Qualitative and semiquantitative [18F]FDG-PET/CT analysis in patients with IE: a strategy to enhance specificity?

Abstract

Abstract Aim The 2015 ESC Guideline for the diagnosis and management of infectious endocarditis (IE) included molecular imaging procedures such as radiolabeled leukocytes and [18F]FDG-PET/CT as major criteria to establish the diagnosis of IE. Whereas in case of radiolabeled leukocytes the criteria for defining positive and negative scan are well established, in case of [18F]FDG-PET/CT a consensus of the criteria that define a positive scan is still lacking. Materials and methods In this work we prospectively evaluated a series of 90 patients (M:F=60:30, mean age 67±16 years, median age 72 years, range 18–88) that performed [18F]FDG-PET/CT for suspected IE between January 2016 and January 2019 in our center to define which pattern of uptake and semiquantitative parameters (SUVmax, SUVmean, valve/liver, valve/lung and valve/mediastinum uptake ratio) present the highest diagnostic accuracy for IE. PET/CT results were correlated with transthoracic (TTE) or transesophageal (TEE) echocardiography, blood culture, the Duke criteria. The new ESC criteria were also evaluated. Results Globally PET/CT presented sensitivity=96% and specificity=75%, NPV= 97% and PPV=68%. A focal pattern of uptake was found 17/30 patients with confirmed IE (mean SUVmax=5.29±1.8, median SUVmax=5.5) whereas whereas a diffuse pattern of uptake was found in 11 patients with confirmed IE (mean SUVmax=8.1±3.5, median SUVmax=7.6). No significant differences between the valve/liver ratio uptake and the valve/mediastinum uptake ratio were found between positive and negative patients with either focal or diffuse pattern of uptake. Interesting, in presence of diffuse valve uptake both SUVmax and the valve/lung uptake ratio were significantly higher in patients with a final diagnosis of IE, including the group of patients under antimicrobial treatment (>90%). PET/CT confirmed its ability to identify extracardiac sites of [18F]FDG uptake in presence of septic embolism (35%) and in patients with other diseases (vasculitis, vascular prosthesis infections, mediastinitis and cancer, particularly in patients with IE sustained by Streptococcus Gallolyticus). Conclusions Our results suggest pattern of [18F]FDG uptake more specific for valve infection is a focal uptake or a diffuse uptake with a SUVmax>5 and a valve/lung ratio >8. Funding Acknowledgement Type of funding source: None