Real-world outcomes with pembrolizumab in patients with treatment-naive advanced/metastatic NSCLC in the UK: multicentre retrospective observational study

Abstract

Introduction: Anti-PD1 inhibitor pembrolizumab became standard of care therapy for previously untreated advanced/metastatic NSCLC with PD-L1 ≥50% following publication of results from KEYNOTE-001 and KEYNOTE-024 trials, having demonstrated improved overall survival and tolerability over chemotherapy. We evaluated the realworld efficacy and tolerability of first-line pembrolizumab in UK patients. Methods: Multicentre retrospective observational study. Primary endpoint: progression-free survival (PFS). Key secondary endpoints: time to PD-L1 report, overall survival (OS), objective response rate (ORR), prognostic factors associated with PFS and OS, and treatmentrelated toxicities. Results: 219 patients treated with first-line pembrolizumab between 07/2016 and 01/2018 at 27 centres were included (Table 1). Median time from diagnosis of advanced disease to PDL1 report was 18 days. Median time from PD-L1 report to start of pembrolizumab treatment was 23 days. After median follow-up of 5.7 months, there were 90 events of progression or death. ORR was 56.6% (KN-024: 44.8%) with disease-control rate of 76.4%. Median PFS was 8.2 months (KN-024: 10.3mo). 20 patients continued pembrolizumab beyond first documented progression, with median time to further PD of 2.8 months. Median OS was not reached (KN- 024: NR; KN-001: 35mo). 6-month and 1-year OS rates were 73.8% (KN-024: 80.2%) and 68.2% (KN-024: 70.3%), respectively. In the multivariate analysis, poor performance status and presence of a confirmed mutation or unknown mutational status were associated with increased risk of death (p=0.024 and p=0.044). 22.8% patients experienced grade ≥3 treatment-related toxicity (KN-024; 26.6%). 38% experienced any-grade immune-related toxicities (KN-024: 29.2%), the most common being hypothyroidism (7.3%), rash (5.9%) and hyperthyroidism (3.7%). 77 patients (35.2%) required a dose delay and 58 (26.5%) required immunosuppressant therapy.Conclusion: Real-world efficacy and safety of first-line pembrolizumab are comparable to published trial data. Prolonged times to PD-L1 report and start of pembrolizumab treatment likely reflect complex access routes prior to UK NICE approval.