Predictors of Remission in Adults with Immune Thrombocytopenia Treated with Romiplostim

Background: In adults with ITP, thrombopoietin receptor agonists (TPO-RA) are often assumed to be lifelong therapy. However, some patients achieve treatment-free remission (TFR) by maintaining hemostatic platelet counts after discontinuing TPO-RA, despite not replacing them with other ITP medications. Predictors of safely discontinuing romiplostim have varied across studies. In an analysis of 8 romiplostim studies, 27 patients achieved TFR; more remitters than nonremitters had ITP for <12 months but there were no unequivocal predictors of TFR (Bussel et. al. Hematology 2016;21:257). In a phase 2 study with forced romiplostim taper after 12 months, 24 (32%) of 75 patients achieved TFR; a higher platelet count in the first 2 months was associated with TFR (P<.05; Newland et. al. Br J Haematol 2016;172:262). Differences between remitters and nonremitters for baseline patient characteristics and time to first platelet response were not statistically significant. Analysis of romiplostim treatment at 2 centers found that 12 (28%) of 43 patients achieved TFR; splenectomized patients were more likely than nonsplenectomized patients to achieve TFR (P=.05; Marshall et. al. Haematologica 2016;101:e476).

Aim: This multivariate analysis integrated data across multiple romiplostim studies in adults with ITP to explore predictors of TFR.

Methods: Data were pooled for 911 romiplostim-treated patients in 13 ITP studies conducted from 2002 to 2014. All patients failed first-line treatments before study enrollment. Secondary thrombocytopenia patients were excluded. Romiplostim treatment was discontinued per dosing rules; 1 study included a forced taper after 1 year of romiplostim. Typically, the romiplostim dose was reduced for platelet counts >200×10⁹/L and withheld, then reduced, for platelet counts >400×10⁹/L. Concomitant ITP medication could be reduced for platelet counts >50×10⁹/L. TFR was defined as a ≥6-month treatment-free period with platelet counts consistently ≥50×10⁹/L. Univariate analysis with logistic regression examined potential predictors of TFR, including age, sex, ITP duration, prior splenectomy, platelet count (baseline and first 4 weeks), bleeding in the first 6 months, and baseline concurrent therapy. Multivariate analysis with logistic regression evaluated significant predictors from the univariate analysis.

Results: TFR was achieved by 61 ITP study patients who received romiplostim. For remitters vs nonremitters, median baseline age was 53 years in both cohorts; 57% vs 61% were female; and 93% vs 88% were Caucasian (Table 1). Median ITP duration at study baseline was 0.5 years among remitters and 3.5 years among nonremitters. Prior splenectomy occurred in 21% of remitters and 36% of nonremitters. The number of prior ITP treatments was ≤3 for 74% of remitters and 39% of nonremitters. Baseline platelet count was <30×10⁹/L for 80% of remitters and 78% of nonremitters. Median platelet counts on treatment during the romiplostim studies were often higher among remitters than nonremitters, despite increasing median doses of romiplostim over time for nonremitters (Figure). During the first 6 months on study, 30% of remitters and 43% of nonremitters had a bleeding event. Prior splenectomy and bleeding in the first 6 study months were significant predictors of lower odds of TFR in the univariate model but not in the multivariate model. In the multivariate model, shorter ITP duration predicted significantly higher odds of TFR, either <3 months (newly diagnosed) vs >12 months (chronic; odds ratio [OR], 4.275; P<.0001) or 3-12 months (persistent) vs chronic (OR, 2.171; P=.0408; Table 2). One limitation of this analysis was 19 of 61 patients entering TFR were from one study of newly diagnosed and persistent patients; the protocol included a forced romiplostim taper. Number of previous treatments was not included in the multivariate analysis.

Conclusions: In this integrated analysis of 911 adult ITP patients across 13 studies, shorter ITP duration at baseline (≤12 months) was an independent predictor of TFR. Neither previous splenectomy (which predicted TFR in 1 study) nor bleeding were independent predictors of TFR in this much larger analysis. In total, these results imply that earlier use of romiplostim in adults with ITP could be associated with greater likelihood of achieving TFR. Future studies should also explore biologic variables as predictors of TFR.

View largeDownload slide

View largeDownload slide

Close modal