JCO PO author Dr. Bryson Katona shares insights into his JCO PO article, “Outcomes of the IMMray PanCan-dTM test in High-Risk Individuals Undergoing Pancreatic Surveillance: Pragmatic Data and Lessons Learned.” Host Dr. Rafeh Naqash and Dr. Katona discuss IMMray PanCan-d:  A Blood-based Test for Early Detection of Pancreatic Cancer.

TRANSCRIPT

Dr. Rafeh Naqash: Hello and welcome to JCO Precision Oncology Conversations, where we bring you engaging conversations with authors of clinically relevant and highly significant JCO Precision Oncology articles. I'm your host, Dr. Rafeh Naqash, Social Media Editor for JCO Precision Oncology and Assistant Professor at the OU Stephenson Cancer Center. 

Today, we are thrilled to be joined by Dr. Bryson Katona, Director of Gastrointestinal Cancer Genetics Program and Assistant Professor of Medicine at University of Pennsylvania Abramson Cancer Center, and also lead author of the JCO Precision Oncology article, “Outcomes of the IMMray PanCan-d tm Test in High Risk Individuals Undergoing Pancreatic Surveillance: Pragmatic Data and Lessons Learned.” 

At the time of this recording, our guest disclosures will be linked in the transcript. 

One of the other unique things about this article is that this will be a concurrent podium presentation at the 2023 CGA-IDC Meeting and also concurrent publication in JCO Precision Oncology. Bryson, welcome to our podcast and thank you for joining us today.

Dr. Bryson Katona: Thank you so much for the invitation and I'm looking forward to discussing the findings of our recent study.

Dr. Rafeh Naqash: So, one of the unique things about this topic or this publication is surveillance, which we often talk about a lot in lung cancer, breast cancer, colorectal cancer, but not so much in pancreas cancer. For a listener who might be a community oncologist or for a trainee out there, fellow, who may not be as specialized in cancer genetics, what are the kinds of individuals that you would screen for pancreatic cancer? What are the high risk populations where there's a decent pretest probability that you could catch something early? And what are the implications of catching something early in people with pancreas cancer?

Dr. Bryson Katona: The whole field of pancreatic cancer screening and high risk individuals, which we'll refer to as pancreatic surveillance, has just been an area of increasing growth and increasing data over the last five years or so. For individuals that we would consider pancreatic cancer surveillance in, really, we try to target those individuals that have a lifetime risk of pancreatic cancer of about 5% or more. And so those fall in two main groups. So one of those groups are individuals that would have what we would define as familial pancreatic cancer. That's where there's a very strong family history of pancreatic cancer without a known genetic susceptibility. And so by strong family history, what we typically refer to is having two members of that family with pancreatic cancer. And those two members of the family have to be directly related to one another, and your patient would then have to be directly related to one of those individuals. So, say your patient has a father and a brother or a mother and a maternal grandfather. Those would kind of fit the criteria. 

The other big group of patients that we consider doing pancreatic cancer surveillance in are those with genetic susceptibility to pancreatic cancer. And this can come in multiple different forms. Some more of the common cancer risk syndromes, such as Lynch Syndrome, and carriers of a BRCA1 or BRCA2 mutation. Typically in the presence of one family member with pancreatic cancer, those individuals could be eligible for surveillance. But there are other rarer syndromes as well, such as carriers of an ATM or PALB2 mutation, individuals with Peutz-Jeghers syndrome or hereditary pancreatitis as well. 

Dr. Rafeh Naqash: My understanding of these mutations is they have to be germline in these individuals. Could you shed a little more light on germline mutations, what they are, how potentially to recognize them? Because in the current setup, at least, most patients with advanced cancer get next generation sequencing, which is not the same as germline testing. So, for somebody who sees these reports, I do early phase clinical trials primarily and do a lot of sequencing, and find a lot of interesting things which lead to germline testing in individuals, could you shed some light on the differences between NGS testing and germline testing, since you do this day in and day out? And in what scenarios would you test for germline things that could lead to downstream implications for individuals? 

Dr. Bryson Katona: Oftentimes, this is a point of some confusion or uncertainty. And so when we think about germline testing, basically we're looking for gene mutations that are going to be present in every cell throughout the body. When patients who have a tumor undergo somatic sequencing of that tumor, of course, we expect to find lots of gene mutations in the cancer itself. But then, typically, if there is a mutation in all cells in the body or i.e., a germline mutation, that mutation will oftentimes be picked up in the cancer as well. And so I think when you're looking at somatic sequencing reports, sometimes some of these mutations that you see in there in cancer predisposition genes may not be a cancer specific gene mutation, but may be germline and may merit kind of more additional testing. That's always one way that we can identify individuals that if one of these gene mutations is picked up on the somatic sequencing, they can get referred. 

Now, if we think about the population who may not have cancer yet but may have a strong family history, in those cases, those individuals are typically getting referred directly for germline testing since they don't have any tumor or anything like that that has been tested. So that also is another opportunity where, ideally, we would identify all these gene mutations in the ideal world just because we, of course, like to know about these before an individual actually does develop any cancer.

Dr. Rafeh Naqash: I think the important point you brought out is doing this before the individual develops cancer is a huge opportunity for us in the cancer field to improve outcomes for certain cancers. As you pointed out in your paper, like pancreas cancer, survival outcomes in the later stages are not that great, and we haven't had any significant breakthroughs in treatments. Could you tell us a little bit more about how pancreas cancer screening could help change some of those aspects where you could see an opportunity where individuals who get screened timely, diagnosed timely, could also benefit in the long term? 

Dr. Bryson Katona: For pancreatic cancer, really, we know that if it can be detected at stage 1, that's really our only opportunity to meaningfully intervene and impact survival. And so when we talk about screening for pancreas cancer, basically our goal is that we either want to find very high risk lesions that are almost at the stage of cancer, or potentially stage 1 pancreatic adenocarcinomas. We know that for stage 1 disease, especially stage 1a, a five year survival is over 80%. And once you get to stage 2 disease, the five-year survival decreases dramatically. And so I think that if we're going to screen patients, really, our goal should be to detect these cancers at stage 1, where there's an opportunity for surgical intervention. Even stage 2 pancreatic cancers or any that are not surgically resectable, our ability to cure these patients essentially evaporates, and so really, screening should be done to detect things at a time when we can surgically intervene. 

Dr. Rafeh Naqash: In your program, from a practical standpoint at UPenn, is it the cancer geneticist, or is it the medical oncologist who makes the referral and determines need? Or how do you go about it so that other programs or individuals who might be listening to it would understand how a screening program for pancreas cancer actually works? 

Dr. Bryson Katona: I pretty much serve as the referral point for all of our pancreatic cancer early detection studies. We have a great collaborative group here and a lot of our oncologists who may be following these carriers with genetic susceptibility, oftentimes, if the patient is very convinced and wanting to go forward with screening, the oncologist will just order the screening tests themselves, and then me and my team will enroll individuals in studies once they come in for screening. So patients who are interested in screening oftentimes will come in for a full office appointment to really discuss the pros and the cons of screening and decide if this is something that they want to pursue themselves. Regardless of how patients get into the process, whether they get directly referred from oncologists or they come through our high risk pancreatic clinic through which I see patients, it's important that these individuals be offered the opportunity to enroll in pancreatic cancer early detection studies as really capturing data on all individuals who are getting screened is really our only hope of continuing to move this field forward.

Dr. Rafeh Naqash: I totally agree, and I think, as you mentioned in the manuscript, EUS, MRIs are decent tests, but nevertheless, one is invasive, the other one is probably not cheap, may not even be available outside of North America to a large extent. So you definitely need something that's easy, less invasive, perhaps can be implemented in a broader scale. 

Now, from a payer standpoint in the current landscape, do payers actually reimburse for pancreas cancer screening? Is that something that is easy to come by, or do you actually end up requiring, like, peer to peer prior auth, which is something that has become a theme, an unfortunate theme in the oncology space these days. What is your experience with the pancreas cancer screening aspect? 

Dr. Bryson Katona: I think that the landscape has been getting much more favorable towards payer reimbursement. Since the National Conference of Cancer Network, or NCCM, guidelines have included pancreatic cancer screening in one of their high risk guidelines, individuals who meet those criteria, we really haven't been having any issues getting their screening covered. Now, of course, there can be copays and deductibles and things like that that may still provide a financial barrier, but the insurance companies have fortunately been much more amenable to covering the screening procedures. 

But you make a great point that doing an annual EUS or MRI is a pretty involved endeavor. It's expensive, it's time intensive. And in the pancreatic cancer screening space, what the studies have shown is that you probably have to screen 100 or a little bit over 100 individuals in order to find a high risk lesion. So you end up doing a lot of screening to identify a very small minority of individuals who will develop a high risk lesion. And I think that's important as well because even in these high risk populations, pancreatic cancer is still not an incredibly common cancer. It's still diagnosed at a fairly low rate. So, any individual who is undergoing screening, I think they have to be ready to embark on a long journey where they're sticking to this annually, but in the majority of cases, we will not identify a high risk lesion.

Dr. Rafeh Naqash: And from a genetic susceptibility, germline susceptibility standpoint, is there an age cut off where you mandate that these individuals should have pancreas cancer screening? Let's say if somebody's a BRCA1 or BRCA2, or is it irrespective of the age?

Dr. Bryson Katona: We know pancreas cancer is a cancer that affects older individuals. And so typically for the gene mutation carrier, so for like the BRCA1s and 2s and Lynch syndromes, those individuals in whom we're doing screening, usually we say to start at around age 50. 50 may still be a little bit on the early side, and the only reason we would ever consider starting earlier than 50 was if there was a very young pancreatic cancer in the family. One of the more difficult discussions to have with patients is when should they stop screening? And we don't have a really great answer for that, but I think that each year that we do screening, we always address that or think about those factors with patients. 

Technically, a patient who would not be a surgical candidate, if a pancreatic cancer were to be identified, that individual probably should continue to screen, but that ends up being a very individualized decision between the patient and the provider. Now, apart from the BRCA1 and 2 and Lynch and those type of high risk patients, there is the familial pancreatic cancer patients where there's no known genetic mutation. Those individuals, we typically start around age 50 as well. Some of the rare syndromes, such as Peutz-Jeghers syndrome, and carriers of a CDKN2A mutation, we would start at earlier ages, but those syndromes are overall fairly rare. 

Dr. Rafeh Naqash: Going to another aspect of biomarkers in the pancreas cancer setting, something that we regularly, routinely do. In fact, I was on service a couple of weeks back. A fellow of mine saw a patient with pancreas mass, he ordered a CA 19-9, and it was kind of elevated. You mentioned some very important aspects of limitations associated with this CA 19-9. Could you elucidate a little more about some of those aspects?

Dr. Bryson Katona: Definitely. So finding a biomarker, a blood based biomarker that we could use for pancreatic cancer early detection would really be the holy grail and would solve so many issues that come up in the field. And so CA 19-9 has definitely been the one that has been studied the most in this space. With CA 19-9, we know that it's a good marker for pancreas cancer once it develops. The issues that arise in the pre cancer stage when it's used as more of a screening type tool, is that, one, 10% to 20% of patients actually are Lewis antigen null and don't express CA 19-9. And so you have a pretty decent amount of at risk individuals who are not going to express CA 19-9 to begin with. 

And then I think the other thing that makes it very difficult is the differing baseline levels of CA 19-9 that can be present. An individual who has a CA 19-9 that's checked and say it's 5 or 10 above the upper limit of normal, that could just be their baseline level or that could be something of concern. And oftentimes in those situations, it's a lot of worry and angst for the patient, but then also a lot of worry and angst for the provider as well, and individuals just don't quite know what the next best step is for evaluation in that case. 

There is some data that's come out of Hopkins recently showing that there are a few genes where there are certain genetic mutations that may alter one's level, baseline level of CA 19-9. And so that may be a potential way to personalize or individualize what someone's baseline CA 19-9 level should be. But that's, of course, still kind of in the experimental stage and not something that's ready for widespread practice yet.

Dr. Rafeh Naqash: And the other biomarker, as you've based the study around is the IMMray PanCan-d ™ test. Seems like a composite biomarker. Could you tell us a little more about what constitutes this biomarker and what are the different parameters associated with it? 

Dr. Bryson Katona: The IMMray PanCan-d test, really just to give a little history on it. So this was the first commercially available pancreas cancer specific blood-based test. This was commercially available to patients starting in late 2021 and through the summer of this year 2023. This particular test looked at 18 protein biomarkers and in addition also factored in the level of CA 19-9. And basically what it would result in, it would give four possible results. It would either give a positive or a negative result, but also had the potential to give a borderline result. And then the fourth possibility was via test not performed result. And those were primarily for the individuals who were CA 19-9 non-expressors. Again, this was really the first commercially available pancreas screening blood based test that was available.

Just for listeners, the company has, based on, I guess, their first 18 months worth of data, they temporarily pulled it off of the market to reevaluate some of the parameters. But while it was on the market, we were fortunate enough to collect this real world data. And I think that the data that we collected through this study really allowed us to learn some really important lessons about how a pancreas specific blood based biomarker, how it should work, and what are some important components of this biomarker that really need to be addressed in future versions of this test, but then also in future blood based pancreatic cancer biomarkers that are developed down the road as well. 

Dr. Rafeh Naqash: Bryson, can you tell us a little bit about the inclusion exclusion for this study, the kind of individuals that you incorporated in the study, and then we can go on to some of the results?

Dr. Bryson Katona: Sure. So for this study, what we did was we ran an IMMray PanCan-d test on individuals who were coming in for their normal, their regular pancreatic cancer surveillance. So these were high risk individuals, lifetime risk of pancreatic cancer of 5% or higher who were already enrolled in a surveillance program. And so we had had baseline imaging of their pancreas. And in addition to those individuals who were offered this test, who were undergoing surveillance, we also spiked in several patients that had a known diagnosis of pancreatic cancer. And this was done such that the company was blinded to the results of who had a known diagnosis of pancreatic cancer. So what this did was it provided us really with the first real world data on the use of this test in clinical practice and also, because we utilized some spike in samples, allowed us to utilize and calculate some performance metrics of this particular pancreatic cancer early detection test. 

Dr. Rafeh Naqash: And some of the important metrics that you mentioned in the manuscript is this test having a very high negative predictive value, but a low to medium positive predictive value, partly from some of the testing aspects related to some of the results. So could you tell us a little bit more why the negative predictive value was so high, but the positive predictive value was not as high? And what are the things that could be done in the near future to perhaps improve upon the positive predictive value?

Dr. Bryson Katona: Yeah, you're right. As you do mention, the negative predictive value was very strong. The issue with the positive predictive value came out of the fact that this test had a borderline result that was used. And when you have the option of not just a positive result or a negative result, but you have a test that can offer this borderline result, how this borderline result is characterized can really alter the performance characteristics. So just to give some numbers, about 7% of the patients who had the test performed came back with a borderline result. 

Now, if people were comfortable with just brushing off the borderline result and not doing anything additional, then that would be fine, that wouldn't be a problem. But I think most providers, and probably most patients as well, if they see a borderline result, they're probably not going to be okay with just brushing it off and they're going to want to either do some additional testing, whether that be additional blood testing or potentially a sooner imaging study. And so, although it's a borderline, it's not a true positive, it's in the borderline region, I think it's probably going to raise a level of concern enough that most people are probably going to act on it like it may actually be a positive test. And so how those were characterized, the borderlines, it really altered the performance characteristics. And if you considered those borderlines to actually be positive tests, given that people are probably going to want to act on those results differently, it made the positive producing value of the IMMray PanCan-d very suboptimal. 

And I think that probably is one of the most important lessons that we learned from looking at this data is that any future pancreatic cancer early detection tests probably should not have a borderline result and should be designed so that it either comes back as a positive or a negative without having kind of that ambiguity of a borderline result in the middle.

Dr. Rafeh Naqash: I think this is an aspect that touches every aspect, every side of biomarker assessment. One size does not fit all. There's a lot of borderline results we get. I do a lot of immunotherapy. Of course, immunotherapy biomarkers is a huge thing and so far we have not been able to narrow down something that is just one sole marker.  

Now from a pancreas cancer standpoint, I think in your data set you had a few individuals with IPMNs, a few with pancreatic adenocarcinomas, and a couple, I think, with neuroendocrine tumors. Do you think that actually impacts the type of the lesion, actually impacts the outcome or the positive predictive value associated with the test?

Dr. Bryson Katona: We did find a couple incidental pancreatic neuroendocrine tumors in these patients. The test itself was not designed. It was designed and trained on pancreatic adenocarcinomas. And so the tests in the patients with pancreatic neuroendocrine tumors did not come back as positive or borderline. And then in terms of IPMNs, IPMNs are so incredibly common. I mean, most series have shown that probably around 30% of individuals who are undergoing pancreatic cancer surveillance have an IPMN that's identified. And so again, I don't think it was trained to necessarily pick those up. Although talking to patients about the frequency of IPMNs on surveillance is really important because I always counsel patients that you have about a 30% chance or that something's going to be found in your pancreas. Granted, most of these are inconsequential and just things that we continue to monitor, but they are found incredibly frequently.

Dr. Rafeh Naqash: And one of the other aspects I guess, is you mentioned this as a limitation in your manuscript, is the fact that the study had a majority of individuals that were of Caucasian ancestry or ethnic ancestry. So from an ethnicity standpoint or racial diversity standpoint, is there a plan to incorporate other ethnicities in a bigger, broader data set to validate whether a similar pattern of high NPV negative predictive value could be identified in non Caucasian individuals also?

Dr. Bryson Katona: That's a great point and at this point, unfortunately, it's not an issue that is specific to this particular study, but it is one that affects really the entire literature in the pancreatic cancer early detection space. Diversity amongst individuals who are undergoing pancreatic cancer surveillance enrolled in these pancreatic cancer surveillance studies is really limited. So most of the data that we have on outcomes of pancreatic cancer surveillance really in all of the major publications is primarily limited to individuals who identify as white. And so this is an area where the whole field I think needs to think of ways and create new innovative ways to get more diverse populations into surveillance. And so again, I think until we fix the bigger problem of getting more diverse populations into surveillance, we won't have populations big enough to validate this particular test on those individuals.  

So I think one other point that is really important that came out of this data was having a pancreatic cancer early detection test that requires CA 19-9 to be expressed is also a major limitation. In this particular study that we did, we found that about 14% of our high risk individuals got a ‘test not performed’ result back because they were not CA 19-9 expressors. And with 10 to 20% of the general population not expressing CA 19-9, that really limits the individuals that would potentially be eligible for this type of a test. And so I think one of the other lessons that we learned from this is that any type of pancreatic cancer early detection test that is developed in the future, I think it ideally needs to be developed such that it can be run regardless of CA 19-9 expression. Otherwise you run the risk of developing a test that up to a fifth of at risk individuals may not even be able to benefit from.

Dr. Rafeh Naqash: I think those are very important points, and as the biomarker field advances, I think taking into account, one, different races ethnicities because biomarker expression could be different based on genetic ancestry, and then taking into account the fact that, for example, for pancreas cancer, as you mentioned specifically, CA 19-9 may not be a universally expressed biomarker. So trying to overcome some of those limitations by designing things that are more universal is probably the way to go and I guess more to come in this field. So Dr. Katona, what is next for this project to perhaps expand on this pilot proof of principle project into a bigger, broader population or a validation cohort of some subtype?

Dr. Bryson Katona: Immunovia, who is the company that developed the IMMray PanCan-d test, they're working to retool this test to hopefully release a second generation test in the near future. And I think that the hope will be that the second generation test will not have some of the issues such as dependence on CA 19-9 and use of a borderline result that this first version had. So we'll certainly be excited to help test and validate a second line test once it comes out. 

And I think that just the pancreatic cancer biomarker field is an exciting one. I know there's several other commercial tests that are also currently being developed a little bit earlier in the development stage, but it's exciting to see so much effort and research being put into this area of just really great need. And so I'm really excited to see where the field goes over the next couple of years. 

Dr. Rafeh Naqash: A couple of quick questions about yourself. Tell us a little bit about your journey, what led you into this specific field and how have things gone for you as an early career investigator in the last several years?

Dr. Bryson Katona: Great. Yeah, I'm a physician scientist, so I see high-risk individuals in the clinic. I have a basic science lab and I run a lot of clinical research and early detection studies. Really, I knew I was interested in digestive cancers very early on and as a gastroenterologist, I found that I could probably have the most impact in the early detection space. And so ever since I was a fellow, this has been kind of my area of interest and expertise and I think taking the highest risk individuals, so those with strong family histories, those with genetic predisposition, I think you really have the opportunity to make the biggest impact with early detection strategies. And so that was one thing that always really motivated me in this field.

Dr. Rafeh Naqash: Bryson, thank you so much for indulging with us. Thank you for choosing JCO Precision Oncology as one of the destinations, the final destination for your work and hopefully, more to see from you and your group in the near future. Congratulations on your concurrent publication presentation and the podcast, and thank you again for joining us.

Dr. Bryson Katona: Thank you so much for the opportunity. It's been a great discussion.

Dr. Rafeh Naqash: Thank you for listening to JCO Precision Oncology Conversations. Don't forget to give us a rating or review and be sure to subscribe so you never miss an episode. You can find all ASCO shows at asco.org/podcast. 

The purpose of this podcast is to educate and to inform. This is not a substitute for professional medical care and is not intended for use in the diagnosis or treatment of individual conditions.  

Guests on this podcast express their own opinions, experiences, and conclusions. Guest statements on the podcast do not express the opinions of ASCO. The mention of any product, service, organization, activity, or therapy should not be construed as an ASCO endorsement.