KRAS Status and Efficacy in the First –Line Treatment of Patients with Metastatic Colorectal Cancer (mCRC) Treated with FOLFIRI with or without Cetuximab: The Crystal Experience

Reviewer: Eric Shinohara MD, MSCI
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 23, 2008

Presenter: Van Cutsem E.
Presenter's Affiliation: University Hospital Gasthuisberg, Leuven, Belgium
Type of Session: Plenary


  • Cetuximab is a chimeric antibody which blocks epidermal growth factor (EGF) signaling by binding to the EGF receptor (EGFR). 
  • The EGFR has several downstream targets, including the PI3K/Akt pathway and the MAP kinase pathway (via k-ras). 
    • K-ras is a guanine nucleotide-binding protein (g-protein) which is activated when bound to GTP. GTP can then be hydrolyzed to GDP turning off k-ras activity.
    • However, if k-ras is mutated such that it is unable to hydrolyze GTP, k-ras can remain constitutively active.  
    • Because k-ras is downstream from the EGFR, it makes sense that mutations in k-ras (which cause it to be constitutively active) would interfere with EGFR inhibition (such as with cetuximab), as the mutated k-ras is no longer dependent on signaling from the EGFR.
    • Conceptually, this may be why EGFR inhibitors are more effective in cancers with wild type (WT) k-ras.
  • Several small retrospective studies have suggested that cetuximab is effective in cancers with WT k-ras, but has little to no effect in cancer patients with mutated k-ras.
  • The CRYSTAL trial (Van Custem E., et al., ASCO Annual Meeting , 2007) was a large, randomized, phase III trial which investigated the use of FOLFIRI (regimen described below in methods) versus FOLFIRI plus cetuximab as first line treatment in EGFR expressing advanced colorectal cancers. Results demonstrated an increase in both response rate (RR), median progression free survival (PFS), and curative surgery rates in the arm with cetuximab. 
    • The present study is a retrospective subset analysis of patients from the CRYSTAL study for whom k-ras data is available. 

Materials and Methods

  • There were archived tissue blocks available for 587 patients from the initial 1,198 patients who were enrolled in the CRYSTAL trial from August 2004 to October 2005.
  • Chemotherapy was administered as follows:
    • FOLFIRI (arm A)
      • Given every 2 weeks: irinotecan 180 mg/m2, folonic acid (FA) 400 mg/m2, 5-FU bolus 400 mg/m2, 5-FU infusion 2,400 mg/m2 over 46 hours
    • FOLFIRI plus cetuximab (arm B):
      • FOLFIRI given in identical fashion as arm A.
      • Cetuximab: 400 mg/m2 (initial dose) then 250 mg/m2/week
  • DNA was directly isolated from patient slides. The most common sites of k-ras mutation on codons 12 and 13 were assayed for using quantitative polymerase chain reaction (qPCR). K-ras status was evaluable in 540 patients.
  • The endpoints for this subset analysis were RR and PFS.  Analyses were performed as in the initial CRYSTAL trial.


  • For review, the results from the initial CRYSTAL study were as follows: 
    • Median PFS was significantly longer for arm B (8.9 months) compared to arm A (8 months) (p=0.036).
    • RR was also significantly increased by cetuximab (arm B 46.9% vs. arm A 38.7%, p=0.005).
  • Mutated k-ras was found in 35.6% of patients and WT k-ras in 64.4% of patients, in keeping with prior studies.
    • The distribution of WT and mutant k-ras was similar between arm A and the arm B.
    • There was no significant difference in patient demographics between patients with mutant and WT k-ras. 
  • Outcomes for patients with WT k-ras improved with the addition of cetuximab:
    • Median survival was 9.9 months for patients in arm B versus 8.7 months for patients in arm A (p=0.017) with a hazard ratio (HR) of 0.68 (95% CI 0.051-0.934).
    • 1-year PFS was 45% for patients in the arm B and 25% for arm A.
    • RR for patients in arm B was 59% versus 43% in arm A (p=0.0025)
  • Outcomes for patients with mutant k-ras were not affected by the addition of cetuximab:
    • Median survival was 7.6 months for arm B versus 8.1 months for arm A with a HR of 1.07 (95% CI of 0.71-1.61).
    • RR was 36% for arm B versus 40% for arm A (p=0.46).
  • Response to FOLFIRI was similar for patients with WT and mutant k-ras.
  • The duration of treatment was longer in patients with WT k-ras, likely reflecting their better outcome and longer survival.
  • There was no difference in toxicity seen between patients with WT and those with mutant k-ras. 
  • In the original CRYSTAL study, there was more grade 3 or greater toxicity in the cetuximab arm. This held true for patients with WT and mutant k-ras in the present study (both groups had more grade 3 or higher toxicity with cetuximab).

Author's Conclusions

  • The findings from the present retrospective subset analysis are similar to those found by the OPUS study (Bokemeyer C., et al., ASCO Annual Meeting, 2007). The OPUS study found that in patients with WT k-ras, there was a RR of 61% in patients who received cetuximab versus 37% in those who did not, which is similar to the CRYSTAL study findings (59% versus 43%).
  • Overall cetuximab in addition to FOLFIRI improved PFS. However, this benefit was limited to patients who had WT k-ras.
  • Cetuximab and FOLFIRI is an effective treatment regimen; however, from the results of this study, it is clear that patient selection is important.
  • The authors state that k-ras is one of the first molecular markers that is effective in dictating patient treatment. They believe that testing for k-ras mutations should be implemented in all patients who are to receive cetuximab or panitumumab.

Clinical/Scientific Implications

  • It is possible that WT k-ras may have be a prognostic factor rather than a predictive factor of therapy. If k-ras was a prognostic factor then one would expect that, regardless of the treatment, patients with WT k-ras would have done better. However, given that WT k-ras did not change outcomes in patients treated with FOLFIRI alone, it seems that WT k-ras is indeed a predictor of outcome.
  • It is important to keep in mind that this is a retrospective subset analysis. However, the quality of this subset analysis is good. The WT and mutant k-ras arms were well balanced with similar demographics and tumor characteristic between the two groups. The authors also used appropriate assays for the most common k-ras mutations. These factors suggest that the results from this study are valid.
  • Supporting this study are additional data from several other large studies which have suggested that cetuximab is of benefit only in patients with WT k-ras. Hence, we have a consistent body of retrospective data, across thousands of patients, suggesting that cetuximab is only effective in k-ras WT patients. How then do we apply this information? First, we have to find appropriate ways to separate patients into those with WT k-ras and those with mutant k-ras. If an assay is used to detect mutant k-ras, it must be very specific, as a false positive could potentially deny a k-ras WT patient therapy that would be effective for him or her. Similarly, any assay for WT k-ras must be very sensitive. It may be possible to increase our ability to appropriately define non-responders by identifying and assaying for other negative predictors for response to cetuximab. Potential negative predictors include inactive PTEN and increased activity of PI3K, which may predict for a worse outcome. Possible genetic markers for other positive predictors, such as skin rash, could also allow further refinement in our ability to select patients. In the future, biomarkers should be developed during the early stages of a clinical trial such that they can be used and potentially validated as the trial expands. Ongoing trials with cetuximab and panitumumab likely need to be amended to include k-ras testing
  • Based on these data, k-ras status should be assayed in patients who are candidates for cetuximab or panitumumab therapy, as it appears that only those with WT k-ras will benefit. This will spare k-ras mutant patients unnecessary toxicity. It is also important to tell patients with k-ras mutation that traditional chemotherapies remain effective.

View Patient Summary