Mean Platelet Volume as a Risk Factor for Deep Vein Thrombosis In An Ethnically Diverse Population

Abstract 4211

Platelet size is thought to reflect reactivity; Mean platelet volume (MPV) was recently reported as a possible predictor for VTE, but it is not clear whether ethnic origin impacts on this risk factor. King's serves an ethnically diverse community and to assess whether MPV is a predictor of VTE in our population, we conducted a retrospective analysis of consecutive patients referred to our DVT service between January 2007 and October 2009. Patients with a confirmed first lower limb DVT (provoked n=153, unprovoked n=110) were included as subjects and controls (n=151) were derived from consecutive patients referred with objective exclusion of a DVT and no previous history of VTE, active cancer or surgery in the previous 6 weeks. All patients had a full blood count at presentation analysed on an automated analyser (using optical light scatter for MPV) within 4 hours of collection. There was no difference in mean age (54.7 vs 54.8), smoking status or ethnic group (51% vs 52.3% white, 38.4% vs 33.8% black and 10.6% vs 13.9% other) between subjects and controls respectively. Males accounted for 47.5% of subjects and 27.2% of controls. DVTs were unprovoked in 41.8% with 13.7% associated with surgery, 7.6% cancer, 10.6% pregnancy or hormone therapy. The remainder (25.1%) were secondary to cast, trauma, immobilisation or travel. Mean MPV was significantly higher in subjects than controls (8.17 vs 7.79, p=0.001) with a more marked difference in those with unprovoked DVT compared with controls (8.28 vs 7.79, p<0.001). The platelet count was lower in the DVT group (median, range 270, 21–812 vs 293, 31–642 p=0.027), with a more marked difference in those with unprovoked DVT (median, range 250, 21–584 vs 293, 31–642 p<0.001). Relative risk associated with MPV > 9.18 (90^th centile) was 1.26 (95% CI 1.08– 4.76, p=0.01) and increased to 1.59 (1.18-2.1, p=0.008) in those with unprovoked DVT. Relative risk associated with platelet count <210 (10^th centile) was 1.21 (1.02-1.43, p=0.06) and increased with unprovoked DVT to 1.70 (1.3-2.2, p=0.002). An inverse correlation between MPV and platelet count was confirmed (-0.305, p<0.001). Logistic regression was undertaken to investigate effect of MPV, platelet count, age and smoking status. MPV was the only significant risk factor for DVT with odds ratio 1.39 (1.14-1.68). For unprovoked DVT, both MPV and platelet count contributed to risk with odds ratio of 1.36 (1.06-1.74, p=0.015) and 0.997 (0.994-1.0, p=0.037) respectively. Further analysis was undertaken to compare MPV in white (provoked 84, unprovoked 50, controls 79) and black (provoked 55, unprovoked 46, controls 51) subgroups. There was no difference in mean age between white and black subjects or controls. Interestingly, in the black subgroup 73.9% of males had an unprovoked DVT compared with 26.1% of females. This gender difference was not seen in the white subgroup (unprovoked 37.9% males, 36.8% females) and was not explained by the presence of pregnancy or hormone use (18 vs 18.4% black vs white females). There was no significant difference in MPV or platelet count between white and black subjects or white and black controls. There remained a significant difference between white subjects and white controls mean MPV (8.1 vs 7.7, p=0.014) accentuated in the unprovoked subgroup (8.3 vs7.7, P=0.007); median platelet count was only significantly lower for unprovoked DVT compared to controls (251.5, 21–509 vs 285, 31–687, p=0.02). MPV was also significantly higher in black subjects compared to controls (8.3 vs 7.8, p=0.011), and platelet count was significantly lower (256, 129–811 vs 293, 138–642 p=0.032). MPV was no different between unprovoked DVTs and controls, however the effect of platelet count was accentuated (244.5, 167–584 vs 293, 138–642 p<.001). Logistic regression confirmed male gender as the only predictive factor for unprovoked VTE in the black subgroup (OR 5.8, 95% CI 2.36–14, p<0.001); neither MPV nor platelet count contributed to DVT risk. Limitations include the retrospective nature of the study, number of subjects, unavailable body mass indices and the discrepant gender distribution between controls and subjects. In summary, MPV is a risk factor for DVT in both white and black populations, though this link appears to hold true for unprovoked DVT in white populations only.