Neurocognitive Decline Among Long-Term Survivors of Pediatric Hodgkin Lymphoma: A Report from the St. Jude Lifetime Cohort

Background: Survivors of pediatric Hodgkin lymphoma (HL) are at risk for persistent neurocognitive impairments. However, as existing data are cross-sectional, the risk of progressive neurocognitive decline with aging and dementia among survivors of HL remains unknown.

Methods: Participants included 238 long-term survivors (at least 5 years from diagnosis) of pediatric HL in the St. Jude Lifetime Cohort (mean[SD] current age 38[8] years, 23[8] years post-diagnosis). Participants completed objective neurocognitive testing (intelligence, attention, processing speed, memory) at two time points at least 1 year apart. Neurocognitive scores were converted into age-adjusted Z-scores (μ=0, σ=1.0) using national normative data. A reliable change index (RCI) was used to identify a clinically meaningful change while accounting for potential practice effects. This was calculated by subtracting the baseline score from the follow up score and dividing by the standard error of the difference (from a normative sample). Participants were defined as having a reliable decline [RCI≤ -1.64], reliable improvement [RCI≥1.64], and stable [-1.645 < RCI < 1.645]. Treatment history was abstracted from medical records and chronic health conditions (at baseline) were prospectively ascertained and systematically graded according to a modified version of the NCI Common Terminology Criteria for Adverse Events v4.03. Multinomial logistic regression examined the odds of a reliable decline or reliable improvement (relative to remaining stable [referent]) associated with treatment exposures or having a chronic health conditions (≥ grade 2).

Results: The mean time (SD) between assessments was 3.3 (1.4) years. Many survivors experienced a reliable decline (RD) on tests of sustained attention (24%) and attention span (16%). On measures of memory function, 21% experienced a RD in short-term memory while 13% experienced RD in verbal learning and long-term verbal recall. For measures of executive function, 16% experienced a RD in cognitive flexibility and approximately 10% experienced RD on tests of cognitive switching, working memory, and verbal fluency. Treatment exposures such as chest radiation, anthracyclines, and vincristine were not associated with an increased odds of experiencing a RD. High-dose methotrexate is a known risk factor for cognitive impairment, however the prevalence was too low (n=6) for analysis and instead we examined frequency of change groups. All participants treated with high-dose methotrexate remained stable on measures of attention span and verbal fluency, while half experience a RD on verbal fluency. The presence of any cardiovascular condition (≥ grade 2) was associated with a 2-fold increased odds of experiencing a RD, rather than remaining stable, on a test of attention span (OR 2.2 [95%CI 1.01, 4.8]; p=0.048). Neuropathy was associated with an increased odds of RD on cognitive switching (4.9 [1.4, 18.2]; p=0.016) and verbal fluency (3.5 [0.8, 13.6]; p=0.09).

Conclusions: Over a relatively brief interval, a subgroup of long-term survivors of pediatric HL experience significant neurocognitive decline in the domains of attention, memory, and executive function. Cardiovascular conditions and neuropathy were associated with this decline. These data suggest that management or prevention of chronic health conditions may help to preserve the neurocognitive functioning of long-term HL survivors. Future work is ongoing to identify other modifiable risk factors for cognitive decline.