Abstract
Objective:
Repair of renal collecting system (CS) during laparoscopic partial nephrectomy (LPN) requires advanced skills and nevertheless prolongs renal ischemia time. We assessed tumor parameters that may predict CS transection and thus improve preoperative planning.
Methods:
Data were prospectively collected for 184 consecutive LPN cases performed at our institution between 2002 and 2008 by a single senior surgeon. Twelve patients were excluded because of open conversion and seven because of missing data. Among the rest (n = 165), CS was transected in 115 (61%). Tumor parameters (radiographic appearance—solid vs. cystic, size, polar location, and depth) were evaluated with univariate and multivariate logistic regression analysis. Classification and Regression Tree analysis was applied to define the optimal cutoff for tumor size.
Results:
In univariate analysis, tumor size (odds ratio [OR] 2.8, 95% confidence interval [CI] 1.8, 4.3; p < 0.0001) and tumor appearance (solid: OR 2.1, 95% CI 1.1, 4.3) achieved statistical significance, while tumor depth (endophytic: OR 3.1, 95% CI 0.8, 11.0; p = 0.08) trended toward significance. In multivariate analysis, size (p < 0.0001) and solid tumor appearance (p = 0.006) were independent predictors. In Classification and Regression Tree analysis, 2.5 cm was found to be the optimal cutoff for the tumor size.
Conclusions:
The odds of CS transection during LPN triple with each additional centimeter in tumor size, are 10-fold higher for tumors >2.5 cm, and are almost twice higher for solid tumors, compared with cystic. These findings may be useful in LPN planning.