Predictive factors of romiplostim response in patients with refractory aplastic anemia: data from two clinical trials.

The thrombopoietin receptor agonist (TPO-RA) romiplostim has been shown to be efficacious in the treatment of aplastic anemia (AA) refractory to immunosuppressive therapy. However, there are few reports on predictors of TPO-RA efficacy. We analyzed patient clinical and baseline laboratory data from two clinical trials that enrolled patients with refractory AA, a phase 2 dose-finding study (NCT02094417) and a phase 2/3 study (NCT02773290), to determine predictors of romiplostim efficacy. Data from 66 patients (35 from the phase 2 dose-finding study and 31 from the phase 2/3 study), median age 47.0 (range 20-78) years, were analyzed. Of these, 51.5% (34/66) had non-severe AA, 86.4% (57/66) had received prior therapy with anti-thymocyte globulin plus cyclosporine, and 80.3% (53/66) were transfusion-dependent. At 27 weeks, no patients achieved a complete response, 57.4% (35/61) achieved a partial response, 49.0% (24/49) became red blood cell transfusion-independent, and 56.8% (21/37) became platelet transfusion-independent. Univariate logistic regression analysis identified duration from diagnosis (P = 0.040), baseline reticulocyte count (P < 0.001), and platelet count (P < 0.001) as predictors of response to romiplostim at 27 weeks. Receiver operating characteristic curve analysis identified a cutoff value of 30.77 × 10/L at baseline as the threshold reticulocyte count to predict response to romiplostim in patients with refractory AA (sensitivity: 82.9%, specificity: 73.1%). Of note, the initial dose included in this analysis was lower than that used in real-world clinical practice, as data from the phase 2 dose-finding study were used. This study identified factors predicting response to romiplostim at 27 weeks.Trial registration: ClinicalTrials.gov NCT02094417 and NCT02773290.