Introduction

Anemia of chronic kidney disease (CKD) is common as a result of relative erythropoietin (EPO) deficiency and inadequate iron supply for effective erythropoiesis [1]. Anemia in the CKD population is associated with diminished quality of life, cognitive impairment [2], cardiovascular morbidity, hospitalizations [3], and mortality [4]. As the prevalence of end-stage renal disease continues to rise, the management of anemia represents a growing economic burden. The introduction of the bundled payment system for dialysis since January 2011, as well as, the lowering of hemoglobin targets effective June 2011, have encouraged efficient use of erythropoiesis-stimulating agents (ESAs) and iron with a consequent increase in ferritin levels [5].

Anemia Management Using Erythropoiesis-Stimulating Agents

ESAs have revolutionized the management of anemia since their availability in 1989. The need for blood transfusions has been reduced and ESAs have improved the quality of life in chronic kidney disease [6]. However, cardiovascular safety concerns have limited ESA use and lowered hemoglobin targets to approximately 11 g/dL with consequent increase in blood transfusions [7]. High doses (average >10,095 units/week) of recombinant erythropoietin (rhEPO) have been associated with an increased risk of cardiovascular events, regardless of the hemoglobin level achieved within the first 4 months of treatment [8]. The adverse cardiovascular outcomes may be associated with the magnitude of the ESA dose, as well as, comorbid conditions associated with ESA hyporesponsiveness.

An additional recognized burden of ESA use is hypertension. Compared to placebo, ESA treatment doubles the relative risk for hypertensive adverse events in predialysis and hemodialysis patients (unadjusted percentages are 13.6 % for placebo and 30.3 % for ESA) [6, 913]. Although no large randomized controlled trials have been conducted to investigate the impact of ESAs on blood pressure as a primary end point, an increase in blood pressure or antihypertensive medications use has been observed in the majority of uncontrolled studies [1424]. Meta-analyses have also demonstrated a link between ESA treatment and hypertension [2529]. A statistically significant difference in hypertensive adverse events in the low versus high hemoglobin target ESA treatment arms was evident in the meta-analysis by Krapf and Hulter [23]. The ESA-associated hypertension appears to involve endothelium-dependent and endothelium-independent vasodilatory impairment [30], rather than a simple erythropoiesis-mediated effect on blood rheology consequent to increased erythrocyte mass [31].

Hypoxia-Inducible Factor System

Anemia management may be altered with the addition of a new class of drugs that modulate the hypoxia-inducible factor-prolyl hydroxylase (HIF-PHD) axis. The search to untangle the molecular mechanisms involved in the increased EPO levels observed in individuals living at high altitude ultimately led to the discovery of HIF-PHD pathway, which mediates the body’s systemic and local response to oxygen deprivation. The HIF-PHD pathway is involved in many biological processes, including erythropoiesis, iron homeostasis, angiogenesis, as well as cellular growth and survival [32].

HIF, a heterodimer consisting of alpha and beta subunits, was first identified in 1992 [33]. The alpha subunit is the functional limiting factor and exists in three forms, HIF-1 alpha, HIF-2 alpha, and HIF-3 alpha. HIF-1 alpha and HIF-2 alpha bind to DNA sequences known as hypoxia response elements (HREs) to initiate the expression of target genes. On the other hand, HIF-3 alpha is devoid of a DNA-binding domain and serves as a negative regulator of the hypoxia-inducible cell responses [34]. Prolyl hydroxylase domain enzymes (PHD), which exist in four isoforms (PHD-1, PHD-2, PHD-3, PHD-4-TM), and factor inhibiting HIF (FIH) regulate the activity of HIF [3540]. PHD requires iron, ascorbate, alpha ketoglutarate, and molecular oxygen to hydroxylate the HIF alpha subunit at one or two proline residues, which permits binding of von Hippel-Lindau tumor suppressor protein (pVHL), followed by ubiquitination and proteasomal degradation [41•].Hydroxylation of C-terminal asparagine residues of HIF alpha subunit by FIH prevents recruitment of transcriptional coactivators and thereby inhibits HIF activity [38]. Oxygen saturation alters the activity of PHD family and FIH, resulting in rapid degradation or stabilization of HIF, under normoxic and hypoxic conditions, respectively.

HIF regulates both non-coding and coding transcriptional responses [42] and more than 500 HIF binding sites have been identified [43•]. HIF-2 alpha has a selective cell expression and is the primary regulator of renal EPO synthesis and iron metabolism [4450]. HIF-1 alpha is expressed in all cells and is primarily involved in the regulation of angiogenic factors and glycolytic enzymes [51, 52]. PHD-2, most abundantly expressed, is the predominant isoform involved in HIF alpha degradation in several cell types, while PHD-3 is more selective for HIF-2 alpha [53, 54]. In contrast, PHD-1 is not hypoxia sensitive and appears to be minimally involved in the maintenance of oxygen homeostasis. PHD 1, 2, and 3 regulate erythropoiesis to varying extents and through distinct HIF pathways but PHD 2 alone is the main suppressor of erythropoiesis via HIF-alpha degradation [55]. The role of PHD-4-TM in hypoxia-dependent regulation of erythropoiesis remains to be established in humans but appears to contribute to the regulation of EPO production and erythropoiesis in mice [56]. A deficiency of both PHD-1 and PHD-3 may be necessary for an erythropoietic response of hepatic origin, while a deficiency of PHD-2 is sufficient to stimulate renal EPO production [55]. Neither PHD-1 nor PHD-3 knockout produced detectable erythrocyctosis [55]. Therefore, a pan-PHD inhibitor may be necessary to reactivate hepatic EPO production [57]. The loss of PHD-3 results in abnormal development and hypofunction of the sympathoadrenal system, with hypotensive manifestation in mice [58].

Potential Advantages and Disadvantages of HIF-PHD Manipulation

Erythropoietin-Producing Cells

EPO is a glycoprotein hormone that is primarily synthesized by the peritubular interstitial fibroblasts (renal EPO producing cells; REPC) of the kidneys in postnatal life. The transdifferentiation of REPCs into myofibroblasts has been postulated to explain the mechanism by which REPC lose their capacity for EPO synthesis in CKD. Nonetheless, myofibroblasts retain functional plasticity and their EPO-producing capacity is recoverable [59]. Recently and with more advanced techniques, intercalated cells of the cortical nephrons have been identified as the site of basal EPO production [60]. Moreover, juxtaglomerular renin-producing cells and mesangial cells can also be induced to express EPO [61•]. The deletion of pVHL, which allows HIF levels to increase, converts renin-producing cells into EPO-producing cells in mice [62, 63]. Similarly, interstitial mesenchymal cells of adult kidneys may also differentiate into EPO-producing fibroblasts when exposed to a hypoxic environment [64]. Extra-renal sites of EPO production include hepatocytes [65], neurons [66], glial cells [67], and osteoblasts [68].

In addition, indoxyl sulfate has been associated with a reduced nuclear fraction of HIF-alpha and HRE-mediated luciferase activity in HepG2 cells [69] suggesting that uremic toxins in CKD population may interfere with HIF-erythropoiesis pathway. These findings emphasize the possible usefulness of PHD inhibitors in the ESRD population who have uremia with atrophic and fibrosed kidneys.

Diurnal Pattern of Erythropoietin Secretion

A diurnal variation in serum EPO concentration consisting of higher afternoon serum EPO levels and lower night-time EPO levels is present in healthy individuals [70]. This diurnal pattern of EPO response results from the interaction between HIF pathway and the circadian clock [71]. Circadian rhythm disturbances have been reported in renal disease and rhEPO administration may not produce a similar diurnal pattern of serum EPO levels. Therefore, there exists a potential to correct or enhance the diurnal rhythm of EPO secretion through HIF-PHD manipulation.

Erythropoietin Receptors and Supraphysiological Levels of Erythropoietin

The physiological serum concentrations of endogenous EPO range from 4 to 27 mU/mL but increase by 100- to 1,000-folds in response to hypoxia and anemia [72]. There are two different receptors for EPO; one exhibits high affinity while the other possesses low affinity for EPO. Although, the low affinity receptor mediates non-erythropoietic, tissue protective effects of EPO, it requires a high dose of exogenous EPO which creates a prothrombotic state via production of highly reactive platelets and vascular endothelium activation [73•]. In addition, higher concentrations of endogenous EPO are associated with incident heart failure in adults aged 70–79 years [74]. This highlights the potential advantage of PHD inhibitors which lead to small incremental increases in endogenous EPO levels which are near physiological range and likely to stimulate the high affinity receptor responsible for hematopoiesis. However, the clinical impact of sustaining physiological levels of endogenous EPO with PHD inhibitors on cardiovascular health remains to be determined.

HIF, Hepcidin, and Iron Homeostasis

Nearly 10 % of hemodialysis population exhibits ESA resistance/hyporesponsiveness, which is often caused by iron deficiency [7577]. Although iron deficiency can be corrected with iron supplementation, the inflammatory profile of chronic kidney disease predisposes to a state of functional iron deficiency. Hepcidin, a small peptide synthesized in the liver, serves as a key culprit in the development of functional iron deficiency. Hepcidin blocks iron export and utilization via induction of degradation of the only known iron exporter, ferroportin [78, 79•]. The expression of hepcidin is influenced by the rate of erythropoiesis, hypoxia, inflammatory status, and iron levels [8082]. Although a direct repressor effect of HIF on hepcidin was initially suggested from experiments conducted by Peyssonnaux et al. [83], HIF is now believed to influence hepcidin expression indirectly, via erythropoietic drive [84, 85] HIF-2 alpha upregulates ferroportin [86, 87] as well as duodenal cytochrome B reductase-1 and divalent metal transporter-1 enzyme which are critical for intestinal iron uptake [86, 8890]. On the other hand, HIF-1 alpha participates in the regulation of transferrin [91], transferrin receptor [92], ceruloplasmin [93], furin [94], and hemeoxygenase-1 [95], which are essential for iron homeostasis. However, neither HIF-1 alpha nor HIF-2 alpha appear to play a role in macrophage-mediated iron recycling [96].

Although, an inverse correlation has been reported between hepcidin concentration and rhEPO dose [97], intravenous iron sucrose administration to hemodialysis patients leads to an increase in hepcidin levels [98]. Hepcidin levels correlate with carotid pulse wave velocity [99] and carotid intima media thickness [100], as well as predict cardiovascular events [101] in the hemodialysis population. Moreover, hepcidin levels serve as an independent predictor of systolic blood pressure among healthy men [102]. A correlation has also been demonstrated between hepcidin levels and arterial plaques in postmenopausal women [103]. In contrast, low levels of hepcidin were associated with aortic stiffness [104]. However, Akhtar et al. showed that hyperlipidemia upregulates endothelial HIF-1 alpha which promotes atherogenic monocyte recruitment [105]. Elevated levels of hepcidin in diabetic CKD population independently predict mortality and are associated with progression of CKD [106]. Therefore, HIF stabilization therapy may increase the availability of iron for effective erythropoiesis. The implications of HIF axis in the coordination of iron supply for erythropoiesis is depicted in Fig. 1.

Fig. 1
figure 1

Schematic representation of the roles of HIF axis in iron homeostasis and erythropoiesis. Solid and dash arrows reflect positive and negative regulation, respectively. Furin: paired basic amino acid cleaving enzymes, matriptase-2: serine protease, HJV: hemojuvelin, sHJV: soluble hemojuvelin, PDGF-BB: platelet derived growth factor-BB, CREB-H: cAMP response element-binding protein-H, TWSG-1: twisted gastrulation-1, GDF-15: growth differentiation factor-15, EFRE: erythroferrone, TFR-1: transferrin receptor-1, RES: reticuloendothelial system, DCYT-B: duodenal cytochrome-B, DMT-1: divalent metal transporter-1, KLF-1: Kruppel-like factor-1, GATA-1: GATA binding protein-1 (globin transcription factor-1), PU.1: PU.1 transcription factor. Ferroportin, a rate-limiting factor for iron absorption, recycling, and mobilization, is upregulated by HIF-2 alpha. The expression of enzymes that are essential for intestinal uptake such as duodenal cytochrome B reductase-1 and divalent metal transporter-1 enzyme are also upregulated by HIF-2 alpha. Hepcidin sequesters iron in tissues, lowers serum iron levels, and reduces intestinal iron uptake. Hemojuvelin is necessary for hepcidin expression. Matriptase-2 suppresses hepcidin expression via downregulation of hemojuvelin. HIF-1 alpha upregulates furin, which generates soluble hemojuvelin by cleaving hemojuvelin. Soluble hemojuvelin suppresses hepcidin expression. In addition, hypoxia reduces hepcidin expression via platelet derived growth factor-BB, which downregulates cAMP response element-binding protein-H. Transferrin is an iron-binding glycoprotein that delivers to all tissues, including active sites of erythropoiesis. Proerythroblasts produce transferrin receptor-1 and erythroblasts produce erythroferrone, growth differentiation factor-15, and twisted gastrulation-1, which are all negative regulators of hepcidin. These factors form a communication system between sites of active erythropoiesis and hepcidin, thus ensuring an adequate supply of iron

While there may be multiple benefits associated with HIF-PHD manipulation, there are safety concerns to be addressed.

Tumor Growth and Proliferative Retinopathy

Intratumoral hypoxia often leads to overexpression of HIF. The HIF-1 alpha-regulated glycolytic enzymes and pro-angiogenic factors such as vascular endothelial growth factor (VEGF) hold important implications in tumor biology. In addition, HIF-1 alpha may promote tumor metastasis via induction of transforming and epidermal growth factors [107].

VEGF is essential for angiogenesis, in physiological and pathological conditions. VEGF not only promotes angiogenesis and increases vascular permeability but also affects tumor stem cell function as well as tumor initiation [108•]. Moreover, VEGF is also implicated in intraocular neovascular diseases, including proliferative diabetic retinopathy [109]. Therefore, the impact of HIF stabilization on VEGF is being evaluated in clinical trials.

Tissue Fibrosis

HIF-1 alpha upregulates the expression of DNA methylating enzymes which are implicated in the profibrotic state that result from hypoxic insult [110]. HIF-1 alpha also plays a role in myocardial fibrosis following infarction [111]. A growing body of research has also depicted a role of HIF-1 alpha in hepatic fibrosis [112]. In addition, HIF-1 alpha is involved in ischemic renal injury [113]. However, endothelial HIF-2 alpha appears renoprotective against ischemic insult [114]. On the contrary, HIF-1 stabilization attenuated myocardial fibrosis and enhanced right ventricular contractility in mice exposed to hypoxic environment [115]. Similarly, Kido et al. suggest that HIF-1 alpha is involved in limiting infarction size and cardiac dysfunction following myocardial infarction [116]. Nonetheless, the HIF pathway is involved in tissue repair and fibrosis [117].

Osteoporosis

Both HIF-1 alpha and HIF-2 alpha appear to decrease osteoclastic activity via upregulation of osteoprotegerin [118, 119]. In contrast, a HIF-1 alpha dependent increase in osteoclastic bone resorption mediated via angiopoietin-like 4 has been suggested by Knowles et al. [120]. Moreover, in a mouse model of osteoarthritis, mechanical overload led to osteoclastogenesis via HIF-1 alpha-dependent osteoprotegrin repression [121]. Notwithstanding, estrogen may prevent HIF-1 alpha stabilization and HIF-1 alpha-dependent activation of osteoclasts may contribute to estrogen deficiency-related osteoporosis [122]. The exact role of HIF pathway in bone homeostasis remains poorly understood.

HIF System and Pulmonary Hypertension

The response of the systemic circulation to hypoxia is vasodilation, while the pulmonary circuit responds with vasoconstriction. Hypoxic pulmonary vasoconstriction is enhanced in humans with gain-of-function mutations in HIF-2 alpha [123]. The loss-of-function mutation in one allele encoding HIF-1 alpha was partially protective against chronic hypoxia-induced pulmonary hypertension, reducing vascular modeling and attenuating increases in right ventricular pressure and hypertrophy in mice [124, 125]. Moreover, one of the HIF target genes encodes for endothelin-1 (ET-1) which is a potent vasoconstrictor and promoter of smooth muscle cell proliferation [126]. Chronic hypoxia increases the plasma concentration of ET-1 in the lung and up-regulates its receptors [127, 128]. A vicious cycle is created because ET-1 also increases HIF-1 alpha expression in pulmonary artery smooth muscle cells [129, 130]. In addition, a heterozygous deficiency of HIF-2 alpha attenuates right ventricular pressure and vascular remodeling. Overexpression of HIF-2 alpha has been implicated in the development of pulmonary hypertension in humans [123, 131] and mouse [132] studies. The increased predisposition to pulmonary hypertension in Chuvash polycythemia (a genetic disorder associated with elevated levels of HIF and erythrocyte mass) and in animal models with homozygous mutation of pVHL supports a role of HIF in the pathogenesis of pulmonary hypertension [133, 134]. The pathogenesis of chronic hypoxia-induced pulmonary hypertension potentially involves a feedforward mechanism which includes enhanced HIF-2 alpha expression, leading to increased synthesis of ET-1, which increases expression and transcriptional activity of HIF-1 alpha in pulmonary artery smooth muscle cells [135]. In addition, endothelial HIF pathway has been shown to increase endothelial cell expression of connective tissue growth factor, vascular permeability, and pulmonary artery smooth muscle cell proliferation in mice. Moreover, HIF pathway involves certain proinflammatory substances which may also contribute to the development of pulmonary hypertension [136]. Recently, an increased expression of HIF-1 alpha was evident in explanted lung tissue belonging to patients suffering from group III pulmonary hypertension [137]. The implications of HIF-PHD manipulation on pulmonary pressures are of interest in the CKD population due to the relatively high prevalence of pulmonary hypertension. The impact of PHD inhibitors on the development or progression of pulmonary hypertension should be monitored as studies proceed.

Hypoxia Mimetic Agents Under Clinical Investigation

Fibrogen

Several pharmaceutical companies have filed new drug applications for the use of HIF-PHD inhibitor in the treatment of anemia. Fibrogen has completed phase II trials on FG-2216 [138]. However, human clinical trials investigating FG-2216 have not been conducted since 2007 when a case of death due to fulminant hepatitis was reported in a phase II trial, in spite of FDA’s approval to resume clinical investigations. A second-generation HIF-PHD inhibitors, FG-4592 (roxadustat), with an improved pharmacokinetic and pharmacodynamic profile has now reached phase III trials. The phase IIa trial investigating multiple doses of FG-4592 with BIW or TIW dosing frequency in CKD patients not on dialysis revealed a dose-related increase in hemoglobin from baseline while suppressing hepcidin levels [139]. Ninety-six CKD patients not on dialysis who did not receive rhEPO for 12 weeks prior to randomization, participated in a phase IIb trial of 16–24 weeks of treatment with FG-4592, in varying doses (50–140 mg) and frequencies (QW, BIW, TIW) [140]. Hemoglobin increases of >1 g/dL were observed in 96 % of the patients while 93 % achieved a hemoglobin target of 11–13 g/dL [140]. A maximum mean hemoglobin increase of 2.09 g/dL from baseline was observed at week 17 across all FG-4594 dose groups [140]. In an open-label, phase IIb trial involving rhEPO-naïve incident dialysis patients, FG-4592 produced a mean hemoglobin increase of ≥2.0 g/dL within 7 weeks and reduced hepcidin levels [141]. FG-4592 reduced hepcidin levels in two CKD studies and in two dialysis studies [142]. A post hoc analysis found a more profound reduction in hepcidin levels in incident dialysis patients that received FG-4592 and either no iron or oral iron than those that received FG-4592 and intravenous iron [143]. The analysis of four phase II studies involving CKD and dialysis patients revealed an increase in soluble transferrin receptor, which supports an increased iron supply for erythropoiesis [144]. Total cholesterol was reduced in FG-4592 treatment groups compared to placebo or ESA treatment groups in phase II studies involving CKD patients and studies consisting of dialysis patients [145]. FG-4592 also showed an improvement in the health related quality of life in 141 CKD patients and 55 hemodialysis patients [146].

Glaxo–Smith–Kline

Glaxo–Smith–Kline is investigating a HIF-PHD inhibitor drug candidate, GSK1278863. Multiple phase I studies exploring pharmacokinetics and pharmacodynamics of GSK1278863 in humans have been completed. Recently, the results of phase IIa trial consisting of a 4-week treatment with GSK1278863 demonstrated a dose-dependent increase in hemoglobin. A mean increase of 1 g/dL was achieved in 5-mg treatment arm at 4 weeks in predialysis rhEPO-naive population and a dose-dependent decrease in hepcidin levels was observed [147]. In addition, ferritin levels decreased at 4 weeks while transferrin levels and total iron-binding capacity was increased in the 5-mg GSK1278863 group. The mean hemoglobin concentration was maintained in hemodialysis population after switching from rhEPO in the 5-mg treatment arm but not with lower doses of 0.5 and 2 mg of GSK1278863. The hepcidin levels were not reduced in the 5-mg GSK 1278863 group but rather an increase was seen in the 0.5 and 2 mg GSK 1278863 groups. A decreasing trend of ferritin levels was evident with increasing dose strength of GSK 1278863. Intra-subject variability in response to GSK1278863 and in vascular endothelial growth factor levels was evident. GSK1278863 increased or maintained hemoglobin without producing supraphysiological plasma levels of endogenous EPO [147].

Akebia Therapeutics

Akebia Therapeutics is developing a predominantly HIF-2 alpha-stabilizing PHD inhibitor, AKB-6548 (vadadustat). The EPO levels were increased at 8 and 12 h following the administration of a single dose of 500 mg of AKB-6548 in a phase IIa open-label study which included 22 CKD patients [148]. EPO levels returned to baseline after 24-h post AKB-6548 exposure and hepcidin levels were decreased [148]. In another phase IIa, 28-day study involving 10 CKD patients, AKB-6548 increased hemoglobin from a baseline of 9.91 ± 0.63 to 10.54 ± 0.89 g/dL while ferritin decreased from a baseline of 324 ± 199.2 to 271.7 ± 181.3 ng/mL [148]. A Phase IIa multidose trial of 42-days treatment with varying doses of AKB-6548 produced a significant increase in hemoglobin in all dose groups [149]. It also demonstrated a significant increase in total iron binding capacity and a reduction in hepcidin levels. Similarly, a phase IIb study showed that once daily 450 mg of AKB-6548 maintained a mean baseline hemoglobin of 10.5 g/dL over 20 weeks in CKD patients not on dialysis [150]. A linear dose-exposure relationship of AKB-6548 was evident, irrespective of renal function [151]. In another study of hemodialysis patients, the hemoglobin was effectively maintained over 16 weeks after switching from rhEPO therapy to AKB-6548 and hepcidin levels were reduced [152]. AKB-6548 has shown a potential to restore an enhanced diurnal rhythm of EPO production and it has not been associated with VEGF elevation [149, 153].

Bayer

Bayer Healthcare is currently evaluating a pan-HIF PHD inhibitor, BAY 85-3934 (molidustat). A dose-dependent increase in endogenous EPO levels following BAY 85-3934 administration in rats with gentamicin-induced renal anemia has been reported [154]. Daily treatment with 5 mg/kg of BAY 85-3934 in rodents counteracted the decline in mean packed cell volume (PCV) produced by peptidoglycan-polysaccharide (PG-PS) challenge [154]. EPO expression was significantly increased in the kidney following treatment with BAY 85-3934 but did not resolve inflammatory anemia induced by PG-PS. Treatment with BAY 85-3934 also increased mean PCV in a disease model of subtotal nephrectomy [154]. BAY 85-3934 is absorbed rapidly and produces a dose-dependent increase in endogenous EPO levels and an increase in reticulocyte count following 37.5- and 50-mg dose in healthy men [155].

The findings from clinical trials evaluating PHD inhibitors are encouraging but long term monitoring will be necessary to enhance our understanding of the various effects of HIF-PHD axis manipulation. The impact of PHD inhibitors on total iron-binding capacity, ferritin, hepcidin, and VEGF levels is summarized in Table 1.

Table 1 The impact of PHD inhibitors on ferritin, total iron-binding capacity, hepcidin, and VEGF levels

HIF System and Systemic Arterial Pressure

The HIF-mediated transcriptional cascade involves many genes that participate in vasomotor control [136]. A significantly lower blood pressure compared to matched controls has been reported in patients suffering from Chuvash polycythemia. [156].

Nitric oxide (NO) serves as a physiological vasodilator and the balance between NO, sympathetic, and renin–angiotensin systems determines the peripheral vascular resistance. Indeed, hypertensive individuals have lower bioavailability of NO than normotensives [157, 158]. Increasing L-arginine bioavailability leads to increase in NO concentration which produces vasodilation, thereby decreasing arterial pressure [159, 160]. Recent evidence suggests that the balance between HIF-1 alpha and HIF-2 alpha is critical for the maintenance of systemic arterial pressure [161•]. In macrophages, the activity of HIF-1 alpha and HIF-2 alpha is influenced by Th1 and Th2 cytokines which regulate NO homeostasis [162]. Th1 cytokine-induced transcription of HIF-1 alpha leads to an increased expression of NO synthase 2 (NOS2) which results in increased NO production [163165]. On the other hand, Th2 cytokine-dependent transcription of HIF-2 alpha induces arginase-1 which depletes L-arginine, thereby reducing NO production, indirectly [162]. HIF-regulation of NO homeostasis is illustrated in Fig. 2.

Fig. 2
figure 2

The potential role of HIF axis in NO homeostasis [162169]

NO suppresses HIF-1 alpha activity during hypoxia in Hep3B cells [166] while enhancing its activity in hypoxic colon carcinoma cells [167]. Under normoxic conditions, NO stabilizes HIF-1 alpha [168, 169]. It is speculated that an alteration in the HIF-1 alpha and HIF-2 alpha expression pattern may not only be involved in idiopathic hypertension but also for many of the fibrotic vascular events that are common in hypertensive individuals since arginase enzyme indirectly affects collagen synthesis [161•]. The equilibrium of HIF-1 alpha and HIF-2 alpha is also a prerequisite for oxygen sensing by the carotid body and adrenal medulla and for their role in the maintenance of cardio-respiratory homeostasis [170]. Chronic intermittent hypoxia causes dysequilibrium in HIF-1 alpha and HIF-2 alpha activity and leads to an imbalance of antioxidant and pro-oxidant enzyme expression, with a consequent increase in reactive oxygen species that alter the chemosensory reflex and contribute to the development of hypertension [171•]. Additionally, increased endothelial HIF-1 alpha expression play a role in glomerular injury and progression of hypertensive kidney disease [172].

Almost half of the hypertensive population suffers from salt sensitive hypertension which predisposes to a greater risk of end organ damage [173176]. The salt sensitivity of blood pressure is inversely correlated with renal function [177]. HIF-1 alpha regulates the expression of nitric oxide synthase (NOS), hemeoxygenase-1 (HO-1), and cyclooxygenase-2 (COX-2) and these enzymes upregulate in response to a high-salt diet, whereas this response is diminished in salt-sensitive hypertensive animal models [178184]. High-salt intake inhibits renal medullary PHD-2 expression and activity in normotensive rats but not in salt-sensitive hypertensive rats [185]. Inhibition of NOS, HO-1, and COX-2 impair sodium excretion and increase salt sensitivity of arterial blood pressure. Moreover, HIF-1 alpha induction or PHD-2 gene silencing or inhibition, promotes natriuresis and attenuation of salt sensitive hypertension in Dahl S rats [186188]. Of note, inhibition of HIF-1 alpha mediated gene transcription in rats did not result in hypertension in the absence of a high salt challenge. These findings support an important role for the HIF axis in the regulation of systemic arterial pressure. It appears to exert its effect on vascular tone by regulating the generation of NO.

The Impact of Investigational HIF-PHD Inhibitors on Systemic Blood Pressure

Emerging but limited evidence supports a small blood pressure-lowering effect of PHD inhibitors. BAY 85-3934 lowered blood pressure in healthy Wistar rats and cynomolgus monkeys [154]. In fact, the systolic blood pressure was significantly lower in the 5 mg/kg BAY 85-3934 group compared to control and rhEPO treated group. The effect of BAY 85-3934 on mean systolic blood pressure was comparable to enalapril, but without compensatory increase in prorenin levels [154].

FG-2216 improved hemoglobin levels in a rat remnant kidney model without increasing systolic blood pressure in contrast to the rhEPO group in which hemoglobin increased along with exacerbation of hypertension [189]. Two episodes of moderate exacerbation of hypertension were reported in one patient participating in the 4-week phase IIa trial of FG-4592 [139] but no safety signals were detected from ambulatory blood pressure monitoring. A mean reduction of 2.6 ± 9.6 mmHg in blood pressure from baseline was observed in the phase IIb trial of 16 and 24 weeks of treatment with FG-4592 [190]. Adverse events of hypertension were reported in 7 % of FG-4592 treated patients, which is lower than the rates reported in similar ESA-treated populations [191, 192]. Moreover, the mean blood pressure remained unchanged in 107 diabetic CKD patients not on dialysis (subset population of phase IIb trial) treated with varying doses and frequencies of FG-4592 over 16 and 24 weeks [193]. However, in another phase IIb trial with open-label design, the most frequent treatment emergent adverse event (10 %) was hypertension which necessitated an increase or change of antihypertensive medication [141].

In a phase IIa dose escalation study, treatment with AKB-6548 in 10 CKD patients for 28 days was accompanied by a small reduction in mean blood pressure [148]. No significant change in blood pressure was observed in 91 patients who received AKB-6548 in the phase IIa trial [194].

Long-term studies regarding the impact of PHD inhibitors on arterial pressure and cardiovascular endpoints will be clarified in the ongoing phase III trials.

Conclusions

A new era of HIF stabilizers is cautiously welcomed but warrants appropriately designed studies to evaluate cardiovascular risks as well as the potential risks of CKD progression and pulmonary hypertension.