- Email: [email protected]
Minimally Invasive Approaches to Aortic Valve Surgery: Brigham Experience
Minimally Invasive Approaches to Aortic Valve Surgery: Brigham Experience Betty S. Kim, MD,* Edward G. Soltesz, MD, MPH,† and Lawrence H. Cohn, MD‡ Aortic valve surgery is a proven and effective therapy for severe aortic stenosis and insufficiency. Conventional aortic valve surgery is performed with a full sternotomy, cardiopulmonary bypass, and replacement of the diseased aortic valve. Unlike minimally invasive (or “off-pump”) coronary artery bypass, minimally invasive aortic valve surgery still requires cardiopulmonary bypass but refers primarily to smaller incisions and access. Minimally invasive approaches to aortic valve surgery have evolved over the past decade and have become the standard in institutions that perform large-volume minimally invasive cardiac surgery. The upper hemisternotomy has become our standard approach to isolated aortic valve surgery. It is a safe and effective technique with a similar morbidity and mortality to conventional aortic valve surgery. Patients derive clear benefits from this minimally invasive approach including less pain, shorter length of hospital stay, and faster return to preoperative function levels. Semin Thorac Cardiovasc Surg 18:148-153 © 2006 Elsevier Inc. All rights reserved. KEYWORDS minimally invasive aortic valve surgery, upper hemisternotomy, cardiac surgery
A
ortic valve surgery is a proven and effective therapy for severe aortic stenosis and insufficiency. Thousands of aortic valve replacements are performed annually, making it one of the most common cardiac surgical procedures performed. The conventional technique of aortic valve replacement (AVR) consists of a full sternotomy, cardiopulmonary bypass, and the subsequent replacement of the diseased aortic valve. In the early 1990s, the clear success of laparoscopic cholecystectomy compared with traditional open surgery awakened surgeons= and patients= interests in minimally invasive approaches to cardiac surgery. With this impetus in 1996, Cosgrove and Sabik first reported a minimally invasive approach to AVR using a right parasternal incision.1 Other minimally invasive approaches followed and were developed in keeping with the general trend toward less invasive procedures. Although cardiopulmonary bypass is still necessary, the smaller incisions with these approaches are far less traumatic to patients and in fact result in less postoperative pain, less blood loss, and quicker return to preoperative function
*Division of Cardiac Surgery, Brigham and Women’s Hospital, Boston, MA. †Cardiothoracic Surgery, Brigham and Women’s Hospital, Boston, MA. ‡Harvard Medical School, Boston, MA. Address reprint requests to Lawrence H. Cohn, MD, Division of Cardiac Surgery, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115. E-mail: [email protected]
148
1043-0679/06/$-see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1053/j.semtcvs.2006.07.007
levels.2 More importantly, these minimally invasive approaches have not compromised safety and good outcome of the traditional AVR. Over the past decade, the upper hemisternotomy approach has evolved as the most adaptable technique since it does not deviate much from traditional cannulation and myocardial protection techniques, and therefore, is familiar to most surgeons. This article will review the various techniques of minimally invasive AVR described to date in the literature as well as review our series of 901 patients at the Brigham and Women’s Hospital who underwent minimally invasive AVR between July 1996 and January 2006.
Minimally Invasive Approaches to AVR A Medline search of published series of minimally invasive aortic valve replacements between 1996 and 2006 describes four different minimally invasive approaches to aortic valve surgery: right parasternal, transverse sternotomy, right thoracotomy, and upper hemisternotomy (Fig. 1). The common goal is to reduce the size of the incision and avoid the morbidity of a full sternotomy, thereby decreasing surgical trauma and pain. The right parasternal and transverse sternotomy approaches are no longer utilized; due to their historical value, though, these techniques will also be described. In 1996, Cosgrove described the right parasternal approach to AVR.1 A 10-cm right parasternal incision extending
Minimally invasive approaches to aortic valve surgery
149
Figure 1 Minimally invasive approaches to the aortic valve. (A) Conventional full sternotomy. (B) Right parasternal incision. (C) Right anterior thoracotomy. (D) Upper hemisternotomy.
from the second costal cartilage to the fifth costal cartilage was made; the third and fourth costal cartilages were excised, and the right internal thoracic artery (RITA) was ligated. The right pleural cavity was then entered just lateral to the edge of the sternum, and with the pericardium incised, the ascending aorta and the right atrium were exposed. The common femoral artery and vein were exposed and cannulated so the tip of the venous cannula lay in the superior vena cava for adequate drainage. Clamping the commissure sutures to the surrounding drapes elevated the aortic root for improved exposure. After the aortic valve replacement, right-sided chest tubes were placed, and the parasternal incision was closed following intercostal nerve blocks.1 The disadvantages of this approach included occasional cases of less than optimal aortic valve exposure and an unstable anterior chest wall with obligatory sacrifice of the RITA. Additionally, a second incision in the groin was necessary for femoral cannulation. Importantly, conversion to full median sternotomy was not practical, if needed.3 Soon thereafter in 2000, Gillinov and coworkers reported that they have evolved away from the parasternal incision to the more accepted hemisternotomy, which they concluded was the incision of choice for minimally invasive aortic valve surgery.4 The transverse sternotomy approach was popularized early in the evolution of minimally invasive aortic valve surgery and involved transection of the sternum at the third intercostal space (ICS). After both internal thoracic arteries were ligated, a retractor was placed in this space to expose the aortic root, superior vena cava, and the right atrium. The aorta and right atrium could then be directly cannulated. After the aortic valve was replaced, the sternotomy was approximated with two interrupted sternal wires. The disadvantages of this approach included the obligatory sacrifice of both internal thoracic arteries with potential wound-healing
problems and loss of viable conduit for future coronary bypass surgery. Conversion to median sternotomy was also difficult. Only small series have been reported using this technique since some authors found unacceptable morbidity and mortality rates with this approach.5 In 1997, Bennetti and coworkers described the right thoracotomy approach using a 6- to 7-cm anterolateral thoracotomy incision in the third intercostal space. With use of a specially adapted wound spreader, the rib space was opened without removal of rib or costal cartilage. After incision in the pericardium, the aortic root, right atrium, and right superior pulmonary vein were exposed and available for cannulation for bypass and venting.6 The disadvantages of this approach included significant postthoracotomy pain and difficult conversion to median sternotomy. Nonetheless, Kort and colleagues in 2001 published their success in using a 5-cm right anterior thoracotomy approach in the second or third interspace. The aorta was directly cannulated in about half their patients, and the venous drainage catheter was placed into the right atrium from the femoral vein under transesophageal echocardiography (TEE) guidance. Direct external cross-clamping was applied before the aortotomy. They reported reduced postoperative pain and stress with their approach.7 Gundry and coworkers in 1998 described the partial ministernotomy approach. A 6- to 9-cm skin incision was used, and the sternum was divided not only in the midline but also with a T-incision into an intercostal space. The internal thoracic arteries were left intact. A small Finochetti retractor was used for exposure and the pericardial edges were sewn to the skin, bringing the heart closer to the surface. The aorta and right atrium were directly cannulated and a retrograde cardioplegia catheter was advanced into the coronary sinus using TEE guidance. The advantages of this approach were
150 clear: excellent exposure of the necessary cardiac structures for cannulation with standard cannulas, retractors, and myocardial protection strategies. In addition, conversion to full sternotomy was easy.8 A modification of this mini-sternotomy has since evolved that does not require transection of the sternum, but rather only an inverted L sternal incision extending into the right fourth intercostal space as described by Svensson and Tam and colleagues.9,10 This has become the preferred approach at our institution. Essentially all recent comparative studies report that minimally invasive approaches to aortic valve surgery, primarily the upper hemisternotomy and right thoracotomy approaches, are as safe and effective as conventional aortic valve surgery using a full-median sternotomy. Importantly, patients derive a variety of tangible benefits from these new surgical approaches. Patients have less pain, better cosmetic result, less use of blood products, and a shorter hospital stay. All these benefits translate into lower hospital costs.2,11-16 An obvious advantage to these approaches is that only the necessary structures for aortic valve surgery are exposed, making future reoperations safer by leaving most surfaces of the heart, especially the right ventricle, undisturbed. More recent series report that the upper hemisternotomy approach has become the preferred technique of exposure and more widely accepted by most surgeons.17
Upper Hemisternotomy for Minimally Invasive AVR Every patient undergoing isolated aortic valve surgery is evaluated for the potential use of a minimally invasive approach. Patients who are obese and have a wide anterior-posterior diameter, scoliosis, or significant chest wall deformity are generally excluded from a hemisternotomy due to difficulty with exposure. Patients with concomitant coronary artery disease have been excluded due to the need for a full sternotomy for a combined valve and coronary artery bypass (CABG) procedure. However, lately, if there is single- or double-vessel coronary artery disease amenable to percutaneous stenting, these patients undergo a hybrid-type procedure: coronary stenting hours before their minimally invasive AVR.18 Reoperations are not contraindicated nor is a patent left internal mammary artery graft from a previous CABG, as myocardial protection can be achieved with hypothermia.19,20 External defibrillator pads and TEE are used routinely in every patient before the incision. An approximate 6- to 8-cm upper midline skin incision is made in the chest, and the upper sternum is separated in the midline with a sternal saw. The right fourth ICS is scored and the upper sternum is subsequently separated horizontally with an oscillating saw taking care to avoid the RITA. A Koros CAB sliding retractor (T Koros Surgical Instruments Corp., Moorpak, CA) is used to expose the upper mediastinum. The pericardium is opened in the usual fashion and pericardial stitches are sutured to the deep dermis, allowing for elevation of the ascending aorta (Fig. 2). The ascending aorta is then palpated for disease; an epiaortic scan is used when there is suspicion
B.S. Kim, E.G. Soltesz, and L.H. Cohn for significant aortic calcification or atheromatous disease. The distal ascending aorta is cannulated in the usual fashion. Venous drainage is provided by a 24-F right-angled cannula placed into the right atrial appendage providing this cannula will not obstruct the aortic valve exposure. Otherwise, a percutaneous femoral venous cannula (23- to 27-F BioMedicus cannula) is placed, preferably into the right femoral vein, given its straighter course into the IVC. The tip of the venous cannula is guided with the TEE to the right atrium–SVC junction. Vacuum-assisted drainage on cardiopulmonary bypass allows for use of smaller cannula sizes. After directly cross-clamping the ascending aorta, antegrade cardioplegia is given through the aortic root or directly into the coronary ostia once the aorta is opened (Fig. 3). If retrograde cardioplegia is used, the retrograde catheter is placed in the tip of the RAA and advanced into the coronary sinus under TEE guidance. If a vent is needed, the right superior pulmonary vein can be quickly exposed in a decompressed heart. Otherwise, a small 12-F wire-reinforced catheter can be directly placed into the left ventricle through the aortic annulus. Continuous carbon dioxide insufflation can be advantageous in minimizing intracardiac air and shortening de-airing at the end of the case. After completion of the aortic valve operation, with the heart still decompressed on cardiopulmonary bypass, two small flexible chest tubes are placed from the epigastrium to lay in the mediastinum, being careful to protect the heart during passage. Bipolar ventricular wires are placed on the muscular right ventricle and brought out through the right fourth ICS. Adequate de-airing is confirmed by TEE. Once separated from cardiopulmonary bypass and after hemostasis, the upper sternum is approximated with four to five stainless steel wires, and the skin is closed in the usual fashion.
Brigham Experience and Results The Brigham and Women’s Hospital as an institution has performed 901 minimally invasive AVRs between July 1996 and January 2006. The demographics of patients demonstrate that 546 patients were male (61%) with an average age of 62 (range 25 to 94) and 355 patients were female (39%) with an average age of 69 (range 24 to 95). Additionally, 157 patients (17%) were over 80 years old. One hundred seventeen patients were reoperations. Seven hundred sixteen patients (79%) were in NYHA Class 2 or 3 with 40 patients in NYHA Class 4. The etiologies of aortic valve disease were calcific in 591 (66%), bicuspid in 94 (10.4%), rheumatic in 32 (3.5%), annuloectasia in 33 (3.6%), endocarditis in 31 (3.4%), and structural valve degeneration and myxomatous in the remainder. All types of aortic valves were implanted, including Carpentier Edwards pericardial valves in 567 (63%), mechanical valves in 194 (22%), homografts in 72 (8%), Mosaic/Hancock valves in 45 (5%), and aortic valve repairs and stentless valves in the remainder. Our earlier experience with different aortic valve exposures included 37 cases using the right parasternal incision and 3 using the right thoracotomy incision. We now have experience with 861 cases of minimally invasive AVR performed through an upper hemisternotomy approach. Intraoperatively, the aver-
Minimally invasive approaches to aortic valve surgery
151
Figure 2 (A) Six centimeter upper hemisternotomy incision. (B) Operative field with Koros retractor, ascending aorta with percutaneous femoral venous cannula, and cross-clamp with antegrade cardioplegia catheter. (C) After aortotomy, unobstructed view of aortic valve with exposed coronary ostia.
age aortic cross-clamp and cardiopulmonary bypass times for first time AVRs were 79 and 109 minutes, respectively. Of note, these times are significantly shorter than those we reported in 2001 with our first 365 minimally invasive-AVR patients (91 and 124 minutes).21 Early deaths occurred in 18 patients (1.9%). In 24 patients (2.7%), a reoperation for bleed was required. The stroke rate was 2.4% (22 patients), while the incidence of new postoperative atrial fibrillation was 22% (201 patients). In 40 patients (4.4%), a pacemaker and/or automatic implantable cardiac defibrillator was required. Wound infection occurred in 12 patients (1.3%). There were four patients (0.4%) who had early (⬍3 months) reoperative AVRs for endocarditis. Other morbidities included myocardial infarction and multisystem organ failure. Of all first-time minimally invasive AVRs, 53% of patients did not receive any blood products. Of those who did require postoperative transfusion, the average red blood cell transfusion was 2 units. For reoperative AVRs, the average red blood cell
transfusion was 3 units. The average length of stay for patients without complications was 6.9 days (range 3 to 52); most patients staying more than 20 days were in fact awaiting rehabilitation bed placement. Most patients (71%) went home; only 26% of patients required a stay in rehabilitation. Of the 157 octogenarians, 62% of patients went to a rehabilitation facility. According to the Society of Thoracic Surgery National Database Fall 2005 Executive Summary, the unadjusted operative mortality for conventional aortic valve replacements is approximately 3.5%.22 Our 30-day operative mortality is 1.9%, better than the national average. Furthermore, the majority of our first-time minimally invasive AVR patients did not receive any blood products. Most of the patients went home instead of to a rehabilitation facility after an accepted length of stay of 6 to 7 days. Even the octogenarians generally did well after minimally invasive AVR with about 40% of them returning home.
B.S. Kim, E.G. Soltesz, and L.H. Cohn
152
Figure 3 (A) Upper hemisternotomy incision for re-operative aortic valve surgery. (B) Right axillary artery cannulation with side graft arterial cannulation. (C) Limited mediastinal dissection with unobstructed view of the aortic valve and untouched previous coronary grafts.
Discussion Our updated series underscores the safety of the upper hemisternotomy approach to aortic valve replacement. The morbidity and mortality are comparable to conventional aortic valve replacements performed through a full median sternotomy. The upper hemisternotomy approach allows adequate exposure to perform aortic valve replacements for all etiologies and all types of implantation, with reasonable cross-clamp and bypass times, and with minimal alteration in cannulation techniques. Therefore, the learning curve for the surgeon is short. Studies have shown that patients benefit from these minimally invasive approaches; they are not merely pleasing to the public. There is less pain, less use of blood products, greater percentage of discharges home, and quicker return to preoperative function levels.
This technique of minimally invasive aortic valve exposure has also been extrapolated to surgery for more complex aortic valve pathology including valve conduits and root replacements. Furthermore, the upper hemi-resternotomy has been used in 117 reoperative cases at our institution. In 2000, Byrne and coworkers published an early series of these patients and highlighted their excellent results.20 Preoperative computed tomographic angiograms with 3D reconstruction have helped in surgical planning for patients with patent coronary grafts. Right axillary or femoral artery and percutaneous femoral venous cannulation are performed before upper hemi-resternotomy. Mediastinal dissection is kept to a minimum, enough to expose the aortic valve. Since there is no dissection and attempted exposure of previous coronary grafts, including LIMA grafts, reoperative aortic valve surgery through this approach is safer.19 Since the LIMA pedicle is
Minimally invasive approaches to aortic valve surgery not dissected in such situations, myocardial protection must be adequately achieved with moderate hypothermia (20 to 22°C). Minimally invasive aortic valve surgery has undergone a decade of refinements since its inception in 1996 and has now become a respected and safe option for most all patients requiring aortic valve or ascending aortic surgery. The familiar, easily adaptable upper hemi-sternotomy approach should be considered an option in all patients presenting for isolated aortic valve surgery, first time and reoperative. Even patients with concomitant coronary artery disease should not be immediately dismissed from the minimally invasive AVR option, as hybrid coronary stenting followed by minimally invasive AVR can be accomplished.18 Particularly, elderly patients with significant comorbidities, who cannot tolerate a prolonged combined valve/CABG procedure, may in fact do better with the hybrid approach. Similarly, patients with previous chest irradiation might do better with a hemisternotomy and limited mediastinal dissection. The upper hemisternotomy for aortic valve surgery has become the standard approach at the Brigham and Women’s Hospital. Direct in-line vision over the aorta and the aortic valve and expeditious conversion to full sternotomy if necessary make this approach more easily accepted by surgeons who are learning to perform minimally invasive aortic valve surgery.23 We predict that this technique will become the standard of care for patients undergoing aortic valve surgery in whom there is no concomitant coronary artery disease.
References 1. Cosgrove DM, Sabik JF: Minimally invasive approach for aortic valve operations. Ann Thorac Surg 62:596-597, 1996 2. Cohn LC, Adams DH, Couper GS, et al: Minimally invasive cardiac valve surgery improves patient satisfaction while reducing costs of cardiac valve replacement and repair. Ann Surg 226(4):421-428, 1997 3. Cosgrove DM, Sabik JF, Navia JL: Minimally invasive valve operations. Ann Thorac Surg 65:1535-1538, 1998 4. Gillinov AM, Banbury MK, Cosgrove DM: Hemisternotomy approach for aortic and mitral valve surgery. J Card Surg 15:15-20, 2000 5. Bridgewater B, Steyn RS, Ray S, et al: Minimally invasive aortic valve replacement through a transverse sternotomy: a word of caution. Heart 79:605-607, 1998
153 6. Benetti FJ, Mariani MA, Rizzardi JL, et al: Minimally invasive aortic valve replacement. J Thorac Cardiovasc Surg 113:806-807, 1997 7. Kort S, Applebaum RM, Grossi EA, et al: Minimally invasive aortic valve replacement. Echocardiographic and clinical results: Am Heart J 142: 476-481, 2001 8. Gundry SR, Shattuck OH, Razzouk AJ, et al: Facile minimally invasive cardiac surgery via ministernotomy. Ann Thorac Surg 65:1100-1104, 1998 9. Svensson LG, D=Agostino RS: “J” incision minimal-access valve operations. Ann Thorac Surg 66(3):1110-1112, 1998 10. Tam RK, Almeida AA: Minimally invasive aortic valve replacement via hemisternotomy: a preliminary report. Eur J Cardiothorac Surg 14(1): S134-S137, 1998 11. Sharony R, Grossi EA, Saunders PC, et al: Propensity score analysis of a six-year experience with minimally invasive isolated aortic valve replacement. J Heart Valve Dis 13:887-893, 2004 12. Dogan S, Dzemali O, Wimmer-Greinecker G, et al: Minimally invasive versus conventional aortic valve replacement: a prospective randomized trial. J Heart Valve Dis 12:76-80, 2003 13. Bonacchi M, Prifti E, Giunit G, et al: Does ministernotomy improve postoperative outcome in aortic valve operation? A prospective randomized study. Ann Thorac Surg 73:460-465, 2002 14. Sharony R, Grossi EA, Saunders PC, et al: Minimally invasive aortic valve surgery in the elderly: a case-control study. Circulation 108:II43II47, 2003 (suppl 1) 15. Cohn LH: Minimally invasive valve surgery. J Card Surg 16(3):260265, 2001 16. Leshnower BG, Trace CS, Boova RS: Port-access-assisted aortic valve replacement: a comparison of minimally invasive and conventional techniques. Heart Surg Forum 9(2):E560-E564, 2006 17. Mihaljevic T, Cohn LH, Unic D, et al: One thousand minimally invasive valve operations: early and late results. Ann Surg 240(3):529-534, 2004 18. Byrne JG, Leacche M, Unic D, et al: Staged initial percutaneous coronary intervention followed by valve surgery (“hybrid approach”) for patients with complex coronary and valve disease. J Am Coll Cardiol 45:14-18, 2005 19. Byrne JG, Karavas AN, Filsoufi F, et al: Aortic valve surgery after previous coronary artery bypass grafting with functioning internal mammary grafts. Ann Thorac Surg 73:779-784, 2002 20. Byrne JG, Karavas AN, Adams DH, et al: Partial upper re-sternotomy for aortic valve replacement or re-replacement after previous cardiac surgery. Eur J Cardiothorac Surg 18(3):282-286, 2000 21. Sepic J, Cohn LH: Minimally Invasive Aortic Valve Surgery in Minimally Invasive Cardiac Surgery (ed 2). Totowa, NJ, Humana Press Inc., 2004, pp 293-307 22. STS National Database. Fall 2005 Executive Summary. Duke Clinical Research Institute, 2005 23. Chitwood WR: Minimally invasive aortic valve surgery: what is “port access”? Am Heart J 142:391-392, 2001 (editorial)
A
ortic valve surgery is a proven and effective therapy for severe aortic stenosis and insufficiency. Thousands of aortic valve replacements are performed annually, making it one of the most common cardiac surgical procedures performed. The conventional technique of aortic valve replacement (AVR) consists of a full sternotomy, cardiopulmonary bypass, and the subsequent replacement of the diseased aortic valve. In the early 1990s, the clear success of laparoscopic cholecystectomy compared with traditional open surgery awakened surgeons= and patients= interests in minimally invasive approaches to cardiac surgery. With this impetus in 1996, Cosgrove and Sabik first reported a minimally invasive approach to AVR using a right parasternal incision.1 Other minimally invasive approaches followed and were developed in keeping with the general trend toward less invasive procedures. Although cardiopulmonary bypass is still necessary, the smaller incisions with these approaches are far less traumatic to patients and in fact result in less postoperative pain, less blood loss, and quicker return to preoperative function
*Division of Cardiac Surgery, Brigham and Women’s Hospital, Boston, MA. †Cardiothoracic Surgery, Brigham and Women’s Hospital, Boston, MA. ‡Harvard Medical School, Boston, MA. Address reprint requests to Lawrence H. Cohn, MD, Division of Cardiac Surgery, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115. E-mail: [email protected]
148
1043-0679/06/$-see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1053/j.semtcvs.2006.07.007
levels.2 More importantly, these minimally invasive approaches have not compromised safety and good outcome of the traditional AVR. Over the past decade, the upper hemisternotomy approach has evolved as the most adaptable technique since it does not deviate much from traditional cannulation and myocardial protection techniques, and therefore, is familiar to most surgeons. This article will review the various techniques of minimally invasive AVR described to date in the literature as well as review our series of 901 patients at the Brigham and Women’s Hospital who underwent minimally invasive AVR between July 1996 and January 2006.
Minimally Invasive Approaches to AVR A Medline search of published series of minimally invasive aortic valve replacements between 1996 and 2006 describes four different minimally invasive approaches to aortic valve surgery: right parasternal, transverse sternotomy, right thoracotomy, and upper hemisternotomy (Fig. 1). The common goal is to reduce the size of the incision and avoid the morbidity of a full sternotomy, thereby decreasing surgical trauma and pain. The right parasternal and transverse sternotomy approaches are no longer utilized; due to their historical value, though, these techniques will also be described. In 1996, Cosgrove described the right parasternal approach to AVR.1 A 10-cm right parasternal incision extending
Minimally invasive approaches to aortic valve surgery
149
Figure 1 Minimally invasive approaches to the aortic valve. (A) Conventional full sternotomy. (B) Right parasternal incision. (C) Right anterior thoracotomy. (D) Upper hemisternotomy.
from the second costal cartilage to the fifth costal cartilage was made; the third and fourth costal cartilages were excised, and the right internal thoracic artery (RITA) was ligated. The right pleural cavity was then entered just lateral to the edge of the sternum, and with the pericardium incised, the ascending aorta and the right atrium were exposed. The common femoral artery and vein were exposed and cannulated so the tip of the venous cannula lay in the superior vena cava for adequate drainage. Clamping the commissure sutures to the surrounding drapes elevated the aortic root for improved exposure. After the aortic valve replacement, right-sided chest tubes were placed, and the parasternal incision was closed following intercostal nerve blocks.1 The disadvantages of this approach included occasional cases of less than optimal aortic valve exposure and an unstable anterior chest wall with obligatory sacrifice of the RITA. Additionally, a second incision in the groin was necessary for femoral cannulation. Importantly, conversion to full median sternotomy was not practical, if needed.3 Soon thereafter in 2000, Gillinov and coworkers reported that they have evolved away from the parasternal incision to the more accepted hemisternotomy, which they concluded was the incision of choice for minimally invasive aortic valve surgery.4 The transverse sternotomy approach was popularized early in the evolution of minimally invasive aortic valve surgery and involved transection of the sternum at the third intercostal space (ICS). After both internal thoracic arteries were ligated, a retractor was placed in this space to expose the aortic root, superior vena cava, and the right atrium. The aorta and right atrium could then be directly cannulated. After the aortic valve was replaced, the sternotomy was approximated with two interrupted sternal wires. The disadvantages of this approach included the obligatory sacrifice of both internal thoracic arteries with potential wound-healing
problems and loss of viable conduit for future coronary bypass surgery. Conversion to median sternotomy was also difficult. Only small series have been reported using this technique since some authors found unacceptable morbidity and mortality rates with this approach.5 In 1997, Bennetti and coworkers described the right thoracotomy approach using a 6- to 7-cm anterolateral thoracotomy incision in the third intercostal space. With use of a specially adapted wound spreader, the rib space was opened without removal of rib or costal cartilage. After incision in the pericardium, the aortic root, right atrium, and right superior pulmonary vein were exposed and available for cannulation for bypass and venting.6 The disadvantages of this approach included significant postthoracotomy pain and difficult conversion to median sternotomy. Nonetheless, Kort and colleagues in 2001 published their success in using a 5-cm right anterior thoracotomy approach in the second or third interspace. The aorta was directly cannulated in about half their patients, and the venous drainage catheter was placed into the right atrium from the femoral vein under transesophageal echocardiography (TEE) guidance. Direct external cross-clamping was applied before the aortotomy. They reported reduced postoperative pain and stress with their approach.7 Gundry and coworkers in 1998 described the partial ministernotomy approach. A 6- to 9-cm skin incision was used, and the sternum was divided not only in the midline but also with a T-incision into an intercostal space. The internal thoracic arteries were left intact. A small Finochetti retractor was used for exposure and the pericardial edges were sewn to the skin, bringing the heart closer to the surface. The aorta and right atrium were directly cannulated and a retrograde cardioplegia catheter was advanced into the coronary sinus using TEE guidance. The advantages of this approach were
150 clear: excellent exposure of the necessary cardiac structures for cannulation with standard cannulas, retractors, and myocardial protection strategies. In addition, conversion to full sternotomy was easy.8 A modification of this mini-sternotomy has since evolved that does not require transection of the sternum, but rather only an inverted L sternal incision extending into the right fourth intercostal space as described by Svensson and Tam and colleagues.9,10 This has become the preferred approach at our institution. Essentially all recent comparative studies report that minimally invasive approaches to aortic valve surgery, primarily the upper hemisternotomy and right thoracotomy approaches, are as safe and effective as conventional aortic valve surgery using a full-median sternotomy. Importantly, patients derive a variety of tangible benefits from these new surgical approaches. Patients have less pain, better cosmetic result, less use of blood products, and a shorter hospital stay. All these benefits translate into lower hospital costs.2,11-16 An obvious advantage to these approaches is that only the necessary structures for aortic valve surgery are exposed, making future reoperations safer by leaving most surfaces of the heart, especially the right ventricle, undisturbed. More recent series report that the upper hemisternotomy approach has become the preferred technique of exposure and more widely accepted by most surgeons.17
Upper Hemisternotomy for Minimally Invasive AVR Every patient undergoing isolated aortic valve surgery is evaluated for the potential use of a minimally invasive approach. Patients who are obese and have a wide anterior-posterior diameter, scoliosis, or significant chest wall deformity are generally excluded from a hemisternotomy due to difficulty with exposure. Patients with concomitant coronary artery disease have been excluded due to the need for a full sternotomy for a combined valve and coronary artery bypass (CABG) procedure. However, lately, if there is single- or double-vessel coronary artery disease amenable to percutaneous stenting, these patients undergo a hybrid-type procedure: coronary stenting hours before their minimally invasive AVR.18 Reoperations are not contraindicated nor is a patent left internal mammary artery graft from a previous CABG, as myocardial protection can be achieved with hypothermia.19,20 External defibrillator pads and TEE are used routinely in every patient before the incision. An approximate 6- to 8-cm upper midline skin incision is made in the chest, and the upper sternum is separated in the midline with a sternal saw. The right fourth ICS is scored and the upper sternum is subsequently separated horizontally with an oscillating saw taking care to avoid the RITA. A Koros CAB sliding retractor (T Koros Surgical Instruments Corp., Moorpak, CA) is used to expose the upper mediastinum. The pericardium is opened in the usual fashion and pericardial stitches are sutured to the deep dermis, allowing for elevation of the ascending aorta (Fig. 2). The ascending aorta is then palpated for disease; an epiaortic scan is used when there is suspicion
B.S. Kim, E.G. Soltesz, and L.H. Cohn for significant aortic calcification or atheromatous disease. The distal ascending aorta is cannulated in the usual fashion. Venous drainage is provided by a 24-F right-angled cannula placed into the right atrial appendage providing this cannula will not obstruct the aortic valve exposure. Otherwise, a percutaneous femoral venous cannula (23- to 27-F BioMedicus cannula) is placed, preferably into the right femoral vein, given its straighter course into the IVC. The tip of the venous cannula is guided with the TEE to the right atrium–SVC junction. Vacuum-assisted drainage on cardiopulmonary bypass allows for use of smaller cannula sizes. After directly cross-clamping the ascending aorta, antegrade cardioplegia is given through the aortic root or directly into the coronary ostia once the aorta is opened (Fig. 3). If retrograde cardioplegia is used, the retrograde catheter is placed in the tip of the RAA and advanced into the coronary sinus under TEE guidance. If a vent is needed, the right superior pulmonary vein can be quickly exposed in a decompressed heart. Otherwise, a small 12-F wire-reinforced catheter can be directly placed into the left ventricle through the aortic annulus. Continuous carbon dioxide insufflation can be advantageous in minimizing intracardiac air and shortening de-airing at the end of the case. After completion of the aortic valve operation, with the heart still decompressed on cardiopulmonary bypass, two small flexible chest tubes are placed from the epigastrium to lay in the mediastinum, being careful to protect the heart during passage. Bipolar ventricular wires are placed on the muscular right ventricle and brought out through the right fourth ICS. Adequate de-airing is confirmed by TEE. Once separated from cardiopulmonary bypass and after hemostasis, the upper sternum is approximated with four to five stainless steel wires, and the skin is closed in the usual fashion.
Brigham Experience and Results The Brigham and Women’s Hospital as an institution has performed 901 minimally invasive AVRs between July 1996 and January 2006. The demographics of patients demonstrate that 546 patients were male (61%) with an average age of 62 (range 25 to 94) and 355 patients were female (39%) with an average age of 69 (range 24 to 95). Additionally, 157 patients (17%) were over 80 years old. One hundred seventeen patients were reoperations. Seven hundred sixteen patients (79%) were in NYHA Class 2 or 3 with 40 patients in NYHA Class 4. The etiologies of aortic valve disease were calcific in 591 (66%), bicuspid in 94 (10.4%), rheumatic in 32 (3.5%), annuloectasia in 33 (3.6%), endocarditis in 31 (3.4%), and structural valve degeneration and myxomatous in the remainder. All types of aortic valves were implanted, including Carpentier Edwards pericardial valves in 567 (63%), mechanical valves in 194 (22%), homografts in 72 (8%), Mosaic/Hancock valves in 45 (5%), and aortic valve repairs and stentless valves in the remainder. Our earlier experience with different aortic valve exposures included 37 cases using the right parasternal incision and 3 using the right thoracotomy incision. We now have experience with 861 cases of minimally invasive AVR performed through an upper hemisternotomy approach. Intraoperatively, the aver-
Minimally invasive approaches to aortic valve surgery
151
Figure 2 (A) Six centimeter upper hemisternotomy incision. (B) Operative field with Koros retractor, ascending aorta with percutaneous femoral venous cannula, and cross-clamp with antegrade cardioplegia catheter. (C) After aortotomy, unobstructed view of aortic valve with exposed coronary ostia.
age aortic cross-clamp and cardiopulmonary bypass times for first time AVRs were 79 and 109 minutes, respectively. Of note, these times are significantly shorter than those we reported in 2001 with our first 365 minimally invasive-AVR patients (91 and 124 minutes).21 Early deaths occurred in 18 patients (1.9%). In 24 patients (2.7%), a reoperation for bleed was required. The stroke rate was 2.4% (22 patients), while the incidence of new postoperative atrial fibrillation was 22% (201 patients). In 40 patients (4.4%), a pacemaker and/or automatic implantable cardiac defibrillator was required. Wound infection occurred in 12 patients (1.3%). There were four patients (0.4%) who had early (⬍3 months) reoperative AVRs for endocarditis. Other morbidities included myocardial infarction and multisystem organ failure. Of all first-time minimally invasive AVRs, 53% of patients did not receive any blood products. Of those who did require postoperative transfusion, the average red blood cell transfusion was 2 units. For reoperative AVRs, the average red blood cell
transfusion was 3 units. The average length of stay for patients without complications was 6.9 days (range 3 to 52); most patients staying more than 20 days were in fact awaiting rehabilitation bed placement. Most patients (71%) went home; only 26% of patients required a stay in rehabilitation. Of the 157 octogenarians, 62% of patients went to a rehabilitation facility. According to the Society of Thoracic Surgery National Database Fall 2005 Executive Summary, the unadjusted operative mortality for conventional aortic valve replacements is approximately 3.5%.22 Our 30-day operative mortality is 1.9%, better than the national average. Furthermore, the majority of our first-time minimally invasive AVR patients did not receive any blood products. Most of the patients went home instead of to a rehabilitation facility after an accepted length of stay of 6 to 7 days. Even the octogenarians generally did well after minimally invasive AVR with about 40% of them returning home.
B.S. Kim, E.G. Soltesz, and L.H. Cohn
152
Figure 3 (A) Upper hemisternotomy incision for re-operative aortic valve surgery. (B) Right axillary artery cannulation with side graft arterial cannulation. (C) Limited mediastinal dissection with unobstructed view of the aortic valve and untouched previous coronary grafts.
Discussion Our updated series underscores the safety of the upper hemisternotomy approach to aortic valve replacement. The morbidity and mortality are comparable to conventional aortic valve replacements performed through a full median sternotomy. The upper hemisternotomy approach allows adequate exposure to perform aortic valve replacements for all etiologies and all types of implantation, with reasonable cross-clamp and bypass times, and with minimal alteration in cannulation techniques. Therefore, the learning curve for the surgeon is short. Studies have shown that patients benefit from these minimally invasive approaches; they are not merely pleasing to the public. There is less pain, less use of blood products, greater percentage of discharges home, and quicker return to preoperative function levels.
This technique of minimally invasive aortic valve exposure has also been extrapolated to surgery for more complex aortic valve pathology including valve conduits and root replacements. Furthermore, the upper hemi-resternotomy has been used in 117 reoperative cases at our institution. In 2000, Byrne and coworkers published an early series of these patients and highlighted their excellent results.20 Preoperative computed tomographic angiograms with 3D reconstruction have helped in surgical planning for patients with patent coronary grafts. Right axillary or femoral artery and percutaneous femoral venous cannulation are performed before upper hemi-resternotomy. Mediastinal dissection is kept to a minimum, enough to expose the aortic valve. Since there is no dissection and attempted exposure of previous coronary grafts, including LIMA grafts, reoperative aortic valve surgery through this approach is safer.19 Since the LIMA pedicle is
Minimally invasive approaches to aortic valve surgery not dissected in such situations, myocardial protection must be adequately achieved with moderate hypothermia (20 to 22°C). Minimally invasive aortic valve surgery has undergone a decade of refinements since its inception in 1996 and has now become a respected and safe option for most all patients requiring aortic valve or ascending aortic surgery. The familiar, easily adaptable upper hemi-sternotomy approach should be considered an option in all patients presenting for isolated aortic valve surgery, first time and reoperative. Even patients with concomitant coronary artery disease should not be immediately dismissed from the minimally invasive AVR option, as hybrid coronary stenting followed by minimally invasive AVR can be accomplished.18 Particularly, elderly patients with significant comorbidities, who cannot tolerate a prolonged combined valve/CABG procedure, may in fact do better with the hybrid approach. Similarly, patients with previous chest irradiation might do better with a hemisternotomy and limited mediastinal dissection. The upper hemisternotomy for aortic valve surgery has become the standard approach at the Brigham and Women’s Hospital. Direct in-line vision over the aorta and the aortic valve and expeditious conversion to full sternotomy if necessary make this approach more easily accepted by surgeons who are learning to perform minimally invasive aortic valve surgery.23 We predict that this technique will become the standard of care for patients undergoing aortic valve surgery in whom there is no concomitant coronary artery disease.
References 1. Cosgrove DM, Sabik JF: Minimally invasive approach for aortic valve operations. Ann Thorac Surg 62:596-597, 1996 2. Cohn LC, Adams DH, Couper GS, et al: Minimally invasive cardiac valve surgery improves patient satisfaction while reducing costs of cardiac valve replacement and repair. Ann Surg 226(4):421-428, 1997 3. Cosgrove DM, Sabik JF, Navia JL: Minimally invasive valve operations. Ann Thorac Surg 65:1535-1538, 1998 4. Gillinov AM, Banbury MK, Cosgrove DM: Hemisternotomy approach for aortic and mitral valve surgery. J Card Surg 15:15-20, 2000 5. Bridgewater B, Steyn RS, Ray S, et al: Minimally invasive aortic valve replacement through a transverse sternotomy: a word of caution. Heart 79:605-607, 1998
153 6. Benetti FJ, Mariani MA, Rizzardi JL, et al: Minimally invasive aortic valve replacement. J Thorac Cardiovasc Surg 113:806-807, 1997 7. Kort S, Applebaum RM, Grossi EA, et al: Minimally invasive aortic valve replacement. Echocardiographic and clinical results: Am Heart J 142: 476-481, 2001 8. Gundry SR, Shattuck OH, Razzouk AJ, et al: Facile minimally invasive cardiac surgery via ministernotomy. Ann Thorac Surg 65:1100-1104, 1998 9. Svensson LG, D=Agostino RS: “J” incision minimal-access valve operations. Ann Thorac Surg 66(3):1110-1112, 1998 10. Tam RK, Almeida AA: Minimally invasive aortic valve replacement via hemisternotomy: a preliminary report. Eur J Cardiothorac Surg 14(1): S134-S137, 1998 11. Sharony R, Grossi EA, Saunders PC, et al: Propensity score analysis of a six-year experience with minimally invasive isolated aortic valve replacement. J Heart Valve Dis 13:887-893, 2004 12. Dogan S, Dzemali O, Wimmer-Greinecker G, et al: Minimally invasive versus conventional aortic valve replacement: a prospective randomized trial. J Heart Valve Dis 12:76-80, 2003 13. Bonacchi M, Prifti E, Giunit G, et al: Does ministernotomy improve postoperative outcome in aortic valve operation? A prospective randomized study. Ann Thorac Surg 73:460-465, 2002 14. Sharony R, Grossi EA, Saunders PC, et al: Minimally invasive aortic valve surgery in the elderly: a case-control study. Circulation 108:II43II47, 2003 (suppl 1) 15. Cohn LH: Minimally invasive valve surgery. J Card Surg 16(3):260265, 2001 16. Leshnower BG, Trace CS, Boova RS: Port-access-assisted aortic valve replacement: a comparison of minimally invasive and conventional techniques. Heart Surg Forum 9(2):E560-E564, 2006 17. Mihaljevic T, Cohn LH, Unic D, et al: One thousand minimally invasive valve operations: early and late results. Ann Surg 240(3):529-534, 2004 18. Byrne JG, Leacche M, Unic D, et al: Staged initial percutaneous coronary intervention followed by valve surgery (“hybrid approach”) for patients with complex coronary and valve disease. J Am Coll Cardiol 45:14-18, 2005 19. Byrne JG, Karavas AN, Filsoufi F, et al: Aortic valve surgery after previous coronary artery bypass grafting with functioning internal mammary grafts. Ann Thorac Surg 73:779-784, 2002 20. Byrne JG, Karavas AN, Adams DH, et al: Partial upper re-sternotomy for aortic valve replacement or re-replacement after previous cardiac surgery. Eur J Cardiothorac Surg 18(3):282-286, 2000 21. Sepic J, Cohn LH: Minimally Invasive Aortic Valve Surgery in Minimally Invasive Cardiac Surgery (ed 2). Totowa, NJ, Humana Press Inc., 2004, pp 293-307 22. STS National Database. Fall 2005 Executive Summary. Duke Clinical Research Institute, 2005 23. Chitwood WR: Minimally invasive aortic valve surgery: what is “port access”? Am Heart J 142:391-392, 2001 (editorial)