The predictive value of C-reactive protein and erythrocyte sedimentation rate for 18F-FDG PET/CT outcome in patients with fever and inflammation of unknown origin

Hans Balink, Nic J. G. M. Veeger, Roel J. Bennink, Riemer H. J. A. Slart, Frits Holleman, Berthe L. F. van Eck-Smit, Hein J. Verberne

Research output: Contribution to journalArticleAcademicpeer-review

27 Citations (Scopus)

Abstract

The objective of this study was to determine the predictive value of C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) to a positive fluorine-18 fluorodeoxyglucose ((18)F-FDG) PET/computed tomography (CT) result in patients with inflammation of unknown origin and fever of unknown origin. Individual data of 498 patients were retrieved from three retrospective studies. Receiver operating characteristic derived areas under the curve were used to assess (18)F-FDG PET/CT versus age, CRP, and ESR. The discriminative value of age, CRP, and ESR related to (18)F-FDG PET/CT was examined using the net reclassification improvement (NRI). A diagnosis was established in 331 patients; (18)F-FDG PET/CT had a diagnostic accuracy of 89%. (18)F-FDG PET/CT had the highest area under the curve (0.89, P <0.001). The addition of (18)F-FDG PET/CT to a diagnosis prediction model including age, CRP, and ESR resulted in an NRI of 42% (P <0.001). In the same model with CRP values below 20 mg/l or ESR values below 20 mm/h, the NRI was 64% (P <0.001) and 29% (P=0.059), respectively. In 30 of 91 patients with CRP less than 10 mg/l, a diagnosis could be established; (18)F-FDG PET/CT was 100% true negative only in patients with CRP levels less than 5 mg/l. In patients with fever of unknown origin or inflammation of unknown origin, compared with elevated ESR levels, elevated CRP levels more often indicate a true positive (18)F-FDG PET/CT outcome.In addition, (18)F-FDG PET/CT, compared with CRP and ESR, shows the highest discrimination of patients with possible disabling disease
Original languageEnglish
Pages (from-to)604-609
JournalNuclear Medicine Communications
Volume36
Issue number6
DOIs
Publication statusPublished - 2015

Cite this