NCOG-24. CHARACTERIZING NEUROTOXICITY ASSOCIATED WITH SYSTEMIC CHEMOTHERAPY IN PATIENTS WITH BREAST CANCER USING DOMAIN-SPECIFIC INTRA-INDIVIDUAL NEUROCOGNITIVE VARIABILITY

Abstract BACKGROUND Systemic chemotherapy is associated with neurotoxicity and neurocognitive functioning (NCF) changes. Intra-individual neurocognitive variability (IIV) is an index of NCF spread/dispersion across neuropsychological tests and can be elevated in patients with breast cancer (BC) after systemic chemotherapy. This study investigated longitudinal IIV changes within NCF domains after chemotherapy to determine whether: (1) IIV for NCF domains are more sensitive to changes than traditional mean-based metrics, and (2) whether changes in domain-specific IIV are associated with changes in mood disturbance or quality of life (QOL). METHODS Prior to and during/shortly after chemotherapy, 37 women with BC were administered a standardized NCF battery including tests within domains of attention/speed of information processing (A/SoP), learning/memory (L/M), and executive functions (EF). Within each domain, the coefficient of variation (CoV) was computed using the standard deviation of normatively-adjusted T-scores while adjusting for intra-individual mean. Patients were assessed for QOL using Functional Assessment of Cancer Therapy-Breast and mood using the STAI and BDI. RESULTS Paired samples t-tests revealed that mean NCF performances declined for L/M (t[36])=-4.44,p<.001,Hedge’s g=.70) and improved for A/SoP (t[36]=3.21,p=.003,g=.29), while CoV increased/worsened for L/M (t[36]=3.82,p<.001,g=.76) but did not change for A/SoP (t[36]=-0.05,p=.959,g=.01). There was no change in EF mean or CoV (ps >.10). No domain mean or IIV change was associated with mood changes (ps>.10). Mean NCF changes were not associated with QOL (ps >.10). Increased/worsened L/M CoV was associated with improved FACT-Breast scale QOL (ρ=.40,p=.020). CONCLUSIONS Both mean- and IIV-based NCF worsened for learning/memory. Discrepancies between mean and IIV change for attention/speed of information processing may be due to differential susceptibility to practice effects. Unexpected associations between learning/memory IIV changes and QOL require further investigation. Future studies are needed to examine NCF dispersion changes across treatment status and in other cancer populations.