Microvascular Research 94 (2014) 73–79
Contents lists available at ScienceDirect
Microvascular Research journal homepage: www.elsevier.com/locate/ymvre
Chronic overcirculation-induced pulmonary arterial hypertension in aorto-caval shunt Alessio Rungatscher a,⁎, Daniele Linardi a, Elisabetta Milani a, Grazia Ucci b, Elena Nicolato c, Flavia Merigo c, Beatrice Salvetti c, Alessandro Mazzucco a, Giovanni Battista Luciani a, Giuseppe Faggian a a b c
Department of Surgery, Division of Cardiac Surgery, University of Verona Verona, Italy Division of Cardiology, University of Novara Novara, Italy Department of Neurological and Movement Sciences, University of Verona Verona, Italy
a r t i c l e
i n f o
Article history: Received 1 January 2014 Revised 9 May 2014 Accepted 15 May 2014 Available online 23 May 2014 Keywords: Pulmonary arterial hypertension Vascular remodeling Right ventricle overload Volume overload Ventricular vascular coupling Ventricular interdependence Aorto-caval shunt Rats Pressure-volume loops Experimental model
a b s t r a c t Pulmonary arterial hypertension is a common complication of congenital heart defects with left-to-right shunts. Current preclinical models do not reproduce clinical characteristics of shunt-related pulmonary hypertension. Aorto-caval shunt was firstly described as a model of right ventricle volume overload. The pathophysiology and the possible determination of pulmonary arterial hypertension of different periods of shunt exposure are still undefined. A method to create standardized, reproducible aorto-caval shunt was developed in growing rats (260 ± 40 g). Three groups of animals were considered: shunt exposure for 10 weeks, shunt exposure for 20 weeks and control (sham laparotomy). Echocardiography and magnetic resonance revealed increased right ventricular end diastolic area in shunt at 10 weeks compared to control. Hemodynamic analysis demonstrated increased right ventricular afterload and increased effective pulmonary arterial elastance (Ea) in shunt at 20 weeks compared to control (1.29 ± 0.20 vs. 0.14 ± 0.06 mmHg/μl, p = 0.004). At the same time point, the maximal slope of end-systolic pressure– volume relationship (Ees) decreased (0.5 ± 0.2 mmHg/ml vs. 1.2 ± 0.3, p b 0.001). Consequently, right ventricular–arterial coupling was markedly deteriorated with a ≈50% decrease in the ratio of end-systolic to pulmonary artery elastance (Ees/Ea). Finally, left ventricular preload diminished (≈ 30% decrease in left ventricular end-diastolic volume). Histology demonstrated medial hypertrophy and small artery luminal narrowing. Chronic exposure to aorto-caval shunt is a reliable model to produce right ventricular volume overload and secondary pulmonary arterial hypertension. This model could be an alternative with low mortality and high reproducibility for investigators on the underlying mechanisms of shunt-related pulmonary hypertension. © 2014 Elsevier Inc. All rights reserved.
Introduction Left-to-right shunting congenital heart diseases are the common cause of secondary pulmonary arterial hypertension (PH). The status of the pulmonary vascular structure and extent of PH are important determinants of feasibility of corrective procedures and long-term survival after cardiac operation (Badesch et al., 2009).
Abbreviations: Ea, effective pulmonary arterial elastance; Ees, maximal slope of end systolic pressure–volume relationship; PH, pulmonary hypertension; RV, right ventricle; IVC, inferior vena cava; MRI, magnetic resonance imaging; VM, ventricle mass; EDV, end-diastolic volume; ESV, end-systolic volume; SV, stroke volume; EF, ejection fraction; TAPSE, systolic tricuspid annular excursion; LV, left ventricle; P–V, pressure–volume; PRSW, preload recruitable stroke work; ESPVR, end-systolic pressure–volume relationship; EDPVR, end-diastolic pressure–volume relationship. ⁎ Corresponding author at: Department of Surgery, Division of Cardiac Surgery, University of Verona, Piazzale Stefani 1, 37126 Verona, Italy. Fax: +39 0458123308. E-mail address: [email protected] (A. Rungatscher).
http://dx.doi.org/10.1016/j.mvr.2014.05.005 0026-2862/© 2014 Elsevier Inc. All rights reserved.
During past decades, animal models have deduced numerous cellular and molecular mechanisms of PH. Although much insight has been gained from “classic” PH models, such as hypoxic PH or monocrotalineinduced PH, most of these studies are hampered by the fact that these models do not truly reflects the typical features of PH in respect of histopathology and progressive course (Dickinson et al., 2013). Indeed, while drugs targeting the signaling pathways involved in the pathogenesis of PH have been confirmed to partially or completely reverse hypoxia and monocrotaline-induced PH in rats, a disappointing discrepancy has emerged in recent clinical trials (Bai et al., 2011). Hence, there is an increasing interest in models that more reliably reflect the complex disease PH. Because increased pulmonary blood flow appears to be crucial factor in pathogenesis, particular emphasis is given in models of increased pulmonary blood flow that lead to development of pulmonary vascular remodeling. Continuous attempts have been made to simulate the pulmonary vasculopathy by surgically creating peripheral or central systemic-to-
74
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
pulmonary communication (Fullerton et al., 1996; Reddy et al., 1995; Wang et al., 2010; Xiong et al., 2012). Central shunts prepared in large animals closely simulate the pathophysiologic characteristics of congenital intracardiac shunts; however, the high-level surgical requirements, relatively greater mortality, limited patency in inducing remodeling lesions and inadequacy of biochemical reagents for large animals have impeded the widespread adoption of these models. Thus, a satisfactory shunt-related model that could be used to determine the underlying molecular mechanism of pulmonary vascular remodeling due to in vivo exposure to high blood flow is still needed. One pre-tricuspid model, the aorto-caval shunt, is relatively widely applied and was firstly described by Garcia and Diebold (1990) later modified by others (Ocampo et al., 2003). The aorto-caval shunt model was used by several authors to study right ventricle (RV) response to acute, sub-acute and chronic volume overload (Szabo et al., 2006; Wang et al., 2003; Yerebakan et al., 2010), while other authors used this model to study the effects of increased pulmonary blood flow on pulmonary circulation (Nishimura et al., 2003; Van Albada et al., 2005). However, results reported in literature are conflicting, especially when the model is used to produce secondary pulmonary hypertension, and the exact mechanism, which leads to development of pulmonary hypertension in this model, is not clear. The aim of the present study was to demonstrate with morphologic, functional and hemodynamic analysis the possible induction of secondary PH as a consequence of different periods of RV volume overload and pulmonary overcirculation. Methods Male Sprague–Dawley rats (Charles River Laboratories, Calco, Italy) were used (age, 8 weeks; weight, 260 ± 40 g). All experiments were performed in adherence to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (National Institutes of Health Publication No. 85-23, revised 1996). The protocol was approved by the University of Verona authority for experimental research (C.I.R.S.A.L. Interdepartmental Centre of Experimental Research Service). Experimental protocol Rats were randomly divided into 3 groups: the first group (n = 10) underwent shunt procedure and was observed for 10 weeks after procedure (shunt 10W); thereafter, shunt was closed and non-invasive morphometric and hemodynamic evaluations were performed. The second group (n = 10) underwent the same surgical procedure but was observed for 20 weeks before shunt closure and subsequent evaluation (shunt 20W). The third group (n = 10) served as control, underwent a sham laparotomy operation and then was observed for 20 weeks (sham). All surgical procedures and invasive and non-invasive evaluations were conducted under general anesthesia (0.25% pentobarbital sodium, 40 mg/kg, intraperitoneal injection). At the end of the programmed observation period, rats were sacrificed and the samples for histology were collected. Surgical creation of the shunt A midline abdominal incision was made, and a descending aorta above renal bifurcation was cleared of adjacent tissues. There a 3-0 Ti-Cron suture (Covidien, Mansfield, MA, USA) was positioned around the aorta. The infra-renal portions of aorta and inferior vena cava (IVC) were exposed at a site where the two vessels share a common fascia. At this site, an 8-0 polypropylene (Ethicon Inc., Johnson & Johnson, NJ, USA) purse suture was applied on aorta. The supra-renal portion of abdominal aorta was then occluded with a 3-0 suture wire (Ethicon Inc., Johnson&Johnson, NJ, USA) to control for bleeding. The shunt was
then created with use of an 18-gauge Neo Delta Ven 2 catheter (Delta Med Mantova, Italy) inserted between the purse suture on anterior surface of aorta, carefully puncturing the posterior aortic wall up to the adjoining IVC with needle, and then advancing catheter into IVC. The purse suture was then tightened against the catheter and tied upon its withdrawal, while the suture occluding aorta was released. Before catheter withdrawing a second suture, a 5-0 polypropylene (Ethicon Inc., Johnson&Johnson, NJ, USA) was applied, creating a loop around the catheter by sticking together the ends of the suture with surgical glue. This permitted the closure of the shunt before hemodynamic analysis at the end of the observation period. After catheter withdrawal, the success of the procedure was confirmed by observing mixing of arterial and venous blood in IVC with distension and pulsations in the vessel (Fig. 1). Rats in sham group underwent same procedure except for shunt. During preliminary experiments, the measure of the catheter for shunt creation was determined in order to have Qp/Qs greater than 2. In detail, after shunt creation blood samples (0.5 ml) were obtained from pulmonary artery, aorta and superior vena cava, Qp/Qs was calculated by the formula Qp/Qs = (% saturation in the aorta − % saturation in the superior vena cava) / (% saturation in pulmonary vein − % saturation in pulmonary artery). When saturation in aorta was N95%, saturation in pulmonary vein was considered 100%, while when saturation in aorta was b95%, saturation in pulmonary vein was considered 95%. Magnetic resonance After the established observation period for each group, rats underwent a quantitative magnetic resonance imaging (MRI) to assess structural and functional parameters of the heart in vivo. All MRI experiments were carried out using a Biospec Tomograph System (Bruker, Karlsruhe, Germany) equipped with a 4.7-T, 33-cm bore horizontal magnet (Oxford Ltd, UK). Rats were placed in prone position into a 7.2-cm transmitter/receiver birdcage coil. Electrocardiographic and respiration rate recordings were continuously monitored by a physiological monitor for animals compatible with magnetic fields (PC-SAM, SAII, NY, USA). The imaging session started with acquisition of 3 Fast Gradient Echo (FLASH) images in order to identify long and short axis of the heart. Then cine-FLASH images were acquired in the short axis of the heart with electrocardiographic and respiration gating. Several contiguous slices were acquired in order to cover whole cardiac volume. Parameters of cine-FLASH acquisitions were as follows: repetition time = RR − interval / number of frames, with typically 20–22 frames acquired depending on the cardiac frequency and on the minimum repetition time of the sequence. The number of frames was calculated as follows: number of frames = cardiac period / repetition time. Other parameters were as follows: field of view = 6 × 6 cm2; repetition time = 10 ms; echo time = 3.6 ms; slice thickness = 1.5 cm; number of averages = 4; matrix size = 192 × 192 with an in-plane space resolution of 312 × 312 μm2. All images were manually analyzed by using software tools available in ParaVison 4.0 (Bruker, Karlsruhe, Germany). Volumetric measurements, such as thickness of the ventricular wall and intraventricular septum, ventricle mass (VM), end-diastolic volume (EDV), endsystolic volume (ESV), stroke volume (SV) and indices of global ventricular function such as ejection fraction (EF), were obtained. VM was derived from the corresponding myocardial volumes obtained from the difference between outer and inner volumes and assuming a myocardial density of 0.94 g/cm−3. SV was defined as the difference between EDV and ESV. EF was defined as the ratio between SV/EDV × 100. Echocardiography Transthoracic two-dimensional, M-mode and Doppler (pulse wave and continuous wave) imaging was performed using a Vivid Q GE cardiovascular ultrasound system, with a 10-s transducer (from 5 to 7.5 MHz).
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
75
Fig. 1. Surgical creation of aorto-caval shunt. (A) The infra-renal portion of the aorta (Ao) and inferior vena cava (IVC) are shown. An 8-0 polypropylene purse suture was applied on the anterior wall of the aorta. An 18-gauge catheter was inserted between the purse suture on the anterior surface of the aorta and then advanced into the IVC. (B) The success of the procedure was confirmed by observing mixing of arterial and venous blood in the IVC with distension and pulsations in the IVC.
The analysis of RV free wall thickness and RV end diastolic area was performed in a parasternal short axis view at the level of the papillary muscles. From an apical 4-chamber view, an estimate of systolic tricuspid annular excursion (TAPSE) was also obtained. A continuous-wave Doppler of pulmonary blood flow was obtained in parasternal short axis view, this time at the level of aortic valve. For the purpose of this study, the development of a mid-systolic notch in pulmonary artery waveform was considered as a reliable echocardiographic sign of increased pulmonary arterial pressure, in agreement with prior evidence (Jones et al., 2002). Pressure–volume analysis Rats were placed on a heating pad to maintain normothermia. Trachea was cannulated with a 16-G soft venous cannula and mechanically ventilated with 100% oxygen at 70 cycles/min with a tidal volume of 4 body weight (model 683, Harvard Apparatus, Holliston, MA). Right carotid artery was isolated and closed cranially. A 2-Fr miniaturized, combined catheter-micromanometer (model SPR 838, Millar Instruments, Houston, TX) was inserted into right carotid artery and then advanced into left ventricle (LV). The correct position of the catheter was checked by the shape of pressure waveform (from a typical artery shape when placed in carotid artery and then in aorta to a ventricular shape when placed into LV). The signals were continuously recorded at a sampling rate of 1000/s using an ARIA pressure–volume (P–V) conductance system (Millar Instruments) coupled to a PowerLab/4SP A/D converter (AD Instruments; Mountain View, CA) and a personal computer. The volume calibration of conductance system was performed as described previously (Pacher et al., 2004). Systolic pressure, end-diastolic pressure, mean arterial pressure, maximal peak systolic pressure increment (dP/dtMAX) and diastolic pressure decrement (dP/dtMIN), EF, SV, EDV, time constant of ventricular relaxation (Tau, according to the Weiss and Glantz method) (Weiss et al., 1976) and stroke work were computed using a cardiac P–V analysis program (PVAN 3.2, Millar Instruments). To obtain load-independent parameters such as preload recruitable stroke work (PRSW) and end-systolic pressure–volume relationship (ESPVR) and end-diastolic pressure–volume relationship (EDPVR), P–V relations were measured by transiently occluding the inferior vena cava under the diaphragm with a small vascular clamp (Nakano et al., 1990). Maximal slope of ESPVR defined ventricle systolic elastance (Ees). The combined catheter-micromanometer (model SPR 838, Millar Instruments, Houston, TX) was then inserted into RV through jugular vein, and the same parameters previously described were recorded. In addition, effective pulmonary arterial elastance (Ea computed as RV end systolic pressure/SV) was calculated as index of pulmonary vascular
load. The catheter was then slowly advanced into RV outflow tract and then in pulmonary artery, pulmonary artery systolic and diastolic pressure were recorded.
Histology After rats had been sacrificed, lungs were rapidly perfused with formalin through pulmonary artery, inflated through trachea and removed. Three pieces of lung tissue from different lobes were excised and immersed in 10% formalin and sectioned. Hematoxylin–eosin staining was subsequently performed. Pulmonary artery morphology was observed and the percentages of muscularized and non-muscularized Table 1 Pressure–volume, echocardiographic and magnetic resonance parameters. Sham Pressure–volume parameters mPAP (mmHg) 21.09 sPAP (mmHg) 28.78 dPAP (mmHg) 11.45 RV EDP (mmHg) 5.05 RV dP/dtMAX (mmHg/s) 1500 1980 RV dP/dtMIN (mmHg/s) RV PRSW (mmHg) 34.05 RV Tau (ms) 16.12 LV EDP (mmHg) 6.05 8120 LV dP/dtMAX (mmHg/s) LV dP/dtMIN (mmHg/s) 5980 LV PRSW (mmHg) 96.65 LV Tau (ms) 13.82
± ± ± ± ± ± ± ± ± ± ± ± ±
1.52 9.34 2.60 0.60 160 465 9.33 4.30 1.40 930 370 10.33 4.60
Shunt 10W
Shunt 20W
24.81 72.23 16.45 13.88 1810 992 28.25 26.20 12.80 4030 3892 81.20 18.00
43.62 105.34 56.83 32.04 975 1045 19.01 34.45 17.03 3128 2890 38.04 42.10
± ± ± ± ± ± ± ± ± ± ± ± ±
4.30 10.02⁎ 4.10 1.91⁎ 498 498⁎ 6.15 3.82⁎ 4.21⁎ 312⁎ 288⁎ 6.10 8.81
± ± ± ± ± ± ± ± ± ± ± ± ±
2.85⁎,§ 23.40⁎,§ 9.71⁎,§ 11.22⁎,§ 155⁎ 98⁎ 3.55⁎ 8.72⁎ 2.30⁎ 101⁎,§ 97⁎,§ 8.95⁎ 9.73⁎
Echocardiographic parameters 0.23 ± 0.90 RV EDA (cm2) RV WT (cm) 0.20 ± 0.05 TAPSE (mm) 2.25 ± 0.05
0.35 ± 0.06⁎ 0.35 ± 0.05⁎ 1.97 ± 0.05
0.36 ± 0.04⁎ 0.33 ± 0.05⁎ 1.93 ± 0.25
Magnetic resonance parameters RV M/BW 0.3 RV SV (μl) 251 RV EF (%) 65 LV M/BW 1.3 LV SV (μl) 330 LV EF (%) 62
0.5 322 59 1.7 333 56
0.4 245 53 1.4 302 50
± ± ± ± ± ±
0.09 5 2 0.4 5 3
± ± ± ± ± ±
0.1 11⁎ 6 0.3 6 7
± ± ± ± ± ±
0.07 7⁎,§ 2⁎ 0.4 18 5⁎
mPAP = mean pulmonary artery pressure; sPAP = systolic pulmonary artery pressure; dPAP = diastolic pulmonary artery pressure; RV = right ventricle; EDP = end-diastolic pressure; PRSW = preload recruitable stroke work; LV = left ventricle; sPAP = systolic pulmonary artery pressure; EDA = end-diastolic area; WT = free wall thickness; TAPSE = tricuspid annular plane systolic excursion; M/BW = mass normalized on corresponding body weight; SV = stroke volume; EF = ejection fraction. Data are presented as mean ± SEM. ⁎ p b 0.05 vs. sham. § p b 0.05 vs. shunt 10W.
76
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
pulmonary arteries of the total amount of small pulmonary arteries were calculated. Statistical analysis Data are presented as means ± standard deviations. Differences between groups were determined using two-way analysis of variance (ANOVA) and Bonferroni's multiple comparison tests with statistical significance set at p b 0.05. Results All rats survived surgical procedure of shunt creation (Fig. 1). During the planned follow-up, 4 animals died among shunt 20W group. A postmortem examination ascertained postoperative heart failure in all the four cases, with relevant pleural and pericardial effusion. Qp/Qs was 2.12 ± 0.54 in shunt 10W, 1.98 ± 0.41 in shunt 20W and 1 ± 0.12 at 10 and 20 weeks in the control (sham). In all rats with shunt, Qp/Qs was significantly increased (p b 0.05) compared to control and gave evidence of shunt patency. Right ventricle Table 1 illustrates P–V parameters and echocardiographic and MRI data. RV EDV and ESV increased in shunt groups with a significant thickness in RV free wall. This was confirmed by an augmentation of RV mass, expressed as normalized to corresponding body weight.
Functional evaluation demonstrated a significant increase in RV enddiastolic pressure in shunt groups, in parallel with a progressive reduction in RV systolic functions parameters (dP/dtMAX and PRSW), which was significant at 20 weeks. dP/dtMIN augment and a corresponding diminishing in Tau were suggestive of impaired RV diastolic function (Fig. 3). Right ventricle–arterial coupling The maximal slope of RV ESPVR, i.e., RV Ees, was lower in shunt 20W compared to both shunt 10W and control (p b 0.01). Pulmonary artery Ea increased in shunt 20W (p b 0.01) and was consistent with an increase in mean pulmonary arterial pressure (p b 0.01). Therefore, the Ees/Ea ratio was affected, demonstrating an uncoupling of RV contractility and RV afterload (Fig. 3). Ventricular interdependence Both RV and LV masses, expressed as normalized to corresponding body weights, as well as cardiac chambers volumes increased in shunt groups (Table 1). LV EDV and ESV significantly increased in shunt 20W. EDPVR, PRSW and dP/dtMAX demonstrated an impaired systolic function in shunt 20W (Figs. 2 and 3). LV dP/dtMIN decreased significantly in both shunt groups and LV Tau was higher in shunt 20W (p b 0.05). These modifications suggested impaired diastolic ventricular function and a compromised ventricular relaxation.
Fig. 2. Magnetic resonance imaging and pressure–volume loops. Morphologic and functional evaluation by MRI (first line) and conductance pressure–volume catheter (second line) in representative cases of sham, shunt at 10 weeks and 20 weeks.
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
77
Fig. 3. Hemodynamic analysis. (A–B) Right ventricle afterload was expressed by pulmonary artery effective elastance (PA Ea) and mean pulmonary artery pressure (MPAP). (C–D) Right ventricle systolic and diastolic functions measured by systolic elastance (RV Ees) and time constant of ventricular relaxation Tau (RV Tau), respectively. (E) Right ventricle–arterial coupling defined as the ratio between ventricle systolic elastance and effective pulmonary artery elastance (Ees/Ea). (F) Ventricular interdependence was measured by left ventricle end-diastolic volume (LV EDV).
Histological analysis Cross-sectional slices of the ventricles demonstrated biventricular dilatation and augmentation of RV free wall in shunt 20W (Fig. 4). Small pulmonary artery morphology detected by hematoxylin– eosin staining revealed a non-significant medial hypertrophy in shunt 10W compared with sham. However, after 20 weeks of overcirculation exposure, a significant thickening of tunica media with muscular hypertrophy was evident. Masson's trichrome stain revealed low fibrosis rate. In some cases of shunt 20W, it was possible to demonstrate a complete occlusion of the intralobular arteries due to concentric intimal hyperplasia. No advanced plexiform lesions, sign of severe and irreversible PH were observed (Fig. 4).
Discussion In the present study, aorto-caval shunt rat model was demonstrated to closely reproduce the aberrant hemodynamic status and characteristic morphologic changes observed in the lungs of patients with PH.
Historically, the most widely used animal models of PH have been chronic hypoxia and the monocrotaline-induced PH rodent model. Although these models have added enormously to the understanding of the mechanisms of pulmonary vascular remodeling in PH, they are truly limited by different mechanisms of pathogenesis and disease phenotype compared to clinical setting (Dickinson et al., 2013). In these animal models, recent trials have shown that mono-therapy or combination therapy using targeted drugs exerted preventive, or even reversible, efficacy on hypoxia/monocrotaline-induced PH (Itoh et al., 2004). When administered to patients with severe PH, however, no drugs resulted in significantly lower mortality. This discrepancy has led researchers to reconsider the inadequacy and limitations of hypoxia/monocrotaline-induced PH (Bauer et al., 2007; Zaiman et al., 2005). With PH research shifting from a concept of mere vasoconstriction toward a mechanism of angioproliferation, the role of disturbed blood flow is seen as being pivotal (both clinically and experimentally) in PH development. Therefore, the need for a satisfactory and simple shunt-related model led to numerous attempts to simulate the pulmonary vasculopathy by surgically creating peripheral artero-venous or central systemic-to-pulmonary communications with different and
78
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
Fig. 4. Histology. (A–B) Progressive bi-ventricular dilatation with right ventricle hypertrophy in a case of shunt at 20 weeks (B) compared to sham operated rat (A). (C–D) Small pulmonary artery morphology detected by hematoxylin–eosin staining in sham (C) and shunt 20W (D). (E–F) Percentage of non-muscularized pulmonary artery (E) and muscularized pulmonary artery (F) in small pulmonary artery.
conflicting results (Bauer et al., 2007; Itoh et al., 2004; Tenmark and McMurtry, 2005; Zaiman et al., 2005). Many animal studies inducing increased pulmonary blood flow have used restrictive shunts resulting in the increased muscularization of the pulmonary arteries but with no rise in pulmonary arterial pressure (Black et al., 2003; Rondelet et al., 2012). This is in analogy to the clinical observation that patients with untreated pre-tricuspid shunts, such as atrial septal defect, in general develop advanced pulmonary vascular remodeling only in 5%–15% of the cases and only after two to three decades. In contrast, patients with untreated non-restrictive, posttricuspid shunts, such as ventricular septal defects, develop progressive PH in 1–2 years in 80%–100% of the cases and, eventually, Eisenmenger
syndrome, if left without surgical treatment (Van Albada and Berger, 2008). The aorto-caval shunt model is known to be a reproducible, simple and rapid method to develop high output heart failure and cardiac hypertrophy. Lam et al. (2005)) firstly observed the signs of pulmonary hypertension in rats with aorto-caval fistula. Although volume overload was previously demonstrated by Garcia and Diebold (1990), the observation of 4 weeks was too short to induce pulmonary vascular remodeling. The progressive development of PH during chronic exposure to aorto-caval shunt was demonstrated by the present study. For the first time, a detailed hemodynamic evaluation and a state-of-the-art imaging
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
were applied to address ventricular–vascular interactions in a rat pulmonary artery overflow model. Echocardiogram and MRI showed a progressive increase in RV volume and end-diastolic area. The presence of a mesosystolic notch in Doppler profile of pulmonary blood flow in shunt at 20W was suggestive of increased systolic pulmonary pressure as demonstrated by Jones et al. (2002). With pressure–volume catheter, PH was confirmed by the direct measurement of increased pulmonary pressure and increased effective pulmonary Ea, which is the most accurate index of pulmonary vascular load (Morimont et al., 2008) and is related to the increase of pulmonary artery stiffness (Weiss et al., 1976). Further evidence of development of a progressive pulmonary vascular remodeling related to hypertension in rats was provided by histological analysis of lungs. A progressive thickening of muscular lamina after 10 and 20 weeks of shunt was demonstrated. In the group of 20-week overload exposure, we found a complete occlusion of the intralobular arteries due to concentric intimal thickening (Fig. 4). The coupling between pulmonary artery Ea (index of arterial load) and RV Ees (index of contractility) is defined by the Ees/Ea ratio, and it is an effective index of the mechanical performance of RV and the dynamic modulation of pulmonary circulation. In particular, Ea increased and Ees decreased, leading to uncoupling the right side of cardiovascular apparatus. This reflected an imbalance between myocardial oxygen consumption and mechanical energy request to perform hemodynamic work. Evaluation of the Ees/Ea ratio is nowadays recognized as an adjunctive perspective for understanding the pathophysiology of altered hemodynamic profiles, and for guiding therapeutic strategies and testing the effectiveness of treatments. Moreover, we addressed the interdependence between RV and LV. In particular, the overloaded RV had a direct impact on LV performance through serial ventricle interactions (such as failure to produce anterograde filling of LV) and parallel ventricular interactions such as the disturbance of leftward shifting of the septum. This ventricular interdependence was demonstrated by observation of leftward shift of interventricular septum (MRI) that leaded to impaired LV preload (EDV). In the present work, aorto-caval shunt generated Qp/Qs ≈ 2, which usually describes a moderate pre-tricuspid shunt, rarely generating PH in humans. This could limit the present model for translation to clinic. Nonetheless, clinical observations are gathering showing how longterm exposure to atrial level shunts may produce even severe PH (Bradley et al., 2013; Goetschmann et al., 2008; Luciani et al., 2008). In addition, atrial level shunt, with or without severe PH, and PH per se have since been shown to cause LV dysfunction by a variety of mechanisms, including ventricular remodeling, decreased myofiber preload and ventricular interdependence (Haeck et al., 2014; Hardegree et al., 2013; Walker et al., 2004). Finally, a possibility exists that severe PH observed in present experimental setting may be explained by species-specific differences, thereby representing a possible limitation to translation of results to human pathology. However, a period of 20 weeks relative to an average life span of 2 years represents a remarkably long exposure and could explain the development of pulmonary hypertension in this animal model. Conclusion Chronic exposure to aorto-caval shunt is a reliable model to produce RV volume overload and secondary PH. Timing of overcirculation exposure has a pivotal role in the development of PH. Therefore, this model, with an adequate time of follow-up, could be an alternative with low mortality and high reproducibility for investigators on the underling mechanisms of shunt-related PH. References Badesch, D.B., Champion, H.C., Sanchez, M.A., et al., 2009. Diagnosis and assessment of pulmonary arterial hypertension. J. Am. Coll. Cardiol. 54, S55–S66.
79
Bai, Y., Sun, L., Hu, S., et al., 2011. Combination therapy in pulmonary arterial hypertension: a meta-analysis. Cardiology 120, 157–165. Bauer, N.R., Moore, T.M., McMurtry, I.F., 2007. Rodent models of PAH: are we there yet? Am. J. Physiol. Lung Cell. Mol. Physiol. 293, L580–L582. Black, S.M., Mata-Greenwood, E., Dettman, R.W., et al., 2003. Emergence of smooth muscle cell endothelin B-mediated vasoconstriction in lambs with experimental congenital heart disease and increased pulmonary blood flow. Circulation 108, 1646–1654. Bradley, E.A., Chakinala, M., Billadello, J.J., 2013. Usefulness of medical therapy for pulmonary hypertension and delayed atrial septal defect closure. Am. J. Cardiol. 11, 1471–1476. Dickinson, M.G., Bartelds, B., Borgdorff, M.A.J., et al., 2013. The role of disturbed blood flow in the development of pulmonary arterial hypertension: lessons from preclinical animal models. Am. J. Physiol. Lung Cell. Mol. Physiol. 305, L1–L14. Fullerton, D.A., Mitchell, M.B., Jone, D.N., et al., 1996. Pulmonary vasomotor dysfunction is produced with chronically high pulmonary blood flow. J. Thorac. Cardiovasc. Surg. 111, 190–197. Garcia, R., Diebold, S., 1990. Simple, rapid and effective method of producing aortocaval shunts in the rat. Cardiovasc. Res. 24, 430–432. Goetschmann, S., Dibernardo, S., Steinmann, H., et al., 2008. Frequency of severe pulmonary hypertension complicating “isolated” atrial septal defect in infancy. Am. J. Cardiol. 102, 340–342. Haeck, M.L., Höke, U., Marsan, N.A., et al., 2014. Impact of right ventricular dyssynchrony on left ventricular performance in patients with pulmonary hypertension. Int. J. Cardiovasc. Imaging 30, 713–720. Hardegree, E.L., Sachdev, A., Fenstad, E.R., et al., 2013. Impaired left ventricular mechanics in pulmonary arterial hypertension: identification of a cohort at high risk. Circ. Heart Fail. 6, 748–755. Itoh, T., Nagaya, N., Fujii, T., et al., 2004. A combination of oral sildenafil and beraprost ameliorates pulmonary hypertension in rats. Am. J. Respir. Crit. Care Med. 169, 34–38. Jones, J., Mendes, L., Rudd, M.A., et al., 2002. Serial noninvasive assessment of progressive pulmonary hypertension in a rat model. Am. J. Physiol. Heart Circ. Physiol. 283, H364–H371. Lam, C.F., Peterson, T.E., Croatt, A.J., et al., 2005. Functional adaptation and remodeling of pulmonary artery in flow-induced pulmonary hypertension. Am. J. Physiol. Heart Circ. Physiol. 289, H2334–H2341. Luciani, G.B., Viscardi, F., Pilati, M., et al., 2008. Age at repair affects the very long-term outcome of sinus venosus defect. Ann. Thorac. Surg. 86, 153–159. Morimont, P., Lambermont, B., Ghuysen, A., et al., 2008. Effective arterial elastance as an index of pulmonary vascular load. Am. J. Physiol. Heart Circ. Physiol. 294, H2736–H2742. Nakano, K., Sugawara, M., Ishihara, K., et al., 1990. Myocardial stiffness derived from endsystolic wall stress and logarithm of reciprocal of wall thickness. Contractility index independent of ventricular size. Circulation 82, 1352–1361. Nishimura, T., Faul, J.L., Berry, G.L., et al., 2003. Effect of a surgical aortocaval fistula on monocrotaline-induced pulmonary hypertension. Crit. Care Med. 31, 1213–1218. Ocampo, C., Ingram, P., Ilbawi, M., et al., 2003. Revisiting the surgical creation of volume load by aorto-caval shunt in rats. Mol. Cell. Biochem. 251, 139–143. Pacher, P., Mabley, J.G., Liaudet, L., et al., 2004. Left ventricular pressure–volume relationship in a rat model of advanced aging-associated heart failure. Am. J. Physiol. Heart Circ. Physiol. 287, H2132–H2137. Reddy, V.M., Meyrick, B., Wong, J., et al., 1995. In utero placement of aorta pulmonary shunts: a model of postnatal pulmonary hypertension with increased pulmonary blood flow in lambs. Circulation 92, 606–613. Rondelet, B., Dewachter, C., Kerbaul, F., et al., 2012. Prolonged overcirculation-induced pulmonary arterial hypertension as a cause of right ventricular failure. Eur. Heart J. 33, 1017–1026. Szabo, G., Soos, P., Bährle, S., et al., 2006. Adaptation of the right ventricle to an increased afterload in the chronically volume overloaded heart. Ann. Thorac. Surg. 82, 989–995. Tenmark, K.R., McMurtry, I.F., 2005. Vascular remodeling versus vasoconstriction in chronic hypoxic pulmonary hypertension: a time for reappraisal? Circ. Res. 97, 95–98. Van Albada, M.E., Berger, R.M., 2008. Pulmonary arterial hypertension in congenital cardiac disease—the need for refinement of the Evian–Venice classification. Cardiol. Young 18, 10–17. Van Albada, M.E., Schoemmaker, R.G., Kemna, M.S., et al., 2005. The role of increased pulmonary blood flow in pulmonary arterial hypertension. Eur. Respir. J. 26, 487–493. Walker, R.E., Moran, A.M., Gauvreau, K., et al., 2004. Evidence of adverse ventricular interdependence in patients with atrial septal defects. Am. J. Cardiol. 93, 1374–1377. Wang, W., Liu, R., Cao, G., et al., 2010. A reliable rabbit model for hyperkinetic pulmonary hypertension. J. Thorac. Cardiovasc. Surg. 140, 395–399. Wang, X., Ren, B., Liu, S., et al., 2003. Characterization of crdiac hypertrophy and heart failure due to volume overload in the rat. J. Appl. Physiol. 294, 752–763. Weiss, J.L., Frederiksen, J.W., Weisfeldt, M.L., 1976. Hemodynamic determinants of the time-course of fall in canine left ventricular pressure. J. Clin. Invest. 58, 751–760. Xiong, M., Yao, J.P., Wu, Z.K., et al., 2012. Fibrosis of pulmonary vascular remodeling in carotid artery-jugular vein shunt pulmonary artery hypertension model of rats. Eur. J. Cardiothorac. Surg. 41, 162–166. Yerebakan, C., Klopsch, C., Niefeldt, S., et al., 2010. Acute and chronic response of the right ventricle to surgically induced pressure and volume overload, an analysis of pressure–volume relations. Interact. Cardiovasc. Thorac. Surg. 10, 519–525. Zaiman, A., Fijalkowska, I., Hassoun, P.M., et al., 2005. One hundred years of research in the pathogenesis of pulmonary hypertension. Am. J. Respir. Cell Mol. Biol. 33, 425–431.
Contents lists available at ScienceDirect
Microvascular Research journal homepage: www.elsevier.com/locate/ymvre
Chronic overcirculation-induced pulmonary arterial hypertension in aorto-caval shunt Alessio Rungatscher a,⁎, Daniele Linardi a, Elisabetta Milani a, Grazia Ucci b, Elena Nicolato c, Flavia Merigo c, Beatrice Salvetti c, Alessandro Mazzucco a, Giovanni Battista Luciani a, Giuseppe Faggian a a b c
Department of Surgery, Division of Cardiac Surgery, University of Verona Verona, Italy Division of Cardiology, University of Novara Novara, Italy Department of Neurological and Movement Sciences, University of Verona Verona, Italy
a r t i c l e
i n f o
Article history: Received 1 January 2014 Revised 9 May 2014 Accepted 15 May 2014 Available online 23 May 2014 Keywords: Pulmonary arterial hypertension Vascular remodeling Right ventricle overload Volume overload Ventricular vascular coupling Ventricular interdependence Aorto-caval shunt Rats Pressure-volume loops Experimental model
a b s t r a c t Pulmonary arterial hypertension is a common complication of congenital heart defects with left-to-right shunts. Current preclinical models do not reproduce clinical characteristics of shunt-related pulmonary hypertension. Aorto-caval shunt was firstly described as a model of right ventricle volume overload. The pathophysiology and the possible determination of pulmonary arterial hypertension of different periods of shunt exposure are still undefined. A method to create standardized, reproducible aorto-caval shunt was developed in growing rats (260 ± 40 g). Three groups of animals were considered: shunt exposure for 10 weeks, shunt exposure for 20 weeks and control (sham laparotomy). Echocardiography and magnetic resonance revealed increased right ventricular end diastolic area in shunt at 10 weeks compared to control. Hemodynamic analysis demonstrated increased right ventricular afterload and increased effective pulmonary arterial elastance (Ea) in shunt at 20 weeks compared to control (1.29 ± 0.20 vs. 0.14 ± 0.06 mmHg/μl, p = 0.004). At the same time point, the maximal slope of end-systolic pressure– volume relationship (Ees) decreased (0.5 ± 0.2 mmHg/ml vs. 1.2 ± 0.3, p b 0.001). Consequently, right ventricular–arterial coupling was markedly deteriorated with a ≈50% decrease in the ratio of end-systolic to pulmonary artery elastance (Ees/Ea). Finally, left ventricular preload diminished (≈ 30% decrease in left ventricular end-diastolic volume). Histology demonstrated medial hypertrophy and small artery luminal narrowing. Chronic exposure to aorto-caval shunt is a reliable model to produce right ventricular volume overload and secondary pulmonary arterial hypertension. This model could be an alternative with low mortality and high reproducibility for investigators on the underlying mechanisms of shunt-related pulmonary hypertension. © 2014 Elsevier Inc. All rights reserved.
Introduction Left-to-right shunting congenital heart diseases are the common cause of secondary pulmonary arterial hypertension (PH). The status of the pulmonary vascular structure and extent of PH are important determinants of feasibility of corrective procedures and long-term survival after cardiac operation (Badesch et al., 2009).
Abbreviations: Ea, effective pulmonary arterial elastance; Ees, maximal slope of end systolic pressure–volume relationship; PH, pulmonary hypertension; RV, right ventricle; IVC, inferior vena cava; MRI, magnetic resonance imaging; VM, ventricle mass; EDV, end-diastolic volume; ESV, end-systolic volume; SV, stroke volume; EF, ejection fraction; TAPSE, systolic tricuspid annular excursion; LV, left ventricle; P–V, pressure–volume; PRSW, preload recruitable stroke work; ESPVR, end-systolic pressure–volume relationship; EDPVR, end-diastolic pressure–volume relationship. ⁎ Corresponding author at: Department of Surgery, Division of Cardiac Surgery, University of Verona, Piazzale Stefani 1, 37126 Verona, Italy. Fax: +39 0458123308. E-mail address: [email protected] (A. Rungatscher).
http://dx.doi.org/10.1016/j.mvr.2014.05.005 0026-2862/© 2014 Elsevier Inc. All rights reserved.
During past decades, animal models have deduced numerous cellular and molecular mechanisms of PH. Although much insight has been gained from “classic” PH models, such as hypoxic PH or monocrotalineinduced PH, most of these studies are hampered by the fact that these models do not truly reflects the typical features of PH in respect of histopathology and progressive course (Dickinson et al., 2013). Indeed, while drugs targeting the signaling pathways involved in the pathogenesis of PH have been confirmed to partially or completely reverse hypoxia and monocrotaline-induced PH in rats, a disappointing discrepancy has emerged in recent clinical trials (Bai et al., 2011). Hence, there is an increasing interest in models that more reliably reflect the complex disease PH. Because increased pulmonary blood flow appears to be crucial factor in pathogenesis, particular emphasis is given in models of increased pulmonary blood flow that lead to development of pulmonary vascular remodeling. Continuous attempts have been made to simulate the pulmonary vasculopathy by surgically creating peripheral or central systemic-to-
74
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
pulmonary communication (Fullerton et al., 1996; Reddy et al., 1995; Wang et al., 2010; Xiong et al., 2012). Central shunts prepared in large animals closely simulate the pathophysiologic characteristics of congenital intracardiac shunts; however, the high-level surgical requirements, relatively greater mortality, limited patency in inducing remodeling lesions and inadequacy of biochemical reagents for large animals have impeded the widespread adoption of these models. Thus, a satisfactory shunt-related model that could be used to determine the underlying molecular mechanism of pulmonary vascular remodeling due to in vivo exposure to high blood flow is still needed. One pre-tricuspid model, the aorto-caval shunt, is relatively widely applied and was firstly described by Garcia and Diebold (1990) later modified by others (Ocampo et al., 2003). The aorto-caval shunt model was used by several authors to study right ventricle (RV) response to acute, sub-acute and chronic volume overload (Szabo et al., 2006; Wang et al., 2003; Yerebakan et al., 2010), while other authors used this model to study the effects of increased pulmonary blood flow on pulmonary circulation (Nishimura et al., 2003; Van Albada et al., 2005). However, results reported in literature are conflicting, especially when the model is used to produce secondary pulmonary hypertension, and the exact mechanism, which leads to development of pulmonary hypertension in this model, is not clear. The aim of the present study was to demonstrate with morphologic, functional and hemodynamic analysis the possible induction of secondary PH as a consequence of different periods of RV volume overload and pulmonary overcirculation. Methods Male Sprague–Dawley rats (Charles River Laboratories, Calco, Italy) were used (age, 8 weeks; weight, 260 ± 40 g). All experiments were performed in adherence to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (National Institutes of Health Publication No. 85-23, revised 1996). The protocol was approved by the University of Verona authority for experimental research (C.I.R.S.A.L. Interdepartmental Centre of Experimental Research Service). Experimental protocol Rats were randomly divided into 3 groups: the first group (n = 10) underwent shunt procedure and was observed for 10 weeks after procedure (shunt 10W); thereafter, shunt was closed and non-invasive morphometric and hemodynamic evaluations were performed. The second group (n = 10) underwent the same surgical procedure but was observed for 20 weeks before shunt closure and subsequent evaluation (shunt 20W). The third group (n = 10) served as control, underwent a sham laparotomy operation and then was observed for 20 weeks (sham). All surgical procedures and invasive and non-invasive evaluations were conducted under general anesthesia (0.25% pentobarbital sodium, 40 mg/kg, intraperitoneal injection). At the end of the programmed observation period, rats were sacrificed and the samples for histology were collected. Surgical creation of the shunt A midline abdominal incision was made, and a descending aorta above renal bifurcation was cleared of adjacent tissues. There a 3-0 Ti-Cron suture (Covidien, Mansfield, MA, USA) was positioned around the aorta. The infra-renal portions of aorta and inferior vena cava (IVC) were exposed at a site where the two vessels share a common fascia. At this site, an 8-0 polypropylene (Ethicon Inc., Johnson & Johnson, NJ, USA) purse suture was applied on aorta. The supra-renal portion of abdominal aorta was then occluded with a 3-0 suture wire (Ethicon Inc., Johnson&Johnson, NJ, USA) to control for bleeding. The shunt was
then created with use of an 18-gauge Neo Delta Ven 2 catheter (Delta Med Mantova, Italy) inserted between the purse suture on anterior surface of aorta, carefully puncturing the posterior aortic wall up to the adjoining IVC with needle, and then advancing catheter into IVC. The purse suture was then tightened against the catheter and tied upon its withdrawal, while the suture occluding aorta was released. Before catheter withdrawing a second suture, a 5-0 polypropylene (Ethicon Inc., Johnson&Johnson, NJ, USA) was applied, creating a loop around the catheter by sticking together the ends of the suture with surgical glue. This permitted the closure of the shunt before hemodynamic analysis at the end of the observation period. After catheter withdrawal, the success of the procedure was confirmed by observing mixing of arterial and venous blood in IVC with distension and pulsations in the vessel (Fig. 1). Rats in sham group underwent same procedure except for shunt. During preliminary experiments, the measure of the catheter for shunt creation was determined in order to have Qp/Qs greater than 2. In detail, after shunt creation blood samples (0.5 ml) were obtained from pulmonary artery, aorta and superior vena cava, Qp/Qs was calculated by the formula Qp/Qs = (% saturation in the aorta − % saturation in the superior vena cava) / (% saturation in pulmonary vein − % saturation in pulmonary artery). When saturation in aorta was N95%, saturation in pulmonary vein was considered 100%, while when saturation in aorta was b95%, saturation in pulmonary vein was considered 95%. Magnetic resonance After the established observation period for each group, rats underwent a quantitative magnetic resonance imaging (MRI) to assess structural and functional parameters of the heart in vivo. All MRI experiments were carried out using a Biospec Tomograph System (Bruker, Karlsruhe, Germany) equipped with a 4.7-T, 33-cm bore horizontal magnet (Oxford Ltd, UK). Rats were placed in prone position into a 7.2-cm transmitter/receiver birdcage coil. Electrocardiographic and respiration rate recordings were continuously monitored by a physiological monitor for animals compatible with magnetic fields (PC-SAM, SAII, NY, USA). The imaging session started with acquisition of 3 Fast Gradient Echo (FLASH) images in order to identify long and short axis of the heart. Then cine-FLASH images were acquired in the short axis of the heart with electrocardiographic and respiration gating. Several contiguous slices were acquired in order to cover whole cardiac volume. Parameters of cine-FLASH acquisitions were as follows: repetition time = RR − interval / number of frames, with typically 20–22 frames acquired depending on the cardiac frequency and on the minimum repetition time of the sequence. The number of frames was calculated as follows: number of frames = cardiac period / repetition time. Other parameters were as follows: field of view = 6 × 6 cm2; repetition time = 10 ms; echo time = 3.6 ms; slice thickness = 1.5 cm; number of averages = 4; matrix size = 192 × 192 with an in-plane space resolution of 312 × 312 μm2. All images were manually analyzed by using software tools available in ParaVison 4.0 (Bruker, Karlsruhe, Germany). Volumetric measurements, such as thickness of the ventricular wall and intraventricular septum, ventricle mass (VM), end-diastolic volume (EDV), endsystolic volume (ESV), stroke volume (SV) and indices of global ventricular function such as ejection fraction (EF), were obtained. VM was derived from the corresponding myocardial volumes obtained from the difference between outer and inner volumes and assuming a myocardial density of 0.94 g/cm−3. SV was defined as the difference between EDV and ESV. EF was defined as the ratio between SV/EDV × 100. Echocardiography Transthoracic two-dimensional, M-mode and Doppler (pulse wave and continuous wave) imaging was performed using a Vivid Q GE cardiovascular ultrasound system, with a 10-s transducer (from 5 to 7.5 MHz).
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
75
Fig. 1. Surgical creation of aorto-caval shunt. (A) The infra-renal portion of the aorta (Ao) and inferior vena cava (IVC) are shown. An 8-0 polypropylene purse suture was applied on the anterior wall of the aorta. An 18-gauge catheter was inserted between the purse suture on the anterior surface of the aorta and then advanced into the IVC. (B) The success of the procedure was confirmed by observing mixing of arterial and venous blood in the IVC with distension and pulsations in the IVC.
The analysis of RV free wall thickness and RV end diastolic area was performed in a parasternal short axis view at the level of the papillary muscles. From an apical 4-chamber view, an estimate of systolic tricuspid annular excursion (TAPSE) was also obtained. A continuous-wave Doppler of pulmonary blood flow was obtained in parasternal short axis view, this time at the level of aortic valve. For the purpose of this study, the development of a mid-systolic notch in pulmonary artery waveform was considered as a reliable echocardiographic sign of increased pulmonary arterial pressure, in agreement with prior evidence (Jones et al., 2002). Pressure–volume analysis Rats were placed on a heating pad to maintain normothermia. Trachea was cannulated with a 16-G soft venous cannula and mechanically ventilated with 100% oxygen at 70 cycles/min with a tidal volume of 4 body weight (model 683, Harvard Apparatus, Holliston, MA). Right carotid artery was isolated and closed cranially. A 2-Fr miniaturized, combined catheter-micromanometer (model SPR 838, Millar Instruments, Houston, TX) was inserted into right carotid artery and then advanced into left ventricle (LV). The correct position of the catheter was checked by the shape of pressure waveform (from a typical artery shape when placed in carotid artery and then in aorta to a ventricular shape when placed into LV). The signals were continuously recorded at a sampling rate of 1000/s using an ARIA pressure–volume (P–V) conductance system (Millar Instruments) coupled to a PowerLab/4SP A/D converter (AD Instruments; Mountain View, CA) and a personal computer. The volume calibration of conductance system was performed as described previously (Pacher et al., 2004). Systolic pressure, end-diastolic pressure, mean arterial pressure, maximal peak systolic pressure increment (dP/dtMAX) and diastolic pressure decrement (dP/dtMIN), EF, SV, EDV, time constant of ventricular relaxation (Tau, according to the Weiss and Glantz method) (Weiss et al., 1976) and stroke work were computed using a cardiac P–V analysis program (PVAN 3.2, Millar Instruments). To obtain load-independent parameters such as preload recruitable stroke work (PRSW) and end-systolic pressure–volume relationship (ESPVR) and end-diastolic pressure–volume relationship (EDPVR), P–V relations were measured by transiently occluding the inferior vena cava under the diaphragm with a small vascular clamp (Nakano et al., 1990). Maximal slope of ESPVR defined ventricle systolic elastance (Ees). The combined catheter-micromanometer (model SPR 838, Millar Instruments, Houston, TX) was then inserted into RV through jugular vein, and the same parameters previously described were recorded. In addition, effective pulmonary arterial elastance (Ea computed as RV end systolic pressure/SV) was calculated as index of pulmonary vascular
load. The catheter was then slowly advanced into RV outflow tract and then in pulmonary artery, pulmonary artery systolic and diastolic pressure were recorded.
Histology After rats had been sacrificed, lungs were rapidly perfused with formalin through pulmonary artery, inflated through trachea and removed. Three pieces of lung tissue from different lobes were excised and immersed in 10% formalin and sectioned. Hematoxylin–eosin staining was subsequently performed. Pulmonary artery morphology was observed and the percentages of muscularized and non-muscularized Table 1 Pressure–volume, echocardiographic and magnetic resonance parameters. Sham Pressure–volume parameters mPAP (mmHg) 21.09 sPAP (mmHg) 28.78 dPAP (mmHg) 11.45 RV EDP (mmHg) 5.05 RV dP/dtMAX (mmHg/s) 1500 1980 RV dP/dtMIN (mmHg/s) RV PRSW (mmHg) 34.05 RV Tau (ms) 16.12 LV EDP (mmHg) 6.05 8120 LV dP/dtMAX (mmHg/s) LV dP/dtMIN (mmHg/s) 5980 LV PRSW (mmHg) 96.65 LV Tau (ms) 13.82
± ± ± ± ± ± ± ± ± ± ± ± ±
1.52 9.34 2.60 0.60 160 465 9.33 4.30 1.40 930 370 10.33 4.60
Shunt 10W
Shunt 20W
24.81 72.23 16.45 13.88 1810 992 28.25 26.20 12.80 4030 3892 81.20 18.00
43.62 105.34 56.83 32.04 975 1045 19.01 34.45 17.03 3128 2890 38.04 42.10
± ± ± ± ± ± ± ± ± ± ± ± ±
4.30 10.02⁎ 4.10 1.91⁎ 498 498⁎ 6.15 3.82⁎ 4.21⁎ 312⁎ 288⁎ 6.10 8.81
± ± ± ± ± ± ± ± ± ± ± ± ±
2.85⁎,§ 23.40⁎,§ 9.71⁎,§ 11.22⁎,§ 155⁎ 98⁎ 3.55⁎ 8.72⁎ 2.30⁎ 101⁎,§ 97⁎,§ 8.95⁎ 9.73⁎
Echocardiographic parameters 0.23 ± 0.90 RV EDA (cm2) RV WT (cm) 0.20 ± 0.05 TAPSE (mm) 2.25 ± 0.05
0.35 ± 0.06⁎ 0.35 ± 0.05⁎ 1.97 ± 0.05
0.36 ± 0.04⁎ 0.33 ± 0.05⁎ 1.93 ± 0.25
Magnetic resonance parameters RV M/BW 0.3 RV SV (μl) 251 RV EF (%) 65 LV M/BW 1.3 LV SV (μl) 330 LV EF (%) 62
0.5 322 59 1.7 333 56
0.4 245 53 1.4 302 50
± ± ± ± ± ±
0.09 5 2 0.4 5 3
± ± ± ± ± ±
0.1 11⁎ 6 0.3 6 7
± ± ± ± ± ±
0.07 7⁎,§ 2⁎ 0.4 18 5⁎
mPAP = mean pulmonary artery pressure; sPAP = systolic pulmonary artery pressure; dPAP = diastolic pulmonary artery pressure; RV = right ventricle; EDP = end-diastolic pressure; PRSW = preload recruitable stroke work; LV = left ventricle; sPAP = systolic pulmonary artery pressure; EDA = end-diastolic area; WT = free wall thickness; TAPSE = tricuspid annular plane systolic excursion; M/BW = mass normalized on corresponding body weight; SV = stroke volume; EF = ejection fraction. Data are presented as mean ± SEM. ⁎ p b 0.05 vs. sham. § p b 0.05 vs. shunt 10W.
76
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
pulmonary arteries of the total amount of small pulmonary arteries were calculated. Statistical analysis Data are presented as means ± standard deviations. Differences between groups were determined using two-way analysis of variance (ANOVA) and Bonferroni's multiple comparison tests with statistical significance set at p b 0.05. Results All rats survived surgical procedure of shunt creation (Fig. 1). During the planned follow-up, 4 animals died among shunt 20W group. A postmortem examination ascertained postoperative heart failure in all the four cases, with relevant pleural and pericardial effusion. Qp/Qs was 2.12 ± 0.54 in shunt 10W, 1.98 ± 0.41 in shunt 20W and 1 ± 0.12 at 10 and 20 weeks in the control (sham). In all rats with shunt, Qp/Qs was significantly increased (p b 0.05) compared to control and gave evidence of shunt patency. Right ventricle Table 1 illustrates P–V parameters and echocardiographic and MRI data. RV EDV and ESV increased in shunt groups with a significant thickness in RV free wall. This was confirmed by an augmentation of RV mass, expressed as normalized to corresponding body weight.
Functional evaluation demonstrated a significant increase in RV enddiastolic pressure in shunt groups, in parallel with a progressive reduction in RV systolic functions parameters (dP/dtMAX and PRSW), which was significant at 20 weeks. dP/dtMIN augment and a corresponding diminishing in Tau were suggestive of impaired RV diastolic function (Fig. 3). Right ventricle–arterial coupling The maximal slope of RV ESPVR, i.e., RV Ees, was lower in shunt 20W compared to both shunt 10W and control (p b 0.01). Pulmonary artery Ea increased in shunt 20W (p b 0.01) and was consistent with an increase in mean pulmonary arterial pressure (p b 0.01). Therefore, the Ees/Ea ratio was affected, demonstrating an uncoupling of RV contractility and RV afterload (Fig. 3). Ventricular interdependence Both RV and LV masses, expressed as normalized to corresponding body weights, as well as cardiac chambers volumes increased in shunt groups (Table 1). LV EDV and ESV significantly increased in shunt 20W. EDPVR, PRSW and dP/dtMAX demonstrated an impaired systolic function in shunt 20W (Figs. 2 and 3). LV dP/dtMIN decreased significantly in both shunt groups and LV Tau was higher in shunt 20W (p b 0.05). These modifications suggested impaired diastolic ventricular function and a compromised ventricular relaxation.
Fig. 2. Magnetic resonance imaging and pressure–volume loops. Morphologic and functional evaluation by MRI (first line) and conductance pressure–volume catheter (second line) in representative cases of sham, shunt at 10 weeks and 20 weeks.
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
77
Fig. 3. Hemodynamic analysis. (A–B) Right ventricle afterload was expressed by pulmonary artery effective elastance (PA Ea) and mean pulmonary artery pressure (MPAP). (C–D) Right ventricle systolic and diastolic functions measured by systolic elastance (RV Ees) and time constant of ventricular relaxation Tau (RV Tau), respectively. (E) Right ventricle–arterial coupling defined as the ratio between ventricle systolic elastance and effective pulmonary artery elastance (Ees/Ea). (F) Ventricular interdependence was measured by left ventricle end-diastolic volume (LV EDV).
Histological analysis Cross-sectional slices of the ventricles demonstrated biventricular dilatation and augmentation of RV free wall in shunt 20W (Fig. 4). Small pulmonary artery morphology detected by hematoxylin– eosin staining revealed a non-significant medial hypertrophy in shunt 10W compared with sham. However, after 20 weeks of overcirculation exposure, a significant thickening of tunica media with muscular hypertrophy was evident. Masson's trichrome stain revealed low fibrosis rate. In some cases of shunt 20W, it was possible to demonstrate a complete occlusion of the intralobular arteries due to concentric intimal hyperplasia. No advanced plexiform lesions, sign of severe and irreversible PH were observed (Fig. 4).
Discussion In the present study, aorto-caval shunt rat model was demonstrated to closely reproduce the aberrant hemodynamic status and characteristic morphologic changes observed in the lungs of patients with PH.
Historically, the most widely used animal models of PH have been chronic hypoxia and the monocrotaline-induced PH rodent model. Although these models have added enormously to the understanding of the mechanisms of pulmonary vascular remodeling in PH, they are truly limited by different mechanisms of pathogenesis and disease phenotype compared to clinical setting (Dickinson et al., 2013). In these animal models, recent trials have shown that mono-therapy or combination therapy using targeted drugs exerted preventive, or even reversible, efficacy on hypoxia/monocrotaline-induced PH (Itoh et al., 2004). When administered to patients with severe PH, however, no drugs resulted in significantly lower mortality. This discrepancy has led researchers to reconsider the inadequacy and limitations of hypoxia/monocrotaline-induced PH (Bauer et al., 2007; Zaiman et al., 2005). With PH research shifting from a concept of mere vasoconstriction toward a mechanism of angioproliferation, the role of disturbed blood flow is seen as being pivotal (both clinically and experimentally) in PH development. Therefore, the need for a satisfactory and simple shunt-related model led to numerous attempts to simulate the pulmonary vasculopathy by surgically creating peripheral artero-venous or central systemic-to-pulmonary communications with different and
78
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
Fig. 4. Histology. (A–B) Progressive bi-ventricular dilatation with right ventricle hypertrophy in a case of shunt at 20 weeks (B) compared to sham operated rat (A). (C–D) Small pulmonary artery morphology detected by hematoxylin–eosin staining in sham (C) and shunt 20W (D). (E–F) Percentage of non-muscularized pulmonary artery (E) and muscularized pulmonary artery (F) in small pulmonary artery.
conflicting results (Bauer et al., 2007; Itoh et al., 2004; Tenmark and McMurtry, 2005; Zaiman et al., 2005). Many animal studies inducing increased pulmonary blood flow have used restrictive shunts resulting in the increased muscularization of the pulmonary arteries but with no rise in pulmonary arterial pressure (Black et al., 2003; Rondelet et al., 2012). This is in analogy to the clinical observation that patients with untreated pre-tricuspid shunts, such as atrial septal defect, in general develop advanced pulmonary vascular remodeling only in 5%–15% of the cases and only after two to three decades. In contrast, patients with untreated non-restrictive, posttricuspid shunts, such as ventricular septal defects, develop progressive PH in 1–2 years in 80%–100% of the cases and, eventually, Eisenmenger
syndrome, if left without surgical treatment (Van Albada and Berger, 2008). The aorto-caval shunt model is known to be a reproducible, simple and rapid method to develop high output heart failure and cardiac hypertrophy. Lam et al. (2005)) firstly observed the signs of pulmonary hypertension in rats with aorto-caval fistula. Although volume overload was previously demonstrated by Garcia and Diebold (1990), the observation of 4 weeks was too short to induce pulmonary vascular remodeling. The progressive development of PH during chronic exposure to aorto-caval shunt was demonstrated by the present study. For the first time, a detailed hemodynamic evaluation and a state-of-the-art imaging
A. Rungatscher et al. / Microvascular Research 94 (2014) 73–79
were applied to address ventricular–vascular interactions in a rat pulmonary artery overflow model. Echocardiogram and MRI showed a progressive increase in RV volume and end-diastolic area. The presence of a mesosystolic notch in Doppler profile of pulmonary blood flow in shunt at 20W was suggestive of increased systolic pulmonary pressure as demonstrated by Jones et al. (2002). With pressure–volume catheter, PH was confirmed by the direct measurement of increased pulmonary pressure and increased effective pulmonary Ea, which is the most accurate index of pulmonary vascular load (Morimont et al., 2008) and is related to the increase of pulmonary artery stiffness (Weiss et al., 1976). Further evidence of development of a progressive pulmonary vascular remodeling related to hypertension in rats was provided by histological analysis of lungs. A progressive thickening of muscular lamina after 10 and 20 weeks of shunt was demonstrated. In the group of 20-week overload exposure, we found a complete occlusion of the intralobular arteries due to concentric intimal thickening (Fig. 4). The coupling between pulmonary artery Ea (index of arterial load) and RV Ees (index of contractility) is defined by the Ees/Ea ratio, and it is an effective index of the mechanical performance of RV and the dynamic modulation of pulmonary circulation. In particular, Ea increased and Ees decreased, leading to uncoupling the right side of cardiovascular apparatus. This reflected an imbalance between myocardial oxygen consumption and mechanical energy request to perform hemodynamic work. Evaluation of the Ees/Ea ratio is nowadays recognized as an adjunctive perspective for understanding the pathophysiology of altered hemodynamic profiles, and for guiding therapeutic strategies and testing the effectiveness of treatments. Moreover, we addressed the interdependence between RV and LV. In particular, the overloaded RV had a direct impact on LV performance through serial ventricle interactions (such as failure to produce anterograde filling of LV) and parallel ventricular interactions such as the disturbance of leftward shifting of the septum. This ventricular interdependence was demonstrated by observation of leftward shift of interventricular septum (MRI) that leaded to impaired LV preload (EDV). In the present work, aorto-caval shunt generated Qp/Qs ≈ 2, which usually describes a moderate pre-tricuspid shunt, rarely generating PH in humans. This could limit the present model for translation to clinic. Nonetheless, clinical observations are gathering showing how longterm exposure to atrial level shunts may produce even severe PH (Bradley et al., 2013; Goetschmann et al., 2008; Luciani et al., 2008). In addition, atrial level shunt, with or without severe PH, and PH per se have since been shown to cause LV dysfunction by a variety of mechanisms, including ventricular remodeling, decreased myofiber preload and ventricular interdependence (Haeck et al., 2014; Hardegree et al., 2013; Walker et al., 2004). Finally, a possibility exists that severe PH observed in present experimental setting may be explained by species-specific differences, thereby representing a possible limitation to translation of results to human pathology. However, a period of 20 weeks relative to an average life span of 2 years represents a remarkably long exposure and could explain the development of pulmonary hypertension in this animal model. Conclusion Chronic exposure to aorto-caval shunt is a reliable model to produce RV volume overload and secondary PH. Timing of overcirculation exposure has a pivotal role in the development of PH. Therefore, this model, with an adequate time of follow-up, could be an alternative with low mortality and high reproducibility for investigators on the underling mechanisms of shunt-related PH. References Badesch, D.B., Champion, H.C., Sanchez, M.A., et al., 2009. Diagnosis and assessment of pulmonary arterial hypertension. J. Am. Coll. Cardiol. 54, S55–S66.
79
Bai, Y., Sun, L., Hu, S., et al., 2011. Combination therapy in pulmonary arterial hypertension: a meta-analysis. Cardiology 120, 157–165. Bauer, N.R., Moore, T.M., McMurtry, I.F., 2007. Rodent models of PAH: are we there yet? Am. J. Physiol. Lung Cell. Mol. Physiol. 293, L580–L582. Black, S.M., Mata-Greenwood, E., Dettman, R.W., et al., 2003. Emergence of smooth muscle cell endothelin B-mediated vasoconstriction in lambs with experimental congenital heart disease and increased pulmonary blood flow. Circulation 108, 1646–1654. Bradley, E.A., Chakinala, M., Billadello, J.J., 2013. Usefulness of medical therapy for pulmonary hypertension and delayed atrial septal defect closure. Am. J. Cardiol. 11, 1471–1476. Dickinson, M.G., Bartelds, B., Borgdorff, M.A.J., et al., 2013. The role of disturbed blood flow in the development of pulmonary arterial hypertension: lessons from preclinical animal models. Am. J. Physiol. Lung Cell. Mol. Physiol. 305, L1–L14. Fullerton, D.A., Mitchell, M.B., Jone, D.N., et al., 1996. Pulmonary vasomotor dysfunction is produced with chronically high pulmonary blood flow. J. Thorac. Cardiovasc. Surg. 111, 190–197. Garcia, R., Diebold, S., 1990. Simple, rapid and effective method of producing aortocaval shunts in the rat. Cardiovasc. Res. 24, 430–432. Goetschmann, S., Dibernardo, S., Steinmann, H., et al., 2008. Frequency of severe pulmonary hypertension complicating “isolated” atrial septal defect in infancy. Am. J. Cardiol. 102, 340–342. Haeck, M.L., Höke, U., Marsan, N.A., et al., 2014. Impact of right ventricular dyssynchrony on left ventricular performance in patients with pulmonary hypertension. Int. J. Cardiovasc. Imaging 30, 713–720. Hardegree, E.L., Sachdev, A., Fenstad, E.R., et al., 2013. Impaired left ventricular mechanics in pulmonary arterial hypertension: identification of a cohort at high risk. Circ. Heart Fail. 6, 748–755. Itoh, T., Nagaya, N., Fujii, T., et al., 2004. A combination of oral sildenafil and beraprost ameliorates pulmonary hypertension in rats. Am. J. Respir. Crit. Care Med. 169, 34–38. Jones, J., Mendes, L., Rudd, M.A., et al., 2002. Serial noninvasive assessment of progressive pulmonary hypertension in a rat model. Am. J. Physiol. Heart Circ. Physiol. 283, H364–H371. Lam, C.F., Peterson, T.E., Croatt, A.J., et al., 2005. Functional adaptation and remodeling of pulmonary artery in flow-induced pulmonary hypertension. Am. J. Physiol. Heart Circ. Physiol. 289, H2334–H2341. Luciani, G.B., Viscardi, F., Pilati, M., et al., 2008. Age at repair affects the very long-term outcome of sinus venosus defect. Ann. Thorac. Surg. 86, 153–159. Morimont, P., Lambermont, B., Ghuysen, A., et al., 2008. Effective arterial elastance as an index of pulmonary vascular load. Am. J. Physiol. Heart Circ. Physiol. 294, H2736–H2742. Nakano, K., Sugawara, M., Ishihara, K., et al., 1990. Myocardial stiffness derived from endsystolic wall stress and logarithm of reciprocal of wall thickness. Contractility index independent of ventricular size. Circulation 82, 1352–1361. Nishimura, T., Faul, J.L., Berry, G.L., et al., 2003. Effect of a surgical aortocaval fistula on monocrotaline-induced pulmonary hypertension. Crit. Care Med. 31, 1213–1218. Ocampo, C., Ingram, P., Ilbawi, M., et al., 2003. Revisiting the surgical creation of volume load by aorto-caval shunt in rats. Mol. Cell. Biochem. 251, 139–143. Pacher, P., Mabley, J.G., Liaudet, L., et al., 2004. Left ventricular pressure–volume relationship in a rat model of advanced aging-associated heart failure. Am. J. Physiol. Heart Circ. Physiol. 287, H2132–H2137. Reddy, V.M., Meyrick, B., Wong, J., et al., 1995. In utero placement of aorta pulmonary shunts: a model of postnatal pulmonary hypertension with increased pulmonary blood flow in lambs. Circulation 92, 606–613. Rondelet, B., Dewachter, C., Kerbaul, F., et al., 2012. Prolonged overcirculation-induced pulmonary arterial hypertension as a cause of right ventricular failure. Eur. Heart J. 33, 1017–1026. Szabo, G., Soos, P., Bährle, S., et al., 2006. Adaptation of the right ventricle to an increased afterload in the chronically volume overloaded heart. Ann. Thorac. Surg. 82, 989–995. Tenmark, K.R., McMurtry, I.F., 2005. Vascular remodeling versus vasoconstriction in chronic hypoxic pulmonary hypertension: a time for reappraisal? Circ. Res. 97, 95–98. Van Albada, M.E., Berger, R.M., 2008. Pulmonary arterial hypertension in congenital cardiac disease—the need for refinement of the Evian–Venice classification. Cardiol. Young 18, 10–17. Van Albada, M.E., Schoemmaker, R.G., Kemna, M.S., et al., 2005. The role of increased pulmonary blood flow in pulmonary arterial hypertension. Eur. Respir. J. 26, 487–493. Walker, R.E., Moran, A.M., Gauvreau, K., et al., 2004. Evidence of adverse ventricular interdependence in patients with atrial septal defects. Am. J. Cardiol. 93, 1374–1377. Wang, W., Liu, R., Cao, G., et al., 2010. A reliable rabbit model for hyperkinetic pulmonary hypertension. J. Thorac. Cardiovasc. Surg. 140, 395–399. Wang, X., Ren, B., Liu, S., et al., 2003. Characterization of crdiac hypertrophy and heart failure due to volume overload in the rat. J. Appl. Physiol. 294, 752–763. Weiss, J.L., Frederiksen, J.W., Weisfeldt, M.L., 1976. Hemodynamic determinants of the time-course of fall in canine left ventricular pressure. J. Clin. Invest. 58, 751–760. Xiong, M., Yao, J.P., Wu, Z.K., et al., 2012. Fibrosis of pulmonary vascular remodeling in carotid artery-jugular vein shunt pulmonary artery hypertension model of rats. Eur. J. Cardiothorac. Surg. 41, 162–166. Yerebakan, C., Klopsch, C., Niefeldt, S., et al., 2010. Acute and chronic response of the right ventricle to surgically induced pressure and volume overload, an analysis of pressure–volume relations. Interact. Cardiovasc. Thorac. Surg. 10, 519–525. Zaiman, A., Fijalkowska, I., Hassoun, P.M., et al., 2005. One hundred years of research in the pathogenesis of pulmonary hypertension. Am. J. Respir. Cell Mol. Biol. 33, 425–431.
