|| Checking for direct PDF access through Ovid
The average size of blood platelets determined by mean platelet volume might represent a biologically meaningful parameter in carcinogenesis and potentially serve as a novel prognostic biomarker in renal cell carcinoma.In this retrospective analysis of the records of 652 patients we evaluated the potential prognostic value of mean platelet volume and its ability to improve existing risk assessment tools used in adjuvant clinical trials in nonmetastatic renal cell carcinoma cases. Associations of mean platelet volume with baseline covariates and clinical outcomes (recurrence, and death from renal cell carcinoma and other causes) were assessed with the competing risk estimators of Kaplan-Meier, and Marubini and Valsecchi, respectively. Univariable and multivariable Cox proportional hazard models were constructed. The Harrell c-index was applied to test improvements in the predictive accuracy of the established Leibovich prognosis score.Small platelet volume was associated with large tumors (p = 0.043), high Fuhrman grade (p = 0.001), sarcomatoid components (p <0.0001), histological tumor necrosis (p = 0.044) and vascular invasion (p = 0.022). On univariable and multivariable analyses small platelet volume accurately predicted recurrent renal cell carcinoma (continuously and binary coded) and cancer specific survival. Adding mean platelet volume to the Leibovich prognosis score improved its discriminative performance (c-index = 0.83, p = 0.004).Mean platelet volume represented a highly significant predictor of recurrence and cancer specific death in patients with renal cell carcinoma. This parameter improved the accuracy of the Leibovich prognosis score to better predict long-term outcomes in localized renal cell carcinoma cases after curative surgical resection.