Lexicon Pharmaceuticals, Inc. (LXRX) Earnings Call Transcript & Summary

Good morning, ladies and gentlemen, and welcome to the Lexicon Pharmaceuticals, Inc. Top Line Results for LX9211 Conference Call. [Operator Instructions] This call is being recorded today, Thursday, June 30, 2022. And I would now like to turn the conference over to Mr. Michael O'Kelly. Please go ahead, sir.

Thank you, Michelle. Good morning, and welcome to Lexicon Pharmaceuticals conference call. Joining me today are Lonnel Coats, Lexicon's Chief Executive Officer; Jeff Wade, Lexicon's President and Chief Financial Officer; and Dr. Craig Granowitz, Lexicon's Senior Vice President and Chief Medical Officer. Yesterday afternoon, Lexicon issued a press release reporting positive top line results from our RELIEF-DPN-1 Phase II proof-of-concept study of LX9211 in painful diabetic neuropathy, which is available on our website at www.lexpharma.com. A webcast of this call, along with the slide presentation is available on our website. During this call, we will discuss the study results reported in the release and then use the remainder of our time to answer your questions. Before we begin, let me remind you that we will be making forward-looking statements, including statements relating to the safety, efficacy and therapeutic and commercial potential of LX9211 and other drug candidates. These statements may include characterizations of the RELIEF-DPN-1 study results and other research and development results relating to LX9211 and other drug candidates. These statements may also include expected timing of clinical trials of our drug candidates and the regulatory status and marketing opportunity for those programs. This call may also contain forward-looking statements relating to our growth and future operating results, covering development of our drug candidates, launch and commercialization plans for any approved products, strategic alliances and intellectual property as well as other matters that are not historical facts or information. These risks may cause our actual results to differ materially from those expressed or implied in such forward-looking statements. These risks include uncertainties related to our NDA for sotagliflozin and heart failure, the timing and results of clinical trials and preclinical studies of our drug candidates, success of our commercialization efforts with respect to any approved products, our dependence upon strategic alliances and other third-party relationships, our ability to obtain patent protection for our discoveries, limitations imposed by patents owned or controlled by third parties, and the requirements of substantial funding to conduct our planned research, development and commercialization activities. For a list and a description of the risks and uncertainties that we face, please see the reports we have filed with the Securities and Exchange Commission. I would now like to turn the call over to Lonnel Coats. Lonnel?

Thanks, Mike. It's a wonderful morning for Lexicon, our scientists, clinicians and the patients and principal investigators who participated in the LX9211 painful diabetic neuropathy study. We're so pleased to have announced a successful Phase II clinical trial. The LX9211 low dose was statistically significant and the high dose trended positive, although it plateaued. We purposely tested 2 doses to gain insights into the balance of efficacy and adverse events. Insights were gained from both doses, which suggest our next step is to work on dose selection in preparation for larger trials. I'm excited to say that Lexicon has, once again, shown positive results from our scientific platform. When I arrived at Lexicon about 8 years ago, I quickly picked up on extremely high levels of enthusiasm and confidence within the company for the AAK1 target and the molecules created within our 10-year alliance with BMS, designed to inhibit AAK1 such as LX9211. It didn't take me long to concur with the sentiment of the company. With the support of our Board of Directors, we negotiated an arrangement with BMS, under which this asset became fully under the control of Lexicon. And taken on this challenge, we understood that the industry failure rate in CNS, especially painful diabetic neuropathic is indeed very hot, often due to the notoriously high placebo effects that typically occur in these studies. However, we also understood that, if successful, we will be introducing a new non-opioid mechanism to treat pain for diabetic neuropathy, potentially aiding millions of patients who are woefully underserved by existing therapeutic options. Therefore, we were extremely careful in our design and deliberately patient in our execution of the study, making a successful proof-of-concept outcome that much sweeter. [indiscernible] Jeff, who has been engaged in the program from the start to give an overview of the discovery work that gave us confidence in developing LX9211. Following Jeff will be Dr. Granowitz, who will speak to our previous Phase I results and provide some insights into the successful Phase II outcome. Jeff, I'll turn it over to you.

Thanks, Lonnel. I'm going to go back to go forward and talk about the discovery of this target through our platform and taking it all the way up to human clinical proof of concept. So back to the next slide here. Our foundation as a company was really based on drug discovery that was built on comprehensive screening at physiology and genetics, and we identified this target AAK1 as part of a very broad program to identify new druggable targets out of the human genome. We knocked out about 5,000 targets. We identified about 100 that we thought were of interest. And notably, AAK1 was the target out of all of these 5,000 that had the most compelling phenotype for neuropathic pain. So this is something that we're really excited to now translate to human clinical use. Next slide. The initial discovery was made in the screening [indiscernible] model, which is a screen to identify responses to pain. And in this study, we basically found that the AAK1 knockouts were resistant to Phase II of this assay, which is really related to chronic pain. For things that were [indiscernible] in that chronic pain area, we did a more intensive study, which was called the spinal nerve ligation model, where we ligated a nerve and that made the animals more sensitive to a stimulus that would have pain behavior. And the use in animals, despite the fact that they had this ligated nerve, were resistant to that stimulus and so basically had less pain behavior, and that was an important finding. From there, we went to discover additional things about the target. And we created AAK1 inhibitors and this was part of a collaboration -- a neuroscience collaboration we had with Bristol Myers Squibb, which we entered when their most important drug was Abilify, but subsequently evolved to where we ended up getting this [indiscernible]. But the work that was done in that collaboration established the mechanism of action here is that AAK1 inhibitors reduced spontaneous firing in the dorsal horn neurons of the spinal cord. So it's a central-acting mechanism. Next slide. We also did some work because we felt like it was important to be sure that this did not affect or involved the opioid pathway. So we did work in that setting and concluded that this mechanism of action of AAK1 inhibitors did not involve the opioid pathway, which could be very important for new medications for neuropathic pain. Next slide. What we have also established is the mechanism of action for AAK1 inhibitors in both the alpha-2 adrenergic receptor pathway, which is known to have connections to pain, and that was an important planning as well. Next slide. So from these molecules, we had multiple molecules that we identified as part of this. These molecules ended up showing us the same thing in animal models as we saw in the knockout, and so resistance to pain behavior in various models. I'm going to share a couple of ones with LX9211 specifically among those molecules. This one is the first one that I'll show, which is relating to the study that we had today. So this is in a model of painful diabetic neuropathy where you're inducing diabetes using streptozotocin in animals, and that makes them more sensitive to a stimulus that causes pain behavior. And what we see with LX9211 is a reduction of that pain behavior in these animal models. Next slide. The second one that I'll show you ties into another study that we're doing right now in post-herpetic neuralgia. And, in this case, animal models, the animals were infected with varicella-zoster virus, which is the virus that causes shingles and ultimately, post-herpetic neuralgia and animals in this study resisted a stimulus that would cause pain as -- almost as if they hadn't had varicella-zoster virus at all. Next slide, and we also did this work in comparison with positive controls across a number of these programs. This is just an example of a comparison relative to gabapentin, in which LX9211 outperformed gabapentin in this model also without the kind of motor effects that you would see with gabapentin. So there's the promise for LX9211 as a potent effect on neuropathic pain without the kinds of side effects that you see with gabapentin on cognition [indiscernible]. Next slide. I'm going to turn this over now to -- as we're starting to talk about clinical studies and turn this over to Craig Granowitz to begin here.

Thank you, Jeff. The first slide is a single-dose, ascending dose proportionality PK study. And similar to what we saw in the animal models, we had a very well-behaved drug from the pharmacokinetic perspective. And if you look both at the Cmax and the AUC in the 2 figures in the chart, you can see that the drug is very well behaved over a very wide range of dosing in single dose in humans. You can see the Tmax was achieved in about 8 hours, and the drug has a long half-life of about 7 days. So we felt confident that we had a very well-behaved drug. Also importantly, in the single dose, we found that the most common AEs were mild headache and dizziness, which are consistent signals that we see throughout the program. And that 200 milligrams was really established as a maximum tolerated dose. And again, I think that is another important concept as we go into the study that we're going to talk about in detail today is that we picked the dose that was really the maximum tolerated dose, which was established early on in a single rising dose PK study. If you move forward to the next slide, Mike, thank you, is -- what you see here is an intensive PK that was done both on day 1 and day 14. That's why you see those 2 spikes. All the other days, the PKs were done at the trough level of PK, so it's the lowest dose during that intensive PK period. The idea behind this study, which was carried forward into the proof-of-concept Phase II efficacy study is to get blood levels that were the target drug levels on day 1. So as you look at the dosing, it was 10x the maintenance dose provided on day 1. And what you can see is, first of all, a very nice, linear proportionality of dosing, both in terms of Cmax and AUC, high dose from 2.5 milligrams of maintenance doses up to 20 milligrams of maintenance dose. And you can also see that those dose levels were achieved very quickly based on that tenfold loading dose on day 1. So what we had anticipated in a lot of the modeling work was actually demonstrated in the rising multiple dose and the single dose study. Well-behaved trial. We could achieve target drug levels on day 1. Those were maintained consistently throughout. There was no additional significant accumulation. And you can see again, after the day 14 that there is a decline that, again, is dose-proportional and consistent with the elimination kinetics that had been established in the animal models and was part of the modeling. If you go forward to the next slide, Slide 15, the summary of all of this data that Jeff summarized really 15 years' worth of preclinical knockout animal model work and the rising single and multiple dose studies are we believe that this was an innovative approach to treat neuropathic pain broadly. LX9211 is potent, highly selective and a small molecule inhibitor that could be delivered once daily of AAK1. The preclinical data did demonstrate efficacy across a range of animal models. We achieved CNS penetration. The mechanism of action is independent of the opioid pathway and seemingly separate from gabapentin and the drug was well tolerated in safety and toxicity studies, both in animal and in limited human exposure. The Phase I clinical program showed a favorable safety profile and convenient once-daily dosing. And as a reminder, this LX9211 has achieved fast track designation from the FDA for painful diabetic neuropathy. If we move on to the actual what is called the RELIEF-DPN Phase II proof-of-concept study, again, the idea behind this overall as we chose the maximum tolerated dose, [indiscernible] program and then picked the dose of 50% of that to demonstrate proof of concept in neuropathic pain. As Lonnel mentioned at the outset, conducting neuropathic pain studies are notoriously difficult in human subjects. There tends to be tremendous subjectivity and variability across these trials. And based on all of the animal data and the Phase I data picking the 20-milligram maintenance dose with the 200-milligram first dose and then a 10-milligram maintenance dose with a 100-milligram first loading dose was what was determined. But again, the idea behind the study was to demonstrate efficacy or effectiveness in a category that is notorious for giving ambiguous signals. The study design was randomized, double-blind, placebo-controlled, and it was a parallel group study that was run in about 40 centers in the United States. There were 3 arms, a placebo arm, and then matched 2 experimental arms. The first was what we'll call the 10-milligram or low dose, with a 100-milligram day 1 only loading dose, and a 20-milligram maintenance dose with a 200-milligram day 1 loading dose of LX9211. The primary endpoint was the change from baseline to week 6 in a standard recognized model of pain, or the average daily pain score ADPS, and that has an 11-point patient identified numeric rating scale. There were approximately or a little over 300 patients enrolled in the trial in 40 -- sorry, 40 U.S.-only clinical sites. And again, the idea behind the trial was to push the edge of tolerability and to determine what the bounds of efficacy or effectiveness are in the study, and those were achieved. If you move on to the inclusion and exclusion criteria, I think these are pretty standard for these kinds of trials in terms of what was included and what was excluded. These patients, as a reminder, all had diabetes of long-standing nature and their A1Cs were what I would call relatively well controlled with an A1C below 11 at screening. As a reminder, they needed to be on a stable regimen of their type 1 or type 2 diabetes medications for greater than 1 month prior to screening. As you look at the exclusion criteria, again, these were relatively standard. We wanted to exclude patients that had major depressive disorder or other significant psychiatric disorders as well as substance abuse disorders, any kind of mechanical or device interventions or painful diabetic neuropathy, no use or ongoing use of opioid medications within 2 months, and no NSAID use within 2 weeks prior to screening. Importantly, the patients could be on gabapentin or those related medications at baseline, and we'll come back to that. If you move forward to the next slide, you can see there were 319 patients enrolled. There are roughly 106 patients per group because the randomization was 1:1:1. The mean age was about 62 years. There was a significant representation of women in the study, with a split of roughly 60-40. The patients, not surprisingly, being majority type 2 diabetes, had a BMI in what would be considered the overweight or mildly obese, with the BMI of 32. The baseline ADPS was 6.55. Again, just to put a little more color on that, 75% of patients were in the moderate pain category, 25% were considered severe based on the ADPS scaling, and that again was similar across groups at baseline. As I mentioned, and very importantly, there was allowable the use of underlying other medications, 1 medication for painful diabetic neuropathy. The vast majority of those were gabapentin, but what were allowable was gabapentin, pregabalin or duloxetine. About half the patients were on that. We did not pre-stratify the patients for that, and there was a slight imbalance where more of those patients in the placebo group happen to be on the underlying medications. From the standpoint of -- from an ethical perspective, there was the allowance of rescue medication that was usable in the study that consisted of acetaminophen. And again, over time, more of the patients in the placebo group were rescued with acetaminophen than the active drug treatment groups. Moving forward to the next slide, I want to go first through the efficacy results, then the safety results. These are largely incorporated in the press release. We achieved the primary endpoint of the study demonstrating a significant -- statistically significant reduction in ADPS at week 6 compared to placebo in the low-dose arm, with the results that plateaued with similar results in the high-dose arm. The reduction of patients from their own baseline was 1.39 in the low dose, 1.27 in the high dose. [indiscernible] reduction of ADPS in the placebo group. So that is certainly a significant response in the placebo group. And regardless, we still demonstrated efficacy in the low dose and the plateauing at the high dose. Importantly, separation from placebo was seen at week 1 in both dose arms and the effect size was consistent across a number of preidentified criteria including age, gender, importantly, concurrent use of DPN medication. So the effect size was similar whether or not the patients were on gabapentin and was consistent across underlying baseline pain score. So again, results were similar benefit, whether they had moderate or severe pain at baseline. Importantly as well is that we did have PROs, or patient reported outcome metrics. The most commonly one is called the Global Impression of Change or PGIC. That measures overall patient well-being. So as opposed to the pain score, which is specifically targeted to measure intensity of pain throughout the day, the PGIC is an overall, what I would call, integrated patient perception of therapy benefit or not. And that, again, was also statistically significant in patients who were on LX9211 compared to placebo. Adverse events in the trial were more frequent in the LX9211 treatment arms and more common and more frequent in the higher dose arm, consistent with our expectation going into the study of the 200-milligram dose being at the very limit of tolerability based on underlying data. There were more dropouts in the higher dose than the lower dose group, which may account for why the ADPS at week 6 was lower in the high dose versus the low-dose group. Nearly all -- essentially all of the adverse events reported in the study were mild to moderate. There were no severe adverse events or serious adverse events that were attributed to the drug. And there were no deaths reported in the study attributed to the drug. There were 2 deaths reported in the study. One was related to complications of diabetes and the other was related to COVID. With that, I'll turn it back over to Jeff.

Thank you, Craig. So just key conclusions here. So the study -- results of this study support, first of all, the AAK1 inhibition is a potential new mechanism of action for neuropathic pain. This is the first principle for this mechanism, which we identified, our scientists identified as a potential mechanism of action for neuropathic pain. That's now been translated to a human outcome, which is a major advance. It also supports the advancement of LX9211 development in painful diabetic neuropathy, and we're excited to move this program forward as well. We are beginning to plan for further development. We've been doing some work in preparation for this, and we're continuing that planning. The next step in diabetic -- painful diabetic neuropathy is going to be designing and picking doses for further clinical development in those studies, and we're going to -- in that regard, we're going to be looking at the -- thoroughly at all of the data from this study and using [indiscernible] those decisions. But also we're going to have discussions with potential partners. We've had a tremendous amount of interest in this program from [indiscernible]. It is very rare that you can identify a new CNS target. And so the level of interest probably shouldn't be surprising. But that is going to be part of our planning process as well as talking to partners about this program. Also, just as a reminder, LX9211 is also under evaluation in RELIEF-PHN1. That's a study in patients with post-herpetic neuralgia. We showed you some of the preclinical data that supports that as well. And we're targeting top line results from that study around the end of the third quarter. We are planning on publishing a full analysis of the results of this study in an upcoming medical conference and in a peer-reviewed journal. So be looking for that hopefully later this year. So this is a big achievement for the company. We're really bringing the benefits of our science in our innovation to patients, starting with the platform that got us started and then going through the process of coming up with a new drug against those targets and then now it's a human clinical proof of concept. So the next stage is going to be later-stage development. We're very excited to be able to move this program forward and excited to have established a new mechanism of action for neuropathic pain treatment from the discoveries that our scientists made. So we'll turn this over to questions and answers, and I look forward to your questions.

[Operator Instructions] Your first question comes from Yigal Nochomovitz of Citigroup.

Hi, Lonnel and Jeff and team, thanks very much for going through the data with us. Just to start out a few low-hanging fruit questions. I don't think you actually mentioned what the P values were. If you could let us know what the P values were for the low dose versus placebo and the high dose versus placebo? And then just with respect to the 2/3 improvement on the 11-point ADPS scale, could you just help us put that in context in terms of how clinically meaningful you believe [indiscernible] improvement is? Is placebo corrected? And then how reproducible do you think that is in the next study?

I'll let Craig answer the second question about clinical relevance and the data. We chose not to put the P values in because we are getting ready to submit this to a journal and for publication, hopefully, we'll be able to get something done by the end of the year. So there is a lot of really meaningful data that we're going through the study to harness, and we think the best place to do that is put it into a publication in a medical meeting. So stay tuned for that. Craig, I'll turn it over to you for the second part of his question in terms of clinical meaningfulness.

Thank you, Lonnel, and thank you, Yigal, for the question. Yigal, it's a very important point that you're asking, and I'm going to answer it a couple of different ways. The first is, what patients feel in terms of their well-being is the change from their own baseline. And the fact that there was a change in baseline of 1.37 in the low-dose group is very meaningful. What's generally considered clinically meaningful in this category is a 1-point reduction in score. I think that data is backed up by the fact of seeing the clinical benefit that quickly within 1 week. As you know, often with CNS drugs, it takes quite some time to see that benefit -- now one of the things that we'll explore further in our further clinical development is, will that effect be seen with or without that 10x day 1 dose. But certainly, we are achieving drug levels that are meaningful either way within 6 days. So we have the rapid onset. We have a significant reduction in the patient's own pain score, and the fact that the PGIC is not only effective compared to the patient's baseline, but was significantly different from placebo is another factor in our view that shows that this drug is one that is seen as beneficial to the patients. So I think there's really 3 lines of evidence that make us confident that this drug is clinically meaningfully important. The first is, the change in the average daily pain score; the second is the rapidity of onset of the drug benefit; and the third is that more qualitative patient assessment of their own well-being was also significantly different. So I think those are really 3 important attributes of what I think most patients would be looking for. And just as a reminder, there were more patients in the placebo group that were on the gabapentin at baseline and more of the patients required the use of that rescue medication or the acetaminophen during treatment. And I think those are also important considerations to think about in terms of the robustness of the clinical benefit of this treatment to patients.

Okay. Got it. Just -- so just a couple of quick follow-ups on that. Just to clarify, I know you don't want to disclose the P values, but can you say what the threshold was? I assume it was a 0.05 threshold, not a 0.01. And then also, how many weeks did it take for the -- to start to see [indiscernible] separation from placebo in the low dose? Did you do that analysis?

We saw the separation in week 1. So it was like right off the bat. We saw separation in both doses actually at week 1. And we haven't really talked about the standard for the P value, but it's pretty -- I'd say it was astringent. We definitely have evidence of proof of concept in both of these doses. But it does appear that the lower dose, the combination of the effect being basically as high or higher with a better tolerability profile that one was the one that clearly succeeded the best.

Your next question comes from Joseph Stringer of Needham & Company.

A couple from us. The first one is on the comments around effect size in patients who were either on concomitant meds, such as -- which I think you mentioned was largely gabapentin. I think you mentioned that was in the 40 percentile range. Just curious, you said you saw a similar treatment effect between patients who were -- I guess, correct me if I'm wrong, between patients who are on background med and those who are not. Just curious to get your thoughts on why -- would you have expected sort of a an additive effect with -- on the -- with the patients on background [indiscernible] with LX9211? First question. And then second 1 is around the apparent similar treatment effect with a low dose and high dose, they were similar, numerically higher in the low dose. Just curious if you think you're reaching sort of target saturation there? Or how you think about the similar effect size there? And then I guess, last question, and squeeze one in, it's around just general strategy around the pain program is advancing this program sort of built into your current cash guidance, cash runway guidance?

Joey, thank you. I'll come back to question 3 about the overall strategy. Questions 1 and 2. Craig, I'll turn it over to you.

Thank you, Lonnel. Thank you, Joey, for your questions. Yes, it's a great observation around the similar clinical benefit, whether or not the patients were on concomitant DPN meds. As you rightly identified, about 90% of the patients that were on a concomitant med were on gabapentin. And as you identified, about 50% of patients overall in the study were on a concomitant med. So I hope that's clear enough. About half the patients overall with an imbalance with significantly more on the placebo just by randomized chance were on a concomitant med and nearly all the patients that were on concomitant med were on gabapentin. So I hope that's clear. If you look at the mechanism of action, and Jeff didn't have the chance to go through all of that, while we are acting on the same pathway, at least within the central nervous system of the spinal cord, it's not that surprising to us that the benefit is seen above and beyond what you would see with gabapentin, which is acting both what I would call subtentorial and supratentorial if you think about the tentorium as the separation between the brain and the spinal cord. So our drug, while getting into the CNS, seems to be acting primarily in the spinal cord at that dorsal root ganglion within the spinal cord, which is part of the CNS, but not in the brain itself. And as you know, gabapentin and opioids act primarily above the tentorium or in the brain, even though gabapentin also seems to be acting at the dorsal root. And while we are acting on that same adrenergic pathway, we do know that the mechanism, whereby AAK1 is involved in reducing neuronal firing is different from gabapentin. AAK1 seems to be involved somehow in inhibiting what's called clathrin-coated endocytosis on the cell surface. How that is specifically related to the down regulation of pain is something that is actively being studied, but that is a distinct and separate mechanism of action of gabapentin or from opiates. The issue -- I hope -- so I hope, Joey, that answers question 1.

Yes. Thank you.

The next question, Lonnel, if I should go ahead and answer that one as well as the issue of similar effect on low dose and high dose. And, Joey, I think that really is a reflection probably of overall drug exposure because there were more dropouts in the high dose. And while we still are doing a lot of statistical work, remember, we just got the study unblinded a few days ago, and we continue to do work on that. It is probably a reflection of the higher dropout rate and lower drug exposure across the entire dosing interval. As you know, we did the study based on intent to treat. So whether or not -- this was not a completers analysis. So whether or not the patients were on treatment, whether you -- whatever you were randomized and got the first dose, that was the group that you were in throughout the entire analysis period. We believe that based on the PK that we showed today and our understanding of this drug that we have a lot of opportunity to maximize the benefit-risk ratio of this drug. As you know, based on the other classes that most commonly encountered side effects are a consequence of blocking those same neurons. So dizziness, headache, nausea, are related to blocking the same neuronal pathway that is involved in mitigating the sensation of pain within the central nervous system. So it wasn't a surprise for us. The good news is that we have a lot of [indiscernible] that we can modulate to further optimize the dosing and hopefully magnify the benefit versus the side effect trade-off over time, and that really is the area that we are most interested and focused on, on some of the upcoming studies.

Craig, I think that's well stated on both of those. I'll just take question 3 around strategy. The interest in LX9211 has always been high. The inbound interest has always been high on this asset. I think most have been waiting for whether there's going to be a signal in the translation of a signal in humans. The preclinical data has always been very impressive. And so all the historical work has been impressive. But I think [indiscernible] have been waiting for a signal. That signal has now been achieved. So when we look down the road, it's our objective certainly to partner this compound because it is not our intent to build a commercial organization that goes into primary care, where you have broad indication, broad opportunities. That will be better suited in the hands of a partner. It is also not our ambition to go outside the United States. The implications for a compound like this to meet this maximum potential requires the ability to push the compound outside the United States. So our strategy is very simple. We've now shown proof of concept. I think the evidence here is clear. It's not ambiguous. It is very clear. I fully expect we'll be in dialogue about advancing this compound and some kind of partnership in the very near future. That's always been our strategy, and we're now at a point where we can execute it with a greater degree of certainty.

Your next question comes from Yasmeen Rahimi of Piper Sandler.

Congratulations to the really exciting data, especially in a field where we have not had any innovation. Two questions for you. Maybe the first one is, Craig, you had mentioned the dropout rate, but maybe didn't quantify them. Could you maybe give us an idea of whether there was an imbalance between dropout rates between the 2 dose group versus placebo. Specifically, I assume given that patients saw an improvement at week 1 that maybe that was less dropouts on the treatment arm? That's question 1. And then question number 2 for you is, could you maybe give us [indiscernible] for what are other key efficacy secondary endpoints that were assessed in the study? I respect your wishes that the data will be saved for a scientific conference. But if you could just kind of give us ballpark, some ideas of what were some key secondary endpoints, and whether you were able to pick up stat sig in those key secondaries as well, that could be really helpful for us?

Yasmeen, a great couple of questions. I will first say that there are secondary endpoints, and we're still doing work around that. But the study was powered for the primary endpoint. It wasn't powered so much for those secondaries. But when we do publication, then you'll see much more of that going forward. I will turn the first part of your question over to Craig.

Yes. Thank you, Lonnel. And, Yasmeen, thank you for your questions. On the dropout rates, Yasmeen, the dropout rates were significantly higher in both the dosing arms and particularly the higher dose arm. And again, that was not unexpected based on how we designed the trial. So we really designed the trial to be what would be the maximum tolerated dose, and that was at 200/20 based on both the single rising and multiple dose studies. I think, importantly, what we've seen, like in many cases, is humans had never been dosed for more than 2 weeks with this drug prior to the start of the trial. And what we've seen in both the studies, both the and EPN and PHN trials is that the dropout rates are dropping over time as the study runs as I think both the study types and the patients get more comfortable with the study. And I think that's an important component to keep in mind because, as I mentioned, nearly all of the adverse events were mild to moderate. A lot of it is around expectation setting also of what to expect and what to do. And the fact that in both studies, over the course of the trial, again, I want to separate the course of the trial versus the individual patient course. But over the course of the trial, the dropout rates overall in the study go down. With respect to the benefit, what you really see is most of the adverse events that are reported seem to be reported early. However, that's obviously compounded because there were opportunities in the trial to do dose interruptions. And with the dropouts, there's obviously going to be a survivor bias in that regard is that there's going to be fewer dropouts later because the patients that were really sensitive and didn't want to tolerate therapy already have exited from the study. Now again, all those patients will be counted in the overall efficacy results, but you'll no longer be collecting adverse events on those patients because they're not experiencing adverse events if they're not on the drug -- on the study drug. But the adverse events do seem to be the majority are reported relatively early in the trial. So I hope that answers those questions.

Your next question comes from Jessica Fye of JPMorgan.

Just a couple of kind of clarifiers for me. I appreciate you providing the baseline average daily pain scores across the overall study population. Can you provide that per each arm?

We can say that they were -- that was well balanced among the arms. I don't have the per arm numbers, but they -- it was well balanced.

And I think you said more patients were on background pain medicine in placebo arms and the active arms in addition to using some more Tylenol as the study progressed. Can you quantify that imbalance? Did you say it was statistically significant?

Craig, do you want to answer that?

Yes, I'm happy to. Thank you, Jessica, for your question. So again, Jessica, that was not something that was we put any statistics on. Some studies, and again, often not in Phase II but certainly in Phase III, you would stratify patients at baseline for underlying medication use, if that was going to be allowed. Being smaller in a Phase II study, we didn't have those as stratification criteria. So the patients randomize as they randomized. A little more than half of the patients around 55% at baseline in the placebo group were on underlying pain meds. It was quite a bit lower numerically in both of the study arms and lowest of around 35% to 40% in the highest dose arm. So again, we can't put statistics on that. Was it statistically significantly more? That wasn't something you could put statistics on. But numerically, there were more patients that were on gabapentin in the placebo than in the low dose and the high dose, and they were the fewest patients, just again numerically on gabapentin in the highest dose group. And again, that's just how the randomization occurred. We're obviously blinded. We didn't know what the patients were on and the patients didn't know what drug arm they were going to be on. That was just sort of how it unfolded. And to your point about the rescue medication, again, it was all blinded, but the fact that more patients in the placebo group used the pain medication, the rescue, again, I think, speaks to the impact of the clinical benefit of the drug. Again, it's suggestive, not definitive. We didn't put statistics on it. There were no formal tests on it. But again, it gives us more confidence in the activity and the beneficial activity of the LX9211 compared to placebo.

Okay. Got you. I also have a couple of questions on tolerability in light of what sounds like higher dropouts on the experimental arm despite what looks like some signal of improved pain scores. So I'm curious, what were the rates of dizziness and nausea by arm?

We'll get into that when we do the publication, but it was definitely -- it's meaningfully higher in the high-dose arm than in the drug arm and it was higher than the low-dose arm than in placebo.

Okay. And I'm curious if you guys have taken a look at how many side effects were early and maybe associated with the Cmax at that loading dose relative to the lower ongoing dose levels?

Jessica, great question. There's a lot of the work that we're going to have to do to determine how we're going to dose going forward into Phase III. So one, to your point, is what impact the loading dose may have had on the tolerability, but also what the overall effect was. And then how do we mimic what we believe will be the -- we get the efficacy quickly in how we dose without getting the AEs that we may have seen early. So that's a lot of work that we're going to have to do in our next step in preparation for a larger trial in Phase III potentially.

Got it. And just to clarify on what you guys see as the next step [indiscernible]? I understand that partner would likely be involved in kind of making some decisions here as well. Do you guys think more of Phase IIb work is in order to confirm the best dose? Or do you think this product could be poised to go right into some pivotal trials?

No, I do think Phase IIb, that's my opinion. But certainly, we'll be talking to others about that. I do think -- what Craig and Jeff showed earlier, we have a range of doses that we can deploy. And I think we are learning quite a bit from these data that will give us some clues of how we maximize the benefit and try to avoid the AEs that we saw, particularly in a high dose. And even if we did some work around the high dose, we may be able to neutralize some of the AEs. So there's some work that we can do in IIb to optimize dosing to increase effect and reduce the AEs. Because from a commercial point of view, if we're able to achieve that, we'll have an even bigger commercial runway.

[Operator Instructions] Your next question comes from Julian Harrison of BTIG.

Congrats on the data. First, I'm curious how these results inform intended positioning versus current standard of care? And then second, wondering how we should be thinking about [indiscernible] considering it's only about high loading, high maintenance regimen?

Not sure if I got that.

I'm not sure I got the second part of the question. Okay. So just a couple of things. So I think that mean an interesting part about this is this was on top of standard of care. So a lot of the patients were on standard of care therapies already, and we saw a benefit over and above that. And the ones that weren't, they may not have been on the standard of care because the standard of care didn't work for them. So it's -- I think this is a pretty high bar when we're talking about just taking patients as they are and showing an incremental benefit. So this is a pretty compelling result. The other thing that I think is important to remember is that the current standard of care therapies have not always succeeded in this particular indication. In fact, they've relatively frequently failed to show an effect in this particular indication because it's variable. And that's true even when you get the Phase III type studies. So I think this is a proof-of-concept study. It is a first in human patient study, and we demonstrated convincing evidence we think of the benefit of inhibiting this new mechanism. But it's in a tough area to succeed. I think that PHN is a different indication. I would say that this has to improve our use in terms of what we are likely to see in PHN. The PHN study is a smaller sample size. So I would say that. But we think that this is evidence that lines up with the preclinical evidence that we have with this mechanism should have benefit across many different types of neuropathic pain. And we're just looking at the first one of those. And if it succeeded in that first study, we think it's pretty compelling.

Yes. I think what I would add is that from a PHN point of view, we will learn, and we are learning from DPN that will inform how we complete our PHN work. So I would say stay tuned to that relative to what we learn from the DPN study. To Jeff's point, it's always important to remember that you're trying to beat -- to your point, you're trying to beat standard of care, but in the cases we've already outlined, this was put on top of standard of care, you had benefit and to what Craig has already talked about. If you want to own standard of care, it has benefit. The hurdle rate, we all know, is very, very high in this space. And this new target with this new compound has now overcome that hurdle rate where many others have failed. We have a lot more work to do, I would say, to understand how best we dose it. As I said before, I think we still have an opportunity to maximize to get greater benefit in the absence of some of the AEs we saw. So we need to do that work, be patient in doing that work so we can maximize the opportunity on the other side. So stay tuned, I would say.

There are no other questions. I will now turn the call back over to Lonnel Coats for closing remarks. Please go ahead, sir.

Well, thank you. I just want to take this moment and just thank all of my colleagues here at Lexicon for having confidence and faith in the target that they found early on in the process, convinced me, convinced the board, we went forward. We -- it's been 7 years to take it from the early stage into the clinic. Lots of risk here. I've had a robust experience in CNS in my previous life. And I've always warned that there's a high risk here of failure when we started this program. But we also recognize that if we were to show a signal, there is a high reward on the other side because this field is littered with failures and the placebo effect is the 1 thing that undo compounds in this area. And it's so refreshing to see that both of these doses that we chose showed remarkable benefit and it did it early. There's a lot more we need to learn about what we are now seeing in our data, and we're going to go through that data is going to inform us how we go into later-stage development. And last, but not least, it's always been a strategy given the size of the opportunity with a compound like this, the markets in which it has an opportunity to enter into, the global nature of the indications that we will be seeking, it does require for us to engage in the conversations for those who've been waiting to have the conversation with us to partner it. And then the last piece I'll make is that we've talked about 2 indications. One is DPN that we've now shown -- we've shown a strong signal in. PHN, we'll know the results by -- hopefully, by the end of the third quarter. And we're still doing work, I would say, that we have the potential to enter this drug in a clinic with an additional indication. So this is a remarkably important compound for the company and all the people who worked on it and all the stakeholders who believed in it. This is the first of what I think will be the start of a significant event for the company, and I look forward to having more conversations with you about it as we communicate more as we know more. Thank you for joining us, and have a great weekend.

Ladies and gentlemen, this does conclude your conference call for today. We would like to thank you for participating and ask that you please disconnect your lines.