Circling Back: The Effects of Emerging Molecular Understanding on Treatment Decisions in Colorectal Cancer

Joleen M. Hubbard, MD

Axel Grothey, MD

Points Made in Commentaries on “Colorectal Cancer: How Emerging Molecular Understanding Affects Treatment Decisions” Are Addressed by Authors Hubbard and Grothey

In the February issue of Oncology, Drs. Meera Sridharan, Joleen Hubbard, and Axel Grothey of the Mayo Clinic published a paper that reviews the recent work that has been done in attempting to understand colorectal cancer at the molecular level, and thus improve our ability to effectively use molecular targeted agents in these patients. The authors discuss how the latest findings affect medical decision making in metastatic colorectal cancer.

As with all Oncology articles, the review by Drs. Sridharan and colleagues was accompanied by two commentaries that presented different perspectives on this body of work. Because the authors of these commentaries made some especially interesting points, we wanted to provide Dr. Sridharan, Dr. Hubbard, and Dr. Grothey with the chance to address some of these.

Both Dr. Hubbard and Dr. Grothey are able to join us today, and I’m looking forward to hearing what they have to say. So let’s start with the first commentary, which was by Dr.

Nilofer Azad and Dr. Luis Diaz of the Sidney Kimmel Cancer Center at Johns Hopkins.

This commentary calls into question, to a degree, the value of RAS/RAF mutation testing. The authors question whether the testing for an extended list of RAS/RAF mutations that is commonly performed now in patients with advanced colorectal cancer really leads to greater benefit. They make the point that we are merely selecting patients better – simply identifying those patients who have all along been the drivers of the benefits observed in EGFR antibody trials (although they do concede that such testing at least spares patients with mutated tumors the adverse effects of such therapy).

Oncology: What are your thoughts on these authors’ perspective on RAS/RAF mutation testing? Dr. Grothey, would you like to take this one?

Dr. Grothey: Yes, more than happy to. So first of all, it’s interesting to see that when I listened to “merely selecting patients better” – this is actually one of the strong points I’d like to make, and actually one of the hallmarks of our idea of individualizing cancer care, is selecting patients better for a certain treatment intervention.

Number one was the idea that we can spare them unnecessary toxicities if no benefit, no chance for a benefit on a certain agent. On the other hand, we also try to avoid potential detrimental effects. And I think with comprehensive RAS testing, which I think is making inroads into the US too-it’s already been adopted in Europe-and we actually have a chance to prevent detrimental effects from patients with mutant RAS tumors, NRAS, KRAS tumors, when exposed to EGFR antibodies and chemotherapy, potentially  having a detriment in outcome.

This has been quite consistent in various presentations and publications, in combination with panitumumab and FOLFOX, but also cetuximab and FOLFIRI. There’s a suggestion that patients with mutant RAS tumors might actually have a detrimental effect in terms of progression-free survival.

So selecting patient data for certain treatment options is exactly the holy grail of individualized therapy. And we shouldn’t minimize this value for patients in terms of less toxicity, financial toxicity too, of course; and the potential in decreased detrimental effects, or preventing detrimental effects for these patients.

Oncology: Okay. Thanks very much. Dr. Azad and Dr. Diaz also express the opinion in their commentary that some of the most compelling recent data in advanced colorectal cancer involves neither RAS nor RAF family mutations – but rather PIK3CA mutations and the benefit seen from aspirin or NSAID therapy in patients who have these mutations. Would you agree with that assessment? Do you think that this is one of the most promising lines of current research in colorectal cancer? Dr. Hubbard?

Dr. Hubbard: Well, I think that there is some intriguing data about PIK3CA. But I think the jury is really still out. For instance, the prior analysis looked at patients with PIK3CA mutations, and actually showed retrospectively a large benefit from aspirin use in these patients with these mutations.

However, the largest study of this kind, investigating the benefit of aspirin in PIK3CA mutations was actually just presented at the ASCO GI Symposium at the end of January [2014]. And this actually contradicted some of the earlier findings. And for patients with stage 2 or stage 3 disease, they actually did not see a significant benefit from the use of aspirin.

Those patients with metastatic disease did appear to have some benefit in overall and cancer-specific survival. So these data contradict the prior data in this. So I still think it warrants further investigation in a prospective manner. But I agree. It’s interesting. The jury’s still out, really.

Oncology: Okay. Dr. Grothey, did you have anything to add?

Dr. Grothey: That mainly ties into, as we said in the earlier part of this discussion. We’d like to identify patients who can benefit or not benefit from certain treatment approaches using molecular markers. So in the end it really ties into the idea of conducting appropriately designed, well-conducted studies that can validate the initial treatment paradigms in retrospective analysis.

So, whether PIK3CA mutations are going to guide us in our treatment for, let’s say, NSAID, aspirin, or celecoxib, I think the jury’s still out there. But as Dr. Hubbard said, I think we need to go further, and this ties into the whole trend in oncology right now to not treat all patients alike, but try to identify what are the subgroups for different treatment interventions.

Oncology: Okay. Let’s move on now to the second commentary. The authors of this commentary, Dr. Laura Tenner and Dr. Bert O’Neil of the Melvin and Bren Simon Cancer Center at Indiana University, express the view that, in colorectal cancer, molecular targeted agents have fallen far short of what was once hoped for them. They note that the most effective agent in the treatment of metastatic colorectal cancer is still fluorouracil, and that most of the survival benefit that has been achieved in the last 10 years has come from optimizing the use of chemotherapeutic agents.

So first let me ask -- would you agree with that statement about cytotoxics being the source of most of the recent increased survival benefit in colorectal cancer? Dr. Grothey? Or Dr. Hubbard? Go ahead!

Dr. Hubbard: I do agree that a large amount of benefit that we see with metastatic colorectal cancer are from 5-FU–based regimens, and including combining with oxaliplatin and irinotecan. And really optimizing their use has really pushed survival up.

But I think if we looked back at the history over the past 20 years, of survival in metastatic colon cancer, with the addition of the newer cytotoxics and then the monoclonal antibodies, we went from, you know, a survival of about 12 months from 5-FU alone, to a survival well over two years when we combined cytotoxic agents with the monoclonal antibodies.

So I think we all would like there to be more benefit from the monoclonal antibodies. There’s no question. But I think the monoclonal antibodies have really helped us optimize how we use the cytotoxic therapies, for instance, in a maintenance fashion, etc, that really has been pushing the envelope for what we’ve seen for survival.

Oncology: Okay, thank you. All right. I’d like now to get your thoughts on the three principal reasons that Drs. Tenner and O’Neil cited for the rather dim view that they seem to have of molecular targeted agents for colorectal cancer. And these three reasons are as follows:

• One, that they help a relatively small proportion of patients.

• Two, that their adverse effects are not less than those of cytotoxics, as had been hoped, and that because the mechanisms underlying the adverse effects of these new agents are poorly understood, using these drugs makes treatment more complex.

• And the third reason they cite is that there are too few good biomarkers to help properly select the patients who will actually benefit from these agents – and that as a result, these very expensive drugs are being used indiscriminately, with little overall survival benefit, and at great cost to both the patients and the healthcare system.

So what would you say to each of these claims?

Dr. Grothey: I could only partially agree to these statements. So first of all, there are agents like, for instance, VEGF inhibitors; they probably benefit a much larger proportion of patients than we think. In fact, we haven’t seen a patient population that has no distinct benefits from these agents. We don’t know exactly who benefits most or who benefits less; we don’t have a biomarker. But I believe in-tying into the point that Dr. Hubbard made-in terms of using VEGF inhibitors longer-term for maintenance therapy, this actually does drive survival, and can improve survival for a larger proportion of patients than early commentaries actually believed here.

Now with regard to the adverse effects, I strongly agree in some ways that we probably underestimated the toxicities from some of these targeted agents. You know when you talk about targeted agents-you know, the targeted effects on the tumor but less, you know, on the host-and this is clearly not true: all drugs that work have side effects or potential side effects.

Now the question is, can you really compare the side effects of cytotoxic agents with side effects of targeted agents. A drug, for instance, like bevacizumab-probably beyond the initial two or three months-is clearly less toxic than drugs like oxaliplatin. When you use a drug like oxaliplatin, the longer you use it, the more toxicity you will encounter, particularly neurotoxicity.

So there are clearly distinctions in terms of what toxicities you’re looking at, when these toxicities occur, etc. But I think we have all realized that even targeted agents come at a price of toxicity and of course financial toxicity, as outlined in the comment.

Now, I do agree also that some of these adverse mechanisms are poorly understood. For instance, a drug like regorafenib, which is used in the last-line setting, we know it causes or can cause significant hand-foot-skin reaction. We don’t know exactly what drives this hand-foot-skin reaction, so we are, we have no distinct interventions or more specific ways to treat it. It would be very helpful to do that.

Now, in terms of biomarkers and selecting agents, you know this is not only important for the targeted agents; I would like to see biomarkers for cytotoxic agents because I would like to be able to see whether the patients will be able to benefit from irinotecan versus oxaliplatin as first-line therapy. There are some patients who might be 5-FU– resistant, and we shouldn’t even bother with using fluoropyrimidines. And so this is an issue that goes way beyond targeted agents.

Now we are trying to be better, and everything we’ve talked about before really highlights the importance of biomarkers. And for agents that target tumor cells directly, like EGF receptor antibodies, and I think other future drugs that really interfere with the tumor biology, I strongly believe we have a higher chance to find biomarkers like mutations in certain signaling pathways or patterns of disease, or heat expression profiles that will guide us in therapeutic decisions.

But for agents that mainly target the tumor-host interaction, or the host itself, like anti- VEGF therapies, etc, it will be much harder to identify biomarkers.

So everybody screens for biomarkers. And of course, this is a key future area of research, actually an ongoing area of research, to optimize the patient selection for the use of these targeted agents, making agents that they work in a more defined patient population and with a higher benefit and eventually lower costs for healthcare systems. And that is, we are not there yet for most of the agents, not just in colorectal cancer but in a lot of other diseases, in particular when we talk about anti-VEGF therapies.

Oncology: Well thank you. Dr. Hubbard, did you have any points you wanted to make here?

Dr. Hubbard: No. I think Axel said it very nicely. I also have to disagree in parts of the point “a” that they help a relatively small proportion of patients. I think when you think about the EGFR inhibitors, and now with all RAS mutations occurring in about 60% of patients, you know 40% of the 100,000 people with newly diagnosed colon cancer is a lot of patients. So I do think they probably benefit more patients than we realize.

Oncology: Well thank you both very much for being with us today. I think there’s some really good information here that our readers will really appreciate.