Serelaxin as a potential treatment for renal dysfunction in cirrhosis: Preclinical evaluation and results of a randomized phase 2 trial

Abstract

Background: Chronic liver scarring from any cause leads to cirrhosis, portal hypertension, and a progressive decline in renal blood flow and renal function. Extreme renal vasoconstriction characterizes hepatorenal syndrome, a functional and potentially reversible form of acute kidney injury in patients with advanced cirrhosis, but current therapy with systemic vasoconstrictors is ineffective in a substantial proportion of patients and is limited by ischemic adverse events. Serelaxin (recombinant human relaxin-2) is a peptide molecule with anti-fibrotic and vasoprotective properties that binds to relaxin family peptide receptor-1 (RXFP1) and has been shown to increase renal perfusion in healthy human volunteers. We hypothesized that serelaxin could ameliorate renal vasoconstriction and renal dysfunction in patients with cirrhosis and portal hypertension.

Methods and findings: To establish preclinical proof of concept, we developed two independent rat models of cirrhosis that were characterized by progressive reduction in renal blood flow and glomerular filtration rate and showed evidence of renal endothelial dysfunction. We then set out to further explore and validate our hypothesis in a phase 2 randomized open-label parallel-group study in male and female patients with alcohol-related cirrhosis and portal hypertension. Forty patients were randomized 1:1 to treatment with serelaxin intravenous (i.v.) infusion (for 60 min at 80 μg/kg/d and then 60 min at 30 μg/kg/d) or terlipressin (single 2-mg i.v. bolus), and the regional hemodynamic effects were quantified by phase contrast magnetic resonance angiography at baseline and after 120 min. The primary endpoint was the change from baseline in total renal artery blood flow. Therapeutic targeting of renal vasoconstriction with serelaxin in the rat models increased kidney perfusion, oxygenation, and function through reduction in renal vascular resistance, reversal of endothelial dysfunction, and increased activation of the AKT/eNOS/NO signaling pathway in the kidney. In the randomized clinical study, infusion of serelaxin for 120 min increased total renal arterial blood flow by 65% (95% CI 40%, 95%; p < 0.001) from baseline. Administration of serelaxin was safe and well tolerated, with no detrimental effect on systemic blood pressure or hepatic perfusion. The clinical study's main limitations were the relatively small sample size and stable, well-compensated population.

Conclusions: Our mechanistic findings in rat models and exploratory study in human cirrhosis suggest the therapeutic potential of selective renal vasodilation using serelaxin as a new treatment for renal dysfunction in cirrhosis, although further validation in patients with more advanced cirrhosis and renal dysfunction is required.

Trial registration: ClinicalTrials.gov NCT01640964.

Conflict of interest statement

PCH has received consultancy fees from Novartis and GORE. JAF has received consultancy fees from Merck and Novartis and research grant funding from GlaxoSmithKline. TS, JP, DY and ND are currently employees of Novartis and own either Novartis stocks or shares. TS, RS, JP, DY, and ND contributed to the Phase 2 trial of Serelaxin as Novartis associates but played no role in the independent preclinical research. RS is currently an employee of GlaxoSmithKline. DJW is a member of a Data Safety Monitoring Board for AbbVie for a phase 3 trial with atrasentan in renal disease (institutional) and a non-executive Board member of the Medicines and Healthcare products Regulatory Agency. DJW's research is supported by BHF, MRC and Wellcome Trust. JSD owns Biogen, Vertex, Regulus and BMS stock. JSD has patents for therapies to treat acute kidney injury.

Figures

Fig 1. Rat models of advanced cirrhosis, portal hypertension, and renal dysfunction.

Portal pressure (PP; A), renal blood flow (RBF; B), and glomerular filtration rate (GFR; C) in 16-wk CCl4 and olive oil (OO) control rats (n = 6–11). Representative H&E-stained kidney (scale bar 50 μm) showing minor tubular epithelial cell vacuolation (arrows) without significant necrosis after 16 wk of CCl4 (D). PP (E), RBF (F), and GFR (G) in bile duct ligation (BDL) and sham-operated (sham) control rats (n = 4–8). Representative H&E-stained kidney (scale bar 50 μm) showing necrotic cells within the tubule lumen and loss of the normal circumferential epithelial cell population (arrows) 4 wk after BDL (H). Data presented as mean ± standard error of the mean, analyzed by one-way ANOVA with post hoc Bonferroni correction (*p < 0.05; **p < 0.01; ***p < 0.001).

Fig 2. Vascular responses in isolated extrarenal renal arteries from cirrhotic rats.

Concentration–response curves to acetylcholine (ACh; A and E), phenylephrine (PE; B and F), and sodium nitroprusside (SNP; C and G) in isolated extrarenal renal arteries from 16-wk CCl4 and 4-wk bile duct ligation (BDL) rats (n = 8–12). Data presented as mean ± standard error of the mean (SEM), relative to preconstricted value, analyzed by two-way ANOVA with post hoc Bonferroni correction (*p < 0.05; ***p < 0.001; NS, not significant). Nitric oxide synthase (NOS) activity in whole kidney extracts from 16-wk CCl4 (D) and 4-wk BDL (H) rats (n = 6–7). Data presented as mean ± SEM, analyzed by unpaired t-test (*p < 0.05). OO, olive oil.

Fig 3. Effect of 72-h serelaxin infusion on renal perfusion and renovascular responses in cirrhotic rats.

Relative Rxfp1 transcripts (normalized to 18S rRNA) in whole kidney extracts from 16-wk CCl4 (A) and 4-wk bile duct ligation (BDL) (E) rats (n = 3–6). Renal blood flow (RBF; B and F), mean arterial pressure (MAP; D and H), and glomerular filtration rate (GFR; C and G) in CCl4 and BDL rats after 72-h s.c. serelaxin or vehicle (n = 5–8). Data presented as mean ± standard error of the mean (SEM), analyzed by unpaired t-test (*p < 0.05; **p < 0.01; ***p < 0.001; NS, not significant). Concentration–response curves to acetylcholine (ACh; I), phenylephrine (PE; J), and sodium nitroprusside (SNP; K) in the presence of serelaxin or vehicle in 16-wk CCl4 rats (n = 5–8). Data presented as mean ± SEM, analyzed by two-way ANOVA with post hoc Bonferroni correction (*p < 0.05; **p < 0.01; ***p < 0.001). OO, olive oil.

Fig 4. Effect of sustained serelaxin infusion on AKT/eNOS/NO signaling in cirrhotic rat kidney.

Quantification of p-eNOS/eNOS and p-AKT/AKT in whole kidney extracts from 72-h serelaxin- or vehicle-treated 16-wk CCl4 and 4-wk bile duct ligation (BDL) rats (n = 4) (A–F). Data presented as mean ± standard error of the mean (SEM), analyzed by unpaired t-test (*p < 0.05; **p < 0.01; ***p < 0.001). NOS activity in whole kidney extracts from CCl4 (G) and BDL (H) rats treated with serelaxin or vehicle (n = 6–8). Data presented as mean ± SEM, analyzed by unpaired t-test (*p < 0.05; **p < 0.01). Renal blood flow (RBF; I) and glomerular filtration rate (GFR; J) in 16-wk CCl4 rats co-treated with L-NG-nitroarginine methyl ester (L-NAME) (n = 4–8). Data presented as mean ± SEM, analyzed by one-way ANOVA with post hoc Bonferroni correction (*p < 0.05; **p < 0.01; ***p < 0.001). OO, olive oil.

Fig 5. Effect of acute serelaxin treatment on renal blood flow and tissue oxygenation in CCl4 cirrhotic rats.

Renal blood flow (RBF, A) and mean arterial pressure (MAP, B) responses to acute i.v. serelaxin (4 μg) or vehicle in 16-wk CCl4 rats (n = 5–7). Measurement of velocity time integral (C) and renal resistive index (D) following acute i.v. serelaxin (4 μg) or vehicle (n = 6–8). Deoxygenated hemoglobin levels (R2*) in renal medulla in 8-wk (E) and 16-wk (F) CCl4 rats at baseline, 30 min, and 60 min following acute i.v. serelaxin (4 μg) or vehicle (n = 5–8). Data presented as mean ± standard error of the mean, analyzed by two-way ANOVA (*p < 0.05; **p < 0.01; ***p < 0.001; NS, not significant) with post hoc Bonferroni correction to compare individual CCl4 time points with respective vehicle controls (#p < 0.05; ##p < 0.01; ###p < 0.001).

Fig 6. Flow chart of participants in the study.

Fig 7. Differential effects of serelaxin and terlipressin treatment on splanchnic and extra-splanchnic hemodynamics.

Blood flow changes in total renal artery (left + right renal artery; A), superior abdominal aorta (B), superior mesenteric artery (C), portal vein (D), and hepatic artery (E); mean arterial pressure (MAP; F); and renal vascular resistance (G) following serelaxin infusion (60 min at 80 μg/kg/d then 60 min at 30 μg/kg/d) or terlipressin (2-mg i.v. bolus). Data presented as geometric mean ± 95% CI, analyzed by paired t-tests for within-group comparison of baseline and post-treatment flow measurements (n = 20; *p < 0.05; ***p < 0.001; NS, not significant).

Fig 8. Effect of serelaxin infusion on pharmacokinetics and plasma biomarkers in patients with cirrhosis and portal hypertension.

Serum serelaxin concentration measured by ELISA pre-dose (0 min), at 60 min and 120 min post-initiation of serelaxin infusion, in recovery period (~60 min after cessation of serelaxin), and at 4-wk follow-up visit (A). Data presented as mean ± standard deviation (n = 20). Plasma nitrate (B), endothelin-1 (ET-1) (C), and matrix metalloproteinase-9 (MMP-9) (D) measured by ELISA pre-dose (0 min) and at 120 min after initiation of serelaxin infusion. Data presented as geometric mean ± 95% CI (n = 20). NS, not significant.