Introduction: Functional volume loss (FVL) is a significant predictor of kidney function decline after partial nephrectomy (PN). Here, we sought to assess two different methods for quantifying FVL post-PN: imaging-based tissue segmentation (TS) vs pathological analysis.
Methods and results: From a single surgeon series, we performed a retrospective analysis of 42 patients who underwent PN for a cT1 renal mass between 2015 and 2017. The association between TS and pathological analysis at a median follow-up of 6 months (range: 3-9 months) was evaluated using Spearman's correlation. The association between pathological analysis, TS analysis, and estimated glomerular filtration rate (eGFR) decline at 6 months was evaluated using a multivariable linear mixed-effects models. For pathological analysis, dimensions of the specimen and tumor were extracted from pathology reports. FVL was calculated as [specimen volume (Length*Width*Height*π/6) - tumor volume (Length*Width*Height* π/6)]. For TS analysis, preoperative cross-sectional imaging was used (MRI n = 20; CT n = 22). FVL was calculated as [(overall kidney volume) - (tumor volume) - (cyst volume of renal cysts >1 cm)]. Postoperative functional volume was subtracted from preoperative functional volume to assess FVL post-PN for TS method.
Results: eGFR significantly decreased from baseline to postoperative 6 months (-5.1 mL/min/1.73 m2; p = 0.004). Even though there was a correlation between the two methods (coefficient = 0.245, p < 0.001), pathological analysis underestimated volume loss (32.2 mL vs 5.76 mL, p < 0.001). In multivariate linear regression analysis, TS analysis was significantly associated with a decline in eGFR (β = 0.084, 95% CI = -0.02, 0.15; p = 0.012), whereas pathological analysis was not (β = 0.02, 95% CI = -0.24, 0.28; p = 0.87).
Conclusion: Pathological analysis underestimates parenchymal volume loss. Only imaging-based TS method is associated with change in eGFR post-PN.
Keywords: functional outcome; partial nephrectomy; tumor volume; volume assessment.