaired health-related quality of life and physical function in childhood cancer survivors (10, 14, 16, 44), this is the first study to our knowledge to demonstrate its association with anxiety, depression, obesity, unemployment, and lower educational attainment. We recognize survivors with neuromuscular dysfunction may represent a population with poor overall health that could confound this finding, though our analyses controlled for other severe/disabling chronic health conditions, so we do not believe this is the case. In addition, associations persisted in a sensitivity analysis excluding survivors with CNS tumors, cranial or spinal radiotherapy, or lower limb amputation, who are particularly at risk for adverse health outcomes (45). There are many reasons survivors with neuromuscular dysfunction may be at risk for these outcomes. Neuromuscular dysfunction can lead to social isolation that may contribute to emotional distress (20). Impairment in fine motor skills necessary for schoolwork early in life may play a role in the association of neuromuscular dysfunction with lower educational attainment and SES (8, 19). The association of lower SES with neuromuscular dysfunction is particularly relevant since physical therapy can improve neuromuscular dysfunction (46, 47), however individuals with lower SES are less likely to receive this intervention (48–50). Future interventions for survivors with neuromuscular dysfunction should consider these potential socioeconomic barriers. We also found survivors with neuromuscular dysfunction were more likely to experience frailty, an increasingly identified phenotype in childhood cancer survivors that is characterized by a loss of physiologic reserve typically seen in older adults (34, 51). Frailty is associated with emotional distress and inability to work in adults, and it is possible it contributed to these outcomes in our study (52). In addition, in childhood cancer survivors increased duration of chronic conditions is associated with frailty (34), further highlighting the need for surveillance and early interventions for survivors with neuromuscular dysfunction to mitigate adverse outcomes.
While our study supported previous findings that exposure to platinum and radiotherapy is associated with late neuromuscular dysfunction (14, 53, 54), the association of other treatment factors with neuromuscular dysfunction was less straightforward. An unexpected finding was that vinca alkaloid exposure was not associated with neuromuscular dysfunction among all survivors, despite its well-described association with chemotherapy-induced peripheral neuropathy (7, 11, 17). This may be partially due to the contribution of central neurologic deficits or musculoskeletal conditions to neuromuscular dysfunction (55). We found when excluding survivors that could have causes of neuromuscular dysfunction outside of their chemotherapy treatment exposure (i.e., survivors with a history of CNS tumors, cranial or spinal radiotherapy, or lower limb amputations), survivors ≥10 years of age at diagnosis exposed to vinca alkaloids had an increased prevalence of neuromuscular dysfunction compared to those with no vinca alkaloid exposure. This supports previous research that increasing age at diagnosis is a risk factor for vinca alkaloid-induced peripheral neuropathy (3). The lower prevalence of neuromuscular dysfunction in survivors who were <10 years of age at diagnosis and exposed to vinca alkaloids (compared with survivors with no vinca alkaloid exposure) was an unexpected finding we cannot fully explain. It is possible survivors diagnosed early in life are more likely to adapt to any dysfunction caused by their treatment exposure, and therefore less likely to have it diagnosed in adulthood. In addition, the association of neuromuscular dysfunction with vinca alkaloids may have a threshold effect, only occurring at higher cumulative dosages (4, 18). Our study did not have available cumulative vinca alkaloid exposure data for all participants so we could not examine this. It is also possible vinca alkaloids are associated with another protective exposure that we could not measure in this population. Finally, the treatment factors we examined did not fully explain the nearly 10-fold increased prevalence of neuromuscular dysfunction in survivors compared with siblings. Other factors we could not examine, such as access to early rehabilitation may have also contributed to dysfunction (46). Treatment and non-treatment related risk factors for developing neuromuscular dysfunction should continue to be explored prospectively.
Additional potential limitations in this study should be considered when interpreting these data. We analyzed neuromuscular dysfunction in a dichotomous way and could not grade the severity of these conditions. Therefore, we could not determine whether patterns of severity changed over time, or whether associations with treatment and health and socioeconomic outcomes varied by severity of conditions. In addition, participants were asked to report whether they ever experienced neuromuscular dysfunction and when it first occurred, however we could not determine the duration of these conditions or whether they resolved with time. Studies assessing durability of dysfunction are needed. Our primary outcome relied on self-report of provider-diagnosed neuromuscular dysfunction, and did not have medical record confirmation, leading to potential over or under estimation of the true prevalence. It is also possible that survivors with anxiety and depression were more likely to report neuromuscular dysfunction that represented somatic complaints. Because our outcomes of interest were based on provider-diagnosed conditions we do not believe this was the case. Self-report of chronic health conditions was valuable for this study as it allowed for collection of data from a large, representative sample of survivors across multiple institutions (2, 26). Finally, we could not differentiate whether dysfunction was due to underlying musculoskeletal conditions, or peripheral or central neurologic deficits. Although this information would be helpful to better characterize risk factors for dysfunction, our findings still demonstrate that nearly a quarter of survivors report some form of neuromuscular dysfunction that is associated with other meaningful health outcomes.
Conclusions
Overall, we found childhood cancer survivors across cancer diagnoses are more likely to experience neuromuscular dysfunction than a sibling comparison group as many as 20 years postdiagnosis. Within childhood cancer survivors, neuromuscular dysfunction is associated with anxiety, depression, obesity, frailty, activity limitation, unemployment and lower educational attainment. Although further studies are needed to elucidate the mechanisms of neuromuscular dysfunction, interventions to prevent and treat neuromuscular dysfunction are warranted in this population.
Authors' Disclosures
R.L. Rodwin reports grants from NIH/NCI and grants from William O. Seery Mentored Research Award for Cancer Research, Bank of America, N.A., Trustee during the conduct of the study. Y. Yasui reports grants from NCI during the conduct of the study. W.M. Leisenring reports grants from NIH during the conduct of the study. R.M. Howell reports grants from St. Jude Children's Research Hospital during the conduct of the study. K.R. Krull reports grants from National Cancer Institute during the conduct of the study. E.J. Chow reports grants from Abbott Labs outside the submitted work. G.T. Armstrong reports grants from NIH during the conduct of the study. K.K. Ness reports grants from NIH and other support from ALSAC during the conduct of the study. No disclosures were reported by the other authors.
Authors' Contributions
R.L. Rodwin: Conceptualization, visualization, methodology, writing–original draft, writing–review and editing. Y. Chen: Conceptualization, data curation, software, formal analysis, visualization, methodology, writing–review and editing. Y. Yasui: Conceptualization, formal analysis, supervision, methodology, writing–review and editing. W.M. Leisenring: Conceptualization, formal analysis, supervision, methodology, writing–review and editing. T.M. Gibson: Conceptualization, methodology, writing–review and editing. P.C. Nathan: Conceptualization, methodology, writing–review and editing. R.M. Howell: Conceptualization, methodology, writing–review and editing. K.R. Krull: Conceptualization, methodology, writing–review and editing. C. Mohrmann: Conceptualization, methodology, writing–review and editing. R.J. Hayashi: Conceptualization, methodology, writing–review and editing. E.J. Chow: Conceptualization, supervision, methodology, writing–review and editing. K.C. Oeffinger: Conceptualization, supervision, methodology, writing–review and editing. G.T. Armstrong: Conceptualization, supervision, methodology, writing–review and editing. K.K. Ness: Conceptualization, methodology, writing–review and editing. N.S. Kadan-Lottick: Conceptualization, supervision, methodology, writing–original draft, writing–review and editing.
Acknowledgments
This work was supported by the NCI at the NIH (CA55727, to G.T. Armstrong, principal investigator) and support to St. Jude Children's Research Hospital from the Cancer Center Support (CORE) grant (CA21765, to C. Roberts, principal investigator) and the American Lebanese-Syrian Associated Charities (ALSAC). R.L. Rodwin was supported by the NIH under the Yale Cancer Prevention and Control Training Program (T32 CA250803) from the NCI, as well as the Yale Pediatric Scholar Program and the William O. Seery Mentored Research Award for Cancer Research, Bank of America, N.A., Trustee.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.