Association between dehydroepiandrosterone-sulfate and attention in long-term survivors of childhood acute lymphoblastic leukemia treated with only chemotherapy

Abstract

Long-term survivors of childhood acute lymphoblastic leukemia (ALL) are at risk for neurocognitive impairment, as well as compromised hypothalamic-pituitary-adrenal (HPA) function. Dehydroepiandrosterone-sulfate (DHEAS) is an adrenal androgen commonly used as a marker of HPA function. In the general population, a low level of DHEAS has been associated with poorer cognition. At ≥2years post-treatment, we examined the association of DHEAS with attention outcomes in 35 male and 34 female long-term survivors of childhood ALL (mean[standard deviation] age at evaluation 14.5[4.7] years; 7.5[1.9] years post-diagnosis) who were treated with only chemotherapy and without prophylactic cranial irradiation. Male survivors with low-normal levels of DHEAS had worse performance than male survivors with high levels of DHEAS on multiple measures of attention (all P's<0.05). However, association between DHEAS and attention measures were not found in female survivors. Our results suggest that survivors of ALL who suffer from partial but persistent adrenal insufficiency may be at risk for neurocognitive deficits. This finding should be validated in a larger prospective study, with attention to sex differences in the potential impact of adrenal insufficiency on neurocognitive outcomes.

Keywords: Attention; Chemotherapy; Childhood acute lymphoblastic leukemia; Dehydroepiandrosterone sulphate; Neurocognitive function.

Copyright © 2016 Elsevier Ltd. All rights reserved.

Figures

Figure 1. Association between Attention Scores and DHEAS

Caption: Attention performance is represented by the CPT raw scores; a higher score is indicative of worse attention performance. P denotes the statistical significance for comparing attention performance between survivors in the bottom tertile of DHEAS and survivors in other tertiles, adjusted for Tanner Stage, age at evaluation and time since diagnosis, using general linear modelling. Error bars represent the 95% confidence interval of the adjusted raw scores. Associations are tested in male and female survivors separately. Results show that male survivors in the bottom tertile of DHEAS level demonstrated worse attention performance than those in other tertiles. Association between DHEAS and attention measures were not found in female survivors. Rates of impairment (%) are presented above the error bars for each group. Impairment is defined as those with age-adjusted Z-scores falling below the bottom 10th percentile of the population norms.