In colorectal cancer, all KRAS mutations may not be created equal: A post-EMCC update

Issue: October 25, 2011

ByTanios Bekaii-Saab, MD

At ASCO 2011, during a presentation from Sabine Tejpar, MD, PhD, data from a retrospective analysis that she presented suggested a heterogeneous treatment effect from cetuximab according to KRAS mutation status.

It appeared that codon 13 mutant KRAS (KRAS G13D mutation) was an adverse prognostic finding similar in extent to the presence of BRAF mutations, so that patients did less well on chemotherapy alone if they had this mutation. What were surprising in Dr. Tejpar’s results were the relative effects of cetuximab (Erbitux, ImClone) that were parallel to that of wild-type KRAS, although the absolute effects were lesser. This data seems to be strengthened by findings from a study of a metastatic colorectal cancer cell line, where the addition of KRAS codon 12 cells but not KRAS G13D cells reduced sensitivity to cetuximab.

In sum, Tejpar’s data suggested that tumors with KRAS G13D appear to be still susceptible to cetuximab, whereas other codon types not only appear to predict no benefit but also predict a deleterious effect, especially in oxaliplatin-based therapies. Now, based on findings at EMCC 2011 in Stockholm just last month, it appears tumor KRAS mutational status have a more complex impact on response to cetuximab and panitumumab (Vectibix, Amgen), both monoclonal antibodies that block the epidermal growth factor receptor, in patients with metastatic colorectal cancer than previously thought.

A change in response

At EMCC, Marc Peeters, MD, presented data evaluating individual codon 12 and 13 mutant KRAS alleles as potential prognostic and predictive biomarkers of response to panitumumab in patients with metastatic colorectal cancer (mCRC). The data were extracted from three phase 3 studies with panitumumab as a single agent (Study 20020408) or in combination with either FOLFOX (Study 20050203) or FOLFIRI (Study 20050181). The results of Peeters’ findings suggest that, collectively, the presence of the previously described KRAS G13D mutation may be neither prognostic nor predictive for the use of panitumumab in mCRC, which seems to be contradictory to the findings with cetuximab presented by Tejpar at ASCO 2011.

Tanios Bekaii-Saab, MD
Tanios Bekaii-Saab

However, when looking more closely at the data with panitumumab and comparing it to the data with cetuximab, patients seemed to have a trend toward benefit from the addition of the monoclonal antibodies (panitumumab or cetuximab) to an irinotecan-based regimen in the presence of a KRAS G13D mutation. The same comparative observation is not seen when the agents are given alone or combined with an oxaliplatin-based regimen. This observed disparity is very interesting given that both anti-EGFR antibodies seem to exert a more consistent overall differential synergism with irinotecan-based regimens compared with oxaliplatin-based regimens across lines of therapy.

Small benefits, big costs

Previously published studies produced the conclusion that wild-type KRAS is a prerequisite for benefit from cetuximab or panitumumab. Consistently across studies, the presence of a KRAS G12 mutation is predictive of lack of response to anti-EGFR antibodies. However, the presence of a KRAS G13D mutation may be associated with a modest benefit with the addition of anti-EGFR antibodies to only irinotecan-based regimens. But, given that there is an inconsistent prognostic effect of the KRAS G13D mutation in mCRC, the cost/benefit question may be even more difficult to address in a randomized trial, especially that among all KRAS mutations the G13D mutation accounts for about 8% to 10% of mCRC tumors.

It is also possible that other KRAS mutations (codons 61, 146 and others) may define different tumor behaviors, but this will be even more difficult to show in prospective studies given the limited number of tumors expressing other codons. Given the complexity of the EGFR signaling system, strong consideration should be made for gene expression profiling rather than binary reporting of mutational status (present vs. absent) in future prospective studies.

In conclusion, the data available to us as of today suggest that the presence of a KRAS mutation should exclude all patients from receiving an EGFR inhibitor as a single agent or in combination with oxaliplatin-based therapies in mCRC. The same is true for irinotecan-based regimens, except that the clinical significance of the presence of a KRAS G13D mutation remains unclear when considering an EGFR inhibitor.

Although these data are still preliminary because of their retrospective nature, a case can be made for further subtyping KRAS mutations by codon when considering the use of an EGFR monoclonal antibody inhibitor combined with irinotecan-based therapies, given the consistency of the findings across studies. Of course, prospective validation of results in this particular setting is still needed, and future studies with anti-EGFR antibodies should include and stratify for KRAS G13D mutations.

For the practicing clinician today, this will not change the paradigm of how EGFR inhibitors are selected. As such, and outside of a clinical trial, the presence of any KRAS mutation should exclude patients from receiving EGFR inhibitors.

Tanios Bekaii-Saab, MD, is Medical Director, Gastrointestinal Oncology, and Associate Professor of Medicine and Pharmacology at The Ohio State University – James Cancer Hospital. He is also a member of the HemOnc Today Editorial Board. Disclosure: Dr. Bekaii-Saab reports receiving consulting fees from BMS and Amgen.

For more information:

De Roock W. JAMA. 2010;304:1812-1820.

De Roock W. Lancet Oncol. 2010;11:753-762.

Di Nicolantonio F. J Clin Oncol. 2008;26:5705-5712.

Frattini M. Br J Cancer. 2007;97:1139-1145.

Loupakis F. Br J Cancer. 2009;101:715-721.

Peeters M. Abstract 33LBA. Presented at: EMCC 2011; Sept. 23-27; Stockholm, Sweden.

Tejpar S. J Clin Oncol. 2011 29:(suppl 4; abstr 630).

Wheeler DL. Nat Rev Clin Oncol. 2010; 7(9): 493-507.