In total, 52 patients were eligible and enrolled in the present study. The median age of all patients was 61 (range, 46–84) years. The majority of the patients were men (88.5%) and exhibited ECOG performance status of 1 (75.0%) (Table 1).
Arm A (sintilimab-TP) included 32 (61.5%) patients and arm B (sintilimab-GP) 20 (38.5%). In arm A, 20 patients received nab-paclitaxel and 12 received paclitaxel. Baseline demographic and disease characteristics exhibited no significant differences between the groups (Table 1). The overall median follow-up for this analysis was 12.1 months. A median of 4 doses (range, 1–20) of arm A and 6 doses (range, 3–34) of arm B was used. As of April 31 2021, a total of 17 patients (53.1%) in the sintilimab-TP group and 7 patients (35.0%) in the sintilimab-GP group were continuing treatment.
The confirmed ORR was 59.4% (95% CI, 41.4–77.4) in arm A and 40.0% (95% CI, 16.5–63.5) in arm B (P = 0.24; Table 2 and figure 1). In addition, 6.4% of the patients who received sintilimab-TP and 5.0% of those who received sintilimab-GP exhibited disease progression (Table 2 and Fig.1). The DCR was 93.6% in arm A and 95.0% in arm B. The median TTR was 2.0 months (range: 0.8–6.7) with arm A and 1.9 months (range: 0.7–3.0), respectively (Table 2).
A total of 28 events of progression or death were noted and the median PFS was estimated to 10.9 months (95% CI, 5.0 to 16.7) in arm A and 7.5 months (95% CI, 4.0–10.9) in arm B (HR for PFS, 0.64; 95% CI, 0.30–1.4; P = 0.24; Fig.2a). The 12-month PFS rate was 44.6% with TP and 27.1% with GP. Across subgroups that were analyzed, the two treatment groups exhibited similar PFS benefits (figure 2B). However, male patients tended to benefit from arm A over arm B (median PFS, 13.9 months [95% CI, 7.4– 20.5] vs. 7.5 months [3.7–11.2]; HR, 0.51 [95% CI, 0.23–1.12]; P = 0.093; Fig.2b).
The median OS was 20.1 months (95% CI, 13.6–26.6) in arm A and 16.3 months (95% CI, 2.9–29.6) in arm B (HR, 0.69; 95% CI, 0.26–1.84; P = 0.46) (Fig.3a), with a total of 19 events in the treatment population. The OS rate at 1 year was 72.9% in arm A and 53.3% in arm B. Similar OS benefits were observed in all prespecified subgroups of the two treatment groups (Fig.3b).
AEs of any cause occurred in all patients in the two treatment groups (Table 3). AEs of grade 3 or higher occurred in 37.5% of the patients in arm A and 55.0% of the patients in the arm B. The AEs led to the discontinuation of all treatment components in 3.1% and 10.0% of arms A and B, respectively and to the discontinuation of sintilimab in 12.5% and 10.0% of arms A and B, respectively. No toxicity-related deaths occurred. The most common AEs in arm A were anemia (68.8%), decreased white blood cell count (37.5%), increased alanine aminotransferase (31.3%) and decreased neutrophil count (25.0%), whereas the most common AEs in arm B were anemia (75.0%), decreased white blood cell count (40.0%), decreased neutrophil count (30.0%) and decreased platelet count (20.0%).
Immune-mediated adverse events (irAEs) occurred in 12 of 32 patients (40.6%) in arm A and in 9 of 20 patients (45.0%) in arm B; grade 3 or higher AEs occurred in 3 patients (9.1%) and one patient (5.0%), respectively. A higher number of hematological AEs was noted in arm B than in arm A, while the number of cases with hyperthyroidism, myositis, peripheral neuropathy and immune-mediated pneumonitis was higher in arm A than that in arm B.