Abstract 685: A social epigenomic investigation of racial disparity in pulmonary impairment among aging survivors of childhood cancer

Background: Prior research suggests that social determinants of health (SDOH) may influence health through an epigenetic mechanism. However, the social epigenomic approach has not yet been applied to childhood cancer survivors, a population at high risk for chronic health conditions. We aim to investigate how SDOH factors contribute to racial disparity in pulmonary impairment with survivors from the St. Jude Lifetime Cohort Study.

Methods: DNA methylation (DNAm) profile was generated with EPIC BeadChip using blood derived DNA. SDOH factors included educational attainment and personal income self-reported using a survey, and socioeconomic area deprivation index (ADI) geocoded using full home addresses. Clinically assessed pulmonary impairment included pulmonary diffusion deficits (PDD), restrictive pulmonary deficits (RPD) and obstructive pulmonary deficits (OPD). Epigenome-wide association study (EWAS) was performed to evaluate the association between DNAm at each CpG and each SDOH factor. Mediation analysis was conducted by treating each SDOH-associated CpG as a mediator, SDOH factor as an exposure, and specific pulmonary condition as the outcome. Genetically inferred races, i.e. survivors of European ancestry (EA) and African ancestry (AA), were considered.

Results: The study included 258 AA (median time from cancer diagnosis [MTD]=25.2 years, interquartile range [IQR]=19.9-32.1 years) and 1,618 EA survivors (MTD=27.3, IQR=21.1-33.7 years). Compared to EA survivors, AA survivors had lower educational attainment (P<0.0001), lower personal income (P<0.0001), and higher ADI (P<0.0001). Compared to EA survivors, incidence of PDD (25.2% in AA vs 18.2% in EA, P=0.03) and RPD (14.2% in AA vs 7.5% in EA, P=0.002) were significantly higher in AA survivors, whereas OPD were comparable between groups (9.8% vs 13.1%, P=0.21). However, the racial disparity in PDD became nonsignificant after adjusting for SDOH. EWAS identified 130 SDOH-CpG associations at epigenome-wide significance (P<9×10-8) including 88 for educational attainment, 23 for personal income, and 19 for ADI. Thirteen CpGs, commonly associated with all three SDOH factors, resided at pleiotropic loci featuring cigarette smoking genes, e.g. CPOX and AHRR among others. Three independent SDOH-associated CpGs (cg04180924, cg1120500, and cg27470486) had a significant combined mediation effect for educational attainment (%mediation = 48.9%), and a single mediator cg08064403 had a significant mediation effect for personal income (25.9%) and ADI (24.1%) on PDD risk.

Conclusions: DNAm signatures, many resembling the effect of tobacco use, are associated with educational attainment, personal income, and ADI. Our findings suggest that these DNAm are potential mechanistic mediators for the effects of SDOH factors on PDD risk. Through this mechanism, poor SDOH factors in AA survivors led to racial disparity in PDD.

Citation Format: Nan Song, Jin-ah Sim, Qian Dong, Yinan Zheng, Lifang Hou, Zhenghong Li, Chia-Wei Hsu, Haitao Pan, Heather Mulder, John Easton, Emily Walker, Geoffrey Neale, Carmen L. Wilson, Kirsten K. Ness, Kevin R. Krull, Deo Kumar Srivastava, Yutaka Yasui, Jinghui Zhang, Melissa M. Hudson, Leslie L. Robison, I-Chan Huang, Zhaoming Wang. A social epigenomic investigation of racial disparity in pulmonary impairment among aging survivors of childhood cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 685.

Longitudinal Evaluation of Neuromuscular Dysfunction in Long-term Survivors of Childhood Cancer: A Report from the Childhood Cancer Survivor Study

BACKGROUND

Children treated for cancer are at risk for neuromuscular dysfunction, but data are limited regarding prevalence, longitudinal patterns, and long-term impact.

METHODS

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.

RESULTS

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).

CONCLUSIONS

Neuromuscular dysfunction is prevalent in childhood cancer survivors, continues to increase posttherapy, and is associated with adverse health and socioeconomic outcomes.

IMPACT

Interventions are needed to prevent and treat neuromuscular dysfunction, especially in survivors with platinum and radiation exposure.

Long-term neurocognitive and quality of life outcomes in survivors of pediatric hematopoietic cell transplant

PURPOSE

Pediatric patients who undergo hematopoietic cell transplant (HCT) are at risk for neurocognitive impairments, which can impact quality of life. Given limited long-term studies, we aimed to characterize the late neurocognitive outcomes in a cohort of pediatric HCT survivors.

METHODS

Eligible survivors (HCT at age < 21 year and ≥ 1 year post-HCT) completed a 60-question survey of neurocognitive function and quality of life, which included the Childhood Cancer Survivor Study Neurocognitive Questionnaire (CCSS-NCQ) and the Neuro-Quality of Life Cognitive Function Short Form (Neuro-QoL). Analyses of risk factors included univariate comparisons and multivariable logistic regression.

RESULTS

Participants (n = 199, 50.3% female, 53.3% acute leukemia, 87.9% allogeneic transplants) were surveyed at median age of 37.8 years (interquartile range [IQR] 28.5-48.8) at survey and median 27.6 years (IQR 17.0-34.0) from transplant. On the CCSS-NCQ, 18.9-32.5% of survivors reported impairments (Z score > 1.28) in task efficiency, memory, emotional regulation, or organization, compared with expected 10% in the general population (all p < 0.01). In contrast, survivors reported average Neuro-QoL (T score 49.6±0.7) compared with population normative value of 50 (p = 0.52). In multivariable regression, impaired Neuro-QoL (T score < 40) was independently associated with hearing issues (OR 4.97, 95% CI 1.96-12.6), history of stroke or seizure (OR 4.46, 95% CI 1.44-13.8), and sleep disturbances (OR 6.95, 95% CI 2.53-19.1).

CONCLUSIONS

Although long-term survivors of pediatric HCT reported higher rates of impairment in specific neurocognitive domains, cognitive quality of life was perceived as similar to the general population. Subsets of survivors with certain co-morbidities had substantially worse neurocognitive outcomes.

IMPLICATIONS FOR CANCER SURVIVORS

While the long-term impact of pediatric HCT can include neurocognitive deficits, survivors report average cognitive quality of life.

Health-related and cancer risk concerns among siblings of childhood cancer survivors: a report from the Childhood Cancer Survivor Study (CCSS)

OBJECTIVE

To characterize the prevalence and predictors of concerns regarding future health and cancer risk among siblings of childhood cancer survivors.

METHODS

This study reports longitudinal data (baseline and follow-up) from 3969 adult siblings (median age = 29 [range 18-56] years) of long-term survivors of childhood cancer (median time since diagnosis 19.6 [9.6-33.8] years). Self-reported future health and cancer risk concerns (concerned vs not concerned) were assessed. Demographics and health data reported by both the siblings and their matched cancer survivors were examined as risk factors for health concerns using multivariable logistic regression.

RESULTS

Percentage of siblings reporting future health and cancer risk concerns, respectively, decreased across decade of survivors' diagnosis: 1970s (73.3%; 63.9%), 1980s (67.2%; 62.6%), and 1990s (45.7%; 52.3%). Risk factors associated with future health concerns included sibling chronic health conditions (grade 2 Odds Ratio [OR]=1.57, 95% CI: 1.12-2.20; grades 3-4 OR=1.86, 95% CI: 1.18-2.94; compared to less than grade 2). Risk factors associated with future cancer concerns included sibling chronic health conditions (grade 2 OR=1.43, 95% CI: 1.05-1.94; grades 3-4 OR=1.64, 95% CI: 1.09-2.47; compared to less than grade 2).

CONCLUSIONS

Sibling concerns regarding future health and cancer have diminished in recent decades. There are subgroups of siblings that are at-risk for future health and cancer risk concerns.

IMPLICATIONS FOR CANCER SURVIVORS

Routine screening of concerns in at-risk siblings of survivors of childhood cancer may benefit the siblings of cancer survivors. These individuals may benefit from early interventions during diagnosis and treatment of their siblings.

Associations between diet and chronic disease burden in adult survivors of childhood cancer in the St. Jude Lifetime Cohort (SJLIFE) study

e22017

Background: Nutritional intake can impact health. Data describing associations between diet or specific dietary components and chronic health conditions in long-term survivors of childhood cancer are limited. This study evaluated the associations between specific components of diet and chronic disease burden in childhood cancer survivors. Methods: Adult survivors (≥5 years) of childhood cancer participating in SJLIFE (n = 2822), who completed a clinical evaluation and the Block Food Frequency Questionnaire were included. 168 chronic conditions were graded using a modified Common Terminology for Adverse Events. Conditions were summed within grade category, multiplied by weights (1, 2, 3, and 8 for grades 1-4, respectively) and standardized (z-scores) with higher scores indicating more disease burden. Adherence to the 2015 Dietary Guidelines for Americans was scored with the Healthy Eating Index-15 (HEI). Intake of key food groups and nutrients was estimated and divided into quartiles for analysis. Multivariable logistic regression, adjusted for sex, age, and race was used to compare dietary intake and higher disease burden (defined as a z-score > 0). Results: Survivors (median [range] survival time 23 [10-49] years, diagnosis age 7 [0-24] years, 47.6% female, 16.9% non-white, 38.0% leukemia, 19.5% lymphoma, 10.3% CNS tumor, 32.2% other solid tumor) had an average of 10.0 (95% CI 9.7-10.3) conditions by 30 years of age and had a mean score on the HEI of 57.9±12.2 (out of 100 where the population average is 59). Scores on the HEI were not associated with disease burden z-score. Compared to those in the lowest quartile, survivors who consumed in the highest quartile of dark green vegetables (OR 0.77, 95% CI 0.61-0.97), legumes (OR 0.74, 95% CI 0.59-0.92), and whole grains (OR 0.75, 95% CI 0.60-0.95) were less likely to have a disease burden z-score > 0, whereas those who consumed in the highest quartile of white potatoes (OR 1.31, 95% CI 1.05-1.64), refined grains (OR 1.26, 95% CI 1.01-1.58), and added sugars (OR: 1.27, 95% CI 1.01-1.56) were more likely to have a disease burden z-score > 0. Conclusions: Specific components of diet are associated with chronic disease burden in long term survivors of cancer. Dietary interventions may have potential to decrease chronic disease burden in this population.

Effects of metabolic syndrome on cognitive outcomes in long-term survivors of childhood cancer

12013

Background: Childhood cancer therapy increases risk for cognitive impairment and other chronic conditions, which also may impact cognition. We assessed the unique impact of metabolic syndrome (MetS) on cognition in survivors participating in the St. Jude Lifetime Cohort Study. Methods: Participants included 4058 survivors of childhood cancer (53.9% female; mean [SD] age 30.1 [10.5] years at evaluation; 22.6 [10.1] years from diagnosis) who completed clinical evaluation and cognitive testing. MetS criteria followed Adult Treatment Panel III guidelines (at least 3 of: hypertension, high triglycerides, abdominal obesity, low high-density lipoprotein [HDL], high fasting glucose). Multivariable log-binomial regression models assessed risk of cognitive impairment associated with MetS stratified by survivors who did (n = 2301) or did not (n = 1757) receive central nervous system (CNS)-directed therapy. Mediation analysis assessed effects of MetS and physical activity in cranial radiotherapy (CRT)-associated cognitive impairment. Models were adjusted for age, sex, follow-up time and treatment exposures. Results: MetS was associated with increased risk of impaired attention (relative risk [RR] 1.34 95% confidence interval [CI] 1.07-1.66), processing speed (RR 1.25 CI 1.11-1.41) and executive function (RR 1.18 CI 1.01-1.37) in survivors with CNS-directed therapy and academic achievement (RR 1.84 CI 1.18-2.89), attention (RR 1.43 CI 1.10-1.87), and processing speed (RR 1.46 CI 1.21-1.75) in those without CNS-directed therapy. MetS components associated with cognitive impairment included abdominal obesity (memory RR 1.34 CI 1.13-1.59; processing speed RR 1.41 CI 1.24-1.59; executive function RR 1.21 CI 1.05-1.39) and low HDL (intelligence RR 1.26 CI 1.06-1.49; attention RR 1.27 CI 1.03-1.57; processing speed RR 1.17 CI 1.01-1.35; executive function RR 1.20 CI 1.05-1.37) in survivors with CNS-directed therapy. In survivors treated without CNS-directed therapy hypertension (academic achievement RR 1.49 CI 1.18-1.88; intelligence RR 1.34 CI 1.02-1.76; attention RR 1.42 CI 1.12-1.79; memory RR 1.45 CI 1.14-1.84; processing speed RR 1.30 CI 1.08-1.55; executive function RR 1.32 CI 1.08-1.62) and abdominal obesity (academic achievement RR 1.71 CI 1.07-2.72; processing speed RR 1.23 CI 1.02-1.49; executive function RR 1.38 CI 1.09-1.75) were associated with impairment. In mediation analyses, direct effects of CRT were identified, as were indirect effects through physical activity (processing speed β = 0.035 p < 0.01; attention β = 0.03 p < 0.01; executive function β = 0.172 p < 0.01). Conclusions: MetS increases risk of cognitive impairment in survivors, particularly abdominal obesity, hypertension and low HDL. Physical activity appears to partially mediate impact of CRT on cognitive outcomes and is an important target for interventions to lower impairment risk.

Financial hardship in adult survivors of childhood cancer: A report from the Childhood Cancer Survivor Study (CCSS)

10026

Background: The impact of childhood and adolescent cancer on the long-term financial outcomes of survivors is poorly understood. We compared financial hardship between survivors and siblings enrolled in the CCSS and identified survivors at elevated risk. Methods: Survivors treated for cancer at age < 21 years in 1970-1999 and siblings responded to a survey (23 binary-response questions) at age ≥26 years administered in 2017-2019. Principal component analysis with promax rotation extracted 3 factors with eigenvalues > 1 and KR-20 reliability coefficients > 0.7, retaining items with factor loadings > 0.4. These factors were behavioral hardship (8 items, e.g., forgone needed medical care), material hardship/financial sacrifices (8 items, e.g., problems paying medical bills) and psychological hardship (3 items, e.g., worry about having enough money to pay rent/mortgage). Factor scores were calculated by adding the item responses and dividing by their standard deviation. Multiple linear regression examined the association of sociodemographic and cancer treatment variables with factor scores. Results: Among 3349 survivors (49% male; median age [range] 40.2 [26.0-67.4] years) and 976 siblings (42% male, median age 46.5 [ 26.1-69.2] years), survivors were more likely to report being sent to debt collection (29.5 vs 21.4%), problems paying medical bills (20.0 vs 11.9%), foregoing needed medical care (13.3 vs 7.7%) and worry/stress about paying their mortgage (32.8 vs 23.2%) or having enough money to buy nutritious meals (25.0 vs 16.2%), all P < 0.001. Survivors reported greater hardship than siblings on all 3 factors: behavioral hardship (standardized mean score 0.51 vs 0.36), material hardship/financial sacrifices (0.63 vs 0.44), psychological hardship (0.69 vs 0.44), all P < 0.001. Behavioral hardship was increased by female gender (regression coefficient [ꞵ] 0.17, 95% CI 0.10-0.25), < high school (ꞵ 0.45, CI 0.12-0.79) or < college (ꞵ 0.18, CI 0.09-0.26) education, no (ꞵ 1.14, CI 0.93-1.35) or public (ꞵ 0.23, CI 0.10-0.35) health insurance, being divorced/separated (ꞵ 0.28, CI 0.10-0.46) and ≥250mg/m2 anthracycline chemotherapy (ꞵ 0.09, CI 0.00-0.19). The same variables were significantly associated with the other two hardship factors, but total body irradiation and cranial radiation also contributed to the risk of material hardship/financial sacrifices, and ≥8g/m2 cyclophosphamide equivalent dose and cranial radiation contributed to psychological hardship. Conclusions: Survivors of childhood and adolescent cancer are at elevated risk for financial hardship as compared to sibling controls. Those at highest risk can be defined using a combination of sociodemographic and treatment variables. This information can be used to inform targeted intervention strategies to reduce the risk of poor financial outcomes in this vulnerable population.

Neurocognitive outcomes in survivors of early adolescent and young adult (eAYA) hematologic cancers from the Childhood Cancer Survivor Study (CCSS)

10029

Background: Neurocognitive impairment in eAYA hematologic cancer survivors has not been well described, despite intensive neurotoxic therapies. We examined prevalence and risk for such impairment in hematologic cancer survivors diagnosed during eAYA compared to a younger age. Methods: We identified 1,213 eAYA (diagnosed at 15-21 years; median [range] follow-up age 40 [30-54]) and 4,538 childhood (diagnosed at <15 years; median age 30 [17-48]) survivors of ALL (n= 301 vs 3274), AML (n= 77 vs 424), and Hodgkin lymphoma (HL; n= 835 vs 840) from the CCSS (diagnosed 1970-1999) who completed the Neurocognitive Questionnaire. Impairment was defined as a score >90% of normative data in task efficiency (TE), organization (Org), memory (Mem), and emotional regulation (ER) domains. 1,014 age-matched siblings were controls. Treatment by diagnosis group, chronic health conditions, health status and health behaviors were examined as risk factors for neurocognitive impairment using multivariable logistic regression. Adjusted odds ratios (ORs) and corresponding 95% CI are reported. Results: Prevalence of neurocognitive impairment (≥1 impaired domain) was similar for eAYAs and childhood survivors of HL (31.0% vs 29.6%, p=0.54) and AML (36.4% vs 40.3%, p=0.51), although eAYA AML survivors were more likely to have impaired Mem (OR=2.3, 95% CI 1.0-5.4). eAYA ALL survivors were less likely to have neurocognitive impairment than childhood ALL survivors (28.2% vs 38.5%, p<.001) due to lower risk for impaired TE (OR=0.7, 95% CI 0.4-1.0) and Org (OR=0.5, 95% CI 0.4-0.9). No factors, including cranial radiation (RT), explained the rate differences. Treatment by diagnosis group (including cranial RT in ALL, chest RT in HL, and salvage therapy use) was not consistently associated with neurocognitive impairment in eAYA survivors. However, anthracycline dose ≥120mg/m2 was a risk factor for impaired ER (OR=6.0, 95% CI 2.0-17.9) only in eAYA ALL survivors. Presence of a neurologic health condition was associated with impairment in all 4 domains in eAYA (ORs ranged 1.7-2.9) and childhood cancer survivors (ORs ranged 1.9-5.3). eAYA survivors with a respiratory condition were more likely to have impaired TE (OR=2.1, 95% CI 1.2-3.8). Being in good general health was associated with less impairment across all 4 domains for eAYA (ORs ranged 0.2-0.4) and childhood survivors (ORs ranged 0.3-0.5). eAYA survivors who never smoked were less likely to have impaired ER (OR=0.4, 95% CI 0.2-0.6) than those who smoke. Conclusions: Survivors of hematologic cancers diagnosed during eAYA are susceptible to neurocognitive impairment at rates similar to those diagnosed at younger ages. Having comorbidities and being in fair/poor general health are risk factors for impairment. Higher anthracycline exposure in ALL survivors diagnosed during eAYA was the only therapy associated with impairment rates.

Health-related unemployment trends among survivors of childhood cancer: A report from the Childhood Cancer Survivor Study (CCSS)

10048

Background: The impact of treatment era and chronic health conditions on health-related unemployment among childhood cancer survivors has not been studied. Methods: Childhood cancer survivors (age ≥25 years) enrolled in the CCSS (3,420 diagnosed in the 1970s, 3,564 in the 1980s, and 2,853 in the 1990s) were matched 1:5 on sex, race/ethnicity, census bureau division, age, and year of survey to the Behavioral Risk Factor Surveillance System (BRFSS), a nationally representative population. Among survivors, health-related unemployment was defined as self-reported unemployment due to illness/disability and for BRFSS participants as self-reported inability to work. To standardize follow-up, health-related unemployment was assessed either in 2002-05 or 2014-16 for both cohorts. Sex stratified standardized prevalence ratio (SPR) and relative SPR (rSPR) with 95% confidence intervals (CI) for health-related unemployment were estimated using multivariable generalized linear models, with BRFSS background rates to assess the impact of treatment era and moderate to severe health conditions (per the Common Terminology Criteria for Adverse Events). Results: Prevalence of health-related unemployment in survivors (median age 9 years [range 0-20] at diagnosis and 33 years [25-54] at follow-up) was significantly higher compared to BRFSS participants (females: 11.3% vs 3.7%; SPR 3.0, 95% CI 2.7-3.3; males: 10.5% vs 3.0%; SPR 3.5, 95% CI 3.1-3.9). Health-related unemployment risks declined among survivors in more recent decades (ptrend< 0.001) for females: 1970s SPR 3.8, 95% CI 3.2-4.5, 1980s SPR 2.9, 95% CI 2.5-3.5, 1990s SPR 2.5, 95% CI 2.1-3.0; and males: 1970s SPR 3.6, 95% CI 2.9-4.4, 1980s SPR 3.8, 95% CI 3.1-4.7, 1990s SPR 3.0, 95% CI 2.5-3.7. Among survivors, multivariable models identified associations between presence of specific health conditions and elevated health-related unemployment (Table) adjusting for all statistically significant health conditions, race/ethnicity, treatment era, age at survey, and diagnosis. Among females, rSPR for endocrine conditions differed between 1970s and 1990s (interaction p = 0.04); fewer significant health conditions remained in the final model for males. Conclusions: While prevalence for health-related unemployment has declined over time, childhood cancer survivors remain at higher risk compared to the general population. These elevated risks are associated with chronic health conditions and affect female survivors more than male survivors.[Table: see text]

Accelerated cognitive decline in adult survivors of pediatric central nervous system (CNS) tumors: A report from the Childhood Cancer Survivor Study (CCSS)

10049

Background: Survivors of pediatric CNS tumors may be at elevated risk for accelerated cognitive decline as they age through adulthood relative to the general population, which may be an early risk factor for dementia. Methods: Longitudinal analysis of 512 CNS tumor survivors (52.3% female, mean [SD] 30.6 [7.1] years at T1) and 232 siblings (57.8% female, mean [SD] 34.2 [8.4] years at T1) from the CCSS was conducted using the Neurocognitive Questionnaire (NCQ) to assess task efficiency, emotional regulation, organization and memory at two timepoints separated by a mean of 11.6 [0.7] years. Impairment in each NCQ domain was defined as a score ≥ 90th percentile of the CCSS sibling distribution at each survey, with decline defined as moving from unimpaired at T1 to impaired at T2. Treatment exposures were abstracted from medical records. Chronic health conditions were self-reported at T1 and graded according to CTCAE v4.3. Relative risk of decline for group, treatment and chronic condition predictors was estimated using generalized linear models with robust variance estimates. Mediation analysis examined direct effects of treatments and mediating effects of chronic conditions. All models were adjusted for age, sex, and race. Results: At T1, survivors demonstrated higher frequency of impaired memory (24.5% vs. 6.5%, p < 0.001), emotional regulation (14.3 % vs. 5.6%, p < 0.001), task efficiency (43.3% vs. 13.8%, p < 0.001) and organization (17.7% vs. 10.8%, p = 0.015) than siblings. Among those unimpaired at T1, more survivors vs. siblings declined in memory (34.7% vs. 7.8; RR 4.2, 95% CI 2.6-6.9), emotional regulation (15.5% vs. 5.0%; RR 2.8, 95% CI 1.5-5.3), task efficiency (22.7% vs. 7.0%; RR 2.9, 95% CI 1.7-5.2), and organization (14.5% vs. 2.9%; RR 4.9, 95% CI 2.1-11.0) by T2. Decline in survivor memory was associated with exposure to craniospinal irradiation (RR 1.9, 95% CI 1.3-2.8) and focal irradiation (RR 1.6, 95% CI 1.1-2.3) compared with no radiation, and exposure to Ara-C (RR 1.7, 95% CI 1.0-2.8) and cyclophosphamide (RR 1.7, 95% CI 1.01-2.8). Independent of therapy, serious/disabling or life-threatening cardiopulmonary conditions at T1 predicted future decline in memory (RR 1.5, 95% CI 1.02-2.2) and organization (RR 2.0, 95% CI 1.1-3.6), with the presence of 2 or more cardiopulmonary conditions associated with even higher risk (memory RR 2.6, 95% CI 2.0-3.1; organization RR 3.4, 95% CI 1.1-10.5). Chronic conditions did not mediate associations between treatment exposures and cognitive decline. Conclusions: CNS tumor survivors are at elevated risk for impairment and accelerated cognitive decline compared to siblings. Cranial radiation, Ara-C, cyclophosphamide, and cardiopulmonary morbidity are risk factors for decline. Survivors with these exposures/conditions may benefit from interventions to prevent additional future cognitive decline.

Mortality among five-year survivors of childhood cancer: Results over five decades of follow-up in the Childhood Cancer Survivor Study

10013

Background: Adult survivors of childhood cancer are at greater risk for late mortality compared to the general population due to cancer and its treatment. Risk factors, patterns and specific causes of late mortality across the lifespan are not well established. Methods: All-cause, cause-specific, and health-related late mortality (HRM; excludes death from primary cancer and external causes) > 5 years from diagnosis were evaluated in survivors diagnosed < 21 years of age between 1970-1999. Cause of death was based on ICD codes from the National Death Index through December 2017. Cumulative mortality, mortality rates and standardized mortality ratios (SMRs) with 95% confidence intervals (CIs) were estimated, overall and in 5- and 10-year survival periods. Results: Among 34,230 survivors (median time from diagnosis 29.1 years, range 5.0 - 48.0) the 40-year cumulative mortality was 23.3% (95% CI 22.7 - 24.0). Of 5,916 deaths, 3,061 (51.2%) were attributable to health-related causes including subsequent neoplasm (n = 1,458), cardiac (n = 504), and pulmonary causes (n = 238). All-cause mortality by time from diagnosis demonstrated a U-shaped distribution: 10.1 deaths/1000 person-years at 5-9 years, largely due to recurrence of the primary cancer, decreasing to 4.1 at 15-19 years before increasing to 18.5 at 40-48 years, attributable to an increasing mortality rate from HRM (2.3 at 5-9 years; 17.0 at 40-48 years). For the interval 5-9 years from diagnosis, survivors had an 18.1-fold (95% CI 17.3-18.9) higher risk of death from any cause, and a 13.1-fold (11.9-13.4) higher risk for HRM when compared to the general population. Although the SMRs declined with duration of follow-up, survivors had a 4-fold higher risk of death overall, attributable to a more than 4-fold increased risk of HRM. HRM 40-48 years from diagnosis was largely attributable to an increased risk of death due to subsequent neoplasm (SMR 6.0, 95% CI 4.9-7.2), cardiac (3.9, 2.9-5.0) and pulmonary (5.6, 3.6-8.4) causes. Cause-specific mortality remained markedly elevated at 40-48 years from diagnosis: CNS malignancy (SMR 11.7, 95% CI 5.4-22.3), benign meningioma (171.3, 34.4-500.5), valvular heart disease (39.8, 21.2-68.1), cardiomyopathy (10.4, 4.5-20.5), stroke (7.9, 4.6-12.6), and renal failure (5.6, 1.8-13.2). HRM was significantly higher among the youngest group of survivors (0-4 years at diagnosis), non-Hispanic blacks and those who received radiation to the brain, chest or total body, or who were exposed to anthracycline, alkylating or platinum chemotherapy. Conclusions: After five decades, aging survivors consistently remain at higher risk of all-cause mortality compared to the general, aging population, primarily due to a persistent 4-fold increased risk of HRM. Continued late-effects surveillance and reduction of therapies associated with long-term morbidity and increased mortality is essential.

Accelerated aging and mortality in long-term survivors of childhood cancer: A report from the St. Jude Lifetime Cohort (SJLIFE)

10045

Background: Survivors of childhood cancer have functional limitations and health-related morbidity consistent with an accelerated aging phenotype. We characterized aging using a Deficit Accumulation Index (DAI) which examines the accumulation of multiple aging-related deficits readily available from medical records and self-report. DAI’s are used as surrogates of biologic aging and are validated to predict mortality in adult cancer patients. Methods: We included childhood cancer survivors (N = 3,758, mean age 30 [SD 8], 22 [9] years post diagnosis, 52% male) and community controls (N = 575, mean age 34 [10] 44% male) who completed clinical assessments and questionnaires and who were followed for mortality through December 31st, 2018 (mean follow-up 6.1 [3.1] years). Using the initial SJLIFE clinical assessment, a DAI score was generated as the proportion of deficits out of 44 items related to aging, including chronic conditions (e.g. hearing loss, hypertension), psychosocial and physical function, and activities of daily living. The total score ranged 0 to 1; scores > 0.20 are robust, while moderate and large clinically meaningful differences are 0.02 and 0.06, respectively. Linear regression compared the DAI in survivors and controls with an age*survivor/control interaction and examined treatment associations in survivors. Cox-proportional hazards models estimated risk of death associated with DAI. All models were adjusted for age, sex, and race. Results: Mean [SD] of DAI was 0.17 [0.11] for survivors and 0.10 [0.08] for controls. 32% of survivors had a DAI above the 90th percentile of the control distribution (p < 0.001). After adjustment for covariates, survivors had a statistically and clinically meaningfully higher DAI score than controls (β = 0.072 95%CI 0.062, 0.081; p < 0.001). When plotted against age, the adjusted DAI at the average age of survivors (30 years) was 0.166 (95% CI 0.160,0.171), which corresponded to 60 years of age in controls, suggesting premature aging of 30 years. The mean difference in DAI between survivors and controls increased with age from 0.06 (95% CI 0.04, 0.07) at age 20 to 0.11 (95% CI 0.08, 0.13) at age 60, consistent with an accelerated aging phenotype (p = 0.014). Cranial radiation, abdominal radiation, cyclophosphamide, platinum agents, neurosurgery, and amputation were each associated with a higher DAI (all p≤0.001). Among survivors, a 0.06 increase in DAI was associated with a 41% increased risk of all-cause mortality (HR 1.41 95%CI 1.32, 1.50; p < 0.001). Conclusions: Survivors of childhood cancer experience significant age acceleration that is associated with an increased risk of mortality; longitudinal analyses are underway to validate these findings. Given the ease of estimating a DAI, this may be a feasible method to quickly identify survivors for novel and tailored interventions that can improve health and prevent premature mortality.

Determinants of symptom clusters and associations with health outcomes in childhood cancer survivors: A report from the St. Jude Lifetime Cohort (SJLIFE)

10046

Background: Childhood cancer survivors experience concurrent symptoms, but associations with health outcomes are unknown. We characterize symptom clusters among adult survivors of childhood cancer in SJLIFE and tests associations with health-related quality of life (HRQL) and clinically assessed physical and neurocognitive performance. Methods: This cross-sectional study includes survivors diagnosed when <18 years of age, ≥10 years off-therapy, and ≥18 years of age at evaluation. Survivors rated 37 symptoms over 10 domains (cardiac, pulmonary, sensory, motor, nausea, pain, fatigue, memory, anxiety, depression), representing 3 broader symptom groups (physical, somatic, psychological). They also underwent a rating of HRQL (SF-36 PCS/MCS) and testing of physical performance (quantitative sensory, motor, endurance, mobility) and neurocognition (processing speed, executive function, attention, memory problems). Latent class analysis determined survivors with distinct symptom burden. Polytomous logistic regression identified risk factors of symptom clusters; multivariable regression tested associations of symptom clusters with health outcomes. Results: Among 3,085 survivors, mean [SD] age at evaluation was 31.9 [8.3] years, time from diagnosis was 28.1 [9.1] years, 49.7% were female, 37.1% were treated for leukemia and 33.0% for solid tumors. Four groups of survivors with distinct symptom burden were found: Cluster 1 (52%, low prevalence in all 3 symptom groups); Cluster 2 (16%, low in physical, moderate in somatic, high in psychological); Cluster 3 (18%; high in physical, moderate in somatic, low in physiological); and Cluster 4 (14%, high in all 3 symptom groups). Compared to the lowest symptom burden (Cluster 1), survivors with highest burden (Cluster 4) were significantly more likely to be female (OR 2.5; 95%CI 1.9, 3.4), have below a high school education (OR 7.7; 95%CI 4.5, 13.3), no insurance (OR 1.5; 95%CI 1.1, 2.3) and previous exposure to corticosteroids (OR 1.8; 95%CI 1.0, 3.0). High physical, moderate somatic and low psychological symptom burden (Cluster 3) was associated with below high school education (OR 2.7; 95%CI 1.4, 5.0), exposure to platinum agents (OR 2.2; 95%CI 1.4, 3.7) and brain radiation ≥30Gy (OR 4.0; 95%CI 2.3, 6.9) in contrast to Cluster 1. Survivors in Cluster 4 had the poorest PCS, MCS, physical and neurocognitive outcomes vs in Clusters 2 or 3, whereas those in Cluster 1 had the best outcomes (F-values for 4 clusters: 291.4 [PCS], 269.2 [MCS], 61.5 [physical], 36.9 [neurocognitive], p-values <0.001; effect sizes for Clusters 4 vs 1: 0.4-2.0 [4 outcomes]). Conclusions: Nearly 50% of survivors belong to symptom clusters with ≥1 moderate/high burden groups, associated with the socio-demographic and treatment exposures. Survivors in the highest symptom burden cluster had the poorest HRQL and functional outcomes.

Predicting decreased health-related quality of life (HRQL) in adult survivors of childhood cancer: A report from the Childhood Cancer Survivor Study (CCSS)

10043

Background: This study examines temporal patterns in HRQL among adult survivors of childhood cancer, and socio-demographic, lifestyle and health status predictors of decline in HRQL. Methods: Adult survivors of childhood cancer (4755, 55.2% female, 86.9% non-Hispanic white) completed baseline (T0) and follow-up (T1 in 2003, T2 in 2014) surveys (median[SD] age 32.4[7.5] at T1, time since diagnosis to T1 23.0[4.5], T1-T2 interval 11.7[0.6] years). Socio-demographic (e.g., age, sex, educational attainment, annual family income), lifestyle (physical inactivity, smoking) and health status predictors were collected at T0 and T1. Chronic conditions graded ≥2 by CTCAE defined as presence, and mental and cognitive status with ≥1SD from norms defined as poor. SF-36 Physical and Mental Component Summary (PCS/MCS; mean 50/SD 10) at T1 and T2 classified HRQL as optimal (≥40) or suboptimal ( < 40). Multivariable logistic regression identified risk factors (T0, T1 or status change T0-T1) of decreased HRQL (i.e., optimal to suboptimal) using a backward selection method (p < 0.1), adjusting for sex, race, age at T1 and years between T1-T2. The sample was randomly split into training (80%) and test (20%) datasets to develop and validate prediction models; Area Under the ROC Curve (AUC) evaluated model performance. Results: From T1-T2, 8.1% and 8.3% of survivors reported decreased PCS and MCS. AUCs of training/test models were 0.75/0.74 for decreased PCS and 0.72/0.68 for decreased MCS. Risk factors at T0 or T1 predicting decreased PCS included female sex (OR 1.67, 95%CI 1.25-2.24), younger age (OR 1.04, 95%CI 1.02-1.06), < college/vocational education (OR 1.59, 95%CI 1.02-2.46), family income < $20,000 (OR 2.00, 95%CI 1.21-3.30), obesity (OR 1.97, 95%CI 1.32-2.92), chronic health conditions (neurologic OR 2.47, 95%CI 1.69-3.60; musculoskeletal OR 2.27, 95%CI 1.42-3.64; endocrinological OR 2.25, 95%CI 1.44-3.52; gastrointestinal OR 1.89, 95%CI 1.32-2.69; pulmonary OR 1.66, 95%CI 1.06-2.59; cardiovascular OR 1.53, 95%CI 1.14-2.06) and depression (OR 1.79, 95%CI 1.20-2.67). Risk factors at T0 or T1 predicting decreased MCS included unemployment (OR 1.68, 95%CI 1.19-2.38), smoking (OR 2.03, 95%CI 1.37-3.00), physical inactivity (OR 1.48, 95%CI 1.05-2.09), poor mental health (depression OR 4.29, 95%CI 2.44-7.55; somatization OR 1.63, 95%CI 1.05-2.53) and poor cognitive status (task efficiency OR 1.90, 95%CI 1.34-2.68; organization OR 1.67, 95%CI 1.12-2.48). Conclusions: Nearly 10% of childhood cancer survivors have significant late-onset decline in HRQL. Chronic health conditions predict decreased physical HRQL, whereas smoking, physical inactivity and poor mental health predict decreased mental HRQL. Interventions targeting modifiable lifestyle and health conditions should be considered to prevent decreased HRQL for childhood cancer survivors.

Late health outcomes in survivors of Wilms tumor: A report from the St. Jude Lifetime (SJLIFE) cohort study

10038

Background: We aimed to characterize late health, neurocognitive, and physical performance outcomes among survivors of Wilms tumor. Methods: Wilms tumor survivors (n = 280), ≥ 5 years from diagnosis, participating in SJLIFE were clinically assessed along with a community control sample (n = 625). Health outcomes were graded per a modified Common Terminology Criteria for Adverse Events (grade 1 [mild] to grade 4 [life threatening]). Standardized neurocognitive testing was graded using age-adjusted z-scores. Aerobic function (six-minute walk), mobility (timed up and go) and flexibility (sit and reach) were assessed. Associations between treatment exposures and prevalent conditions were examined by multivariable logistic regression, adjusted for current age, sex and race. Results: Survivors (59% female, 73% white), median age 3 years (range 0-15) at diagnosis and 31 years (9-58) at evaluation, were comprehensively evaluated on the St. Jude campus. Two-thirds (67%) were treated with doxorubicin (median dose 175 mg/m2 range 52-490), 167 (60%) received abdominal radiation (median dose 12 Gy range 8.8-61.2) and 25% received chest radiation (12 Gy range 9-44). By age 40 years, survivors averaged 12.7 (95% confidence interval [CI] 11.7-13.8) grade 1-4 and 7.5 (CI 6.7-8.2) grade 2-4 conditions, compared to 4.2 (CI 3.9-4.6) and 2.3 (CI 2.1-2.5), respectively, among controls. The most prevalent medical conditions (grade ≥ 2) are reported in the table. Abnormal glucose metabolism was associated with abdominal radiation (relative risk [RR] 5.1 CI 1.4-19.0); restrictive pulmonary defects with chest radiation (RR 24.0 CI 3.2-180); and cardiomyopathy (RR 15.6 CI 1.9-128), pulmonary diffusion (RR 4.5 CI 1.3-15.1), and chronic kidney disease (RR 4.5 CI 1.3-16.1) with doxorubicin exposure. Survivors had a three-fold higher risk (standardized incidence ratio 3.5, CI 2.2-6.6) for subsequent neoplasms. Impairments (grade ≥ 2) in executive function (20% vs. 12%), attention (17% vs. 9%), memory (21% vs. 10%), and processing speed (20% vs. 8%) were more frequent in survivors than controls (p < 0.05). Impairments in aerobic function (13.6%), mobility (13.6%), and flexibility (11.1%) were higher than expected (p < 0.01). Significant associations were not identified between treatment exposures and neurocognitive or physical performance outcomes. Conclusions: Systematic clinical assessment identified a significant burden of chronic health conditions and previously unrecognized neurocognitive and physical performance limitations in survivors of Wilms tumor.[Table: see text]

Pain and functional outcomes in adult survivors of childhood cancer: A report from the St. Jude Lifetime Cohort study

BACKGROUND

Although survivors of childhood cancer are at risk of chronic pain, the impact of pain on daily functioning is not well understood.

METHODS

A total of 2836 survivors (mean age, 32.2 years [SD, 8.5 years]; mean time since diagnosis, 23.7 years [SD, 8.2 years]) and 343 noncancer community controls (mean age, 35.5 years [SD, 10.2 years]) underwent comprehensive medical, neurocognitive, and physical performance assessments, and completed measures of pain, health-related quality of life (HRQOL), and social functioning. Multinomial logistic regression models, using odds ratios and 95% confidence intervals (95% CIs), examined associations between diagnosis, treatment exposures, chronic health conditions, and pain. Relative risks (RRs) between pain and neurocognition, physical performance, social functioning, and HRQOL were examined using modified Poisson regression.

RESULTS

Approximately 18% of survivors (95% CI, 16.1%-18.9%) versus 8% of controls (95% CI, 5.0%-10.9%) reported moderate to very severe pain with moderate to extreme daily interference (P < .001). Severe and life-threatening chronic health conditions were associated with an increased likelihood of pain with interference (odds ratio, 2.03; 95% CI, 1.62-2.54). Pain with daily interference was found to be associated with an increased risk of impaired neurocognition (attention: RR, 1.88 [95% CI, 1.46-2.41]; and memory: RR, 1.65 [95% CI, 1.25-2.17]), physical functioning (aerobic capacity: RR, 2.29 [95% CI, 1.84-2.84]; and mobility: RR, 1.71 [95% CI, 1.42-2.06]), social functioning (inability to hold a job and/or attend school: RR, 4.46 [95% CI, 3.45-5.76]; and assistance with routine and/or personal care needs: RR, 5.64 [95% CI, 3.92-8.10]), and HRQOL (physical: RR, 6.34 [95% CI, 5.04-7.98]; and emotional: RR, 2.83 [95% CI, 2.28-3.50]).

CONCLUSIONS

Survivors of childhood cancer are at risk of pain and associated functional impairments. Survivors should be screened routinely for pain and interventions targeting pain interference are needed.

Association between obesity and neurocognitive function in survivors of childhood acute lymphoblastic leukemia treated only with chemotherapy

BACKGROUND

Neurocognitive impairment and obesity are common adverse sequelae in survivors of childhood acute lymphoblastic leukemia (ALL); however, the association has not been investigated.

METHODS

Neurocognitive function was evaluated once in survivors of ALL who were at least 8 years old and 5 years from their diagnosis. In a cross-sectional analysis, the associations with the body mass index (BMI) category and Z score were examined. A longitudinal analysis used the overweight/obesity area under the curve (AUC), which was determined via the trapezoidal rule by a sum of the integrals defined by the BMI Z score at each time point and the time intervals of the BMI measurement.

RESULTS

For 210 survivors, the median BMI Z score at diagnosis was 0.17, which increased to 0.54 at the end of induction and to 0.74 at the neurocognitive assessment. In the cross-sectional analysis, overweight/obese survivors scored significantly lower than others on the measures of executive function (cognitive flexibility, planning, verbal fluency, working memory, and spatial construction; all P < .05), attention (attention span and risk taking; all P < .05), and processing speed (visual motor coordination, visual speed, and motor speed; all P < .05). In the longitudinal analysis, when the treatment period was subdivided into 4 time periods (induction, consolidation, early maintenance, and late maintenance), a greater overweight/obesity AUC during induction therapy was associated with worse cognitive flexibility (P = .01) and slower motor speed (P = .02), which persisted throughout the treatment.

CONCLUSIONS

Overweight/obesity was significantly associated with neurocognitive impairment during long-term follow-up, and this association started early in treatment for ALL. Novel early interventions to provide cognitive training and prevent weight gain are required for patients at risk.

Persistent variations of blood DNA methylation associated with treatment exposures and risk for cardiometabolic outcomes in long-term survivors of childhood cancer in the St. Jude Lifetime Cohort

BACKGROUND

It is well-established that cancer treatment substantially increases the risk of long-term adverse health outcomes among childhood cancer survivors. However, there is limited research on the underlying mechanisms. To elucidate the pathophysiology and a possible causal pathway from treatment exposures to cardiometabolic conditions, we conducted epigenome-wide association studies (EWAS) to identify the DNA methylation (DNAm) sites associated with cancer treatment exposures and examined whether treatment-associated DNAm sites mediate associations between specific treatments and cardiometabolic conditions.

METHODS

We included 2052 survivors (median age 33.7 years) of European ancestry from the St. Jude Lifetime Cohort Study, a retrospective hospital-based study with prospective clinical follow-up. Cumulative doses of chemotherapy and region-specific radiation were abstracted from medical records. Seven cardiometabolic conditions were clinically assessed. DNAm profile was measured using MethylationEPIC BeadChip with blood-derived DNA.

RESULTS

By performing multiple treatment-specific EWAS, we identified 935 5'-cytosine-phosphate-guanine-3' (CpG) sites mapped to 538 genes/regions associated with one or more cancer treatments at the epigenome-wide significance level (p < 9 × 10-8). Among the treatment-associated CpGs, 8 were associated with obesity, 63 with hypercholesterolemia, and 17 with hypertriglyceridemia (false discovery rate-adjusted p < 0.05). We observed substantial mediation by methylation at four independent CpGs (cg06963130, cg21922478, cg22976567, cg07403981) for the association between abdominal field radiotherapy (abdominal-RT) and risk of hypercholesterolemia (70.3%) and by methylation at three CpGs (cg19634849, cg13552692, cg09853238) for the association between abdominal-RT and hypertriglyceridemia (54.6%). In addition, three CpGs (cg26572901, cg12715065, cg21163477) partially mediated the association between brain-RT and obesity with a 32.9% mediation effect, and two CpGs mediated the association between corticosteroids and obesity (cg22351187, 14.2%) and between brain-RT and hypertriglyceridemia (cg13360224, 10.5%). Notably, several mediator CpGs reside in the proximity of well-established dyslipidemia genes: cg21922478 (ITGA1) and cg22976567 (LMNA).

CONCLUSIONS

In childhood cancer survivors, cancer treatment exposures are associated with DNAm patterns present decades following the exposure. Treatment-associated DNAm sites may mediate the causal pathway from specific treatment exposures to certain cardiometabolic conditions, suggesting the utility of DNAm sites as risk predictors and potential mechanistic targets for future intervention studies.

Negligible Effects of the Survey Modes for Patient-Reported Outcomes: A Report From the Childhood Cancer Survivor Study

PURPOSE

This study compared the measurement properties for multiple modes of survey administration, including postal mail, telephone interview, and Web-based completion of patient-reported outcomes (PROs) among survivors of childhood cancer.

METHODS

The population included 6,974 adult survivors of childhood cancer in the Childhood Cancer Survivor Study who completed the Brief Symptom Inventory-18 (BSI-18), which measured anxiety, depression, and somatization symptoms. Scale reliability, construct validity, and known-groups validity related to health status were tested for each mode of completion. The multiple indicators and multiple causes technique was used to identify differential item functioning (DIF) for the BSI-18 items that responded through a specific survey mode. The impact of the administration mode was tested by comparing differences in BSI-18 scores between the modes accounting for DIF effects.

RESULTS

Of the respondents, 58%, 27%, and 15% completed postal mail, Web-based, and telephone surveys, respectively. Survivors who were male; had lower education, lower household income, or poorer health status; or were treated with cranial radiotherapy were more likely to complete a telephone-based survey compared with either a postal mail or Web-based survey (all P < .05). Scale reliability and validity were equivalent across the 3 survey options. One, 2, and 5 items from the anxiety, depression, and somatization domains, respectively, were identified as having significant DIF among survivors who responded by telephone (P < .05). However, estimated BSI-18 domain scores, especially depression and anxiety, between modes did not differ after accounting for DIF effects.

CONCLUSION

Certain survivor characteristics were associated with choosing a specific mode for PRO survey completion. However, measurement properties among these modes were equivalent, and the impact of using a specific mode on scores was minimal.

Longitudinal Trajectories of Neurocognitive Functioning in Childhood Acute Lymphoblastic Leukemia

OBJECTIVE

Children with acute lymphoblastic leukemia (ALL) are at risk for neurocognitive deficits, and examining individual variability is essential to understand these risks. This study evaluated latent longitudinal trajectories and risk factors of neurocognitive outcomes in childhood ALL.

METHODS

There were 233 participants with ALL who were enrolled on a phase 3, risk-stratified chemotherapy-only clinical trial (NCT00137111) and who completed protocol-directed neurocognitive assessments [47.6% female, mean (SD) = 6.6 (3.7) years]. Measures of sustained attention, learning/memory, and parent ratings of attention were completed during and after treatment. Longitudinal latent class analyses were used to classify participants into distinct trajectories. Logistic regression was used to identify predictors of class membership.

RESULTS

Within the overall group, attention performance was below age expectations across time (Conners Continuous Performance Test detectability/variability, p < 0.01); memory performance and parent ratings were below expectations at later phases (California Verbal Learning Test learning slope, p < 0.05; Conners Parent Rating Scale, Revised attention/learning, p < 0.05). Most participants (80-89%) had stable neurocognitive profiles; smaller groups showed declining (3-6%) or improving (3-11%) trajectories. Older age (p = 0.020), female sex (p = 0.018), and experiencing sepsis (p = 0.047) were associated with greater attention problems over time. Lower baseline IQ was associated with improved memory (p = 0.035) and fewer ratings of attention problems (p = 0.013) over time.

CONCLUSIONS

Most patients with ALL have stable neurocognitive profiles. Smaller groups have significant impairments shortly after diagnosis or have worsening performance over time. A tiered assessment approach, which includes consideration of individual and clinical risk factors, may be useful for monitoring neurocognitive functioning during treatment and survivorship.

Change in Pain Status and Subsequent Opioid and Marijuana Use Among Long-Term Adult Survivors of Childhood Cancer

We evaluated pain status change and associations with subsequent opioid/marijuana use among 1208 adult survivors of childhood cancer. Pain status and opioid/marijuana were self-reported at baseline and follow-up evaluation (mean interval = 4.2 years). Over time, 18.7% of survivors endorsed persistent/increasing significant pain; 4.8% and 9.0% reported having used opioids and marijuana at follow-up. Persistent/increased (vs none/decreased) pain, persistent/increased (vs none/decreased) anxiety, and lack of health insurance increased odds of subsequent opioid use by 7.69-fold (95% confidence interval [CI] = 3.71 to 15.95), 2.55-fold (95% CI = 1.04 to 6.24), and 2.50-fold (95% CI = 1.07 to 5.82), respectively. Persistent/increased (vs none/decreased) depression increased odds of subsequent marijuana use by 2.64-fold (95% CI = 1.10 to 6.33).