Durvalumab with or without tremelimumab vs platinum-based chemotherapy as first-line treatment for metastatic non-small cell lung cancer: MYSTIC

Background: Durvalumab (D), a human IgG1 mAb against PD‐1 and CD80, has shown clinical activity in patients (pts) with non‐small cell lung cancer (NSCLC). Tremelimumab (T) is a human IgG2 mAb against CTLA‐4. D+T has previously also shown durable responses in metastatic NSCLC (mNSCLC). MYSTIC (NCT02453282) was an open‐label, Phase 3 trial of first‐line treatment with D vs platinum‐based doublet chemotherapy (CT) and D+T vs CT inmNSCLC. Methods: Eligible pts had mNSCLC; were immunotherapy/chemotherapy‐naïve; and had no EGFR sensitising mutation or ALK rearrangement. Tumour cell (TC) PD‐L1 expression (≥25% vs<25%) and histology were stratification factors. Patients were randomised (1:1:1) to D (20 mg/kg i.v. q4w); D+T (D: 20 mg/kg i.v. q4w; T: 1 mg/kg i.v. q4w [up to 4 doses]); or CT (intended up to 6 cycles; pemetrexed maintenance permitted in eligible pts) until disease progression. Primary endpoints were overall survival (OS) for D vs CT and OS and progression free survival (PFS; blinded independent central review [RECIST v1.1]) for D+T vs CT in pts with PD‐L1 TC expression ≥25% defined by the VENTANA PD‐L1 (SP263) assay. Data cutoffs were 4 Oct 2018 (OS and safety) and 1 Jun 2017 (PFS). Results: 1118 pts were randomised. Baseline characteristics were balanced. Efficacy findings are presented for the 488 pts with PD‐L1 TC ≥25%. Median OS was 16.3 vs 12.9 months for D vs CT (HR 0.76 [97.54% CI, 0.564, 1.019]; p=0.036) and 11.9 vs 12.9 months for D+T vs CT (HR 0.85 [98.77% CI, 0.611, 1.173]; p=0.202). Median PFS was 3.9 vs 5.4 months for D+T vs CT (HR 1.05 [99.5% CI, 0.722, 1.534]; p=0.705). 39.5% pts in the CT arm received subsequent immunotherapy after treatment discontinuation vs 6.1%and 3.1% pts in the D andD+T arms. Incidence of Grade 3/4 treatment‐related AEs was 14.6%, 22.1% and 33.8% with D, D+T and CT, respectively. Efficacy based on additional PD‐L1 cutoffs will be presented. Conclusions: In pts with mNSCLC, while statistical significance was not achieved for primary OS and PFS endpoints, first‐line D demonstrated clinically meaningful improvement in OS vs CT (PD‐L1 TC ≥25%). Safety data were consistent with the known safety profiles of D+/‐T. Further analyses are ongoing.