PurposeSternal wound infections (SWI) are complications of sternotomy and can be divided into deep SWI (DSWI) and superficial SWI (SSWI). In recent years, the use of 18F-Fluorodeoxyglucose (FDG)-positron emission tomography (PET)/computed tomography (CT) in diagnosing infections and inflammation has expanded significantly, with a growing number of clinical indications. This study assesses FDG-PET/CT diagnostic role in DSWI detection, the evolution of FDG uptake intensity in patients without DSWI and the potential biomarkers for DSWIs prediction.MethodsWe conducted a single center prospective study of consecutive patients referred for suspected SWI post-median sternotomy. Gold standard diagnosis was established by chart review of clinical follow-up, surgical findings, and cultures. To characterize the time between sternotomy and imaging, participants were subsequently subdivided into recent (< 3 months) or remote surgery (≥ 3 months) groups.Results44 FDG-PET/CT scans, 12 (27%) of which had DSWI according to the gold standard, were collected and analyzed. 20 studies were assigned to the recent group, and 24 studies to the remote surgery group. Sensitivity and specificity of FDG-PET/CT for detection of DSWI were 67% and 66%, respectively and an accuracy of 66% was obtained. Positive and negative predictive values were 42% and 84%, respectively. The NPV was higher in the remote surgery group (100%) compared to the recent surgery group (73%). SUVmax of the median sternal wound was significantly higher in the DSWI (9.3 ± 2.3) than the non-DSWI group (7.1 ± 3.0) (p = 0.025). There was however significant overlap of SUVmax between the two groups. CRP, WBC counts, and PCT levels were not significantly different between the DSWI and non-DSWI groups (p ≥ 0.34).ConclusionFDG-PET/CT has modest sensitivity and specificity for the detection of DSWI post-sternotomy. FDG-PET/CT results must take into account time since surgery; when PET/CT is performed more than 3 months following surgery, a negative scan can exclude DSWI with a high level of certainty.
Read full abstract