Matthew Davids, MD, MMSc:We’ve been talking a lot about venetoclax in CLL [chronic lymphocytic leukemia], but of course venetoclax is also approved now in AML [acute myeloid leukemia]. If you could, talk a little about the frontline approval for venetoclax and how that’s impacted therapy for AML patients.
Harry Erba, MD, PhD:Interestingly, venetoclax was tested in a small phase I with expansion study as a single agent in patients with relapsed-refractory AML, and we did not see in that study the robust activity that was seen in CLL as a single agent. This has been published by Marina Konopleva, MD, from The University of Texas MD Anderson Cancer Center. There was about a 20% response rate. Interestingly, though, in that very small experience, they had a handful of patients withIDH-mutated AML and saw a much higher response rate. So there was some level of activity. In the days of chemotherapy, that drug might have died at that point in AML because of a low response rate. But the investigators persisted and started combining the drug with HMAs [hypomethylating agents] and low-dose AraC [cytarabine].
And it brings up the question, why there? I think part of the issue is that 1 of our greatest unmet needs is how to improve on outcomes for patients who are not fit to get intensive chemotherapy. I’ll come back to that thought about combining a BCL2 inhibitor maybe with intensive chemotherapy. But at least the first studies were done with these less intensive chemotherapy regimens. It seemed like a good place to start because if you’re going to look for a signal combined with a backbone that has really very little activity, azacitidine, decitabine, low-dose AraC, have complete remission rates in older previously untreated AML of about 20%. If you add in CRIs [incomplete blood count recovery], you might get it up to 30% but with very low response rates. That was the rationale for combinations with hypomethylating agents and to low-dose AraC.
There were separate phase Ib studies done. One set of studies were with both azacitidine and decitabine at the standard dosages, the MD Anderson dosages of decitabine 20 mg/m2for 5 days and azacitidine 75 mg/m2for 7 days. And the venetoclax was combined together with it and then in sequence with it.
The way the ramp-up was done is on day 1 it was 100 mgthis is with HMAs—200 mg on day 2; and 400 mg on day 3, a much quicker ramp-up than what’s done in CLL. As long as we’re talking about the other study as well, low-dose cytarabine, the ramp-up there went up to 600 mg with a similar very rapid ramp-up of the drug. So let’s focus on the HMA combination with venetoclax. What did they see?
First of all, let’s talk about the population because I think that’s really important when we’re thinking about where to apply this regimen and what safety issues might come up. The population was older patients who are not fit for chemotherapy, so they had to be over 75 years old or have had some other comorbidity that precluded intensive chemotherapy.
It’s interesting that in the study, the white blood cell count couldn’t be over 25,000 at the start of the treatment because of the risk of tumor lysis syndrome. But only 10% of people in the study actually required hydroxyurea to get the white blood cell count that low. In fact, when you look at the baseline marrows, about 25% to 30% of these patients had what we used to call MDS [myelodysplastic syndrome]: 20% to 30% blast.
I think 1 of the first cautionary notes is we’re not talking about the proliferative AMLs, the white blood cell counts of 100,000 that are difficult to control when you’re looking at the efficacy and safety of this combination. In the HMA-venetoclax study, they also excluded patients with prior exposure to HMAs. They were included in the low-dose AraC studies, and I’ll come to that.
This combination has caught our attention very quickly for 2 major reasons. No. 1, the response rates are much higher than we’ve seen with single-agent venetoclax or a single-agent HMA. If you look at CR [complete response] and CRh [partial blood count recovery], the FDA has defined CRh as a platelet count over 50,000 mm3and ANC [absolute neutrophil count] over 500, as opposed to the usual 100,000 and 1000, but still a very good count recovery. CR plus CRh is 61% with decitabine or azacitidine and venetoclax.
If you look instead at CR and CRI, it’s 70%. If you look across subsets, you see very high response rates above 50% in all subsets, including TP53, and in theIDH-mutated subset it was 90% with azacitidine and 100% with decitabine. Very small numbers but very high response rates. Echoing back to the single-agent activity of venetoclax in the IDH-mutated study that Marina Konopleva published. So it has very high response rates.
The second thing that has caught our attention is how rapid these responses occur. Marrows were done at cycle 1, and the median time to response was 1 or 2 months. If there was cytoreduction by cycle 1 it was recommended that after cycle 2 another marrow be done. Most of the responses, not all but most, are seen within the first couple of cycles. That clearly catches our attention, and it really suggests that these aren’t additive, that there’s some synergistic activity of using these 2 drugs in the schedules of a week of chemotherapy, decitabine-azacitidine, with 28 days of venetoclax.
Now a phase III study has been completed. The accrual completed late in 2018. In that study was azacitidine with venetoclax versus azacitidine with placebo. Why azacitidine? There are decitabine lovers out there, so what’s wrong with decitabine? Nothing is wrong with decitabine. When you look at the phase Ib data, there’s to my eye a suggestion of a little bit more febrile neutropenia, a little more myelosuppression with decitabine. Many of us say that when we use decitabine as opposed to azacitidine. But they had to pick 1 drug, and I think that’s why they picked azacitidine.
Transcript edited for clarity.