Introduction:

Partial nephrectomy (PN) has emerged as the reference standard management of localized renal masses, yet the use of PN in T3a renal cell carcinoma (RCC) is controversial, due to concerns about potential of increased procedure-related complications and oncological risk for T3a renal masses. Herein we sought to evaluate outcomes of patients treated with partial and radical nephrectomy (RN) in pT3aN0M0 RCC utilizing a propensity score match model. 

Methods:

We conducted a multi-institutional retrospective analysis of pT3aN0M0 RCC patients who underwent PN or RN. Type of surgery performed was made by individual surgeons in the context of shared decision making. Patients with clinical metastasis at time of surgery were excluded. A propensity score match (PSM) in a 1:1 ratio was conducted, within a caliper width of 0.1, using age at surgery, sex, body mass index (BMI), tumor histology, tumor size, cT stage, presence of positive surgical margin (SM), and tumor necrosis. Primary outcome was all-cause mortality/overall survival (ACM/OS), and secondary outcomes were cancer-specific mortality/ cancer-specific survival (CSM/CSS), recurrence/progression free survival (PFS) and new onset de novo eGFR<45mL/min/1.73m² (CKD-S3b). Kaplan Meier analysis (KMA) and multivariable analysis (MVA) via Cox regression were fitted to delineate survival outcomes and their predictors. Multivariable logistic regression (MLR) was fitted to elucidate predictors of CKD-S3b at last follow up. 

Results:

After PSM 345 patients, were matched to the RN [n=161 (46.6%)] and the PN [n=184 (53.3%)] group; median follow up time 36 (IQR 14-70.5) months. MVA revealed, age [hazard ratio (HR)1.03, p=0.022], tumor necrosis (HR 3.79, p<0.001), SM (HR 2.97, p=0.015), cT3 vs. cT1-T2 (HR 2.32, p=0.048) and intraoperative transfusion (HR 2.35, p=0.013) associated with ACM, while RN was not (p=0.1). MVA for CSM revealed, cT3 vs. cT1-T2 (HR 3.80, p=0.025), and SM (HR 4.49, p=0.027) associated with CSM, while RN was not (p=0.1). MVA for recurrence revealed, SM (HR 2.59, p=0.015) as independent risk factor, while RN was not (p=0.4). KMA comparing PN vs. RN revealed 5-year OS of 87.5% vs. 81.3% (p=0.2, Figure); 5-year CSS of 96.1% vs. 89.9% (p=0.09, Figure) and 5-year PFS of 76.4% vs. 73.3% (p=0.9, Figure). MLR revealed RN (OR 2.71, p=0.003 ) associated with development of new onset CKD-S3b.

Conclusion:

In a PSM model of pT3aN0M0, PN exhibited oncological equipoise while reducing risk of development of eGFR<45mL/min/1.73m.²  PN may be considered as an alternative to RN in T3a RCC when prioritization of functional preservation is indicated. Further investigation is requisite. 

Funding: N/A