Methods
All colorectal resections for cancer performed at the Surgical Clinic of the University of Brescia in the period 1994–2007 were retrospectively analyzed. Every patient entered into a scheduled program of clinical and instrumental follow-up. Our object was the comparison between patients under 50 years (Group A), and a sample of 85 randomly selected patients (which follow-up was recently updated for another study), more than 50 (included) years old (group B). The following parameters were recorded: age, sex, clinical presentation (symptomatic or not), operative treatment (resection of a colonic segment or extended surgery, including at least two segments or subtotal/total colectomy), cancer site (proximal to the transverse colon included, or distal), histotype (mucinous or not), staging (T1–2 or T3–4), grading (G1–2 or G3), disease-free and overall survival rates.
In group A, immunohistochemical analysis on histological samples derived from preoperative endoscopic biopsies and/or removed pieces was performed, to determine the loss of expression of 3 MMR proteins (MLH1, MSH2 and MSH6). After retrieving the paraffin blocks, 8 micron thick sections were obtained and stained with hematoxylin and eosin. The slides were placed in a microwave oven for 5 minutes in jet mode (full power) twice, and 3 more times for 5 minutes at 750 W power and lastly incubated for 40 minutes. Anti-MLH1 Mouse antibody clone G168–15 BD, at 1:30 dilution (Biosciences CompanyTM, Toronto, Ontario, Canada, code 551092), anti-MSH2 Mouse antibody, clone G216–1129, at 1:20 dilution (Biosciences CompanyTM, Toronto, Ontario, Canada, code 556349) and the Anti-MSH6 Mouse antibody, clone 44 (Zymed Laboratories Inc.TM, San Francisco, California, USA, code 08–1374) have been employed. For antigen unmasking the EDTA buffer at pH 8, 0.1 M was used. Data about MMR were expressed as dichotomous variables (yes/no), even though in 8 cases a diagnostic doubt was due to an insufficient but not entirely absent expression of the protein. The cases with preserved expression were considered normal, those without expression (as well as those with diagnostic doubt) were not. In the present work it was not taken into account what has been previously demonstrated: the absent expression of MSH6 is associated in 5% of cases with impaired quality (but with normal immunohistochemical expression) of MSH2, in which case it would be useful to proceed with its genetic analysis. Furthermore, in the present work we did not investigate micro satellite instability (MSI) and gene sequences of MMR.
Patients with complete or partial lack of expression (group A1) were compared to patients with preserved expression (group A2) of one or more MMR protein, concerning clinical, pathological and survival features. In patients belonging to group A1, the personal and family history was deeply assessed, up to the previous second generation (brethren, parents, uncles, grandparents and grandparent's brethren) and first subsequent generation (children and brethren's children), in accordance with the Amsterdam II criteria; for patients in which such informations were not clearly available from the analysis of medical supplies, telephone interviews were conducted. All the relatives in the 4 generations and the probands were investigated for every type of cancer, including previous and metachronous ones; colorectal polyps were considered as neoplasm if high-grade dysplasia was documented. All the interviews were done in 2010, from March to July, follow-up information being recorded up to 12/12/2009.
Statistical analysis was performed by Chi square test for dichotomous variables and Student's t test for continuous variables. Survival curves were constructed with the actuarial model, and comparison between curves was carried out by the method of Mantel-Haenzn.
This study was approved by the University Institutional Board Ethics Committee (Department of Clinical and Experimental Sciences).