Minimally Invasive Aortic Valve Surgery

The surgical approach to aortic valve surgery has traditionally involved a median sternotomy with cannulation of the ascending aorta and right atrium for the conduct of cardiopulmonary bypass (CPB) [1]. Concerns regarding hospital stay, cost containment, and managed care have prompted less invasive surgical approaches for coronary artery bypass surgery. Recently, a new approach to aortic valve surgery has been used to facilitate early tracheal extubation, ambulation, and hospital discharge. We present two cases that describe the surgical approach as well as perioperative complications and the anesthetic considerations for managing these patients. Case Reports Case 1 A 41-yr-old (53 kg, 163 cm) woman with aortic insufficiency secondary to a faulty bicuspid valve presented for aortic valve surgery. Coronary angiography revealed normal coronary arteries. Preoperative transthoracic echocardiography revealed severe aortic insufficiency and moderate dysfunction (ejection fraction 40%) of the left ventricle. The patient received lorazepam 1 mg orally and glycopyrrolate 0.2 mg intramuscularly 30 min preoperatively. Hemodynamic monitoring included systemic and pulmonary artery catheters and a transesophageal echocardiography (TEE) probe. Anesthesia was induced with diazepam in 5-mg increments until loss of consciousness (total 30 mg), fentanyl 250 micro g and pancuronium 7 mg intravenously, followed by intubation with a 35-Fr, left-sided double-lumen endotracheal tube (DLT). Defibrillator pacemaker pads were placed over the left anterior and posterior thorax. The patient remained supine with a roll placed along the thoracic spine. Anesthetic maintenance consisted of halothane (0.5%-1.5%) supplemented with additional fentanyl (750 micro g). A right parasternal incision was performed from the lower edge of the second intercostal space to the superior edge of the fifth intercostal space and the third and fourth costal cartilages were excised. The right lung was deflated, the right pleura entered, and the pericardium opened to expose the ascending aorta. The left common femoral artery and vein were cannulated for CPB. Heparin (20,000 U) was administered, CPB was initiated, and an aortic valve repair was performed. Normothermia was maintained during CPB with an aortic crossclamp time of 29 min and a total pump time of 37 min. After the aorta was unclamped, sinus rhythm returned spontaneously. The patient was separated from CPB and protamine 350 mg was administered without incident. TEE revealed trivial aortic regurgitation, normal ventricular function, and an antegrade maximal aortic valve velocity of 2.6 m/s. Before closing, the surgeon performed intercostal nerve blocks with 20 mL of 0.25% bupivacaine. At the conclusion of surgery, the patient was reintubated with a single-lumen tube and transported to the intensive care unit (ICU) in stable condition. Total duration of the operation was 3 h 30 min with extubation occurring 2 h after arrival at the ICU. The patient experienced only mild discomfort which was well controlled with intravenous morphine and she was transferred to the floor 4 h after admission to the ICU. The remainder of her postoperative course was unremarkable, and she was discharged to her home on postoperative Day 3. Case 2 A 57-yr-old (73 kg, 173 cm) man with a history of severe aortic stenosis, noninsulin-dependent diabetes mellitus, and 15 pack-years of smoking presented for aortic valve replacement. Preoperative transthoracic echocardiography revealed severe aortic stenosis (peak gradient, 117 mm Hg), normal left ventricular function with moderate to severe hypertrophy, and a mildly dilated ascending aorta (sinus of Valsalva, 3.8 cm, ascending aorta 3.6 cm). Cardiac catheterization revealed normal coronary arteries and a heavily calcified aortic valve with an area of 0.6 cm2. The patient received lorazepam 1 mg orally 30 min preoperatively. Hemodynamic monitoring included systemic and pulmonary artery catheters, and a TEE probe was placed. Anesthesia was induced with diazepam in 5-mg increments until loss of consciousness (total 20 mg) along with pancuronium 7 mg and fentanyl 1 mg, and intubation was performed with an 8.0 single-lumen tube. Defibrillator-pacemaker pads were placed over the left anterior and posterior thorax. The patient remained supine with a roll placed along the thoracic spine. Anesthetic maintenance consisted of halothane (0.5%-1%) supplemented with fentanyl (500 micro g). A right parasternal incision was performed with cannulation of the left femoral vessels for CPB as described in Case 1. Heparin (25,000 U) was administered and CPB was initiated. An aortic homograft was inserted during normothermic CPB (crossclamp time 93 min, total pump time 112 min). After the aorta was unclamped, atrial-ventricular sequential pacing via epicardial wires was used for separation from CPB. TEE revealed initial mild left ventricular dysfunction (posterior wall hypokinesis), moderate to severe right ventricular dysfunction, no aortic insufficiency, and a large pocket of air in the apex of the left ventricle. The patient's thorax was shaken by the surgeon in an attempt to dislodge intracardiac air. Protamine 300 mg was administered without incident. After an intercostal nerve block performed by the surgeon, the patient was taken to the ICU in stable condition (mean arterial pressure, 70 mm Hg; cardiac output, 4.2 L/min; central venous pressure, 18 mm Hg). Shortly after arrival (30 min) at the ICU, he experienced cardiac arrest (ventricular fibrillation-ventricular tachycardia). He was resuscitated and returned to the operating room with multiple episodes of ventricular tachycardia. The right parasternal incision was reopened. However, additional surgical exposure was required, and a median sternotomy was performed. TEE showed severe left ventricular dysfunction (the septal and inferior walls were hypokinetic), moderate right ventricular dysfunction (the free wall was hypokinetic), and no aortic insufficiency. Heparin (30,000 U) was administered and CPB was initiated after cannulation of the ascending aorta and right atrium (total pump time, 20 min). Anesthetic management consisted of scopolamine 0.4 mg and pancuronium 10 mg prebypass with isoflurane (0.5%-1%), fentanyl 1 mg and midazolam 5 mg during CPB. A saphenous vein graft was placed to the right coronary artery (the surgeon was concerned with the previous right coronary artery reanastamosis to the aortic homograft) and an intraaortic balloon pump (IABP) was inserted into the right femoral artery. Separation from CPB required atrial pacing, IABP counterpulsation (1:1), and infusions of epinephrine 2 micro g/min, norepinephrine 2 micro g/min, and lidocaine 2 mg/min. TEE prior to leaving the operating room revealed normal left and right ventricular function and no aortic insufficiency. Protamine 500 mg was administered in a continuous infusion without incident, and the patient was taken to the ICU in stable condition. A total of 8 U of packed red cells and 12 U of platelets were transfused intraoperatively. The IABP was removed on postoperative Day 1 and the trachea was extubated on postoperative Day 2 without sequelae. The patient was neurologically intact and transferred to the floor on postoperative Day 2. Hospital discharge occurred on postoperative Day 6. Discussion Recently, a new strategy for aortic valve surgery consisting of a right anterior minithoracotomy has been used at our institution. A similar approach (left anterior minithoracotomy) has previously been used for coronary artery bypass surgery in selected patients [2,3]. This approach may reduce hospital and ICU length of stay, reduce respiratory therapy and nursing efforts, and facilitate patient care by allowing early extubation and reducing postoperative discomfort, enabling earlier mobilization and discharge [4]. At the time of this writing, 32 patients have had aortic valve surgery with this approach at our institution. There was no significant difference in total ICU length of stay (20.7 hours for median sternotomy versus 20.9 hours for minithoracotomy). Minithoracotomy patients were tracheally extubated earlier (28% vs 0% at four hours postoperatively) and had a hospital stay of two days less when compared to those receiving the median sternotomy approach. This reflects a direct cost savings of 26.7% (Data on file; Cleveland Clinic Cardiothoracic Anesthesia Cardiac Database, 1996), which was achieved by decreasing the postoperative stay despite similar stays in the ICU. In addition, the incidence of major organ system morbidity, wound infection, or hospital readmission within 30 days was similar in both groups. The anesthetic technique chosen reflects a goal of rapid tracheal extubation but provides amnesia and postoperative analgesia. This requires titration of intravenous anesthetics and increased use of inhalation techniques compared to traditional opioid-based anesthesia. In addition, the patients must be normothermic upon arrival to the ICU (facilitated by the use of warm CPB), hemodynamically stable without bleeding, and without residual neuromuscular paralysis. At our institution, postoperative hemodynamic stability, satisfactory analgesia, and return of motor strength is observed for 60 minutes before tracheal extubation is considered. At our institution, these patients all receive systemic and pulmonary artery catheters. Initially, airway management consisted of insertion of a DLT for one-lung ventilation to facilitate surgical exposure. However, as our surgical experience has increased, the use of a DLT has decreased. After the airway is secured, defibrillator-pacemaker pads (Zoll Medical Corp., Burlington, MA) are placed. These pads allow for defibrillation, cardioversion, and pacing, as it is difficult for the surgeon to introduce defibrillation paddles or pacemaker wires with this approach. A pulmonary artery catheter with pacing capabilities can also be used, but care must be taken to maintain proper lead placement during the bypass procedure with surgical manipulation of the aorta and heart. Intraoperative TEE can help assess the adequacy of valve repair and left ventricular air removal. Intracardiac air may cause ventricular fibrillation after the aortic cross-clamp is removed and can require numerous countershocks. After successful defibrillation, bradycardia requiring pacing may also occur. The patient is transported to the ICU with the defibrillator-pacemaker pads in place, should resuscitation become necessary. The second case demonstrates complications which can occur with this approach. This limited surgical exposure can prevent the surgeon from adequate removal of intracardiac air, control of mediastinal bleeding or evacuation of cardiac tamponade, especially in emergency situations. These situations may require median sternotomy for hemostasis or rapid return to CPB. Additional morbidity can occur if these complications arise after the trachea is extubated. We have described a different surgical approach for aortic valve surgery. Potential advantages of this approach may include a smaller incision with less postoperative discomfort and reduced hospital stay. The paramedian approach may also be beneficial for patients with pathology that increases the risk of soft tissue damage during median sternotomy (e.g., ascending aortic aneurysm, chronic obstructive pulmonary disease, or right ventricular dilatation). However, this approach provides limited surgical access and exposure. It may also hinder control of bleeding and removal of air from the ventricle, and may require median sternotomy to facilitate surgical repair. The requirement of cannulation of the femoral vessels for CPB may be a contraindication in selected patients (e.g., morbid obesity, aortoiliac disease) [5]. In addition, the anesthesiologist cannot directly observe myocardial contractility and ventricular filling, as is possible with median sternotomy cases, and must rely on hemodynamic monitoring and TEE to guide separation from CPB. In addition, the sacrifice of the right internal thoracic artery as a potential future conduit may be a disadvantage for patients who return for coronary artery bypass grafting. Recently, there has been concern regarding the safety and superiority of less invasive coronary artery and thoracic surgery [6,7]. These issues focus on the quality of the revascularization through a limited-access surgical incision, complications related to routine femoral vessel cannulation for CPB (e.g., retroperitoneal bleeding, leg ischemia, neural injury), speed of recovery, and length of hospital stay compared to traditional surgical techniques. More experience is necessary to determine whether the results of aortic valve repair through a paramedian incision are reproducible and similar to a median sternotomy approach, whether this surgical approach is adequate to deal with patients who return to the operating room for postoperative bleeding or tamponade, and whether this approach is more cost-effective or truly "less invasive" than a median sternotomy. Although this approach currently is controversial, continued use will allow for current and future guidelines, along with safety and cost analysis, to be developed.