Prognostic factors and predictive models in renal cell carcinoma: a contemporary review.

CONTEXT: The natural history of renal cell carcinoma (RCC) is highly unpredictable. Small renal masses may be accompanied by metastatic disease. Conversely, patients with locally advanced disease may enjoy long-term disease-free survival. OBJECTIVE: To review the status of prognostic factors in RCC. EVIDENCE ACQUISITION: A literature review was performed using the PubMed, MEDLINE, and Cochrane databases for articles published as of February 15, 2010. Electronic articles published ahead of print were also considered. Search was limited to the English language. Search was conducted using the following keywords: renal cell carcinoma, molecular, tissue, markers, blood, urine, progression, prognosis, risk factor, and survival. Studies were selected according to the relevance of the study, the number of patients included, originality, actuality, and clinical applicability of the results. EVIDENCE SYNTHESIS: Four areas of prediction were examined: (1) new RCC diagnostics, (2) RCC grade and stage at diagnosis, (3) disease progression, and (4) disease-specific mortality. All identified reports represented either case series or controlled studies. Although a large number of markers were identified, only a few were validated. Several prognostic factors were integrated in predictive or prognostic models. CONCLUSIONS: Several prognostic factors can help discriminate between favourable and unfavourable RCC phenotypes. Of those, several clinical, pathologic, and biologic markers have been tested and validated, and they are used in predictive and prognostic models. Nonetheless, the search continues, especially for informative markers predicting the response to targeted therapies.