Stent valve implantation in conventional redo aortic valve surgery to prevent patient prosthesis mismatch

Transcription

1 Interactive CardioVascular and Thoracic Surgery 24 (2017) doi: /icvts/ivw397 Advance Access publication 31 December 2016 ADULT CARDIAC Cite this article as: Ferrari E, Franciosi G, Clivio S, Faletra F, Moccetti M, Moccetti T et al. Stent valve implantation in conventional redo aortic valve surgery to prevent patient prosthesis mismatch. Interact CardioVasc Thorac Surg 2017;24: Stent valve implantation in conventional redo aortic valve surgery to prevent patient prosthesis mismatch Enrico Ferrari a, *, Giorgio Franciosi a, Sara Clivio b, Francesco Faletra c, Marco Moccetti c, Tiziano Moccetti c, Giovanni Pedrazzini c and Stefanos Demertzis a a b c Cardiac Surgery Unit, Cardiocentro Ticino Foundation, Lugano, Switzerland Cardiac Anaesthesia Unit, Cardiocentro Ticino Foundation, Lugano, Switzerland Cardiology Department, Cardiocentro Ticino Foundation, Lugano, Switzerland * Corresponding author. Cardiocentro Ticino Foundation, Via Tesserete 48, 6900 Lugano, Switzerland. Tel: ; fax: ; enricoferrari@bluewin.ch (E. Ferrari). Received 22 August 2016; received in revised form 27 October 2016; accepted 1 November 2016 Abstract OBJECTIVES: The goal was to show the technical details, feasibility and clinical results of balloon-expandable stent valve implantation in the aortic position during conventional redo open-heart surgery in selected obese patients with a small aortic prosthesis and severe patient prosthesis mismatch. METHODS: Two symptomatic overweight patients (body mass index of 31 and 38), each with a small aortic prosthesis (a 4-year-old, 21- mm Hancock II biological valve and a 29-year-old, 23-mm Duromedic mechanical valve), increased transvalvular gradients (59/31 and 74/ 44 mmhg) and a reduced indexed effective orifice area (0.50 and 0.43 cm 2 /m 2 ) underwent implantation of two 26-mm balloonexpandable Sapien 3 valves during standard on-pump redo valve surgery. RESULTS: Using full re-sternotomy, cardiopulmonary bypass and cardioplegic arrest, the two balloon-expandable stent valves were implanted under direct view using a standard aortotomy, after prosthesis removal and without annulus enlargement. Aortic cross-clamp times were 162 and 126 min; cardiopulmonary bypass times were 178 and 180 min; total surgical times were 360 and 318 min. At discharge, echocardiograms showed transvalvular peak and mean gradients of 13/9 and 23/13 mmhg and indexed effective orifice areas of 0.64 and 1.08 cm 2 /m 2. The 3-month echocardiographic follow-up showed transvalvular peak and mean gradients of 18/9 and 19/ 11 mmhg and indexed effective orifice areas of 0.78 cm 2 /m 2 and 0.84 cm 2 /m 2, with improved symptoms (New York Heart Association class 1). CONCLUSIONS: Implantation of a balloon-expandable stent valve during redo aortic valve surgery is feasible in selected cases and prevents patient prosthesis mismatch in obese patients without need for aortic annulus enlargement. Moreover, in the case of stent valve degeneration, this approach permits additional valve-in-valve procedures with large stent valves and prevents re-redo surgery. Keywords: Aortic valve replacement Transcatheter aortic valve replacement Redo cardiac surgery Patient prosthesis mismatch INTRODUCTION Patient prosthesis mismatch (PPM) (defined as moderate for an indexed aortic area of cm 2/ m 2 and severe when the area is <0.65 cm 2 /m 2 ) occurs in symptomatic patients with biological or mechanical aortic valve prostheses who have small functional indexed effective orifice areas (ieoa) and high transvalvular gradients [1 2]. Small ieoa can be due either to a malfunctioning prosthesis or pannus (mechanical valves) or to degeneration (biological valves) or can be determined by a significant increase in the patient s weight after surgical aortic valve implantation of a small valve in a narrow aortic annulus. Some patients can have both conditions. The treatment of choice is valve replacement during on-pump redo open-heart surgery [3 5]. However, when the patient is overweight or obese, it is necessary to implant a large valve to minimize the risk of residual PPM. In redo valve surgery, the technical strategies used to enlarge the aortic annulus (to position a larger valve) present an important challenge for the surgeon and can significantly increase the bypass and the ischemic times, which increases surgical risk [6, 7]. The choice of the new prosthesis is also debatable because a mechanical valve still carries the risk of re-redo surgery and because the valve-in-valve technique is not applicable for a mechanical prosthesis. Therefore, in selected overweight or obese patients undergoing redo aortic valve surgery for PPM, the use of VC The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 320 E. Ferrari et al. / Interactive CardioVascular and Thoracic Surgery Table 1: Patient characteristics, valve haemodynamics and surgical outcomes Patient 1 Patient 2 Age, years Gender Male Female Height/weight/body surface area, cm/kg/m 2 182/101/ /105/2.1 Body mass index Logistic EuroSCORE II, % Concomitant coronary disease Yes No Previous stroke No No Hypertension Yes No Pacemaker No No Atrial fibrillation Yes (paroxysmal) No Diabetes, type II Yes No Date of previous aortic valve surgery Type of prosthesis/size 4-year-old biological Medtronic Hancock II, 21 mm 29-year-old mechanical Baxter Duromedic, 23 mm Functional preoperative EOA, cm Preoperative ieoa, cm 2 /m Preoperative peak/mean gradients, mmhg 59/31 74/44 Preoperative LVEF, % Sapien 3 valve size, mm Concomitant procedure Mitral valve replacement No Functional postoperative EOA at discharge, cm Postoperative ieoa at discharge, cm 2 /m Postoperative peak/mean gradients, mmhg 13/9 23/13 Postoperative LVEF, % Paravalvular leak None Trace Cardiopulmonary bypass/aortic cross-clamp time, min 178/ /126 Complications Respiratory failure-prolonged intubation None Postoperative electrocardiogram SR SR Intensive + intermediate care unit stay, days 26 4 Hospital stay, days 30 8 ieoa: indexed effective orifice area; LVEF: left ventricular ejection fraction; SR: sinus rhythm. biological valves without annulus enlargement and allowing for additional transcatheter valve-in-valve placement in case of early degeneration seems the best surgical option with the lower surgical impact for the patient [8 10]. But a large valve is mandatory [11]. To the best of our knowledge, the use of balloon-expandable Sapien 3 transcatheter valves (Edwards Life Sciences, Irvine, CA) in open-heart redo aortic valve surgery has never been suggested before, and it carries the advantage of not requiring annulus enlargement despite the implantation of a large valve, thereby minimizing the risk of PPM in future valve-in-valve procedures. We report our experience with 2 consecutive symptomatic obese patients with small-aortic valve prostheses and PPM treated with surgical implantation of two 26-mm Sapien 3 valves in a standard on-pump redo setting. METHODS Patient selection Two overweight patients (101 and 105 kg, respectively, with a body mass index of 31 and 38, respectively), aged 73 and 55 years, already operated for aortic valve replacement with biological and mechanical prostheses (a 21-mm Medtronic Hancock II porcine valve implanted in 2012, at 69 years of age) and a 23- mm Baxter Duromedic mechanical valve implanted in 1987 (at 26 years of age) presented to our hospital with continuously increasing dyspnoea (New York Hospital Association [NYHA] class III). Echocardiograms showed good ventricular function with a normal valve prosthesis (except for a subvalvular pannus in the Duromedic valve, which slightly reduced the effective orifice area) and new moderate-to-severe mitral regurgitation in 1 patient. However, small functional ieoa (0.50 and 0.43 cm 2 / m 2 ) and high transvalvular gradients were measured across the two aortic valve prostheses, indicating 2 cases of PPM in obese patients (body mass index of 31 and 38, respectively, with considerable weight increase during the past few years). Patients baseline characteristics, valve haemodynamics and clinical outcomes are listed in Table 1. During the heart team meeting, standard redo open-heart surgery for aortic valve replacement with large valves was considered the best therapeutic option in both cases. However, we discussed the option of implanting an openheart Sapien 3 transcatheter heart valve to guarantee at first a larger valve size without concomitant annulus enlargement and then the use of valve-in-valve procedures with large transcatheter valves in case of premature Sapien 3 valve degeneration. The alternative option of a sutureless Perseval (LivaNova, London, UK) valve was discussed, but, at that moment, this valve was still not available in our unit. A mitral valve replacement with a bioprosthesis was also planned for 1 patient before the aortic valve replacement with the stent valve. We decided not to attempt to repair the mitral valve to avoid the risk of a residual intraoperative leak requiring mitral valve re-exploration with a Sapien 3 valve in place in the aortic position. Both

3 E. Ferrari et al. / Interactive CardioVascular and Thoracic Surgery 321 patients agreed to have the procedure and signed informed consent forms. Surgical strategy and technique Re-sternotomy and central cannulation for cardiopulmonary bypass were performed in the operating theatre. After aortic crossclamping and injection of antegrade del Nido cardioplegia solution, a transverse aortotomy was performed 2 cm above the right coronary ostia. The two previously implanted valves were carefully explanted without damaging the aortic valve annulus (Fig. 1). Then, progressive sizes of Hegar dilators (equivalent to millimetres, in contrast to prosthetic valve sizers whose numbers are not always equivalent to millimetres and can introduce errors) were used to measure the anatomic diameters of the aortic annulus: 23 and 25 mm, respectively. Therefore, we opted to implant 26-mm Sapien 3 stent valves in both cases. The delivery system with the crimped valve was placed into the aortic root in the transaortic-transcatheter orientation, meaning with the valve inflow close to the yellow nose cone of the delivery system (Fig. 2A). The middle point of the Sapien 3 Figure 1: The 23-mm Duromedic mechanical valve with a subvalvular pannus that was removed from the patient. was placed at the level of the aortic annulus, and we tried to maintain the anatomic alignment of the three valve cusps with the three Valsalva sinuses. Under direct vision, the balloon was slowly inflated to maintain the valve in the right position and coaxial with the aorta (Fig. 2B). After 5 s of full inflation (Fig. 2C), the delivery system was removed and the valve positioning checked: In both cases, the stent valve was placed appropriately without covering the coronary ostia (Fig. 2D). To secure the valve to the aorta and prevent migration (the Sapien valve is designed for placement in heavily calcified aortic annuli with strong radial forces), three 4-0 polypropylene stitches (U-fashion) with external pledgets were placed across the aortic wall and across the metal frames at the top of each Sapien 3 commissure (Fig. 3). In 1 patient, the planned mitral valve replacement with a stented bioprosthesis was performed before the Sapien 3 valve was positioned. Then, the aortotomies were sutured and the cross-clamps were removed after careful deairing. In the patient with a measured 25-mm annulus diameter, the presence of a mild anterior paravalvular leak required a new short period of aortic cross-clamping and cardioplegic arrest to reopen the aorta and perform a Sapien 3 reballooning with one more millilitre of liquid in the balloon catheter. The intraoperative transoesophageal echocardiographic follow-up showed good gradients and preserved left ventricular function without paravalvular leaks (Fig. 4). Patients were weaned from cardiopulmonary bypass in sinus rhythm and without complications. The two procedures were terminated in the standard way with patients transferred to the intensive care unit in stable condition. RESULTS The procedure was completed successfully in the 2 patients without complications. Because the surgeon had a direct view of the standard surgical field during open-heart surgery, contrast injections and fluoroscopic guidance were not necessary for valve placement, positioning and deployment. Total procedural time Figure 2: (A) The Sapien 3 valve (Edwards Lifesciences, Irvine, CA) is prepared and crimped in a transaortic-transcatheter orientation with the inflow close to the tip of the balloon catheter. (B) After the previously implanted valve is removed, the Sapien 3 is positioned in the aortic annulus such that the commissure corresponds with the anatomy of the native valve. (C) The balloon is slowly inflated to maintain the position. (D) The final result with the Sapien 3 valve in place.

4 322 E. Ferrari et al. / Interactive CardioVascular and Thoracic Surgery Figure 3: The Sapien 3 valve with the schematic view of the aortic valve annulus (red lines) and the location of the three 4-0 polypropylene U-shaped stitches (blue lines) placed across the metal frame at the top of the three valve commissures to fix the valve to the aortic wall. Figure 4: Postoperative echocardiogram showing the Sapien 3. was 360 and 318 min, respectively. The valve positioning and delivery in the landing zone were not demanding and were based on our previous experience positioning and implanting the Sapien valve during transaortic transcatheter aortic valve replacements [12]. Postprocedural transoesophageal echocardiograms showed perfect placement of the Sapien 3 valves with no signs of PPM or paravalvular leak. One patient was rapidly extubated and remained in the intensive and intermediate care units, respectively, for a total of 4 days. The second patient experienced postoperative respiratory failure complicated by pneumonia requiring a long period of intubation (10 days). Then, after a long stay in the intensive and intermediate care units, he was transferred to the ward on postoperative day 26 in good clinical and haemodynamic condition. At discharge, the transthoracic echocardiogram showed good valve functioning with transvalvular peak and mean gradients of 13/9 and 23/13 mmhg, a functional EOA of 1.46 and 2.24 cm 2, an ieoa of 0.64 and 1.08 cm 2 /m 2, left ventricular ejection fractions of 60% and 55%, absence of paravalvular leak in 1 patient and a trace in the other one (Fig. 5). Both patients were transferred to the rehabilitation unit in good haemodynamic condition and with improved clinical symptoms. At the 3-month follow-up, the patients underwent both clinical and echocardiographic examinations: both were in stable Figure 5: (A) Aortic transvalvular mean and peak gradients before and after the procedure. (B) ieoa before and after the procedure. condition with NYHA functional class I. The prosthesis showed good functioning with transvalvular peak and mean gradients of 18/9 and 19/11 mmhg, functional EOA of 1.7 and 1.7 cm 2, ieoa of 0.78 and 0.84 cm 2 /m 2, absence of paravalvular leak in 1 patient and a trace in the other one, and good left ventricle ejection fractions (55% and 56%, respectively) (Fig. 5). At the 6-month follow-up, both patients were still in stable clinical condition with NYHA functional class I. DISCUSSION Aortic PPM is determined by several factors and can be due to the association of a small prosthesis implanted in an overweight or obese patient and requires redo open-heart aortic surgery to replace the small valve prosthesis with a larger one. This approach has some limitations and important considerations are required: (i) to prevent residual PPM, the new valve has to be larger; (ii) to implant a large prosthesis, a surgical technique for annulus enlargement is needed, which increases the surgical time and, sometimes, involves surgical risk associated with redo surgery; (iii) to prevent further re-redo surgery (in the case of valve failure), biological valves should be given priority because valvein-valve techniques are easy to perform and they guarantee good haemodynamic results [9, 13].

5 E. Ferrari et al. / Interactive CardioVascular and Thoracic Surgery 323 In addition, the use of stent valves in open-heart redo surgery can be considered by the modern heart team because it involves implantation of a large valve (compared to similarly sized standard stented bioprostheses) without enlarging the surgical annulus and offers the advantage of the valve-in-valve technique instead of re-redo valve surgery in cases of premature failure. The procedure is not technically demanding, but it should be performed by experienced surgeons who are members of the heart team and are familiar with the transcatheter program, because they are familiar with the newest generation of transcatheter valves and delivery systems and with the necessary orientation and positioning of the valve during the procedure. In our limited experience, the chosen balloon-expandable stent valve has the advantage of a short, simple delivery system, low valve profile and a symmetrical expansion during the ballooning that guarantees precise coaxial alignment, which happens during transaortic-transcatheter valve replacements [12]. Moreover, Sapien-in-Sapien strategies are already well defined with proven good results. Other available self-expandable stent valves were not considered for these redo cases because the height of the valve would have interfered with the aortotomy suture line. In cases where the valve-in-valve options are not suitable, a standard surgical bioprosthesis has the advantages of valve-invalve treatments but a 25-mm stented bioprosthesis, for example, has a small inner diameter compared to the 26-mm Sapien valve; therefore, it can only accept a 23-mm stent valve (with risk of new PPM), whereas the Sapien-in-Sapien can be performed with same-sized valves, at least during the first attempt [9 11, 13]. Another option that should always be considered in redo aortic valve surgery is the sutureless Perceval valve (LivaNova, London, UK). This valve is a self-expandable bioprosthesis available in four sizes that can easily be placed in the aortic position during redo aortic valve surgery after the removal of the failed prosthesis [14]. The technology used is similar to that used with Sapien valves (self-expandable nitinol stents) because it is an intra-annular implantation and uses radial forces to expand the annulus. However, this valve is not always available in all cardiac surgery units, and we should be afraid of performing further valve-in-valve procedures and of the risk of residual PPM because, so far, only 1 case has been reported of a valve-in-valve Perceval bioprosthesis. In contrast, the Sapien-in-Sapien option, as already mentioned, is an established technique that allows for same stent valve size implantation, at least during the first attempt [9 11, 13, 15]. The rapid-deployment, Edwards Intuity valve (Edwards Lifesciences, Irvine, CA) system is another valuable option but it is supra-annular and is more comparable to a standard stented aortic bioprosthesis. Moreover, implanting this valve after annulus enlargement is not recommended by the industry. Alternatively, surgical valves that can also be placed in small aortic annuli are the stentless valves (i.e. the Freedom Solo; LivaNova, London, UK): they feature larger orifice areas compared to a comparably sized stented bioprosthesis but, in redo aortic surgery, a Sapien valve can be placed much more easily and faster. In fact, after the prosthesis is removed, the annulus tissue can be partially destroyed, is not homogeneous and carries the risk of a bad haemodynamic result after implantation of the stentless valve. Moreover, not all hospitals are familiar with stentless valves. At the time, we noted only the trace of a paravalvular leak in 1 patient who received the Sapien valve implant in non-calcified annuli, with no haemodynamic consequences or haemolysis. Patients will be followed up with transthoracic echocardiograms every year to identify new or worsening leaks. In conclusion, this report shows that selected obese patients with aortic PPM can be safely treated with a biological large, low-profile transcatheter valve implant during standard openheart on-pump redo surgery with the 2-fold advantage of no need for aortic annulus enlargement and an efficient valve-invalve therapeutic option if needed. Conflict of interest: Enrico Ferrari is a consultant for Edwards Lifesciences. REFERENCES [1] Dahou A, Mahjoub H, Pibarot P. Prosthesis-patient mismatch after aortic valve replacement. Curr Treat Options Cardiovasc Med 2016;18:67. [2] Pibarot P and Dumesnil JG. Prosthesis-patient mismatch: definition, clinical impact, and prevention. Heart 2006;92: [3] Walther T, Rastan A, Falk V, Lehmann S, Garbade J, Funkat AK et al. Patient prosthesis mismatch affects short- and long-term outcomes after aortic valve replacement. Eur J Cardiothorac Surg 2006;30: [4] Vahanian A, Alfieri O, Andreotti F, Antunes MJ, Baron-Esquivias G, Baumgartner H, Borger MA, Carrel TP et al. Guidelines on the management of valvular heart disease (version 2012): the Joint Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS). Eur J Cardiothorac Surg 2012;42:S1 44. [5] Mohty D, Dumesnil JG, Echahidi N, Mathieu P, Dagenais F, Voisine P, Pibarot P. Impact of prosthesis-patient mismatch on long-term survival after aortic valve replacement: influence of age, obesity, and left ventricular function. J Am Coll Cardiol 2009;53: [6] Castro LJ, Arcidi JM Jr, Fisher AL, Gaudiani VA. Routine enlargement of the small aortic root: a preventive strategy to minimize mismatch. Ann Thorac Surg 2002;74:31 6. [7] Pibarot P, Clavel MA, Dahou A. Patient and procedure selection for the prevention of prosthesis-patient mismatch following aortic valve replacement. EuroIntervention 2015;11 Suppl W:W doi: / EIJV11SWA31. [8] Vahanian A, Alfieri OR, Al-Attar N, Antunes MJ, Bax J, Cormier B et al. Transcatheter valve implantation for patients with aortic stenosis: a position statement from the European Association of Cardio-Thoracic Surgery (EACTS) and the European Society of Cardiology (ESC), in collaboration with the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur J Cardiothorac Surg 2008;34:1 8. [9] Dvir D, Webb JG, Bleiziffer S, Pasic M, Waksman R, Kodali S et al. Transcatheter aortic valve implantation in failed bioprosthetic surgical valves. JAMA 2014;312: [10] Ferrari E, Marcucci C, Sulzer C, von Segesser LK. Which available transapical transcatheter valve fits into degenerated aortic bioprostheses? Interact CardioVasc Thorac Surg 2010;11:83 5. [11] Ferrari E. Transapical aortic 0 valve-in-valve 0 procedure for degenerated stented bioprosthesis. Eur J Cardiothorac Surg 2012;41: [12] Ferrari E, Muller O, Demertzis S, Moccetti M, Moccetti T, Pedrazzini G, Eeckhout E. Transaortic transcatheter aortic valve replacement through a right mini-thoracotomy with balloon-expandable Sapien 3 TM valve. Multimed Man Cardiothorac Surg doi: /mmcts/mmw011. [13] Phan K, Zhao DF, Wang N, Huo YR, Di Eusanio M, Yan TD. Transcatheter valve-in-valve implantation versus reoperative conventional aortic valve replacement: a systematic review. J Thorac Dis 2016;8:E [14] Santarpino G, Pfeiffer S, Concistrè G, Fischlein T. REDO aortic valve replacement: the sutureless approach. J Heart Valve Dis 2013;22: [15] Durand E, Tron C, Eltchaninoff H. Emergency transcatheter aortic valve implantation for acute and early failure of sutureless perceval aortic valve. Can J Cardiol 2015;31:1204.e13 15.