Association between Platelet-Associated Immunoglobulin G Levels and Response to Corticosteroid Therapy in Patients with Newly Diagnosed Immune Thrombocytopenia

Abstract

Objective: Platelet-associated immunoglobulin G (PA-IgG) refers to IgG attached to the surface of platelets, while the immature platelet fraction (IPF) reflects the state of platelet production in bone marrow. Since PA-IgG and IPF are increased in patients with immune thrombocytopenia (ITP), reflecting amounts of platelet antibodies and compensatory platelet production, respectively, we hypothesized that these laboratory findings may provide useful markers for predicting treatment response in patients with ITP. We therefore retrospectively investigated associations between levels of these markers at diagnosis and response to first-line therapy in patients with ITP. Methods: Forty-three patients diagnosed with ITP at Oita Kouseiren Tsurumi Hospital between May 2010 and November 2018 were included. Patients were divided into 2 groups based on response to corticosteroid as first-line therapy. Laboratory findings were compared between responders and nonresponders. Results: Median PA-IgG was 285 ng/10⁷ cells (range, 45.5–18,200 ng/10⁷ cells), and median IPF was 15.5% (range, 5.4–62.1%). Median levels were higher than the respective upper limits of normal range (PA-IgG, 0–46 ng/10⁷ cells; IPF, 1.1–9.5%). First-line therapy was performed using standard-dose prednisolone (0.5–1.0 mg/kg/day) in 32 patients and high-dose dexamethasone (40 mg/day, 4 days) or methylprednisolone (125–1,000 mg/day, 3–4 days) in 11 patients. Twenty-four patients (55.8%) responded to first-line therapy. In univariate analysis, type of corticosteroid (p = 0.17) tended to differ between groups but did not differ significantly, and no difference in IPF level was apparent between responders (15.35%; range, 5.4–41.5%) and nonresponders (16.7%; range, 6.3–62.1%; p = 0.15). PA-IgG was significantly higher among nonresponders (430 ng/10⁷ cells; range, 101–18,200 ng/10⁷ cells) than among responders (254.5 ng/10⁷ cells; range, 45.5–470 ng/10⁷ cells; p = 0.004). Multivariate analysis revealed PA-IgG was independently associated with response to first-line therapy (odds ratio, 1.000; 95% confidence interval, 1.000–1.010; p = 0.029). Conclusion: Our data suggested that PA-IgG at diagnosis could offer a useful predictor of response to first-line corticosteroid therapy for ITP.