Abstract
PURPOSE: Immunotherapy alone or in combination with chemotherapy has become the standard of care for medically fit patients who have advanced non-small cell lung cancer (NSCLC) without a driver mutation. The optimal treatment for patients with poor performance status remains an active area of investigation. PATIENTS AND METHODS: Patients with advanced NSCLC and an Eastern Cooperative Oncology Group performance status (PS) of 2 were randomized to single-agent pembrolizumab at 200 mg every 3 weeks (Arm A) or the same dose of pembrolizumab combined with weekly carboplatin area under the curve 1 and paclitaxel 25 mg/m(2) (Arm B). The co-primary outcomes were differences in response rates between the Arms and comparison of each Arm to a historical control. Progression free survival (PFS) and overall survival (OS) were secondary outcomes using intent to treat analyses. Optional blood samples were obtained at baseline and after 2 cycles of treatment, and immune cells were measured using flow cytometry. A subset of post-treatment blood samples were analyzed using single-cell sequencing. RESULTS: 43 patients enrolled with 20 patients evaluable for response in each arm. All enrolled patients were included in survival analyses. Therapy was generally well tolerated with no treatment-related deaths in either arm. Both Arms exceeded the predefined historical control response rate of 10 % (Arm A = 35 %, P = 0.0002; Arm B = 45 %, P < 0.0001). Response rates favored Arm B but the difference was not statistically significant (P = 0.75). Median PFS and OS were not significantly different between the two arms. To compare survival outcomes of either pembrolizumab-based therapy to historical control of platinum-based chemotherapy for patients with PS 2, the two arms were combined. Median PFS was similar to historical control of platinum-based chemotherapy (4.6 months vs 4.6 months historical control). However, overall survival favored pembrolizumab-based therapy at 12 months (44 % vs 31 % historical control, P = 0.062) and 24 months (33 % vs 11 % historical control, P = 0.0001). Twenty patients provided optional blood samples for biomarker analyses. Consistent with prior reports, a numerically longer (but not significant) OS was observed in patients with low regulatory T cells (CD4 + FoxP3 + ) at baseline (14.5 vs 4.6 months, P = 0.068). Abundance of myeloid derived suppressor cells (CD14 + HLA DR-) at baseline did not correlate with clinical outcomes. Single-cell sequencing identified several significant differences in gene expression profiles within the CD14 + cell population for responding and non-responding patients treated with chemoimmunotherapy. CONCLUSIONS: For patients with poor performance status, adding very low dose chemotherapy to pembrolizumab did not significantly improve clinical outcomes compared to pembrolizumab alone. Patients receiving either of these pembrolizumab-based regimens demonstrated better long-term survival when compared to historical outcomes of platinum-based chemotherapy for the PS 2 population.