Initial Results from SELECT-AML-1, a Phase 2 Study of Tamibarotene in Combination with Venetoclax and Azacitidine in RARA-Positive Newly Diagnosed AML Patients Ineligible for Standard Induction Chemotherapy

Abstract

RARA-positive AML is a novel genomically-defined subset with an actionable target for treatment with tamibarotene, an oral and selective RARα agonist (McKeown, Cancer Discovery 2017). Approximately 30% of non-APL AML patients are RARA-positive based on overexpression of the RARA gene in peripheral blood blasts as determined by an RT-PCR based assay (de Botton, ESH 2019). In patients with newly diagnosed (ND) AML ineligible for standard induction chemotherapy, tamibarotene/azacitidine led to a CR/CRi rate of 61% in RARA-positive patients with a rapid onset of response. The combination was well-tolerated with no increase in myelosuppression compared to azacitidine (aza) alone (de Botton, ASH 2020). Responses were observed irrespective of mutations or cytogenetic risk. Approximately one-third of ND AML patients ineligible for standard induction therapy do not respond to venetoclax (ven) and nearly all eventually relapse (DiNardo, NEJM 2020). Translational data suggest that RARA-positive patients are enriched for monocytic features based on a positive Monocytic Expression Score (MES) that may be associated with ven resistance (Fiore, ASH 2020), suggesting that the blood-based RARA biomarker test selects for patients who may respond to tamibarotene and who may also be less likely to respond to ven/aza alone. SELECT-AML-1 (NCT04905407) is a Phase 2, open-label, multi-center study in the US and France comparing the activity of tamibarotene/ven/aza to ven/aza in treatment-naive RARA-positive ND AML patients ineligible for standard induction therapy. The 3-part study includes a safety lead-in of the tamibarotene/ven/aza combination, a randomized efficacy study versus ven/aza, and a salvage arm for ven/aza only patients with primary endpoints of safety, CR/CRi rate, and overall response rate, respectively. Following the safety lead-in, patients will be randomized 1:1 to receive tamibarotene/ven/aza or ven/aza. In the salvage arm, patients randomized to ven/aza only who experience progressive disease, treatment failure, or relapse, will have tamibarotene added to their ven/aza regimen. Patients currently enrolled in the safety lead-in are treated with starting doses of aza at 75 mg/m2 IV/SC daily on Days 1-7, ven on Days 1-28 per the ven USPI/SMPC, followed by tamibarotene at 6 mg twice per day PO on Days 8-28 of each 28-day cycle. Enrollment is ongoing. As of 22 June 2022, six patients have enrolled to date, including four who are response-evaluable. Baseline characteristics of the response-evaluable patients include 75% male; median age 57 (55-74); median bone marrow blasts 55% (39-100%); cytogenetically normal (n=2), inversion 16 (n=1), del 5q and -7 (n=1); and molecular genetic testing normal (n=3), mutations of IDH2, BCOR, SRSF2, CSF3R, RUNX1 (n=1). The MES was high in 3 of 4 (75%) response evaluable patients consistent with a prior report in RARA-positive AML (Fiore, ASH 2020). All 4 response-evaluable patients achieved a best response of CR/CRi (2 CR, 2 CRi). The onset of response was rapid with all patients achieving an initial response after Cycle 1 (2 CR, 1 CRi, 1 MLFS). Median duration of treatment was 68 days (21-110). Two patients discontinued, both in remission, one due to death from neutropenic sepsis and the other by physician decision. The adverse event (AE) profile of the triplet was consistent with reports of ven/aza in AML. AEs experienced by 2 or more patients (all grades/causality) included febrile neutropenia, pneumonia, anxiety, cough, diarrhea, fatigue, decreased appetite, myalgia, decreased neutrophil count, and decreased platelet count. The only serious AE (all causality) occurring in more than one patient was febrile neutropenia. A high response rate was observed in early results of tamibarotene/ven/aza in RARA-positive ND AML patients ineligible for standard induction chemotherapy with 100% (4 of 4) response evaluable patients achieving CR/CRi, 3 of whom may have features of ven-resistance based on high MES. The triplet demonstrated an initial safety profile similar to ven/aza. Data from the safety lead-in will inform triplet dosing for the randomized portion of the study.