18F-FDG PET/CT In The Characterization And Management Of Continuous Flow Left Ventricular Assist Device Infection

Introduction Implantable continuous flow left ventricular assist devices (LVADs) are now more prevalent in end-stage heart failure treatment both as a bridge-to-transplant and destination therapy. With increasing use of LVADs, device-related infections are a common complication and are linked to worse outcomes yet diagnosis remains challenging. Implementation of 18F-fluorodeoxyglucose (18F-FDG) PET/CT as a diagnostic modality is emerging as an important tool in this challenging population. Hypothesis The purpose of the study was to describe our experience with 18F- FDGPET/CT in diagnosing and characterizing LVAD related infections. Methods This was a single-center case series of 36 LVAD recipients who underwent 18F- FDG PET/CT scans due to suspected device infection. Follow-up was obtained for up to 3 years. PET/CT was analyzed qualitatively (presence/absence of uptake) and quantitatively (SUVmax) in four LVAD components (driveline entry, driveline subcutaneous pathway, pump pocket and outflow tract). Data on the following outcomes was collected: anti-microbial course, surgical exploration/debridement, pump exchange or transplant. Results Mean age of the patients was 56 ± 13.7 years; 39% were female. Patients were divided into three clinical groups: LVAD-specific infection with positive blood cultures (Group 1), LVAD-specific infection with positive wound cultures but negative blood cultures (Group 2), or pain/fever of unknown origin with negative blood cultures and swab (Group 3). 30% of patients were Group 1, 61% of patients were Group 2, and 8% of patients were Group 3. 18F- FDG PET/CT showed various patterns of involvement of the four LVAD components (Figure 1). The highest average SUVmax of patients in category 1 was 9.76±4.23 and seen in patients with a pump pocket infection, in category 2 was 8.1±3.4, seen in driveline entry infection and in category 3 was 7.39±.5.2 seen in subcutaneous driveline path. Patients in category 1 were more likely to exhibit a pump pocket infection. Driveline smears revealed Staphylococcus or Pseudomonas strains as the underlying pathogen in most cases (32% and 23% respectively, with the remainder of patients demonstrated E.coli, Corneybacterium, Serratia, and fungal organisms). Pocket infection was associated with pseudomonal infection while limited infection of the driveline was associated to MSSA. Post PET/CT, 8 patients were transplanted, 3 had a pump exchange, and 3 patients had debridement or other forms of surgical exploration, while the remainder had the course of anti-microbial therapy altered. Conclusions 18F- FDG PET/CT is a useful technique for establishing the site, pattern and extent of the LVAD infection, which has critical implications for management and patient outcome. Implantable continuous flow left ventricular assist devices (LVADs) are now more prevalent in end-stage heart failure treatment both as a bridge-to-transplant and destination therapy. With increasing use of LVADs, device-related infections are a common complication and are linked to worse outcomes yet diagnosis remains challenging. Implementation of 18F-fluorodeoxyglucose (18F-FDG) PET/CT as a diagnostic modality is emerging as an important tool in this challenging population. The purpose of the study was to describe our experience with 18F- FDGPET/CT in diagnosing and characterizing LVAD related infections. This was a single-center case series of 36 LVAD recipients who underwent 18F- FDG PET/CT scans due to suspected device infection. Follow-up was obtained for up to 3 years. PET/CT was analyzed qualitatively (presence/absence of uptake) and quantitatively (SUVmax) in four LVAD components (driveline entry, driveline subcutaneous pathway, pump pocket and outflow tract). Data on the following outcomes was collected: anti-microbial course, surgical exploration/debridement, pump exchange or transplant. Mean age of the patients was 56 ± 13.7 years; 39% were female. Patients were divided into three clinical groups: LVAD-specific infection with positive blood cultures (Group 1), LVAD-specific infection with positive wound cultures but negative blood cultures (Group 2), or pain/fever of unknown origin with negative blood cultures and swab (Group 3). 30% of patients were Group 1, 61% of patients were Group 2, and 8% of patients were Group 3. 18F- FDG PET/CT showed various patterns of involvement of the four LVAD components (Figure 1). The highest average SUVmax of patients in category 1 was 9.76±4.23 and seen in patients with a pump pocket infection, in category 2 was 8.1±3.4, seen in driveline entry infection and in category 3 was 7.39±.5.2 seen in subcutaneous driveline path. Patients in category 1 were more likely to exhibit a pump pocket infection. Driveline smears revealed Staphylococcus or Pseudomonas strains as the underlying pathogen in most cases (32% and 23% respectively, with the remainder of patients demonstrated E.coli, Corneybacterium, Serratia, and fungal organisms). Pocket infection was associated with pseudomonal infection while limited infection of the driveline was associated to MSSA. Post PET/CT, 8 patients were transplanted, 3 had a pump exchange, and 3 patients had debridement or other forms of surgical exploration, while the remainder had the course of anti-microbial therapy altered. 18F- FDG PET/CT is a useful technique for establishing the site, pattern and extent of the LVAD infection, which has critical implications for management and patient outcome.