|Home | About | Journals | Submit | Contact Us | Français|
Group II pulmonary hypertension commonly occurs in the setting of a pressure overloaded left ventricle that is also conducive to the development heart failure with preserved ejection fraction. Population trends and a high prevalence of underlying causative conditions, such as essential hypertension or aortic stenosis, have increased awareness of the pressure overloaded left ventricle and associated Group II pulmonary hypertension. Patients often exhibit poor exercise tolerance and signs of heart failure indistinguishable from systolic heart failure; but effective medical treatments in this area have been lacking. Recent pre-clinical work has shed light on how the down-regulated nitric oxide – cyclic GMP pathway (within the myocardium and pulmonary vasculature) contributes to the pathophysiology of these associated conditions. This article will discuss the impact of the nitric oxide – cyclic GMP pathway on the pathogenesis of the pressure overloaded left ventricle and Group II pulmonary hypertension, and will also introduce the potential therapeutic value of modulating this pathway.
Group II pulmonary hypertension (PH) or ‘PH owing to left heart disease’ represents an integral aspect to the 4th World PH Symposium’s classification scheme . Within this category, PH develops from chronic left-sided heart conditions, such as left ventricular myocardial dysfunction (systolic, diastolic, or mixed types) and/or valvular heart disease due to aortic or mitral valve dysfunction. Given the high prevalence of left-sided heart disease, particularly in Western countries, Group II PH represents the largest segment of the PH population .
A common element to many left heart conditions is the pressure overloaded left ventricle (LV), resulting from chronically increased LV afterload. To compensate for conditions of increased afterload, the LV undergoes hypertrophic remodeling, which favorably lowers wall stress (Figure 1). Over time, the remodeling process unfavorably alters LV filling characteristics through impaired relaxation (in early diastole) and decreased compliance or increased stiffness (in mid to late diastole), which leads to diastolic dysfunction . Diastolic dysfunction may occur in isolation or in combination with valvular disease or systolic dysfunction and can be established through echocardiography (mitral inflow patterns or tissue Doppler imaging) or sophisticated hemodynamic evaluation (LV pressure-volume curves) [3, 4]. When diastolic dysfunction is present in the context of classic heart failure manifestations (i.e. exertional dyspnea, orthopnea, pulmonary edema, pleural effusions), diastolic heart failure or heart failure with preserved ejection fraction (HFpEF) are appropriate clinical designations . In such situations, LV filling pressures are often elevated (i.e. left ventricular end-diastolic (LVEDP) or pulmonary capillary wedge pressure (PCWP) ≥ 15 mm Hg); and once the trans-pulmonary gradient (TPG, normally ≤ 12 mm Hg) is factored, it is understandable for abnormal pulmonary artery pressures (mean pressure ≥ 20 mm Hg) to occur [6, 7]. Over time, the left atrium reacts to elevated LV filling pressures by dilating and being less compliant, leading to elevated atrial pressures that are transmitted to the pulmonary veins causing pulmonary venous hypertension (Group II PH).
Unlike Group I PH (Pulmonary Arterial Hypertension) conditions that persist in spite of normal left-sided filling pressures and are noteworthy for a proliferative arteriopathy resulting from imbalances in vascular tone, platelet function, and cellular proliferation, Group II PH is driven by elevated “down-stream” pressures in the LV and left atrium. In some instances, pulmonary venous hypertension can lead to significant remodeling of pre-capillary arterioles and an exaggerated or “reactive” form of PH (i.e. TPG > 12 and PVR > 3 Wood units) [8, 9]. If culpable co-morbid conditions (e.g. chronic obstructive lung disease, obstructive sleep apnea syndrome, or chronic thromboembolic disease) are excluded, these cases of “reactive” PH are speculated to either result from an (undefined) predisposing genetic factor “triggered” by chronically elevated down-stream pressures or, less likely, an independent and coexisting proliferative arteriopathy akin to idiopathic pulmonary arterial hypertension.
Two clinical conditions that often lead to a pressure overloaded LV and share many pertinent features are hypertensive cardiomyopathy (HCM) and aortic valve stenosis (AS). First, the pressure overload caused by systemic hypertension or valve stenosis leads to LV hypertrophy and diastolic dysfunction. Second, pulmonary venous hypertension is frequently encountered in both entities and can be associated with an exaggerated rise in the pulmonary artery pressures and right ventricular dysfunction, which has important clinical ramifications. Finally, LV systolic dysfunction is typically not present until the very late stages and is less correlated with the pulmonary vascular changes .
HFpEF as a clinical presentation of the pressure overloaded LV (Figure 1) has become an extremely important clinical entity in the last twenty years. Two single-center experiences with a pooled population of 9000 patients found 30–50% of hospitalized patients with congestive heart failure (CHF) had preserved ejection fractions [11, 12]. Furthermore, a Japanese study noted no long-term survival differences between patients hospitalized for CHF with and without preserved ejection fractions (~70% at 3 years) .
As described earlier, HFpEF and Group II PH have a tight pathophysiologic relationship -- one that is increasingly recognized in the clinical arena. While most forms of left-heart disease can be associated with Group II PH, a key determinant is the presence and degree of diastolic dysfunction [14, 15]; for example, neither the magnitude of depressed LV ejection fraction in systolic heart failure nor the valve area in AS are as powerful determinants of PH as the degree of associated diastolic dysfunction in those respective conditions [10, 14, 15]. Further strengthening the relationship between HFpEF and PH, one recent study reported 70–85% of patients with HFpEF have PH and the prevalence remained high after excluding co-morbidities independently associated with PH . With regards to AS, two retrospective series of > 600 patients (each) reported pulmonary artery systolic pressures > 60 mm Hg in 16–20% of their cohorts [17, 18]. When present in cases of HFpEF, PH confers an increased risk for dying even when adjusting for age, while significant PH negatively impacts peri-operative mortality in patients undergoing aortic valve replacement [16, 19].
When the pressure overloaded LV state leads to HFpEF, patients experience reduced exercise capacity from stress-related cardio-pulmonary interactions, including blunted augmentation of stroke volume and decreased lung compliance due to pulmonary congestion from rising LV filling and left atrial pressures [20–23]. Unfortunately, little is known of the sequence and time-course between this dynamic and repetitive process to a phase of persistent pulmonary hypertension. But, clinical experience has shown that the combination of significant PH (and right ventricular dysfunction) in the setting of HFpEF influences symptomatology and hinders the use of selective pulmonary vasodilators, which may have deleterious effects in the context of abnormal LV filling.
Treatment of HFpEF has been particularly vexing because, unlike systolic heart failure, there is a paucity of conclusive data from randomized-controlled trials. Two trials of angiotensin II receptor blockers (CHARM-preserved: candesartan vs. placebo; & I-PRESERVE: irbesartan vs. placebo) and the PEP-CHF trial (perindopril vs. placebo) failed to show a mortality benefit with their respective interventions; and, reduction in rates of cardiovascular hospitalization was mixed. A recently published trial also failed to demonstrate improvement in exercise capacity after 12 months of enalapril (versus placebo) in a small cohort of HFpEF patients . In the absence of evidence-based guidelines, management of HFpEF has been driven by physiologic principles in attempts to alleviate symptoms and improve functional status. Key recommendations include diuretics to optimize volume status and avoid pulmonary congestion, anti-hypertensives to reduce LV afterload and minimize myocardial ischemia, rate control agents in the setting of tachyarrhythmias, and anti-anginal medications when appropriate; rhythm control and coronary revascularization are also advised in appropriate circumstances [5, 25]. Meanwhile, medical therapy for AS and associated PH has been non-existent, partly out of concerns with decreasing afterload and potentially inducing hypotension in patients with limited ability to augment cardiac output because of a fixed stenosis. However, systemic afterload reducers (e.g. ACE inhibitors and nitroprusside) have been safely given to patients with severe AS and are well tolerated [26–30].
Recent preclinical work has highlighted the emerging role of the nitric oxide – cyclic guanosine monophosphate (NO-cGMP) pathway in the pressure-overloaded LV and associated PH, as NO-cGMP downregulation in the heart and the pulmonary vasculature has deleterious consequences. The role of NO-cGMP signaling in the pressure-overloaded LV and Group II PH is the focus of remainder of this article.
In patients with a pressure-overloaded LV and associated Group II PH, numerous signaling pathways are altered in both the heart and pulmonary vasculature. Whereas the alteration of many signaling pathways may be unique to either the heart or pulmonary vasculature, downregulation of the NO-cGMP pathway seems to be a common finding in both locations (Figure 2). Efforts to better understand this pathway’s downregulation will likely yield new insights with potential therapeutic implications.
A variety of stimuli lead to the production of NO in the vascular endothelium by nitric oxide synthase (NOS) [31, 32]. NOS functions as a protein dimer; physical or functional uncoupling of the enzyme leads to reduced NO and increased reactive oxygen species production. Once generated, NO diffuses across cell membranes to the cytoplasm of adjacent smooth muscle cells, where it activates soluble guanyl cyclase (sGC) and increases intracellular levels of cGMP, which is hydrolyzed by phosphodiesterase type 5 (PDE5) in the pulmonary circulation and the myocardium. Several targets of cGMP have been identified; most notable for this review is cGMP-dependent protein kinase (PKG). Phosphorylation of PKG has numerous downstream effects, including decreasing vascular tone and attenuating cellular proliferation and hypertrophy [33–35].
cGMP is also generated as natriuretic peptides (ANP, BNP, and CNP) bind transmembrane guanyl cyclase (particulate guanyl cyclase), which has an intracellular domain that hydrolyzes guanosine triphosphate (GTP) to cGMP. Recent evidence has highlighted that spatial distribution and signaling of cGMP is not uniform ; this may underlie the differential effects of cGMP generated via natriuretic peptides/particulate guanyl cyclase vs. nitric oxide/soluble guanyl cyclase. This functional compartmentalization of cGMP also has ramifications for therapeutic targets.
The NO-cGMP signaling pathway plays an important role in the myocardium; downregulation of this pathway has deleterious structural and functional myocardial effects, whereas upregulation can attenuate or reverse unfavorable remodeling and dysfunction that occur in response to a variety of stimuli. This has been demonstrated in models of ischemia/infarction [37, 38], adrenergic stimulation , angiotensin II stimulation [40–42], and doxorubicin cardiotoxicity . The role of this pathway in the pathophysiologic response of the ventricle to pressure overload has also been investigated; downregulation of its signaling is associated with hypertrophic remodeling (increased fibrosis and myocyte hypertrophy) and myocardial dysfunction (both diastolic and systolic). To simulate the pressure overloaded LV found in systemic hypertension or AS, the transaortic constriction (TAC) model has been developed and entails surgical banding of the proximal aorta in mice, creating a pressure gradient and increased afterload for the left ventricle.
There is some controversy regarding the role of NOS in ventricular remodeling induced by pressure overload. Some have found that NOS−/− mice exhibit increased LV hypertrophic remodeling and reduced LV function with aortic banding compared to WT mice [44, 45]. Buys showed that restoration of NOS to the heart of NOS−/− mice attenuated LV hypertrophy and dysfunction . In contrast, Takimoto reported more LV hypertrophy, dilation, and fibrosis in WT compared to NOS−/− mice . These conflicting results may have been due to differences in the tightness of the banding and/or differences in the production of reactive oxygen species (ROS).
Takimoto did show that WT mice exhibited a marked increase in ROS production in the setting of pressure overload. ROS can be generated by multiple sources, including NADPH oxidases, xanthine oxidase, mitochondria, and uncoupled NOS . Relevant to the NO-cGMP signaling pathway, Takimoto showed that the increased ROS generated in the setting of pressure overload was associated with NOS uncoupling. Taking it further, they noted a decline in tetrahydrobiopterin (BH4) levels in the WT mice in the setting of pressure overload; BH4 is a necessary cofactor for NOS to generate NO. When mice were treated with BH4, NOS coupling was restored, ROS generation was suppressed, and hypertrophic remodeling and dysfunction were blunted. Recently, Silberman reported that in a mouse model of mild hypertension and diastolic dysfunction (without cardiac hypertrophy or systolic dysfunction), there was an increase in oxidative stress that was associated with NOS uncoupling and depletion of BH4; treatment with BH4 both prevented or reversed diastolic dysfunction, highlighting the important role of NOS coupling in diastolic dysfunction . When considering therapeutic strategies targeting oxidative stress, it will likely be important to target the source(s) of oxidative stress most relevant for the underlying pathology; for the pressure overloaded LV this may mean targeting NOS uncoupling.
Early research in human heart pathophysiology suggested that PDE5 was insignificant—either not expressed at all or expressed at only low levels [50–52]. In the setting of pressure overload, however, PDE5 expression and activity appears to be induced. Takimoto showed that in TAC mice, 60% of the total cGMP-esterase activity was attributable to PDE5, which reflected increased PDE5 activity in comparison to control mice (35–45%). With this increase in PDE5 activity in banded mice, PDE5 inhibition with sildenafil more than doubled downstream PKG-1 activity. Treatment with sildenafil significantly blunted hypertrophic remodeling, fibrosis, and ventricular dysfunction. In a treatment reversal model, when treatment with sildenafil was delayed in the TAC mice until hypertrophy had been established—which has relevance for treatment of established hypertrophy in the pressure overloaded human heart—the hypertrophy and fibrosis reversed toward baseline levels. They subsequently showed that in TAC mice who have already developed LV hypertrophy, dilation, and dysfunction due to pressure overload, sildenafil prevented further remodeling and dysfunction in the face of ongoing pressure overload . Intracellular calcium handling was also improved and associated with improved contractility and relaxation . Along these lines, Kruger recently demonstrated that PKG-1 can phosphorylate titin (a key determinant of myocardial stiffness) and reduce passive myocardial stiffness , which could have a favorable effect on diastolic function. Finally, Nagendran showed that PDE5 protein expression is markedly upregulated in the hypertrophied human right and left ventricles, indicating a possible therapeutic target in humans .
While PDE5 inhibition is becoming an attractive therapeutic option for the pressure overloaded ventricle, two important findings should be highlighted. The first reinforces the functional compartmentalization of cGMP pools. In two related studies, Takimoto showed that adrenergic stimulation of the heart is regulated by cGMP signaling—specifically a nitric oxide-synthesized/PDE5-hydrolyzed pool—via PKG. PDE5 inhibition suppressed adrenergic-stimulated contractility, but this effect was lost in NOS3−/− mice or in control mice with inhibitors to NOS, sGC, or PKG-1. In contrast, ANP had no effect on adrenergic-stimulated contractility; ANP increased cGMP levels 5-fold, but had no effect on PKG-1 activity. Although these studies were performed in the setting of adrenergic stimulation, they emphasize an important point for identifying therapeutic targets for the pressure overloaded left ventricle: drugs that modulate cGMP synthesis and hydrolysis may likely have differential effects depending on which pool(s) of cGMP and therefore which downstream molecule(s) are affected.
The second important finding to recognize is the relationship between oxidative stress and PDE5 inhibition. Kass showed that in WT mice, aortic banding led to increased ROS generation and this was mechanistically linked—at least in part—to NOS uncoupling . They also showed that PDE5 inhibition in TAC mice led to reduced ROS generation and restored physical and functional coupling of NOS . Recently, Lu provided further support for a link between oxidative stress and PDE5 . They found a strong correlation between markers of oxidative stress and PDE5 protein levels in patients with end-stage dilated cardiomyopathy. Like Kass, they also found that PDE5 inhibition with sildenafil diminished TAC-induced oxidative stress, LV hypertrophy, and LV dysfunction. Both the increased NO production (as opposed to ROS generation) via a coupled NOS as well as the decreased degradation of cGMP via PDE5 inhibition would serve to upregulate NO-cGMP signaling and presumably have favorable effects on LV structure and function in the setting of pressure overload. In short, the interplay between oxidative stress and PDE5 activity has been clearly established; clarifying the precise mechanisms of this relationship is not only likely to enhance our understanding of the pathophysiology of heart failure, but provide valuable insights to guide therapeutic intervention.
The NO-cGMP pathway is also an important regulator of the pulmonary vasculature and highly relevant to understanding the pathophysiology of pulmonary arterial hypertension . Vascular tone, cell proliferation, and apoptosis are modulated by NO-cGMP signaling and downregulation of this pathway is associated with vasoconstriction, in situ thrombosis, and vascular remodeling [57–59]. This pathway plays a role in other forms of pulmonary hypertension, including our focused area of Group II PH. Melenovsky studied patients with symptomatic heart failure, left ventricular dysfunction (EF < 35%), with either low PVR or high PVR (≤ 200 or >200 dyn•s•cm−5) . Those with high PVR had reduced transpulmonary cGMP release, but had similar transpulmonary BNP uptake. The subjects with high PVR were administered 40 mg of oral sildenafil, which led to a significant decrease in pulmonary artery pressure and PVR, and increase in pulmonary artery compliance and cardiac index; these changes were accompanied by an increase in transpulmonary cGMP release, but no change in transpulmonary BNP uptake. Of note, the sildenafil-induced improvement in PVR was similar regardless of whether subjects had been labeled with “irreversible” or “reversible” pulmonary hypertension based on a prior pulmonary vasodilation test with prostaglandin E1.
The pressure overloaded LV and Group II PH represent a “coupled” pathophysiologic process in at least two ways: 1) they often occur together clinically with the pressure overloaded LV causing pulmonary venous hypertension; and 2) the NO-cGMP signaling pathway is downregulated in both processes leading to deleterious consequences. As such, therapies directed at upregulating the NO-cGMP pathway may have favorable effects on both aspects of this coupled pathophysiology. Targeting this pathway may have favorable effects on the pulmonary vasculature and LV structure/function via both a hemodynamic effect (afterload reduction through reduced systemic vascular tone) and direct myocardial effects through LV relaxation and reverse tissue remodeling (altering hypertrophic, proliferative, oxidative stress, and fibrosis pathways). Therapeutic strategies include: increasing endogenous NO production (through NO donors, increasing NOS levels and/or activity, or improved NOS coupling), stimulating sGC, increasing cGMP (keeping in mind that the particular “pool” of cGMP may be important), and decreasing degradation of cGMP by inhibiting PDE5.
To date, most of the interventions affecting the NO-cGMP pathway in patients with Group II pulmonary hypertension have involved PDE5 inhibition (e.g. sildenafil) in patients with systolic heart failure. Lepore demonstrated acute drops in PA pressure, PCWP, and PVR with a commensurate rise in cardiac output after administration of sildenafil (or inhaled nitric oxide) in 11 symptomatic subjects with baseline LVEF < 40%, mean PA pressure ≥ 25 mmHg, and elevated PVR; the magnitude and duration of effects were more substantial by combining agents . In another study of 13 symptomatic subjects with LVEF < 35%, Lewis observed improvements in resting and exercise hemodynamics (i.e. PA pressures, PVR, SVR, and cardiac index) and exercise capacity (VO2 max) after a single dose of sildenafil; notably, only patients with baseline mean PA pressure > 25 mmHg benefited . With longer-term sildenafil administration, Guazzi randomized 46 subjects to sildenafil or placebo for 6 months and demonstrated significant improvement in PA pressures and exercise capacity, as well as improved systemic endothelial function (measured by brachial arterial flow-mediated dilation) . Lastly, Lewis treated 34 patients with LVEF < 40% and mean PA pressure > 25 mm Hg with varying doses of sildenafil (or placebo) for 12 weeks and noted significant improvement in peak VO2 (~2.0 ml/kg/min placebo-adjusted treatment effect), 6 minute walk distance (62 meters vs. 29 meters), as well as PVR and PVR/SVR at rest and peak exercise . Importantly, the benefit in 6 minute walk distance persisted for at least 6 months. The sildenafil group also reported improved quality-of-life and better New York Heart Association functional classification. In total, these studies confirm sildenafil’s benefits in patients with systolic heart failure and elevated PVR through preferential pulmonary vasodilation, which translated into improved exercise capacity and quality-of-life.
Unfortunately, clinical investigations aimed at upregulating NO-cGMP signaling have been unexplored in patients with pressure-overloaded left ventricles and preserved systolic function. Investigation in this group of patients presents unique challenges: accurate and consistent case identification, inhomogeneous sub-types of patients with potential variable treatment responses, and relevant co-morbid conditions that may mute the benefits of improved LV function. Additionally, endpoints of study may need refining, as “survival benefit” may be impractical for this generally older population; and, benefits in exercise capacity and quality-of-life should not be underestimated as goals of therapy. Finally, the critical period for assessing response to NO-cGMP modulation may need to be longer than what is typically afforded to vasodilatory compounds, as modulation of the pathway has effects on myocardial hypertrophy and remodeling, which (intuitively) would take longer to materialize to a clinically discernible level.
The ongoing RELAX study (Evaluating the Effectiveness of Sildenafil at Improving Health Outcomes and Exercise Ability in People with Diastolic Heart Failure: ClinicalTrials.gov identifier: NCT00763867) is investigating the benefits to exercise capacity after 6 months of sildenafil (versus placebo) in individuals with diastolic heart failure (LVEF >50%). Other agents that modulate the NO-cGMP pathway (Figure 2) at different levels may also have potential benefits. Cicletanine® is an anti-hypertensive agent (approved in Europe) that has garnered interest for Group I PH due to its potential to increase NO-cGMP pathway activity through an unclear mechanism. Tetrahydrobiopterin® has already been investigated in a small Phase I study of PAH patients and found to be well-tolerated along with a hint of improved exercise capacity . Meanwhile, riociguat® has a down-stream effect through direct sGC stimulation and is being investigated in pulmonary arterial hypertension, chronic thromboembolic pulmonary hypertension, and pulmonary hypertension associated with systolic heart failure. Nebivolol® and celiprolol® are selective β1-receptor blockers that also activate NOS and have anti-oxidant properties through stimulation of other beta receptors, potentially benefitting the pressure overloaded LV and resultant pulmonary hypertension [66–68].
The pressure overloaded LV and accompanying Group II PH are commonly observed clinical conditions associated with chronic symptoms and poor outcomes. Recent preclinical work has highlighted the relevance of NO-cGMP signaling for both of these processes, with downregulation of NO-cGMP signaling having deleterious consequences. Targeted interventions aimed at upregulating this pathway may favorably alter the underlying coupled pathophysiology in the myocardium and the pulmonary vasculature and may thereby improve clinical outcomes.
BRL is the recipient of a Midwest Affiliate American Heart Association Clinical Research Program Grant (09CRP210070) and is also supported by NIH/NCRR Washington University-ICTS Grant (KL2 RR024994 and UL1 RRo24992). The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH.
The authors would like to acknowledge Dr. Michael Rich for his review and commentary.
BRL has no relevant disclosures.
AUTHOR CONTRIBUTIONS: BRL and MMC equally contributed to all phases of the development of this manuscript, including conceptualization of outline, literature search, writing, and editing.
DISCLOSURES: MMC has had consulting relationships with Actelion, Gilead, United Therapeutics and has received research support from Pfizer, Actelion, Lilly – ICOS, Gilead, United Therapeutics, and Norvartis.
Brian R Lindman, Division of Cardiology, Department of Internal Medicine, Washington University School of Medicine, St. Louis, MO (USA)
Murali M Chakinala, Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Washington University, St. Louis, MO (USA)