Fractionated gemtuzumab ozogamicin (GO) with high-dose cytarabine (HiDAC) based induction chemotherapy is safe and highly effective for newly diagnosed favorable risk AML. Free

Several randomized trials establishedbenefit of GO when added to intensive chemotherapy for newly diagnosed AML, particularly for cytogenetically favorable-risk disease. Since the most benefit was seen with GO given on days 1, 4, and 7 (fractionated GO, GO3) in the context of 7+3, we subsequently conducted a single-arm phase 2 clinical trial of cladribine, high-dose cytarabine (HiDAC), G-CSF, and mitoxantrone (CLAG-M) + GO3. This trial indicated delayed relapse and improved survival over CLAG-M alone for patients (pts) with ELN 2017 defined favorable-risk disease and established HiDAC-based induction chemotherapy + GO3 as a standard for these pts. Here, we retrospectively analyzed all pts receiving HiDAC+GO3 to better understand the impact of individual cytogenetic/molecular abnormalities on response and survival.

We identified all adults with AML (n=139) or MDS (n=3) who received either CLAG-M or FLAG-IDA + GO3 via IRB-approved institutional database. Demographics, pathologic characteristics, treatment-related mortality (TRM) scores, disease responses, liver injury (AST/ALT ≥2.5x upper limit of normal or bilirubin ≥2 mg/dL), overall survival (OS), and relapse were obtained from medical records. Multiparameter flow cytometry (MFC)-defined measurable residual disease (MRD) and receipt of hematopoietic cell transplantation (HCT) was also recorded. Fisher exact and log-rank tests were used to compare quantitative/categorical variables and to time to events, respectively.

159 pts received CLAG-M (n=137) or FLAG-IDA (n=22) + GO3 between Aug 2018 and May 2025. Median [range] age was 59 [19-78] yrs. Most were fit as defined by a TRM score of <13.1 (86%, median=4) and/or ECOG performance status (PS) of 0-1 (n=113, 70%). Pts with ELN 2022 favorable risk (n=118, 74%) included those with mutations in NPM1 without FLT3 mutations (n=64, 50%), translocation of core binding factor [CBF; n=41, 26%], and CEBPA (n=9, 6%). The remaining 41 pts (26%) NPM1 and FLT3 mutations (intermediate risk).

5 pts (4%) died within 28 days. Liver injury during induction occurred in 6/159 (4%) pts, all nonfatal. No relationship was seen between liver injury and presenting ECOG PS, TRM, or whether therapy was administered as part of a clinical trial.

Complete remission (CR) or CR with incomplete hematologic recovery (CRi) was observed in 148/159 pts (93%); 88 pts (56%) achieved MRD-negative CR (CR_MRD-). Compared to intermediate-risk pts, those with favorable risk more frequently achieved CR (104/118 [88%] vs 28/41 [68%]; p=0.0069) and CR_MRD- (69/118 [59%] vs 19/41 [46%]; p=0.021) and had longer OS (HR=2.61, ref = favorable, 95% CI 1.35-5.02, p=0.004).

Among favorable risk pts, CR/CRi responses were more frequent among those with NPM1 (60/64 [94%]) or CBF (40/41 [93%]) abnormality than CEBPA mutation (6/9 [67%], p=0.047), acknowledging the limitation of this comparison by small cohort sizes. Rates of CR_MRD- were similar between pts with NPM1 (41/64 [66%]) CBF (22/41 [54%]), or CEBPA (6/9 [67%]) abnormalities, and there were no statistically significant differences in relapse-free survival (RFS) or OS among these favorable-risk pts.

FLT3-ITD PCR assay was positive in 17 pts. Among these, 12 obtained a CR (70%), but CR_MRD- was only achieved in 8/17 (47%). Compared to FLT3-ITD negative favorable-risk pts , OS among FLT3-ITD pts was shorter (p=0.01) and associated with earlier relapse (p=0.01) despite administration of a FLT3 inhibitor (FLT3i) in 12/17 pts. A trend towards earlier relapse remained despite treatment with FLT3i compared to other pts (p=0.08).

We investigated whether favorable-risk pts could achieve long term survival without the need for subsequent HCT. We conducted time-to-event analyses considering HCT, relapse, or death as events. HCT was performed in 22/127 (17%) pts, 12 pts received HCT in first remission. Reasons for HCT included inability to complete consolidation (n=2), molecular relapse (n=1), presence of mutations in BCOR/BCORL1, SRSF2, or PTPN11 (n=2 each), or MRD+ remissions (n=9). Only 1 favorable-risk pt died following HCT, and overall 75/118 (63%) remain event-free (median follow-up, 28.6 months).

Pts with favorable-risk AML and MDS/AML who received HiDAC-based induction chemotherapy +GO3 had low early mortality, infrequent liver injury, high rates of CR, and excellent survival outcomes, confirming this therapeutic strategy is well-tolerated and highly effective for favorable risk AML.