Identification of Patient-Specific Anti-Apoptotic Molecules As Therapeutic Targets in Poor Prognosis Acute Lymphoblastic Leukemia (ALL)

Relapse and refractory ALL shows dismal prognosis despite recent progress in intensive chemotherapy and development of molecular targeting agents. In this study, we aimed to identify vulnerabilities in genetically-diverse ALL and to find additional therapeutic targets in Philadelphia chromosome-positive (Ph+) ALL and chronic myeloid myeloid leukemia (CML) to overcome BCR-ABL tyrosine kinase inhibitor (TKI) resistance.

To this end, we performed in vitro drug screening using 23 human ALL samples (9 Ph+ ALL, 5 MLL-ALL, 1 ALL with ETV-RUNX1, 5 B-ALL, not otherwise specified and 4 T-ALL), 10 CML samples and eight mixed phenotype acute leukemia (MPAL) samples (four B-Myeloid and four T-Myeloid). Based on our recent findings in AML (Hashimoto, Saito et al., Nature Cancer 2021), we treated leukemia-engrafting cells with small molecules targeting anti-apoptotic molecules (BCL-2, MCL-1 or BIRC) or molecules associated with cell division (AURKB) in addition to tyrosine kinase inhibitors.

Among these compounds, BIRC inhibitor and venetoclax exhibited great efficacy. Responsiveness to each compound was: 28 of 41 for BIRC inhibitor (68.3%), 24 of 40 for venetoclax (60%), 11 of 37 for MCL-1 inhibitor (29.7%) and 4 of 37 for AURKB inhibitor (10.8%). We found differential sensitivity between T-ALL/MPAL T-myeloid and CML. Seven out of eight T-ALL and MPAL T-myeloid samples were highly sensitive to venetoclax (87.5%), while nine out of 10 CML samples were responsive to BIRC inhibition (90%). On the other hand, among Ph+ ALL/MPAL and Ph- B-ALL/MPAL B-myeloid samples, sensitivities to BIRC inhibitor and venetoclax were variable.

To identify determinants of sensitivity to compounds, we examined the relation between mutational profile and in vitro leukemia elimination through targeted DNA sequencing for 79 lymphoid and myeloid malignancy-associated somatic mutations. Consistent with our previous study, CBL-mutated leukemia showed higher sensitivity to BIRC inhibitor (four of five cases) compared with venetoclax (two of five cases). Among genes related to RAS signaling pathway, KRAS mutations were most frequent (n=6). While five of six KRAS-mutated cases were BIRC inhibitor sensitive (83.3%), three of six cases were sensitive to venetoclax (50%). For cases with mutations in BCR-ABL1 kinase domain, the sensitivity to BIRC inhibitor and venetoclax was variable.

Finally, we went on to setup in vivo experiments to elucidate if targeting the patient-specific vulnerabilities resulted in potent therapeutic efficacy against patient leukemic cells. We created patient-derived xenograft (PDX) models of 5 Ph+ ALL/MPAL, 2 Ph- B-ALL/MPAL, 2 T-ALL/MPAL and 2 CML cases. For Ph+ ALL/MPAL and CML, BIRC inhibitor and/or venetoclax combined with dexamethasone (DEX) and TKI achieved effective in vivo elimination of leukemic cells as predicted by in vitro experiments. Combination therapy showed almost complete elimination of Ph+ leukemic cells even in the presence of T315I mutation. For Ph- B-ALL/MPAL and T-ALL/MPAL, preliminary in vivo experiments showed additional inhibition of BIRC and BCL-2 resulted in more profound reduction of leukemic cells in BM compared with DEX alone (Combination: 2.1% [0.25-8] of hCD45+ leukemic cells, n=13 vs. DEX alone: 67.2% [59-76.5], n=11, median [IQR], p<0.001).

We found targeting anti-apoptotic molecules in combination with DEX and/or TKI eradicated human genetically-diverse ALL, MPAL and CML cells both in vitro and in vivo. Inhibition of BIRC and BCL-2 overcame glucocorticoid resistance of Ph- ALL. Altogether, our results may offer precision medicine approach and contribute to improvement of clinical outcome in treatment-resistant ALL, CML and MPAL.