Multiparametric magnetic resonance imaging-derived radiomics for the prediction of disease-free survival in early-stage squamous cervical cancer.

OBJECTIVE: To conduct multiparametric magnetic resonance imaging (MRI)-derived radiomics based on multi-scale tumor region for predicting disease-free survival (DFS) in early-stage squamous cervical cancer (ESSCC). METHODS: A total of 191 ESSCC patients (training cohort, n = 135; validation cohort, n = 56) from March 2016 to September 2019 were retrospectively recruited. Radiomics features were derived from the T2-weighted imaging (T2WI), contrast-enhanced T1-weighted imaging (CET1WI), diffusion-weighted imaging (DWI), and apparent diffusion coefficient (ADC) map for each patient. DFS-related radiomics features were selected in 3 target tumor volumes (VOIentire, VOI+5 mm, and VOI-5 mm) to build 3 rad-scores using the least absolute shrinkage and selection operator (LASSO) Cox regression analysis. Logistic regression was applied to build combined model incorporating rad-scores with clinical risk factors and compared with clinical model alone. Kaplan-Meier analysis was used to further validate prognostic value of selected clinical and radiomics characteristics. RESULTS: Three radiomics scores all showed favorable performances in DFS prediction. Rad-score (VOI+5 mm) performed best with a C-index of 0.750 in the training set and 0.839 in the validation set. Combined model was constructed by incorporating age categorized by 55, Federation of Gynecology and Obstetrics (Figo) stage, and lymphovascular space invasion with rad-score (VOI+5 mm). Combined model performed better than clinical model in DFS prediction in both the training set (C-index 0.815 vs 0.709; p = 0.024) and the validation set (C-index 0.866 vs 0.719; p = 0.001). CONCLUSION: Multiparametric MRI-derived radiomics based on multi-scale tumor region can aid in the prediction of DFS for ESSCC patients, thereby facilitating clinical decision-making. KEY POINTS: • Three radiomics scores based on multi-scale tumor region all showed favorable performances in DFS prediction. Rad-score (VOI+5 mm) performed best with favorable C-index values. • Combined model incorporating multiparametric MRI-based radiomics with clinical risk factors performed significantly better in DFS prediction than the clinical model. • Combined model presented as a nomogram can be easily used to predict survival, thereby facilitating clinical decision-making.

The value of the apparent diffusion coefficient in differentiating type II from type I endometrial carcinoma.

BACKGROUND: Diagnostic type II endometrial carcinoma (EC) is considered more aggressive and has a poorer prognosis than type I EC; differentiation between them is helpful for preoperative clinical decision-making. However, the diagnostic value of the apparent diffusion coefficient (ADC) in differentiating them remains unclear. PURPOSE: To investigate the value of ADC in differentiating type II EC from type I EC. MATERIAL AND METHODS: Ninety-four patients with EC who underwent diffusion-weighted imaging (DWI) were retrospectively included and divided into type I and type II subgroups, based on the postoperative pathologic results. We analyzed the clinical characteristics, conventional magnetic resonance imaging manifestations, and ADC mean values (ADCmean), ADC minimum values (ADCmin), and ADC max values (ADCmax). Receiver operating characteristic (ROC) curve analysis was further used to assess the predictive performance. RESULTS: The ADCmean, ADCmin, and tumor size differed significantly between the two subtypes. The area under the ROC curve (AUC) for ADCmean and ADCmin was 0.787 (95% confidence interval [CI] = 0.692-0.88) and 0.835 (95% CI = 0.751-0.919) for predicting type II EC, respectively. The optimal cut-off value of ADCmean for prediction was 0.757 × 10-3 mm2/s with a sensitivity of 91%, a specificity of 58%, and an accuracy of 74%, while for ADCmin was 0.637 × 10-3 mm2/s with a sensitivity of 82%, a specificity of 73%, and an accuracy of 75%. CONCLUSION: EC with lower ADCmean and ADCmin values derived from DWI, and a larger size, are indicative of type II EC.

Development and external validation of a nomogram for predicting overall survival of patients with non-endometrioid endometrial cancer: A population-based analysis.

Objectives: The main objective of this study was to identify the key predictors and construct a nomogram that can be used to predict the overall survival of individuals with non-endometrioid endometrial cancer. Methods: A total of 2686 non-endometrioid endometrial cancer patients confirmed between 1988 and 2018 were selected from the Surveillance, Epidemiology, and End Results database. They were divided into a training cohort and an internal validation cohort. Independent risk factors were chosen by Cox regression analyses. A predictive nomogram model for overall survival was constructed based on above factors. A Chinese cohort of 41 patients was collected to be an external validation cohort. Results: Eight variables were estimated as independent predictors for overall survival. A nomogram was established using these factors. The C-index for predicting the overall survival of patients with non-endometrioid endometrial cancer from the nomogram was 0.734, 0.700, and 0.767 in training, internal, and external validation cohort, respectively. Calibration plots and decision curve analysis showed that the nomogram was valuable for further clinical application. Conclusion: We constructed a nomogram which can be used as an effective tool to predict the 3- and 5-year overall survival of Non-endometrioid endometrial cancer patients.

MRI-based nomogram for differentiation of ovarian fibrothecoma and broad ligament myoma.

Currently, there are no effective approaches for differentiating ovarian fibrothecoma (OF) from broad ligament myoma (BLM). This retrospective study aimed to construct a nomogram prediction model based on MRI to differentiate OF from BLM. The quantitative and qualitative MRI features of 41 OFs and 51 BLMs were compared. Three models were established based on the combination of these features. The ability of the models to differentiate between the two cancers was assessed by ROC analysis. A nomogram based on the best model was constructed for clinical application. The three models showed good performance in differentiating between OF and BLM. The areas under the curve (AUC) of the models based on quantitative and qualitative variables were 0.88 (95% CI: 0.79-0.96) and 0.85 (95% CI: 0.76-0.93), respectively. The combined model designed from the significant variables exhibited the best diagnostic performance with the highest AUC of 0.92 (95% CI: 0.86-0.98). Calibration of the nomogram showed that the predicted probability matched the actual probability well. Analysis of the decision curve demonstrated that the nomogram was clinically useful. Relative T1 value, stone paving sign, enhancement patterns, and ascites were identified as valuable predictors for identifying OF or BLM. The MRI-based nomogram can serve as a preoperative tool to differentiate OF from BLM.

MRI-Based Radiomic Model for Preoperative Risk stratification in Stage I Endometrial Cancer.

Introduction: Preoperative risk stratification is crucial for clinical treatment of endometrial cancer (EC). This study aimed to establish a model based on magnetic resonance imaging (MRI) and clinical factors for risk classification of EC. Materials and Methods: A total of 102 patients with pathologically proven Stage I EC were included. Preoperative MRI examinations were performed in all the patients. 720 radiomic features were extracted from T2-weighted images. Least absolute shrinkage and selection operator (LASSO) regression model was performed to reduce irrelevant features. Logistic regression was used to build clinical, radiomic and combined predictive models. A nomogram was developed for clinical application. Results: The radiomic model has a better performance than the model based on clinical and conventional MRI characteristics [AUC of 0.946 (95% CI: 0.882-0.973) vs AUC of 0.756 (95% CI: 0.65, 0.86)]. The combined model consisting of radiomic features and tumor size showed the best predictive performance in the training cohort with AUC of 0.955 in the training and 0.889 in the validation cohorts. Conclusions: MRI-based radiomic model has great potential in prediction of low-risk ECs.

Preoperative MRI and immunohistochemical examination for the prediction of high-risk endometrial cancer.

BACKGROUND: Magnetic resonance imaging (MRI) and immunohistochemical (IHC) examination provides useful information for the risk stratification of endometrial cancer (EC). However, the use of the combination of MRI and IHC for the prediction of high-risk EC is controversial. The aim of this study was to evaluate the value of preoperative MRI and IHC examination in prediction of patients with high-risk EC. METHODS: This retrospective case-control study was conducted from January 1, 2018 to May 1, 2021 at two hospitals. A primary cohort (n=102) comprised patients with histologically confirmed EC in one hospital between January 1, 2018 and May 31, 2020. An additional external cohort (n=35) comprising patients with histologically confirmed EC in a different hospital from January 1, 2020 to May 1, 2021 was included for validation. Imaging features including tumor size, tumor margin, relative T2 value, tumor signal intensity on diffusion-weighted imaging (DWI), T1-weighted imaging (T1WI), T2-weighted imaging (T2WI) were determined from preoperative MRI images. IHC markers including ER, PR, p53 and Ki67 were determined through IHC staining of preoperative curettage specimen. Patients were divided into high-risk and low-intermediate- risk group based on the final histological results. Differences between categorical and numerical variables were assessed using chi-square test and independent-sample t-test, respectively. Multivariate binary logistic regression analyses were used for construction of the prediction model A fusion prediction model was constructed by combining MRI features and IHC markers. The predictive performance of the model was then validated using the external cohort. RESULTS: Imaging and IHC markers were significantly associated with risk ranks. Model 1 based on MRI features showed an area under the curve (AUC) of 0.822 [95% confidence interval (CI), 0.741-0.903] whereas Model 2 based on IHC markers showed an AUC of 0.894 (95% CI, 0.829-0.960). Notably, model 3 integrating independent MRI and IHC risk factors demonstrated good calibration and high differentiation ability with an AUC of 0.958 (95% CI, 0.923-0.993), and showed good discrimination with an AUC of 0.84 (95% CI, 0.677-0.942) using the external validation set. CONCLUSIONS: This study proposes a comprehensive predictive model comprising MRI and IHC features as a powerful tool for preoperative risk stratification to assist in clinical decision-making for EC patients.