Pediatric Patients with Relapsed/Refractory Acute Lymphoblastic Leukemia Harboring Heterogeneous Genomic Profiles Respond to Venetoclax in Combination with Chemotherapy

Introduction: Venetoclax (VEN) is an orally administered selective inhibitor of the B-cell lymphoma 2 (BCL2) antiapoptotic protein. VEN combined with chemotherapy (CTx) has demonstrated encouraging responses in a variety of hematologic malignancies. Here, we present safety, efficacy, and preliminary genomics results from our phase 1 study of VEN administered with CTx in pediatric patients (pts) with relapsed/refractory (R/R) acute lymphoblastic leukemia (ALL).

Methods: This open-label, phase 1, 2-part study enrolled pts <25 years old with different tumors including ALL (NCT03236857). In part 1, 5 pts with ALL received a weight- or age-adjusted adult-equivalent dose (AED) of 800 mg VEN PO daily after a 3-day ramp-up to mitigate tumor lysis syndrome risk; CTx could be added after 21 days of monotherapy (monoTx). In part 2, 20 pts with ALL received 800 mg AED VEN daily with CTx on day 4 following the ramp-up. Primary and secondary endpoints of the study included safety of VEN monoTx and preliminary efficacy of VEN ± CTx, respectively. Exploratory objectives included analysis of predictive biomarkers and minimal residual disease (MRD). Whole exome sequencing analysis was performed on baseline blood or bone marrow samples; gene fusion data were acquired through site-reported cytogenetic entries at screening. Highly recurrent genetic alterations in pediatric ALL were analyzed across a range of functional classifications. RNA sequencing analysis was performed on pretreated blood or bone marrow samples to evaluate gene expression of BCL2 family members.

Results: As of June 2020, 25 pts with ALL were enrolled and received VEN + CTx; CTx regimens included dexamethasone and/or vincristine and/or peg-asparaginase (D/V/P, n=16) or cytarabine and/or etoposide and/or peg-asparaginase (C, n=9). Pts had a median of 3.5 (range 1.0–9.0) prior therapies; the median duration of VEN therapy was 2.1 mo (range 0.4–4.6). Most common reasons for VEN discontinuation were progressive disease (28%; D/V/P: n=1; C: n=6) or continuation to transplant (28%; D/V/P: n=6; C: n=1). Eight (32%; D/V/P: n=4; C: n=4) pts had fatal adverse events (AEs), all unrelated to VEN. Most common grade 3/4 AEs were febrile neutropenia (52%) and anemia (44%); 1 pt (D/V/P)completed ramp up, but reported laboratory tumor lysis syndrome with CTx combination.

A best objective response rate (ORR) of 56% was seen with VEN + D/V/P: 4 complete responses (CR), 4 CR with incomplete marrow recovery (CRi), and 1 CR without platelet recovery (CRp). ORR of 11% (CRi: n=1) was seen with VEN + C (Table). In pts who achieved CR/CRi/CRp, MRD <10⁻⁴ leukemic cells were observed in 28% of all pts (VEN + D/V/P: n=6; VEN + C: n=1).

Genetic analysis of evaluable pts (VEN + D/V/P: n=12; VEN + C: n=6) demonstrated a highly heterogeneous genomic landscape (Figure A) including a range of somatic mutations and structural variants, comprising translocations, rearrangements, and chromosomal abnormalities. Analysis of mutations found in ≥2 pts revealed that responses were seen in VEN + D/V/P-treated pts with CREBBP, KMT2A, KMT2D, RB1, PTPN11, and PDGFRB mutations. Of the pts who had CREBBP mutations, 2/3 obtained CR. Pts with mutations in RB1, PTPN11, and PDGFRB (n=2 each) also achieved CR or CRi. Of the pts with KMT2A or KMT2D mutations detected by our sequencing platform, 1/2 and 2/4 achieved CR, respectively. Furthermore, of the 6 pts with KMT2A translocations identified locally, 3 obtained CR with VEN + D/V/P. Baseline gene expression profiling revealed that BCL-xL and MCL1 levels were slightly higher in comparison to BCL-2 (Figure B). There was no significant correlation between these gene expression levels and response. No mutations were detected in BCL2, BCL-xL, or MCL1.

Conclusions: VEN + CTx was well tolerated, with no unexpected toxicities in pediatric pts with R/R ALL. Promising preliminary efficacy was observed in these refractory pts receiving VEN + D/V/P (ORR 56%). Our genomic profiling suggests that responses with VEN + CTx occur in pediatric ALL with a variety of mutations, including those with KMT2A rearrangements. However, due to the limited sample size and the overall heterogeneity, further investigation in a larger pt cohort is warranted to determine which mutations confer resistance or sensitivity to VEN.