Background: A recent multi-center study showed how estimated glomerular filtration rate (eGFR) and cancer-specific mortality (CSM) are linearly and inversely related in organ-confined renal cell carcinoma (RCC) whenever the eGFR decreases below specific thresholds. We addressed our previous work limitations related to heterogeneity and missing data, and explored the relationship between eGFR and CSM also in locally advanced RCC.
Materials and methods: All patients with RCC treated with either partial or radical nephrectomy from 1990 to 2018 at a single institution and with complete data on renal function were included. eGFR was managed as a time-dependent variable. The relationship between eGFR and CSM was analyzed using a Fine and Gray multivariable competing risks framework. Subdistribution hazard ratios (SHRs) were calculated accounting for deaths from other causes.
Results: Multivariable competing risks analysis showed a "piecewise" relationship between eGFR and CSM, with an inverse linear correlation for eGFR values below 85 mL/min. Below this breakpoint, a significant relationship existed between eGFR and CSM in both clinical (SHR, 1.27; P < .001) and pathologic (SHR, 1.27; P = .001) models in stage I to II RCC subgroup. Conversely, no significance was recorded in this subgroup when considering eGFR values above 85 mL/min. In the stage III to IV subgroup, no significant relationships were recorded, regardless of eGFR values. The retrospective design with inherent biases in data collection represents a limitation.
Conclusions: In patients undergoing surgery for stage I to II RCC, preservation of renal function over "safety limits" is protective from CSM.
Keywords: Cancer specific mortality; Estimated glomerular filtration rate; Partial nephrectomy; Prognosis; Radical nephrectomy.
Copyright © 2019 Elsevier Inc. All rights reserved.