 Future drug development of active agents in the later-line setting will focus on 2 things. Number one, identifying patients molecularly to find subgroups for a targeted intervention. I’m looking at low-frequency mutations and alterations, like 

 mutations, which do exist in colorectal cancer but at lower frequency, for instance, in than pancreatic cancer, prostate cancer, and breast cancer. I’m looking at other alterations like fusions, 

 alterations, etc. Everything we know from other tumors should be tested in colorectal cancer. Even if we just identify 2% to 5% of patients with these selected mutations that are targetable, it helps these 2% to 5% of patients.

The other point is, of course, we want the immunotherapy to work because immunotherapy is a very powerful tool that can meet sustained response, really affect the clinical course of patients dramatically. I strongly believe that with further insights and escape mechanisms, immunosuppressive effects of cancers, we will find combination therapy. It is clearly not a single agent. It is combination therapies that will augment immune responses, will make immunotherapy work in MSS [microsatellite stable] colon cancer. It is not one agent that I would single out right now, so that’s the road forward. There are dozens of trials ongoing right now, and this shows me, with various combinations, that no one knows exactly the right answer because preclinical models are not perfect in predicting what’s happening in a host like a human body in terms of immunotherapy. But I have not given up hope.

 I’m very excited about the future because I think you have just started to take advantage of the newest molecular technology profiling platforms, such as NGS [next-generation sequencing]. We will identify more unique targets and may potentially fragment out more colon cancer, which we have started with the 

. That will expand, and 1 of the major new strategies is that we are so focused on a single mutation—

—that we have not even learned to understand what it means when you have a 

 mutation. Does that change the efficacy of certain target drugs? I think we are just in the beginning of understanding the complexity of genetic signatures because we have been stuck to 1 mutation or the other and have not moved on to that. Machine learning, artificial intelligence will help us understand it. I’m very excited about that.

There are a lot of new drugs coming, and the challenge is how can we select the patients? These molecular profiling technologies developing are very helpful. I’m also very excited about the use of liquid biopsies to monitor real-time molecular changes, identifying novel targets and potentially use better treatments, taking advantage of the activated mechanism resistant pathway you see in the liquid biopsies. I’m very optimistic. I’m sure we did push further the limit and increase overall survival for our patients. 

We will see in the next couple of years improvements in the 

-mutant population. We now have SOS inhibitors. We have certain vaccines and Polo kinase inhibitors, which seemed to be promising in a subset of 

 mutant. I’m very hopeful that this will move forward and we will finally have a more effective treatment for 

 mutant. With the deep understanding of immune signaling, we will get better immunotherapy combinations, and we have now more sophisticated vaccines and the entrance of oncolytic viruses. I’m very optimistic that we will also make significant progress for the MSS patients, who at the moment are very limited to use clinical trials with immunotherapy combinations.

 Clearly, we have done a very important series of steps forward that better derive prognosis from a management of metastatic colorectal cancer. Twenty years ago we had very few drugs and the expected median survival was between 6 and 12 months for most patients. Now we have extended 3 or 4 times the expected survival time for most patients. What we need to do is understand more the tumor heterogeneity of metastatic colorectal cancer. We need to apply more chemotherapy, chemotherapy combinations, sequencing of therapy in different lines of treatment, and rechallenge with specific drugs and biologic agents against specific molecular targets and immunotherapy in a more selective way. Basically, the big challenge in the treatment of metastatic colorectal cancer is to apply the enormous knowledge that we are accruing on the genetics and epigenetics of this cancer at the individual patient level, so that it can be followed during evolution of disease by liquid biopsy testing. All this information is a big challenge because this should be applied to selecting specific therapies on an individual basis. The big challenge is really to apply precision medicine in the treatment of metastatic colorectal cancer.

Videos

Future Directions in Later Line Therapy for mCRC

IMblaze370 gives us a lot of lessons to be learned because everyone was excited about the design of the study, based on pretty limited initial data, a phase 1b study, which looked at the combination of cobimetinib-atezolizumab, a MEK inhibitor plus a PD-L1 inhibitor in the later-line setting in about 20 patients with MSS [microsatellite stability]. Most of them had MSS tumors, and we saw 4 responses: 3 were really MSS tumors, and 1 we didn’t know what the MSI [microsatellite instability] status was. This created a lot of excitement, initially. Everyone wanted to make immunotherapy work in these otherwise immunotherapy-refractory colorectal cancers, and there was some preclinical rationale. There was the idea of recruiting T cells into the tumor through the MEK inhibition, making these T cells be more sustainable, have durable activity, etc. They have a lot of data that led to the idea of running a phase 3 study. Well, the phase 3 study eventually accrued very rapidly. It was an international effort to randomize patients to atezolizumab-cobimetinib versus atezolizumab alone versus regorafenib as the control arm, which was used because it was considered an easy arm to beat, based on toxicity profile and limited activity. The study eventually came out negative. There was no difference in outcome between the combination of atezolizumab-cobimetinib or atezolizumab alone compared with regorafenib. In fact, regorafenib outperformed expectations and now is being cited for being associated with overall survival of about 8.5 months in the later-line, phase 3 setting. That’s encouraging to know we have an active agent, even in modern times, that held its ground in a phase 3 setting.

The lessons learned are several. First, we move very rapidly from limited data in phase 1b to expanding into a phase 3 study, and we need to be very careful with future endeavors to have better signals. This is also important right now for the excitement people have based on some limited data on regorafenib plus nivolumab. Are the data from Japan in gastric cancer and colorectal cancer really enough to warrant a phase 3 study? Particularly when, now, single-arm studies from other institutions do not seem to be able to completely replicate these data. It’s a very important point to be careful in the development to see what patients could benefit from this combination.

This other part is that cobimetinib-atezolizumab not only did not beat regorafenib but is also more toxic. A lot more patients, 2 times more patients, went off therapy with cobimetinib-atezolizumab, compared with regorafenib, because of adverse effects. So it’s not just that this drug didn’t work. We might have also potentially harmed some patients, at least with adverse effects. All this shows how important it is to have good data that allow us to move combination therapy, immunotherapy combinations, into a later-line setting, into phase 3. People are more careful now. On the other hand, everyone recognizes that if we can make MSS colorectal cancer immunogenic, that would be a home run. It’s the holy grail of immunotherapy because it’s a very common and very immunotherapy-refractory cancer so we all want to make sure we find, eventually—and I’m sure we will—find ways to treat MSS cancers with immunotherapy.

There are several possibilities in metastatic colorectal cancer to try to enhance the immune response in the vast majority of metastatic colorectal cancer patients whose tumors are microsatellite stable. These tumors we know could be either immune desert or immune inflamed. Both conditions are not the best conditions for the conventional I/O like anti–PD-1 or anti–PD-L1 antibodies to work. In the past, we had some hope that inhibiting MEK pathway will enhance in these patients the activity of PD-L1 antibodies. This strategy unfortunately failed. What has been now tried to be done is a dual approach. One is in RAS/RAF wild-type patients, trying to combine anti-

 monoclonal antibodies and most likely cetuximab, also because it is an ADCC [antibody-dependent cell-mediated cytotoxicity] potential immunoglobulin inducer with anti–PD-1 or anti–PD-L1 antibodies.

There are some studies that are ongoing also in colorectal cancer. Some are ongoing with these combinations in head and neck squamous cell carcinoma, and some have been done also in non–small cell lung cancer. Another opportunity has been to try to combine immuno-oncology drugs, such as PD-1 or PD-L1 inhibitors, with antiangiogenic drugs. For example, we know that the combination of small molecules that block angiogenesis—such as the ones that are used in first-line renal cancer, combined with anti–PD-1 or anti–PD-L1 antibodies, such as pembrolizumab or avelumab—has activity, has become the standard therapy for these patients. We know, for example, that combining bevacizumab and atezolizumab is another PD-L1 antibody that is becoming the first-line potential preferred therapy for hepatocellular carcinoma. There is a strong rationale combining either anti-

 monoclonal antibodies or tyrosine kinase antiangiogenic small molecules with PD-1 or PD-L1 antibodies in metastatic colorectal cancer. With a report last year of interesting data in MSS patients with refractory disease of combining regorafenib and nivolumab. At the current annual congress at ASCO some data have been presented in an abstract and then in a presentation, like the combination of regorafenib with a PD-L1 antibody, avelumab. The rationale behind these combinations is that there’s a broad spectrum of action of regorafenib. It blocks several tyrosine kinases that are important for endothelial cells and several cells that are involved within the microenvironment of the tumor, who in some way interact with the T-effector cells. This inhibition could increase the activity of PD-1 or PD-L1 antibodies. I hope we will have more data in the near future because, unfortunately, the bottom line—especially after the KEYNOTE-177 randomized phase 3 trial—is that we know that for MSI-high patients, pembrolizumab, a PD-1 antibody, is very effective. However, these patients are only 5% of metastatic colorectal cancer patients. What can we offer them in terms of immunotherapy to the other 95%? This is a very important topic, in which the combination with other molecular-targeted agents—such as EGFR inhibitors in select patient populations with RAS/RAF wild type or with broad spectrum antiangiogenic drugs. In terms of regorafenib, it is very interesting and is worth to be addressed in new studies.

Key Trials in mCRC: I/O Combination Therapies

I have no doubt in MSI high immunotherapy is the most successful treatment. I think the question is not immunotherapy versus chemotherapy. The question is what immunotherapy do we choose? Right now we have 2 options: pembrolizumab or nivolumab-ipilimumab combination. I would never give nivolumab alone because of all the data. There is no side-by-side comparison, but the combination seems to be consistently better. When you choose ipilimumab at the right dosage, with 1 mg every 6 weeks, the toxicity is not different from nivolumab alone, so I’m very comfortable with a nivolumab-ipilimumab combination. The problem is that we don’t have the statistical powers, since comparisons are not done, to tell if the combination is better than pembrolizumab. Is it the same? Just looking at the data presented at ASCO this year, 30% progression of disease is the best response, with the first CT scan. It is a little worrisome because we don’t see that with nivolumab-ipilimumab, also recognizing that this is a much smaller study. But the 30% progression was seen in other later-lines of treatment with KEYNOTE-164. I think this gives us a lot of ideas to better understand what the best immunotherapy combination is for this patient combination.

Turning a cold tumor into a hot tumor is the holy grail of developing immunotherapy right now, and this goes beyond the question of MSI high versus MSS. There are some data that we’ve seen over time across the landscape of solid tumors that sometimes combinations of checkpoint inhibitors, PD-1 and PD-L1 inhibitors, and chemotherapy can turn an otherwise cold tumor into a hotter tumor, likely by releasing neoantigens and tumor antigens in the environment. A combination of immunotherapy plus chemotherapy in PD-L1–low tumor it seems to be more effective than just immunotherapy alone.

Then we’ve learned that combination of immunotherapy, like checkpoint inhibitor combination, PD-1 antibodies plus CTL4 antibodies, ipilimumab-nivolumab, can be interesting and can potentially turn some of the colder tumors into hotter tumors. This is not the 1-size-fits-all treatment for all cancers. I do believe, from all the data we’ve seen, we need an additional modulation. It could be that different patients, even in 1 cancer type, need different approaches to activate the immune system. The idea right now is to really generate an immune profile of tumor before treatment, so you know what angle we need to use to turn these tumors from cold into hotter tumors, recognizable by the immune system. This could be very individualized treatment approach.

The way to move forward with immunotherapy in microsatellite stable tumors is speedy combination treatments. This could be with other immune system–targeting drugs, like OX40 agonists and many different ones. There is approved synergism between the anti-VEGF strategies and immune checkpoint. The data in HCC with bevacizumab and atezolizumab are amazing. It is approved in first-line, where is has the highest response rate. The anti-VEGF strategies with immune checkpoint makes a lot of sense. The reason is that anti-VEGF really modulates the polarization of the macrophages, which plays a significant role in the immune response pathway. VEGF is a part of an immune response through the macrophage’s polarization from M1 to M2. Here, I think there is a potential understanding that this combination may be very helpful. The most data we have are on regorafenib-nivolumab from Japan that showed a 36% response rate in this patient population. This is unheard of in refractory patients. What is even more exciting is when you look at the gastric cancer cohort, there is a 44% response rate. Of these patients, there are patients who had previously been treated with immune checkpoint inhibitors who responded to this combination. I think you can overcome resistance to immune checkpoint with regorafenib-nivolumab combinations, which gives us some clue that this may be a promising combination for patients who are usually not candidates for immunotherapies. We still need to learn who are the patients who benefit from this combination, because 1 of the discussion points coming out of regorafenib-nivolumab from Japan is that the patient selection in Japan may be a little different. Patients there have less tumor burden, maybe have fewer tumors involved in the liver and more in the lung. These are the patients who may benefit more. We are going to learn much more. There are many clinical trials going on with TKIs [tyrosine kinase inhibitors] of the VEGF pathway and immune checkpoints. Within the next year, I’m sure we’ll be learning much more the efficacy and hopefully also about which patients will benefit the most.

Role of I/O Combination Therapies in mCRC

 In terms of immunology, it’s very important to discuss what we consider a microsatellite unstable tumor. As you know, we can define these in 2 different kinds of assays. One is related to immunohistochemistry, which looks at lack of proteins that are produced when genes are mutant or altered, and so a tumor that is MSI high lacks some specific proteins that are specific DNA-repair proteins. The other is by genetic analysis, and this can be done most likely using next-generation sequencing [NGS] panels. We can measure gene mutations in these genes that account for the genetic alterations. Both assays have some pros and some cons; both assays would hence work.

My feeling is that the more we are using extended gene mutation, rearrangement, deletion panels by next-generation sequencing, the more this will be used, especially because the extended knowledge about next-generation sequencing approaches through liquid biopsy. Even in plasma we can get this information. We don’t need to go back to the primary tumor to extract DNA from the tissue and worry about all the caveats that can some way hamper the real readout and the real results. In my opinion, immunohistochemistry could play an important role, especially in the expert hands of hepatologists. More and more, we will see data coming out by NGS analysis and more and more by liquid biopsy use of NGS analysis.

 At this point in time, in 2020, every single patient with colorectal cancer, independent of stage and age, needs to be tested for the MSI status. This could be done by immunohistochemistry for the mismatch repair proteins, because mismatch repair deficiency—meaning the absence of 1 protein expression and 4 proteins for testing—is really kind of synonymous with MSI high and microsatellite instability. Microsatellite instability was routinely tested with pCR [pathologic complete response] reaction of 5 different loci. Nowadays, we get most of our MSI status through next-generation sequencing, the comprehensive panels that also includes the MSI status, even in liquid biopsies. We have various ways to test for these, for MSI status as mismatch repair deficiency. Unfortunately, when you look at FDA approval of pembrolizumab and nivolumab, for instance, they don’t care how we test for this phenomenon of MSI-high mismatch repair deficiency. They even allow local testing, to confirm sometimes in clinical trials. No one can tell me to test the patient for MSI. I actually think, based on the fact that pembrolizumab, for instance, is approved for all mismatch repair deficient MSI-high cancers anywhere, every single cancer—at least solid tumor—should be tested for mismatch repair deficiency, or MSI high.

The new presentation of KEYNOTE-177 shows that pembrolizumab is significantly better than chemotherapy, but this trial also gave us some interesting insights. We already knew that MSI-high tumors are more refractory to chemotherapy. This is a treatment option that I have no doubt moving to first-line. I was a little surprised when the data represented that 30% had progression of disease as the best response. That is a little scary because you don’t want to have progression in a big tumor burden patient with MSI-high tumors, missing the option of going to other chemotherapy options. This data also raised the question: What is the role of CTLA4 in the combination with immune checkpoint inhibitors? As you know, we presented the update on nivolumab-ipilimumab at ASCO this year, and we showed incredible efficacy with a response rate of 69%, and 13% had a complete remission. We didn’t see this 30% progression of disease. Now, it’s a much smaller study, so it’s limited. However, this data should raise the question, what is the role of a CTLA4 in combination with the immune checkpoint, versus an immune checkpoint alone?

Now, nivolumab-ipilimumab is already approved in refractory patients and is actually included in the NCCN Guidelines in first-line if a patient is not a candidate for chemotherapy. I have used this many times in first line because of our ongoing results and efficacy, so I’m very comfortable doing that. I think the pembrolizumab data in KEYNOTE-177 are very provocative and should be further investigated. When you look at the other data on KEYNOTE-164, they are actually very consistent. There are 20% to 30% progression of disease. The first 6 months, the chemotherapy is better than the immunotherapy in these patients, and we need to really understand the best combination for these patients in first-line.

 So far, immunotherapy has only clear evidence in microsatellite instable tumors, meaning mismatch repair deficient tumors. We just saw data at ASCO for pembrolizumab in first-line, the KEYNOTE-177 study, using immune checkpoint inhibitors in first-line as a first step and will augment the activity later by adding chemotherapy and other checkpoint inhibitors. MSS tumors, which constitute 95% of the metastatic disease, have been frustratingly refractory to approaches with immune checkpoint inhibitors. There are some MSS tumors that have high tumor mutation burden. For instance, for a patient with 

 mutations, we generate an ultramutated phenotype. Those patients have a chance to respond to immune checkpoint inhibitors, but they are less than 1%. I’ve seen a handful of patients in 2 years in our breast cancer center, where we routinely test for 

 mutations, so we can’t really rely on that. The future of immunotherapy in MSS tumors is likely a combination of immune checkpoint inhibitors plus a VEGF inhibitor plus something else. I mean, the data we’ve seen so far, the most interesting preclinical and clinical data, come from PD-1 antibodies plus a VEGF inhibitor plus a kinase inhibitor, to a certain degree. We have seen data of pembrolizumab and lenvatinib, for instance, in MSS endometrial cancer that are very intriguing. We have seen some preliminary data of regorafenib plus nivolumab; in the Japanese study, mainly lung metastases responded. We’ve seen at ASCO data of regorafenib plus avelumab, which showed higher-than-expected disease-control rate. I believe there is a future, but I don’t know exactly where this future is going. We need to clearly validate all preclinical findings, translational findings, in clinical trials because modulating the immune system is tricky. It’s not clear that you can really translate data that you generate in an animal model into a human organism.

Using I/O in MSI-H Tumors in mCRC

 When we initially looked at patients who actually got off regorafenib and then were in good shape and wanted to be treated, even outside clinical trials, we observed that when we re-challenge with agents that had been used before regorafenib, even patients who progressed on these regimens before regorafenib, might have a response stable disease after regorafenib. I remember 2 patients who triggered my interest in that, who had progressed on the fluoropyrimidine plus bevacizumab before regorafenib and responded really beautifully after regorafenib and quite long lasting. We actually discussed this and looked at preclinical data and clinical data. We did a joint analysis between the Mayo Clinic, [The University of Texas] MD Anderson [Caner Center], and the USC [Norris Comprehensive Cancer Center] and did actually collect patients where there was this idea of a clinically recent sensitization effect. We don’t know whether this is just based on the fact that regorafenib as a nonchemotherapy drug gives patients time off cytotoxic therapy, and just through this time effect allows clones to reemerge that eventually are sensitive to chemotherapy. It could just be a time effect. The other question would be, Is it a biological effect? Because regorafenib is a multikinase inhibitor, it is involved in a lot of biological phenomena. That really characterized cancer cells, including chemosensitivity, that adding regorafenib to chemotherapy, 5-FU [5-fluororacil]–based, irinotecan-based, even EGFR antibodies, suggests there is some additive-synergistic effect. But I do believe we need to really work on these animal models more, and it’s not easy to design the right animal models because regorafenib’s activity could be related to various different things, including angiogenesis inhibition and microenvironment modulation, pathway inhibition within tumor cells, and also modulation of the immune system. All these factors are not very easy to test in an animal model, and animal models might not be fully predictive of what’s happening in humans.

A very important question is when we decide to use either regorafenib or TAS-102 [trifluridine, tipiracil hydrochloride], which is the best to use before and which is the best to use after? I guess we have no evidence in the literature so far that is suggesting that 1 sequence is superior to the other. In other words, in my opinion, my preference is more based on which are the treatment goals, which are the adverse effects that I would like to avoid for the patient rather than what to expect in terms of activity or efficacy. Therefore, I treat with regorafenib followed by TAS-102 or vice versa based more on what is the natural history of the disease, how the patient was treated, and which lines of treatment were given before. I also treat in terms also of adverse effects in treatment, rather than thinking of giving regorafenib before TAS-102 is better or giving TAS-102 before regorafenib is better. I don’t think we have these data, and I think we will never have these data in the general population of patients.


That’s an ongoing question, what is the best 1 in third-line? When you look at the efficacy profiles, they’re very similar. In the future that may be answered with new data coming out with TAS-102 and bevacizumab showing nice overall survival data as well as regorafenib with immune checkpoint inhibitors. We may learn much more in the future about what combinations we should use. In general, what we learned with regorafenib is that it is important to use it in earlier-lines of treatment as well as in younger, fit patients compared with TAS-102. That’s what we use as a direction to make the most effective treatment decision and most convenient treatment decision for the patient.

A possibility then is combining TAS-102, which is a fluoropyrimidine, with regorafenib, which is an antiangiogenic agent. The question is, is it worth it to do the 2 drugs together, or is it better to do the 2 drugs in sequence? It’s very complex to give an answer. First, we need the randomized phase 3 data really to show if combining the 2 of them is better than giving each alone followed by the other. Because, for sure, we are putting together 2 different types of toxicities with 2 drugs that most likely will be both cytostatic rather than cytotoxic. We have to see if combining the antiangiogenic properties of regorafenib and antimetabolite properties of TAS-102 is worth it in terms of enhanced activity.

 Regorafenib has been used mainly as a single-agent treatment, and it’s being recognized as one of the standards of care. Now, regorafenib could potentially land itself for combination therapy with an agent that does not have overlapping toxicity, like TAS-102. TAS-102 has a very different toxicity profile from regorafenib; they do not overlap. There is a single-arm, phase 1b study for early experiences with the combination of regorafenib and TAS-102, and the data are very preliminary so far. I do believe this is an interesting combination that could move forward, but we definitely need to have randomized data to show that. Other combinations are regorafenib in combination with immune checkpoint inhibitors, like nivolumab. I believe regorafenib is 1 of those drugs that can augment immune response through various mechanisms, such as macrophage population shift, that needs to be explored further. We have preliminary data that are very interesting.

Sequencing in mCRC: Later-Line Therapies

It’s important to say that only approved drugs for third- and fourth-line are either regorafenib or TAS-102 [trifluridine, tipiracil hydrochloride]. The ideal sequence is not known, but sequences should be considered equally active or effective. In terms of rechallenge with anti-EGFR monoclonal antibodies in third-line, the question is still very intriguing and open. We have some respective data and some small-size prospective proof-of-concept trials that point out the possibility that in third- or fourth-line, when we re-treat or re-challenge with an anti-EGFR monoclonal antibody, a patient whose tumor was 

 wild-type and had a very good response to chemotherapy plus anti-

 monoclonal antibody in first-line, we will get again an effect in that treatment. Therefore, this is an option we consider but not an option that is standard of care. In my opinion, although this is a valid option, we should still regard them as experimental to explore it in appropriate clinical trials.

This is a more interesting discussion because with the development of liquid biopsies, we can monitor the progression of disease based on a molecular makeup. We now know that we have potentially the choice of re-challenging treatment we have used in earlier lines of treatment in this case first-line. Liquid biopsies would be helpful to discuss re-challenging the cetuximab, also TAS-102 and regorafenib. We have learned from many clinical trials that it seems like regorafenib will work better in earlier lines of treatment. This is an important factor. It seems that it works better in younger patients, and it works better in patients who are fit. TAS-102 seems to remain efficacious also in later-lines of treatment, and maybe overall it is a bit better tolerated in patients who have the best performance status. That all goes into the equation of what you want to do. In the third-line, I also consider if there is a tumor burden or tumor symptoms associated with tumor progression that I could re-challenge with chemotherapy. For example, when I gave FOLFOX [5-fluororacil, leucovorin calcium, oxaliplatin] in the first-line and then switched to FOLFIRI [5-fluororacil, leucovorin calcium, irinotecan] for a significant time of treatment, I would consider FOLFOX if they need tumor shrinkage, because reintroduction of FOLFOX has shown that you can have response rate from 10% to 20%, which is very difficult to reach with the single-agent TAS-102 or single-agent regorafenib. Here you can reintroduce cetuximab combination with the FOLFOX depending on your previous use of this treatment. Today, with introduction of liquid biopsy and better understanding than mechanism resistance, we have a choice to revisit treatments we have given in earlier-lines of treatment, along the new approved medications such as TAS-102 and regorafenib.

Anticancer agents are given either intravenously or orally, some of them also subcutaneously. For all drugs that are not given orally, usually for a short access to the hospital, that we in Europe call day hospital access, are administered to patients. In terms of fourth-line, both TAS-102 and regorafenib could be provided as oral drugs. This is good and could add to the compliance of the patients but requires very close monitoring of the patient to be sure that the patient is taking the right amount of pills, the right days, with the right day off or rest periods. More important, when we treat patients with TAS-102 or regorafenib, we should frequently see the patient in ambulatory care during treatment to be sure that we can control, as early as possible, potential adverse effects and check if therapy is being active and effective. Usually in our practice, when patients are put in third- or fourth- line with the regorafenib or TAS-102, patients obviously have therapy at home but they come every week to the hospital for ambulatory care. This is not very easy everywhere. Before the COVID-19 pandemic, this was normal in most Italian hospitals. With the COVID-19 pandemic, the situation has been different region by region in Italy and Europe. I still think it is important that the patients who are treated with anticancer drugs given at home orally should be closely monitored by an expert nurse or by a medical oncologist.

We have continued to do chemotherapy, and we did not adjust or make decisions or step down on our chemotherapy. Patients in second- and third-line are dependent on the efficacy of treatment, and a significant delay or undertreatment may not be in their benefit. It has to be all risk and benefits, and for us, the benefits of chemotherapy outweighed in almost all patients. That led to the decision that we continue with the treatment we think is the best without making adjustments for COVID-19.

Therapeutic Agents in Relapsed/Refractory mCRC

 When we compare the drugs available that we consider for the third-line setting, number one, of course, is EGFR antibody. I think they have the most solid data either as a single agent or in combination with chemotherapy, based on prior studies. We have also refined the patient population that can benefit, such as 

 wild-type tumors, left-sided tumors. This is a standard of care, using pemtumomab, cetuximab, either as single agent or in combination with irinotecan, for instance. They have pretty good efficacy, and we can expect to see some tumor shrinkage when we select the right patient—again, left-sided, 

Regorafenib and TAS-102 [trifluridine, tipiracil hydrochloride] are newer drugs. They are not restricted by any sidedness or molecular profile, so we have no data about whether patients benefit only when they have 

 wild type in the later-line setting. The CORRECT study, which was a pivotal study for regorafenib, randomized patients in a third- and fourth-line setting after failure of standard therapy. Patients on regorafenib versus placebo showed an improvement in overall survival 2:1. The hazard ratio was 0.77, which means a 23% reduction of death events in the study was mainly related to disease-control rate, some patients had prolonged disease control. The main issue with regorafenib was the adverse-effect profile, at least with the dosing schedule that was used, 160 mg initially, which we later addressed in a randomized comparison. However, it definitely showed efficacy as a single agent in the later-line setting.

TAS-102 was developed very similarly, about 2 years later after regorafenib. A 2:1 randomization versus placebo, international study. One hundred of 500 patients were pretreated with regorafenib, and there was no difference in outcome for TAS-102 whether the patients had received regorafenib or not, which is an interesting piece of information. Overall survival end point was reached with a hazard ratio of 0.68, so a 32% reduction of death events on the study. The progression-free survival curves of TAS-102 and regorafenib are completely overlapping, you can do a PowerPoint experiment really projecting them on top of each other and you wouldn’t know which line is which. We have very similar data. Disease control was achieved in about 45% to 50% of patients, really no objective responses, and patients normally tolerate TAS-102 better than regorafenib because there are fewer subjective adverse effects. The TAS-102 adverse effect is mainly neutropenia, which now with the coronavirus epidemic is something we don’t like. There are recommendations in terms of sequencing from regorafenib to TAS-102, which is 1 of the questions, to use regorafenib first and reserve TAS-102 for a later time point.

Regorafenib is active as a treatment option in the later-line setting in colorectal cancer and as a cytostatic agent. We want to keep patients on therapy as long as possible, because that is really how this drug will be most active: by prolonged exposure. The problem with regorafenib—especially in a high dose when it was used with 160 mg per day, like in the pivotal registration study—is that a lot of patients will encounter subjective adverse effects like fatigue and hand-foot skin reaction, which is an inflammatory response to the drug from their palms and feet. These adverse effects come early on in the treatment of these patients. We actually see these adverse effects come within the first 2 weeks of therapy, which is different, for instance, than the hand-foot syndrome with capecitabine, which comes over time with more of a delayed effect. When we learned how to use regorafenib, 1 of the important points to recognize was that adverse effects come early and that the first 2 weeks are really the critical time point to monitor adverse effects, make those adjustments, and really get patients used to this drug, because we know the longer they get treated, the better they tolerate it. So that’s an interesting phenomenon. That’s why we ran a randomized comparison trying to optimize the regorafenib dosing administration by comparing the standard 160-mg dose, which is 4 pills a day, 3 weeks on, 1 week off, with starting with half the dose, 80 mg, 2 pills a day, for the first week and then getting a toxicity assessment from patients. If they tolerate 80 mg, we will go on 120 mg per day for week 2 and then really try to go 160 mg, the so-called standard dose, in week 3. Then patients have a week break and come back to cycle 2. Then you can ask them, “Which dose level do you think you really tolerate?” This is the dose level patients will be continued on. This randomized comparison with over 100 patients compared the escalating dosing strategy with the standard starting dose, 160 mg. When you run a study like that, you need to have the right end point because safety and toxicity is only 1 aspect of that. You also want to make sure you are not harming patients, in terms of reducing efficacy, by lowering the starting dose. The end point we chose was how many patients will go to cycle 3, because at 2 cycles, 8 weeks, you got a scan, and patients would continue to cycle 3 only if they tolerated the treatment and had stable disease on the scan. That was the primary end point, and yes, this study met its primary end point. Actually, quite surprisingly, there was about a 20% difference between patients who continued to have started on 160 mg compared with patients who escalated the dose who were able, eventually, to continue treatment into the third cycle. There was a trend—not statistically, not powered for that—toward more improved survival with the lower starting dose, presumably because patients could stay on therapy longer, and their quality of life remained stable. But for patients who started with a higher dose, there was a different quality of life in all our parameters in week 2. I believe the standard of care right now in colorectal cancer dosing regorafenib is starting with half the oral standard dose, 80 mg, and ramping it up slowly. This is embraced by NCCN [National Comprehensive Cancer Network] Guidelines and has really changed the way how we perceive regorafenib. It’s turned from a perceived toxic, not very effective agent, into an agent that can be used in a community practice, even by physicians who don’t have a lot of experience with managing the adverse effects. I believe it really gave regorafenib a second chance.

CORRECT, RECOURSE, and ReDOS Trials in mCRC

For the continuum of care of medical treatment for patients with metastatic colorectal cancer, to offer them all the opportunities to extend lifespan, we must offer all the appropriate therapeutic options, also beyond second line. Therefore, after the first 2 lines, basically incorporate FOLFIRI [folinic acid, fluorouracil, irinotecan] and/or FOLFOX [folinic acid, fluorouracil, oxaliplatin] in either sequence and incorporate all the potential biologic agents. This includes anti-EGFR monoclonal antibodies and antiangiogenic drugs, as well as including the combination of BRAF inhibitor and anti-EGFR therapy in patients with a 

In third-line treatment, unless patients are entered in an early phase clinical trial with investigational drugs for specific mutations, patients are candidates for treatments that are not based on any molecular selections. We have basically 2 options that are completely different from each other, but can be used in sequence. The drug regorafenib, the multikinase, mostly antiangiogenic small molecule that is an oral drug, and another oral drug, TAS-102 [trifluridine/tipiracil]. Obviously, TAS-102 [trifluridine/tipiracil] is a chemotherapy drug and has mostly chemotherapy [adverse] effects, including bone marrow toxicity. Whereas, regorafenib has only specific [adverse] effects that are typical of small molecules of that class of agents and oral antiangiogenic agents. Therefore, especially for patients in the first lines of treatment who have been treated with very aggressive chemotherapy regimens, for example, with triplet chemotherapy or for long periods, I would like to spare them from further bone marrow toxicity by chemotherapy. I prefer to use the regorafenib first, and then eventually at progression, TAS-102 [trifluridine/tipiracil]. For some patients in which the angiogenic drugs have been toxic in the previous lines of treatment, or the antiangiogenic treatment has been very short because of rapid progression, maybe TAS-102 [trifluridine/tipiracil] would be a better option before going to regorafenib. I will say that, in general, we have no evidence that TAS-102 [trifluridine/tipiracil] is better than regorafenib, or vice versa. We have no evidence that sequencing the regorafenib followed by TAS-102 [trifluridine/tipiracil], or vice versa, is a preferred option. What we know is that we should use both agents in sequence in third and fourth line and that in most cases in practice, when patients have been treated intensively with chemotherapy in the first lines, regorafenib will be, for the mechanism of action and for the adverse effects, the best potential option.

An important question is how many patients after failing second line will go to third line? I would say that between 30% and 60% of patients after we start with the first line, will end up starting an appropriate third line of treatment. Among those patients that start a third line, about half of them will start a following fourth line of therapy.

When you look at the official data, it's always shocking how low they are. In my practice, maybe because it's a referral center, basically almost 100% go to a third-line treatment. Very few don’t go to third line, when they are 

 mutated, have very aggressive extensive disease, are elderly, or have a significant issue with liver function. But the vast majority goes into third line. In fact, I think our dilemma is we don't have enough treatment after third and fourth line because these patients are still in relatively good shape, and with increasing effectiveness of our first- and second-line treatments, this patient population moving to higher lines of therapy may get bigger.

 As patients move from line to line to line of therapy throughout the continuum of care, the more limited our treatment options will become, eventually, the more potentially affecting the patient's performance status, overall health, and they might be fatigued from just getting a lot of treatments; their health outcome and their health perspective has completely changed. When we look at later lines of therapy, we're not talking about the immunogenic tumors where we have great treatment options with immune therapies. We’re talking about the garden variety, 95% of patients, who have no new, let’s say, immune checkpoint inhibitor option, and are linked to regorafenib and TAS-102 [trifluridine/tipiracil] as treatment options.

The goal of therapy has changed. We're not looking at shrinkage of the tumor, long-term progression-free survival, and maintenance therapies, etc. Our goal is instead to control the disease and maintain quality of life for patients. I think that's a critical point. When patients initially come to see me in first line, everyone wants to be cured, and we all know that this is not possible. The longer patients stay on therapy, the more they realize that this is not possible, and that goals of care are changing. My goal is really to have them live as long as possible and as well as possible, and we talk about this in patient interaction. When I introduce, a third- or fourth-line setting based on progression of disease, if the patient has good performance status, willing to continue therapy, then I introduce the goal of care, which is that I want to control the cancer as long as possible and I want you to have good quality of life. Patients are intrigued by the idea of getting oral agents, which make them more independent, but you need to be quite aware of the toxicities of these oral agents, with the idea that we don't want to compromise quality of life.

 It is very important to talk to the patient before we start first line and explain to them that we have a series of therapeutic options that we are trying to use in the best sequential way that we know, according to clinical practice, based on evidence-based medicine. Therefore, when we start the first line, we try to explain to the patient that we will try to make their disease as chronic as possible and try to control the disease for as long as possible. Therefore, any passage from one line to another for us is not seen as a failure but as a possibility or an opportunity to explore another combination or another therapy with a different mechanism of action that we hope will be effective. What we say to the patients before starting third line is that with either regorafenib or TAS-102 [trifluridine/tipiracil], whichever one we pick, we explain the adverse effects, and we hope that at least 1 or 2 of them will get a good advantage in terms of progression-free survival. Especially for regorafenib, I could say to the patient that a subgroup of them, maybe 1 out of 5, or 1 out of 6 could experience a very prolonged control of disease progression that could last even several months.

 I think the treatment goal depends on the extent of disease and organ involvement. If you have organ-limited disease, like liver or lung, I see if there is a chance that the treatment goal could be cure or curative resections. If this is a widespread disease, the treatment goal would be palliative, and it's very clear why we do chemotherapy. It's very important to share that with the patients, that the treatment goal is to extend their life, but also guaranteeing that quality of life. It's not only extending life but also preventing complications that can be caused by the tumor, like obstruction, and bleeding, and pain. I think that is often not really discussed in detail enough that patients understand that it's not only about living longer, but also living better.

Later-line therapy in colorectal cancer, in patients with too much progression on 5-FU [fluorouracil], oxaliplatin, irinotecan, bevacizumab, so the VEGF inhibitors, and if they are 

 wild type, EGFR antibodies, the drugs that we have available are 2 oral agents as standard of care. It is regorafenib, which is a multikinase inhibitor, not chemotherapy, and TAS-102 [trifluridine/tipiracil], which is a combination of fluorothymidine, and a degradation inhibitor of trifluridine. These 2 drugs are available, and they have very similar efficacy in phase 3 trials. These drugs were developed in the later-line setting, 2:1 randomization against placebo, and studied in the Western population and Asian population.

Development of these agents was much the same. Both have clear demonstration of overall survival benefit, both agents do not induce response to a large degree, and the response rates we see are around 1% or lower. The emphasis of both agents is disease control rate, so we can control the disease in about 45% to 50% of patients. In those patients, we have data on regorafenib, patients have control of disease after 2 months from when we get the first scan. If they don't have progression of disease, they have a 50% chance to be alive 12 months from now, which I think is a meaningful parameter for patients.

Later Lines of Therapy for mCRC

 When we plan a continuum of care for the patient with metastatic disease, we should keep in mind what are the therapeutic options for second, third, and fourth line if we choose A or B as first-line therapy. Basically, the destiny of subsequent lines is in some way dictated by first-line choice. Therefore, for all 

 wild-type patients, in first line, especially if patients have a left-sided primary tumor, the first-line choice is an anti-EGFR monoclonal antibody, either cetuximab or panitumumab, plus FOLFIRI [folinic acid, fluorouracil, and irinotecan] or FOLFOX [folinic acid, fluorouracil, and oxaliplatin]. In this case, the second line upon progression of disease will be the alternate chemotherapy regimen—so FOLFIRI if FOLFOX was first line, or vice versa. After failure on a combination of FOLFIRI or FOLFOX in the first line with cetuximab or panitumumab, in the second line we incorporate bevacizumab with a chemotherapy doublet, or if using FOLFIRI, we will be incorporating aflibercept.

-mutated tumors, the choices are again either FOLFIRI, or FOLFOX , or FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan], in the way I explained it before, in combination with bevacizumab. Upon progression, if a patient was treated with FOLFIRI, we will utilize antiangiogenic treatment beyond progression with bevacizumab; or with aflibercept, substituting FOLFOX with FOLFIRI for example. More complex is what we choose if we’re treating a first-line patient with FOLFOXIRI. In this case, the second line will be decided also based on chemotherapy adverse effects in terms of bone marrow toxicity if the patient can be treated again, either with irinotecan or oxaliplatin in combination with 5-FU [fluorouracil] in the second line.

 mutated now, the second-line first option for us is a combination of the anti–BRAF-mutated inhibitor encorafenib, and anti-EGFR monoclonal antibody cetuximab. So encorafenib/cetuximab is the preferred second line, regardless of the chosen first line, in a patient with a 

 mutation. So although this is not yet used in Europe and is used mostly within clinical studies, I hope in a short time this will change, and finally we will also have approval in Italy.

 I think it all, of course, depends what you use in first line. I think it's very easy when you use the FOLFOX/cetuximab in first line for left-sided colon cancer. The next best effective treatment is FOLFIRI with bevacizumab. If you use FOLFOX with bevacizumab or FOLFIRI with bevacizumab, the question is if this is a wild-type or a mutant 

, I think the next subsequent treatment would be FOLFIRI with bevacizumab after FOLFOX with bevacizumab. I think it's more complicated for patients who failed on FOLFOXIRI with bevacizumab, and what the next line of treatments are, because you basically failed on the standard of care cytotoxic and the targeted agents approved for this situation. I often go to a regorafenib combination. We have now a clinical trial with regorafenib and pembrolizumab as a second-line treatment, as we learned regorafenib may be more beneficial in earlier lines of treatment, and this combination had very promising results presented by Dr. Shota Fukuoka, at ASCO [the American Society of Clinical Oncology annual meeting] last year.

 The reason to move from first line to second line therapy is clear progression of disease, that is indisputably likely the most important reason, the other one being tolerability of therapy. The emergence of biomarkers that have become available also plays a role. For instance, if a patient has not been tested for MSI [microsatellite instability] status up front, which should not have happened, and we see this patient is barely hanging in there with chemotherapy, has toxicities, but now has an MSI-high tumor, then I would of course switch to immune checkpoint inhibitor therapy, either single agent or combination with a CDK-4 [cyclin-dependent kinase 4] antibody. There are various determinants that help us: progression of disease, toxicity, adverse effects, tolerability of therapy, and then of course, the emergence of molecular markers.

: I think depending on the first-line treatment, for second line you choose the most effective treatment, FOLFOX to FOLFIRI, or FOLFIRI to FOLFOX. In wild type you switch from cetuximab to bevacizumab. Most of the time in second line, it depends on the progression of disease, the tumor burden, how aggressive you want to be. The most used and the most effective treatment is switching to the other backbone of chemotherapy. As mentioned before, the dilemma is a little bit more difficult when deciding what to move to after you use FOLFOXIRI with bevacizumab. In the case of wild type, this is often not an issue for our practice because we use the cetuximab first line. For my colleagues who use bevacizumab first line, they could choose a cetuximab combination in second line.

 Regorafenib has so far mainly been used as a single agent in the later-line therapy. There have been some attempts to combine it with chemotherapy like FOLFIRI in the second-line setting. That was a randomized, phase 2 study, FOLFIRI plus either regorafenib or placebo. The study overall, showed improvement in progression-free survival, no difference in overall survival. Most recently at ASCO, there was an analysis based on 

 mutation status, and it appeared that regorafenib did not improve progression-free survival, at least not statistically significantly, in patients with 

-mutated tumors. This is something that is not unique to regorafenib, there are a lot of subgroup analyses of existing trials, even looking at VEGF inhibitors, that are not perfectly solid in terms of whether bevacizumab or other VEGF inhibitors would add activity to 

-mutated tumors. This must be taken with a grain of salt because these are unplanned subgroup analyses that are narrowing down the patient population. There might be other confounding factors that we haven't realized. At this point, since we’re not using regorafenib in combination with FOLFIRI in second-line therapy, I would use this as hypothesis generating. I would not translate these data, and I would not restrict the use of bevacizumab in the 

Second-Line Treatment for mCRC

 Your choice of first-line treatment sets in motion a sequence of events. When we start with, let's say, an oxaliplatin-based regimen, we want to use an irinotecan-based regimen in the second line. The choice of the biologic agent is determined by molecular factors, 

 mutation analysis, MSI [microsatellite instability] now with immune checkpoint inhibitors, but also adverse effects. We know that left-sided tumors have a chance to respond to EGFR-antibody therapy, and I consider them, for most patients, standard of care. So, cetuximab, panitumumab, added to backbone chemotherapy, either a chemotherapy doublet or even a chemotherapy triplet, and then we would use bevacizumab in a second and potentially even in the third-line setting, with bevacizumab beyond progression.

These sidedness questions really determine, in a way, the sequence of events and how many lines of therapy we have available. In the right-sided tumor, even if it is 

 wild type, I would not use an EGFR antibody in the first line. I might, however, consider the EGFR antibody after failure of all standard therapy, which would then include regorafenib and TAS-102 [trifluridine/tipiracil], because there are very likely patients with right-sided tumors who do benefit from EGFR antibody therapies. We have not been able to characterize those patients yet, but I would not withhold treatment when a right-sided tumor patient’s back is against the wall when I believe EGFR antibodies have not been used before.

What we have learned over the past 20 years is that metastatic colorectal cancer is a very heterogeneous disease, depending on which are the most important or relevant genetic alterations that drive tumor development and progression. Depending also on where the tumor arises, if it’s a localized primary tumor in the right side or in the left side, because these are usually accompanied by different genetic alterations and so different potential genetic drivers, and different potential molecular targets for specific therapies. Therefore, before first line for metastatic disease, what we do is try to characterize, as much as possible, the genetic makeup of the tumor in a patient.

In large comprehensive cancer or academic cancer centers, this is achieved usually by a next-generation sequencing approach. In most cases in the big countries in Europe, most patients, before starting first-line [therapy], are characterized in their tumor, at least for the presence of mutations in the 

 genes, presence or absence of mutations for the 

 gene, and more recently, for DNA repair deficiency or proficiency. Therefore, microsatellite stability or microsatellite instability is measured in these patients.

To summarize my answer, I will say that at least for the large cancer centers in Europe, 

 mutations, MSI status, primary tumor location, and FDG [fluorodeoxyglucose] amplification status are routinely measured before starting first-line treatment. These define what is our choice because usually what we decide to use first line is a chemotherapy doublet like FOLFIRI [folinic acid, fluorouracil and irinotecan] or FOLFOX [folinic acid, fluorouracil and oxaliplatin] that is combined with an appropriate biologic agent, such as an antiangiogenic drug, most likely bevacizumab, or an anti-EGFR monoclonal antibody for 

 wild-type patients, mostly if they have a tumor localized on the left side of the colon, with cetuximab and panitumumab. For MSI-high patients in Europe, this is different from what happens in the United States, unfortunately PD-1 or PD-L1 inhibitors have not yet been registered by our authority, EMA [European Medicines Agency]. So we cannot use them outside of clinical trials, either in first line or subsequent lines. I’m sure this will change soon in Europe after the presentation at the ASCO [American Society of Clinical Oncology] 2020 annual meeting of the results of the KEYNOTE-177 study, in which pembrolizumab showed a very important progression-free survival advantage over standard chemotherapy in first line of MSI-high patients.

 mutation, after the BEACON data reported at the end of last year, clearly what is known now is that for second- and third-line patients, a combination of BRAF selective inhibitor encorafenib, and cetuximab, the anti-EGFR monoclonal antibody, is the primary option for second line. Whereas, what is the primary option for first-line patients with a 

 mutation is still under debate, unless these patients have an MSI-high status in which I think immunotherapy plays a major role. After first line, all patients are candidates for second line. However, unfortunately, a proportion of them after progression are not fit for second line. What happens in general in Europe is that between 50% and 80% of patients after first-line failure are suitable and fit for the second-line treatment.

 I think the decision for first-line treatment is really the most important decision you'll make for your patients. I think it is critical to understand the molecular makeup. You need to know 

 status, microsatellite instability, because that will determine your choice of first-line treatment. For 

 wild-type and left-sided colon cancers, for us, it would be a combination of chemotherapy with an EGFR inhibitor. I like FOLFOX [folinic acid, fluorouracil and oxaliplatin] with cetuximab. If this would be a mutation of 

 where you cannot use an EGFR inhibitor, you would use bevacizumab. Here it comes to a decision about how aggressive you want to be with your chemotherapy backbone. If dealing with a big tumor burden, borderline resectable, young, fit patients, I would consider a modified FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan] regimen. I do the same for right-sided colon cancer with wild-type 

, because right-sided colon cancers are significantly associated with poor outcomes, and here I step up the aggressiveness with a chemotherapy backbone.

: An important issue for treatment of patients with first-line metastatic colorectal cancer is, which is the best chemotherapy backbone. In my opinion and according to the guidelines, like the ESMO [European Society for Medical Oncology] guidelines, a chemotherapy doublet should be offered to all patients regardless of age, unless these patients are unfit for therapy. A major question is do we need to intensify doublet chemotherapy, either FOLFIRI [folinic acid, fluorouracil and irinotecan] or FOLFOX [folinic acid, fluorouracil and oxaliplatin] to FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan]? The answer is, in some patients who are fit, especially with bad prognostic characteristics, maybe the choice of FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan] as a backbone for first-line treatment is a good choice. In our practice, we usually use FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan] instead of FOLFIRI [folinic acid, fluorouracil and irinotecan] or FOLFOX [folinic acid, fluorouracil and oxaliplatin] for fit patients with 

 mutations in which we would like to have a rapid tumor shrinkage. In this case, we combine it with bevacizumab. Also for fit patients with 

 mutations, for which we require important tumor shrinkage, we use.

In these cases, we don’t know if FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan]/bevacizumab is better than FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan] alone. But in our experience, FOLFOXIRI [folinic acid, fluorouracil, oxaliplatin and irinotecan] as a chemotherapy backbone, as compared to FOLFIRI [folinic acid, fluorouracil and irinotecan] or FOLFOX [folinic acid, fluorouracil and oxaliplatin], is reserved for fit patients with 

 mutations where we need a rapid and sustained tumor shrinkage.

Initiating Therapy for mCRC 

: Treatment of metastatic colorectal cancer has been improving since we started to do a continuum of care. That means we have the possibility of using a series of different antineoplastic drugs together, and more importantly in sequences, that allow us to treat patients for long periods to control disease progression, obtain tumor shrinkage, and extend the survival in these patients while preserving quality of life.

Therefore, continuum of care with different lines of treatment became the standard treatment for most patients with metastatic colorectal cancer. This continuum of care includes whole surgical resection or locoregional treatment, whenever necessary, of some metastatic sites, and mostly to liver metastases. Continuum of care has changed metastatic colorectal cancer from a disease that used to have a median survival time of between 6 months and 12 months only 20 years ago, to a disease in which most patients can experience a survival time across all lines of treatment, globally, of 30 to 40 months. In some groups, extension of survival is even longer.

All patients should get all the lines of treatment that are necessary to control the disease for as long as possible. In Europe, most patients are treated either in comprehensive cancer centers or in regional hospitals in which a multidisciplinary team approach allows these patients to get all the therapeutic options within a public health care system. Therefore, in most regional health care systems in Europe, when patients get the diagnosis of metastatic colorectal cancer, they are referred to a specific team of experts that will define the sequence of therapies, and if necessary, also the introduction of surgery for liver metastases when it is appropriate. There is not too much difference between treatment options if a patient goes to a large, big comprehensive cancer center or a smaller regional community hospital, at least for the big countries in Europe.

 We've known for a long time that patients benefit from getting access to all active agents, and this is still linked to this question of lines of therapy; first line, second line, with maintenance therapy in between. There's this perception that patients in a community setting might get fewer lines of therapy that can influence the outcome. I'm not sure about that because it's really not about academic versus community, it's more about the patient population. From my experience in the Mayo Clinic and now being in a more academically phrased community setting, there might be different patients, performance status, access to health care, etc, that might be a more determining factor than academic versus community. The goal is to keep patients exposed to all lines of treatment based on molecular parameters that we test up front. For instance, a patient with a 

 wild type situation has the chance to get 1 line of therapy more just being exposed to EGFR-antibody therapy.

Unfortunately, even in an academic or high-end community setting, not every patient will go from first line, to second line, third line treatment. I think we probably lose patients, about 20% from each line to line to line of therapy, and that is reality. That is something we are likely never going to change.

I think the unmet needs for colorectal cancer are significantly defined by the molecular characterization of colorectal cancer. We have come a long way to use and integrate biomarkers to decide the best treatment options. We all now look for 

 mutations and microsatellite instability, which changed the selection of treatments we discuss with our patients. I think one of the biggest unmet need is the 

mutated group, we still do not have treatments for this patient population when they have failed first-line treatment. I think there are some promising new targets being developed, however, we do not have standard of care treatment options for this patient population. I think another very important one is the microsatellite stable [group], where we have not successfully integrated immunotherapy for these patients to try to turn a cold tumor into a hot tumor to offer immunotherapies in combination with other targeted drugs to see significant success.

I think one of the biggest unmet needs is the Wnt pathway; 95% of colon cancers have activation in this pathway, and we still have not achieved any new drug development with any success or promising results in that arena. I think that would be a big step forward, and particularly, we would also target the colon cancer stem cells, which are ultimately responsible for the failures of our therapies.

Fortunato Ciardiello, MD, PhD, University of Campania

Articles