Shields Sheds Light on Complex mCRC Landscape

Anthony Shields, MD, PhD

Anthony Shields, MD, PhD

Immunotherapy and targeted therapy have shown promise in the treatment of select subsets of patients with metastatic colorectal cancer (mCRC) who harbor genetic alterations, but the question of whether this benefit can be extended into a broader patient population remains unanswered, said Anthony Shields, MD, PhD.

Prior research has suggested that patients with microsatellite stable and mismatch repair—proficient mCRC are not responsive to immunotherapy, and a recent attempt to overcome this challenge fell short. In the phase III IMblaze370 trial, investigators compared the use of the PD-L1 inhibitor atezolizumab (Tecentriq) alone or in combination with the MET inhibitor cobimetinib (Cotellic) to regorafenib (Stivarga) in patients with unresectable locally advanced mCRC who received ≥2 prior regimens of chemotherapy.

The study failed to show improved overall survival (OS) with atezolizumab compared with regorafenib. Median OS was 8.9 months in those who received the combination compared with 8.5 months in patients treated with regorafenib. Those who received atezolizumab alone experienced a median OS of 7.1 months.

Although this trial failed to meet its primary endpoint, Shields, a professor in the Department of Oncology and the Molecular Imaging and Diagnostics Program at Barbara Ann Karmanos Cancer Institute, stressed that many ongoing trials are combining immunotherapy with different modalities in an effort to induce greater responses in these patients.

In an interview during the 2019 OncLive^® State of the Science Summit™ on Gastrointestinal Malignancies, Shields shared recent efforts being made with immunotherapy research in CRC, stressed the importance of molecular testing, and shed light on the use of liquid biopsies in the space.

OncLive: What are the recent updates in relapsed/refractory mCRC?

Shields: Over the last couple of years, we've certainly had new treatments—particularly in the targeted therapy space with HER2 and BRAF—although this is in just a small subset of patients. The HER2 data look very promising in terms of patients with advanced disease and HER2-positive cancers, which represents a small percent of our patients. At this point, there are really only phase II data available, but there are ongoing phase III trials, which we hope will be equally as successful and lead to FDA-approved therapies.

What did the IMblaze370 trial tell us about the use of immunotherapy in the space?

We know that immunotherapy works exceedingly well in patients with mismatch repair deficient (dMMR) tumors. Unfortunately, [that accounts for] only about 5% of [all] patients [with CRC], so we don't get to use immunotherapy as much as we would like to. Investigators are exploring all kinds of combinations to try to enhance [response]; one of which was explored in this trial, which combined a MEK inhibitor [with atezolizumab]. It certainly sounded promising, and in the phase II data we saw a 10% to 15% response rate [to the combination]. We were all ready to adopt this combination, thinking it was going to be a great winner. We thought this was going to allow us to offer immunotherapy to patients with mismatch repair—proficient mCRC. Unfortunately, when we compared the combination with regorafenib, we found that it was really no better [than the standard]. I must say, we were all pretty surprised by this, but the data were clear.

With that being said, investigators are trying every such combination in the world. I am just opening a trial with pembrolizumab (Keytruda) plus an Interleukin-2—enhancing drug. It's going to take a while because it wasn't the quick hit we thought it would be, but immunotherapy will eventually find a way into this patient population. That's basically what IMblaze370 told us. There are many combinations out there exploring various targeted agents, dual immunotherapies, and bispecific antibodies. There are hundreds of trials out there that look interesting.

At Barbara Ann Karmanos Cancer Institute, we have numerous phase I and II trials looking at various immunotherapies. Nobody has jumped up yet and said that one approach stands out and that they've had a patient respond remarkably well, but we are trying every which way to do this.

For the patients who do respond to immunotherapy, how much of a benefit do you see?

For some of the patients, [this approach] makes an amazing difference for a short period of time; it doesn't necessarily last forever. Then there are some patients for whom immunotherapy has done an amazing job for years. I've got a patient with hepatocellular carcinoma who was in an early anti—PD-L1 trial and still has a shrinking tumor after 5 years. This patient has even been off therapy for a few years. When you get wins like that, you think, “Immunotherapy is going to change your practice.”

Unfortunately, we don't see responses like this as much as we'd like [in this space]. We clearly need better biomarkers to identify the best agents for each patient, and we need better ways to monitor these patients to ensure the therapy is working. The other problem is, the toxicity can be overwhelming sometimes, and sometimes we don't know whether it's the disease, an infection, or immunotherapy [that is causing the symptoms]. I've had patients in the last few months who came in with pneumonitis. How do we tell if this is pneumonia or [an immune-related adverse event (AE)?] Patients can get a wide range of AEs that are unpredictable and really hard to diagnose.

What is the standard of care in newly diagnosed mCRC?

We obviously are looking at several targeted agents, but right now these are mainly in the second and third lines [of treatment]. The most difficult thing for the community oncologist is that we have a number of drugs and it isn't necessarily clear how to sequence them. When do you pull back on one of the drugs? If you start [a patient on] a regimen like FOLFOXIRI and bevacizumab (Avastin), the data suggest that you [should] start cutting back after a few months and give them a maintenance regimen. When you [should] do that and who you [should] do that in is still unanswered. You have to personalize [treatment] for each patient.

Part of [the treatment decision should be] based on the feelings of the patient. You might have a patient who says, "That previous therapy really beat me up. I want to stop everything now and take a break." If their disease is under control and they've had a decent response, I'm happy to stop everything, sometimes. Other times it's like, "We've clearly done okay, but you still have pretty active disease, so I'd like to continue on maintenance."

How is genetic testing being used in patients with mCRC?

We're doing it basically in everyone now—even patients with early-stage disease; [in those patients,] we are at least looking for dMMR. This may not change my practice in a stage I patient, but it certainly changes my monitoring and how we monitor the family. At a minimum, we do MMR testing in all patients. In [those with] stage IV disease, we are running big [gene] panels now. These [panels] cost a few thousand dollars, but that's equivalent to 1 dose of my drugs these days. If you tell me a new drug that works or doesn't work, it's worth it by far. Those panels keep getting better, and we just instituted full transcriptome. I expect within the year we'll be doing full sequencing of DNA. We are looking at many potential biomarkers, and 1% of the time I'll find something really rare that is transformative. NTRK or RET fusions can make a difference in how we treat someone.

What is the significance of tumor sidedness?

I must say, when investigators first started analyzing this a few years ago, everyone was surprised. But then we looked back at old studies and found the same thing: left-sided CRC behaves much better than right-sided CRC. There's a really marked improvement in survival for [patients with] left-sided CRC. How they responded to treatment varied greatly. For instance, EGFR inhibitors did not work very well in those with right-sided tumors. The incidence of MMR was higher in right-sided tumors and BRAF was higher in right-sided tumors, so we have some more options there.

There is also just an array of different mutations that one finds. For example, if you look at the BRAF wild-type patients, the improvement you see with EGFR inhibitors still varies from the right to the left side. Therefore, what is it? We are starting to analyze those patients and their tumors and we’re finding that there are secondary mutations driving the cancer. However, we are just starting to try to figure that out. It's clear that anatomically the right- and left-sided tumors are different, but [we are seeing that they are] biologically and genomically different, too.

What is the importance of liquid biopsies?

We are still trying to figure that out. Liquid biopsies are the next up-and-coming thing, and we are just starting national trials evaluating their use in CRC. There are several trials suggesting that those with positive liquid biopsies or circulating tumor DNA (ctDNA) do much worse if they have stage II or stage III disease. We are designing trials to be more aggressive in the treatment of those patients, although some have argued that if [a patient is] stage II and they have positive ctDNA, they really have stage IV disease. There are questions here about [disease] biology that still need to be answered.

The other problem is the assays are quite variable. There are several companies looking at these assays, but cross-comparisons have not been regularly done. You can get different answers with different assays. We want to make sure that we are [choosing appropriately] when we are using these assays. However, we still don't know what the “right” assay is. [These biopsies] will be transformative in the long-run, either as a prognostic indicator or in that looking at a plasma sample [will be] much more efficient. There are many questions like that, so it's a very active field of investigation.

Bendell J, Ciardiello F, Tabernero J, et al. Efficacy and safety results from IMblaze370, a randomised phase III study comparing atezolizumab + cobimetinib and atezolizumab monotherapy vs regorafenib in chemotherapy-refractory metastatic colorectal cancer. Ann Oncol. 2018;29(suppl 5; abstr LBA-004). doi: 10.1093/annonc/mdy208.003.