J A Husby1, Ø O Salvesen2, I J Magnussen3, J Trovik4, L Bjørge4, H B Salvesen4, I S Haldorsen5. 1. Department of Radiology, Haukeland University Hospital, Jonas Liesvei 65, Postbox 7800, 5021 Bergen, Norway; Section for Radiology, Centre for Cancer Biomarkers, Department of Clinical Medicine, University of Bergen, 5020 Bergen, Norway. Electronic address: jenny.hild.aase@helse-bergen.no. 2. Unit for Applied Clinical Research, Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology, Postbox 8905, 7491 Trondheim, Norway. 3. Department of Radiology, Haukeland University Hospital, Jonas Liesvei 65, Postbox 7800, 5021 Bergen, Norway. 4. Department of Obstetrics and Gynaecology, Haukeland University Hospital, 5021 Bergen, Norway; Centre for Cancer Biomarkers, Department of Clinical Science, University of Bergen, 5020 Bergen, Norway. 5. Department of Radiology, Haukeland University Hospital, Jonas Liesvei 65, Postbox 7800, 5021 Bergen, Norway; Section for Radiology, Centre for Cancer Biomarkers, Department of Clinical Medicine, University of Bergen, 5020 Bergen, Norway.
Abstract
AIM: To explore possible correlations between tumour apparent diffusion coefficient (ADC), morphological tumour volume, and clinical and histological characteristics in endometrial carcinomas and to evaluate interobserver agreement for preoperative staging by MRI and for ADC measurements. MATERIALS AND METHODS: Preoperative conventional MRI including diffusion-weighted imaging (DWI) was performed in 105 endometrial carcinoma patients. Three radiologists independently reviewed the images for the presence of deep myometrial invasion, cervical stromal invasion, and lymph node metastases, and measured tumour ADC in regions of interest (ROIs). ADC values were analysed in relation to histomorphological characteristics and tumour volume. Kappa coefficients (κ) and intraclass correlation coefficients (ICC) for interobserver agreement for MRI staging results and ADC measurements, respectively, were calculated, and receiver operating characteristic (ROC) curves for identification of deep of myometrial invasion were generated. RESULTS: Mean tumour ADC was significantly lower in tumours with deep myometrial invasion (ADC = 0.75 × 10(-3) mm(2)/s) compared to tumours with superficial or no myometrial invasion (ADC = 0.85 × 10(-3) mm(2)/s; p < 0.001). ADC was negatively correlated to tumour size (p = 0.007). The interobserver agreement was fair (κ = 0.32) for depth of myometrial invasion, good for cervical stromal invasion (κ = 0.66), and moderate for lymph node metastases (κ = 0.54), and the interobserver variability for ADC value measurements was low (ICC = 0.60). CONCLUSION: Tumour ADC measurements may in the future provide an adjunct tool, aiding in the preoperative identification of high-risk patients with deep myometrial infiltration.
AIM: To explore possible correlations between tumour apparent diffusion coefficient (ADC), morphological tumour volume, and clinical and histological characteristics in endometrial carcinomas and to evaluate interobserver agreement for preoperative staging by MRI and for ADC measurements. MATERIALS AND METHODS: Preoperative conventional MRI including diffusion-weighted imaging (DWI) was performed in 105 endometrial carcinomapatients. Three radiologists independently reviewed the images for the presence of deep myometrial invasion, cervical stromal invasion, and lymph node metastases, and measured tumour ADC in regions of interest (ROIs). ADC values were analysed in relation to histomorphological characteristics and tumour volume. Kappa coefficients (κ) and intraclass correlation coefficients (ICC) for interobserver agreement for MRI staging results and ADC measurements, respectively, were calculated, and receiver operating characteristic (ROC) curves for identification of deep of myometrial invasion were generated. RESULTS: Mean tumour ADC was significantly lower in tumours with deep myometrial invasion (ADC = 0.75 × 10(-3) mm(2)/s) compared to tumours with superficial or no myometrial invasion (ADC = 0.85 × 10(-3) mm(2)/s; p < 0.001). ADC was negatively correlated to tumour size (p = 0.007). The interobserver agreement was fair (κ = 0.32) for depth of myometrial invasion, good for cervical stromal invasion (κ = 0.66), and moderate for lymph node metastases (κ = 0.54), and the interobserver variability for ADC value measurements was low (ICC = 0.60). CONCLUSION:Tumour ADC measurements may in the future provide an adjunct tool, aiding in the preoperative identification of high-risk patients with deep myometrial infiltration.