Corticosteroids and Thrombotic Risk in Immune Thrombocytopenia: A Systematic Review and Meta-Analysis

Background/Objectives: Corticosteroids are the standard first-line treatment for immune thrombocytopenia (ITP), a disorder that itself carries an excess risk of thrombosis, yet the evidence on corticosteroid-associated thrombotic risk in ITP has not been systematically synthesized. We aimed to estimate the association between corticosteroid exposure and venous (VT) and arterial thrombosis (AT) in ITP. Methods: Following PRISMA 2020 (PROSPERO CRD420261504746), we searched PubMed/MEDLINE, Scopus, Cochrane CENTRAL, LILACS, ClinicalTrials.gov, and WHO ICTRP from 2000 to 28 September 2026. Adjusted hazard ratios (HRs) were synthesized by study and outcome; crude odds ratios (ORs) were pooled with random-effects (REML) models within two subgroups (European cohorts; Asian hospital-based series). Risk of bias was assessed with JBI tools (and ROBINS-E for the adjusted cohorts) and certainty of evidence with GRADE. Results: Twelve studies (18,059 participants) were included, ten in the meta-analysis. Adjusted evidence came almost entirely from a French nationwide cohort (10,039 adults), in which time-varying corticosteroid exposure was associated with VT (HR 3.55, 95% CI 2.74–4.58; approximately 11 additional hospitalizations per 1000 person-years of exposure) but not with AT (HR 1.15, 0.94–1.41); an Italian cohort with ever-use exposure gave imprecise estimates (VT 2.2, 0.5–10.4; AT 5.3, 0.9–40.6). Crude ORs were 3.48 (1.73–7.01) in two European cohorts and 0.82 (0.59–1.14) in seven Asian hospital-based series, none of which adjusted for corticosteroid exposure (P for subgroup difference < 0.001). Two randomized trials comparing corticosteroid regimens reported no thrombotic events (0/328; exact 95% CI 0–1.1%). Certainty was low for VT and very low for the other outcomes. Conclusions: Low-certainty evidence, dominated by a single nationwide cohort, suggests that current corticosteroid exposure is associated with an approximately 3.5-fold higher hazard of VT in adults with ITP, without a consistent arterial association. Retrospective series that compare prior use without a defined exposure window may be less informative about the risk associated with current exposure. Corticosteroid exposure may reasonably be considered in VT risk assessment in ITP.