Development and validation of a diagnostic model based on contrast-enhanced CT to identify clear cell renal cell carcinoma in solid small renal masses: a multicentre study
Author Block: J. Han, Y. Tao, Y. Zhang; Zhuhai/CN
Purpose: The aim of this study was to develop and validate a diagnostic model based on contrast-enhanced CT for identifying clear cell renal cell carcinoma (ccRCC) in small renal masses (SRMs).
Methods or Background: This retrospective multicentre study enrolled patients with pathologically confirmed SRMs. Data from three centres were used as training set (n=229), with data from one centre serving as an independent external validation set (n=81). Univariate and multivariate logistic regression analyses were used to screen independent risk factors for ccRCC and build the classification and regression tree (CART) model. Three radiologists were asked to diagnose the SRMs in each case independently based on professional experience and re-evaluated using the CART model.
Results or Findings: A total of 71% (220/310) of SRMs were ccRCC. Enhancement pattern, early dark cortical band (EDCB), the ratio of lesion to normal cortex attenuation (L/C) in the corticomedullary phase, non-enhancing phase L/C and sex were used to develop the CART model. In the testing cohort, the AUC and accuracy of the CART model were 0.903, 85.1%. The accuracy of radiologists was 67.9%,58.0%, and 56.8%. With the CART model support, the accuracy of radiologists improved to 86.4%,84.0%,79.0%. Interobserver agreement was significantly improved with the use of model aids (0.323 vs 0.654, P<0.001).
Conclusion: The CART diagnostic model can identify ccRCC in SRMs and help radiologists make the diagnosis, potentially reducing the number of unnecessary biopsies.
Limitations: First, since it was a retrospective study, the existing selection bias may affect the results, and further prospective verification is needed in future work. Second, the data were collected from different institutions, and the scanning protocols were different. To minimise the impact of internal and external factors on the results, the quantitative data were standardised, and an independent external validation set was established.
Funding for this study: This work was supported in part by the National Key Research and Development Program of China under Grant Nos. 2023YFE0204300, in part by the National Natural Science Foundation of China under Grant No: 81801809, 82371917, 81830052, 81971691, 12126610, 62371476; in part by the Guangzhou Technology Program of Agriculture and Social Development of Key Research and Development Scheme under Grant No: 2023B03J1237, in part by the Basic and Applied Basic Research Foundation of Guangdong Province under Grant No: 2020A1515010572, and in part by the Zhuhai Basic and Applied Basic Research Foundation under Grant No: ZH22017003200001PWC.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the institutional review boards of the Fifth Hospital of Sun Yat- sen University.