Copy Number Variation of PROS1 and Protein S Deficiency Are Rare in a Population-Based Case-Control Study On Venous Thrombosis (MEGA study).

Abstract 3166

Poster Board III-106

Protein S (PS) deficiency increases the risk of venous thrombosis (VT) in family studies, but it has been difficult to estimate the risk in population-based studies. Moreover, the genetic basis of PS deficiency is only poorly understood as in about 50% of PS deficient families no mutations are found by exon-targeted sequencing of the PS gene PROS1. Further, the laboratory diagnosis of PS deficiency is complicated by physiological variation due to age and sex, as well as interference by pregnancy, use of hormones and medication and vitamin K deficiency. We recently showed with the multiplex ligation-dependent probe amplification (MLPA) methodology, that whole or partial deletion of PROS1 explained PS deficiency in around 40% of point mutation-negative French PS deficient families. Here, we analyzed PS deficiency in the MEGA (Multiple Environmental and Genetic Assessment) study to verify whether gross gene deletions or duplications are also common in unselected individuals with venous thrombosis. The MEGA study included 5000 patients with a first episode of deep VT or pulmonary embolism and 5000 controls. Questionnaires and DNA samples were taken for every participant and blood was collected in around half of patients and controls. Total PS (tPS) was measured in plasma by ELISA. For analysis, vitamin K users were excluded yielding 2198 patients and 2904 controls. The mean tPS (±SD) was 102.6 U/dL (±19.1) in patients and 100.8 U/dL (±19.7) in controls. OR and 95% confidence interval were computed to estimate the relative risk of thrombosis. PS was not a risk factor for thrombosis in the MEGA study when deciles were used to calculate OR neither when lowest 2.5% or 1% PS levels was examined (data not shown). We then randomly analyzed DNA samples from 2270 participants out of 10000 individuals (1395 patients and 875 controls) by MLPA to look for copy number variations (CNVs). In only one individual an abnormal MLPA pattern was found and the test was repeated confirming results. This individual, a female patient of 66 years who was not using estrogens, was heterozygous for a deletion of the complete PROS1 gene. Her tPS was 64.3 U/dL, which was below the 2.5th percentile (tPS = 65.3 U/dL) of the total MEGA control group. To check if the frequency of CNVs in PROS1 was indeed low in the MEGA, we zoomed in on PS deficiency by selecting DNA from individuals with low tPS (cuttof: mean tPS -2SD of the total MEGA control group), not excluding vitamin K users. This selection yielded 404 individuals, out of which 198 had been previously analyzed in the first round. MLPA was performed for the remaining 206 individuals and no new CNV was found in PROS1. In conclusion, PS deficiency was not a risk factor for VT in this population-based study and deletions or duplications in PROS1 were not common. This is in contrast with previous data from family-based studies but in accordance with the fact that the genetic basis of PS deficiency and its risk for VT outside of a family setting was also not evident in other studies.