NeoGenomics, Inc. (NEO) Earnings Call Transcript & Summary

Good morning, and thank you for joining us for this morning's Breast Cancer Tumor Board. We are so glad that you have joined us today. Before we get started, just a few housekeeping items. You are being muted upon entry. Questions are welcome and encouraged. If you do have any questions for our panelists, please submit via the Q&A on the bottom tool bar. In addition, shared cases are what makes this program successful and we encourage submissions. Should you have a case that you would like to present at one of our Tumor Boards, please send an e-mail to tumorboards@neogenomics.com. The programming today is brought to you by the Medical Affairs Department at NeoGenomics, whose aim is to foster education and promote the understanding of cancer within the global scientific community at large through a program that enrich medical and scientific knowledge of cancer diagnostics that point the way to advancing patient care. We would like to invite you to join us for a 3-part speaker series, featuring GI oncologists discussing MRD use in CRC on October 31, November 1 and November 2. Registration is required. You can use the QR code that is on your screen right now. Additionally, a link will be provided in the chat section shortly. In addition to our Breast Cancer Tumor Boards, we also have a Lung Series Tumor Boards. Link for the lung series will be provided in the chat as well, and you can also use the QR code on your screen. Our host today is Dr. Fernando Lopez-Diaz, Director of Clinical Science at NeoGenomics, Dr. Lopez-Diaz, welcome.

Good morning, everybody.

And we are so privileged to welcome back as our moderator today, Dr. Razelle Kurzrock. Dr. Kurzrock is the Chief Medical Officer at the WIN Consortium for Personalized Cancer Therapy. She is known for founding and building the largest Phase I clinical trial department in the nation and world, while at the University of Texas' MD Anderson Cancer Center. A central theme of this program was the precision medicine strategy. Dr. Kurzrock is from Toronto, Canada and has over 800 publications and has been named to the list of the most highly cited scientists in the world in the 25 World Leaders in Precision Medicine. Dr. Kurzrock, as always, we are so honored to have you joining us today. Thank you so much for your time.

Thank you for having me.

And we have 3 really exciting cases to discuss today. So I'm going to turn it over to Dr. Lopez-Diaz for the first case.

Okay. Well, thank you so much for the introduction. So with no more ado, we're going to start with our first case. And the 1 note that we'll make on these series of cases in -- we have chosen some common topic, which is something we've discussed briefly before, and we have addressed some specific points about male breast cancer. In this case, in this opportunity, we have chosen to present some male breast cancer cases, but you will see how this can be really a very interesting topic for discussion and for rare cases, where suspected breast cancer might lead us somewhere else, and we'll have Dr. Kurzrock was representing very interesting cases, and we'll discuss 2 actually of those, which are cases of cancer of unknown primary origin. The first patient is a male patient, 66 years old. And he presented with invasive ductal carcinoma. Please, next slide. And this patient was referred for an invasive lesion in breast. And as we can see here, the molecular profile of this patient has presented an ARID1A truncating mutation at position -- sorry, at position 1,409, which is actually in the middle of the protein, ARID1A is the protein that is about 2,285 amino acids. And in this position, is very frequently mutated. Most mutations in ARID1A are truncating mutations. Also, the patient presented a truncating alteration in P10, which is also a loss of function mutation, and it presented, at the same time, a P53 truncating mutation, loss of function mutation in the DNA binding domain. There were no alterations detected by FISH, and which include ERBB2 and P10 FISH were performed and not detected. The patient presented micro satellites were stable and did not express PD-L1 in terms of immunotherapy markers, but have an intermediate tumor mutation burden. And I think if you can move down a little bit, I'm trying to remember, we had the notes for what was the value of -- it presented with 7.4 mutations per megabase. Okay. So Dr. Kurzrock, maybe you can now discuss the case.

Yes. So thank you for presenting such an interesting case and so much information there. So the first thing is this is a man with breast cancer and male breast cancer is rare, less than 1% of breast cancers, but it does occur. And it usually has a worse prognosis maybe because it's often diagnosed later, perhaps just not enough awareness that men can get breast cancer. Like this patient, it's often ER-positive. This patient was ER-positive. And like this patient, it's often HER2 negative. Now about 60% of cases have androgen receptor as well, and that's something that I think we should think about testing in both female and male breast cancer because about -- probably about 30% of women, 15% to 30% also have AR positivity. Now with the alterations, I think there's a lot of interesting material here. The first thing I'll talk about is a [ RIG-1a ]. A RIG-1a is a chromatin remodeling gene, and therefore, it does a lot of different things. But the literature suggests that sensitizes to several different therapies. One is 2 PI3 kinase inhibitors. So something like PI3 kinase AKT inhibitors -- PI3-kinase AKT mTOR inhibitor. So 1 possibility is a drug like everolimus, which is an mTOR inhibitor, and obviously approved for breast cancer. It may also sensitize to PARP inhibitors. So the literature is not straightforward there, but there is some significant literature suggesting sensitization. So one could think about a drug like olaparib or talazoparib. And then finally, there's literature, some of which we've written, that ARID1A sensitizes to immunotherapy. So a drug like pembro could be considered. So the second alteration the patient has is a P10 loss and P10 loss also sensitizes to PI3-kinase AKT mTOR inhibitors. And I would pick an mTOR inhibitor, again, like everolimus because there's drugs like Alpelisib which is a PI3-kinase inhibitor, which is approved for breast cancer, but it specifically targets the alpha unit of the PI3K. And P10 loss activates both the alpha and the beta unit. So a drug like Alpelisib would not be expected to be effective. Now I would think about drugs like temsirolimus, everolimus and copanlisib. So my putting this together, the patients ER-positive. I think the simplest regimen that matches his alterations is everolimus and exemestane. So everolimus and ER hormone modulator. So I think that's pretty straightforward that would target both the ARID1A and the P10. As I mentioned, we could still think about a PARP inhibitor in the future olaparib. And we could still think about immunotherapy because ARID1A sensitizes to immunotherapy. Now what is the factors against immunotherapy? Well, first, the PD-L1 was negative. And second, the TMB, although it's a little bit, it's in the intermediate range at 7.4, it's not really in the of 10 or greater, where pembrolizumab was approved. But I want to point out 1 thing that I think is very important. Tumor mutational burden is a surrogate for the number of neoantigens that are produced. So every mutation will produce neoantigens with the immune system once reawakened with immunotherapy recognizes. So the higher the mutational burden, the more neoantigens you have, and therefore, the better the response to immunotherapy. Now this patient has a TMB of 7.4, which is a little bit -- it's into the intermediate range, but it's not above 10. But very interestingly, the patient has frameshifts. And the patient has frame ships in both ARID1A and P53. And the data isn't out there yet. But conceptually, a frameshift should cause more neoantigens. So it's not a single base pair change. But you're shifting the entire reading of the protein. And therefore, you should be getting many neoantigens. And then that would mean that even if your TMB is intermediate, your neoantigen load may actually be higher. So that is something that is yet to be validated, but I think it's something to think about. My first choice would be everolimus in a hormonal -- ER hormonal margin later like exemestane. So I'm going to just pause there, and see if there's any -- if there's any questions.

There is 1 question already coming in. And so somebody is asking if expression is looked at. And so just to be clear, in the clinical setting, gene expression is not widely used. But I don't know, Dr. Kurzrock, in your clinic, and I know that you have typically much advanced access to differential testing and transcription testing in your clinic is something that you consider and you're an advocate for. Can you tell us a little bit about that?

Yes. So if we have patients, we do access transcription. And sometimes it's less clear how valuable that is compared to the genomics. But we do think it's something that is definitely worth looking at. Now from the frameshift point of view, however, this would be a bioinformatical calculation of how many neoantigens are produced. And to my knowledge, this has not yet been done. But if you think about it, it makes sense because when you have a mutation, you have a single nucleotide change that will translate to a single amino acid change. And therefore, it's a single new neoantigen. It's actually more than that because usually anywhere from 10 to 20 amino acids are presented by the MHC region. But if you have a frameshift, everything is shifted. So now you have potential for a lot more neoantigens. This can be calculated in a bioinformatical view. And I want to emphasize, this is theoretical. It hasn't been done before. We did notice in the clinic, again, not validated that patients with frameshifts seem to do unusually well with immunotherapy. So that would support the -- that they generate more neoantigens. And I'll give you 1 example. Indels, insertion deletions, in a way are like frameshifts, because they shift the entire reading of the protein. And in renal cell carcinoma, immunotherapy is extremely effective but the tumor mutational burden is low. So you could say, why is renal cell carcinoma a kind of exception where you get immunotherapy is approved, it's effective and the TMB is low? Well, the reason might be, again, it's not validated, but I'm putting it out there is that TMB is a surrogate for neoantigens. It's just a surrogate for how many neoantigens you may be presenting that provoke the immune system. And in renal cell carcinoma, your tumor mutational burden is low. You don't have a lot of mutations, but you have the highest level of insertion deletions, indels, and that would be expected, just like frameshifts ships to create a lot of neoantigens. And so maybe the indels are why renal cell carcinoma responds so well.

Yes. I think that what we have fantastic set of questions already and reading them. One quick addition is that we had to remember that very likely the ARID1A itself alterations is going to lead to more additional mutations and frameshifts. So basically, ARID1A is a chromatin remodel, but it's involved in DNA repair. And when DNA repair processes, specifically base excision repair mechanisms are altered, frameshifts will be more likely to occur because there is incorrect connection between the ARID1A and [indiscernible] Polymer [ inhibitor ]. So we had 1 more question about germline testing. The answer to that question is no when this patient had not been referred for germline testing that we know of. And another question is if the patient had anthrax fusion data present. And no, it was not order at the moment this patient was tested, which, at the same time, speaks for the need. I mean, test for anthrax fusions do exist -- have existed for a while by medical -- by many laboratories, including ours. But it's not typically very much order. As specific in breast cancer, it's sometimes not really considered as a possibility.

Fernando, would an anthrax fusion come out on your NGS even if you did...

Yes, we have panels for fusions that -- and even in the breast cancer panels that our laboratory offers, which is a subset of fusions, we already have Anthrax 1, 2 and 3 and many other fusions that are known to be actionable in breast cancer.

Is there any immunotherapy clinical trial actively looking at the frameshifts? No, there isn't, and not actively looking at indels. So I think that this is an area that I think is extremely interesting, but it hasn't been validated yet. It makes sense from a bioinformatic point of view calculating out the neoantigens, but there is no, to my knowledge, validation. About the germline, just looking at the somatic alterations, the only 1 that might become in or could occur in a patient like this is the P53 whether the patient has Li-Fraumeni syndrome. The most common germline alteration in men with breast cancer is BRCA. And this patient doesn't have a BRCA in their somatic panel. So even though it's not a germline [indiscernible], I don't think we're going to pick up BRCA if we didn't see it in the tissue panel.

Yes. That's correct. And regarding the neoantigens, I think that just a quick comment of what's coming as technologies are becoming more available, we've been doing genomics for a few years now in general, and it widely applies in the clinical setting. And neoantigens will [ count ] to service much more often when we get proteomics moved forward into the clinical setting. Because the one thing that it's known to be behind is the knowledge of the new proteins that are generated by all these mutations. So we typically often talk about oncogenes and tumor suppressors, so mutations either remove the function or exacerbate a non-function by activating constitutively a given protein. But many of these mutations, specifically those that involve frameshifts or indels, they create new protein domains that can interact with novel proteins. And this is an unchartered territory until we get more proteomic-based technology being more [indiscernible]. Okay. I think we can probably move into our next -- are there more questions, first of all? Let me see.

No, we have answered all of the questions in the Q&A. Just a reminder, questions are welcome and encouraged. So feel free to submit any questions you have via the Q&A tool.

So thank you. So before we move to the next patient, I thought I would talk about cancer of unknown primary in the molecular era, because it's relevant to the next patient. So the next patient, which will be presented right after this, will be a man with breast cancer, who, the question was, does he have non-small cell lung cancer or does he have breast cancer? And I think the conclusion in the end was that he had breast cancer. But this is a classic cup or cancer of unknown primary dilemma, where does the cancer originate. And it's especially relevant to a man where you might not think of breast cancer in your differential diagnosis. So let's move on. And I'm going to particularly concentrate on some work with carcinoma of unknown primary in the molecular era. So carcinoma of unknown primary is almost the ideal cancer to evaluate by genomics because we really don't know where -- which organ is that cancer. By definition, we don't know where the cancer is coming from. And furthermore, a lot of these patients have a really poor prognosis, probably the 50% survival is something like a year. And furthermore, the tissue of origin testing studies so far have not really shown a really remarkable improvement in survival in trying to figure out the tissue of origin. So we've looked at this in our own studies as a molecular disease. We actually look them in all of cancer molecular disease, but I think it's especially well-suited to carcinoma of non-primary. Next slide. And what you can see is we did a study of 1,739 patients with CUP evaluated by liquid biopsy. And then we looked how many of these patients have Level 1, Level 2 or R1, which is resistance, validated alterations based on the onco KB classification. And I think it turns out to be really interesting, because Level 1 is 26% of patients, have a Level 1 alteration. Level 2 alterations is 4%, and R1 is 17%. So altogether, about 47% of patients really have a very evidence-based alteration. Now in our view, most of these patients have an actionable alteration. But even when you take the highest level of evidence, which is FDA recognized or onco KB recognized is 47% of patients have what we would consider a high level of evidence alteration. Next slide, please. And Level 1 is obviously the highest level in onco KB. And what you can see here is what are the Level 1 alterations. And just looking at this, you can see common alterations that we all know there are drugs that may impact them, alterations like EGFR or B2 amplification, BRCA1 and BRCA2, microsatellite unstable, [ met ] skipping, alk and FGR kit and so forth. Next slide, please. So we did a clinical study where we looked at our patients that were matched. And this is a smaller group of patients, of course. This is a group of 40 patients with cancer of unknown primary from UCSD, and looked at patients that were highly meant. They were treated on a molecular basis, the physician could choose the treatment, but they were presented in molecular tumor board and molecular tumor board gave recommendations and then the physician could choose the treatment. And then we measure the degree of matching, which we dichotomized into a high match, which is over 50% and a low match which is less than or equal to 50%. And let me just explain this. The degree of matching is roughly the number of alterations matched over the total number of alterations. So if you only have 1 pathogenic alteration and you match that, you're going to get a score of 100%. If you have 10 pathogenic alterations and you only match 1 of them, you're going to get a score of 10%. So this assumes that if you have multiple co-drivers and you don't match a lot of them, you only match 1 or 2, you're just not going to do that well. And that's exactly what I think this graphic shows. What you see is the median PFS was 10.4 versus 2.5 months for highly matched versus low match patients, the median survival was 13.4 versus 5.7 months. And when we looked at clinical benefit rate, which included complete remission, partial remission and stable disease for at least 6 months, it was 83% in the well-matched patients and only 25% in the poorly matched patients. And so really, there was improvement in all the outcome parameters with increased degrees of matching. Again, this is CUP patients based on data from UCSD. Next slide. So this is just an overview of CUP treatment strategy. What we usually do is a patient presents with a cancer of unknown primary. We do all the regular imaging, IHC, tissue of origin, assays, and if we figure out where the primary cancer comes from, the patient gets site-specific therapy. If we can't figure out where the cancer comes from, we generally do molecular profiling and then give molecularly-based therapy. Now since I run a precision medicine clinic to be totally open, we kind of give everybody preferentially molecular therapy. But this outlines the usual way of trying to deal with the cancer of unknown primary. Next slide, please. And so I'm just going to give you 1 example of a patient before we move on to the presented patient that came from our study. This patient was a 62-year old man with a poorly differentiated cancer of unknown primary. And his liquid biopsy showed an unusual KRAS mutation, but this is a deleterious mutation that activates the met pathway. And like the patient previously presented, the patient also had an ARID1A alteration. Tissue NGS had an insufficient sample, so this was done on a liquid biopsy. Next slide. And so we match the patient. The KRAS activates the met pathway. We've been matching the patient to MEK inhibitor trametinib. And ARID1A, as we discussed with the previous patient, does a lot of different things. But one of the things that it does is it may sensitize to PARP inhibitors. So the patient also got olaparib. And he was consented on our IPREDICT protocol, which is a protocol that matches patients to drugs in this way. Next slide, please. So this is what you can see in the treatment. He tolerated the treatment well. The tumor on the -- compared to the left side to the right side at 6 weeks, a pretty dramatic change. Next slide, please. And at 16 weeks, you can see the tumor is almost gone. And this lasted for a little bit over 2 years. Progression-free survival was 26 months. So for a completely molecular match therapy, I think, well-tolerated, chemotherapy-free. This was a really excellent response. Next slide. Okay. So I don't know if there's any questions about that or you'd like us to move on to the next case.

Let's see if we have questions by the audience. I see that there are no. I'll ask you 1 question. You commented, let's start with some of a typical dilemma. When you have a cancer of unknown primary, how often is that you'll have insufficient tissue for the patient to be biopsied?

The truth is that usually you have a sufficient tissue. And in this patient, because it's a superficial lesion, the physician could have obviously gone in and biopsied it again. So the physician didn't go in and biopsied it again because he had a liquid biopsy with results that were actionable. So he went in and treated the patient. And within 6 weeks, the patient had a dramatic response. So there wasn't much to biopsy again. But I think it's a minority of patients that will have inadequate tissue. And I want to emphasize again, in this patient had the physician wanted to with a superficial lesion, he could have gone in and biopsied it again.

Thank you. That's...

We do have -- yes, we have some questions coming in for the Q&A. Dr. Kurzrock, could you please elaborate on LB solutions for the cancer of unknown primary cases?

I'm assuming liquid -- LB as liquid biopsy.

Liquid biopsy.

Yes. So I think liquid biopsies are very important. And to me, they're complementary to tissue. Tissue and liquid give you -- they give overlapping answers, but not 100% overlapping. Usually, about 85% overlapping. And people have questioned when there are differences between the tissue and the liquid if this is a technical issue, I absolutely believe that in the vast majority of cases, it's not a technical issue. It's that the assays are looking at different things. The tissue is looking at the piece of tissue biopsied. It's usually more sensitive. Liquid is dependent on very small amounts of DNA that is shed into the bloodstream. But it also has advantages. It has the advantage that it may be shed DNA from multiple sites, not just from 1 site. So personally, we try to do both tissue and liquid in our patients. If for unknown primary, if there are several mutations and known targeted therapy, how do you decide which therapy to choose first? So that is a really good question. In our clinic and also in our publications, we have done customized combinations of therapy. So we've learned how to do that safely by initially reducing the dose. And I would point you to, as an example, our publication called iPredict, which is the iPredict trial that we published in Nature Medicine in 2019. We actually have several publications on this, but that might be the easiest one to find and read. So Dr. Sicklick, S I C K L I C K, is the first author on that, and I'm the last author, and that points to how we put the drugs together. Now even putting the drugs together, however, some patients have too many alterations to target all of them. So how are we going to decide what we target? And quite frankly, there's no perfect way to decide. We do have some rules that I personally like that I'll mention. One is that if there is a fusion, we will try to preferentially target that because we believe that fusions are strong drivers. And anything we believe, we published or other people have published. So this is not -- I'm saying, we believe based on the literature. The other thing that we will prioritize is immunotherapy. If there are biomarkers for immunotherapy, we will often prioritize those biomarkers because we've all seen patients with immunotherapy, which is not perfect, but we've all seen patients who go into complete remission appear to be carried even though they had widely metastatic disease. So we do prioritize immunotherapy if there is a biomarker. And then the last thing we have to consider is, is there a clinical trial for this patient? Is the patient eligible for the clinical trial? If there's no clinical trial with the drugs that we want to give, are we going to be able to obtain the drugs off-label? Is insurance going to pay for it? Will the company provide the drugs? So yes. So that's something. So outside of a clinical trial, how do you get approval for targeted drugs if not FDA approved? If they're not FDA-approved, that's very difficult, although you could do compassionate use. If they're FDA-approved and you want to use them off-label, we do have a way that we've set up in our clinics, which involves appealing to insurance. And if that fails, appealing to the drug company. A question of a 73-year-old patient with relapsed AML and has ECH2 mutation. I am having issues getting approval for tazemetostat, any suggestions? Tazemetostat is very difficult to get. We have a few patients that are similar, and the insurance denied the tazemetostat and the drug company gave it to us. And more recently, the drug company said that they're not giving out any more drug. And how can we refer patients to your clinic? I'm happy to see a patient. If you contact NeoGenomics, they can connect you with me. And if the patient wants to travel to see me, I'm happy to see the patient. Okay.

Okay. Since we -- let's move to the next case, but thank you. This has been a wonderful discussion. I think that we're having a lot of interesting questions. Next slide, please. Okay. So these patients are, yes, this patient is also a 66 years old patient. And in this case, this patient presented with suspected breast cancer, actually the lesion was a lesion on the bronchia in the lung. And it presented as a potential lung cancer but suspected breast cancer, and it has been tested, and the order that came to our laboratory was for the breast cancer profiling, and it presented a BRCA1 alteration [ truncating ] mutation, amino acid 143, that's in the [ Terminal ] Domain of protein and a P53 alteration in the DNA binding domain, which is also, in this case, it's a frameshifts, now that we are talking about frameshifts, and which truncates the protein for amino acid later and with a full frameshift, as you can see here. The patient presented a deletion of P10 detected by FISH. It had expression of PD-L1 with microsatellite stable and an intermediate tumor mutation burden. The TMB was of 8.5 mutations per megabase. And in our case, in this test, the TMB of 8.5 is really in the lower end of the intermediate side. Next slide, please. Do we have -- I think that -- yes, that's what I wanted to see. Patient is ER-negative, PR-negative. If we can go one, yes, and here HER2 negative. It's a triple negative patient. Next slide. Okay. Dr. Kurzrock, if you want to discuss the patient, maybe if we can go back to the slide, the first slide.

Yes, I think it would be good to go back. I mean, I remember what was wrong, but I think in the audience will be good to keep that. So great patient, great case in so many ways. So the first thing about this is that reason for referral non-small cell lung cancer, suspected breast cancer, malignant neoplasm of unspecified part of the -- of unspecified bronchus or lung. So this is exactly what I was referring to of cancer of unknown primary. The patient has something in the lung, and then you're thinking -- so what is this? Is this a lung cancer? This is a man, so you're not likely to think of a breast cancer. But then obviously, on pathologic review, somebody thought, hey, this may be a breast cancer. And that's why we talked about cancer of unknown primary. So let's look at some of the alterations. You didn't mention this, but you had previously told me that the PD-L1 CPS in this patient was 70%. So I just want you...

Yes. I think that we didn't mention the percentage. Yes, 70% that's very high.