Moleculin reports 37% blinded CRc in venetoclax-failed AML patients from MIRACLE trial

Moleculin Biotech announced on July 31, 2026 updated preliminary blinded results from Part A of its pivotal Phase 2/3 MIRACLE trial (MB-108) of Annamycin in combination with cytarabine, referred to as AnnAraC, for relapsed or refractory acute myeloid leukemia. With 62 subjects evaluable to date, the preliminary blinded complete remission (CR) rate is 24% and the composite complete remission (CRc) rate is 37%.¹

Of those 62 subjects, 30, or 48%, had previously received a venetoclax-based regimen, and among that subgroup, the preliminary blinded CR and CRc rates were 23% and 37%, respectively, essentially identical to the evaluable population as a whole.¹ The company noted that a retrospective analysis of 41 patients with relapsed or refractory AML following failure of frontline venetoclax plus a hypomethylating agent found that 3 of 24 patients who received salvage therapy, or 13%, achieved complete remission or complete remission with incomplete hematologic recovery, and median overall survival from relapse or refractory disease onset was 2.4 months.¹

Across three successive blinded analyses, at 30, 45 and 62 evaluable subjects, the blinded CRc has remained within a narrow band of approximately 37% to 40%, while the proportion of subjects entering the trial after failure of a first-line venetoclax-based regimen has risen from 31.1% in the n=45 population to 48% today.¹

Enrollment stands at 74 of 90 subjects, with the company expecting to treat the 90th subject in September 2026 and unblinding of comprehensive Part A data anticipated in the December 2026 to February 2027 timeframe.¹ The blinded figures are expected to be lower than the unblinded Annamycin-arm results reported at the June 2026 interim analysis, in which CRc reached 50% and 57% in the two Annamycin arms versus 29% for control.¹ The company cautioned that the venetoclax-failure subgroup data are descriptive only, based on 30 subjects, include control-arm subjects, and that the trial is not powered for subgroup comparisons.¹