Independent prognostic value of eosinophil and mast cell infiltration in colorectal cancer tissue

Overall peritumoural inflammatory cell infiltration is a prognostic variable in solid tumours, but the survival-related impact of the individual cell types within the infiltrate has still not been fully evaluated and compared with the conventional disease classification. In the present study, the prognostic value of individual white cell counts in the peritumoural inflammatory infiltrate in colorectal cancer was assessed. Intra-operative tumour tissue samples from 584 patients undergoing elective surgery for colorectal cancer were included. None of the patients received pre- or post-operative adjuvant chemotherapy. Tissue blocks were cut from the periphery of the tumours and embedded in paraffin. All blocks included both tumour tissue and normal bowel tissue. Serial sections of 4 microm were analysed for tumour tissue inflammatory cell infiltration using a computer- and video-assisted microscope, which allowed semi-automated quantification of cells within a fixed area. Total white cells and individual counts of eosinophils, neutrophils, mast cells, lymphocytes, and plasma cells were evaluated in every tumour specimen. Stratification into four groups with similar numbers of events was used to dichotomize the cell counts with respect to survival. The median observation period was 61 (49-75) months. In a multivariate analysis including Dukes' stage, gender, age, peri-operative blood transfusion, tumour location, and counts of specific inflammatory cells, only advanced Dukes' stage ( p< 0.0001), high age ( p=0.0003), and tumour location in the rectum predicted poor survival, while high counts of eosinophils ( p=0.006) and mast cells ( p=0.02) predicted good survival. Tumour-associated eosinophilia and mastocytosis appear to be independent prognostic variables in colorectal cancer. Future studies should investigate the potential biological role of tumour tissue eosinophils and mast cells in the modulation of tumour growth.