The value of 18F-FDG PET/CT for the diagnosis of device related infections in patients with a left ventricular assist device - a dual centre study

Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): Partnership UMCG Siemens for building the future of Health (PUSH) Background Left ventricular assist devices (LVADs) are increasingly used for the treatment of end-stage heart failure. LVADs improve quality of life and long-term survival, but device specific infections remain cumbersome. These infections can lead to life threatening complications and are difficult to diagnose with conventional radiological imaging. In this study, FDG-PET/CT accuracy for the diagnosis of LVAD specific infections was evaluated with specific focus on potential confounders and the additive value of semi-quantitative analysis. Materials/methods All LVAD recipients with an available FDG-PET/CT from September 2013 until August 2020 in two LVAD centres in the Netherlands, with a suspicion of driveline and/or central device infection were included. FDG-PET/CT was performed according to European Association of Nuclear Medicine (EANM) guidelines. Potential confounders were documented and assessors were blinded to the clinical context of included patients. Assessment of FDG-PET/CT was performed visually and semi-quantitatively with 6 regions of interest alongside the driveline and central device components. The final clinical diagnosis of either driveline infection or central device infection, based on findings during surgical intervention or multidisciplinary consensus, was used as the reference for diagnosis. Results In total 38 patients (average age 55 years, 84% males) were evaluated for a total of 55 episodes of suspected device-specific infection. Clinical evaluation established driveline infection in 28 cases, central device infection in 7 and combined infection in 4. Visual analysis obtained  a sensitivity and specificity of 0.83 and 0.75 respectively, in differentiation between infected and non-infected drivelines. Visual analysis of FDG-PET/CT of central device components showed excellent sensitivity: 1.0, but suffered from poor specificity: 0.23.  Semi-quantitative analysis using a SUVmax was comparable to visual analysis for establishing driveline infections, with a sensitivity and specificity of 0.80 and 0.84 respectively, while for central device infections, semi-quantitative analysis using a SUVratio outperformed visual analysis, with a sensitivity and specificity reaching 0.88 and 0.90 respectively. Conclusions FDG-PET/CT is a valuable tool for the assessment of device-specific infections in LVAD recipients. Semi-quantitative analysis can significantly increase diagnostic accuracy of FDG-PET/CT for the analysis of the central device components and should be considered in cases where the diagnosis cannot be rejected based on visual analysis.