SAN DIEGO, CA—Long-term use of romiplostim in myelodysplastic syndrome (MDS) patients for up to 3.5 years (five years including therapy on prior studies) produced platelet responses in 83% of patients, according to an open-label extension study presented at the 53rd American Society of Hematology Annual Meeting and Exposition.

Thrombocytopenia, an independent risk factor for survival, occurs in 40–65% of MDS patients. Romiplostim works to increase platelet production. In a previous Phase 1/2 study, 46% of MDS patients demonstrated a durable platelet response following romiplostim therapy (Kantarjian et al, Journal of Clinical Oncology 2010, 28:437–444). The primary endpoint of this extension study was adverse event incidence and development of antibodies to romiplostim and/or thrombopoietin; secondary endpoints were bleeding event incidence, platelet transfusions, and platelet response duration. To evaluate the long-term hematological improvement and sustained clinical benefit of romiplostim, patients who entered into the extension study and continued treatment for >2 years were compared in a post hoc analysis to those who did not enter the extension phase or had <2 years exposure to romiplostim.

All patients in this trial completed a prior romiplostim study, had a platelet count ≤50×109/L and no evidence of disease progression. Out of 72 enrolled patients, 60 had received romiplostim or placebo, seven had received romiplostim with decitabine, and five had received romiplostim with lenalidomide. Median age was 71.5 years and median baseline platelet count was 27.5×109/L (Q1‑Q3: 14–42×109/L). MDS subtypes included: refractory anemia (RA; n=22), RA with ringed sideroblasts (RARS; n=1), RA with excess blasts type 1 (RAEB-1; n=6), RAEB-2 (n=1), refractory cytopenia with multilineage dysplasia (RCMD; n=25), RCMD-RS (n=2), and unclassifiable (MDS-U; n=15). International Prognostic Scoring System (IPSS) status at prior study baseline: low (n=22), int-1 (n=44), int-2 (n=4), and unknown (n=2).  Median duration of MDS (until last contact or acute myeloid leukemia [AML] progression) was 3.1 years (Q1-Q3: 1.7–5.2 years).

Patients received 250mcg, 500mcg, 750mcg, 1,000mcg, or 1,500mcg of romiplostim weekly or biweekly based on previous dosages received. Dosages were adjusted every four weeks based on platelet counts. Treatment ceased if no response was evident after four weeks at a dose of 1,000mcg/week. The median treatment duration during the extension period was 28 weeks (range 2–181). Patients who received romiplostim in prior studies had additional exposure for a median of 52 weeks (range 7–74). The median average weekly dose was 750mcg.

From Week 3 onwards, the median platelet count was ≥50×109/L. Following the International Working Group’s 2006 criteria, 60 patients (83%) had a platelet response. The median time to first platelet response was 2.1 weeks (Q1–Q3: 1.1–3.7 weeks), while the median platelet response duration was 20 weeks (Q1–Q3: 6.5–72 weeks).  Five cases of progression to AML occurred. (See Table). Soon after the data cutoff for the information presented, all patients stopped romiplostim therapy and moved into the long-term observation part of the study.

The most common adverse events were epistaxis (32%), cough (25%), fatigue (24%), anemia (21%), and arthralgia (21%). Eleven deaths occurred, four of which were on study (cardiac arrest + intestinal obstruction, congestive heart failure [CHF], progressive muscle dystrophy, pulmonary fibrosis in a patient with a history of chronic obstructive pulmonary disease and CHF). The last death was considered to be related to romiplostim by the researchers. Seven of the deaths were post-study (four due to AML, one due to respiratory causes, one due to cerebral hemorrhage, one due to unknown causes). The annual rate of AML or death was 14.3% (95% CI: 8.1–25.2%).

Fifty-two (72.2%) patients reported ≥1 bleeding event(s); the incidence rate was 23.3/100 patient-weeks.  The proportion of patients with significant bleeding events decreased over time; 23 patients (32%) reported ≥1 clinically significant bleeding event.  Platelet transfusions occurred in 32 (44%) patients, all occurring the first 48 weeks of romiplostim therapy.

Among patients in this trial, AML progression occurred at expected rates, concluded Pierre Fenaux, MD, PhD of Avicenne Hospital, Bobigny, France. The post hoc analysis indicated that a subset of younger patients, more likely to be RA, and who had lower baseline blasts, were more likely to have better platelet responses observed early as after two weeks of treatment. However, conclusions to be drawn from these results are still limited, and thus, further trials will be needed to better characterize the subset of patients who will benefit from romiplostim therapy.