From the Human Genome Project to 48 Hours: The Rapid Evolution of Tumor Biology

Most treatments and diagnostics once took years and hundreds of millions of dollars. Today, it costs less than $100,000 to sequence a genome and open the door to tailored cancer therapies that could save lives. But understanding tumor biology is only half the battle. The real challenge lies in interpreting complex genetic data quickly, accurately, and at scale.

In this episode of AVBCC’s Value-Based Voices, James Hamrick, MD, MPH, Chair, Precision Oncology Alliance at Caris Life Sciences, teams up with John Fox, MD, MHA, Senior Medical Director for the Americas, Illumina, to explore how AI-driven bioinformatics is making genomic insights accessible and actionable. They reveal concrete strategies to overcome systemic hurdles such as inconsistent testing, reimbursement issues, and integrating insights into busy clinical workflows, that currently hinder widespread, equitable adoption of precision medicine.

Discover:

• How rapid DNA and RNA sequencing is transforming tumor profiling
• The critical role of bioinformatics and AI in interpreting complex genomes for real-world use
• Why comprehensive multi-omic panels and gene signatures are changing treatments and clinical trials
• The systemic, policy, and reimbursement barriers slowing progress, and what can be done to break them down
• How timely, precise testing can reduce costs, improve survival, and personalize treatment

This isn’t just about technology, it's about unlocking a new era of value-based, patient-centered cancer care that can save lives and resources. 

Resources & Links:

• Illumina
• National Comprehensive Cancer Network (NCCN)
• Caris Life Sciences
• Connect with Dr. James Hamrick:
• LinkedIn
•  Connect with Dr. John Fox:
• LinkedIn

Contact Value-Based Voices

- Follow AVBCC on LinkedIn  

- View our podcast lineup 

- Contact us at info@avbcc.org

Thanks for listening! 

Transcript

James Hamrick (00:00.3)
Welcome to Value-Based Voices, a podcast from the Association for Value-Based Cancer Care. Each episode dives into the shifting terrain of cancer care in the United States, exploring what value means in today’s clinical, policy, and patient-centered environments. Our mission is to spark informed dialogue, promote transparency, and equip every stakeholder—from providers to payers to patients—with the insights they need to navigate cancer care with clarity, confidence, and purpose.

James (00:33.358)
So let’s get into it. First of all, I’ll introduce myself. My name is James Hamrick. I’m a medical oncologist. I’ll be your host today. I currently serve as the chair of the Precision Oncology Alliance at Caris Life Sciences, a molecular diagnostics and AI company. Today, we’re going to be talking with John Fox—he’ll introduce himself in just a minute—about some of the incredible progress that’s been made in understanding tumor biology and optimizing cancer care in the diagnostic space, along with some of the hurdles, challenges, and opportunities we’ve got in that space.

So, John, welcome this morning, and I’ll let you go ahead and introduce yourself.

John Fox (01:12.706)
Sure, thanks, James. John Fox. I’m a pediatrician in epidemiology. I’ve spent 20-plus years working for health plans developing medical policies, but for the last four years, I’ve been the senior medical director in market access for Illumina, which is the world’s largest manufacturer of sequencing equipment.

James (01:29.186)
Thank you, John. And that’s really impressive. I’m old enough in my career to have witnessed—I remember the Human Genome Project and the incredible effort that that took from so many in the scientific community to sequence the first genome. It took many years and many, many, many dollars. It’s just amazing to see what we can do now. Maybe walk us through a little bit about what’s possible now in terms of understanding tumor biology with sequencing of DNA and RNA.

John (01:59.118)
Okay, so I’m not an oncologist—I’m just a pediatrician—so I’m not sure I’m going to hit the mark. But it was 2003 when we first sequenced the human genome, and I think the publicly reported cost of that was $100 million. Fast forward 22 years—that’s for one genome, yeah—and fast forward now, and we can sequence 200 genomes in under 48 hours at a cost of about—well, I probably shouldn’t say the cost—but significantly less than $100,000. Three to four orders of magnitude less than that first genome.

James (02:30.412)
Right, yeah. And I think it’s really fascinating. Part of the learnings I’ve had on my journey is there’s the sequencing, which is sort of a science fiction miracle—to be able to do it at that scale.

John (02:40.302)
3.2 billion base pairs.

James (02:56.722)
The other piece, though, is when you sequence, you have a sequence of A’s, C’s, and G’s. And what many people don’t totally understand is that there’s interpretation and analysis of that—what we call bioinformatics. That’s really a part that is very nuanced and still has tons of opportunity, but also tons of challenges.

John (03:13.762)
Yeah, and you pointed out in your introduction that Caris is an AI company. So, is Illumina, we’re using AI to simplify the interpretation and leave the final touches to laboratory physicians or others. That has really simplified the process and sped up getting information into the hands of clinicians.

James (03:34.45)
Yep. So with my clinical hat on, during my career I was able to witness the first sort of rationally designed targeted therapy—what many people would say is a drug called imatinib for chronic myeloid leukemia. I had a patient in my fellow’s clinic who was alive and being treated by James Hamrick as a lowly intern because of this amazing drug. I didn’t take any credit for it—someone else had actually treated the patient—but it was amazing to see.

Now we’re seeing this explosion of innovation. But let’s talk a little bit about how, as with any major leap forward in science—whether in diagnostics or therapeutics—not everyone can benefit right away. You’ve done a lot of work in the field of appropriate use of testing. I’d love for you to sketch out for listeners what we mean by that—how this technology lands and doesn’t necessarily get deployed and implemented everywhere.

John (04:26.798)
Well, first of all, it’s safe to say that with any new technology, our ability to implement it lags behind our ability to execute it. Meaning the process today of running whole genome sequencing or whole transcriptome sequencing is pretty much cookie cutter. Interpretation adds some complexity, but implementation into clinical practice—ensuring patients benefit in terms of improved outcomes—varies widely across the country.

For example, in non-small cell lung cancer, about 50% of patients who should be tested aren’t. The consequences are profound. There are about five papers suggesting that patients who don’t get appropriate testing and targeted therapy have a survival deficit of between 8 and 14 months. If we had a new drug that improved life expectancy by that amount, everyone would be on it.

The issue is we don’t have systems to ensure physicians order the test, receive results, interpret them, and get patients on the right therapy. It’s changing, but slowly—and it’s tragic that we have the ability and yet not every patient has access to these results.

James (05:33.058)
Yeah.

James (06:01.89)
You really see that play out. I spend a lot of time traveling around the country, seeing cancer care in large health systems, academic centers, and community practices, and what you say absolutely rings true. What do you see as some of the barriers? For someone unfamiliar, it might seem plug-and-play—but what are the cultural or systemic barriers?

John (06:30.702)
In 2022, Helen Sattuck and Darrell Pritchard from the Personalized Medicine Coalition published a paper outlining barriers in non-small cell lung cancer: not ordering tests, not sending samples, not getting results in time, or not getting them at all. There are also reimbursement barriers.

Tests can cost $3,000 to $4,000, and if physicians aren’t sure they’ll be reimbursed, they might order something else. Ironically, many insurers only cover panels of fewer than 50 genes, which don’t capture all biomarkers recommended by NCCN.

The fundamental question is: what are the appropriate therapies? Not testing for all relevant biomarkers leaves you in an information void.

James (07:50.894)
That’s absolutely true...

James (07:50.894)
I mean, it’s absolutely true. So, you often find yourself in a position where you really want to, from a scientific standpoint, of course, find out what are the actionable alterations. Is there another line of therapy that might extend the life of this patient or even save this patient’s life?

Even if they were static—for a disease like non-small cell lung cancer, there are seven to 12 biomarkers. You can quibble about what the science is to back them, but guidelines supported, they move very quickly. I know at Caris, and in my practice, when I’ve seen patients, there was a time between having the tissue sample or the blood sample available—usually tissue, because that’s the hard one to get—you really want to be as comprehensive as you can from the outset.

Because it tells you—it gives you the best picture of what’s going to happen with the patient. At the same time, I understand the reasoning behind saying, well, if you’re a payer, why would you want to pay for hundreds of alterations when there are only eight that are actually on the NCCN guidelines?

There’s an operational conundrum as well. It would be one thing if you could walk in, press a button, and get the one next biomarker that’s most valuable automatically delivered for a fixed price all the time. That’s just not the way the world works.

John (09:13.932)
Yeah. And then in the FDA approval process for companion diagnostics, you can add additional biomarkers if that gene was already in the test. But, for example, if you didn’t have RNA in your test, you can’t add that to your DNA-based test without voiding the companion diagnostic approval.

The challenge with diagnostic tests, as you point out, is you simply can’t add an additional biomarker. You have to go through a whole analytic and clinical validation again. And so the notion of a large panel is that you’re going to future-proof your test in the sense that things that might come in the future are already built into the test.

James (09:51.892)
Yeah, I mean, that is sort of the approach. I know at Caris that’s the approach we’ve taken—to be comprehensive from the outset: whole exome on the DNA side, whole transcriptome on the RNA side, and even moving toward whole genome.

And we’re not the only ones, but I think that is this philosophical approach of future-proofing—saying, look, it’s better to do the entire run because you may find things that, even for matching for a clinical trial…

Or now that patients are living years and years with these diseases—thanks to advancements in diagnostics and therapy—several new biomarkers may come about. And then, of course, there’s the other piece of having informed patients who have the best understanding they can of what’s going on.

John (10:40.03)
Yeah, you brought up the topic of clinical trials. One of my jobs is to work with payers to convince them of the value of comprehensive testing.

One of the things we point out is that if you want to know all the treatment options for many cancers now, clinical trials are a first- or second-line option. And if you don’t test for those, you don’t know.

If you wait to test until the patient has progressed, the likelihood of enrolling in that trial is very low because the patient wants to get treatment. We can’t wait six to eight weeks to find out eligibility, get them vetted, and enrolled.

So, the notion of a comprehensive profile allows for identification of both FDA-approved drugs and identification of patients for clinical trials.

James (11:30.412)
Yeah, I don’t know the exact number, but I’ve seen reports that something like 60-plus percent of oncology trials are biomarker guided.

Any medical oncologist who’s seen patients exhaust standard therapy—and then, if you send them to a phase one center, the first thing they say is, “We need a comprehensive panel.” Because we can’t even begin to understand trial opportunities without that.

So you mentioned value, though. It’s a challenge to demonstrate that by understanding tumor biology as early as possible, we’ll make better decisions and ultimately drive better value—keeping patients out of the hospital, reducing toxicity, fewer therapy changes, and better outcomes.

What have you seen in terms of actually demonstrating value?

John (12:53.582)
Yeah. The studies I cited earlier showed patients on the wrong therapy have a survival deficit. There’s also financial waste.

If you look at a PD-L1–based triplet therapy, that’s about $30,000 a month. If someone is on that for six months before we recognize it’s not working, that’s $200,000 spent on the wrong therapy.

So there is real value in ensuring patients are on the right therapy. The irony is we’re debating spending $800 to $3,000 on a diagnostic test when the real cost is in the therapy itself.

James (13:53.784)
Yeah, that makes sense. Another interesting area is minimal residual disease. Doing the right test at the right time can actually lead to de-escalation in therapy.

Some patients can avoid unnecessary treatment, toxicity, and cost if they’re negative on MRD testing. Others may need escalation if positive.

So what do you see in terms of adoption of this logic?

John (14:50.958)
Right. The prototype was the Oncotype DX test in breast cancer. It identified patients who didn’t benefit from chemotherapy, saving both cost and toxicity.

That’s the promise of MRD—identifying patients who don’t need therapy. And payers adopted Oncotype DX quickly because it reduced costs.

I think we’ll see broader adoption of MRD as more studies demonstrate value, especially where it reduces unnecessary treatment.

James (16:41.526)
Yeah, and Oncotype DX represents another element—gene signatures rather than single biomarkers.

Now, with comprehensive testing and AI, we’re seeing signatures like MSI, TMB, and others evolve. They help guide complex decisions like first-line therapy.

That’s another proof point for comprehensive testing.

John (18:02.906)
Yes, and the challenge is explaining complex science in terms payers understand.

We have the technology and interpretation, but physicians still need tools to know when to order tests.

We’re working on solutions to answer three questions:

1. What biomarkers are recommended?
2. Which tests cover them?
3. Will the payer cover it?

If physicians had that, it would remove a huge barrier.

James (19:38.792)
I love the practicality. Integrating this into a busy oncologist’s workflow is key.

John (20:01.1)
AI-enabled EMRs may help. If they improve outcomes and physician experience, adoption will follow—even if costly.

James (20:42.19)
Absolutely.

John (21:04.364)
It gives me hope that new technologies will enable precision medicine at scale.

James (21:18.702)
Very exciting—and critical to ensure equitable access.

John, thank you so much for sharing your insights.

John (21:17.486)
You’re welcome. This has been a great conference.

James (21:18.702)
I really enjoyed it. Meetings like AVBCC inspire you to go back and do something differently.

This wraps up this episode of Value-Based Voices. A heartfelt thank you to John Fox and to our listeners.

If you found today’s discussion thought-provoking, be sure to subscribe wherever you get your podcasts. Stay tuned for more episodes spotlighting the voices shaping oncology.