Harvard Expert Sees Potential in BCL-2 Combos

Anthony Letai, MD, PhD

The prospect of developing anticancer strategies that target the apoptotic pathway most likely lies in combinations involving members of the B-cell lymphoma-2 (BCL-2) protein family, according to Anthony G. Letai, MD, PhD, a leading investigator in the field.

In his laboratory at Dana-Farber Cancer Institute in Boston, Massachusetts, where he is an associate professor of medicine, Letai focuses on investigating apoptotic dysfunction’s role in tumor maintenance. In particular, he is interested in understanding the interactions among the BCL-2 protein family members, which include myeloid cell leukemia-1 (MCL-1). Letai, who also is an associate professor at Harvard Medical School, has led efforts to translate BH3 mimetics that target BCL-2 family members to the clinic.

OncLive: How does apoptosis fit into the hallmarks of cancer?

In an interview with OncologyLive^®, Letai discussed the complex nature of apoptosis and efforts to target the process.Letai: Both the older and newer constructs of Hanahan and Weinberg’s “Hallmarks of Cancer” list “resisting cell death” as a hallmark. The concept is that a lot of the bad things cancers do, such as overexpress oncogenes, invade and metastasize out of their native locations, and proliferate relentlessly, should provoke apoptotic signaling (apoptosis is a prominent form of programmed cell death) that should kill the cancer cells. The fact that some cancer cells escape this level of control suggests that they have selected for evasion of programmed cell death. Indeed, many mouse genetic models of impaired apoptosis display accelerated oncogenesis, supporting programmed cell death as 1 level of control of cancer.

However, some have taken this to mean that cancer cells in established tumors are more resistant to programmed cell death than normal cells are. This is probably incorrect. In fact, most cancers are more sensitive to cell death signaling than most normal tissues are. This is the main reason conventional chemotherapy, targeting the ubiquitous elements of DNA and microtubules, ever demonstrates a therapeutic index.

What is the role of MCL-1 in that hallmark ability?

Cancer cells may select for blocks in apoptosis that enable it to escape endogenous death signaling induced by oncogenesis, but there is no mechanism for them to foresee future exposure to chemotherapy and select for the extra apoptotic blocking that resistance to these agents would require. This often results in cancer cells that survive, but just barely. Of course, this principle exists on a broad spectrum, with leukemias and other blood cancers being the most primed for apoptosis, in concert with their broad chemosensitivity. Many solid tumors, however, exhibit more profound blocking, which yields less chemosensitivity.MCL-1 is one of the proteins of the BCL-2 family that regulate apoptosis. It is a so-called antiapoptotic protein, as it opposes commitment to apoptotic cell death by binding and sequestering proapoptotic proteins. Theoretically—and demonstrated in mouse models—high levels of MCL-1 expression can facilitate tumorigenesis. In human tumors, amplification of the MCL-1 locus is one of the more common somatic genetic abnormalities.

How is MCL-1 being targeted for anticancer therapy?

However, what is more important from a therapeutic perspective is whether or not the tumor cell is dependent on MCL-1 function to stay alive, and this is a phenotype that is not easily identified by somatic genetic alterations. MCL-1— dependent tumor cells are good candidates for targeting with MCL-1–inhibiting drugs.The most direct way is via small molecules that compete for the pocket in MCL-1 that is required to bind the BH3 domain of proapoptotic proteins. These so-called BH3 mimetic drugs inhibit the homodimerization that is necessary for MCL-1 function. BH3 mimetics can displace proapoptotic proteins that are already bound by MCL-1, allowing them to progress with commitment to programmed cell death. Right now, the companies that are furthest along with BH3 mimetic small-molecule antagonists of MCL-1 include Novartis (in a partnership with Servier), AstraZeneca, and Amgen.

How might MCL-1 inhibitors be used in hematologic malignancies? Do you think they might be effective in solid tumors?

An alternative way to target MCL-1 is via CDK9 (cyclin-dependent kinase 9) inhibition. CDK9 regulates transcriptional elongation, and its inhibition selectively depletes cells of proteins with a short half-life. MCL-1 is such a protein and is depleted by CDK9 inhibition. Of course, this is a “dirtier” way to inhibit MCL-1 function—only clinical experience will reveal if it is more or less effective, or more or less toxic, than BH3 mimetic MCL-1 inhibition.BCL-2 is an antiapoptotic cousin of MCL-1. Based on work my lab and others have done and based on clinical experience, it looks like chronic lymphocytic leukemia is pretty homogeneously dependent on BCL-2 and sensitive to BCL-2 inhibition. B-cell acute lymphoblastic leukemia and blastic plasmacytoid dendritic cell neoplasm (a rare blood cancer) are likely quite similar.

So far, we have not identified a blood cancer that is quite so homogeneously dependent on MCL-1. What we see more often are significant subsets of disease, like acute myelogenous leukemia or myeloma that is MCL-1 dependent. I think this is where we will first see their single-agent activity, and clinical trials are ongoing. I think that clinical response rate will improve when treatment is driven by predictive biomarkers of MCL-1 dependence. Also, as we have seen for BCL-2 inhibition, I expect that response rates and durability of response will significantly increase as MCL-1 inhibitors are brought into combinations.

I think that when this is done, MCL-1 inhibitors, as I think for BCL-2 inhibitors, will find very broad application across hematologic malignancies. I am very excited about the possibility of combining BCL-2 and MCL-1 inhibitors. There may be toxicity issues to sort out, but I believe that whoever finds a way to combine these 2 will have an extremely powerful combination therapy with very broad applicability.

For solid tumors, it is simply harder to get them to undergo apoptosis, as they are generally less primed for apoptosis. Therefore, I suspect there will be limited single-agent activity of MCL-1 inhibitors. However, as in blood cancers, I think that the use of good predictive biomarkers, as well as incorporation into combinations, will facilitate penetration of MCL-1 inhibitors (and BCL-2 inhibitors) into solid tumors. I expect that progress in solid tumors will lag behind that in blood cancers, as it will be much easier to accumulate the necessary proof-of-principle data in the latter, which are more primed for apoptosis.