The association between an electronic health record (EHR)–embedded frailty index and patient-reported outcomes among patients with metastatic non–small-cell lung cancer on immunotherapy.

Abstract

335 Background: While frailty is a well-established predictor of overall mortality among patients with metastatic non-small cell lung cancer (mNSCLC), its association with patient-reported outcomes is not well-characterized. We assessed quality of life and symptom burden by frailty status among mNSCLC patients receiving immunotherapy. We used an electronic health record (EHR) frailty index (eFI) as a marker of frailty. In addition, we assessed self-perceived goals of treatment by frailty status. Methods: In a cross-sectional study, patients from a single academic medical center with mNSCLC who were on first or second-line immunotherapy or chemoimmunotherapy completed the European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC-QLQ-C30) and the National Cancer Institute Patient Reported Outcomes version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE). In addition completed the Prognosis Treatment and Perceptions Questionnaire. We computed means of EORTC-QLQ-C30 subscales and proportions reporting at least moderate PRO-CTCAE symptoms, and compared results across three levels of frailty status using the eFI (which uses 54 variables from the EHR to calculate a score from 0-1), which categorizes frailty status as either fit (eFI ≤0.10), pre-frail (0.10 < eFI≤0.21), or frail (eFI > 0.21). P-values for comparisons of continuous variables were derived from F-tests from ANOVA models, while those for comparisons of categorical variables were derived from chi-square or Fisher’s exact tests. Results: Sixty patients (mean age 62.5 years, 75% Caucasian, 60% women, 57% receiving single-agent immunotherapy) participated. Most patients were pre-frail (68%), with few categorized as fit (18%) or frail (n=15%). The EORTC-QLQ-C30 global health score was similar among frailty categories (mean 62.6). Among EORTC functional subscales, only physical function differed by frailty status (mean 83.9 fit vs 74.8 pre-frail vs 60.0 frail, p-value 0.04). We found no differences in any of the EORTC-QLQ-C30 symptom subscales across frailty categories. On the PRO-CTCAE measures, only self-reported pain differed among the three groups, with pain being higher among frail patients (75% of frail reporting at least moderate pain, vs 51.5% of pre-frail and 18.2% of fit; p=0.046). No differences in self-perceived goals of treatment or likelihood of cure emerged by frailty status. Conclusions: Frail patients identified by the eFI with mNSCLC have higher rates of pain and physical functional impairments but similar perspectives on likelihood of cure as those without frailty. This highlights the importance of symptom monitoring and frailty screening along with prognostic discussions in those with mNSCLC. Clinical trial information: NCT03741868 .