L Raco, I Sim*, DJ Wheatley, PR Belcher

University of Glasgow, Departments of Cardiac Surgery and Medicine and & Therapeutics*.

Objectives: To assess the thromboembolic potential of prosthetic heart valves, by measurement of platelet aggregation, coagulation variables and Doppler arterial high-intensity signals (HITS).

Methods: 25 juvenile sheep underwent mitral valve replacement with a mechanical or biological prosthesis, respectively anticipated to possess high and low thrombogenicity. 20 survivors were observed at 1, 3 and 6 months for:

1. Platelet aggregates: single platelet counting in EDTA (for disaggregation of platelet aggregates) and EDTA-Formalin (to capture platelet aggregates) and in blood smears, before and after intravenous administration of the antiplatelet agent ICI 170809.

2. Coagulation Markers: fibrinogen, von Willebrand Factor (vWF), thrombin-antithrombin III (TAT).

3. Circulating Microemboli: recording of HITS in the carotid arteries by Doppler Ultrasound. The number of HITS was recorded before and after the infusion of ICI 170809. Our method enabled us to obtain reliable and reproducible data without general anaesthesia.

4. Macroscopic embolization: necropsy at 6 months.

5.The biological valves were X-rayed.

6. The biological valves were also tested in a pulse duplicator.

Results: All animals retained sinus rhythm. Anticoagulation was not used. There were two valve-related and three other deaths.

1. There were always more circulating platelet aggregates with mechanical valves than with bioprostheses (p<0.01). ICI 170809 consistently reduced these in both valve types (p<0.025).

2. Coagulation Markers: Fibrinogen and TAT did not change. vWF fell with both valves (p<0.025).

3. Circulating Microemboli: HITS were more frequent at 3 (p<0.044) and 6 months (p<0.006) in mechanical valves when compared with biological valves, but were not altered by ICI 170809.

4. Necropsy: multiple renal infarcts were seen in 4 of the 6 biological valve survivors, but not in the 9 with mechanical prostheses (p<0.006).

5. X-ray of the biological valves: showed calcification and tissue overgrowth on the valve leaflets.

6. Pulse duplicator: showed varying valve degeneration and malfunction of the biological prostheses.

Conclusions: HITS neither correlated with number of circulating platelet aggregates nor occurrence of peripheral embolism. Platelet aggregate numbers did not correlate with occurrence of peripheral embolism. Surprisingly, in this juvenile sheep model, the bioprostheses was associated with a higher incidence of peripheral tissue damage due to thromboembolism, probably due to the accelerated degeneration of biological prostheses in this animal model.