AIM: to identify possible risk factors for the development of metachronous lesions in colorectal cancer (CRC) which would allow to establish a post-surgical individual prognostic index. PATIENTS AND METHODS: three hundred eighty-two surgically treated CRC were reviewed. We compared the incidence of metachronous lesions in 40 variables concerning patient clinical data and initial neoplastic findings. An individual risk index for metachronicity was drawn up including those variables which presented significant differences in multivariate logistic regression, dividing patients into three groups. RESULTS: variables with prognostic value for metachronicity were distal cancer location: OR = 2.30 (1.03-5.13), alcohol intake: OR = 2.20 (1.08-4.48), presence of synchronous adenomas: isolated: OR = 2.47 (1.03-4.48), multiple: OR = 4.26 (1.78-10.17), and presence of synchronous advanced adenoma: OR = 2.91 (1.52-12.60). Tumor MUC-5 expression proved to have a protective role: OR = 0.23 (0.08-0.66). An individual risk score was established considering these variables and patients could be classified into three groups, with a discrimination power for metachronicity of p < 0.0000001. Classification in high and low risk groups had a sensitivity = 75.32%, specificity = 84.21%, positive predictive value = 75.34%, negative predictive value = 92.31% and global diagnostic accuracy = 80.75%. CONCLUSIONS: the identification of risk factors for the development of metachronous lesions allow to calculate, at the time of surgical treatment, an individual prognostic index and to classify patients into three different risk groups. In high and low risk groups, both specificity and accuracy were acceptable for the prognosis of metachronous lesions, being remarkable the negative predictive power of our classification, which could become relevant when planning a different endoscopic follow up of these patients.
AIM: to identify possible risk factors for the development of metachronous lesions in colorectal cancer (CRC) which would allow to establish a post-surgical individual prognostic index. PATIENTS AND METHODS: three hundred eighty-two surgically treated CRC were reviewed. We compared the incidence of metachronous lesions in 40 variables concerning patient clinical data and initial neoplastic findings. An individual risk index for metachronicity was drawn up including those variables which presented significant differences in multivariate logistic regression, dividing patients into three groups. RESULTS: variables with prognostic value for metachronicity were distal cancer location: OR = 2.30 (1.03-5.13), alcohol intake: OR = 2.20 (1.08-4.48), presence of synchronous adenomas: isolated: OR = 2.47 (1.03-4.48), multiple: OR = 4.26 (1.78-10.17), and presence of synchronous advanced adenoma: OR = 2.91 (1.52-12.60). Tumor MUC-5 expression proved to have a protective role: OR = 0.23 (0.08-0.66). An individual risk score was established considering these variables and patients could be classified into three groups, with a discrimination power for metachronicity of p < 0.0000001. Classification in high and low risk groups had a sensitivity = 75.32%, specificity = 84.21%, positive predictive value = 75.34%, negative predictive value = 92.31% and global diagnostic accuracy = 80.75%. CONCLUSIONS: the identification of risk factors for the development of metachronous lesions allow to calculate, at the time of surgical treatment, an individual prognostic index and to classify patients into three different risk groups. In high and low risk groups, both specificity and accuracy were acceptable for the prognosis of metachronous lesions, being remarkable the negative predictive power of our classification, which could become relevant when planning a different endoscopic follow up of these patients.
Authors: Charles J Kahi; C Richard Boland; Jason A Dominitz; Francis M Giardiello; David A Johnson; Tonya Kaltenbach; David Lieberman; Theodore R Levin; Douglas J Robertson; Douglas K Rex Journal: Am J Gastroenterol Date: 2016-02-12 Impact factor: 10.864