ASH 2022 | Investigating the combination of pivekimab sunirine, azacitidine & venetoclax in R/R AML

Pivekimab is a CD123 antibody-drug conjugate. This works by delivering a toxin called IGN, which causes single-stranded DNA breaks into the leukemic cells. And CD123 has been a target that we have evaluated in AML now for almost 20 years. It is expressed heavily on AML mature blasts, as well as on the leukemic stem cells, but it also is expressed on the endothelium and hematopoietic stem cells to a lower level. And the good thing about the pivekimab is that the toxin used here cause single-stranded DNA breaks, which is anticipated to have less toxicity to normal hematopoietic stem cell component. So we initially did a study a few years ago of single-agent pivekimab CD123 antibody-drug conjugate in relapsed/refractory AML. And we found that the response rate there was about 30, 35%, and it was higher in relapsed patients compared to refractory patients, but the duration of responses were short, only about three months.

So we did not think that this was sufficient for developing a single agent, especially in a competitive AML landscape where a number of drugs are being approved. We then decided to try the combinations and looked at aza-venetoclax with pivekimab, given that aza-ven had just been approved and was being used commonly, both in the frontline setting where it’s approved, but also in this salvage. And with this three-drug combination, that’s the data we reported the ASH, we actually looked at six different iterations of how we can combine it. The venetoclax, we looked at eight days per cycle, 14 days per cycle, 21 days per cycle. And the Pivekimab, we looked at two doses, 15 and 45 micrograms. These were the two lower doses that gave us responses in the single agent previous study. And we found that the 14 days of venetoclax, seven days of azacitidine with pivekimab 45 microgram was the optimal dose that was well tolerated, did not result in prohibitive myelosuppression, but also did result in a good activity.

So that’s the dose that we’re taking forward, after spending a lot of time, almost one and a half, two years on finding the optimal dosing and duration of venetoclax. We see that among the entire population, it was actually a reasonably sized study, 91 patients were enrolled, and the overall response rate was close to around 45% with a CRC rate of about 25, 30%, which is quite encouraging and better than what has been reported for HMA ven and salvage in the past. But among the recommended Phase II dose, we do see that there was a slightly higher true CR rate, and especially when you split the patients by those who were prior venetoclax-exposed versus those who were prior venetoclax-naïve, there we did see a clear signal among the prior venetoclax-naïve patients. The overall response rate was 55% with a CRC rate of close to 40, 45%, and a true CR rate of almost 25%. And this is very high numbers for a relapsed/refractory population.

So we think that in the prior venetoclax-naïve, there could be a path to developing this three-drug combination. And then we also saw, interestingly, in the three mutated patients, there was a signal for response with nine out of 11 FLT3-mutated patients responding and many of them having a true CR/CRh. So based on this, we have now moved the study to the frontline setting, which is going to be the eventual registration pad. And in the frontline three-drug combo, we are very early, but among the first 10 patients, we’ve had six responders, five true CRs, and one true PR without any signs of very severe myelosuppression. But we hope to enroll 50 to 70 patients in the next year. The study’s now open around 30, 35 sites across the US, and I think we will get a good idea looking at the CR rates, the MRD-negativity rates, duration of response, early mortality, whether this three-drug regimen has some potential to be taken forward in randomized studies in AML.