Cognitive function (CF) and prognostic information understanding in advanced cancer patients (ACP)

9516 Background: The role of advanced cancer patients' (ACP) cognitive function (CF) and its relationship to understanding of a terminal prognosis has never been formally evaluated. Methods: ACP CF was evaluated among a population of terminally-ill patients enrolling in phase I trials at our institution using a neuropsychological battery designed to assess several domains of decisional capacity: Memory (Hopkins Verbal Learning HVLT); Executive Functioning (Verbal Fluency and Trail-making A/B); Language (Boston Naming-short); Attention (Digit Span); Comprehension (Auditory Comprehension & WAIS comprehension). Semi-structured interviews of ACPs also evaluated md-pt communication regarding prognosis, and included the Hospital Anxiety and Depression Scale (HADS), BDI-II, and the FACT-COG. Results: To date, 110 ACP enrolling in phase I trials have been interviewed: median age 60 (23–83); 66% male; 88% Ca; 62% married; 71% >high school education; 52% GI dx. 59% of ACPs reported having a discussion regarding life expectancy, and 55% stated that the physician gave them a prognostic timeframe regarding the amount of time left to live. ACP who stated the physician did not provide them a timeframe had measurable deficits in CF as indicated by Z scores for HVLT immediate recall (-1.3 ± .84 v -.74 ± 1.1, p=.03); total recall (-1.9 ± 1.2 v -1.2 ±1.5, p=.02); delayed recall (-1.7 ± 1.6 v -1.1 ±1.6, p=.04); language (.26 ± 1.3 v .78 ± .68, p=.03); Trails B (-1.6 ± 2.5 v -.61 ± 2.2, p=.04) and WAIS comprehension ss scores (14 ± 2.9 v 16 ± 3.0, p=.06). These ACP tended to exhibit more depressive symptoms (12 ± 10 v 6 ± 3, p=.04) and had lower scores for perceived cognitive impairment (96 ± 25 v 105 ±16, p=.04); impact on quality of life (26 ± 7 v 28 ± 5, p=.03); and FACT-COG total (152 ± 31 v 164 ± 22, p=.05). Conclusions: ACP enrolling in phase I trials who could not recall specific prognostic information had measurable cognitive impairment as compared to those ACP who could recall such information. Our data indicate that CF may play a role in ACP communication and/or understanding of prognostic information. [Table: see text]