Confirmation of deficient mismatch repair (dMMR) as a predictive marker for lack of benefit from 5-FU based chemotherapy in stage II and III colon cancer (CC): A pooled molecular reanalysis of randomized chemotherapy trials

4008 Background: Patients (pts) with CC demonstrating high-frequency MSI (MSI-H) have a stage-independent improved survival compared to pts with microsatellite- stable (MSS) tumors. We have previously published that MSI status is a predictive marker for lack of response to 5-FU-based chemotherapy (rx) (Ribic, NEJM 2003). dMMR by immunohistochemisty for MMR proteins is an almost perfect predictor for MSI status. We sought to confirm the value of dMMR as a predictor of survival benefit from adjuvant rx in stage II & III CC pts in an independent dataset drawn from randomized clinical trials. Methods: MSI or IHC analyses were performed on tumors from pts enrolled in trials conducted by the NCCTG (N=135), GIVIO (N=183), and ECOG (N=23) that have not been used in previous MSI or dMMR analyses. All trials randomized pts with stage II and III CC to either 5-FU based rx (either 5-FU + levamisole or 5-FU + leucovorin, n=166)) or no post-surgical rx (n=175). XX microsatellite loci from the National Cancer Institute panel were used to amplify genomic DNA and determine MSI status (GIVIO); IHC testing for hMLH1 and hMSH2 was used for the NCCTG and ECOG trials. Patients with MSI-H tumors or negative IHC staining were classified as dMMR;, the remainder were considered to have proficient MMR (pMMR). Median follow-up on living pts was 6.4 years with a primary outcome of overall survival (OS). Results: 341 tissue specimens were examined, of which 47 (13.8%, 20 treated, 27 untreated) exhibited dMMR. Adjuvant rx had a significant beneficial effect on OS (HR = 0.69, p= 0.047) and DFS (HR = 0.59, p = 0.004) in pts with pMMR tumors. However, pts with dMMR tumors receiving 5-FU rx had no trend toward improved OS (HR = 1.26, p = 0.68) or DFS (HR=1.41, p = 0.53) compared to those randomized to no rx. Results were maintained in multivariate models adjusted for stage and age. Conclusions: Stratification of pts according to MMR status provides a more tailored approach to the use of adjuvant rx in CC. Our data suggest that in a patient being considered for 5-FU alone rx (i.e a stage II pt), MMR status should be assessed and considered in rx decision making. No significant financial relationships to disclose.