Transthoracic Pulmonary Artery Denervation for Pulmonary Arterial Hypertension

Introduction

Pulmonary artery hypertension (PAH) is a severe clinical syndrome characterized by progressive pulmonary vascular remodeling and increased mean pulmonary artery pressure (mPAP) and pulmonary vascular resistance.^1,2 Currently, targeted drugs to treat PAH are limited to a few vasodilators; however, it is difficult to reverse the progression of disease in the pulmonary arteries and right ventricle in patients with PAH. Vasodilators have not been demonstrated to significantly decrease mortality during long-term follow-up.³ Therefore, there is an urgent need for new management approaches targeting the PAH disease progression processes.

See cover image

As early as 1980, it was demonstrated that surgical removal of adventitial tissue surrounding the main pulmonary artery (PA) reduced acute increases in mPAP, as shown in a canine model with disease induced by balloon dilation of the main PA.⁴ However, that study did not reveal whether surgical denervation of the PA reversed or attenuated progressive PAH-related pathophysiological changes. They performed PA denervation before the development of PAH, which differed from the actual series of clinical events, thus limiting the clinical translation of their surgical denervation procedure. Recently, with the development of catheter radiofrequency ablation technology, Chen and his colleagues demonstrated that percutaneous pulmonary artery denervation (PADN) improved hemodynamics in patients with PAH.^5,6 However, because of the influence of the anatomic distribution of sympathetic nerves (SNs), PA wall thickness, and radiofrequency ablation energy delivery, the range of SN histological injury caused by PADN was limited.⁷ Therefore, we hypothesized that transthoracic pulmonary artery denervation (TPADN) procedures, in which the main SNs around the PA could be completely and accurately removed under direct visualization after thoracotomy, could achieve complete denervation of the PA and thus reverse or attenuate PAH progression.

Herein, we explored the distribution and composition of SNs around human and rat PAs, which might provide important evidence in the clinical translation of TPADN. Furthermore, we investigated whether TPADN could attenuate PA remodeling, improve right ventricular (RV) function, and affect their underlying mechanisms.

Materials and Methods

Data supporting our findings are available from the corresponding author upon reasonable request. Major resource details can be found in the article and in the online-only Data Supplement.

Experimental Design

The research protocol was approved by the Institutional Animal Care and Use Committee of Fuwai Hospital, Chinese Academy of Medical Sciences, and was performed in accordance with the Guide for Animal Care for animal study and the Declaration of Helsinki conventions on human studies.

Thirty-five healthy male Sprague-Dawley rats (weight, 250–280 g; Beijing Vital River Laboratory Animal Technology Co, Ltd, China) underwent the first hemodynamic measurement at the beginning of study and then were randomized into 2 groups: controls

(single subcutaneous injection of physiological saline, 60 mg/kg, n=10) and tests (single subcutaneous injection of monocrotaline (Sigma-Aldrich, St Louis, MO) dissolved in sterile saline, 60 mg/kg, n=25). During the fourth study week, the second hemodynamic measurement was obtained from the control and test groups. Test group rats with mPAP >25 mm Hg (n=20) were randomly reassigned into sham (n=10) and TPADN groups (n=10). The TPADN group underwent the TPADN procedure, in which the main SNs around the PA were completely and accurately removed under direct visualization after thoracotomy, while the sham group underwent thoracotomy. In the sixth week, hemodynamics, RV function, morphological changes in the PA and RV tissues, and changes in the sympathetic nervous system (SNS) and renin-angiotensin-aldosterone system (RAAS) neurohormone-receptor axes were measured. Another 13 rats were used for the SNs study. Detailed experimental designs are shown in Figure 1. We obtained adipose and connective tissues from around the main PA trunk and its bifurcation from heart transplant patients without PAH to explore the distribution and composition of SNs around the human PAs. The patient information is shown in the online-only Data Supplement (Table I in the online-only Data Supplement).

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

In addition, we also adopted a sugen/hypoxia-induced PAH model to investigate whether TPADN could attenuate PAH progression after the establishment of PAH. The detailed experimental process is shown in Figure I in the online-only Data Supplement.

Distribution and Composition of SNs Around Rat and Human Pulmonary Arteries

The PA and surrounding adipose and connective tissues from rat and human samples were subjected to immunohistochemical staining with tyrosine hydroxylase (TH) to explore SN distributions. Immunofluorescence staining of TH, neuropeptide-Y (NPY), and calcitonin-gene–related peptide (CGRP) was used to analyze SN composition.

Three-Dimensional Reconstruction of SNs and PAs

Referring to a previous research method,⁸ all TH-stained images of SNs around the PAs were converted from RGB (red, green, blue) to grayscale and re-encoded using Adobe Illustrator software (AI-cc 2018; Adobe Systems, San Jose, CA) to determine the relative positional relationship between the SNs and PAs. The images were imported into 3-dimensional imaging software (Autodesk 3ds Max2018) to reconstruct the PAs and SNs in 3-dimension.

TPADN Procedure

Rats were routinely anesthetized with isoflurane (induction, 4.0% in 1:1 O₂/air mix; maintenance, 2.0% in 1:1 O₂/air mix), intubated (16-G Teflon tube), and attached to a small animal ventilator (Type: ALC-V8-SLB; Shanghai Alcott Biotech Co, Ltd, China) at the following settings: breathing frequency 75/min and inspiratory/expiratory ratio 1:1. After thoracotomy in the left third intercostal space, the adipose and connective tissues around the trunk and branches of the main PA were exfoliated using microsurgical techniques. The main pulmonary artery trunk and its bifurcation and the proximal regions of the left and right PAs were the key stripping areas for TPADN. After stripping the main SN around the PA, a RENOV tissue patch (Beijing Qing Yuan Wei Ye Bio-tissue Engineering Co, Ltd, China), which is an acellular dermal matrix implant used as a nerve barrier, was placed over the surgical area to avoid sympathetic reinnervation after TPADN. Detailed surgical diagrammatic sketches are shown in Figure 3A through 3F. We performed TH staining on the surgically excised tissues to demonstrate whether TPADN successfully achieved denervation of the PA (Figure 3G and 3H). The TPADN procedure video can be found in the online-only Data Supplement.

• Download figure
• Download PowerPoint

Measurement of Hemodynamic Parameters and PA Trunk Diameter

The rats were routinely anesthetized with intraperitoneal injections of sodium pentobarbital (50 mg/kg). During the procedure, the right external jugular vein was isolated, and a self-made 2-F right heart catheter was inserted after the incision was made. The catheter was connected to the Biofunctional Experimental System (PowerLab Data Acquisition and Analysis System, Australia) through a pressure sensor to monitor pressure changes. Under waveform guidance, the catheter was advanced through the superior vena cava, right atrium, right ventricle, and finally into the PA trunk. The hemodynamic parameters, including mPAP, pulmonary arterial systolic pressure, pulmonary artery diastolic pressure, right ventricular systolic pressure (RVSP), and mean right ventricular pressure (mRVP), were measured. Rat systemic blood pressure measurements were performed using rat tail artery assays. In addition, we measured the PA trunk diameter of each group with ultrasound (Visual Sonics Vevo 2100 Micro-Ultrasound, small animal imaging system).

Assessment of PA Remodeling and Proliferation Activity

Elastic-Van Gieson staining and immunofluorescence double-staining of von Willebrand Factor and α-smooth muscle actin were performed on lung tissue sections to assess the relative medial wall thickness (MWT) and degrees of PA muscularization in different PAH rat models (monocrotaline model: Figure 4F and 4G; Sugen/hypoxia model: Figure IIC and IID in the online-only Data Supplement). Lung tissue sections were subjected to immunofluorescence staining with Ki67 to assess PA proliferation activity (Figure 4H; monocrotaline model). According to previous research methods,⁸ the relative wall thickness of the PA was calculated as a percentage of the MWT (%MWT):

• Download figure
• Download PowerPoint

%MWT=([external diameter−internal diameter]/external diameter)×100

The degree of muscularization of the PA was divided into 3 groupings: full muscularization, partial muscularization, and nonmuscularization. PA proliferation activity was assessed by counting cells with positive (Ki67) signals per total area.

Assessment of RV Hypertrophy, Fibrosis, and Inflammation

RV hypertrophy was evaluated by calculating the RV cardiomyocyte cross-sectional area (RVCSA) and RV hypertrophy index (RVHI): (RVHI=RV/(LV+S)). RV paraffin sections were stained with hematoxylin and eosin and wheat germ agglutinin to calculate the RVCSA in different PAH rat models (monocrotaline model: Figure 5A and 5B; Sugen/hypoxia model: Figures IIIA and IIIB in the online-only Data Supplement). RV tissue sections were stained with Picrosirius red and evaluated under polarized light to identify their degrees of cardiac fibrosis in different PAH rat models (monocrotaline model: Figure 5C; Sugen/hypoxia model: Figure IIIC in the online-only Data Supplement).⁹ The degree of RV fibrosis was quantified as the percentage of collagen-positive tissue. RV inflammation was measured by leukocyte infiltration; 5 randomly selected areas per ventricle were used to count the number of positive CD45-nuclei per section area (Figure IVC and IVD in the online-only Data Supplement, monocrotaline model).

• Download figure
• Download PowerPoint

Cardiac Function Measurement and Treadmill Test

Cardiac magnetic resonance imaging was performed in 3 rat groups 2 weeks after TPADN to assess cardiac function, which was determined according to the RV end-diastolic volume, RV end-systolic volume, RV ejection fraction, LV end-diastolic volume, LV end-systolic volume, and LV ejection fraction.

According to the previous method,¹⁰ exercise capacity was tested by measuring maximal distance run on a motorized treadmill (Simplex II Instrument; Columbus Instruments, Columbus, OH). The initial treadmill speed was 10 m/min and increased 5 m/min every 5 minutes for 30 minutes or until the rat fatigued. Two weeks after TPADN, the treadmill distances in each group were measured.

Measurement of Changes of Plasma Neurohormone, Serum Inflammatory Cytokines, and Oxidative Stress Level in Lung Tissue

Commercially available ELISA kits (Jiancheng, Nanjing, China) were used to measure plasma norepinephrine, NPY, aldosterone, and angiotensin II, serum inflammatory cytokines, the level of malondialdehyde, and the activity of (SOD) superoxide dismutase.

Expression and Signaling of Neurohormonal Receptors in Lung and RV Tissues

The density and expression of α1-AR (α1-adrenergic receptor), NPY1R (neuropeptide-Y I type receptors), AT1R (angiotensin II type 1 receptors), and MRs (mineralocorticoid receptors) in lung tissue, as well as β1-AR (β1-adrenergic receptor) and AT1R in RV tissue, were assessed after immunofluorescence staining and Western blotting. The expressions of downstream mediators that regulate signaling of neurohormonal receptors were measured via Western blot.

Immunohistochemistry

The immunohistochemical staining procedure was performed with reference to a previous method.^8,11 Briefly, paraffin sections (4–5 μm) were de-paraffinized, hydrated, and repaired with citrate buffer under high pressure for 15 minutes. The sections were blocked with 3% endogenous peroxide at 25°C for 10 minutes. After rinsing, the sections were incubated with a primary antibody at 4°C overnight. Next, the sections were rewarmed at 25°C for 45 minutes and incubated with a secondary antibody at 25°C for 30 minutes. The specimens were then stained with DAB (3,3′-diaminobenzidine) chromogenic reagent. The hematoxylin counterstaining, hydrochloric acid alcohol differentiation, gradient alcohol dehydration, and xylene soaking were performed on the sections. Finally, the sections were sealed with neutral balsam reagent. The main primary antibody data can be found in the online-only Data Supplement (Major Resources Table in the online-only Data Supplement).

Immunofluorescence

The immunofluorescence staining procedure was performed with reference to a previous method.¹² Briefly, paraffin sections (4–5 μm) were de-paraffinized, hydrated, and repaired with EDTA under high pressure for 15 minutes. Next, the sections were blocked with 5% BSA blocking buffer at 25°C for 60 minutes. After rinsing, the sections were incubated with primary antibody at 4°C overnight. The sections were rewarmed at 25°C for 45 minutes and incubated with secondary antibody at 25°C for 60 minutes. Finally, the sections were incubated with 4′, 6-diamidino-2-phenylindole at 25°C for 10 minutes and sealed. The main primary antibody data can be found in the online-only Data Supplement (Major Resources Table in the online-only Data Supplement).

Western Blot

Western blotting was performed as described previously.¹³ Tissue lysates containing equal amounts of protein were extracted with radioimmunoprecipitation assay buffer containing protease inhibitor (Roche Diagnostics, Risch-Rotkreuz, Switzerland). These protein samples were electrophoresed on a sodium dodecyl sulfate-polyacrylamide gel and transferred to nitrocellulose membranes using an iBlot 2 dry blotting system (Thermo Fisher

Scientific, Waltham, MA). The membrane was blocked in 5% skim milk powder in tris-buffered saline with Tween 20 (100 mmol/L Tris, pH 7.5, 0.9% NaCl, 0.1% Tween-20) for 60 minutes and incubated with primary antibody at 4°C overnight. The membrane was then washed 3 times with tris-buffered saline with Tween 20 and incubated with secondary antibody at 25°C for 60 minutes. After rinsing, enhanced chemiluminescent substrate (Millipore Corporation, Burlington, MA) was used to develop the membranes. GAPDH actin was used as a loading control. AlphaView software (ProteinSimple, San Jose, CA) was used to analyze the band intensities. The main primary antibody data can be found in the online-only Data Supplement (Major Resources Table in the online-only Data Supplement).

Statistical Analysis

The statistical analysis was performed using SPSS 24.0 (IBM Corporation, Armonk, NY). Data are expressed as means±SD. All data have been tested for normality and homogeneity of variance and were consistent with a positive distribution. For data with homogeneity of variance, unpaired Student t test was used to compare the 2 groups, and one-way ANOVA followed by Bonferroni multiple comparisons was used to evaluate the differences among ≥3 groups. For data without homogeneity of variance, Welch t test with Bonferroni correction was used to compare the 2 groups, and one-way ANOVA followed by Dunnett T3 multiple comparisons was used to evaluate the differences among ≥3 groups. P<0.05 was considered statistically significant.

Results

Distribution and Composition of SNs Around Rat and Human PAs

As shown in Figure 2A, there were abundant SNs in the adipose and connective tissues around the rat PA adventitia. According to the analysis of SN area around PAs (Figure 2B), and a 3-dimensional model of the PA and its surrounding SNs (Figure 2C and 2D), we found that the area of SNs in the posterior and right walls of the main PA trunk and its bifurcation was greater than that of anterior and left walls (all P<0.01). From this, we learned that the SNs were mainly distributed in the posterior and right walls of the main PA trunk and its bifurcation. In addition, we further analyzed the difference in SN distribution between normal and PH rats. First, as shown in Figure VA in the online-only Data Supplement, the SN area around PA trunk and bifurcation in PAH rats is much larger than that in normal rats. Moreover, the difference in SN area is greater in the posterior and right wall area of the PA trunk and its bifurcation (Figure VB and VC in the online-only Data Supplement). Also, we demonstrated that the composition of SNs around the rat PA mainly contained TH-positive nerve fibers, NPY-positive nerve fibers, and CGRP-positive nerve fibers, whereas the TH-positive area and NPY-positive area were far greater than the CGRP-positive area (Figure 2E and 2F). We found that the mean individual TH-positive/CGRP-positive area ratio was 12.3±1.9 in the PA trunk and 8.3±1.5 in the PA bifurcation, while the mean individual NPY-positive/CGRP-positive area ratio was 5.3±1.1 in the PA trunk and 3.5±0.8 in the PA bifurcation (all P<0.01 for TH-positive area versus CGRP-positive area, NPY-positive area versus CGRP-positive area).

We also explored the SN distribution and composition around the human PA (Figure 2G through 2L). We found that the SN composition around the human PA included many TH-positive and NPY-positive nerve fibers and a small number of CGRP-positive nerve fibers (Figure 2K and 2L). The mean individual TH-positive/CGRP-positive area ratio was 17.5±2.8 in the PA trunk and 21.7±3.4 in the PA bifurcation, while the mean individual NPY-positive/CGRP-positive area ratio was 5.8±1.4 in the PA trunk and 8.4±1.8 in the PA bifurcation (all P<0.01 for TH-positive area versus CGRP-positive area, NPY-positive area versus CGRP-positive area).

TPADN Procedure

A conceptual map of the TPADN procedure is shown in Figure 3A through 3C. Figure 3A and 3B show that there were abundant adipose and connective tissues around the main PA trunk and bifurcation before TPADN. After TPADN, Figure 3C demonstrates that these tissues are completely stripped. The actual TPADN procedure is shown in Figure 3D through 3F. Figure 3D shows that the PA is separated after thoracotomy in the left third intercostal space. Figure 3E demonstrates the adipose and connective tissues around the main PA trunk, and bifurcation were exfoliated completely. And Figure 3F showed that the surgical tissue patch implants are placed to cover the surgical areas. Meanwhile, we found that stripped adipose and connective tissues after TPADN contained abundant SNs (Figure 3G and 3H), which shows that the TPADN procedure successfully denervated the PA. Two weeks after TPADN, compared with the sham group, there was no sympathetic reinnervation around the surgical areas of the rat PA in the TPADN group (Figure VI in the online-only Data Supplement).

Effects of TPADN on Hemodynamics and PA Dilatation in Monocrotaline-Induced PAH

The hemodynamic variables among the 3 groups are shown in Table and Figure 4A through 4E. There were no significant differences in baseline hemodynamic parameters among the 3 groups at the beginning of the experiment. After 4 weeks of monocrotaline injections, the mPAP, RVSP, pulmonary arterial systolic pressure, pulmonary artery diastolic pressure, and mRVP of the sham and TPADN groups were significantly higher than those of the control group (Table; all P<0.01), indicating that the PAH rat model was successfully constructed. However, 2 weeks after TPADN, the hemodynamic variables in the TPADN group were significantly decreased, while the variables in the sham group remained constantly increased (Table; all P<0.01). Compared with the sham group, the mPAP, RVSP, and other hemodynamic variables in the TPADN group were significantly decreased (Table; mPAP: 20.5±3.1 mm Hg versus 36.5±2.4 mm Hg; RVSP: 34±3.2 mm Hg versus 62.3±4.5 mm Hg, respectively; all P<0.01). We observed that heart rates and systemic blood pressures among the 3 groups were not significantly different (Table; Table II in the online-only Data Supplement; all P>0.05). Representative RV and PA waveforms of each group in monocrotaline-induced PAH model can be found in Figure VIIA in the online-only Data Supplement.

Besides hemodynamics, PA dilatation is one of the consequences of pulmonary arterial hypertension (PAH), which is useful for identifying patients with PAH. As a result, we measured the PA trunk diameter of each group with ultrasound to observe the effect of TPADN on PA dilatation in PAH. The results are shown in Figure VIII in the online-only Data Supplement. Compared with the normal group, the PA trunk diameter of the sham group was significantly increased (P<0.01), suggesting PA dilatation in PAH rats. However, after TPADN, compared with the sham group, the PA trunk diameter were significantly decreased in the TPADN group (P<0.01), indicating that TPADN surgery can attenuate the PA dilatation in PAH rats.

Taken together, these results demonstrate that TPADN improved the hemodynamics and attenuated PA dilatation in monocrotaline-induced PAH rats.

Effects of TPADN on PA Remodeling in Monocrotaline-Induced PAH

As shown in Figure 4I and 4J and Table, in the sixth study week, the PA full-muscularization rate in the TPADN group was significantly reduced, while the nonmuscularization rate was significantly increased compared with the sham group. Meanwhile, the percentage of MWT in the TPADN group was significantly reduced (Figure 4J; full muscularization: 22%±4.5% versus 41.7%±6.3%; nonmuscularization: 56.7%±4.4% versus 31%±6.5%; Figure 4I; MWT: 30.4%±2.0% versus 39.5%±2.9%; all P<0.01). These results indicate that TPADN significantly attenuated PA remodeling in monocrotaline-induced PAH rats. In addition, we also demonstrated that TPADN inhibited abnormal PA proliferation in PAH rats (Figure 4H and 4K).

Effects of TPADN on RV Hypertrophy, RV Fibrosis, RV Inflammation, and Cardiac Function in Monocrotaline-Induced PAH

We evaluated the effects of TPADN on RV remodeling and dysfunction in PAH rats from 4 aspects. First, the RVHI and RVCSA in the TPADN group were significantly lower than those in the sham group (Table and Figure 5D and 5E; RVHI: 34.0%±1.2% versus 46.9%±1.3%; RVCSA: 234.5±15.6 μm² versus 535.1±18.7 μm²; all P<0.01), suggesting that TPADN decreased RV hypertrophy in PAH rats. Second, after TPADN, the percentage of collagen-positive tissue areas in the TPADN group was reduced compared with that of sham group (Figure 5C and 5F; P<0.01), indicating that TPADN attenuated RV fibrosis in PAH. Third, as shown in Figure IVC and IVD in the online-only Data Supplement, after TPADN, the number of positive CD45-nuclei per unit area in the TPADN group was significantly decreased than that of sham group, indicating that TPADN attenuated RV inflammation in PAH. Finally, 2 weeks after TPADN, compared with the sham group, the RV end-diastolic volume and RV end-systolic volume were decreased while RV ejection fraction was significantly increased in the TPADN group (Table and Figure 5G through 5K; RV end-diastolic volume: 0.41±0.019 mL versus 0.56±0.034 mL; RV end-systolic volume: 0.15±0.022 mL versus 0.28±0.027 mL; RV ejection fraction: 63.8%±5.25% versus 49.9%±2.28%; all P<0.01). Furthermore, we also explored the effect of TPADN on left ventricular function in PAH rats. Two weeks after TPADN, compared with the sham group, the LV end-diastolic volume, LV end-systolic volume and LV ejection fraction were significantly increased in the TPADN group. (Table; all P<0.01). Taken together, these results demonstrated that TPADN improved cardiac function in PAH.

Effects of TPADN on Exercise Capacity of Monocrotaline-Induced PAH Rats and RV Function Under Exercise Load

To observe whether TPADN could improve the exercise capacity of PAH rats, we performed treadmill test and measurement of RV function under exercise load. As shown in Figure IXA through IXE in the online-only Data Supplement, compared with the sham group, the RV end-diastolic volume and RV end-systolic volume were decreased while the RV ejection fraction was significantly increased in the TPADN group (all P<0.01), indicating that TPADN significantly improved the RV function of PAH rats under exercise load, thus significantly improving their exercise ability (Figure IXF in the online-only Data Supplement).

Effect of TPADN on Plasma Neurohormone, Serum Inflammatory Cytokines, and Oxidative Stress Level of Lung Tissue in Monocrotaline-Induced PAH

Because many studies reported that plasma neurohormone levels of SNS and RAAS were elevated in PAH, we measured the changes in plasma neurohormone levels in each group.

As shown in Figure X in the online-only Data Supplement, after TPADN, the expressions of plasma norepinephrine, NPY, angiotensin II, and aldosterone were significantly decreased in the TPADN group compared with the sham group (all P<0.01), suggesting that TPADN inhibited neurohormonal overactivation of SNS and RAAS in PAH.

Next, as inflammation plays an important role in the initiation and progression of PAH, we measured changes in serum inflammatory cytokines. As shown in Figure IVA and IVB in the online-only Data Supplement, serum IL-1β (interleukin-1β) and IL-6 (interleukin-6) levels were significantly lower in the TPADN group compared with the sham group (all P<0.01), indicating that TPADN surgery reduced circulating levels of inflammatory cytokines in PAH rats.

Oxidative stress is closely related to the progression of PAH disease. We examined changes in oxidative stress in local lung tissue. As shown in Figure XI in the online-only Data Supplement, compared with the sham group, malondialdehyde levels in the lung tissue of the TPADN group were significantly decreased, while SOD activity was significantly increased (all P<0.01), indicating that TPADN surgery may attenuate PAH by reducing the oxidative stress level of lung tissue in PAH rats.

Effects of TPADN on the Expression and Signaling of Neurohormonal Receptors of Local Tissues in Monocrotaline-Induced PAH

As shown in Figure 6A through 6I, compared with the sham group, the density and expression of α1-AR, NPY1R, AT1R, and MR in PA were downregulated in the TPADN group. In addition, the density and expression of β1-AR in RV tissues were upregulated (Figure XIIA, XIIC, XIIE, and XIIF in the online-only Data Supplement), while the density and expression of AT1R in RV tissues were downregulated in the TPADN group (Figures XIIB, XIID, XIIE, and XIIF in the online-only Data Supplement). We also tested changes in downstream neurohormonal receptor regulatory mediators. The results showed that downstream regulatory mediator expression, such as for phospho-PLCγ1 (phospholipase Cγ1), phospho-PKD/PKCμ (protein kinase D/protein kinase Cμ), phospho-ERK1/2 (p44/42 mitogen-activated protein kinase), phospho-p38 MAPK (p38 mitogen-activated protein kinase), and phospho-G protein–coupled receptor kinase-2, were significantly decreased in the TPADN group compared with the sham group (lung tissue: Figure 6J and 6M; RV tissue: Figure XIIG and XIIH in the online-only Data Supplement). These results indicate that TPADN inhibited the abnormal expression and signaling of neurohormonal receptors in local tissue.

• Download figure
• Download PowerPoint

Effects of TPADN on Hemodynamics, PA Remodeling, RV Hypertrophy, and RV Fibrosis

in Sugen/Hypoxia-Induced PAH

In addition, we also adopted a sugen/hypoxia-induced PAH model to investigate severe form of PAH reversal by TPADN. First, the hemodynamic variables among 3 groups are shown in Figure IIA and IIB in the online-only Data Supplement and Table III in the online-only Data Supplement. Five weeks after Sugen5416 injection, RVSP and mPAP of the sham and TPADN groups were significantly higher than those of the control group (Figure IIA and IIB in the online-only Data Supplement; all P<0.01), indicating that the PAH model was successfully constructed. However, 5 weeks after TPADN (at the tenth study week), the hemodynamic variables such as mPAP and RVSP were significantly decreased in the TPADN group, compared with the sham group (Figure IIA and IIB in the online-only Data Supplement, mPAP: 28.2±3.4 mm Hg versus 52.8±3.5 mm Hg; RVSP: 47.5±3.9 mm Hg versus 94.2±6.3 mm Hg, respectively; all P<0.01). Generally, these results demonstrated that TPADN improved the hemodynamics of rats with PAH.

Moreover, we investigated the effects of TPADN on PA remodeling in sugen/hypoxia-induced PH model. At the tenth study week, compared with the sham group, the percentage of MWT in the TPADN group was significantly reduced (Figure IIE in the online-only Data Supplement, MWT: 37.5%±1.8% versus 57.1%±2.3%; all P<0.01). Meanwhile, the PA full-muscularization rate in the TPADN group was significantly reduced, while the nonmuscularization rate was significantly increased, compared with the sham group (Figure IIF in the online-only Data Supplement, full-muscularization rate: 30.4%±5.2% versus 55.8%±5.6%; nonmuscularization: 47.1%±5.8% versus 20.4%±3.3%; all P<0.01). In summary, these results suggested that TPADN significantly attenuated PA remodeling in PAH rats.

Finally, we evaluated the effects of TPADN on RV remodeling in PAH rats from 2 aspects. First, at the tenth study week, the RVHI and RVCSA in the TPADN group were significantly lower than those in the sham group (Figure IIID and IIIE in the online-only Data Supplement and Table III in the online-only Data Supplement, RVHI: 41.2%±2.4% versus 61.8%±2.9%; RVCSA: 325±18.7 μm² versus 650.1±24 μm², all P<0.01), suggesting that TPADN decreased RV hypertrophy in PAH rats. Second, compared with the sham group, the percentage of collagen-positive tissue areas in the TPADN group was significantly decreased (Figure IIIC and IIIF in the online-only Data Supplement), indicating that TPADN attenuated RV fibrosis in PAH rats.

Taken all together, we demonstrated that TPADN could attenuate PAH progression in rats with sugen/hypoxia-induced PAH.

Discussion

In our study, we demonstrated that (1) there were abundant SNs in adipose and connective tissues around rat and human main PA trunks and bifurcations; (2) TPADN improved hemodynamics and cardiac function and attenuated PA remodeling, RV hypertrophy, and fibrosis in rats with PAH; (3) therapeutic effects of TPADN on disease progression in PAH might be associated with inhibiting SNS and RAAS neurohormonal overactivation and regulating abnormal expression and signaling of neurohormonal receptors in local tissue.

Previous studies showed that SNs around canine PAs were mainly distributed in the PA adventitia.¹⁴ To the best of our knowledge, we were the first to demonstrate abundant SNs in the adipose and connective tissues around human and rat PA trunks and bifurcations, providing important anatomic evidence for careful selection of surgical sites during clinical translation of TPADN. We found that SN components around rat PAs included numerous TH-positive nerve fibers, NPY-positive nerve fibers, and a small number of CGRP-positive nerve fibers, which were also identified in SNs around the human PA. When the SNS was stimulated, the SNs released a mixture of neurohormones, resulting in upregulation of multiple neurohormone plasma levels. These neurohormones contributed to PAH progression by interacting with receptors in the local tissues. These results partially explain the reason why previous treatments targeting single neurohormones or their receptors in the SNS and RAAS did not alleviate or reverse PAH. However, in our research, we found that TPADN delayed disease progression by inhibiting the upregulation of multiple neurohormones in the SNS and RAAS, thus making TPADN a promising treatment for PAH.

Previous studies demonstrated that percutaneous PADN attenuated pulmonary vascular remodeling and RV dysfunction in experimental PAH models.^8,15 Because of limitations in SN anatomic distributions and the output power in radiofrequency ablation, percutaneous PADN caused nearly no injuries to SNs around thicker-walled main PAs. This finding is consistent with a previous study that reported radiofrequency energy delivery caused limited histological damage to SNs around thicker-walled main or proximal PAs.⁷ Therefore, we believe that the TPADN procedure, in which the main SNs around the PA trunk and its bifurcation could be completely and irreversibly removed under direct visualization, could broaden the range and degree of SN damage and achieve complete denervation of the PA. Some studies have reported that sympathetic reinnervation occurred in a previously denervated lung in experimental models. Sympathetic reinnervation may be associated with sympathetic axon growth mediated by nerve growth factor secreted by proliferating PA smooth muscle cells (PASMC).^16–18 Percutaneous PADN did not completely damage all SNs around the PA adventitia. Therefore, unimpaired or less impaired sympathetic axons could extend to surgical areas under nerve growth factor stimulation and affect long-term percutaneous PADN effects. The surgical tissue patch implants composed of acellular dermal matrix as an implantable physical barrier were widely used for preventing Frey’s syndrome after parotidectomy.^19,20 One of the mechanisms of Frey’s syndrome is abnormal reinnervation and abnormal connections between 2 different nerves. Injuries or defects in parotid fascia after parotidectomy can expose postganglionic parasympathetic nerve fibers that innervate salivary glands and promote secretions. These nerve fibers might make abnormal connections with SNs that innervate skin sweat glands and blood vessels, leading to abnormal secretions by the sweat glands. However, tissue patch implants composed of acellular dermal matrix could effectively hinder abnormal reinnervation and misconnection of these nerve fibers, thus preventing Frey’s syndrome after parotidectomy. Similarly, to prevent sympathetic reinnervation in surgical areas after TPADN, we used surgical tissue patch implants to cover the surgical areas after SN stripping. We also demonstrated that, 2 weeks after TPADN, there was no sympathetic reinnervation around the operated areas of the rat PA in the TPADN group compared with the sham group. This suggests that TPADN might avoid postoperative sympathetic reinnervation.

A vast number of preclinical and clinical studies have demonstrated that RAAS activity in PAH was elevated, which was closely associated with the progression of PAH.^21,22 Specifically, increased plasma angiotensin II and aldosterone, respectively, stimulated the AT1R and MR in PASMC. Stimulation of AT1R and MR could activate NADPH oxidase to increase the level of oxidative stress, and thus decreased the expression of eNOS (endothelial nitric oxide synthase) and bioavailable nitric oxide (NO·) in local lung tissue. Thereby, it ultimately lead to abnormal vasoconstriction, dysregulated cell proliferation, and hypertrophy that contribute to PA remodeling. Intriguingly, our results indicated that the beneficial effect of TPADN on PA remodeling could be partly explained by reductions in plasma angiotensin II and aldosterone levels, downregulated expressions of AT1R and MR, as well as inhibition of downstream NADPH oxidase/eNOS signaling pathways of these neurohormone receptors (Figure 6 and Figure XIII in the online-only Data Supplement). Some studies have reported that AT1R and MR-mediated cell proliferation, hypertrophy, and migration were important contributors to RV hypertrophy and fibrosis.^23,24 Upregulation of ERK1/2 phosphorylation, a downstream mediator of AT1R, played an important role in RV remodeling and dysfunction. In this study, we showed that the beneficial effect of TPADN on RV remodeling and dysfunction in PAH could be relevant to the decreased expression of AT1R in RV cardiomyocytes, as well as the downregulation of ERK1/2 phosphorylation. This was consistent with Chen’s studies, which showed that percutaneous PADN improved RV function in PAH by regulating local RAAS activity.¹⁵ Taken together, we demonstrated that TPADN might delay PAH progression by hindering excessive activation of the RAAS neurohormone-receptor axis.

In addition to the neurohormonal activation of RAAS in PAH, SNS activation also played a key role in PAH progression.^21,22 Specifically, increased plasma norepinephrine and NPY, respectively, stimulated α1-AR and NPY1R in PASMC. Stimulation of α1-AR and NPY1R resulted in the activation of PLC (phospholipase C) and PP2B (phosphatase 2B ), which promoted translocation of nuclear factor of activated T cells (NFAT) to the nucleus and transcription of cyclin D1. Eventually, it resulted in dysregulated cell proliferation and perivascular fibrosis, leading to PA remodeling.^22,25 In addition, we found that the activation of PLC could activate the downstream phospho-PKD, phospho-ERK, and phospho-p38 signaling pathways of neurohormone receptors, which was closely related to the proliferation and migration of PASMC. However, until now, there was no direct evidence indicating whether surgical or percutaneous PADN could reverse or attenuate PAH progression by inhibiting the SNS neurohormone-receptor axis. Therefore, we investigated the effect of TPADN on the SNS neurohormone-receptor axis. Intriguingly, our data demonstrated that TPADN to attenuate PA remodeling might be explained partially by reductions in plasma norepinephrine and NPY levels, downregulation expression of α1-AR and NPY1R in PASMC, and suppression of downstream PLC/PP2B and ERK/P38 signaling pathways of these neurohormone receptors (Figure 6 and Figure XIII in the online-only Data Supplement). Some studies reported that GRK2-mediated downregulation of β1-AR expression in RV cardiomyocytes and increased β1-AR desensitization impaired inotropic reserves in RV failure with PAH.¹⁰ According to our research, TPADN improved RV function by upregulating the expression of β1-AR in RV cardiomyocytes and suppressing β1-AR desensitization mediated by the activation of GRK2 signaling pathways. Generally, we demonstrated that TPADN could attenuate PAH progression by inhibiting excessive activation of the SNS neurohormone-receptor axis.

PAH is a fatal clinical syndrome with a poor long-term prognosis. Surgeons are continuously exploring new surgical approaches to PAH. It was reported that pulmonary artery banding successfully reversed PAH in a 19-year-old woman.²⁶ Although no explanation of the surgical mechanism could be given in that clinical case, according to our findings, the therapeutic effect of pulmonary artery banding might be associated with denervation of the PA caused by intraoperative pulmonary arterial dissociation. In the future, with a help of cutting-edge thoracoscopic and near-infrared technologies for intraoperative imaging of thoracic SNs,²⁷ the minimally invasive TPADN procedure can efficiently and accurately achieve PA denervation, thus providing a promising treatment option for patients with PAH.

Study Limitations

Previous studies reported that PADN could result in systemic hypotension. However, we only investigated whether systemic hypotension occurred in rats after TPADN during the experimental period. We did not monitor the long-term dynamic blood pressure changes in rats. Further research is required to verify the safety and long-term efficacy of the TPADN procedure in different PAH models. We validated the efficacy of TPADN using 2 kinds of PAH models (monocrotaline-induced and Sugen-induced), but we also need to validate its efficacy in human tissue specimens with PAH.

Conclusions

We demonstrated that TPADN improved hemodynamics and cardiac function in rats with PAH and attenuated PAH progression by inhibiting overactivation of the SNS and RAAS neurohormone-receptor axes. Furthermore, there were abundant SNs in the adipose and connective tissues around the rat and human main PA trunks and bifurcations, which provides important evidence for the clinical translation of TPADN.

Acknowledgments

H. Zhang and Z.-C. Jing conceived and designed the study. Y. Huang, Y.-W. Liu, H.-Z. Pan, and P.-H. Wang performed the animal studies. Y. Huang and L. Xiang analyzed the data. X.-L. Zhang and J. Li collected the human samples. Y. Huang, J. Yang, and J. Meng performed the pathological experiments. Y. Huang and H. Zhang drafted the article. H. Zhang is the principal investigator, obtained funding, and assisted with the drafting of the article. All authors reviewed, contributed to, and approved the final article. We greatly appreciate the help in experimental design from Prof Shao-Liang Chen from the Department of Cardiology, the First Affiliated Hospital of Nanjing Medical University.

Footnotes