Taein An1, Chan Kyo Kim1,2,3. 1. Department of Radiology and Center for Imaging Science, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea. 2. Department of Medical Device Management & Research, SAIHST, Sungkyunkwan University, Seoul, South Korea. 3. Department of Digital Health, SAIHST, Sungkyunkwan University, Seoul, South Korea.
Abstract
OBJECTIVES: Accurate pre-operative prediction of risk stratification using a non-invasive imaging tool is clinically important for planning optimal treatment strategies, particularly in early-stage endometrial cancer (EC). This study aimed to investigate the utility of apparent diffusion coefficient (ADC) histogram analysis in evaluating the pathological characteristics and risk stratification in patients with Stage I EC. METHODS: Between October 2009 and December 2014, a total of 108 patients with surgically proven Stage I EC (endometrioid type = 91; non-endometrioid type = 17) excluding stage ≥II that underwent preoperative 3T-diffusion-weighted imaging without administration of contrast medium were enrolled in this retrospective study. Risk stratification was divided into four risk categories based on the ESMO-ESGO-ESTRO Guidelines: low, intermediate, high-intermediate, and high risk. The ADC histogram parameters (minimum, mean [ADCmean], 10th-90th percentile, and maximum [ADCmax]) of the tumor were generated using an in-house software. The ADC histogram parameters were compared between patients with endometrioid type and non-endometrioid type, between Stage IA and IB, between histological grades, and evaluated for differentiating non-high risk group from high risk group. Inter-reader agreement for tumor ADC measurements was also evaluated. Statistical analyses were performed using the Student's t-test, Mann-Whitney U test, receiver operating characteristics (ROC) analysis, or intraclass correlation coefficient (ICC). RESULTS: In differentiating endometrioid type from non-endometrioid type EC, all ADC histogram parameters were statistically significant (p < 0.05). In differentiating histological grades, 90th percentile ADC and ADCmax showed significantly higher values in tumor Grade III than in tumor Grade I-II (p < 0.05). In differentiating superficial myometrial invasion from deep myometrial invasion, all ADC histogram parameters were statistically significant (p < 0.05), except ADCmax. In differentiating non-high risk group from high risk group, ADCmean, 75th-90th percentile ADC, and ADCmax were statistically significant (p < 0.05). For predicting the high risk group, the area under the ROC curve of ADCmax was 0.628 and the highest among other histogram parameters. All histogram parameters revealed moderate to good inter-reader reliability (ICC = 0.581‒0.769). CONCLUSION: The ADC histogram analysis as reproducible tool may be useful for evaluating the pathological characteristics and risk stratification in patients with early-stage EC. ADVANCES IN KNOWLEDGE: ADC histogram analysis may be useful for evaluating risk stratification in early-stage endometrial cancer patients.
OBJECTIVES: Accurate pre-operative prediction of risk stratification using a non-invasive imaging tool is clinically important for planning optimal treatment strategies, particularly in early-stage endometrial cancer (EC). This study aimed to investigate the utility of apparent diffusion coefficient (ADC) histogram analysis in evaluating the pathological characteristics and risk stratification in patients with Stage I EC. METHODS: Between October 2009 and December 2014, a total of 108 patients with surgically proven Stage I EC (endometrioid type = 91; non-endometrioid type = 17) excluding stage ≥II that underwent preoperative 3T-diffusion-weighted imaging without administration of contrast medium were enrolled in this retrospective study. Risk stratification was divided into four risk categories based on the ESMO-ESGO-ESTRO Guidelines: low, intermediate, high-intermediate, and high risk. The ADC histogram parameters (minimum, mean [ADCmean], 10th-90th percentile, and maximum [ADCmax]) of the tumor were generated using an in-house software. The ADC histogram parameters were compared between patients with endometrioid type and non-endometrioid type, between Stage IA and IB, between histological grades, and evaluated for differentiating non-high risk group from high risk group. Inter-reader agreement for tumor ADC measurements was also evaluated. Statistical analyses were performed using the Student's t-test, Mann-Whitney U test, receiver operating characteristics (ROC) analysis, or intraclass correlation coefficient (ICC). RESULTS: In differentiating endometrioid type from non-endometrioid type EC, all ADC histogram parameters were statistically significant (p < 0.05). In differentiating histological grades, 90th percentile ADC and ADCmax showed significantly higher values in tumor Grade III than in tumor Grade I-II (p < 0.05). In differentiating superficial myometrial invasion from deep myometrial invasion, all ADC histogram parameters were statistically significant (p < 0.05), except ADCmax. In differentiating non-high risk group from high risk group, ADCmean, 75th-90th percentile ADC, and ADCmax were statistically significant (p < 0.05). For predicting the high risk group, the area under the ROC curve of ADCmax was 0.628 and the highest among other histogram parameters. All histogram parameters revealed moderate to good inter-reader reliability (ICC = 0.581‒0.769). CONCLUSION: The ADC histogram analysis as reproducible tool may be useful for evaluating the pathological characteristics and risk stratification in patients with early-stage EC. ADVANCES IN KNOWLEDGE: ADC histogram analysis may be useful for evaluating risk stratification in early-stage endometrial cancer patients.
Authors: N Bharwani; M E Miquel; A Sahdev; P Narayanan; G Malietzis; R H Reznek; A G Rockall Journal: Br J Radiol Date: 2011-09-06 Impact factor: 3.039
Authors: Roberto Vargas; J Alejandro Rauh-Hain; Joel Clemmer; Rachel M Clark; Annekathryn Goodman; Whitfield B Growdon; John O Schorge; Marcela G Del Carmen; Neil S Horowitz; David M Boruta Journal: Gynecol Oncol Date: 2014-02-16 Impact factor: 5.482
Authors: N Colombo; C Creutzberg; F Amant; T Bosse; A González-Martín; J Ledermann; C Marth; R Nout; D Querleu; M R Mirza; C Sessa Journal: Ann Oncol Date: 2015-12-02 Impact factor: 32.976