Robotic cardiac surgery has evolved during the last decade and allowed surgeons to perform coronary artery bypass grafting; mitral, tricuspid, and aortic valve procedures; atrial septal defect closure; and epicardial or endocardial ablations for treatment of atrial fibrillation. Robotic technology has especially been used in the treatment of mitral valve disease, and recent clinical advances have made robotic totally endoscopic mitral valve repair feasible especially in degenerative valve disease.1–3
However, robotic technique may be challenging when the valve pathology is not suitable for repair and replacement is inevitable. The purpose of this study was to report a case series of robotic mitral valve replacement in patients with severe rheumatic mitral disease with a follow-up of 18 months.
From March 2010 to June 2013, a total of 61 patients underwent robotic cardiac procedures. Robotic procedures were performed using the da Vinci Si HD surgical systems (Intuitive Surgical, Inc, Sunnyvale, CA USA).
All patients underwent transthoracic echocardiography, coronary angiography, and vascular ultrasound imaging of the femoral vessels before the operation. The patients at high risk for peripheral artery disease underwent computed tomographic angiographic examination of the peripheral vessels. Patients with extensive coronary artery disease, severe peripheral vascular disease, extensive mitral annular calcification, and prior median sternotomy or right thoracotomy were excluded. No other exclusion criteria were used. Informed consent of the patients and local ethical committee approval were received.
Eighteen (28.5%) of the total robotic operations were mitral valve replacement with or without an additional cardiac procedure. Rheumatic disease was the underlying pathology in all patients. The mean (SD) follow-up period was 18 (10) months.
Anesthesia and Patient Positioning
After induction of general anesthesia with 2 to 4 mg/kg of sodium thiopental, 0.1 mg/kg of midazolam, 5 to 10 μg/kg of fentanyl, and 0.1 mg/kg of vecuronium intravenously, a double-lumen endotracheal tube was placed along with a left radial artery catheter and a multiplane transesophageal echocardiography probe. External defibrillator pads were placed properly. A chest roll was placed under the right shoulder, the right arm was placed below the lower limit of the operating table, and the table was rotated 20 to 30 degrees right-side up position. The incision sites were marked (Fig. 1A). Anesthesia was maintained with sevoflurane in air and oxygen along with a total of 5 to 8 mg/kg of intravenously administered fentanyl in divided doses.
Cardiopulmonary Bypass and Port Implantation
Peripheral cannulation was performed with a catheter placed via the right internal jugular vein, the common femoral artery, and the right common femoral vein, as described before.1–3 A 3-cm anterolateral thoracotomy incision was made in the right fourth intercostal space. The incision was approximately 3 cm lateral to the nipple. It was retracted with a small soft tissue retractor. Carbon dioxide insufflation was applied. The settings of the CO2 insufflator were set with a flow rate of 6 L/min. A camera port was placed through this incision in most of the patients. Alternatively, a 1-cm incision for the camera port, 2 cm laterally to the right nipple, and a 3-cm incision for anterolateral thoracotomy, 1 cm laterally to this camera port incision, could be made. The right arm port, the left arm port, and the left atrial retractor port were placed as described previously3 (Fig. 1A). After the port implantation, the robotic arms were connected to the ports (Fig. 1B). Cardiopulmonary bypass was instituted. The pericardium was opened, and pericardial edges were suspended on stay sutures. The ascending aorta was cross-clamped with a transthoracic clamp. Careful attempt was applied to position this clamp coming from the upper side of the junction of the atrium and the superior vena cava. The heart was arrested using cardioplegia of histidine-tryptophan-ketoglutarate Bretschneider solution (Custodiol; Bretschneider HTK-solution, Koehler Chemie, Bensheim, Germany).4 A single dose of 2 L of this solution was delivered into the aortic root with a transthoracic cannula through the thoracotomy in most of the patients. Transesophageal echocardiography confirmation of cross-clamping and cardioplegia delivery was done in every patient. An alternative method for cardioplegia delivery might be the aortic root cannula inserted through the second or the third intercostal space parasternally lateral to the internal mammary artery. This cannula can be used for both cardioplegia and venting.
After the heart was arrested, the left atrium was opened through a classic left atrial incision posterior to the interatrial groove. The exposure of the mitral valve was established by properly placing the left atrial retractor.
The common features regarding the valve pathology were nonpliable leaflet morphology (the anterior leaflet was affected in all patients, the posterior leaflet was affected in 77% of the patients additionally) and subvalvular calcifications in changing degrees. None of the patients had extensive annular calcifications. The patients with extensive annular calcifications that might require decalcification or debridement procedures with annular reconstruction were excluded.
The valve was excised with a curved scissor and long tip forceps in routine fashion. In the patients with heavily calcified valve tissue, the valve was grabbed with a prograsper instead of the long tip forceps because of the difficulty of handling the nonpliable and grossly stiff valve tissue. Whenever the subvalvular apparatus was suitable, the chordal structures were preserved; however, in most of the cases, this was not possible because of heavy calcifications.
St. Jude Medical (Medtronic, Minneapolis, MN, USA) or OnX (On-X Life Technologies Inc, Austin, TX, USA) mechanical heart valves were used for replacement. The valve prosthesis was implanted in routine fashion by using 12 to 14 stitches of 2/0 Ethibond pledgets sutures (Ethicon, Somerville, NJ, USA). All sutures were passed first from the annulus and then from the prosthesis, so all the sutures were placed on the sewing ring of the prosthesis before it was lowered in place. A small suture setter was placed to keep the sutures in order. The valve was removed from its holder before placing into the thoracic cavity to enable the deployment through a small incision. All knots were secured with a knot pusher (Figs. 1A–D). With this technique, the valve prosthesis with a soft sewing cuff may be more practically implanted when compared with the ones with a hard sewing cuff because the knot can be buried into the cuff easily and does not loosen rapidly in soft cuffs. Alternatively, an automatic mechanical knotting system, Cor-Knot (LSI Solution, Victor, NY, USA), or knotting the sutures inside by the robotic arms can be preferred.
Additional procedures were performed in classic fashion if needed. The atriotomy was closed using a premade loop suture, and the heart was deaired, and cross-clamp (CC) is removed. After adequate hemostasis was achieved, the robotic arms were removed from the chest, and a small flexible drainage tube was placed in the pericardium and one chest tube was placed in the right pleural space, all through existing port incisions. After decannulation, heparin was reversed, and all incisions were closed in layers.
Baseline characteristics are reported as mean (SD) and median for continuous variables as well as number (percentage) for categorical variables.
The mean (SD) age and Logistic EuroSCORE of the patients were 51.2 (11) years and 4.1% (4%), respectively. Demographic variables are shown in Table 1. Seven patients (41.1%) had an additional cardiac procedure (Table 2). The additional procedures were also performed with CC. No operative and hospital mortality were observed. The mean (SD) CC time and cardiopulmonary bypass time were 116 (30) and 178 (54) minutes, the mean (SD) drainage was 430 (350) mL, the mean (SD) intubation time was 9.4 (7) hours, the rate of the patients extubated within 6 hours or less was 94.4%, and the mean (SD) intensive care unit stay time was 30 (12) hours. Sixteen of the patients (88.8%) were discharged from the intensive care unit within the first 24 hours postoperatively. During the intensive care unit stay, one patient (5.5%) needed inotropic support. There was one early reoperation for bleeding (5.5%). One (5.5%) intensive care unit readmission and one (5.5%) hospital readmission were observed. No device-related complications were observed. During the midterm follow-up, there were no valve dysfunction at the echocardiographic examinations, no mortality, and no need for reoperation or reintervention (Table 3).
Rheumatic mitral disease may cause severe and complex lesions with calcification that complicate valve reconstruction and reduce the durability of valve repair. For this reason, rheumatic etiology and calcified lesions have been described as exclusion criteria for robotic mitral surgery.3,5 Thus, the main topic of robotic mitral surgery has been the valve repair so far.1,3,5,6
However, robotic mitral valve replacement can safely be performed with acceptable early results, applying a similar setup used in robotic mitral valve repair. Furthermore, additional procedures for tricuspid valve, atrial septum, and atrial fibrillation as usually seen in a patient with severe mitral valve disease can also be managed at the same stage safely.
A 3-cm anterolateral thoracotomy and use of a soft tissue retractor (instead of a working port) to allow a valve prosthesis deployment are the main differences of operative setup when compared with robotic mitral repair.
Implementing the camera port through the already opened 3-cm thoracotomy enables the directional port movement during the operation and facilitates the patient-side surgeons’ view of exposure and mobility, which are important during knot tying from outside.
The knot tying from outside the thoracic cavity, through the thoracotomy, was performed with the knot pusher using special endoscopic tying methods for knot safety. The knot was started with straight and double throws, and further ones were single, consecutively straight and reverse. Recently, we started to use Cor-Knot (LSI Solution, Victor, NY USA) to tie the knots during robotic operations, which have resulted in a shorter operative time. With this device, up to 10 minutes of less time for annuloplasty in robotic mitral repairs was reported7; our experience for mitral valve replacement is a saving of 15 minutes. Moreover, we started to use sutures with handmade loops at one end for atrial closure.8 This eliminated the need for knot tying at both ends of the atrial incision, which further decreased the operative time.
One important point regarding the patient selection is the valve pathology. Because the strength of the robotic devices would not be suitable for these processes such as decalcification or annular debridement, patients with extensive annular calcification were excluded in the group. These patients would probably be the best candidate for open surgery or minimally invasive techniques without a robot. In our series, only the patients with nonpliable leaflets with subvalvular calcifications in changing degrees without extensive annular calcification were referred to robotic surgery; others were operated on either with open technique or with minithoracotomy.
The overall mean CC time seems to be high in our series (118 minutes). However, there are seven patients with additional major procedures that contribute to the CC time. We started to use cardioplegia of histidine-tryptophan-ketoglutarate Bretschneider solution (Custodiol), by which up to 2 hours of ischemia is reported to be tolerated well during minimally invasive procedures.9 Moreover, the mean CC time was shortened with surgical experience, the latest mean CC time being approximately 85 minutes for the last six cases. This finding is consistent with the report of other groups also.10 This emphasizes that, with a well-trained robotic team and after a substantial learning curve, optimal results can be achieved.
This operation can also be performed with minimally invasive video-assisted techniques. Both techniques require a learning period. The main benefit of robotics to a mini video-assisted mitral valve replacement may be a smaller thoracotomy as well as better visualization and exposure of the valve. Moreover, the suture handling in difficult anatomies may be more feasible with a robot.
The team approach is extremely important during the procedure. The operations were performed by two surgeons, the console surgeon assisted by the patient-side surgeon and two well-trained operating room nurses. Both surgeons were capable of performing the operation safely, and both nurses were capable of setting up the entire equipment and assisting the patient-side surgeon perfectly, leading to improvement of operative safety and efficiency.
The present results suggest that robotic mitral valve replacement is feasible with early encouraging results. There has been no attempt to repair rheumatic mitral valves in our clinic; however, this may be a subject to another study.
It may be assumed that stepwise progression of robotic technology and procedure development will continue to make robotic operations simpler and more efficient, which will encourage more surgeons to take up this technology and extend the benefits of robotic surgery to a larger patient population.
Robotic mitral valve replacement for severe rheumatic mitral disease is technically feasible. Early results are encouraging. Patient selection criteria for robotic mitral valve surgery may be expanded to include valve replacements.
1. Lehr EJ, Rodriguez E, Chitwood WR. Robotic cardiac surgery. Curr Opin Anaesthesiol
. 2011; 24: 77–85.
2. Leff JD, Enriquez LJ. Robotic-assisted cardiac surgery. Int Anesthesiol Clin
. 2012; 50: 78–89.
3. Suri RM, Burkhart HM, Rehfeldt KH, et al. Robotic mitral valve repair for all categories of leaflet prolapse: improving patient appeal and advancing standard of care. Mayo Clin Proc
. 2011; 86: 838–844.
4. Bretschneider HJ. Myocardial protection. Thorac Cardiovasc Surg
. 1980; 28: 295–302.
5. Rehfeldt KH, Mauermann WJ, Burkhart HM, Suri RM. Robot-assisted mitral valve repair. J Cardiothorac Vasc Anesth
. 2011; 25: 721–730.
6. Cheng W, Fontana GP, De Robertis MA, et al. Is robotic mitral valve repair a reproducible approach? J Thorac Cardiovasc Surg
. 2010; 139: 628–633.
7. Nifong LW, Alwair H, Parker D, et al. Significant reduction in operative times using Cor-Knot in robot-assisted mitral valve repair. Paper presented at: Annual Scientific Meeting of the International Society for Minimally Invasive Cardiothoracic Surgery; June 12–15, 2013; Prague, Czech Republic.
8. Kiliç L, Şenay Ş, Ümit Güllü A, Alhan C. Leyla loop: a time-saving suture technique for robotic atrial closure. Interact Cardiovasc Thorac Surg
. 2013; 17: 579–580.
9. Seeburger J, Borger MA, Falk V, et al. Minimal invasive mitral valve repair for mitral regurgitation: results of 1339 consecutive patients. Eur J Cardiothorac Surg
. 2008; 34: 760–765.
10. Gao C, Yang M, Xiao C, et al. Robotically assisted mitral valve replacement
. J Thorac Cardiovasc Surg
. 2012; 143 (suppl): S64–S67.
This is a small case series of robotically assisted mitral valve replacement in 18 patients with severe rheumatic disease. It is unique in that most prior reports have focused on mitral valve repair and patients with degenerative valve disease. The authors provide a number of helpful technical tips for the reader for operating on these sometimes challenging patients. This report documents the feasibility of this approach but obviously does not establish an advantage of the robotic approach over either conventional or a nonrobotic right minithoracotomy minimally invasive surgery.