Perspectives in Targeted Therapy for Colon Cancer with Scott Paulson, MD
AJMC®: Can you summarize how treatment of colon and colorectal cancer begins with surgical therapy, progresses to surgery with chemotherapy, and then also includes, potentially, targeted therapies?
Paulson: In general, the management of local disease, like early-stage cancers, involves a combination of surgery and possibly chemotherapy, depending on exactly what point it’s at. Targeted therapy up to this point has no role in that setting whatsoever. So, early-stage curable disease is managed with a combination of surgery and chemotherapy. With metastatic disease, it depends on several factors, but often we manage it using a combination of chemotherapy and surgery. The role of targeted therapy in curable disease often includes management of, for instance, cancer that spreads into the liver and/or lungs. It is a bit unclear, but certainly multiple different targeted therapies are integrated with that as well.
AJMC®: What is the current role of genetic testing in colorectal cancer? Which tests are done clinically?
Paulson: At present, the standard of care would be testing for mutations such as KRAS, NRAS, and BRAF. There are a lot of other tests that are done in multiple different centers, and personally, I often perform many of those tests. However, in colon cancer I actually tend to keep that fairly limited and try to be mindful because a lot of the therapeutic implications are really summed up in that first KRAS, NRAS, BRAF analysis. For patients who require more expanded options right up front, often I will do a focused panel of HER2 [human epidermal growth factor receptor 2] analysis and a couple of additional molecules as well, just to try to get a better genetic profile. But for the first pass, it is going to still start with what the NCCN [National Comprehensive Cancer Network] guidelines say, and standard of care.
In colorectal cancer, from the payer perspective, coverage generally does not extend beyond KRAS, NRAS, and BRAF testing. Particularly with colorectal cancer, we tend to get a lot more questions if we are trying to do a more expanded panel up-front. As a result, I tend to stick to the standard of care first, if it is ordained by NCCN, and then move forward from there. For patients who understand the potential financial implications of expanded genetic analysis, I am actually much more inclined to do that than look at HER2 analysis, as well as a couple of other choice molecules as well. There are still cost implications to the patient, and they have to be willing to accept that they may be financially liable for a lot of extended molecular testing. I, personally, do not think it is necessarily right that we are getting too overprotective of a couple of gene samples in a disease that is exceptionally costly. At times, having a genetic profile of a patient with cancer is extremely useful. At present, it is not necessarily something that is always recognized by the payers.
AJMC®: When it comes to use of various chemotherapeutic regimens, when you are choosing a regimen and you are choosing candidates for surgical therapy, what kind of criteria do you use? What are some of the interesting edge cases in terms of selecting patients for different modes of treatment?
Paulson: Therapy for colon cancer has gotten so much better, and appropriate selection of patients has improved significantly enough. The only edge cases are patients who are either unwilling or physically unable to tolerate combination chemotherapy. These are generally elderly patients who are well into their late 70s and early 80s, who don’t have a lot of functional reserve, and they may be jeopardized if put on combination chemotherapy.
The other edge cases that you will run across are patients who are just flat out unwilling to undergo chemotherapy. They’ve either had family members who’ve gone through it, they’ve seen toxic effects from it, or they’re just interested in different holistic approaches. These would also be patients who I would not consider to fall under the standard treatment algorithm. However, this can be a little challenging because, from a payer perspective, you’re not going to be able to just do whatever you want. You still have to follow some degree of established treatment protocol. Those tend to be a little more challenging if patients are simply not interested in following standard of care.
AJMC®: As far as use of chemotherapy in combination with targeted regimens, how do you make treatment decisions about using those treatments in combination, and how are those discussions coordinated with patients? What are challenges from a managed care perspective in terms of using targeted treatments?
Paulson: Insurance is not as much of an issue in coverage of targeted therapy as long as you’re following established guidelines, provided that you’re not treating KRAS-mutant patients with EGFR inhibitors or doing anything that’s considered inappropriate treatment. I don’t think you frequently run across insurers flat out refusing to pay for targeted therapies that are more expensive. Some very selected, very managed plans will only allow you a certain amount of cetuximab every 3 months. I have seen that, where you have to be judicious about your use of it. That‘s exceptionally rare, though. I would say as long as you’re following standard procedures, you don’t really run into many challenges with the payers. FOLFIRI [folinic acid, fluorouracil, and irinotecan] is actually very cheap, FOLFOX [folinic acid, fluorouracil, and oxaliplatin] is also fairly cheap relative to the targeted therapies, but a little bit more expensive.
So, getting and securing combination chemotherapy is dependent on the targeted therapies, and is very, very rarely a challenge. But I don’t tend to run into that much of an issue except from exceptionally managed plans when using targeted therapy, provided you’re following guidelines.
AJMC®: When you are selecting, for instance, an EGFR inhibitor or a VEGF inhibitor, how do some of the indications for use of those treatments vary? For instance, panitumumab varies in its indication for use versus cetuximab: Are there any limitations that you run into as far as the specific indication for the product?
Paulson: You mentioned very specific examples. Essentially, [these are] what I would view as similar medications. There are certainly established medical data in different settings for each of them. When I look at those data, I find that each drug could essentially be interchanged in most clinical settings. I feel that the payers certainly agree with that as well. They are both extremely costly. So, that choice is often just dictated by whatever is most appropriate for the situation.
Panitumumab and cetuximab are similar enough drugs that I feel like a lot of the evidence in both of their settings can be applied to either drug. I think the VEGF inhibitors are a little bit more challenging, just because there tends to be a much broader difference in price. Ramucirumab is a much more expensive drug at present, and bevacizumab and ziv-aflibercept are less expensive. So, I think focusing on those different drugs is still important, certainly when you’re looking at the second-line setting.
When it comes to choosing the differences between the VEGF inhibitors that are used in combination with chemotherapy, that’s a little more important than choosing between the 2 or choosing between the 3. Whereas, when it comes to panitumumab and cetuximab, their prices do vary somewhat, and I think it’s better to go with a more cost-effective option, certainly. But it’s less of a clinical piece than, for example, the difference between ramucirumab and the rest.
AJMC®: How do some of the differences in the trials with targeted therapies factor in to making a decision as far as selecting patients for a specific treatment? Or is it more of a selection based on general classes?
Paulson: If you had asked me 3 years ago, I would have said that you have to follow the cookbook recipe to an absolute fine point, and that if it’s not specifically provided for by the indication, then you absolutely cannot vary it. I’m starting to call that into question. I’m not saying that I’ve certainly changed my perspective on it. I like to follow the evidence as it is displayed. However, we’re running into situations where the cost of drugs continues to increase and we need to be mindful of finding more cost-effective options, and be able to integrate them into patient care in what we would feel is an appropriate way, even if it doesn’t necessarily meet the indication. So, I’d say that that’s evolving.
Another thing that’s evolving as far as choice of targeted agents and how we sequence everything is right-sided versus left-sided data, where I don’t think the entire molecular picture of the tumor is quite complete yet. I hope to see further data coming out on that, which will probably be able to help us choose patients extremely effectively right from the start, and treat them in a sequence where we can expand survival. We’re still trying to figure out exactly which patients are best served by an upfront EGFR inhibitor versus a late EGFR inhibitor, and how long we can keep people on vascular endothelial growth factor inhibitors. These questions, I don’t think, are perfectly answered yet, but I do believe that they’re going to continue to evolve. Throughout all of that, we have expensive drugs that are being tagged on to therapy at each point. Insurers are trying to make smart, sound decisions about how to prescribe the low-cost targeted therapy in a reasonable way, even if it doesn’t stick exactly to what we’ve done in the study.
AJMC®: Regarding left-sided versus right-sided tumors, 1 piece of data comes from a subanalysis of the CRYSTAL study, in which patients with left-sided tumors had much better response than patients with right-sided tumors. What are some possible reasons for that, and what does that say about targeting and subgrouping patients more precisely than current practice?
Paulson: You have to pay attention to it. What does it mean on a molecular level, is the key question, in my opinion. There are people who are working hard to answer those questions for us, thankfully. I think it’s an important distinction to look at, and what it tells you is that right-sided tumors just genetically tend to have more of a bad actor behind them. This does affect the sequencing of an EGFR inhibitor. It’s already a questionable piece to move immediately into the front line, and I think it’s much tougher to start off a completely wild-type patient with an EGFR inhibitor for their right-sided tumor. It’s important when making these decisions. I’d be much more interested to know what’s happening with those patients on a molecular level. Patients with left-sided tumors do exceptionally well on EGFR inhibitors. Why are they so susceptible to that? Those are the questions we want to answer. You can still have genetically right-sided tumors that appear on the left side of the colon, and vice versa. You can’t always define the tumor by drawing a line down the middle of the colon, although there is certainly a marked statistical difference.
AJMC®: What is the role of KRAS testing, and how early can you test a patient for KRAS so that targeted therapy can be used as soon as possible?
Paulson: I’m glad you brought that up, because I think that the landscape is continuing to change. In terms of pathology, it quite explicitly states that as soon as they have a patient’s tissue and see that it has metastatic colon cancer, the pathologist is essentially empowered to reflexively test those against an appropriate metric, such as KRAS, NRAS, and BRAF. The final step would be to test for microsatellite instability (MSI), because we want to try to get that information into the hands of oncologists as quickly as we can.
MSI testing is another important piece to therapy. Those results should be considered by oncologists as they contemplate chemotherapy. It’s gotten fast enough now where you can often get those results in a timeframe that’s effective, where the patient doesn’t need to start treatment. So, if someone is profoundly systematic and has a large burden of disease that needs to start chemotherapy right away, sometimes the oncologist can’t necessarily sit back and wait for those test results or collect test results on tumor tissue that was collected too long ago, or repeat a biopsy. Sometimes you have to make the decision based on the information you have on hand. However, for the most part, our pathologists have been very engaged in that, and we’re trying to encourage our pathology departments around our state [Texas] to also be engaged on that, because knowing your RAS status, knowing your BRAF status, and knowing the microsatellite instability status and mismatch repair protein testing of your particular tumor, is critical for how you look at the global treatment of that patient.
AJMC®: What are some important takeaways with regard to microsatellite instability status, and what does it affect, in terms of treatment decisions?
Paulson: For now, it is evidenced by the FDA’s approval of the use of pembrolizumab for any microsatellite-unstable cancer, which very closely follows the rapid approval in colon cancer based on a handful of patients, that if you use these drugs in MSI-high patients, they can be complete game changers.
Another important implication is that there’s a large national study currently going on for front-line MSI-high patients. The accrual in the United States has been difficult, in that most oncologists are very comfortable starting those patients on a front-line chemotherapy regimen, rather than looking at a randomized study for them. And so, there are some implications, and they’re important to look at, and it’s certainly exceptionally important to know the long-term status, even if you don’t decide to try to enroll the patient in the trial in the front line. This is because it opens up more immunotherapy options for these patients, whereas, otherwise, they would not have an option.
AJMC®: What are some of the treatment options for patients with metastasis, whether it occurs in the brain, lung, or liver? What will influence your treatment decisions? Which types of cases in patients with metastatic cancer will pose limitations to targeted therapy?
Paulson: As it relates to targeted therapy, I don’t think the picture is very clear—we are getting mixed messages from different trials. Some of the EPOCH data would suggest that using an EGFR inhibitor led to a trend in reduced survival in patients who were planning to undergo surgery in any operative setting, whereas other data suggest that higher response rates improve resectability outcomes. So, there are mixed data regarding the efficacy of targeted therapy in the planned surgical patient.
How does that affect my decision? If I’m looking at some of these data and I know that they’re surgically resectable up front, and the patient has synchronous disease and an isolated liver metastasis that needs surgery, I would actually treat that patient with just chemotherapy up front; I don‘t necessarily apply any targeted therapy in addition to it. We tend to try to abbreviate the amount of systemic therapy that these patients get, to limit significant toxicity to the liver as they’re going to undergo liver resection. In lung, where you’re getting less toxicity from a lot of the chemotherapeutic agents, there’s certainly a lot more flexibility. Frequently, I’ll treat those patients with a VEGF inhibitor as well, but there are much less data to guide physicians around targeted therapy selection for lung metastasis protection. It’s a less common scenario, and therefore, it’s a lot less compiled.
Right now, as far as choice of targeted therapy for a resectable patient, I don’t think there are enough data to guide management; it has to be taken on a case-by-case basis, looking at the subtotal of everything. Is the tumor left-sided or right-sided? Should we be changing our tune on how we use targeted therapy based on that? I don’t know. I don’t have a specific algorithm. I tend to treat these patients who I know are going to go to surgery with chemotherapy alone up front. In this case, the use of targeted therapy is based on very specific clinical indicators. It’s not possible to have an algorithm for that; it’s a complex approach.
AJMC®: How does some of the uncertainty there potentially leave the door open for differences between specific agents within targeted therapy classes, if at all? Are there any possibilities there, in the future, for finding more specific populations that are appropriate for a given treatment, or do you think it’s more likely that we have to take a more pragmatic approach now, and that will continue to be the case?
Paulson: I think, hopefully, we will eventually find the right population. That’s where I think we need to head with it. Which left-sided and wild-type tumor with an isolated hepatic metastasis is really going to benefit from an EGFR inhibitor to make them a suitable resection candidate? We need to nail down the genetics of it. We have conflicting data from multiple different studies that have been done over the past number of years. I think it’s going to be a matter of figuring out a better profile of these patients, and having those data so that you can make sound decisions. For now, we just have to take it on a case-by-case basis, trying to see what’s best for that patient. A lot of the choices of agents can be dictated by patients who know what their toxicity profiles are and what they could handle and tolerate on a day-to-day basis. You never know when you’re going to run into a patient who will absolutely not tolerate losing their hair or getting a skin rash, and then all of a sudden the FOLFIRI plus cetuximab combination is out of the question.
AJMC®: In terms of choosing therapy, it sounds like there are a lot of variables, not least of which is the patient’s preferences and expectations. How you go about setting expectations for therapy for patients within the metastatic setting?
Paulson: It’s a challenging thing, because even patients who you do get to resection have a very high recurrence rate. Someone who will have an isolated liver lesion removed along with their primary tumor, they have metastatic disease, and especially if that’s showing up right at the time of diagnosis, even with a complete resection, the chances of recurrence are still more likely than not. That’s challenging because, if a patient is going to be going through a major amount of toxic chemotherapy followed by an exceptionally large surgery, you’re at least hoping that you can get to the point where you’re not going to have any cancer left in the body. So, the question becomes: How do we manage those expectations? We try to be as open and honest as we can, and then try to put that aside and go for as big a win as we can possibly try to get out of a tough situation. For the most part, that works. I think as soon as a patient hears that they have stage IV cancer, they realize what they’re up against and they also don’t want to get so aggressive that all quality of life is completely stripped, and that can happen. So that requires frequent check-ins and reassurance, but some people develop pretty debilitating neuropathy and skin rashes, with chemotherapy and targeted therapy. In addition, surgical morbidity is certainly not small if they’re having a large liver resection.
AJMC®: What are some key takeaways that managed care professionals needs to take into consideration to help oncologists and patients? How can they help more in making it easier for all parties involved?
Paulson: I think a more relaxed approach to molecular testing needs to be considered. I’m not saying that everyone should have full-genome analysis of their tumor, but I do think it would be beneficial to have a much more relaxed approach to looking at more targeted agents and expanding the normal sequencing of what we practice as standard of care, which at present is NRAS, KRAS, BRAF, and MSI testing. The reason is, if you look at the cost of doing an extended genome analysis on a tumor, the cost can be in the thousands, but that’s still about one-third of the cost of 1 month of some of the vascular endothelial growth factor inhibitors, and certainly the EGFR inhibitors. If you look at how you can more effectively select these patients, you’re going to drill down more and more so that you’re treating patients properly with targeted therapy, that is a wise decision. I’m not someone who recommends that everyone get their tumors sequenced right up front, but from a payer perspective, it does make sense when you look at the relative cost of having that information versus the cost of just a single dose of cetuximab or panitumumab; it makes sense because you’re going to start to understand and learn things more. And if you get that more and more out there in the hands of physicians, we will learn more. That may not be what the payer wants to hear, but it’s certainly something that I think makes sense in the long run. The biggest cost is the cost of drugs, and being practical about that and being better able to select the patients who will benefit from extremely expensive drugs, is important.
If I can get 1 piece of advice out there, I would say that we need to be more practical about genetic testing upfront, and less restrictive rather than more.
AJMC®:What do you see as the future of treatment of metastatic colon cancer and colorectal cancer? How do you see it changing over the next 5 to 10 years?
Paulson: Back to what I was saying regarding left-sided and right-sided tumor data, and all of the increasing understanding that we have of advancing genetic testing, better molecular characterization of tumors, where you’re selecting out BRAF mutations—I think you’re going to start to get patients with metastatic colon cancer placed into a number of different subsets.
From a clinical perspective, there are many very different types of colon cancer. There are some that are exceptionally aggressive; you can’t do anything about them. There are some patients for whom you can manage disease for a number of years. We’re going to be able to better understand why we do 5 to 10 years of management with 1 patient, whereas with another patient everything that we give them doesn’t affect the disease. I think you’re going to start seeing that you have multiple different buckets of different types of colon cancer, and that you’re going to be better able to link up patients with targeted agents. I think immunotherapy and MSI are scratching the surface. A small percentage of colon cancer patients qualify for immunotherapy and account for up to 5% to 15% of the total cancer population of 140,000 people in colorectal cancer, and that’s not a small amount. I think the future is going to be interesting, and that the typical approach of FOLFOX plus a targeted intervention, to FOLFIRI plus another targeted agent, and then onto oral agents, will not necessarily be the model. I would tell you that it’s probably on the horizon within the next 5 to 10 years.