Low Expression of Two Genes Could Predict Response to Frontline Oxaliplatin in mCRC

June 23, 2015

Low expression of the genes ERCC1 and TS was associated with substantial improvements in overall survival following front-line treatment with an oxaliplatin-based chemotherapy regimen for patient with metastatic colorectal cancer.

John Paul Shen, MD

Low expression of the genesERCC1andTSwas associated with substantial improvements in overall survival (OS) following front-line treatment with an oxaliplatin-based chemotherapy regimen for patient with metastatic colorectal cancer (mCRC), according to a 41-patient study published inPLOS ONE.

“Our study is small, retrospective and all of the patients were located at a single medical center, but it demonstrates that it’s possible to use molecular diagnostics to identify subgroups of patients more likely to respond to a given treatment,” co-first author John Paul Shen, MD, senior clinical fellow and postdoctoral fellow, said in a statement. “Given this proof-of-principle, it’s our hope that molecular biomarkers will be included in future prospective clinical trials in metastatic colorectal cancer."

Response to front-line chemotherapy is a strong predictor of OS in patients with mCRC, the authors of the study noted. At this time, there is not currently a tool available to predict response to chemotherapy regimens commonly used in the frontline setting.

In the study,ERCC1andTSlevels were assessed using the ResponseDX: Colon assay in 41 patients with de novo mCRC. A “normal” threshold value was established, in order to stratify patients into high or low groups, based on previous clinical trials. ForTS,a normal expression level was indicated as 4.0 and forERCC1the normal level was 1.73.

In theERCC1-low group (n = 33), the median OS was 36.0 months compared with 10.1 months in those with highERCC1expression (HR = 0.29; 95% CI, 0.10-0.84,P= .000009). The median time to treatment to failure following first-line chemotherapy was 14.1 vs. 2.4 months, in theERCC1low and high arms, respectively (HR = 0.17; 95% CI, 0.05-0.58,P= .00053). Patients with lowTSexpression (n = 29) experienced a median OS of 36.0 months compared with 14.8 months in theTShigh group (HR = 0.25, 95% CI, 0.074-0.82,P= 0.022).

Combined low expression ofERCC1/TSwas predictive of response in patients treated with FOLFOX (n = 27). In those with low expression of both genes the response rate was 91% versus 40% in those with high expression (P= 0.03). However, this same observation was not seen with FOLFIRI (n = 14; 71% vs. 71%), suggesting that the predictive effect could be related to oxaliplatin.

To validate the findings, the researchers looked at the TCGA pan-cancer database. Across the full dataset (n = 6202),ERCC1andTSwere not found to impact OS. Again, in a cohort of 62 patients with mCRC, expression of the two genes did not demonstrate a statistical impact (HR = 0.98;P= .96). However, in those treated with oxaliplatin,ERCC1was associated with improved outcomes.

In 27 patients from the TCGA database treated with oxaliplatin, those with lowERCC1expression had a median OS that was not yet reached compared with 39.5 months in those withERCC1-high tumors (HR = 0.22; 95% CI, 0.05-1.5;P= .13). There was a trend toward improvement in theTSlow group; however, this was not statistically significant (HR = 0.77;P= .77).

“Several large trials compared oxaliplatin and irinotecan head-to-head and concluded that the response rate is about equal. How an oncologist bases his or her treatment decision can be based on experience, comfort level prescribing and the patient’s health,” senior author Paul Fanta, MD, assistant clinical professor of medicine and oncologist at UC San Diego Moores Cancer Center, said in a statement. “But in reality, the two drugs are very different. For any individual patient, one might be better than the other. As an oncologist, how do I know which is better for my patient? That’s where this study comes in.”

Given the small retrospective nature of the study, further study will need to validate the two potential biomarkers in a randomized fashion. "In the future, the prognostic and predictive value ofERCC1andTSexpression should be examined the context of a larger prospective clinical trial," the authors noted.