Material and methods

Between June, 1994 and December, 2002, 150 consecutive patients with echocardiographic diagnosis of isolated patency of the arterial duct were referred to our Institution for video assisted thoracoscopic closure. Excluded from our series were patients with associated cardiac lesions requiring simultaneous surgical treatment, and premature infants.

The median age at operation was 45 months, with a range from 3 to 161 months, and mean weight was 18 kg, with a range from 4.2 to 73 kg. In all cases, diagnosis was made with cross-sectional and colour-Doppler echocardiographic imaging. The size of the duct was measured at its aortic origin, with values ranging from 1.5 to 9 mm.

Intravenous fentanyl was administered at a dose of 2 μg/kg to induce anaesthesia, and to facilitate endotracheal intubation. Anaesthesia was then maintained using Propofol at from 5 to 15 μg/kg. The patients were continuously monitored during the procedure by electrocardiography, pulse oxymetry, capnography and non invasive measurement of arterial blood pressure.

With the patients placed in right lateral decubitus position, three incisions are made around the left scapula. The first is an incision of 5-mm at the middle axillary line in the fourth intercostal space to permit retraction of the lung. The second, of 10 mm, is made beneath the angle of the scapula in the fifth intercostal space for insertion of the electro-cautery hook and the device used to apply the clip. The third incision, again of 5 mm, is made on the posterior axillary line in the fifth intercostal space for introduction of fiberoptic videoscope (Figs 1 and 2).

Figure 1. The patient is shown in right sided decubitus position. S: scapula; 1-2-3: the marks for the skin incisions.

Figure 2. Surgical instruments for thoracoscopy. CA: clip applier; C: 10 mm titanium clips; D: disposable dissector; E: electrocautery hook; G: grasper for parietal pleura; P: device used for lung retraction (“endopeanuts”); T: 5 to 10 mm trocars for introduction of instruments into the chest; V: 5 mm fiberoptic videoscope.

After gentle retraction of the lung, the mediastinal pleura is dissected with the electro-cautery hook above the aorta, starting about 2 cm below the aortic isthmus, leading toward the origin of the subclavian artery. Should a left hemiazygos vein be present, it is divided. The patent arterial duct is then isolated from the surrounding tissue with sharp and blunt dissection. Care is made to identify and to avoid the recurrent laryngeal nerve, and a single titanium clip 10 mm long is squeezed tight on the aortic junction of the duct. After checking the correct position of the clip, and the integrity of the left recurrent laryngeal nerve, complete closure of the duct is then confirmed in the operating room using cross-sectional and colour-Doppler echocardiographic imaging. In case of residual shunting, a second clip is applied. The skin is then closed, and a chest tube inserted, still with the patient lying in right lateral decubitus position. Extubation is carried out in the operating room and, a few hours later, after a control chest X-ray, the chest tube is removed. Antibiotic prophylaxis consisted of two doses of Cefazoline at a dose of 50 mm/kg at induction, and then 25 mm/kg. Relief for pain was achieved in all with non steroidal anti inflammatory medications given every 6 h.

The following morning, after a routine clinical examination, and sampling of blood, the patient is routinely discharged home. All patients were followed-up with a clinical examination cross-sectional echocardiographic study 1 year after the procedure.

Results

Thoracoscopic closure proved possible in 139 patients (93%). The mean time for the operation was 37 min, with a range from 14 to 89 min. In the remaining 11 patients (7%), we converted to a mini-thoracotomy. This was earlier in our experience, and was undertaken in 6 because of unfavourable anatomy, with posterior displacement of the duct, in 3 because of bleeding at the time of application of the clip, in 1 because of apparent lymphatic leakage during dissection of the duct, and in the final patient because of residual shunting detected at the intra-operative echocardiographic control.

In 4 patients, we applied a second clip because of residual shunting detected at the intraoperative cross-sectional echocardiographic examination.

There were no major haemorrhages requiring blood transfusion, and no hospital deaths. The mean period of hospital stay was less than 24 h in 61 patients, less than 48 h in 88 patients, and 21 days in 1 patient because of a persistent chylothorax detected after the surgical correction. Other complications included palsy of the left recurrent laryngeal nerve in 1 patient, with partial regression at 6 months follow-up.

There was no recurrence of patency of the arterial duct in any patient at the follow-up control 1 year after the surgical correction.

Discussion

Video assisted thoracoscopic closure of the patent arterial duct is an effective answer to the modern trend to seek repair of congenital cardiac malformations using minimally invasive techniques. Using this technique, results can be achieved that are similar to those obtained with conventional surgery,⁷ but avoiding the late complications linked to a left thoracotomy such as scoliosis, winged scapula, deformities of the chest wall, mammary disfigurement, and rib fusion with respiratory compromise.^8–10

A limiting factor could be the size of the titanium clip currently available for closure, at 10 mm. In our experience, however, we have yet to encounter a patent arterial duct too large to be completely occluded by such a clip. In fact, while large ducts are usually closed in small infants, bigger children usually present with ducts of smaller size.

Elective conversion to thoracotomy, needed a few times in our early experience, did not introduce any additional morbidity when compared to an initial thoracotomy. Doubts have been expressed about the technique because of supposed difficulties in controlling sudden bleeding should there be disruption of the patent duct.¹¹ In our experience of this complication in 3 patients, the availability of an open thoracotomy tray allowed us immediate vascular control, again with no incremental morbidity. We did not find any residual shunting at the follow-up echocardiographic studies, our experience being in keeping with the results reported by Laborde et al.⁵

Early in our experience we did encounter some problems, such as one case of palsy of the recurrent left laryngeal nerve, and one instance of chylothorax. No further complications of this kind have occurred in our last 78 cases.

Percutaneous catheter closure of the patent duct,⁶ of course, is now advocated as an alternative therapy.^{12, 13} With this approach, nonetheless, residual shunting across the duct is reported in up to one-quarter of patients at 6 weeks follow-up, persisting in up to one-fifth of cases at 6 months follow-up. The magnitude of the residual shunt is directly proportional to the size of the arterial duct.¹³ Migration or embolization of the device also occurs, and can necessitate surgical removal.¹³ There can also be haemolysis from the peri-prosthetic residual shunting, and prophylaxis against bacterial endocarditis is obviously necessary. In addition, transcatheter closure is only feasible in patients weighing more than 10 kg. Other complications have been reported, including injuries to the femoral artery, creation of artero-venous fistulas, the need of transfusion, and mild stenosis of the left pulmonary artery from protrusion of multiple coils.^{12, 13}

Along with others,^{2, 5} therefore, we are convinced that video assisted thoracoscopic closure of the arterial duct is the procedure of choice not only in children with a body weight of less than 10 kg, but also at an older age. The technique combines the advantages of safe and complete extraluminal occlusion with an excellent clinical and aesthetic result, minimal postoperative pain, and a short stay in hospital at low cost.