Endothelin Receptor Antagonist Ambrisentan, Sodium-Glucose Cotransporter 2 Inhibitor Henagliflozin, and Their Combination in IgA Nephropathy

Visual Abstract

Abstract

Key Points

• Proteinuria responses to endothelin receptor antagonists and sodium-glucose cotransporter 2 inhibitors were not correlated.

• Ambrisentan-henagliflozin combination therapy reduced proteinuria similarly to ambrisentan monotherapy but exceeded henagliflozin alone.

• Combination therapy resulted in markedly less fluid retention and weight gain compared with endothelin receptor antagonist monotherapy.

Background

Sodium-glucose cotransporter 2 (SGLT2) inhibitors and endothelin receptor antagonists (ERAs) can reduce proteinuria and may slow kidney progression in IgA nephropathy. Their potential synergistic efficacy and safety are unclear.

Methods

This open-label randomized crossover trial enrolled adults with IgA nephropathy, eGFR >30 ml/min per 1.73 m², and 24-hour urine protein-creatinine ratio (UPCR) ≥0.44 g/g or 24-hour urine protein ≥0.5 g, despite maximal renin-angiotensin system blockade. Treatments were given in random order (ambrisentan 5 mg/d, henagliflozin 10 mg/d, and combination), each for 4 weeks followed by a 4-week washout. Primary end point was correlation between treatments in 24-hour UPCR changes. Secondary end points included eGFR change. Safety end points included fluid retention (body weight change).

Results

Sixty-five patients were enrolled, with a mean baseline eGFR of 69 ml/min per 1.73 m² (SD=23) and a median 24-hour UPCR of 0.8 g/g (0.6–1.3). No correlation was found in 24-hour UPCR reduction between single-agent therapy and combination therapy (ambrisentan, r=0.13, P = 0.97; henagliflozin, r=0.04, P = 1.00). After treatment, the mean reduction in 24-hour UPCR was −44% with combination therapy (95% confidence interval [CI], −52 to −34), which was similar to ambrisentan alone (−48%; 95% CI, −56 to −39) but significantly superior to henagliflozin alone (−21%; 95% CI, −33 to −6). The mean reduction in eGFR from baseline was −4.7 ml/min per 1.73 m² (95% CI, −6.9 to −2.6) with combination therapy, which was significantly greater than with ambrisentan (−0.5 ml/min per 1.73 m²; 95% CI, −2.8 to 1.9) but comparable with henagliflozin (−3.5 ml/min per 1.73 m²; 95% CI, −5.8 to −1.2). Less fluid retention occurred with combination treatment versus ambrisentan alone.

Conclusions

Combining an ERA with an SGLT2 inhibitor further reduced proteinuria versus SGLT2 inhibitor alone, with less fluid retention than ERA alone.

Clinical Trial registry name and registration number:

Chinese Clinical Trial Registry (ChiCTR2400080435).

Introduction

IgA nephropathy is the most common cause of primary glomerular disease globally. Proteinuria is associated with progressive kidney function decline and kidney failure.^1–4 The 2025 Kidney Disease Improving Global Outcomes guidelines recommend renin-angiotensin system inhibitors as first-line supportive therapy for proteinuria reduction; however, many patients still experience disease progression despite optimized treatment, highlighting the unmet need for novel therapeutic strategies.^2,5,6

Recent years have seen the emergence of several supportive treatments, notably sodium-glucose cotransporter 2 (SGLT2) inhibitors and endothelin receptor antagonists (ERAs). Sparsentan, a novel dual endothelin-angiotensin receptor antagonist, significantly reduced proteinuria by 40% compared with irbesartan at week 110 in the Efficacy and Safety of Sparsentan Versus Irbesartan in Patients with IgA Nephropathy trial.⁷ Similarly, in the Atrasentan in Patients with IgA Nephropathy (ALIGN) study, the selective endothelin A (ETA) receptor antagonist atrasentan produced a 36% greater reduction in proteinuria after 9 months versus placebo.⁸ However, their clinical utility may be limited by fluid retention.

In the IgA nephropathy subgroup of the Dapagliflozin and Prevention of Adverse Outcomes in Chronic Kidney Disease trial, dapagliflozin reduced albuminuria by 26% and attenuated the mean eGFR decline (−3.5 versus −4.7 ml/min per 1.73 m²/yr with placebo).⁹ Moreover, the diuretic action of SGLT2 inhibitors enables them to mitigate ERA-induced fluid retention.

The Atrasentan and Renal Events in Patients with Type 2 Diabetes and Chronic Kidney Disease trial demonstrated that adding an SGLT2 inhibitor to ERA therapy reduced body weight and enhanced albuminuria reduction compared with ERA alone,¹⁰ and their combination exhibited favorable safety profiles.¹¹ In the Zibotentan in Combination with Dapagliflozin Compared with Dapagliflozin in Patients with Chronic Kidney Disease trial, the zibotentan-alone arm was terminated due to higher rates of fluid retention; by contrast, low-dose zibotentan combined with dapagliflozin reduced albuminuria with acceptable tolerability and safety.¹² Nevertheless, this combination therapeutic strategy has not yet been evaluated in the context of IgA nephropathy.

We hypothesized that, in patients with IgA nephropathy, individual changes in proteinuria during treatment with an ERA and during treatment with an SGLT2 inhibitor would be significantly correlated, reflecting shared and patient-specific responsiveness across these mechanistically distinct therapies. To explore this, we selected ambrisentan as a potent and selective ETA receptor antagonist demonstrating more than 4000-fold ETA over endothelin B selectivity, based on its well-defined pharmacology and prior evidence of rapid proteinuria reduction (41% within 4 weeks) in IgA nephropathy.^13,14 Henagliflozin was used as the representative SGLT2 inhibitor.¹⁵ The proof-of-concept trial aimed to mechanistically evaluate the proteinuria-lowering efficacy and safety of ERA, SGLT2 inhibitor, and their combination, thereby addressing the critical knowledge gap in IgA nephropathy management.

Methods

Study Design and Patients

We conducted a prospective randomized open-label crossover clinical trial at Peking University First Hospital. The study was approved by the local ethics committee (2024-004) and registered with the Chinese Clinical Trial Registry (ChiCTR2400080435). It was conducted in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki. All participants provided written informed consent before enrollment. Eligible participants were adults with biopsy-proven primary IgA nephropathy. The inclusion criterion for 24-hour urine protein-creatinine ratio (UPCR) was revised to ≥440 mg/g (or 24-hour urine protein ≥500 mg) after the first 4 months of recruitment to align with recent evidence and expedite enrollment.^3,16 Additional inclusion criteria included eGFR >30 ml/min per 1.73 m², systolic/diastolic BP ≤130/80 mm Hg, and stable use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers at maximally tolerated dose for ≥4 weeks before randomization. Key exclusions were secondary IgA nephropathy, recent immunosuppressive therapy, heart failure, hospitalization for fluid retention, or B-type natriuretic peptide (BNP) >200 pg/ml. The full inclusion and exclusion criteria are listed in Supplemental Table 1.

Procedures

Eligible patients entered a run-in phase to stabilize renin-angiotensin system blockade. Those already on stable doses provided two 24-hour urine samples collected 2 weeks apart. Participants were randomly assigned via stratified block randomization, stratified by baseline eGFR (≥45 or <45 ml/min per 1.73 m²) and 24-hour UPCR (≥3 or <3 g/g); details of randomization and allocation concealment are provided in Supplemental Section 1. Participants received each of the following 4-week treatments in randomly assigned order: ambrisentan (5 mg daily), henagliflozin (10 mg daily), or their combination, separated by 4-week washout periods. The 4-week duration for both treatment and washout was selected based on prior evidence that the full pharmacologic effects of SGLT2 inhibitors and ERAs are established within this time frame.^7,17–19 Dose adjustments were permitted per investigator's judgment. Medications were dispensed at the beginning of each phase, and concomitant therapies were required to remain unchanged throughout. Clinic visits occurred at baseline and week 4 of each period, with a safety call at week 2. A final follow-up was conducted 4 weeks post-treatment.

Measurements

At each visit, blood and urine samples were collected for laboratory analyses, including eGFR (CKD-Epidemiology Collaboration equation), UPCR from 24-hour urine collection, 24-hour protein excretion (urine protein), and urinalysis. Vital signs, adverse events, and medication adherence were recorded. Medication adherence was calculated as (number of tablets taken/number of tablets prescribed)×100%, where the prescribed number was based on the individual titration schedule.

End Points

The primary end point was the correlation between individual changes in 24-hour UPCR from baseline across treatments. Secondary end points included percentage changes in 24-hour UPCR and 24-hour urine protein and absolute changes in systolic BP and eGFR from baseline during each treatment period. Exploratory outcomes included achieving prespecified proteinuria reduction thresholds (≥30%, ≥50%, or complete remission), along with subgroup analyses and a post hoc analysis adjusting for systolic BP changes. Safety end points included adverse events of special interest, such as hypotension, AKI, and fluid retention (defined as edema with weight gain >2 kg, BNP >300 pg/ml, or heart failure symptoms), and the longitudinal assessment of hemoglobin and body weight.

Statistical Analyses

A sample size of 46 provided 80% power (α=0.05) to detect a Pearson correlation coefficient (r) ≥0.4 between log-transformed UPCR responses across treatments. The null hypothesis assumes no correlation in UPCR response among ambrisentan, henagliflozin, and their combination. A Pearson correlation of 0.4 is considered to be clinically relevant.^20,21 To account for an anticipated 20% dropout rate, a sample size of 60 was planned. The sample size would provide 90% power (α=0.05) to detect a 30% difference in log-transformed UPCR (SD=0.432) between combination and henagliflozin alone. Ultimately, 65 participants were randomized, slightly exceeding the target due to a higher enrollment and lower screen failure rates than expected.

Baseline characteristics were summarized using descriptive statistics. Continuous variables were presented as means with SDs or medians with interquartile ranges (IQRs), as appropriate. All analyses followed the intention-to-treat principle. Correlation of UPCR responses across treatments was evaluated using Pearson r. Longitudinal UPCR measurements and treatment effects were analyzed using a mixed-effects linear model with sequence, period, treatment, and baseline log-UPCR as fixed effects and a patient-specific random intercept. UPCR values were log-transformed before analysis. Treatment effects were expressed as geometric mean percentage changes in 24-hour UPCR, calculated using the formula 100×(exp [least squares mean change]−1), with corresponding 95% confidence intervals (CI) derived by back-transforming the log-scale limits. This model was applied uniformly to primary and secondary efficacy outcomes. We assessed the assumptions of the crossover design by testing for period and sequence (carryover) effects using likelihood ratio tests. The proportions of participants achieving ≥30% or ≥50% reduction in 24-hour UPCR were compared using logistic regression, with odds ratios (OR) and 95% CIs estimated using the profile likelihood method. All P values were two-sided, with significance defined as P < 0.05. Bonferroni correction was applied to the primary correlation and prespecified combination versus monotherapy comparisons, with adjusted P values reported. All other analyses were unadjusted.

Prespecified subgroup analyses (e.g., age, gender, diuretic use, and baseline parameters) were conducted by testing treatment-by-subgroup interactions in separate linear mixed-effects models. Owing to the enrollment imbalance across the four prespecified randomization strata (with three strata containing zero, two, and five participants, respectively), the proteinuria subgroup analysis was performed by dichotomizing participants based on the median baseline 24-hour UPCR (0.8 g/g). These analyses are exploratory. All statistical analyses were performed using R software, version 4.4.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Patient Disposition and Baseline Characteristics

Between April 2024 and January 2025, 103 patients were screened and 65 were randomized. Two participants withdrew consent, leaving 63 participants who completed the study (Figure 1). All 65 randomized participants were included in the full analysis set, which constituted the intention-to-treat population. Baseline characteristics are presented in Table 1. Of all 65 participants, 39 (60%) were women, with a mean age of 40 years, a mean eGFR of 69 ml/min per 1.73 m² (SD=23), and a median 24-hour UPCR of 0.8 g/g and 24-hour urine protein 1.1 g/d (0.6–1.3). The mean medication adherence rate was 98% (SD=5%).

Figure 1.

Trial profile.

Table 1.

Demographic and baseline characteristics of the 65 randomized patients

Characteristics	Value
Age, yr, mean (SD)	40 (9)
Sex, male, n (%)	26 (40)
Time since kidney biopsy, mo, median (IQR)	44 (6–86)
BMI, kg/m2, mean (SD)	24.9 (3.3)
Systolic BP, mm Hg, mean (SD)	117 (10)
Diastolic BP, mm Hg, mean (SD)	76 (9)
eGFR, ml/min per 1.73 m2, mean (SD)	69 (23)
24-h UPCR, g/g, median (IQR)	0.8 (0.6–1.3)
24-h urine protein, g/24 h, median (IQR)	1.1 (0.8–1.6)
Concomitant medications, n (%)
Renin-angiotensin system inhibitors	64 (98)
Hydroxychloroquine	15 (23)
Mineralocorticoid receptor antagonist	15 (23)
Diuretic	13 (20)
Histology on kidney biopsy, No./total No. (%) a
Mesangial hypercellularity (M)
M0	17/53 (32)
M1	36/53 (68)
Segmental glomerulosclerosis (S)
S0	26/53 (49)
S1	27/53 (51)
Endocapillary hypercellularity (E)
E0	14/53 (26)
E1	39/53 (74)
Tubular atrophy or interstitial fibrosis (T)
T0	37/53 (70)
T1	14/53 (26)
T2	2/53 (4)
Crescents (C)
C0	16/46 (35)
C1	28/46 (61)
C2	2/46 (4)

BMI, body mass index; IQR, interquartile range; UPCR, urine protein-creatinine ratio.

^aThe M, E, S, and T scores were all not available for 12 patients, and the C score was not available for 19 patients.

Primary Outcome

No significant period or sequence effects were observed (P = 0.61 for both), validating the crossover design. Individual responses in 24-hour UPCR reduction varied substantially across treatments (Figure 2). There was no significant correlation between responses to ambrisentan and henagliflozin (r=−0.25; adjusted P = 0.14) nor between the combination therapy and either monotherapy (versus ambrisentan: r=0.13, adjusted P = 0.97; versus henagliflozin: r=0.04, adjusted P = 1.00). Among the six patients unresponsive to ambrisentan, four and five responded to henagliflozin and the combination, respectively. Conversely, of the 24 patients unresponsive to henagliflozin, 22 and 20 responded to ambrisentan and the combination, respectively.

Figure 2.

Primary outcome: correlations in percent 24-hour UPCR changes from baseline. (A) Correlation between ambrisentan and henagliflozin, (B) correlation between ambrisentan and ambrisentan-henagliflozin, and (C) correlation between henagliflozin and ambrisentan-henagliflozin. P values were adjusted for multiple comparisons using the Bonferroni method. UPCR, urine protein-creatinine ratio.

Secondary Outcomes

Baseline 24-hour UPCR was comparable before each treatment period. At week 4, the mean percentage change in 24-hour UPCR from baseline was −48% (95% CI, −56 to −39) with ambrisentan, −21% (95% CI, −33 to −6) with henagliflozin, and −44% (95% CI, −52 to −34) with the combination (Figure 3). The between-group difference in proteinuria reduction was NS for combination therapy versus ambrisentan alone (8%; 95% CI, −12 to 33) but was significant versus henagliflozin alone (−29%; 95% CI, −42 to −13). Upon treatment cessation, 24-hour UPCR rebounded to levels not significantly different from baseline across groups. A similar pattern was observed for 24-hour urine protein. At week 4, the ambrisentan-henagliflozin combination resulted in a mean change of −27% (95% CI, −39 to −13). The difference versus ambrisentan alone was 7% (95% CI, −14 to 32) and versus henagliflozin alone was −28% (95% CI, −42 to −11).

Figure 3.

Secondary outcome: changes in 24-hour UPCR during treatment with ambrisentan, henagliflozin, and ambrisentan-henagliflozin. The figure shows the percent change in 24-hour UPCR from baseline. The error bars indicate the 95% CI. CI, confidence interval.

At baseline, the mean eGFR was comparable among the three treatment groups, averaging around 69 ml/min per 1.73 m². After 4 weeks, eGFR change from baseline was −0.5 ml/min per 1.73 m² (95% CI, −2.8 to 1.9) with ambrisentan, −3.5 ml/min per 1.73 m² (95% CI, −5.8 to −1.2) with henagliflozin, and −4.7 ml/min per 1.73 m² (95% CI, −6.9 to −2.6) with the combination. The mean differences in eGFR change during combination therapy were −4.3 ml/min per 1.73 m² (95% CI, −7.0 to −1.5) versus ambrisentan and −1.2 ml/min per 1.73 m² (95% CI, −4.0 to 1.6) versus henagliflozin, respectively. eGFR increased 4 weeks after discontinuation of each treatment, returning toward baseline levels (Figure 4).

Figure 4.

Secondary outcome: changes in eGFR during treatment with ambrisentan, henagliflozin, and ambrisentan-henagliflozin. The figure shows the absolute changes from baseline in eGFR. The error bars indicate the 95% CI.

The mean systolic BP at baseline was 118 mm Hg (SD=9), 118 mm Hg (SD=10), and 116 mm Hg (SD=10) at the start of the ambrisentan, henagliflozin, and combination treatment periods, respectively. After 4 weeks, the mean changes in systolic BP from baseline were −1.9 mm Hg (95% CI, −4.5 to 0.8) for ambrisentan, −1.4 mm Hg (95% CI, −4.0 to 1.2) for henagliflozin, and −1.0 mm Hg (95% CI, −3.6 to 1.5) for the combination. The changes in systolic BP were modest and not significantly different among the three treatment groups.

Exploratory Outcomes

After 4 weeks of treatment, 41 (63%), 14 (22%), and 39 (60%) patients achieved a ≥30% reduction in 24-hour UPCR with ambrisentan, henagliflozin, and the combination therapy, respectively. The efficacy of the combination therapy was evaluated against each monotherapy. The OR for achieving this end point with combination therapy was 0.88 (95% CI, 0.43 to 1.78) versus ambrisentan and 5.46 (95% CI, 2.56 to 12.20) versus henagliflozin. For a ≥50% reduction, the numbers of patients were 19 (29%), three (5%), and 18 (28%), respectively. The OR for the combination was 0.93 (95% CI, 0.43 to 1.99) compared with ambrisentan and 7.83 (95% CI, 2.47 to 34.85) compared with henagliflozin (Figure 5). Complete proteinuria remission (24-hour urine protein <0.3 g) occurred in eight patients (12%) each with ambrisentan and combination therapy, compared with three patients (5%) with henagliflozin.

Figure 5.

Exploratory outcome: proportion of patients and OR for achieving a ≥30% and ≥50% reduction in 24-hour UPCR during treatment with ambrisentan, henagliflozin, and ambrisentan-henagliflozin. The bars show the proportion of patients achieving 30% or 50% 24-hour UPCR reduction from baseline. OR, odds ratio.

Subgroup analyses revealed consistent effects of all treatments on 24-hour UPCR across predefined subgroups, with no significant treatment-by-subgroup interactions (all P_interaction > 0.05; Supplemental Figure 1). Furthermore, a post hoc analysis confirmed that the treatment effect on 24-hour UPCR was maintained after adjustment for concurrent systolic BP changes. Specifically, the ambrisentan-henagliflozin combination reduced UPCR by −30% (95% CI, −43 to −14) versus henagliflozin alone, whereas no significant difference was observed versus ambrisentan alone (6%; 95% CI, −13 to 30).

Safety Outcomes

Body weight decreased with henagliflozin alone but increased with ambrisentan alone and combination therapy from comparable baselines (Figure 6). Compared with their respective baselines, weight increased by 0.6 kg (95% CI, 0.2 to 1.1) with ambrisentan, changed by −0.4 kg (95% CI, −0.8 to 0.1) with henagliflozin, and increased by 0.3 kg (95% CI, −0.1 to 0.7) with combination therapy. Although body weight after washout did not differ significantly from baseline in any group, the reduction observed with henagliflozin alone appeared more sustained than the transient increase seen with ambrisentan monotherapy. BNP analyses were limited by incomplete direct measurements and conversion from N-terminal pro-B-type natriuretic peptide (NT-proBNP) in some participants. Findings were directionally similar in sensitivity analyses restricted to participants with complete BNP data (Supplemental Figure 2 and Supplemental Section 2).

Figure 6.

Safety outcome: changes in body weight during treatment with ambrisentan, henagliflozin, and ambrisentan-henagliflozin. The figure shows the absolute changes from baseline in body weight. The error bars indicate the 95% CI.

Treatments had opposing effects on hemoglobin levels. Hemoglobin levels decreased with ambrisentan and increased with henagliflozin, with the combination yielding an intermediate effect. From comparable baseline levels (ambrisentan: 141.3 g/L, SD=19.0; henagliflozin: 141.3 g/L, SD=18.1; combination: 140.6 g/L, SD=18.2), the mean change from baseline was −9.5 g/L (95% CI, −12.1 to −6.9) with ambrisentan, 2.8 g/L (95% CI, 0.2 to 5.4) with henagliflozin, and −4.6 g/L (95% CI, −7.0 to −2.2) with the combination, which differed significantly from both monotherapies. Levels rebounded after treatment cessation, and no elevations of NT-proBNP >2000 pg/ml or BNP >300 pg/ml occurred.

Adverse events are summarized in Table 2. Among the 65 participants, adverse events occurred in 44 (68%) during ambrisentan, 25 (38%) during henagliflozin, and 36 (55%) during combination therapy. Adverse events of special interest include hypotension (nine [14%] with ambrisentan; four [6%] with henagliflozin; seven [11%] with combination therapy); fluid retention (three [5%] with ambrisentan; zero [0%] with henagliflozin; one [2%] with combination therapy); and one case of AKI during henagliflozin treatment, assessed as related to coronavirus disease 2019 infection. Notably, peripheral edema was most frequent with ambrisentan (11 [17%]), occurring at more than double the rate observed with combination therapy (five [8%]), and was infrequent with henagliflozin (two [3%]). Two participants withdrew consent: one due to elevated hepatic enzymes and the other due to increasing proteinuria. Four additional participants discontinued a treatment phase, with three during ambrisentan treatment due to fluid retention, hypotension, and headache, respectively, and one during combination therapy due to hypotension, and all subsequently entered the next planned phase and completed the trial.

Table 2.

Safety outcome: summary of adverse events

Variables	Ambrisentan (%)	Henagliflozin (%)	Ambrisentan-Henagliflozin (%)
AEs	44 (68)	25 (38)	36 (55)
Death	0 (0)	0 (0)	0 (0)
AEs leading to drug discontinuation
Fluid retentiona	1 (2)	0 (0)	0 (0)
Hypotension	1 (2)	0 (0)	1 (2)
Headache	1 (2)	0 (0)	0 (0)
Elevated hepatic enzyme	0 (0)	0 (0)	1 (2)
AEs of clinical interest
AKI	0 (0)	1 (2)	0 (0)
Hypotension	9 (14)	4 (6)	7 (11)
Fluid retentiona	3 (5)	0 (0)	1 (2)
Edema
Peripheral edemab	11 (17)	2 (3)	5 (8)
Fluid retentiona	3 (5)	0 (0)	1 (2)

Data are number of patients (%). AE, adverse event.

^aFluid retention was defined as edema accompanied by weight gain >2 kg, B-type natriuretic peptide >300 pg/ml, or heart failure symptoms.

^bPeripheral edema was defined as edema not meeting the criteria for fluid retention.

Discussion

In this randomized crossover trial, responses to the ERA ambrisentan and the SGLT2 inhibitor henagliflozin showed no significant correlation (r =−0.25, P = 0.14), indicating independent mechanisms, thereby supporting that patients insufficiently responsive to one agent may benefit from the other. In these high-risk patients with IgA nephropathy, both treatments reduced proteinuria, with combination therapy achieving reductions comparable with ERA monotherapy. SGLT2 inhibition attenuated the fluid-retaining properties of ERAs, reducing edema and weight gain.

The dissociation between proteinuria responses to ambrisentan and henagliflozin provides evidence for distinct mechanisms between ERAs and SGLT2 inhibitors, which is further underscored by their divergent effects on eGFR. Henagliflozin caused a prompt, reversible eGFR reduction, consistent with a hemodynamic effect from lowered intraglomerular pressure.^22,23 This process involves modulation of tubuloglomerular feedback, increasing sodium delivery to the macula densa and triggering afferent arteriolar vasoconstriction.²⁴ By contrast, ambrisentan reduced proteinuria substantially (−48%) with minimal eGFR change, aligning with other ERA trials,^17,25 indicating that its benefits are mediated largely through nonhemodynamic, tissue-protective pathways, including podocyte stabilization, inhibition of mesangial activation, anti-inflammatory, antifibrotic effects, and preservation of the glomerular endothelial glycocalyx.^26,27 These mechanistic insights support a “rotate-before-combine” strategy and underscores the need for biomarkers to guide personalized pathway-specific therapy in IgA nephropathy.

In this proof-of-concept trial, henagliflozin and ambrisentan were used as prototypical inhibitors of SGLT2 and endothelin pathways, respectively. The proteinuria reduction with henagliflozin in our trial, at –21%, is consistent with findings from other SGLT2 inhibitor trials. Dapagliflozin reduced proteinuria by 20% in the Rotation for Optimal Targeting of Albuminuria and Treatment Evaluation trial, and by 15% in the Effects of the SGLT2 Inhibitor Dapagliflozin on Proteinuria in Non-Diabetic Patients with Chronic Kidney Disease trial.^22,24 Empagliflozin achieved a 29% reduction in the Combination Effect of Finerenone and Empagliflozin in Participants with Chronic Kidney Disease and Type 2 Diabetes Using an Albumin-to-Creatinine Ratio End Point trial.²⁸ The 5 mg dose of ambrisentan was selected based on its prior efficacy in IgA nephropathy (41% proteinuria reduction in 147 patients) and its established safety in pulmonary arterial hypertension.¹⁴ In our trial, ambrisentan achieved a 48% reduction at 4 weeks, comparable with ERA agents developed for IgA nephropathy, including sparsentan (−50% in Efficacy and Safety of Sparsentan Versus Irbesartan in Patients with IgA Nephropathy) and atrasentan (−38% in ALIGN).^8,29 These results validate our selection of both drugs for probing their respective pathways.

Analyses of previous trials have indicated the potential for complementary effects between ERAs and SGLT2 inhibitors. Given the diuretic properties of SGLT2 inhibitors, their combination with ERAs represents a rational approach to mitigating fluid-related adverse effects. In the Atrasentan and Renal Events in Patients with Type 2 Diabetes and Chronic Kidney Disease trial post hoc analysis, adding an SGLT2 inhibitor to atrasentan in patients with type 2 diabetes and CKD reversed atrasentan-associated weight gain (1.2 kg between-group difference), further reduced albuminuria by 28%, and favorably lowered BNP.¹⁰ Similarly, in Zibotentan in Combination with Dapagliflozin Compared with Dapagliflozin in Patients with Chronic Kidney Disease, dapagliflozin plus low-dose zibotentan enhanced albuminuria reduction versus dapagliflozin alone, whereas the zibotentan-only group was discontinued due to excess fluid retention (47% versus 14% with dapagliflozin alone).¹² In a recent prospective IgA nephropathy trial, 61 (47%) participants were receiving SGLT2 inhibitors at baseline. The proteinuria-reducing effect of SC0062 and its impact on body weight and NT-proBNP were consistent regardless of baseline SGLT2 inhibitor use.¹⁷ In the ALIGN trial, IgA nephropathy patients already on stable SGLT2 inhibitors who received atrasentan add-on therapy achieved a 37% greater reduction in proteinuria after 36 weeks versus those receiving add-on placebo.⁸ However, previous studies were limited by the lack of a dedicated IgA nephropathy population, permitted baseline SGLT2 inhibitor use, and no direct ERA control. Our trial demonstrates that initial ERA and SGLT2 inhibitor combination therapy delivers the full efficacy of the ERA while mitigating its fluid-retaining effects, thereby enabling a more favorable therapeutic window in IgA nephropathy.

The lack of additional proteinuria reduction with combination therapy over ERAs alone could be interpreted as insufficient duration for SGLT2 inhibitor effects to fully manifest. This interpretation is consistent with the Combination Effect of Finerenone and Empagliflozin in Participants with Chronic Kidney Disease and Type 2 Diabetes Using an Albumin-to-Creatinine Ratio End Point trial, where the albuminuria-lowering effect of SGLT2 inhibitors progressed over 6 months,²⁸ and the Dapagliflozin and Prevention of Adverse Outcomes in Chronic Kidney Disease trial, which showed continued albuminuria reduction over 4 months in IgA nephropathy.⁹ Nevertheless, our study was adequately powered to detect the prespecified difference—a nearly 30% reduction in 24-hour UPCR with combination therapy versus henagliflozin alone—consistent with our initial sample size calculation.

The combination of ERAs and SGLT2 inhibitors exhibited a favorable safety profile. Compared with ERA monotherapy, combination therapy reduced the incidence of peripheral edema and fluid retention by over 50% and attenuated ERA-associated weight gain. We also observed a trend toward a sustained weight-reducing effect of SGLT2 inhibitors after treatment discontinuation. Taken together, these findings implicate body fat reduction beyond fluid loss in SGLT2 inhibitor–associated weight loss. As weight management is integral to lifestyle intervention in IgA nephropathy,^16,30 SGLT2 inhibitors may represent a more appropriate option than other diuretics for ERA combination therapy.

All three treatments showed a nonsignificant trend toward reducing BP. The proteinuria-lowering effects remained significant after systolic BP adjustment, supporting a renoprotective mechanism independent of BP reduction. Henagliflozin increased hemoglobin, likely due to plasma volume contraction and stimulation of hematopoiesis, consistent with other SGLT2 inhibitors.^31,32 Ambrisentan lowered hemoglobin, consistent with other ERAs, an effect likely attributable to hemodilution.^8,12 Coadministration produced an intermediate hemoglobin level, reflecting counterbalancing effects between the two drug classes.

This study has several limitations. First, the sample size was limited. Second, the primary outcome was a surrogate end point requiring further trials to confirm clinical efficacy and safety of combining ERAs with SGLT2 inhibitors. Third, the 4-week follow-up was insufficient to determine whether proteinuria reduction is sustained, although published evidence supports durability.^7,8,12,17,23 Longer parallel-group trials are needed to evaluate durability and potential late additive benefits of combination therapy. Fourth, the open-label design may have introduced potential bias in adverse event reporting. Fifth, natriuretic peptide assessment was not standardized, with NT-proBNP converted to BNP in some participants, potentially introducing measurement variability. Finally, ambrisentan was used here as a pharmacologic probe; its administration (including the 5 mg dose) does not constitute a clinical recommendation for IgA nephropathy, unlike ERAs specifically in development for this disease. Any clinical application would require validation in dedicated outcome trials.

In conclusion, both ERAs and SGLT2 inhibitors reduced proteinuria in patients with IgA nephropathy. The lack of correlation between their individual effects suggests a potential for therapy rotation. Although combining ambrisentan and henagliflozin did not significantly enhance proteinuria reduction relative to ambrisentan alone, it led to significant decreases in body weight and in the incidence of peripheral edema and fluid retention. These results support the combination therapy of ERAs and SGLT2 inhibitors as a promising strategy to delay IgA nephropathy progression, warranting validation of its kidney protection in larger, long-term prospective trials.

Supplementary Material

jasn-37-2016-s001.pdf ^{(1.4MB, pdf)}

jasn-37-2016-s002.pdf ^{(2.9MB, pdf)}

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/F801.

Author Contributions

Conceptualization: Pei Chen, Qinlan Chen, Yang Li, Lijun Liu, Jicheng Lv, Sufang Shi, Jinwei Wang, Hongyu Yang, Hong Zhang, Xujie Zhou, Sainan Zhu.

Data curation: Pei Chen, Qinlan Chen, Ling Guo, Jicheng Lv.

Formal analysis: Pei Chen, Qinlan Chen, Ling Guo, Jicheng Lv, Jinwei Wang.

Funding acquisition: Jicheng Lv, Hong Zhang.

Investigation: Pei Chen, Qinlan Chen, Yang Li, Lijun Liu, Jicheng Lv, Sufang Shi, Hongyu Yang, Xujie Zhou.

Methodology: Pei Chen, Qinlan Chen, Ling Guo, Yang Li, Lijun Liu, Jicheng Lv, Sufang Shi, Jinwei Wang, Hongyu Yang, Hong Zhang, Xujie Zhou, Sainan Zhu.

Project administration: Yunfei Bao, Yuting Cai, Pei Chen, Qinlan Chen, Rong He, Yang Li, Lijun Liu, Muqing Liu, Yanjie Liu, Jicheng Lv, Xiaoyuan Ning, Sufang Shi, Qian Wang, Hongyu Yang, Qiong Zhang, Xujie Zhou, Sainan Zhu.

Resources: Jicheng Lv, Hong Zhang.

Software: Pei Chen, Qinlan Chen, Jinwei Wang, Sainan Zhu.

Supervision: Yunfei Bao, Pei Chen, Qinlan Chen, Ling Guo, Yang Li, Lijun Liu, Jicheng Lv, Sufang Shi, Jinwei Wang, Hongyu Yang, Hong Zhang, Xujie Zhou.

Validation: Pei Chen, Qinlan Chen, Ling Guo, Lijun Liu, Jicheng Lv, Sufang Shi, Xujie Zhou.

Visualization: Pei Chen, Qinlan Chen, Ling Guo.

Writing – original draft: Pei Chen, Qinlan Chen.

Writing – review & editing: Ling Guo, Yang Li, Lijun Liu, Jicheng Lv, Sufang Shi, Jinwei Wang, Hongyu Yang, Hong Zhang, Xujie Zhou, Sainan Zhu.

Funding

This study was supported by National Natural Science Foundation of China (81925006), Capital’s Funds for Health Improvement and Research (2024-1-4073), and National Key Research and Development Program of China (2024YFC2511000).

Declarative Statements

This study includes clinical experimentation and received Institutional Review Board or Ethics Committee approval. All patients provided written informed consent. This study includes clinical experimentation and complies with the Declaration of Helsinki.

Data Availability Statements

Original data generated for the study will be made available upon reasonable request to the corresponding author. Data Type: Clinical Trial Data. Reason for Restricted Access: Due to local privacy regulations, the data cannot be deposited in a persistent repository. Access to the data underlying this manuscript is restricted and can only be made available via the data sharing processes upon agreement with the corresponding author, in compliance with relevant privacy requirements.

Supplemental Material

This article contains supplemental material online, published as provided by the authors, at http://links.lww.com/JSN/F802.

Supplemental Section 1. Randomization procedures and allocation concealment.

Supplemental Section 2. BNP analysis with NT-proBNP imputation and sensitivity results.

Supplemental Table 1. Detailed eligibility criteria.

Supplemental Figure 1. Exploratory outcome: Percent change from baseline in 24-hour UPCR after 4 weeks of treatment with ambrisentan, henagliflozin, and ambrisentan-henagliflozin in patient subgroups defined by baseline characteristics.

Supplemental Figure 2. Safety outcome: changes in BNP during treatment with ambrisentan, henagliflozin, and ambrisentan-henagliflozin.

Statistical Analysis Plan.