Biomarkers of Ineffective Erythropoiesis Predict Response to Luspatercept in Patients with Low or Intermediate-1 Risk Myelodysplastic Syndromes (MDS): Final Results from the Phase 2 PACE-MDS Study

Background: Luspatercept is a fusion protein (modified activin receptor IIB-IgG Fc) being investigated for the treatment of anemias with ineffective erythropoiesis. MDS patients have increased Smad2/3 signaling in the bone marrow, leading to ineffective erythropoiesis. Luspatercept inhibits Smad2/3 signaling and promotes late-stage erythroid differentiation, thereby correcting ineffective erythropoiesis.  

Aims: This completed, 3-month, phase 2, multicenter, open-label study evaluated the effects of luspatercept on anemia in patients with low/int-1 risk MDS (IPSS classification). Study outcomes include erythroid response of increased hemoglobin (Hb) in low transfusion burden (LTB) patients (<4 RBC units/8 weeks), reduced transfusions in high transfusion burden (HTB) patients (≥4 RBC units/8 weeks), transfusion independence, safety, PK, and PD biomarkers.

Methods: Inclusion criteria include age ≥ 18 yr, anemia defined as being HTB or LTB with Hb <10.0 g/dL, EPO >500 U/L or nonresponsive/refractory to ESAs, no prior azacitidine or decitabine, and no current treatment with ESA, G-CSF, GM-CSF, or lenalidomide. Luspatercept was administered once every 3 weeks by SC injection in sequential cohorts (n=3-6 each) at dose levels ranging from 0.125 to 1.75 mg/kg for up to 5 doses with a 3-month follow-up. An expansion cohort with a starting dose level of 1.0 mg/kg was enrolled with individual patient dose titration allowed. Patients completing this study were eligible to enroll into the PACE-MDS extension study. 

Results: Enrollment is complete for 27 patients in the dose escalation cohorts and 31 patients in the expansion cohort (total n=58). Preliminary safety and efficacy data (as of 17-Apr-2015) were available for 49 patients (22 female/27 male, 17 LTB/32 HTB). Final 3-month data for all 58 patients will be presented. Median age was 71 years, 61% had prior ESA therapy, and 18% had prior lenalidomide. Forty (83%) patients had ≥15% ring sideroblasts (RS+) in the bone marrow. The median baseline Hb for LTB patients was 8.7 g/dL (range 6.8-10.1 g/dL). The median number of RBC units transfused over 8 weeks prior to treatment for HTB patients was 6 units (range 4-14 units). Luspatercept was generally well-tolerated.

In patients (n=40) receiving 0.75-1.75 mg/kg, 48% of patients responded per IWG HI-E (Hb increase ≥1.5 g/dL for LTB patients or reduction of ≥ 4 RBC units/8 weeks for HTB patients).   Higher response rates were observed in RS+ patients, including patients with EPO ≥ 200 U/L; patients with splicing factor (SF) mutations present (primarily SF3B1) had 60% response rate (see Table).  Of the patients who received RBC transfusions prior to luspatercept treatment (range 2-14 units/8 weeks), 11 of 30 (37%) patients were transfusion-free for ≥8 weeks during the 12 week treatment period.  Neutrophil responses (IWG HI-N) were seen in 4 of 8 (50%) patients with baseline neutrophil count <1.0 X 10^9/L.

Patient Subgroup	IWG HI-E Response Rate (0.75-1.75 mg/kg)
RS positive1	19/35 (54%)
EPO < 200 U/L	14/23 (61%)
EPO ≥ 200 U/L	5/12 (42%)
RS negative1	0/4 (0%)
SF mutation2 present	18/30 (60%)
SF mutation2 absent	1/9 (11%)

                                                     ¹RS: Ring sideroblasts ≥15% in bone marrow; data available for n=39
                                                     ²SF: Splicing factor mutations: SF3B1, SRSF2, U2AF1, or ZRSR2; data available for n=39

The relationship of erythroid response to bone marrow erythroid cellularity, morphology and characteristics, comprehensive gene mutation profile, GDF15, and other biomarkers of ineffective erythropoiesis will be presented.

Conclusions: Based on these data, treatment with luspatercept in lower risk MDS patients at therapeutic dose levels for 3 months led to a high response rate for increased Hb levels or decreased transfusion burden, including transfusion independence, with a favorable safety profile.  Ring sideroblasts, splicing factor gene mutations and other biomarkers of ineffective erythropoiesis may help define a MDS population most likely to respond to luspatercept treatment.  Phase 3 studies in patients with anemia due to lower risk MDS are planned.