POPULATION-BASED COMPARISON OF CANCER SPECIFIC MORTALITY AFTER LOCAL TUMOR ABLATION OR NON-ACTIVE TREATMENT FOR T1A KIDNEY CANCER: A COMPETING RISK ANALYSIS

INTRODUCTION & OBJECTIVES: The management of non-surgical candidates diagnosed with kidney cancer includes local tumor ablation[LTA] and absence of local treatment [ALT]. However, data regarding cancer-specific mortality [CSM] after LTA or ALT are lacking. Ourhypothesis stated that LTA might result in a lower cancer specific mortality relative to [ALT] after adjustment for competing other-causemortality [OCM].MATERIAL & METHODS: A Surveillance Epidemiology and End Results-Medicare-linked retrospective cohort of 1860 T1a N0 M0 kidneycancer patients who underwent LTA or ALT between 2000 and 2009 was abstracted. ALT was defined as absence of either LTA or anyother surgical treatment during the first 6 months after diagnosis. The outcome of the study was CSM defined as death due to kidneycancer. Analyses consisted of propensity-score matching to reduce potential measured differences between LTA and ALT patients.Subsequently, regression models [MVA] adjusted for competing risk of OCM, patient (age) and cancer characteristics (tumor size, histologyand grade) were fitted to address the impact of LTA vs. ALT on CSM.RESULTS: Median follow-up was 30 (IQR 14-52) months among survivors. Fewer patients (30%; n=553) were treated with LTA, while most(70%; n=1307) received ALT. With respect to ALT patients, individuals treated with LTA were younger (77 vs. 78; p<0.001), more frequentlywhite (84 vs. 78%; p=0.005), more frequently married (59 vs. 52%; p=0.02) and more frequently of high socio-economic status (54 vs. 45%p=0.001). In patients treated with LTA, surgical approach was open, laparoscopic and percutaneous in 6, 52 and 42% of the cohort,respectively. Following a 1:1 ratio propensity score matching for all the covariates, 553 LTA and 553 ALT patients remained. The meanstandardized differences of patient characteristics between the two groups were <10%, indicating a high degree of similarity in thedistribution of all the covariates in both populations. All subsequent analyses were based on the post-propensity score matched cohort. Inpatients treated with LTA, the 5-years CSM and OCM rates were 3.7 and 22%, respectively. In patients treated with ALT the 5-years CSMand OCM rates were 11 and 37%, respectively. After adjustment for patient and cancer characteristics and for competing risk of OCM, LTAwas associated with lower risk of CSM (HR 0.52; CI 0.27-0.98; p=0.045). Conversely, age (HR 1.04; CI 1-1.08; p=0.049), and tumor size(HR 1.04; CI 1-1.08; p=0.03) were associated with higher risk of CSM.CONCLUSIONS: OCM represents the leading cause of death in non-surgical candidates diagnosed with T1a kidney cancer. Afteradjustment for competing OCM, and for patient and cancer characteristics, LTA results in lower CSM compared to ALT. Despite ourattempt to control for all the measurable baseline differences between LTA and ALT patients, a limitation of the study could be related topotential non-measurable bias for treatment selection.