Jamie Lewis (Host): What if biomarkers in your blood could tell doctors you have a specific type of cancer well before they turn to other methods? Dr. David Bartlett is chair of the Allegheny Health Network Cancer Institute and an internationally recognized cancer surgeon. As a pioneer in abdominal cancer treatment and a prolific researcher, he joins me today to talk about cancer biomarkers and how artificial intelligence might help doctors discover a blood test.

 This is AHN MedTalks, a podcast from Allegheny Health Network. I'm your host, Jamie Lewis. Hello, Dr. Bartlett. Thank you for joining me.

Guest: Hello, Jamie. Thank you for having me.

Jamie Lewis (Host): Let's start with this. How would cancer diagnosis and treatment change if we had a blood test to identify all cancers?

Guest: it would be very transformational to be able to diagnose cancer early to be able to screen for cancers, and to know which type of cancer that you might have so that we can identify which test would pinpoint that cancer. So it would really be a game changer and we could, get rid of some of the more cumbersome screening tools that we currently have and have screening that's applicable to all patients.

So it would really transform the way we manage cancer.

Jamie Lewis (Host): sounds extremely exciting. And currently, which kind of cancer blood tests exist?

Guest: we have cancer tests that can help us monitor whether a patient's responding to treatment or not which are specific proteins. related to that cancer cancer. We have tests like PSA for prostate cancer that can help us in screening for Certain types of cancers, but they're very limited and the accuracy and specificity and sensitivity is very limited.

So there's a lot that we can do to make this better. and the, technology is exploding around this right now. So a very exciting area of cancer research.

Jamie Lewis (Host): What are the technologies that exist for identifying cancer in the blood?

Guest: So, all cancers release different things into the blood as they're growing. So the cancer cells are dying and they're releasing DNA into the blood. They're releasing RNA into the blood. They're releasing proteins, lipids, different things. And that creates a signature that can be identified. And so that's where the technology is going.

It's... identifying these very, very amounts of DNA, RNA, protein, lipids in the blood that become a signature for cancer. And the technologies have gotten to where we can pick these things up even though they're very dilute in the bloodstream itself, but we can find them and identify them and then create a signature to diagnose cancer.

Jamie Lewis (Host): Okay, so we mentioned artificial intelligence earlier. How does artificial intelligence help us discover a blood test for cancer,

Guest: So artificial intelligence and machine learning is what can put together all of the variable data that we can collect from a single patient's blood, and then define that that's a signature for a specific type of cancer. So we're looking at. Thousands of unique proteins. We're looking at hundreds of mutations of the DNA.

We're looking at hundreds to thousands of RNA fragments. In the blood that we can identify, but how do we put all of that information together to define that this is a signature for a specific type of cancer. That's where machine learning comes in. And as we get more things that we can measure in the blood.

We can continue to learn more about and be more specific in terms of identifying what type of cancer it is and trying to identify these things as early as possible. The earlier that we pick these things up, the better chance that we can cure cancer. And so, we could really make a dent in cancer mortality by using machine learning to identify these signatures early on.

Jamie Lewis (Host): Talk to me more about those challenges. What are the roadblocks to discovering a successful blood test for cancer?

Guest: Well, there's a lot of challenges. First of all, we have to be able to get just thousands and maybe even millions of samples from patients. And these studies can be hard and can be very expensive to run, but these studies are being run. But we need the samples to then create these signatures. And so that's one of the challenges.

The other challenge is combining all of the technology. So every company has its own niche in terms of what they're going to look at to create that signature. But how can we combine all of the information so that we're getting as much information as we can from the bloodstream to identify these signatures.

So, taking these different technologies out of the silos and putting them together to create the signature, I think, is also a barrier. And then finally, the clinical data itself. A lot of times we have samples to look at, but we don't really know which patients developed cancer in six months, 12 months, 18 months versus those that never developed cancer.

And so really identifying, What happens to those patients over time is a challenge. So we have to have just a lot of data, and all of that can be a challenge to obtain.

Jamie Lewis (Host): Well, how about at AHN, what have we done to move the field forward?

Guest: So what we're doing is just trying to put all of that together, as I said, not in kind of siloed technologies, but trying to take all of the technologies that exist and create or discover what the best signature will be for cancer. And what we're asking of our patients is everybody that walks through the door, we're asking them to sign up on an IRB approved protocol that would allow us to collect all of their data, to collect their tissue and to collect blood.

over time during their cancer journey so that every month we collect tubes of blood that we can then do our analysis on. And we've created core facilities in proteomics, lipidomics, in genomics, in tissue arrays, in A lot of cores we can then rely on to analyze the samples and so combining the analysis of the sample as well as the clinical data that we have in terms of what happens to the patients and what their diagnosis is, we put all that together and we have the data analyst that will do the machine learning and artificial intelligence for identifying these signatures.

So as with many Institutions, we're just trying to be ahead of the curve in identifying this signature for blood.

Jamie Lewis (Host): Well, Dr. Bartlett, it's been such a pleasure speaking with you today. Thank you so much for helping us peer into the future of cancer treatment.

Guest: Thank you, my pleasure.

Jamie Lewis (Host): I'm Jamie Lewis, and this has been AHN MedTalks. For more information about Dr. David Bartlett and to find out more about the research he conducts, visit findcare. ahn. org slash david l bartlett. Thanks for listening.