CT texture analysis for the differentiation of papillary renal cell carcinoma subtypes

The Use of Radiomic Tools in Renal Mass Characterization

The incidence of renal mass detection has increased during recent decades, with an increased diagnosis of small renal masses, and a final benign diagnosis in some cases. To avoid unnecessary surgeries, there is an increasing interest in using radiomics tools to predict histological results, using radiological features. We performed a narrative review to evaluate the use of radiomics in renal mass characterization. Conventional images, such as computed tomography (CT) and magnetic resonance (MR), are the most common diagnostic tools in renal mass characterization. Distinguishing between benign and malignant tumors in small renal masses can be challenging using conventional methods. To improve subjective evaluation, the interest in using radiomics to obtain quantitative parameters from medical images has increased. Several studies have assessed this novel tool for renal mass characterization, comparing its ability to distinguish benign to malign tumors, the results in differentiating renal cell carcinoma subtypes, or the correlation with prognostic features, with other methods. In several studies, radiomic tools have shown a good accuracy in characterizing renal mass lesions. However, due to the heterogeneity in the radiomic model building, prospective and external validated studies are needed.

Advances in Imaging-Based Biomarkers in Renal Cell Carcinoma: A Critical Analysis of the Current Literature

Cross-sectional imaging is the standard diagnostic tool to determine underlying biology in renal masses, which is crucial for subsequent treatment. Currently, standard CT imaging is limited in its ability to differentiate benign from malignant disease. Therefore, various modalities have been investigated to identify imaging-based parameters to improve the noninvasive diagnosis of renal masses and renal cell carcinoma (RCC) subtypes. MRI was reported to predict grading of RCC and to identify RCC subtypes, and has been shown in a small cohort to predict the response to targeted therapy. Dynamic imaging is promising for the staging and diagnosis of RCC. PET/CT radiotracers, such as ¹⁸F-fluorodeoxyglucose (FDG), ¹²⁴I-cG250, radiolabeled prostate-specific membrane antigen (PSMA), and ¹¹C-acetate, have been reported to improve the identification of histology, grading, detection of metastasis, and assessment of response to systemic therapy, and to predict oncological outcomes. Moreover, ⁹⁹Tc-sestamibi and SPECT scans have shown promising results in distinguishing low-grade RCC from benign lesions. Radiomics has been used to further characterize renal masses based on semantic and textural analyses. In preliminary studies, integrated machine learning algorithms using radiomics proved to be more accurate in distinguishing benign from malignant renal masses compared to radiologists' interpretations. Radiomics and radiogenomics are used to complement risk classification models to predict oncological outcomes. Imaging-based biomarkers hold strong potential in RCC, but require standardization and external validation before integration into clinical routines.

The value of CT features and demographic data in the differential diagnosis of type 2 papillary renal cell carcinoma from fat-poor angiomyolipoma and oncocytoma

PURPOSES

To determine the CT features and demographic data predictive of type 2 papillary renal cell carcinoma (PRCC) that can help distinguish this neoplasm from fat-poor angiomyolipoma (fpAML) and oncocytoma.

METHODS

Fifty-four patients with type 2 PRCC, 48 with fpAML, and 47 with oncocytoma in the kidney from multiple centers were retrospectively reviewed. The demographic data and CT features of type 2 PRCC were analyzed and compared with those of fpAML and oncocytoma by univariate analysis and multiple logistic regression analysis to determine the predictive factors for differential diagnosis. Then, receiver operating characteristic (ROC) curve analysis was performed to further assess the logistic regression model and set the threshold level values of the numerical parameters.

RESULTS

Older age (≥ 46.5 years), unenhanced lesion-to-renal cortex attenuation (RLRCA) < 1.21, corticomedullary ratio of lesion to renal cortex net enhancement (RLRCNE) < 0.32, and size ≥ 30.1 mm were independent predictors for distinguishing type 2 PRCC from fpAML (OR 14.155, 8.332, and 57.745, respectively, P < 0.05 for all). The area under the curve (AUC) of the multiple logistic regression model in the ROC curve analysis was 0.970. In the combined evaluation, the four independent predictors had a sensitivity and specificity of 0.896 and 0.889, respectively. A corticomedullary RLRCNE < 0.61, irregular shape, and male sex were independent predictors for the differential diagnosis of type 2 PRCC from oncocytoma (OR 15.714, 12.158, and 6.175, respectively, P < 0.05 for all). In the combined evaluation, the three independent predictors had a sensitivity and specificity of 0.889 and 0.979, respectively. The AUC of the multiple logistic regression model in the ROC curve analysis was 0.964.

CONCLUSION

The combined application of CT features and demographic data had good ability in distinguishing type 2 PRCC from fpAML and oncocytoma, respectively.

Differential Diagnosis of Type 1 and Type 2 Papillary Renal Cell Carcinoma Based on Enhanced CT Radiomics Nomogram

Objectives

To construct a contrast-enhanced CT-based radiomics nomogram that combines clinical factors and a radiomics signature to distinguish papillary renal cell carcinoma (pRCC) type 1 from pRCC type 2 tumours.

Methods

A total of 131 patients with 60 in pRCC type 1 and 71 in pRCC type 2 were enrolled and divided into training set (n=91) and testing set (n=40). Patient demographics and enhanced CT imaging characteristics were evaluated to set up a clinical factors model. A radiomics signature was constructed and radiomics score (Rad-score) was calculated by extracting radiomics features from contrast-enhanced CT images in corticomedullary phase (CMP) and nephrographic phase (NP). A radiomics nomogram was then built by incorporating the Rad-score and significant clinical factors according to multivariate logistic regression analysis. The diagnostic performance of the clinical factors model, radiomics signature and radiomics nomogram was evaluated on both the training and testing sets.

Results

Three validated features were extracted from the CT images and used to construct the radiomics signature. Boundary blurring as an independent risk factor for tumours was used to build clinical factors model. The AUC value of the radiomics nomogram, which was based on the selected clinical factors and Rad-score, were 0.855 and 0.831 in the training and testing sets, respectively. The decision curves of the radiomics nomogram and radiomics signature in the training set indicated an overall net benefit over the clinical factors model.

Conclusion

Radiomics nomogram combining clinical factors and radiomics signature is a non-invasive prediction method with a good prediction for pRCC type 1 tumours and type 2 tumours preoperatively and has some significance in guiding clinicians selecting subsequent treatment plans.

Novel Imaging Methods for Renal Mass Characterization: A Collaborative Review

CONTEXT

The incidental detection of localized renal masses has been rising steadily, but a significant proportion of these tumors are benign or indolent and, in most cases, do not require treatment. At the present time, a majority of patients with an incidentally detected renal tumor undergo treatment for the presumption of cancer, leading to a significant number of unnecessary surgical interventions that can result in complications including loss of renal function. Thus, there exists a clinical need for improved tools to aid in the pretreatment characterization of renal tumors to inform patient management.

OBJECTIVE

To systematically review the evidence on noninvasive, imaging-based tools for solid renal mass characterization.

EVIDENCE ACQUISITION

The MEDLINE database was systematically searched for relevant studies on novel imaging techniques and interpretative tools for the characterization of solid renal masses, published in the past 10 yr.

EVIDENCE SYNTHESIS

Over the past decade, several novel imaging tools have offered promise for the improved characterization of indeterminate renal masses. Technologies of particular note include multiparametric magnetic resonance imaging of the kidney, molecular imaging with targeted radiopharmaceutical agents, and use of radiomics as well as artificial intelligence to enhance the interpretation of imaging studies. Among these, 99mTc-sestamibi single photon emission computed tomography/computed tomography (CT) for the identification of benign renal oncocytomas and hybrid oncocytic chromophobe tumors, and positron emission tomography/CT imaging with radiolabeled girentuximab for the identification of clear cell renal cell carcinoma, are likely to be closest to implementation in clinical practice.

CONCLUSIONS

A number of novel imaging tools stand poised to aid in the noninvasive characterization of indeterminate renal masses. In the future, these tools may aid in patient management by providing a comprehensive virtual biopsy, complete with information on tumor histology, underlying molecular abnormalities, and ultimately disease prognosis.

PATIENT SUMMARY

Not all renal tumors require treatment, as a significant proportion are either benign or have limited metastatic potential. Several innovative imaging tools have shown promise for their ability to improve the characterization of renal tumors and provide guidance in terms of patient management.

Prediction of histologic grade and type of small (< 4 cm) papillary renal cell carcinomas using texture and neural network analysis: a feasibility study

OBJECTIVE

To predict the histologic grade and type of small papillary renal cell carcinomas (pRCCs) using texture analysis and machine learning algorithms.

METHODS

This was a retrospective HIPAA-compliant study. 24 noncontrast (NC), 22 corticomedullary (CM) phase, and 24 nephrographic (NG) phase CTs of small (< 4 cm) surgically resected pRCCs were identified. Surgical pathology classified the tumors as low- or high-Fuhrman histologic grade and type 1 or 2. The axial image with the largest cross-sectional tumor area was exported and segmented. Six histogram and 31 texture (20 gray-level co-occurrences and 11 gray-level run-lengths) features were calculated for each tumor in each phase. Feature values in low- versus high-grade and type 1 versus 2 pRCCs were compared. Area under the receiver operating curve (AUC) was calculated for each feature to assess prediction of histologic grade and type of pRCCs in each phase. Histogram, texture, and combined histogram and texture feature sets were used to train and test three classification algorithms (support vector machine (SVM), random forest, and histogram-based gradient boosting decision tree (HGBDT)) with stratified shuffle splits and threefold cross-validation; AUCs were calculated for each algorithm in each phase to assess prediction of histologic grade and type of pRCCs.

RESULTS

Individual histogram and texture features did not have statistically significant differences between low- and high-grade or type 1 and type 2 pRCCs across all phases. Individual features had low predictive power for tumor grade or type in all phases (AUC < 0.70). HGBDT was highly accurate at predicting pRCC histologic grade and type using histogram, texture or combined histogram and texture feature data from the CM phase (AUCs = 0.97-1.0). All algorithms had highest AUCs using CM phase feature data sets; AUCs decreased using feature sets from NC or NG phases.

CONCLUSIONS

The histologic grade and type of small pRCCs can be predicted with classification algorithms using CM histogram and texture features, which outperform NC and NG phase image data. The accurate prediction of pRCC histologic grade and type may be able to further guide management of patients with small (< 4 cm) pRCCs being considered for active surveillance.