The ASCEND-ND trial: study design and participant characteristics

ABSTRACT

Background

Anaemia is common in chronic kidney disease (CKD) and assessment of the risks and benefits of new therapies is important.

Methods

The Anaemia Study in CKD: Erythropoiesis via a Novel prolyl hydroxylase inhibitor Daprodustat-Non-Dialysis (ASCEND-ND) trial includes adult patients with CKD Stages 3–5, not using erythropoiesis-stimulating agents (ESAs) with screening haemoglobin (Hb) 8–10 g/dL or receiving ESAs with screening Hb of 8–12 g/dL. Participants were randomized to daprodustat or darbepoetin alfa (1:1) in an open-label trial (steering committee- and sponsor-blinded), with blinded endpoint assessment. The co-primary endpoints are mean change in Hb between baseline and evaluation period (average over Weeks 28–52) and time to first adjudicated major adverse cardiovascular (CV) event. Baseline characteristics were compared with those of participants in similar anaemia trials.

Results

Overall, 3872 patients were randomized from 39 countries (median age 67 years, 56% female, 56% White, 27% Asian and 10% Black). The median baseline Hb was 9.9 g/dL, blood pressure was 135/74 mmHg and estimated glomerular filtration rate was 18 mL/min/1.73 m². Among randomized patients, 53% were ESA non-users, 57% had diabetes and 37% had a history of CV disease. At baseline, 61% of participants were using renin–angiotensin system blockers, 55% were taking statins and 49% were taking oral iron. Baseline demographics were similar to those in other large non-dialysis anaemia trials.

Conclusion

ASCEND-ND will define the efficacy and safety of daprodustat compared with darbepoetin alfa in the treatment of patients with anaemia associated with CKD not on dialysis.

Graphical Abstract

Open in new tabDownload slide

INTRODUCTION

Anaemia frequently accompanies advanced chronic kidney disease (CKD), mainly affecting patients with kidney failure requiring dialysis [1]. However, it is however, also common in patients with Stage 4 or 5 CKD [i.e. estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m²] and interventions are often required to increase or maintain haemoglobin (Hb) levels in this population. Furthermore, most people with kidney failure will develop anaemia before they start dialysis and anaemia is associated with poor quality of life and high rates of mortality and morbidity [2, 3]. Data regarding the risks and benefits of treatments for anaemia in people with CKD not requiring dialysis (CKD-ND) are therefore important.

Although treatment with recombinant human erythropoietin (rhEPO) and its analogues was initially focussed on patients with kidney failure requiring dialysis, it was rapidly extended to patients with earlier stages of CKD and has become the standard of care in this population [4–6]. The effect of normalizing Hb on cardiovascular (CV) and other outcomes in patients with CKD has been assessed in large, randomized trials, but these did not demonstrate evidence of benefit. Indeed, some of these trials showed an increase in specific CV events [7–10] with higher Hb targets, possibly related to high doses of exogenous rhEPO and its analogues [11–13].

Hypoxia-inducible factor prolyl hydroxylase inhibitors (HIF-PHIs) have been developed to stimulate erythropoiesis through the inhibition of HIF-prolyl hydroxylase (PHD) enzymes [14]. It has been suggested that these oral agents provide a more physiological approach to treat anaemia. HIF-PHIs may have particular advantages for patients with CKD-ND along with patients receiving peritoneal dialysis or kidney transplant recipients for its ease of use (oral therapy), as well as with the removal of injection burden and cold storage requirements with current therapies.

Daprodustat (previously GSK1278863) is a HIF-PHI that is being developed to treat anaemia in CKD. Initial Phase 3 clinical trials in Japan demonstrated that daprodustat is effective at correcting and maintaining Hb and it appears to be well-tolerated [15].

Here we describe the ASCEND-ND (Anaemia Studies in CKD: Erythropoiesis via a Novel PHI Daprodustat-Non-Dialysis) Phase 3 trial, designed to assess the efficacy and safety of daprodustat compared with darbepoetin on Hb and CV outcomes and examine the baseline characteristics of randomized participants.

MATERIALS AND METHODS

Study design

ASCEND-ND is a global, randomized, open-label (steering committee- and sponsor-blinded), parallel-group, active-controlled, event-driven Phase 3 trial comparing the efficacy and safety of daprodustat with darbepoetin alfa in patients with CKD-ND (ClinicalTrials.gov: NCT02876835; EudraCT: 2016-000542-65). The study was approved by the ethics committees or institutional review boards at each participating institution and was conducted according to the recommendations of Good Clinical Practice and the Declaration of Helsinki.

The ASCEND-ND study timeline was divided into four periods: a screening period, a placebo run-in period, a treatment period and a follow-up period (Figure 1). The 4-week screening period permitted assessment of eligibility based on clinical and laboratory assessments, while the 4-week run-in period was used to establish the tolerance of and adherence to placebo tablets and study procedures. Participants receiving prior erythropoiesis-stimulating agents (ESAs) continued these during the screening and run-in periods.

Participants were randomized to oral daprodustat or subcutaneous darbepoetin alfa (1:1) if they successfully completed the run-in period. Thereafter, the treatment period was divided into a stabilization phase, from Day 1 to Week 28, with dose titration to achieve the prespecified Hb target range (10–11 g/dL) and a maintenance phase to maintain Hb, from Week 28 to the end of the study; randomized treatment was discontinued at the end-of-study visit.

Participants attended routine follow-up visits at least every 4 weeks during Year 1 of the study and at least every 12 weeks thereafter until the end-of-study visit was completed. Participants were asked to complete an off-treatment follow-up visit after discontinuing randomized treatment. Serum, plasma and urine samples were collected at baseline, Week 28 and Week 52 for future analysis of biomarkers.

Eligibility criteria

Eligibility was determined at Week −8, with a subset of entry criteria reconfirmed at Day 1 (randomization). Eligible patients had CKD Stage 3–5, were not currently receiving dialysis or scheduled to start dialysis within 90 days after study start, had either a screening Hb 8–10 g/dL if they were not receiving ESAs or a screening Hb of 8–12 g/dL if they were receiving ESAs, demonstrated adherence to daprodustat placebo tablets during the run-in period, were iron-replete (ferritin >100 ng/mL and transferrin saturation >20%) and were able to provide informed consent. The key inclusion and exclusion criteria are provided in Table 1 and complete entry criteria are outlined in Supplementary data, Table S1.

Study treatments and management strategies

Daprodustat and darbepoetin alfa dosing strategies, along with those for iron therapy, are detailed in Table 2. A protocol-mandated rescue algorithm was in place to minimize the risk of an inadequate Hb response for an extended period and to enable consistency in the application of rescue therapy across the study (Table 3).

Objectives and endpoints

The trial has two primary objectives: to compare the effects of daprodustat and darbepoetin alfa on Hb response (efficacy outcome) and CV events (safety outcome) using non-inferiority margins agreed to in consultation with regulatory agencies (see Statistical analysis section). Hb response will be evaluated as the mean change in Hb from baseline to the evaluation period (EP; Weeks 28–52). The CV safety outcome is the first adjudicated major adverse CV event [MACE; i.e. the composite of all-cause mortality, non-fatal myocardial infarction (MI) or non-fatal stroke]. All deaths and potential non-fatal CV events are adjudicated by an external, independent and blinded endpoints committee led by the Duke Clinical Research Institute (Durham, NC, USA) in collaboration with George Clinical (Sydney, NSW, Australia). The principal secondary endpoints, which include superiority assessment of effects on MACEs and CKD progression, are listed in Table 4 along with other secondary endpoints (also to be tested for superiority).

Randomization and stratification

Participants were stratified by baseline use of an ESA, by region and by participation in the ambulatory blood pressure monitoring substudy. The region groupings used for stratification were developed to enable balance across different parts of the world with similar standards of care (see Supplementary data, Table S2 for details).

Following stratification, participants were randomized 1:1 to receive oral daprodustat or darbepoetin alfa. A central randomization approach was used to help protect against bias due to the open-label design.

Statistical analysis

A sample size of 4500 was originally planned for this event-driven trial, based on a target of 945 adjudicated first MACE events. With the non-inferiority margin defined as a hazard ratio (HR) of 1.2, this provided ∼90% power to establish non-inferiority for time to first adjudicated MACE, assuming a true underlying 3% lower relative risk of a MACE in favour of daprodustat (i.e. a true underlying HR of 0.97). In addition, this also provided 80% power for non-inferiority, assuming the true underlying risk of a MACE is the same in both groups (i.e. a true underlying HR of 1.00). In July 2020, prior to completion of recruitment and study unblinding, the MACE non-inferiority margin was changed to 1.25, resulting in 664 MACEs being required to maintain 90% power. The change was made after discussion with regulatory authorities and with approval from the academic-led steering committee and the Independent Data Monitoring Committee. The rationale for the non-inferiority margin change was to accelerate study closeout in the context of the coronavirus disease 2019 (COVID-19) pandemic and to align with the non-inferiority margin used in other HIF-PHI clinical studies [16]. Additionally, due to the COVID-19 pandemic, a decision was made to end the study screening in September 2020, which resulted in <4500 participants being randomized.

The study size also provided >99% power to perform the intention-to-treat (ITT) non-inferiority test for the treatment difference of mean change in Hb between baseline and EP (daprodustat − darbepoetin alfa) based on a non-inferiority margin of −0.75 g/dL. Multiple imputations using a missing-at-random assumption will be used to impute missing Hb values.

Conditional on the co-primary endpoints achieving non-inferiority at the one-sided 2.5% level, statistical testing will progress to the principal secondary endpoints (Table 4), including superiority for MACEs and CKD progression. CKD progression is defined as a 40% decline in eGFR from baseline (confirmed 4–13 weeks later) or end-stage kidney disease, which is defined as initiating chronic dialysis (≥90 days or adjudicated as intended to be chronic if the duration of dialysis is <90 days); dialysis being indicated but not initiated (as reported by the investigator and confirmed by adjudication); or kidney transplantation. Statistical testing for the principal secondary endpoints will be adjusted for multiplicity using the Holm–Bonferroni procedure [17].

Descriptive statistics in the form of number and percentage of participants or median and 25th (P25) and 75th (P75) percentiles are provided for baseline variables. Baseline values are presented for the ITT population, overall and by baseline ESA use.

Study oversight

ASCEND-ND was developed in collaboration with steering committees, which provided academic and scientific leadership as well as oversight during the study, as previously described for the ASCEND-D trial [18]. An external, independent Clinical Events Classification group blinded to the treatment assignment, led by the Duke Clinical Research Institute in collaboration with George Clinical, is responsible for adjudicating predefined events (all-cause mortality, MI, stroke, hospitalization for heart failure, thromboembolic events and select progression of CKD events). Committee members and their respective affiliations are presented in Supplementary data, Table S3.

Comparison with other large ND anaemia trials

To assess comparability to prior anaemia studies in the ND population, we compared baseline characteristics from ASCEND-ND participants with baseline characteristics from participants enrolled in similar anaemia trials in a CKD-ND population. Included in the comparison were the HIF-PHI trials with roxadustat (OLYMPUS; NCT02174627) [19] and vadadustat (PRO₂TECT; NCT02648347) [20], as well as the placebo-controlled TREAT (NCT00093015) [21] study of darbepoetin alfa.

RESULTS

ASCEND-ND is being conducted at 506 research sites in 39 countries. The study completed recruitment in December 2020; country-level/region participant distribution is listed in Figure 2. In total, 30% of participants originated in Europe, the Middle East and Africa (EMEA); 26% in North America (predominantly the USA); 15% in Latin America and 28% in the Asia Pacific region.

Screening, run-in, and randomization

A total of 10 498 patients were screened, including patients who were rescreened, of whom 6626 (63%) did not meet entry criteria and were not randomized. The reasons for screening failure are listed in Supplementary data, Table S4.

A total of 3872 participants were randomized. One additional participant was randomized but had not provided valid informed consent, so was removed from the total of randomized participants.

Participant characteristics

Baseline characteristics are summarized in Table 5. The randomized participants had a median age of 67 years and 56% were female. The majority of participants were white (56%); 27% self-identified as Asian, 10% as Black, 23% as Hispanic or Latino and 5% as American Indian or Alaskan native background. Of note, within the USA, 33% of participants were black (post hoc). At randomization, 81% of participants had Stage 4 or 5 CKD, 57% had diabetes and 37% had a history of CV disease.

The median Hb at randomization was 9.9 g/dL, eGFR was 18 mL/min/1.73 m² and participants were iron-replete (transferrin saturation 29% and median ferritin 271 µg/L). Fifty-three percent of participants were not receiving ESAs prior to the study. Participants receiving ESAs were predominantly receiving epoetin (58%) or darbepoetin (31%). Oral iron was prescribed for 49% of participants at baseline; 12% received intravenous iron. Sixty-one percent of participants were using renin–angiotensin system blockers.

Although ESA usage at entry was well balanced across most of the baseline characteristics, there was a notable difference of less ESA use in the Black population at entry (5% of ESA users were Black compared with 13% of ESA non users). Participants receiving ESA therapy at entry had higher median Hb levels (10.4 versus 9.5 g/dL); other clinical and laboratory parameters as well as other concomitant medications by ESA status at entry are listed in Table 5.

ASCEND-ND compared with other large ND anaemia trials

Participants enrolled in ASCEND-ND generally had similar demographic characteristics as the participants in other HIF-PHI ND trials [19, 20], as well as TREAT [21]. ASCEND-ND and the PRO²TECT ND [20] trials had an active comparator, whereas the OLYMPUS [19] and TREAT [21] trials compared the respective agents with placebo (Table 6).

Participants in the OLYMPUS trial [19] of roxadustat were somewhat younger than those in the other trials, but the distribution of kidney function appears similar in the trials that have reported these data. TREAT [21] was conducted entirely among people with diabetes, whereas slightly more than half of participants in the HIF-PHI trials had diabetes. Ferritin and transferrin levels were lower in TREAT [21] than in the HIF-PHI trials [19, 20], but eGFR and Hb levels were higher. Other concomitant medications, clinical characteristics and laboratory parameters were similar across the various trials.

DISCUSSION

Anaemia is common in patients with advanced CKD and its treatment remains an area of considerable uncertainty. ESA therapy is widely used to maintain Hb levels in patients with CKD not requiring dialysis; however, parenteral treatment can be inconvenient and painful. HIF-PHIs such as daprodustat potentially offer a more convenient oral alternative and have demonstrated promising safety and efficacy profiles in studies reported to date [15, 22]. The ASCEND-ND trial, as part of the broader group of ASCEND trials including prevalent (ASCEND-D) and incident (ASCEND-ID) dialysis studies, will specifically define the benefit–risk profile of daprodustat when used in patients with anaemia of CKD.

The ASCEND-ND trial compares oral daprodustat with darbepoetin alfa because ESA therapy is currently the standard of care in this population and is recommended in relevant guidelines [4–6]. A difficulty in the design of a trial like ASCEND-ND is blinding participants to the randomized treatment given that the active comparator is administered parenterally. In view of the practical complexities of a double-blind, double-dummy design for a trial of this magnitude, ASCEND-ND has an open-label, sponsor-blinded, design. Throughout the study, the steering and adjudication committees also remained blinded to treatment assignment [23]. Although we acknowledge that the selected ESA comparator (darbepoetin alfa) and the specific dose steps and frequency of administration may differ from local ESA protocols, a protocol-mandated dose adjustment algorithm was used for both treatment groups to prevent differential treatment of each group as had occurred in earlier daprodustat clinical trials [24].

Implementation of a global Hb target range of 10–11 g/dL also enabled uniform treatment of anaemia for participants in all regions. This range accommodates most anaemia guidelines and ESA labelling worldwide, except for the USA, where labelling is more restricted [25]. Other standardizations included a single comparator with defined dose steps and frequency, as well as common iron management protocols, and an anaemia rescue algorithm that was used for both treatment groups. For anaemia rescue, early intervention with intravenous iron and/or transfusions was allowed to help increase Hb before declaring a patient had met the rescue endpoint, which in turn led to permanent discontinuation of randomized treatment. These standardizations will help ensure a more unbiased comparison between the randomized treatment groups.

The trial uses co-primary outcomes that reflect the major questions regarding the potential role of daprodustat in clinical practice. The first is to demonstrate that it is an effective treatment for the anaemia of CKD and is not inferior to ESA therapy as the current standard of care. The second co-primary outcome assesses the CV safety of daprodustat, defined as non-inferiority to ESA for the outcome of a MACE. If non-inferiority is confirmed, the potential for daprodustat to achieve superiority in reducing the risk of MACE and CKD progression will also be assessed. The more physiological circulating levels of erythropoietin likely to be achieved with daprodustat have the potential to lead to better CV outcomes. However, this was not found in pooled ND studies of another HIF-PHI, vadadustat, in which a HR of 1.17 [95% confidence interval (CI) 1.01–1.36] for CV safety did not meet the prespecified non-inferiority criterion of 1.25 for this agent compared with darbepoetin alfa [26]. Another programme assessing the HIF-PHI roxadustat included a pooled CV safety analysis across three placebo-controlled trials with a MACE HR of 1.10 (95% CI 0.96–1.27) being observed [27]; a smaller active-controlled study (Dolomites; NCT02021318) was not powered for CV safety but reported a HR of 0.81 (95% CI 0.52–1.25) [28]. At this time, it is not known what will be seen with other HIF-PHIs.

The optimal Hb target in the CKD population treated with ESAs has been discussed extensively in the scientific kidney community and the lower Hb target—as assessed in both arms of the daprodustat ND trial—should provide a useful comparison of the two therapies. Additionally, the progression of CKD in both treatment arms will be assessed.

ASCEND-ND was designed to allow generalization of the results to the broadest possible CKD population by ensuring that participants are representative of the overall CKD population. Sites were selected from around the world in order to achieve a broad geographical representation in recruitment. A placebo run-in period was established to confirm adherence to an oral medication and to minimize the withdrawal of consent post-randomization that occurred in a prior daprodustat study [29]. While this helped ensure that participants were adherent with randomized treatment, it may limit the general applicability of the results, particularly considering that the oral route of administration is a major advantage of the potential treatment. The baseline characteristics of randomized participants are very similar to those of previously reported trials in participants with anaemia related to CKD, allowing some degree of comparison across trials and comparators (active or placebo).

ASCEND-ND has a number of strengths. It is one of the largest prospective randomized cardiovascular outcomes trials in this field. ASCEND-ND recruited a diverse population with respect to race. Participants include members of minority groups such as those who self-identified as Black, Native American and Latino. Furthermore, one-third of participants within the USA were Black. This diversity of the study population will be important in generalizing the safety and efficacy of daprodustat.

In conclusion, ASCEND-ND has enrolled 3872 participants who are generally representative of patients with anaemia of CKD not on dialysis. The study will test the hypothesis that daprodustat is non-inferior to the comparator darbepoetin alfa for two co-primary endpoints—Hb efficacy and CV safety—as well as a range of secondary efficacy outcomes. The results from this trial, expected in late 2021, will define the benefits and risks of a potential new treatment option for anaemia in patients with CKD.

ACKNOWLEDGEMENTS

Editorial support with writing assistance, assembling figures, grammer editing and referencing was provided by Jonathan Plumb, of Fishawack Indicia, part of Fishawack Health, and was funded by GlaxoSmithKline.

FUNDING

ASCEND-ND is funded by GlaxoSmithKline.

DATA AVAILABILITY STATEMENT

Anonymized individual patient data and study documents can be requested for further research from www.clinicalstudydatarequest.com.

AUTHORS’ CONTRIBUTIONS

V.P. wrote the first draft of the manuscript and contributed to the design, interpretation of data, supervision, management of the research and writing and critical review of the manuscript. All authors contributed to the design, interpretation of data, management of research and writing and critical review of the manuscript. All authors affirm that authorship is merited based on the International Committee of Medical Journal Editors authorship criteria. V.P. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

CONFLICT OF INTEREST STATEMENT

V.P. reports consultancy agreements with AbbVie, Bayer, Boehringer Ingelheim, Chinook, GlaxoSmithKline, Janssen, Pfizer, Astellas, AstraZeneca, Bayer, Baxter, Bristol-Myers Squibb, Durect, Eli Lilly, Gilead, Janssen, Merck, Mitsubishi Tanabe, Mundipharma, Novartis, Novo Nordisk, Pharmalink, Relypsa, Retrophin, Roche, Sanofi, Servier and Vitae; research funding from Pfizer (supplied drug and seed funding for TESTING trial) and GlaxoSmithKline; honoraria from AbbVie, Bayer, Boehringer Ingelheim, GlaxoSmithKline, Janssen, Pfizer, Astellas, AstraZeneca, Baxter, Bristol-Myers Squibb, Chinook, Durect, Eli Lilly, Gilead, Merck, Mitsubishi, Tanabe, Mundipharma, Novartis, Novo Nordisk, Pharmalink, Relypsa, Retrophin, Roche, Sanofi, Servier and Vitae; serves as a scientific advisor on steering committees for trials funded by AbbVie, AstraZeneca, Bayer, Boehringer Ingelheim, Chinook, Eli Lilly, Gilead, GlaxoSmithKline, Janssen, Novartis, Novo Nordisk and Retrophin; and is a board director for George Clinical, George Institute, Garvan Institute, Mindgardens Network, Children's Cancer Institute and Victor Chang Cardiac Research Institute. K.C. reports consultancy fees from GlaxoSmithKline. V.J. reports consultancy fees from GlaxoSmithKline. K.L.J. reports consultancy fees from GlaxoSmithKline and is an advisory board member for Akebia. R.D.L. reports grants and personal fees from Bristol-Myers Squibb and Pfizer; personal fees from Boehringer Ingelheim and Bayer AG and research grants from Amgen, GlaxoSmithKline, Medtronic PLC and Sanofi Aventis. I.C.M. reports research grants, consultancy fees and honoraria from GlaxoSmithKline and Vifor Pharma. J.J.V.M. reports personal fees from Abbott, Hikma, Sun Pharmaceuticals and Servier and his employer has received fees from Alnylam Amgen, AstraZeneca, Bayer, Bristol-Myers Squibb, Cardurion, Cytokinetics, Dal-Cor, GlaxoSmithKline, Ionis, Novartis, Pfizer and Theracos. G.T.O. reports personal fees from Roche Mexico, Johnson & Johnson, Vifor and AbbVie. S.S. reports grants and consultancy fees from Alnylam, AstraZeneca, Bayer, Bristol-Myers Squibb, Cytokinetics, Gilead, GlaxoSmithKline, Lilly, MyoKardia, Novartis, Respicardia, Sanofi Pasteur and Theracos; grants from Bellerophon, Celladon, Eidos, Ionis, Lone Star Heart, Mesoblast, National Institutes of Health/National Heart, Lung, and Blood Institute and Neurotronik and consultancy fees from Akros, Amgen, Arena, Cardior, Cardurion, Corvia, Daiichi-Sankyo, Ironwood, Merck Sharp & Dohme, Roche, Takeda, Quantum Genetics, AoBiome, Janssen, Cardiac Dimensions, Tenaya, Dinaqor, Tremeau, CellProThera and Moderna. C.W. reports consultancy fees from Akebia, Astellas, AstraZeneca, Bayer, Chiesi, FMC Idorsia, Mundipharma, GlaxoSmithKline, Merck Sharp & Dohme, Reata, Gilead, Tricida, Vifor, Lilly and Takeda and grants and consultancy fees from Boehringer Ingelheim, Sanofi Genzyme and Shire. S.S.W. reports personal fees from Public Health Advocacy Institute, CVS, Roth Capital Partners, Kantum Pharma, Mallinckrodt, Wolters Kluewer, GE Healthcare, GlaxoSmithKline, Mass Medical International, Barron and Budd (versus Fresenius), Johnson & Johnson, Venbio, Strataca, Takeda, Cerus, Pfizer, Bunch and James, Harvard Clinical Research Institute, Oxidien, Sironax, Metro Biotechnology, Biomarin and Bain and grants and personal fees from Allena Pharmaceuticals. D.C.W. reports honoraria and/or consultancy fees from AstraZeneca, Amgen, Astellas, Bayer, Boehringer Ingelheim, GlaxoSmithKline, Jansen, Merck Sharp & Dohme, Mundipharma, Napp, Pharmacosmos, Reata, Tricida and Vifor Fresenius. A.W. reports personal fees from Roche, Bayer, Fresenius and Medice. A.K.S. reports consultancy fees from GlaxoSmithKline. A.B., B.C., A.R.C., R.D., T.L.D., L.K., A.M.M. and L.T. are employees of and stockholders in GlaxoSmithKline.

Notes

This trial is registered with ClinicalTrials.gov: NCT02876835; EudraCT: 2016-000542-65; sponsor protocol number: 200808

REFERENCES