Insight Molecular Diagnostics Inc. (IMDX) Earnings Call Transcript & Summary

Good afternoon, and welcome to the Oncocyte KOL Event Series, Part 2 on DetermaIO in Triple-Negative Breast Cancer. [Operator Instructions] As a reminder, today's event is being recorded and a replay will be made available on the Oncocyte website following the conclusion of the event. I would now like to turn the call over to your host, Ronnie Andrews, President and Chief Executive Officer of Oncocyte. Please go ahead.

Thank you very much, Sarah, and thanks, everyone, for joining us on our second call in this many weeks. Very excited today to have 2 very spectacular researchers that will be speaking to you today, Dr. Giampaolo Bianchini, who, as you know, authored and just gave a terrific presentation at ESMO on DetermaIO in triple-negative breast cancer; and Dr. Priyanka Sharma, who also will be speaking to some research and some work that she's done in the world of breast cancer and immune therapy. You will be also introduced today to Rob Seitz. Rob Seitz is the Head of Immune Oncology here at Oncocyte, and Rob will be giving a brief introduction. We will be going through forward-looking statements today. So please know that the forward-looking statement will be in effect. If you think about how the immune oncology fits in our portfolio, our mission statement really says a lot about the purpose of where Oncocyte is going and why IO becomes important, and that's that we're trying to provide answers to critical questions today that are unanswered by current technologies. And clearly, the immune oncology decision point is one that's important today, but certainly will grow in importance over the next 3 to 5 years. And having an effective biomarker that can identify quiescent to hot tumors or immunomodulated tumors is going to be very, very important in the industry to ensure the appropriate use of this drug class. And so if you go to the next slide, please. So our strategy since I got here 2 years ago was really been to put in play a comprehensive set of tests for treatment decision and then follow that with monitoring applications that allow us to look at patients over time and longitudinally monitor the efficacy of the therapy and then ultimately, monitor disease-free survival patients for recurring tumors. Today, though, in the treatment decision buckets, we're going to be focused just on the question, should I give an immune oncology therapy? Our DetermaIO product was a product that we acquired 1.5 years ago from a company called Insight Genetics in Nashville, Tennessee. A lot of work has been done since we acquired it to bring it through the clinical trial process. And certainly, today's speakers have been very helpful to us in that process. And so next slide, so now I'd like to do is just turn the call over to Rob Seitz. Rob is our Head of Immune Oncology here at Oncocyte. And Rob will be taking you through a few slides on the DetermaIO application. And then from there, he'll turn it over to our guest speakers. I'll come back at the end and certainly take some questions and answers for you. Thank you. Rob?

Thank you, Ronnie. What I want to do is to give a brief background on how the test came to be and on the various tumors that we have tested DetermaIO in. Those tumors comprise over 900 patients, and we've listed them there. We have 2 non-small cell lung cancer cohorts, 2 triple-negative cohorts, a bladder cohort and a very small renal cell cohort, and I'll go over some of this data with a little bit more detail. But first, DetermaIO as a test is a 27 gene assay, gene expression assay that we can run either by qPCR and that is definitely the ultimate intention for the clinic as a CAP/CLIA approved test, which we've done. But it can also be done as an NGS test where we just take data from whole transcriptome and pull out the 27 genes. There seems to be very good agreement when we were developing that sort of test. And in no case have we ever changed the algorithm, the way we get the call of DetermaIO between the 2 assays. Furthermore, we also try to stratify the patients into positive and negative. And as we have gone across all of these different tumor types and all of these different immune checkpoint inhibitors the algorithm and the definition of positive and negative, the point on that algorithm that we say qualifies you as being DetermaIO positive has not changed across any of these cohorts. We try to give a call that is clinically relevant to the physician of positive or negative. I'm excited about the fact that we've gone back to triple negative because it is in effect coming around full circle. In 2011, pretty groundbreaking work done by Brian Lehmann and Jennifer Pietenpol out of Vanderbilt, where they took roughly 2,200 genes and created a way of looking at triple-negative breast cancer with 6 different subtypes. That algorithm was in-licensed by Insight Genetics, the company that Oncocyte acquired roughly 2 years ago or 1.5 years ago. And while we were there, it was turned into a 101-gene algorithm with the idea that, that would be easier to use in the clinic in a more practical way. We still have a great deal of interest in TNBC type, but we also, around 2016 drawing on other work that was done by the Lehmann Group and some of our own work began to realize that the bottom 3 subtypes, this IM, M and MSL were very likely to be giving us information about immune checkpoint inhibitors, not just in triple negative, but also in other tumors as well. And so Insight Genetics as a whole began to refocus on this question of immune checkpoint inhibitors. And this again, is really what we feel makes DetermaIO different. A lot of algorithms out there have genes that are designed to be IM or immune hot. What makes us different is the fact that we look at immune cold genes as well. This MSL signal, which we think is cancer-associated fibroblasts and the M group, which is the epithelial to mesenchymal transition. We just see numerous advantages by looking at the entire tumor immune microenvironment and looking at both genes that are sensitivity and genes that are resistance. And now to quickly recap some of the data that I showed on that first slide, we've given data in an advanced non-small cell community cohort made up of patients where DetermaIO had a very clear separation of curves. We tested against PD-L1 and it was independent of that. We also asked is it working against pembro and nivo? Patients in this cohort were roughly split 50-50, and it was working with both of them. We asked if it was working with adenocarcinoma histology versus squamous cell carcinoma. And again, it was working. And we also ask questions such as where the biopsy was taken and it continued to give us robust results. At AACR this year, we actually showed our first data in overall survival. This is atezolizumab, the same drug that Dr. Bianchini will be talking about in his talk, but this is in metastatic urothelial carcinoma, a very late-stage set of patients treated with Genentech's drug. They put this data online and we were able to download it and run our assay against it. And what you can see there is DetermaIO is clearly separating these curves, particularly around median survival. There's quite a separation between the DetermaIO positive patients and DetermaIO negative patients. And once again, coming back to full circle, the other triple-negative cohort, the one that we're not talking about in the larger part of the presentation, is a study that we have done with Dr. Naoto Ueno and Lajos Pusztai of MD Anderson and Yale, respectively. I actually have down at the bottom currently impressed that should be updated because it printed -- it went online by the Journal tonight. So this is now a published peer-reviewed paper in cancers with First Author Iwase. And what we saw there was a group of patients that were treated in the neoadjuvant setting with durvalumab. So now the fourth drug, durvalumab, AstraZeneca's immune checkpoint inhibitor and showed that we had a clear significant response -- clearly could identify responders who had been treated to durvalumab. What you see over the -- to the side is also showing a figure that we look at in the continuous score. I mentioned we had an algorithm and a cut point. We use the cut point as our primary endpoint, the first thing we look at, but we also like to look at does that continuous data seem to be separating the patients in a way that makes sense. And that's what that's showing over there. So we came full circle back to triple negative. That's largely what we're going to be talking about tonight. It is really -- we started with triple negative, went into lung cancer, went into bladder cancer, and now we're coming back to triple-negative cancer. And with that, I'd like to introduce 2 of our speakers. This is Priyanka Sharma, at the University of Kansas. She's actually been working with Insight Genetics for over 7 years. And Dr. Giampaolo Bianchini, who has been working with us ever since we reached out to him 1.5 years ago and it has been an absolutely very positive collaboration I hope to say for both sides. And with that, I will turn it over to Dr. Sharma.

Thank you, Rob, for the kind introduction and for the invitation to be part of this session. I am going to try to advance my slide. Is the slide advancing?

Yes, it has advanced.

Okay. Let's see. You all see the first slide, the title slide, unmet need for predicting immunotherapy response in early-stage triple-negative breast cancer?

Yes, we do.

Okay. Just want to make sure we're all starting from the same time point. So what I'm going to do today in the next 10, 15 minutes is just talk a little bit about the landscape of treatment for early-stage triple-negative breast cancer and how immunotherapy fits into it and what are the current knowledge gaps for us to make efficient use of this new modality. So we all know that checkpoint inhibitor therapy has revolutionized the care for many cancers and now is finally making its way into breast cancer. Breast cancer is a little late to the game for immunotherapy. I'm going to talk briefly about rationale for immunotherapy in triple-negative breast cancer, our current landscape for chemotherapy choices and leave you with some future directions for the next speaker to cover. So among all breast cancer subtypes, triple-negative breast cancer has the highest potential of benefiting from immunotherapy, this subtypes enriched in TILs and TILs stand for tumor infiltrating lymphocytes and enrichment of TIL is associated with better outcome and response to chemotherapy. PD-L1 is expressed primarily in the infiltrating immune cells in breast cancer and blocking this PD-1 PD-L1 can augment T-cell response. Chemotherapy is standard of care for triple-negative breast cancer and can have several immunogenic effects as pointed on the right side of the slide. Thus the combination of chemotherapy and immunotherapy makes a lot of biological rationale for breast cancer. We can't really talk about the role of immunotherapy in triple-negative breast cancer before we lay some groundwork on what is the standard of care before the advent of immunotherapy. So standard of care for early-stage breast cancer is polychemotherapy and that typically includes some sort of anthracycline, taxane-based chemotherapy. The early breast cancer trial is metanalysis has demonstrated for us that compared to the really old generation of chemotherapy, CMF-based chemotherapy, the risk of death can be reduced substantially by using this polychemotherapy regimen that includes DNA damaging therapy and taxane. Platinum agents have been around for triple-negative breast cancer for a few years now, mainly being studied in new adjuvant setting and new adjuvant refers to when we give treatment prior to surgery, which is increasingly being the way to treat triple-negative breast cancer as a response to treatment helps us guide future adjuvant therapy for these patients. So multiple trials assess platinum in new adjuvant setting for triple-negative breast cancer. And most of these studies included addition of a platinum to a standard anthracycline taxane chemotherapy backbone, all of them universally demonstrated that additional platinum improved pathologically complete response rate. But of course, when we add another agent, it increases toxicity. So there is increase in PCR but also added toxicity, mainly hematological toxicity with addition of another drug. There are some studies, including data from our institution that have looked at non-anthracycline platinum regimen of a taxane and carboplatin. And these data also demonstrate very nice and encouraging pathological response rates. Recently at ESMO, finally, we now see results of long-term outcomes with additional platinums. So it has been known that you add platinum to chemotherapy, neoadjuvant chemotherapy improves pathological response, but all of us were awaiting that does this increase in pathological response translate into improvement in recurrence and survival for these patients. So this data finally comes from a study called BrighTNess, and BrighTNess was a randomized Phase III study that randomized patients in 3 arms, the standard chemotherapy arm of anthracycline taxane, cyclophosphamide, second arm where the carboplatin was added and a third arm where a low dose veliparib was added to carboplatin. And what we see -- what we saw at ESMO was a very robust improvement in event-free survival with the addition of carboplatin to anthracycline taxane backbone with an absolute improvement of 11%. No EFS benefit for addition of veliparib was seen, and this really isn't very surprising since we also knew from the PCR data from this study that addition of veliparib did not improve PCR over and above what carboplatin is providing. So now moving on to checkpoint inhibition and where do they fit in and reviewing just some of the recent data for checkpoint inhibition. There are many trials that have assessed addition of checkpoint inhibitor to chemotherapy. I'll start with the German GeparNuevo study, which assessed durvalumab addition to an anthracycline nab-paclitaxel chemotherapy. And what we saw that in terms of PCR, there was a numeric but not statistically significant improvement in PCR with addition of durvalumab to chemotherapy. This study did not include any adjuvant immunotherapy, which is relevant because some of the other trials that I'll discuss included adjuvant immunotherapy after surgery. We saw at ASCO, the iDFS data from GeparNuevo. So even though the improvement in PCR with the addition of durvalumab to chemotherapy was modest. This study did translate into improvement in invasive disease-free survival at a median follow-up of about 3.5 years, the invasive disease, 3-year invasive disease-free survival went from 77% to 85%, with a hazard ratio of 0.48. There was also improvement in distant disease-free survival and improvement in overall survival. Again, this study was not -- the survival and iDFS were not the primary endpoint of PCR was, so this is all secondary endpoint analysis. On subgroup analysis, it was noted that the benefit of durvalumab on iDFS really was seen across all subgroups and was not dependent on TILs or PD-L1 status. So neither TILs or PD-L1 status are really predicted for who would benefit from durvalumab. The next study and the one that has led to approval of immunotherapy here in the United States, at least, is the KEYNOTE-522. This is a randomized Phase III trial, which assessed the addition of pembrolizumab to a 4-drug chemotherapy backbone of carbo-paclitaxel followed by AC. The study also includes adjuvant pembrolizumab for a total 1 year of pembrolizumab therapy for all patients. The study had co-primary endpoints of both pCR and event-free survival. The initial pCR data that we saw showed a statistically significant improvement in pCR in the pembrolizumab arm compared to chemotherapy. And again, in this study, as in GeparNuevo, the PD-L1 status did not predict who would benefit from pembro. Overall, tumors with PD-L1 positivity had a higher pCR rate but across both arms. And these are the EFS results at the interim analysis 4 that showed significant improvement in EFS in the pembrolizumab arm compared to the chemotherapy arm, with an absolute improvement of about 8%. Again, if you look at subgroups, there does not seem to be a specific subgroup that is not deriving benefit with confidence intervals, wide for the smaller subgroups. And on the right, you see the EFS by pCR, showing that patients with pCR have really good outcome and the absolute improvement in outcomes beyond what is achieved with pCR by additional immunotherapy after surgery may not be substantial. But again, the study was not powered to answer that question, and this should be a question that should be addressed in another study. Atezolizumab has also been assessed in the new adjuvant setting in the IMpassion031 study, similar design to some of the other designs that we've seen, chemotherapy with or without atezolizumab. This trial does not include a platinum in the chemotherapy backbone, but does include adjuvant atezolizumab, just like KEYNOTE-522 study. This study also showed improvement in pathological complete response rate with the addition of atezolizumab to chemotherapy. Again, PD-L1 status really didn't predict for would benefit from addition of the immune checkpoint inhibitor. The NeoTRIP study, which we'll hear a little bit more about in the next talk, I'm just going to briefly touch on the design, again, a preoperative study that assessed atezolizumab addition to carboplatin paclitaxel given for 8 cycles before surgery with additional anthracycline-based therapy after surgery for all of the patients. The primary end point of this study is actually EFS and not pCR. So that's one different thing about the study. And another aspect that's different about this study is the inclusion of really high-risk patients. So almost 90% of patients in this study had no positive disease and almost half at T3 or T4 tumors, overall, a higher-risk patient population. This study showed numeric but not statistically significant improvement in pCR with the addition of adhesive to chemotherapy. I'd like to point out the really robust pCR rate in just the chemotherapy arm of the study, where the inclusion criteria were for very high-risk patients. So pCR of almost 50% with just chemotherapy, where almost all patients have no positive disease is quite encouraging. There was some hint that PD-L1 positivity is associated with higher delta for pCR improvement in this study. So this table summarizes the 4 trials that I just covered and it's a busy table, and I don't expect you all to review all of the details, but it kind of points out some of the differences between the various studies. Out of these studies, KEYNOTE-522 was the only study that had EFS as its primary endpoint that we have seen the results from the NeoTRIP EFS data is still outstanding. And in the 2 studies that have reported the long-term survival, we are seeing the improvement in long-term EFS, with the addition of immunotherapy to a chemotherapy backbone. The studies have included different types of chemotherapy backbone, some included platinum, some didn't, some included anthracycline and some didn't and different durations of therapy and the adjuvant IO therapy also deferred. So overall cross-trial comparisons might be hard, but I think we're seeing trends that point in the same direction from all of these studies. As we talk about incorporating even checkpoint inhibition in our current standard of care for early-stage triple-negative breast cancer, we really can't talk about it without addressing these side effects. So immune-related incidents in early stage at TNBC is higher than what we have seen in our metastatic triple-negative breast cancer trials. Here, we're seeing grade 3 to 5 immune-related AEs that range from 14% to 15%. And this is consistent across all of the trials that we've seen data from. So of course, early recognition and prompt management is important and management guidelines are available at ASCO and NCCN websites. This slide just shows the immune-related AEs from 522 study and see about 15% of patients developed this during the combined phase in the adjuvant phase once the patients are off chemotherapy and just receiving immunotherapy by itself. This rate is a little bit lower. And the most frequent immune AEs are the ones that we know about. Also note that some of these can be permanent, which is an issue for our patients with early-stage breast cancer we're going to cure. Similar to 522, about 15% grade 3 to 5 immune-related AEs were noted in IMpassion031 in which atezolizumab was a checkpoint inhibitor. So in conclusion, addition of immune checkpoint inhibitor improves pCR and EFS and iDFS. We've seen this data from 2 studies now. And neo plus adjuvant pembrolizumab plus chemotherapies FDA approved in the United States for patients with high-risk early-stage triple-negative breast cancer. We do see improvement in outcomes, but this is probably leading to over treatment of many patients. We see a pCR of about 55%, 54% of chemotherapy alone, especially if 4-drug regimen is used and about an 80% 4-year EFS with the 4-drug chemotherapy regimen from BrighTNess. So essentially, we're curing about 80% of these patients, which is chemotherapy alone. Immune-related AEs in early stage, at least the incidence is much higher than what we are used to seeing in metastatic setting. The long-term toxicities of these drugs also need to be assessed, and we will see more data as these early-stage trials mature. Of course, this treatment is costly. So that has to be kept in mind. The role of chemotherapy backbone is still debatable in terms of what are the best partners for inclusion when we include a checkpoint inhibitor. In terms of future research, there are many outstanding questions in terms of how we would apply these drugs, especially in patients who have a pCR do they need more therapy after they've had pCR for dose with residual disease after they've already had an immune checkpoint inhibitor, it doesn't make sense to keep doing more of the same. The patients derive benefit from it. Should we use this approach for everybody or only in patients who don't have a pCR but standard chemotherapy, SWOG 1418 will answer that question. And as we're adding on more drugs can be deescalate the chemotherapy backbone. The current regimen now includes 5 drugs for early-stage triple-negative breast cancer, and there are studies that are looking at a deescalated chemotherapy backbone. And more importantly, quest for biomarkers that can identify patients that derive the preferential benefit from addition of immune checkpoint inhibitor. Thank you.

Thank you, Dr. Sharma for this terrific overview on the present landscape of triple-negative breast cancer. Now it's a pleasure to introduce you to the data that we present just 1 week ago at ESMO. For the predictive value of gene expression profile, we will not see now the results for the dynamic, but we will focus on the result that we generate with the IO score and TNBC types. So as Dr. Sharma has presented, the NeoTRIP trial is a Phase III trial with a primary endpoint event-free survival and secondary endpoint pCR rate. Overall, around 280 patients with high-risk triple-negative breast cancer that has been treated with 8 cycle of carboplatin nab-paclitaxel plus or without atezolizumab. The per protocol population, that is the one that we use for all the translational analysis include 258 patient in which we include only patient available for pCR. And in this population, we observed a numerical increase of pCR of 4.64% that was not statistically significant. So what we did is to perform RNA sequencing, actually, all core biopsies being collected prospectively at baseline so 100%. And we were able to generate successfully RNA sequencing in 93.8% of the population that correspond to 242 patients. Then this is a post-hoc study where we evaluated on pretreatment sample, the association with pCR defined as pCR in breast and axilla is the most broadly set definition of pCR of the IO score, as was mentioning before by Rob. This is the IO score calculated by RNA sequencing data, which has been as presented and associated, we have benefit from immune checkpoint in several diseases. This is -- probably is very important because we know that there are common mechanism of action of immune checkpoint. So if a biomarker work, it's not surprising, they work in different diseases because it's related to the drug, not really on the backbone of the biology necessarily. And then we also assessed as secondary analysis. So it was not the primary analysis, the role of TNBC types defined by the 101-gene score algorithm. So I want also to stress a way we work. So the RNA sequencing has been generated in Oncocyte's lab, blinded to any clinical outcome information. The RNA sequencing has been QC in my lab. The IO score binary, I would like to stress, as presented, IO score is mainly a binary score, so the primary analysis refer to the binary value of the scores, but the IO score is also continuous score. So you will see as secondary analysis, the analysis using IO score as a continuous marker and then the TNBC type. All this score has been generated by the Oncocyte lab because it's patent the code to generate the score, blinded again to any clinical outcome information. The result has been sent to my group, and we performed the association of the score of the IO score and TNBC type. We have PD-L1, stromal TILs and pCR, the most important association, according to a prespecified standard operating procedure. So again, to highlight this is important, the primary analysis is related to the IO score used as a binary score positive versus negative. So this is the first consideration that we've confirmed, 43% of tripe negative was IO score positive with a very good balance among the 2 arms. We found a strong association with PD-L1 and stromal TILs. This is not a new information, was already been observed previously. Probably this is the most important finding that we have. As I mentioned, the binary IO score was the primary endpoint of this study. As you see in the graph in azure and pink, you have the pCR rate in the IO negative. You see 40% in the Atezo arm, 44% in CT arm, no difference at all. Instead, when you look at the IO positive group, you have a delta increase of pCR of 20% going from 51% in chemotherapy to 71% in the atezolizumab arm. On the right, you see the results for the logistic regression analysis when you look at the predictive value of the biomarker within each arm. So it appears clear that the IO score negative and positive at an odd ratio 3.64 with a p 0001 (sic) [ 0.001 ] in the CT/Atezo arm and the predictive value was essentially absent in the chemotherapy arm. The most important test in randomized trial to demonstrate that the difference that you observe according to the biomarker and the treatment arm is not a finding that can be explained by chance is what is called the test of interaction. Test of interaction has a p 0.029, so it is statistically significant. Just to remind, the test of interaction is really demanding. So it's unusual in a trial that is not very large to find a statistically significant test of interaction and those to notice that the significance is achieved including in the model, PD-L1 as dichotomic variable and stromatives. We also test as exploratory analysis, the value of IO score as a continuous score. On the left, you see the level of expression of IO score in the pCR group and residual disease group of Atezo arm. And on the right of the box plot, you see the same analysis performed in the chemo arm. We got some tests, you see that there is a higher expansion overall of this continuous score in the pCR group compared to residual disease only in the chemotherapy atezolizumab. So mirroring what we are seeing when we look at the score as a dichotomic score. On the right, you see an analysis that use the spleen to describe the association with the probability to achieve a pCR where the IO score as a continuous score in CT/Atezo arm, the azure line and in CT arm, the pink line. And it is here that if you go above a certain threshold, you have continuously increase of the life achieve at pCR in CT/Atezo compared to the CT. And probably, there is also an incremental effect going for higher level of the score. And above, you see just the result of this analysis according to the logistic progression. So this is a comparison that was not, I would say, planned, but because with RNA sequencing, we assess the association with pCR at baseline of 151 selected tumor intrinsic and extrinsic gene signature. What we noticed and was very interesting observation that none of the immune-related signature that come from HALLMARK, the Consensus cluster, Nanostring outperformed IO score at the baseline. So the only 2 immune-related signatures that were associated with pCR in atezolizumab, again, we didn't find an association with the chemo arm have a p-value that is between 0.01 and 0.05. So it was underperforming compared to the IO score. And all the other immune-related signature was not significant at all. TNBC type, here, you see the distribution of TNBC and there was fairly distributed according to the treatment arm. Also TNBC type are associated with PD-L1 as you see, the LAR and the mesenchymal has the higher rate of PD-L1 negativity. Just to remind, PD-L1 in the NeoTRIP has been defined using the VENTANA SP142 with IC more or equal 1% of the area as cut-off as usually is adopted as companion for atezolizumab. On the right, you see the stromal TILs. And again, mesenchyma was the fully infiltrated subtype. These are not new finding, but interesting itself. Here, you see the rate of pCR according to the TNBC type across the 2 arms. All the test of interaction were not significant. So we cannot claim and this is a difference with the IO score that the TNBC type are able to demonstrate a different probability to benefit from mutual point. I would like to stress that a not statistically significant test of interaction doesn't mean that there is no difference in the effect. Just we cannot measure the effect. And the TNBC type was most predictive of pCR. So that's another very important and probably very clinical useful information, even not completely new and in some ways, the confirmation is basal-like 1 has the highest pCR rate and the last subtype as the lowest pCR rate. So in conclusion, the IO score, but not TNBC type is predictive of atezolizumab benefit over CT alone. The significant test of interaction confirmed that this predictive value is there even after adjustment for PD-L1 and stromal TILs in the model. And as I mentioned, the IO score outperformed all the other immune-related incidents. This highlight was Rob described by the development and the discovery of this score is not capturing just the immune component of the tumor, but also tumor-related features like EMT and probably some other environment-related feature like cough that are not captured or described by the typical immune-related features. So -- and these are due acknowledgment of all who contribute to the NeoTRIP trial. We have a lot of supporter and of course, also Oncocyte who provided support for this very nice and relevant translational study.

Thank you very much for the presentation. [Operator Instructions] Okay. So our first question comes from Bruce Jackson at Benchmark. Bruce, you may unmute your line. Bruce, can you hear us? Right, moving on to the next analyst. So we have Mike Matson of Needham. Mike, you may unmute your line. Mike, you are still on mute.

Sorry. Sorry about that. So I guess my question would really be for both of the physician speakers. The data looks pretty good here. Would it really have to happen for you to start to use the term IO in your daily practice. Do you need to see more data? Do we need reimbursement guidelines et cetera?

So if I may I'll take that on since we actually have immune therapy approved here to use in the United States. So for any test to make selection or deselection of treatment, especially for early-stage curative cancers, the gold standard is for it to be studied in a prospective trial. So hypothetically, when we're thinking of a trial design, one would design a trial where patients would have this test upfront as part of research, not routine care. And those that are positive then would be assigned to receive chemo plus immunotherapy because we think that, that's where the greatest benefit is and those that are negative would receive some other form of therapy. So we have to pan out in a prospective trial where the differential benefit is seen prospectively only in patients where biomarker positive.

Dr. Bianchini, if you don't mind, I know Dr. Sharma has a U.S. perspective. I think given that our effort is global, maybe you can give your thought.

I think that this is something that we have to keep in mind for the biomarker development. When you use what is called retrospective prospective approach was very well described by Richard Simon in 2009 in the JCI paper. You provide a good level of evidence that usually require at least retrospective prospective confirmatory trial of the same quality. So you should theoretically validate this in an independent trial. So I believe when Richard Simon take this position, some stress that the prospective trial are the only one that generated Level 1 evidence. But he said that if there are the conditions -- and one of the condition for instance, was that in a randomized trial, you have to have at least 85% of the randomized population to assess a biomarker. So that's the first consideration. The second one is the backbone of chemo. Here, the backbone of chemo is different from the one that is used in the KEYNOTE-522. That is the only treatment that probably will be approved. And I would not be in favor to translate directly what we found in the NeoTRIP to a trial with a different chemotherapy backbone. So that's another reason to say before thinking to any potential clinical utility. I, for sure, a validation trial will be needed. But these, I would say, stress more the potential of the biomarker, more than what I will do tomorrow in the clinic. So I see this in this direction. It's more an understanding of what the biomarker can provide. And this has actually demonstrated that the biomarker work for the purpose. And other things I completely agree with Sharma is the clinical application, especially when you have a curative setting. And I -- if you use something to deny a treatment opportunity, you have to be sure that, of course, you are not denying a curative option.

So the next question will come from Bruce Jackson at Benchmark.

So fabulous presentation, I've got one question and one follow-up. So the basic question is to what do you attribute the success of the DetermaIO test? Rob mentioned that you're measuring both hot and cold tumor attributes. Dr. Bianchini mentioned that you're measuring some tumor-related attributes. Maybe you could just like go into a little bit more detail about what do you think the advantage of this test is?

Dr. Bianchini, you want to take a shot at that?

Yes. As every test, first, obviously, how exactly work the test is, I would say, patented I don't have really that kind of information. But what I can tell you because there are release data is to describe which are the gene special features that are associated with the IO score high and low. And essentially, this is -- so we know that the tumor and immune system are in equilibrium. They interact. And of course, if you just picture in the immune system without describing what do the tumor or the opposite. It's like having just a vision of the moon is what we can see from the half. But there is another side. And I believe that biomarkers that are able to capture both tumor escape feature because, of course, it's the tumor that is driving the immune escape, it's not the immune system. The new system is following something. The immune system tried to attack the tumor, and the tumor is react, orchestrating way to escape. So they make absolute sense that with a biomarker, you have to capture how the tumor try to escape the immune system and how good was to escape. That's why you also need immune infiltration measure. I don't know if I answer your question.

Yes. Let me add some color, Bruce. I think one of the things that makes this unique is, as Dr. Bianchini said, we -- as you know, we reduced this from a 2,000-gene classifier. And if our own Dr. Ross, who's a molecular pathologist, he would spend a little time explaining the importance of molecular subclassification at a broader gene expression phase and then reducing the practice, these relevant RNAs that come from both the immune modulation side as well as the quiescent or mesenchymal transition side. And as you know, we iterated algorithm after algorithm to get to a point where we believe we have covered and now I think with the Rob Seitz's work and the heat maps where we've looked at 10 solid tumor types with the same threshold and have been able to see very similar results across all 10 tumor types. I think we're pretty confident that we found the right 27 RNAs that give us the right mix of relevant information between quiescent and immune modulation. So it's a -- more studies to come to support that. But Rob, I saw you jump on, so if you wanted to add some color commentary.

Obviously I was just going to say, I really do like that analogy of only seeing one side of the moon because we said that what makes us different is everybody seems to be looking at what's making the tumor ready for an immune checkpoint inhibitor. This is what we think makes DetermaIO different is what's making the tumor look for an escape. And I think that's a very good way to look at it, and I'm going to steal Dr. Bianchini's description next time I give a talk.

Thanks.

And it's in the -- it is showing the differential benefit for immunotherapy. Thus far in breast cancer, many markers have been studied, and they all kind of showed general sensitivity to whatever treatment you're giving. So a group is more sensitive compared to another. It doesn't help you preferentially select one or another agent within that milieu of poly treatment. So I think that's another advantage.

Okay. That's very helpful. And I just have one follow-up, if that's okay. The data in triple-negative breast has been really remarkable. You've also done all the subtype analysis. Can I just get a brief commentary on the extensibility of the test into the other solid tumor types? You've already got a fair amount of data. Do you think that you've got everything pretty well covered to move into those other tumor types?

Yes, I'll take that one. We've actually have publicly announced that we're doing a collaboration with the clinical trial with the GONO Group, likewise here in Italy in colon cancer, and it's a randomized trial, and they introduced some very interesting data we thought at ESMO as well. But one of the things they said was it's important to move into translational studies by which they mean at least in part, biomarker study. So we're going to be moving into that area as well. And then we are looking -- and a lot of these are still confidential, but pretty much everything you saw I mentioned previously is areas of future -- are areas of active work we're doing. I mean we're not leaving a single stone unturned when it comes to this.

So the next question comes from Thomas Flaten at Lake Street Capital Markets.

Hey Ronnie, I was wondering if you could maybe reflect you back to the first question about a prospective validation study. Your thoughts on that and how you would think about that against the numerous tumor types and the numerous potential underlying immunotherapies that you might be looking at. Could you just provide some context for your thinking along those lines?

Sure. And Rob, add some color commentary here because Rob is one point for all this work. But clearly, the -- he mentioned the GONO Group and PSA in colorectal cancer, that's obviously one -- another shot on goal. As you guys have heard me mention on previous calls, we are working with pharma companies where we are involved in and working on some pilots to become part of their prospective CDx work as a companion diagnostic. And so Tom really to attack this across multiple solid tumors, we have to have partners. I mean we can pick the 1 or 2 that we can fund and go after, which we're doing. But obviously, to have partners across the globe that can help us. So obviously, again, we appreciate the work that Dr. Bianchini has done. We've got the team in GONO. We've got other things that are brewing alike that, that will give us multiple shots on goal. So short of having a pharma-type budget to do studies, we have to be a little more creative. So we're doing the best we can. The good news is there's -- I think, as Rob said, we've got significant momentum across multiple tumor types now with multiple groups. We talked to SWOG and obviously, there's interest there. I mean there's interest in a lot of the groups that we work with. So I can't -- I want to compromise some of that by hedging it, but a lot of that stuff is ongoing, and you'll hear us talk more about it at earnings call.

And for Dr. Bianchini, so we haven't seen the event-free survival data yet. Is it your intention to go back and essentially redo the study with that endpoint as opposed to pCR?

Absolutely, yes, when they will be mature because the trial is event-free survival as primary endpoint. The final analysis is event driven. So that means that it depends on how much recurrence we will have, we will have the data sooner or later. But of course, all this analysis will be repeated with the event-free survival data when they will be available.

Got it. And then a final one for Rob. So we've talked about PD-1s and PD-L1s exclusively. Can you maybe speculate for us about applicability to some of the next-generation immune therapies like the LAG-3 TIGIT that family of drugs? Would this be equally applicable there? Would -- just curious to get your thinking on it, maybe I'll leave it at that.

Rob? Sorry, Dr. Bianchini, please, if you're able to go, otherwise, I'll let Rob take it, so.

I missed a piece of the question, so I don't know if it was for me or for others. Sorry.

Rob, do you want to take that question?

I'm sorry. I was making sure I didn't get feedback from his answer. Could you repeat the question very quickly?

I was wondering if you could speculate about the applicability of IO to other classes of immune therapy, especially the next-generation ones like the LAG-3s, TIGIT, et cetera.

Okay. So I will not outrun the evidence out on this because we've not done the study. So I won't outrun the evidence. What I will say is that we -- it simply depends upon how similar the micro environment -- the microenvironments rightness is for a LAG-3 versus a PD-L1. And what do I mean by that? I mean we have always said that this really is not a mechanism of action test. It is really a -- is the tumor microenvironment right for a PD-1 inhibitor, our first study in non-small cell lung cancer said, yes, it was. We then said, is it right for a PD-L1 inhibitor. And the data we've gotten in TNBC 2 studies and bladder cancer has said, yes, it is. So the next question is, is it right for LAG-3? And we won't know that until we do it, but we're not measuring something that says we're looking at what LAG-3 actually does. We're actually saying is the tumor microenvironment. So that's when it comes to LAG-3. TIGIT, I guess, X4 is somewhere out there. Whatever is out there, I want to -- we never know until you do the experiment. But again, we're not trying to say we're not trying to measure what LAG-3 is actually doing in the tumor. We're trying to say, is the tumor ready for LAG-3. So far, it's worked with PD-1 drugs and PD-L1 drugs.

Yes. I think Rob, let me just piggyback on that because we had an online question that kind of is a follow-on to Thomas' question, and that is can we speak to the traction we've had with the developers of these types of immune therapies. And I think the answer there is because the test as we're developing it and now that we've proven across the solid tumors has a linear -- sort of a linear curve to full immune modulation, which is more monotherapy oriented to no immune status at all, completely quiescent to what else can you give there. But in the middle, there's a lot of interest, as you might imagine, because of the prevalence of pembro and some of the other larger first-to-market drugs. These companies are very interested in finding that sort of intermediate patient that does need a complementary therapy. And therefore, as Rob says, even though we can't specifically today say TIGIT versus a VEGF or versus an IL-15, the idea is we can find those patients, therefore, there's interest in us working with those types of companies to get involved in their trials, Thomas. And so that's where the pilots are really focused.

The next question comes from Mark Massaro at BTIG.

Okay, great. Can you hear me?

We can, Mark.

Excellent. This question is for Dr. Bianchini. It looks like you compared DetermaIO to 151 tumor intrinsic and extrinsic gene signatures. Can you expand on what these signatures are and how you came up with them? And just maybe the impact that you believe this result may mean for the test itself?

Just to recap, I -- it's very tough to summarize what are within this signature. If you look at the ESMO slide, you find the reference of -- because the HALLMARK, the Consensus cluster are really predefined. We actually just calculate the signa score for each of the signature, and we run them as pCR. If you go to the ESMO presentation, you will find several of the analysis. I just would like to point out that I don't say that the IO provide more information to all the 151 signature because that was not a purpose. We actually -- we did have the aim to compare the IO compared to all the others, but was an interesting observation that because we have a huge amount of immune signa to macrophage 2, 1, B-cell, T-cell memory and so on, was really surprising that no one of these immune signature outperformed the IO score. And this highlighted this is not -- even it's correlated with PD-L1 and TILs. This is not just an immune-related signature. It's something different. That is, I would say, the purpose. If for your curiosity, you ask me what come out as more associated with resistance. For instance, we found like angiogenesis, some stromal-marker, EMT, some immune-related feature, some metabolic-related feature. Probably they are captured by the IO score, but we really didn't make a comparison because that was not, I would say, the purpose. Also because if you make comparison, you have to adjust for multiple testing. And because we test 151 gene, indeed, if you go through the presentation, you will see that we apply the typical for discovery rate adjustment because we make many tests, many assumptions. Instead, the IO score was one shot, one biomarker, one test, one observation. It could have been positive or not. That is a huge difference.

Okay, great. Just my last question here. It's for you, Ronnie. I guess it's a business question. I think you've talked about wanting to launch DetermaIO in the clinic here in the fourth quarter of this calendar year. I guess you've now accumulated quite a bit of data on DetermaIO across disease types. Do you think you need to wait for the EFS readout in mid '22 before submitting for Medicare reimbursement? I'm just trying to get a sense for how much evidence do you think you need to build and when the clock can start maybe for you to -- for us to think about getting Medicare?

Yes. Realistically great question, Mark. And by the way, I've got a couple of those questions online as well. So I'll roll them all into one answer. I think for us, to be candid, to become standard of care, we've got -- it will be a never-ending stream of data and studies and things that we'll do over the next 3 to 5 years. But for now, what we're going to do is focus on the indications where we have enough data to petition game here in the United States and start thinking about reimbursement for those that we have powered. And so the goal is still to launch the test to early adopters. We have a significant number of sites that are very interested from early adopters here in the United States. And those will be very important for us because these are, as you might imagine, some of the key opinion leaders in the industry around different disease areas, breast, obviously being one, but certainly lung and bladder, where we've got significant amount of data. And then we hope to build a lot -- much more stronger data case for the ones where we have good indication that it works, but we don't have enough data to submit for reimbursement. So our goal is still Q1 next year to be engaging with the CMS, get some feedback from them and then hopefully have a package in a dossier to go to them by late either first quarter or second quarter, sometime and then hopefully get in the cycle for the fall. So that's still on track for that. As you know, Mark, we are launching the DetermaTX assay, which is already -- has a blanket LCD in the United States for over 50 genes. And so the goal is to -- and we still get feedback every day that we need to be providing those together. So I think that the idea will be to provide those together. And obviously, the TX will be reimbursed. So as we look at revenue next year, I mean, candidly, TX will be the one that would drive any revenue next year. We expect IO, if we get reimbursement from CMS that would be in the second half and probably late second half of next year. I would say this that because immune therapy is not necessarily a CMS-only patient, in other words, these patients could be younger, the idea here would be that we would get some private payers interested just because of the health economic benefit to them from, I think you guys see the numbers I see. And I don't know, Mark, I don't think you wrote the analyst report, but somebody wrote an analyst report recently that talked about $125 billion of immune therapy spend in the United States alone by 2025. And that right now, based on a 30% durable response on average across the solid tumor types that you're wasting $70 billion on drugs that aren't going to impact the patients. So I think from where I sit and the colleagues I have the blessing of working with at ASCO, et cetera, there is a concern of overtreatment and the idea of knowing how to manage patients better, especially as these drugs move out of salvage treatment, late stage into earlier treatment. And you heard Dr. Sharma today, I mean there is utility of these drugs earlier stage, and we want to obviously provide a tool that allows physicians to know is that monotherapy at early stage is a common door therapy or will it work at all, et cetera. So a lot of work ahead of us across all the solid tumor types, but we're going to put more wood behind the arrow, if you will, for the ones that we can get to market next year. All right. Sarah, do we have any more online questions?

No, Ronnie. That concludes the Q&A session.

Well, listen, let me take a moment and just say thank you, Dr. Bianchini. Thank you, Dr. Sharma. Again, we know it's late in Italy. So Dr. Bianchini, we thank you. And Dr. Sharma, we know you have a busy clinic, so we appreciate you taking the time out today to present. We appreciate everyone's attention. We thank you for the thoughtful questions. And again, more to come, but we appreciate your attendance. Thank you.