Tanios Bekaii-Saab, MD, outlines clinical trials that have defined the current landscape of metastatic colorectal cancer and what research is underway.
Tanios Bekaii-Saab, MD
Dosing strategies and novel treatment regimens are 2 of many areas being explored in the colorectal cancer (CRC) eld, while paying close attention to molecular subgroups, explained Tanios Bekaii-Saab, MD.
For example, Bekaii-Saab said that phase II results of the ReDOS dose-optimization study demonstrate that weekly dose escalation of 80/120/160 mg of regorafenib (Stivarga) is the favored treatment for patients with metastatic CRC (mCRC). Additionally, data need confirmation on moving the agent earlier in treatment, which may provide more benefit to patients, he said.
Moreover, the superiority of receiving regorafenib prior to cetuximab (Erbitux) versus cetuximab prior to regorafenib was shown in findings from the REVERCE study in patients with mCRC, he explained.
“They certainly need to be reproduced in a Western-patient population, but it was intriguing to see the same trends between CONCUR and REVERCE that moving regorafenib a little bit earlier may make a difference,” said Bekaii-Saab, a medical oncologist at Mayo Clinic.
Immunotherapy continues to be explored beyond the microsatellite instability—high (MSI-H) population, as well, though has proven to be more challenging to implement.
In an interview during the 2018 OncLive® State of the Science Summit,™ Bekaii-Saab, who chaired the meeting, outlined clinical trials that have defined the current landscape of mCRC and what research is underway.Bekaii-Saab: These patients go through first and second- line therapies, and sometimes through third-line therapy. The question becomes, “What do we do for patients beyond second-line therapy?” Previously, we had very few options, if any, for these patients. Now we have 2 oral options with regorafenib and TAS-102 (Lonsurf). These agents are different from each other. Regorafenib is a multikinase inhibitor; it’s a multitargeted agent. TAS-102 is a more traditional cytotoxic agent; it belongs to the family of fluoropyrimidines. It’s not 5-FU [5-fluorouracil] or capecitabine; it belongs to that family that seems to work when the others fail.
The knowledge we have about these 2 agents comes from separate studies that have never been compared in a head-to-head study. What we know about their activity primarily came from their comparison to best supportive care. [The agents] follow very similar pathways.
The RECOURSE study led to the approval of TAS-102, which looked at TAS-102 versus placebo. The study suggested a survival benefit; there was some progression-free survival (PFS) benefit. Toxicities with this agent are primarily gastrointestinal (GI) and hematologic events. We do see some early neutropenia in the first 4 weeks and some febrile neutropenia. About 4% of patients may need to be hospitalized.
On the other hand, regorafenib was approved based on the CORRECT trial [results], which randomized patients to regorafenib versus best supportive care and placebo. This study also showed a PFS benefit. The toxicities are a little bit different given the mechanism of action. We see a little bit more hand-foot syndrome in approximately 15% of patients; that can be quite significant. That’s something we don’t see with TAS-102. You also see some fatigue, very low nausea, diarrhea, and hypertension because of the VEGF activity.
In terms of toxicities, the differences between the 2 agents are primarily hand-foot syndrome, hypertension, and neutropenia. Fatigue is about the same. The GI toxicities are a little bit worse with TAS-102. The overall outcome seems to be very similar across the 2 studies. The next question is how to sequence these agents. We don’t have any randomized trials, so we have to rely on practical experience as well as whatever we can infer from the studies that were done.
The RECOURSE study included about 20% of patients who were previously exposed to regorafenib. This exposure to regorafenib did not seem to change the outcome. We don’t have data with TAS-102 followed by regorafenib.
There are also the CONCUR and TARA studies. CONCUR looked at regorafenib and TARA examined the use of TAS-102. Both studies had very similar designs as RECOURSE and CORRECT, but in Asian-only populations. Those studies examined patients with less exposure to pre-biologics because of lack of accessibility. In CONCUR and CORRECT, the delta value was much wider for regorafenib versus no treatment. On the other hand, TARA showed that earlier use of TAS-102 did not seem to change the delta value as much as in RECOURSE. There could be a hint that moving regorafenib up or not waiting until the end of the line to administer the agent may be more beneficial to patients.
There’s also another hint from a study called REVERCE that was run out of Japan. That study randomized patients to receive regorafenib before cetuximab or cetuximab before regorafenib. Interestingly, the survival, which was the primary endpoint of the study, was superior when regorafenib was given before cetuximab. This was statistically significant, [although] it was a small randomized phase II study.
The biggest challenge with both oral medications is compliance. There are also issues with determining the right dose—perhaps a little bit less with TAS-102 than with regorafenib. With regorafenib, physicians used a lot of different schedules. Some would do every other week, others would start with 80 mg and stop at 120 mg. Others chose 120 mg. About 30% of physicians gave 160 mg and then de-escalated as needed.
We examined the optimal dose in a prospective randomized fashion in a phase II study called ReDOS. We designed the study so that patients were randomized to either arm A, which was dose escalation from 80 mg to 120 mg to 160 mg on a weekly basis, or arm B, which was the control arm of 160 mg. At the start of every week, patients in arm A received a dose escalation of 40 mg. Whatever dose the patient finished cycle 1 with would be the dose the patient started cycle 2 on. Patients in arm B received 160 mg every day for 21 out of 28 days and then were de-escalated as needed.
The primary endpoint of the study was a composite of efficacy and toxicity. How many patients could go through 2 cycles of chemotherapy and start cycle 3? Patients started cycle 3 as long as there were no major toxicities that prevented them from continuing and their CT scan didn’t show progression.
The study was powered to see about a 15% difference between arm A and arm B. We had close to a 20% difference between arms. It resulted in a positive study with a statistically significant P value, favoring the dose-escalation schema in arm A from 80 mg to 120 mg to 160 mg. Interestingly, survival was numerically superior with arm A. Patients who went from 80 mg to 120 mg to 160 mg saw a survival close to 9.5 months. Patients who received 160 mg of regorafenib saw a survival of 5.9 months, which is pretty much what you see with the CORRECT study.
Arm B looked very similar to what you would expect from 160 mg [of regorafenib]. It was very interesting and surprising to see the superiority in the dose-escalation arm, though the P value was not statistically significant. That said, this was not the primary endpoint. The PFS was also a little bit better with the dose escalation. Progressive disease was the main reason why patients in arm A versus arm B did not continue through cycle 3 and beyond.
More patients progressed on arm B than arm A—almost double the number—which explains why the study was positive. The dose-escalation strategy seems to provide more control of disease and longer survival. The quality of life is also non-compromised in the dose-escalation strategy versus the full dose.
The NCCN guidelines took over the results of the study, and included the dose-escalation schema as an option for patients with refractory mCRC. In my viewpoint, this is now the favored [treatment] based on the positive results. When a physician proceeds with regorafenib, they should proceed with the dose-escalation strategy given all the findings from ReDOS.
For TAS-102, there are no studies that have looked at dosage. It tends to be less problematic than regorafenib. In practice, there are a lot of different dosing strategies depending on the physician and patient. We know that oral medications tend to be challenging from a compliance standpoint; you have less control. When you give intravenous drugs, you’re controlling the situation. You need to see patients on regorafenib every week in the first cycle and then every 4 weeks after the first cycle. With TAS-102, you want to see patients every couple of weeks in the first 2 cycles and then every cycle after that.When we look at the immunotherapy landscape today, PD-1/ PD-L1 inhibitors have been limited to MSI-H tumors, and they have some good responses. The problem is that it is only 4% of patients, so what do we do for 96% of patients who did not seem to benefit from single-agent PD-1 inhibitors? Why this is the case? The difference in tumors that are MSI-H versus microsatellite stable is tumor mutational burden (TMB). MSI-H tumors have very high TMB. These tumors were hypermutated and very visible, or very immunogenic, and are more likely to respond to PD-1 inhibitors.
Microsatellite stable (MSS) tumors are like immune deserts.
You can’t see the tumor-infiltrating lymphocytes (TILs). The quest now is, “How do we increase the accessibility of these TILs to where they should be, so PD-1 inhibitors can do their job?” One way to do that that has been shown preclinically is through MEK inhibitors, such as cobimetinib. The initial study with cobimetinib plus atezolizumab focused primarily on MSS tumors. These patients had a response rate of a little bit more than 10%. Some patients had significant survival—not at the level of what you expect with PD-1 inhibitors, but certainly above and beyond what you expect for a PD-1 or MEK inhibitor alone. This could change the landscape if the study is positive.
There are other efforts underway in phase I/II trials looking at strategies to change the immune desert into an immune-infiltrated disease. Those [trials] primarily focus on the MAPK/ERK pathway, but there are also other hints that the Wnt/beta-catenin pathways could also be manipulated through certain agents to increase the lymphocytic infiltrate. This would make those tumors more likely to respond to PD-1 inhibitors. It will be exciting if we can expand the advantages of immune therapy into MSS CRC.If you asked me this question 2 or 3 years ago, I would say, “You really have to be very selective with who gets genomic sequencing.” Now, everyone should get MSI testing. MSI is a surrogate biomarker for TMB. All these platforms are reporting TMB with meaningful cutoffs. High and perhaps intermediate [cutoffs] respond to immune-therapy agents, whereas the low [cutoffs] do not. Even in the absence of MSI information, we’re able to find out TMB and whether patients should receive a PD-1 inhibitor or not.
There are also other elements in the genomic analysis that are very important and very useful for us. For KRAS and RAS, we argue for limited testing. We understand a little bit more that patients with BRAF mutations have a worse prognosis with FOLFOX or FOLFOXIRI. Bevacizumab (Avastin) may make more sense in the first-line setting.
Then, we have the HER2 amplification, which is found in about 3% to 4% of patients. These seem to be exquisitely responsive to HER2 inhibition. There are a number of studies that are underway looking at HER2 inhibition; they are accessible to patients. In other studies, we are exploring the low yield mutations that are linked to agents like FGFR inhibitors. We have a number of options for our patients with CRC and other solid tumors in the GI space that mandate genomic sequencing and understanding of the genomic landscape of the tumor when we meet the patient.