Our inability to predict thromboembolic events after prosthetic valve surgery.

Background and Aim of the Study: Thromboembolic and bleeding complications detract from outcome for patients with prosthetic heart valves. The study aim was to investigate whether measurement of coagulation activation markers and transcranial Doppler ultrasound microembolic signals (MES) could identify patients at subsequent higher risk of thromboembolism or bleeding events.
Methods: A total of 526 patients (mean age 66 years; 266 males, 260 females) who underwent elective valve replacement surgery was enrolled between April 1999 and October 2002. Clinical assessment and blood sampling for coagulation activation markers was performed preoperatively and at three and 12 months postoperatively. Transcranial Doppler MES were recorded in the first 144 patients. Status was reviewed between 21st April and 9th June 2005, with 99.4% follow up.
Results: Among patients, 62% had an aortic valve replaced, and mechanical valves constituted 60% of all implants. The mean follow up was 3.61 years; total follow up was 1,899.2 patient-years (pt-yr). In total, 115 patients died, while 61 experienced a total of 80 thromboembolic events: linearized event rates were 3.94% (mechanical valves) and 4.4% (bioprostheses). There was no difference between mitral and aortic implants, or among bileaflet, tilting-disc mechanical and porcine valves. Atrial fibrillation was not influential. Coagulation activation markers were not associated with thromboembolic events, except for an elevated von Willebrand factor (vWF), which was associated with a five-fold increase in embolic event rate. Fifty-one patients experienced 59 bleeding events; eight patients experienced multiple events. Linearized event rates were 3.37% (mechanical valves) and 2.49% (bioprostheses). The INR was suboptimal in 44-58% of patients. Transcranial Doppler MES were not associated with blood coagulation markers or thromboembolic events.
Conclusion: Coagulation activation markers (except vWF) and MES did not predict thromboembolic events in valve replacement patients. Thromboembolic and bleeding event rates for West of Scotland patients generally exceeded reported rates: suboptimal anticoagulation appeared common and most likely influenced thromboembolic and bleeding event rates more than any other factor.