A Phase I Study of Autophagy Inhibition With Hydroxychloroquine and Oral Decitabine With Venetoclax After Hypomethylating Agent and Venetoclax Failure in Acute Myeloid Leukemia.

Relapsed / refractory (R/R) AML remains a therapeutic dead end, with real-world series reporting a median overall survival of under six months, even with intensive salvage approaches. Although the hypomethylating-agent/BCL-2 inhibitor doublet of decitabine (now available as the oral cedazuridine-boosted formulation) plus venetoclax yields encouraging remission rates, most responders ultimately relapse, underscoring an urgent need for resistance-modifying strategies. Converging pre-clinical and translational data show that AML blasts upregulate cytoprotective autophagy to evade BCL-2-mediated apoptosis; pharmacologic blockade of autophagosome-lysosome fusion with chloroquine analogues reverses this escape and synergizes with both venetoclax and cytotoxic backbones. Hydroxychloroquine, the only orally available autophagy inhibitor with a long safety record, therefore represents a pragmatic, mechanism-based addition to an all-oral decitabine / venetoclax regimen. The proposed Phase I study is significant because it will be the first to co-target aberrant DNA methylation, intrinsic apoptosis, and adaptive autophagy in R/R AML, aiming to convert transient responses into durable remissions while preserving outpatient convenience. Success would not only fill a critical therapeutic gap for patients who otherwise face dismal prognoses but also provide clinical proof-of-concept that autophagy inhibition can extend the utility of venetoclax-based regimens across hematologic malignancies.

Venetoclax (VEN), a selective BCL-2 inhibitor, in combination with a hypomethylating agent (HMA) or low-dose cytarabine, has re-defined first-line therapy for older or unfit adults with acute myeloid leukemia (AML).In VIALE-A, azacitidine + VEN yielded an approximately 65% composite complete remission (CR + CRi) rate and a median overall survival (mOS) of 14.7 months versus 8 months with azacitidine alone. Despite these gains, the majority of responders ultimately progress; up to 42% relapse by two years in long-term follow-up of VIALE-A, and real-world datasets mirror these findings.[ Outcomes after VEN failure are dismal. A 2025 multicenter Spanish series reported an mOS of only 2.3 months following relapse or primary refractoriness to VEN-HMA, with a salvage treatment rate of 30% and an overall response rate (ORR) of 23% among those treated. Smaller US single-institution cohorts and ASH abstracts have described similar post-VEN mOS figures of 6 months and highlight the absence of a standard salvage approach.Thus, venetoclax-resistant relapsed/refractory (R/R) AML represents a high-mortality population with an urgent unmet need.

AML stem and progenitor cells are often in a state referred to as "primed" for mitochondrial, or intrinsic, apoptosis-meaning they are close to the threshold of undergoing programmed cell death. This apoptotic readiness is governed by the balance between pro-apoptotic and anti-apoptotic members of the BCL-2 protein family within the mitochondria. In many AML subtypes, particularly in leukemic stem-like cells, survival is heavily dependent on the anti-apoptotic protein BCL-2, which functions by binding to and sequestering key pro-apoptotic effectors such as BIM and BAX. BIM is a BH3-only activator protein, and BAX is a pore-forming effector that oligomerizes to disrupt the mitochondrial outer membrane, a critical step in the initiation of apoptosis.

VEN, a selective BCL-2 inhibitor, disrupts this protective interaction by displacing BIM from BCL-2. Once liberated, BIM is able to activate BAX and/or BAK, leading to mitochondrial outer membrane permeabilization (MOMP). This event triggers the release of cytochrome c into the cytosol, which binds APAF1 and forms the apoptosome, culminating in the activation of caspase-9 and downstream executioner caspases such as caspase-3 and caspase-7, the hallmark steps of intrinsic apoptosis. The result is rapid and irreversible cell death in BCL-2-dependent AML blasts.

Clinical and translational studies have shown that venetoclax sensitivity varies by AML genetic subtype. The highest response rates are observed in AML with NPM1 or IDH1/2 mutations, which often exhibit strong BCL-2 dependence and low expression of alternative anti-apoptotic proteins, such as MCL-1. Similarly, secondary AML with spliceosome mutations (e.g., SRSF2, U2AF1, or SF3B1) demonstrates heightened apoptotic priming and increased venetoclax responsiveness. Conversely, monocytic AML subtypes tend to upregulate MCL-1 and BCL2A1, reducing reliance on BCL-2 and conferring intrinsic resistance to venetoclax. Thus, understanding the apoptotic dependencies of AML subtypes has become critical in tailoring venetoclax-based therapies.