QOS-32EFFECT OF SEIZURE MORBIDITY ON NEUROCOGNITIVE OUTCOME, QUALITY OF LIFE, AND SOCIAL ATTAINMENT IN ADULT SURVIVORS OF CHILDHOOD CANCERS

Abstract

Objective: To correlate seizure morbidity with functional outcomes in adult survivors of childhood cancer. Background: Adult survivors of childhood cancer are at risk for seizures and reduced functional outcomes. Direct impact of seizures on functional outcomes in cancer survivors has not been well explored. Methods: We examined seizure characteristics among 2,022 cancer survivors (48.3[percnt] female; median [range] age=31.5[18.4-65.9] years; time since diagnosis=23.6[10.4-51.1] years). Seizure frequency and seizure medications following cancer diagnosis were abstracted from medical records, and reviewed by board-certified neurologists. Survivors completed neurocognitive testing and surveys to assess health-related quality of life (HRQOL) and social attainment. Neurocognitive performance was converted to age-adjusted z-scores (M=0, SD=1.0) based on population norms. Multiple logistical regression and Poisson regression with robust variance were used to correlate seizure characteristics and functional outcomes, stratified by CNS tumor diagnosis and adjusting for age, gender, and prior cancer therapy. Results: Seizures following diagnosis/treatment were identified in 232 (11.5[percnt]) survivors, and were most frequent among survivors of CNS tumors (29.9[percnt]) and leukemia (12.6[percnt]) compared to lymphomas (4.1[percnt]) and solid tumors (4.5[percnt]). Seizures were persistent in 36.2[percnt] of survivors at long-term follow-up. Adjusting for age, gender, CRT and anti-metabolites, seizures were associated poorer performance on cognitive flexibility (CNS tumor effect size [ES]=-0.74, p=0.02; non-CNS tumor ES=-0.68, p<0.001), fluency (non-CNS tumor ES=-0.37, p<0.001) and short-term memory (CNS tumor ES=-0.63, p=0.001; non-CNS tumor ES=-0.46, p<0.001). Attention and processing speed had similar effects. HRQOL was generally not associated with seizures. Less than full-time employment was associated with seizures in survivors of CNS (ES=1.09, p=0.03) and non-CNS (ES=1.11, p=0.001) tumors. Among seizure characteristics, whether or not seizures resolved was the single most common predictor of poor outcome. Conclusions: Seizures in cancer survivors impact neurocognitive outcome and employment, independent of prior cancer therapy, and should be a focus of clinical management.