EORTC-1416-LCG/ETOP 8-15 – PEARLS/KEYNOTE-091 study of pembrolizumab versus placebo for completely resected early-stage non-small cell lung cancer (NSCLC): Outcomes in subgroups related to surgery, disease burden, and adjuvant chemotherapy use

8512 Background: At the second interim analysis (IA2) of the triple-blind, phase 3 PEARLS/KEYNOTE-091 study (NCT02504372), pembrolizumab significantly improved DFS compared with placebo in patients (pts) with completely resected stage IB (T ≥4 cm) to IIIA NSCLC per AJCC v7, regardless of PD-L1 expression (N = 1177, HR 0.76, 95% CI 0.63-0.91, P = 0.0014). We present DFS in subgroups related to surgery, disease burden, and adjuvant chemotherapy use. Methods: Pts had pathologically confirmed, completely resected stage IB (T ≥4 cm) to IIIA NSCLC of any PD-L1 expression and ECOG PS 0-1. Systematic complete or lobe-specific mediastinal lymph node dissection was recommended; minimally, the subcarinal and 1 lobe-specific lymph node must have been examined. Adjuvant chemotherapy of ≤4 cycles was given as indicated by local guidelines. Eligible pts were randomized 1:1 to pembrolizumab 200 mg or placebo Q3W for 18 doses (̃1 y). Treatment effects on DFS were assessed in prespecified subgroups of with and without adjuvant chemotherapy and in exploratory subgroups defined by surgery type, pN stage, tumor size, no. of adjuvant chemotherapy cycles, and adjuvant regimen; only subgroups of > 50 pts were analyzed. Data cutoff for IA2 was September 20, 2021 (median time from randomization to cutoff, 35.6 mo). Results: By surgery type, the HR (95% CI) for DFS was 0.78 (0.64-0.96) for lobectomy (n = 925), 0.85 (0.43-1.69) for bilobectomy (n = 92), and 0.71 (0.40-1.24) for pneumonectomy (n = 127). For subgroups based on nodal status, HR (95% CI) for DFS was 0.63 (0.46-0.86) for pN0 (n = 490), 0.77 (0.57-1.03) for pN1 (n = 456), and 1.00 (0.71-1.41) for pN2 (n = 231). By tumor size, and irrespective of nodal status, the HR (95% CI) for DFS was 0.91 (0.69-1.20) for size ≤4 cm (n = 491) and 0.70 (0.55-0.89) for size > 4 cm (n = 685). The HR (95% CI) for DFS was 0.73 (0.60-0.89) in pts who received adjuvant chemotherapy (n = 1010) and 1.25 (0.76-2.05) in those who did not (n = 167). Among pts who received adjuvant chemotherapy, HR (95% CI) for DFS by number of cycles was 0.59 (0.28-1.26) for 1-2 (n = 67) and 0.74 (0.61-0.91) for 3-4 (n = 943); by regimen, it was 0.74 (0.55-0.98) for cisplatin + vinorelbine (n = 491), 0.51 (0.31-0.83) for carboplatin + vinorelbine (n = 151), 1.21 (0.73-1.98) for carboplatin + paclitaxel (n = 135), 0.65 (0.30-1.40) for cisplatin + gemcitabine (n = 57), and 0.68 (0.41-1.14) for other regimen (n = 176). Conclusions: Pembrolizumab generally improved DFS versus placebo regardless of type of surgery, lymph node involvement, tumor size, and type and extent of adjuvant chemotherapy in pts with completely resected stage IB (T ≥4 cm) to IIIA NSCLC. These data support the benefit of pembrolizumab as adjuvant therapy for early-stage NSCLC following complete resection and, if indicated, adjuvant chemotherapy. Clinical trial information: NCT02504372.