Value of PET/CT for the management of inflammatory syndromes of undetermined origin in the Internal Medicine department of the University Hospital of Amiens

Introduction: Inflammation of unknown origin (IUO) is a common concern in the internal medicine departments. The aim of our study was to examine the diagnostic value of PET/CT for the investigation of patients with IUO in an internal medicine department. Materials and methods: We performed a retrospective cohort study in the internal medicine department of Amiens University Hospital in the patients for whom a PET/CT was performed, between October 2004 and April 2017. Patients were included if PET/CT has been prescribed for the exploration of IUO. The patients were classified according to the result: the PET/CT made the diagnosis, the PET/CT participated in the diagnosis (by targeting an area to biopsy or feeding relevantly the diagnostic argumentation), the PET/CT was negative or the PET/CT gave false positive result. Results: From the 763 patients in the cohort, 144 were included. The median age was 67.7 years [IQR55.8-75.8], with 56% men. The diagnoses finally retained were varied: 13.2% of infectious diseases, 33% of noninfectious inflammatory diseases, 16% of tumoral diseases, and 8.3% of other pathologies. However, IUO remained unexplained in 29.2% of the patients by the end of the hospital work-up with a subsequent favourable evolution in half of the cases. Fever was present in 43% of the cases, and failure to thrive in 40%. Before the prescription of PET/CT: 81% of the patients had a thoracoabdominal CT scan, 30% a colonoscopy, 43% a cardiac ultrasound examination, 27% a bone marrow biopsy, 27% a temporal artery biopsy and a 15% dental x-ray. From the 144 patients who had their PET/CT scan for IUO: the PET/CT made the diagnosis on its own in 13.2% of cases; the PET/CT participated in the diagnosis in 26% of cases; the PET/CT was negative in about 43.7% of cases. It was therefore noted that PET/CT misorient the management in 17% of situations. Contrariwise, when the PET/CT fixed and was helpful, it divided the diagnosis period between the first consultation and the date of the diagnosis by 2.7 (on average 142 days vs 63 days, p = 0.0006). And it divided by 3.1 the diagnostic delay between the date of completion of the PET/CT and the date of diagnosis (47 days vs 16 days, p = 0.007). When the PET/CT fixed (relevant result or not) it divided the median time to diagnosis after PET/CT by two (25 days [6-55] when the test fixed versus 50 days [9- 236] when it does not fix, p = 0.0323). Its median completion time was 34 days [17-85]. The result of PET/CT ended by a hypermetabolic location to biopsy in 30% of cases. In univariate analysis, PET/CT seemed to be more in patients without fever (p = 0.0467) and with general impairment (p = 0.007). When thoracoabdominopelvic CT was not contributory, PET/CT was in 21.5% of cases. Conclusion: PET/CT seems to be useful for the diagnostic workup of IUO and seems to shorten the diagnostic delay. However, its false positive results may have serious consequences. (author)