AUTHORS
Amy Ji, Ermin Li, PhD2, Kehan Ren, PhD3
AFFILIATIONS
1 Columbia University, New York, New York, United States
2 Stony Brook University, Stony Brook, New York, United States
3 Stony Brook University, Stony Brook, New York, United States
ABSTRACT
Azacitidine plus venetoclax (Aza/Ven) has become an important therapeutic option for patients with high-risk myelodysplastic syndromes (MDS) and a standard treatment for newly diagnosed acute myeloid leukemia (AML) in patients ineligible for intensive induction chemotherapy. However, clinical responses remain highly heterogeneous, and there is no routinely available functional assay to predict individual patient responses before treatment initiation. Although ex vivo drug sensitivity testing has shown promise, conventional liquid-culture assays typically require several hundred thousand primary CD34+ cells, limiting their application in MDS, where diagnostic bone marrow aspiration samples often yield only limited numbers of CD34+ hematopoietic stem and progenitor cells (HSPCs). Furthermore, liquid-culture systems lack the three-dimensional stromal and extracellular matrix context that can influence leukemia-cell survival and therapeutic response. We previously developed a human induced pluripotent stem cell–derived bone marrow organoid that supports engraftment and maintenance of patient-derived HSPCs within a three-dimensional human marrow microenvironment. We therefore evaluated whether this bone marrow organoid engraftment platform could serve as a low cell-input functional drug sensitivity assay.
As a proof of concept, bone marrow–derived CD34+ cells from three newly diagnosed or Aza/Ven-naive AML patients were labeled with CellVue Far Red and engrafted into fully differentiated human bone marrow organoids (day 18 to day 21 of differentiation). After 48 hours of engraftment, organoids were treated with vehicle control (0.1% DMSO) or azacitidine (1 μM) plus venetoclax (200 nM) for an additional 48 hours. Two engrafted organoids were pooled to generate each biological replicate, and three replicates were analyzed per treatment group, corresponding to six engrafted organoids per condition. The pooled organoids were dissociated with collagenase, stained with Annexin V and DAPI, and analyzed by flow cytometry. Viability was defined as the percentage of Annexin V−/DAPI− cells within the CellVue+ engrafted population. All patients subsequently received Aza/Ven therapy, permitting retrospective comparison of ex vivo drug sensitivity with clinical response. Clinical response was assessed from post-treatment bone marrow blast percentages, and classified as response when blasts decreased to <5%, partial response when blasts remained between 5% and 20%, or resistant disease when blast percentages were unchanged or increased relative to pretreatment levels.
Ex vivo drug sensitivity closely mirrored clinical response across all three patients. The organoids engrafted with CD34+ cells from patient 1, who achieved a strong clinical response with reduction of marrow blasts from 80% to <5% after the first treatment cycle, demonstrated the greatest reduction in viable engrafted cells following Aza/Ven exposure, with a 72.2% decrease relative to vehicle-treated controls. In contrast, organoids generated from patients 2 and 3, whose disease was clinically resistant after the first and second treatment cycles, respectively, showed only modest decreases in viability of 25.9% and 22.3%. Thus, the rank order and magnitude of ex vivo drug sensitivity were concordant with the subsequent clinical outcomes of all three patients. Importantly, the organoid assay required only approximately 6 × 104 patient-derived CD34+ cells per treatment condition, compared with approximately 7 × 105 cells required by conventional liquid-culture assays. The platform therefore reduced primary-cell requirements by more than ten-fold while assessing drug response within a three-dimensional human bone marrow microenvironment containing stromal and extracellular matrix components.
These findings establish the feasibility of a human bone marrow organoid engraftment platform for low-input functional assessment of Aza/Ven sensitivity. Although AML samples were used here to demonstrate proof of concept and enable correlation with clinical response, the principal advantage of the platform is its ability to perform drug testing with limited numbers of patient-derived CD34+ cells. This feature supports future prospective application in MDS and measurable residual disease settings, where insufficient cell availability has been a major barrier to functional precision medicine.
DISCLOSURES
Amy Ji: No relevant conflicts of interest to declare.