At the American Society of Clinical Oncology (ASCO) Annual Meeting in Chicago, Jon Wetzel, Chief Operating Officer (COO), and Leif Honda, Chief Innovation Officer (CIO) of TriMetis, a life sciences technology company, sat down with Oncology Data Advisor to talk about using artificial intelligence (AI) and computer-assisted pathology to revolutionize tissue-based research.

This podcast episode was recorded live at the 2022 ASCO Annual Meeting in Chicago by Oncology Data Advisor and ConveyMed. 

Oncology Data Advisor: Great. Thank you. So, let's talk about the role of quality and biospecimens in cancer treatment. Would you like to tell us about that?

Mr. Wetzel: Sure. So, I've been doing biospecimens for 20-plus years now, and what we've discovered over the years is finding a high-quality biospecimen is extraordinarily difficult for researchers. When biospecimens come on in and they're reviewed by pathologists, they're reviewed for clinical diagnostics. Was it an adenocarcinoma, or what was it from a patient care perspective? But the dilemma is that the same sample needs to be looked at from a research perspective, and the researchers look at something entirely different. They need that piece of data, but they also need information about the specific biospecimen, like what the tumor percentage is or whether there is necrosis on it.

That's where it gets really difficult from a pathology aspect. The difficulty is that pathologists have gone to medical school and everything for about 12, 14 years and now they're being asked to count beans in a jar to find out the percentages of tumor and necrosis. It's very difficult to do, it's very time consuming, and what we've learned over the years is they really don't like doing it. It has to be done because it's done for every single next-generation sequencing (NGS) test that gets performed, but it's really difficult and really time consuming.

Mr. Honda: Yeah, I would agree. I think that the interesting point of the whole thing is that when you don't have good-quality specimen, it actually affects everything downstream. If you want to conduct your research, you want to know if there's percent tumor necrosis, but if you don't have those things, it holds up all the research. It takes time to collect these samples from these patients that are very specific. The interventions are very specific time points on specific therapies. If you don't get them at the right point and you don't have good specimen, it essentially fails or delays the rest of the study.

If you need 10, 20, 100, 1,000 samples in order to get your therapy into the next level, you really need a really good high-quality specimen. Unfortunately, if that's not collected from the beginning, it's all for naught.

Oncology Data Advisor: Thank you. Is there anything else you would like to share about this?

Mr. Wetzel: One of the things that we are always looking at is actually using artificial intelligence to do that, I guess, bean counting in a way. Because like I said, it's a very time-consuming job. For NGS companies that are doing 300, 1,000 samples a day, that's basically a pathologist looking through a microscope, counting cells for an eight-hour day. We're always looking at deploying AI, artificial intelligence, to automate that process. I'm Lean Six Sigma Black Belt, so any time I can find something that does not add value, I'm going to remove it.

Oncology Data Advisor: Thank you much for explaining all of this. Anything else you'd like to share about being here at ASCO?

Mr. Honda: Well, I think it's interesting. I think it's an emerging place for us to start talking about this, and the use of AI is going to become much more prominent. I mean, quality control is just the first step in this and the research process, but then eventually it starts to grow throughout as far as tightening your diagnostics and determining what patients you want to select and things of that nature. We started off with something very simple, and then we're seeing these companies evolve and start to do very complex things that the human eyes aren't really trained to do, that the brain isn't really good at calculating.

Computers are really good at counting, they're really good at sorting and those types of things. We want to use those tools to help us when we're talking about the development of therapeutics, so that when we get that patient and we get that opportunity to give them the right therapy, it's targeted and it's precise in that they have a chance of beating their cancer.

Oncology Data Advisor: Great. Well, thank you so much for talking with us about this. It's very interesting to hear.

Thank you for listening to this podcast recorded live at the 2022 ASCO Annual Meeting by Oncology Data Advisor and ConveyMed. For more expert perspectives on the latest in cancer research and treatment, be sure to subscribe to the podcast at conveymed.io and oncdata.com. Don't forget to follow us on social media for news, exclusive interviews, and more.

About Mr. Wetzel and Mr. Honda

Jon Wetzel and Leif Honda have 49 years of combined experience in the life sciences industry, and each have created multiple life science companies. Mr. Wetzel has held senior research positions in academia and industry. Mr. Honda has held multiple senior roles in life sciences companies. Each is an inventor or co-inventor and hold several patents. Most recently, Mr. Wetzel and Mr. Honda cofounded FoundationBio, which is now a part of TriMetis.

For More Information

TriMetis (2022). Available at: Trimetisls.com

Transcript edited for clarity. Any views expressed above are the speakers' own and do not necessarily reflect those of Oncology Data Advisor.