Integrated FDG-PET/MRI enhances the diagnostic ability for distinguishing between type I and II endometrial cancers

Objectives: Endometrial cancers have been divided into estrogen-dependent type I and the less common clinically aggressive estrogen-independent type II. The recent introduction of integrated PET/MRI has demonstrated the advantages of simultaneous PET and MR imaging with higher soft-tissue contrast, multiplanar image acquisition, functional imaging capability, and lower radiation exposure than PET/CT. The aim of this study was to evaluate the capability of integrated FDG-PET/MRI to distinguish between type I and II endometrial cancers.

Methods: 18 patients with endometrial cancer (11 type I including 8 G1 and 3 G2 endometrioid adenocarcinomas, and 7 type II including 2 G3 endometrioid adenocarcinomas, 1 carcinosarcoma and 4 serous carcinomas) underwent FDG-PET/MRI before the initiation of treatment. Among them, 14 patients underwent simultaneous high-resolution, small field-of-view DWI. FDG SUV, apparent diffusion coefficient (ADC), and SUV-to-ADC ratio were compared between type I and II cancers. The diagnostic accuracy for discriminating type II from type I cancers was compared for each PET/MRI parameter using receiver-operating-characteristic (ROC) analysis.

Results: Although FDG SUV was not significantly different between type I and II cancers, significant differences were observed in ADC and SUV-to-ADC ratio between them. Type II cancer showed a significantly lower ADC (0.54 ± 0.27 x 10^-3) and a significantly higher SUV-to-ADC ratio (27.4 ± 6.1 x 10⁴) than did type I cancer (0.89 ± 0.11 x 10^-3, p=0.004, and 15.6 ± 3.9 x 10⁴, p=0.001, respectively). The comparison of area under the curve in ROC analyses indicated that the SUV-to-ADC ratio was the most accurate diagnostic parameter for predicting type I and II cancer (SUV-to-ADC ratio 0.96, ADC 0.84, SUV 0.57). The optimal SUV/ADC cutoff value of 20.5 x 10³ (s/mm²), determined by ROC analysis, revealed 80% sensitivity, 100% specificity, and 93% accuracy, which was much better than those for FDG SUV.

Conclusions: Pretreatment SUV-to-ADC ratio obtained by simultaneous PET and MR imaging is useful for predicting type I and II endometrial cancer.