Surgical quality surrogates do not predict colon cancer survival in the setting of technical credentialing: A report from the prospective COST trial

OBJECTIVE: We tested the hypothesis that the 12 lymph node (LN) count and other surgical variables would not predict survival in a setting where surgical techniques were standardized and surgeons were credentialed and audited. BACKGROUND: The National Quality Forum has adopted the 12-node minimum as a surgical quality metric due to the strong association between node count and survival. METHODS: We performed a secondary analysis of data from the Clinical Outcomes of Surgical Therapy (COST) multicenter randomized trial testing laparoscopic versus open colectomy. Surgeons were credentialed and video-audited for adherence to technical standards. Patients with noninvasive and stage IV disease were excluded, leaving 787 subjects (267 stage I, 284 stage II, and 236 stage III). Median age was 70 years and 50% were male. The overall 5-year survival was 77.2%. RESULTS: Five-year overall and disease-free survival was not influenced by LN count (< 12 vs 12), sex, tumor location (right vs left vs sigmoid), surgical technique (laparoscopic vs open), total bowel length, proximal margin, distal margin, radial margin, or mesenteric length (P > 0.05 for all). Univariate predictors of survival included age and cancer stage, and these remained significant in the multivariate model. Across all stages of disease, after adjusting for age and stage, LN count did not predict overall or disease-free survival (P = 0.60). CONCLUSIONS: Despite the known association between LN count and survival, we could not confirm an association between surgical surrogates and cancer outcomes. We postulate that standardization, credentialing, and monitoring may be more important than traditional surgical quality surrogates.