Background: It is widely accepted that autoantibodies directed against platelet glycoproteins (GP) are a major pathophysiological mechanism in immune thrombocytopenia (ITP), but little clinical data is available demonstrating an association between platelet antibodies and platelet counts.

Objective: We hypothesized that if platelet antibodies are clinically relevant, number of targeted glycoproteins and antibody concentration should be associated with the extent of thrombocytopenia.

Methods: Platelet antibodies were identified in a direct, GP-specific test that detects antibodies against GP IIb/IIIa and GP Ib/IX. Using laboratory data from 12,335 thrombocytopenic patients with and without GP-specific platelet antibodies, we conducted a large retrospective cohort study.

Results: We identified 1,469 adults with GP-specific platelet antibodies in our database with complete entries. Compared to 10,866 adults without antibodies, patients with antibodies had significantly lower median platelet counts (54 G/l [IQR, 29-89] vs. 85 G/l [IQR, 52-123], P<0.0001). Patients with antibodies against two GPs had significantly lower platelet counts than patients with antibodies against one GP (47 G/l [IQR, 26-81] vs. 62 G/l [IQR, 32-99], P<0.0001 for GP IIb/IIIa and 58 G/l [IQR, 32-99], P=0.0004 for GP Ib/IX). Increasing antibody levels correlated with decreasing platelet counts for anti-GP IIb/IIIa (R2=0.69; rho -0.84), and anti-GP Ib/IX (R2=0.57; rho -0.6).

Conclusions: The presence of autoantibodies against GP IIb/IIIa or GP Ib/IX is associated with lower platelet counts. More glycoproteins targeted by autoantibodies and increasing antibody levels predict lower platelet counts. Platelet antibodies appear to be of clinical relevance.