Molecular Testing for Advanced Colorectal Cancer - Episode 6
Transcript:Daniel G. Haller, MD, FACP, FRCP: For the RAS wild-type patient, the patient who might be able to receive an EGFR inhibitor, the use of these drugs has changed since they were approved more than 10 years ago. And the actual use of these drugs, especially in the United States, is different from their use in Europe. Part of that is based on where studies were done and how the drugs were initially tested. Bevacizumab, the VEGF agent, was developed from the top down. It was developed as a first-line agent and as a second-line agent, E3200, so that people got used to using it as first-line therapy. They did it in trials; they learned about it first. It was approved in 2004.
For the EGFR agents, they were tested as salvage therapies. In people’s minds, they saw bevacizumab as first-line; the EGFR agents are after the single agents, which bevacizumab is not; it has a predictive marker. And they were developed against best supportive care. Panitumumab was in the Canadian trial as a last line of therapy. It was a very potent teaching lesson for American doctors, which is that the EGFR agents come last, bevacizumab comes first.
In Europe, most of the first-line studies using the EGFR agents were done there; they weren’t done in the United States. The only one that was a first-line agent in the United States was the PACE trial with panitumumab, which unfortunately was combined with bevacizumab, and that study showed a negative result. The US clinicians’ exposure to the EGFR agents in first-line were either negative or not at all. And there was very little to change people’s minds until some of the more recent studies from Europe, like PRIME and CRYSTAL, using EGFR agents in first-line, were published and widely read. In fact, cetuximab in first-line therapy and panitumumab have only been approved in the US for the last few years. So these are really recent introductions.
Educating people about when to consider first-line EGFR agents has to be derived from both 80405 and the FIRE-3 trial. In FIRE-3, they didn’t make any distinctions. Everyone seemed to do better with first-line cetuximab, although the practices of second and third and further line therapies in Europe are very different from the US, where we have more drugs and we use them more often.
The one point of agreement between the two studies, between FIRE-3 and 80405, is that the response rate was clearly higher in both of the trials, even if survival was not. Response matters in a number of situations. Response matters if people have symptoms related to bulk disease. It may not matter if someone has seven tiny asymptomatic pulmonary metastases. In fact, you might not need a biologic in that situation. You probably could get along with single agent capecitabine, especially in the elderly.
But people who have bulky disease or have BRAF-mutant tumors with big liver metastases or peritoneal disease, these people need response. They need it soon, they need it up front. You can’t rely on fourth-line therapy because they don’t get to fourth-line therapy. Your first treatment is your most important. In other patients, you can make it up in further line therapies. In the BRAF-mutant patients, unfortunately 10%, if they get to second-line therapy, they’re lucky. And third-line is an allusion for many of them.
The other group of patients where response matters are the people who are borderline-resectable liver metastases patients. We know from the new EPIC trial, for example, that in the clearly resectable patients, cetuximab did not add to FOLFOX chemotherapy, but in the borderline patient, it might. The 80405 study is unique in that more than 100 patients on that study were entered with the intent to do some form of resective surgery, including liver resections. This was the first time people were entered who were potentially resectable, and stated to be so, at entry.
If I remember the numbers correctly, 132 people underwent some form of resection, typically liver resection, and there was an imbalance: 62% got the EGFR agent, 38% got bevacizumab. So it was nearly 2-to-1 in favor of cetuximab. Now to be fair, surgeons were not blinded or doctors were not blinded as to what drug the patient had. In their mind, their bias was they could resect someone, plus they had the EGFR agent, then they might have done more surgery.
It’s probably not only that. I do think that with a higher response rate, the EGFR agent, cetuximab in this case, just simply led to more resections. This also explains one of the differences between the two trials. If 132 patients had resection, if I remember at the time of the last report, their 5-year survival was extraordinarily good; it was like 60%. So all of a sudden you’re diluting out by putting in potentially curable patients, you’re diluting out the potential results in the total trial on overall survival. That could be one of the big differences between the results for survival in the 80405 and FIRE-3 trials.
Charles S. Fuchs, MD: Right now, we have some really interesting studies that are comparing the biologics in metastatic colon cancer, and more or less, they tell us that you get similar benefit no matter whether you pick a VEGF inhibitor or an EGFR inhibitor. But we have to do better than that, and the way to do better is to get in-depth sequencing data and even more in-depth gene expression data. The variety of technologies that are now becoming available, I guarantee you, are going to tell you which of those drugs should be used frontline. What I hope over the next 3 to 5 years is we’re going to get that data as part of routine clinical practice and that we’re going to make decisions where we can say to the patient confidently, “This is the right drug to use because that’s where this field needs to go.”
Transcript Edited for Clarity