focused on metastatic colorectal cancer. Throughout this presentation, my colleagues and I will review strategies for managing 4 clinical cases based on recent evidence as demonstrated in clinical trials.

I&rsquo;m Dr Shubham Pant, an associate professor in the Department of Investigational Cancer Therapeutics and an associate professor in the Department of Gastrointestinal Medical Oncology, Division of Cancer and Medicine, at the University of Texas MD Anderson Cancer Center in Houston, Texas.

Also joining us in this discussion are Dr Richard Kim, associate member of the Gastrointestinal Oncology Department at Moffitt Cancer Center and associate professor of oncology at the South Florida College of Medicine in Tampa, Florida; Dr Michael Morse, a professor of medicine and professor in the Department of Surgery and a member of the Duke Cancer Institute in Durham, North Carolina; and Dr Gregory Riedlinger, assistant professor of pathology in the Division of Translational Pathology at Rutgers Cancer Institute of New Jersey, Robert Wood Johnson Medical SchoolRutgers University in New Brunswick, New Jersey. Thank you for joining us. Let&rsquo;s get started with our first case.

-mutant metastatic colorectal cancer. This is the history and physical: A 73-year-old Caucasian male sought medical attention after experiencing bloody diarrhea. His past medical history is remarkable for hypertension, which was managed with losartan. He&rsquo;s active and can perform daily activities without restriction. A colonoscopy showed a fungating mass in the descending colon, which was biopsied.

The pathology findings for this patient revealed invasive, poorly differentiated adenocarcinoma. Molecular testing showed

V600Emutant and was microsatellite stable.

The CEA [carcinoembryonic antigen] was 6.4 ng/mL. The CT [computed tomography] of his chest, abdomen, and pelvis showed a 9.5-cm mass in the descending colon, small bilateral lung nodules, and a 3-cm mass in the right lobe of the liver. His surgical consult for metastasectomy indicated unresectable disease.

First-line treatment with FOLFIRI [folinic acid, fluorouracil, and irinotecan] and bevacizumab was initiated. Imaging at 3 months showed stable disease in the liver and lung nodules. After 6 months of therapy, the patient complained of increased fatigue, requiring the need for frequent rest. A CT scan showed an increase in the size of the liver nodule and the appearance of 4 new small liver lesions.

I&rsquo;m going to go through some of the data related to

-mutant metastatic colorectal cancer. The first was presented by Scott KopetzSWOG S1406. The SWOG S1406 trial was a randomized trial in patients with

V600Emutant colorectal cancer. The arms were cetuximab with irinotecan, which is considered the standard of care. This was with or without vemurafenib as an add-on in the treatment arm. This was, again, for patients with a PS [performance status] of 0/1. The important thing to mention is that these patients had received 1 or 2 prior systemic therapies for metastatic or locally advanced unresectable disease. We saw a median progression-free survival benefit of 4.4 months versus 2 months with a hazard ratio of 0.42. The

value was statistically significant: .0002. The overall response rate was 16% versus 4%, and there was a better disease control rate: 67% versus 22%.

The other trial, which is a precursor to the new trial that is going on, is the COLUMBUS trial. This looked at

-mutant melanoma with a combination of encorafenib and binimetinib, which is also relevant in colorectal cancer, as the next few slides will show us.

This looked at the progression-free survival of encorafenib plus binimetinib versus vemurafenib in melanoma, and this combination has been approved by the FDA for

V600E patients with metastatic melanoma. This is relevant to our patients with colorectal cancer because of the BEACON trial that is currently ongoing, which we saw earlier results on at this year&rsquo;s ASCO [American Society of Clinical Oncology] Gastrointestinal Cancers Symposium.

This is a phase III randomized controlled trial of patients with

V600Epositive metastatic colorectal cancer and progression on 1 or 2 prior regimens of therapy. This is a combination of encorafenib, which is a BRAF inhibitor; binimetinib, which is a MEK inhibitor; and cetuximab, which is an EGFR inhibitor, versus irinotecan/cetuximab or FOLFIRI/cetuximab.

The presentation was on the trial safety lead-in of 30 patients. Even in these patients, there was an overall response rate of 48% and a median progression-free survival of 8 months. Most of the grade 3/4 adverse events were fatigue, UTI [urinary tract infection], increased AST [aspartate aminotransferase], and increased blood creatine phosphokinase. About 3% of patients had grade 3/4 rash.

Case 1: BRAF-Mutated Metastatic Colorectal Cancer

-mutant colorectal cancer patient who has failed frontline therapy. Greg, can you tell us about the importance of genomic testing in colorectal cancer?

Genomic testing has become standard of care for metastatic colon cancer. The NCCN [National Comprehensive Cancer Network] guidelines now recommend that all metastatic colon cancer patients be tested for

testingnot only exon 2 but also exon 3 of the

genes, because those are shown to correlate with lack of response to anti-EGFR therapies.

Wonderful. Richard, do you test your patients? When that frontline metastatic colorectal cancer patient comes to you, what all do you tend to test for?

I think molecular profiling is a must for any patient who&rsquo;s diagnosed with advanced colorectal cancer. It&rsquo;s recommended by the NCCN and ASCO [American Society of Clinical Oncology] guidelines. The typical molecular profile we do up front looks at the

status to see if they&rsquo;re a candidate for an EGFR drug. Otherwise, we test for

has some significance. First, it is a prognostic marker. We know that a patient with a

mutation tends to have a very poor prognosis. Their median survival is not your typical 2 to 3 years. It is probably less than thatanywhere from 18 to 20 months. For those patients, we tend to be very aggressive. The third type of testing we do is for immunotherapy&mdash;MSI [microsatellite instability] status.

and then we do MSI testing. Mike, in your practice, when this result comes back and you see that the patient has a

V600E mutation, what type of discussion do you have with your patient? What do you tell them? Do you say, &ldquo;You&rsquo;ve got a

mutation. I&rsquo;m going to start you now?&rdquo; Do you wait until the second line or third line? How do you treat your patients who have

V600E-mutated colorectal cancer if you&rsquo;ve tested them in the frontline, and you know that the mutation is there?

One of the challenges that we face right now with all of this molecular testing is trying to explain it to our patients. This is terminology that people typically haven&rsquo;t heard before. The first thing is just to try to get people to understand their disease and what the meaning of these results is. For the average nonscientific person, we&rsquo;ll explain that they have a certain mutation in their cancer, which has some prognostic significance, as Richard pointed out. There are therapies that can now be used to target that. However, as we all know, the data, as you presented, are in people who had 1 or 2 prior therapies. Even with the NCCN guidelines, people really should get a standard therapy in the first line and then be considered for something that targets the

genes, subsequently. Basically, we explain that we&rsquo;re going to give a standard therapy, but if that doesn&rsquo;t work, we have optionseither on study or off study&mdash;for targeting that mutation.

In the first line, Richard, that&rsquo;s what you treat them withlike a chemotherapy with a backbone of&hellip;

mutation has a poor prognosis. We want to be aggressive. These are not the patients who are going to get first-, second-, third-, or fourth-line therapy. If they&rsquo;re young and they&rsquo;re healthy enough to tolerate chemotherapy, I tend to be aggressive. Based on the TRIBE study, I would technically go with FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin, and irinotecan] plus bevacizumab. It will be my first-line choice, assuming they&rsquo;re able to tolerate it.

Then, once they fail that, like Mike said, that&rsquo;s when you would try to bring in some of the BRAF inhibitors plus an EGFR drug. Once again, the data that you just presented show a response rate of 48%. If you look at the patient who actually progressed after 1 night of therapy, the response rate is actually higher60%. In the second-line setting, 60% response is pretty high. So, in those patients who failed FOLFOXIRI plus bevacizumab as second-line therapy, based on the BEACON study&mdash;once again, it was a very small study&mdash;I would tend to try to get the BRAF inhibitor plus the EGFR drug, if possible.

Mike, you talked about the NCCN guidelines. I understand that it has been incorporated into the NCCN guidelines&hellip;?

Yes, based on the SWOG data. The NCCN guidelines really talk about using the combination that was used in that studyirinotecan, cetuximab, and vemurafenib. Obviously, we eagerly await the results of the combination of encorafenib, binimetinib, and anti-EGFR therapy to see what the future may bring for these patients. Right now, the challenge is always the insurance and whether insurance will cover these off-label uses. I think we have a stronger case now that the SWOG randomized trial was actually reported out.

So what we are saying is, it&rsquo;s not FDA approved, but it is part of the NCCN guidelines.

Case 1: Genomic Testing & NCCN Guidelines for Metastatic CRC

Richard, the SWOG S1406 study was irinotecan, vemurafenib, and cetuximab. BEACON is a combination with a MEK inhibitor. Do you think that makes a differencelike BRAF, MEK, and EGFR inhibition? There&rsquo;s no chemotherapy backbone to that.

Right. If you look at the preclinical rationale, patients with

, and MAP [mitogen-activated protein] kinase overactivated. In the melanoma data, you show that you could use a MEK inhibitor plus a BRAF inhibitor and have a very good outcome. In colon cancer, that&rsquo;s probably not true because of the feedback activation to

. To overcome the resistance, the importance is probably not in the chemotherapy but in the EGFR drug.

The BEACON study doesn&rsquo;t add chemotherapy to the backbone of the targeted therapy, but, if you look at the rationale, I think it is important to add the EGFR drug on top of the MEK and BRAF inhibitors. So the triplet regimenI think there&rsquo;s a good rationale behind it. Obviously we&rsquo;ll wait to see the results of the BEACON study, but so far so good, in terms of what has been presented.

Before, these patients had very poor options. Right now, what would you treat them with, Mike? What&rsquo;s your frontline regimen for patients who come in with a

As Richard pointed out, because they respond so poorly and have a very short progression-free survival with standard treatmentsuch as this patient, getting FOLFIRI [folinic acid, fluorouracil (5-FU), and irinotecan] plus bevacizumab&mdash;if somebody can tolerate it, I would agree that FOLFOXIRI [folinic acid, 5-FU, oxaliplatin, and irinotecan] is probably preferred. Now, I don&rsquo;t think that the TRIBE trial that used FOLFOXIRI plus bevacizumab necessarily proves that it is a regimen uniquely suited to

-mutated tumors. However, people tended to do better. If they had a

-mutated tumor, they tended to do better with that combination versus the 2-drug chemotherapy regimen. This suggests that a very aggressive chemotherapy combination is probably needed for people with a very aggressive cancer.

Can you tell us a bit about the TRIBE study, Richard?

TRIBE was a randomized study comparing FOLFOXIRI plus bevacizumab versus FOLFIRI plus bevacizumab. The progression-free survival of FOLFOXIRI plus bevacizumab was over 12 months. This is one of the rare studies for which you have a progression-free survival that&rsquo;s greater than 1 year.

In the subset analysis, they looked at patients with

-mutant cancer. In that study, the patients who got FOLFOXIRI plus bevacizumab tended to do better compared with those who received FOLFIRI plus bevacizumab, so there was a slight advantage. Once again, this was not the primary endpoint of the study, but, based on that, it seems like because this type of patient tends to be more aggressive, tumor-wise, we tend to go with a more aggressive approach. There are other data that show that the triplet regimen plus bevacizumab seems to be better than the doublet plus bevacizumab in the small subset population.

Mike, when you use it, what is the regimen?

It&rsquo;s 5-FU, leucovorin, irinotecan, and oxaliplatin. This is not to be confused with FOLFIRINOX, which combines the same drugs but with slight differences in the dosing. The challenge for many people, like this older individual, is obviously that there is greater toxicity. There is more of a risk for neutropenia. There is more of a risk for diarrhea and so on. So it&rsquo;s not ideal for every patient. Like anything else, you have to look at comorbidities and performance status before using it.

So if you were not using FOLFOXIRI, Richard, what would you use in the frontline? Let&rsquo;s say they are not a candidate, in your clinical judgment, for FOLFOXIRI. What else would you use?

In the first-line setting, I still think chemotherapy is the answer. If the patient is not a candidate for the triplet regimen, I would probably still go with either FOLFOX or FOLFIRI plus bevacizumab. I would not use EGFR drugs. A lot of data coming out suggest that patients with

-mutant tumors tend to be resistant to EGFR drugs. So I would definitely not use it in the first-line setting. I think either FOLFOX or FOLFIRI plus bevacizumab will be my choice in patients who cannot tolerate FOLFOXIRI.

For our physicians/clinicians out there, we&rsquo;ll talk a lot more about tumor sidedness, microsatellite instability [MSI], and patients who don&rsquo;t have any molecular alterations later on in our discussion. But specifically for patients with

-mutant colorectal cancer, what I&rsquo;m hearing is, if a patient comes to us, we have to sit down and explain to them that there are no data for treating them in the frontline with a better drug that inhibits the BRAF pathway. What we do is give them FOLFOXIRI with bevacizumab, or FOLFOX or FOLFIRI with bevacizumab, depending on the tolerability in the clinical scenario. If they then feel that we should try, at least in the second-line setting, to think about combining themsimilar to the SWOG study&hellip;We eagerly await the BEACON results to see if we can get a great combination therapy for our patients.

I would agree. You mentioned 2 key points. First, you have to test for it. If you don&rsquo;t know about the

status, you can&rsquo;t determine whether this is an option for them. I still see patients who get

testing done. Secondly  and we&rsquo;ll be talking about MSI later &mdash; a certain fraction of patients who have

mutations are also MSI-high, and we&rsquo;ve managed those people differently.

Richard, before we end, what about atypical

mutation. I think one-fifth of that percentage are probably not

V600E. Those have been looked at in small studies with, obviously, very small numbers. So the prognostic value is unclear. Actually, there are data that suggest that non

V600Emutant patients tend to have a better prognosis than patients who have

But in terms of whether there&rsquo;s a predictive value to EGFR inhibitors or not is unclear. This has not been studied yet. Right now, I think it&rsquo;s too premature. I think we could say that, at least from the data we have, there is probably a better prognosis. However, the predictive nature of this non

V600E mutation to the EGFR drug or this BRAF combination or MEK inhibitor is unclear because the trial only included patient with

Greg, it&rsquo;s also important, as a pathologist, that you look at the second part of all mutations. Just because it says

doesn&rsquo;t mean that you don&rsquo;t need to look at the second part of it.

mutations have been described. We know of the canonical

is signaling as a monomer. And then there are other

-activating mutations that theoretically function as a dimer through

. And then there are the so-called class 3

mutations, which are actually inactivated or are kinase dead. Knowing the differences between those and the different treatment options is obviously important.

Thank you so much. This has been a great discussion. So our take-home points are: Test for these mutations that we&rsquo;ve talked about. Don&rsquo;t just go and look at

and go &ldquo;I can treat this with a BRAF inhibitor,&rdquo; because it&rsquo;s not approved for that. It&rsquo;s only approved for

V600E mutations. In the frontline, you can choose FOLFOXIRI with bevacizumab, or FOLFIRI or FOLFOX with bevacizumab, or chemotherapy, depending on the clinical status of the patient.

In the second-line setting, once they&rsquo;ve failed, based on the SWOG data and the NCCN [National Comprehensive Cancer Center] guidelines, we can try to combine it with a BRAF inhibitor, such as vemurafenib. We are really waiting on the results of the BEACON trial, but this is an exciting era for our patients. Two years back, we really did not have anything for our patients. It&rsquo;s a great step forward. Thank you so much.

Case 1: Treatment of BRAF-Mutated Colorectal Cancer

The second case I&rsquo;ll be presenting is on a patient with advanced colon cancer with no molecular alteration. This is a 35-year-old gentleman who was referred to a gastroenterologist for symptoms of acute abdominal pain and intermittent constipation. His physical examination shows some abdominal distension, decreased bowel sounds, and pain in the left side of his abdominal area. A CT scan shows focal enhancement and some thickening of the sigmoid descending colon junction with pericolic fat stranding and 3 small liver lesions that are greater than 1 cm.

A scope was done, and a biopsy confirmed well-differentiated adenocarcinoma. The molecular testing was done as per the guidelines, and this patient was found to be

wild type. Consequently, the patient was started on FOLFOX [folinic acid, fluorouracil (5-FU), and oxaliplatin] plus cetuximab as palliative chemotherapy.

Now I&rsquo;ll be discussing some of the data that was recently presented at the ESMO [European Society for Medical Oncology] Congress. This explored the benefits of upfront primary tumor resection according to sidedness in metastatic colorectal cancer. This is a retrospective analysis of the MACRO-2 and the PLANET randomized trials.

wild type and had upfront resection of their tumors. In patients with stage IV disease, upfront tumor resection was associated with improved overall survival. However, it was significant only in right-sided tumors. The median overall survival in patients with right-sided tumors that got upfront resection was 20.9 months, versus 10.6 months. The patients with left-sided tumors had a better overall survival regardless of upfront resection or not. As you can see, the hazard ratio was 0.4 with a

value of .006. This tells us that for left-sided tumors, upfront resection didn&rsquo;t really matter.

Now we&rsquo;re going to look at the primary tumor location as an independent prognostic marker from molecular features for overall survival in patients with metastatic colorectal cancer. This was an analysis of the CALGB/SWOG 80405 study.

This study determined that in 728 patients, sidedness was an independent prognostic factor when adjusted for age, gender, CMS [consensus molecular subtypes], MSI [microsatellite instability], and

status. Patients with left-sided tumors did well overall. The overall survival was close to 33 months compared with 20 months in patients with right-sided tumors. The question is, is sidedness also a predictive marker for possible biologic therapy?

In this study, in patients who got bevacizumab plus chemotherapy, there was a 38-month overall survival in the left-sided tumor group compared with patients who got cetuximab-based therapy, who had a slight improvement in overall survivalabout 40 months. So there was an improvement of about 2 months in overall survival. However, in the right-sided tumor group, patients who got bevacizumab plus chemotherapy had an overall survival of 34 months. Patients treated with cetuximab had a worse overall survival&mdash;18.4 months.

Next I&rsquo;ll be discussing the VALENTINO study, which is maintenance therapy with panitumumab. This is a first-line study of FOLFOX plus panitumumab, followed by 5-FU/leucovorin plus panitumumab, or single-agent panitumumab as a maintenance therapy in a patient with

wild-type metastatic colorectal cancer. This was presented at our meeting this year. Patients with

wild-type unresectable metastatic colorectal cancer were enrolled in the study. A total of 229 patients were randomized to FOLFOX plus panitumumab as induction therapy for 8 cycles. When their disease was stable or they showed a partial response, they were randomized to a maintenance arm with panitumumab alone or panitumumab plus 5-FU/leucovorin. They continued on until progression or unacceptable toxicity. The primary endpoint of this study was noninferiority of progression-free survival of arm B versus arm A.

For secondary endpoints, they looked at safety, response rates, duration of response, and overall survival. At the median follow-up of close to 14 months, the 10-month progression-free survival of the panitumumab arm was 52.8 months, versus 62.8 months with the combination of 5-FU plus panitumumab. The median progression-free survival favored the combination of 5-FU and panitumumab by close to 3 months10.2 months versus 13 months. The overall response rate was very similar. During maintenance therapy, the grade 3 adverse events that occurred were diarrhea, mucositis, handfoot skin reaction, and neutropenia. Grade 1 or 2 adverse events were 75% in arm B versus 64% in arm A.

Case 2: Metastatic Colon Cancer, No Molecular Alterations

In the review of this case, we have a very young gentleman with the diagnosis of a left-sided tumor that is

wild type. Obviously, part of the molecular profiling that is missing is MMR [mismatch repair] status. I would probably check for that because the patient is young, even though it doesn&rsquo;t seem like the patient has any family history of colon cancer.

In the case of a left-sided tumor, there is some controversy. What do you do for a patient who has a left-sided tumor with

wild-type disease? Do you use EGFR-based therapy or VEGF-based therapy? This is a question that I&rsquo;ll discuss with the patient. Based on the PEAK study, the FIRE-3 study, and the CALGB 80405 study that I just presented, there seems to be a benefit in using the anti-EGFR drug over the anti-VEGF drug in left-sided tumors.

However, the main issue with using EGFR therapy in the first-line setting is toxicity and the question of maintenance therapy. Therefore, I would have a discussion with the patient about the possible overall survival benefit when using the anti-EGFR drug but would also discuss the side effects. We know that with

, there is a risk for paronychia if you are on the EGFR drug for a long time. So I will have a long discussion with the patient, and then we will decide whether to go with an anti-EGFR drug or an anti-VEGF drug.

Richard, thank you so much. This is a case for which we have a lot to discuss. Greg, first I&rsquo;ll come to you. In these patients who have wild-type disease, should we look for other mutations or not? Microsatellite instabilitydefinitely. I think you would all agree that we should look for that. Is there anything else that we should look for in this patient?

It&rsquo;s a little concerning that the patient is 35 years old. This is a younger patient. So certainly, as is the routine standard of care, you would want to make sure that you looked at the mismatch repair proteins. You also want to make sure there&rsquo;s a good family history. You may even, potentially, refer the patient for genetic counseling, just to make sure that there&rsquo;s not some type of syndrome going on. There could be additional testing there. In terms of anything you know past that point, there&rsquo;s generally smaller percentages of potentially actionable mutations that larger gene panels, such as next-generation sequencing-based testing, may be able to pick upthese types of alterations. If you&rsquo;re going to do these types of tests, the patient really needs to be managed at an academic medical center, where they&rsquo;ll potentially be equipped to deal with some of these rare alterations.

Greg, you brought up a great point. I&rsquo;m going to ask you a question that we sometimes get asked. Can you tell us the difference between germline and somatic mutation testing? How are they different? Nowadays, we do a biopsy and we pick up something on next-generation sequencing. Can you explain the difference between those?

Germline testing essentially tests every cell in the patient&rsquo;s body for a specific DNA alteration. Somatic testing generally refers to cancer-specific testing. A lot of places are doing so-called tumor-only testing. They don&rsquo;t necessarily have a germline sample. With somatic testing, you may also be picking up the germline alterations. There are different algorithms that go into some of these tests, where they can sometimes try to infer that something may be germline versus somatic only in the cancer cells. That is never 100% accurate. It can be used only to give you an idea. In terms of somatic testing, you generally want 20% tumor cell nuclei to ensure that you don&rsquo;t have false-negative results. So heterozygous mutation frequency of 10% in 20% tumor nucleimost of these tests have a sensitivity to at least 5%, so you wouldn&rsquo;t really worry about missing things as long as you have that amount of tumor and it&rsquo;s a solid tumor specimen.

Mike, I&rsquo;ve seen younger patients with colorectal cancer in my clinic. What do you counsel them on? Do you send them for genetic testing? I think that&rsquo;s a very important point that you brought up, Greg, because sometimes these patients may show up and probably do need to be referred to see a genetic counselor.

We are seeing an increase in younger patients, younger than age 50 now. That&rsquo;s certainly been cause for concern because screening recommendations may need to be altered. In fact, there are now recommendations to lower the age of screening for colon cancer to age 45. So this is an important consideration when we discuss the ramifications for patients themselves, as well as their family members. I agree that in somebody with an unusual presentation, if there is a family history or if they are particularly younger in age, we should certainly offer a visit with a genetic counselor to discuss other options.

At what point would you think it&rsquo;s important to test for what is probably the next most common mutation? It&rsquo;s really not a mutation but an amplification of

We do a lot of larger panel testing. I&rsquo;m not sure, off the top of my head, what the percentages are. You can probably correct me, but I believe it&rsquo;s usually somewhere around 2%.

amplification. There&rsquo;s an FDA-approved drug for that. There are also other mutations in some of the other cases that are at similar, at low levels, to that. The advantage to doing

testing is that it&rsquo;s cheaper, with immunohistochemistry, to just run that test if you&rsquo;re concerned about that, as opposed to doing a large panel. However, these large panels will pick up

Case 2: Rare Alterations and Germline vs Somatic Testing in CRC

Richard, you talked about resection, but will this retrospective data change how you treat metastatic cancer? Would you resect the primary tumor on the right side based on this data?

Well, that&rsquo;s retrospective data. It&rsquo;s very hypothesis generating. We all know that as standard practice, for a patient who is symptomatic, is bleeding, or has near obstruction, we would take out the primary tumors for a clinical benefit. I think that&rsquo;s clear. But in patients who are asymptomatic, whether you would take out the primary tumor or not, there has been conflicting datamostly retrospective. The standard practice, at least at our institution, in patients who are asymptomatic, is not to take out the primary tumor.

Now, this data is very intriguing. The right-sided tumor patients tend to do poorly anyway, so taking out their primary tumor may make a difference. But based on that, I don&rsquo;t think we can change the standard of care. As you know, there is a trial going on in Europe right now. This is a randomized trial of chemotherapy plus or minus surgical resection upfront. That randomized study will tell us, once and for all, what the right answer is. Based on the retrospective data I just presented, I will probably not take the patient to surgery upfront.

Now we come to the big question: sidedness. In this day and age, with fancy next-generation sequencing and everything else, does sidedness make a difference? Mike, if somebody comes in with a left-side tumor, what do you talk to them about? Let&rsquo;s say they don&rsquo;t have any mutations. They&rsquo;re

-negative. They are microsatellite stable. There is no

mutation. What are your current recommendations?

First, we want to make sure that we&rsquo;re all on the same page, in terms of sidedness, in what we&rsquo;re really talking about. The data are strongest when you talk about the left side splenic flexure, all the way through the sigmoid colon. And for the right-sided cases, most of the data [are] for the ascending colon. The data for the transverse colon are often lumped in with the right-sided data, based more on embryologic considerations. Obviously we&rsquo;d like to know the biologic explanation for the sidedness, and obviously there are molecular differences in the tumors on the left and right sides. We will hopefully be able to use that information to make decisions in the future. Right now, what we have is just for clinical observation. The prognosis for the tumors on the left side is better than those on the right side. At least for right-sided tumors, there&rsquo;s a big differential in response with anti-EGFR therapy.

In terms of counseling somebody regarding therapy, I think it&rsquo;s easy for the right-sided tumors because there is such a clear difference. The NCCN [National Comprehensive Cancer Network] guidelines recognize that if you&rsquo;re going to use anti-EGFR therapy, at least in the first line, it should be used only in people who have left-sided primary tumors. Whether or not you would choose an anti-EGFR therapy or bevacizumab for a left-sided tumor is still an open question, and again, it&rsquo;s reflected in the NCCN guidelines. They don&rsquo;t specify one or the other. I think you can see that there is a small difference between anti-EGFR therapy and bevacizumab in left-sided tumors, but it is a very small difference. As Richard pointed out, there are concerns about tolerability of the different regimens. You really have to discuss this with a person before making a blanket decision about what you&rsquo;re going to treat them with.

That&rsquo;s a great point. So the first point is tolerability, right? What can it do? Are there other comorbidities that you would take into consideration for treating a patient? Let&rsquo;s say they were like this patienta healthy, fit, 35-year-old with no mutations. Richard, what would you do if they came to you with a right-sided tumor?

I think it is pretty straightforward. In young patients who have a right-sided tumor that is

wild type, the data clearly show that the anti-EGFR drug is not as effective as anti-VEGF therapy. So for a right-sided tumor, I would probably use either FOLFOX [folinic acid, fluorouracil (5-FU), and oxaliplatin] or FOLFIRI [folinic acid, fluorouracil, and irinotecan] plus bevacizumab. That would be my choice for someone with a right-sided tumor.

For patients with left-sided tumors, once again, depending on which trial you look at, if you look at the FIRE-3 study, the benefit of the EGFR drug is about 8 months&rsquo; difference. In the CALGB 80405 study, the benefit is a little less. If you look at the ESMO [European Society for Medical Oncology] guidelinesit&rsquo;s a different country, obviously&mdash;for left-sided

wild-type disease, the recommendation is to use the EGFR drug. They don&rsquo;t even put the VEGF drug on there.

The NCCN guidelines sort of give you a choice. And so it&rsquo;s a discussion to have with the patient. Personally, putting somebody on EGFR therapy for a long time is very tough on the patient. You see the rash, the paronychia that occurs. That&rsquo;s probably the only reason I would stay away from it. But I would have a discussion with the patient. I would talk about the possible survival benefit and the potential toxicity that he or she may get from the EGFR drug. So I would say in the first-line of therapy for a left-sided tumor, it&rsquo;s fifty-fifty in terms of what I would do. It all depends on the discussion with the patient.

So what you are saying is that the Americans differ from the Europeans? The Americans don&rsquo;t think you should use it for right-sided tumors, right? And the Europeans suggest to definitely use the anti-EGFR therapy for left-sided tumors, just to be clear. It&rsquo;s actually very interesting. They&rsquo;re not exactly the same, but they&rsquo;re similar, in a way, in what they&rsquo;re recommending.

In the first line, it&rsquo;s sort of dry, in terms of not using anti-EGFR therapy for a right-sided tumor. The question is, what about in the second- and third-line settings? If you have

wild-type disease in a right-sided tumor and you give FOLFIRI or FOLFOX plus bevacizumab and you progress on it, we understand that. But then what do you do?

Well, I&rsquo;d like to see more data. There are some data for later-line therapy. Even there, I think they question whether there is activity, which means there is a real biologic explanation for this. As more data come out, it may be the case that we find that right-sided tumors do not respond to anti-EGFR therapy. However, currently, because there are inadequate data to answer that and because we have only so many drugs to treat people with, we will inevitably offer them anti-EGFR therapy, even for a right-sided tumor.

I agree with that. There are some data in the second line for anti-EGFR therapy. It might not work for right-sided tumors. In the third line, there may be a slight overall survival benefit. It&rsquo;s unclear. Therefore, I would not use it. I would use this as my last resource. If out of options, if this is a right-sided tumor and I have nothing to offer, and if there are no trials, that is when I would use it. But as you said, I would be very hesitant to use it in the first- or second-line setting. So that would be the last resource of treatment.

Case 2: Sidedness in mCRC

Coming back to maintenance therapy, Richard, the case that you discussed was on maintenance therapy but with the biologic. First there was OPTIMOX1 and OPTIMOX2, which looked at maintenance therapy. Mike, what are your thoughts on maintenance therapy? Do you use the biologic or not? What do you use as maintenance therapy in your patients?

Clearly, the chemotherapy component, as we&rsquo;ve seen time and time again, is an important part of maintenance therapy if you&rsquo;re going to use it. As you pointed out, the OPTIMOX trials tell us that there is utility to continuing a chemotherapy drug. Oxaliplatin is obviously impractical to give long term, so it&rsquo;s natural that we&rsquo;re going to stop at 6 or 8 doses.

As far as adding on the biologic, we have data with bevacizumab plus chemotherapy that consistently show that there&rsquo;s a progression-free survival benefit. In 1 study, there was an overall survival benefit. This was small, but there was a benefit to adding bevacizumab to chemotherapy. The reverse is not so clear, in using just the biologic. I think we see the same thing here in the VALENTINO trial, which is really not so much a study of the benefityes or no&mdash;of maintenance anti-EGFR therapy. It&rsquo;s really, again, the question of, is a chemotherapy component important here? I think it proves, once again, that it is.

In general, the biologic adds only a small amount of additional toxicity. My opinion is that a biologic with chemotherapy is a bit better, in terms of progression-free survival. I would continue to use the biologic if we&rsquo;re going to continue the chemotherapy.

Great point. What about you, Richard? A biologic with VEGF has its own side effects, but a biologic with anti-EGFR as maintenance has its own side effects, as you said. Patients can get rash and fatigue, and there can be other issues. How do you tackle this problem with patients?

I agree with Mike. The concept of maintenance therapy came about because of oxaliplatin. We cannot do oxaliplatin forever, so we need to put up some maintenance therapy. There is a lack of data on the use of bevacizumab alone. Bevacizumab alone should not be used when there are data suggesting that you could possibly use the combination, or the 5-FU [fluorouracil] plus bevacizumab. There is clear data, as you said, of some progression-free survival benefit and some overall survival benefit in that setting. There are data that demonstrate that you can probably use 5-FU plus bevacizumab in the setting as maintenance therapy.

However, EGFR maintenance therapy is a question. I just presented VALENTINO data, which really do not answer the question of maintenance therapy. Rather, they just explore whether 5-FU plus panitumumab is better than panitumumab alone in maintenance therapy. The trial, to me, was a negative study. However, that sort of contradicts the MACRO2 study, which was a trial using cetuximab as maintenance therapy. You may know that the trial looked at FOLFOX [folinic acid, fluorouracil, and oxaliplatin] with cetuximab. I don&rsquo;t remember exactly how many cycles, but after a certain number of cycles they were randomized to the continuation of FOLFOX/cetuximab or cetuximab alone. It was a noninferiority study, and it met its endpoint. Cetuximab was noninferior to FOLFOX with cetuximab.

So we have conflicting data here. With panitumumab, you have to give 5-FU. You could probably give cetuximab alone. I think that&rsquo;s the problem in using cetuximab as a first-line therapy. I don&rsquo;t know what to do for maintenance therapy. So it&rsquo;s still a controversial area. If you&rsquo;re a purist, there are data that suggest that you could use cetuximab alone as maintenance therapy. However, if you use panitumumab in the first-line setting, from the VALENTINO study that I showed you, you&rsquo;d probably have to use 5-FU plus panitumumab because that seems to be better than panitumumab alone.

But the chemotherapy is probably important. I think most of us, in clinical practice, do continue with a chemotherapy backbone and think about continuing the biologic on top of that.

This has been a great discussion. I just want to highlight the take-home points. For asymptomatic patients, the current data demonstrate that there&rsquo;s really not much benefit in doing upfront resection in these patients. Obviously, it&rsquo;s based on the clinical outcome of the patient and what their clinical situation is. In regard to tumor sidedness, the treatment options for left-sided tumors is anti-EGFR therapy or bevacizumab. On the right side, it is suggested not to use anti-EGFR therapy. I think that is very clear. So the right side is more clear than the left side. They&rsquo;re still trying to figure out the left side. However, if you have a left-sided tumor, you&rsquo;re going to do much better than if you have a right-sided tumor.

What to do in the second line is still unclear. We don&rsquo;t have enough data. You could continue with an anti-EGFR therapy in the second-line or, possibly, third-line setting in these patients.

Looking at maintenance therapy, probably through the OPTIMOX data, you should continue with a chemotherapy backbone. Potentially, you could add on and continue with the biologic. You could also use single-agent cetuximab or the combination of 5-FU/panitumumab. That&rsquo;s a lot to digest, but I think this was a great discussion. Thank you.

Case 2: Maintenance Therapy in mCRC

For the third case, I&rsquo;d like to present a woman with metastatic colorectal cancer that was

-mutated. In March of 2016, she presented, otherwise healthy, on a referral to a gastroenterologist for a colonoscopy, which showed a 2-cm mass in the sigmoid colon that was invading the muscularis propria into the adjacent pericolorectal tissue. A biopsy confirmed it was poorly differentiated and was felt to be stage III. Molecular testing showed it to be positive for a

exon 2 codon 12 mutation, but there were no mutations in

, or PIK3CA, and this was microsatellite stable. Shortly thereafter, she underwent a resection, with complete resection of disease. Following that, she received 6 months of adjuvant CAPEOX [capecitabine plus oxaliplatin].

In May of 2017, at her 12-month follow-up, it was noted that her CEA [carcinoembryonic antigen] level was 146. A CT scan of the abdomen, pelvis, and chest showed multiple small bilateral lung nodules, less than 5 mm. Her performance status [PS] remained at 0. She was then initiated with FOLFOX [folinic acid, fluorouracil, and oxaliplatin] plus bevacizumab, which she was tolerating well, and she subsequently continued maintenance bevacizumab.

Six months further, her CEA level was now normalized to 2.2 and a CT scan showed a reduction in the lung lesions. Subsequently, she began complaining of a dry cough and shortness of breath on exertion. Her CEA had increased to 25.3. A CT scan of the chest, abdomen, and pelvis showed that there was progression of 2 lung lesions in the right pulmonary lobe. FOLFIRI [folinic acid, fluorouracil, and irinotecan] was initiated while bevacizumab was continued.

Unfortunately, the CEA continued to rise and was now 99. A CT scan of the chest, abdomen, and pelvis showed progression of lung disease, a new lesion in the liver that was in close proximity to the bile duct, and diffuse peritoneal disease. Her performance status had now declined to a PS of 1. Nonetheless, she continued to work but sometimes had to leave early because of fatigue. She&rsquo;d now like to know her options for additional therapy. She would like to avoid infusion-based therapy and its associated toxicities.

Let&rsquo;s first look at some data around next-generation sequencing. This study asked whether it was actually beneficial or just an economic burden in the community oncology practice setting. This was a retrospective chart analysis of cancer patients treated at a large community practice. The data were collected on 209 patients who had next-generation sequencing in 2015 and 2016. Thirty-one of these patients had colorectal cancer. The question was whether there was a change in management based on the NGS testing. The results demonstrated that there was, in fact, not a change in management. In some cases, there was no change in management because patients had a poor performance status and, therefore, were not going to be candidates for additional therapies. But in the large majority of cases, there was a lack of a target or a drug that could be used to treat those patients.

The next study is a therapeutic one. This is the regorafenib pivotal trial, the CORRECT trial, which was a multicenter, randomized, placebo-controlled phase III clinical trial that compared regorafenib plus best supportive care with placebo in patients who had progressed on a standard chemotherapy such as fluorouracil, oxaliplatin, irinotecan, anti-VEGF therapy, and, if appropriate, anti-EGFR therapy.

The results showed that there was a median overall survival of 6.4 months for regorafenib versus 5 months for placebo, which was statistically significant. Of course, adverse events were seen in a large number of patients in both arms of the study. The most common grade 3 adverse events were handfoot skin reaction, fatigue, diarrhea, hypertension, and rash. Because of these toxicities, the question has been asked: Is there an alternate strategy for initiating regorafenib that might be more tolerable for patients and allow them to continue on the drug longer? So this study, which was performed in the ACCRU Network, was a randomized phase II trial that looked at 2 different dosing strategies for regorafenib in patients with refractory metastatic colorectal cancer. It was called the ReDOS trial. In this study, patients were randomized to start regorafenib at 80 mg a day and increase weekly to eventually reach the full dose, 160 mg, if they had no toxicities, or to start at the standard dose of 160 mg. Both strategies gave patients the standard 3-weeks-on, 1-week-off regimen of regorafenib.

There was a further randomization that looked at whether preemptive clobetasol or reactive clobetasol would also alter the handfoot skin reaction. In this study, it was called the palmar-plantar erythrodysesthesia syndrome. However, the primary endpoint for the study was the proportion of patients who had completed 2 cycles of regorafenib and were able to initiate the third cycle.

Interestingly, and this was not a primary endpoint of the trial, there was a trend toward better overall survival for the patients who initiated the regorafenib at 80 mg and escalated to tolerance compared to those patients who started at the full dose. The interpretation of this was that starting at a lower dose and escalating was certainly no worse for patients; it might even be better. Regarding toxicities, the severest of toxicities were less common in patients who had the full dose.

The take-home message of this study was that dose escalation of the regorafenib was certainly not inferior to using a full dose to start, and more patients were able to get to the third cycle of regorafenib if they used the dose-escalation strategy. From a tolerability standpoint, it was well tolerated to use the dose escalation strategy.

I&rsquo;d also like to address the other oral therapy that is available for patients who have progressed through the standard fluoropyrimidine, irinotecan, oxaliplatin, anti-VEGF therapy, and, if appropriate, anti-VEGF therapyTAS-102 [trifluridine/tipiracil].

In the RECOURSE trial, a randomized phase III clinical trial, patients with advanced colorectal cancer with a good performance status of 0 or 1 who had received at least 2 prior regimens of standard chemotherapy and were refractory to or were failing all the of the following therapiesfluoropyrimidine, irinotecan, oxaliplatin, anti-VEGF therapy, and anti-EGFR therapy&mdash;if they were

wild-type, they were randomized to TAS-102, which is a combination of trifluridine/tipiracil, plus best supportive care versus placebo plus best supportive care. The primary endpoint was overall survival and secondary endpoints were progression-free survival, safety, and tolerability.

The results of the study showed that the overall survival favored TAS-102, with a 7.2-month median overall survival compared to a 5.2-month median survival for placebo, with a hazard ratio of 0.69. The toxicities of this drug are primarily related to bone marrow and neutropenia. Thirty-eight percent of patients had grade 3/4 neutropenia with TAS-102 versus, of course, none with best supportive care. Also, 18% of patients had grade 3/4 anemia and 5% had grade 3/4 thrombocytopenia with TAS-102.

Case 3: KRAS-Mutated Metastatic Colon Cancer

In summary, this is a case of a patient with

-mutated metastatic colon cancer who had progressed on first-line FOLFOX [folinic acid, fluorouracil, and oxaliplatin] plus bevacizumab [Avastin], then proceeded to second-line FOLFIRI [folinic acid, fluorouracil, and irinotecan] plus bevacizumab, and is now in the third-line setting and has the options of 2 different oral agents.

Thank you for the case discussion, Mike. Greg, we saw this interesting thing about NGS testing. When do you normally get the NGS testing for colorectal cancer [CRC] from your oncologist?

It really depends. We&rsquo;ve been doing a lot of Foundation Medicine, the commercial NGS testing, for a while. I think the oncologists in our area are fairly comfortable with that test. So we do a lot, typically when the patients either aren&rsquo;t responding, have refractory disease, or maybe don&rsquo;t have MSI [microsatellite instability]high disease or some other potentially targetable mutation through your normal frontline types of therapies.

What&rsquo;s the difference between IHC [immunohistochemistry], PCR [polymerase chain reaction], and NGS testing, or expanded NGS testing?

Specifically, in colon cancer, usually when they&rsquo;re talking about immunohistochemistry, they&rsquo;re doing immunohistochemistry specifically for the mismatch repair [MMR] proteins. It&rsquo;s just an antibody-based test that looks for the protein expression of MLH1, MSH2, MSH6, and PMS2, which are typically the 4 most frequently mismatch repairaltered genes. And then, a PCR is just polymerase chain reaction. Traditionally, that had looked more at single genespotentially what people were doing for

And then, there is something which people usually refer to as next-generation sequencing. Some people don&rsquo;t like that. They prefer the term &ldquo;massively parallel sequencing.&rdquo; It&rsquo;s essentially millions, or even billions, of PCR reactions that are occurring simultaneously on a chip. It allows you to interrogate many gene regions and, potentially, multiple patient samples all in the same reaction.

There has been a lot of discussion in the gastrointestinal world about liquid biopsy in the ctDNA [circulating tumor DNA]. What are your thoughts about testing from the tumor versus ctDNA?

MD Anderson Cancer Center has kind of been the pioneer in nonsmall cell lung cancer. Traditionally, I think people are most comfortable, at least up front, with doing some type of tissue-based test to identify the mutation. Obviously like in non&ndash;small cell lung cancer, where there are certain resistant mutations and it can be a bit more difficult to get a second biopsy, I think we&rsquo;ve seen a lot more of the plasma circulating tumor DNA. You&rsquo;re looking for those specific mutations. We&rsquo;ve seen certain instances for which a tissue biopsy isn&rsquo;t feasible. In those cases, liquid biopsy has been used as an alternative.

Rich, I&rsquo;ll come to you. This is a great clinical question. We talked about this before. A new patient comes in. You have a biopsy. Are you sending it for all NGS at that time, just an NGS panel, or do you test for what we talked about before:

In my practice, we may have a lack of tissue, right? The tissue is always an issue. When you see the patient, you have tissue. If you send a test for

, MMR, you&rsquo;re using all the tissue up. Let&rsquo;s say they progress after first-line therapy. You&rsquo;re out of tissue. Then you have to do another biopsy or track some tissue down from outside of the hospital.

In our practice, we like to get everything at once. We have an internal panel of 200 assay mutations. You could send to Foundation Medicine, where they have 400-assay panels as well. I like to do everything up front for the reasons that we talked about: to test for

. It could also test for MSI status as well. And there are other subtle-like fusions out there that don&rsquo;t commonly occur in colon cancer, such as

, which we see in about 1% or less of patients. The nice thing, even though it&rsquo;s very rare, is that they are targetable, actionable fusions. There are drugs out there, on a trial basis, that you can get off label. Some are approved for

fusions. The response rate can be as high as 75%. They are durable. This could be life saving for certain patients.

Therefore, in my practice, I like to get everything up front. I have a battle plan, whether you have a

fusion, there is a therapy to possibly use in the second-line setting. If you have a

mutation, you don&rsquo;t want to do that. So, in my practice I like to do everything up front, if possible.

Rare is rare, right? In the community practice, when they&rsquo;re doing it, these tests are always expensive versus doing an IHC and testing for a couple of genes. Let&rsquo;s say the performance status goes down in these patients and you cannot use this testing. What would you recommend? What is your practice? Would you do everything up front or would you say, &ldquo;Hey, I&rsquo;m going to test for these 4? And then, after my first or second line of therapy, if the performance status is good and they look good clinically, that&rsquo;s the time I would come in and maybe do an NGS. I don&rsquo;t test in the beginning, and the patients don&rsquo;t get it because their disease is uncontrollable.&rdquo;

The theme, today, has been that this is personalized. Not every patient should get complete NGS testing up front. It&rsquo;s not going to be appropriate for a lot of people. However, there are some subtleties. Physicians should know what&rsquo;s actually being tested. A good example is with

. I always assumed that when you sent that test off we were getting exon 2, 3, and 4. In reality, we weren&rsquo;t. We were getting exon 2, predominantly. And so, for us to get all of the possible mutations, it actually involves getting next-generation sequencing. For that reason alone, we find it valuable to do it up front.

It also gives you a baseline. Most of these mutations are not going to go away. They&rsquo;ll presumably be there later on. It then allows you to use a liquid biopsy, as you pointed out, subsequently. If somebody is being treated with, for example, anti-EGFR therapy, we have sometimes seen mutations develop that render the therapy no longer effective. The only way you can realistically follow for that is with a liquid biopsy. And so, I think knowing what the mutational status is up front does allow you to plan for later. If you find there is

amplification, you know that&rsquo;s going to be an option.

The mantra in gastrointestinal oncology has been this continuum of care and being able to sort of anticipate what therapies you&rsquo;re going to have available for the person over time. Having all of the information available up front allows you to really do the continuum of care in the best possible way.

Case 3: NGS Testing in Patients With CRC

fusions, and there&rsquo;s a new kid on the block. Are you testing for all of the

fusions? We&rsquo;ve seen very high response rates in the

We&rsquo;re generally not doing that up front. Up front, we&rsquo;re just doing the

. And then, if the patient becomes refractory, we&rsquo;re sending to Foundation Medicine. What has generally been reported in colon cancer is NTRK1 and NTRK3. And then, obviously, as we kind of talked about with

, it&rsquo;s smaller percentages. One take-home message for places that aren&rsquo;t routinely sending all of these up front is, it seems like these rearrangements tend to be found more often in older patients with microsatellite instabilityhigh tumors. They also tend to be

-negative. That might be 1 subgroup to send or to think more about.

That&rsquo;s an excellent point. Richard, I&rsquo;ll come back to you. For our patient, let&rsquo;s say there are no targetable mutations. You do NGS [next-generation sequencing] testing. You may use FOLFOX [folinic acid, fluorouracil, and oxaliplatin], FOLFIRI [folinic acid, fluorouracil, and irinotecan], bevacizumab, or cetuximab. Then you come to the third-line setting. What do you normally choose as your third-line agent? How do you pick among the agents?

In the third-line setting, when you fail standard chemotherapy, there are 2 oral oncolytics available. One is TAS-102, and the other is regorafenib. There&rsquo;s no head-to-head comparison so we cannot say one is better than the other. They both showed an overall survival benefit compared to placebo. The mechanism is different. Regorafenib is a TKI [tyrosine kinase inhibitor] that blocks multiple pathways. TAS-102 is a combination of trifluridine [Viroptic] and tipiracil. The trifluridine actually gets incorporated into DNA, preventing DNA synthesis. Tipiracil is a thymidine phosphorylase inhibitor. So the mechanism is totally different.

Management of Metastatic Colon Cancer

Case 1: BRAF-Mutated Metastatic Colorectal Cancer

Case 1: Genomic Testing & NCCN Guidelines for Metastatic CRC

Case 1: Treatment of BRAF-Mutated Colorectal Cancer

Case 2: Metastatic Colon Cancer, No Molecular Alterations

Case 2: Rare Alterations and Germline vs Somatic Testing in CRC

Case 2: Sidedness in mCRC

Case 2: Maintenance Therapy in mCRC

Case 3: KRAS-Mutated Metastatic Colon Cancer

Case 3: NGS Testing in Patients With CRC

Case 3: Third-Line Therapy for Metastatic CRC

Case 3: TAS-102 Dose Adjustment in mCRC

Case 4: Metastatic Colon Cancer, MSI-High

Case 4: MSI Testing and Tumor Mutation Burden in mCRC

Case 4: Standard of Care and Second-Line Therapy for mCRC