9513 Background: TIM-3 and PD-1 are markers of T-cell suppression that are upregulated in melanoma. AMBER (NCT02817633) is evaluating cobolimab (TSR-022/GSK4069889), an anti-TIM-3 therapy, monotherapy or with PD-1 inhibitors, including dostarlimab, in pts with solid tumors. Methods: This multicenter, open-label study was conducted in 2 parts: dose escalation (Parts 1 A–D and F–H) and cohort expansion (Parts 2 A–D). Part 1C and exploratory cohort 1E (reported here) included pts with advanced/metastatic melanoma; prior therapies, except for immunotherapies, were permitted. Pts received cobolimab (100 [1C only], 300, or 900 mg IV) with dostarlimab (500 mg IV) Q3W. Part 1C primary endpoints were safety, tolerability, and recommended Phase 2 dose. Objective response rate (ORR; complete [CR] or partial [PR] response per RECIST v1.1) was a secondary endpoint in 1C and the primary endpoint in 1E (ad hoc efficacy analysis reported). An integrated safety analysis for all pts (Parts 1 and 2) receiving cobolimab with dostarlimab, regardless of tumor type or cobolimab dose, is reported here. Results: 28 pts were enrolled in 1C (n=10) and 1E (n=18). Most pts (n=23; 82.1%) had cutaneous disease of the skin. One pt had anorectal mucosal disease and 3 pts in the 900-mg cohort had uveal melanoma. Most pts (67.9%) had an ECOG PS=0. At data cut-off (May 19, 2021), treatment was ongoing in 5 pts. In the integrated safety analysis of pts who received cobolimab 100 mg (n=41), 300 mg (n=167), or 900 mg (n=69) with dostarlimab, treatment-related treatment-emergent AEs (TR-TEAEs) occurred in 53.7%, 57.5%, and 59.4%, respectively. The most common TR-TEAEs (any grade, ≥10% in 100 mg, 300 mg, or 900 mg groups, respectively) were fatigue (22.0%, 13.2%, 24.6%), rash (9.8%, 5.4%, 11.6%), diarrhea (4.9%, 6.0%, 10.1%), and dyspnea (2.4%, 0%, 10.1%). Grade ≥3 TR-TEAEs occurred in 12.2% (100 mg), 10.8% (300 mg), and 20.3% (900 mg); serious TR-TEAEs occurred in 7.3%, 7.8%, and 11.6%, respectively. No pts died due to TR-TEAEs; 2.4% (100 mg), 4.2% (300 mg), and 7.2% (900 mg) discontinued due to TR-TEAEs. ORR and disease control rate (DCR: stable disease [SD] ≥16 weeks, PR, or CR) are shown in the Table. Twelve pts achieved a PR and an immune-related (ir)PR (1 in 100 mg; 8 in 300 mg; 3 in 900 mg groups). Three pts achieved SD (2 in 300 mg; 1 in 900 mg groups); 8 pts had irSD (1 in 100 mg; 4 in 300 mg; 3 in 900 mg groups). Conclusions: Cobolimab with dostarlimab showed preliminary clinical responses in pts with advanced/metastatic melanoma and an acceptable safety profile across advanced cancers. Funding: GSK (213348). Clinical trial information: NCT02817633. [Table: see text]

1092 Background: Lurbinectedin (L) is a selective inhibitor of oncogenic transcription that leads to cell apoptosis and shows antitumor activity against homologous recombination repair-deficient cell lines. A previous phase II study (Cruz et al. JCO 2018;36:3134-3143) demonstrated antitumor activity in patients (pts) with pretreated metastatic breast cancer (median of 1 prior advanced chemotherapy line) and BRCA1/2-mutated tumors with L 3.5 mg/m2 or 7.0 mg flat dose (equivalent to 4.0 mg/m2) every three weeks [q3wk]). This report focuses on the outcomes in the BRCA1/2-associated breast cancer cohort of a phase II Basket multitumor trial. Methods: This phase II study evaluated L 3.2 mg/m2 1-hour intravenous (i.v.) infusion q3wk in a cohort of 21 female pts with pretreated BRCA1/2-associated breast cancer. The primary efficacy endpoint was ORR according to RECIST v1.1. Secondary endpoints included duration of response (DoR), progression-free survival (PFS), OS and safety. Results: Median age was 45 years (range, 29-73 years). Hormone receptor (HR)+ disease was observed in 76.2% of pts, triple negative disease in 19.0% and HER2+ in 9.5%. BRCA1 and BRCA2 were reported in 47.6% and 52.4% of pts, respectively. Median number of prior lines of chemotherapy for advanced disease was 2 (range, 0-3 lines). Prior poly(ADP-ribose) polymerase inhibitors and platinum compounds had been administered to 23.8% and 47.6% of pts, respectively. Confirmed partial response (PR) was observed in six pts (ORR = 28.6%; 95% CI, 11.3-52.2%). Lurbinectedin was active in both BRCA mutations: four PRs in 11 pts (36.4%) in BRCA2 and two PRs in 10 pts (20.0%) in BRCA1. Median DoR was 8.6 months, median PFS was 4.1 months and median OS was 16.1 months. Stable disease (SD) was observed in ten pts (47.6%), including three pts with unconfirmed response in a subsequent tumor assessment (ORR unconfirmed = 42.9% [95%CI, 21.8-66.0]). Clinical benefit rate (PR + SD≥4 months) was 76.2% (95% CI, 52.8-91.8%). The most common grade 3/4 toxicity was neutropenia (42.9%; grade 4, 23.8%; with no febrile neutropenia). Conclusions: This phase II study met its primary endpoint and confirmed the activity of L in pretreated BRCA1/2-associated breast cancer pts. L 3.2 mg/m2 1-hour i.v. infusion q3wk showed an acceptable, predictable and manageable safety profile. Considering the exploratory aim of this trial as well as previous results in other phase II study, further development of L in this indication is warranted. Clinical trial information: NCT02454972.

e20632 Background: Lurbinectedin is a selective inhibitor of oncogenic transcription. The US FDA granted accelerated approval for lurbinectedin as monotherapy (3.2 mg/m2 q3wk) for relapsed small cell lung cancer (SCLC) based on a phase 2 basket trial (NCT02454972). The ATLANTIS trial (NCT02566993) investigated lurbinectedin 2.0 mg/m2 plus doxorubicin 40.0 mg/m2 q3wk versus topotecan or CAV in relapsed SCLC. The CORAIL trial (NCT02421588) evaluated lurbinectedin (3.2 mg/m2 i.v. q3wk) compared to pegylated liposomal doxorubicin or topotecan in patients with platinum-resistant ovarian cancer. The elderly population is usually under-represented in clinical trials despite its interest due to the increased frailty and high comorbidity. Methods: Data from Basket and CORAIL trials were included in an integrated safety analysis to evaluate lurbinectedin safety profile at the approved regimen (3.2 mg/m2 q3wk). This post hoc safety analysis compares the outcomes of lurbinectedin in elderly population (n = 207 patients ≥65 years old; Basket n = 113 and CORAIL n = 94) versus topotecan (n = 75 patients ≥65 years old; ATLANTIS n = 45 and CORAIL n = 30). Results: Treatment-related adverse events (AEs) and laboratory abnormalities regardless of relationship occurring in > 10% of patients, were: Compared with topotecan, lurbinectedin showed lower frequency of treatment-related grade ≥3 AEs (53.1% vs. 85.3%; p < 0.01) and grade ≥3 serious adverse events (SAEs) (19.8% vs. 34.7%; p = 0.01) and a trend for less AEs leading to treatment discontinuation (5.3% vs. 8.0%; p = 0.40) or deaths (1.4% vs. 5.3%; p = 0.08). Use of supportive treatments was significantly lower with lurbinectedin: G-CSF (17.9% vs. 29.3%; p < 0.05), red blood cells transfusions (20.8% vs. 56.0%; p < 0.01) or erythropoietin (2.4% vs. 14.7%; p < 0.01). Conclusions: Lurbinectedin is well tolerated in elderly patients and has a predictable and manageable safety profile. No new safety signals were observed. With the limitations of cross-trial comparisons, lurbinectedin shows a clear advantage in terms of hematological toxicity compared with topotecan in patients ≥ 65 years old. Furthermore, fewer elderly patients treated with lurbinectedin had severe AEs, SAEs, treatment discontinuations and deaths compared to elderly patients treated with topotecan. Clinical trial information: NCT02454972 , NCT02566993 , NCT02421588 . [Table: see text]