Preliminary results from RIVER-81, a phase 2 study of romaciclib (RVU120) + venetoclax in patients with acute myeloid leukemia failing first-line venetoclax + hypomethylating agent (HMA)

Abstract Background: First-line venetoclax (VEN) in combination with hypomethylating agents (HMA) is the standard of care for elderly patients with acute myeloid leukemia (AML), who are unfit for intensive chemotherapy. Despite improved overall survival (OS), approximately 70% of patients experience primary refractory or relapsed (R/R) disease, which is associated with a median OS of <3 months. Romaciclib (RVU120), a first-in-class CDK8/CDK19 inhibitor that showed single-agent activity in R/R AML in a Phase 1 trial (NCT04021368), has also demonstrated strong preclinical synergy with VEN in VEN-resistant AML models (Rzymski et al. Cancer Res, 2020). Mechanistically, romaciclib enhances VEN-induced apoptosis via caspase-dependent MCL-1 cleavage and broad transcriptomic reprogramming suppressing resistance-associated pathways like IL6/JAK/STAT3 and PI3K/AKT/MTOR. Based on this biological rationale, the Phase 2 RIVER-81 study (NCT06191263) is evaluating the efficacy and safety of romaciclib + VEN in patients who have failed front-line VEN+HMA; the study includes dose-optimization cohorts. Aims: The primary endpoint of RIVER-81 is the composite complete remission (CCR) rate, including complete remission (CR), CR with partial hematological recovery (CRh), and CR with incomplete count recovery (CRi), with or without minimal residual disease (MRD). Secondary endpoints include transfusion independence (TI), duration of response (DoR), OS, percentage of patients bridged to hematopoietic stem cell transplantation (HSCT), and pharmacokinetics (PK). Changes in BCL-2, BCL-XL, and MCL-1 in leukemic cells, and inhibition of STAT5 phosphorylation, are assessed as pharmacodynamic (PD) markers by flow cytometry. Methods: RIVER-81 is an ongoing study that comprises three parts. Part 1 is a 3+3 dose-escalation phase testing up to seven cohorts to establish the recommended dose (RD) of romaciclib + VEN for Part 2. DLTs are assessed at the end of Cycle 1. PK assessments confirm systemic exposure and support dose proportionality. Parts 2 and 3 will enroll up to 80 patients at the RD to assess responses per ELN 2022 criteria. Romaciclib + VEN are administered in 21-day cycles until progression. In Part 1 (Cohorts 1-3) and Stage 1 of Part 2, romaciclib was tested at 125-250 mg every other day (EOD) on days 1-13; VEN was given at 100-400 mg QD for 14 days. In Part 1 Cohorts 4 and 6, romaciclib was escalated to 150 and 200 mg QD, respectively, with VEN 400 mg QD for 14 days. All patients provided informed consent. Results: As of July 11, 2025, 48 patients were treated. Median age was 76 years (range: 47–86); 21 patients harbored high-risk mutations (ASXL1, TP53, NRAS). No DLTs were observed through Cohort 3 (romaciclib 250 mg EOD + VEN 400 mg QD), selected for Stage 1 of Part 2, and in Cohort 4 (romaciclib 150 mg QD + VEN 400 mg QD). No DLTs have been reported to date in ongoing Cohort 6 (romaciclib 200 mg QD + VEN 400 mg QD). PK analyses showed dose-proportional increases in Cmax and AUC for romaciclib, with no unexpected deviations from exposure targets. PD analysis of pSTAT5 confirmed robust target engagement. The most common adverse event (AE) was Grade 1-2 nausea (54%), followed by vomiting (42%), decreased appetite (25%), anemia (23%), febrile neutropenia (23%), and pneumonia (21%). There were 69 serious AEs (SAEs), with 5 assessed as at least possibly drug-related (acute cardiac failure x2, septic shock, gait disturbance, nausea). Twenty-eight patients were evaluable for response across Parts 1 and 2. One of 2 evaluable patients in Cohort 2 achieved CR. In Stage 1 of Part 2, 3 of 13 (23%) achieved CRi. In Cohort 4, 3 of 6 (50%) achieved CR (n=2) or CRi (n=1), the CRx rate in the intent-to-treat population was 43% (3/7). In Cohort 6, 1 CRi and 1 substantial blast reduction were reported in 2 evaluable patients; both are ongoing. Five of the 8 responding patients remain on treatment, with response durations ranging from 0.6 to 7 months. Four additional patients in Cohort 6 have not yet undergone disease evaluation. Conclusions:Romaciclib in combination with VEN shows promising anti-leukemic activity in a subset of patients with historically poor prognosis. The observed CR/CRi responses suggest that romaciclib may help overcome VEN resistance. The current data support continuation of the study. Enrollment and long-term follow-up are ongoing.