Prevalence and Clinical Significance of Elevated Antiphospholipid Antibodies in Adults with Immune Thrombocytopenic Purpura.

Background: The clinical and pathogenic significances of the presence of antiphospholipid antibodies (APA) [anticardiolipin antibodies (aCL) and lupus-like anticoagulant (LA)] in patients with immune thrombocytopenic purpura (ITP) remain controversial.

Objective: To determine the prevalence and clinical significance of positive APA in a large monocentric cohort of adults with newly diagnosed ITP.

Methods: aCL (IgG and IgM) and LA were searched at time of diagnosis in 260 adults with ITP and a platelet count < 50×10⁹/L over a 15 years period. Thirty-five patients with definite systemic lupus, 5 with primary antiphospholipid syndrome and 4 with HIV infection were excluded and the analysis was performed on 216 patients.

Results: APA (aCL and LA) were detected at ITP diagnosis in 55 patients (25%) including: IgG-aCL in 43 patients (20%) with a level between 15-39 UI/L in 31 and ≥40UI/l in 12; IgM-aCL in 11 patients (5%); and LA in 16 patients (7%). LA was associated with IgG-aCL in 14 patients, IgM-aCL in 1 patient, and was isolated in 1 patient. Positive test for LA was correlated with the presence of highly positive IgG-aCL (≥40UI/L) (p=0.001). Age, gender, initial platelet count, acute or chronic outcome of ITP were not associated with the presence of APA (aCL and LA), aCL and LA status except for age which was lower in LA positive group (mean age 22 years vs 42 years in LA negative group, p=0.002). After a median follow-up of 31 months (range: 1 to 185 months), 15/216 (7%) patients experienced a thrombosis (arterial = 3, deep venous thrombosis and/or pulmonary embolism = 12); 5 of whom (33%) had positive aCL (with a level ≥ 40 UI/L in 4). In 4 of these 5 patients, aCL were associated with positive LA. In multivariate analysis, thrombosis events were associated with age [Hazard ratio = 1.5 (IC95% : 1.1–2.0], presence of LA [Hazard ratio = 6.5 (IC95% : 1.7–25.5] or presence of IgG-aCL at high level (≥40UI/L) [Hazard ratio = 6.1 (IC95% : 1.6–23.7]. In contrast, there was no association of thrombosis events with sex, platelet count at diagnosis, acute or chronic evolution (≥ 6 months), antinuclear antibodies or treatments used.

Conclusions: Prevalence of APA (25%) in adults with newly diagnosed ITP is lower in our cohort than previously observed in other studies. Though the incidence of thrombosis is significantly increased in patients with LA and/or highly positive aCL, this risk seems too low to recommend a systematic testing in patients with ITP and no previous history of thrombosis.