|Home | About | Journals | Submit | Contact Us | Français|
To reduce the incidence of thromboembolism, a cloth-covered version of the Starr—Edwards aortic valve was introduced in 1968 (Figure 1)1. A new complication then arose—systemic embolism due to cloth tears1,2,3.
A woman aged 76 reported progressive memory loss over the past year. The onset had been insidious with initial loss of short-term memory followed by difficulties in social interaction and visual hallucinations. In 1971 she had had an aortic valve replacement with a cloth-covered Starr—Edwards prosthesis for rheumatic disease. There was no history of stroke or transient ischaemic attacks. On examination her abbreviated mental test score was 1/10; she had expressive dysphasia but no other focal neurodeficit. She was normotensive and in controlled atrial fibrillation. Prosthetic heart sounds were heard along with an ejection systolic murmur.
Blood tests were unremarkable. Her records showed that, since the valve replacement, she had almost always been adequately anticoagulated. Electrocardiography confirmed atrial fibrillation. Computed tomography of the brain showed a degree of cerebral atrophy. Transthoracic echocardiography showed subnormal left ventricular function with a dilated left ventricle and reduced septal movement. There were no thrombi or vegetations. On transoesophageal echocardiography the prosthetic valve appeared normal and no thrombus was seen in the left atrium. Magnetic resonance imaging was not possible because of her prosthetic valve. Cerebral perfusion scintigraphy and single-photon emission tomography showed a defect in the left frontoparietal cortex and thinning in the occipitoparietal perfusion consistent with vascular dementia. Several falls and infections punctuated her clinical course and she died suddenly in a nursing home.
At necropsy there were multiple areas of microcystic degeneration around the lateral ventricles and one larger (1 cm) area of infarction in the left cerebellar hemisphere. There were mild atherosclerotic changes in the aorta and carotid arteries. On examination of the Starr—Edwards valve in situ, the cloth covering the struts was seen to be frayed and broken down, hanging from the cage. There were no vegetations on the valve, and the atria and its appendages were clear. Brain tissue was not examined microscopically.
Complications related to the ball-in-cage heart valves include infective endocarditis, haemolytic anaemia, systemic arterial embolization, thrombotic occlusion of the prosthesis and paraprosthetic leak2. The one hazard unique to the cloth-covered prosthetic valve is wear of the cloth covering the struts, leading to systemic cloth fibre emboli1,2,3 and haemolytic anaemia4.
To the best of our knowledge, there have been no previously reported cases of dementia associated with a damaged prosthetic heart valve. Though we did not look for microscopic evidence of material from the valve within the multiple infarcts, the absence of vegetations, of clots and of severe atherosclerosis makes the damaged prosthetic valve the most likely source of embolic material. Hypertension and diabetes being absent, small vessel disease causing multiple lacunar infarcts is unlikely.
In patients with cloth-covered prosthetic valves, cloth wear should be suspected as a cause of recurrent embolic events when anticoagulation has been adequate3. Exclusion of other sources of systemic emboli and infective endocarditis is necessary. Transthoracic and transoesophageal echocardiography may be useful in detecting cloth tears, which have to be differentiated from vegetations and thrombi5,6. Some patients with sequelae of cloth wear have benefited from replacement with porcine xenografts2.
We thank Dr Mike Davis for help with the patient and Dr P J S Dunn for conducting the necropsy.