Ivosidenib in refractory or relapsed IDH1-mutated Acute Myeloid Leukemia patients in real life settings: The ivoobs observational study from the french AML intergroup ALFA/filo

Abstract Background IDH1 R132 mutations are found in about 5-10% of AML at diagnosis. Ivosidenib (IVO) is an oral, targeted, small-molecule inhibitor of the mIDH1 enzyme, approved for IDH1mut newly diagnosed AML aged ≥75 years or who are ineligible for intensive induction chemotherapy. In relapse/refractory settings (R/R), IVO monotherapy yielded promising complete remission or complete remission with partial hematologic recovery (CR/CRi) rate of 30.4% associated with a median overall survival (OS) of 8.8 months (Di Nardo, NEJM, 2018). However, there are very few data regarding IVO use outside clinical trials. In this study, we aimed to evaluate the efficacy and safety of IVO in R/R AML patients in real life settings. Method IVOOBS (NCT06377579) is a retrospective, non-interventional, multicentric study including patients from 32 French centers with newly diagnosed or R/R IDH1mut AML treated with IVO through a compassionate use program. Here, we focused on R/R patients treated with IVO, either as monotherapy or in combination with other therapies between January 2017 and February 2024. The primary objective was OS. Secondary objectives were response rate (ELN2022 criteria), and toxicity. Overall response (ORR) was defined as patients reaching CR, CRi, or CRh at any time. Results Overall, 127 patients were included. Secondary AML were observed in 15.8% (post-MDS/MPN=16, t-AML=4). Median number of previous lines prior IVO onset were 1 [IQR:1-3]; 82 patients (67%) received intensive chemotherapy as first line,36 (29% ) were pre-exposed to VEN prior IVO initiation, while 20 (15.8%) patients received IVO as post hematopoietic stem cell transplantation (HSCT) salvage treatment. 93 patients received IVO monotherapy, 26 in combination with azacitidine (AZA) and 8 with venetoclax (VEN) +/- AZA (defined as AZA/IVO (+/-VEN) group) Most frequent co-mutations were NPM1 (32%), RUNX1 (23%), ASXL1 (21%) and BCOR (18%). Clinicians reported differentiation syndrome (DS) of any grade in of 12 patients (9.5%) (8 with IVO and 4 with IVO+AZA (+/-VEN)). QTc prolongation and febrile neutropenia was observed in 8 (7%) and 15 (12%) patients respectively. Regarding grade 3-4 hematological adverse events (AE), neutropenia, thrombocytopenia and anemia occurred in 4%, 14% and 18% respectively. There were 2 grade 5 AE related to IVO (1 pneumocystis carinii pneumonia and 1 DS) both in IVO monotherapy treated patients. ORR (CR/CRi/CRh) rates was 45.9% (35.8%/9.2%/0.9%), with 46.6% showing no response and 4.6% MLFS (6 patients died prior evaluation). Median time to best response was 2.8 months. Median time on IVO treatment was 6.2 months. ORRs were 39%, 48.6% (p=0.55) with a median time to achieve ORR of 3 months, and 2.8 months in IVO and IVO/AZA (+/-VEN) treated patients, respectively. 65% (68/102) and 75% (55/73) of patients were transfusion independent at 3 and 6 months after IVO initiation, respectively. Prior VEN exposition did not significantly influence ORR probability: (36.1% (13/36) in VEN pre-exposed vs 44.4% (40/90) in VEN naïve (p=0.39). Co-mutations at AML diagnosis did not influence ORR probability including MAPK/RTK mutations. In multivariate analysis for ORR, only higher platelets at IVO onset (HR=1.05, p=0.01) and HU use (HR=0.13, p=.002) were independently associated with ORR. In the 22 patients who received HU to manage initial leukocytosis, ORR rate was 13.6% compared to 53.5% for those without HU requirement (p<0.001). In responding patients, 22.8% (13/57) were bridged to HSCT after a median time of 4.2 months. Only 2 patients relapsed post-transplant. After a median follow-up of 13.9 months, median OS (mOS) of the entire cohort was 14 months. mOS with IVO monotherapy and AZA/IVO (+/-VEN) was 13.2 and 20.2 months, respectively (p=0.16). VEN exposure pre-IVO did not significantly influence OS compared to VEN-naïve patients (HR=0.90, p=0.69). In multivariate analysis, only higher platelets (HR=0.96, p=0.015) and HU use (HR=2.95, p<.001) were independently associated with OS. Conclusion In this real-life study, IVO compares favourably with previously reported prospective studies in R/R settings, with a manageable safety profile. HU use for proliferative disease at IVO onset is associated with a lower response rate and inferior outcome.