Abstract
Endothelin A and B receptors, together with sodium-glucose cotransporter-2 (SGLT-2) channels are important targets in improving endothelial function and intervention with inhibitors has been the subject of multiple mechanistic and clinical outcome trials over recent years. Notable successes include the treatment of pulmonary hypertension with endothelin receptor antagonists, and the treatment of heart failure and chronic kidney disease with SGLT-2 inhibitors. With distinct and complementary mechanisms, in this review, we explore the logic of combination therapy for a number of diseases which have endothelial dysfunction at their heart.
Introduction
Maintaining endothelial health is crucial to avoidance of cardiovascular, renal and hepatic events and endothelial dysfunction is an important target of pharmacological intervention. A number of therapeutic strategies target the renin–angiotensin–aldosterone system (RAAS), although until the discovery of endothelin-1 (ET-1) in 1987, potentially the most important pathway to target was neglected. ET-1 is a 21 amino acid potent vasoconstrictor produced predominantly by endothelial cells. In total three isoforms of endothelin have been identified, designated ET-1, ET-2 and ET-3 [1].
ET-1 exerts its effects through both the endothelin A (ETA) and endothelin B (ETB) G-protein coupled receptors. ETA receptors are primarily found on vascular smooth muscle cells (VSMSs), cardiac myocytes, hepatic stellate cells and the renal glomeruli and vasculature. ETA receptor agonism is thought to be the primary driver for vasoconstriction, cell growth, adhesion and migration, fibrosis, inflammation and the stimulation of aldosterone secretion [2]. ETB is predominantly expressed on endothelial cells and to a lesser extent on VSMCs. Although activation of ETB on VSMCs leads to vasoconstriction, activation on endothelial cells leads to transient vasodilation and growth inhibitory effects. In addition, the ETB receptor plays the major role in clearance of ET-1 by binding and internalising the peptide, primarily in the lungs [3].
ET-2 and ET-3 are beyond the scope of this review. ET-2 may play an important role in ovarian physiology, vascular inflammation, and in the pathogenesis of primary sclerosing cholangitis [4–6]. ET-3 has a role in the development of neural crest cells and related pathologies include enteric neuronal diseases such as Hirschprung’s disease [7].
In human-isolated vessels in vitro, ET-2 has a similar vasoconstrictor potency to ET-1 [8]. However, this vasoconstrictor action of ET-2 is blocked equally well by ETA antagonists, so even if there is an as yet unmeasured contribution of ET-2 to a particular pathological condition, it would also be blocked. It is also well established in human vessels that ET-3 is much less potent that ET-1. ET-3 does not activate ETA receptors at physiological or pathophysiological concentrations found in the plasma [8].
It might seem surprising, but up to now clinical trials of endothelin antagonists have concentrated on orphan indications, with initial marketing authorisations granted in the pulmonary arterial hypertension space [9], especially given evidence that elevated ET-1 as a hormone initiating, driving and worsening a host of diseases (Figure 1). Most recently sparsentan (DUPLEX; NCT03493685 and PROTECT; NCT03762850) and atrasentan (ALIGN; NCT04573478) have explored the efficacy of endothelin antagonism in the treatment of rare kidney diseases [10–12]. Others have been trialed in the treatment of subarachnoid induced vasospasm (clazosentan) [13–15], and recently approved for the treatment of a subset of patients with hypertension in the US (aprocicentan) [16]
Endothelin-1 overabundance and signaling through endothelin receptors is implicated in a multitude of vascular diseases across major organs
At least some of the reticence with respect to wider use of these agents has been their adverse event profile. These adverse events has included the need for close monitoring of liver safety (e.g., bosentan and sparsentan) [17,18], and signs of sodium and water retention caused by increased vascular permeability and vasodilation, which is well described in conjunction with endothelin antagonism. Most recently both atrasentan and zibotentan have demonstrated reassuring liver safety data across diverse populations [19,20]. Both atrasentan and zibotentan are selective ETA receptor antagonists, with zibotentan being the most selective [21].
SGLT-2 inhibitors, another class of agents which exert potent effects in restoring endothelial function have demonstrated cardio- and reno-protective properties and have broad marketing authorisations for the treatment of Type 2 diabetes, chronic kidney disease and heart failure (irrespective of ejection fraction).
Considerable recent interest with respect to drug development has centred on whether combination of SGLT-2 inhibitors with ETA receptor antagonists may be additive with respect to positive effects on endothelial function and whether SGLT-2 inhibitors may successfully modulate sodium and water retention associated with endothelin antagonist therapy.
This mini-review examines evidence to support combination therapy with endothelin antagonism and SGLT-2 inhibition across a range of conditions both respect to efficacy and modulation of endothelin antagonist related adverse effects.
Chronic kidney disease
The role which ET-1 plays in the pathophysiology and progression of chronic kidney disease (CKD) is established in animal models although translatability to humans is the subject of debate. On top of the well described effects of ET-1 in increasing blood pressure, it has a key role in driving cell proliferation, podocyte dysfunction, inflammation and fibrosis, and these effects may play as important a role in driving CKD progression [22].
Both atrasentan and sparsentan have shown benefit as monotherapy in delaying CKD progression. Atrasentan is an ETA receptor antagonist thought to be selective for ETA receptors at clinically relevant concentrations [8], sparsentan is a dual ETA and angiotensin II receptor antagonist [23]. Atrasentan was first studied in patients with diabetic nephropathy in the RADAR (NCT01356849) and SONAR (NCT01858532) clinical trials. These studies demonstrated a robust 33.6% decrease in urine albumin-creatinine ratio (UACR) versus placebo, and SONAR demonstrated a 35% reduction in the composite renal outcome [24]. Earlier and more robust UACR response has been shown in the SONAR study to be associated with better renal outcomes, which suggests that the potent effects of endothelin antagonism in reducing glomerular pressure and inflammation are important for beneficial outcomes [25]. Sparsentan treatment of IgA nephropathy over 110 weeks resulted in a 40% reduction in UACR and a difference in eGFR decline as measured by chronic slope of 1.1 ml/min per 1.73 m2 per year versus the ACE inhibitor irbesartan [11]. The DUPLEX study examined the effect of sparsentan versus irbesartan in the treatment of focal segmental glomerulosclerosis (FSGS) over 108 weeks, partial remission rates were 42% for sparsentan and 26% in the irbesartan group. Unfortunately, data on GFR slope reported by Rheault et al. did not report a significant effect on GFR slope overall where no significant difference was seen in either total or chronic slope for sparsentan versus irbesartan [10]. Efficacy of sparsentan in FSGS in combination with an SGLT-2 inhibitor is currently unclear [26]. A small study, called SPARTACUS (NCT05856760), is currently ongoing and will evaluate the effect of sparsentan in combination with an SGLT-2 inhibitor in the FSGS population [27].
Despite showing an impact on GFR progression in the treatment of CKD, both atrasentan and sparsentan are associated with sodium and water retention. Atrasentan in the SONAR study was associated with a 4.3% excess of fluid related events versus placebo, and a small excess of heart failure events (1% excess of hospital admissions for heart failure) in spite of excluding participants who experienced a substantial fluid retention during an enrichment period before randomisation [24]. In the sparsentan IgA nephropathy study (PROTECT), a 3% excess of peripheral oedema was seen for sparsentan versus irbesartan [11]. Early fluid retention may at least in part be related to increased vascular permeability [28] and dilatation of large lower limb veins, which are known to be rich in ETA receptors, leading to a compensatory rise in sodium and water retention [29]. Diuretic therapy is recognised to only partially deal with this, at the potential expense of prerenal impairment.
Intriguingly, the SONAR study showed a lower number of events of non-fatal stroke in favour of atrasentan, although it was a low risk population for cerebrovascular disease (0.6% for atrasentan and 2% for the comparator group) [24]. These events likely to be predominantly driven by an effect on microvascular, (lacunar) infarcts, and may be confirmed by greater exposure to endothelin antagonists in at risk populations.
Running almost parallel in time to phase 3 ETA receptor antagonist CKD trials were trials investigating the SGLT-2 inhibitors, canagliflozin (CREDENCE; NCT02065791) [30], dapagliflozin (DAPA-CKD; NCT03036150) [31] and empagliflozin (EMPA-KIDNEY; NCT03594110) [32] for the same indication. These trials over a follow-up period of more than 2 years have demonstrated between a 28 and 39% reduction in events comprising a combined renal endpoint [33–35]. Although SGLT-2 inhibitors exert their primary effects on sodium and glucose reabsorption within the proximal tubule of the kidney, they are associated with notable improvement in endothelial function which also potentially translates into a positive impact on heart failure events and cardiovascular death [36–40]. It is also postulated that SGLT-2 inhibition may represent a better option for combination with respect to managing fluid overload. This hypothesis was supported by a publication by Ohara et al. which suggested that a greater diuretic response to dapagliflozin is seen in individuals with greater levels of extracellular water, leading to a normalisation of water balance [41]. A preclinical study by AstraZeneca demonstrated the combination potential of the two molecules zibotentan and dapagliflozin to achieve balanced fluid status versus zibotentan alone [42]. A small post-hoc analysis of 14 patients in SONAR showed the combination of atrasentan and a SGLT-2 inhibitor achieved a beneficial weight difference versus atrasentan alone of 1.2 kg, and a 27.6% incremental benefit in reducing UACR [43].
In 2023, Heerspink et al. published in the first data in the Lancet from a CKD trial which randomised patients to combination therapy with zibotentan and dapagliflozin versus dapagliflozin alone for 12 weeks (ZENITH-CKD; NCT04724837) [44]. This trial demonstrated a 27% benefit on urinary albumin excretion for the combination using the 0.25 mg dose of zibotentan versus dapagliflozin alone, without any significant increase in fluid retention at this dose [45], strengthening the proposed combination mechanism of action in CKD (Figure 2). This successful phase 2b trial has paved the way for the ZENITH-High Proteinuria phase 3 study in 1500 patients with CKD and elevated urinary albumin excretion which is now underway (NCT06087835).
Combining the benefits of the SGLT-2 inhibitor dapagliflozin with the endothelin receptor a antagonist zibotentan delivers additive therapeutic efficacy with an acceptable safety profile
Image constructed from data produced by ZENITH-CKD, DAPA-CKD and DAPAMECH Trials.
Image constructed from data produced by ZENITH-CKD, DAPA-CKD and DAPAMECH Trials.
Liver cirrhosis
Liver cirrhosis occurs as a consequence of chronic liver inflammation and characterised by severe fibrosis. Liver cirrhosis is progressive and results in increased resistance to portal blood inflow. The increased vascular resistance is because of deranged hepatic architecture caused by the fibrosis and increased hepatic vasoconstriction. Subsequently, splanchnic vasodilation causing RAAS activation and sodium and water retention result in increased blood flow to the liver which further increase the portal pressure. Portal blood pressure is complex to measure directly, hence an indirect measure of portal pressure, the hepatic venous pressure gradient, or HVPG has become the gold-standard surrogate for measurement of portal pressure.
Circulating endothelin levels and splanchnic endothelin production in humans are higher in liver cirrhosis and levels are associated severity of cirrhosis and elevated HVPG [46,47]. In humans, ETA and ETB receptors are present on hepatic stellate cells (HSCs) and hepatocytes, while ETB receptors are present on liver sinusoidal endothelial cells (LSECs) and Kupffer cells. ET-1 is thought to be synthesised by activated LSECs in the cirrhotic liver to act on activated HSCs in a paracrine fashion to cause sinusoidal constriction, which would be blocked by ETA selective antagonists [48]. A clinical study has shown that a single dose of ambrisentan, an ETA receptor antagonist, significantly reduced HVPG in patients with established clinically significant portal hypertension (CSPH) [49]. Thus, there are reasons to believe that ETA receptor inhibition would reduce the increased intrahepatic vasoconstriction associated with CSPH. Moreover, ETA receptor antagonism has been demonstrated in animal experiments to result in anti-fibrotic effects [50,51], which may play a role for long-term treatment of cirrhosis.
SGLT-2 inhibitors block glucose and sodium reabsorption in the proximal tubule of the kidney. This results in osmotic diuresis counteracting the effects of RAAS activation and leading to normalisation of the extracellular fluid volume [52,53].
This mode of action has the potential to be a useful therapy to reduce cirrhotic ascites. Unlike standard diuretic therapies, an SGLT-2 inhibitor is both less likely to exacerbate pre-existing hyponatremia or provoke worsening renal function as indicated in several exploratory studies [54–57].
In addition, there is evidence that SGLT-2 inhibitors could reduce progression of liver fibrosis, making them a potential ideal combination with ETA-receptor antagonists for therapeutic benefit in cirrhosis [58,59], whilst balancing ET-1 associated risks of sodium and water retention [60].
In light of the strong scientific rationale the results of the ZEAL study (NCT05516498) are eagerly awaited. The 2-part ZEAL study is designed to evaluate a range of doses of zibotentan in combination with 10 mg of dapagliflozin in patients who have compensated and decompensated cirrhosis up to Child Pugh B classification. The treatment period is 6–16 weeks, and the primary endpoint is HVPG, which is the surrogate with the closest links to outcomes in cirrhosis [61].
Microvascular angina
Fewer than half of patients with symptoms of angina have evidence of obstructive coronary artery disease demonstrated on angiography. A substantial proportion of these patients with ‘normal’ coronary arteries have objective evidence of ischaemia, a group now termed INOCA patients (Ischaemia with Normal Coronary Arteries). The majority of INOCA patients have either microvascular angina or coronary artery vasospasm [62].
Microvascular remodelling with loss of capillaries is heavily implicated in the pathogenesis of microvascular angina. Functional microvascular angina can also be demonstrated in response to physiological and pharmacological stressors [63].
Microvascular angina is both associated with increased levels of ET-1 and prolonged exposure to elevated ET-1 leads to deleterious effects on small vessel function [64].
Against this background attempts have been made to study the effects of ETA receptor antagonists in the microvascular angina population. These studies have however been hindered by limited doses of the agents being available, (namely higher doses used in previous cancer programs). One 72 patient trial of atrasentan 10 mg versus placebo showed positive effects on cardiovascular risk factors including blood pressure, lipids and glucose. However, two-thirds of the patient population randomised to atrasentan suffered significant oedema [65]. Another trial examined the effects of darusentan on myocardial perfusion at rest as measured by PET. It showed that darusentan significantly increase myocardial blood flow [66]. A very small study including 18 patients investigating zibotentan 10 mg in coronary slow flow phenomenon, a condition related to microvascular dysfunction, showed a positive impact on angina frequency, but severity of episodes and quality of life were not changed [67]. The latest trial of zibotentan 10 mg in microvascular angina, called PRIZE (NCT04508998), was designed to test two hypotheses. Firstly, did zibotentan 10 mg impact on exercise time in microvascular angina, and secondly did the presence of the PHACTR1 minor G allele SNP impact on the effect of the intervention [68]. The rationale for the PRIZE study is that patients with coronary microvascular angina carrying the G allele of rs9349379 is associated with higher circulating ET-1 levels. Patients homozygous for the GG genotype had the highest risk perfusion defects on stress perfusion, microvascular dysfunction on invasive testing, cardiac magnetic resonance imaging and reduced exercise tolerance due to angina [69]. The PRIZE study did not show an effect of zibotentan on exercise tolerance but did confirm the earlier findings of endothelin antagonist trials which showed an impact on resting myocardial blood flow and cardiovascular risk markers.
Several studies have investigated the effects of dapagliflozin or empagliflozin on myocardial blood flow. Empagliflozin has been shown to inhibit coronary microvascular dysfunction and reduce pericyte loss in db/db mice [70], and a study of coronary microvascular dysfunction in patients with Type 2 diabetes has shown a positive impact on cardiovascular risk markers although it did not significantly change measures of coronary flow reserve [71]. Data on dapagliflozin were more positive, with the DAPAHEART trial demonstrating a significant impact on myocardial flow reserve in 16 patients with Type 2 diabetes [72].
These data raise the possibility of a potentially additive effects of the combination in the treatment of microvascular angina, opening the door to a lower dose of endothelin antagonist in combination with SGLT-2 inhibitor to enhance efficacy and minimise the adverse effects of fluid retention, potentially opening the door to an effective therapy in this underserved population.
COVID-19
The logic of endothelin antagonist and SGLT-2 inhibitor combination in the treatment of COVID-19 is well described in a review from 2021 by Fisk et al [73]. ET-1 is a potent vasoconstrictor and pro-inflammatory, and is significantly up-regulated in acute respiratory distress syndrome (including in COVID-19). ET-1 plasma levels increased significantly during the acute phase in patients hospitalised with COVID-19 including those who died and those developing acute myocardial or kidney injury [74]. The increased ET-1 levels is likely to be the result of ET-1 released by endothelial injury as multiple biomarkers of endothelial injury are increased in patients with COVID-19 [75]. High ET-1 levels are thought to be a strong driver for pulmonary inflammation and also to drive deleterious changes in pulmonary haemodynamics seen with significant respiratory infection. As discussed earlier, given that ETA receptors mediate vasoconstriction in response to ET-1, and ETB receptors are responsible for ET-1 clearance, a selective ETA antagonist may have significant advantages in this population with respect to improving respiratory and cardiovascular outcomes.
SGLT-2 inhibitors, both empagliflozin and dapagliflozin, have been trialed in the treatment of acute COVID-19 [76]. DARE-19 (NCT04350593), the trial comparing dapagliflozin 10 mg to standard of care alone achieved a relative risk reduction of 20% in a combined endpoint of worsening organ dysfunction or death, although this did not reach statistical significance. Empagliflozin in the RECOVERY trial (NCT04381936) showed no impact on 28 day mortality. These trials were conducted at different timepoints in the COVID-19 pandemic, the DARE-19 study occurring first. At the time DARE-19 was conducted corticosteroids and anticoagulation had become a part of routine practice. At the time the RECOVERY trial was conducted with empagliflozin, vaccination and antibodies targeting the spike protein of the virus were used. It’s clear that both drugs were well tolerated, although neither trial population was apparently large enough to give conclusive evidence for an effect on outcomes [77].
Combination therapy utilising an ETA receptor antagonist and an SGLT-2 inhibitor may offer advantages with respect to prevention of fluid related events linked to endothelin antagonism, and potential additive effects in protecting the microvasculature. The TACTIC-E study (NCT04393246) utilising ambrisentan 5 mg and dapagliflozin 10 mg explores this hypothesis and publication is awaited [78].
Studies in long-COVID have demonstrated persistent immune activation, elevated ET-1 and long-term defects in lung perfusion in patients who have significantly reduced exercise tolerance and symptoms of breathlessness [79,80]. These studies raise the possibility of endothelin antagonism in combination with SGLT-2 inhibitors for organ protection and reducing the risk for fluid retention as an attractive combination treatment of long-COVID.
Resistant hypertension
Multiple endothelin receptor antagonists have been explored as treatments for systemic hypertension. Aprocitentan is the latest of these agents to be explored. Patients could enter the study if their sitting systolic blood pressure was greater than 140 mmHg, despite stable dosing with three or more oral agents [81]. High dose aprocitentan resulted in a reduction in 24 h systolic blood pressure of 5.9 mmHg. This was however at the expense of mild-to-moderate oedema or fluid retention among 18% of the participants, likely limiting its utility as a treatment for hypertension.
The anti-hypertensive effect of ET-A receptor blockade is caused by both vasodilation and reduced production of aldosterone by the adrenal cortex [82].
The mechanism by which SGLT-2 inhibitors regulate blood pressure is unclear, although they don’t significantly impact on aldosterone levels. Over the duration of the DAPA-CKD study this translated to a reduction in systolic blood pressure of 2.9 mmHg [83]. Combination therapy has been recently tested in the ZENITH-CKD study. After 12 weeks therapy with zibotentan 1.5 mg and dapagliflozin 10 mg, a reduction in office based systolic blood pressure of 7.6 mmHg was seen vs dapagliflozin alone, compared with 3.6 mmHg in the 0.25 mg zibotentan combination group. With respect to adverse events of special interest, 7% of patients in the high dose combination arm had fluid retention or peripheral oedema AESIs versus 5% in the low dose group. These data raise the possibility that the combination may be at least as effective in the treatment of hypertension versus aprocitentan, with much lower rates of oedema.
Conclusion
The focus on the benefits of endothelin antagonist therapy was targeted towards pulmonary hypertension, given the fluid retaining adverse effects of therapy and potential hepatic side-effects of early compounds at the doses deployed. We’re entering a new era as studies with atrasentan and zibotentan confirm the benefit of endothelin-A antagonism on endothelial function without any evidence of liver harm. The prospect of combination with SGLT-2 inhibition raises the possibility of unlocking further benefit through a complementary mode of action (Figure 3) in improving endothelial function and balancing sodium and water retaining effects of endothelin antagonist therapy.
Effects of endothelin antagonism and SGLT-2 inhibition on blood vessels
cyclic GMP, cyclic guanosine monophosphate; eNOS, endothelial nitric oxide synthase; ET-1, endothelin-1; ETA, Endothelin A receptors; ETB, Endothelin B receptors; mTOR, mammalian target of rapamycin; NO: nitric oxide; SGLT-2i, sodium-glucose cotransporter-2 inhibitor.
cyclic GMP, cyclic guanosine monophosphate; eNOS, endothelial nitric oxide synthase; ET-1, endothelin-1; ETA, Endothelin A receptors; ETB, Endothelin B receptors; mTOR, mammalian target of rapamycin; NO: nitric oxide; SGLT-2i, sodium-glucose cotransporter-2 inhibitor.
Data Availability
This is a review article, the data availability statement is therefore not applicable
Competing Interests
P.A., P.J.G, R.I.M, L.B. and J.O. are AstraZeneca employees. S.K. has no competing interests. A.P.D. is a member of the scientific advisory boards of Janssen, holds a research grant from AstraZeneca, ENB Therapeutics and Pharmaz.
Funding
This work was supported by AstraZeneca; S.K. is funded by UKRI-MRC Secondment Award [grant number MR/W003538/1].
CRediT Author Contribution
Phil Ambery: Conceptualization, Data curation, Writing—original draft, Writing—review & editing. Peter J. Greasley: Writing—original draft, Writing—review & editing. Robert I. Menzies: Writing—original draft, Writing—review & editing. Lena Brynne: Writing—original draft, Project administration, Writing—review & editing. Spoorthy Kulkarni: Writing—original draft, Writing—review & editing. Jan Oscarsson: Writing—original draft, Writing—review & editing. Anthony P. Davenport: Writing—original draft, Writing—review & editing.
Abbreviations
- ACE
angiotensin-converting enzyme
- AESI
adverse events of special interest
- CKD
chronic kidney disease
- COVID-19
Coronavirus disease (SARS-CoV-2 virus)
- CSPH
clinically significant portal hypertension
- cyclic GMP
cyclic guanosine monophosphate
- db/db
mutation of the diabetes (db) gene encoding for the ObR
- eGFR
estimated glomerular filtration rate
- eNOS
endothelial nitric oxide synthase
- ET-1
endothelin-1
- ET-2
endothelin-2
- ET-3
endothelin-3
- ETA
endothelin A
- ETB
endothelin B
- FSGS
focal segmental glomerulosclerosis
- GFR
glomerular filtration rate
- HSC
hepatic stellate cell
- HVPG
hepatic venous pressure gradient
- IgA
immunoglobulin A
- INOCA
Ischaemia with Normal Coronary Arteries
- LSEC
liver sinusoidal endothelial cell
- mTOR
mammalian target of rapamycin
- NO
nitric oxide
- PET
positron emission tomography
- PHACTR1
phosphatase and actin regulator 1 gene
- RAAS
renin–angiotensin–aldosterone system
- SGLT-2
sodium-glucose cotransporter-2
- SGLT2i
sodium-glucose cotransporter-2 inhibitor
- SNP
single nucleotide polymorphism
- UACR
urine albumin-creatinine ratio
- VSMC
vascular smooth muscle cell