Venetoclax for Hypodiploid ALL: Novel Therapy for Bad Biology

Acute lymphoblastic leukemia (ALL) is a heterogenous disease that can be subdivided into favorable and less favorable genetic subtypes based on sentinel genetic lesions. The overall cure rates for children with B-cell ALL (B-ALL) exceeds 85 percent in many countries; however, survival remains relatively poor for certain less favorable genetic subtypes, including Philadelphia chromosome–like (Ph-like), t(17;19), and hypodiploid ALL.¹ 

For some patients with unfavorable genetic lesions, intensification of therapy can improve outcomes to levels nearing those with favorable genetics. For example, internal amplification of chromosome 21 (iAMP21), which occurs in approximately 2 percent of childhood B-ALL, was initially shown to portend worse outcome.²  The Medical Research Council cooperative group investigated treating iAMP21 B-ALL patients as very high risk regardless of other clinical and response risk factors, improving five-year event-free survival (EFS) from 29 percent to 78 percent.³  Other cooperative groups observed similar results, showing that intensive chemotherapy mitigates poor prognosis of iAMP21.¹  While this approach has improved outcomes for some unfavorable ALL genetic subtypes, it has not worked for others.

Loss or gain of chromosomes, or aneuploidy, is common in B-ALL blasts. Up to 30 percent of children with B-ALL have high hyperdiploidy (51-67 chromosomes), which carries a favorable prognosis.¹  In contrast, 2 to 3 percent of children with B-ALL have hypodiploid ALL (<44 chromosomes), which has inferior outcomes. Two recent manuscripts from the Children’s Oncology Group and Ponte di Legno Group summarized outcomes from 131 and 306 patients, respectively.^4,5  They found outcome was poor in patients with both favorable and unfavorable disease response (minimal residual disease negative and positive, respectively, at end of induction) and remained poor regardless of treatment intensity, as outcomes were inferior even with hematopoietic stem cell transplantation in first remission (CR1). These data reinforce the notion that therapy intensification does not overcome disease biology in some B-ALL subtypes, and new agents or approaches are needed. Improved understanding of disease biology may lead to rational targeted approaches.

In 2013, Dr. Linda Homfeldt and colleagues performed comprehensive genomic profiling of 124 cases of hypodiploid ALL, finding it could be subdivided based on disease biology.⁶  Near-haploid B-ALL (24-31 chromosomes) frequently has alterations affecting receptor tyrosine kinase and Ras signaling (71%), as well as alterations in IKZF3 (13%). Low-hypodiploid (32-39 chromosomes) B-ALL often has alterations in TP53 (91.2%), IKZF2 (53%), and RB1 (41%). These data provide a compelling rationale for potential targeted approaches for hypodiploid ALL, as some hypodiploid cell lines with dysregulated PI3K or Ras signaling were sensitive to PI3K inhibitors.

Dr. Ernesto Diaz-Flores and colleagues recently performed biochemical characterization of hypodiploid B-ALL to investigate which pathways were dysregulated and study how aneuploidy and specific genetic alterations may impact cell signaling. The authors performed phosphoproteomics in a panel of hypodiploid cell lines and primary patient samples, measuring more than 20 different proteins from numerous signaling pathways that are frequently dysregulated in ALL. They confirmed that hyperactive PI3K and RAS signaling were common in hypodiploid ALL. They also made the novel observation that high levels of BCL-2 were frequently observed in hypodiploid ALL. Using publicly available datasets, they confirmed that increased expression of BCL-2 is often present in hypodiploid ALL. This increased expression was found to occur at the mRNA and protein level without identifiable mutations or genetic amplifications. Moreover, they found that increased expression of BCL-2 usually occurred with high levels of other proapoptotic proteins such as BID or BAD, and with low levels of BCL-XL.

Based on these results, they tested the activity of the BCL-2 inhibitor venetoclax (ABT-199) in multiple preclinical models, including xenografted human B-ALL in NOD-scid IL2Rgamma^null mice. Venetoclax was highly effective in hypodiploid ALL, demonstrating marked antileukemic responses in xenografts generated from eight different patients with hypodiploid B-ALL. They also found that a signature of either high BCL-2/BIM or high BCL-2/BAD with low levels of MCL-1 and BCL-XL predicted sensitivity to venetoclax.