What BriaCell's Trial Design Says About Real-World Cancer Research

A conversation between BriaCell President and CEO William Williams, MD, FACP, and Clinical Leader Executive Editor Abby Proch

Developing a last-line cancer treatment is a complex space to occupy — as stressful and high stakes as it is hope-filled and invigorating. So goes the experience of William Williams and his team at BriaCell, which is developing BRIA-IMT, a genetically modified breast cancer cell line designed to stimulate a targeted immune response against breast cancer cells.

Understanding the role it plays in treating some of the most high-unmet-need patients, BriaCell has designed its Phase 3 trial to include heavily pretreated patients with advanced metastatic or locally recurrent breast cancer, including those across major breast cancer subtypes and with ECOG performance status up to 2. The study includes BRIA-IMT alone as well as in combination with an immune checkpoint inhibitor against treatment of physician’s choice.

In this interview, Williams discusses how the company approached eligibility criteria for a high-need patient population, why the protocol allows broader enrollment than many late-stage cancer trials, and how real-world considerations shaped the use of treatment of physician’s choice as the control arm.

Clinical Leader: How did you determine the inclusion-exclusion criteria to get the best data and the best outcomes for your patients?

William Williams, MD: This is not the first time somebody has tried this type of approach in cancer. The usual method is to treat patients early in the course of the disease because their immune system gets beat up as they go through rounds of chemotherapy. But our scientific founder, Dr. Charles Wiseman, documented several patients where bulky disease disappeared even though they'd been very heavily pretreated. The common wisdom about these types of approaches doesn't necessarily hold true. Patients with severe, heavily pretreated cancer can respond to these interventions.

We've seen patients who have been on numerous lines of therapy have very good responses. One woman, for example, had 13 prior regimens that failed. She had a big tumor behind her eye, causing her eye to bulge forward, and tumors on the outside lining of the brain and the adrenal gland. With our therapy, the tumor behind the eye completely disappeared. The other ones improved.

In our Phase 2 study, we got some very encouraging results: survival that looks like it's about twice what would've been expected based on what is published in the literature. It’s very encouraging data despite the gamut of treatments these patients have seen.

Some of the inclusion-exclusion criteria are likely boilerplate, but some might be highly specific or niche. Were any criteria hotly debated or considered unique?

When you develop drugs for breast cancer, you usually pick a subtype— hormone receptor positive breast cancer, HER2-positive breast cancer, or triple-negative breast cancer. Those are the three big buckets that breast cancer falls into.

We looked at our mechanism of action, because our BRIA-IMT cell line expresses multiple different breast cancer antigens and induces a very broad immune response. And we've seen good activity in all three major subtypes of breast cancer. So, we went to the FDA and said we would like to look at all comers in breast cancer, but we are going to stratify enrollment so that patients get randomized into our therapy or treatment of physician's choice based on these subtypes.

The other thing that I think is actually more important for the patients is what's called ECOG performance status. ECOG is typically graded from zero to five. Zero means you're fully functional, four means you're completely bedridden and five is dead. Typically, cancer studies, even in late-stage cancer, allow ECOG 0 or 1. One means there is some functional decline, but it's not marked. You can completely take care of yourself and do most of your activities. ECOG 2 is a little bit more serious, where patients have to rest for up to 50% of the time. So, it's a more advanced disease population.

We've allowed ECOG 2 patients into our study. We've noticed that our therapy can induce an immune response, even in these ECOG 2 patients. Also, these are the patients who really need something. They're completely out of options and at the end of the line. You wouldn't want to give them something that's got a lot of side effects, and their limited life expectancy is going to be spent with chemotherapy-type side effects. But our therapy has been very well tolerated. The main side effect we see is local irritation at the inoculation sites. So, we're allowing inpatients all the way up to ECOG 2 into our study, even though that's not typical.

Do you track the types of treatments patients have received prior to their enrollment, and if so, what is the most important history to capture?

We do track it, absolutely. The treatment for breast cancer depends on the subtype. Hormone receptor positive patients, for example, initially get treated with hormonal therapy, which makes sense. And then they get CDK4/6 inhibitors, which is considered to be a targeted therapy. HER2-positive patients get antibody therapy targeting the HER2 molecule, which includes multiple different antibodies and ADCs, and that's very effective in many of those patients. Triple-negative breast cancer traditionally has been treated with chemotherapy because there's not a lot else that can be offered. But more recently, ADCs have also proved to be effective in triple-negative and in hormone receptor positive cancers. Immune checkpoint inhibitors are also used in triple-negative, but only in combination with chemotherapy.

Now, our patients usually have failed all of these therapies, including ADCs and checkpoint inhibitors. We're looking at those patients in particular. We think that's a very interesting group because they have the greatest unmet medical need. We are going to be looking at all the prior therapies to see if there's a pattern there that could inform us which patient population we're most effective in.

Because you've allowed the different subtypes to be enrolled in this study, does that make patient recruitment any less challenging?

Several investigators have said to us that they really want our study because they don't have any other options for these types of patients. We're clearly in a patient population that has a huge unmet medical need, and then the fact that we don't limit it to one subtype or another does clearly help with recruitment. We have a built-in interim analysis, and if it's positive, we can file our licensing application at that point. But if it falls short, we will look at the different subtypes and see if there's a subtype that we're working best in and then narrow our study after that.

Some patients will be in treatment arms, and some will have the physician's choice. How variable is that from provider to provider, and how does that impact your data collection and analysis?

It varies depending on the type of breast cancer. Although at this stage where they've already failed everything else, the main therapy still available is chemotherapy. So, almost all patients on the treatment of physician's choice arm are getting chemotherapy.

The disadvantage is you don't have a homogeneous comparative group, but the advantage is that it's a real-world comparison, which makes it more relevant to clinical practice. Functionally, there's really no other way to do this study because it's impossible to specify exactly what you want investigators to give these patients. We have to give maximal latitude to our investigators.

Is there a terminology difference between physician's choice and standard of care?

The NCCN puts out guidelines for the treatment of cancer, and they go through everything from initial surgery to the very late lines of therapy. They define the standard of care, so to speak. We are at a stage of treatment where, if you read the NCCN guidelines for our patients, they have a very long laundry list of chemotherapies and say, "You can try this or this or this or this, or maybe these two in combination."

Technically, treatment of physician's choice is a standard of care-type of treatment in our trial, but they do mean slightly different things. Treatment of physician's choice just means it's up to the treating physician. Theoretically, that could be something that's not standard of care, so we give guidance based on the NCCN guidelines of what would be reasonable to try.

The design allows for BRIA-IMT to be studied as a monotherapy and as a combination therapy. What do you hope to learn by studying each?

The FDA asked us to add a separate arm of monotherapy to get more data, and we did that. The two main arms are the Bria-IMT regimen with an immune checkpoint inhibitor versus treatment of physician's choice. Each has 177 patients. The monotherapy arm was limited to 50 patients, and it's pretty much done with at this point. The FDA really wanted to have more safety information, primarily on the BRIA-IMT regimen alone, so they could parse out what side effects might be coming from the checkpoint inhibitor versus just the Bria-IMT.

What advice would you give to sponsors trying to develop balanced inclusion-exclusion criteria, given the high-need patient population?

The first thing I would say is don't be afraid of the ECOG 2 patients. ECOG 2 is really a measure of the activity of the patient. And if you have a patient who's a little bit more depressed as opposed to somebody who's more optimistic, that can change the ECOG status, even if they have the same level of disease. These patients really need something. They need some hope. So, do not be afraid of including ECOG 2 patients, even though they do have a limited life expectancy.

And then in terms of general trial design issues, consider including a heterogeneous control population. The FDA is very open to this. You'd love to have a homogeneous comparative group so you can get great statistical analysis, but that's not the real world. The real world is that therapy options are changing all the time. Different physicians are going to use different things. And so, if you really want to see how your therapy stacks up against what they would otherwise get in the real world, having a treatment of physician's choice arm is very reasonable.

The drawback to that approach is that you really can't do a blinded study. You can't make a placebo for every possible therapy that's out there. Now, of course, if we were a Big Pharma, we might've made a dozen different placebos and done a truly blinded study, but it would've been very inconvenient for the patients. Our treatment is every three weeks. Others might be every month or every couple of weeks, so it would be very complicated. Ours is a real-world comparison, and I think that's the major advantage.

About The Expert:

William V. Williams, MD, FACP, is the president and CEO of BriaCell Therapeutics Corp. Dr. Williams has over 35 years of industry and academic expertise, including clinical management in multinational pharmaceutical companies. He earned his BSc. in Chemistry and Biotechnology from MIT and medical doctorate from Tufts University School of Medicine. At the University of Pennsylvania, he developed novel methods of bioactive peptide design, ran a research program in receptor biology, helped develop DNA vaccines, and introduced novel therapies into the clinic. At GlaxoSmithKline, he brought several molecules into the clinic, ran an international biomarker laboratory, and worked on regulatory approvals for drugs, including Hycamtin, Navelbine, Bexxar, Zofran, Tykerb, and Boniva. At Incyte Corporation, he established proof of concept for drugs in several disease areas. He also introduced Jakafi and Olumiant into the clinic, working on them through post-approval.  Dr. Williams is the named author on over 130 peer-reviewed publications, over 15 patents, and numerous INDs and NDAs.