5th Annual BMRP Investigator Meeting - Abstract
Clinical, Pathological and Genotypic Variations in UC-CRC: Preliminary Findings
Megan M. Garrity-Park1,b, William J. Sandborn1, Edward V. Loftus, Jr.1, S. Breanndan Moore2 and Thomas C. Smyrk3,a
1Department of Medicine, Division of Gastroenterology and Hepatology, 2 Department of Lab Medicine and Pathology, Division of Transfusion Medicine, and 3Department of Lab Medicine and Pathology, Division of Anatomic Pathology, Mayo Clinic Rochester (Minnesota, U.S.A.)
Introduction: Patients with chronic ulcerative colitis (CUC) are at increased risk for developing colorectal cancer (CRC). Although this risk has been associated with increased duration (≥10 years) and extent of disease (panulcerative), many patients with similar duration and extent of CUC do not develop CRC. We therefore investigated what puts certain patients at greatest risk for CUC-CRC by examining HLA (immune regulatory) as well as TNF-α (inflammatory mediator) genes. Methods: We reviewed the medical chart on 270 CUC-CRC patients seen at the Mayo Clinic (1980-2006) to confirm that the date of CUC diagnosis was ≥10 years prior to CRC. Slides were recalled on all qualifying cases and reviewed to eliminate Crohns and indeterminate colitis. This yielded 114 CUC-CRC cases for this study. CUC patients who did not develop CRC were then identified as controls. A large database was created to catalog all clinical, demographic, social, familial and medical history as well as the pathology of the archived colon specimen for future study comparisons. Sections were cut for immunohistochemistry (IHC) and genomic DNA extraction. IHC was performed following standard protocols. A genomic DNA extraction technique was developed to yield the best quality and quantity of DNA possible from formalin fixed, paraffin embedded tissues. HLA typing using a commercially available SSO kit was performed after extensive optimization to allow for analysis of DNA extracted from tissues archived for >20 years. TNF-α primers were designed to analyze five known promoter single nucleotide polymorphisms (SNP) as well as to investigate the areas surrounding each SNP. Results: CUC-CRC cases and CUC-no CRC controls were matched with regard to age, race, gender, duration and extent of disease (all p>0.05). Loss of HLA-DR expression was found in 52% of tumors with the majority of those (86%) showing complete loss. Multiple aggressive pathologies were observed in CUC-CRC tumors including mucinous and serrated lesions both adjacent to and distant from the primary tumor. Mucinous pathology was associated with a decrease in HLA-DR expression and increased frequency of the DR1 allele. When compared to CUC-no CRC controls, CUC-CRC cases were more likely to be carriers of the DR17 or DR13 alleles but were less likely to carry the DQ5 allele. Although analysis is ongoing, it appears that at least one TNF-α SNP will be significantly different between the CUC-CRC and CUC-no CRC groups. Conclusions: This is the first study to indicate that two Chromosome 6 genes (TNF and HLA) may be involved in predisposing some CUC patients to CRC. Furthermore, the frequency of HLA-DR loss in CUC-CRC tumors lends insight into the mechanism of CUC-CRC carcingenesis in general. Finally, because CUC is a heterogeneous disease, requiring careful study design and patient selection, the database and DNA bank we created will provide an excellent resource for future investigations into CUC-CRC.
aPrincipal Investigator, bCo-Investigator and Presenter