Akebia Therapeutics, Inc. (AKBA) Earnings Call Transcript & Summary

Good day, ladies and gentlemen, and welcome to Akebia Therapeutics Investor Briefing Webcast. As a reminder, this webcast is being recorded. I would now like to introduce your host for today's event, Kristen Sheppard.

Thank you, and welcome to Akebia's Investor Briefing Webcast to discuss the global Phase III data for vadadustat for the treatment of anemia due to chronic kidney disease, presented at the American Society of Nephrology Clinical Meeting yesterday and also earlier today. Please note that the slides for today's event are available on our Investor Relations website. For your convenience, a replay of today's webcast will also be available on our website shortly after we conclude today's event. Joining me for today's event is John Butler, our Chief Executive Officer; Dr. Steven Burke, our Chief Medical Officer; and Dr. Glenn Chertow, Professor of Medicine, Chief of the Division of Nephrology at Stanford University and the Co-Chair of the Independent Executive Steering Committee for both INNO2VATE and PRO2TECT. Following our discussion, we'll have a question-and-answer session, where the team will respond to what we believe are some of the most frequently asked questions about the data. So thank you to all of you who shared your questions with me in advance. Before we begin, I'd like to remind everyone that this webcast includes forward-looking statements. Each forward-looking statement contained in this webcast is subject to risks and uncertainties that could cause actual results to differ materially from those described in these statements. Additional information regarding these factors is described in the forward-looking statements section of the slides for today's webcast and in the press releases we issued earlier today and yesterday as well as in the Risk Factors and Management's Discussion and Analysis sections of our most recent quarterly and annual reports filed with the SEC. The forward-looking statements on this webcast speak only as of the original date of this webcast, and we do not undertake any obligation to update or revise any of these statements. With that, I'd like to now introduce our CEO, John Butler. John?

Thank you, Kristen, and welcome, everyone. It's great to have you all join us fresh off what continues to be a very exciting and busy virtual ASN Clinical Meeting for Akebia. Our presence at ASN this year is very strong. We had 13 abstracts accepted with a tremendous set of data presentations from our global Phase III program of vadadustat. The opportunity to present these data at this clinical meeting is a testament to the amazing work of our research and development team and their principal investigators and their staff, and speaks volumes to the robust data emerging from our global Phase III program. It also underscores the renal community's increasing interest in HIF-PHIs as a potential treatment for anemia due to CKD. For our event today, Dr. Glenn Chertow will review the Phase III data that were presented at ASN. Following his presentation, Dr. Steven Burke will discuss the newly presented data beyond what we shared in our top line discussions. And of course, as Kristen mentioned, we'll spend time answering questions. But first, I'd like to frame these important data. INNO2VATE, 1 of our 2 global Phase III programs, evaluated vadadustat in nearly 4,000 incident and prevalent patients on dialysis with anemia due to CKD. As we reported in May, this program has been successful. Vadadustat met the primary and key secondary endpoints for efficacy and also met the primary safety endpoint, marking a significant milestone for Akebia, and we believe an even more significant milestone for people on dialysis and their care teams. The INNO2VATE data were presented during an oral session at ASN yesterday by Dr. Kai-Uwe Eckardt, Dr. Chertow's Co-Chair on the INNO2VATE Executive Steering Committee. These positive data further demonstrate the efficacy and cardiovascular safety of vadadustat in dialysis. As you'll hear from Dr. Chertow today, the greatest strength of these data is the consistency across efficacy and across the MACE endpoints. With respect to the primary and key secondary efficacy endpoints, the INNO2VATE data show that vadadustat achieved noninferiority to darbepoetin. The data also show vadadustat's ability to achieve and maintain patients in the target hemoglobin range, while minimizing excursions above that target range. We're very pleased with these data. Upon approval, we expect these data will be highly informative to physicians and dialysis providers as they make important decisions about patient care. In terms of the primary and key secondary safety endpoints, the newly presented positive data from INNO2VATE show vadadustat achieved noninferiority compared to darbepoetin in an analysis of MACE as well as expanded MACE, cardiovascular MACE, cardiovascular mortality and all-cause mortality. Again, we're very pleased with these results. Simply put, these data build on the positive top line data we reported for INNO2VATE in May, filling out a picture of vadadustat for patients on dialysis that we believe is clear, consistent and compelling. We remain confident that vadadustat has a straightforward path to approval in dialysis globally. As you'll hear today, with data in hand, we're continuing our work on the NDA and expect to file as early as possible next year. Upon approval, we believe we have a significant opportunity to treat the nearly 500,000-plus adult patients on dialysis in the U.S., including both incident and prevalent dialysis patients. We believe this could translate into a potential multibillion-dollar market opportunity in the U.S. alone. Turning now to our PRO2TECT program in nondialysis. We're obviously pleased that PRO2TECT was accepted for a late-breaker presentation at ASN, which Dr. Chertow gave earlier today. This was a critical step in allowing us to put forth additional analyses of PRO2TECT's cardiovascular safety results beyond the top line data we reported last month. As we previously announced, vadadustat met the PRO2TECT program's primary and key secondary efficacy endpoints, but missed the primary safety or MACE endpoint. Today's PRO2TECT presentation includes newly presented prespecified analysis from PRO2TECT that show regional differences in cardiovascular outcomes observed in the program, consistent with well-known differences in the standard of care, and more specifically, the target hemoglobin model. As you'll hear today, we believe these analyses demonstrate that vadadustat has no clinically meaningful increase in cardiovascular risk in U.S. nondialysis patients treated to a target hemoglobin level of 10 to 11 grams per deciliter, which is consistent with U.S. treatment guidelines. And as you'll see in Dr. Burke's presentation, these subgroup analyses show similar cardiovascular safety results across both PRO2TECT and INNO2VATE in U.S. patients treated to a target hemoglobin level of 10 to 11. We expect the treatment of nondialysis patients will be a review issue for the FDA. But we continue to believe that these are important analyses, and when viewed together with the totality of data from our Phase III program, we believe they will inform that review. So with that, I'm pleased to introduce the Co-Chair of the Independent Executive Steering Committee for both INNO2VATE and PRO2TECT, Dr. Glenn Chertow, who will now review the Phase III vadadustat data presented at ASN. Welcome, Dr. Chertow, and thank you for joining us today.

Thank you so much. Thank you, John and Steve. And I'd also like to thank my Co-Chair, Dr. Eckardt and the members of our Executive Steering Committee as well as the investigators and, of course, the participants in the trials, without whom we would have no results at all. So I will have the privilege of presenting some data from the 2 programs, INNO2VATE and PRO2TECT. And I'll be starting with INNO2VATE, the 2 trials that were conducted in dialysis-dependent chronic kidney disease. As you know, vadadustat is an investigational agent, an oral HIF-PHI that's under development for the treatment of anemia of CKD. In earlier phase studies, vadadustat raised and maintained hemoglobin concentrations within a target range in both patient populations. The long-term safety and efficacy of vadadustat are unknown. And so our objectives of both programs and, first, again, the INNO2VATE program, is to evaluate the long-term safety and efficacy of vadadustat compared to the active controlled darbepoetin alfa, a commonly prescribed erythropoiesis-stimulating agent, or ESA, in dialysis-dependent chronic kidney disease. And we have 2 trials: one that was a much larger trial of prevalent dialysis-dependent chronic kidney disease, patients who were already on ESAs; and what we term as incident dialysis-dependent chronic kidney disease, that is patients who were just starting dialysis who had limited or no ESA exposure. This slide shows a schematic of the study design. One can see that the patients were randomized 1:1 to either oral vadadustat or subcutaneous or IV darbepoetin. There were 4 study phases after screening, either conversion phase in the patients who are already treated with ESAs or a correction phase for the patients who had limited or no ESA exposure. Then the maintenance phase, which was the bulk of the trial, at least the bulk of the evaluation, weeks 24 to 52, and we had 2 primary evaluation -- or 2 evaluation periods of primary evaluation period and a secondary evaluation period during weeks 24 through 36 and weeks 40 through 52. And these were the periods of time during which efficacy was assessed. And then we had long-term treatment. As you know, this was an event-driven trial focused on cardiovascular safety. So the patients were maintained on their randomized treatment or they were intended to be maintained on their randomized treatment until the accrual of sufficient events to reach our goal target number. And we'll go over some of these details in just a moment. Now the safety analysis was prespecified as a pooled analysis across both trials, the much smaller incident DD-CKD trial and the much larger prevalent DD-CKD trial. And the primary safety endpoint was time to the first adjudicated MACE, or major adverse cardiovascular events. Now in the case of this study, MACE was defined as all-cause mortality or nonfatal MI or stroke. And there was a noninferiority margin prespecified not to exceed 1.25. The efficacy analysis was conducted separately for each trial in the primary efficacy period -- primary evaluation period rather, which was weeks 24 through 36, and the noninferiority margin was 0.75 gram per deciliter in the hemoglobin concentration. As one can see from the next slide, the baseline characteristics were very well balanced across populations in terms of their demographics as well as their underlying kidney disease and cardiovascular disease. And this slide shows the main trial results. When one looks across the vadadustat and darbepoetin group for the MACE safety endpoint, one sees a hazard ratio, 0.96, with a confidence limit that ranges from 0.83 to 1.1, which fell below the -- or within, rather, the noninferiority margin. And all of the other outcomes, expanded MACE, which included MACE plus hospitalizations for heart failure or nonvascular access-related thromboembolic events or cardiovascular MACE or the classic MACE, which is cardiovascular mortality, nonfatal MI or nonfatal stroke. And in each case, we fell within the noninferiority margin with vadadustat relative to the darbepoetin alfa active control. This slide shows the forest plot by prespecified subgroups. The typeface is rather small, but one can see based on region, baseline hemoglobin, the age, sex, race ethnicity, presence or absence of diabetes, et cetera. One can see that the results were consistent across all prespecified subgroups. This slide shows a summary of treatment-emergent adverse events and the treatment-emergent adverse events occurring in 10% or more of patients in each group. So there's a very long list of adverse events that occur in patients on dialysis. They're not exhaustively shown here. But here, we show the much more common adverse events, and one can see that they're actually quite balanced across both groups with the -- anyone can see here the very much larger numbers in the prevalent DD-CKD trial, nearly 1,800 in each treatment arm and just under 200 in each treatment arm in the incident DD-CKD groups. The efficacy endpoints are shown here. Now these are confirmatory, as you know earlier trials had shown efficacy at raising and maintaining hemoglobin concentrations. And in each case, during the primary and the secondary efficacy periods, we saw hemoglobin differences that were vadadustat versus darbepoetin that fell within the noninferiority margin. So to summarize the results from the INNO2VATE program, that is the program in dialysis-dependent CKD. Vadadustat was noninferior to darbepoetin alfa with respect to cardiovascular safety and hematological efficacy in patients on dialysis. Just to summarize a couple of strengths and limitations. The strength is a very large sample size with a relatively long follow-up time. 3,923 patients. All told, the patient population was diverse by multiple factors, and we had very broad inclusion criteria, meaning patients weren't excluded commonly for comorbid conditions or other aspects of their health or dialysis care. And maybe most importantly, all MACE safety endpoints were adjudicated by a committee that was blinded to the treatment assignment. Some of the limitations which can affect, for instance, adverse event reporting was the fact that investigators and patients were not blinded to treatment assignment. So while we would have liked to look at some patient-centered endpoints and patient-reported outcomes, perhaps, they would have been unblinded because, of course, one therapy is intravenous or subcutaneous, the darbepoetin, and another is oral, the vadadustat. And although follow-up time was relatively long, many patients with DD-CKD or what we often refer to as end-stage kidney disease require treatment of anemia for many years and even in some instances, for one or more decades. So we have not yet accrued 10 or 12 or 15 years of safety data with vadadustat, though there is a sizable number of patient years of follow-up. So with that, I'll move forward to the PRO2TECT program, which this was the global Phase III clinical trials for patients with nondialysis-dependent chronic kidney disease, what's sometimes referred to as CKD. And this slide shows also the study schema. Similar in that patients were randomized 1:1 to either vadadustat or darbepoetin alfa. We similarly had a correction or conversion model, for instance, the patients who were ESA untreated, that is they were not treated with epoetin alfa, darbepoetin or other ESAs entered in one trial. And the other trial was for patients who were established on ESA therapy and converted from their dose of ESA to an equivalent or roughly equivalent dose of darbepoetin. Or if they were on darbepoetin, they continued on their stable dose of darbepoetin. We, again, had for efficacy 2 periods at weeks 24 through 36 and weeks 40 through 52 for the evaluation of efficacy. And similar -- similarly to the INNO2VATE programs, we had longer-term follow-up during which we were accruing events. And so the duration of the study depends not on a predetermined study length, but rather, how long does it take to determine a sufficient number of events in order for us to demonstrate or not the noninferiority with respect to safety. And the key eligibility criteria are listed here. The patients were adults. They had to have not received red blood cell transfusions within 8 weeks before randomization. And some clear differences in practice, the baseline hemoglobin was less than 10 for the patients who were ESA-untreated. For the patients who are ESA-treated, we allowed enrollment based on what was generally the practice and the targets in the different regions. Now in the U.S. The -- generally, the practice is to dose ESAs toward the lowest available dose that allows for non-need for transfusion. And in 2011, many of you may recall, the FDA changed the label from a target-based dosing strategy to a dosing strategy based on using the lowest possible dose to avoid transfusion. Now we obviously can't predict when someone might have a drop in hemoglobin. But in general, for patients with nondialysis-dependent CKD patients are not started on ESAs when they have very mild anemia, say, hemoglobin of 10.5 to 12, but rather tend to be started on ESAs at lower levels, aiming again to prevent the need for transfusion and to improve cardiovascular health and well-being. In regions outside the U.S., there tends to be more liberal prescription of ESAs. And the prescriptions are more targeted and the hemoglobin target is higher. And so with respect to ESA use, the 2 studies were quite different. The patients in the ESA untreated had either limited or no ESA exposure and no ESAs within the 8 weeks before randomization. And the patients in the ESA-treated trial were currently on ESAs, and we made certain that patients had adequate iron stores before being treated with either of the 2 agents, vadadustat or darbepoetin. And as for INNO2VATE, the safety analyses were pooled and were cardiovascular focused. The primary safety endpoint was, again, as an INNO2VATE, time to first adjudicated MACE, again, defined all-cause mortality or nonfatal MI or stroke. The same noninferiority margin of 1.25. And for efficacy, similarly, efficacy assessed within each of the 2 trials separately, focusing on the primary evaluation period and also looking at the secondary evaluation period with the same noninferiority margin of 0.75 gram per deciliter in hemoglobin. The baseline characteristics were well matched, as one might expect. These trials were more equivalently sized as opposed to INNO2VATE, where we had the incident and the prevalent trials being very different sizes. Here, we had roughly similar-sized trials of patients who were ESA-untreated or ESA-treated. The -- there was a, as one would expect, altered distribution of region since there were more patients in the U.S. who were untreated with ESAs and more patients outside of the U.S. who were treated with ESAs. And this slide shows the primary safety analysis. And I would just like to highlight the fact that these were the prespecified safety analyses, but not those that provide the most precision, which I will show a bit later. So here, we see that the primary safety endpoint MACE had a hazard ratio of 1.17, and that did not fall -- with the confidence, limits did not fall within the 1.25 limit that was prespecified. So we cannot conclude here based on these results that vadadustat is noninferior to darbepoetin when considering the entire population broadly. And one can see the results for expanded MACE, all-cause mortality, cardiovascular MACE. We're all to the right of 1.0. So in terms of numerical differences, there were more events in the vadadustat relative to the darbepoetin groups and the confidence limits varied based on really the number of events and the precision of the estimate. Now I mentioned the issue of precision. I would point out that initially, we had prespecified adjustment for age as a dichotomous variable greater than or less than 65. And that's a means by which we often present data, for instance, in a forest plot to dichotomize groups for variables that are continuous like baseline dose or dialysis vintage or age. Now because age is so important in terms of risk of death, in particular, and cardiovascular events, including MI and stroke as well, we lose precision when we dichotomize age. And one can gain precision when one considers and adjusts for age as a continuous variable in modeling. And that's because patients who are 65 or 66 have a different risk of death than patients who are 78 or 79. And likewise, patients below the 65-year age range have different risks of death and myocardial infraction and stroke. That is, patients were eligible for inclusion, even as young as 19 or 20 years of age, but those patients' risks are quite low relative to the risks of a patient at age 60 or 64. And when one adjusts for age as a continuous variable, providing more precision, one can see that the confidence limits narrow. But still, the point estimates fall above 1.0. And still, even with more precision here, we cannot conclude that vadadustat is noninferior to darbepoetin. Now here, we see some very provocative and extremely important results. On the left panel, we see as one of the prespecified subgroup analyses, the over 1,700 patients treated in the PRO2TECT trials, both trials. These are safety data, where the trials are pooled. We see in the U.S., where the target hemoglobin concentration was 10 to 11, we see that the curves are virtually overlapping. Whereas in Europe and other regions outside the U.S., we see the cumulative incidents curves begin to separate, and one can see that there is a higher risk of MACE in the patients outside the U.S. who are treated to a hemoglobin target of 10 to 12. And we see this numerically here using age as a continuous variable for precision. We see that within the U.S., we see the point estimates and the confidence limits. And we see the higher point estimates and the higher upper ends of the confidence limits in the regions outside the U.S., where the target hemoglobin concentrations were higher. Now as far as efficacy, we can see really parallel curves in the ESA-untreated patients. Both vadadustat and darbepoetin yielded an increase in the hemoglobin concentration over the first 12 weeks and then maintenance of hemoglobin thereafter, with virtually nil differences between the 2 groups. And likewise, in the ESA-treated population, that is patients who were previously on ESAs who were either maintained on their darbepoetin dose or switched, for instance, from epoetin alfa to darbepoetin alfa. And the vadadustat group, we see the hemoglobins maintained over the course of the trial. These are truncated after 128 weeks. The treatment-emergent adverse events, again, are very similar across the 2 populations with respect to any serious treatment-emergent adverse events or any drug-related serious treatment adverse events. And I would just point out that these are a bit difficult to interpret because we're comparing an existing drug well known to the community, that is darbepoetin, with a new drug that is less well-known to the community. So for instance, when one experiences -- when a participant experiences a common adverse effect like diarrhea, then one might be more apt to report an adverse effect from a drug that is unknown or unfamiliar to the patient and the investigator. But nevertheless, the treatment-emergent adverse events were similar across both groups. So to conclude the data from PRO2TECT. As I pointed out earlier, Vadadustat in the nondialysis-dependent CKD population, vadadustat did not meet the prespecified noninferiority criterion compared to darbepoetin alfa. Adjusted for age as a continuous variable, the hazard ratios and the upper bounds of the confidence limits were attenuated. Though they did not meet the 1.25 prespecified criterion. Cardiovascular risk was similar between the 2 treatment arms in the U.S., where the hemoglobin target was 10 to 11 gram per deciliter but was higher in patients randomized to vadadustat in regions using a hemoglobin target of 10 to 12 gram per deciliter. With respect to efficacy, the vadadustat was clearly noninferior to darbepoetin alfa and maintaining target-range hemoglobin concentrations during both the primary and secondary evaluation periods. And so with that, I will pass the baton as it were to my friend and colleague, Dr. Steven Burke.

Thank you, Glenn, for your review of the clinical data, as well thank you for your continued commitment to the vadadustat clinical development program. As John indicated at the start of the call, I'd like to take some time to put the newly presented data in context with the top line data we reported back in May and in September. First, I'd like to highlight the positive data from INNO2VATE in dialysis. As Glenn discussed, the data for vadadustat in dialysis are clear and consistent across both efficacy and safety endpoints. Vadadustat's effect on hemoglobin was noninferior to darbepoetin alfa and sustained across both incident and prevalent dialysis patients. This result was achieved while minimizing excursions above the target hemoglobin range. With respect to cardiovascular safety, vadadustat achieved noninferiority compared to darbepoetin in an analysis of MACE, expanded MACE, cardiovascular MACE, cardiovascular mortality and all-cause mortality. We're very pleased with these results, which, again, were consistent across regions and in multiple prespecified subpopulations. The actual MACE event numbers for INNO2VATE and PRO2TECT are summarized in Slide 27. As can be seen here, the majority of MACE events were due to all-cause mortality with a smaller number of events due to nonfatal myocardial infarction and nonfatal stroke. It is our belief that our INNO2VATE data support a straightforward path to approval for vadadustat in dialysis. We look forward to working with the FDA and regulators around the world to make this therapy available to patients on dialysis, subject to approval. I will now turn to the late-breaking data presentation of PRO2TECT, our global Phase III program in patients with nondialysis-dependent CKD. The PRO2TECT program was stratified by geographic region, baseline hemoglobin and New York Heart Association Heart Failure Class. The drivers for the stratification by region were the differences in baseline characteristics and well-known differences in standards of care in the U.S., Europe and the rest of the world, most notably hemoglobin targets. As you know, in the U.S., the hemoglobin treatment target is 10 to 11, whereas 10 to 12 in Europe and much of the rest of the world. As with INNO2VATE, the Cox model for MACE included the following covariates: study; region; baseline hemoglobin is a continuous variable; heart failure score; race; sex; diabetes; cardiovascular disease; and age, dichotomized as above or below 65 years. And as previously reported, the PRO2TECT program missed the primary safety endpoint. These results are shown in the left column on Slide 28. We are pleased to show that in the U.S. strata, in the middle column, with the hemoglobin treatment target of 10 to 11 that there was no clinically meaningful increase in cardiovascular risk compared to darbepoetin alfa, with a hazard ratio for MACE of 1.06 and a 95% confidence interval of 0.87 to 1.29. The hazard ratios for expanded MACE and all-cause mortality in the U.S. were 1.02 and 0.92, respectively. All the lower bounds of the confidence intervals were below 1. The increased risk of MACE in the global population was driven by the ex U.S. strata in the right column that had a hemoglobin target of 10 to 12. For MACE, expanded MACE mortality, the lower bounds of the confidence interval were above 1. However, it should be noted that the U.S. and ex U.S. analyses are based on fewer events, and so the confidence intervals are wide. The overall study sample size was selected based on a desire to record 631 MACE events, which would provide 80% power to show noninferiority, assuming a hazard ratio of 1. Thus only MACE, expanded MACE and all-cause mortality had enough events to maintain significant power. Power drops further when examining the subsets of U.S. and ex U.S. patients. To address this, the Cox model was run using age as a continuous variable, which increases power and precision since age is strongly associated with MACE. As you see on Slide 29, when using age as a continuous variable, the hazard ratios were attenuated and the confidence intervals narrowed in the global and U.S. populations in the left and middle columns. In the U.S., the hazard ratios were 1.01, 0.99 and 0.86 for MACE, expanded MACE and all-cause mortality, and the upper bounds of the confidence intervals were all below the noninferiority margin of 1.25. The Kaplan-Meier curves for MACE in the U.S. and ex U.S. strata are displayed in Slide 30. There was a clear difference by region. Side 31 summarizes the U.S. MACE data from INNO2VATE, PRO2TECT in the combined population from both programs with a hemoglobin target of 10 to 11. The totality of the U.S. data in over 4,000 patients was supportive of the cardiovascular safety of vadadustat as compared with darbepoetin alfa. The ex U.S. data is presented in Slide 32. The INNO2VATE data summarized in the left column is supportive of the cardiovascular safety of vadadustat as compared to darbepoetin alfa and ex U.S. dialysis-dependent patients titrated to a hemoglobin target between 10 and 12. To explore the PRO2TECT data ex U.S., we conducted a post hoc analysis that divided patients into 2 groups based on the average hemoglobin achieved in the primary evaluation period, weeks 24 to 36, and examined MACE, expanded MACE and mortality endpoints since there were a reasonable number of events. Since the median of hemoglobins are approximately 11 in each treatment group, this created 2 subgroups of approximately 800 patients, as shown on Slide 33's Kaplan-Meier curves for MACE. The left panel is for those with an achieved hemoglobin less than or equal to 11. The right panel is for those with achieved hemoglobin above 11. As shown on Slide 34, in those ex U.S. patients who achieved a hemoglobin less than or equal to 11, vadadustat was not associated with an increased risk in MACE, expanded MACE or mortality consistent with the U.S. results. The lower bounds of the confidence interval were all below 1. In contrast in the group with achieve hemoglobin above 11, the hazard ratios were increased and the lower bounds of the confidence interval were near or above 1. We were happy to have Dr. Chertow present the PRO2TECT data today, and we look forward to submitting the totality of the Phase III data to support FDA's review of vadadustat for the treatment of nondialysis patients. In summary, the INNO2VATE program in incident and prevalent patients met the primary and key secondary efficacy endpoints and the primary and key secondary safety endpoints of noninferiority for MACE. We believe that our INNO2VATE data support a straightforward path to approval for vadadustat in dialysis patients. We look forward to working with the FDA and regulators around the world to make this therapy available to patients on dialysis, subject to approval. I'll now turn the call over to John. John?

Thank you, Glenn and Steve. Of course, the major milestone of our vadadustat development program will be the submission of a high-quality NDA as early as possible next year, and we're laser-focused on that. We believe we have an extensive data package from over 20 clinical studies in over 8,000 subjects. The next step in this NDA process is completing our pre-NDA meeting with the FDA, which remains on track to occur before the end of this year. Drs. Chertow, Eckardt and the Executive Steering committee have submitted the INNO2VATE manuscript for publication and are already working on the PRO2TECT manuscript to submit for publication as well. As both INNO2VATE and PRO2TECT were very large studies with over 7,300 patients in total, we expect to continue to generate data that we intend to present and publish over time. Kristen, I believe we're now ready to move to questions.

Thank you. We're now going to move to a number of questions that have been asked of us recently regarding this data. The first question is for Dr. Chertow. Dr. Chertow, can you take a moment to explain some of the differences between the dialysis and the nondialysis patient populations?

Yes, I'll be happy to. So the dialysis patient population and the nondialysis requiring chronic kidney disease population are really distinct. Patients requiring dialysis are much more debilitated, much weaker, much more commonly affected by anemia and other comorbid conditions. The patients with chronic kidney disease who are not yet on dialysis tend not to be as either anemic or as challenged with respect to their anemia. For instance, a patient on dialysis -- virtually all patients on dialysis are currently treated with ESA therapy, somewhere in the range of 90%. And if patients are not treated with ESAs, their hemoglobins tend to drop rather rapidly and over the course of several weeks or months will require blood transfusions. Patients with nondialysis requiring chronic kidney disease have a very variable requirement for anemia treatment. Some of them become quite anemic and require treatments. Others have very mild anemia and can have their anemia corrected rather easily with small doses of medications, whether darbepoetin or potentially vadadustat. There are many other differences in the populations, which I can get into in more detail. But with respect to anemia, the patient's dialysis tend to be much more severely affected.

Thank you, Dr. Chertow. We're now going to move to a question for Dr. Burke. Dr. Burke, does the PRO2TECT study and analysis have any implications for the use of vadadustat in Japan, where it is commercially available for both dialysis and nondialysis patients?

No. I don't think there's any implication for Japan. There are multiple regional differences. Japan is a distinct patient population with different genetic makeup, diet, lifestyle and different medical practices that result in lower prevalence of atherosclerotic vascular disease and mortality rates. As an example, the Japanese Phase III study in nondialysis patients presented at ASN last year, there were 300 patients who were randomly assigned equally to vadadustat or darbepoetin alfa and followed for 1 year. There was only 1 death reported, in this case, in a patient randomized to darbepoetin alfa. In contrast, on the PRO2TECT program, approximately 10% of patients experience death in the first year, consistent with expectations. If these populations were the same, we may have expected 30 deaths in the Japanese study, which we clearly did not have. This is probably why PMDA doesn't require or request MACE studies because the underlying CD risk in this patient population is so low compared to that of the rest of the world. So no, I don't believe this has any implications for Japan.

The next question is for John. John, can you give us a sense of how you think the newly presented data will impact the FDA's review of vadadustat for dialysis in nondialysis patient populations?

So the FDA has always viewed these as 2 separate populations. So let me answer the question that way. In the dialysis patient population, as you saw from the INNO2VATE data presented today, we believe that compelling results equal straightforward path. And that's how we feel about the path for vadadustat in dialysis. The U.S. data for PRO2TECT and INNO2VATE MACE was very similar, and we think that should make everyone more confident in the path forward for dialysis. Now on nondialysis. We're pleased to have the opportunity to present the analysis we discussed last month on regional differences in cardiovascular outcomes. We do expect that the treatment of nondialysis patients will be a review issue for the FDA, but we continue to believe that these are important analyses when viewed with the totality of the data from the Phase III program, and that they may help inform that review. We remain on track for a pre-NDA meeting with the FDA before the end of the year, and we look forward to those discussions.

Great. Thank you. John, given the focus on dialysis, can you remind us about vadadustat's potential market opportunity within dialysis?

The U.S. market for dialysis is very significant. There's over 500,000 patients on dialysis, 90-plus percent of them are treated with an ESA today. And when you consider the data that we generated in INNO2VATE, it was extremely clear and we think extremely compelling as well. It's also a very unique market opportunity, given the reimbursement of the bundle payment. And with TDAPA, the transitional drug add-on payment adjuster that's now in place, it really encourages the adoption of innovative products, which clearly vadadustat is going to fall into that category. So we're excited about the opportunity that we have in what today is a $2 billion market.

Great. And Dr. Burke, we're going back to you. Looking outside the United States, what are Akebia's plans for regulatory filings in Europe?

We are working very closely with our colleagues at Otsuka to prepare the NDA. And then once the NDA is filed, we'll file an MAA as soon as possible thereafter.

Great. And lastly, John, what are the upcoming milestones we should be expecting for Akebia?

Sure. Well, it's going to be busy for us moving forward. As I mentioned, we remain on track for a pre-NDA meeting with FDA before the end of this year. We expect to submit our NDA for both dialysis and nondialysis patients as early as possible next year. We've submitted to INNO2VATE for publication in a peer-reviewed journal, and we're working now to submit PRO2TECT for publication as well. And then in addition, we look forward to updating on progress in Japan, where vadadustat is already launched for both dialysis and nondialysis population.

Well, thank you, everyone. We're quickly running out of time here. So I'm going to turn it back over to you, John, to wrap up today's webcast.

Great. Thanks, Kristen. And a special thank you to Dr. Chertow for joining us today. We hope you all have found this discussion informative, and we look forward to speaking with you again very soon on our third quarter financial results call. Thanks very much.