Real‐world experience with luspatercept and predictors of response in myelodysplastic syndromes with ring sideroblasts

Abstract

Myelodysplastic syndromes with ring sideroblasts (MDS-RS) is a subcategory of MDS with ≥15% (or ≥5% in the presence of SF3B1 mutation) of bone marrow erythroid precursors; additional diagnostic criteria include the absence of ≥5% bone marrow or ≥1% peripheral blood myeloblasts, Auer rods, and diagnostic criteria for MDS with isolated del(5q).1 Prognosis in MDS-RS is more favorable than other MDS subtypes with estimated median survival of 46 months and a relatively low risk of leukemic transformation.2 Currently available drugs in MDS-RS do not modify the natural history of the disease and are primarily directed at anemia and include erythropoiesis stimulating agents (ESAs), lenalidomide, hypomethylating agents (HMAs), and luspatercept.3 ESA response rates in lower-risk MDS range from 30% to 60%, with median duration of response estimated at 23 months4; transfusion-dependent patients or those with serum erythropoietin (Epo) level above 200 IU/L are less likely to respond to ESAs.5 HMAs, used as second-line agents, have been shown to produce transfusion independence in approximately 30%–40% of patients with low-/intermediate-risk MDS without a del(5q).6 Lenalidomide works best in del(5q) MDS but has also been shown to reduce red cell transfusion needs in about 56% of lower-risk MDS patients with a normal karyotype7; transfusion independence ≥8 weeks was obtained in 26.9% and 2.5% of patients in the lenalidomide and placebo groups, respectively, in a randomized, phase-3 clinical trial utilizing lenalidomide or placebo in transfusion-dependent, lower-risk MDS patients without del(5q).8 Luspatercept is a recombinant activin receptor type IIB fusion protein that traps TGF-β superfamily ligands and derails SMAD2/3 signaling allowing late-stage erythroblast differentiation and erythroid maturation.9 Luspatercept is currently FDA approved for use in ESA-refractory adult patients with transfusion-requiring low/intermediate-risk MDS-RS, based partly on data received from the phase-3 MEDALIST clinical trial,10 which compared the efficacy of luspatercept to placebo in lower-risk MDS-RS patients who had transfusion dependence and were refractory, intolerant, or ineligible for ESA. A total of 153 individuals were given luspatercept at a dose of 1 mg/kg subcutaneously every 21 days, while 71 patients were given placebo. The primary and secondary endpoints for the trial were defined as transfusion independence for ≥8 or ≥12 weeks, respectively. The primary endpoint was achieved in 58 (31.9%) patients treated with luspatercept and 10 (13.2%) placebo, whereas the secondary endpoint was achieved in 51 (33.3%) patients on luspatercept and 9 (11.8%) placebo.10 Reported side effects of luspatercept included fatigue, diarrhea, asthenia, nausea, and dizziness. In the current study, we describe our experience in a consecutive series of patients with MDS-RS who received luspatercept therapy outside of a clinical trial setting and after the FDA approval date of April 2020. The current study was conducted under an institutional review board approved minimum risk protocol that allowed retrospective collection and analysis of data from patients who were seen at the Mayo Clinic and received treatment with luspatercept at the discretion of their treating physician. Conventional criteria were used for diagnosis and disease classification.1 Response criteria used were mostly in line with previous publications10 and included transfusion independence for at least 8 weeks for transfusion-dependent patients and an increase in hemoglobin by 1.5 g/dL lasting for at least 8 weeks, in transfusion-independent patients. Transfusion dependency was assigned to those patients who had received at least 3 units of packed red cells in the 8 weeks prior to therapy with luspatercept. Response assignments were double-checked in an independent fashion to assure accuracy. Conventional statistical methods were applied using JMP Pro 16.0.0 software (SAS Institute, Cary, NC). Baseline characteristics of the 39 study patients (median age 75 years; 77% males) and treatment information are detailed in Table 1 and Tables S1 and S2; information on karyotype, mutations, serum Epo level, and revised international prognostic scoring system (IPSS-R) risk assignment were obtained from the time of initial diagnosis; all other variables including hemoglobin level, leukocyte count, absolute neutrophil count (ANC), absolute monocyte count (AMC), absolute lymphocyte count (ALC), platelet count, and age were obtained at the time of initiation of luspatercept therapy. As outlined in Tables 1, 87% of patients were IPSS-R very low/low risk category, at time of initial diagnosis, and harbored SF3B1 (83%), DNMT3A (31%), or TET2 (31%) mutations, and abnormal karyotype (39%). Table 1 also includes hemoglobin level (for transfusion-independent patients), leukocyte count, ANC, AMC, ALC, and platelet count recorded at time of treatment with luspatercept. Prior therapy included ESAs (95%), HMAs (38%), and lenalidomide (26%). Thirty-one (80%) patients were transfusion-dependent at the time of luspatercept initiation. Serum Epo at the time of diagnosis was documented in 25 patients (median 85 IU/L; range 20–653). Nine (23%) patients displayed single-lineage and 30 (77%) multi-lineage dysplasia. The median luspatercept starting dose was 1 mg/kg (0.8-1 mg/kg) (Table 1). The median treatment duration was 4 months (range 1–20). Luspatercept dose was increased in 18 (55%) patients due to inadequate response; the median increase in dose was 1.33 mg/kg (range 1.25–1.75 mg/kg). The median duration from initiation of therapy with luspatercept to last follow-up was 12 months (range 2–21). Overall, 7 (18%) patients met criteria for anemia response, including 5 (16%) of 31 transfusion-dependent patients, by way of achieving ≥8 weeks transfusion-free period, and 2 (25%) of 8 patients who were transfusion-independent at the time of treatment initiation (Table S1). All 5 (16%) patients who remained transfusion-independent for 8 weeks sustained their response for at least 12 weeks. Median response duration for transfusion-dependent patients was 6 months (range 3–19) and the duration of response for the two transfusion-independent patients were 3 and 7 months (Table S1). Among the total seven responding patients, four lost their response in 3–7 months (Table S1). In addition to the anemia response, among the seven treatment responders, only one patient displayed platelet increase from grade 1 thrombocytopenia to normal levels. There were no changes in five patients and one patient experienced worsening of counts in both ANC and platelets. Additional individual patient-specific details are outlined in Table S2. In univariate analysis, borderline significance was detected for association of anemia response to luspatercept with higher serum Epo level (p = .15; N informative = 25) and abnormal karyotype (p = .05; N informative = 36), both obtained at the time of diagnosis, and higher AMC (p = .17; N informative = 35) and ALC (p = .1; N informative = 36), both obtained at the time of treatment initiation. Optimal cutoff levels were subsequently determined by receiver operating characteristic curve (ROC) analysis that confirmed significant response prediction for serum Epo level >80 IU/L (p = .01), ALC ≥1.8 × 109/L (p = 0.005), and AMC ≥0.5 × 109/L (0.03). Serum Epo level >80 IU/L retained its significance on multivariable analysis of variables obtained at the time of diagnosis; among 25 informative cases, none (0%) of 10 patients with serum Epo level ≤80 × 109/L responded versus 5 (33%) of 15 with serum Epo level >80 × 109/L (p = 0.01) responded. Multivariable analysis of variables obtained at the time of luspatercept treatment initiation identified ALC ≥1.8 × 109/L as the sole predictor of response; 3 (75%) of 4 patients with ALC ≥1.8 × 109/L responded versus 3 (9.4%) of 32 with lower ALC. None of the other patient characteristics including age, gender, neutrophil count, platelet count, SF3B1, ASXL1, or DNMT3A mutations predicted response to therapy. At the time of last follow-up, treatment was discontinued in 23 (59%) patients because of no response (n = 17), loss of response (n = 1), side effects (n = 3), leukemic transformation (n = 1), and hospice placement due to metastatic lung cancer (n = 1). At the time of this writing, 16 (41%) patients remain on active therapy, with only 3 (8%) maintaining response for a duration of 3, 6, and 19 months; of note, 3 patients continued treatment despite losing response after 3, 5 and 7 months. The most common luspatercept-associated adverse events (of any grade) included fatigue, dizziness, abdominal pain, muscle aches, bone aches, dyspnea on exertion, palpitations; these side effects were observed in 9 (23%) patients (Table 1). The current study highlights the limited value of luspatercept in securing durable anemia response in MDS-RS patients treated in the context of routine clinical practice. The size (18% response rate) and transient nature (6 months median duration of response) of treatment response raise questions regarding the cost-effective value of the drug; as of January 2022, the estimate for maximum billable to Medicare is $11 219 per 75 mg dose to cover medication, preparation, and administration (CMS Drug Spending. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Information-on-Prescription-Drugs. Accessed 15 Feb 2022). While many may currently have full coverage, current controversy with white bagging practices may push large portions of these expenditures to Medicare Part D where annual out-of-pocket expenses exceed $7000, making this unaffordable without patient assistance funding. It would be inappropriate to compare our observations with those reported in the MEDALIST trial (38% achievement of 8-week transfusion independence) for obvious reasons that include differences in study design, response criteria used, dosing schedules, and heterogeneity in patient population; the latter point considers the fact, as was shown in the current study, that real-world patients had received multiple drugs including HMA and lenalidomide before being offered luspatercept. Regardless, our observation that only 7% of the patients receiving luspatercept in routine clinical practice maintained their response at the time of their last follow-up is noteworthy; incidentally, the median duration of response in the MEDALIST study (8 months) was comparable to that of the current study (6 months). Our novel observations regarding predictors of response (i.e., increased serum Epo at diagnosis and ALC at time of treatment) require further validation but offer the opportunity for identifying patients who may derive benefit from the drug, as is the case when selecting appropriate patients for ESA use.3 Furthermore, future studies should consider laboratory correlative studies that account for such possible correlations. Interestingly, we have recently reported on the adverse prognostic effect of lymphocytopenia in MDS-RS,11 and the observation from the current study regarding its potential impact on response to luspatercept therapy further strengthens the contribution of host immunity to disease biology in MDS-RS.12 In regards to our observation on the association between luspatercept-induced anemia response and higher serum Epo level, it is tempting to speculate that serum Epo level in MDS-RS directly correlates with the degree of ineffective erythropoiesis, and thus identifies patients who are more likely to respond to drugs targeting ineffective erythropoiesis. Regardless, current deficit in terms of effective anti-anemia drugs in low-risk MDS, including MDS-RS, underscores an unmet need that might be addressed faster by targeting the underlying neoplastic clone7, 13, 14 rather than the associated ineffective erythropoiesis.15, 16 The latter approach has proven more successful in anemias associated with beta-thalassemia.17 Finally, the observations from the current study might also undermine the potential value of luspatercept in the treatment of anemias associated with other myeloid neoplasms, including MDS/MPN-RS-T and myelofibrosis. However, we caution against making definitive conclusions regarding the value of luspatercept in MDS-RS, based on the current study, considering the smaller number of informative patients (n − 39), as opposed to that of the MEDALIST trial (n = 153).10 The authors declare no conflicts of interest. Faiqa Farrukh and Ayalew Tefferi wrote the manuscript including preparation of Tables. Faiqa Farrukh and Deandra Chetram collected patient data that were further quality checked by Ayalew Tefferi and Naseema Gangat. Faiqa Farrukh and Ayalew Tefferi performed analysis and interpretation of patient data. Ayalew Tefferi, Aref Al-Kali, James Foran, Mrinal Patnaik, Talha Badar, Kebede Begna, Christopher Hook, William Hogan, Abhishek Mangaonkar, and Naseema Gangat participated in patient care. Rong He contributed to pathology review. Kristen B. McCullough provided drug cost estimates and oversaw drug prescription and administration. All authors reviewed and approved the manuscript. The data that support the findings of this study are available from the corresponding author. The data that support the findings of this study are available from the corresponding author. Table S1. Luspatercept treatment details and response data among 39 patients with myelodysplastic syndromes with ring sideroblasts (MDS-RS) Table S2. Characteristics of 39 patients with myelodysplastic syndromes with ring sideroblasts (MDS-RS) at start of therapy with luspatercept. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts