Risk Factors of Thrombosis in Adults with Primary Immune Thrombocytopenia. a French Nationwide Cohort Study

Background: Immune thrombocytopenia (ITP) is associated with an increased risk of venous and arterial thrombosis (VT and AT, respectively) as compared with the general population. However, the impact of thrombosis risk factors and of ITP treatments, particularly of thrombopoietin-receptor agonists (TPORAs), is not well known in the routine clinical practice.

Aim: The objective of this cohort study was to assess the risk factors of VT and AT in adults with primary ITP, including ITP treatments.

Methods: The population was the cohort of all incident primary ITP adults in France during 2009-2015 built within the national health insurance database (French Adult Immune Thrombocytopenia - FAITH - cohort; NCT03429660). Incident ITP patients were identified using a validated algorithm combining drug exposures and diagnosis codes according to the international classification of diseases, version 10 (ICD-10). Risks of first hospitalization with a validated primary discharge diagnosis code of VT and AT (coded with the ICD-10) were assessed separately. Cox regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CI). Variables included in multivariable models were: age, sex, history of AT and of VT, diabetes, cardiovascular disease, chronic kidney disease, chronic liver disease, cancer; exposures to antihypertensive, lipid-lowering, antiplatelet, anticoagulant drugs and ITP treatments including splenectomy were modeled as time-dependent variables.

Results: The cohort included 7225 adult patient with incident primary ITP: 3807 (52.7%) were ≥60 year-old, 3199 (44.3%) were males, 692 (9.6%) had a history of cardiovascular disease, 937 (13.0%) had diabetes. During the follow-up, 5737 (79.4%) were exposed to corticosteroids, 3364 (46.6%) to intravenous immunoglobulin (IVIg), 995 (13.8%) to TPORAs, and 755 (10.4%) were splenectomized. During the follow-up (23 852 patient-years in total; mean follow-up: 39.5 months), 174 patients had a hospitalization with a primary discharge diagnosis of VT and 333 of AT, leading to incidences of 7.4 (95% CI: 6.4-8.6) and 14.4 (95% CI:12.9-16.0)/1000 patient-years, respectively. In multivariable Cox models, the most important risk factors for VT were higher age (≥60 years vs. <40 years: HR: 2.22, 95% CI: 1.39-3.53), a history of VT (HR: 4.38, 95% CI: 1.07-18.02), splenectomy (HR: 3.22, 95% CI: 2.06-3.03), exposure to IVIg (HR: 2.30, 95% CI: 1.41-3.75), corticosteroids (HR: 3.29, 95% CI: 2.39-4.53) and TPORAs (HR: 3.16, 95% CI: 2.04-4.88). All classical baseline cardiovascular risk factors listed above as covariables were associated with the risk of AT. The HRs for AT were 0.97 (95% CI: 0.59-1.61) for splenectomy, 1.05 (95% CI: 0.80-1.40) for corticosteroids, 2.35 (95%CI: 1.58-3.50) for IVIg, 1.25 (95%CI: 0.46-3.37) for danazol and 1.31 (95%CI: 0.84-2.06) for TPORAs. It is of note that among the 25 patients who had a VT while treated by TPORA, 18 (72.0%) were>50 year-old, 14 (56.0%) were women, 6 (24.0%) were splenectomized, 9 (36.0%) were concomitantly exposed to corticosteroids and 3 (12.0%) to IVIg; only 3 women aged<50 years had no additional risk factor. Among the 21 patients who had an AT while treated by TPORA, 18 (85.7%) were>50 year-old, 15 (71.4%) were men, 8 (38.1%) were splenectomized, 5 (23.8%) were concomitantly exposed to corticosteroids and one to IVIg; only one 48-year-old man had no additional risk factor for AT.

Conclusions: Baseline risk factors for VT and AT were highly associated with VT and AT occurrence in adults with primary ITP. Splenectomy, corticosteroids, IVIg and TPORAs were risk factors for VT. Most patients who had a thrombosis while treated by TPORA had additional risk factors. These findings help choosing a tailored treatment strategy for a given patient depending on his/her risk profile for VT and AT.