This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Google Scholar Articles by Garassino, M. C. Articles by Torri, V. PubMed Articles by Garassino, M. C. Articles by Torri, V.

Should KRAS Mutations Be Considered an Independent Prognostic Factor in Patients With Advanced Colorectal Cancer Treated With Cetuximab?

Marina Chiara Garassino, Gabriella Farina

Department of Oncology, Fatebenefratelli and Oftalmico Hospital, Milan, Italy

Antonio Rossi

Division of Oncology, S.G. Moscati Hospital, Avellino, Italy

Olga Martelli

Department of Oncology, San Giovanni Hospital, Rome, Italy

Valter Torri

Department of Oncology, Mario Negri Institute, Milan, Italy

To the Editor:

Lievre et al¹ analyzed 89 metastatic colorectal cancer (CRC) patients to investigate the prognostic value of KRAS mutations on response to cetuximab and survival. The great majority of patients (97.7%) received, as salvage therapy, cetuximab plus irinotecan, alone or in combination with fluorouracil and folinic acid, after failure of an irinotecan-based regimen. A KRAS mutation was present in 27% of the patients and was associated with resistance to cetuximab. An objective response of 0% versus 40% was reported among the 24 and 65 patients with and without mutations, respectively. A median progression-free survival of 2.1 versus 7.2 months (P = .0001), and a median overall survival of 10.1 versus 14.3 months (p 0.026) were reported, respectively. The authors concluded, "These results confirm the high prognostic value of KRAS mutations on response to cetuximab and survival in metastatic CRC patients treated with cetuximab."¹ This retrospective analysis gives the opportunity to make a few considerations.

In these reported series of patients, only two (2.3%) received cetuximab therapy alone. Therefore, it is not possible to draw any conclusion regarding the prognostic value of KRAS mutations to cetuximab alone; KRAS mutations could be a negative prognostic value on response to cetuximab or to chemotherapy, or to both. These doubts arise from the absence, to our knowledge, of a detailed analysis reporting prognostic correlations between KRAS mutations and chemotherapy alone for metastatic CRC. Several studies have tried to determine whether the presence of KRAS mutations correlates with response to adjuvant therapy or survival, but only conflicting results were reported,² and none of them addressed this issue properly. The prognosis of metastatic CRC patients does not seem to be influenced by KRAS mutations, but few data are available,^3,4 and they refer mostly to only old-generation, fluorouracil-based regimens. Regarding this last issue, the study might provide data concerning the correlation between KRAS mutations and response or duration of disease control during the previous lines of chemotherapy. These further data may be of great interest, in particular for a better evaluation of the association between KRAS mutations and chemotherapy, which could consequently help in a clearer evaluation of data related to cetuximab.

Moreover, the authors pooled the patients of the present series with the ones of a previous study for a total of 114 patients confirming KRAS status as an independent prognostic factor associated with overall and progression-free survival. But the authors did not discuss the fact that the research of KRAS mutations was performed on metastatic tissue in 31 patients (27%), whereas it was performed on the primary site for the others. The implicit assumption that the evaluation of the KRAS mutations can be performed independently from primary or metastatic site should be supported by correlation analysis.

In our opinion, the answer to the title question cannot be given by the analysis of Lievre et al. At present, KRAS mutations could be considered as a negative prognostic factors of outcome to every treatment and not only to cetuximab. Case series or "post hoc," not previously planned analyses may be hypothesis generators, but only a properly planned randomized controlled study specifically aimed to investigate the possible prognostic and predictive role of KRAS is warranted to definitively answer this question.

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The author(s) indicated no potential conflicts of interest.

NOTES

published online ahead of print at www.jco.org on April 28, 2008

REFERENCES

1. Lièvre A, Bachet JP, Boige V, et al: KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab. J Clin Oncol 26:374-379, 2008[Abstract/Free Full Text]

2. McDermott U, Longley DB, Johnston PG: Molecular and biochemical markers in colorectal cancer. Ann Oncol 13:235-245, 2002 (suppl 4)[Medline]

3. Andreyev HJ, Norman AR, Cunningham D, et al: Kirsten ras mutations in patients with colorectal patients: The multicenter "RASCAL" study. J Natl Cancer Inst 90:675-684, 1998[Abstract/Free Full Text]

4. Andreyev HJ, Norman AR, Cunningham D, et al: Kirsten ras mutations in patients with colorectal patients: The "RASCAL II" study. Br J Cancer 85:692-696, 2001[CrossRef][Medline]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Google Scholar Articles by Garassino, M. C. Articles by Torri, V. PubMed Articles by Garassino, M. C. Articles by Torri, V.