Pulmonary artery denervation for pulmonary arterial

hypertension

Andrew Constantine MBBS MA & Konstantinos Dimopoulos MD MSc PhD FESC

From the Adult Congenital Heart Centre and National Centre for Pulmonary Hypertension,

Royal Brompton Hospital, London, UK & the National Heart and Lung Institute, Imperial

College London, UK.

Correspondance to:

Professor Konstantinos Dimopoulos

Adult Congenital Heart Centre

Royal Brompton and Harefield NHS Foundation Trust

Sydney Street, London SW3 6NP, UK

Tel +44 2073 528121

E- mail: k.dimopoulos02@gmail.com

Manuscript word count:

3481 words (5529 including references)

Conflicts of interest:

Dr Constantine has received a personal educational grant from Actelion Pharmaceuticals (a

Janssen Pharmaceutical company). Professor Dimopoulos has received nonfinancial support

from Actelion Pharmaceuticals; and has been a consultant to and received grants and personal

fees from Actelion Pharmaceuticals, Pfizer, GlaxoSmithKline, and Bayer/MSD.

Abstract

Pulmonary arterial hypertension remains a progressive, life-limiting disease despite optimal

medical therapy. Pulmonary artery denervation has arisen as a novel intervention in the

treatment of pulmonary arterial hypertension, and other forms of pulmonary hypertension,

with the aim of reducing the sympathetic activity of the pulmonary circulation. Pre-clinical

studies and initial clinical trials have demonstrated that the technique can be performed safely

with some positive effects on clinical, haemodynamic and echocardiographic markers of

disease. The scope of the technique in current practice remains limited given the absence of

well-designed, large-scale, international randomised controlled clinical trials. This review

provides an overview of this exciting new treatment modality, including pathophysiology,

technical innovations and recent trial results.

Key words

Pulmonary artery denervation; pulmonary hypertension; pulmonary arterial hypertension

Abbreviations

6MWD – 6-minute walk distance

6-OHDA – 6-hydroxydopamine

BMPR-II, bone morphogenetic protein receptor type 2

BNP – Brain natriuretic peptide

Ca2+, calcium ion

CHD – congenital heart disease

CMR – cardiac magnetic resonance imaging

CTEPH – chronic thromboembolic pulmonary hypertension

DHMCT – dehydrogenized monocrotaline

EM – electron microscopic

IV – intravenous

mPAP – mean pulmonary arterial pressure

PA – pulmonary artery/arterial

PAH – pulmonary arterial hypertension

PADN – pulmonary artery denervation

PCWP – pulmonary capillary wedge pressure

PE – pulmonary embolism

PH – pulmonary hypertension

PVR – pulmonary vascular resistance

RAAS – renin-angiotensin-aldosterone-system

RHC – right heart catheterisation

RV – right ventricle/ventricular

SAE – severe adverse event

SNA – sympathetic nervous system activation

TGF-β – transforming growth factor beta

TxA2 – thromboxane A2

TPG – transpulmonary gradient

WHO – World Health Organisation

WU – Wood Unit(s)

Introduction

Pulmonary arterial hypertension (PAH) is characterised haemodynamically by the presence

of pre-capillary pulmonary hypertension (PH), with a mean pulmonary arterial pressure

(mPAP) of ≥25 mmHg, a pulmonary artery wedge pressure of ≤15 mmHg, and a pulmonary

vascular resistance (PVR) of >3 Wood units (WU) (1), although a mPAP of 21-24 mmHg is

abnormal and a cut off of >20 mmHg has recently been proposed for the haemodynamic

definition of PH (2). PAH is a clinical diagnosis consisting of various underlying clinical

entities, including idiopathic, heritable, drug or toxin-induced or associated with underlying

systemic disease. This latter group is formed mainly of patients with connective tissue disease

or congenital heart disease (CHD). Endothelial dysfunction with inappropriate

vasoconstriction, pulmonary vascular remodelling and in-situ thrombosis has been identified

as major pathophysiological mechanisms in all types of PAH. As disease develops and

progresses, the cross-sectional area of the distal pulmonary vessels is reduced, causing a rise

in PVR and pulmonary arterial (PA) pressure, which impact on right ventricular function.

Without therapy, the progressive increase in afterload and strain on the right ventricle (RV) is

associated with significant morbidity and mortality.

Several pharmacological therapies have been developed for PAH, focusing on the endothelin,

nitric oxide and prostacyclin pathways (3,4). These therapies have improved clinical

outcomes, and are nowadays often used in combination to maximise benefit (5). However,

many patients can have an inadequate response to treatment and fail to reach therapeutic

targets despite optimal medical therapy. In others, the disease may progress after an initial

positive response and a period of stability on PAH therapies. While appropriate candidates

with PAH failing medical therapy should be assessed for lung or heart-lung transplantation

(6,7), novel treatment strategies are needed to prolong longevity and improve quality of life

prior to transplantation, given the limited availability of donor organs. Moreover, there are

many patients not suitable for transplantation, who rely on treatment escalation and novel

therapeutic options for prolonging their life and limiting their symptoms. Developments in

our understanding of the cellular and molecular mechanisms of disease in PAH, beyond the

vasomotor pathways that are the targets of current vasodilator therapy, have highlighted the

role of emerging pathways amenable to therapeutic intervention (Figure 1) (8). Non-

pharmacological techniques are already in use for the treatment of pulmonary hypertension

(PH). Exercise-based rehabilitation programmes are safe and effective in PH, and can result

in clinically relevant improvements in exercise capacity (9). Invasive treatment strategies

include percutaneous or surgical creation of pulmonary-to-systemic shunts aimed at

offloading the RV and enhancing cardiac output (atrial septostomy, or a Potts shunt between

the descending aorta and left PA) (10–12), and balloon pulmonary angioplasty for patients

with CTEPH (13,14). Various forms of endovascular autonomic system modification have

been proposed in animal models of PH, including sympathetic ganglion blockade and

catheter-based renal denervation (15–18). In recent years, the autonomic function of the

pulmonary vasculature has attracted further attention and most recently this has culminated in

numerous human trials of PA denervation (PADN) (19–25). This treatment has the potential

to add considerably to the management of PH, where novel therapies are in short supply. This

review follows the evolution of the technique, from the initial experiments in animal models

to the latest multi-centre trials. We assess the state of the evidence supporting the adoption of

PADN in clinical practice and review the unanswered questions in this young field.

Autonomic dysregulation in pulmonary hypertension

The pulmonary vasculature receives a rich autonomic nerve supply, with sympathetic

(predominantly), parasympathetic and sensory nerve fibres (26). The anatomy of the

innervation of the lungs and pulmonary vasculature is well documented. Over the past half a

century, our understanding of the role of the autonomic nervous system in health has

increased significantly, with elucidation of central and local reflexes and their interaction

with humoral and local vasoactive mechanisms of control. The role of the nervous system in

disease states, however, is not well understood.

The sympathetic nerve supply to the pulmonary vessels arises from neuronal cell bodies in

the middle and inferior cervical ganglia and the first 5 thoracic ganglia (27). The post-

ganglionic fibres of these nerves meet parasympathetic nerve fibres to form the anterior and

posterior plexi at the carina. From here, nerve fibres enter the lungs forming a peribronchial

plexus, which innervates the bronchial tree, and a periarterial plexus, which runs in the

adventitial layer and innervates the pulmonary vasculature. Although the density and extent

of the sympathetic nerve supply varies significantly between species (for example, adrenergic

nerve fibres are absent in the intrapulmonary arteries of the rat) (28) in humans the

periarterial plexus extends to small PAs of <100µm diameter (29). As in other medium-to-

large vessels, β2-adrenoceptors predominate (30) with evidence of co-existing β1- (31) and α-

adrenoceptors. In the parasympathetic system, descending pre-ganglionic nerves arise from

the brainstem and pass as the pulmonary branches of the vagus nerve to the pulmonary plexus

at the lung roots, where they congregate with postganglionic sympathetic fibres. The

parasympathetic fibres either synapse with the cells of the ganglion or continue in the walls

of the bronchial and arterial trees to the target organs, where they synapse with post-

ganglionic neurons. Of note, the adipose and connective tissues surrounding the pulmonary

trunk are also richly innervated with sympathetic nerve fibres (32). These sympathetic nerve

fibres are in close proximity to the cardiac autonomic supply and the right phrenic nerve,

especially in PAH patients with passive PA dilatation. Thus, any therapy that targets the

sympathetic supply to the lungs is at risk of adversely affecting the nerve supply to adjacent

structures.

In health, pulmonary vascular tone is carefully regulated by a complex network of neural and

humoral factors, which interact with the haemodynamic loads imposed by the systemic

circulation. This results in optimal ventilation-perfusion matching in the pulmonary

circulation, both at rest and during exertion, aimed at maintaining normoxia and oxygen

delivery to peripheral tissues (26). Central and local reflex mechanisms, mediated by

autonomic efferents, are partly responsible for this tight regulation of pulmonary vascular

tone. For example, stimulation of chemoreceptors in the carotid or aortic bodies has a

variable effect on PVR depending on experimental conditions (33,34). It has also been shown

that distension of the large PAs produces vasoconstriction of the distal PAs, mediated by a

local pulmonary reflex involving baroreceptors in the PA bifurcation and along the branch

PAs, with afferent fibres in the adventitia of the large vessels and effector fibres in the muscle

layer (35).

The role of these mechanisms in disease states, including interactions with local mediators,

such as nitric oxide, endothelin and thromboxane, and humoral mechanisms remains unclear.

Indeed, there is overwhelming evidence that autonomic dysregulation is not the major

pathophysiological driver of PAH; rather endothelial dysfunction, excessive proliferation of

PA smooth muscle and a resultant vasoconstriction are the major disease mechanisms. As

pulmonary vascular disease progresses, however, the elevation in PVR and the increased

afterload on the RV lead to an initial compensatory sympathetic nervous system activation

(SNA) and an increase in renin-angiotensin-aldosterone-system (RAAS) activity. As in left

ventricular failure, chronic activation of these systems may be detrimental over a longer

timeframe (36). Chemoreflex mediated SNA (37), reduced heart rate variability (38,39), and

increased cardiac SNA (40), all markers of chronic sympathetic activation, have been

reported in PAH. Increased circulating plasma noradrenaline levels are a variable finding,

having been reported by some investigators (40,41), but not others (37,41,42). SNA is

associated with clinical deterioration in PAH (43), and atrial septostomy (44) was able to

reduce SNA in these patients. Preclinical studies of β-adrenergic receptor blockade in two rat

models of PH (45,46) have provided evidence of attenuated RV remodelling and

improvement in RV function. In a clinical study of beta-blockade in idiopathic PAH,

however, bisoprolol reduced heart rate along with cardiac index and 6-minute walk distance

(6MWD) (47). The possibility of negative inotropy and chronotropy associated with these

agents in patients with already reduced and relatively fixed stroke volumes is a concern (48).

The international PH guidelines do not recommend beta-blockade for PAH unless this is

required by co-morbid conditions e.g. coronary artery disease (1).

RAAS activation in PAH has been a more recent subject of research. Clinical studies have

demonstrated increased plasma levels of renin and angiotensin I and II, along with signs of

local upregulation of angiotensin I receptor expression and signalling in PA smooth muscle

cells in PAH patients (49). Pharmacological modulation of the RAAS, by angiotensin

antagonism, and mineralocorticoid receptor antagonism, has been trialled successfully in

several animal models of PAH. These interventions have led to reduced muscularisation of

small PAs, RV afterload reduction and improved RV ventricular-arterial coupling (50–52).

However, these have not translated into benefit in human trials and the use of angiotensin-

converting enzyme inhibitors and angiotensin-2 receptor antagonists is not recommended in

patients with PAH unless required by comorbidities (1).

Pulmonary artery denervation in animal models of PH

While SNA cannot be targeted pharmacologically in humans, interventional options focusing

on interrupting the sympathetic nerve supply to the PAs may provide a feasible alternative. In

1980, Juratsch et al. (53) performed experiments on a canine model of acute PH, produced by

balloon distension of the main PA. Surgical or chemical denervation of the PA significantly

reduced acute elevations in mPAP produced in this model. This formed the basis of

intervention to disrupt the autonomic nerve supply to the lung.

More recently, percutaneous PADN has been demonstrated in a canine model of acute PH

(54) and animal models of chronic PH. These studies have demonstrated that the innervation

of the pulmonary vasculature stems from large nerve bundles densely packed in the adventitia

of vessels around the PA bifurcation, as described above. It is therefore theoretically possible

to achieve PADN using high-frequency alternating current applied to these sites. These

studies also provided histological evidence of the effects of PADN on nerve endings,

associated with significant improvements in pulmonary haemodynamics, reduced

muscularisation of the PAs and lower RV mass (55,56). In a porcine model of PH secondary

to left heart disease induced by aortic banding, surgical and chemical PADN were associated

with reduced muscularisation of pulmonary arterioles, haemodynamic improvements and

changes in the concentration of adrenoceptors within pulmonary tissue compared to controls

(57).

Some investigators have questioned the ability of percutaneous PADN, as opposed to a

surgical approach, to achieve sufficient denervation (32,58). In an elegant translational study

comparing surgical and percutaneous PADN, Garcia-Lunar et al. (58) reported that PADN by

radiofrequency ablation produced incomplete denervation, with only focal damage to

adventitial nerve fibres. By contrast, surgical PADN was associated with histological

evidence of widespread fibrotic nerve fibres with absent nerve axons in the PA adventitia. It

is obvious that adequate denervation will depend on numerous parameters including the

distribution and depth of the sympathetic nerves in individual patients, the thickness of the

PA wall and the energy that is delivered to these structures.

Pulmonary artery denervation techniques

Multiple techniques for PADN have been described. Surgical PADN via left lateral

thoracotomy, using surgical bipolar radiofrequency clamps applied to the PA bifurcation and

proximal left and right PA branches has been described in pigs (58). Although this process

has been shown to produce more complete denervation than a transcatheter approach,

exposing PAH patients, especially those who have not responded adequately to medical

therapy, to a general anaesthetic and thoracic surgery is a risky and unattractive strategy.

Human studies have, therefore, mainly employed 2 different catheter-based approaches:

radiofrequency ablation (19) and high-energy endovascular ultrasound (Figure 2) (20). In

both, the catheter is positioned in the main PA via a peripheral vein, and energy is applied to

the nerve fibres in the adventitia of the PA wall. PADN trials to date have utilised empirical

strategies for denervation, based on anatomical (fluoroscopic) landmarks. More recently, it

has been suggested that targeted PADN, aimed at areas of autonomic nerve activity, may

improve the efficacy of PADN, while reducing the risks associated with extensive ablation of

the pulmonary trunk (59). In atrial fibrillation ablation, combined computed tomography

nuclear imaging with cardiac cadmium-zinc-telluride cameras (e.g. D-SPECT) have been

used to map ganglionated plexi within the heart, allowing directed denervation of these

structures; this could be applied to PADN. An alternative technique to determine ablation

sites has recently been described in a case study of PADN (21): autonomic nervous responses

were induced by high-output burst electrical stimulation and the anatomical site was mapped

onto a pre-procedural 3-dimensional computed tomography image. Sites where stimulation

induced bradycardia (suggestive of autonomic supply modification) were targeted, whereas

those inducing cough or diaphragmatic twitching were avoided. Mapping of the autonomic

activity around the PAs and the response to PADN may also provide further valuable

information on the mechanisms of action of this technique and the role of the autonomic

nervous system in PAH.

Clinical trials of pulmonary artery denervation

There is preliminary evidence of safety and efficacy of PADN in humans, including patients

with pre-capillary PH (PAH) , combined pre- and post-capillary PH and a handful of patients

with PH secondary to left heart disease and CTEPH (19–25).

In 2013, Chen et al. presented the first-in-man experience with PADN. The PADN-1 study

recruited 13 patients with idiopathic PAH not responding adequately to medical therapy (19).

Eight patients who refused the procedure formed the “control” arm. This was then extended

into a phase II clinical trial reported in 2015 (22), which included 66 patients with PH of

various aetiologies (39 patients with PAH, 18 patients with PH secondary to left heart

disease, and 9 patients with CTEPH) and no control arm. PADN was undertaken by

endovascular radio-frequency ablation. Chest pain was common during the procedure,

reported by 71% of the study participants. At 12 months of follow-up, the authors reported

remarkable improvements in clinical, echocardiographic and haemodynamic variables, with a

94m increase in 6MWD, a reduction in mPAP from 41 to 36mmHg and a favourable change

in RV function (Tei index decreased from 0.63 to 0.39). This preliminary finding needs to be

replicated in larger clinical trials. This study raised some questions regarding the trial

population and design. Many patients in the phase II study were on background medication

that do not represent the current standards of care; 89% of those with PH secondary to left

heart disease were on a prostacyclin analogue. Patients undergoing PADN were withdrawn

from medical PAH therapy, while the vast majority of patients continued on long-term

oxygen therapy. In an accompanying editorial, Galiè and Manes (60) suggest that the

continuation of oxygen therapy may have blunted the hypoxic vasoconstrictive reflex that is

important in maintaining pulmonary ventilation/perfusion matching, and hence may have

influenced the results.

The preliminary results of a phase I multi-national safety study (TROPHY1) were recently

reported (61). A total of 14 patients from 5 study centres in Europe and Israel were enrolled

in this open-label study. PADN was performed in PAH patients who were receiving

combination therapy, using high-frequency, intravascular ultrasound at and around the PA

bifurcation. At 4 months there were no procedure-related serious adverse events (defined as

PA arterial perforation, aneurysm, dissection, stenosis and thrombus formation, haemoptysis,

or death). The investigators also studied secondary efficacy endpoints. An acute change in

pulmonary haemodynamics was not observed following PADN as opposed to previous trials.

At follow-up catheterisation, a very small but statistically significant change in PVR (9.4 vs.

8WU, p<0.01) was accompanied by a fall in mPAP (52.7mmHg to 44.5mmHg, p=0.01)

without significant changes in 6MWD. Since this was a non-randomised study consisting of

only 14 patients, larger, controlled trials are needed to fully appreciate the safety and efficacy

of this procedure. The lack of an acute effect followed by significant longer-term effects,

however, raises questions about the mechanisms underlying PADN using this procedure.

PADN was also recently tested in patients with combined pre- and post-capillary PH in the

PADN-5 study (23). In this single-country multi-centre trial, patients presenting with new-

onset heart failure were randomised to PADN (n = 48) or sham denervation plus sildenafil

therapy (n = 50). Exercise capacity improved markedly in the intervention arm, with the

6MWD increasing by 83m compared to 15m in the sildenafil arm (p<0.001). There was a

concomitant fall in PVR (6.4±3.2 to 4.2±1.5WU, p<0.001) and PA wedge pressure (22.2±6.6

to 16.1±6.2mmHg, p=0.01) in the PADN arm. The results of this ground-breaking trial

should be interpreted keeping in mind various aspects of the study design. Patients were

recruited following an acute hospital contact and uptitration of new heart failure medication

took place during the study period. The use of sildenafil for patients with post-capillary PH is

not standard practice and only acts to detract from the purity of the control group. The role

that pulmonary vasoconstriction plays in combined pre- and post-capillary PH and the effect

of PADN on PA wedge pressure in this population again raise intriguing questions

concerning the mechanism of action of PADN.

These pioneering trials have been successful in demonstrating the safety and feasibility of

this procedure, with positive, yet not conclusive, signals of efficacy. These studies raise as

many questions as they answer, which need to be appropriately addressed in large rigorously

designed multi-national randomised controlled trials investigating robust efficacy end-points,

before PADN can be recommended for routine clinical use.

Unanswered questions and future research challenges

Many novel therapies for PAH have proven efficacious in animal models and small clinical

trials, but not in human studies and have not, thus, entered clinical practice. PADN has shown

promise as a new treatment modality, but carefully designed trials are necessary to further

elucidate the technique’s mechanisms of action, long-term safety and efficacy, and its role in

the current management strategy for PAH and other types of PH. Even if PADN does reverse

sympathetic-dependent vasoconstriction, how can it positively influence the severe, fixed,

obstructive lesions that are typically observed on lung histology in PAH? PADN may have

effects beyond regulating vasomotor response, but these remain to be elucidated.

Evidence to date suggests that PADN does not only act via vasoconstriction of the pulmonary

vasculature. The cardiac autonomic nerves and parasympathetic supply to the lungs are

located close to sympathetic nerves at the level of the main PAs, especially in patients with

PA dilatation, commonly encountered in PAH. This raises the question of whether PADN

modified cardiac autonomic innervation and haemodynamics apart from its effect on the

pulmonary vasculature. The change in heart rate following PADN is not routinely reported,

but negative chronotropy following PADN has been described in some patients (22);

parasympathetic denervation may provide an alternative explanation for the benefit of the

procedure on pulmonary vascular remodelling and RV function (62), although this finding is

not consistent. Furthermore, the significant reduction in PA wedge pressure reported in the

PADN-5 study is also difficult to explain through the effect of PADN on pulmonary

haemodynamics. Differences between species and models of pulmonary hypertension mean

that caution is required when extrapolating the effects of PADN from animal models to

human subjects. The effects of PADN on left and right heart haemodynamics, combined with

methods to study the anatomy and physiology of the autonomic system during PADN, should

be a focus of future studies and may allow refinement of current techniques.

The longevity of PADN using different techniques also remains unknown. In other scenarios,

such as following heart or lung transplantation or arterial switch procedure for transposition

of the great arteries, there is often complete and immediate external cardiac or pulmonary

denervation. Following heart transplantation, for example, parasympathetic vagal neurons

and post-ganglionic sympathetic nerve fibres are transected as they pass from the stellate

ganglion to the myocardium. There is evidence of variable degrees of delayed re-innervation,

occurring in 40-70% of recipients in the late post-transplant period (> 1-year post-transplant)

assessed using Iodine-123 metaiodobenzylguanidine imaging and immunohistochemistry.

Reinnervation can have an impact on functional variables such as exercise tolerance (63,64).

To examine whether delayed reinnervation occurs following PADN, longer term follow-up

data is needed; to date, only one study has shown a sustained effect at 12 months (22).

Future clinical trials of PADN should be adequately powered multi-centre multi-national

ventures based in specialist PH centres. Strict inclusion and exclusion criteria need to be

applied, recruiting patients with similar pathophysiology (e.g. idiopathic PAH only), stable

on combination PAH therapy to avoid bias from rapidly deteriorating disease or escalation of

other concomitant treatments. The control arm should be carefully selected and should only

differ from the active arm in terms of the PADN procedure itself, preferably using sham

procedures. All patients in the trial should be receiving standard treatment as per international

guidelines, to make the results interpretable and applicable worldwide.

Conclusions

PADN has generated some excitement in the PH community, with early positive signs in

non-randomised studies. However, questions remain on the short and long-term efficacy of

PADN, its mechanism of action and appropriate patient selection and timing. There is still a

long way to go before PADN can become part of routine clinical care.

References

1. Galiè N, Humbert M, Vachiery J-L, Gibbs S, Lang I, Torbicki A, et al. 2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension: The Joint Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS)Endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC), International Society for Heart and Lung Transplantation (ISHLT). European Respiratory Journal. 2015 Oct;46(4):903–75.

2. Simonneau G, Montani D, Celermajer DS, Denton CP, Gatzoulis MA, Krowka M, et al. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. European Respiratory Journal [Internet]. 2018 Jan 1 [cited 2019 Nov 19]; Available from: https://erj.ersjournals.com/content/early/2018/10/11/13993003.01913-2018

3. Humbert M, Lau EMT, Montani D, Jaïs X, Sitbon O, Simonneau G. Advances in therapeutic interventions for patients with pulmonary arterial hypertension. Circulation. 2014 Dec 9;130(24):2189–208.

4. McLaughlin VV, Shah SJ, Souza R, Humbert M. Management of Pulmonary Arterial Hypertension. Journal of the American College of Cardiology. 2015 May 12;65(18):1976–97.

5. Bai Y, Sun L, Hu S, Wei Y. Combination therapy in pulmonary arterial hypertension: a meta-analysis. Cardiology. 2011;120(3):157–65.

6. Fadel E, Mercier O, Mussot S, Leroy-Ladurie F, Cerrina J, Chapelier A, et al. Long-term outcome of double-lung and heart-lung transplantation for pulmonary hypertension: a comparative retrospective study of 219 patients. Eur J Cardiothorac Surg. 2010 Sep;38(3):277–84.

7. Granton J, Mercier O, De Perrot M. Management of severe pulmonary arterial hypertension. Semin Respir Crit Care Med. 2013 Oct;34(5):700–13.

8. Schermuly RT, Ghofrani HA, Wilkins MR, Grimminger F. Mechanisms of disease: pulmonary arterial hypertension. Nature Reviews Cardiology. 2011 Aug;8(8):443–55.

9. Morris NR, Kermeen FD, Holland AE. Exercise-based rehabilitation programmes for pulmonary hypertension. Cochrane Database Syst Rev. 2017 19;1:CD011285.

10. Kerstein D, Levy PS, Hsu DT, Hordof AJ, Gersony WM, Barst RJ. Blade balloon atrial septostomy in patients with severe primary pulmonary hypertension. Circulation. 1995 Apr 1;91(7):2028–35.

11. Blanc J, Vouhé P, Bonnet D. Potts shunt in patients with pulmonary hypertension. N Engl J Med. 2004 Feb 5;350(6):623.

12. Law MA, Grifka RG, Mullins CE, Nihill MR. Atrial septostomy improves survival in select patients with pulmonary hypertension. Am Heart J. 2007 May;153(5):779–84.

13. Feinstein JA, Goldhaber SZ, Lock JE, Ferndandes SM, Landzberg MJ. Balloon pulmonary angioplasty for treatment of chronic thromboembolic pulmonary hypertension. Circulation. 2001 Jan 2;103(1):10–3.

14. Sugimura K, Fukumoto Y, Satoh K, Nochioka K, Miura Y, Aoki T, et al. Percutaneous transluminal pulmonary angioplasty markedly improves pulmonary hemodynamics and long-term prognosis in patients with chronic thromboembolic pulmonary hypertension. Circ J. 2012;76(2):485–8.

15. Na Sungwon, Kim Ok Soo, Ryoo Sungwoo, Kweon Tae Dong, Choi Yong Seon, Shim Hyo Sup, et al. Cervical Ganglion Block Attenuates the Progression of Pulmonary Hypertension via Nitric Oxide and Arginase Pathways. Hypertension. 2014 Feb 1;63(2):309–15.

16. Qingyan Z, Xuejun J, Yanhong T, Zixuan D, Xiaozhan W, Xule W, et al. Beneficial Effects of Renal Denervation on Pulmonary Vascular Remodeling in Experimental Pulmonary Artery Hypertension. Rev Esp Cardiol (Engl Ed). 2015 Jul;68(7):562–70.

17. Hu W, Yu S-B, Chen L, Guo R-Q, Zhao Q-Y. Renal sympathetic denervation prevents the development of pulmonary arterial hypertension and cardiac dysfunction in dogs. Kaohsiung J Med Sci. 2015 Aug;31(8):405–12.

18. Liu Q, Song J, Lu D, Geng J, Jiang Z, Wang K, et al. Effects of renal denervation on monocrotaline induced pulmonary remodeling. Oncotarget. 2017 Feb 7;8(29):46846–55.

19. Chen S-L, Zhang F-F, Xu J, Xie D-J, Zhou L, Nguyen T, et al. Pulmonary Artery Denervation to Treat Pulmonary Arterial Hypertension: The Single-Center, Prospective, First-in-Man PADN-1 Study (First-in-Man Pulmonary Artery Denervation for Treatment of Pulmonary Artery Hypertension). Journal of the American College of Cardiology. 2013 Sep 17;62(12):1092–100.

20. Rothman A. Therapeutic intravascular ultrasound pulmonary artery denervation for the treatment of pulmonary arterial hypertension (TROPHY1): a multicentre, international, open-label trial. EuroPCR 2019; 2019 May 22.

21. Fujisawa T, Kataoka M, Kawakami T, Isobe S, Nakajima K, Kunitomi A, et al. Pulmonary Artery Denervation by Determining Targeted Ablation Sites for Treatment of Pulmonary Arterial Hypertension. Circ Cardiovasc Interv. 2017;10(10).

22. Chen S-L, Zhang H, Xie D-J, Zhang J, Zhou L, Rothman AMK, et al. Hemodynamic, functional, and clinical responses to pulmonary artery denervation in patients with pulmonary arterial hypertension of different causes: phase II results from the Pulmonary Artery Denervation-1 study. Circ Cardiovasc Interv. 2015 Nov;8(11):e002837.

23. Zhang H, Zhang J, Chen M, Xie D-J, Kan J, Yu W, et al. Pulmonary Artery Denervation Significantly Increases 6-Min Walk Distance for Patients With Combined Pre- and Post-Capillary Pulmonary Hypertension Associated With Left Heart Failure: The PADN-5 Study. JACC Cardiovasc Interv. 2019 Feb 11;12(3):274–84.

24. Kiuchi MG, Andrea BR, da Silva GR, Coelho SBP, Paz LMR, Chen S, et al. Pulmonary artery ablation to treat pulmonary arterial hypertension: a case report. Journal of Medical Case Reports. 2015 Dec 16;9(1):284.

25. Zhang H, Zhang J, Xie D-J, Jiang X, Zhang F-F, Chen S-L. Pulmonary artery denervation for treatment of a patient with pulmonary hypertension secondary to left heart disease. Pulm Circ. 2016 Jun;6(2):240–3.

26. Barnes PJ, Liu SF. Regulation of pulmonary vascular tone. Pharmacol Rev. 1995 Mar;47(1):87–131.

27. Richardson JB. Nerve supply to the lungs. Am Rev Respir Dis. 1979 May;119(5):785–802.

28. Bradley DE, McNary WF, el-Bermani AW. The distribution of acetylcholinesterase and catecholamine containg nerves in the rat lung. Anat Rec. 1970 Jun;167(2):205–7.

29. Kai Y. Study on the distribution of symatheic nerves in the lung using the Falck-Hillarp’s fluorescent method. Bull Chest Dis Res Inst Kyoto Univ. 1969 Mar;2(2):225-245 passim.

30. Carstairs JR, Nimmo AJ, Barnes PJ. Autoradiographic visualization of beta-adrenoceptor subtypes in human lung. Am Rev Respir Dis. 1985 Sep;132(3):541–7.

31. Boe J, Simonsson BG. Adrenergic receptors and sympathetic agents in isolated human pulmonary arteries. Eur J Respir Dis. 1980 Aug;61(4):195–202.

32. Huang Yuan, Liu Yi-Wei, Pan Hai-Zhou, Zhang Xiao-Ling, Li Jun, Xiang Li, et al. Transthoracic Pulmonary Artery Denervation for Pulmonary Arterial Hypertension. Arteriosclerosis, Thrombosis, and Vascular Biology. 2019 Apr 1;39(4):704–18.

33. Aviado DM, Ling JS, Schmidt CF. Effects of anoxia on pulmonary circulation: reflex pulmonary vasoconstriction. Am J Physiol. 1957 May;189(2):253–62.

34. Fitzgerald RS, Dehghani GA, Sham JS, Shirahata M, Mitzner WA. Peripheral chemoreceptor modulation of the pulmonary vasculature in the cat. J Appl Physiol. 1992 Jul;73(1):20–9.

35. Osorio J, Russek M. Reflex changes on the pulmonary and systemic pressures elicited by stimulation of baroreceptors in the pulmonary artery. Circ Res. 1962 Apr;10:664–7.

36. de Man FS, Handoko ML, Guignabert C, Bogaard HJ, Vonk-Noordegraaf A. Neurohormonal Axis in Patients with Pulmonary Arterial Hypertension. Am J Respir Crit Care Med. 2013 Jan 1;187(1):14–9.

37. Velez-Roa S, Ciarka A, Najem B, Vachiery J-L, Naeije R, van de Borne P. Increased sympathetic nerve activity in pulmonary artery hypertension. Circulation. 2004 Sep 7;110(10):1308–12.

38. McGowan CL, Swiston JS, Notarius CF, Mak S, Morris BL, Picton PE, et al. Discordance between microneurographic and heart-rate spectral indices of sympathetic activity in pulmonary arterial hypertension. Heart. 2009 May;95(9):754–8.

39. Wensel R, Jilek C, Dörr M, Francis DP, Stadler H, Lange T, et al. Impaired cardiac autonomic control relates to disease severity in pulmonary hypertension. Eur Respir J. 2009 Oct;34(4):895–901.

40. Mak S, Witte KK, Al-Hesayen A, Granton JJ, Parker JD. Cardiac sympathetic activation in patients with pulmonary arterial hypertension. Am J Physiol Regul Integr Comp Physiol. 2012 May 15;302(10):R1153-1157.

41. Nootens M, Kaufmann E, Rector T, Toher C, Judd D, Francis GS, et al. Neurohormonal activation in patients with right ventricular failure from pulmonary hypertension: Relation to hemodynamic variables and endothelin levels. Journal of the American College of Cardiology. 1995 Dec 1;26(7):1581–5.

42. Richards AM, Ikram H, Crozier IG, Nicholls MG, Jans S. Ambulatory pulmonary arterial pressure in primary pulmonary hypertension: variability, relation to systemic arterial pressure, and plasma catecholamines. Br Heart J. 1990 Feb;63(2):103–8.

43. Ciarka A, Doan V, Velez-Roa S, Naeije R, van de Borne P. Prognostic Significance of Sympathetic Nervous System Activation in Pulmonary Arterial Hypertension. Am J Respir Crit Care Med. 2010 Jun 1;181(11):1269–75.

44. Ciarka A, Vachièry J-L, Houssière A, Gujic M, Stoupel E, Velez-Roa S, et al. Atrial Septostomy Decreases Sympathetic Overactivity in Pulmonary Arterial Hypertension. Chest. 2007 Jun 1;131(6):1831–7.

45. Bogaard HJ, Natarajan R, Mizuno S, Abbate A, Chang PJ, Chau VQ, et al. Adrenergic receptor blockade reverses right heart remodeling and dysfunction in pulmonary hypertensive rats. Am J Respir Crit Care Med. 2010 Sep 1;182(5):652–60.

46. de Man FS, Handoko ML, van Ballegoij JJM, Schalij I, Bogaards SJP, Postmus PE, et al. Bisoprolol delays progression towards right heart failure in experimental pulmonary hypertension. Circ Heart Fail. 2012 Jan;5(1):97–105.

47. van Campen JSJA, de Boer K, van de Veerdonk MC, van der Bruggen CEE, Allaart CP, Raijmakers PG, et al. Bisoprolol in idiopathic pulmonary arterial hypertension: an explorative study. Eur Respir J. 2016;48(3):787–96.

48. Peacock A, Ross K. Pulmonary hypertension: a contraindication to the use of {beta}-adrenoceptor blocking agents. Thorax. 2010 May;65(5):454–5.

49. de Man FS, Tu L, Handoko ML, Rain S, Ruiter G, François C, et al. Dysregulated renin-angiotensin-aldosterone system contributes to pulmonary arterial hypertension. Am J Respir Crit Care Med. 2012 Oct 15;186(8):780–9.

50. Rouleau JL, Kapuku G, Pelletier S, Gosselin H, Adam A, Gagnon C, et al. Cardioprotective effects of ramipril and losartan in right ventricular pressure overload in the rabbit: importance of kinins and influence on angiotensin II type 1 receptor signaling pathway. Circulation. 2001 Aug 21;104(8):939–44.

51. Okada M, Harada T, Kikuzuki R, Yamawaki H, Hara Y. Effects of telmisartan on right ventricular remodeling induced by monocrotaline in rats. J Pharmacol Sci. 2009 Oct;111(2):193–200.

52. Maron BA, Zhang Y-Y, White K, Chan SY, Handy DE, Mahoney CE, et al. Aldosterone inactivates the endothelin-B receptor via a cysteinyl thiol redox switch to decrease pulmonary endothelial nitric oxide levels and modulate pulmonary arterial hypertension. Circulation. 2012 Aug 21;126(8):963–74.

53. Juratsch CE, Jengo JA, Castagna J, Laks MM. Experimental Pulmonary Hypertension Produced by Surgical and Chemical Denervation of the Pulmonary Vasculature. Chest. 1980 Apr 1;77(4):525–30.

54. Chen S-L, Zhang Y-J, Zhou L, Xie D-J, Zhang F-F, Jia H-B, et al. Percutaneous pulmonary artery denervation completely abolishes experimental pulmonary arterial hypertension in vivo. EuroIntervention. 2013 Jun 22;9(2):269–76.

55. Zhou L, Zhang J, Jiang X-M, Xie D-J, Wang J-S, Li L, et al. Pulmonary Artery Denervation Attenuates Pulmonary Arterial Remodeling in Dogs With Pulmonary Arterial Hypertension Induced by Dehydrogenized Monocrotaline. JACC Cardiovasc Interv. 2015 Dec 28;8(15):2013–23.

56. Rothman AMK, Arnold ND, Chang W, Watson O, Swift AJ, Condliffe R, et al. Pulmonary artery denervation reduces pulmonary artery pressure and induces histological changes in an acute porcine model of pulmonary hypertension. Circ Cardiovasc Interv. 2015 Nov;8(11):e002569.

57. Zhang H, Yu W, Zhang J, Xie D, Gu Y, Ye P, et al. Pulmonary artery denervation improves hemodynamics and cardiac function in pulmonary hypertension secondary to heart failure. Pulm Circ. 2019 Jun;9(2):2045894018816297.

58. Garcia-Lunar I, Pereda D, Santiago E, Solanes N, Nuche J, Ascaso M, et al. Effect of pulmonary artery denervation in postcapillary pulmonary hypertension: results of a randomized controlled translational study. Basic Res Cardiol. 2019 11;114(2):5.

59. Dimopoulos K, Ernst S, McCabe C, Kempny A. Pulmonary Artery Denervation: A New, Long-Awaited Interventional Treatment for Combined Pre- and Post-Capillary Pulmonary Hypertension? JACC Cardiovasc Interv. 2019 Feb 11;12(3):285–8.

60. Galiè N, Manes A. New Treatment Strategies for Pulmonary Arterial Hypertension: Hopes or Hypes?∗. Journal of the American College of Cardiology. 2013 Sep 17;62(12):1101–2.

61. Rothman AMK, Vachiery JL, Howard L, Lang I, Avriel A, Jonas M, et al. Pulmonary artery denervation for the treatment of pulmonary arterial hypertension: preliminary results of the TROPHY 1 Study. Eur Heart J [Internet]. 2018 Aug 1 [cited 2019 Nov 26];39(suppl_1). Available from: https://academic.oup.com/eurheartj/article/39/suppl_1/ehy564.P567/5081677

62. Kuebler WM, Friedberg MK. Letter by Kuebler and Friedberg Regarding Article, ‘Pulmonary Artery Denervation by Determining Targeted Ablation Sites for Treatment of Pulmonary Arterial Hypertension’. Circ Cardiovasc Interv. 2018;11(2):e006148.

63. Parry DS, Foulsham L, Jenkins G, Wharton J, Marron K, Banner N, et al. Incidence and functional significance of sympathetic reinnervation after cardiac transplantation. Transplant Proc. 1997 Mar;29(1–2):569–70.

64. Buendia-Fuentes F, Almenar L, Ruiz C, Vercher JL, Sánchez-Lázaro I, Martínez-Dolz L, et al. Sympathetic Reinnervation 1 Year After Heart Transplantation, Assessed Using Iodine-123 Metaiodobenzylguanidine Imaging. Transplantation Proceedings. 2011 Jul 1;43(6):2247–8.

Figure legends

Figure 1: Known and emerging pathways involved in the pathogenesis of PAH (left), with

associated putative therapeutic targets (right). Current evidence-based therapies target

dysregulation of vascular tone through modulation of vasoactive mediators, promoting

vasodilatation. Abbreviations: BMPR-II, bone morphogenetic protein receptor type 2; Ca2+,

calcium ion; PVR, pulmonary vascular resistance; RAAS, renin-aldosterone-angiotensin

system; TGF-β, transforming growth factor beta.

Figure 2: Schematic of the mode of action of two available PADN catheters. Catheter cross-

sections are shown in black within the pulmonary artery, along with the fluoroscopic

appearance of the catheter tips (insets). In ‘A’, the circular tip of a dedicated 7.5F temperature

sensing and ablation catheter is displayed. Radiofrequency ablation is performed sequentially,

through each of the 10 pre-mounted electrodes, resulting in a circumferential energy

distribution (grey arrows). In ‘B’, a 6F multidirectional intra-vascular ultrasound catheter is

shown. This system (TIVUS, Cardiosonic) uses high-frequency, high-intensity

multidirectional ultrasound to thermally damage the target tissue (golden arrows) and does

not require contact with the vessel wall.

Figure 1

Figure 2

Table 1

Summary of selected pre-clinical and clinical studies of pulmonary artery denervation.

Reference Study type PH model Treatment n (PADN/control)

Mean age

Follow-up

Major inclusion criteria Endpoints Key findings

Pre-clinical

Juratsch et al. 1980 (53)

Prospective case series

Canine, balloon inflation in main PA

Surgical denervation or chemical denervation using IV 6-OHDA

13 / - - 24 hours - PAP, PVR Surgical and chemical PADN reduced or abolished the elevation in PAP produced by balloon distension of main PA. Effects of balloon distension not affected by cervical vagotomy.

Chen et al. 2013 (54)

Prospective case series

Canine, left pulmonary distal basal trunk or interlobar artery occlusion

Percutaneous PADN (7.5F catheter with circular tip, 10 electrodes, fluoroscopy-guided RF ablation)

20 / - - Acute - Invasive haemodynamic measurements

Occlusion of the left pulmonary interlobar artery induced a rise in PA and right ventricular pressure.PADN abolished the PAP response to balloon occlusion.

Rothman et al. 2015 (56)

Prospective non-randomized, sham-controlled trial

Porcine, TxA2 challenge pre- and post-PADN

Percutaneous PADN (6Fr spiral catheter, 1 electrode, fluoroscopy-guided RF ablation)

5 / 3 - Acute - Invasive haemodynamic measurements, changes in microscopy and histological staining

PADN reduced the mPAP and PVR response to the TxA2 challenge.PADN induced acute microscopic and histological changes (intimal disruption and thrombus, reduced medial thickness, altered adventitial architecture, reduced expression of nerve-associated S100 protein)

Zhou et al. 2015 (55)

Prospective randomized, sham-controlled trial

Canine, intra-atrial N-dimethylacetamide or DHMCT

Percutaneous PADN (7Fr catheter, fluoroscopy-guided RF ablation)

10 / 10(SN

substudy20 / 15)

- 14 weeks - Group A: Haemodynamic measurements, PA remodelling Group B: SN injury, SN conduction velocity, EM measurements

PADN reduced mPAP and right atrial pressure, increased cardiac output, and was associated with less RV hypertrophy.PADN induced SN demyelination, axon loss and slowing of SN conduction velocity. PADN was associated with reduced muscularisation of small PAs and

attenuated upregulation of growth factors, induced by DHMCT

Zhang et al. 2018 (57)

Prospective randomized, sham-controlled trial

Rat model, supracoronary aortic banding

Surgical (longitudinal damage to vessel nerves) and chemical (10% phenol applied to nerve surface) PADN

6 / 7 - 4 weeks; 6 months

- Haemodynamic and echocardiographic indices, histological staining

Surgical and chemical PADN improved PA haemodynamics, RV functional indices, and markers of PA relaxation compared to sham.PADN upregulated -adrenoceptor and downregulated -adrenoceptor expression.

Huang et al. 2019 (32)

Prospective randomized, sham-controlled trial

Rat model, IV monocrotaline

Surgical PADN 10/10 - 2 weeks post-

PADN

- Invasive haemodynamics, histological staining, plasma neurohormone, cytokine and neurohormone receptor levels, exercise tolerance

PADN group had lower mPAP, less PA and RV remodelling, and improved RV function.PADN attenuated overactivation of the SN system and reduced expression of neurohormone receptor levels

Garcia-Lunar et al. 2019 (58)

Prospective randomized, sham-controlled trial

Porcine model, pulmonary vein banding

Surgical and percutaneous PADN

6 / 6 (6 healthy subjects

underwent percutaneous

PADN)

- Up to 3 months

- Haemodynamic measurements, CMR, histological staining, plasma neurohormone levels

Surgical PADN did not improve mPAP or PVR compared to sham procedure at any follow-up. PADN was not associated with any benefit in RV anatomy or function.Percutaneous PADN produced focal damage to adventitial fibres compared to the transmural PA lesion produced by surgical PADN

Clinical

Chen et al. 2013 (19)

Non-randomised, non-blinded controlled phase I trial

- Percutaneous PADN (see pre-clinical study by Chen et al. 2013)

13 / 8 40 years 3 months Idiopathic PAH without optimal response to current medical therapy (defined as reduction in mPAP <5mmHg, increment of 6MWD<50m)

1º: ∆mPAP and 6MWD2º: Adverse clinical events

Greater fall in mPAP (PADN:55±5 to 36±5mmHg vs. control:53±5 to 50±5mmHg, p<0.001) and increase in 6MWD (PADN: 324±21 to 491±38m vs. control: 358±30 to 364±38m, p<0.01) in PADN vs. control group.Chest pain during procedure in all. No procedure-related SAEs

Chen et al. 2015 (22)

Open-label phase II trial

- Percutaneous PADN (see Chen et al.

66 / - 52 years 1 year PH at RHC (WHO group 1, 59%; group 2, 27%;

Change in haemodynamic,

Improvements in 6MWT, WHO functional class, BNP,

2013) group 4, 14%). CTEPH included after surgical management

functional and clinical markers; PAH-related clinical events

echocardiographic and RHC measurements.Chest pain during procedure in 71%. Temporary sinus bradycardia in 1 patient. No procedure-related SAEs

Zhang et al. 2019 (23)

Randomised, sham-controlled trial

- Percutaneous PADN (see Chen et al. 2013)

48 / 50 63 years 6 months Combined pre- and post-capillary PH (mPAP≥25mmHg, PCWP>15mmHg, PVR>3WU)

1º: ∆6MWD2º: ∆PVR, occurrence of PE, clinical worsening

Greater increase in 6MWD in PADN than control group (PADN:351±106 to 435±108m, control:344±86 to 359±93m, p<0.001)Greater average reduction in PVR (HR 4.7 95%CI 2.1-10.9, p<0.001) and lower rate of clinical worsening in PADN than control group (HR 2.7, 95%CI 1.2-6.1, p=0.02)

Case reports and conference abstracts were not included in this table. Abbreviations: 6MWD – 6-minute walk distance; 6-OHDA – 6-

hydroxydopamine; BNP – Brain natriuretic peptide; CMR – cardiac magnetic resonance imaging; CTEPH – chronic thrombo-embolic

pulmonary hypertension; DHMCT – dehydrogenized monocrotaline; EM – electron microscopic; IPAH – idiopathic pulmonary arterial

hypertension; IV – intravenous; (m)PAP – (mean) pulmonary artery pressure; PA – pulmonary artery; PADN – pulmonary artery denervation;

PCWP – pulmonary capillary wedge pressure; PE – pulmonary embolism; PH – pulmonary hypertension; PVR – pulmonary vascular

resistance; RHC – right heart catheterisation; SAE – severe adverse event; SN – sympathetic nerve; TxA2 – thromboxane A2; TPG –

transpulmonary gradient; WHO – World Health Organisation.