OR01: Universal screening of colorectal cancers for Lynch syndrome by reflex immunohistochemistry – a single centre 9 year experience

E. Ryan 1, 2, B. Creavin 1, 2, R. Geraghty 1, D. Gibbons 1, S.T. Martin 1, R. Kennelly1, P.R. O’Connell 1, 2, K. Sheahan 1, 2, D.C. Winter 1, 2

1. Centre for Colorectal Disease, St. Vincent’s University Hospital, Elm Park, Dublin 4, Ireland
2. School of Medicine and Medical Sciences, University College Dublin, Belfield, Dublin 4, Ireland

Background

Up to 15% of colorectal cancer (CRC) exhibit microsatellite instability (MSI), where errors in replication go unchecked due to deficient mismatch repair (dMMR). Many institutions now advocate universal tumor screening via either polymerase chain reaction for MSI or immunohistochemistry (IHC) for dMMR to help identify Lynch syndrome. The majority of MSI is caused by MLH1 promoter hypermethylation and occurs sporadically. BRAF mutation may be used as a proxy marker for MLH1 promoter hypermethylation. We describe a series of 1753 consecutive CRCs subjected to reflex IHC for dMMR in our institution, and the resulting follow up with BRAF and germline testing.

Design

Retrospective analysis of the pathological features of CRCs identified on a prospectively maintained database at St. Vincent’s University Hospital (SVUH) from January 2004 to December 2013 inclusive. IHC was used to identify mismatch repair (MMR) status. Molecular analysis for the BRAFV600E mutation hotspot was carried out selectively on MLH1 deficient tumours. Tumours with absent MLH1 and who were BRAF mutation negative and tumours with absent MSH2, MSH6 or PMS2 were referred for genetic testing.

Results

Of the 1753 CRCs in the SVUH database during the study period 1753 had IHC for dMMR. 13.69% (n=183) exhibited dMMR. The median age was 69 (range 26-96). 22.4% (n=39) had a genetics referral during the study period. BRAF testing was undertaken in 29.5% (n=54) individuals. Once MLH1 absent tumours with BRAF mutation were excluded 27% (41/151) of patients with dMMR were referred for genetic testing. 1.2% (n=21) of the total CRCs screened were diagnosed with Lynch syndrome. 60.1% (n=110) of MMRd tumours were not referred. Those not referred or who declined further testing had a median age of 77 years.

Conclusions

Universal IHC for Lynch syndrome results in increased detection of dMMR tumours. BRAF testing can reduce the number of genetic referrals if performed in specimens with MLH1 loss. Additional testing for MLH1 hypermethlation on BRAF wild type tumours may distinguish sporadic from germline MSI cases. When Universal IHC for Lynch syndrome is introduced there is a need for appropriate planning and resources to fully investigate abnormal IHC resultsto ensure patients are appropriately referred for genetic testing.

Top