Randomized, double-blind, placebo-controlled phase 3 study of bardoxolone methyl in patients with diabetic kidney disease: design and baseline characteristics of the AYAME study

ABSTRACT

Background

Diabetic kidney disease (DKD) is the leading cause of end-stage kidney disease (ESKD), but currently available treatments do not improve kidney function or prevent the initiation of dialysis/kidney replacement therapy. A previous study demonstrated that bardoxolone methyl improves the estimated glomerular filtration rate (eGFR), but the study was prematurely terminated because of an imbalance in heart failure between treatment groups. The subsequent phase 2 TSUBAKI study demonstrated no incidence of heart failure and an improved eGFR and GFR as determined by inulin clearance in DKD patients.

Methods

This randomized, double-blind, placebo-controlled multicentre phase 3 study was designed to assess the efficacy and safety of bardoxolone methyl in DKD patients with an eGFR ≥15.0–<60.0 ml/min/1.73 m² and a urinary albumin:creatinine ratio (UACR) ≤3500 mg/g but without risk factors for heart failure. The primary endpoint is the time to onset of a ≥30% decrease in the eGFR or ESKD. Randomized patients (1:1) have been under treatment with once-daily oral bardoxolone methyl (5, 10 or 15 mg by intrapatient dose adjustment) or placebo for at least 3 years.

Results

The mean age of the 1013 patients is 65.9 years, 21.5% are female, the mean eGFR is 37.84 ml/min/1.73 m² and the median UACR is 351.80 mg/g.

Conclusions

Appropriate patients are enrolled in this study. This study will investigate the long-term efficacy and safety of bardoxolone methyl in DKD patients covering a wider range of eGFR (≥15.0–<60.0 ml/min/1.73 m²) and albuminuria (≤3500 mg/g) compared with previous studies.

Graphical Abstract

Graphical Abstract.

KEY LEARNING POINTS.

What is already known about this subject?

• Bardoxolone methyl improves the estimated glomerular filtration rate (eGFR) in patients with chronic kidney disease (CKD) with type 2 diabetes but causes an imbalance in heart failure between treatment groups.

• A previous phase 2 study, in which CKD patients with type 2 diabetes without risk factors for heart failure were treated with bardoxolone methyl for 16 weeks in Japan, showed that bardoxolone methyl improved the GFR as determined by inulin clearance and the eGFR, and no patient developed heart failure

• The long-term efficacy and safety of bardoxolone have not yet been evaluated in a large-scale study.

What this study adds?

• The AYAME study will verify that bardoxolone methyl (an activator of nuclear factor, erythroid 2-like 2) suppresses the onset of renal endpoints compared with placebo without safety issues in patients with G3/G4 diabetic kidney disease (DKD) who have no previous risk factors for heart failure

What impact this may have on practice or policy?

• The AYAME study will reveal the long-term efficacy and safety of bardoxolone methyl in patients with various stages of DKD because it includes a wider range of eGFR and albuminuria categories than previous studies.

INTRODUCTION

Diabetic kidney disease (DKD) is well known as the leading cause of end-stage kidney disease (ESKD) [1]. Several treatment options for DKD, such as renin–angiotensin system inhibitors, anti-hyperglycaemic medications [including sodium–glucose cotransporter 2 (SGLT2) inhibitors and glucagon-like peptide-1 (GLP-1) receptor agonists] and mineralocorticoid receptor antagonists, have been shown to prevent a decline in renal function and delay the initiation of dialysis in DKD patients [2]. However, these options have not been demonstrated to improve renal function or prevent the initiation of dialysis [3, 4].

Oxidative stress, inflammation and metabolic disorders promote diabetes, cardiovascular disease, chronic kidney disease (CKD) and DKD [5–8]. Bardoxolone methyl is an activator of nuclear factor erythroid 2-related factor 2 (Nrf2) and inhibits kelch-like ECH-associated protein 1 (Keap1), resulting in activation of the Keap1–Nrf2 pathway [9–12]. Thus bardoxolone methyl, which has antioxidative and anti-inflammatory effects, is a potential candidate for treatment of DKD. Indeed, in 2018, bardoxolone methyl was designated as a target drug by the SAKIGAKE Project Team in Japan [13] and is expected to improve the renal function of DKD patients. Previous studies have demonstrated improvements in the estimated glomerular filtration rate (eGFR) in CKD patients treated with bardoxolone methyl [14–16]. However, higher numbers of deaths, heart failure events, serious adverse events and adverse events leading to treatment discontinuation were demonstrated in stage G4 CKD patients with type 2 diabetes mellitus (T2DM) in the bardoxolone methyl group than in the placebo group in the Bardoxolone Methyl Evaluation in Patients With Chronic Kidney Disease and Type 2 Diabetes (BEACON) study, which was prematurely terminated [15]. In a subsequent study, a baseline B-type natriuretic peptide (BNP) level >200 pg/ml and a history of hospitalization for heart failure were identified as risk factors for heart failure [17].

Following the identification of these risk factors, the Phase 2 Study of Bardoxolone Methyl in Patients With Chronic Kidney Disease and Type 2 Diabetes (TSUBAKI) study demonstrated no major safety concerns in patients without risk factors for heart failure [16]. In addition, it demonstrated improvement in the GFR in patients with stage G3 CKD with T2DM as determined by inulin clearance at week 16. However, the 16-weeks treatment with bardoxolone methyl and small sample size in the TSUBAKI study did not allow the investigators to comprehensively profile the safety and efficacy of bardoxolone methyl.

One of the challenges in drug development for CKD/DKD is that costly large-scale studies with long-term follow-up are required to evaluate efficacy endpoints such as doubling of serum creatinine and the onset of ESKD as a hard endpoint. In consideration of this difficulty, the Kidney Disease: Improving Global Outcomes Controversies Conference and the Research Working Group for Establishing Guidelines for Clinical Evaluation of Chronic Kidney Disease of the Japanese Society of Nephrology and the Japan Diabetes Society concluded that surrogate renal endpoints may be adopted in assessing the status of CKD, including a 30% or 40% decrease in the eGFR within 2 or 3 years depending on the stage and progression of CKD [18, 19].

On the basis of these previous studies and recent guidelines, a phase 3 study was designed to evaluate the efficacy and safety of at least 2 years of treatment with once-daily oral bardoxolone methyl in DKD patients. We herein present the study design and baseline characteristics of the enrolled patients.

MATERIALS AND METHODS

The study protocol and all amendments were approved by the independent ethics committees at each study site. The study is being implemented in accordance with the International Council for Harmonisation Guidelines for Good Clinical Practice, the Declaration of Helsinki, the Pharmaceutical and Medical Device Act and applicable local regulations. All patients provided written consent before the start of the study. The study is registered at ClinicalTrials.gov (NCT03550443).

Study design and patient eligibility

This randomized, double-blind, placebo-controlled multicentre phase 3 study in DKD patients (AYAME study) is being performed to demonstrate the long-term efficacy and safety of once-daily multiple oral doses of bardoxolone methyl (5, 10 or 15 mg according to the intrapatient dose adjustment design). Eligible patients are ≥20–<80 years of age and have an eGFR ≥15.0–<60.0 ml/min/1.73 m² and a urinary albumin:creatinine ratio (UACR) ≤3500 mg/g without a BNP >200 pg/ml or a history of heart failure (Table 1). This study is designed to be adaptive in two ways: the study duration may be extended if the number of events of the primary efficacy endpoint is fewer than anticipated and the sample size may be increased before the end of enrolment if there is an obvious decrease in the statistical power to assess the primary endpoint.

Table 1:

Main patient eligibility criteria and masked parameters.

Inclusion criteria

1. Patients with stage G3/G4 DKD

2. Age ≥20–<80 years at the time of consent

3. Mean eGFR ≥15.0–<60.0 ml/min/1.73 m2 at screening

4. UACR ≤3500 mg/g at screening

5. Treatment with an ACE inhibitor and/or an ARB for at least 8 weeks before screening, with no change in dosage or medication

Exclusion criteria

1. Diabetes mellitus that is neither type 1 nor type 2

2. Progression of CKD mainly attributed to a non-diabetic cause

3. History of renal transplantation or confirmed plan to undergo renal transplantation

4. SBP >160 mmHg or DBP >90 mmHg

5. Haemoglobin A1c >10.0% at screening

6. Serum albumin ≤3.0 g/dl at screening

7. Cardiovascular disease defined as any of the following: • Congenital heart disease (e.g. atrial septal defect, ventricular septal defect or patent ductus arteriosus) categorized as NYHA class ≥II • Acute coronary syndrome (unstable angina pectoris or acute myocardial infarction) within 1 year until screening • Valvular heart disease requiring treatment • Hereditary or mixed primary cardiomyopathy • Acute or chronic constrictive pericarditis • Unstable paroxysmal or chronic atrial fibrillation (e.g. frequent attacks or poorly controlled heart rate) • Severe bradyarrhythmia (e.g. sinus node dysfunction or atrioventricular block of second degree or higher) not treated with a pacemaker • Tachyarrhythmia (including supraventricular arrhythmia and ventricular arrhythmia) not stabilized by treatment • Cerebrovascular accident, including transient ischemic attack within 1 year until screening • Class III or IV NYHA congestive heart failure

8. History of heart failure (e.g. a prior diagnosis of congestive heart failure defined as NYHA class III or IV or a history of hospitalization for heart failure)

9. BNP >200 pg/ml at screening

10. Left ventricular fractional shortening <22% on echocardiography

11. Ongoing or worsening of symptoms/signs of uncontrolled peripheral oedema

12. Abnormal hepatobiliary function (total bilirubin >1.5 times the ULN, AST or ALT of >1.5 times the ULN, ALP > ULN) at screening

13. Change of hypoglycaemic agents (other than insulin), antihypertensive medications (other than ACE inhibitors and ARBs) or their dosing regimen within 4 weeks until screening

14. Dialysis, acute kidney injury or rapid improvement or worsening of CKD within 8 weeks until screening

15. Diagnosis or a history of treatment of a malignancy within 5 years of screening

16. Ineligible for initiation, addition or change of diuretics when diuretic therapy is needed after initiation of study drug administration

Masked parameters

AST, ALT, γ-glutamyl transpeptidase, serum creatinine, eGFR, urinary albumin, urinary protein, urinary creatinine, UACR, UPCR, and plasma bardoxolone methyl concentrations

ACE, angiotensin-converting enzyme; ALP, alkaline phosphatase, ALT, alanine aminotransferase; ARB, angiotensin II receptor blocker; AST, aspartate aminotransferase; BNP, B-type natriuretic peptide; DBP, diastolic blood pressure; SBP, systolic blood pressure; NYHA, New York Heart Association; UACR, urinary albumin:creatinine ratio; UPCR, urinary protein:creatinine ratio; ULN, upper limit of normal.

The study comprises a screening period of up to 4 weeks, treatment period of approximately 3–4 years and post-treatment observation period of 16 weeks (Fig. 1). Patients were originally planned to be treated with either bardoxolone methyl or placebo for 2 years after the treatment initiation date of the last enrolled patient. However, only 1.3% of the planned sample size of 700 patients was expected to be treated with the study drug for 3 years based on the actual patient enrolment status. With such a low percentage of patients being treated for 3 years, a 30% or 40% decrease in the eGFR within 3 years may not be assessed, which raises the concern that the clinical significance of bardoxolone methyl may not be fully evaluated. On the basis of the estimation with the current enrolment status, the majority of the 700 patients are expected to be treated for ≥3 years if patients are treated for ≥2.5 years from the treatment initiation date of the last enrolled patient; this will enable us to evaluate a 30% or 40% decrease in the eGFR within 3 years. Accordingly, a study protocol amendment to treat patients for 2.5 years after the treatment initiation date of the last enrolled patient was finalized in May 2021. This was subsequently amended to 3 years in December 2021 so that the efficacy and safety of bardoxolone methyl can be evaluated in patients who will be treated for at least 3 years.

Figure 1:

AYAME study diagram. (a) Patients visit the study sites 2 weeks after initiation of the study treatment and every 4 weeks thereafter during the treatment period for examinations other than the following cases. Patients visit the study sites 2 weeks after a dose increase or resumption of treatment. Patients visit the study sites within 7 days after the day of examination if the aspartate aminotransferase or alanine aminotransferase level exceeds three times the respective upper limit of normal or the alkaline phosphatase level exceeds two times the upper limit of normal. If any of the following applies, a patient will visit the study site within 7 days: body weight gain ≥3.5% associated with volume overload or symptoms/signs of volume overload suspected by telephone interview, B-type natriuretic peptide level >300 pg/ml or use of a potent CYP3A4 inhibitor at the time of initiation or resumption of study treatment or a dose increase during the study treatment. (b) Patients visit the study sites every 4 weeks after termination of the study treatment.

The dose adjustment criteria for bardoxolone methyl and the study withdrawal criteria are summarized in Table 2. The starting dose of bardoxolone methyl is 5 mg, which will be increased up to 15 mg if the eGFR declines from baseline and there are no safety issues. The dose will be promptly reduced if patients on 10 or 15 mg meet any of the dose reduction criteria. The patients will visit the study sites 2 weeks after the initiation of the study treatment and then every 4 weeks during the treatment period and post-treatment observation period for examinations. Additional tests will be performed if any safety concerns arise. Patients who are withdrawn will be requested to make study visits by following the schedule to the greatest extent possible. Safety data will be monitored by a separate independent data monitoring committee (IDMC) specializing in liver disease in addition to the regular IDMC. The IDMC specializing in liver disease comprises hepatologists who periodically review the hepatobiliary-related safety in an open-label manner and make recommendations about the appropriateness of study continuation.

Table 2:

Dose adjustment for bardoxolone methyl and withdrawal criteria.

Dose increase Daily dose of bardoxolone methyl will be increased at the scheduled visits when the patients meet all of the dose increase criteria.

1. Decrease in eGFR at the last previous visit compared with the eGFR at baseline

2. No ≥3.5% increase in body weight associated with volume overload compared with the last previous scheduled visit

3. Eligible for initiation, addition or change of diuretic therapy

4. No onset or worsening of symptoms/signs of uncontrollable peripheral oedema

5. ≥4 weeks after the day of dose reduction, if the dose of bardoxolone methyl has been reduced

6. ≥4 weeks after the day of treatment resumption, if study treatment has been resumed after interruption

7. Judged by the investigator or subinvestigator that the dose of bardoxolone methyl can be increased

Note that the dose of bardoxolone methyl should be a maximum of 10 mg for patients receiving a potent CYP3A4 inhibitor.

Dose reduction The dose of bardoxolone methyl should be promptly reduced if patients on 10 or 15 mg meet any of the dose reduction criteria.

1. A ≥3.5% increase in body weight associated with volume overload compared with the last previous scheduled visit

2. An increase in the BNP level to ≥2 times the level at baseline with a value of >300 pg/ml If patients meet the above criterion while on bardoxolone methyl 5 mg, it will be interrupted according to the discretion of the investigator or subinvestigator.

3. Onset or worsening of symptoms/signs of peripheral oedema as an early manifestation of acute volume overload noted during physical examination or telephone interview

4. Judged by the investigator or subinvestigator that the dose of bardoxolone methyl should be reduced for reasons other than those stated above

Dose interruption The dose of bardoxolone methyl should be promptly interrupted if patients on 10 or 15 mg meet any of the dose interruption criteria. The duration of interruption should be 12 weeks at the maximum; if interruption of >12 weeks is required, bardoxolone methyl should be permanently discontinued.

1. A ≥3.5% increase in body weight associated with volume overload compared with the last previous scheduled visit

2. An increase in the BNP to ≥2.5 times the level at baseline with a value >400 pg/ml

3. Onset or worsening of symptoms/signs of peripheral oedema as an early manifestation of acute volume overload noted during physical examination or telephone interview

4. Inability to check the patient's condition after treatment for a certain period of time for patient's personal reasons, as judged by the investigator or subinvestigator

5. An increase in a hepatobiliary parameter defined as any of the following parameters in the results measured at the central laboratory and the unavailability of an additional test for confirmation for these parameters within 7 days after the first test: AST or ALT >3 times the ULN or ALP >2 times the ULN

6. Confirmed concrete actions to be taken for renal transplantation (such as hospitalization for renal transplantation and use of immunosuppressants)

7. Judgment by the investigator or subinvestigator that the bardoxolone methyl treatment should be interrupted for reasons other than those stated above

Treatment resumption Study treatment will be resumed when any of the following criteria are met at a scheduled visit and when the investigator or subinvestigator considers it appropriate. Treatment should be resumed at the dose before interruption or at a lower dose.

1. In the case of dose interruption 1, recovery to a <3.5% increase in body weight compared with the body weight measured on the last visit day before interruption with the use of medication (e.g. diuretics)

2. In the case of dose interruption 2 or because of an increase in the BNP level to ≥2 times the level at baseline (with a value >300 pg/ml) while on 5 mg of bardoxolone methyl, a decrease in the BNP level to ≤300 pg/ml with the use of medication (e.g. diuretics)

3. In the case of dose interruption 3, resolution or resolving of the symptoms/signs with the use of medication (e.g. diuretics)

4. In the case of dose interruption 4, judgment by the investigator or subinvestigator that the safety can be checked

5. In the case of dose interruption 5, a decrease in the hepatobiliary parameter to the normal range in an additional test conducted ≥8 days after the first test

6. In the case of dose interruption 6, followed by the postponement of renal transplantation: for patients who have received immunosuppressants for renal transplantation, the study treatment should be resumed within 12 weeks after interruption, with a washout period of at least 8 weeks after the discontinuation of immunosuppressants.

7. In other cases wherein the investigator or subinvestigator judges that study treatment can be resumed

Treatment discontinuation Treatment with bardoxolone methyl must be discontinued if patients meet any of the following criteria. When the data obtained at the central laboratory after examination at a study visit are found to meet any of criteria 6–10, instructions to discontinue bardoxolone methyl must be made immediately by telephone.

1. The patient wishes to withdraw from the bardoxolone methyl

2. The patient is found to not meet the eligibility criteria after the start of the study

3. The patient cannot undergo necessary observations and examinations for personal reasons

4. The patient becomes pregnant or intends to become pregnant

5. The investigator or subinvestigator determines that the patient should be withdrawn from bardoxolone methyl because of an adverse event experienced by the patient.

6. AST or ALT level >5 times the ULN

7. AST or ALT level >4 times the ULN for two consecutive measurements

8. AST or ALT level >3 times the ULN with total bilirubin >2 times the ULN

9. AST or ALT level >3 times the ULN with pyrexia, rash or >5% increase in eosinophils

10. ALP level >4 times the ULN with pyrexia, rash or >5% increase in eosinophils

11. More than 12 consecutive weeks of interruption of bardoxolone methyl according to the above interruption criteria after initiation of bardoxolone methyl administration

12. ESKD is judged to be present by the investigator or subinvestigator

13. The investigator or subinvestigator determines that the patient should be newly prescribed SGLT2 inhibitors or GLP-1 receptor agonists

14. Hospitalization for renal transplantation

15. The investigator or subinvestigator determines that the patient should be withdrawn from bardoxolone methyl for a reason other than those stated above

Withdrawal (early termination) Patients who meet any of the following criteria must be withdrawn from the study.

1. The patient wishes to withdraw censent from the study

2. The patient does not meet the eligibility criteria after the start of the study

3. The patient cannot undergo necessary observations and examinations for personal reasons

4. The investigator or subinvestigator determines that the patient should be withdrawn from the study because of an adverse event experienced by the patient

5. The patient becomes pregnant or intends to become pregnant

6. The patient undergoes maintenance dialysis for 4 weeks

7. The eGFR remains at ≤6 ml/min/1.73 m2 for 4 weeks

8. The investigator or subinvestigator determines that the patient should be newly prescribed SGLT2 inhibitors or GLP-1 receptor agonists

9. The patient undergoes renal transplantation

10. The patient develops reduced muscle mass with severe physical inability

11. The investigator or subinvestigator determines that the patient should be withdrawn from the study for a reason other than above

ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BNP, B-type natriuretic peptide; GLP-1, glucagon-like peptide-1; SGLT2, sodium-glucose cotransporter 2; ULN, upper limit of normal.

After initiation of the study treatment, patients, investigators, medical staff and sponsor will remain blinded to the laboratory parameters measured at the central laboratory until all patients complete the post-treatment observation period (Table 1).

Randomization

Enrolled patients have been centrally randomized (1:1) to receive bardoxolone methyl or placebo using an interactive web response system with the minimization method using baseline characteristics in the order of UACR (<30, ≥30–<300, ≥300–<2000 and ≥2000 mg/g), eGFR (<30.0, ≥30.0–<45.0 and ≥45.0 ml/min/1.73 m²) and haemoglobin A1c level (<7.5% and ≥7.5%) for each type 1 diabetes mellitus (T1DM) and T2DM patient. The number of patients with a UACR ≥2000 mg/g measured at screening should be ≤15% of the target patient number.

Endpoints

Efficacy endpoints

The primary efficacy outcomes will be evaluated using the composite renal endpoint of the time to onset of a ≥30% decrease in the eGFR from baseline or ESKD (Table 3). In this study, eGFR is calculated using the following equation, which is more accurate for the Japanese population: eGFR (ml/min/1.73 m²) = 194 × serum creatinine^−1.094 × age^−0.287 (× 0.739 for female patients) [20]. The efficacy will also be evaluated by the key secondary endpoint (i.e. time to onset of a ≥40% decrease in the eGFR from baseline or ESKD); other secondary endpoints, including time to onset of a ≥53% decrease in the eGFR from baseline or ESKD and change in the eGFR from baseline at each evaluation time point; and exploratory endpoints. In this study a 53% decrease in the eGFR from baseline will be used because it is equivalent to doubling of the serum creatinine level used as the conventional ESKD surrogate endpoint when converted to the percentage change in the eGFR based on the eGFR estimation equation of the Japanese Society of Nephrology [19]. An independent Kidney Event Assessment Committee (Kidney EAC), comprising specialists in the field, has been established to independently evaluate the events of the efficacy endpoint–related ESKD in a blinded manner. ESKD is defined as initiation of maintenance dialysis, renal transplantation, a sustained eGFR ≤6 ml/min/1.73 m² for ≥4 weeks and other cases of ESKD as determined by the Kidney EAC. Evaluation of the efficacy endpoints includes only events judged as events of efficacy endpoints by the Kidney EAC. The evaluation period of all efficacy endpoints is the treatment period and observation period.

Table 3:

Primary, secondary and exploratory efficacy endpoints and definition of ESKD.

Efficacy endpoints
Primary composite endpoint
Time to onset of a ≥30% decrease in eGFR from baseline
Time to ESKD
Secondary endpoints
Important secondary composite endpoint
Time to onset of a ≥40% decrease in eGFR from baseline
Time to ESKD
Other secondary endpoints
Time to onset of a ≥53% decrease in eGFR from baselinea or ESKD
Time to onset of ESKD
Change in eGFR from baseline at each evaluation time point
Exploratory endpoints
Quality of life (Health Survey Short Form-36 and European Quality of Life 5-Dimension Questionnaire)
Definition of ESKD
Initiation of maintenance dialysis
Renal transplantation
Sustained eGFR ≤6 ml/min/1.73 m2 for ≥4 weeks
Other ESKD cases as determined by the Kidney Event Assessment Committee

^aEquivalent to doubling of the serum creatinine level used as the conventional ESKD surrogate endpoint when converted to the percentage change in the eGFR based on the eGFR estimation equation of the Japanese Society of Nephrology.

Safety endpoints

Safety endpoints are adverse events, time to onset of cardiac events [heart-related death (cardiac death), non-fatal myocardial infarction, heart failure requiring hospitalization, sudden death of unknown cause, unstable angina pectoris requiring hospitalization and serious arrhythmia] and changes in laboratory data, vital signs, echocardiography, 12-lead electrocardiography and chest X-ray (Table 4). An independent Heart EAC that comprises specialists in the field has been established to independently evaluate the above-mentioned cardiac events in a blinded manner and only those judged as events of safety endpoints by the Heart EAC will be included in the assessment in this study. The evaluation period of all safety endpoints is the treatment period and observation period.

Table 4:

Test items and schedule for cardiovascular risk monitoring.

Characteristics	Screening 2 (baseline assessment)	Week 2 of study treatment	Every 4 weeks of study treatment	Every 24 weeks of study treatment	Termination of study treatment	Post-treatment examination (every 4 weeks after termination of study treatment)
Weight, vital signs	X	X	X	X	X	X
Laboratory test (blood test)	X	X	X	X	X	X
Laboratory test (urinalysis)	X	X	X	X	X	X
Electrocardiography, echocardiography, chest radiography	X			X	X

Vital signs: blood pressure in a sitting position, pulse rate in a sitting position and body temperature (axillary).

Blood tests: creatinine, eGFR, albumin, BUN, BNP, NT-pro BNP, and troponin T.

Urinalysis: UACR and UPCR.

BNP, B-type natriuretic peptide; BUN, blood urea nitrogen; NT-pro BNP, N-terminal fragment of pro-BNP; UACR, urinary albumin:creatinine ratio; UPCR, urinary protein:creatinine ratio.

Statistical analyses

Sample size

This study started with a target sample size of 700 patients (350 patients in each group) on the basis of an expected hazard ratio (HR) of 0.6 for the bardoxolone methyl group versus the placebo group and accumulative incidence of the primary endpoint estimated conservatively in the Chronic Kidney Disease Japan Cohort (CKD-JAC) study [21] of 30% for the placebo group to demonstrate prolongation of the time to a ≥30% decrease in the eGFR from baseline or ESKD after treatment with bardoxolone methyl compared with placebo in the logrank test at a significance level of 5% (two-tailed) with ≥90% power [14, 15]. Patient enrolment in the AYAME study began on 25 May 2018 (treatment started on 30 May 2018) and a decrease in the statistical power was predicted before enrolling 700 patients. The recalculated incidence of ‘the event’ estimated in the CKD-JAC study, based on the distribution of patients enrolled in this study at the completion of enrolment of 700 patients, was 25.3%. Based on the incidence of ‘the event’ of the primary endpoint at that point in time, the discontinuation/withdrawal rate and the fact that the CKD-JAC is an epidemiologic study, the 2-year incidence of ‘the event’ in the placebo group in this study was conservatively set at 20%. When the enrolment period and the follow-up period in this study was set at 1.5 years and 2 years, respectively, the 2-year incidence of ‘the event’ in the placebo group was 20%, the expected HR for the bardoxolone methyl group versus the placebo group was 0.6 and the rate of discontinuation/withdrawal of patients before the onset of ‘the event’ was 30%; the number of patients required for a power of 90% in the logrank test at a significance level of 5% was calculated to be 950 patients (475 per group). Accordingly, an increase in the number of patients to 950 was decided by the sponsor based on consensus between the EAC and sponsor.

Patient populations

The efficacy endpoints will be evaluated in the full analysis set, which includes all randomized patients except those who are not being treated with any study drug and those who have no evaluable eGFR data after the start of treatment with the study drug. The safety endpoints will be evaluated in the safety analysis set, which includes all randomized patients except those who are not being treated with any study drug.

Analyses of endpoints

Unless stated otherwise, categorical data will be summarized using incidence and proportion and continuous data will be summarized using summary statistics by treatment group. The summary statistics will include the number of patients, mean, standard deviation (SD), minimum, median and maximum.

Primary analysis will be conducted using a stratified logrank test with stratification factors, baseline eGFR (<30.0 and ≥30.0 ml/min/1.73 m²) and UACR (<300 and ≥300 mg/g) to verify the superiority of bardoxolone methyl to placebo. As a sensitivity analysis, the HR between treatment groups and its 95% confidence interval (CI) will be estimated using a Cox regression model with the treatment group as a factor and with the baseline eGFR (<30.0 and ≥30.0 ml/min/1.73 m²) and UACR (<300 and ≥300 mg/g) as covariates. In addition, the HR between treatment groups and its 95% CI will be estimated using a Cox regression model with the treatment group as a factor for subgroup analysis (baseline eGFR <30.0 and ≥30.0 ml/min/1.73 m² and UACR <300 and ≥300 mg/g).

The following demographic and other baseline characteristics will be summarized: sex, age (<65 or ≥65 years as well as descriptive statistics), height, weight at baseline, body mass index (BMI) at baseline (<23 or ≥23 kg/m² as well as descriptive statistics), waist circumference, eGFR at baseline, CKD stage by eGFR at baseline, type of diabetes mellitus, UACR at baseline (<30, ≥30–<300, ≥300–<2000 and ≥2000 mg/g as well as descriptive statistics), BNP at baseline, glycated haemoglobin (%) at baseline, systolic blood pressure at baseline and diastolic blood pressure at baseline. The geometric mean will also be calculated with respect to the UACR at baseline.

Current status

Patient enrolment was completed on 25 June 2019 (treatment started on 26 June 2019) with 1323 patients from 139 study sites. Study treatment has been ongoing in a double-blind manner. The data cut-off date was 27 May 2021.

RESULTS

In total, 1120 patients were eligible for pre-enrolment and 1020 were enrolled and randomized to receive bardoxolone methyl or placebo. Of these patients, 1013 started the study treatment (Fig. 2A). Once patient recruitment started, we continuously enrolled patients throughout the screening period (Fig. 2B).

Figure 2:

(A) Patient disposition and (B) trend in cumulative numbers of patients who started the study treatment. ^aSome patients were ineligible for multiple reasons. BNP, B-type natriuretic peptide; DBP, diastolic blood pressure; SBP, systolic blood pressure; UACR, urinary albumin:creatinine ratio.

The mean age of the 1013 patients is 65.9 years (SD 9.4) and the majority are male [795 patients (78.5%)] (Table 5). The mean BMI is 26.28 kg/m² (SD 4.24), systolic/diastolic blood pressure is 133.09 mmHg (SD 14.56)/73.92 mmHg (SD 10.01), eGFR is 37.84 ml/min/1.73 m² (SD 12.64), UACR is 712.39 mg/g (SD 834.22), BNP is 30.21 pg/ml (SD 28.65) and haemoglobin A1c concentration is 6.96% (SD 0.83). Among the 1013 patients, 982 (96.9%) have T2DM. Additionally, 363 (35.8%) had an eGFR of category G3a, 323 (31.9%) had category G3b and 327 (32.3%) had category G4. The UACR was <30 mg/g in 163 (16.1%) patients, ≥30–<300 mg/g in 314 (31.0%) patients, ≥300–<2000 mg/g in 428 (42.3%) patients and ≥2000 mg/g in 108 (10.7%) patients. All patients were treated with angiotensin-converting enzyme inhibitors and/or angiotensin II receptor blockers as of the data cut-off date.

Table 5:

Patients’ baseline characteristics.

Characteristics	Values (N = 1013)
Sex (male), n (%)	795 (78.5)
Age (years)	65.9 ± 9.4
Height (cm)	163.98 ± 8.45
Weight, kg	70.78 ± 13.13
BMI (kg/m2)	26.28 ± 4.24
Waist circumference (cm)	93.43 ± 10.62
eGFR (ml/min/1.73 m2)	37.84 ± 12.64
CKD stage by eGFR, n (%)
G3a	363 (35.8)
G3b	323 (31.9)
G4	327 (32.3)
Type of diabetes mellitus, n (%)
1	31 (3.1)
2	982 (96.9)
UACR (mg/g), n (%)	712.39 ± 834.22
<30	163 (16.1)
≥30–<300	314 (31.0)
≥300–<2000	428 (42.3)
≥2000	108 (10.7)
BNP (pg/ml)	30.21 ± 28.65
HbA1c (%)	6.96 ± 0.83
Systolic blood pressure (mmHg)	133.09 ± 14.56
Diastolic blood pressure (mmHg)	73.92 ± 10.01

Data are presented as mean ± standard deviation unless stated otherwise.

BNP, B-type natriuretic peptide; HbA1c, glycated haemoglobin; UACR, urinary albumin:creatinine ratio.

DISCUSSION

This study is designed to evaluate the long-term efficacy and safety of bardoxolone methyl with surrogate renal outcomes as composite efficacy endpoints in DKD patients. Patient enrolment has been completed with an expected ≥90% power and treatment has begun. Compared with the off-drug period of 12 weeks in the TSUBAKI study, that in the AYAME study is 16 weeks, enabling us to assess ESKD and the eGFR for a longer off-drug period for evaluation of the efficacy endpoints. The TSUBAKI study has demonstrated an increased BNP and UACR in patients treated with bardoxolone methyl [16]. Therefore we will evaluate the BNP and UACR in the same way as in the TSUBAKI study (i.e. during the treatment period and throughout the 16-weeks post-treatment observation period) to ensure the safety of patients.

To ensure patient safety, vigilant safety measures are implemented in the AYAME study in addition to excluding patients with known risk factors for heart failure (Table 6) [17]. First, different criteria are applied to adjust the daily bardoxolone methyl dose compared with the criteria used in the BEACON [15] and TSUBAKI studies [16]. A fixed dose of 20 mg and a dose titration scheme was used in the BEACON study and the TSUBAKI study, respectively, whereas an intrapatient dose adjustment design (dose can be increased only when the eGFR is decreased and can be reduced if there is a safety concern) is used in this study. The starting dose is 5 mg and can be increased up to 15 mg, whereas it was 20 mg in the BEACON study [15]. Furthermore, the estimated exposure to bardoxolone methyl by body weight shows that patients in the AYAME study are likely exposed to bardoxolone methyl less than those in the BEACON study, and the exposure by body weight will be similar to that in the BEACON study even if patients are treated with 15 mg. After completing the study, the safety of 15 mg bardoxolone methyl (i.e. bardoxolone methyl exposure similar to that in the BEACON study by body weight) can be evaluated. Second, in the AYAME study, patients’ body weights are carefully monitored and the dose is reduced or interrupted if a volume overload–related ≥3.5% increase in body weight is observed. Third, the IDMC has been monitoring patients’ safety during the entire study period. Furthermore, an IDMC specializing in liver disease has been established and monitors patient safety in detail.

Table 6:

Safety measures implemented in the AYAME and BEACON studies.

Safety measure	AYAME	BEACON [15]
Patient exclusion criteria	BNP >200 pg/ml or history of heart failure
Bardoxolone methyl daily dose Mean (range) daily dose per body weight	Maximum 15 mg, intrapatient dose adjustment design Bardoxolone methyl daily dose 5 mg: 0.07 mg/kg (0.04–0.13); 10 mg: 0.14 mg/kg (0.08–0.25); 15 mg: 0.21 mg/kg (0.12–0.38)	Fixed dose at 20 mg Bardoxolone methyl daily dose 20 mg: 0.21 mg/kg (0.10–0.43)
Safety monitoring	• A telephone interview is conducted 2 weeks after the day of each scheduled visit and 1 week after the additional test by the investigator/subinvestigator or a clinical research coordinator checks the patients’ health conditions and whether any change in body weight or signs of volume overload have occurred • Patients measure and record their body weight at home every day after initiation of the study treatment until the end of treatment. In the event of a ≥3.5% increase in body weight associated with volume overload or suspected symptoms/signs of volume overload, the patient should contact the investigator, subinvestigator or clinical research coordinator by telephone

BNP, B-type natriuretic peptide.

Previous large-scale clinical studies have assessed the efficacy of study drugs in DKD patients using renal outcomes as efficacy endpoints, but there are major differences in the study designs compared with the AYAME study (Table 7). As a composite primary endpoint, a doubling of serum creatinine, ESKD or renal/cardiovascular death was used in previous studies [the Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation (CREDENCE) [22, 23] and the Study Of diabetic Nephropathy with AtRasentan (SONAR)] [24, 25]. The guidelines indicate that a 30% or 40% decrease in the eGFR within 2 or 3 years may be adopted as a surrogate endpoint in DKD and other types of CKD [18, 19]. Accordingly, more recent studies such as Finerenone in Reducing Kidney Failure and Disease Progression in DKD (FIDELIO-DKD) [26] and Dapagliflozin and Prevention of Adverse outcomes in Chronic Kidney Disease (DAPA-CKD) [27, 28] used a ≥40% and 50% decline in the eGFR, respectively. In the guidelines [19], a ≥30% decrease in the eGFR is also considered appropriate as a surrogate endpoint for a drug that shows positive acute effects continuing over a long term. Because bardoxolone methyl shows a positive acute effect on the eGFR [14–16], a ≥30% decrease in the eGFR appears to be appropriate in this study. To determine the feasibility of the primary endpoint in assessing DKD status, we have included secondary endpoints by which the trend in the primary endpoint will be confirmed. The use of surrogate endpoints is expected to decrease the sample size and the duration of follow-up needed to obtain a high enough number of events to draw solid conclusions.

Table 7:

Comparison of main patient eligibility criteria and baseline characteristics between the AYAME study and previous studies with a composite kidney endpoint as the primary outcome in patients with CKD.

	AYAME	CREDENCE [22, 23]	DAPA-CKD [27, 28]	FIDELIO [26]	SONAR [24, 25]
Study design	N = 1013	N = 4401	N = 4304	N = 5674	N = 2648a
Study drug	Bardoxolone methyl	Canagliflozin	Dapagliflozin	Finerenone	Atrasentan
Regions	Japan	Asia, Africa, Europe, North America, Oceania, South America	Asia, Europe, North America, South America	Asia, Africa, Europe, North America, Oceania, South America	Asia, Africa, Europe, North America, Oceania, South America
eGFR range (ml/min/1.73 m2)	15–60	30–90	25–75	25–75	25–75
ACR range (mg/g)	≤3500	300–5000	200–5000	30–5000	300–5000
Patients with T1DM	Included	Excluded	Excluded	Excluded	Excluded
Composite renal primary endpoint	≥30% decline in eGFR or ESKD	ESKD, doubling of Cr or renal/CV death	Sustained ≥50% decline in eGFR, ESKD or renal/CV death	Sustained ≥40% decline in eGFR or renal death	Doubling of Cr or ESKD
Baseline characteristics
Patients (n)	Overall (1013)	Overall (4401)	Type 2 diabetes (2906)	Overall (5674)	Atrasentan responders (2648)
eGFR (ml/min/1.73 m2), mean	37.8	56.2	43.8	44.3	43.8
UACR (mg/g), median	351.80	927	1016.5	851	802

^aAtrasentan responders only.

^bPatients with T2DM only.

Cr, creatinine; CV, cardiovascular.

Another major difference between the AYAME and other studies is found in the patients’ eligibility criteria and therefore the characteristics of the enrolled patients. Because the safety of bardoxolone methyl in patients with stage G3/G4 CKD with T2DM was demonstrated in the TSUBAKI study [16], we decided to include patients with an eGFR ≥15.0–<60.0 ml/min/1.73 m² (Table 7). Consequently, the mean baseline eGFR in the AYAME study is indeed lower than that in other studies (37.8 versus 43.8–56.2 ml/min/1.73 m²), which indicates that the patients included in this study are closer to initiating dialysis than those in previous studies. However, the UACR in the AYAME study is limited to ≤3500 mg/g but has no lower limit, allowing patients with a lower UACR to be included (Table 7). According to the CKD nomenclature by GFR and albuminuria categories [29], this study covers patients with GFR categories G3–G4 and albuminuria categories A1–A3, unlike previous studies that did not include patients with category A1 (Fig. 3). Recent studies have identified a subset of patients with T1DM or T2DM who have non-proteinuric DKD, which is characterized by an eGFR <60 ml/min/1.73 m² and a UACR <300 mg/g [30–32]. Therefore, this study includes a higher percentage of non-proteinuric DKD patients than previous studies [22–28]. Because DKD comprises heterogeneous subpopulations of pathological characteristics, the AYAME study is expected to provide clinical evidence that will lead to a specific patient-targeted large-scale study.

Figure 3:

Distribution of patients in the AYAME study according to the CKD nomenclature by the Kidney Disease: Improving Global Outcomes National Kidney Foundation classification of CKD [29]. Data are presented as n (%).

In conclusion, this AYAME study is the first long-term, large-scale study since termination of the BEACON study to evaluate the efficacy and safety of bardoxolone methyl in DKD patients.

FUNDING

The AYAME study is supported by Kyowa Kirin.

AUTHORS’ CONTRIBUTIONS

All authors have made substantial contributions to this manuscript and were involved in critically revising it for intellectual content and accuracy. All authors have approved the final manuscript. M.N., H.H. and T.A. were responsible for conceptualization, methodology, reviewing and editing, supervision and project administration. H.T., T.I. and K.M. were responsible for conceptualization, methodology, writing the original draft, reviewing and editing and visualization. M.K. was responsible for conceptualization, methodology, formal analysis, writing the original draft, reviewing and editing and visualization. Y.S., H.W., T.W., K.U., I.N. and N.K. were responsible for resources, reviewing and editing and supervision. T.K. was responsible for reviewing, editing and supervision.

DATA AVAILABILITY STATEMENT

The data underlying this article cannot be shared publicly because the study is currently ongoing.

CONFLICT OF INTEREST STATEMENT

M.N. has received research grants, consulting fees and lecture fees from Kyowa Kirin and is vice president of the Japanese Society of Internal Medicine, president of the Asian Pacific Society of Nephrology and president-elect of the International Society of Nephrology. H.T., T.I., K.M. and M.K. are employees of Kyowa Kirin. Y.S. received grants from Kyowa Kirin during the study and has received honoraria from Kyowa Kirin. H.W. has received institutional grants from Takeda, Boehringer Ingelheim, Kissei, Novo Nordisk, Mitsubishi Tanabe, Lifescan Japan, Sumitomo Dainippon, Kyowa Kirin and Merck Sharp & Dohme; has received honoraria from Mitsubishi Tanabe, Sumitomo Dainippon, Sanwa Kagaku, Takeda, Sanofi, Kowa, Merck Sharp & Dohme, Boehringer Ingelheim, Eli Lilly and Novo Nordisk; and is a board member of the Japanese Diabetes Society. T.W. has received grants from Shiseido, Kyowa Kirin, Astellas, Chugai, Daiichi Sankyo, Kissei, Kowa, Mitsubishi Tanabe, Asahi Kasei Pharma, Mochida, Takeda, Merck Sharp & Dohme, Otsuka, Sanofi, Terumo, Torii and Hayashidera Medinol as well as honoraria from Astellas, AstraZeneca, Baxter, Bayer Yakuhin, Chugai, Daiichi Sankyo, Eli Lilly Japan, Kissei, Kowa, Kyowa Kirin, Mitsubishi Tanabe, Miyarisan, Takeda, Merck Sharp & Dohme, Nippon Boehringer Ingelheim, Ono, Otsuka, Sanofi, Sanwa Kagaku, Taisho Pharma, Teijin Pharma, Terumo, Torii, Vifor (International) AG and Fujirebio. K.U. has received honoraria from Takeda, Sumitomo Dainippon, Novo Nordisk and Mitsubishi Tanabe. N.K. has received consulting fees from AstraZeneca, Boehringer Ingelheim and Kyowa Kirin; has received honoraria from Kyowa Kirin, AstraZeneca, Otsuka, GlaxoSmithKline, Astellas, Boehringer Ingelheim, Novartis and Daiichi Sankyo; and has a leadership or fiduciary role in other boards for AstraZeneca, GlaxoSmithKline, Boehringer Ingelheim, Novartis and Gilead Sciences. T.K. has received grants from Nippon Boehringer Ingelheim, Eli Lilly Japan, Kyowa Kirin, Merck Sharp & Dohme, Daiichi Sankyo, Novo Nordisk, Sanofi, Takeda, Astellas, Ono, Mitsubishi Tanabe, Sumitomo Dainippon and Asahi Mutual Life Insurance Company and has received honoraria from Merck Sharp & Dohme, Takeda, Mitsubishi Tanabe, Astellas, Abbott Japan, Terumo, Teijin Pharma, Ono, AstraZeneca, Sumitomo Dainippon, Sanofi, Eli Lilly Japan, Nippon Boehringer Ingelheim, Novo Nordisk, Bayer Holding, Novartis, Daiichi Sankyo, Fujifilm Toyama Chemical, Kowa, Kyowa Kirin and Medtronic Japan. T.A. received grants from Kyowa Kirin during the study; has received consulting fees from Astellas, GlaxoSmithKline, Bayer, Japan Tobacco, Torii, Sanwa Kagaku, Nipro and Kyowa Kirin; and has received honoraria from Astellas, Bayer, Torii, Kissei, Mitsubishi Tanabe, Ono, Chugai, Fuso and Kyowa Kirin. The remaining authors have nothing to disclose.