The Diagnostic Yield of [18F]FDG-PET/CT in a Heterogeneous In-Patient Population with Suspected Infection or Inflammation Is Comparable to Findings in Patients with Classic Fever of Unknown Origin.

Abstract

Suspected infection or inflammation of unknown origin in in-patients remains challenging. Literature on [18F]FDG-PET/CT is abundant in classic fever of unknown origin (FUO), but evidence is complex and may not always reflect clinical reality. This study explores the application of [18F]FDG-PET/CT in a diverse clinical population of in-patients with suspected infection not defined by stringent FUO-criteria. Retrospective chart review of consecutive in-patients who underwent [18F]FDG-PET/CT in the workup of suspected infection or inflammation from 1 July 2022 to 31 December 2022 was conducted. We evaluated indications, diagnostic yield, and clinical impact of [18F]FDG-PET/CT, and compared the findings of [18F]FDG-PET/CT and stand-alone CT. Univariate logistic regression assessed associations between [18F]FDG-PET/CT outcome and clinical parameters. Receiver operating characteristic curve (ROC) analysis evaluated diagnostic performance. 77 patients met the inclusion criteria. [18F]FDG-PET/CT established a diagnosis in 35% of cases, ruled out focal infection in 26%, and thus was helpful in 61% of patients. It prompted 72 additional examinations resulting in seven incidental diagnoses, including two cancers. Antibiotic treatment was changed in 26% of cases. Regression analysis found white blood cell counts (WBC) associated with true positive outcomes. [18F]FDG-PET/CT was compared to stand-alone CT findings, and was concordant in 69% of cases. Results were comparable to findings in more classic FUO. [18F]FDG-PET/CT was clinically helpful in 61% of cases but also prompted many additional examinations with relatively few clinically important findings. WBC count was a predictor of true positive outcome. CT and [18F]FDG-PET/CT were discordant in 31%, of cases, especially in cases of endocarditis and spondylodiscitis.