This interview has been edited for clarity.
TwinStrand Biosciences technology, duplex sequencing, can be applied to many diverse areas. The three main buckets the company focuses on are genetic toxicology, cell and gene therapy, and minimal residual disease in cancer. IPM spoke with the company’s CEO, Jesse Salk, to learn more about how duplex sequencing can be used in the clinic, the hurdles to getting it there, and how practicing medicine informs his business decisions—and vice versa.
IPM: Jesse, you are not only the CEO of TwinStrand, but you also work as a practicing oncologist. Can you explain how, in both of those roles, you think about Minimal Residual Disease or MRD, and the current thinking in the field?
Salk: Let me start by saying that the M in MRD has become somewhat controversial, with a lot of people trying to avoid the term “minimal”. The reason being, if the M stands for minimal, and that is the amount you’re measuring, it is not really that minimal. You’re still talking about millions of cells. Some people prefer the term measurable residual disease. I don’t like that term because it implies that truth is what you can measure. So, if you’re looking for the presence of something, and you’re using two different technologies to measure it, and one picks it up and the other doesn’t—it implies that the truth is what you can measure rather than what it actually is.
Other people use the term molecular residual disease which adds a little more precision, but it doesn’t say much about it.
If I had my way, I would stick with just residual disease, but that M is really stuck in there.
IPM: Regarding cancer detection, what does TwinStrand focus on?
Salk: We have a major focus on acute myeloid leukemia (AML) right now. It’s not a very common cancer, but it is quite aggressive. It is a disease we can truly cure if we thread the needle just right–balancing just enough therapy without overdoing it. We can measure residual disease levels far higher than any other technology, detecting less than one-in-a-million. Now, we are supporting multiple prospective clinical trials with drug companies, and we are the main molecular technology measuring residual disease in a large National Institute of Health (NIH) program for thousands of AML patients that are getting treated with many different approaches. Another AML MRD consortium that we participate in was announced in February through the Foundation for the National Institutes of Health (fNIH), and entails a partnership between industry, academia and regulatory bodies to develop new diagnostic standards.
I’m excited because we’ve had a lot of traction in this area, and we recently presented some strong clinical data at the ASH Annual Meeting. And we plan to publish more studies. But we’ve shown that we are able to predict outcomes with significantly greater accuracy and precision than the standard of care—which is flow cytometry. And, our technology is more broadly applicable across laboratories, as opposed to being a very difficult thing to set up and do.
IPM: When do you predict TwinStrand’s technology making a move into the clinic?
Salk: In terms of AML MRD, I’m convinced that our approach is going to become a standard of care before too long. In terms of MRD more broadly across all oncology, I think that will soon follow, but will require more complex operationalization given that the mutational drivers in most cancers are more diverse than in AML. Right now, it’s logistically challenging, or even impractical, to do whole genome sequencing on a tumor and a normal sample from a person at the time of diagnosis to define that cancer’s specific mutations. But with new sequencing technologies that offers a faster turnaround time, and the ability to print oligos locally, it will be practical to go from whole genome, tumor-normal sequencing, to a one-in-a-million residual disease assessment in a few days from a clinically realistic amount of material in a cost-efficient way. I’m not saying that this will happen today. But I think that everything points to that being imminently realistic.
IPM: How do you know that measuring residual disease with such high accuracy actually makes a difference?
Salk: It’s critically important to have the clinical data to show this. We have to be able to show that measuring with amazing accuracy correlates with predicting X and Y and Z for patients. That’s essential. No amount of cleverness or accuracy substitutes for empirically showing that what you measure correlates with what you really care about—which is relapse or survival or response to drugs.
I put in a lot of time working with major academic groups, drug companies, regulators, key opinion leaders, and using not just our technical knowledge but our clinical knowledge, to think about how we can generate the most meaningful data for affecting patient outcomes.
In residual monitoring, it’s only useful if there’s something actionable about it. That said, there are an increasing number of things that are actionable. One can say, if I see this, I am going to put someone in a clinical study or do a more aggressive therapy. On the flip side, it is important to say, if I don’t see this, and I can be confident that that correlates with a very low risk of relapse,
I will avoid something that is very toxic.
For example, in AML, we do bone marrow transplants on most patients who can tolerate it. And in some populations that has up to a 25%, 30% mortality rate it. It cures more than it harms, but it’s not benign. And if I could say, we can avoid doing this to people who we can molecularly prove are already cured, that would meaningfully change their care.
You align a lot of interests. One always has to keep in mind the five P’s: physician, patient, payer, pharma, and policymaker. It’s nice to be operating in an area where a new technology aligns everybody’s interest. With all five of those groups—everybody gets something. Everybody wins. Who can’t get behind making cancer treatment more effective and less toxic?
IPM: What are the plans for your product?
Salk: Right now, our assay is for a Research Use Only, although in some CLIA labs, including ours, it is being used by clinical pathologists to create laboratory developed tests. When we have a large amount of strong clinical outcome data, I imagine we’ll probably go down a regulatory pathway. But the idea is to first have a robust technology that’s very consistent and can work in the real world—not just in a lab of experts. Our goal is to take something complex and make it an easy tool for people to use so that they can focus on advancing science and medicine, instead of spending their time fighting with technical details. Rather than having to be secretive and hide our data, we genuinely want our tools broadly used to renerate open-source clinical research information. We can do good science and good medicine yet have a viable business by setting other people up for success and then helping them and working with them.
I’m not in academics anymore, but now I can work with a ton of super smart, engaged people and help them do what they’re trying to do. I think this is how we will move the needle on cancer care and cancer research in a way that’s more than any single one company or academic lab could alone.
IPM: How will your technology become adopted in hospitals?
Salk: The adoption is relatively easy. There’s no special equipment needed beyond that which NGS-capable labs already have. It comes in the form of kits with a set of reagents and access to cloud-based analysis software. So, anyone that has a sequencer can set this up. And that’s exactly the idea.
Historically, scaling up NGS assays has been a big barrier. People either sent something to a one-stop outside reference lab that controls everything, or they had to build everything from scratch, which is incredibly complicated. So being able to get things out quickly that are easy for an average user to work with, without needing special equipment or to train a bioinformatician, has been very successful.
Residual disease monitoring can follow a patient longitudinally over time. Speaking from the perspective of an oncologist, we commonly do CT scans about every three months post-treatment for many types of solid tumor monitoring. It’s a big market opportunity to move that process to a more objective, molecular, method. When I hear “market opportunity,” I think opportunity to affect change and to improve patient care. It’s something that I think could easily benefit around half (or more) patients with cancer.
We have to do the rigorous clinical studies first to prove utility, but I’m convinced that this piece is going to be a very important one. I think we’ll see the practice of oncology fundamentally change over the next decade to incorporate routine MRD monitoring.
IPM: Jesse, you have a unique vantage point as a CEO and a practicing oncologist at the VA medical center. Can you speak a bit as to how wearing both hats guides your decisions about where your technology is going and how it will be implemented to help patients?
Salk: It is easy to look at data in a laboratory and say, “Oh, this makes sense. We should do that.” But the reality is that things are much more complicated on the ground and it’s rarely black-and-white. There are cancer patients who have a lot of medical comorbidities, where there is a real risk of giving an aggressive therapy. You can make the situation worse if not exceptionally careful. It’s not as simple as saying, “there’s a standard of care so let’s just do that.” Physicians must make infinitely contextualized decisions that will result in the least harm and the most likelihood of success and is in line with a patient’s goals and values. And a lot of times, there is no easy tie breaker. I generally know what I would like to do, but there may be a significant risk and the exact balance in each patient is often frustratingly hard to estimate. Having more information to put into that otherwise subjective equation is much more important, I think, than people who do pure laboratory science recognize.
My patients at the VA have a lot going on. Very few have no other medical problems, in addition to their cancer. Many have complex social situations or live far away. So, for me to be able to avoid an unnecessary or unhelpful treatment or direct them to a better one could vastly improve their quality (and length) of life.
But being able to give a good reason to do something that is rooted in objective data about their personal situation, not just population averages, helps people make decisions. It helps me make recommendations and it helps provide a level of confidence that we’re collectively making a good decision. That is the most significant thing that I get out of practicing medicine. It is being able to see how gray and challenging some of these decisions are in the real-world environment.
If you could come into clinic with me, as a fly-on-the-wall, and see what we discuss, I think it would be very, very different than what a lot of people imagine. There is so much social and life context that goes into it. Patients ask, how is this going to affect my job? How is it going to affect my family? How should I tell somebody? Is my hair going to fall out? It’s not, what is the hazard ratio of the Kaplan-Meier curve in the New England Journal of Medicine paper? The patients in those trials are a lot more idealized in terms of comorbidities. My colleagues and I hardly treat anybody that looks like a patient in a clinical trial, so we have to do a lot of extrapolating.
Practicing medicine grounds me and it keeps me nimble. But the combination can be funny. Here’s a little anecdote of the sometimes bizarre juxtaposition. On the day that we were closing a big, maybe $50 million financing at TwinStrand, I had to fill out a ton of paperwork to coordinate the wire transfer. As I was carefully counting the number of zeros, I got a page that one of my patients needed their constipation medication refilled. Even though sometimes it’s hard jumping back and forth, I’m grateful for the regular reminders of what the most important things are to focus on. It not only puts things in perspective, it helps me make better decisions.