Roginolisib, an oral, highly selective and allosteric modulator of phosphoinositide 3-kinase inhibitor delta (PI3Kδ) in patients with uveal melanoma and advanced cancers.

Abstract

9597 Background: The highly selective, oral allosteric modulator of PI3Kδ, roginolisib (IOA-244), blocks the activity of PI3Kδ-dependent signaling, in both tumor cells and Tregs. Methods: Continuous daily dosing of roginolisib at 10, 20, 40 and 80 mg in patients (pts) with pretreated solid tumours and Follicular Lymphoma (FL) were evaluated (Part A). Part B consists of a dose confirmation cohort in uveal melanoma (UM) pts of 2 parts, B.1 and B.2. Primary objective: safety of the anticipated biologically effective dose (BED). Secondary objectives: PK; PD (e.g., inhibition of CD63 expression on basophils, changes in immune cell subsets in peripheral blood); RECIST 1.1 or Lugano-based responses; PFS and OS. Exploratory studies: phenotypic changes in circulating immune cells by mass cytometry (i.e., CyTOF); response assessments by radiomics; analysis of circulating protein signatures via Olink proteomics. Results: Part A Solid Tumour (completed): Sixteen pts were treated in 4 cohorts of 4 pts each. Pts characteristics: uveal melanoma (9/16; 56%), cutaneous melanoma (5/16; 31%) and pleural mesothelioma (2/16; 13%). Only Grade 1 and 2 AEs by CTCAE v.5 were observed, including 2 cases of transient (lasting < 24 hrs) diarrhoea and 2 of AST/ALT elevation. Part A FL Cohort (completed): 8 pts were treated at 20 mg (n=4) and 80 mg (n=4). Grade 3 toxicities were transient and observed in 2/8 (25%) pts. 2/4 pts (50%) at the 80 mg had PR as per Lugano criteria. Treatment-emergent adverse events (TEAEs) did not result in study drug discontinuation, immune-related toxicity, or dose-limiting toxicity in either the solid tumor or hematological patients. UM pts (Part A and Part B)>: 29 pts (Part A: 9 pts; Part B.1: 7 pts, Part B.2: 13 pts), of which 23 were treated at 80 mg QD (RP2D). Mean time on treatment: 10.7 mo (range: 1.5-39.6 mo). ORR (RECIST 1.1): PR: 1/29 (3%); SD: 21/29 (72%). Median OS for Part A: 20.8 mo (5/9 with 4 pts alive); Part B.1 20.1 (4/7 with 3 pts alive); Part B.2 not determined; 1-year OS rate: Part A 66.7%; Part B.1 71%; Part B.2 immature. Exploratory Investigations: At 3.7 mo (Cycle 5), there were 2 groups: one with SD (13/29; 45%) and the other with PD (16/29; 55%). At 6 mo, this split continued with 38% (11/29) showing SD and 62% having PD (18/29). Volumetric changes matched durable SDs while having higher sensitivity in the 15 analyzed patients. Interestingly, tumor burden reduction in mediastinal lymph nodes was indicative of SD and 70% of pts had volumetric spleen size reduction. Volumetric and radiomic changes were matched with peripheral changes in plasma proteins and immune cell changes. Conclusions: Roginolisib is well tolerated at 80 mg, demonstrates efficacy, and induces phenotypic changes in circulating immune cells. At the RP2D, the most robust anti-tumour activity was observed in FL (50% PR) and UM pts (3% PR; 72% SD). Clinical trial information: NCT04328844 .