Published in:
01-03-2016 | Original Article
Inclusion of cytological features in tumor grading improves prognostic stratification of patients with colorectal cancer
Published in: International Journal of Colorectal Disease | Issue 3/2016
Login to get accessAbstract
Purpose
Tumor grade is a traditional prognostic parameter in colorectal cancer. Remarkably, however, there is still no generally accepted consensus how to perform tumor grading. In this study, we systematically compared the prognostic value of traditional grading based upon histological features, that is, gland formation alone with grading based upon both histological and cytological features, such as nuclear pleomorphism and anaplasia (“alternative grade”).
Methods
Three hundred eighty-one tumors of randomly selected patients were retrospectively reviewed. Traditional and alternative tumor grades were related to various clinicopathological features and to progression-free and cancer-specific survival applying both univariate and multivariate testing.
Results
Traditional and alternative tumor grades were significantly associated with T and N classification, tumor size, lymphovascular invasion, as well as both progression-free and cancer-specific survival. In Cox’s proportional hazards regression models, the alternative grade was superior to the traditional tumor grade and was significantly associated with progression-free survival (hazard ratio 1.57, 95 % confidence interval 1.04–2.35; p = 0.031), independent of patients’ age and gender, T and N classification, and lymphovascular invasion. Likewise, patients with tumors with high alternative grade were more likely to die of disease (hazard ratio 1.30, 95 % confidence interval 0.85–2.00), but this difference was not statistically significant (p = 0.22).
Conclusions
Tumor grade based upon both histological and cytological features was superior to grade based upon histological features alone and proved to be an independent prognostic parameter. Thus, tumor grade based upon both histological and cytological features may help to improve prognostic stratification and may thereby affect clinical decision-making and patient management.