A Systematic Literature Review and Meta-Analysis of the Risk of Thromboembolic Events in Patients with Immune Thrombocytopenia (ITP) in Observational Studies.

A recent systematic review and meta-analysis of randomized controlled trials of thrombopoietin-receptor agonists (TPOr), romiplostim and eltrombopag, in adult thrombocytopenic patients found a numerically but non-statistically significant increase in the occurrence of thromboembolism (TE) among treated patients compared to those not treated with TPOr.¹ To put these results into context and to estimate the baseline risk of TE in ITP patients, this study will summarize results of observational studies that reported the risk of venous thrombembolism (VTE) or arterial thrombembolism (ATE) in patients with ITP relative to comparable populations without ITP.

We searched MEDLINE and EMBASE via Ovid and the Cochrane database for articles that included both terms relating to ITP (i.e., immune thrombocytopenia, idiopathic thrombocytopenia) and TE (i.e., thromboembolism, thrombosis, embolism). Articles were restricted to research on humans, published in English, January 1996 to July 2012. Only observational studies were included. We abstracted measures of relative risk (or rate ratio) comparing the incidence of TE in ITP patients with that in a comparable non-ITP population. The pooled relative risk and 95% confidence intervals (CIs) were calculated by taking a weighted average of individual study results using both fixed and random effects models using the META² module for STATA 10.

Five observational studies met the inclusion criteria: two from Denmark and one each from Sweden, United Kingdom, and United States; all published 2010–2012. All studies completed patient follow-up before the commercial availability of TPOr. Studies varied in the duration of follow-up, choice of comparison group, and specific events reported. Some studies excluded patients with previous events. The relative risk of any VTE among patients with ITP compared to that in a non-ITP population ranged from 1.6 to 2.9 based on 3 studies that reported VTE. The pooled relative risk of any VTE was 1.9 (95%CI 1.4, 2.7) using a fixed-effect model and similar estimates using a random effect model (test for heterogeneity, p = 0.3). The relative risk of any ATE among patients with ITP compared to that in a non-ITP population ranged from 1.3 to 1.6 based on 3 studies that reported ATE. The pooled relative risk of any ATE was 1.5 (95%CI 1.3, 1.8) for both the random and fixed effects models (test for heterogeneity, p = 0.7).

Five population-based observational studies have been published recently comparing the risk of thrombembolism among ITP patients to populations without ITP. The results of a meta-analysis showed nearly a 2-fold increased risk of VTE and a 50% increased risk of ATE among the general population of ITP patients not treated with TPOr. These ITP patients were likely to have less severe disease than those in experimental trials where entrance criteria may require significant thrombocytopenia, prior bleeding episodes, or failure of a previous therapy. However, the demonstrated elevated risk of TE among patients with ITP should be considered when evaluating the risk of TE ascribed to ITP treatments.

Langeberg:Amgen: Employment, Equity Ownership. Eisen:Amgen: Employment, Equity Ownership. Schoonen:Amgen: Employment, Equity Ownership. Gamelin:Amgen: Employment, Equity Ownership. Stryker:Amgen: Employment, Equity Ownership.