LAUSR

BACKGROUND Historically, partial nephrectomy (PN) showed no benefit on other-cause mortality (OCM) in elderly patients with small renal masses. OBJECTIVE To test the effect of PN versus radical nephrectomy (RN) on OCM, cancer-specific mortality (CSM), as well as 30-d mortality in patients with nonmetastatic T1a renal cell carcinoma (RCC), aged ≥75 yr old. DESIGN, SETTING, AND PARTICIPANTS Within the Surveillance, Epidemiology and End Results registry (2004-2014), we identified surgically treated patients with nonmetastatic pT1a RCC aged ≥75 yr. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS We relied on propensity score (PS) matching to reduce the effect of inherent differences between PN and RN. After PS matching, cumulative incidence, multivariable competing-risks regression (CRR) and logistic regression models were used. LOESS plots graphically depicted the relation between nephrectomy type and OCM after adjustment for all the covariates. Landmark analyses at 6 mo tested for immortal time bias. RESULTS AND LIMITATIONS Of all 4541 patients, 41.6% underwent PN. After 1:1 PS matching, 2826 patients remained. In multivariable CRR models, lower OCM rates were recorded in PN patients (hazard ratio [HR]: 0.67, confidence interval [CI]: 0.54-0.84; p<0.001). LOESS plots showed lower OCM rates after PN across all examined ages. Lower CSM rates were also recorded in PN patients (HR: 0.64, CI=0.44-0.92; p=0.02). Landmark analyses rejected the hypothesis of immortal time bias. Finally, PN did not result in different 30-d mortality rates (odds ratio: 1.87; CI: 0.79-4.47; p=0.2) versus RN. Data are retrospective. CONCLUSIONS PN results in lower OCM in elderly patients with pT1a RCC. Moreover, PN does not contribute to higher CSM or 30-d mortality in patients aged ≥75 yr. In consequence, PN should be given strong consideration, even in elderly patients. PATIENT SUMMARY Partial nephrectomy (PN) may protect from renal insufficiency, hypertension, and other unfavorable health outcomes, even in elderly patients. This protective effect results in lower other-cause mortality. Moreover, PN benefits are not undermined by higher cancer-specific mortality or 30-d mortality.