Targeted Treatment Approaches for Colorectal Cancer (CRC) — Thierry André, MD

PROF ANDRÉ: My name is Thierry André. I’m a medical oncologist working in Saint Antoine Hospital in Paris. It’s a pleasure to be with you for this seminar with focus on GI cancer.

The first study is a study published in JCO at the end of the year. It’s important because it is for locally advanced colon cancer. And it is with pembrolizumab for MSI colon cancer or other GI cancer with no metastases. It’s locally advanced. And it’s a proof of concept that pembrolizumab works in this situation. It’s a lot of complete response for patients receiving pembro with surgery after pembro, median 3 cycles. It’s a retrospective trial. It’s not a prospective trial. And the reason, it’s different populations and the decision depends of each patient at surgery after 3 cycles of pembro. The vast majority were in complete pathological response. And the other half is without surgery, but with complete response in radiology or on endoscopy. Not all, some, the majority had efficacy.

But during the study course and subsequent follow-up, some patients with progressive disease. But it is a proof of concept for locally advanced unresectable GI cancer. Majority of colon and rectal cancers that are anti-PD-1, pembro can, before surgery, obtain a complete histological response. And some patients have the opportunity to avoid surgery and to have a watch-and-wait approach if endoscopy and radiology shows a complete response. A lot of questions after these concepts. How many cycles before surgery? Watch and wait or surgery? And what predictive factor we have right now because not all respond. We have some patients with progression in this situation.

DR LOVE: So I was kind of surprised when I saw in the initial slide 2 patients with pancreatic cancer because, I don’t know, I haven’t heard much about MSI-high pancreatic cancer. How often do you see it? And I noticed that both patients didn’t do too well.

PROF ANDRÉ: MSI, it’s approximately, there’s now some data, it’s approximately probably between 1 and 2% of pancreatic cancer. So diagnosis is not simple because by pCR, it’s difficult because the diagnosis is done by a needle biopsy with not a lot of material, but with immunochemistry, it’s probably easier. And now, in my opinion, it’s really necessary to have immunochemistry for pancreatic cancer because if it’s MSI, it’s really an incredible chance for the patient to have access to immunotherapy. And in this publication, there’s 2 patients with efficacy of pembro in this situation.

DR LOVE: Yeah, but I mean these 2 patients progressed immediately. Have you seen patients have good responses to IOs with pancreatic?

PROF ANDRÉ: Yes, sure. It’s some publication about metastatic pancreatic cancer and it’s some data with pembro with nivo. And I published a paper with dostarlimab, and there really is a rate of response for metastatic disease is exactly the same independent of the localization and in pancreatic cancer with dostarlimab is probably the 2 trials, one is pembro with maybe a little less response rate compared to the other localization. But with dostarlimab for pancreatic cancer, we have approximately 20 patients. I don’t remember exactly the number. But the response rate is around 40%, exactly the same that we have for other localization.

DR LOVE: Yeah, I think that’s really interesting. Do you think that most investigators know that? And do you think that most oncologists in general practice know that, that you can see MSI-high pancreatic?

PROF ANDRÉ: No. Definitely, no. For pancreas, it’s really immunochemistry because it’s a very particular tumor and at this time, pCR didn’t work well for the diagnostic of MSI in pancreas. Maybe in the future, NGS, but with few material it’s probably complex immunochemistry and really it’s important for metastatic disease and to have the immunochemistry for the diagnosis of dMMR in pancreatic cancer. It’s not advised because it is very rare, but we have patients in my center — we have patients treated with immuno-oncology for MSI pancreatic cancer, for dMMR pancreatic cancer to be clear.

DR LOVE: Very interesting.

PROF ANDRÉ: Okay. We have presented at ESMO a Phase II trial, the NICHE-3 trial with the combi of nivo plus relatlimab. It’s difficult to pronounce. It’s an anti-LAG3 for, it’s not really locally advanced, it’s for colon cancer nonmetastatic. The problem with MSI tumors nonmetastatic, it’s really tough to have the TN staged by CT. And on this slide, it’s 74% of patients and plus. But the reality is it’s not the pathologic report. And we know that MSI has a lot of nodes and nodes are not — so it is a correlation between the CT and the pathologic report is not so good is the reason. It’s CT classification of colon cancer. The thing is the response is really excellent. It’s like NICHE-1 and NICHE-2. It’s a very high rate of complete histological response at surgery, but not so different of nivo/ipi.

The summary of all the NICHE trials and also the trial with immuno-oncology in nonmetastatic colon cancer. NICHE-1, 67% of complete histological response with nivo/ipi. NICHE-3, it’s a little bit more but it’s a very little trial. It’s difficult to tell if it’s better. We have a Chinese trial with another anti-PD-1, so toripalimab, we have also but it is very little number. And in NICHE, we have a very little rate of complete pathological response on MSS and it’s relatively interesting. And it’s very interesting to know that colon cancer MSI, at the very high rate of complete histological response before surgery in this situation. So complexity, it is not possible to have the TNM classification by the CT. And if you treat the colon cancer by Immuno-Oncology before the surgery, you treat at lot of Stage II. And Stage II is cure, Stage II MSI is cure by surgery. It’s not true that — in all the NICHE, on CT it is a lot of N+, but we know that N+ on CT is not N+ in the surgical material. And for this reason, it’s complex because it’s clear that it works, but it’s not the standard of care to have immuno-oncology before surgery for MSI colon cancer. And it’s a lot of question about what we will do right now with these results with the fact that it’s not possible by CT to define if it’s a Stage II or a Stage III. Stage II don’t need immuno-oncology. Stage III, it is probably a way to avoid chemo after the surgery. And we need a clinical trial to answer that question. And at this time, it’s an incredible proof of concept of efficacy. It’s really a demonstration that for locally advanced, when the surgeon that it’s probably difficult to have resection of the tumor. So immuno-oncology is really necessary before surgery for tumor with the opportunity to have resection by surgery. And it’s more complex and we need more data before to give immuno-oncology before surgery.

Moving on to the KEYNOTE-177. So KEYNOTE-177 was published in New England Journal of Medicine, was published in Lancet Oncol for overall survival. It’s a study with 2 primary endpoints, PFS and OS. PFS was, as results show, at 0.6 in the paper of the New England. Overall survival was published in Lancet but was nonsignificant. And it’s a lot of explanation because in this trial, the curves are pretty good, but because it’s a crossover with the opportunity to have pembrolizumab if the patients progress after chemo. And there is a presentation of the research with 5 year follow-up. And for the first time, overall survival is, with hazard ratio with limits at 0.99. It’s not a p-value because it’s new and it’s not good to have a lot of p-value, but it’s clear that overall survival seems better if we begin by the pembrolizumab. But it’s also important to see in this trial that if patients begin by chemo, pembro works and immuno-oncology works well in this trial. 63% of patients received immuno-oncology in second line, if received chemo in first line. And it’s very good results.

Concerning PFS hazard ratio of 0.6, not very different. We have it at the beginning. Adverse events are less with pembro compared to chemo. So Grade 3-5 toxicity was 21.6% with pembro and it’s 67.1% with chemo. Published in Lancet Oncol is the quality of life, and it’s really better with pembro compared to chemo. And really, these results now, it’s clear that pembro is an option in first line. It’s the standard of care in first line. It works well. And it’s a study with a better overall survival of all studies previously published in metastatic colorectal cancer. OS is better. That was a study published by the French team concerning the surgery of metastatic disease, not for MSI but for all metastatic colorectal cancer when the liver metastases can be removed. And it’s probably the first time we have the hope to cure metastatic disease with medical therapy. Because 34% of the patients are without progression at 5 years and it’s really a huge progress in colorectal cancer disease.

We will move on to the next step. The next step is 8HW. We have demonstrations that monotherapy with anti-PD-1 works. So the second step is to see if the combo of nivo plus an anti-CTLA4, anti-PD-L1 and anti-CTLA4 works well. The design of the 8HW, it’s a little bit complex because it’s for colorectal metastatic patients, MSI. But it’s the possibility to include patients in first line but also in advanced line if you received chemo before. And it’s a randomization between 3 arms, nivo alone, nivo plus ipilimumab, ipilimumab is only for the full first cycle, and the third arm is chemo +/- bev or cetux. And it’s a study with 2 endpoints, a primary endpoint for the nivo plus ipi compared to chemo only in first line, and a second endpoint comparing nivo versus nivo/ipi for patients included in all line. And I presented at ESMO GI, the first endpoint is a comparison between nivo/ipi versus chemo in first line. It’s, 303 patients were included in this trial. And the population is well balanced between 2 arms. The important thing is the fact that the primary endpoint is for the population with confirmation of MSI-high or dMMR status in central testing, inclusions were local, but the endpoint would be there with central testing. It’s important things because in this trial, around 15% in the arm of nivo/ipi and 17 in the arm chemo was not confirmed MSI. And we have data about the fact that on this percentage 10% were MSS and pMMR. It’s really misdiagnosed patients and it’s relatively important between 5 and 6%, we don’t have the method yet to confirm it is MSI or dMMR.

Results are amazing because it’s rare to have in oncology a hazard ratio at 0.1. Remember with pembro, it was 0.6. There, it’s 0.1 with now follow-up of 21 months. An important thing is all the subgroup analyses also show good benefit of the combi compared to the chemo with the toxicity of the immunotherapy, less important with immuno-oncology compared to chemo. There, the percentage of Grade 3/4 is really more important with chemo, 48% Grade 3/4 and 23% with nivo/ipi with a classical effect of nivo/ipi concerning the immune-mediated effect with the fact that we have in this trial 2 toxic deaths in the nivo/ipi arm in relation with the immune toxicity, 1 cardiac toxicity, and 1 pneumo toxicity. It’s not possible to compare the toxicity of the CheckMate 8HW to the KEYNOTE-177 because it’s complex to have this kind of comparison. But it’s clear that with these results, the question is to clearly an option in first line for metastatic colorectal cancer. And the question for the future is which patients for the combi, which patients for the immunotherapy? And certain parts of this trial is really important because it’s a comparison between nivo and nivo/ipi. And we have to answer this question.

The next study was the Phase II. This Phase II was published 2 times, well known in the US because nivo/ipi, it’s an option for MSI in first line. It is a study done before the result of the Phase III with pretty good results. It’s a very high overall response rate, 71%. And there, it’s the results with more longer follow-up. I think it’s 64 months with the curve of nivo/ipi in first line, pretty good. You see the plateau after, and it shows the demonstration that probably we can speak about a cure of metastatic disease with immuno-oncology with a pretty good curve of overall survival. The only thing is the schedule of nivo/ipi is not the same that it was done in previous publication concerning the combi of nivo/ipi in patients refractory to chemo or with the schedule presented in the 8HW. It’s a schedule with nivo every 2 weeks and ipi every 6 weeks but not only 4 cycles, it’s for the totality. And those are the schedule of nivo/ipi and a question right now, what is the best schedule? It’s clear with the fact that we have a Phase III, probably. The schedule will be more nivo/ipi with ipi low dose only for the first cycle but it’s also another schedule.

DR LOVE: Could I just mention one question which is kind of looking outside of colon cancer, I think endometrial maybe, but what about chemo plus IO or chemo plus IO/anti-CTLA4? It seems like most of these studies don’t look at that.

PROF ANDRÉ: It’s a good question. But it’s clear that the development in colon cancer was to develop immuno-oncology in MSI. For example, in gastric cancer, it was to develop immuno-oncology plus chemo and to look what happens for the MSI. It’s clear with the results of nivo plus ipi in first line with the fact that at 2 years we have 70% of the patients with control of the disease without chemo. We don’t want to go back and to give a combi of nivo plus ipi plus chemo. The problem will be in the future to probably have a predictive factor to try to find the patient who is — because it is the minority, it’s not the majority — to find the patient who is really risk of relapse or progressive disease on immuno-onco to maybe have other option and other option is maybe the combi of immuno-oncology and chemo.

So another manuscript about BEACON mCRC study. It’s clear that at this time, for BRAF, it’s a prognostic factor for colon cancer. It’s important to know that approximately 25% to 30% of the patient BRAF with a mutation V600E is MSI. If it’s MSI the therapy is immuno-oncology because the results are pretty good. But if the patient is MSS, at the time of the BEACON, it’s not clear, it probably, or showed some MSI there, but it’s clear that things are moving. So standard of care for patient BRAFV600E, after first line, is a combination of encorafenib plus cetuximab. And this publication is actualization of the toxicity and indicate that the therapy is well tolerated with few nephrotoxicity but it’s really now standard of care and it’s important to know the toxicity of this combination but this combination is not chemo. It’s targeted therapy and so it’s really — so safety is good and it’s good news for the patient.

The previous, the BEACON study, is in second line and we have now that in first line. It’s a study ANCHOR. It is a Phase II of BRAF mutated V600E with the combination of encorafenib, binimetinib, and cetuximab. In BEACON it was a randomization between encora/cetux, encora/binimetinib and cetux. And it’s not clear that binimetinib adds something. It’s some toxicity but don’t add a lot of efficacy and for this reason, the standard of care is encorafenib plus cetuximab.

In first line, the combination of the 3 therapy give good response rate. The primary endpoint was met. 47.4% of response rate. The primary endpoint was met but the PFS is 5.8 months and it’s a little bit disappointing because we were hoping better. OS is 18.3 months and for this reason it was not planned to have a Phase III because RECIST PFS, it will be probably complex to have better results compared to chemo with the standard of care, FOLFIRINOX plus bev — or FOLFOX plus bev. And for these reasons, this combination is a possibility if already, it’s an option for patient noneligible to receive standard first line therapy.

We move on right now in a new target, HER2. HER2, it’s a very exciting target because it’s a very important target in medical oncology, breast but also gastric cancer. And we have now data about some compound in colon cancer. In colorectal cancer HER2 is positive in 3% approximately of colorectal cancer and it’s more frequently in left colon cancer and in rectal cancer compared to right colon cancer. So the diagnosis is made by immunochemistry+++ or ++ and if the case is ++ need FISH-positive to have access to the — probably, with efficacy results here after with compound or by amplification by next generation sequencing, NGS. We have now this study for patients with refractory disease with opportunity to treat with tucatinib plus trastuzumab or with tucatinib alone and it’s a randomized Phase II with really a tyrosine kinase inhibitor anti-HER2 combined with trastuzumab in colorectal cancer. These patients, 100, more than 100 patients were included and it is a publication in Lancet Oncol, important, because it’s really a — in US it’s fairly possible to have access to the combi of some compound trastu and also tucat — not tucatinib but lapatinib. In Europe it’s no labeling for anti-HER2 compound at this time.

The credits are very impressive. If we look, the combi, the response rate is 38.1% with some rare complete response and passive response but PFS at 8.2 months and the median OS at 24.1 months. It’s pretty high for population refractory to chemo. We have, in this trial, it was a study with 1 arm, with tucatinib alone. Tucatinib alone is probably not sufficient because the response rate is low, only 3%. But clearly this trial demonstrates interest of this combination. In US it’s labeling in patients refractory to chemo for this population. No labeling in Europe. The most common treatment-emergent adverse events with this combi were diarrhea but with low rate of Grade 3-4, also fatigue, and also nausea. But it’s really interesting because it’s really a little part of the, a little number of patients but with efficacy.

The problem is the diagnosis of HER2 and it’s a different way to diagnose. It’s immunochemistry and so it’s clear that we have the 3+, we have the 2+ positive, and we have the negative, negative 1. And it’s clear that if we have immunochemistry 3+, the chance to have a high response rate is higher if we have only immunochemistry 2+ with FISH positive and the response rate is very low if it’s negative. And it’s important to now move on with this combi, this drug, in first line. And the MOUNTAINEER-03 trial began. It’s a randomization for HER2 positive patients, immunochemistry of 3+ or 2+, FISH positive, randomization between the combi of tucatinib/trastuzumab with FOLFOX compared to the standard of care, FOLFOX plus bev or cetuximab in this situation. It’s a very important trial.

The other compound in HER2 is, like in breast, is trastuzumab deruxtecan and we have also Phase II published in Nature. It’s actualization of previous that are published, I think it’s in — I don’t remember, but it’s already in Lancet Oncol, a very good journal. And it’s actualization of the results with this compound for patients who are refractory to chemo. It has 3 cohorts. A cohort with HER2 3+ or 2+, FISH-positive, cohort HER2, FISH-negative, and a cohort HER2 with only 1+ or HER2-negative. And it’s clear that the results are with interest for the appropriation of patients with immunochemistry with 3+ or 2+, FISH-positive, with response rate of 45.3% with no efficacy in cohort with HER2 2+ FISH-negative or HER2 1+ with the classical toxicity of these compounds, primarily interstitial lung disease with symptom of pneumonitis but with time it is very important to follow these patients to try to object nodes very early of the pneumonitis to stop the therapy in the case of patients who receive this therapy.

After the proof of concept with a Phase II, we have a randomized Phase II because the dose is not so clear and it was important to know what is the best dose and it was a randomization between the — in metastatic colorectal cancer refractory to chemo between the dose of 5.4 mg/kg, if a dose with 6.4 mg/kg is better. It’s a randomized Phase II. The primary endpoint is overall response rate. And the results are not better with increasing the dose to 6.4 and so the response rate is pretty good in — with the dose like in previous trial of 5.4 mg/kg with response rate of 37.8 response rate and the curve of PFS with median PFS there, 5.8 months, but not better if we increase the dose.

DR LOVE: I’m kind of surprised that T-DXd is not approved and being used in colorectal cancer. Do you think the data is adequate right now to actually use it?

PROF ANDRÉ: In Europe it’s difficult to have labeling without randomization between another therapy. It’s tough. But it’s clear that in US it’s possible and in Europe we have the opportunity to use — for example, in France, with some program, but the labeling is not there but we have access to the drug with a very particular way towards the drug but it’s clear that for HER2 it’s a very interesting compound with very good results in this population. If the immunochemistry is 3+, a little bit less. It’s 2+ positive. And so important thing is it works, if patient received previously trastuzumab, and it’s really important, with time, the toxicity, the management of the pneumonitis is better and it’s relatively important because it is a really problematic toxicity of this drug.

DR LOVE: So, I’m kind of curious from indirect comparisons — I know that’s something that’s difficult to do but people do it anyhow. I’m curious at this point how you would compare the efficacy from what we’ve seen indirectly of tucatinib/trastuzumab versus T-DXd?

PROF ANDRÉ: Difficult to have this kind of comparison. And it’s pretty, yeah, it seems the same kind of efficacy. Three, 2, 2 possibility with response rate. It’s not very different, very, Neil. PFS it’s a little bit better with tucatinib and trastuzumab but it’s difficult to compare because it’s not the same population.

DR LOVE: The other question I had is, I was kind of surprised that tucatinib only had a 3% objective response rate. What’s the objective response rate to trastuzumab alone?

PROF ANDRÉ: We don’t have response of this question in previous trial. We have only the response with the combi of trastu with lapatinib. I don’t know. But I think tucatinib, 3% of response is low. But it’s some disease control. And it’s not no efficacy. It’s some efficacy increased by the combination of trastuzumab but if we go back to the slide concerning the tucatinib alone, yeah, it’s low but it’s some control of the disease with tuca alone. Some efficacy. The results are not on this slide but I don’t remember exactly, but it’s some efficacy. It’s not no efficacy of tucatinib. And the PFS is, I don’t remember. It’s probably around 3 months, and 3 months it’s not nothing in this kind of, in this situation.

Okay, the next presentation is another new target. It’s a KRAS G12 mutation. It’s, like HER2, approximately 3% of colorectal metastatic disease and we have now sotorasib. It’s an inhibitor of the KRAS G12C mutation. And we have published in New England the study comparing sotorasib plus panitumumab. It’s 2 different doses of sotorasib, high dose that is a lower dose compared to the standard of care. It’s for patients refractory to standard of care, not all standard of care but the chemo, fluoropyrimidines, irinotecan, and oxaliplatin. And results show that the high dose of sotorasib have a better response rate compared to low dose and the response rate is 26% with the schedule of sotorasib 960 mg with pani and this is really interesting data. There is a response rate better than the standard of care. And if we look at the curve of PFS, the curve of PFS with the combi of pani plus sotorasib is better than the curve of investigator choice.

And the hazard ratio is pretty good, 0.49 for the higher dosed and with relatively durable therapy. So skin toxicity is really — the myelotoxicity of the combi in relationship with pani, but it’s really a new therapy, a new combination for patients with the G12C mutation.

We have other anti-G12C compounds. It was combined with cetuximab adagrasib. It’s the same kind of results, but it’s not the same comparison. It’s proved that it’s better to combine the inhibitor with anti-EGFR compared to the compound alone. The response rate is better with the combi compared to the mono. Better that we see that were produced with the compound, but it’s difficult to compare, to have indirect comparison. So PFS is pretty good at 6.9 with the combi but it’s also pretty good with the monotherapy but it’s a very little number of patients. For this reason, the OS is difficult to interpret. It’s lower in the combi compared to the mono but the confidence interval is very large. And for this reason, probably too early and too little population to conclude. The percentage of Grade 3/4 is low, 16%, and toxicity, it’s some skin toxicity without Grade 5.

It is now time to conclude and it’s really, like in all primary, we try to follow the lung with the fact now it’s a lot of targeted therapy in that cancer. In colon cancer, each year, these figures change and a new compound and it’s really 20 — the year 2023 was with a lot of new compounds. And it’s clear that now to treat a metastatic colorectal cancer, we need NGS, we need immunochemistry for HER2, we need immunochemistry for MSI, and we have a lot of possibility. It’s clear that MSI is a little bit particularity because it’s really for the patients, a hope of probably cure with metastatic disease because we have a PFS now with pembro at 5 years at 32%. We have research with the combi of nivo/ipi with longer follow-up with 70% have control at 2 years. Really, it’s important to have MSI at the beginning of the disease because if the patient is MSI, it’s clear that we need immuno-oncology. Careful with the misdiagnosed. Because misdiagnosed, if you give immunotherapy to a patient with MSS, it won’t work. And it’s clear that it’s better to have immunochemistry and pCR or NGS to be sure that it’s not misdiagnosed. And in some cases, it’s rare, but if you have a normal immunochemistry, no loss of protein, it’s possible to have a little number of patients MSI-high. And for this reason, it’s better to have the best technique, immunochemistry, and pCR — or to have immunochemistry and pCR or NGS for the diagnosis of MSI tumor.

In first line, it’s good progress. In advanced line, we have now a lot of other compounds, encora and cetux in BRAF V600E. We have in first line, panitumumab for patients with wild type disease. I think it’s very important. It’s a publication of the Japanese team. It’s clear that if the patient is RAS wild type, necessary to have anti-EGFR in first line because overall survival is improved. And we have HER2 with two compounds, tuca/trastu or trastu/deruxtecan for patients with disease refractory to chemo. We have also very rare the fusion of NTRK. It’s important because it’s some MSI. And if we have a patient with progressive disease with immuno-oncology and MSI, please test fusion of NTRK because it’s some there. And if it’s the case, we have inhibitor of NTRK, larotrectinib. And we have a paper in Lancet Oncol.

And we have now also the sotorasib compound. We have 2 compounds, adagrasib and sotorasib for the 3% of patients with this mutation. It’s really the facts we need. To have NGS we need to have immunochemistry to choose a good therapy for the patients who have — really the good therapy for the patients depending on the target.

Focus on MSI because I think it’s one of the really huge progresses in this disease to remember that 5% of metastatic patients are MSI and are dMMR with amazing results with immune checkpoint inhibitor, pembro or nivo plus ipilimumab. It is 2 options in first line we have now to — yeah, the question is, if we give nivo/ipi for the patients in this situation or if it’s some for the mono and some for the combi? It’s a question. Probably, we need to have the results of the second part of the 8HW to answer this question. We have also to probably factor to decide between mono and to the combi and it’s a lot of work to try to answer this question. And it’s a lot of questions about local disease. If it’s no metastases disease we have more cases because it’s 10% to 15% of nonmetastatic colorectal cancer, 2% of nonmetastatic colorectal cancer with amazing results with dostarlimab with the opportunity to avoid chemo, to avoid radiotherapy, to avoid surgery for these patients. And we have the proof of concept for locally advanced colorectal cancer. And it’s clear we have to define a better strategy for resectable colorectal cancer, for resectable colon cancer. And it’s on a trial to try to answer this question because at this time, it’s not clear.

Thank you for your attention. It was a pleasure to be with you for this summary of the year.

Other Considerations in the Management of CRC — Arvind Dasari, MD, MS

DR DASARI: Hello. I am Arvind Dasari. I’m an Associate Professor in the Department of Gastrointestinal Medical Oncology at MD Anderson in Houston, Texas. I will be talking about the management of colorectal cancer in this Year in Review session.

We’ll first start in the neoadjuvant setting in the management of colon cancer. The Phase III NeoCol study compared neoadjuvant chemotherapy versus up-front surgery in patients with locally advanced colon cancer. This trial enrolled clinically Stage T3 with invasion depth of about 5 mm or more or T4, with or without lymph node involvement, and obviously no metastatic disease.

The trial enrolled close to 250 patients who were randomized 1:1 to up-front surgery versus neoadjuvant chemotherapy followed by surgery. The neoadjuvant chemotherapy was CAPOX for 3 cycles or FOLFOX for 4 cycles, and further adjuvant therapy in both arms was based on the pathological stage.

The primary endpoint of the study was disease-free survival, and there were several secondary endpoints.

First starting with the patients who were enrolled onto the trial and actually looking at the group that went to surgery immediately without neoadjuvant therapy. The reason I picked this group is to see whether the staging that is done by CT is accurate and how well it matches up with the pathological stage. And I pulled out some of the key numbers on the left bottom panel, where you can see that there are some discrepancies between the clinical staging versus the pathological staging. And similar concerns were also raised on the FOxTROT trial that also asked a similar question, so that’s something to keep in mind, that CT is not really accurate at determining T or N stage.

Looking at the primary endpoint of disease-free survival there was no improvement in disease-free survival with the addition of neoadjuvant chemotherapy. However, as to be expected, the patients who received neoadjuvant chemotherapy actually required less adjuvant chemotherapy, and there was also downstaging of the tumors with neoadjuvant chemotherapy.

So takeaways. Neoadjuvant chemotherapy does not really improve disease-free survival as compared to standard up-front surgery followed by adjuvant chemotherapy, we do need to be cautious about the risk of overtreating these patients, and CT scans may not be very accurate in determining the actual stage. This is a study that was designed many years ago, so these patients did not have MSI testing or ctDNA testing done. So those are some of the limitations of this study.

So what is the takeaway from this study? I think this and the FOxTROT study show that neoadjuvant chemotherapy is safe and can be done when needed, and which would these patients be? These would be patients with very locally advanced disease, T4, T4b lesions, to try and downstage their tumors to make resection easier. But there’s no clear benefit to really recommend to doing this on a routine basis in most patients that we see every day.

Keeping with the theme of neoadjuvant therapy and looking at neoadjuvant chemotherapy, but moving a little lower down looking at rectal cancer patients, the PROSPECT trial evaluated preoperative chemotherapy with selective chemoradiation in patients who did not have adequate response to preoperative chemotherapy versus the standard of care, that is chemoradiation, for all these patients.

The key inclusion and exclusion criteria are listed below. So these patients had fairly early stage mid to upper rectal tumors, so these patients were candidates for sphincter-sparing surgery, they should not have required an APR, and they should have had a T2N+ or T3N- or + tumors. However, the lymph node involvement had to be limited, with less than or equal to 4 pelvic lymph nodes and less than or equal to 1 cm in short axis.

And these patients were randomized to the standard of care of chemoradiation followed by surgery and adjuvant chemotherapy versus up-front chemotherapy with FOLFOX for 6 cycles and then restaging. And adequate response was defined as at least 20% improvement in the tumor, and those patients who had at least 20% improvement went to surgery skipping chemoradiation, and those who did not had pelvic chemoradiation prior to surgery.

Looking at the patient characteristics who were enrolled onto this trial, like I mentioned, the tumors were mid to upper rectum, so the median distance from the anal verge was 8 cm for these patients. The vast majority of patients had T3 disease, and about 40% or so were node-negative.

This trial was — as I mentioned, looked at the selective use of chemoradiation in patients who did not have adequate response with neoadjuvant chemotherapy. And in fact, only 9% of the patients who received chemotherapy actually received chemoradiation, suggesting that it’s pretty effective to do neoadjuvant chemotherapy. And the trial was with noninferiority design and showed that the selective use of chemoradiation preceded by chemotherapy was noninferior to up-front chemoradiation in everybody.

The disease-free survival rate at 5 years was comparable between the 2 arms at around 81% versus 79%.

Furthermore, looking at other key outcomes, such as the risk of local recurrence, it was very, very low in both arms. As you can see, the local recurrence-free rate was over 98%. Overall survival was also very high and comparable between the 2 arms.

What was really intriguing in this study was the path CR rates even with just chemotherapy was around 22% and with chemoradiation was 24%, so again very comparable between the 2 arms.

So in summary, what this study showed and validated was that in patients with mid and upper rectal tumors with cT2/3 and without extensive lymph node involvement up-front FOLFOX may be done to omit radiation safely and effectively.

And this trial actually validated, as I mentioned, what we’ve already been doing for a few years, so it’s very reassuring to see that.

So takeaway. I just wanted to summarize very quickly the current approach to locally advanced rectal adenocarcinoma, taking into account not just this study but also a few other studies that we’ve heard of over the last 2 to 3 years. Every patient with locally advanced rectal cancer should be accurately staged, and in most patients in the US that would be with a pelvic MRI, although an EUS may also be used, and these patients should also get microsatellite testing done. In patients who are MSI high it’s very reasonable to use immunotherapy, and in fact that’s in NCCN Guidelines currently.

In patients who have upper rectal tumors with very early-stage disease, say T2 or maybe early T3, without any lymph node involvement, perhaps you could think about up-front surgery and adjuvant chemotherapy as needed, or these patients could also be treated with chemoradiation.

In patients who have upper or mid tumors who have slightly more advanced disease, say very early node-positive disease, and that is also T2 to T3, you could think about the PROSPECT approach that we’ve just talked about with chemotherapy and selective chemoradiation.

In other patients with very — more advanced disease, such as T4 tumors, N2 disease, or other high-risk features such as EMVI, lateral pelvic node involvement, or circumferential radial margin involvement, we could think about total neoadjuvant therapy for these patients, where we deliver both radiation as well as chemotherapy up front.

And then patients, especially those with low-lying tumors who may require an APR, if we have a complete clinical response with this approach, we could think about watch and wait and omit surgery.

Moving on to the use of circulating tumor DNA in patients with resectable colorectal cancer, we had another update from the CIRCULATE-Japan group looking at ctDNA dynamics in colorectal cancer patients.

Before I go over the results I wanted to refresh your memory about what the CIRCULATE-Japan study was. So this is actually a group of 3 different trials. The study that’s called the GALAXY study is the screening part of the study, where patients with Stage II through IV resectable colorectal cancer are enrolled at the time of diagnosis, and they are monitored with ctDNA. And the assay that’s used is a tumor-informed assay, Signatera. And they get this assay done at the time of diagnosis and serially through any therapies that they may get in the neoadjuvant, post-surgical adjuvant setting, and through surveillance.

And patients who are ctDNA positive on the GALAXY study, some of them are enrolled onto interventional trials, including the VEGA trial that is trying to ask the question whether we could safely de-escalate the adjuvant chemotherapy in patients who are ctDNA negative, and the ALTAIR study that is asking the question if we should escalate therapy in patients who are ctDNA-positive after surgery.

These studies are still pending results, but the updates that we’ve had over the last year to 2 years are from the observational GALAXY cohort who have not been enrolled onto either of these interventional trials.

So the latest update we had at GI ASCO earlier this year included close to 3,000 patients, and they updated some of their prior data with this larger data set and also with a longer median follow-up.

And some of these findings that were shown were actually — kind of confirmed what we already know from the GALAXY presentations, as well as other studies. For instance, they showed that patients who are ctDNA positive after surgery had much worse outcomes as measured by disease-free survival, with a hazard ratio of 12.

And they also showed that patients who are ctDNA positive when they get adjuvant chemotherapy a proportion of these patients actually clear their ctDNA. And this was shown in prior presentations, but what they showed of significance this time is that in these patients who cleared their ctDNA not everybody had a sustained clearance. Some patients appeared to have cleared their ctDNA for a period, and then they turned ctDNA-positive.

So when you look at the outcomes of these patients, and you compare them to the patients who never cleared their ctDNA, what we see is that the median disease-free survival is very, very good for patients who’ve had a sustained ctDNA clearance. That’s the blue line that you see. And the patients who had transient clearance of their ctDNA, so their disease-free survival did improve as compared to those without clearance. For instance in this cohort the median was 9 months in the transient clearance cohort versus 3 and a half months in those with no clearance.

While this seems somewhat promising, what is really critical to look at is when you look at the 24-month disease-free survival. So how are they doing after a longer follow up? And what you see is that in the sustained clearance outcomes are great, 24-month disease-free survival is over 90%, but in the transient clearance and the no clearance group it’s actually only 2% to 3% of the patients who have disease-free survival. So clearly it’s only the sustained clearance group that are doing well with longer follow-up.

And they also provided some additional data about the dynamics as to when these patients turn positive with regards to their ctDNA after transiently clearing with adjuvant chemotherapy. And what they showed was that virtually everybody had a recurrence in terms of their ctDNA by 18 months post-surgery.

The other piece of data that they showed and confirmed from prior presentations is the fact that in patients who’ve say finished their adjuvant chemotherapy and are undergoing surveillance, during surveillance if they turn ctDNA-positive, similar to what we’ve seen before, their outcomes are also pretty dismal with regards to disease-free survival, with a hazard ratio of 42, corresponding to a 24-month disease-free survival rate of over 90% in the patients who are negative versus about 6.5% in patients who are ctDNA positive during surveillance.

The BESPOKE study was a study conducted in the US, and also an observational cohort, also used Signatera, and with a similar overarching design where patients with resectable colorectal cancer would be enrolled and would have serial evaluation for ctDNA from the time of diagnosis through any adjuvant chemotherapy and during surveillance.

And the results also were similar to what we noted in the GALAXY trial in that the ctDNA status after surgery was very strongly prognostic, and patients who were ctDNA positive had a much higher risk of recurrence. What this study did also provide was that it gave us some benchmarks as to the proportion of patients who may be ctDNA positive after surgery according to the TNM stage. In Stage II patients it’s about 5% to 6%. In Stage III patients it’s about 20%.

And what they also showed was that there was a proportion of patients in whom ctDNA cleared with adjuvant chemotherapy who were initially ctDNA-positive, but similar to the GALAXY study it is critical to see whether the ctDNA clearance is sustained versus transient because the ones that are — have transient clearance have a recurrence of their ctDNA, as well as their disease, at around 15 to 24 months. And also similar to the GALAXY study, even during surveillance patients who are ctDNA positive have worse outcomes with a much, much higher risk of recurrence overall.

A novel piece of information that the BESPOKE trial provided was also the patients perspective. I mean we’re ordering these tests, but how do the patients feel about these tests, and what are their thoughts? Overall there seems to be good engagement and acceptance from patients with regards to these tests, with improvement in their anxiety about cancer recurrence, empowering them in trying to make the right treatment decisions, leading to kind of more acceptance and wanting to do these tests in the future.

So the key takeaways from both these studies, to reiterate what we’ve already discussed, ctDNA positivity has very strong negative prognostic effects in the postoperative and surveillance settings, and in patients who are ctDNA-positive, it’s the sustained ctDNA clearance cohort that does well, with over 90% disease-free survival at 2 years. Patients who have transient ctDNA clearance have maybe a slight improvement in the median DFS, but long-term outcomes continue to be somewhat limited.

So I think the main takeaway is all of these observation studies have shown that ctDNA is a very strong prognostic marker, but we really need some interventional trials to establish clinical utility in terms of how to use these tests, how can we better manage these patients who are ctDNA-negative or -positive. And that’s the stand that the NCCN Guidelines also take, where they say that ctDNA is emerging as a prognostic marker. However, there is currently insufficient evidence to recommend routine use of ctDNA assays in this setting outside of a clinical trial, and participation in clinical trials is encouraged.

DR LOVE: Do you not use circulating DNA in your own practice outside a trial or no?

DR DASARI: Yeah. It is being ordered because the test is covered by CMS, and so insurance companies pay for it. But we really don’t know how to use the data from these tests, right? So if you have a Stage III colon cancer patient who’s ctDNA negative, are we comfortable today not giving that patient adjuvant chemotherapy? And similarly, in a patient who say is ctDNA positive during surveillance after they’ve finished adjuvant chemotherapy, and they turn positive, sure they have a high risk of recurrence, but what do we do? I mean how do we manage that patient? We don’t know. And that’s what the NCCN is trying to say.

DR LOVE: Well, what about node-negative?

DR DASARI: So I think the DYNAMIC trial suggested that Stage II low-risk patients who are ctDNA negative, perhaps in those patients we could think about safely deescalating therapy. But the NCCN Guidelines, to your point, has not put that in there yet to use.

DR LOVE: That’s really interesting. I’m surprised I didn’t know that. But I see so many people doing it, it’s hard to believe.

DR DASARI: Indeed. I think that’s kind of the disconnect between having an assay that’s available, but the data are still catching up in terms of how to use it.

DR LOVE: I think the 1 situation, though, I would wonder about would be, again, Stage II with a negative ctDNA. I think a lot of people could be on the fence, and I don’t see why you wouldn’t apply it. I mean you want more data?

DR DASARI: So low-risk Stage II I think is 1 situation where I’ll have a discussion with the patient and say look, it’s low risk, and if ctDNA comes back negative I think it’s reasonable to omit chemotherapy. So that’s my practice, yes.

And another setting would be where say if a patient is at high risk of toxicities from adjuvant chemotherapy then maybe try to get the ctDNA test to help determine the decisions.

DR LOVE: That’s very interesting.

DR DASARI: And I just wanted to highlight an ongoing clinical trial here in North America, the CIRCULATE-North America, which is enrolling patients with Stage II high-risk patients or Stage III patients with colon cancer who will be evaluated with the Signatera assay, and they would be put in 1 of 2 cohorts. If they’re ctDNA negative the cohort is asking the question of whether we can safely de-escalate adjuvant chemotherapy in these patients, randomizing them to immediate chemotherapy versus serial surveillance with ctDNA and starting adjuvant chemotherapy if and when they turn positive. If they are ctDNA positive the question being asked is can we escalate over the current standard of care and do a more intensive adjuvant chemotherapy to improve these — improve outcomes of these patients, who we just saw have in a good proportion of patients, only have transient clearance and subsequent recurrence.

Moving on to the INTERCEPT trial. This is a study that came out of MD Anderson that looked at this question of clinical utility of these ctDNA assays. And this observational cohort enrolled patients who had Stage II through IV resectable colorectal cancer and similar to the other studies that we’ve seen also used Signatera, and this was obtained serially from the time of diagnosis, postoperatively, during and after adjuvant chemotherapy, and during surveillance up until the time of recurrence.

And if and when the patients turn positive the patients would have evaluation to see if they had radiographic evidence of recurrence. And those who did not have evidence of recurrence an attempt was made to enroll them onto interventional clinical trials geared towards these patients who are ctDNA positive but without any radiographic evidence.

In the presentation at ASCO 2023 around 1,100 patients’ data were used, and kind of a key take home message was that of the patients who were ctDNA positive during surveillance about 50% of them actually had evidence of radiographic recurrence when you do scans in these patients. And these scans could have been kind of already scans that were ordered as part of surveillance or were triggered by the positive ctDNA assay. So this kind of suggests that perhaps the current generation of assays are probably not as sensitive as we would like them to be. Maybe they’re detecting these recurrences a little too late. So it’s only about 50% of these patients are truly MRD positive, that is ctDNA positive but radiographically negative. And these patients who are MRD positive could be enrolled onto ongoing clinical trials, and about 60% of these patients were enrolled onto ongoing clinical trials to evaluate novel therapies in this space.

So conclusions and takeaways. When ctDNA-positive, I think it’s important to do scans to look for any overt metastatic disease. What needs to be evaluated is if we keep doing these assays serially are we able to pick up patients who if they have radiographically overt disease would they at least have oligometastatic disease that we can kind of treat and hopefully cure with locoregional therapies. And in patients who are truly MRD positive, because we don’t know what to do and how to treat these patients, these patients should be encouraged to enroll onto clinical trials. And that’s a huge, huge unmet need, and hopefully in the near future we’ll have more sensitive assays and also data on how to manage these patients.

Moving on to the metastatic setting and starting with the PARADIGM trial, this trial asked the question of in patients with RAS wild-type metastatic colorectal cancer who are receiving first-line chemotherapy, here it was FOLFOX, would panitumumab or bevacizumab be the better choice as the biologic agent. This was a Phase III study conducted in Japan, and as I said enrolled patients with RAS wild-type metastatic colorectal cancer undergoing first-line chemotherapy, and patients were randomized to panitumumab plus FOLFOX or bevacizumab plus FOLFOX.

After the study was initiated data around the predictive implications of sidedness came out and that right-sided tumors would not respond to EGFR antibodies, and so the study was amended only to have left-sided tumors subsequently after that.

So when you look at the overall study population about 75% were left sided and about 25% were right sided. And the statistical design was hierarchical, where they looked at the primary endpoint of overall survival in the left-sided population first, and if that was positive then they would look at it in the entire study population that included the right-sided tumors as well.

So with regards to the outcomes, overall survival did improve with the addition of panitumumab in first-line therapy as compared to bevacizumab, with stratified hazard ratio of 0.82 in the left-sided tumors and with a stratified hazard ratio of 0.84 in the overall patient population.

Furthermore, response rates also seem to have improved, from about 68% to about 80% in the left-sided patient population and from about 67% to around 75% in the overall patient population. And this translated to improved R0 resection rates in these patients with the addition of panitumumab from about 10% to 11% to around 17% to 18%. And as shown by multiple other prior studies there was really no benefit in the right-sided patient population with the addition of panitumumab.

They then followed up this large Phase III study with a well thought out biomarker study where they evaluated ctDNA at baseline from these patients, and they had ctDNA from the majority of patients, in the ctDNA marker study they had a little over 730 patients, and they looked at the alterations in ctDNA in these patients. Remember that in the original study design they only looked at RAS wild type, so they did not go beyond RAS to look at other alterations.

So what they found with this additional analysis was that overall about 10% of the patients had BRAF V600E mutation, and as to be expected the majority of these were from the right side. In addition, about 6% to 8% had additional RAS mutations, and these were in KRAS, as well as in NRAS, and this perhaps reflects some tumor heterogeneity that is now picked up on the tumor biopsies, and about 3% were MSI high.

And the subsequent analysis showed that these prior improvements that we saw in terms of overall survival was limited to the patients who were microsatellite stable and RAS and BRAF wild type, both in the left-sided tumors, as well as in the overall study population. So addition of panitumumab did not really help in the other cohorts. And similarly the overall response rate, as well as R0 resection improvements were limited to these patients who were microsatellite stable and RAS/BRAF wild type. And even after adjusting for all these factors there’s no benefit for right-sided tumors.

So conclusions and takeaways from the PARADIGM study. First-line therapy with EGFR monoclonal antibodies in metastatic colorectal cancer patients with left-sided RAS/RAF wild-type tumors is associated with improved response rate, R0 resections, as well as overall survival.

In my practice, and this is in the guidelines as well, in addition to looking at RAS and RAF wild type and also sidedness, it is also important to make sure that we are ensuring that the patients do not have HER2/neu-amplified tumors because we know that this is a mechanism of resistance as well. So in patients who are left sided and fulfil all these criteria I do have a discussion with my patients about starting with an EGFR antibody over bevacizumab.

DR LOVE: Can you try to track the biology of how HER2 would interact with this EGFR pathway?

DR DASARI: So the HER2 pathway when amplified I think causes resistance to the EGFR antibodies by really bypassing the EGFR receptor activity.

DR LOVE: That’s interesting.

DR DASARI: And so one of the reasons to really consider using EGFR antibodies in front line is because increasingly we’re accumulating data that this resistance to EGFR therapies is reversed over time. So if you start with a tumor that has sensitivity to EGFR antibodies, that’s the panel on the top left, and you treat them with EGFR-containing regimens, over time they may develop resistance, and this could be because there are the development of clones that have genomic alterations that portend resistance to these antibodies. So at that time, when they’ve developed resistance, if you take them off of the EGFR-containing therapies and put them on other therapies, that is in a sense you’re giving them an anti-EGFR treatment holiday, the EGFR antibody sensitivity clones repopulate, and so you could go back and rechallenge these patients with these therapies later.

And there has been work that was done that looked at both the types of alterations that happen, and what was shown was that these alterations that happened are mostly in the RAS/RAF/MAP kinase pathway. And what was noted was that there tends to be multiple alterations happening at the same time, so it’s not limited to just say HER2/neu amplification or RAS mutation. It could be 3 or 4 different alterations. In fact, the median number of alterations were 4 in a study that looked at this. However, what is important to know is that there is an exponential decay of these alterations over time with a half-life of around 4 to 4 1/2 months after discontinuation of EGFR therapy.

So there were several trials, Phase II, small sized, that really looked at this question of rechallenge with EGFR antibodies that were reported over the past few years. And the study that was presented at ESMO last year did a pooled analysis of 4 of these studies with a total of about 114 patients, and they confirmed what was noted individually in these patients and that there is some response, with a response rate of around 17 to 18%. And remember, these are truly refractory patients who progressed through pretty much everything, so response rate of 17% is reasonably significant, with a medial progression-free survival of 4 months and overall survival of about 13 months or so.

So takeaway. EGFR rechallenge is a very reasonable option but after checking ctDNA for resolution of any resistance mechanisms that these tumors may have developed with prior therapy.

There was a study that looked at kind of acquired alterations on first-line chemotherapy with EGFR antibody, that is cetuximab, in advanced colorectal cancer. So this would be somewhat similar to the PARADIGM study, and the authors used the ctDNA samples from the CALGB/SWOG 80405 study.

To refresh your memory, the SWOG 80405 study had patients who had metastatic colorectal cancer, first line, who were KRAS wild type, at the time of doing the study the implications of extended RAS were not known, and patients received first-line chemotherapy. The chemotherapy backbone was per treating physician’s choice, and patients were then randomized to receive either cetuximab or bevacizumab. A little over 1,100 patients were randomized with a primary endpoint of overall survival.

So in the current study what the authors did was that they identified about 130 patients who had ctDNA samples and at baseline had extended RAS wild type and also did not have other alterations that may predict resistance to EGFR antibodies. And then they looked at kind of serial samples with ctDNA and looked at how the ctDNA profile changed with therapy, both in the cetuximab arm, as well as the bevacizumab arm. And so they looked at the alterations for each mutation, for instance KRAS, NRAS, BRAF, and so on. They also looked at amplifications as well.

And then they kind of did a pooled analysis of all these kind of alterations, which is what I marked with the blue arrows there, so the cetuximab arm and the bevacizumab arm. So these were the number of — the proportion of patients who had alterations on serial sampling. And they compared that to historical data from later lines of therapy as to how alterations occur with EGFR therapy in later lines.

What they found was that to their surprise in these first-line patients actually these tumors were not developing a whole lot of alterations with therapy with cetuximab. In fact, it appeared that perhaps there was a slight trend towards the patients who are receiving bevacizumab receiving — having more alterations with therapy in the MAP kinase pathway, which is surprising.

This is very much in discordance with what we noted previously and also as we can see in this figure where the patients on later lines of therapy with EGFR antibodies have a much higher chance of developing resistance alterations. This is very much a hypothesis-generating study in that they only had 130 patients but nevertheless raises the intriguing possibility that resistance to anti-EGFR therapies is different, the underlying mechanisms, in earlier lines of therapy versus later lines of therapy. Why that is, how that happens, we don’t know, but still requires more validation and further investigation.

Moving to the refractory setting, the SUNLIGHT trial was a Phase III study TAS-102 with or without bevacizumab in the third-line setting in patients with metastatic colorectal cancer. In this global study that was open label patients with metastatic colorectal cancer who had received at least 2 prior regimens with progression were randomized to TAS-102 alone or TAS-102 in combination with bevacizumab. The primary endpoint was overall survival. And about 490 patients were enrolled onto this trial.

Just a key pertinent criteria with regards to the patients who were enrolled. About 75% of the patients received prior anti-VEGF antibody, and about only 25% actually received anti-VEGF antibody in both of the prior lines.

This study did meet its primary endpoint with improvement in overall survival with median overall survival improving from about 7 1/2 months to around 10.8 months, corresponding to hazard ratio of 0.6. Similarly, there was improvement in median progression-free survival, as well, with a hazard ratio of 0.44.

With regards to the adverse effects, these were to be expected, keeping in line with the known side effect profile of TAS-102 that tends to cause myelosuppression and bevacizumab that tends to cause vascular adverse effects, especially hypertension.

One of the key secondary endpoints was looking at time to worsening of performance status. That was defined as performance status of 2 or more. The eligibility criteria was limited to those with 0 and 1. And what they showed was that the time to worsening of ECOG PS improved with TAS-102 plus bevacizumab from about 6 months to 9 months, with a hazard ratio of 0.54.

They then looked at the subset of patients who had their PS maintained, so these were patients who did not have worsening of their PS, and these were about 90% of the patients who at discontinuation still continued to have a PS of 0 to 1. And what they showed was that the drug was very effective in these patients, as well, with both overall survival and progression-free survival being better in these patients.

So what does this mean clinically? I think what this tells us is that these patients who have a good PS, they could go on to get other therapies thereby improving their overall survival.

So takeaways. TAS-102 plus bevacizumab significantly improved overall survival and progression-free survival compared to TAS-102 alone in the third line. Toxicities were consistent with prior experience with these agents. And the key takeaway, I’d say, at this point, based on these data, TAS-102 plus bevacizumab should be chosen over TAS-102 unless there are strong contraindications for anti-VEGF therapy, and in fact has since been approved by the FDA for third-line setting.

Moving on to the FRESCO-2 trial. So this was a trial in the refractory setting that evaluated fruquintinib. Fruquintinib is an oral tyrosine kinase inhibitor of VEGF receptors 1, 2, and 3. And eligible patients for this trial had received all available therapies, and this included cytotoxic chemotherapy, an anti-VEGF antibody, and as appropriate EGFR antibodies, as well as immunotherapy or BRAF-directed therapies. In addition to all this they should have also received prior TAS-102 and/or regorafenib. Close to 700 patients were randomized 2:1 to fruquintinib versus placebo with a primary endpoint of overall survival.

Key kind of factors with regards to the patients who were enrolled. They received a median or 4 prior lines of therapy, and almost everybody received a prior anti-VEGF antibody, and prior TAS-102 in about 52%, both agents in about 40%, and prior regorafenib in about 10%.

The study did meet its primary endpoint with improvement in overall survival from about 4.8 months to 7.4 months, with a hazard ratio of around 0.86 and improvement in progression-free survival from about 1.8 months to 3.7 months with a hazard ratio of about 0.32.

With regards to the toxicities, these were consistent with the anti-VEGF activity of this drug, with hypertension, asthenia being the most common side effect. The typical side effects that we tend to associate with regorafenib, such as elevated LFTs, hand-foot syndrome, were not very common with this agent.

And there was a subsequent subgroup analysis that was done that looked at the efficacy and activity of this drug based on the number of prior lines of therapy. As I mentioned, these patients were very heavily pretreated, and so the subgroup analysis looked at patients who were less heavily pretreated, that is less than — 4 or fewer lines of therapy versus those who had received more than 4 lines of therapy. And overall the efficacy was consistent in both these cohorts and improved both overall survival, as well as progression-free survival, irrespective of prior number of lines of therapy or type of treatment.

So in conclusion, fruquintinib did improve overall survival and progression-free survival in patients with refractory metastatic colorectal cancer and toxicities consistent with prior experience with this agent as well as other anti-VEGF therapies, and overall appears to be well tolerated. The FDA did approve fruquintinib both for third line, as well as fourth line. That was based on this data, as well as another trial that was done in China that was conducted in the third line. So the FDA took into account both trials and approved it for both third- and fourth-line setting and should be considered as an option for eligible patients.