Could computed tomography (CT)-based radiomics enhance preoperative risk assessment for patients with clear cell renal cell carcinoma (ccRCC)?
For the retrospective study, published recently in Insights into Imaging, researchers explored radiomic models, a clinical model and a fusion model (combining radiomic and clinical factors) for predicting preoperative grading and survival in patients with ccRCC. The cohort was comprised of 284 patients in the training, 71 patients in the internal validation group and 40 patients in an external validation cohort, according to the study.
The researchers found that the radiomic model that combined internal tumor area (IAT) with peritumoral area of the tumor at 5 mm (PAT 5 mm) had a 90 percent AUC and 100 percent sensitivity for predicting ccRCC grading in external validation testing in comparison to 61 percent and 23 percent, respectively, for a clinical model.
The radiomics models that incorporated peritumoral features (PAT 3 mm and PAT 5 mm) also demonstrated higher predictive capabilities than the radiomics model that relied solely on intratumoral features (IAT). External validation testing revealed a 70 percent AUC and 46 percent sensitivity for the IAT model in contrast to 83 percent and 85 percent, respectively, for the PAT 3 mm model and the aforementioned 90 percent and 100 percent, respectively, for the PAT 5 mm model, according to the study authors.
“Our results revealed that the PAT model (PAT 3 mm and PAT 5 mm) had higher predictive values than that of the IAT model. This finding confirms that incorporating incremental information from both the internal tumor components and peritumoral features into a combined model can significantly enhance predictive performance,” wrote lead study author Xiaoxia Li, M.D., who is affiliated with the Department of Radiology at Zhongshan Hospital and Fudan University in Xiamen, China, and colleagues.
Noting that conventional imaging struggles to capture the microenvironment surrounding a tumor and that ccRCC research, at times, overlooks the peritumoral microenvironment, the study authors emphasized the significance of considering peritumoral characteristics in inratumoral grading.
“Our study revealed an association between the histological characteristics of perirenal fat invasion and radiomics features. This finding provides evidence that the radiomics features obtained from the PAT region can reflect the biological behavior of tumors,” maintained Li and colleagues. “Therefore, when delineating ROIs, it is important not to overlook areas within the tumor and those outside the tumor, as this is associated with tumor proliferation and heterogeneity.”
Three Key Takeaways
- Enhanced predictive performance. The radiomic models that incorporated peritumoral features (PAT 3 mm and PAT 5 mm) showed significantly higher predictive capabilities for ccRCC grading compared to models that used only intratumoral features (IAT). The PAT 5 mm model demonstrated a 90 percent AUC and 100 percent sensitivity in external validation testing, outperforming the clinical model which had a 61 percent AUC and 23 percent sensitivity.
- Importance of peritumoral characteristics. The study highlighted the importance of including peritumoral characteristics in radiomic models, as conventional imaging often fails to capture the microenvironment surrounding a tumor. Incorporating peritumoral features can enhance the predictive performance of radiomic models, providing a more comprehensive assessment of tumor behavior and malignancy.
- Clinical implications of radiomic features. The study found a significant association between radiomic features from the peritumoral area and the biological behavior of tumors. Specifically, higher kurtosis values in the PAT 5 mm model indicated greater cell density and potential tumor heterogeneity, which are associated with increased malignancy. This suggests that detailed radiomic analysis can offer valuable insights into tumor proliferation and heterogeneity, aiding in better preoperative risk assessment.
The researchers added that a first-order texture feature involving kurtosis values had the most significant impact in the PAT 5 mm model.
“A higher kurtosis value indicates greater cell density within the tumor, implying a uniform and densely structured tumor. … Higher values of kurtosis may indicate increased tumor heterogeneity, which has a potential correlation with tumor malignancy. Our research supports this relationship, as we have discovered a close association between higher kurtosis and tumor malignancy grade,” added Li and colleagues.
(Editor’s note: For related content, see “Meta-Analysis Assesses Impact of PSMA PET/CT for Staging of Renal Cell Carcinoma,” “Multicenter CT Study Shows Benefits of Emerging Diagnostic Model for Clear Cell Renal Cell Carcinoma” and “Can a CT-Based Radiomics Model Enhance Risk Stratification for Clear Cell Renal Cell Carcinoma?”)
In regard to study limitations, the authors noted the retrospective design and small sample size with respect to cases involving chromophobe and papillary RCC.
