Surgical treatment of severe pulmonic stenosis under cardiopulmonary bypass in small dogs



Objectives: The aim of this study was to report the long-term outcome of the surgical palliation of pulmonic stenosis in dogs.

Methods: The subjects comprised three female and six male dogs, mean (±sd) age: 23 (±25) months, mean (±sd) weight: 3·4 (±2·1) kg, diagnosed with severe pulmonic stenosis and right ventricular hypertrophy, with an average preoperative pressure gradient of 153 (±43) mmHg on echocardiography.

Results: The pressure overload with severe pulmonic stenosis was reduced by valvotomy, i.e., open pulmonary valve commissurotomy, with/without biomembrane patch grafting, under cardiopulmonary bypass. The postoperative pressure gradient at 1 to 7 days was significantly decreased to 65 (±39) mmHg (P<0·05). The reduced pressure gradient was maintained at 58 (±38) mmHg at final follow-up.

Clinical Significance: Open valvotomy, pulmonary valve commissurotomy and biomembrane patch grafting were effective in reducing obstruction in severe pulmonic stenosis in dogs.


Congenital pulmonic stenosis is caused by a congenital stricture of the right ventricular outflow tract, narrowing of the pulmonary ostium or fusion of the leaflets of the pulmonary valve, and infundibular hypertrophy (Tidholm 1997); this is a common congenital heart disease in dogs (Fingland and others 1986). The stenosis may be supravalvular, valvular or subvalvular, with pulmonary valvular stenosis being the most common (Fingland and others 1986). One type of pulmonary valvular stenosis is caused when commissural fusion of thin to moderately thickened valve leaflets occurs, and creates a dome-shaped valve. However, valvular stenosis caused by pulmonary valve dysplasia is different from such a fused valve, which has a central orifice. Valvular dysplasia involves markedly thickened valve leaflets, with or without annular hypoplasia and fusion of the commissures (Estrada 2009, Oyama and others 2009).

Dogs with mild pulmonic stenosis survive for years without related clinical signs of any consequence (Ristic and others 2001). However, there is general agreement that dogs with severe resting pressure gradients (>80 mmHg) between the right ventricle and the pulmonary artery are at increased risk of syncope, cardiac arrhythmia or sudden death (Fingland and others 1986). Therefore, surgical intervention is usually recommended for these dogs.

We have previously reported several successful cardiac surgeries under cardiopulmonary bypass (CPB) in small-breed dogs and cats (Uechi and others 2011, in press, Yamano and others 2011). Surgical techniques for reducing severe pulmonic stenosis include open and closed patch grafting pulmonary commissurotomy and valvotomy, using pulmonary bypass techniques (Orton and others 1990, Hunt and others 1993, Staudte and others 2004). Recently, pulmonary balloon valvuloplasty has become the primary treatment for valvular pulmonic stenosis in dogs as well as in humans (Bright and others 1987, Bussadori and others 2001). However, balloon valvuloplasty is not constructive in supravalvular or subvalvular pulmonic stenosis in dogs (Ristic and others 2001). Further, dysplastic valves are more difficult to dilate effectively (Ware 2008). In addition, in small dogs (weight <3 kg), it is difficult to position the balloon accurately in the narrow segment during valvuloplasty (Tanaka and others 2009); this is also true in dogs with coronary artery abnormalities (Bussadori and others 2001, Johnson and others 2004).

In humans, surgical relief of pulmonic stenosis achieves long-term reduced pressure gradients and freedom from cardiac events (Peterson and others 2003). In dogs, positive short-term results have been demonstrated with surgical repair for pulmonic stenosis (Tanaka and others 2009); however, no long-term follow-up studies on the effectiveness of these repair techniques are available. The aim of this study was to report the long-term outcome of the surgical palliation of pulmonic stenosis in dogs.

Materials and methods


We retrospectively analysed nine dogs that had been treated consecutively at the Nihon University Animal Medical Center between 2006 and 2009. Data were retrospectively collected by reviewing the medical records, including clinical signs and postoperative course.

A diagnosis of pulmonic stenosis had been established in all nine dogs by physical examination, ECG, radiography, and two-dimensional and Doppler echocardiography. Doppler measurement of the systolic pressure gradient across the pulmonary valve was assessed by the modified Bernoulli equation. According to the pressure gradient, dogs were classified into the following categories: mild stenosis, 10 to 49 mmHg; moderate stenosis, 50 to 80 mmHg; and severe stenosis, more than 80 mmHg (Bussadori and others 2001). The decision to apply surgical palliation for pulmonic stenosis was made based on the severity of the pressure gradient, the presence of moderate to severe right ventricular hypertrophy and/or clinical signs.

The type of valvular stenosis (type A or B) was classified based on the results of echocardiography, as well as direct visual observation of the valve leaflets during surgery. Type A was diagnosed when mild to moderately thickened leaflets was observed, with evidence of commissural fusion and systolic doming. Type B was diagnosed when severely thickened leaflets were observed (Locatelli and others 2011).

Doppler echocardiography studies were repeated at various times for each case at 1 to 7 days, 1 to 3 months and at 6 months to 3 years postoperatively. A single echocardiographer (MU) consistently performed all the measurements, thus precluding interobserver variability.

Anaesthesia and surgery

The dogs were anaesthetised using 0·025 mg/mL, im atropine sulphate premedication (atropine sulphate injection®; Mitsubishi Tanabe Pharma Corporation), 0·1 mg/kg, iv, fentanyl (Fentanyl®; Daiichi Sankyo Company), 0·2 mg/kg, iv, midazolam (Dormicum®; Astellas Pharma Inc.) and 20 mg/kg, iv, cefazolin sodium (Cefamezin®; Astellas Pharma Inc.). Anaesthesia was induced in each dog with 5 μg/kg, iv, fentanyl and 4 mg/kg, iv, propofol (Rapinovet®; Schering-Plough Corporation) and maintained by inhalation of 1·5 to 2·5% isoflurane and 100% oxygen (2·0 l/min). Respiration was maintained at 12 breaths/min with 100% O2 using a ventilator (7900 SmartVent™; GE Yokogawa Medical Systems Ltd).

During CPB, anaesthesia was maintained by 0·4 μg/kg/min, iv, fentanyl and 0·2 mg/kg/min, iv, propofol. During the operation, the left femoral artery and vein were catheterised for time-lapse measurements of arterial (systolic, diastolic and mean) and central venous pressures. Heart rate, arterial oxygen saturation, end-tidal CO2, isoflurane concentration, rectal temperature and oesophageal temperature were continuously monitored. The volume of urine was also monitored by a catheter in the bladder.

Cardiopulmonary bypass was provided by a heart-lung machine (NAPS-III®; Senko Medical Instrument), with a small extracorporeal circuit (Terumo), oxygenator and heat exchanger (Baby-RX®; Terumo). The CPB circuit was primed with 5 mL/kg d-mannitol (20% Mannitol®; Kowa Company), 2 mL/kg sodium bicarbonate (7% Meylon®; Otsuka Pharmaceutical Factory, Inc.), 500 U/head heparin sodium (Novo-Heparin®; Mochida Pharmaceutical Co.) and Ringer’s acetate solution (Veen-F®; Kowa Company). Fifty millilitres of whole blood was replaced with priming solution in dogs weighing less than 4 kg.

Thoracotomy was performed with a median sternotomy in case 1. In the other eight cases, thoracotomy was performed through the fourth intercostal space after an intercostal nerve block with bupivacaine hydrochloride (Marcain®; Astrazeneca).

A measure of 400 U/kg heparin sodium was administered, and the activated clotting time (ACT) was measured to confirm that it was ≥300 seconds prior to CPB cannulation. The CPB cannula was inserted into the carotid artery, and the jugular vein was cannulated for venous return in all cases. In cases 1 to 3, the vena cava was also cannulated through the right atrium. Then, partial CPB was initiated. The blood flow was set at 100 mL/kg/min by the extracorporeal circulation pump. Oesophageal and rectal temperatures were monitored. Each dog was cooled to a rectal temperature of 25 to 28°C.

In cases 1 to 3, after the aorta was occluded using an arterial clamp, 10 mg/kg cardioplegic solution (Miotecter®; Mochida Pharmaceutical Co.; cooled to ≤4°C, Na+: 120 mEq/L, K+: 20 mEq/L, Cl: 160·4 mEq/L, Mg2+: 32·0 mEq/L, Ca2+: 2·4 mEq/L, HCO: 10 mEq/L) was immediately and rapidly infused into the coronary artery via a catheter extended into the root of the aorta. A measure of 10 mL/kg cardioplegic solution was administered every 20 minutes.

In cases 4 to 9, intracardiac procedures were performed on the beating heart. The reservoir blood level was constantly maintained at a stable level during the pulmonary artery procedure to prevent “air block”.

In case 1, a polytetrafluoroethylene (PTFE) patch was placed along the atrial septal defect (ASD) and secured by a simple continuous suture with 6-0 Prolene sutures (Proline® 8725H; ETHICON). Immediately after the complete closure of the ASD, the right atrium was sutured. Since echocardiography confirmed that ASD was mild in case 2, it was not treated. Once the main pulmonary artery incision was made, the hypoplastic valve was identified and the thickened or fused leaflet was resected with scissors. The reduction in the stenosis was confirmed by passing an inflated tracheal tube, which had a diameter equivalent to that of the pulmonary artery, from the main pulmonary artery through the pulmonary valve.

In cases 2 and 3, a patch graft was affixed over the right ventricular outflow tract, using continuous 6-0 Prolene sutures, to dilate the lumen; for this purpose, we used biomaterial allografts, i.e., glutaraldehyde-fixed canine tunica vaginalis obtained at the time of castration. In the seven other cases, the pulmonary artery incision was sutured with continuous 6-0 Prolene sutures. When haemostasis was required during suturing on the pulmonary artery, bovine dermis-derived atelocollagen (Integran®; Nippon Zoki Pharmaceutical Co.), which is an absorbable local haemostat, was used.

In case 1, perioperative defibrillation was required (Defibrillator FC-1760, Fukuda Denshi), and resulted in the sinus rhythm returning immediately.

After the CPB pump was stopped, 6 mg/kg protamine sulphate (Novo-Protamine Sulfate®; Mochida Pharmaceutical) was administered by slow iv injection. After the ACT was reduced to less than 200 seconds, a thoracic tube was put in place prior to chest closure, and the cervical wound was closed routinely.

Postoperative management was conducted in the intensive care unit. Air and fluid in the thoracic cavity were removed through the thoracic tube. Antibiotics, 20 mg/kg, iv (Cefamezina®; Astellas Pharma Inc. ) every 8 hours were administered for 1 week. The dogs were discharged from the hospital 3 to 11 days postoperatively.

Statistical analyses

Friedman’s test was used to assess differences in the pressure gradient preoperatively and at 1 to 7 days, 1 to 3 months, and 6 months to 3 years postoperatively. A P-value of less than 0·05 was considered statistically significant.


The baseline characteristics of the dogs in this study are summarized in Table 1. The mean (±sd) age of the nine dogs was 23 (±25) months (range: 3–83 months); three were females and six males (one neutered). The mean (±sd) body weight was 3·4 (±2·1) (range: 1·5–8·7) kg. Of these nine, five dogs weighed less than 3 kg (Table 1). ASD was diagnosed in cases 1 and 2 (Table 1).

Table 1. Signalment and clinical data in pulmonary stenosis
CaseAge (Mon)BreedSexBW (kg)Present.C. murmurComp.
  1. Mon months, BW body weight, Present presentation, Comp complications, M male, F female, Asympt asymptomatic, ASD atrial septal defect, PR pulmonary regurgitation, TR tricuspid regurgitation

138M. dachshundF4·2Asympt.6ASD, PR
23Shih-tzuM3·2Asympt.6ASD, PR
483M. PinscherM2·9Collapse5TR
57MixM2·9Asympt.5PR, TR
713Yorkshire terrierF1·9Exercise intolerance5PR
918French BulldogM8·7Asympt.3PR
Mean ±sd23 ±25  3·4 ±2·1   

Surgical summary

Seven dogs underwent valvotomy or open pulmonary valve commissurotomy, while cases 2 and 3 underwent open valvotomy and biomembrane patch grafting. Case 3 also underwent relief for supravalvular stenosis. The mean (±sd) cross-clamp time was 26 (±3) minutes, and the mean (±sd) CPB time was 77 (±8) minutes (Table 2). Eight dogs showed rapid recovery from anaesthesia [mean (±sd) recovery time: 159 (±75); range: 80–275 minutes] and were able to stand the next day. One dog died during the operation.

Table 2. Outline of the operation
CaseCPBTechnique (The number of valve leaflet cut)DP (min)DCA (min)LBT (°C)
  1. CPB cardiopulmonary bypass, DP duration of perfusion, DCA duration of cardiac arrest, LBT lowest body temperature

2PartialValvotomy (2) Patch graft1365422·8
3PartialValvotomy (2) Patch graft1427125·7
5PartialValvotomy (2)68025·0
6PartialValvotomy (1)587028·7
8PartialPulmonary commissurotomy69023·9
9PartialValvotomy (1)152028·6
Mean ±sd  162 ±16565 ±1024·8 ±3·3

Clinical signs

Before the operation, six dogs had demonstrated no adverse clinical signs, while three dogs showed syncopal episodes or exercise intolerance. Clinical signs in these three dogs resolved after surgery.

Case 6 could not be examined postoperatively because the owner lived at a great distance from the hospital. However, we confirmed the dog’s condition by interviewing the local veterinarian who examined the dog.

All nine dogs had a grade III–VI/VI systolic murmur with its point of maximum intensity in the left heart base (Table 1).


All cases showed stenosis at the level of the pulmonary valve, while one also had supravalvular stenosis. Type B pulmonary valvular stenosis was noted in four of the nine dogs (44·4%). The mean (±sd) systolic Doppler gradient before surgery was 153 (±43) (range: 108–242 mmHg; Table 3). Poststenotic dilatation of the main pulmonary artery was observed in all dogs. Cases 1, 2, 5 and 7 to 9 had pulmonary regurgitation, while cases 4 to 6 had tricuspid regurgitation. Septal flattening was noted in cases 1, 2, 5 and 7. On postoperative echocardiographic examination, there was no remaining evidence of ASD in case 1.

Table 3. Pre- and postoperative pulmonary pressure gradients and clinical signs in severe pulmonic stenosis
 Type of stenosisType of vulvular stenosisPeak pressure gradient (mmHg)Clinical sign
Pre1–7 days (RR %)1–3 mon. (RR %)Final (time) (RR %)PrePost
  1. mon Month, RR Reduction rate, V Valvular pulmonic stenosis, SUP Supravalvular pulmonic stenosis, y Year, N None, C Collapse, EI Exercise intolerance, NE Not examined

  2. *P<0·05. There was a significant difference compared to preoperative values

1VType B17077 (55)84 (51)87 (1 y) (49)NN
2VType B12024 (80)39 (68)115 (3 y) (4)NEI
3SUP/VType B132120 (9)43 (64)42 (2 y) (65)NN
4VType A11310 (91)36 (69)29 (2 y) (74)CN
5VType B24295 (61)84 (65)77 (6 mon) (68)NN
6VType B140CN
7VType A17651 (71)NENEEIN
8VType A17980 (55)102 (43)51 (1·5 y) (55)NN
9VType B108NENE4 (1 y) (96)NN
Mean ±sd  153 ±4365 ±39* (60 ±26)65 ±29 (60 ±11)58 ±38 (59 ±28)  

The mean peak systolic Doppler gradient was significantly reduced after surgery (Table 3; P<0·05). In case 2, narrowing of the right ventricular outflow tract due to hypertrophic cardiomyopathy was observed, and a rise in pressure gradient was detected. Echocardiography after surgery for case 6, in which the owner resided far from the hospital and was consequently unable to participate in the follow-up visits, was limited to the duration of the hospital stay.


Preoperative thoracic radiographs showed right-sided heart enlargement in case 2. The postoperative cardiothoracic ratio and vertebral heart size (VHS) were not changed in all cases.


This study reported the long-term outcome of the surgical palliation of pulmonic stenosis in dogs. The pressure overload with severe pulmonary valvular stenosis and supravalvular pulmonic stenosis in our cases was reduced by valvotomy and open pulmonary valve commissurotomy with/without biomembrane patch grafting in the nine dogs; however, one dog died perioperatively.

All our cases showed type B pulmonary valvular stenosis, or were dogs weighing less than 3 kg. In such cases, it is considered difficult to completely palliate the pulmonary valve defect using balloon valvuloplasty and to insert a catheter into the stenotic region. We considered that surgery using CPB would be beneficial and, therefore, proceeded directly to CPB in these dogs.

Immediate and late success in pulmonary balloon valvuloplasty is defined as greater than 50% reduction in the pressure gradient from baseline, or a final pressure gradient below 80 mmHg (Bussadori and others 2001, Johnson and Martin 2004). In this study, there was an immediate average reduction of 60% in the pressure gradient following surgery. The reduced pressure gradient was maintained at least 6 months to 3 years after the operation. When cases of type B pulmonary valvular stenosis were considered separately, they showed an average reduction of 57·1% in the pressure gradient at 6 months to 3 years after operation. Balloon valvuloplasty is known to reduce 30 to 45% of the pressure gradient, and maintain this reduction for 6 months to 9 years (Ristic and others 2001, Johnson and others 2004). Depending on the type of valvular pulmonic stenosis, the median Doppler gradient is known to reduce from 115 to 45 mmHg (in type A pulmonary valvular stenosis) and from 132 to 72 mmHg (in type B pulmonary valvular stenosis) immediately (24 hours) after pulmonary balloon valvuloplasty. At 1 year after pulmonary balloon valvuloplasty, the median Doppler gradient in type A and B has been reported to be 50 mmHg (56·5% average reduction) and 86 mmHg (34·8% average reduction), respectively (Locatelli and others 2011).

Our present findings revealed that surgical palliation of pulmonic stenosis under CPB was effective in type B pulmonary valvular stenosis and supravalvular stenosis, which are not suitable for treatment by balloon valvuloplasty. This may be because the pulmonary valve was modified and palliated directly under visual observation. Incomplete reduction of obstruction with poor results has been reported to be due to residual gradients at the infundibular level (Fontes and others 1988, Ray and others 1993, Gupta and others 2001). In our study, cases 5 and 8 continued to show a moderate residual stenosis, whereas cases 1 and 2 showed severe stenosis. We consider that these dogs might require further removal of the pulmonary valve leaflets and/or annulus circumference incision; alternatively, patch grafting on the right ventricular outflow or a valved conduit should be considered. In case 2, we observed a reduced pressure gradient over 1 year; however, subsequently, left ventricular hypertrophy was observed by echocardiography. The reason for this pathological change was not clear, i.e., whether it was related to the surgery or represented the natural course of the disease. In case 3, the pulmonary pressure gradient was only slightly reduced from 132 to 120 mmHg at 1 to 7 days after surgery, but this decreased to 42 mmHg at 2 years postoperatively. This temporary stenosis after surgery may be explained by surgical inflammation that influenced the pulmonary valve annulus area (Estrada and others 2006); further, it is possible that muscular hypertrophy affects the results at 6 months to 3 years postoperatively (Miller and Sisson 1998).

Although three dogs were symptomatic before the surgery, the clinical signs in all three dogs disappeared following surgery. The remaining five dogs remained asymptomatic even after surgery. We, therefore, consider that surgical reduction of the pressure gradient was effective for treating pulmonic stenosis; however, one dog succumbed due to haemorrhage from the suture line after patch grafting. The patch-grafting suture should, therefore, be performed with particular care, because poststenotic dilatation leads to thinned out arterial walls, which may lead to enlargement of the suture holes.

We compared the effects of surgical operation with those of balloon dilatation, and observed that the rate of restenosis was higher following balloon dilatation (Peterson and others 2003). Further, in the study by Ristic and others (2001), 2 of the 11 dogs showed evidence of restenosis after balloon valvuloplasty on follow-up. In this study, the pressure gradient in case 4 increased slightly after the operation. The increase in the pressure gradient in case 2 was distinguishable from restenosis, as it may have been influenced by hypertrophic cardiomyopathy. No case showed any remarkable restenosis. The rate of restenosis was thus lower following surgical repair as compared to that after balloon dilatation; this is similar to the findings in humans (Peterson and others 2003).

Greater relief of the right ventricular outflow gradient is associated with increased right ventricular dimensions, probably because greater pulmonary regurgitation is induced (Masura and others 1993, Gupta and others 2001). These facts suggest that it may be important to correct both stenosis and regurgitation after pulmonary valve surgery. Further studies are needed to clarify whether residual stenosis, restenosis and pulmonary regurgitation would affect long-term prognosis. New surgical techniques might aid in resolving these problems in dogs. Our study was limited to a certain extent by the small sample size and the difficulty in pursuing surgical follow-up in some cases.

In conclusion, open valvotomy, pulmonary valve commissurotomy and biomembrane patch grafting were effective procedures for reducing obstruction in severe pulmonic stenosis. These surgical procedures reduced the clinical signs and improved long-term survival, indicating that they were beneficial in treating pulmonic stenosis. Further studies are needed to clarify whether residual stenosis, restenosis and pulmonary regurgitation would affect long-term prognosis.

Conflict of interest

None of the authors of this article has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.