1

Introduction

Percutaneous aortic valve replacement (PAVR) is a new approach for aortic valve disease. This revolutionary technology has brought hope for patients with severe aortic stenosis who require valve replacement but are at too high a risk of complication to undergo conventional thoracotomy. The first percutaneous aortic valve replacement was performed by Cribier in 2002 . The initial results have been promising but the procedure is not free of complications. The most common are procedural and device related . The incidence of infective endocarditis (IE) with surgical aortic valve replacement is reported to be 13.3% [ ]; percutaneous aortic valve replacement being a developing modality has yet to report in sufficient numbers. We present what we believe to be the first reported case of PAVR endocarditis according to Duke’s criteria in a patient presenting 2 months post procedure.

2

Case

A 64-year-old lady presented to the emergency assessment unit in October 2005 with gradual onset shortness of breath, palpitations and signs of left ventricular failure. She had a 20-year history of myasthenia gravis and had undergone three thoracotomies for resection of thymoma and three episodes of radiotherapy to her chest during that period. Her maintenance medication included azathioprine and low dose steroids since the diagnosis of myasthenia gravis.

On examination, she was tachycardic, with a blood pressure of 110/60 and had an ejection systolic murmur in keeping with the diagnosis of aortic stenosis. Her electrocardiogram showed atrial tachycardia and echocardiogram showed mixed aortic valve disease with the gradient of 49 mmHg and mild aortic regurgitation. Her left ventricular function was mildly impaired. She was treated successfully medically and reverted to normal sinus rhythm. However over the next twelve months she continued to suffer from breathlessness and chest pain despite optimal medical therapy. Further investigation confirmed severe ostial coronary artery disease (95% left main stem, 85% right coronary artery (RCA)) — probably related to her radiotherapy — and important mixed aortic valve disease with dominant stenosis but at least moderate aortic regurgitation. Computed tomography scanning also demonstrated that her superior vena cava, brachiocephalic and subclavian vein were blocked and descending aorta was adherent to the back of the sternum. Cardiothoracic surgical opinion was that she was not suitable for thoracotomy due to lack of venous access for bypass and extensive scarring of her chest wall resulting from previous surgery and radiotherapy as described above. Therefore, an alternative nonsurgical approach to relieve her cardiac symptoms was sought. As a first procedure, she had percutaneous stenting to her ostial coronary artery disease (both right and left) deploying bare metal stents. One month later, in 2007, she had successful percutaneous aortic valve replacement using a Core valve system, with significant improvement in her clinical condition. She was not given prophylactic antibiotics. However, 2 months after intervention, she presented to her general practitioner with fever and general malaise; she was treated empirically with amoxicillin for an upper respiratory tract infection without prior blood cultures. She continued to be pyrexial with rigors at night with malaise after finishing a course of amoxicillin. There was no precipitating factor such as dental or invasive procedures. At hospital review, she was apyrexial with no peripheral stigmata of IE. Examination did not reveal any focus of infection and no new cardiac murmurs were detected. Her C-reactive protein (CRP) was elevated at 160 mg/l and the white cell count (WCC) was 18.4 10 ^ ⁹ /l with neutrophilia of 16.4 10 ^ ⁹ /l. She was anaemic [Hb=10.4g/dl (normocytic normochromic and she had dipstick positive haematuria)].The 12-lead electrocardiogram showed a normal PR interval, and chest X-ray showed clear lung fields and the outline of aortic valve replacement (AVR) strut. Neither transthoracic nor transoesphageal echocardiography showed evidence of vegetations, although there was echo-free space within the wall of the ascending aorta where the stents of the core valve were seen. The aortic valve was well sited and functioning well. Several sets of blood cultures were taken to investigate for possible IE, and she was admitted to hospital for assessment and observation.

Two days after her admission, Gram-negative rods in pairs were found to be growing in multiple sets of blood cultures.

Cultures yielded an oxidase-positive, catalase-positive, Gram-negative bacillus which gave a positive tributyrin test and produced B-lactamase. These tests suggested a provisional identification of Moraxella sp., and this was supported biochemically by API NH (bioMerieux), Further typing of the isolate to species level was undertaken by the Laboratory of Healthcare associated Infection, of the Health Protection Agency (London), who confirmed the organism to be Moraxella nonliquefaciens .

The patient was treated with high dose intravenous ceftriaxone (2 g bd for 48 h and then 2 g once daily) on the advice of the consultant microbiologist, who was able to confirm that the M. nonliquefaciens was fully susceptible to this antibiotic. Due to lack of central venous access, a femoral Hickman line was inserted for administration of antibiotic. The patient showed remarkable improvement symptomatically and her inflammatory markers settled (CRP=30 mg/l white blood count (WBC)=8.2 10 ^ ⁹ /l. It was considered important to aim for 6 weeks of intravenous antibiotics but at 4 1/2 weeks her inflammatory markers started to rise. The Hickman line site was infected and so removed. She was changed to oral ciprofloxacin to which the isolate was also susceptible. She remains clinically well 3 years later.