Objective: Tumor grade employed for colorectal cancer has long been based on the degree of differentiation, which is difficult to judge objectively. The aim of this study was to determine whether the extent of the poorly differentiated component (POR) could be a valuable criterion for a grading system.
Patients and methods: A total of 1075 patients with advanced colorectal cancer were pathologically reviewed. POR was newly defined as a region in which a cancer has no glandular formation, irrespective of a mucin-producing or invasive pattern, and we quantitatively classified the POR into 6 degrees using the microscopic field of an objective lens as a standard.
Results: Survival analyses of the extent of POR demonstrated that a 3-category grading system provides the most efficient survival stratification. Grade III was applied to tumors (n = 339) for which the POR fully occupied the microscopic field of a 40x objective lens. For tumors having a smaller POR, cancer clusters without a gland structure composed of > or = 5 cancer cells ("clusters") were counted in the microscopic field of a 4x objective lens, where "clusters" were observed most intensively. Tumors with < 10 "clusters" were classified as grade I (n = 161), and those with > or = 10 "clusters" as grade II (n = 575). Patients classified as grade I demonstrated a very favorable prognosis, with a 99.3% cancer-related 5-year survival rate, whereas the survival was 86.0% for grade II and 68.9% for grade III (P < 0.0001 in each group). Multivariate analysis demonstrated that the grades of POR function as an independent prognosticator, as do T-stage and N-stage.
Conclusions: The grading system utilizing POR is distinctive in terms of the simplicity of judgment based on its quantification and the ability to determine which patients will likely be cured by surgery alone. It will aid in selecting postoperative treatment strategies.