Prognostic value of FDG PET/CT in the evaluation of hospitalized patients with FUO

Objectives The aim of this retrospective study was to evaluate the prognostic performance of FDG PET/CT in hospitalized patients with a recent history of referring-physician-identified fever of unknown origin (FUO).

Methods The Johns Hopkins inpatient PET/CT report database was queried using “fever of unknown origin” and “FUO” as the key words and including scans from 2003 to 2012 in this IRB-approved analysis. A total of 151 patients were identified by these key words who underwent FDG PET/CT to help identify the underlying cause of fever. Seven patients were excluded due to insufficient follow-up data, leaving 144 evaluable patients (48.6% male and median age of 54 years). Each scan impression was assessed on a 5-point scale: 1 - no FUO cause identified, 2 - FUO cause not likely identified, 3 - equivocal identification of FUO cause, 4 - FUO cause probably identified, 5 - FUO cause definitely identified. Receiver-operator characteristic (ROC) analyses were applied to determine whether an equivocal score should be classified as positive or negative. Survival data were obtained from a variety of sources including the patient’s medical record and the social security death index. Clinical follow-up information was used to score each case as a true-positive, true-negative, false-positive, or false-negative and the outcomes of each of these groups were compared.

Results Based on ROC analysis, a score of 3 was used as the positive cutoff for PET/CT scores (i.e. a score of 蠅 3 was considered positive). Kaplan-Meier statistics revealed a significant hazard ratio of 4.7 (p<0.0001) for patients with a positive PET/CT scan. Median survival of patients with a positive PET/CT scan was 7.4 months and 34.8 months with a negative PET/CT scan. Survival of patients with a true-positive scan was significantly lower than any other group with a median survival of 6.9 months versus 19.7, 23.2, and 27.8 months for patients with false-negative, false-positive, and true-negative scans, respectively (p<0.0001). Oncologic and non-oncologic cases were also analyzed separately. A negative oncologic scan was associated with a significant survival advantage over a positive oncologic scan (median survival of 27.6 months versus 4.8 months, P<0.0001). Survival of patients with a positive non-oncologic scan did not differ significantly from survival of patients with a negative non-oncologic scan.

Conclusions A positive FDG PET/CT scan is associated with significantly worse prognosis than a negative FDG PET/CT scan in hospitalized patients with a recent history of fever of unknown origin.