Epigenetic Age Acceleration and Neurocognitive Function Among Long-Term Survivors of Pediatric Hodgkin Lymphoma: A Report from the St. Jude Lifetime Cohort

Introduction: Long-term survivors of pediatric Hodgkin lymphoma (HL) are at elevated risk for cardiopulmonary morbidity, cognitive impairment, and premature mortality. This pattern of late effects suggests accelerated aging, which may signal risk for premature onset of dementia.⁷ The relation between biological aging and neurocognitive function has not been examined in survivors of pediatric cancer.

Methods: This study included 215 survivors of pediatric HL (50% female, mean[sd] 39[9] years old, 25[9] years from diagnosis) and 282 community controls (51% female, mean[sd] 36[10] years old) of European ancestry from the St. Jude Lifetime Cohort. Participants completed a comprehensive neuropsychological battery and provided a blood sample. Genome-wide methylation data were generated with peripheral blood mononuclear cell-derived DNA using Infinium Methylation EPIC BeadChip arrays. Epigenetic age was calculated according to Levin's clock and epigenetic age acceleration (EAA) was defined as the residual from regressing the epigenetic age on chronological age (higher EAA suggested older biological age relative to chronological age). Linear regression, adjusted for sex, compared EAA in HL and controls, and compared neurocognitive z-score in those in the third or second tertile of EAA to the first tertile, adjusted for sex, age at diagnosis, and high-dose methotrexate.

Results: Survivors of HL (mean EAA 95% confidence interval [CI] 64.2 (3.6, 4.8) had significantly greater EAA compared to community controls (control mean EAA 95%CI -3.5 (-4.1, -2.9); mean difference b =7.7 years, 95% CI 6.8 to 8.5; p<0.001). Among HL, EAA was associated with worse visual-motor processing speed, with those in the second tertile performing, on average, 0.42 standard deviations worse (b = -0.42, 95% CI -0.71 to -0.13, p=0.005) and those in the third tertile performing 0.55 standard deviations worse (b = -0.55, 95%CI -0.84 to -0.25, p<0.001) than HL survivors in the first tertile (Table 1). Those in the second and third tertiles also performed worse on short-term memory compared to those in the first tertile (second b = -0.42, 95% CI -0.74 to -0.10, p=0.011; third b = -0.59, 95% CI -0.90 to -0.27, p<0.001). The third tertile of EAA was also associated with worse performance in verbal learning (p=0.007) and long-term verbal recall (p=0.005). Lastly, those in the second or third tertiles performed 0.4 standard deviations worse, on average, than the first tertile on a measure of executive function (verbal fluency; second b = -0.40 95%CI -0.76 to -0.03 p=0.035; third b = -0.39, 95%CI -0.75 to -0.03, p=0.036).

Conclusions: Survivors of pediatric HL experience significant EAA that is associated with deficits in visual-motor processing speed, short- and long-term memory, and verbal fluency. EAA may be a useful biomarker to not only identify survivors at risk for accelerated cognitive aging but also to gauge preclinical response to interventions designed to improve cognitive function.

No relevant conflicts of interest to declare.