The Causality between Immune Thrombocytopenia and Thrombosis : A Bidirectional Two-Sample Mendelian Randomization Study

Background: Immune thrombocytopenia (ITP) is an autoimmune hemorrhagic disease characterized by low platelet count and bleeding manifestations. However, observational studies have shown that ITP patients have a significantly higher incidence of arterial and venous thromboembolism compared with the normal population, despite the low platelet counts. Treatments options, complications, and inflammatory status of ITP patients are considered to contribute the thromboembolism formation. This study aimed at investigating the causal associations of ITP with arterial and venous thromboembolism.

Methods: A bidirectional two-sample Mendelian randomization (MR) analysis was designed by using summary-level data from large genome-wide association studies performed in European individuals. Inverse variance weighting with multiplicative random effect method was used to obtain the primary causal estimates, and weighted median, weighted mode, and MR egger regression were replenished as sensitivity analyses to test the robustness of the results. We also collected data from 302 ITP patient hospitalized in Qilu Hospital during October 2016 to September 2018 to observe the incidence of thrombotic events.

Results: Among the 302 ITP patients included in our retrospective study, 49 patients (16.2%) presented with coronary atherosclerosis, 12 patients with cerebral infarction, and 2 patients with deep venous thrombosis. In the MR analysis, we found no significant causal effects of ITP on coronary atherosclerosis (odds ratio [OR]: 1.000034, P = 0.74), arterial thrombosis (OR: 1.030618, P = 0.07), ischemic stroke (OR: 0.9999557, P = 0.07), myocardial infarction (OR: 0.9999557, P = 0.77), and venous thromboembolism (OR:1.0011961, P = 0. 82) including deep venous thrombosis (OR: 0.9998565, P = 0.02), pulmonary embolism (OR: 0.999987, P = 0.69), and portal venous thrombosis (OR:0.997588, P = 0.96). In the other direction, there were also no significant causal effects of thrombotic events on ITP.

Conclusion: This MR analysis demonstrated no significant causal associations of ITP with arterial or venous thromboembolism in both directions. Although the previous observational studies indicated a positive association between ITP and thromboembolism, the external effects may be dominant. Further investigations are still needed for the underlying pathogenesis of thrombosis in ITP.

No relevant conflicts of interest to declare.