OBJECTIVE: To evaluate whether assessing the anatomical characteristics of renal masses increases the accuracy of prediction of tumour pathology in small renal masses (SRMs). PATIENTS AND METHODS: We retrospectively reviewed 1129 consecutive patients who underwent extirpative surgeries for a clinical T1 renal mass, for which the preoperative aspects and dimensions used for an anatomical (PADUA) classification were available. Multivariate logistic regression analyses of demographic and anatomical characteristics were performed. Nomograms to predict malignancy and high grade pathology were constructed using a basic model (age, sex and tumour size), and an extended model (anatomical characteristics incorporated into the basic model), and the area under the curve (AUC) between models was compared. RESULTS: Age, sex and tumour size were significantly associated with malignancy and high grade pathology in the T1 and T1a category (except sex for high grade pathology in T1a tumours). Exophytic rate (T1 and T1a) and renal sinus or urinary collecting system involvement (only T1a) were also significant predictors of high grade pathology. Nomograms using the extended model for malignancy showed an insignificant AUC increase compared with those using the basic model (T1, from 0.771 to 0.780, P = 0.149, and T1a, from 0.803 to 0.819, P = 0.055). For high grade pathology, the extended model achieved a significant AUC increase (from 0.595 to 0.643, P = 0.014) in the T1a category, but the AUC for both T1 and T1a tumours showed merely modest competence (0.654 and 0.643, respectively). CONCLUSION: Age, sex and tumour size are the primary predictors of tumour pathology of SRMs, and incorporating other anatomical characteristics has only a limited positive effect on the accuracy of prediction of pathological outcomes.
OBJECTIVE: To evaluate whether assessing the anatomical characteristics of renal masses increases the accuracy of prediction of tumour pathology in small renal masses (SRMs). PATIENTS AND METHODS: We retrospectively reviewed 1129 consecutive patients who underwent extirpative surgeries for a clinical T1 renal mass, for which the preoperative aspects and dimensions used for an anatomical (PADUA) classification were available. Multivariate logistic regression analyses of demographic and anatomical characteristics were performed. Nomograms to predict malignancy and high grade pathology were constructed using a basic model (age, sex and tumour size), and an extended model (anatomical characteristics incorporated into the basic model), and the area under the curve (AUC) between models was compared. RESULTS: Age, sex and tumour size were significantly associated with malignancy and high grade pathology in the T1 and T1a category (except sex for high grade pathology in T1a tumours). Exophytic rate (T1 and T1a) and renal sinus or urinary collecting system involvement (only T1a) were also significant predictors of high grade pathology. Nomograms using the extended model for malignancy showed an insignificant AUC increase compared with those using the basic model (T1, from 0.771 to 0.780, P = 0.149, and T1a, from 0.803 to 0.819, P = 0.055). For high grade pathology, the extended model achieved a significant AUC increase (from 0.595 to 0.643, P = 0.014) in the T1a category, but the AUC for both T1 and T1a tumours showed merely modest competence (0.654 and 0.643, respectively). CONCLUSION: Age, sex and tumour size are the primary predictors of tumour pathology of SRMs, and incorporating other anatomical characteristics has only a limited positive effect on the accuracy of prediction of pathological outcomes.