Objective: To identify clinical variables that can accurately predict the presence of distant metastases in patients with renal cell carcinoma (RCC).
Patients and methods: Age, symptom classification, tumour size and the prevalence of distant metastases at diagnosis before nephrectomy were available for 5376 patients with pathologically confirmed RCC. The data of 2660 (49.5%) patients from 11 centres were used to develop a multivariable logistic regression model-based nomogram predicting the individual probability of distant metastases. The remaining data from 2716 (50.5%) patients from three institutions were used for external validation.
Results: In the development cohort, 269/2660 (10.1%) had distant metastases, vs 285/2716 (10.5%) in the external validation cohort. Symptom classification and tumour size were independent predictors of distant metastases in the development cohort; age was not an independent predictor. A nomogram based on symptom classification and tumour size was 85.2% accurate in predicting the individual probability of distant metastases in the external validation cohort.
Conclusion: Although distant metastases might be easily identifiable in some patients, their diagnosis might be a challenge in others. The current nomogram provides a simple, user-friendly and, most importantly, an accurate tool aimed at predicting the probability of distant metastases in patients with RCC.