BCL‐2 and MCL1 co‐inhibition enhances glucocorticoid therapy in preclinical models of infant KMT2A‐rearranged acute lymphoblastic leukaemia Authors: Mawar Karsa, Ayu Karsa, Dayna Spurling, Lin Xiao, Emma Ronca, Angelika Bongers, Weiman Gao, Andrew J Gifford, Ali El‐Ayoubi, Laurence C Cheung, Rishi S Kotecha, Richard B Lock, Charis E Teh, Guillaume Lessene, Murray D Norris, Andrew Wei, Michelle Haber, Michelle J Henderson, Donia M Moujalled, Klaartje Somers Published: 2026 To the Editor, Translocation of the mixed lineage leukaemia (MLL/KMT2A) gene is present in approximately 10% of acute leukaemias, including 80% of infant acute lymphoblastic leukaemia (ALL), and is associated with poor survival outcomes. 1 These rearrangements define a biologically distinct and clinically high‐risk subset of leukaemia, characterised by aggressive disease and inferior responses to conventional therapy. Although the bispecific T‐cell engager blinatumomab shows strong efficacy in infants with lysine methyltransferase 2A (KMT2A)‐rearranged (KMT2A‐r) leukaemia, 2 their outcomes remain poorer than those of older children, underscoring the need for therapies that better target the disease's underlying biology. There is growing evidence that the B‐cell lymphoma 2 (BCL‐2) family pro‐survival protein myeloid leukemia cell differentiation protein‐1 (MCL1) plays a key role in KMT2A‐r acute leukaemia disease progression and treatment resistance. MCL1 is a direct target gene of KMT2A fusion proteins 3 and is expressed at higher levels in infants than in older children with ALL. 4 Its silencing reduces KMT2A‐r leukaemia cell survival in vitro, and elevated MCL1 expression in infant ALL correlates with resistance to prednisolone, 4 a glucocorticoid widely used in frontline regimens, and the BCL‐2 inhibitor venetoclax. 5 However, pharmacological inhibition of MCL1 in infant KMT2A‐r leukaemia has yet to be comprehensively tested. To explore MCL1 as a therapeutic target, we evaluated the selective BCL‐2 Homology 3 (BH3)‐mimetic MCL1 inhibitor, S63845, 6 in preclinical models of KMT2A‐r acute leukaemia. To determine if S63845 selectively inhibited the viability and growth of KMT2A‐r leukaemia cell lines in vitro, the effect of the drug on the viability of a panel of cell lines with (KMT2A‐r) and without the translocation (KMT2A‐wildtype [KMT2A‐wt]) was assessed (Table S1). S63845 suppressed leukaemia cell survival and growth in a concentration‐dependent manner (Figure 1A,B) by inducing apoptosis (Figure 1C). Notably, KMT2A‐r leukaemia cell lines were significantly more sensitive to S63845 (50% Inhibitory Concentration (IC50) between 9 nM and 218 nM) compared to KMT2A‐wt leukaemia cell lines, which included acute myeloid leukaemia (AML), B‐cell ALL and T‐cell ALL cell lines (IC50 from 232 nM to >1.5 μM) (Figure 1A). This differential sensitivity supports a role for MCL1 in KMT2A‐r leukaemia cell survival. Strikingly, the four cell lines with IC50 for S63845 equal to or less than 50 nM (PER‐485, PER‐703, PER‐494 and KOPN‐8) were all derived from infants with KMT2A‐r ALL and the two most sensitive lines, PER‐485 and PER‐703 (IC50 of 9 nM and 10 nM, respectively) are characterised by co‐expression of lymphoid and myeloid markers. 7 To determine whether this vulnerability of cultured KMT2A‐r leukaemia cell lines extended to short‐term patient‐derived ex vivo models, we tested cultures from infant KMT2A‐r ALL patient‐derived xenografts (PDXs) (Table S2), where S63845 again potently reduced viability of leukaemic cells (IC50 between 10 nM and 228 nM) (Figure 1D). No significant correlations were observed between BCL‐2, MCL1 or B‐cell lymphoma‐extra large (Bcl‐XL) protein levels and S63845 IC50, although a trend was noted for a positive correlation between Bcl‐XL messenger ribonucleic acid (mRNA) expression and S63845 IC50 (Figure S1). At the transcript level, BCL‐2 and MCL1 mRNA showed no correlation with IC50, whereas Bcl‐XL mRNA correlated positively, indicating that higher Bcl‐XL expression is associated with reduced sensitivity to S63845 (Figure S1). Comparative analysis between KMT2A‐r and KMT2A‐wt leukaemias revealed that KMT2A‐r cells exhibited significantly lower Bcl‐XL protein levels (Figure S1). These findings suggest that the increased in vitro sensitivity of KMT2A‐r leukaemia cells to MCL1 inhibitor S63845 may reflect reduced dependence on Bcl‐XL compared to KMT2A‐wt cells. S63845 kills leukaemia cells and synergises with venetoclax. (A) Cytotoxicity curves (left and middle panels) of KMT2A‐rearranged (KMT2A‐r) and KMT2A‐wild‐type (KMT2A‐wt) leukaemia cell lines after a 72‐h treatment with S63845, as measured in resazurin reduction‐based cytotoxicity assays performed at least three times independently. IC50 values were determined by non‐linear regression. Right panel: Mean IC50 values were compared between KMT2A‐r and KMT2A‐wt cells by unpaired t‐test. ****p < 0.0001. (B) S63845 dose‐dependently decreases KMT2A‐r leukaemia cell growth as determined by live cell counting assays based on trypan blue exclusion (relative to seeding density on day 0 = 100%). Data represent mean of two (PER‐485, PER‐703, KOPN‐8) or three (PER‐494) independent runs. For each time point, mean live cell proportions between S63845 and vehicle‐treated cells were compared by unpaired t‐tests. (C) Increase in percentage of Annexin V‐positive cells (sum of Annexin V+ 7AAD– and AnnexinV+ and 7AAD+ cells after treatment, subtracted by sum of Annexin V+ 7AAD– and AnnexinV+ and 7AAD+ cells measured after vehicle treatment) measured by flow cytometry following treatment with S63845 for 24–72 h. The significance of increases in % of Annexin V+ cells was determined by one‐sample t‐tests. Asterisks represent significance levels of p‐values: *p < 0.05; **p < 0.01; ***p < 0.001. (D) Cytotoxicity curves of infant KMT2A‐r ALL PDXs following 72‐h ex vivo treatment with S63845. Mean viability data from at least three independent experiments (except for MLL‐7: n = 2) are shown. (E) Leukaemia cell lines were treated with S63845 combined with dexamethasone in a 6 × 6 matrix assay with incremental increases in concentrations of both drugs. Cell viability was measured by resazurin reduction‐based assays at 72 h. The plots display the synergy distribution according to the Bliss Independence model as visualised by Combenefit software. Representative plots of at least two independent runs are shown. (F) Leukaemia cells were treated with a sublethal concentration of S63845 combined with increasing concentrations of venetoclax for 72 h (in three independent experiments), followed by viability measurements by resazurin reduction‐based assays. (G) Leukaemia cell lines were treated with S63845 combined with venetoclax in a 6 × 6 matrix assay and cell viability after 72 h. Presented plots are representative for synergy distribution according to the Bliss Independence model as visualised by Combenefit software for at least two independent runs. (H) Western blot for cleaved‐PARP‐1, cleaved‐caspase 3, caspase 9 and actin (as loading control) on lysates from PER‐485 and PER‐494 cells following a 6‐h treatment with S63845 (PER‐485: 5 nM; PER‐494: 100 nM) and/or venetoclax (Ven) (PER‐485: 6 μM; PER‐494: 4 nM). Drug concentrations were selected based on viability and apoptosis data. Representative blots of two independent runs are shown. Error bars represent the standard error of the mean (standard error of mean [SEM]). (I) Triple combination matrix plots for dexamethasone‐resistant (PER‐485) and ‐responsive (KOPN‐8) KMT2A‐r leukaemia cell lines. The plots display the dose response after 72‐h treatment with S63845, Ven and dexamethasone (Dex) as measured in resazurin reduction‐based cytotoxicity assays performed at least three times independently (left panel) and the synergy score of each treatment group according to the Bliss model (right panel). Mean synergy scores between groups were compared by one‐way ANOVA followed by multiple comparisons. Asterisks represent significance levels of p‐values: *p < 0.05; **, p < 0.01; ns, not significant. To achieve meaningful survival extensions in aggressive cancers such as KMT2A‐r leukaemia, single‐agent therapy is rarely sufficient, and rational drug combinations are required. As MCL1 has been reported to mediate resistance to both venetoclax and glucocorticoids, 4 , 5 we evaluated the efficacy of the combination of S63845 with venetoclax or the glucocorticoid dexamethasone (Dex). In in vitro viability assays, S63845 and Dex displayed additive or slightly synergistic effects against KMT2A‐r leukaemia cells, with synergy assessed by the Bliss independence (Figure 1E) or ZIP model (Figure S2). However, the addition of a sublethal concentration of S63845 strongly potentiated venetoclax activity in both venetoclax‐resistant (PER‐485, THP‐1) and venetoclax‐responsive (PER‐703) KMT2A‐r leukaemia cells and reduced the amount of venetoclax needed to achieve a 50% viability inhibition by up to 10 000‐fold (Figure 1F). Conversely, increasing concentrations of venetoclax resulted in a dose‐dependent increase in sensitivity to S63845 across multiple KMT2A‐r leukaemia cell lines, supporting a functional shift towards MCL1 dependency upon BCL‐2 inhibition (Figure S3). Consistent with these findings, synergy assays confirmed strong synergy between S63845 and venetoclax in KMT2A‐r leukaemia cell lines (Figure 1G; Figure S4) and the combination further enhanced apoptosis induction, evidenced by cleaved caspase 3, 9 and Poly (ADP‐ribose) polymerase 1 (PARP‐1) (Figure 1H). Given that both BCL‐2 and MCL‐1 are known to contribute to glucocorticoid resistance, we next performed triple combination synergy assays in glucocorticoid‐sensitive (KOPN‐8) and glucocorticoid‐resistant (PER‐485) KMT2A‐r leukaemia cell lines. Dex and venetoclax displayed synergy across both cell lines, which was further enhanced by the addition of the MCL1 inhibitor S63845, resulting in robust triple combination synergy (Figure 1I; Figure S5). Collectively, these data further establish MCL1 as a mediator of venetoclax resistance in KMT2A‐r leukaemia and support an additional synergistic contribution of MCL1 inhibition to the efficacy of venetoclax combined with Dex. Building on these promising in vitro findings, we next assessed whether S63845 was able to limit leukaemia progression and enhance the activity of venetoclax and/or Dex in PDXs derived from infant KMT2A‐r ALL cases. 8 Leukaemia‐engrafted NOD scid gamma (NSG) or non‐obese diabetic (NOD)/severe combined immunodeficiency (SCID) mice were dosed with single, dual combination or triple combination regimens of S63845, venetoclax and Dex with previously established drug administration schedules. 6 , 9 Dosing S63845 twice weekly at 12.5 mg/kg or higher, combined with venetoclax (25 mg/kg) induced lethal tumour lysis syndrome (TLS), characterised by rapidly induced lethargy, hind limb paralysis and laboured breathing in 67% of treated mice (8/12 mice), with histopathological evidence of debris accumulation in kidney and lung (Figure S6). In contrast, TLS was not observed following either monotherapy or with concurrent treatment with Dex (10 mg/kg). Notably, TLS was previously reported for the combination of venetoclax with S63845 in B‐ALL xenograft mouse models. 9 Although it indicates a potent anti‐leukaemic effect of the S63845 and venetoclax combination, the rapid lethality of TLS prevented assessment of its impact on long‐term leukaemia progression. Therefore, we next tested S63845 at a reduced dose of 6.25 mg/kg that did not induce TLS when combined with venetoclax (at 25 mg/kg), in two infant KMT2A‐r ALL PDX models (MLL‐7 and MLL‐14). S63845 monotherapy did not exert a significant effect on disease progression or survival (Figure 2A,B; Table S3). Even at higher doses previously shown to be effective in other blood cancer models (25 and 50 mg/kg), 6 , 9 only minor attenuation of leukaemia growth or minimal survival benefit was achieved (Figure S7). Venetoclax significantly delayed leukaemia progression in MLL‐7 PDXs, albeit only modestly extending survival, but not in MLL‐14 PDXs (MLL‐7: p = 0.0124; MLL‐14: p = 0.0913, Table S3). In both MLL‐7 and MLL‐14 PDX models, addition of S63845 to venetoclax did not extend survival significantly more than venetoclax alone (MLL‐7: p = 0.074; MLL‐14: p = 0.1779, Figure 2A,B; Table S3). Combination of S63845, venetoclax (Ven) and dexamethasone (Dex) delays leukaemia progression in patient‐derived xenograft (PDX) models of infant KMT2A‐rearranged acute lymphoblastic leukaemia (ALL). (A, B) In vivo engraftment level (% human CD45+ cells in peripheral blood (PB), left panel) and survival (right panel) of (A) MLL‐7 and (B) MLL‐14 PDX‐engrafted mice (combined from two independent experiments) treated with vehicle, S63845, Ven, Dex or combinations thereof according to the following schedules: S63845, 6.25 mg/kg intravenously, 1–2×/week for up to 3 weeks; Ven, 25 mg/kg orally, 3–5×/week for up to 4 weeks; Dex, 10 mg/kg intraperitoneally, 3–5×/week for up to 3 weeks. For combination arms, each drug was administered according to their individual regimen. Treatments were suspended when animal weight decreased >10% until weight stabilisation. Gehan–Wilcoxon was used to compare survival between treatment groups. Asterisks represent significance levels of p‐values: *p < 0.05, **p < 0.01; ***p < 0.001. (C) Proportion of MLL‐7 or MLL‐14 PDX‐engrafted mice achieving specified objective response measure (ORM) scores: progressive disease 1 (PD1), progressive disease 2 (PD2), stable disease (SD), partial response (PR), complete response (CR) and maintained complete response (MCR). To explore mechanisms underlying the limited in vivo activity of S63845 (alone or with venetoclax), we performed pharmacodynamic analysis through western blotting on splenocytes from treated mice. Expression of the pro‐apoptotic BIM protein showed a consistent downward trend 16 h after a single S63845 dose (25 mg/kg) and again at end‐point in S63845 and S63845 + venetoclax cohorts (observed in 2/3 mice) without a consistent effect on Bcl‐XL levels (Figure S8A,B). A reduction in the levels of BIM, an activator that facilitates MCL1/BCL‐2 inhibitor‐induced apoptosis, is a possible explanation for the decreased in vivo sensitivity to S63845 ± venetoclax in these models. While Dex alone initially depleted leukaemia cells in peripheral blood and delayed leukaemia progression, mice relapsed rapidly once treatment was halted, mirroring clinical relapse in infant ALL patients (Figure 2A,B). While the addition of S63845 did not extend Dex‐induced survival (Figure 2A,B; Table S3), the combination of venetoclax and Dex significantly prolonged survival in MLL‐7 and MLL‐14 PDXs by 64% and 49%, respectively, compared to Dex monotherapy (MLL‐7: p = 0.0002; MLL‐14: p = 0.0011, Figure 2A,B; Table S3). Most importantly, the triple treatment nearly doubled survival relative to Dex alone (MLL‐7: 68 vs. 37 days; MLL‐14: 61 vs. 32 days) while being well tolerated (Figure S9). The triple combination regimen was significantly more effective than combined venetoclax and Dex treatment in MLL‐14 PDXs (p = 0.0227) (Table S3), while a similar trend observed for MLL‐7 was not significant. Increasing the dose of S63845 in the triple combination to 25 mg/kg did not increase the drug combination's efficacy in the MLL‐14 PDX, even though it was well tolerated (Figure S10). Importantly, these treatment regimens did not result in sustained depletion of major haematopoietic stem cell populations, with recovery of white blood cells, red blood cells and platelets over time, indicating preservation of normal haematopoiesis and supporting the long‐term safety of these combinations on the haematopoietic system (Figure S11). Next, we assessed treatment response in the PDX models according to Objective Response Measure (ORM) criteria (MCR, maintained complete response; CR, complete response; PR, partial response; PD2, progressive disease 2; PD1, progressive disease 1), modelled after the principles used in the clinic to assess therapeutic efficacy by evaluating leukaemia blood levels in the mice. 10 S63845 failed to induce responses (PR, CR or MRC) in any of the treated mice (Figure 2C). Similar results were obtained for venetoclax in MLL‐14‐engrafted mice, while one of seven MLL‐7 PDX mice achieved PR. In the MLL‐7 model, all mice treated with the triple combination achieved MCR (7/7, 100%), compared to five of eight (62.5%) treated with Dex and venetoclax (Figure 2C). The percentage of MLL‐7‐engrafted mice achieving MCR after venetoclax + Dex treatment versus those achieving MCR after the triple combination of S63845 + venetoclax + Dex is significantly increased (p < 0.0001; Figure S12), supporting a biologically and clinically relevant difference in efficacy in this model. In the MLL‐14 PDX, the triple combination achieved MCR in 7/8 (87.5%) mice versus 3/8 (37.5%) with Dex and venetoclax, confirming the additional benefit of adding S63845 in this more venetoclax‐resistant model. The addition of S63845 to venetoclax + Dex seems more impactful in the venetoclax‐resistant MLL‐14 versus the venetoclax‐responsive MLL‐7 PDX, but this is not associated with higher baseline MCL‐1 or lower Bcl‐XL expression levels in MLL‐14 versus MLL‐7 (Figure S13). Collectively, these results demonstrate that co‐targeting BCL‐2 and MCL1 combined with Dex provides the most durable benefit, with the triple regimen delivering the greatest survival extension and treatment response among regimens tested. Dex is known to induce apoptosis in lymphoid cells through upregulation of pro‐apoptotic BH3‐only proteins, including BIM, thereby enhancing mitochondrial apoptotic priming. In the context of BCL‐2/MCL1 co‐inhibition, this increased apoptotic priming is expected to further sensitise leukaemia cells to BH3‐mimetic–induced cell death. In this study, we observed that S63845 treatment was associated with reduced BIM expression in vivo, which may contribute to the diminished sensitivity to BH3‐mimetic therapy. The addition of Dex may help overcome this limitation by restoring or enhancing pro‐apoptotic signalling, thereby increasing the effectiveness of combined BCL‐2 and MCL1 inhibition. To further evaluate the clinical potential of the triple combination of S63845/venetoclax/Dex, we compared its efficacy against that of a previously established induction‐type chemotherapy regimen combining vincristine, Dex and asparaginase (i.e. VXL). 11 S63845/venetoclax/Dex achieved MCR in 7/7 MLL‐7 and 7/8 MLL‐14 mice, on par with response rates induced by the combination of three chemotherapies (Figure S14). Taken together, our data establish dual MCL1/BCL‐2 targeting as a promising strategy to potentiate glucocorticoid therapy in infant KMT2A‐r ALL and potentially other acute leukaemias. While the combination of venetoclax and S63845 induced TLS in mice, the addition of Dex not only prevented this toxicity, potentially by keeping circulating leukaemia levels low, but also produced more durable responses. Clinically, several MCL1 inhibitors 12 , 13 have entered early‐phase trials for haematological malignancies, with TLS reported in adult patients. 12 Our findings suggest that sequential treatment schedules incorporating Dex may improve both the safety and efficacy of dual MCL1/BCL‐2 inhibition. Our data suggest that venetoclax‐resistant KMT2A‐r leukaemias might benefit more from additional MCL‐1 targeting, but given the increasing recognition that KMT2A‐r leukaemia constitutes a highly heterogeneous disease entity, 14 careful clinical trial design, including stringent dose and patient selection, sequential scheduling and proactive monitoring for TLS will be essential to safely translate these strategies into the treatment of infants with KMT2A‐r ALL. AUTHOR CONTRIBUTIONS Mawar Karsa: Methodology; formal analysis; resources; data curation. Dayna Spurling: Methodology. Andrew J. Gifford: Methodology. Angelika Bongers: Methodology. Ali El‐Ayoubi: Methodology. Lin Xiao: Formal analysis; resources. Laurence C. Cheung: Writing – review and editing. Ayu Karsa: Methodology. Murray D. Norris: Writing – review and editing. Weiman Gao: Methodology. Emma Ronca: Methodology. Charis E. Teh: Writing – review and editing. Richard B. Lock: Writing – review and editing. Rishi S. Kotecha: Writing – review and editing. Klaartje Somers: Conceptualization; writing – original draft; supervision; funding acquisition; project administration; formal analysis; data curation; resources; writing – review and editing. Guillaume Lessene: Writing – review and editing. Michelle Haber: Writing – review and editing. Andrew Wei: Writing – review and editing. Michelle J. Henderson: Writing – review and editing. Donia M. Moujalled: Conceptualization; supervision; writing – original draft; project administration; writing – review and editing. CONFLICT OF INTEREST STATEMENT Donia M. Moujalled, Andrew Wei, Charis E. Teh and Guillaume Lessene are employees at the Walter and Eliza Hall Institute, which receives milestone and royalty payments related to venetoclax. The remaining authors declare no competing financial interests. Supporting information Figure S1: Baseline expression of pro‐survival BCL‐2 family proteins in KMT2A‐rearranged (KMT2A‐r) and KMT2A‐wild‐type (KMT2A‐wt) leukaemia cell lines. (A) Protein expression levels of BCL‐2, MCL1 and BCL‐XL as determined by immunoblotting. Representative immunoblot for two independent experiments is shown. The left blot includes KMT2A‐r leukaemia cells, the right blot KMT2A‐wt leukaemia cells, except for RS4;11, a KMT2A‐r leukaemia cell line included in both blots to allow normalisation for quantification across the blots. KMT2A‐r leukaemia cell lines (left panel) are arranged left‐to‐right by decreasing sensitivity to the MCL1 inhibitor S63845. Untreated whole‐cell lysates were immunoblotted with whole‐protein Ponceau staining and β‐actin serving as loading controls. (B) Correlation between relative BCL‐2 family protein levels and S63845 IC50 based on quantification of protein bands by densitometry, normalised to Ponceau. Significance of correlations was determined by Pearson. (C) Correlation between relative BCL‐2 family mRNA levels and S63845 IC50. Each dot represents the mean relative expression relative to the expression in PER‐485 cells (normalised to the geomean of GUSB and HPRT expression as housekeeping genes) calculated based on three independent experiments. Significance of correlations were determined by Pearson. (D) Comparison of relative protein expression levels of BCL‐2 family proteins between KMT2A‐r and KMT2A‐wt leukaemia cells. Expression levels between KMT2A‐r and KMT2A‐wt cells were compared by unpaired t‐tests. (E) Comparison of relative mRNA expression levels of BCL‐2 family proteins between KMT2A‐r and KMT2A‐wt leukaemia cells. Expression levels between KMT2A‐r and KMT2A‐wt cells were compared by unpaired t‐tests. Figure S2: Analysis of matrix synergy assays of S63845 and dexamethasone (Dex) depicted in Figure 1E according to the ZERO Interaction Potency (ZIP) model. The plots display the synergy distribution according to the ZIP model as visualised by Synergy Finder software. Representative data from at least two independent experiments are shown. Figure S3: MCL1 dependency increases upon venetoclax treatment in KMT2A‐r acute leukaemia cells. Upon increasing concentrations of venetoclax, S63845 IC50 decreases. Data depicted are from at least two independent experiments (corresponding to data presented in Figure 1G), with error bars representing SEM. Figure S4: Analysis of matrix synergy assays of S63845 and venetoclax (Ven) depicted in Figure 1G according to the ZERO Interaction Potency (ZIP) model. The plots display the synergy distribution according to the ZIP model as visualised by Synergy Finder software. Representative data from at least two independent experiments are shown. Figure S5: Triple combination matrix synergy assays for dexamethasone‐resistant (PER‐485) and ‐responsive (KOPN‐8) KMT2A‐r leukaemia cell lines. Synergy score of treatment groups according to the Zero Interaction Potency model generated by the Synergy Finder software from at least three independent experiments. Mean synergy scores between groups were compared by one‐way ANOVA followed by multiple comparisons. Error bars represent SEM. Asterisks represent significance levels of p‐values: *p < 0.05; **p < 0.01; ns, not significant. Figure S6: Tumour lysis syndrome and organ pathology in mice treated with the combination of high dose S63845 and venetoclax. Mice (n = 3) were treated with 25 mg/kg S63845, followed by a single administration of 25 mg/kg venetoclax 6 h later. Organs were harvested 1 h later. Representative haematoxylin and eosin sections illustrating organ damage associated with tumour lysis syndrome. (A) Kidney showing accumulation of cellular debris within glomerular capillaries (objective ×600 magnification) indicated by arrow. (B) Lung showing accumulation of cellular debris within a pulmonary artery branch (objective ×400 magnification), indicated by arrow. Images are representative for observations in three mice. Figure S7: No significant survival benefit is observed in response to S63845 monotherapy at higher doses. Engraftment level (left panel) and survival (right panel) of mixed lineage leukaemia (MLL)‐7 PDX‐engrafted mice treated with (A) 25 mg/kg S63845 intravenously twice weekly for 2 weeks, and (B) 50 mg/kg S63845 on 5 consecutive days for 1 week. Figure S8: Attenuated BIM protein expression after in vivo MCL1 inhibition. (A) Western blot for BIM (several isoforms) on lysates from splenocytes isolated from MLL‐7 or MLL‐14 engrafted mice 16 h after one administration of vehicle or 25 mg/kg S63845 (n = 3/group) (left panel). Lanes show individual mice. Quantification of protein bands by densitometry, normalised to β‐actin and to mean expression in the vehicle‐treated mice (right panel). t‐Tests were used to compare expression levels between groups. (B) Western blot on lysates from spleen samples from MLL‐14 engrafted mice (n = 3/group) after reaching end‐point following treatment with vehicle, S63845, venetoclax or the S63845 and venetoclax combination according to the schedule described in the legend of Figure 1. Western blots for BIM (total and upper band) with β‐actin loading control show trends towards lower BIM levels in S63845 and combination arms (observed in 2/3 mice) (left panel). Quantification of protein bands by densitometry, normalised to β‐actin and to mean expression in the vehicle‐treated mice (right panel). Mean expression levels between groups were compared by one‐way ANOVA. ns, not significant. Figure S9: The triple combination of dexamethasone (Dex), venetoclax (Ven) and S63845 is well tolerated. Mean weight of (A) MLL‐7‐ and (B) MLL‐14‐engrafted mice treated with Dex or the triple combination of Dex, Ven and S63845 (S63845 + Ven + Dex) included in the efficacy studies depicted in Figure 2. Error bars represent standard error of the mean. Figure S10: Increasing the dose of S63845 in the triple combination to 25 mg/kg does not increase the drug combination's efficacy in the MLL‐14 PDX. (A) Engraftment level (%human CD45+ cells in peripheral blood (PB), left panel) and survival (right panel) of MLL‐14 PDX‐engrafted mice treated with vehicle, S63845 at 6.25 mg/kg, 12.5 mg/kg or 25 mg/kg intravenously 1–2×/week for up to 3 weeks, or triple combination of venetoclax 25 mg/kg orally, 3–5×/week for up to 4 weeks; dexamethasone 10 mg/kg intraperitoneally, 3–5×/week for up to 3 weeks, and the respective dose of S63845, similar to the single agent schedule. The data presented in this figure for mice treated with drug schedules including 6.25 mg/kg S63845 are also included in Figure 2B. (B) Mean weight of the same cohort of mice. Error bars represent standard error of the mean. Figure S11: Treatment with double and triple combinations of S63845, venetoclax (Ven) and dexamethasone (Dex) does not cause long‐term depletion of haematopoietic stem cell subsets including white blood cells (WBC), red blood cells (RBC) and platelets. NSG mice (4 mice/group) were treated with vehicle, Ven + Dex, S63845 + Ven or the triple combination according to the following schedules: S63845, 6.25 mg/kg intravenously, 1–2×/week for 3 weeks; Ven, 25 mg/kg orally, 3–5×/week for 3 weeks; Dex, 10 mg/kg intraperitoneally, 3–5×/week for 3 weeks. Left panel depicts absolute changes. Right panel shows change in levels compared to day of treatment start. Grey‐shaded area corresponds to treatment period. Error bars represent SEM. Figure S12: The triple combination of S63845/venetoclax (Ven)/dexamethasone (Dex) induces better response rates than Ven/Dex in the MLL‐7 PDX. The percentage of MLL‐7 PDX‐engrafted mice achieving maintained complete response (MCR) in response to treatment with the triple combination of S63845 + Ven + Dex is significantly greater than that achieved by Ven + Dex. Fisher's exact test was performed to compare proportions. Figure S13: Baseline protein expression of pro‐survival BCL‐2 family proteins in infant KMT2A‐rearranged ALL PDXs. Representative immunoblot for two independent experiments. Untreated splenocyte lysates were immunoblotted for BCL‐2, MCL1 and BCL‐XL with whole‐protein Ponceau staining and β‐actin serving as loading controls (left panel). Quantification of protein bands by densitometry, normalised to β‐actin and to MLL‐2 expression levels (right panel). Error bars represent SEM. Figure S14: In vivo response of infant KMT2A‐r ALL PDXs following an induction‐type combination regimen of vincristine, dexamethasone and asparaginase (VXL). Survival curve of (A) MLL‐7 and (B) MLL‐14 PDX engrafted mice treated with vehicle or a combination of vincristine (0.5 mg/kg, once weekly), dexamethasone (5 mg/kg, 5×/week) and asparaginase (1000 KU/kg, 5×/week) intraperitoneally for 2 weeks. Gehan–Wilcoxon was used to compare survival between treatment groups. ***p < 0.001. Table S1: Characteristics of leukaemia cell lines used in this study. Table S2: Characteristics of leukaemia patient‐derived xenografts (PDX) used in this study. Table S3: Summary of in vivo responses of S63845, venetoclax and dexamethasone as single agents and in combination. ACKNOWLEDGEMENTS This study was funded by a project grant from Kids Cancer Alliance to Klaartje Somers. Rishi S. Kotecha and Michelle Haber are supported by Fellowships of the National Health and Medical Research Council of Australia (Rishi S. Kotecha: NHMRC APP2033152, Michelle Haber: APP2017256). Children's Cancer Institute is affiliated with UNSW Sydney and the Sydney Children's Hospitals Network. The authors would like to thank the Animal Facility and Core Services at Children's Cancer Institute for providing valuable technical assistance with the experiments executed during this project. Donia M. Moujalled and Klaartje Somers—Joint senior authors. Contributor Information Donia M. Moujalled, Email: moujalled.d@wehi.edu.au. Klaartje Somers, Email: ksomers@ccia.org.au. DATA AVAILABILITY STATEMENT Data are available upon request. REFERENCES Associated Data Supplementary Materials Figure S1: Baseline expression of pro‐survival BCL‐2 family proteins in KMT2A‐rearranged (KMT2A‐r) and KMT2A‐wild‐type (KMT2A‐wt) leukaemia cell lines. (A) Protein expression levels of BCL‐2, MCL1 and BCL‐XL as determined by immunoblotting. Representative immunoblot for two independent experiments is shown. The left blot includes KMT2A‐r leukaemia cells, the right blot KMT2A‐wt leukaemia cells, except for RS4;11, a KMT2A‐r leukaemia cell line included in both blots to allow normalisation for quantification across the blots. KMT2A‐r leukaemia cell lines (left panel) are arranged left‐to‐right by decreasing sensitivity to the MCL1 inhibitor S63845. Untreated whole‐cell lysates were immunoblotted with whole‐protein Ponceau staining and β‐actin serving as loading controls. (B) Correlation between relative BCL‐2 family protein levels and S63845 IC50 based on quantification of protein bands by densitometry, normalised to Ponceau. Significance of correlations was determined by Pearson. (C) Correlation between relative BCL‐2 family mRNA levels and S63845 IC50. Each dot represents the mean relative expression relative to the expression in PER‐485 cells (normalised to the geomean of GUSB and HPRT expression as housekeeping genes) calculated based on three independent experiments. Significance of correlations were determined by Pearson. (D) Comparison of relative protein expression levels of BCL‐2 family proteins between KMT2A‐r and KMT2A‐wt leukaemia cells. Expression levels between KMT2A‐r and KMT2A‐wt cells were compared by unpaired t‐tests. (E) Comparison of relative mRNA expression levels of BCL‐2 family proteins between KMT2A‐r and KMT2A‐wt leukaemia cells. Expression levels between KMT2A‐r and KMT2A‐wt cells were compared by unpaired t‐tests. Figure S2: Analysis of matrix synergy assays of S63845 and dexamethasone (Dex) depicted in Figure 1E according to the ZERO Interaction Potency (ZIP) model. The plots display the synergy distribution according to the ZIP model as visualised by Synergy Finder software. Representative data from at least two independent experiments are shown. Figure S3: MCL1 dependency increases upon venetoclax treatment in KMT2A‐r acute leukaemia cells. Upon increasing concentrations of venetoclax, S63845 IC50 decreases. Data depicted are from at least two independent experiments (corresponding to data presented in Figure 1G), with error bars representing SEM. Figure S4: Analysis of matrix synergy assays of S63845 and venetoclax (Ven) depicted in Figure 1G according to the ZERO Interaction Potency (ZIP) model. The plots display the synergy distribution according to the ZIP model as visualised by Synergy Finder software. Representative data from at least two independent experiments are shown. Figure S5: Triple combination matrix synergy assays for dexamethasone‐resistant (PER‐485) and ‐responsive (KOPN‐8) KMT2A‐r leukaemia cell lines. Synergy score of treatment groups according to the Zero Interaction Potency model generated by the Synergy Finder software from at least three independent experiments. Mean synergy scores between groups were compared by one‐way ANOVA followed by multiple comparisons. Error bars represent SEM. Asterisks represent significance levels of p‐values: *p < 0.05; **p < 0.01; ns, not significant. Figure S6: Tumour lysis syndrome and organ pathology in mice treated with the combination of high dose S63845 and venetoclax. Mice (n = 3) were treated with 25 mg/kg S63845, followed by a single administration of 25 mg/kg venetoclax 6 h later. Organs were harvested 1 h later. Representative haematoxylin and eosin sections illustrating organ damage associated with tumour lysis syndrome. (A) Kidney showing accumulation of cellular debris within glomerular capillaries (objective ×600 magnification) indicated by arrow. (B) Lung showing accumulation of cellular debris within a pulmonary artery branch (objective ×400 magnification), indicated by arrow. Images are representative for observations in three mice. Figure S7: No significant survival benefit is observed in response to S63845 monotherapy at higher doses. Engraftment level (left panel) and survival (right panel) of mixed lineage leukaemia (MLL)‐7 PDX‐engrafted mice treated with (A) 25 mg/kg S63845 intravenously twice weekly for 2 weeks, and (B) 50 mg/kg S63845 on 5 consecutive days for 1 week. Figure S8: Attenuated BIM protein expression after in vivo MCL1 inhibition. (A) Western blot for BIM (several isoforms) on lysates from splenocytes isolated from MLL‐7 or MLL‐14 engrafted mice 16 h after one administration of vehicle or 25 mg/kg S63845 (n = 3/group) (left panel). Lanes show individual mice. Quantification of protein bands by densitometry, normalised to β‐actin and to mean expression in the vehicle‐treated mice (right panel). t‐Tests were used to compare expression levels between groups. (B) Western blot on lysates from spleen samples from MLL‐14 engrafted mice (n = 3/group) after reaching end‐point following treatment with vehicle, S63845, venetoclax or the S63845 and venetoclax combination according to the schedule described in the legend of Figure 1. Western blots for BIM (total and upper band) with β‐actin loading control show trends towards lower BIM levels in S63845 and combination arms (observed in 2/3 mice) (left panel). Quantification of protein bands by densitometry, normalised to β‐actin and to mean expression in the vehicle‐treated mice (right panel). Mean expression levels between groups were compared by one‐way ANOVA. ns, not significant. Figure S9: The triple combination of dexamethasone (Dex), venetoclax (Ven) and S63845 is well tolerated. Mean weight of (A) MLL‐7‐ and (B) MLL‐14‐engrafted mice treated with Dex or the triple combination of Dex, Ven and S63845 (S63845 + Ven + Dex) included in the efficacy studies depicted in Figure 2. Error bars represent standard error of the mean. Figure S10: Increasing the dose of S63845 in the triple combination to 25 mg/kg does not increase the drug combination's efficacy in the MLL‐14 PDX. (A) Engraftment level (%human CD45+ cells in peripheral blood (PB), left panel) and survival (right panel) of MLL‐14 PDX‐engrafted mice treated with vehicle, S63845 at 6.25 mg/kg, 12.5 mg/kg or 25 mg/kg intravenously 1–2×/week for up to 3 weeks, or triple combination of venetoclax 25 mg/kg orally, 3–5×/week for up to 4 weeks; dexamethasone 10 mg/kg intraperitoneally, 3–5×/week for up to 3 weeks, and the respective dose of S63845, similar to the single agent schedule. The data presented in this figure for mice treated with drug schedules including 6.25 mg/kg S63845 are also included in Figure 2B. (B) Mean weight of the same cohort of mice. Error bars represent standard error of the mean. Figure S11: Treatment with double and triple combinations of S63845, venetoclax (Ven) and dexamethasone (Dex) does not cause long‐term depletion of haematopoietic stem cell subsets including white blood cells (WBC), red blood cells (RBC) and platelets. NSG mice (4 mice/group) were treated with vehicle, Ven + Dex, S63845 + Ven or the triple combination according to the following schedules: S63845, 6.25 mg/kg intravenously, 1–2×/week for 3 weeks; Ven, 25 mg/kg orally, 3–5×/week for 3 weeks; Dex, 10 mg/kg intraperitoneally, 3–5×/week for 3 weeks. Left panel depicts absolute changes. Right panel shows change in levels compared to day of treatment start. Grey‐shaded area corresponds to treatment period. Error bars represent SEM. Figure S12: The triple combination of S63845/venetoclax (Ven)/dexamethasone (Dex) induces better response rates than Ven/Dex in the MLL‐7 PDX. The percentage of MLL‐7 PDX‐engrafted mice achieving maintained complete response (MCR) in response to treatment with the triple combination of S63845 + Ven + Dex is significantly greater than that achieved by Ven + Dex. Fisher's exact test was performed to compare proportions. Figure S13: Baseline protein expression of pro‐survival BCL‐2 family proteins in infant KMT2A‐rearranged ALL PDXs. Representative immunoblot for two independent experiments. Untreated splenocyte lysates were immunoblotted for BCL‐2, MCL1 and BCL‐XL with whole‐protein Ponceau staining and β‐actin serving as loading controls (left panel). Quantification of protein bands by densitometry, normalised to β‐actin and to MLL‐2 expression levels (right panel). Error bars represent SEM. Figure S14: In vivo response of infant KMT2A‐r ALL PDXs following an induction‐type combination regimen of vincristine, dexamethasone and asparaginase (VXL). Survival curve of (A) MLL‐7 and (B) MLL‐14 PDX engrafted mice treated with vehicle or a combination of vincristine (0.5 mg/kg, once weekly), dexamethasone (5 mg/kg, 5×/week) and asparaginase (1000 KU/kg, 5×/week) intraperitoneally for 2 weeks. Gehan–Wilcoxon was used to compare survival between treatment groups. ***p < 0.001. Table S1: Characteristics of leukaemia cell lines used in this study. Table S2: Characteristics of leukaemia patient‐derived xenografts (PDX) used in this study. Table S3: Summary of in vivo responses of S63845, venetoclax and dexamethasone as single agents and in combination. Data Availability Statement Data are available upon request.