Bekaii-Saab Sheds Light on Furthering Biomarker-Driven Treatment in GI Cancers

Tanios Bekaii-Saab, MD

While earlier efforts of precision medicine focused on basket trials, Tanios S. Bekaii-Saab, MD, explained that it’s time for the field of oncology to move on to umbrella trial designs that can better target an individual cancer.

In a breakout session during the 2020 Gastrointestinal Cancers Symposium, Bekaii-Saab, medical oncologist, and medical director of the Cancer Clinical Research Office, vice chair and section chief of Medical Oncology in the Department of Internal Medicine, at Mayo Clinic, presented on basket trials and umbrella trials in biliary cancer, adding that umbrella trials can provide a new proof-of-concept.

“We do these umbrella protocols where we develop a proof-of-concept for FGFR, IDH, HER2, or other targets for biliary cancer, and depending on the alteration we find, the patient will be matched to a treatment,” said Bekaii-Saab. “That's enriching for the disease and enriching for the target.”

Another method for ensuring that patients receive a drug that matches their particular disease characteristics is biomarker testing. According to Bekaii-Saab, every patient with a gastrointestinal (GI) malignancy should be screened for genetic alterations with next-generation sequencing.

However, biomarker testing is not without challenges, Bekaii-Saab explained. Responses seen with a targeted therapy in one disease does not necessarily translate to other diseases, he said, adding that ongoing research is looking into better understanding these differences.

In an interview during the 2020 Gastrointestinal Cancers Symposium, Bekaii-Saab examined the effectiveness of umbrella trials and their importance in aiding precision medicine in GI oncology.

OncLive: Can you provide an overview of your presentation at the 2020 Gastrointestinal Cancers Symposium?

Bekaii-Saab: The way we look at these targets in synchrony has been done in two different ways; we have the basket trials and umbrella trials. Basket trials are those trials that take 1 target, 1 drug, and whatever disease that has the target, and the patient is entered into that arm. The new generation of trials is the umbrella trials, in which there is 1 target, a disease-specific entity that is being tested, and other targets that belong to the same disease. The umbrella trial is a disease-centric mechanism with multiple targets, whereas the basket trials are target-centric, multi-arm studies with a target that is tumor-agnostic.

What are the pros and cons of basket trials and umbrella trials?

With the basket trials, there is a good proof-of-concept about whether an agent works with a specific target, regardless of histology. The problem is that there are very few targets matched with agents that seem to fit the bill; these agents would be PD-1 and NTRK inhibitors, and also agents that [work in MSI-H tumors].

Most targets behave differently in a specific disease. One example is BRAF V600E. In melanoma, BRAF V600E seems to have a good response to single-agent BRAF inhibitors, but it doesn't work the same when we use it as a target in colorectal cancer. The response rate goes from more than 50% to 5%, and [these are] short-lived responses. This is just one example, but there are a lot of targets that [don't have the same activity in GI cancers as they do in other cancers].

It doesn't make a lot of sense to continue down the path of tumor-agnostic drug development. It's better to move our efforts into a disease-specific setting, and that is what the umbrella trials are.

Which GI malignancies have a greater need for umbrella trials?

It becomes urgent to focus on disease-specific trials when we are thinking about rare diseases, such as biliary tract cancer, where you're pulling from a lot of different studies with few patients, or from a basket trial with multiple targets. Then, you get a dilution effect from the patients with biliary tract cancer who enroll. It's like looking for a needle in a haystack, and it makes it difficult to make sense of the results. You may miss some good targets or good agents that may work in that specific setting; that is not a good way to develop agents anymore.

With biliary tract cancer, and others such as pancreatic and colorectal cancer, we want to move to a disease-specific setting. We do these umbrella protocols where we develop a proof-of-concept trial for FGFR, IDH, HER2, or other targets for biliary cancer. Depending on the alteration we find, the patient will be matched to a treatment. This enriches for the disease and enriches for the target.

What steps being taken to move this adoption of umbrella trials forward?

In addition to what umbrellas do in terms of matching the agents, we want to move to a platform, or perhaps have multiple umbrellas under a single platform. In a way, this can help drugs enter in or out [of the treatment landscape]. If the drugs are promising, they go on to larger studies. If they are not promising, they go out of development and get replaced. The platform allows you to interact with translation components, such as biopsies, [disease] modeling, and organoids. This is a way for us to learn a little bit more about the biology [of GI diseases] and refine some of the genetic and molecular predictors of response [to treatment]. Then, we can go back to the umbrella trials and continue developing new agents for new targets. That's the most efficient and logical way to move this precision medicine platform into the future.

For the last 10 years, we spent a lot of time doing tumor-agnostic basket trials and got a lot of information about tumor biology. The next decade needs to focus on these platforms and umbrella trials, because we understand that tumor-agnostic trials only apply to a few targets. For the majority, the target behaves differently and they depend on specific histologies. For that, basket trials are not useful.

Could you discuss the advances with biomarker testing for GI cancers?

Every patient with a GI cancer should be screened for alterations, at least through next-generation sequencing. This is important because of what's approved and what's being studied. When we think about what is approved today, or what's on the way to being approved, we have MSI-H, which is a predictor of significant activity from PD-1 inhibition. We also have NTRK inhibitors for NTRK fusions, and we heard about a review from the FDA for the combination of encorafenib (Braftovi) and cetuximab (Erbitux) from the BEACON CRC trial in BRAF V600E refractory colon cancer. I suspect that this combination will become the standard.

We also understand the value of RAS mutations, and the benefit exists—even for patients who received EGFR inhibitors. Then, we have a number of other promising agents that are emerging. Some of [these agents] are more advanced than others. For HER2 amplification, we presented a study at the 2019 ESMO Congress with tucatinib and trastuzumab (Herceptin), which has shown significant promise with a 52% objective response rate (ORR) and a median progression-free survival close to 9 months. What is interesting from that study is that, for the left-sided patients with HER2 amplification, the ORR is above 60%; those responses tend to be quite durable and meaningful. This [research] is now moving to the next level, with an expansion study of our initial study MOUNTAINEER (NCT03043313).

HER2 is an emerging target, and an important target, because it can predict for the lack of activity from EGFR inhibitors. It also excludes patients from receiving EGFR inhibitors, similar to the BRAF and RAS mutations.

Overall, this is helping us redefine our standards of care in the clinic, but it also gives us opportunities to put more patients on these targeted strategies in colorectal cancer. A lot of them are going through a program called COLOMATE, which is running through our research consortium. This includes EGFR re-challenge, HER2 amplification, FGFR alterations, c-MET amplifications, and a lot of other targets.

The world of GI cancers is moving more towards classifying the cancers according to sight, in addition to molecular and genetic characteristics. That is also true for pancreatic cancer, in which we now know that BRCA1/2 mutations are relevant. BRCA1/2 mutations helps us refine our treatment options for pancreatic cancer.

What challenges exist with using precision medicine in GI oncology?

The world of oncology, overall, is moving more into what we call “precision medicine world,” and we are getting better at picking up targets and going after these targets directly. Quite a few challenges remain with targeted strategies, and one of the main issues is the lack of activity for many patients. We have targets that work in 20% to 30% of patients, but they don't do as well in the other 70%; although we have identified the target that the agent hits directly. In an ideal world, we like 100% of patients to respond to the treatment, but that's not what we're seeing and there are many reasons for that.

One reason is that there is probably co-activation of other targets that may pull resistance. That can be at the onset of disease. We continue to try to find strategies and refine strategies that would allow us to increase the likelihood of a response and ultimately improve outcomes in the future.

Bekaii-Saab TS. Biomarker-based basket trials: a new model for biliary cancer drug development. Presented at: 2020 Gastrointestinal Cancers Symposium; January 23-25, 2020; San Francisco, CA.

<<< View more from 2020 Gastrointestinal Cancers Symposium