American Association for Clinical Chemistry
Better health through laboratory medicine
February 2009 Clinical Laboratory News: Personalized Medicine in Colorectal Cancer Treatment


February 2009: Volume 35, Number 2

Personalized Medicine in Colorectal Cancer Treatment
Should all patients be tested for the KRAS mutation?
By Gina Rollins

A growing body of evidence has demonstrated that patients with metastatic colorectal cancer whose tumor has the wild-type version of the Kirsten-RAS (KRAS) gene are the most likely to benefit from monoclonal antibody therapy that inhibits epidermal growth factor receptor (EGFR). This groundbreaking research has lead influential groups like the National Comprehensive Cancer Network (NCCN) to call for KRAS mutation testing in all patients with metastatic colon cancer prior to starting therapy, an appealing concept in terms of targeting an expensive and potentially toxic agent to those most likely to benefit from it.

However, challenges exist in implementing this new clinical decision-making paradigm. A number of KRAS analysis methods exist, but most require many steps and proficiency not only with the assay but also in tumor specimen selection. In addition, validated methods and standardized testing procedures are lacking. Given these issues, laboratorians “should take the lead in evaluating assays and making a determination of what platform is best for the indication and for their lab,” said Marc Ladanyi, MD, chief of the molecular diagnostics service at Memorial Sloan-Kettering Cancer Center in New York City. “It’s important for the lab community to be leading the discussion.”

Targeting Costly Therapy

KRAS, which was one of the first genes found to be mutated in human cancer, encodes a G-protein that is essential in regulating cell growth and proliferation, as part of the EGFR signaling cascade. EGFR mutations have been implicated in various malignancies, including lung adenocarcinomas and glioblastoma. Anti-EGFR therapies block activation of the receptor, thereby inhibiting downstream events that lead to cancer cell signaling.

A spate of recent research has demonstrated that KRAS status is highly predictive of outcomes in patients with metastatic colorectal cancer who receive anti-EGFR therapies, including the monoclonal antibodies cetuximab (Erbitux, Imclone) and panitumumab (Vectibix, Amgen). The antibodies bind to EGFR on the cell surface, thereby blocking signals that turn on downstream signaling pathways, including the KRAS protein. The FDA approved cetuximab in 2004, and panitumumab in 2006, both as treatments for metastatic colorectal cancer. The therapies are costly— as high as $38,000 for a 4-month course of treatment—so testing to validate that a patient will benefit makes sense.

The recent studies have found that tumors with the normal wild-type form of the KRAS gene responded to treatment with cetuximab and panitumumab, whereas the powerful therapeutics had limited efficacy in patients with KRAS mutations. Indeed, one study found that among patients with mutated KRAS tumors, there was no significant difference between those who were treated with cetuximab and those who received supportive care alone in terms of overall survival or progression-free survival. In contrast, patients with wild-type KRAS tumors who received the same treatment had significantly improved overall survival and progression-free survival in comparison to supportive care (N Engl J Med 2008; 359: 1757–1765).

A New Standard of Care?

Lead author of that study, Christos Karapetis, MBBS., senior consultant medical oncologist and director of clinical research at Flinders Medical Centre in Adelaide, Australia, urged during his presentation at the September 2008 meeting of the European Society for Medical Oncology, that KRAS status testing before administration of EGFR inhibitors should be the standard of care for colorectal cancer treatment. “We had a good rationale for thinking that the presence of the KRAS mutation might predict the patients who would not benefit from treatment,” he said. Karapetis also is senior lecturer at Flinders University.

Similarly, a study involving panitumumab found that progression-free survival in patients with wild-type KRAS tumors was significantly greater than in those with KRAS mutation tumors. The response rate to treatment was 17% among patients with wild-type KRAS tumors, versus 0% among those with KRAS mutations (J Clin Oncol 2008; 26:1626–1634). A co-author of that study, Eric Van Cutsem, MD, PhD, professor of internal medicine at the University of Leuven (Belgium), also recommended during his presentation at the 2008 annual meeting of the American Society of Clinical Oncology (ASCO) that KRAS status testing be performed prior to treatment. “KRAS testing should be routinely conducted in all colorectal cancer patients immediately after diagnosis to ensure the best treatment strategies for the individual patient,” he said. KRAS mutations occur in an estimated 30% to 50% of all colorectal tumors, and more than 3,000 KRAS point mutations in colorectal cancer now have been reported. The most frequent colorectal cancer-related mutations are in codons 12 and 13 in exon 2 of the KRAS gene.

New Guidelines Support KRAS Testing

Based upon the recent evidence about the role of KRAS status in predicting response to EGFR inhibitor therapies, the NCCN updated its guidelines on colorectal cancer in November 2008 to include a recommendation that determining KRAS status of either the primary tumor or a site of metastatis should be part of the pre-treatment work-up for all patients with metastatic colorectal cancer. NCCN also recommended that cetuximab and panitumumab, either as single agents, or in the case of cetuximab, in combination with other agents, be given only for patients with wild-type KRAS tumors.

For now, the FDA has not changed labeling for either cetuximab or panitumumab in relation to KRAS status. Amgen sought FDA approval to require a genetic test for KRAS status prior to administration of panitumumab, and ImClone asked to amend the cetuximab label to include data from clinical trials indicating that wild-type KRAS has been linked to better progression-free survival. However, the FDA Oncologic Drugs Advisory Committee did not vote on the issue during its December 2008 meeting. Instead, discussion centered around the type and amount of data needed to support product labeling using biomarkers. A key consideration was whether biomarker data collected retrospectively could be sufficient for approval decisions.

Testing Challenges

These regulatory and guideline changes, along with evolving scientific evidence, reflect the growing trend of mutation testing being adopted for common cancers, according to Ladanyi. Labs should expect a “big scale up” in activity in this field generally, and with regard to KRAS testing specifically, there are challenges to be addressed, he indicated.

One issue is that KRAS status is not the be-all and end-all in determining a patient’s probable response to EGFR inhibitor therapy. For example, the study of cetuximab co-authored by Karapetis found that some patients with wild-type KRAS did not respond to the drug, and their tumors rapidly progressed. “Additional reliable and easily measured biomarkers are clearly needed,” the authors noted.

In addition, methods of KRAS testing have not been standardized, and there is considerable reported variation, particularly in its sensitivity. A recent review article identified more than 20 different methods of KRAS genotyping, many of which are laboratory-based or for research use only. “Clearly, there is accumulating evidence that the negative predictive power—the specificity—of testing positive for the KRAS mutation is very high. But the sensitivity of that test is not very good,” said Jerry Yeo, PhD, director of the clinical chemistry laboratories and clinical pharmacogenomics program at the University of Chicago. One meta-analysis found a significant difference in specificity based on detection method—0.87% for sequencing methods versus 0.99% for non-sequencing methods—and non-significant differences in sensitivity; 47% for sequencing methods versus 48% for non-sequencing methods (Lancet Oncol 2008; 9:962–72). Yeo considers the relatively low sensitivity to be a concerning limitation of testing. “You might as well toss a coin regarding the sensitivity of the test,” said Yeo, who is also chair of AACC’s molecular pathology division.

Most of the recent studies of KRAS’s predictive role in metastatic colon cancer used direct gene sequencing, according to Ladanyi. “If done intelligently that’s a perfectly acceptable choice, but you need constant interaction with the surgical pathologist to select the best tissue block, and the best area of the block,” he noted. The main concern is that with colorectal cancer there tends to a substantial amount of non-neoplastic tissue within the specimen, and a sufficient amount of invasive tumor cells are needed for analysis. The minimum amount of tumor versus non-tumor area needed depends on the type of KRAS method utilized. Memorial Sloan-Kettering uses direct sequencing and strives to have at least 50% of the specimen composed of tumor, according to Ladanyi.

Two KRAS mutation test kits have met CE-Mark requirements for use in the European Union, and several companies recently launched new tests in the U.S., including TrimGen Corporation and Caris Diagnostics. According to Ladanyi, some of these kits are being marketed directly to oncologists, but he urged laboratorians to take an active role in determining the type of assay used in their institutions.

The Need for Standardization

The need for standardized testing procedures prompted the European Society of Pathology (ESP) to propose guidelines for KRAS mutation analysis and a European quality assurance program to ensure optimal accuracy and proficiency in KRAS mutation testing throughout the European Union (Virchows Arch 2008; 453:417–431). Among other things, the ESP proposed standardized testing in all patients with stage II and stage III colorectal carcinomas, underscored the central role of pathologists in selecting tissue blocks with a sufficient amount of invasive tumor cells, and recommended that labs develop standard operating procedures and testing requirements for KRAS. ESP also recommended the lower detection limit of mutant signal for sensitivity should be 1% of tumor cells for allele-specific PCR and 25% to 30% for direct sequencing. A specific test should be able to detect seven common mutations in codons 12 and 13 of the KRAS gene and not detect mutations in codon 61, according to the guidelines.

Ladanyi expects similar guidelines eventually will be implemented in the U.S., like those issued in 2006 by ASCO and CAP for human epidermal growth receptor 2 testing for breast cancer. “It’s going to happen, but I’m not sure of the timing,” he said. “People may need a little more hands-on experience with the data and a better understanding of how to use the tests.”