Real-World Analysis Shows High, Durable Transfusion Independence Rates With Luspatercept in Lower-Risk MDS

Treatment with the first-in-class erythroid maturation agent luspatercept-aamt (Reblozyl) resulted in high rates of transfusion independence among patients with lower-risk myelodysplastic syndrome (MDS), according to data from a real-world analysis of this heavily pretreated patient cohort.¹

Among response evaluable patients (n = 31), 52% achieved erythroid responses and 48% of those who were transfusion-dependent achieved at least 16 weeks of transfusion independence. The median duration of transfusion independence was 48 weeks; this was comparable between high- vs low- transfusion burden responders. Although 22% of patients (n = 37) lost transfusion independence, it was restored following dose increases in 50% of patients initially receiving less than 1.75 mg/kg of luspatercept.

“Luspatercept achieved high rates of durable transfusion independence in heavily pretreated patients with lower-risk MDS patients; dose escalation restored transfusion independence in 50% [of those] who lost response at submaximal dosages,” Marisa L. Kometas, MD, lead study author and second-year internal medicine resident at UT Southwestern Medical Center in Dallas, Texas, stated in a presentation of the data. “Response was associated with a serum erythropoietin [EPO] level [of] less than 100 mU/mL, but not with baseline transfusion burden.”

Anemia, which is the most common cytopenia in patients with lower-risk MDS, can lead patients to develop transfusion dependence and experience inferior responses with standard-of-care treatment options. To combat this, luspatercept was approved by the FDA for the treatment of anemia in adult patients with very low– to intermediate-risk MDS who were not previously treated with an erythropoiesis-stimulating agent (ESA) in August of 2023.² This regulatory decision was supported by interim findings from the phase 3 COMMANDS trial (NCT03682536), in which luspatercept demonstrated superior efficacy vs epoetin alfa.

Investigators further evaluated real-world responses with luspatercept in patients with lower-risk MDS according to baseline characteristics and prior treatments. These results were shared at the 2024 SOHO Annual Meeting.¹

Study Methodology and Patient Characteristics

The retrospective, single-center, real-world analysis was initially launched at an academic tertiary referral center. Patients with lower-risk MDS who underwent treatment with luspatercept from January 2020 through December 2023 were eligible for study inclusion.

Baseline characteristic data on prior therapies, serum erythropoietin levels, ring sideroblast status, karyotype, and gene mutations were collected. Baseline transfusion burden—which was categorized as no transfusion dependence—low transfusion burden and high transfusion burden were calculated for all patients. Hematologic improvement-erythroid was assessed in those receiving 16 weeks of therapy or more. Mann-Whitney U tests were utilized for continuous variable comparisons and Fisher's Exact Tests were applied to categorical variables. Leukemia-free overall survival (OS) was determined using the Kaplan-Meier method, and group comparisons were performed using log-rank testing.

In the total dataset (n = 37), the median age at the start of luspatercept was 74 years (range, 58-93). When divided into responder (n = 18) and non-responder (n = 13) groups, the median age was 77 years (range, 58-86) vs 82 years (range, 65-93), respectively. Approximately half of all patients (51%) were female and 16% had therapy-related MDS. Patients had previously received an ESA (total dataset, 92%; responders, 94%; non-responders, 100%), a hypomethylating agent (19%; 11%; 23%), lenalidomide (14%; 6%; 23%), or 2 or more prior therapies (22%; 11%; 31%), respectively. Regarding EPO levels, patients had less than 100 mU/mL EPO (32%; 50%; 8%), between 100 and 200 mU/mL (19%; 17%; 31%), between 200 and 500 mU/mL (14%; 6%; 15%), greater than 500 mU/mL (30%; 17%; 46%), or no available data (5%; 11%; 2%).

The median baseline transfusion burden was 3 units (range, 0-11) per 16 weeks prior to treatment in the overall population, 5 units (range, 1-11) in the responder group, and 3 units (range, 1-9) in the non-responder group. Patients were also grouped according to high transfusion burden (41%; 44%; 39%), low transfusion burden (41%; 39%; 54%), or were not transfusion dependent (19%; 17%; 8%).

Morphologic, Cytogenetic, and Prognostic Characteristics

In the overall population, ringed sideroblasts were present in 78% of patients, absent in 19%, and the aspirate quality was too poor for morphology in 13%. Respective values were 94%, 6%, and 0% in the responder group vs 69%, 23%, and 8% in the non-responder group. In total, 73%, 78%, and 77% of patients in the total, responder, and non-responder groups, respectively, harbored a SF3B1 mutation. Notably, 32%, 17%, and 39% of patients across these respective groups displayed a high-risk mutation, such as TP53, ASXL1, CBL, RUNX1, ETV6, or SRSF2). No next generation sequencing (NGS) data were available in 8%, 17%, and 0% of patients from the overall, responder, and non-responder groups, respectively.

Patients enrolled onto the trial showed signs of any cytogenetic abnormality (35%; 22%; 54%), del(5q) (8%; 0%; 23%), and cytogenetic abnormalities other than del(5q) (30%; 22%; 39%). Most patients had low risk scores (78%; 89%; 62%) according to the Revised International Prognostic Scoring System (IPSS-R), followed by intermediate IPSS-R scores of 3.5 or less (11%; 6%; 23%) and IPSS-R scores of 4.0 or greater (11%; 6%; 15%). Patients were also classified as having very low, low, or moderate-low risk (73%; 72%; 69%) according to the Molecular International Prognostic Scoring System (IPSS-M) and moderate-high, high, or very high IPSS-M scores (19%; 11%; 31%). This could not be calculated in 8%, 17%, and 0% of patients in the overall, responder, and non-responder groups, respectively, due to the absence of NGS data.

Additional Efficacy and Safety Data

The median leukemia-free OS was not reached with luspatercept and did not differ between responders vs non-responders. Disease progression to higher-risk MDS or acute myeloid leukemia was observed in 11% of patients. Notably, serum EPO levels less than 100 mU/mL were found to significantly predict responses with luspatercept (P = .02).

Furthermore, patients achieved transfusion independence at equal rates for doses of 1.00, 1.33, and 1.75 mg/kg of luspatercept, respectfully.

Investigators noted that common adverse effects included shortness of breath, falls, fatigue, and hypertension. These precipitated treatment discontinuation in 8% of patients on the study.

“As luspatercept gains prominence as a first-line therapy, further investigation is [needed] to define predictors of response,” Kometas concluded.

References

1. Kometas ML, Hyak J, Hoff F, et al. Real-world efficacy of luspatercept in patients with lower-risk myelodysplastic syndromes/ neoplasms (LR-MDS); a single-center study in a heavily pretreated cohort. Presented at: 2024 SOHO Annual Meeting; September 4-7, 2024; Houston, Texas.
   Poster MDS-199.
2. US FDA approves Bristol Myers Squibb's Reblozyl (luspatercept-aamt) as first-line treatment of anemia in adults with lower-risk myelodysplastic syndromes who may require transfusions. News release. Bristol Myers Squibb. August 28, 2023. Accessed October 1, 2024. https://news.bms.com/news/details/2023/U.S.-FDA-Approves-Bristol-Myers-Squibbs-Reblozyl-luspatercept-aamt-as-First-Line-Treatment-of-Anemia-in-Adults-with-Lower-Risk-Myelodysplastic-Syndromes-MDS-Who-May-Require-Transfusions/default.aspx