Differential diagnosis of hypervascular ultra-small renal cell carcinoma and renal angiomyolipoma with minimal fat in early stage by using thin-section multidetector computed tomography

CECT-Based Radiomic Nomogram of Different Machine Learning Models for Differentiating Malignant and Benign Solid-Containing Renal Masses

Objective

This study aimed to explore the value of a radiomic nomogram based on contrast-enhanced computed tomography (CECT) for differentiating benign and malignant solid-containing renal masses.

Materials and Methods

A total of 122 patients with pathologically confirmed benign (n=47) or malignant (n=75) solid-containing renal masses were enrolled in this study. Radiomic features were extracted from the arterial, venous and delayed phases and further analysed by dimensionality reduction and selection. Four mainstream machine learning algorithm training models, namely, support vector machine (SVM), k-nearest neighbour (kNN), light gradient boosting (LightGBM) and logistic regression (LR), were constructed to determine the best classifier model. Univariate and multivariate analyses were used to determine the best clinical characteristics for constructing a clinical model. The radiomic and clinical signatures were integrated to construct a combined radiomic nomogram model. Receiver operating characteristic (ROC) curves and the area under the curve (AUC) were used to evaluate the performance of the radiomic nomogram, radiomic signature, and clinical model.

Results

Thirteen radiomic features were selected for the development of the radiomic signature. Among the various radiomic models, the LR model demonstrated superior predictive efficiency and robustness, yielding an AUC of 0.952 in the training cohort and 0.887 in the test cohort. The AUC for the clinical model was 0.854 in the training cohort and 0.747 in the test cohort. Furthermore, the radiomic nomogram, which incorporated sex, age, alcohol consumption history, and the radiomic signature, exhibited excellent discriminative performance, yielding an AUC of 0.973 in the training cohort and 0.900 in the test cohort.

Conclusion

The radiomic nomogram based on CECT offers a promising and noninvasive approach for distinguishing malignant from benign solid renal masses. This tool can be used to guide treatment strategies effectively and can provide valuable insights for clinicians.

Renal epithelioid angiomyolipoma: computed tomography manifestation and radiologic-pathologic correlation depending on different epithelioid component percentages

BACKGROUND

Renal epithelioid angiomyolipoma (EAML) is a rare and potentially malignant mesenchymal lesion mainly composed of epithelioid cells. Although some case reports or small case series have been published, the computed tomography (CT) manifestations and radiologic-pathologic correlation depending on different epithelioid component percentages have not been studied before.

OBJECTIVE

To investigate the CT manifestation and radiologic-pathologic correlation between renal EAML and angiomyolipoma (AML) with epithelioid component.

METHODS

The clinicopathologic and imaging data of 53 patients with an original diagnosis of EAML or AML with epithelioid component were retrospectively collected from three hospitals. All tissue specimens were re-sectioned and re-observed under the microscope. Samples were divided into an EAML group (≥ 80% epithelioid component, n = 25) and AML with epithelioid component group (5% ≤ epithelioid component < 80%, n = 28). Two radiologists reviewed the images in consensus, describing and comparing the CT manifestation, including the long diameter of the tumor, morphology, presence of necrosis or cystic change, hemorrhage, fat, calcification, enlarged blood vessels, and dynamic enhancement pattern according to the Hounsfield unit value of each CT phase between the two groups. The radiologic-pathologic correlation depending on the different percentages of epithelioid component were studied.

RESULTS

The long diameter of the tumor, presence of necrosis or cystic change, fat, enhancement pattern, and tumor-to-cortex enhancement ratio of the cortical phase between the two groups were significantly different (z =  - 2.932, P = 0.003; χ² = 18.020, P < 0.001; χ² = 16.377, P < 0.001; P = 0.020; and T =  - 3.944, P < 0.001, respectively). In multivariate logistic regression analysis, the significant predictive factors of EAML included the presence of necrosis or cystic change [odds ratio (OR) 11.864, P = 0.001] and absence of fat (OR 0.095, P = 0.003). Correlation analysis found that the presence of necrosis or cystic change (r = 0.679, P < 0.001) and fat (r =  - 0.603, P < 0.001) were both moderately related to the epithelioid component percentage. The combined model based on the presence of necrosis or cystic change and absence of fat yielded the best diagnostic performance in discriminating EAML and AML with epithelioid component with the highest area under the curve (0.887).

CONCLUSION

EAML has characteristic CT signs; these characteristic CT signs are closely related to the epithelioid component percentage. The presence of necrosis or cystic change and the absence of fat were independent predictors of EAML.

Differentiation of renal angiomyolipoma without visible fat from small clear cell renal cell carcinoma by using specific region of interest on contrast-enhanced CT: a new combination of quantitative tools

Background

To investigate the value of using specific region of interest (ROI) on contrast-enhanced CT for differentiating renal angiomyolipoma without visible fat (AML.wovf) from small clear cell renal cell carcinoma (ccRCC).

Methods

Four-phase (pre-contrast phase [PCP], corticomedullary phase [CMP], nephrographic phase [NP], and excretory phase [EP]) contrast-enhanced CT images of AML.wovf (n = 31) and ccRCC (n = 74) confirmed by histopathology were retrospectively analyzed. The CT attenuation value of tumor (AVT), net enhancement value (NEV), relative enhancement ratio (RER), heterogeneous degree of tumor (HDT) and standardized heterogeneous ratio (SHR) were obtained by using different ROIs [small: ROI (1), smaller: ROI (2), large: ROI (3)], and the differences of these quantitative data between AML.wovf and ccRCC were statistically analyzed. Multivariate regression was used to screen the main factors for differentiation in each scanning phase, and the prediction models were established and evaluated.

Results

Among the quantitative parameters determined by different ROIs, the degree of enhancement measured by ROI (2) and the enhanced heterogeneity measured by ROI (3) performed better than ROI (1) in distinguishing AML.wovf from ccRCC. The receiver operating characteristic (ROC) curves showed that the area under the curve (AUC) of RER_CMP (2), RER_NP (2) measured by ROI (2) and HDT_CMP and SHR_CMP measured by ROI (3) were higher (AUC = 0.876, 0.849, 0.837 and 0.800). Prediction models that incorporated demographic data, morphological features and quantitative data derived from the enhanced phase were superior to quantitative data derived from the pre-contrast phase in differentiating between AML.wovf and ccRCC. Among them, the model in CMP was the best prediction model with the highest AUC (AUC = 0.986).

Conclusion

The combination of quantitative data obtained by specific ROI in CMP can be used as a simple quantitative tool to distinguish AML.wovf from ccRCC, which has a high diagnostic value after combining demographic data and morphological features.