Beyond KRAS in Colorectal Cancer: Additional Markers Ready for Use?

Nick Mulcahy

October 15, 2009

October 15, 2009 — Oncologists should test patients with metastatic colorectal cancer not just for their KRAS status, but also for the status of 3 other biomarkers, according to the authors of a new study from Italy.

The additional testing could nearly double the number of patients who are identified as unlikely to benefit from antiepidermal growth-factor receptor targeted therapy, which is expensive and can be toxic.

Previous research has shown that about 40% of metastatic colorectal cancer patients have the mutated form of KRAS and will not respond to either panitumumab (Vectibix, Amgen) or cetuximab (Erbitux, Bristol-Myers Squibb), note the authors.

Up to 70% of metastatic colorectal cancer patients unlikely to respond can be identified.

But the new study shows that "up to 70% of metastatic colorectal cancer patients unlikely to respond can be identified" when the status of 3 other biomarkers is considered. In addition to testing for KRAS mutations, these researchers also tested for mutations of the BRAF and PIK3CA genes and ascertained the expression of PTEN.

The study was headed by Andrea Satrore-Bianchi, MD, from the Falck Division of Medical Oncology at Ospedale Niguarda Ca'Granda in Milan, Italy, and was published online October 2 in PLoS ONE.

"This is a ground-breaking, hypothesis-generating study," said Wells Messersmith, MD, a gastrointestinal oncology expert from the University of Colorado Cancer Center in Boulder, who was approached by Medscape Oncology for comment. He also said that the Italian group is "leading the world in generating hypotheses about biomarkers in colorectal cancer."

These other biomarkers are not ready for the clinic.

However, Dr. Messersmith continued, "these other biomarkers are not ready for the clinic."

A big part of the problem is that the additional biomarkers have not been definitively evaluated, said Dr. Messersmith. "Practice-changing data must come from large prospective randomized clinical trials," he said.

This new study is a small retrospective analysis of 132 patients who "had been treated in all kinds of different ways," notes Dr. Messersmith.

Nevertheless, Dr. Messersmith said that, of the 3 other markers proposed by the new study, BRAF, along with its oncogenenic V600E mutation, is most likely to be validated. "It is clear. BRAF will eventually be used because all of the data are consistent. Anywhere from 3% to 12% of metastatic colorectal cancer patients have a mutated form of BRAF, and all those patients do poorly," he said.

So, when will oncologists learn about BRAF from a large prospective dataset?

Dr. Messersmith said that Bristol-Myers Squibb and Amgen have the data to generate results about BRAF, but have either not done the analysis or not released the data. "All of us await those results," he said.

What's the Hold Up?

You can test for other biomarkers such as BRAF in an afternoon.

Using the registry trials of cetuximab or panitumumab to test tissue samples of patients for their BRAF status is not difficult, suggested Dr. Messersmith. "Once you isolate the DNA to determine KRAS status, you can test for other biomarkers such as BRAF in an afternoon," he said.

The hold-up in the release of the BRAF data might be the pharmaceutical industry's "potential conflict of interest," he said.

"Further results on biomarkers, including BRAF, may decrease the patient population that is eligible to be treated," he explained.

Testing for KRAS status in patients with metastatic colorectal cancer saves the American healthcare system about $700 million dollars a year, Dr. Messersmith said. Adding testing for BRAF status would further decrease the use of these drugs, and thus provide even more savings.

If I was a shareholder, I would be outraged if my company lost $700 million.

From the perspective of the drug companies, that is money lost. Additional biomarkers will mean more lost revenue, perhaps as much as twice the current amount, he suggested. "If I was a shareholder, I would be outraged if my company lost $700 million," he added.

To date, Bristol-Myers Squibb and Amgen have been "forthcoming" about KRAS data and their employees have been coauthors of papers on the subject, said Dr. Messersmith.

Nevertheless, he said, how quickly the BRAF data are going to be released remains unclear.

At the latest, clinicians will have information on biomarkers in addition to KRAS in a few years. That's because the National Cancer Institute and a European Cooperative Study Group are currently sponsoring and running studies of colorectal cancer and targeted therapies.

"It will be 2 to 3 years before we get results from these trials," said Dr. Messersmith.

More on the Italian Study: Words About PIK3CA and PTEN

The Italian researchers note that information on additional biomarkers is very important, given the fact that less than 20% of patients with wild-type KRAS actually respond to either cetuximab or panitumumab. In other words, there must be other reasons that so many of the wild-type patients do not respond to the therapies, they suggest.

In the new study, the investigators did an analysis of the tumors of a cohort of 132 patients and detected 104 molecular alterations.

Specifically, they detected 35 KRAS mutations (26.5%), 11 BRAF mutations (8.3%), 17 PIK3CA mutations (12.3%), and 41 (out of 114 evaluable) losses of PTEN expression (36.0%). Some patients had more than 1 alteration. The most frequent "overlapping fingerprints" were PTEN loss and KRAS mutation (cooccurring in 13 patients).

The investigators evaluated objective tumor response in the patients, all of whom had been treated with cetuximab or panitumumab.

Among the 106 nonresponsive patients, 74 (70%) had tumors with at least 1 molecular alternation in the 4 markers.

The probability of response was 51% (22/43) among patients with no alterations, they note, was 4% (2/47) among patients with 1 alteration, and was 0% (0/24) among patients with 2 or more alterations (P < .0001).

The investigators also noted that progression-free survival and overall survival were increasingly worse for patients as their number of molecular alterations increased from 0 to 1 to 2 or more.

Dr. Satrore-Bianchi and colleagues also made an important concession: that the molecular alterations could be negative prognostic biomarkers independent of the targeted therapies.

This is why it is important to evaluate these biomarkers in large randomized trials, said Dr. Messersmith. "A negative prognostic biomarker is likely to result in negative outcomes in patients," he said.

"We have to tease out what is prognostic versus what is predictive," he added.

Dr. Messersmith also said that the expression of PTEN is not a good candidate to be a biomarker for colorectal cancer for 2 reasons. First, it is a subjective, graded analysis — unlike mutational status, which "you either have or you don't," he said. Second, the expression of PTEN has a relatively low concordance between primary and metastatic tumors — as low as 70% — which weakens its value as a biomarker, he said.

With regard to PIK3CA, Dr. Messersmith said that, unlike BRAF, there were conflicting results with it as a biomarker in this setting.

Important Afterthought

The discovery of the importance of KRAS and possibly other biomarkers in colorectal cancer illustrates the need for dramatic change in biomedical research, said Dr. Messersmith.

We may need to change trials to being based on the genetics of a tumor instead of being organ-specific.

"You can't spend $800 million to develop a drug that suddenly is not applicable to the entire patient population and in fact may only eventually be applicable to 10% of those patients," he said.

The current drug development model is not sustainable because of the development of biomarkers and the insights of personalized medicine, he suggested.

"We may need to change trials to being based on the genetics of a tumor instead of being organ-specific," Dr. Messersmith offered.

The study authors have disclosed no relevant financial relationships.

PLoS ONE 2009;4:e7287. Abstract

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....