Regorafenib/Nivolumab Shows Modest Activity in MMR Proficient Advanced Refractory CRC

The combination of regorafenib (Stivarga) and nivolumab(Opdivo) yielded modest clinical activityin patients with refractory, mismatch repair (MMR) proficient, advanced colorectal cancer (CRC) based on early results from a phase 1/1b study, according to Richard D. Kim, MD, who added that future efforts should focus on the identification of biomarkers that can predict response to the combination.¹

In the phase 1 portion of the study, investigators examined regorafenib at 3 different dose levels—80 mg, 120 mg, or 160 mg of regorafenib—21-days-on, 7-days-off, plus nivolumab at 240 mg delivered intravenously (IV) once every 2 weeks. In the expanded cohort, nivolumab was given at 240 mg IV once every 2 weeks for 16 weeks and then 480 mg once every week for 4 weeks, along with the maximum-tolerated dose of regorafenib, which was 80 mg.

Results presented during the 2020 ESMO World Congress on Gastrointestinal Cancer showed that of 21 evaluable patients, the best overall response was stable disease, at 66.7% (n = 14). Only 1 partial response (PR) was reported and it was unconfirmed; no complete responses were observed. Additionally, the disease control rate was 71.4%. Six patients (28.6%) experienced disease progression. Moreover, at a median follow-up of 4.7 months, the median progression-free survival (PFS) was 4.3 months, while the median overall survival (OS) was 11 months; the 6-month OS rate was 71.8%, while the 12-month OS rate was 31.9%. The most common grade 3 or higher adverse effect (AE) in both cohorts was rash, at 14.3%.

These results differed from what was seen in the REGONIVO trial, which helped to form the rationale for the phase 1/1b trial, said Kim. In a Japanese patient population with gastric cancer (n = 25) and CRC (n = 25), the median PFS in the CRC cohort was 7.9 months (95% CI, 2.7-10.4) and OS was not reached (MR; 95% CI, 9.8–not reached).²

“Our data didn't quite duplicate what [was seen] in Japan,” said Kim. “It's a small patient population, so there could be many reasons [to explain why]. However, at this time, we have to be very cautious about using this combination, knowing that [the regimen comes with] AEs and may not benefit the patient.”

In an interview with OncLive, Kim, a medical oncologist in the Department of Gastrointestinal Oncology at Moffitt Cancer Center, as well as an assistant professor of oncology at the University of South Florida College of Medicine, further discussed the potential of regorafenib plus nivolumab in patients with advanced CRC and the need for biomarkers to predict response to the regimen.

OncLive: Could you describe the rationale behind using regorafenib and nivolumab in mismatch repair (MMR) proficient, advanced refractory CRC?

Kim: In advanced CRCs, especially in the refractory setting, we know there are 2 oncolytics available: trifluridine/tipiracil (TAS-102; Lonsurf) and regorafenib. We know that for patients with microsatellite instability–high (MSI-H) disease, there's the option of immunotherapy. However, those patients who are mismatch repair deficient (dMMR) or MSI-H represent only 5% of all patients [with CRC]. The other 95% of patients who are MMR proficient or microsatellite stable (MSS) are those with an unmet need; they need better [treatment] options.

The combination of regorafenib and nivolumab came about because we know that regorafenib is a TKI that can actually reduce tumor-associated macrophages by blocking the CSF-1 receptors. You can also block PD-L1 expression while maintaining MHC-1 expression. Therefore, the combination can potentially augment the intratumor immunity of checkpoint inhibitors, such as nivolumab. We know that nivolumab as a single agent will [not elicit a] response in patients who are MSS.

Based on this rationale, [investigators] from the National Cancer Center Hospital East presented data on combining these agents in patients with MSS CRC at last year’s ASCO; these data are now published in the Journal of Clinical Oncology (JCO). The results were provocative; the response rate was up to 30% in patients with MSS disease, and this response seems to be very durable. The dose they recommended [was based on] the toxicity patients experienced, especially the rash that required dose reduction. The recommended dose of regorafenib was 80 mg plus the standard dose of nivolumab.

Could you expand on the previous data reported with the combination?

The Japanese group was the first to examine the combination [in this population]. They had 2 cohorts: 1 was the gastric cohort and the other was the microsatellite refractory colon cancer cohort. In the gastric cohort, they saw very high response rates of 40%. Ours is the second study that's [examining] this combination, but there is a potential plan to perform a larger study.

Interestingly enough, at the 2020 ASCO Virtual Scientific Meeting, a very similarly designed study examined a combination of regorafenib plus avelumab (Bavencio), which is another checkpoint inhibitor, in a similar patient population. In that study, the results are very [comparable] to our study. However, they did find some correlation between T-cell infiltration and response rate. Those are the only clinical data that we have so far. Before we embark on the larger study, I would definitely like to see more of our data to make sure it pans out in terms of the response rate.

By combining [these agents] the expectation is that nivolumab, or the checkpoint inhibitor as a single agent, has a low response rate. A 20% to 30% response rate is reasonable to move to the next step, but a 5% to 10% response rate is very modest. We have to be very careful and try to select those certain patients who will benefit from the combination before we move on to the next step.

Could you speak to the study presented at the 2020 ESMO World Congress on Gastrointestinal Cancer?

Our study is very similar [to the one that was done in Japan]. The main difference is that our group is mostly a US-based population and the study that was published in JCO had a Japanese patient population. We did a phase 1/1b study; the phase 1 portion was a simple dose escalation, which started regorafenib at 80 mg, [then escalated to] 120 mg, and 160 mg, along with a standard dose of nivolumab.

In the phase 1 portion [of the trial], we included a total of 12 patients. Among those patients, we saw 3 dose-limiting toxicities (DLTs); 2 were seen at the 120-mg dose and 1 was at the 80-mg dose. Therefore, based on the design of the study, we felt that in the expanded cohort, a dose of 80 mg of regorafenib should be used in combination with single-agent nivolumab.

Now, in the expanded cohort, we're supposed to get a total of 40 patients; so far, at the time of the presentation, we had a total of 16 patients. We have a total of 28 patients involved in this analysis. Our results were, unfortunately, a little bit different than [what had been reported by] the group from Japan. The [partial] response rate was [4.8%]. Twenty-one patients were evaluable for response, meaning that they had more than 1 baseline scan done. Only 1 patient had a PR, which was not confirmed. About 67% of patients had stable disease. In our study, the median PFS was 4.3 months and the median OS was 11 months, granted that the median follow-up was only 4.7 months.

The toxicity that we saw in the study was very similar to what they saw in the Japanese group. The most common AEs seen were rash, fatigue, hand-foot-skin reaction, and some gastrointestinal toxicity. We'll get more data, but the data we have [are] different [from the data] that were published.

Could you expand on the differences between the 2 studies?

There are many potential reasons behind [these differences]. The first is that the baseline demographic could be different. We had many patients in our study who had KRAS-mutant disease. We also had many patients with right-sided tumors and those who had liver metastases versus lung metastases.

There's also the hypothesis that patients with lung metastases do better because they seem to be more immunogenic than [those with] liver metastases. Once again, our numbers are very small, so it's tough to differentiate those areas. However, in our study, we have seen some mixed responses in patients with liver metastases and lung metastases, where the lung lesions shrunk, but the liver lesions grew. There's some kind of a discrepancy with that, but those are hypothesis-generating at best.

We will need more data before we reach any conclusion, but so far, we see that unlike the group from Japan, our data show very modest efficacy. The safety data are what we expected with the combination. However, we're still waiting to accrue a total of 40 patients in the expanded cohort to gather more efficacy data to come up with a hypothesis. Maybe we will be able to select certain biomarkers or phenotypes of the tumor that can predict the response of the combination of regorafenib and nivolumab.

Was there anything you wanted to add?

In our study, in terms of the inclusion criteria, which is a little bit different than what was already published from the Japanese group, about 85% of patients had a right-sided tumor. We know that those patients tend to have a poor prognosis compared with 14% of patients who had a left-sided tumor. These were heavily refractory patients; 50% received third-line or higher therapy, [while] about 21% of patients received TAS-102. All patients received 5-fluorouracil, oxaliplatin, irinotecan (Onivyde), and bevacizumab (Avastin), and if they were RAS wild-type, they received an anti-EGFR drug, as well; these were very heavily pretreated patients. About 71% of patients were RAS-mutant versus 28% who were wild-type. Why is that important?

There are some data that [suggest] patients who are RAS-mutant may be less immunogenic. Maybe these are patients who are not appropriate for this immunotherapy. Additionally, it seems from other studies that patients with lung metastases do better than those with liver metastases. In our study, about 70% of patients had liver metastases and very few patients had lung metastases only. Many patients had a combination of liver and lung metastases. Some of the data that we've seen, at least from our 28 patients, is that several patients had a mixed response. One patient had a response in the lung, but not in the liver, for example; there are a couple of patients like that.

As of data cutoff, many patients did progress on the combination. The PR rate was only 4.8% and the stable disease was about 66.7% with a disease control rate of 71.4%. However, having said that, if you look at the spider plot, a few patients did have durable, stable disease. For example, 1 patient was on [the regimen] for 56 weeks and another for 32 weeks, despite not meeting the PR criteria. There are a few patients with very durable, stable disease. Why is that? We need to do a better job of taking a look at that.

The medium OS is 11 months, once again, not to compare apples with oranges here, but if you look at the OS of patients who got regorafenib or TAS-102 in those large phase 3 studies, it was around 6 to 7 months. We're comparing a small study with a large study, so I don't want to go there, but at least the median OS seems to be longer than historical numbers. The median follow-up was only 4.7 months and the 12-month OS rates were about 31.9%. A little bit over one-quarter of the patients were living beyond 1 year, which could be a selection bias versus the combination just doing something. Those are the unique findings that we have. In the expanded cohort that we're doing right now, we're doing pre- and post-biopsy on all the patients. Therefore, hopefully, we'll get more qualitative data to figure out which patient is benefiting and why.

Could you expand on the safety profile of the regimen?

Most of the AEs we observed in the 28 patients were grade 1 and 2. The most common grade 3 or higher AE is rash. Three patients in the phase 1 cohort had grade 3 rash and had to be dose-reduced. A few patients had grade 3 hypertension and a few patients had hand-foot-skin reaction. Having said that, the bulk of the patients had grade 1 and 2 AEs and a few patients needed a dose reduction.

Interestingly, it may not be the dose of regorafenib that matters but receiving regorafenib, a TKI that's immunomodulating and changes the microenvironment, at a lower dose and getting immunotherapy [could be] how this combination is working. It isn't clear but it is very interesting to find out that even the patient at the lowest dose of 40 mg experienced some benefit from the combination.

What are the next steps for this research?

Beyond expansion, we are doing pre- and post-biopsy, so we're going to get some translational work to figure out: If [a patient doesn't] respond to this treatment, why are they not responding? If they do respond, certain biomarkers [may] tell us why the patient is responding. If we're going to move on with the study to a larger phase 3 study, hopefully, we could better select the patient population that may benefit from [the regimen].

For example, [will it be best for] a patient with lung metastases, or a patient with a right-sided versus left-sided tumor? Is there a tumor biomarker that can predict who will respond to the combination or not? Those are the things that we have to look into to move on to the next step.

What is your take-home message regarding this research?

Everybody was very excited about the data presented last year from the Japanese group. Some of our colleagues who were running out of options were giving this drug off-label because they had nothing available for their patients. Based on these provocative data, this may be the way to go. However, I want to be very cautious of that, because our data didn’t quite duplicate what they found in Japan.

I would encourage [the field] to wait a little bit longer to see more mature data from our trial or some other data out there before embarking on the combination. Hopefully, we'll learn more about the biomarkers so that we can potentially [identify] certain patient populations who will benefit from the combination.

References

1. Kim R, Sadeghi , Imanirad I, et al. Phase I/IB study of regorafenib and nivolumab in mismatch repair (MMR) proficient advanced refractory colorectal cancer. Presented at: ESMO World Congress on Gastrointestinal Cancer 2020. July 1-4, 2020; Virtual. Oral-20.
2. Fukuoka S, Hara H, Takahashi N, et al. Regorafenib plus nivolumab in patients with advanced gastric or colorectal cancer: an open-label, dose-escalation, and dose-expansion phase Ib trial (REGONIVO, EPOC1603). J Clin Oncol. 2020;38(18):2053-2061. doi:10.1200/JCO.19.03296