Frailty assessed by a 10-item index from comprehensive geriatric assessment (FI-CGA-10) versus oncologist-assessed performance status (PS) in older adults with gastrointestinal (GI) cancer.

Abstract

821 Background: A FI-CGA-10 is a recently developed measure of frailty in the geriatric oncology setting [Oncologist, 26, e1751 (2021)]. Our objective was to compare the multidimensional frailty assessment by FI-CGA-10 with PS assessed by a primary oncologist in older adults with cancer. Methods: This study included 790 older adults with GI cancer who underwent a CGA before cancer treatment decisions at a geriatric oncology service between September 2018 and May 2024, and whose PS was documented in the electronic medical record by a primary oncologist. Fitness and frailty level were evaluated using the FI-CGA-10, which assesses 10 domains: cognition, mood, communication, mobility, balance, nutrition, basic and instrumental activities of daily living, social support, and comorbidity. Deficits in each domain were scored as 0 (no problem), 0.5 (minor problem), and 1.0 (major problem). FI-CGA-10 scores (range 0-1) were calculated by dividing the sum of the scores for each domain by 10 and then categorized as fit (<0.2), pre-frail (0.2–0.35), and frail (>0.35). The strength of the ordinal association between the frailty category and each CGA domain score was assessed using Spearman’s non-parametric correlation coefficient (rho). Results: The median age was 79 years; 60% were male, 62% had GI tract cancer (esophageal, gastric, colorectal), 20% had hepatobiliary cancer, 15% had pancreatic cancer, and 46% had stage 4 disease. Overall (n=790), 25% of patients were fit, 40% were pre-frail, and 35% were frail (Table). Among patients with PS 0-1 (n=633), 30% were classified as fit, 48% as pre-frail, and 22% as frail. Among PS 0-1 patients, Spearman's rho was greater than 0.5 for the association between the frailty category (fit, pre-frail and frail) and cognition, IADL, and mobility domains. In this subcohort (PS 0-1), the proportion of patients with a cognition domain score of 0.5 (mild cognitive impairment) or 1.0 (dementia) was 7%, 30%, and 78% in the fit, pre-frail, and frail groups, respectively; the proportion of those with an IADL domain score of 0.5 (OARS IADL = 12-13) or 1.0 (OARS IADL ≤ 11) was 5%, 46%, and 92%, respectively; the proportion of those with a mobility domain score of 0.5 (0.8 m/s ≤ gait speed < 1.0 m/s) or 1.0 (gait speed < 0.8 m/s) was 45%, 85%, and 98%, respectively. Conclusions: In this study, 70% of older adults with GI cancer whose PS was scored as 0-1 by a primary oncologist were classified as pre-frail or frail when assessed using FI-CGA-10. Among patients with PS 0-1, there were strong correlations between the frailty levels and impairment levels in the cognition, IADL, and mobility domains, indicating that these domains are major contributors to the frailty classification (fit, pre-frail, and frail). All patients (n=790) Fit Pre-frail Frail PS 0 (n=351), n (%) 145 (41) 163 (46) 43 (12) PS 1 (n=282), n (%) 45 (16) 141 (50) 96 (34) PS 2-3 (n=157), n (%) 4 (3) 18 (11) 135 (86)