516P - An audit of screening for loss of mismatch repair (MMR) protein expression in colorectal cancers (CRC) in a tertiary referral centre

Date 30 September 2012
Event ESMO Congress 2012
Session Poster presentation II
Topics Pathology/Molecular Biology
Hereditary syndromes
Basic Scientific Principles
Presenter Karen Cadoo
Authors K. Cadoo1, S. Noonan2, N. McNicholas2, M..J. Kennedy3
  • 1Medical Oncology, St. James Hospital, Dublin, Dublin/IE
  • 2Medical Oncology, St. James Hospital, Dublin, Dublin/IE
  • 3St James's Hospital, IE-8 - Dublin/IE



Lynch syndrome accounts for 2% of CRC, resulting from germline mutations in DNA MMR genes; MSH2, MLH1, MSH6, PMS2 & PMS1. Mutations in MSH2 & MLH1 account for 85% of cases. It is imperative to identify Lynch syndrome patients to aggressively manage their risks. In 2009, the EGAPP working group recommended consideration of genetic testing in all new CRC. To this end, a number of institutions are reflex testing all CRC for loss of MMR protein expression by immunohistochemistry. This is a screening tool to select patients for germline testing as 15% of sporadic CRC will also demonstrate loss of expression. This approach was adopted in our institution in late 2008, with analysis for loss of MSH2, MLH1, MSH6 & PMS2 expression. This audit reviews the data from 2009–2010.


Patients were identified from CRC database & their records reviewed.


There were 271 patients; 9 (3%) had loss of MMR protein expression, 156 (58%) no loss, 106 (39%) not tested. Of the 106 not tested, 38 (36%) had a biopsy only, 15 (14%) had a polypectomy, 11 (10%) had significant response to neo-adjuvant chemoradiation with insufficient tumor for testing at resection, 6 (6%) had other reasons for not testing. 36 (13% of entire cohort) patients were not tested for unclear reasons. Of the 9 who demonstrated loss of expression, median age was 72yr (36–86) & 3 had family history of CRC. 2 patients were referred for genetic assessment, 1 has Lynch syndrome, 1 awaits the outcome of investigations. 1 elderly patient declined referral.


This audit highlights the challenges encountered with the initiation of this screening tool. A number of tumors were not tested, in 13% the reason is unclear, these are being reviewed. In 24% (biopsy & polypectomy & neoadjuvant groups) there was insufficient tumor for testing, raising questions about the validity of the screening test. Of the patients identified with loss, only 2 were referred for genetic assessment. In addition, a cohort of this size would expect to identify 5 Lynch patients, these may exist in the group who demonstrated loss of expression but were not appropriately referred or in the group whose tumors were not tested. Following this audit, processes are being put in place to increase the number of tumors tested and to streamline the further management of patients with a positive screening test.


All authors have declared no conflicts of interest.