BACKGROUND:With gemtuzumab ozogamicin (GO; targeting CD33) and inotuzumab ozogamicin (IO; targeting CD22), 2 antibody-drug conjugates delivering a toxic calicheamicin (CLM) derivative have recently been approved for the treatment of people with acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL), respectively. While effective in some, many patients do not benefit from these ADCs. It is unclear to what degree anti-apoptotic BCL-2 family members are involved in modulating efficacy of CLM-based ADCs, with limited studies coming to differing conclusions. Given the clinical availability of small molecule inhibitors for BCL-2 family proteins (BCLi), here we clarify the impact of BCL-2 family proteins on the anti-leukemic activity of CLM-ADCs.MATERIALS AND METHODS:Human AML and ALL cell lines were engineered to overexpress BCL-2, BCL-XL, and MCL-1 via lentiviral gene transfer. AML and ALL cell lines as well as AML patient samples were exposed to increasing concentrations of GO or IO with or without the BCL-2 inhibitor ABT-199 (venetoclax), the BCL-2/BCL-XL inhibitor ABT-263 (navitoclax), and the MCL-1 inhibitor AZD5991. Dead cells were enumerated by flow cytometry via 4',6-diamidino-2-phenylindole staining after 72 hours. For BH3 profiling of AML patient specimens, thawed AML patient specimen aliquots were exposed to JC-1 mitochondrial dye and BH3 peptides, and peptide-induced depolarization was then calculated as a percent relative to a CCCP positive control, yielding a priming score for each BH3 peptide.RESULTS:At a dose of 1000 pg/ml, GO killing of ML-1 (AML) cells decreased from 56±5% (mean±SEM) in parental cells to 32±7% (p<0.01) and 26±6% (p<0.01) in cells overexpressing BCL-2 and BCL-XL, respectively (all n=3). Similar results were seen in another AML cell line (HL-60). In REH (ALL) cells treated with IO, overexpression of BCL family members also reduced killing - at 500 pg/ml, 59±8% of cells were killed in contrast to 12±1% (p<0.01) of BCL-2-expressing and 11±1% (p<0.01) of BCL-XL-expressing cells, with similar results seen in another ALL cell line (RS4;11). Addition of ABT-199 or ABT-263 at 1 µM modestly increased GO-mediated killing of AML cell lines - for example, ML-1 cells treated with GO at 100 pg/ml, cytotoxicity increased from 41±6% to 57±7% (ABT-199, p<0.01) and 61±8% (ABT-263, p<0.01). The effect of BCLi was more pronounced on IO-mediated killing of ALL cell lines than on GO-mediated killing of AML lines. For example, killing of REH cells treated with IO at 25 pg/ml increased from 39±7% (without BCLi) to 72±8% (ABT-199 1 µM, p<0.01) and 87±9% (ABT-263 1 µM, p<0.01), with similar results seen in RS4;11 cells. BH3 peptide profiling of AML patient specimens treated with GO implicated MCL-1 as a potential additional modulator of AML response to GO. Consistent with this finding, overexpression of MCL-1 reduced leukemia cell death in HL-60 cells treated with GO (GO at 1000 pg/ml, 41±2 % vs. 26±1 %, p=0.01) and RS4;11 cells treated with IO (IO at 100 pg/ml, 76±2% vs. 27±6%, p<0.01). The MCL-1 inhibitor AZD5991 modestly increased the anti-leukemic efficacy of GO in ML-1 cells and AML patient specimens, but more dramatically enhanced IO killing of REH cells (IO at 10 pg/ml, 18±2% without AZD5991 vs. 70±2% with 0.1 µM AZD5991, p<0.01). The triplet combination of GO, ABT-199 and AZD5991 did not improve markedly on the ABT-199/AZD5991 combination in the absence of GO in cell lines or AML patient specimens, though the triplet combination of IO, ABT-199 and AZD5991 showed promising activity: in REH cells treated with 10 pg/ml IO, cytotoxicity was 18±2% without BCLi, 32±8% with ABT-199 0.1 µM, 19±2% with AZD5991 0.01 µM, and 56±14% with the triplet combination (p<0.01 for comparison of triplet combination with IO/BCLi doublet).CONCLUSIONS:Our studies establish an important role of anti-apoptotic BCL-2 family members as resistance factor for CLM-based ADC therapy of acute leukemia. These findings provide the rationale to explore the combination of small-molecule inhibitors of BCL-2 family members with CLM-ADCs as a combination strategy in the clinic to improve the efficacy of GO and, particularly, IO. These therapeutic strategies may incorporate the assessment of the relative contribution of specific BCL-2 family members to an individual cancer patient's disease.DisclosuresJean: Eutropics Pharmaceuticals: Employment. Cardone:Eutropics Pharmaceuticals: Employment, Equity Ownership. Walter:Seattle Genetics: Research Funding; Kite Pharma: Consultancy; Daiichi Sankyo: Consultancy; Jazz Pharmaceuticals: Consultancy; Agios: Consultancy; Amgen: Consultancy; Amphivena Therapeutics: Consultancy, Equity Ownership; Aptevo Therapeutics: Consultancy, Research Funding; Argenx BVBA: Consultancy; Astellas: Consultancy; BioLineRx: Consultancy; BiVictriX: Consultancy; Boehringer Ingelheim: Consultancy; Boston Biomedical: Consultancy; Covagen: Consultancy; New Link Genetics: Consultancy; Pfizer: Consultancy, Research Funding; Race Oncology: Consultancy.