Significance of F-18 FDG PET/MRI in the search for the etiology of inflammation of unclear origin and fever of unknown origin.

PURPOSE: To evaluate the contribution of F-18 FDG-PET/MRI in the search for the etiology of the inflammation of unknown origin (IUO) and fever of unknown origin (FUO). MATERIAL AND METHODS: The study included 104 patients who underwent F-18 FDG-PET/MRI for IUO or FUO. The sensitivity, specificity, predictive values of the PET/MRI findings in relation to the final diagnosis of IUO/FUO were evaluated. A five-point Likert scale was used to semiquantitatively assess the probability of the cause of IUO/FUO based on PET/MRI finding. Furthermore, clinical (fever, arthralgia, weight loss, night sweats, age) and laboratory (C-reactive protein, leukocytes) parameters were monitored and compared with the true positivity rate of PET/MRI. RESULTS: In patients with definitively identified etiology of FUO and IUO, FDG-PET/MRI achieved a sensitivity of 96 %, specificity of 82 %, and positive and negative predictive values of 92 and 90 %. The cause of the IUO was determined in 71 patients (68.3 %). In 33 (31.7 %) patients, the etiology of IUO/FUO remained unknown, while in 25 (75.8 %) of them the symptoms resolved spontaneously and in 8 (24.2 %) patients they persisted without explanation even after 12 months of the follow-up. The most significant parameter in relation to subsequent PET/MRI finding was increased level of CRP, which was present in 96 % of true positive PET/MRI and normal CRP level was present in 56 % of true negative PET/MRI. CONCLUSION: Based on this study, FDG-PET/MRI is a suitable alternative for the investigation of IUO/FUO, this imaging technique has a very high sensitivity and negative predictive value.

The relationship between gadolinium enhancement and [18 F]fluorothymidine uptake in brain lesions with the use of hybrid PET/MRI.

BACKGROUND: To evaluate and compare the diagnostic power of [18F]FLT-PET with ceMRI in patients with brain tumours or other focal lesions. METHODS: 121 patients with suspected brain tumour or those after brain tumour surgery were enroled in this retrospective study (61 females, 60 males, mean age 37.3 years, range 1-80 years). All patients underwent [18F]FLT-PET/MRI with gadolinium contrast agent application. In 118 of these patients, a final diagnosis was made, verified by histopathology or by follow-up. Agreement between ceMRI and [18F]FLT-PET of the whole study group was established. Further, sensitivity and specificity of ceMRI and [18F]FLT-PET were calculated for differentiation of high-grade vs. low-grade tumours, high-grade vs. low-grade tumours together with non-tumour lesions and for differentiation of high-grade tumours from all other verified lesions. RESULTS: [18F]FLT-PET and ceMRI findings were concordant in 119 cases (98%). On closer analysis of a subset of 64 patients with verified gliomas, the sensitivity and specificity of both PET and ceMRI were identical (90% and 84%, respectively) for differentiating low-grade from high-grade tumours, if the contrast enhancement and [18F]FLT uptake were considered as hallmarks of high-grade tumour. For differentiation of high-grade tumours from low-grade tumours and lesions of nontumorous aetiology (e.g., inflammatory lesions or post-therapeutic changes) in a subgroup of 93 patients by visual evaluation, the sensitivity of both PET and ceMRI was 90%, whereas the specificity of PET was slightly higher (61%) compared to ceMRI (57%). By receiver operating characteristic analysis, the sensitivity and specificity were 82% and 74%, respectively, when the threshold of SUVmax in the tumour was set to 0.9 g/ml. CONCLUSION: We demonstrated a generally very high correlation of [18F]FLT accumulation with contrast enhancement visible on ceMRI and a comparable diagnostic yield in both modalities for differentiating high-grade tumours from low-grade tumours and lesions of other aetiology.