Patients with lower risk MDS really represent the lion’s share of patients that we see. For them the most common symptomatic cytopenia they deal with is anaemia. For patients with lower risk MDS with anaemia over time they do become transfusion dependent and the accumulation of iron loading that can occur with them also has with that attendant complications, as you might expect. So for patients with lower risk MDS the standard of treatment to start with is erythroid stimulating agents or recombinant erythropoietin. It has effectiveness but in patients that are red cell transfusion dependent it’s a very lower response rate and a very short duration of response so there really is a need for those patients that are transfusion dependent specifically for some new therapies that are very effective. One of those is luspatercept that we’ve been studying. Luspatercept is an erythroid maturation agent, it is essentially a soluble receptor for activin 2A, then it’s a chimera hooked onto an IgG1 heavy chain for circulation. But it acts as a ligand trap, primarily for the TGF-β superfamily ligands. As a consequence for that, by suppressing those ligands it will inhibit Smad 2/3 signalling thereby allowing erythroid maturation.
So this began in a phase I/II study in Europe and, in fact, in that study there was a higher response rate, transfusion independence rate, in patients that had MDS with ring sideroblasts. That was the rationale to take this forward into a phase III study.
In this study, it was called the MEDALIST trial, it was a randomised double-blind, placebo-controlled trial involving 229 patients. Patients that were eligible had lower risk MDS defined by the IPSS revised criteria, a very low, low and intermediate risk disease. Morphologically they had to have MDS with ring sideroblasts and that could be defined by 15% or greater ring sideroblasts in the bone marrow or 5% or more with an SF3B1 gene mutation. They also had to be transfusion dependent, defined by receiving at least two units or more every eight weeks in the sixteen weeks prior to randomisation. Patients that were excluded from this were patients that had previously received immunomodulatory drugs like lenalidomide or azanucleosides. Those patients, as I mentioned, were randomised in a two to one fashion to receive either luspatercept, beginning at a dose of 1mg/kg as a subcutaneous injection every three weeks, or placebo. The dose could be escalated all the way up to 1.75mg/kg. The overall response assessment occurred after week 24 and patients were reassessed every six months thereafter.
The primary endpoint for the study was red blood cell transfusion independence, defined by a period of eight weeks or longer without receiving transfusions. There were several other key secondary endpoints – responses that occurred, transfusion dependence that occurred greater than twelve weeks, and also, importantly, those that had a major reduction in transfusion, what we call erythroid hematologic improvement. That was defined by at least a four unit or greater reduction in red cell transfusion needs over eight weeks in people that were heavily transfused, and in people with a lower transfusion burden they had to have at least a 1.5g or greater rise in haemoglobin over eight weeks.
What did you find?
The primary endpoint of transfusion independence was achieved in 37.9% of patients on the luspatercept arm and in 11% of patients on the placebo arm. It was very durable; when you look at the duration of transfusion dependence there were 40% of patients that were still transfusion free after one year of therapy. The median rise in haemoglobin in those patients was 2.55g, reaching as high as a maximum of 4.1g/dL in those responders. The other key secondary endpoint, remember these are generally heavily transfused patients, and a hematologic improvement, as I defined earlier, was achieved in 58% of patients on the luspatercept arm. So the vast majority of patients had clinical benefit, whether it’s from transfusion reduction or transfusion dependence and rise in haemoglobin.
What about adverse effects?
This was an extremely well tolerated drug. There was absolutely no difference in the treatment emergent adverse events in placebo compared to luspatercept and no difference in the frequency of evolution to AML as well.
What could be the implications for clinical practice?
Clearly it looks like it’s a very effective agent, particularly in these patients that are MDS with ring sideroblasts with lower risk that are transfusion dependent. It gives us another tool in our armamentarium for patients that, right now, have limited treatment options.
