Impact of Platelet Glycoprotein Polymorphisms upon Clinical Events in Essential Thrombocythaemia Patients Enrolled in the PT1 Trial.

Multiple factors are likely to contribute to the pathogenesis of thrombosis in Essential thrombocythaemia (ET) including platelet number, activation of platelets and leukocytes; the formation of platelet leucocyte aggregates, and circulating prothrombotic and endothelial factors. The normal functional activity of platelets depends on the presence of platelet glycoprotein complexes, which play an important role in platelet adhesion and aggregation. A number of polymorphisms in platelet glycoproteins have been correlated with thrombotic events in hematologically normal patients. Particular polymorphisms of interest include three within the glycoprotein Ibα gene: −5T/C (affecting a Kozak sequence), 1018C/T (T145M, encoding Human Platelet Antigen − 2), a variable number of tandem repeats (VNTR) in a region encoding the macroglycopeptide and one polymorphism C807T, within the glycoprotein Ia gene. To date whether there is a relationship between these platelet glycoprotein polymorphisms and the clinical features of ET has not been determined. In this study samples obtained from 797 ET patients recruited to the 3 large prospective PT–1 trials were genotyped for the polymorphisms of interest and the results were correlated with clinical events including haemorrhagic and arterial or venous thrombotic events in the year prior to diagnosis and following trial entry (median 30 months). The VNTR data were analysed using the sum of the number of repeats of both alleles and treated as a continuous variable using logistic regression for retrospective and a Cox survival model for prospective analyses; other polymorphisms were analysed using the Chi-squared test for the retrospective clinical events, and log-rank survival analysis for the prospective clinical events. Neither the C807T, −5T/C Kozak nor the T145M polymorphism was associated with clinical events in the ET patient cohort. Interestingly the number of VNTR repeats was inversely associated with (p=0.02) rate of arterial events after trial entry in this patient cohort, but not for events in the year prior to diagnosis or post trial entry venous thrombotic or haemorrhagic events. This analysis suggested that a decreasing sum of VNTR were associated with an increased risk of arterial thrombosis, the degree of increase in risk was 75% per repeat (95% confidence interval 1.0–2.9). After multivariate analysis of this data with correction for age, sex, JAK2^V617F and MPL^W515L/K status the effect of the number of VNTR repeats remained significant (p=0.048). These results suggest that the C807T, T145M and −5T/C Kozak polymorphisms do not correlate with clinical events in ET. However a reduction in the number of VNTR within the glycoprotein Ibα gene was progressively associated with arterial events post trial entry (p=0.024) and remained significant on multivariate analysis (p=0.048).