Search

Introduction Checkpoint inhibitors (CPIs) have been approved to treat metastatic NSCLC. Pegilodecakin + CPI suggested promising efficacy in phase 1 IVY, providing rationale for randomized phase 2 trials CYPRESS 1 and CYPRESS 2. Methods CYPRESS 1 (N = 101) and CYPRESS 2 (N = 52) included Eastern Cooperative Oncology Group performance status of 0 to 1 and first-line/second-line metastatic NSCLC, respectively, without known EGFR/ALK mutations. Patients were randomized 1:1; control arms received pembrolizumab (CYPRESS 1) or nivolumab (CYPRESS 2); experimental arms received pegilodecakin + CPI. Patients had programmed death-ligand 1 tumor proportion score of greater than or equal to 50% (CYPRESS 1) or 0% to 49% (CYPRESS 2). Primary end point was objective response rate (ORR) per investigator. Secondary end points included progression-free survival (PFS), overall survival (OS), and safety. Exploratory end points included immune activation biomarkers. Results Median follow-up for CYPRESS 1 and CYPRESS 2 was 10.0 and 11.6 months, respectively. Results for pegilodecakin + pembrolizumab versus pembrolizumab were as follows: ORR per investigator 47% versus 44% (OR = 1.1, 95% confidence interval [CI]: 0.5–2.5); median PFS 6.3 versus 6.1 months (hazard ratio [HR] = 0.937, 95% CI: 0.54–1.625); and median OS 16.3 months versus not reached (HR = 1.507, 95% CI: 0.708–3.209). Results per blinded independent central review were consistent. Treatment discontinuation rate owing to adverse events (AEs) doubled in the experimental arm (32% versus 15%). AEs with grade greater than or equal to 3 treatment-related AEs (62% versus 19%) included anemia (20% versus 0%) and thrombocytopenia (12% versus 2%). Results for pegilodecakin + nivolumab versus nivolumab were as follows: ORR per investigator 15% versus 12% (OR = 1.2, 95% CI: 0.3–5.9); median PFS 1.9 versus 1.9 months (HR = 1.006, 95% CI: 0.519–1.951); and median OS 6.7 versus 10.7 months (HR = 1.871, 95% CI: 0.772–4.532). AEs with grade greater than or equal to 3 treatment-related AEs (70.4% versus 16.7%) included anemia (40.7% versus 0%), fatigue (18% versus 0%), and thrombocytopenia (14.8% versus 0%). Biomarker data suggested activation of immunostimulatory signals of interleukin-10R pathway in pegilodecakin-containing arms. Conclusions Despite evidence of biological effect in peripheral blood, adding pegilodecakin to CPI did not improve ORR, PFS, or OS, in first-line/second-line NSCLC. Pegilodecakin + CPI has been found to have overall higher toxicity compared with CPI alone, leading to doubling of treatment discontinuation rate owing to AEs.

9563 Background: Pembrolizumab (P) is an immune checkpoint inhibitor (CPI) approved to treat 1L advanced NSCLC pts with high PD-L1 expression. PEG + CPI has demonstrated promising efficacy in NSCLC pts in a phase I trial (IVY; NCT02009449; Naing et al., 2019 Lancet Oncol), providing rationale for this randomized phase II trial (CYPRESS 1; NCT03382899). Methods: CYPRESS 1 was an open label phase II trial, for treatment-naïve, ECOG 0-1, PD-L1 high (22C3 clone TPS ≥ 50%), Stage IV NSCLC pts, without known EGFR/ALK mutations. Pts were randomized 1:1 to arm P (received 200mg IV on day 1 of a 21-day cycle) v. arm PEG+P (received P as above + PEG daily of 0.8 mg if weight ≤80kg and 1.6mg if weight> 80 kg up to 35 cycles in each arm). Pts were stratified by tumor histology and must have no prior history of cancer or prior CPI therapy. Primary endpoint was ORR (per RECIST v1.1 by investigator) Secondary endpoints included PFS, OS, and safety. Exploratory endpoints included ORR and PFS by blinded independent central review (BICR). Immune activation biomarkers (baseline and change from baseline) were assessed by serum immunoassay, IHC, and sequencing. Results: As of Dec 6, 2019, 101 pts were randomized to PEG+P (n=51) or P (n=50). Median follow-up time was 10.0 months (95% CI [8.4, 11.1]). Results for PEG+P versus P were: ORR per investigator was 47% v. 44% (p=0.76), ORR per BICR was 53% v. 46%(p=0.78), mPFS per investigator was 6.3 v. 6.1 months with HR = 0.94 (95% CI [0.54, 1.63];p=0.82), mPFS per BICR was 6.4 v. 7.2 months with HR = 1.10 (95%CI [0.62, 1.96]; p=0.74), and mOS was 16.3 months v. not reached with HR = 1.36 (95% CI [0.66, 2.77]; p-value=0.40). Gr ≥3 treatment related adverse events (TRAEs) were 62% for PEG+P versus 19% for P. Gr ≥3 TRAEs with ≥10% incidence included anemia (20% vs. 0%) and thrombocytopenia (12% vs. 2%). Biomarker data on immunostimulatory signals of the IL-10R pathway will be included. Conclusions: Adding PEG to P did not lead to improvement in ORR, PFS, or OS, in 1L advanced NSCLC with high PD-L1 expression. PEG+P arm demonstrated expected safety profile but overall higher toxicity compared to pembrolizumab alone. Clinical trial information: NCT03382899 .