Decline in Smartphone-Assessed Physical Activity Level is Associated With Clinical Outcomes in Phase I/II Clinical Cancer Trials

Background: A decline in physical function may be an early predictor for complications of cancer treatment. This study examined whether repeated objective smartphone measurements of physical activity and exercise capacity in patients with cancer are feasible during early-phase clinical trials (EPCTs) and whether a decline in physical function is associated with clinical outcomes. Methods: Physical activity (steps/day) and exercise capacity (6-minute walk test [6MWT]) were measured with a smartphone before EPCT start (T0) and after 4 weeks (T1) and 8 weeks (T2). Univariable logistic regression analyzed associations between a decline in step count (≥20%), 6MWT distance (≥10%), or deterioration of ECOG performance status (PS) and trial discontinuation at 8 weeks and 90 days. Cox proportional hazards models were used to examine associations with progression-free survival (PFS) and overall survival (OS), adjusting for trial phase (I vs II), cancer type (hematologic malignancy vs solid tumor), and PS (0 vs ≥1). Results: Among 117 included patients, valid step count and 6MWT measurements were available for 96.6% and 76.7% of patients at T0, 74.4% and 53.3% at T1, and 89.7% and 54.4% at T2, respectively. Patients experiencing step count decline between T0 and T1 had higher odds of trial discontinuation at 8 weeks (odds ratio, 8.67; 95% CI, 1.94–61.43), and decline between T1 and T2 was associated with discontinuation at 90 days (odds ratio, 5.20; 95% CI, 1.43–21.14). Step count decline was significantly associated with shorter PFS (hazard ratio, 3.54; 95% CI, 2.06–6.08) and OS (hazard ratio, 2.31; 95% CI, 1.26–4.23). Declines in 6MWT distance or deterioration in ECOG PS were not associated with trial discontinuation or survival. Conclusions: Repeated smartphone measurements of physical activity are feasible in patients participating in EPCTs. Additionally, physical activity decline is significantly associated with trial discontinuation, PFS, and OS. Hence, we envision that objective smartphone measurements of physical activity will contribute to optimal treatment development for patients with cancer.