Clinical Outcomes Using Frontline “Triplet” Regimens for Newly Diagnosed IDH-Mutated Acute Myeloid Leukemia (AML): A Pooled Analysis of Two Phase Ib/2 Clinical Trials

Background: Isocitrate dehydrogenase (IDH) mutations occur in ~20% of patients (pts) with acute myeloid leukemia (AML). Long-term follow up of the VIALE-A trial demonstrated a median overall survival (mOS) of 10.2 and 27.5 months for those with IDH1 and IDH2 mutations, respectively, with composite complete remission (CRc) rates of 66.7% and 86%. Targeted IDH inhibitors (IDHi) such as ivosidenib (IVO) and enasidenib (ENA) are also effective, either as single agents or in combination with azacitidine (AZA). All these regimens are effective, yet not curative, and when relapse occurs it is frequently due to the emergence or expansion of signaling mutations. “Triplet” regimens may more optimally prevent common resistance mechanisms by targeting multiple pathways simultaneously. In this pooled analysis, we report the clinical outcomes and patterns of relapse among pts with newly diagnosed (ND) IDH mutant AML who are not eligible for intensive chemotherapy treated with frontline triplet regimens containing a HMA + venetoclax (VEN) + IDHi.

Methods: Adults with ND IDH-mutated AML treated on IDHi “triplet” trials at our institution were included in this pooled analysis (NCT03471260 and NCT04774393). Pts received either frontline AZA + VEN + IVO or oral Decitabine/Cedazuridine (ASTX727) + VEN + IVO/ENA (arms for IDH1 or IDH2 mutant disease, respectively). OS, event free survival (EFS), and duration of remission (DOR) were assessed by Kaplan-Meier method, with EFS defined by lack of response, relapse, or death. CRc was defined as CR + CRi + CRh and overall response rate (ORR) as CR + CRi + CRh + PR + MLFS within five cycles.

Results: From October 2019 to March 2024, 50 pts with ND IDH-mutated AML were enrolled; 12 pts (24%) received AZA + VEN + IVO and 38 (76%) received ASTX727 + VEN + IDHi. Median age was 71 years (range: 62 – 87) and ELN 2022 risk was favorable for 10 (20%), intermediate for 4 (8%), and adverse for 36 (72%). Twelve pts (24%) had treated secondary AML (tsAML) with prior HMA ± VEN from antecedent MDS or MPN. NPM1 mutations were present in 12 (24%), FLT3-ITD in 3 (6%), K/NRAS in 10 (20%), and TP53 in 3 (6%).

At a median (m) follow up of 18.1 months, the mOS and mEFS have not been reached with a 2-year OS of 82% (95% CI: 70 – 95%), 2-year EFS of 63% (49 – 82%), and mDOR of 27 months. The ORR was 96% (48/50) with CRc of 92% (36/40). Measurable residual disease negativity by flow cytometry was achieved in 78% (36/46) of evaluable pts. Twelve responding pts (24%) transitioned to stem cell transplantation (SCT). For pts who did not undergo SCT, the median number of completed treatment cycles is 6 (IQR: 3 – 15) with 20/36 (56%) of responding pts still on active treatment.

There was no difference in outcomes between pts with IDH1 or IDH2 mutations, treatment with AZA + VEN + IVO versus ASTX727 + VEN + IDHi, or in pts with baseline signaling mutations. However, pts with tsAML experienced significantly worse outcomes compared to non-tsAML with a CRc rate of 75% (9/12) versus 98% (37/38) respectively (p=0.038). tsAML pts experienced mOS of 11 months, mEFS of 7.6 months, and mDOR of 8.5 months; treatment naïve participants experienced a 2-year OS of 94% (95% CI: 84 – 100%) and 92% (95% CI: 79 – 100%) for IDH1 and IDH2 mutant AML, respectively.

Ten of the 48 responding pts relapsed (21%). Five of the 6 pts with IDH1 mutations at baseline had no detectable IDH1 mutation at relapse, similar to 2 of the 4 pts with IDH2 mutations. Only 2 of 9 tested pts developed emergent signaling mutations: one gained a FLT3-ITD mutation along with new ETNK1, GNAS, SETBP1, and TET2 mutations, while the other with baseline NPM1, IDH2, and NRAS mutations gained a KRAS and a second NRAS mutation.

Overall the triplet regimens were well tolerated. 30-day and 60-day mortality were 0% and 2% (1/50), respectively. Grade 3 or higher non-hematologic adverse events were experienced in 10% of pts (5/50), including 2 cases of indirect hyperbilirubinemia and 1 case each of mucositis, QTc prolongation, and differentiation syndrome.

Conclusions: Pts with IDH mutant AML experience excellent outcomes with frontline HMA + VEN + IDHi triplets. There do not appear to be any recurrent mutational patterns at relapse on IDHi triplet regimens, with most IDH1 mutant pts and 50% of IDH2 mutant pts relapsing with an IDH wildtype clone. Given these promising frontline results, prospective studies comparing triplet versus doublet regimens for IDH mutant AML are warranted. Enrollment on both HMA + VEN + IDHi triplet trials is ongoing.