Can Blood TMB Predict Immunotherapy Response?
Tumor mutational burden (TMB) is an important biomarker for predicting response to immunotherapy, but should clinicians rely on blood-based or tissue-based testing? In this episode, Suneel Kamath, MD, discusses research comparing blood and tissue TMB across solid tumors and their ability to predict immunotherapy outcomes. He explains why blood TMB levels are often higher than tissue TMB, why tissue testing remains the stronger predictor of treatment benefit and how disease-specific blood TMB thresholds could improve clinical decision-making in the future.
Advertisement
Cleveland Clinic is a non-profit academic medical center. Advertising on our site helps support our mission. We do not endorse non-Cleveland Clinic products or services. Policy
Recent Episodes
Building Better Cancer Care Through Global Collaboration
In this episode, Dr. Dale Shepard speaks with Dom Kosmina, chief medical physicist and chief operations officer at Radiochirurgia Zagreb in…
Adaptive Radiation Therapy: How AI Is Personalizing Cancer Treatment
Adaptive radiation therapy is changing how radiation is delivered. In this episode, Dr. Ehsan Balagamwala discusses how the Ethos platform u…
Cancer Survivorship in the Era of CAR-T and Bispecific Therapies
As cancer therapies continue to evolve, so do the long-term needs of survivors. In this episode, Dr. Dale Shepard sits down with Christina F…
Video content: This video is available to watch online.
View video online (https://cdnapisec.kaltura.com/p/2207941/sp/220794100/playManifest/entryId/1_9ksurek2/flavorId/1_5f3sgelj/format/url/protocol/https/a.mp4)
Transcript
Dale Shepard, MD, PhD:
Cancer Advances, a Cleveland Clinic podcast for medical professionals, exploring the latest innovative research and clinical advances in the field of oncology.
Advertisement
Cleveland Clinic is a non-profit academic medical center. Advertising on our site helps support our mission. We do not endorse non-Cleveland Clinic products or services. Policy
Thank you for joining us for another episode of Cancer Advances. I'm your host, Dr. Dale Shepard, a medical oncologist and co-director of the Sarcoma Program at Cleveland Clinic. Today, I'm happy to be joined by Dr. Suneel Kamath, a GI medical oncologist here at Cleveland Clinic. He has previously been on this podcast and in fact been here now his fifth time. His previous episode was metabolomic differences between young onset and average onset colorectal cancer, and this and other episodes are still available for you to listen to. Today, he's here to talk about TMB concordance between blood and tissue and immunotherapy response. So welcome.
Suneel Kamath, MD:
Thank you. Hey, good to be back.
Dale Shepard, MD, PhD:
Absolutely. Give us an idea again what you do here at Cleveland Clinic.
Suneel Kamath, MD:
Yes, yes. I am a GI medical oncologist focused mainly on hepatobiliary malignancies, upper GI malignancies as well, but as you can see here, I have a number of different research interests. I dabble in a whole lot of other things on the side, too.
Dale Shepard, MD, PhD:
Excellent. Keeps things interesting, right?
Suneel Kamath, MD:
Absolutely. Yes.
Dale Shepard, MD, PhD:
All right. We're going to talk about TMB, concordance between blood and tissue, immunotherapy response, lots of things. People might be listening in, different backgrounds, so let's just start basics. TMB. What is TMB?
Advertisement
Suneel Kamath, MD:
Yes. TMB is tumor mutation burden, so it's sort of the burden of somatic alterations essentially in a particular tumor. We've known for a while that it is a predictive biomarker for immunotherapy. We definitely think for, for example, the MSI-high population, patients with POLE or POLD1 mutations that the high TMB burdens seen in those particular alterations drives that response to immunotherapy. We also have a tumor-agnostic approval for pembrolizumab based on tissue-based TMB being greater than 10 mutations per megabase. So it's definitely an important, I think, biomarker really across solid tumors.
Dale Shepard, MD, PhD:
Excellent. Then we're going to talk about the difference between tissue testing and blood testing, so give us a little bit of an idea what that entails.
Suneel Kamath, MD:
Yeah, definitely. I think that's really important. Of course we've had tissue-based NGS for many, many years, and also some years later, these circulating tumor DNA assays came about as well, many of them from the same company. So really the same idea, sequencing both DNA and RNA, one in tissue, one in circulating tumor DNA from blood, really to look for the same kind of genomic alterations, biomarkers.
But I think your key here is that what cutoffs might be in terms of variant allele frequencies, what TMB cutoffs there might be can vary between your tissue and blood. Certainly, as an example, we've known that for circulating tumor DNA, ctDNA, there are big variations in terms of site of metastatic disease and tumor type. For example, pancreatic cancers tend to not shed nearly as much as, say, a colorectal cancer or a lung cancer might. As far as site of spread is concerned, liver metastases generally tend to shed a lot more versus on the other end of the spectrum, peritoneal carcinomatosis rarely sheds circulating tumor DNA. So, we know that the blood can have a number of caveats that we do need to pay attention to.
Advertisement
Dale Shepard, MD, PhD:
All right. What led you to think about doing this comparison and looking at the concordance in some solid tumors?
Suneel Kamath, MD:
Yeah, this project was really driven, as I think everything, we do should be, by a patient. I had a gentleman that we had run out of options for, and we did the usual ctDNA roulette, I call it, where we checked ctDNA just to see, is there something we could maybe try to target? As we often do if you run out of standard of care, there's no trial options, let's just see what we can get. For this gentleman, nothing specifically targetable like a HER2 or something was identified, but the TMB was elevated in the blood. I knew it wasn't in his tissue testing because we'd already done that in the past, but I said, "Yeah, let's just try it. It's the only thing we have."
We did pembrolizumab, and it didn't work as I would've expected in that particular histology. It wasn't particularly effective. So, it got me thinking, how much can I really trust this blood TMB? I was already wondering going into it, is the tissue going to be the driver of what happens or the blood? It turned out to be the tissue in this case, and so I said, "Why don't we look into this to see is that true really across the board in other cancers?"
Advertisement
Dale Shepard, MD, PhD:
I guess there's always this question about blood versus tissue testing anyway. We worry about sensitivity with blood tests, but fundamentally the blood test you think is more indicative of the entire tumor burden rather than where you just happen to put a needle. I mean, I'm always torn sometimes, which one should I really be looking at?
Suneel Kamath, MD:
Agree. Yeah, it's a conundrum, you're right, because tissue is that particular part of that particular tumor. We know that there is clonal evolution that happens as things metastasize, those other sites are going to evolve in a different way genomically. So that is the power of ctDNA. You can really capture all of that. It's tricky, I think, for some of these other biomarkers like TMB.
Dale Shepard, MD, PhD:
Patients typically want all of their tumors treated, not the one you just happen to have a needle in.
Suneel Kamath, MD:
Exactly. Yes, absolutely.
Dale Shepard, MD, PhD:
When you did these comparisons, how exactly did you go about comparing?
Suneel Kamath, MD:
Yeah, so what we did was we looked at, our precision medicine group has a really nice database, we looked at patients who had both blood and tissue-based TMB, regardless of vendor that was used to perform it, and then also looked at the subset of patients that had received immunotherapy. This was really any solid tumor patient essentially that was treated here at Cleveland Clinic.
Advertisement
Dale Shepard, MD, PhD:
How many patients were looked at in terms of having both tissue and blood tests?
Suneel Kamath, MD:
We had a total of 221 patients that were included. About 80 of those had received immunotherapy. For the one part, we looked at all 221 for the degree of concordance between blood and tissue TMB, and then of course for those, looking at predictive response immunotherapy, we looked at that subgroup of 80 patients.
Dale Shepard, MD, PhD:
And then what surprised you most?
Suneel Kamath, MD:
What we found, it was known in the literature from a couple of smaller studies that blood TMB tends to be higher than tissue TMB. Most of the studies have shown it's around two to two and a half times higher. We found a similar thing in our study that blood TMB was about 2.3 to 2.4 times higher than tissue TMB.
What I thought was interesting about that, though, was it was very concordant still. So even though it was quite a bit higher, it did match the tissue in a way. If it was eight in the tissue, it would be 24, 26 in the blood. If it was three in the tissue, it'd be more like eight or so. So, it did line up still, which I didn't exactly expect. I though that it might just be a lot higher and there would be a lot more variability there, but the concordance was around 80, 85%, so it actually was pretty tight even though it was consistently higher.
Advertisement
Dale Shepard, MD, PhD:
What does that lead you to think, that we should just have higher cut points? If you look at Pembro approvals, 10. Should we just assume that unless it's a couple of times like 20 or higher, that if we're going to use a blood-based test or, where do we go from a practical standpoint?
Suneel Kamath, MD:
I think we have to find a way to use it. I agree, I think the way forward is to identify higher cutoffs, and really disease specific cutoffs, too. Even if you look at that trial that led to the TMB approval for pembrolizumab, we know that a number of very common diseases were not as well represented in that trial. It wasn't like there were 5,000 patients in it. It was a couple hundred, I think. We know that it's unlikely that that trial really proved a good cutoff, even for tissue, actually. I'm sure there's variability there in blood as well. I think we need to look at those numbers specifically for different tumor types.
I love that there is this narrative out there that we have these biomarkers, they're tissue -gnostic, site of cancer doesn't matter anymore, we're trying to go to this precision approach. I definitely think it's somewhere in the middle. I don't think that we can totally ignore the underlying histology. We see this often in pancreas cancer and esophageal, the same drugs used in melanoma or in breast cancer or lung will not work as well in a lot of GI cancers, even though the target is exactly the same. I think there's going to be some kind of middle ground where finding a cut point that's different and disease specific I think is really the way forward.
Dale Shepard, MD, PhD:
You've mentioned a couple things about different histologies with different shedding of DNA, different histologies with different sort of TMB and how it correlates and things like that. There's not all that much out there about that. I'm not picking on the testing companies, but they just like, "Hey, this is a test. Use it for your cancer." Should there be different guidance? Should we be understanding that a little bit better?
Suneel Kamath, MD:
Yes, we definitely should. That was another really interesting part of doing this project was we of course do the usual literature search, see what's out there already and see where we can build upon, and there was nothing. There's really no foundation in this at all, which is very surprising for how long that approval has been available, how long TMB testing has been available. Multiple companies have thousands of cases, really, with both matched blood and tissue TMB.
I do find it interesting that that hasn't been looked at before. Certainly, there've been certain areas, I know for example for colorectal cancer, there have been some retrospective studies looked at, but these are mostly single center things where they've looked at their own institutional experiences with TMB for blood and tissue, and they have shown, similar to what we're talking about, that you need a much higher blood TMB cutoff, and those studies, at least 20 or higher, sometimes even more.
I think it has been done occasionally, but on a global scale across solid tumors, certainly we may not be able to look at it for every single cancer. Even for companies that are large, maybe they only have 20 or 30 cases with a papillary thyroid cancer, for example, or some of the more rare ones maybe, but certainly for the common ones, colorectal, prostate, breasts, definitely I think we should have enough numbers to look at this.
Dale Shepard, MD, PhD:
Is there a cautionary tale here where, we're kind of a quirky crew sometimes, so we will look for very defined data, but then we lump things in other situations, so if you really are a stickler for detail, if you look at the Pembro approval, it says 10 or higher with an FDA-approved test, which really means a companion diagnostic, not just mean any test, which is really only one company's tissue test. But in common practice, it's like any test goes. I mean, should we be paying a little bit more attention to this?
Suneel Kamath, MD:
Definitely, yes. I think that's a big, big problem, especially as more and more players come into this market. There's so much interest in it and so many people are developing NGS assays, and a ton of other nonmedical factors really go into which ones we use. The cost of them, ease of ordering, insurance approvals, all of these things really are the big drivers for which test we adopt. It's not so much the clinical performance characteristics. I like to think that it would be the medical things, but it really isn't. Even for us, which vendors we're using is mostly based on ease of use and cost factors often.
So yeah, I definitely think that needs to be looked at for individual companies for sure. Also, for both the tissue testing and the blood testing, ours clearly show that looking at time to treatment failure, it was definitely better predicted by high tissue TMB. That's really what you needed. Having a high blood TMB, but if your tissue TMB was low, really didn't predict that great of a response to immunotherapy. So, as we tell our fellows, tissue is the issue, seems it still is for TMB as well until I think we identify better blood-based cutoffs.
Dale Shepard, MD, PhD:
So, in many ways, given the fact that the tissue is so much more predictive, in those cases where you're looking for these options, you have people with metastatic disease, do you think we really should be, if we're going to be doing genomics, do we really do patients more of a service by getting another actual biopsy rather than do the blood test?
Suneel Kamath, MD:
Yeah, I think for now with what we know today, yes, I think getting a repeat biopsy is truly the best way. And it fits the label, as you already mentioned. The label really is based on a single vendor's tissue-based TMB assay, but it's not based on blood TMB. I think until we identify really what those specific cutoffs are based on ctDNA, I think we do really need to rely on tissue.
Whether that's the biopsy you already had in the past... That's another interesting question you raised is how much does this change in the tissue over time, because we would hate to put people through invasive procedures if you're likely to find the same results, so I think that's another great question is, if your biopsy was four years ago, you're getting a fresh one now, is it really going to change over time? I think that's really what we would need to do, assuming that it's in a safe location to be able to do that biopsy.
Dale Shepard, MD, PhD:
But as you mentioned, we know that tumors, there's clonal evolution, so I mean it seems reasonable to believe that there could be acquired mutations that would raise that TMB. And you may not have been able to look at this as based on the numbers, but in cases where there was maybe a low TMB, but the blood test, probably newer, showed a higher TMB, did those tend to be older samples? So back to that point, over time, does the TMB increase to any clue that the blood test might've shown that?
Suneel Kamath, MD:
Yeah, great point. Yes, it typically was, I think on average, these were about a year apart or so, even a year and a half. That is a thought we had as well that maybe it is something that's evolving with time. We just didn't have enough numbers really to have those that were truly lined up to know if it was really just a change in time or was it really blood versus tissue.
Dale Shepard, MD, PhD:
What's next?
Suneel Kamath, MD:
I think our next steps are, we really want to expand this to larger data sets, so we're reaching out to a couple of industry partners to see if they're interested in us looking at their databases, which include thousands of patients, really, and many of whom have received immunotherapy, to see, can we really look at and validate the same results?
Also, I think what's really going to be important there is I think we'll have adequate numbers for some of the more common tumor types, lung cancer, breast cancer, colorectal, prostate, to see if there is a different cut point we can identify with blood TMB. Because certainly we can always get another biopsy, but usually at the point that we'll be thinking about these things, these patients are very sick, we're often sort of facing the choice of, are we going to continue treatment at all or go into a more comfort-oriented approach?
That's not a population that it's a great idea to be doing invasive procedures for, and they're often very sick, and so I think it would be amazing if we could identify a blood-based cutoff that we could really rely upon. That's really the hope is that we can look at a much larger data set, validate our current findings and maybe identify these disease-specific cut points.
Dale Shepard, MD, PhD:
Very good. Well, appreciate you coming around with your insights today for the fifth time, and so we'll look forward to seeing what the sixth topic will be.
Suneel Kamath, MD:
Definitely, yes. Always great to be here.
Dale Shepard, MD, PhD:
To make a direct online referral to our Cancer Institute, complete our online cancer patient referral form by visiting clevelandclinic.org/cancerpatientreferrals. You will receive confirmation once the appointment is scheduled.
This concludes this episode of Cancer Advances. For more podcast episodes, visit our website, clevelandclinic.org/canceradvancespodcast. Subscribe on Apple Podcasts, Spotify, or wherever you listen to podcasts.
Thank you for listening. Please join us again soon.
Recent Episodes
Building Better Cancer Care Through Global Collaboration
In this episode, Dr. Dale Shepard speaks with Dom Kosmina, chief medical physicist and chief operations officer at Radiochirurgia Zagreb in…
Adaptive Radiation Therapy: How AI Is Personalizing Cancer Treatment
Adaptive radiation therapy is changing how radiation is delivered. In this episode, Dr. Ehsan Balagamwala discusses how the Ethos platform u…
Cancer Survivorship in the Era of CAR-T and Bispecific Therapies
As cancer therapies continue to evolve, so do the long-term needs of survivors. In this episode, Dr. Dale Shepard sits down with Christina F…
Other Podcasts You May Love
Beyond Leadership
Hosts Jim Pae and Elizabeth Pugel escort you through a network of thought leaders, sharing world-cla…
Beyond the Pages: CCJM Podcast
Beyond the Pages: CCJM Podcast takes the listener more in-depth into Cleveland Clinic Journal of Med…
Butts & Guts
A Cleveland Clinic podcast exploring your digestive and surgical health from end to end. You’ll lear…
Cardiac Consult
A Cleveland Clinic podcast exploring heart, vascular and thoracic topics of interest to healthcare p…