Abstract
BACKGROUND: MLH1, one of the MMR proteins, is linked to DNA replication, its role being to repair the incorrect DNA sequences and to replace them with proper ones. The loss of the MLH1 gene expression is part of Lynch syndrome which can lead to a series of cancers like colorectal and endometrial ones. The aim of this paper is to correlate the levels of MLH1 in four different bio-logical fluids with clinicopathological features in colorectal cancer patients in order to predict them with high probability. Therefore, a mathematical model with given code in Matlab has been proposed to get the clinicopathological features with high probability by only introducing the values for MLH1 concentrations. All these data can be obtained in a very short time even before surgery which can be very helpful the surgeon and the oncologist. METHODS: Four types of samples (whole blood, saliva, urine and tissue) were analyzed using stochastic microsensors; concentrations of MLH1 were determined and compared with different macroscopic and micro-scopic pathological features to obtain mathematical models for early, non-invasive diagnostic of colorectal adenocarcinoma. RESULTS: There have been established criteria and mathematical models for tumor location, TNM grading system, depth of the tumor, lymphatic, vascular and perineural invasions and the presence of mucus in the tumoral mass. CONCLUSIONS: By using whole blood, saliva and urine samples, the location can be approximated. The proposed mathematical models aimed to allow a minim/noninvasive characterization of the tumor and its location which can help the surgeon and the oncologist to choose faster the personalized treatment.