ASH 2022 | Real-world data on the efficacy of luspatercept in LR-MDS

So at this ASH there were two or three groups that presented real-world data on use of luspatercept after the approval. I think there’s been growing interest in real-world data confirming activities we’ve seen on trials that lead to approval of the drug, so we reported our experience on the use of luspatercept after approval. We treated almost 114 patients at the Moffitt Cancer Center since the approval. Again our patients, majority two-thirds, had MDS ring sideroblasts subtype, where the FDA approval label is. Majority of those patients also harbored SF3B1 mutation, but there were also patients with non-ring sideroblasts and non-SF3B1 treated. Most of our patients, similar to the original study, had a prior ESA treatment. What was unique in our cohort is half of the patients had prior hypomethylating agents used or lenalidomide.

The MEDALIST study did not allow prior exposure to HMA and len, and that’s been a question in the field, do patients respond to this treatment after len or hypomethylating agent failure? We compared our cohort to the original MEDALIST study cohort, again, several similarities pointed out that our cohort included patients that did not have ring sideroblasts, our cohort included patients that had a prior treatment with lenalidomide and hypomethylating agents. The overall response rate was very close to what’s reported in the MEDALIST study, around 40% of our patients responded. We also saw that the major predictor of response was the transfusion burden at baseline. In our study, there were patients that were not transfusion-dependent, while in MEDALIST, all patients had to be transfusion-dependent, but ours at least had two units of blood. But again, we saw the same signal that transfusion burden was predictive of response where we see higher responses in low or non-transfusion-dependent patients.

We also saw that the responses in a way were dose dependent. In terms of non-ring sideroblast patients or SF3B1, we do see a trend or a response higher and among patients with SF3B1 mutation and among patients with ring sideroblasts subtype, 14 patients in our group had this unique MDS, MPN, RST subtype and the responses were very high, approaching 80%. Again, the patients’ serum EPO level predicted responses, but most of those patients that had high serum EPO levels also had a high transfusion burden.

When we looked at the responses where the patients had prior HMA response or not, the responses were less in patients that had HMA failure, 50% versus 30%, however again, the group of patients that had hypomethylating agents failure tended to be a highly transfusion-dependent group, so it could be a reflective of the transfusion burden than the treatment exposure. Similar findings were seen with lenalidomide. So again, I think our data confirms luspatercept’s clinical benefit and safety profile. Responses are dose dependent, low base blood transfusion dependency and SF3B1 mutation correlated with higher response rate, and luspatercept did retain activity after lenalidomide or hypomethylating agents failure. However, a trend of lower responses observed that probably correlates to the higher transfusion burden among those patients.