OBJECTIVE: Suspicious incidental gastrointestinal FDG uptake during positron-emission tomography/computed tomography (PET/CT) examinations can be caused by different diseases, including malignancies. However, differentiation with PET alone is difficult. The aim of this study was to investigate the potential of PET alone, contrast-enhanced CT (ceCT), and low-dose CT (ldCT) in routine PET/CT protocols for differentiation of incidental gastrointestinal lesions. MATERIALS AND METHODS: Sixty patients with incidental gastrointestinal lesions who underwent a routine PET/CT protocol with ldCT and ceCT were retrospectively analysed. The PET lesions were evaluated regarding their FDG uptake patterns and the standard uptake value. The anatomical correlates in both CT protocols were compared in regard to the correct lesion classification with the reference standard endoscopy. RESULTS: Sixty-two lesions were found in 60 patients (17 malignant, 10 premalignant, 5 benign, 13 inflammatory, 17 physiological). The differentiation of the FDG uptake patterns did not enable reliable lesion classification. The positive predictive value for pathology was 0.81 for ceCT in PET/CT and 0.70 for ldCT. Malignancies were detected in 100% of the patients by ceCT vs. 29.4% by ldCT. The false negative rate of ceCT for all pathologies was 31.1%, vs. 68.9% for ldCT. False positive results (17/62) could not be excluded sufficiently by either CT protocol. CONCLUSION: PET/ceCT protocols provide additional benefit especially in detecting gastrointestinal malignancies as a cause of suspicious incidental gastrointestinal FDG uptake. However, since follow-up endoscopy cannot be forgone due to the considerable false negative rate even with ceCT, the addition of ceCT to a routine PET/ldCT protocol cannot be recommended for this purpose.
Adult ; Aged ; Aged, 80 and over ; Contrast Media/diagnostic use/pharmacokinetics ; Female ; Fluorodeoxyglucose F18/*diagnostic use/pharmacokinetics ; Follow-Up Studies ; Gastrointestinal Diseases/*diagnosis/metabolism ; Gastrointestinal Tract/*metabolism/radiography/radionuclide imaging ; Humans ; Male ; Middle Aged ; Positron-Emission Tomography/*methods ; Reproducibility of Results ; Retrospective Studies ; Tomography, X-Ray Computed/*methods

OBJECTIVE: First, to investigate the diagnostic performance of fast T1-weighted sequences for lung nodule evaluation in oncologic magnetic resonance (MR)/positron emission tomography (PET). Second, to evaluate the influence of image acquisition in inspiration and expiration breath-hold on diagnostic performance. MATERIALS AND METHODS: The study was approved by the local Institutional Review Board. PET/CT and MR/PET of 44 cancer patients were evaluated by 2 readers. PET/CT included lung computed tomography (CT) scans in inspiration and expiration (CTin, CTex). MR/PET included Dixon sequence for attenuation correction and fast T1-weighted volumetric interpolated breath-hold examination (VIBE) sequences (volume interpolated breath-hold examination acquired in inspiration [VIBEin], volume interpolated breath-hold examination acquired in expiration [VIBEex]). Diagnostic performance was analyzed for lesion-, lobe-, and size-dependence. Diagnostic confidence was evaluated (4-point Likert-scale; 1 = high). Jackknife alternative free-response receiver-operating characteristic (JAFROC) analysis was performed. RESULTS: Seventy-six pulmonary lesions were evaluated. Lesion-based detection rates were: CTex, 77.6%; VIBEin, 53.3%; VIBEex, 51.3%; and Dixon, 22.4%. Lobe-based detection rates were: CTex, 89.6%; VIBEin, 58.3%; VIBEex, 60.4%; and Dixon, 31.3%. In contrast to CT, inspiration versus expiration did not alter diagnostic performance in VIBE sequences. Diagnostic confidence was best for VIBEin and CTex and decreased in VIBEex and Dixon (1.2 ± 0.6; 1.2 ± 0.7; 1.5 ± 0.9; 1.7 ± 1.1, respectively). The JAFROC figure-of-merit of Dixon was significantly lower. All patients with malignant lesions were identified by CTex, VIBEin, and VIBEex, while 3 patients were false-negative in Dixon. CONCLUSION: Fast T1-weighted VIBE sequences allow for identification of patients with malignant pulmonary lesions. The Dixon sequence is not recommended for lung nodule evaluation in oncologic MR/PET patients. In contrast to CT, inspiration versus expiratory breath-hold in VIBE sequences was less crucial for lung nodule evaluation but was important for diagnostic confidence.
Ethics Committees, Research ; Humans ; Lung* ; Positron-Emission Tomography and Computed Tomography