Abstract
BACKGROUND: The tumor stage, size, grade, and necrosis (SSIGN) score was originally defined using patients treated with radical nephrectomy (RN) between 1970 and 1998 for clear cell renal cell carcinoma (ccRCC), excluding patients treated with partial nephrectomy (PN). OBJECTIVE: To characterize the original SSIGN score cohort with longer follow-up and evaluate a contemporary series of patients treated with RN and PN. DESIGN, SETTING, AND PARTICIPANTS: Retrospective single-institution review of 3600 consecutive surgically treated ccRCC patients grouped into three cohorts: original RN, contemporary (1999-2010) RN, and contemporary PN. INTERVENTION: RN or PN. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: The association of the SSIGN score with risk of death from RCC was assessed using a Cox proportional hazards regression model, and predictive ability was summarized with a C-index. RESULTS AND LIMITATIONS: The SSIGN scores differed significantly between the original RN, contemporary RN, and contemporary PN cohorts (p<0.001), with SSIGN ≥4 in 53.5%, 62.7%, and 4.7%, respectively (p<0.001). The median durations of follow-up for these groups were 20.1, 9.2, and 7.6 yr, respectively. Each increase in the SSIGN score was predictive of death from RCC (hazard ratios [HRs]: 1.41 for original RN, 1.37 for contemporary RN, and 1.70 for contemporary PN; all p<0.001). The C-indexes for these models were 0.82, 0.84, and 0.82 for original RN, contemporary RN, and contemporary PN, respectively. After accounting for an era-specific improvement in survival among RN patients (HR: 0.53 for contemporary vs original RN; p<0.001), the SSIGN score remained predictive of death from RCC (HR: 1.40; p<0.001). CONCLUSIONS: The SSIGN score remains a useful prediction tool for patients undergoing RN with 20-yr follow-up. When applied to contemporary RN and PN patients, the score retained strong predictive ability. These results should assist in patient counseling and help guide surveillance for ccRCC patients treated with RN or PN. PATIENT SUMMARY: We evaluated the validity of a previously described tool to predict survival following surgery in contemporary patients with kidney cancer. We found that this tool remains valid even when extended to patients significantly different than were initially used to create the tool.