Skip to main content
. Author manuscript; available in PMC: 2011 Jan 1.
Published in final edited form as: Eur Urol. 2009 Sep 9;57(1):12–20. doi: 10.1016/j.eururo.2009.09.013

Fig. 4. Recursive partitioning analysis for outcome prediction.

Fig. 4

(a) CCND3 expression level with a cut-off value of 0.4712 was identified as a predictor for recurrence. Thirty-one of 32 patients with recurrence and 10 of 16 patients without recurrence were correctly identified. (b) Expression levels of HRAS, VEGFR2, and VEGF were identified as predictors for progression, with cut-off values of 0.3369, 0.5424, and 0.966, respectively. Thirteen of 16 patients with progression and 30 of 32 patients without progression were correctly identified.