Fdg-Pet/Ct For The Assessment Of Sternal Wound Infection Follo Ng Sternotomy

223 Background: Sternal wound infection (SWI) is a serious complication occasionally seen following cardiac surgery; however, data on the diagnostic accuracy of imaging tests for sternal infections are scarce. FDG-PET/CT is an emerging technique for the diagnosis of soft-tissue and osseous infections which has the potential to play an important role in the diagnosis of SWI. The purpose of this study was to analyze the diagnostic accuracy of FDG-PET/CT for SWI in patients following median sternotomy. Methods: We performed a single-center, retrospective analysis of patients who had undergone median sternotomy and FDG-PET/CT imaging. The gold standard consisted of positive bacterial culture and/or the presence of purulent material at surgery. All studies underwent blinded consensus interpretation. Qualitative patterns of sternal FDG uptake, SUVmax, and associated CT findings were determined for all studies. Patterns of FDG uptake examined included: diffuse low-grade, diffuse high-grade, focal uptake, sternal wire uptake, and soft-tissue extension. Receiver operator characteristic (ROC) analysis was used to calculate the area under the curve (AUC) for each of these imaging features. An imaging scoring system was developed to study the impact on AUC when multiple imaging features were present. The diagnostic performances were studied in both the acute phase (≤6 months between sternotomy and imaging) and chronic phase (>6 months). Results: A total of 40 subjects were identified who had undergone median sternotomy and subsequent imaging with FDG-PET/CT, with 11 confirmed SWI cases. Subjects underwent PET/CT imaging an average of 32.6 ± 60.4 months (range: 8 days to 21 years) following sternotomy. Consensus interpretation was associated with a sensitivity/specificity of 91%/97% (95%CI 58.7-99.8%/82.2-99.9%). Overall, SUVmax at the sternotomy site was higher in infected subjects compared to non-infected subjects (8.5 ± 1.4 vs 5.5 ± 3.8, p = 0.02); however, in the acute subgroup, the difference was not significant (8.4 ± 1.6 vs 7.8 ± 4.4, p = 0.72). With the exception of diffuse uptake, all patterns were associated with high AUCs (sternal wire uptake 0.81; soft-tissue extension 0.86; focal uptake 0.88) while combinations of patterns increased AUC to 0.96. Soft-tissue extension was the strongest predictor of SWI with an odd ratio of 143. Diagnosis based on SUVmax was more accurate in the chronic rather than acute setting (AUC: 0.95 vs 0.66, respectively). Conclusions: This study suggests that interpretation based on imaging patterns yields excellent sensitivity without sacrificing specificity, regardless of the timing between sternotomy and imaging. While diagnosis of SWI based on overall uptake is accurate when imaging is performed >6 months after surgery, the results of this study suggest that infections should be diagnosed on the basis of uptake patterns as the use of SUV cutoffs has limited specifity when time interval between imaging and surgery is ≤6 months.