GenSight Biologics S.A. (SIGHT) Earnings Call Transcript & Summary

Hello, and welcome to GenSight Biologics reflects 78-week results call. My name is Molly, and I'll be your coordinator for today's event. Please note that this call is being recorded. [Operator Instructions] I will now hand you over to your host, Bernard Gilly, Co-Founder and Chief Executive Officer to begin the conference. Thank you.

Thank you, Molly. Ladies and gentlemen, good morning for those of you in North America, and good afternoon for European participants. Thank you for attending our special KOL webcast dedicated to the results of our clinical trial REFLECT. I'm very glad and very grateful to them to introduce our 2 internationally renowned experts who will detail for you the result of this third Phase III trial of LUMEVOQ in the treatment of Leber Hereditary Optical Neuropathy. There -- the first one is Professor José Sahel, is a cofounder of GenSight, is Professor of Ophthalmology at the University of Pittsburgh Medical Center. And he is also the Chair of the UPMC Eye Center. And the second one is Professor Robert Sergott, is Professor of Ophthalmology and is Chief of the neuro ophthalmology service at the Wills Eye Hospital in Philadelphia. Before handing over to them, I would just like to briefly summarize the status of our product, LUMEVOQ. On the next slide, the results of REFLECT that you will see, do confirm for the first time, this is the first Phase III trial -- the third Phase III trial confirms the efficacy and the safety of LUMEVOQ. This product is currently under review by the European Medical Agency for approval, and we indeed expect final approval by the second quarter of 2022. In the meantime, of course, we have started to deploy our commercial structure all across Europe and particularly in the 5 more important European countries, U.K., France, Germany, Italy and Spain. Of course, on the back of this result of REFLECT, we expect to be able to have a pre-BLA meeting with the FDA sometime in the last quarter of this year and that should open up the road for BLA filing by Q2 of 2022 for the U.S. Finally and interestingly, we have a very high and strong demand for -- from patients and from doctors for accessing LUMEVOQ. And we -- to respond to this, we have launched a large expanded access program in U.S. and in Europe. In France, we have a special compassionate plan called ATU that stands for temporary authorization for use. And then on this plan, we can invoice the product to the hospital and it is being reimbursed. And overall, it's important to know that we've been able to already treat more than 50 patients in U.S., Germany, Italy and France in less than 1 year, including this COVID situation, which certainly witnesses the strength of the demand for an early access to the treatment. With this, thank you, and I will hand it over to Professor Sergott.

Thank you, Bernard. And it's indeed a pleasure to share with you these compelling results for a rare disease. And as all of you know, every time we do a clinical trial, we learn more. And in GenSight's program for gene therapy for Leber Hereditary Optic Neuropathy, we learned a lot in RESCUE and REVERSE. The REFLECT results will show you that those results have been confirmed and there's even more learning that has occurred. Leber Hereditary Optic Neuropathy is a rare genetic disease produced by mutations in mitochondrial DNA. In this trial, the mutation 11778 was replaced with intravitreal virus injection into middle of the eye with gene therapy. And this was done with an adeno virus vector. And you can see here that in the first arm of this randomized, double-masked, placebo-controlled trial that one group of patients had the first eye injected and the second eye injected both with the study product. In the other arm of randomization, and we are now going to tell you the results of 98 patients vision loss of less than 1 year, and these came from centers in the United States, Europe and Taiwan. The second eye received LUMEVOQ, the gene therapy in the first eye and then placebo in the second eye. Now this is a very unique design. And just let's spend a minute or 2 on how you're going to see the results and the special parts of this design. The placebo, and that's in the lower right-hand corner of the slide in light green. The color coding is very important to remember when we show you the results about the patients, where they started with their vision and where they ended up with their vision, okay? So this is a trial who -- is designed to -- and a treatment to restore vision in patients who are legally blind. So the placebo injection is a balanced salt solution commonly used in ophthalmic surgery. It does not contain genetic material, it does not contain protein. Very important to remember what the placebo injection includes as we begin to look at data and also its safety. And as in any clinical trial, there are 2 sides of the coin to look at safety and efficacy. And those 2 actually interact. So on the next slide, we will describe to you the subjects studied. As a rare disease we are learning more and more about waivers as we do this study. And I want to talk to you in a minute about perspective here. When I trained in neuro-ophthalmology as a fellow at Bascom Palmer Eye Institute, in 1980, we only check vision with the usual chart that you see when you go for your eye examination. That's not done in clinical trials anymore. We check vision now in a very specialized way that is much more accurate. And when we change techniques, we see different results and when we follow patients closely, we see different results than within the medical literature. And then we learn more about this disease. So every clinical trial is a learning experience. And you may ask, why was 1.5 years chosen to look at the study results. In REVERSE and RESCUE, GenSight's initial pivotal trials in Phase III, we first looked at a year. And then we looked at 1.5 years and more improvement was occurring. So in this trial, we chose 1.5 years. Now these patients are going to continue to be followed. And we will probably be back to you to show you even longer term data. Gene therapy is a new entity. We don't understand everything about its onset, what we would call its kinetics of action, that is how rapidly does it work. And how does the eye, the retina, the optic nerve accept this therapy and how is it integrated into better visual function for the patients. So 108 patients were enrolled, 10 failed to qualify as they were screened. Our intention to treat population that were randomized were a total of 98 patients, 48 in the bilateral arm, 50 in the unilateral placebo arm that you saw. The number of patients available for the primary endpoint analysis was 46 in the bilateral, 48 in the unilateral, 1 patient discontinued and that was not because of an adverse event, either medically or visually or anything with his eye, it was because of failing to keep his appointments for the study visits. So he was lost from the final results. If we can move to the next slide, please. This describes our demographic characteristics of the subjects. We mentioned mitochondrial DNA earlier. We inherit mitochondrial DNA from our mothers. And this mutation that is much more prevalent in men than women. We thought at one time, it was a 100% male disease. That is not true. This is, again, a learning lesson in a rare disease as we get better diagnostic methods and better ways to look at the function of the optic nerve and retina. So there are 3 mutations for labors, the one study here, 11778, 14484 and 3460. So all the results today are from 11778, and this is breakthrough data, in my opinion, and may pave the way for additional treatments for the patients with other mutations. This is precision specific medicine. We're targeting the exact gene that is responsible for patients' blindness. The age of the patients was quite balanced in 31 years, ranging from ages 4 to 73. Again, we used to think this would affect only patients in their teenage years or in their 20s, but with the development of testing for these mutations done by outstanding medical researchers. We can now take a blood test to determine when we see a patient if they do have Leber Hereditary Optic Neuropathy. Here is -- on our next slide, Slide #8, you can see the characteristics of these patients. These are the clinical characteristics. As we move forward in gene therapy as well as other precision therapies that target mitochondria, you're going -- we're going to define these patients even more precisely in their characteristics. For now, we define them, how long do they have loss of vision. So the duration of loss of vision highlighted in yellow was 8.3 months. Then we're going to look at the first eye affected and the second eye affected. So here, we see -- you see the term LogMAR, best -- BCVA is best corrected visual acuity. LogMAR is a special way to record vision that makes our statistical analysis much easier, and we will translate that for you into what we're all used to, that is, is this the big E on the chart, is it less than the big E on the chart, what we call the Snellen visual acuity that was developed by Professor Snellen in the Netherlands in the 19th century. So we see here the first affected eye of 2,800. This is less than the big E on the chart. And the same is true for the second eye here. Second, not yet affected eyes when they entered study, we have, again, a visual acuity of 2,600. So let's move on now to the next slide. And now we're going to look at time from treatment to the endpoint of 1.5 years. So the mean time to treatment was balanced in the 2 groups with a mean of 20 months in the bilaterally treated patients. So remember, they're getting an injection of gene therapy in each eye. The injections are separated usually by a day or 2 or could even be given the same day. So these patients were getting 2 doses of GenSight's medication LUMEVOQ. The other patients, again, treated at the same time are getting only 1 dose and the other eye is being treated with that ophthalmic saline solution. And let's move on now to Slide 10 and then 11. And now we get into the part of the study that will give you the results. Now that we've set a little bit of a background about the disease state, how the trial was conceived and performed, I would now like José Sahel, Chair of Ophthalmology at the University of Pittsburgh and really, the scientific brilliance behind this therapy for patients. José?

Thanks, Bob. So very happy to share these results. So as you heard from Bob, patients were treated once the first eye was affected and the average time was 8 months when patients were treated. The second eye was either already affected or about to be affected. So when you look at the curves, and Bob already told you that the color code is very important. So the red curve corresponds to the first affected eye for patients that are being treated bilaterally. And as you can see, we start at a vision, which is around 1.6 LogMAR. And over the course of the treatment at 18 months, they end up with an improvement of an average of 12 letters, which is quite impressive given the natural history of the disease we know. When you look at the fellow eye of the same patient, the second eye, which may have been affected at the time of treatment or was more recently affected. They start a bit higher, and they end up with a level of vision that is similar to the eye that -- the initial eye that was treated. When you look at the patients that are being treated unilaterally, then these are the curves in green, as you can realize, they start also low, like the first eye of patient treated bilaterally, a small difference, but not significant. And then end up with an improvement of vision, which is around 8 letters, so not 12 letters like we have seen unilateral but under. And the patients that are treated on the fellow eye, the second eye with placebo with saline in the eye injected but with nothing efficient. They start with a higher level of vision, like the second eye in the bilaterally treated patient, and they end up with a similar level of vision. So as we had seen in RESCUE and REVERSE the 2 eyes are trending in the same direction. And there is clearly a dose effect between the unilateral versus bilateral eyes. This is not a statistically significant change, but it's at least one line of differences of vision. Then you may object, given that the patients don't start at the same level, maybe this is the level we start that is influencing the outcome. And when you move to the next curve, which is Slide 12, this has been normalized. So all the vision at baseline is now put at 0. And then you look at the change in vision between the 4 populations. The first group is the red corresponding to bilaterally treated patients in the first eye and it's extremely significant statistically as compared to baseline. This is an improvement of 0.23 LogMAR, which is 12 letters. The second eye of the same patient, which is in yellow, orange yellow. These patients are improving also as compared to baseline, the improvement is around 8 letters, 0.15 LogMAR. It's also statistically significant. When you look at the patient that have been treated on only 1 eye there's LUMEVOQ, again 1 eye there's a placebo, these are in green. The first curve in the strong green is showing the improvement, which is in the range of 0.15, 8, 9 letters. And the patient that got placebo got some improvement, but not as the others. And this is the only group where there is no statistical significance between baseline and 18 months. So clearly, the 3 groups of eyes that got the treatment, got a significant improvement in terms of statistic. And this improvement is less impressive in the placebo eyes. But as we've seen on the previous slide, the eyes are trending in the same direction. So clearly, this is confirming what we had seen already in RESCUE and REVERSE. REVERSE, as a reminder, is more similar to the first eye of this patient, around 8 months between 6 and 12 months, and RESCUE is for some patients, more similar to the second eye at least for some of them. And if you compare the curves that we showed already in several publications in top journals, these effects were significant, but these effects were certainly trending together from the 2 eyes, which explains some of the issues we faced initially that we learn from them and they show clearly that this bilateral treatment is real. Bilateral efficiency is real and that bilateral treatment is providing a better effect as compared to unilateral treatment. This can bring us to the next slide, which is giving the same numbers. So the patients that are treated bilaterally with LUMEVOQ, the first affected eye the improvement is 0.23 LogMAR, 12 letters, very significant. The second affected eye is improving by 8 letters, with 0.15 LogMAR, also significant. The first affected eye treated unilaterally is also improving in the same range, and this is significant again. And the placebo treated eyes have some level of improvement, but this is not statistically significant. So clearly, the results are first confirming what we had seen with LUMEVOQ in RESCUE and REVERSE, but we're also providing some evidence that treating bilaterally probably for a dose effect is providing more benefit to the patient, at least in terms of trend. And when you look at the comparison to what we call a nadir, it's not us, it's a term that is used by many people working with these patients. As you know, patients are continuing to drop vision initially because the time for the treatment to become efficient is not immediate, and they lose vision. And when you look at the difference between nadir, the lowest point of vision and the last time point, then we get above the famous 0.3 LogMAR in terms of improvement, it's 19 letters extremely significant. The second eye is also above 0.3 LogMAR. And this is 16 letters, very significant. The first eye treated with LUMEVOQ can be unilateral also has an improvement of 19 letters, so very strong as compared to nadir. So you have to normalize to nadir to see a difference. And the placebo has an effect as compared to nadir. But as a reminder, the placebo is benefiting from the bilateral diffusion of a vector as we showed and published recently. So all these results are exactly confirming what we had seen, but they are providing even stronger evidence for the activity of the bilateral versus unilateral. So I think I'll hand over again to Bob for the responder analysis, which is also a very important one because you may say that maybe a few patients are driving the response. And as you can see, this is not true. This is really representing the vast majority of the patients in the trial.

Thank you, José. And we will now move to the responder group. This is a planned statistical analysis of the patients who improved. And as we mentioned early on, when we were describing the patients not being able to read the big E on the chart, that is 2,400. Now -- and these patients were 2,800 to 2,600 on average for starting. So now we see the patients back after the injections. Their vision is checked by our technicians, the technicians at every site who are masked to whether the patients had placebo or whether they've had bilateral injections of LUMEVOQ. And so this is all now done without bias. The ophthalmologists to reexamine the patient is not the principal investigator, who first saw the patient, but someone else who's coming in, again, masked to the treatment so that we prevent bias. And if we go back to the color code again, you will see in red, 68% of patients who could not read the big E on the chart. In ophthalmology, we describe that as counting fingers vision. You can see how many fingers we're holding up. Maybe you can't see that. You can only see our hand move back and forth or just perceive light. And 68% of the patients who were off-chart moved to on chart. And as José mentioned, if we follow our color code, remember, the first eye is injected, second eye injected, the one eye injected in the dark green gets a medication effect from that injection, but no effect from the placebo injection. And as José said, when we first saw this bilateral improvement, everyone questioned it, very appropriately questioned it. Was this just the patients trying harder? Well, we've controlled for that. And now we see the same bilateral effect in 3 independent trials at 13 independent centers around the world. So it's very hard to say that, that is just the patients trying harder, especially when you saw the levels of improvement of 15 letters. Our collaborators and who we work closely with, with the EMA, FDA and other regulatory authorities, when you improve by 15 letters, that's doubling your visual angle, you see twice as much. So let's now look at Slide 7. And this is the responder rate eyes and subjects from nadir. And as José very clearly explained the treatment is administered. It takes a while for it to work and the patients dropped their vision and then they regained vision. And I would just direct you again to the responder line, bilateral and unilateral injection, and we see a clinically relevant recovery in 69% of patients bilateral, 62% in unilateral and the odds ratio is 1.3%. Now Slide 18 looks at the same type of data. And this is the improvement of at least 0.3 LogMAR from nadir. That is 3 lines of vision, 15 letters examined at 1.5 years. And again, the results are quite similar, 69% for bilateral, 64% for unilateral. Now on the next slide, #19, we see that if you're on chart, how do you do? Well, let's look at this in terms of as José mentioned, we normalize the curves for all the differences in the patients. And that's why you saw the differences in the mean and the least square mean analysis. Now if we're now going to stratify our patients, are you on chart or off chart? So on chart, you have better vision. And in my clinical practice over the years, I'm still looking for the first disease where it's better to treat it later, whether you're talking about multiple sclerosis or any other condition. Usually, when we treat not so much just in duration of time, but in severe disease, we see better results. And here, in the responders, if they were on chart, we have an 85% rate of improvement and 72% in the unilateral rate and our odds ratio increases to 2.24. The last topic we will discuss before opening this up for questions to the audience is safety. In every clinical trial, we always examine the efficacy, which you just saw and the safety. And the purpose of this is so that our regulatory partners are able to write a label that correctly informs patients and clinicians about what can they expect with treatment with this medication. So on the safety side, let's start with the whole buy. We saw no adverse events systemically that is there is no problems with any effect of the gene therapy on the heart, the lungs, and on the liver, okay? These are all monitored very closely in these trials as they should be. Now let's talk about what happened in the eyes. So here, we have TEAE stands for Treatment-Emergent Adverse Events. And now we're going to categorize them, as you see in the table, by mild, moderate, severe, this is the way they are always characterized for the regulatory agencies. And let's go now across on the first line. First eye, we see 44%, and this was an inflammation in the eye in most cases. What we call an iritis and uveitis. And we can now see again bilateral effect as we go across. And I really want to also concentrate on the placebo eye. And you can see there's also some inflammation there. And as we go down to stratify our results by mild, moderate and severe, you can see this really performs much the same way with the placebo eye. That's the eye that received only saline, did not receive gene material, no mitochondrial DNA, no protein, we see inflammation. This gets back to the controversy of the bilateral effect. Somehow this eye has become inflamed. We can discuss whether that inflammation comes because some of the genetic material got into the bloodstream, but we didn't see a significant change in liver function tests as we would expect. And as you know, GenSight has sponsored work by Professor David Calkins at Vanderbilt in experimental situations with animals that shows the gene and its products can migrate from 1 eye to the other by the optic nerve and even get back to the brain in areas we call the thalamus and the temporal lobe. So now we not only have a bilateral effect on vision, we have a bilateral effect on inflammation. And in my mind, that is confirmatory that we are getting gene product to the uninjected eye. Next slide, please. So the intraocular inflammation was well treated. We treat these patients with topical corticosteroids. Sometimes, there would be -- it's necessary to use a short course of oral corticosteroids. And we also then divide the inflammation into anterior that is in the front of the eye, inner vitreous in the middle of the eye you see that on the left and posterior inflammation, and that's right on top of the optic nerve and retina. And we, in the future, will be looking more closely at this inflammatory reaction to see how that may help us look at responders and nonresponders as well. So to summarize the safety, first and foremost, no adverse effects on any other organ systems. When you inject a protein gene therapy material into an eye, we expect to see the inflammation in the injected eye. What we don't expect to see is the inflammation in the other eye, but we did see that. Both of these were very manageable. There is no problems with controlling the inflammation with a standard way we take care of iritis and uveitis in clinical practice. We'll now move on to the summary slide, and I'll have José talk about this as well. So José, why don't you summarize what the efficacy and safety have shown us.

Yes. Thank you very much, Bob. So as we said, this is first the confirmation of between you already from REVERSE and RESCUE, which is that eyes are improving were treated with LUMEVOQ both as compared to baseline and as compared to nadir where the effect is even stronger. The contour lateral effect in the placebo eye is confirmed also as we had seen in REVERSE and RESCUE. And the value of that is first. It's the first study going exactly in the same direction, which makes me very unlikely, but this will be just a learning effect or just on the unexplained improvement in the untreated eyes. With untreated eyes are certainly treated through the bilateral diffusion of a vector as we showed or potential improvement of time. The second thing, which is this trial is demonstrating is that bilateral treatment looks like it's more beneficial than unilateral treatment and bilaterally treated patients have a better improvement of visual acuity, not statistically significant, but visual line is important, and we know that the trend is going in the direction of continuous improvement. The safety profile is really good. We have 0 strong side effects in this patient, all have been manageable, easily. And the safety profile is comparable between unilateral and bilateral, which is not surprising given the diffusion of unilateral to bilateral second eye. So altogether, this study is not only a confirmation, it's strengthening the value of the treatment for patients. We are talking about the disease that had no treatment where patients couldn't have any hope for improvement, except for a small percentage with mutation that can improve spontaneously, very small number as we showed from comparison to natural history studies. So this is -- so maybe there is hope for this disease and quite strong hope because as Bob was mentioning, it's a doubling of visual line goal. So it's a very significant improvement in vision. Thank you very much. Bernard, yes?

Yes. So we'll take questions. Thank you very much again, Bob and José, this is really detailed information. That's, of course, I realize it's a lot of information. So let me look at whether we -- so we first take our questions, I believe. Is there any question at all?

[Operator Instructions] The first question comes from the line of Dylan van Haaften calling from Bryan Garnier.

So just 2 -- just 3 actually from my side. So the first one would be on the quality of life assessment. I understand there was also a secondary endpoint in this study. I'd like to understand if there is any color you can give there? And also maybe on some of the anatomical markers. I understand that was something that was looked at during REVERSE and RESCUE and maybe now with a bigger pool of patients, maybe there's more inference to be had there? And my third question would be on contrasting the safety against approved LUXTURNA. And if the regulator would have a perception towards the safety there?

Do you want me to start on?

Yes, José, I think you can answer.

Yes. So regarding the safety, well, we are talking about an intravitreal injection here. So the LUXTURNA side effect that were observed or some loss of function over time mostly was relating to the subretinal injection. The safety profile of LUXTURNA was very good. Actually the vector is the same. It's AAV2 also. So there is no -- vector is exactly the same as in LUMEVOQ. The construct is obviously very different. But here we are talking about an intravitreal injection that is a much easier and safer procedure as compared. Then you may wonder about the safety in terms of inflammation. And as we just showed and Bob beautifully commented on that, the safety profile with respect to inflammation is very good also. So it's safe. I don't want to compare 2 diseases that are very different and 2 approaches that are very different, but it's certainly something that is very good. In terms of quality of life, these analyses are ongoing. As you know, we published already the data from the RESCUE and REVERSE and the improvement in quality of life on different metrics, not just one, but around 7 different metrics. So it's showing an extremely good trend in terms of quality of life. Moving from off chart to on chart is a huge experience for patients not being able to read letters, being able to recognize details, being able to move back to seeing that, not perfect vision, but really much better vision. And we have a lot of anecdotes. Everyone seeing this patient can tell about amazing anecdotes about being able to do things they couldn't do anymore. That is showing that the quality of life has improved. On REFLECT, this is ongoing analysis. So this will be done in the next several weeks, and this will be reported and published already everything is going to be published in the peer review journals. And the second question, I guess, relating to Bob expertise, which is looking at the other parameters.

Yes. So thank you for the question about the anatomic measurements, which is optical coherence tomography. And again, we have something that only was really introduced in the clinical practice within the past 10 years. It's now grown into a much better technology than we first had. And for background, with optical coherence tomography, we can measure the number of fibers around the optic nerve, the retinal nerve fiber layer. We can also measure the thickness of the different retinal layers, the retina is like a layer cake, and where the mutation hit is in the retinal ganglion cell layer. And again, this has been a learning experience for us. The OCT data is still being analyzed for REFLECT. But from RESCUE and REVERSE, I can tell you from my impressions of this, and I have interpreted almost all the OCTs in these studies, but without any knowledge of whether the patients had placebo injections or had bilateral injections, the trend we are seeing is that in the placebo eyes, there is continued deterioration that is thinning the retinal ganglion cell layer. In some of the improving eyes, we see increased thickness in the ganglion cell layer. But more importantly and more commonly, at least it's my impression thus far, and we have to wait for the final analysis. We see stabilization that is no further loss of thickness. Remember, this is a structural marker. This is anatomy, as you said. We are -- now have ways to look at the metabolism of the mitochondria, and there will be some further analysis of these metabolic markers probably available for our review within the next 6 months or so. And remember, the mitochondria too, we're learning more about that. When I was in medical school, they were the energy producers only. Now we know mitochondria interact with the smooth endoplasmic reticulum cells. They're very important in calcium metabolism. The mitochondrial DNA can be an immunological activator. So this is just a completely fascinating work. And as we get more technology and more -- and collaborate with more innovative companies like GenSight, we'll get more answers to help these patients. Thank you.

Do we have any further questions?

The next question comes from the line of René Wouters calling from Kempen.

Also a few questions from my end. I would like to start with a question for the KOLs. So can you further comment on how these results compare to what you expect for the natural history of the disease? And also, do you expect that BCVA will continue to improve with longer follow-up?

So maybe we, obviously, analyses are ongoing. Not all the patients were treated exactly the same day. So some patients have a longer follow-up already. So we already have a trend that is going in line with what we have seen with REVERSE and RESCUE, but the improvement is continuing over time. But these are data that are preliminary. So maybe Magali or Bob want to comment on that. But these are data that are continuing to be analyzed. It's going in the same direction that was well observed in REVERSE and RESCUE, which is continuous improvement over time. And this is from the span of patients that were treated. The average is 18 months, but there is a span. Some got longer follow-up already. Magali, maybe?

Yes, absolutely. We get the result at 2 years very soon before the end of this year. And the trends are really positive right now because you know that we have some patients who had visits after 1.5 years. And while these analysis are ongoing and are really supportive about a pronounced effect over time as we have seen with REVERSE and RESCUE.

Yes. And the natural history, again, is when we go back to how much we're learning about this disease, there is some improvement that occur spontaneously much more in the mutation 14484. The 11778 mutations, yes, patients can improve somewhat, but it is rare. I've been doing neuro-ophthalmology now for close to 40 years, and I can count on one hand and have some fingers left over, the number of patients who've returned and said, "I'm better." So there is some spontaneous improvement. But I think what our data shows and everyone's clinical experience is that that is rare. And the current results far exceed the spontaneous improvement rate that we've seen in the past.