MRI- and Histologic-Molecular-Based Radio-Genomics Nomogram for Preoperative Assessment of Risk Classes in Endometrial Cancer

Abstract

High- and low-risk endometrial carcinoma (EC) differ in whether or not a lymphadenectomy is performed. We aimed to develop MRI-based radio-genomic models able to preoperatively assess lymph-vascular space invasion (LVSI) and discriminate between low- and high-risk EC according to the ESMO-ESGO-ESTRO 2020 guidelines, which include molecular risk classification proposed by “ProMisE”. This is a retrospective, multicentric study that included 64 women with EC who underwent 3T-MRI before a hysterectomy. Radiomics features were extracted from T2WI images and apparent diffusion coefficient maps (ADC) after manual segmentation of the gross tumor volume. We constructed a multiple logistic regression approach from the most relevant radiomic features to distinguish between low- and high-risk classes under the ESMO-ESGO-ESTRO 2020 guidelines. A similar approach was taken to assess LVSI. Model diagnostic performance was assessed via ROC curves, accuracy, sensitivity and specificity on training and test sets. The LVSI predictive model used a single feature from ADC as a predictor; the risk class model used two features as predictors from both ADC and T2WI. The low-risk predictive model showed an AUC of 0.74 with an accuracy, sensitivity, and specificity of 0.74, 0.76, 0.94; the LVSI model showed an AUC of 0.59 with an accuracy, sensitivity, and specificity of 0.60, 0.50, 0.61. MRI-based radio-genomic models are useful for preoperative EC risk stratification and may facilitate therapeutic management.

Preoperative prediction of lymphovascular space invasion in endometrioid adenocarcinoma: an MRI-based radiomics nomogram with consideration of the peritumoral region

“…Compared with two-dimensional ROIs, three-dimensional VOIs can improve the specificity of GLCM features for the identification of tumor constituents (24). GLCM and GLSZM have been repeatedly mentioned in previous imaging studies of LVSI status classification and risk classification of EC (11,25,26). Moreover, kurtosis and skewness were associated with LVSI, deep muscular invasion, and high-grade tumors of EC in a previous study (27).…”

Prediction of Deep Myometrial Infiltration, Clinical Risk Category, Histological Type, and Lymphovascular Space Invasion in Women with Endometrial Cancer Based on Clinical and T2-Weighted MRI Radiomic Features

“…If cancer grade is included in our risk model, a classification AUC of 0.84 ( Figure 4 A) can be achieved; however, using LASSO with only radiomic features, the classification accuracy model results in an AUC of 0.67 ( Figure 4 D). Another recent study [ 22 ] showed the low-risk predictive model with an AUC accuracy of 0.74. This accuracy is close to our LASSO results; this is because both methods used the logistic regression method for the classification.…”

“…Further work is needed to upgrade these models to achieve better discrimination accuracy with even larger datasets. Finally, most current studies employed only AUC to assess the classification models [ 8 , 9 , 10 , 20 , 22 ], while AUC has several flaws [ 37 ] such as it being sensitive to class imbalance [ 38 ]. In this study, we included the F1 score and precision–recall curve in our analysis.…”

“…For example, a recent study [ 8 ] adopted a random forest to assess DMI, whilst a second study [ 21 ] applied a logistic regression method. Other studies used a support vector machine (SVM) method [ 20 ] and logistic regression [ 22 ] for risk stratification. These studies did not compare different machine learning methods to achieve the best predictive results.…”