The phase III MATTERHORN trial comparing roxadustat with placebo in patients with transfusion-dependent, lower-risk myelodysplastic syndromes (MDS) failed to meet its primary endpoint with no significant difference between study arms for eight-week transfusion independence. However, post-hoc analysis focusing on the subgroup of patients with MDS with higher transfusion burden demonstrated a significant superiority compared with placebo.
“The company decided that the study failed, but I see it in a different way,” said principal investigator Moshe Mittelman, MD, of Tel Aviv Sourasky Medical Center at Tel Aviv University in Israel.
MATTERHORN enrolled 140 patients with very low-, low-, or intermediate-risk MDS with or without prior erythropoiesis-stimulating agent (ESA) treatment and randomly assigned them 3:2 to oral roxadustat or placebo.
At an interim analysis, about one-third (33.3%) of placebo-treated patients were defined as eight-week transfusion independent compared with almost half (47.5%) of patients assigned to receive roxadustat. The difference was not statistically significant. Based on these data, the study sponsor terminated the study. At the end of treatment, 55.0% of patients in the roxadustat arm and 40.4% of patients in the placebo arm were transfusion-independent responders.
“If we limit our analysis to patients who had high transfusion burden, then the difference is definitely in favor of patients treated with roxadustat,” Dr. Mittelman said. “Moreover, if you limit the analysis to the patients considered to be responders and who had a 1.5 g/dL increase in hemoglobin — such an increase is considered clinically meaningful — this was also significantly better in roxadustat-treated patients.”
Among patients with a baseline transfusion burden of two or more red blood cell (RBC) units every four weeks, 36.1% of patients assigned roxadustat compared with 11.5% of patients assigned placebo achieved eight-week transfusion independence (p=0.047). Additionally, 45.5% of responders in the roxadustat arm had mean hemoglobin concentration increase of at least 1.5 g/dL compared with 17.4% of responders assigned placebo (p=0.044).
According to Dr. Mittelman, the study was hampered by its inclusion criteria, which were too liberal in allowing the inclusion of patients with very low transfusion needs. He illustrated his point by discussing a realistic scenario from the day-to-day treatment of MDS.
“If I see a patient with hemoglobin of six, I will almost always transfuse them; if I see a patient with hemoglobin of 9.5 to 10, I will almost always not transfuse them,” Dr. Mittelman said. “But if I see a patient with a hemoglobin of 8 or 8.5, I will ask them how they feel that day.”
Dr. Mittelman then initiates a doctor/patient decision-making process to decide whether to transfuse.
“If we decide to transfuse, the moment the patient gets one unit, they are transfusion dependent,” Dr. Mittelman said. Indeed, MATTERHORN enrolled patients with a transfusion burden of one to four packed RBC units every eight weeks. “However, when I see that patient two to three months later, we will have that discussion again, and if we decide not to give another transfusion, they are now defined as transfusion independent.”
A patient like this, if placed in the study’s placebo arm, would look like a responder.
“Clinicians know, though, that this response is not the result of the drug or of placebo, but it is the nature and course of the disease,” Dr. Mittelman said.
Consequently, the study was limited by the use of an eight-week screening period to assess baseline transfusion burden instead of the International Working Group (IWG) 2018 recommended 16-week screening period, Dr. Mittelman said. Use of the 16-week period could have resulted “in a recategorization of baseline transfusion burden.”
Dr. Mittelman said he has tried to persuade the roxadustat manufacturer to conduct another trial with use of the 16-week screening period for baseline transfusion burden and to include only patients with a higher transfusion burden.
“I believe this drug works; it may not work for everybody, but it works for some, and the side effects are minimal,” Dr. Mittelman said. “It is also very convenient as an oral pill taken three times a week.”
Two other newly approved drugs in MDS — luspatercept and imetelstat — are given by injection and intravenous infusion, respectively. This is in addition to traditional injected ESAs. Additionally, all three of these drugs have different mechanisms of action. Luspatercept is a recombinant fusion protein that binds TGF-beta ligands to reduce SMAD2 and SMAD3 signaling. Imetelstat is an oligonucleotide telomerase inhibitor. Roxadustat is a hypoxia-inducible factor (HIF)-prolyl hydroxylase (PH) inhibitor.
“Each of these three agents are working on a different place in the production of the RBC pathway,” Dr. Mittelman said. “Where roxadustat will fit into the treatment of MDS in the future … time will tell.”
Any conflicts of interest declared by the authors can be found in the original article.
Reference
Mittelman M, Henry DH, Glaspy JA, et al. Roxadustat versus placebo for patients with lower-risk myelodysplastic syndrome: MATTERHORN phase 3, double-blind, randomized controlled trial. Am J Hematol. 2024;99(9):1778-1789.
Perspective
The introduction of roxadustat into the treatment of patients with MDS was met with great fanfare. After all, the biologic basis of its activity as an HIF-PH inhibitor earned investigators a Nobel Prize in Physiology or Medicine in 2019.
The MATTERHORN trial randomized patients with lower-risk MDS associated with transfusion dependence on RBCs to either receive the prized drug or placebo, with a primary endpoint of achieving transfusion independence lasting eight weeks or longer. Anemia is the most common cytopenia associated with MDS and directly affects patients’ quality of life (QoL). Thus, it is a worthy target of pharmacologic intervention. When a patient requires RBC transfusions regularly, we also start therapy. The patient population meeting the eligibility criteria for this study was entirely appropriate.
The good news is that the drug worked, with almost half the patients achieving transfusion independence lasting eight weeks or greater. The even better news is that the placebo worked, with one-third of the patients enrolled in this trial achieving transfusion independence lasting eight weeks or greater! If I had my druthers, I’d love to get my hands on the placebo that could improve the lives of a large minority of my patients, without any concern of nasty side effects impeding their QoL. As a consequence, there was no significant difference in response rates between roxadustat and placebo.
Unfortunately for my patients and for me, the fault lies not within the star quality of the placebo but within the self-limiting meagerness of the response criteria. Patients with lower-risk MDS may require a unit of blood once every eight weeks or twice every 16 weeks, but they can also easily go an eight-week period without requiring a unit of blood, which leads to the false impression that the patients responded to an intervention. The authors of the study present their data honestly, showing that for patients who required one unit of blood every eight weeks, the “response rate” to roxadustat was guaranteed.
In truth, for these patients, transfusion dependence may have been conflated with a simple transfusion accident. To identify therapies that are truly clinically meaningful, future studies, as the authors point out, should apply the revised IWG 2018 criteria to define a longer 16-week screening period for transfusion dependence and extend the duration of transfusion independence to a minimum 16-week period.
Mikkael A. Sekeres, MD, MS
Chief, Division of Hematology
Professor of Medicine
Sylvester Cancer Center
University of Miami
COI Disclosure: Dr. Sekeres has served on advisory boards for BMS, Kurome, and Schroedinger.
