Clin Res Cardiol (2021)
DOI DOI https://doi.org/10.1007/s00392-021-01843-w

Diagnostic value of the modified Duke criteria in suspected infective endocarditis – the PRO-ENDOCARDITIS study
I. Mahmoud1, I. Dykun1, M. Totzeck1, C. Rischpler2, A. Ruhparwar3, K. Herrmann2, T. Rassaf1, A.-A. Mahabadi1
1Klinik für Kardiologie und Angiologie, Universitätsklinikum Essen, Essen; 2Klinik für Nuklearmedizin, Universitätsklinikum Essen, Essen; 3Klinik für Thorax- und Kardiovaskuläre Chirurgie, Universitätsklinikum Essen, Essen;

Background: 18F-FDG Positron emission tomography (PET) is increasingly performed in patients with suspected infective endocarditis, as it enables detection of areas with increased glucose metabolism that may reflect inflammatory foci. We aimed to determine the clinical characteristics and diagnostic accuracy of PET in patients with suspected infective endocarditis.

Methods: This analysis is based on the cohort of the PRO-ENDOCARDITIS study, a prospective single center observational study on the diagnostic accuracy of clinical pathways for detection of infective endocarditis. All PET examinations were performed as by clinical indication. A clinical event committee adjudicated presence of infective endocarditis. Fischer’s Exact test was used to determine differences in clinical characteristics in patients with and without PET imaging, while the predictive ability was assessed in receiver operating characteristics.

Results: From 261 patients (aged 60.1±16.1 years, 62.8% male) with suspected IE, included in the study, 45 patients (17.2%) received PET imaging. PET was more frequently performed in patients with inconclusive transesophageal echocardiography (29.0%), less frequently when echocardiography was considered typical for infective endocarditis (18.8%) and only rarely in patients without any signs of infective endocarditis in transesophageal echocardiography (14.6%, p=0.035). PET was more frequently performed in patients with prosthetic valves (26.5% vs. 15.9%, p=0.10) and implanted devices (16.2% vs. 26.9%, p=0.16), however, not reaching statistical significance. Overall, positive PET results reached a sensitivity of only 33% with a specificity of 90% as compared to the adjudicated endpoint. The sensitivity improved to 60%, when only including patients with prosthetic valves or implanted cardiac devices. In receiver operating characteristics, addition of PET to the Duke criteria led to a marginal improvement of the prediction of infective endocarditis (0.912 to 0.924, p=0.24).

Conclusion: In patients with suspected infective endocarditis, PET is predominantly performed in patients with inconclusive transesophageal echocardiography. However, negative PET results do not rule out infective endocarditis especially in patients without prosthetic valves or cardiac devices.


https://dgk.org/kongress_programme/jt2021/aP854.html