Diagnosis of peritoneal dissemination: Comparison of ¹⁸F-FDG PET/CT, diffusion-weighted MRI, and contrast-enhanced MDCT

OBJECTIVE. The purpose of this study was to compare the diagnostic performance of ¹⁸F-FDG PET/CT, MRI with and without diffusion-weighted imaging (DWI), and contrast-enhanced MDCT in the detection of peritoneal dissemination of malignant tumors. MATERIALS AND METHODS. We retrospectively evaluated the cases of 107 patients who underwent PET/CT and 130 patients who underwent MRI and contrast-enhanced MDCT. Twenty-six patients who underwent PET/CT and 23 who underwent MRI and contrast-enhanced MDCT were found to have peritoneal dissemination. All images were independently evaluated by two radiologists using a 5-point grading system. The results of PET/ CT, T1- and T2-weighted MRI without DWI, MRI with DWI (b = 1,000 s/mm2), and contrast-enhanced MDCT were compared patient by patient and lesion by lesion by use of receiver operating characteristics analysis. Sensitivity, specificity, and positive predictive value were calculated and compared by use of the chi-square test. RESULTS. Patient by patient, the area under the receiver operating characteristics curve of MRI without DWI (0.88) was significantly less than that of the other modalities (contrast-enhanced MDCT, 0.91; MRI with DWI, 0.93; PET/CT, 0.97). The sensitivity of PET/ CT (94%) was significantly higher than that of MRI without DWI (70%). The specificities of the modalities were not significantly different. In lesion-by-lesion analysis, MRI without DWI had significantly lower sensitivity (56%) than the other modalities (contrast-enhanced MDCT, 76%; MRI with DWI, 84%; PET/CT, 89%). The positive predictive value of PET/ CT (93%) was significantly higher than that of the other three modalities (contrast-enhanced MDCT, 73%; MRI without DWI, 70%; MRI with DWI, 72%). CONCLUSION. PET/CT is the most useful technique for pathologic staging in the care of patients with malignant disease. If PET/CT is not available, DWI can be used as a screening tool.