A single-arm, phase II study of durvalumab (D) and tremelimumab (T) for relapsed/ refractory germ cell tumors (GCT)

5040 Background: Patients (pts) with GCT and progressive disease (PD) after salvage high-dose chemotherapy (HDCT) have dismal outcomes with median progression free survival (PFS) and overall survival (OS) of 1.0 and 4.7 months (mo), respectively (Feldman et al. Cancer 2012). Immune checkpoint blockade (ICB) has resulted in durable benefit in multiple tumor types. We tested D (anti-PD-L1) + T (anti-CTLA-4) for pts with relapsed/refractory GCTs in a phase II trial (NCT03158064). Methods: Eligible pts were ≥ 18 years old with GCT of any primary site who had PD after first-line chemotherapy and HDCT or were not candidates for HDCT or other curative options. Pts received D (1500 mg) and T (75 mg) every 4 weeks (wks) for 4 cycles followed by D (1500 mg) every 4 wks for up to 1 year (schedule 1). The study was later amended to one cycle of D (1500 mg) and T (300 mg) followed by D (1500 mg) every 4 wks for up to 1 year (schedule 2). The primary endpoint was initially overall response rate (ORR) in a Simon’s two-stage design with null and alternate rates of 5% and 20%, respectively. The study was then amended to use a primary endpoint of 16-wk PFS in a new Simon’s two-stage design with null and alternate rates of 5% and 25%. The regimen would be considered promising if 16-wk PFS was achieved in ≥ 3 of 20 pts. Secondary endpoints included PFS, OS, safety, and association of PD-L1 staining with outcome. Results: Between July 2017 and Jan 2022, 29 pts were treated, the last 7 (24%) on schedule 2. Median age was 37 (range, 21-70). Histology was non-seminoma in 25 pts (86%), and primary sites were testis in 20 (69%), mediastinum in 7 (24%), ovary in 1 (3%), and unknown in 1 (3%). Six (21%) pts had late relapse and 21 (72%) received prior HDCT. Median AFP was 310 ng/mL (range, 1–18,087) and HCG 471 mIU/mL (range, 3.3-62,302). Ten pts were treated with the ORR primary endpoint with no responses in stage 1, but 2 pts had tumor reductions of 20% and 22% with PFS of 9.9 and 10.4 mo, respectively, and 1 patient with elevated AFP but no RECIST measurable disease had a PFS of 5.7 mo. Of 19 pts treated with the 16-wk PFS primary endpoint, one had an ongoing partial response (59% tumor reduction) with 33 months follow-up; all others had PD in <16 wks (17 with PD and 1 with stable disease). Thus, the study did not meet its primary endpoint. Among 29 treated pts, the 16-wk PFS was 13.8% (range, 4.3-28.6), and median PFS and OS were 1.4 (95% CI: 4.1, 7.1) and 7.3 mo (95% CI: 3.2, 10.9), respectively. PD-L1 expression data was available for 20/29 pts - 3/13 (23%) with low/negative and 1/7 (14%) with high expression achieved PFS >16 wks. Six (21%) pts had grade 3-4 treatment-related adverse events, 3 (10%) of whom required high-dose steroids. Conclusions: D+T had limited antitumor activity in pts with relapsed/refractory GCT, failing to meet protocol-defined success. However, some pts did appear to benefit, including one exceptional responder. Additional correlative studies are ongoing. Clinical trial information: NCT03158064 .