Multidimensional pharmacogenomic profiling in relapsed/refractory leukemias

Background Ex vivo Drug Response Profiling (DRP) tests patient-derived leukemia cells against panels of therapeutic agents to map sensitivity and resistance, providing a functional complement to genomic profiling. This approach is clinically feasible, can be completed within an actionable timeframe, and has demonstrated predictive value for treatment response in both myeloid and lymphoid leukemias. Methods Within the framework of the AL-TOMICA trial (NCT06626893), we established a regional network across Emilia-Romagna to centralize relapsed/refractory (R/R) acute leukemia (AL) cases. Primary samples underwent high-throughput DRP using a library of nearly 200 compounds, next-generation sequencing (NGS) for recurrent mutations, and bone marrow extracellular vesicle (EV) profiling to assess microenvironmental signaling. These data were integrated with standard diagnostic workups, including cytogenetics, FISH, and multiparametric flow cytometry, to obtain a comprehensive molecular and functional disease characterization. Results Between February 2022 and August 2025, 122 AL patients were enrolled, including AML (n = 104), B-ALL (n = 11), T-ALL (n = 5), and MPAL (n = 2). DRP was successfully performed in 68% of cases; the main reason for assay failure was insufficient blast recovery. In 30% of patients, combined DRP, NGS, and EV analyses were feasible. Unsupervised clustering of normalized IC₅₀ profiles identified six distinct response groups, ranging from chemoresistant to chemosensitive phenotypes. Resistant clusters were enriched in high-risk genotypes, suggesting that ex vivo sensitivity reflects intrinsic biological risk. Functionally, TP53-mutated samples exhibited resistance to MDM2 inhibitors, whereas FLT3-mutated cases showed selective sensitivity to ABL inhibitors. Post-MDS AMLs displayed enhanced response to PI3K/AKT/mTOR inhibition, consistent with known pathogenic mechanisms. EV analysis revealed that CD14 expression was inversely correlated with response to apoptosis-targeting drugs, indicating that microenvironmental monocytic signatures may influence therapeutic resistance. Conclusions The AL-TOMICA study demonstrates that functional precision medicine is feasible at a regional level and enhances therapeutic decision-making beyond genomic data alone. By combining functional, genetic, and microenvironmental profiling, at least one druggable vulnerability was identified in nearly all tested patients, outperforming molecular target–based approaches alone.