Minimally invasive mitral valve surgery through right thoracotomy in patients with patent coronary artery bypass grafts Free

Abstract

We report our institutional experience, with 25 consecutive patients with patent coronary artery bypass grafts (71.8±12.7 years), who underwent video-assisted minithoracotomic approach for mitral valve surgery. The surgical technique includes: right minithoracotomy, femoral cannulation and hypothermic ventricular fibrillation. Mean preoperative EuroSCORE was 10.2±2.4 and mean ejection fraction was 45±9%. Operative mortality was 4% (1/25). No patient required a conversion to sternotomy. Procedures performed were: mitral valve repair in 15 patients (60%), replacement in 10 (40%) and associated tricuspid repair in seven (28%). Mean blood transfusion was 1.2 package/patient. No cardiological, neurological, vascular and wound complications were observed. Postoperative major morbidity includes: severe pulmonary dysfunction in two patients (8%) and acute renal failure in one (4%). Mean ICU and hospital stay were 3.4±2.9 and 10.6±7.9 days. Echocardiographic follow-up (22.8±14.9 months) revealed trace or mild mitral valve regurgitation in all the mitral repair patients. When interrogated, all the surviving patients preferred the minithoracotomic approach rather than the sternotomy. In conclusion, minimally invasive right thoracotomy can be safely performed in patients with functioning coronary bypass grafts requiring mitral valve operation. Low blood transfusion, the avoidance of deep wound infection and the high patient satisfaction are the main advantages of this approach.

1. Introduction

Over the past decade, minimally invasive cardiac surgery has evolved dramatically and has become a standard approach for mitral valve operations in most centers all over the world. Based on our extensive experience with minimally invasive mitral valve surgery [1], we decided to extend this approach to patients with patent coronary artery grafts who required reoperative mitral operation. In the current study we sought to review our institutional experience with this approach and assess its feasibility, safety and effectiveness.

2. Materials and methods

Between November 2003 and March 2008, we have collected prospective data in 796 patients who had undergone mitral valve surgery via right minithoracotomy approach. From this cohort of patients, we identified 25 consecutive patients who had functioning bypass grafts at the time of surgery and so represents the population for this study. No patients were excluded. The preoperative characteristics of the population and their previous cardiac operation are detailed in Table 1 . The mean number of patent bypass grafts was 2.2±0.7. Table 2 details the preoperative coronary grafts status. Mean time to redo surgery was 8.6±8.8 years. In 19 patients the mitral valve operation was the first reoperation and in six patients it was the second. Mean preoperative EuroSCORE was 10.2±2.4 and predicted mortality was 21.6±14%. All patients had severe mitral regurgitation at the time of surgery. Elective surgery was performed in 19 patients and urgent operation in six. In the first 16 patients we have used a standard endotracheal intubation and in the remaining we have shifted to selective single-lung ventilation. All preoperative, operative and in-hospital data were collected prospectively. Data are summarized as mean±S.D.

2.1. Operative procedures

The procedure is performed through a small (5–8 cm) incision in the third right intercostal space. The main working port is created using a dedicated soft tissue retractor and a specially designed rib spreader (Window-Access; Estech, San Ramon, California). Two thoracic ports are inserted in the third and fifth intercostal spaces on the anterior axillary line to allow placement of a videoscope, CO₂ insufflation tube and cardiotomy vent. Cardiopulmonary bypass is instituted by femoral vessels cannulation, with vacuum-assisted venous drainage directly applied on the venous line.

Myocardial protection is achieved by systemic moderate hypothermia (25–28 °C) and ventricular fibrillation. To prevent air embolism, the procedure is performed with the aorta vented and with continuous CO₂ insufflations into the thoracic cavity. In patients requiring an associated tricuspid valve repair, a single two-stage femoral venous cannula (RAP™, Estech, San Ramon, USA) is used (Fig. 1 ).

In patients who have mild to moderate aortic insufficiency we use the Port-Access system with endoaortic balloon clamp and antegrade crystalloid cardioplegia (EndoClamp™, Edwards Lifesciences, Irvine, California).

Surgical ablation of the atrial fibrillation is performed with standard endocardial left atrial lesions using a monopolar radio-frequency probe (Cobra Adhere XL™, Estech, San Ramon, USA) in concomitant with left atrial appendage exclusion.

2.2. Follow-up

Echocardiographic follow-up was performed in all surviving patients from May 2008 to July 2008. At follow-up, patients were also interrogated with a structured questionary to compare their satisfaction with this approach over the previous sternotomy. The questionary consists of three predefined questions regarding the difference between the two approaches in: postoperative pain sensation, how fast the patient regained the expected activity and satisfaction with postoperative esthetic result.

3. Results

No patients required intraoperative conversion to sternotomy. Procedures performed are detailed in Table 3 . Operative mortality was 4% (1/25). One patient, a 78-year-old man, diabetic, with previous aortic valve replacement and CABG, was operated on in secondary septicemia to mitral valve endocarditis. He died from multiple organ failure syndrome on postoperative day 28. Mean duration of surgery was 256±67 min and mean cardiopulmonary bypass time was 149±68 min. Mean ICU stay was 3.4±2.9 days and mean hospital stay was 10.6±7.9 days. Re-exploration for bleeding was necessary in one patient (4%). Mean red blood cell transfusion was 1.1 package per patient (seven patients required postoperative heterologus blood transfusion). Regarding the postoperative morbidity we observed three major complications. Two patients (8%) developed respiratory failure requiring prolonged mechanical ventilatory support (>48 h), and one patient (4%) suffered from acute renal failure requiring ultrafiltration. Particularly significant is the absence of any cardiological, neurological, peripheral vascular and wound complications. However, we observed respiratory complications, such as pleural effusion and pneumothorax, in 13 patients. Of these patients 11 received a standard single-lumen endotracheal intubation during the operation. A morbidity analysis for all patients is shown in Table 4 . Echocardiograpy at hospital discharge revealed absence or minimal residual mitral regurgitation in all the mitral valve repair patients. Mean follow-up was 22.8±14.9 months and 100% complete. Late death occurred in two patients. Causes were sudden death and heart failure. The mean NYHA class at follow-up was 1.7±0.6. There was one late re-intervention at 2.4 years due to mitral prosthesis endocarditis. Echocardiographic follow-up showed trace to moderate (1+/4+) mitral regurgitation in all the patients who had a mitral repair. Follow-up questionary revealed that all the surviving patients preferred the minimally invasive approach when considering perioperative pain, rehabilitation and esthetic results.

4. Discussion

Median resternotomy in patients with functioning coronary artery bypass grafts remains a challenging procedure [2,3]. For this reason, alternative approaches have been advocated in order to reduce the risk of injuries to the grafts during sternal re-entry [4–7]. Re-entry to the chest via a minithoracotomy permits to minimize previous adhesive tissue dissection and reducing the bleeding. In fact, with this access, the right ventricle and the bypass conduits are far away from the working field and were not injured in any of our patients. The low requirements of blood transfusion in our series may be seen as a secondary benefit of this approach. Owing to the inability to isolate and control the ITA grafts through a right minithoracotomy, an adequate myocardial protection strategy represents the key to success in this subset of patients. Many authors reported encouraging results with moderate to profound hypothermia and ventricular fibrillation [8–11] and our results compare favorably to these, as evidenced by the absence of perioperative ischemic myocardial complications and low cardiac output syndrome. In our hands, the presence of a greater than minimal aortic regurgitation represents a contraindication for conducting the operation on ventricular fibrillation. In fact, we have observed that the presence of a backflow through the incompetent aortic valve could severely impair the visualization of the mitral valve and, in so doing, it could render the procedure very difficult. For these patients, an effective option to avoid redo sternotomy is represented by the Port-Access system, with endoaortic balloon clamping and antegrade cardioplegia delivery. Although a significant proportion of antegrade cardioplegia may go into the left ventricle, the use of additional doses of cardioplegia and deeper cooling have permitted to obtain satisfactory myocardial recovery in the patients in which we have adopted this technique.

A potential risk of performing mitral surgery on the fibrillating heart is cerebral air embolism [12]. Conducting the procedure with the aorta vented and the use of continuous carbon dioxide insufflation into the chest, may significantly reduce the incidence of this fearful complication. None of the patients in this series had an echocardiographically detected massive air embolism or suffered perioperative stroke or neurological complications.

The decision to repair a mitral valve should not be influenced by the operative approach, but rather by the pathology of the valve. For this reason an optimal exposure of the mitral valve is crucial during mitral valve surgery. In 23 patients we used the atrial retractor routinely used during primary minithoracotomic mitral valve procedures at our institution. This atrial retractor requires a supporting arm inserted in the chest through a right parasternal hole. However, in patients who had strong substernal adhesion, this maneuver could have been associated with a significant risk of injury to the right mammary peduncle. Hence, we decided to use a new self-expanding, pliable, retractor (MitraXS^®; Cardiolife Research, Lovain la Nueve, Belgium) which is inserted directly from the minithoracotomy and does not require any supporting arm [13] (Fig. 2 ). In our experience it was easy to use and it allowed an excellent valve exposure. In the present series we were able to perform a mitral repair only in 60% of the cases, which is a small percentage in comparison to the standard rate of our center. This is because we mainly dealt with patients who have had a previous mitral procedure which was not suitable for repair, and not because we had a bad exposure of the valve.

In patients requiring associated tricuspid valve surgery, we used a two-stage single femoral venous cannula. This cannula, which has two distal perforated sections separated by a non-perforated segment of 15 cm in length, has been ideated to drain simultaneously both superior and inferior vena cava. Its use allows to conduct the tricuspid procedure without caval veins encircling. Despite right atrial surgery with unsnared vena cava not being new [14], this cannula permits the avoidance of a second jugular venous cannula without obstructing the exposure of the tricuspid valve (Fig. 1). If a backflow from the caval veins is present, a small increase in negative pressure usually results in a clean operative field. Nevertheless, one may argue that air would be entrained in the venous line when the right atrium is open, if the two perforated sections are correctly positioned, reduction of the venous return and air locks in the venous line can be safely avoided. For this reason, a close collaboration between the surgeon and the anesthetist to check the correct positioning of the cannula with trans-esophageal echo, is mandatory.

The data presented here show very low operative morbidity and mortality. The absence of neurological, vascular and wound complications are of particular interest. However, when regarding the operative morbidity we observed that we had to deal with a high incidence of pulmonary complications. In our opinion, this was probably related to the fact that, at the beginning of our experience, we used a standard single lumen endotracheal tube and so we were obliged to displace laterally the right lung, in order to obtain a proper visualization of the heart, before institution of cardiopulmonary bypass. We think that this maneuver might be responsible for the development of the postoperative lung complications, especially in patients with severe preoperative lung dysfunction. For this reason, in the last nine patients of the series we have used a selective double lumen endotracheal tube and we have observed a reduction in the incidence of these complications. Taking into consideration these data, we strongly suggest the use of a selective double lumen intubation to reduce the incidence of respiratory complications in this subset of patients. In conclusion, the low operative mortality and the excellent early echocardiographic results attest to the safety and efficacy of this approach. The low complication rate, low need for blood transfusion, the avoidance of sternal wound infection and the high patient satisfaction represents the major advantages of this approach. The present paper is a description of a case series of 25 patients and the results should be interpreted as such. The small number of patients and the absence of a control group are the main limitations of this study.

References