Cardiac CT in native valve endocarditis

Clinical History

A 54-year-old man had aortic valve endocarditis diagnosed on trans-thoracic echocardiography. He failed to settle on antibiotics with aortic root abscess suspected. He refused to undergo trans-oesophageal echo so a 64-slice cardiac CT examination was performed with excellent imaging of the vegetation and no abscess detected.

Imaging Findings

A 54-year-old man presented with embolic left homonomous hemianopia. On examination, he had an aortic diastolic murmur with elevated white cell count, C-reactive protein and temperature. Trans-thoracic echocardiogram (TTE) confirmed a vegetation on the aortic valve (Fig. 1. a, b) along with aortic regurgitation (Fig. 2). Blood cultures grew streptococcus sanguis. He was commenced on intravenous benzylpenicillin and gentamicin but temperature continued to spike and inflammatory markers remained elevated. An abscess of the aortic valve root was suspected but the patient refused to have an endoscopic trans-oesophageal echo (TOE) to confirm.

He did agree to undergo 64-multidetector cardiac CT (64-MDCT) examination. His heart rate was brought down to <64 bpm with 10mg iv metoprolol. In order to see fine detail from a small vegetation or abscess, a "stent" imaging protocol was used with thin collimated slices that were 0.8mm thick with 0.4mm overlap. Although heart rate was low, retrospective ECG gating was used rather than prospective ECG triggering in case multiple R-R interval reconstructions or cine-CT was required.

The study demonstrated a vegetation located beneath the non-coronary cusp of the aortic valve measuring 10 x 6mm (Figures 3,4 and 5). Images in diastole (75% R-R interval) showed failure of the aortic cusps to fully appose (Figures 4 and 5) causing aortic regurgitation. The aortic root was examined thoroughly with no abscess detected.

64-MDCT answered the clinical query in a situation where TOE is considered the gold standard investigation. Antibiotics were changed but after continued temperature spikes he underwent valve replacement operation.

Discussion

Peri-valvular abscess formation is a feared complication of infective endocarditis. It is associated with increased mortality and is an indication for surgical intervention. It can complicate up to 30% of cases of endocarditis. Clinically, it is suspected if antibiotics fail to suppress temperature and infective markers.

TTE is relatively poor at detecting abscess formation with a sensitivity of about 28%, TOE is superior with a value of 87% and is regarded as the investigation of choice. CT faces several challenges when put into this clinical role; although its spatial resolution is excellent, temporal resolution does not match echocardiography and pyrexic patients may have a fast heart rate!

A further problem is that functional information about valve flow cannot be obtained as easily as in echocardiography or cardiac magnetic resonance imaging but careful observation can often pick up useful clues such as failure of valve cusps to close together (Figures 4 and 5).

One significant advantage of CT is the ability to image the coronary arteries non-invasively since catheter angiography runs the risk of dislodging any vegetation attached to the aortic valve.

Due to the small size of many vegetations, a narrow collimation is advisable - we chose a pre-set stent protocol on our machine.

In a direct head-to-head comparison of CT with TOE in infective endocarditis, both correctly identified 96% of valves (26 out of 27 patients) with vegetations proven at surgery [1].

In terms of abscess detection, CT scored 100% (9 out of 9 patients proven at surgery) versus 89% (8 out of 9 patients proven at surgery) for TOE [1].

Thus CT is worthy of consideration in infective endocarditis if TOE is not possible but care must be taken to obtain high quality images

Differential Diagnosis List

Final Diagnosis

Native aortic valve endocarditis (no abscess)

URL:	https://eurorad.org/case/8217
DOI:	10.1594/EURORAD/CASE.8217
ISSN:	1563-4086