Combinations With PD-1/PD-L1 Agents May Improve Outcomes in mCRC, Studies Show

In an interview with Targeted Oncology, Tanios Bekaii-Saab, MD, discussed the results of the BACCI and MOUNTAINEER studies, which showed promise for the use of PD-1/PD-L1 combinations over single-agent treatment in patients with metastatic colorectal cancer. 

Tainos Bekaii-Saab, MD

New studies are evaluating PD-1 and PD-L1 inhibition in combination with other agents like VEGF and HER2 inhibitors as a new approach for treating patients with colorectal cancer (CRC). These studies aim to prove that combination strategies are more effective than PD-1 and PD-L1 inhibition alone for the treatment of metastatic CRC (mCRC).

The phase II BACCI study of capecitabine and bevacizumab (Avastin) plus atezolizumab (Tecentriq) versus capecitabine (Xeloda) and bevacizumab plus placebo in refractory mCRC, which was presented during the 2019 ESMO Congress, showed that progression-free survival (PFS) significantly improved with the addition of atezolizumab to capecitabine and bevacizumab compared with placebo.1

The randomized, double-blind, multicenter, placebo-controlled study randomized 133 patients 2:1 to the atezolizumab-based combination or the doublet, in an effort to achieve the primary endpoint, PFS, and secondary endpoints of objective response rate (ORR), overall survival (OS), and safety/tolerability, using the modified intention-to-treat analysis (mITT).

Among the patients treated in the atezolizumab arm (n = 82), the median PFS was 4.4 months compared with 3.3 months in the placebo arm (HR, 0.725;P= .051). The ORR was 8.54% with the addition of atezolizumab compared with 4.35% in patients who received placebo (HR, 0.94;P= .4). The investigators reported that no new toxicities arose with the combination therapy versus with capecitabine, bevacizumab, or atezolizumab alone.

Similar results success was seen in the MOUNTAINEER trial, which looked at the combination of trastuzumab (‎Herceptin) and tucatinib (ARRY-380/ONT-380) in patients withHER2-amplified mCRC.

Patients in the multicenter, open-label, single-arm phase II trial received tucatinib (300 mg), and the standard dose of intravenous trastuzumab for 3 weeks. The primary endpoint was ORR.2

Based on the trial results presented at the 2019 ESMO Congress, the ORR was 55% in the 22 evaluable patients. Of this group, 12 patients had a complete response, 5 showed stable disease, and 5 experienced disease progression. The study met its primary endpoint, and the combination of trastuzumab and tucatinib was also considered tolerable.

In an interview withTargeted Oncology, Tanios Bekaii-Saab, MD, professor of medicine, Mayo Clinic, discussed the results of the BACCI and MOUNTAINEER studies, which showed promise for the use of PD-1/PD-L1 combinations over single-agent treatment in patients with mCRC.

TARGETED ONCOLOGY: What was the rationale for combining atezolizumab and bevacizumab in microsatellite stable (MSS) CRC?

Bekaii-Saab: There's a significant amount of evidence that the agents that target VEGF essentially facilitate immune modulation and immune cells access the tumor that would have otherwise been considered cold, such as MSS CRC. Combing VEGF inhibitors, such as bevacizumab, with a PD-1/PD-L1 inhibitor, such as atezolizumab, may have a synergistic effect. That's been shown in a number of preclinical models and has moved to the clinic. We've seen good activity across multiple malignancies when combining the 2, like in hepatocellular carcinoma (HCC) and others. We want to look at this in mCRC with a specific emphasis on those patients where we don't think there's much value from PD-1/PD-L1 inhibition alone.

The study was designed as a large phase II randomized study with about 130 patients who had failed at least 2 lines of standard chemotherapy and who were pre-exposed to bevacizumab and [some to] EGFR inhibitors [as well], [were randomized] to receive capecitabine and atezolizumab plus bevacizumab versus capecitabine and bevacizumab and placebo. This was a placebo-controlled study. The primary endpoint of the study was PFS since the study was blinded and with a placebo.

In our study, we were expecting PFS to be improved with the combination versus the capecitabine plus bevacizumab and placebo. The study reached its primary endpoint, so it was positive for PFS. The hazard ratio for the [patients with] MSS CRC was 0.67. That's impressive. We haven't seen a lot of responses. The responses, as expected, were about 8%. There was a lot of stable disease with some shrinkage, and prolonged end drops in the CA. There was activity there.

The study is positive, and the question is where do we go with this next. The thought is as we go through further analysis, that this may move us closer to a phase III randomized trial to ask a similar question. Also, there's a large amount of data that are coming out in different cancers with similar combinations, and that may continue to set the field.

We're very excited about this. It's a large endeavor that was performed through our research network, Academic and Community Cancer Research United (ACCRU), which helped conduct the studies through multiple institutions. This was an investigator-initiated trial. Hopefully, it will move the research field forward. In terms of the clinical application, it's too soon to tell because this was a phase II randomized study before moving to a larger study. But it certainly establishes a good proof of concept that this strategy is likely an effective strategy in CRC specifically, but also across multiple gastrointestinal (GI) malignancies as we're seeing the results come through.

TARGETED ONCOLOGY: Were the toxicities comparable between arms?

Bekaii-Saab: The toxicities were along the expected lines. There was not much difference between the 2 arms. With atezolizumab, we saw additional higher rates of immune-related events. Other than that, there did not seem to be much significance in terms of difference in toxicity between arms, beyond the expected.

TARGETED ONCOLOGY: In terms of prior exposure to bevacizumab was it a requirement?

Bekaii-Saab: It was required. Unless there was a reason why they shouldn't receive bevacizumab. Most patients in the United States are exposed to bevacizumab, so all the patients in this study were pre-exposed.

TARGETED ONCOLOGY: What key point should community oncologists take away from this study?

Bekaii-Saab: The best part is that this study was large enough to give us a meaningful outcome. What makes these results even more compelling is the fact that we had a placebo arm, and the investigators were blinded to what the patients were on, which makes the results even more compelling.

TARGETED ONCOLOGY: Can you explain the rationale for studying tucatinib in combination with trastuzumab?

Bekaii-Saab: We've seen a number of preclinical and early-phase studies suggesting that double inhibition of HER2 is superior to single inhibition in patients with mCRC that overexpresses HER2.

HER2 is a target that's rare in CRC. It occurs in about 4% of patients. It's rare enough to make studies such as this one difficult to conduct. We've had some hints from 2 studies.

The HERACLES study told us that adding lapatinib (Tykerb) to trastuzumab can induce about a 30% response rate in [patients with] mCRC that overexpresses HER2. We had another study, MyPathway, with trastuzumab and pertuzumab (Perjeta) that also showed a 30%-plus response rate. However, there are some relative toxicities to lapatinib that are non-HER2 targeted than with pertuzumab. With trastuzumab, you can also get significant toxicities.

The thought is to use this oral tyrosine kinase inhibitor (TKI), tucatinib, which is a very specific HER2 inhibitor—it's a much cleaner HER2 inhibitor than lapatinib, it's also more potent and on-target—in addition to trastuzumab. This came from preclinical studies and early clinical promise showing that this type of combination works well in breast cancer and this is now in a phase III study in breast cancer.

We took this combination with one of my co-primary investigators, Dr. John Strickler from Duke who helped develop the study. Then, we looked at it in those patients with HER2-overexpressing mCRC. This again ran through our ACCRU network, which helped run the study through close to 8 institutions. It was a small number of institutions but had very dedicated investigators.

The study showed a response rate close to 50%, which is one of the highest response rates reported in this group of patients. There are now about 34 patients included in the study. At ESMO, we presented the first 25 patients, but the results remain consistent.

The median duration of response is close to a year, and we have one patient who's a complete responder and continues [responding for] close to 2 years. Patients who are crossing the 1-year line, and are continuing, are patients who were pretreated with other chemotherapy. The only thing about this study that may be a little different than others is that we did not mandate EGFR inhibitor failure prior to enrolling in this study. Patients had to fail at least 2 lines of therapy, and that included VEGF inhibitors as per standard in chemotherapy. The reason for not requiring all patients to be exposed to EGFR inhibitors prior to enrolling is the large amount of data that are now showing that patients withHER2amplifications do not respond as well to EGFR inhibitors. It makes sense because HER2 is an escape mechanism for the EGFR blockade. Preclinically, we know that it's true, and clinically, we're seeing the results of 3 separate, large retrospective analyses that are consistently showing that. We decided to make it optional. Some patients were exposed but many others were not.

The percent of responses and the duration of responses were quite significant. We also looked at the biochemical responses with CAs and we've seen similar trends across patients who respond well.

At this time, survival continues to be evaluated but it's reaching close to 2 years. This is in a refractory setting. That's significant activity. We're hoping that our results will move forward to a much larger registrational strategy. We're hoping to establish this combination as one of our standards in the near future.

TARGETED ONCOLOGY: Was the study too small to assess response based on prior exposure to EGFR inhibitors?

Bekaii-Saab: At this point, it is too small of a study to be able to do that. As the study continues to expand, we will have more data regarding prior exposure to EGFR. I suspect that as more knowledge about the lack of activity of EGFR inhibitors starts spreading out, it will become more and more difficult to have patients at least in the United States pre-exposed to EGFR inhibitors when they are moving to a combination such as this.

TARGETED ONCOLOGY: Can you discuss the safety profile for this combination?

Bekaii-Saab: One of the things that we found was that this combination is extremely safe. Since tucatinib itself does not seem to have many off-target toxicities, other than fatigue and diarrhea, which are controllable, the combination of the 2 drugs actually ends up being incredibly well tolerated. As I mentioned, the patients that have responded nicely and continue to tolerate it well have gone back to normal lives and are doing great.

We're very excited. I think this is going to be a transformative study that will push access to HER2-targeted strategies even further for patients with mCRC that overexpress HER2.

TARGETED ONCOLOGY: Were there any other data on GI cancers presented at ESMO that you are particularly hopeful about?

Bekaii-Saab: One of the most interesting studies that was presented and has been most awaited is nivolumab (Opdivo) versus sorafenib (Nexavar) in HCC. This study was designed to show the superiority of nivolumab versus sorafenib based on some promising data in the space. Unfortunately, the study ended up being negative. It came from the heels of another study with pembrolizumab (Keytruda), another PD-1 inhibitor, versus placebo in the second-line setting, which also ended up being negative. It has been quite a series of setbacks in the field of developing PD-1 inhibitors in advanced HCC.

The study showed that the OS has not much improved. Between the 2 arms, the hazard ratio was 0.85. What was more of a surprise was that the PFS was about the same. The hazard ratio of the PFS, which is a good measure across the whole line, was 0.93. There's not much of a difference between the 2 and that was extremely disappointing.

Now, we're left with trying to understand how we can sequence all of these agents that are accessible to us in the United States. Between the PD-1 inhibitors, the TKIs, and ramucirumab (Cyramza) in patients with an alpha-fetoprotein level of more than 400. One could safely say that the best level 1 evidence that we have on hand today is either sorafenib or lenvatinib in the first line. I don't think there's enough data supporting nivolumab, to support its application in the first line, now that we have a negative study. That's gone for now. In the second line, it's either a TKI like regorafenib, or cabozantinib (Cabometyx), or ramucirumab if the alpha-fetoprotein level is greater than or equal to 400, which occurs in about 40% of patients with HCC.

Where does pembrolizumab fit now that we have negative data? It will probably fall into the third line, along with nivolumab.

One of the studies that was also presented at ESMO looked at the combination of bevacizumab with atezolizumab and there are a number of other studies that are looking at either TKI or even ramucirumab with pembrolizumab, or nivolumab. We know that this is an [approach] that's very interesting. We're seeing these hints that combing a VEGF inhibitor with a PD-1 or PD-L1 inhibitor can produce some meaningful effects. The study with bevacizumab and atezolizumab continues to show a high response rate as well as meaningful and prolonged responses. That study has completed comparing atezolizumab plus bevacizumab versus sorafenib and we are expecting some results in the next few months. If the results continue to hold into the phase III data, this may change the landscape of how we utilize these PD-1 and PD-L1 inhibitors in CRC in combination with other agents, primarily VEGF or multikinase inhibitors, as more data come in over the years. That may keep these immune therapeutic strategies relevant in HCC following this major setback that we've seen with the nivolumab and pembrolizumab single-agent studies.

I'm not pessimistic about the future role of the PD-1 and PD-L1 agents. I'm actually quite optimistic that we're going to find the right combination. But clinically for today, I think the nivolumab versus sorafenib, and the pembrolizumab versus placebo study shows us that we are not ready to replace our TKIs or VEGFR2 monoclonal antibodies, such as ramucirumab in patients with high alpha-fetoprotein, with these PD-1 inhibitors. My clinic has actually pushed these PD-1 inhibitors to the third line.


  1. Mettu NB, Twohy E, Ou FS, et al. BACCI: A phase II randomized, double-blind, multicenter, placebo-controlled study of capecitabine (C) bevacizumab (B) plus atezolizumab (A) or placebo (P) in refractory metastatic colorectal cancer (mCRC): An ACCRU network study. Presented at: 2019 ESMO Congress; September 27 to October 1, 2019; Barcelona, Spain. Abstract 533PD.
  2. Strickler JH, Zemla T, Ou FS, et al. Trastuzumab and tucatinib for the treatment of HER2 amplified metastatic colorectal cancer (mCRC): Initial results from the MOUNTAINEER trial. Presented at: 2019 ESMO Congress; September 27 to October 1, 2019; Barcelona, Spain. Abstract 527PD.