Background: Infective endocarditis (IE) remains a difficult to diagnose condition associated with high mortality. ¹⁸ F-fluorodeoxyglucose positron emission tomography/computed tomography ( ¹⁸ F-FDG PET/CT) has recently emerged as another IE imaging modality, although diagnostic accuracy varies across observational studies and types of IE. This meta-analysis assessed the diagnostic performance of ¹⁸ F-FDG PET/CT for IE and its subtypes. Methods: We searched Pubmed, Cochrane, and Embase from January 1980 to September 2019 for studies reporting both sensitivity and specificity of ¹⁸ F-FDG PET/CT for IE. Meta-Disc 1.4 was used to pool data for all cases of IE and its subgroups of native valve IE, prosthetic valve IE, and cardiac implantable electronic devices IE. Results: We screened 2566 records from the search, assessed 52 full-text articles, and included 26 studies totaling 1358 patients (509 IE cases). Pooled sensitivity and specificity (95% CI, inconsistency I-square statistic) were 0.74 (0.70–0.77, 71.5%) and 0.88 (0.86–0.91, 78.5%) for all cases of endocarditis. Corresponding parameters for native valve IE were sensitivity 0.31 (0.21–0.41, 29.4%) and specificity 0.98 (0.95–0.99, 34.4%); for prosthetic valve IE: sensitivity 0.86 (0.81–0.89, 60.0%) and specificity 0.84 (0.79–0.88, 75.2%); and for cardiac implantable electronic devices IE: sensitivity 0.72 (0.61–0.81, 76.2%) and specificity 0.83 (0.75–0.89, 83.6%). Pooled sensitivities and specificities were higher for the 17 studies since 2015 than the 9 studies published before 2015. Conclusions: ¹⁸ F-FDG PET/CT had high specificity for all IE subtypes; however, sensitivity was markedly lower for native valve IE than prosthetic valve IE and cardiac implantable electronic devices IE. It is, therefore, a useful adjunct modality for assessing endocarditis, especially in the challenging scenarios of prosthetic valve IE and cardiac implantable electronic devices IE, with improving performance over time, related to advances in ¹⁸ F-FDG PET/CT techniques.

Background: The growing awareness of cardiovascular toxicity from cancer therapies has led to the emerging field of cardio‐oncology, which centers on preventing, detecting, and treating patients with cardiac dysfunction before, during, or after cancer treatment. Early detection and prevention of cancer therapy–related cardiac dysfunction (CTRCD) play important roles in precision cardio‐oncology. Methods and Results: This retrospective study included 4309 cancer patients between 1997 and 2018 whose laboratory tests and cardiovascular echocardiographic variables were collected from the Cleveland Clinic institutional electronic medical record database (Epic Systems). Among these patients, 1560 (36%) were diagnosed with at least 1 type of CTRCD, and 838 (19%) developed CTRCD after cancer therapy (de novo). We posited that machine learning algorithms can be implemented to predict CTRCDs in cancer patients according to clinically relevant variables. Classification models were trained and evaluated for 6 types of cardiovascular outcomes, including coronary artery disease (area under the receiver operating characteristic curve [AUROC], 0.821; 95% CI, 0.815–0.826), atrial fibrillation (AUROC, 0.787; 95% CI, 0.782–0.792), heart failure (AUROC, 0.882; 95% CI, 0.878–0.887), stroke (AUROC, 0.660; 95% CI, 0.650–0.670), myocardial infarction (AUROC, 0.807; 95% CI, 0.799–0.816), and de novo CTRCD (AUROC, 0.802; 95% CI, 0.797–0.807). Model generalizability was further confirmed using time‐split data. Model inspection revealed several clinically relevant variables significantly associated with CTRCDs, including age, hypertension, glucose levels, left ventricular ejection fraction, creatinine, and aspartate aminotransferase levels. Conclusions: This study suggests that machine learning approaches offer powerful tools for cardiac risk stratification in oncology patients by utilizing large‐scale, longitudinal patient data from healthcare systems.