BACKGROUND: Checkmate 649 (CM 649; NCT02872116), a randomized, open-label, phase 3 study in first-line (1L) gastric cancer/gastroesophageal junction cancer/ esophageal adenocarcinoma (GC/GEJC/EAC), showed statistically significant improvement in overall survival (OS) for nivolumab plus chemotherapy (N+C) vs C. Health care resource utilization (HCRU) during treatment with N+C vs C is reported here.
METHODS: Data on hospitalizations and non-protocol-specified visits (NPSVs) were collected for patients (pts) in CM 649. Frequency of hospitalizations and NPSVs were tabulated, summarized by NPSV types: emergency room, hospital outpatient, physician office, home health care, other. Descriptive statistics were used to summarize duration of hospitalizations. Exposure-adjusted incidence rates (IRs) of hospitalizations were calculated for each treatment arm as the ratio of number of hospitalizations to total time on treatment; incidence rate ratio (IRR) was calculated as the ratio of IR for N+C to C. Bootstrapping methods were used to calculate 95% CIs for the IRs and IRR. Exposure-adjusted IRs and IRR for total days in hospital were calculated using a negative binomial model with days in study included as an offset variable.
RESULTS: 1581 pts were randomized to N+C (n = 789) and C (n = 792). Mean duration of treatment (days) was 264 for N+C, 185 for C. During treatment, 36% of N+C pts and 29% of C pts reported ≥1 NPSV, most commonly physician office and hospital outpatient types in both groups. 40% of N+C pts and 29% of C pts reported ≥1 hospitalization, with 25% (N+C) and 21% (C) experiencing a single hospitalization. IRs (95% CI) for hospitalizations per pt per year of treatment were 0.98 (0.83-1.21) and 0.80 (0.69-0.94) for N+C and C, respectively, with IRR (95% CI) 1.23 (0.97-1.59); P value = 0.10. Among pts reporting ≥1 hospitalization, mean duration was 8.8 and 10.4 days for N+C and C, respectively. Exposure-adjusted IRs (95% CI) for days in hospital per pt per year were 10.08 (8.08-12.58) and 7.62 (6.08-9.55) for N+C and C, respectively, with IRR (95% CI) 1.32 (0.96-1.81); P value = 0.08.
CONCLUSIONS: In CM 649, when adding nivolumab to C, increased HCRU was observed in the N+C cohort compared with C alone. However, when accounting for duration of treatment, pts receiving N+C did not have significantly higher HCRU than pts receiving C alone. These findings, combined with improved OS/PFS and acceptable safety profile support treatment benefit of N+C over C in pts with GC/GEJC/EAC.
