THE IMPACT OF CANCER HISTORY ON COGNITIVE PERFORMANCE AND CONVERSION IN OLDER ADULTS WITH MILD COGNITIVE IMPAIRMENT AND NORMAL COGNITION

Older adults with a cancer history may experience cognitive decline due to cancer-related cognitive impairment (CRCI), amnestic mild cognitive impairment (MCI), and/or Alzheimer's dementia (AD). This analysis examined if histories of invasive cancer or non-melanoma skin cancer (NMSC) are associated with greater cognitive decline or altered trajectories of cognitive aging in participants with normal cognition (NC) and MCI. We analyzed cognitive performance and AD conversion rates in participants enrolled in the Alzheimer's Disease Neuroimaging Initiative (ADNI). We identified:120 participants with NC and invasive cancer history (NC-CA), 96 participants with NC and NMSC history (NC-NMSC), 498 participants with NC and no cancer history (NC-no-CA), 147 participants with MCI and invasive cancer history (MCI-CA), 82 participants with MCI and NMSC history (MCI-NMSC), and 634 participants with MCI and no cancer history (MCI-no-CA). Scores on cognitive tests were transformed; corrected for age, education, and sex; and averaged into four composite scores: Structured-Verbal Memory (SVM), Unstructured-Verbal Memory (UVM), Attention/Processing Speed (APS), and Language (LNG). Linear mixed effects models (LMM), survival analysis, and cox regression corrected for multiple comparisons (p≤0.013) were used for the analysis. Significance threshold for follow-up analyses was set to p≤0.05. MCI-no-CA participants performed significantly worse relative to MCI-CA on UVM (p=0.010) tasks. Decline was present in learning (p=0.009), delayed memory (p=0.022), and recognition (p=0.008). Relative to MCI-CA participants, MCI-no-CA participants converted to AD at a significantly younger age (p=0.005) and were 1.6 times more likely to convert to AD. NC participants did not differ significantly in any cognitive domains or in conversion to MCI/AD regardless of cancer history. NC-NMSC and MCI-NMSC participants did not differ significantly from respective cancer-free comparison groups. Persistent symptoms of CRCI have been document as long as 20-years after cancer diagnosis and treatment. Our results suggest that a percentage of MCI-CA participants enrolled in memory research may be experiencing long-term CRCI as a source of cognitive impairment rather than preclinical AD. Future research in MCI/AD should consider invasive cancer-history as a possible marker for MCI that is slower progressing and less likely to convert to AD.