Revolutionizing Access to Ground-Breaking Clinical Trials for Cancer Patients: Danielle Ralic, Ancora.ai

Danielle Ralic, Ancora.ai: 0:00

People said, I'm out of approved

therapies, I need something, or: 0:01

I'm on my last approved therapy, and I'm not sure it's working, so I want to already see what's coming next. And one person even put, Thank you so much, we never had access to this information.

Alex Jani, X-Health.show: 0:18

Hi, I'm Alex and welcome to the X-Health.show, where I talk to visionaries behind the latest innovations in healthcare. For the eXtra health of the future. From the 15th floor where we're recording, Zurich looks like a lively toy city and on a sunny day, such as today, the Alps emerge on the horizon. Danielle Ralic is the CEO of Ancora.ai. She's a technologist that used to develop analytic solutions for pharmaceutical and healthcare organizations. Now she's on a mission to bring treatments available in breakthrough clinical studies to cancer patients. Ancora.ai is a Swiss startup that developed an AI powered platform where cancer patients and their loved ones can search for available trials. In a matter of minutes. You'll hear about why to consider a clinical trial if you're a cancer patient. How to find the relevant trial. Why it is so frustrating for both those who run them and patience, and how tech bird with willing humans can help. [INTRO ENDS] Danielle, what was the moment that you decided to drop your consultancy job and build a solution for cancer patients searching for the last... well, the last hope clinical trial treatment?

Danielle Ralic, Ancora.ai: 1:30

It was a series of events. But it all started with... I studied information systems with a specialty in healthcare IT because I really saw the potential of technology in healthcare to make people's lives better. And I really think healthcare is one of the most noble industries anyone can work in. And so I spent my career trying to find where is this innovation? Where are we doing technology, or are implementing technology that helps people and really all people, I'm a big proponent of equity as well, when we build these solutions. And I was fortunate enough to work in several cities with multiple big healthcare institutions, kind of learning what technology and healthcare is like and implementing solutions in these settings. But I was just frustrated with the pace. And I was lucky enough to be working and supporting pharmaceutical clients so I had a front row seat to all the innovation that was happening, and specifically oncology or cancer, I was seeing all of these amazing things they were studying, putting in trials. And then I also had a front row seat to reality for many patients, including my family and friends, and seeing their experiences in healthcare. And it was seeing friends getting diagnosed with cancer.

Alex Jani, X-Health.show: 2:46

I'm sorry to hear that.

Danielle Ralic, Ancora.ai: 2:48

I think we all have, unfortunately, these stories. But seeing them not being offered trials and knowing that trials offer this potential in terms of it could help you be better diagnosed, it could help extend your life, it could even cure your cancer. To me, it's a really important option that everyone should be presented. And I kept kind of bumping into this, personally, this problem of it's not happening as often as it should be. It all I guess peaked when I was, from the pharma side or professionally, seeing the challenge they had of recruiting patients, and then personally seeing the lack of outreach or options for these for people that I knew.

Alex Jani, X-Health.show: 3:32

So I mean, the issues on both sides, really. The Pharma cannot recruit enough patients and patients can't find trials.

Danielle Ralic, Ancora.ai: 3:42

Exactly. Not good. An it all peaked when I was hiking. I happen to bump into a man who asked me to take a photo of him and his wife.

Alex Jani, X-Health.show: 3:53

Where was it?

Danielle Ralic, Ancora.ai: 3:54

This was in Patagonia, so quite remote. We just got to talking and he was asking what I was doing. At the time, I was working on analyzing enrollment of a PD-L1 inhibitor and I mentioned that checkpoint inhibitors and he had stopped me, he said, Girl, you're speaking my language! I was like, Oh, are you an oncologist or a researcher? He's like, Oh, no, no, no, I'm just a cancer patient. And I was like, wow, that's really amazing that you know about this. This was just in trials at the time. And he said, Yeah, I'm really fortunate. I I had stage 4 cancer, so far, it's holding. I'm out here hiking with my wife, because we're going to enjoy this as long as we can. And I'm here because my wife's friend's neighbor's cousin, you know, that kind of story, happened to be working at MD Anderson or heard of a trial. And because of that he had access to a clinical trial that extended his life, saved his life. And it was really one of those moments where you see the theoretical science and trials that you've talked about at work. I meet the person or the real life side, and you really see and feel that impact and you think, Gosh, shouldn't everybody have this chance. And that's why I started Ancora. "Ancora" means again or a new hope. And I really think trials do represent that. And again, that's why everyone should have this option.

Alex Jani, X-Health.show: 5:18

How high was that?

Danielle Ralic, Ancora.ai: 5:20

It was really hard hike. I remember thinking like, oh, my gosh, yeah, you should be the poster child for this therapy, because I'm okay, in shape, But it was a pretty high hike, I think it took us four hours to get up there. So very impressive, man.

Alex Jani, X-Health.show: 5:37

He was cleared at the time, basically, thanks to the therapy that he got, because someone knew someone who knew someone. And that...

Danielle Ralic, Ancora.ai: 5:46

shouldn't be happening.

Alex Jani, X-Health.show: 5:48

But it is how it really happens, isn't it. How do people get to know about trials?

Danielle Ralic, Ancora.ai: 5:56

That's exactly right. From the patient side, it's through a set of connections, asking friends and family, asking patients support groups, happening upon maybe a trial being marketed somewhere if you're in the US. And then for physicians, it's what did they remember at a conference. Or do they remember trials at their institution. Or they'll call a colleague and ask them if they happen to know. But hearing the words I'm saying, it's not a very systematic process or a data driven process. And that's why I really think tech flourishes, tech is amazing, and curating information, and helping kind of dissect a lot of a lot of data. And so this is a perfect use case. And I would hope that healthcare would be data driven, systematic, and again, if we can do that, we'll have less of these stories of sheer luck or coincidence. And I always say, if you think about it, then you're leaving your best chance of life up to luck. And it's 2023. We have amazing technology, there's no excuse for not using technology in this case.

Alex Jani, X-Health.show: 7:07

Yeah, totally. And especially that search tools – we use them on a daily basis. You can't probably even calculate how many times a day we use them, right? Why can't we have them for the clinical trials? I read this blog post, a really detailed blog post of Bess Stillman wrote. She's a medical doctor and she was looking for a trial for her husband. Suddenly after a surgery, he developed eight, new tumors within the month. And that basically triggered her to accelerate the search for a clinical trial. She published at some point in this post, what she does, if she wants a certain dress. She went to H&< website, and, long story short, she had a couple of dresses chosen for her based on what she put into this search tool. And she was like, Oh, my God, let the clinicaltrials.gov hire their website developer. Because it doesn't work this way when you go and look for the trail. But we'll come back to that. First, let's maybe go one step back. We already spoke a bit about, why to choose a trial, because it can be, basically, lifesaving. But it's not like it's A trial but we have three phases and some sub phases. Could you speak a bit more about that? What's there and what are the differences between the phases?

Danielle Ralic, Ancora.ai: 8:50

The phases of trials are essentially kind of the steps to getting it approved for the wider population. And so they evaluate them in the steps and the first is starting with safety – it's first inhuman, so it's going from the lab to people, it's usually a smaller population. And there it's really just to see in people, is this tolerable? And is it safe? As the phases increase, there's more and more people being invited to the study because we're trying to make sure the drugs or the new therapies work on different types of people. That's a recent push, thank God, from a lot of regulatory bodies.

Alex Jani, X-Health.show: 9:32

So women actually can have a drug that was actually checked on other women.

Danielle Ralic, Ancora.ai: 9:38

Yes. Because, in fact, we are different. Took them a little too long to realize, but that's okay.

Alex Jani, X-Health.show: 9:44

I was watching your YouTube channel and there was the host, Megan Claire Chase, breast cancer survivor. It took her a couple of years to be diagnosed because she experienced different symptoms that a white woman would.

Danielle Ralic, Ancora.ai: 10:02

Exactly. And, yeah, a lot of the studies are heavily biased towards white men even. There's actually some phenomenal oncologist in Europe trying to research this. And some of the indicators are male oncologists suggest trials to men, but not as often to women. And there's a lot of bias. And this kind of goes back to if the process is based on what you remember, who you know, and things like that, we really introduce a lot of opportunity for bias to hinder who's even invited to these studies. As the representation on the trials is critical, because you want to make sure it's safe for everyone, and that it works as well for everyone.

Alex Jani, X-Health.show: 10:41

And that the dose is appropriate.,

Danielle Ralic, Ancora.ai: 10:43

Right, which is one of the famous ones for Zoloft, where women, after two decades of getting into car accidents because they waked up in the morning, still groggy, realized that women metabolize drugs differently, and this particular one, and that women actually should receive half the dose. But that's why now, there is a big push for diversity on trials, it's taken a little too long to happen, but let's just say we'll celebrate the fact that it's happening. And it's for all those things mentioned. Black women can present differently with symptoms, our bodies might metabolize things differently as well. So it's really, really important that we have these populations to be able to study it. And before it was just white men that were the least risk to your study as well, because women were complicated, you have hormones, and all these things. So you don't want to complicate your study, which might risk the results and might risk your new therapy. But yeah, thankfully, we've learned that it has a lot of, let's say, late stage effects in terms of, could be harmful effects on these different populations that weren't studied. And a lot of drugs were pulled from the market because of that. Thankfully, we have this awareness now and a big push from regulatory bodies saying we really need to do better and we have to invite these populations to studies.

Alex Jani, X-Health.show: 11:59

Right, so we have three phases. What are the differences for the clinical trials?

Danielle Ralic, Ancora.ai: 12:05

Yeah, so there's there's four phases, actually. And the first one is that it's just that safety, a small population. The second is evaluating efficacy: how well is it working? And then we really look at phase 3 for that, is it working better than the standard of care? And this is where it's great if you have a therapy that works, but it has to work better than what we have because obviously, you want people to use it. And for all the right reasons. And so there you'll have the competitor. I think another important thing to note, though, with these studies is that with cancer, I think a lot of the fears we heard from patients were that I don't want to be on a trial where there's a placebo. Because if you have cancer, of course...

Alex Jani, X-Health.show: 12:46

you don't have time to take water with sugar.

Danielle Ralic, Ancora.ai: 12:51

Thankfully, there is a lot of safety involved in trials and how we set them up and what kind of patients are invited, and populations. One of the big things there is, there is no placebo for oncology. It's unethical to do that. That's another thing just to stress for anyone listening. I know that's been brought up many times. They'll always be looking at either the standard of care to compare with the current drug or in the earlier phases, it may just be the drug itself the investigational treatment.

Alex Jani, X-Health.show: 13:44

We have this database, the the US database that actually the whole world is using. Professor George Coukos that I interviewed before says, everyone who has a clinical trial registers the trial in this US database, or they should, or they do it because they want to. And this is clinicaltrials.gov But it has many flows, doesn't it? Coming back to this to this blog post I mentioned, Bess Stillman, it's a series of three very long posts. She started as: "My husband Jake is dying of recurrent metastatic squamous cell carcinoma (R/M HNSCC) and he's exhausted conventional treatments so his only chance at survival is a miracle clinical trial drug". Then she describes her way how to find the clinical study that would work for him. And it's so difficult because it's not like you Google it. So she gives this example: if you change the wording of the search even subtly, you'll get totally different outcomes. She's a medical doctor so she also has an access to doctor groups. She asked several doctors or some doctors offered her help and, like, five doctors delivered completely different five trials. They were not matching. They all got the same history, the patient's history. Let me give you an example. She could search for "squamous cell carcinoma of the head and neck", the different results were for"head and neck squamous cell carcinoma", "squamous cell carcinoma of the tongue","tongue squamous cell cancer". We can see that are similar keywords. And then they they don't... Let me just shorten a bit this story, because what worked for her, she basically found someone suggested, the person that actually does the search professionally. That person spent 15 hours on searching for the trials. She basically figured out the system. Plus, there was also another thing. Her husband read about a trial that completely doesn't mention his his cancer type and actually this could also work. And in this database, they would not find it. There a lot of problems in that. How did you approach them when you wanted to found the startup?

Danielle Ralic, Ancora.ai: 16:59

One important thing, especially in TechCare, I believe, is you need to understand the data, you have

to understand the system. So: 17:05

why was it built, how is it built, and who is putting data in. To understand clinicaltrials.gov, its real intention, or purpose, is to be a registry, a place to list these trials. As you mentioned, it's a popular database, mostly because you want to list your trials and conduct trials in the US to get FDA approval. And that's because your highest price for a drug will likely be in the US. So there is a lot of incentives to run these trials and have them in the US. And in this registry, I will note, there are exceptions. In Europe, the work we do, we actually have to bring in some local trials. We work with organizations like SAKK, here in Switzerland, who helped augment that data for us. And then in Asia, I think the WHO database is a bit stronger. They're all just places to list this. There's no real motivation to keep this data incredibly tidy. It's just a place to put things. That's why you see a lot of variations in terms of even within one cancer, how you talk about it. And I think the other part that adds the complexity is– clinical research is the bleeding edge of medicine. It's what we're just finding out, we're learning together. It's accelerated so much because of the data, we have the technology we have. That acceleration, while it's fantastic, it also means the way we talk about things when we understand things is also accelerating. So if you think of learning concepts, we start talking about these concepts differently very quickly. That's what makes it really difficult to have this these kind of standards in terms of how we're talking about it. Maybe one example, we're working now on nuclear medicine, which can be known as radioligand therapy, radiotheranostics, and it was, I think, seven other different names or categories for this. So if you think of searching through the data, that really blows up trying to find what you're looking for, makes it so much harder. And so that's

the important context: 19:04

understanding what you're getting. And when we first started, actually, someone told me, it wouldn't be possible to use Natural Language Processing on this data. For these reasons of complexity of the jargon. It's also unstructured and shorthand. So for Natural Language Processing, punctuation is very important because it helps us understand how we found the text and what the relationship is to other text. But when it's just bullet points or just sentences combined, and then again, it's just a registry so no one is looking at making sure the formatting is nice, you can really see anything and everything. And the point you mentioned, about looking for specific condition is really interesting one, because of how research has evolved. And one of the interesting thing is, oncology or cancer – we're moving towards precision medicine, where the specific type of cancer isn't the most important criteria anymore. And so to that woman's point, you can be looking for solid tumor trials, which just mean it has to be a solid tumor, not a blood cancer, and it has to have a certain type of molecular profile or certain mutations and certain prior therapies. And so now if you're looking for specifically in clinica trials.gov for head and neck, will it pick up all of the solid tumor trials? I have found not much.

Alex Jani, X-Health.show: 20:25

Exactly. Some of their top search results– they had this "solid tumor" or a specific gene modification.

Danielle Ralic, Ancora.ai: Exactly, exactly. So from our side, we started with just saying, Let's approach it as we can't trust anything we see. So even underlying ontology, they can also mark up what they think the trial is related to, these meta terms. And we tried to use all of them. But we had to put our own weighted ranking system for this logical check at the end, because we have this kind of zero trust approach to it. And we built our own ontology. So we sat down and said, Okay, based on how patients are treated, how do we organize cancer? And we started with that split: 20:36

solid tumors and blood. That's usually a pretty big wall. From there have organizing the data. We use that ontology to take in everything from clinicaltrials.gov and map it to our ontology to be able to overcome some of the solid tumor pieces, neuro endocrine tumors, because of the ontology used in clinicaltrials.gov erroneously brings up melanoma trials, sarcoma trials. In that example, we bring in over 700 trials. And then, with our mapping, we're able to find the 155 trials that are actually an option for neuroendocrine tumor patients. So there's a lot of work just in – first that mapping at a condition level, what really belongs to these diseases, what are real options. And then from there, you begin the eligibility criteria, and then there's quite a bit more work. How do you do this mapping? I can assume that there's actually a real person that does it, or? The first ontology was with our team, and whenever we add a new cancer, we like to do it with a scientific expert group. We're a small team, we can't be experts in every cancer. But there's a lot of smart people out there. When we sat down – neuroendocrine tumors was one of the most recent ones – we had an amazing Patient Advocate, Josh Mailmen, who brought in this amazing expert panels. So we had Thor Halfdanarson from Mayo Clinic, we had representative from Stanford, NIH, Dr. Heidi Rivera. A lot of people who've spent their careers in neuroendocrine tumors, who helped us, This is how I think about this disease, this is what's relevant, and this is what isn't. So we can kind of borrow their knowledge to build that ontology and once we have it built, then of course, we automate the extraction and the mapping of it. But it is something that has to be revisited all the time, just because clinical research is advancing so fast, which is great, because we do need these advancements. But it's not a one

Alex Jani, X-Health.show: 23:09

Let me maybe just say about, you know, and done job. because I tried your platform, it's already live, right? So patients can already go there and try it. And I'll ask some questions on the way because you're already like started, you know, answering some of them. The first step, you select the condition...