Solid and gaseous cerebral microembolization after biologic and mechanical aortic valve replacement: investigation...

Cerebral microembolization is a well-recognized phenomenon after cardiac valve replacement, but the relative proportion of solid and gaseous emboli is uncertain. Particulate microemboli are thought to be the most damaging. With the use of multifrequency transcranial Doppler ultrasound, we compared the number and nature of microemboli in recipients of biologic and mechanical aortic valve prostheses.

The middle cerebral arteries of 60 patients were monitored bilaterally with a new-generation transcranial Doppler ultrasound (Embo-Dop, DWL Elektronische Systeme GmbH, Singen, Germany) that rejects artefacts online and automatically discriminates between solid and gaseous microemboli. All recordings were performed during a 30-minute period 1 day before and at a mean of 5 days and 3 months after isolated aortic valve replacement with a biologic (30, group B) or mechanical (30, group M) prosthesis.

The patients in group B were older, with a mean age of 70.6 +/- 9.7 years versus 55.4 +/- 9.4 years (P < .005) in the patients in group M. Biologic prosthesis recipients were all taking aspirin (no warfarin); patients with mechanical valves were well anticoagulated with warfarin both 5 days and 3 months after surgery. None of the patients had solid microemboli preoperatively. Five days postoperatively, the absolute number of cerebral microemboli was 145 and 594 for total microemboli (P = .001) and 41 and 182 for solid microemboli (P = .002) in groups B and M, respectively. At 3 months, the absolute number was 65 and 608 for total microemboli (P < .001) and 10 and 188 for solid microemboli (P < .001) in groups B and M, respectively. Solid microemboli accounted for 16% of the total microembolic load in group B compared with 31% in group M (P = .05) at 3 months.

Solid cerebral microemboli represent approximately one third of the total cerebral microembolic load after mechanical aortic valve replacement and are detectable in the majority of such patients both 5 days and 3 months after surgery. The neurofunctional consequences of this phenomenon should be carefully assessed.