Potential surrogates for performance status and survival in metastatic non-small cell lung cancer (NSCLC).

Abstract

170 Background: Performance status (PS), a physician-based, subjective assessment of a patient’s symptoms and ambulatory state, is an eligibility criterion for most clinical trials and a key clinical decision tool for treatment of NSCLC. Methods: Patients with metastatic NSCLC were asked to participate in a prospective, observational study of potential prognostic factors that includes demographics, laboratory, patient reported outcomes (PROs) by the patient reported outcomes measurement information system (PROMIS), PS reported by patient, PS assessed by physician, FitBit tracker (includes steps, distance, heart rate (HR), calories, METs), physical challenge and proteomics (Veristratâ). Correlation and regression analyses were performed for the initial 55 patients enrolled prior to COVID restrictions. Tracker data was collected and averaged over the first 7 days with wear > 500 minutes/day. Other data were collected at baseline. Results: We enrolled 55 patients with metastatic NSCLC. There were 35 females and 20 males with a median age of 67 years. A majority of patients were never (38%, n = 21) or former smokers (40%, n = 22); adenocarcinoma (83%, n = 83%) was the most common histology. 30% (n = 17) patients were 2nd or greater line of therapy. 52 were included for survival analysis and 21 expired with median survival 226 days (149, NA). There were minimal gender related differences in the study population except a higher incidence of adenocarcinoma in women (94% vs 65%, p = 0.012 and longer FiTBit assessed daily “very active minutes” (defined as METs > = 6) in men compared to women (4.21 vs 0.43, p = 0.036). Patient- and physician assessed PS were not statistically different, although patients tended underestimated their PS relative to physicians. Physician assessed PS, patient assessed PS, two-minute walk distance, proteomics by Veristratâ, total daily steps, total daily distance covered, and heart rate/activity mismatch were independent predictors of survival. Age, gender, smoking, BMI, histology, and various CBC-based indices were not prognostic. Conclusions: Patient- and physician-assessment of PS do not always match, many alternate measures to PS are prognostic of survival and tracker-based methods are objective and feasible in clinical practice.