Autosomal dominant polycystic kidney disease (ADPKD) is prone to multiple complications, including cyst infection (CyI). 2-Deoxy-2-[18F]fluoro-d-glucose positron emission tomography/computed tomography ([18F]-FDG PET/CT) imaging has proved useful in the diagnosis of renal and hepatic CyI. A 4-point scale comparing the uptake of [18F]-FDG in the suspected infected cyst versus the hepatic physiological background has been recently proposed. We performed an independent validation of this semi-quantitative scoring system. All ADPKD patients hospitalized between January 2009 and November 2019 who underwent an [18F]-FDG PET/CT for suspected CyI were retrospectively identified using computer-based databases. Medical files were reviewed. CyI was conventionally defined by the combination of fever (≥38°C), abdominal pain, increased plasma C-reactive protein levels (≥70mg/L), absence of any other cause of inflammation and favourable outcome after ≥21days of antibiotics. [18F]-FDG uptake of the suspected CyI was evaluated using a 4-point scale comparing the uptake of [18F]-FDG around the infected cysts with the uptake in the hepatic parenchyma. Statistics were performed using SAS version 9.4. Fifty-one [18F]-FDG PET/CT scans in 51 patients were included, of which 11 were cases of CyI. The agreement between the 4-point scale and the gold-standard criteria of CyI was significant [odds ratio of 6.03 for CyI in case of a score ≥3 (P = .014)]. The corresponding sensitivity, specificity, and positive and negative predictive values of [18F]-FDG PET/CT using the 4-point scale were 64% [Clopper-Pearson 95% confidence interval (CI) 30%-89%], 78% (95% CI 62%-89%), 44% (95% CI 20%-70%) and 89% (95% CI 73%-97%), respectively. Our independent validation cohort confirms the use of a semi-quantitative 4-point scoring system of [18F]-FDG PET/CT imaging in the diagnosis of CyI in patients with ADPKD. Considering its performance metrics with high specificity and negative predictive value, the scoring system is particularly useful to distinguish other causes of clinical inflammation than CyI and as such avoid unnecessarily long antibiotic treatment.
Read full abstract