In the past two decades the development of valvesparing

Transcription

1 Valve-Sparing Aortic Root Reconstruction Using In Situ Three-Dimensional Measurements Andras C. Kollar, MD, PhD, Scott D. Lick, MD, and Vincent R. Conti, MD Division of Cardiothoracic Surgery, Department of Surgery, University of Texas Medical Branch, Galveston, Texas Background. The truncated cone overall geometry of the native aortic root, an important factor in maintaining valvular competence, is significantly altered in cases of root aneurysms. We hypothesized that an early trial restoration of valve competence within the native aortic root followed by in situ three-dimensional measurements may lead to a more predictable functional reconstruction. Methods. The operation started with downsizing annuloplasty followed by sinotubular junction plication until full valve competence was observed and tested with the saline squirt test. Subsequent measurements (basal ring and sinotubular junction size, the depth of each sinus of Valsalva) formed the basis of graft sizing and tailoring. Reconstruction was completed with a new proximal suture line technique combining David subannular pledgeted fixation with Yacoub remodeling. Results. Ten patients were operated on during a 3-year period. Intraoperative (nonpressurized) competence by open testing translated into good postoperative valve function seen on transesophageal echocardiography. In situ measurements were done in the last 7 patients, and in 5 of them the restored root geometry was of a reverse cone (sinotubular junction 2 to 4 mm larger than basal ring size). Conclusions. Rebuilding the aortic root based on in situ measurements with a fully competent aortic valve is a conceptually new surgical approach. Our observations suggest that postoperative valve competence, particularly with elongated valve leaflets, may not depend on the normal truncated cone geometry. (Ann Thorac Surg 2009;87: ) 2009 by The Society of Thoracic Surgeons In the past two decades the development of valvesparing operations [1 3] together with recent experimental observations [4, 5] have fundamentally changed our understanding of the aortic root. According to the modern concept advocated by Robicsek and Thubrikar [4], the aortic root should be viewed as a single functional unit containing three semilunar leaflets suspended alongside a coronet-shape annulus and three individually ballooning Valsalva sinuses [6 9]. In the normal aortic root the inflow orifice (basal ring) is slightly larger than the outflow orifice (sinotubular junction [STJ]), creating an overall truncated cone geometry [6, 7, 9, 10], which is considered the ideal three-dimensional arrangement for optimal semilunar valve function [7, 11]. In cases of aortic root aneurysm these geometric relationships are significantly distorted, but valvular regurgitation is not a consistent finding even with larger aneurysms [12]. The presumed remodeling mechanism by which nature adapts to pathologic change seems to be an individualized process, and this added variability between geometric structure and function makes valve-sparing root reconstruction difficult. At present there are two separate valve-sparing concepts: root remodeling developed by Yacoub [1, 2], and reimplantation designed by David and Feindel [3]. However, there are no universally agreed-on Accepted for publication March 17, Address correspondence to Dr Kollar, University of Texas Medical Branch, Department of Surgery, 301 University Blvd, John Sealy Annex 6.120, Galveston, TX ; ankollar@utmb.edu. formulas to determine appropriate restored root geometry, and the recommended graft sizing methods are often estimates based on individual experience [13]. Therefore, the overall success of valve-sparing surgery reflects more the artful skills of the surgeon than the application of established geometric criteria. At the University of Texas Medical Branch in Galveston we have started performing valve-sparing operations emphasizing the concept of a single functional unit. We hypothesized that an early trial restoration of valve competence within the native aortic root followed by in situ three-dimensional measurements may lead to a more predictable functional reconstruction. The findings of this series represent our evolving strategy for rebuilding the functional aortic root unit. Patients and Methods During a 3-year period, 10 male patients (age, 25 to 69 years) presented with aortic root aneurysm containing anatomically intact valve leaflets with varying degrees of incompetence. All 10 patients underwent valve-sparing aortic root replacement with a previously described modified remodeling operation that includes a novel annulus stabilization method as part of the proximal suture line [14]. Retrospective data collection and prospective clinical and echocardiographic follow up were approved by the University of Texas Medical Branch Institutional Review Board by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 1796 KOLLAR ET AL Ann Thorac Surg VALVE SPARING WITH IN SITU AORTIC ROOT MEASUREMENTS 2009;87: Fig 1. Three-dimensional downsizing with the native aortic root intact. Sets of two vertically placed externally pledgeted subcommissural annuloplasty sutures and three mid-sinus level temporary plication sutures are placed 120 degrees apart at the sinotubular junction (see details in text). The details of our surgical technique were published previously [14]. Most importantly we adopted the concept of a single functional unit to the entire aortic reconstruction procedure by repeatedly checking for valvular competence. Our technique is fundamentally different from previously described methods as it starts with a trial restoration of valve competence within the native root. After the aorta is transected above the STJ, we perform a multilevel downsizing procedure before cutting out the Valsalva sinuses (Fig 1). The downsizing maneuvers include externally pledgeted subcommissural annuloplasty sutures [15] that vertically narrow the interleaflet triangles, and three temporary plication sutures 120 degrees apart at the STJ until a fully competent valve is observed and tested with the saline squirt test. Any additional leaflet length adjustment or commissural height bolstering is completed at this point, still within the native aortic root. Our hypothesis was that by measuring the various dimensions of this functionally restored native aortic root unit in situ, then selecting and tailoring the prosthetic graft according to these measurements, the surgical reconstruction would be more accurate and the clinical outcome more predictable. The in situ measurements are (1) basal ring sizing using Hegar dilators before and after root downsizing; (2) STJ sizing with standard valve sizers while the valve is kept competent; (3) marking the location of each commissure on the selected valve sizer s circumference; and (4) individually measuring the depth of all three sinuses of Valsalva. Our concept is that these data accurately describe the three-dimensional geometry (and asymmetry) of the functional aortic root unit in the actual surgical case, and to achieve postoperative valve competence the neoaortic root should be rebuilt to this same geometry. After cutting out the sinuses our technique is essentially a modified remodeling procedure. Graft size selection (Hemashield graft; Boston Scientific, Natick, MA) is based on the measured STJ size (step 2), and any valve or sinus asymmetry is simply matched by tailoring the individual tongues according to the in situ measurements (steps 3 and 4). Our new proximal suture line technique is based on the concept proposed by David and Feindel [3] that the entire aortic annulus should be surgically stabilized to prevent further dilatation, and we have successfully modified their subannular pledgeted suture line technique so that it could be applied to the Yacoub remodeling operation [14]. The commissural aspects of the scalloped annulus are sandwiched between the externally placed subcommissural pledgets, then sutured to the scalloped graft material; the three horizontal segments of the annulus between the commissures are sandwiched between subannular pledgets (from below) and the graft material (above the annulus). Detailed description of individual suture placement was published previously [14]. The final result is that the entire annulus is sandwiched between a layer of graft material and pledgets, but the stabilization line follows the three-dimensional coronet (Fig 2). After repeat competency testing, the operation is completed as a standard root replacement with coronary reimplantations and distal aortic suture line. Fig 2. Integrated aortic root reconstruction. The entire aortic annulus is sandwiched between graft material and pledgets with the suture line following the three-dimensional coronet. (Reprinted with permission from Kollar [14].)

3 Ann Thorac Surg KOLLAR ET AL 2009;87: VALVE SPARING WITH IN SITU AORTIC ROOT MEASUREMENTS 1797 Table 1. Technical Details of the Performed Aortic Root Procedures and Relevant Echocardiographic Data During Follow-up Follow-up TTE AI Grading Length of Follow-up Intraoperative Postrepair AI (by TEE) STJ/Base Ratio (Geometry) Additional Procedures Restored STJ/Graft Size Annulus Downsizing (mm) Aortic Insufficiency Aortic Root Size (cm) Patient No. #1 6.0 (Marfan) 3 28 mm? Type A dissection/hemiarch Moderate Lost... # Yes (not recorded) 28 mm? CABG 1 Trace 3 y Trace/mild # From 29 to mm 1.13 None Trace 2 y Trace/mild #4 6.0 (Marfan) None/trace From 31 to mm 0.96 None Mild 9 mo 3 AI/noncoronary leaflet prolapse (AVR at 9 mo) Trace/mild AI, mild AS Lost after 11 mo Trace AI, mild AS No (mild AS) 24 mm? CABG 1, Alfieri mitral valve repair # AI/AS (bicuspid valve) #6 4.9 (Marfan) 1 2 From 27 to mm 1.0 None Trace 1 y Trace # From 27 to mm 1.08 CABG 2 Trace 1 y Trace/mild #8 4.5 (Marfan) None From 25 to mm 1.14 Two-leaflet resuspension Mild 6 mo Mild/mod # From 29 to mm 1.15 None Trace 2 mo Trace Mild 2 mo Mild # From 27 to mm 1.14 Noncoronary leaflet resuspension AI aortic regurgitation; AS aortic stenosis; AVR aortic valve replacement; CABG coronary artery bypass graft; STJ sinotubular junction; TEE transesophageal echocardiography; TTE transthoracic echocardiography. Results Of the 10 patients, 4 had clinical features of Marfan syndrome (Table 1). The very first patient was the only emergency operation in our series (type A aortic dissection with aneurysm), and 2 other Marfan patients presented with chest pain suggesting aneurysm enlargement. The last Marfan patient (number 8) was asymptomatic, and after a previous pectus repair he had been followed for several years until his aortic root diameter reached 4.5 cm. The 6 non-marfan patients presented with shortness of breath or chest pain, and in 4 of them the symptoms were related to significant aortic insufficiency (AI) with aneurysm size greater than 5.5 cm. In the remaining 2 patients the primary indication for surgery was coronary artery disease with planned elective repair of a 5.3-cm (patient 5) or 5.2-cm (patient 7) root aneurysm with mild AI. Altogether 8 patients had AI (grade 1 to 4 ), and we found no correlations between maximum root dimensions (4.5 to 7.0 cm) and the degree of AI. Nine patients had trileaflet valves and 1 had a bicuspid valve. In 2 patients we performed additional leaflet resuspension using pericardium pledgeted Trusler sutures at the commissures [16], and 4 more patients underwent additional procedures not related to aortic root reconstruction (Table 1). In this small series we had no mortality. Intraoperative postprocedure transesophageal echocardiography confirmed trace to mild AI in 9 patients and moderate AI in the first patient. This young Marfan patient presented with an acute type A aortic dissection in a 6-cm aortic root, and after an uneventful initial recovery he never came back for follow-up; later on we realized that his immigration status was undocumented. The other 9 patients had regular clinical and echocardiographic follow-up to assess aortic root dimensions and valve function. So far serial measurements have confirmed stable and unchanged aortic root dimensions in all patients, with early progression of AI observed in 1 patient (Table 1). In this young Marfan patient the regurgitation was attributable to progressive noncoronary leaflet elongation and prolapse in an asymmetrically enlarged aortic root that was not fully appreciated and not properly addressed at the time of his initial surgery. The surgical benefits of initially leaving the natural aortic root intact, then performing the multilevel downsizing procedure as a trial restoration followed by an early valve competency testing and taking in situ measurements, were only recognized during our second elective case. Seeing a fully competent valve led us to the hypothesis that the neoaortic root should be rebuilt to this same geometry, and in 7 patients we took in situ measurements as described before (except in the patient with a bicuspid valve no basal ring measurements were taken owing to mild valvular stenosis). In 5 of the 7 patients the intraoperatively predicted ideal geometry was of a reverse cone (STJ 2 to 4 mm larger than basal ring size; see Table 1, Fig 3). In our limited experience, observed valve competence on open (nonpressurized) testing with visible good leaflet coaptation translated into

4 1798 KOLLAR ET AL Ann Thorac Surg VALVE SPARING WITH IN SITU AORTIC ROOT MEASUREMENTS 2009;87: Fig 3. Follow-up (2 months) transthoracic echocardiography study of the neoaortic root in patient number 9 (parasternal long-axis view in systole [A] and in diastole [B]). Note the elongated leaflets (arrows) in the open position (A) and showing good coaptation in the midline during diastole (B). The dashed line delineates the overall geometry of the reconstructed root (sinotubular junction greater than basal ring size). predictable good valve function on the intraoperative transesophageal echocardiography and on follow-up transthoracic echocardiography studies (Fig 3). Comment Currently there are two valve-sparing root reconstruction methods, namely reimplantation and remodeling. Both techniques aim to restore normal aortic root geometry while preserving valvular competence, but they follow different surgical strategies. In the first, described by Yacoub [1, 2] and only later designated as remodeling, the focus is on replacing the enlarged Valsalva sinuses by a manually tailored triple-tongued straight tube graft. The technique is anatomically correct in attempting to restore the natural triple-bulge pattern of the aortic root, but the supraannular graft suture technique does not address coexisting aortic annulus pathologic conditions. David and Feindel [3] recognized the importance of circumferential annulus stabilization, particularly in patients with connective tissue disorders, and designed the technique called reimplantation. In this procedure the entire aorticoventricular junction is simply pulled into a straight tube graft (no other graft was available at that time), resulting in external splinting of the aortic root. This maneuver provided excellent long-term root stability, but the tubular graft does not have natural sinus bulges. The functional importance of natural Valsalva sinuses was first emphasized by Leonardo da Vinci 500 years ago [17]. In his anatomic drawings and in the surrounding text he correctly described the aortic root fluid dynamics and illustrated clearly the eddy currents within the aortic sinuses [17, 18]. Recent experimental observations [4, 5] suggest that the compliant Valsalva sinuses provide an additional stress-sharing mechanism for the semilunar leaflets and thus may significantly contribute to the longevity of the natural aortic valve. As a result of these observations, two special root replacement grafts have recently been designed and introduced into clinical practice. The Robicsek-Thubrikar graft [19] has three symmetrical tear-drop shaped artificial extensions attached to the end of a tube graft, and it is designed for the remodeling operation. The de Paulis graft [20] has a circumferential ballooning segment called a skirt mimicking the sinus anatomy, and it is used for reimplantation procedures. Both of these grafts were designed on the assumption that the aortic root contains three leaflets and has geometric symmetry that is reproducible. The overall functional geometry of the natural root is of a truncated cone with three individually bulging sinuses that terminate in well-defined sinus ridges [6, 7, 9]. Together these sinus ridges form the narrowest portion of the aortic root [6, 7, 9], designated as the STJ in the surgical literature. In normal circumstances with no root pathologic disease, there may be relatively predictable mathematic relationships among leaflets, orifice sizes, and STJ dimensions [7, 10], and the overall conal shape of the root is believed to be the most physiologic geometry in maintaining semilunar valve competence [11]. However, the human aortic root rarely has the described precise triple symmetry depicted on Leonardo s drawings. Typically the valve leaflets and the corresponding sinuses are not equal in size [21 24], and there is also a slight (5- to 11-degree) tilt angle between the basal ring plane and the STJ [23, 24]. Any additional pathologic change or abnormal physiology will likely distort this preexisting asymmetry even further. A more recent observation in patients with aortic root aneurysm [12] suggests that there may be a geometric mismatch between anatomically normal leaflet sizes and root dimensions, and the degree of AI did not seem to correlate with any of the above measurements. In addition there are limited data suggesting symmetry or predictable geometry in cases of bicuspid valves, which are often associated with aortic root enlargements and aneurysms [25]. Considering the complex three-dimensional geometry and the usual asymmetry of the natural aortic root, it is

5 Ann Thorac Surg KOLLAR ET AL 2009;87: VALVE SPARING WITH IN SITU AORTIC ROOT MEASUREMENTS 1799 surprising that both remodeling and reimplantation procedures follow the same initial surgical maneuvers (excising all three Valsalva sinuses) with subsequent reconstruction performed on pliable and floppy tissues without any reliable intraoperative testing. From this point of view, both of these techniques are based on assumptions, and the eventual success largely depends on the individual surgeon s skills rather than on empiric data or established guidelines for surgical reconstruction. By initially leaving the Valsalva sinuses intact and performing a trial restoration of valve competence within the native aortic root using our external downsizing maneuvers, we were essentially applying the single aortic root unit functional concept to the circumstances in the operating room. With the sinuses still in place, the old-fashioned water probing used by the pathologists proved to be a reliable intraoperative valve-testing method that could be repeatedly done after each adjustment until good leaflet coaptation and valve competence were observed. Our suggested in situ measurements describe the functional aortic root unit geometry in three dimensions, and in our opinion the neoaortic root should be rebuilt to this same geometry. In our small series valve competence as determined by nonpressurized open valve testing translated into predictable good valve function after the circulation was restored. Additional basal ring size measurements allowed us to describe the neoaortic root geometry that contributed to the reconstruction of a competent valve. So far we have seen good early surgical results with a reverse cone overall root geometry (STJ 2 to 4 mm larger than measured basal ring size). Our surgical experience is limited and does not permit us to make far-reaching conclusions. However, these findings support the idea that the original concept of restoring geometry to normal may not be appropriate to achieve valvular competence, particularly with significantly distorted root structures containing elongated leaflets. Interestingly, the consistent good intermediate surgical results and the long-term satisfactory functional outcomes with the original reimplantation procedure using a straight tube graft [26, 27] may also support our suggested hypothesis. When the entire aorticoventricular junction is pulled into a straight tube and the commissures are resuspended higher up, the ultimate functional geometry will be neither a straight tube nor a truncated cone but rather a reverse cone. At the STJ level there is no circumferential native tissue, and the neoroot outflow orifice equals the graft s internal dimension. When, however, the relatively thick circumferential basal ring consisting of fibrous segments and muscle tissue [9] is pulled into the same tubular structure, the result is a significantly narrowed inflow orifice. This reverse cone geometry concept may not have been recognized by the authors, but it suggests that their initial surgical intuition to restore aortic valve function was correct. Our described integrated root reconstruction technique also includes a new proximal suture line method [14], which allowed us to apply the David circumferential annulus stabilization concept to the Yacoub remodeling. As a result, the annulus is sandwiched between layers of graft material and pledgets, and our suture line follows the three-dimensional coronet. The commissural aspects of the annulus are stabilized between external pledgets and then sutured to the scalloped graft material, and the three intercommissural segments of the annulus are firmly sandwiched between supraannular graft material and subannular pledgets (Fig 2). Essentially, the tongues of the grafts are firmly anchored to the strongest part of the aortic annulus, then the residual sinus wall rim is folded over to the suture line as additional buttress for better hemostasis [14]. In summary, we believe that valve-sparing aortic root surgery should be based on the single functional unit concept [4] and include the following principles: (1) the operation should begin with a trial reconstruction within the native root until perfect valve competence with good leaflet coaptation is observed on open testing; (2) when the valve is fully competent our recommended in situ measurements accurately describe the ideal geometry (and asymmetry) of the restored functional unit; (3) surgical reconstruction should be aimed to match the above geometry; and (4) circumferential annulus stabilization (David concept) should be an integral part of the procedure to produce good long-term outcomes. In our experience, measured root asymmetry could be easily matched with individual tongue tailoring, and by introducing a proximal suture line that reinforces the entire aortic annulus alongside the scallops, we have successfully integrated the Yacoub remodeling operation with the David concept. Longer-term follow-up studies are necessary to assess durability, particularly in Marfan patients. References 1. Yacoub MH, Fagan A, Stassano P, Radley-Smith R. Results of valve conserving operations for aortic regurgitation [Abstract]. Circulation 1983;68: Sarsam MA, Yacoub M. Remodeling of the aortic valve anulus. J Thorac Cardiovasc Surg 1993;105: David TE, Feindel CM. An aortic valve-sparing operation for patients with aortic incompetence and aneurysm of the ascending aorta. J Thorac Cardiovasc Surg 1992;103: Robicsek F, Thubrikar MJ. Role of sinus wall compliance in aortic leaflet function. Am J Cardiol 1999;84: Beck A, Thubrikar MJ, Robicsek F. Stress analysis of the aortic valve with and without the sinuses of Valsalva. J Heart Valve Dis 2001;10: Reid K. The anatomy of the sinus of Valsalva. Thorax 1970;25: Thubrikar M. The aortic valve. Boca Raton, FL: CRC Press, Sutton JP, Ho SY, Anderson RH. The forgotten interleaflet triangles: a review of the surgical anatomy of the aortic valve. Ann Thorac Surg 1995;59: Anderson RH. Clinical anatomy of the aortic root. Heart 2000;84: Kunzelman KS, Grande KJ, David TE, Cochran RP, Verrier ED. Aortic root and valve relationships. Impact on surgical repair. J Thorac Cardiovasc Surg 1994;107: Labrosse MR, Beller CJ, Robicsek F, Thubrikar MJ. Geometric modeling of functional trileaflet aortic valves: development and clinical applications. J Biomech 2006;39:

6 1800 KOLLAR ET AL Ann Thorac Surg VALVE SPARING WITH IN SITU AORTIC ROOT MEASUREMENTS 2009;87: Thubrikar MJ, Labrosse MR, Zehr KJ, Robicsek F, Gong GG, Fowler BL. Aortic root dilatation may alter the dimensions of the valve leaflets. Eur J Cardiothorac Surg 2005;28: Miller DC, Bachet J, Cameron DE, et al. Panel discussion: Session I Ascending aorta. Ann Thorac Surg 2007;83(Suppl): S Kollar A. Valve-sparing reconstruction within the native aortic root: integrating the Yacoub and the David methods. Ann Thorac Surg 2007;83: Kollar A, Hartyanszky I. External subcommmissural annuloplasty to prevent regurgitation in the pulmonary autograft. Interact Cardiovasc Thorac Surg 2003;2: Trusler GA, Moes CA, Kidd BS. Repair of ventricular septal defect with aortic insufficiency. J Thorac Cardiovasc Surg 1973;66: O Malley CD, Saunders JBdeCM, eds. Leonardo da Vinci on the human body. The anatomical, physiological and embryological drawings of Leonardo da Vinci. New York: Random House Gramercy Books, Robicsek F. Leonardo da Vinci and the sinuses of Valsalva. Ann Thorac Surg 1991;52: Zehr KJ, Thubrikar MJ, Gong GG, Headrick JR, Robicsek F. Clinical introduction of a novel prosthesis for valvepreserving aortic root reconstruction for annuloaortic ectasia. J Thorac Cardiovasc Surg 2000;120: De Paulis R, De Matteis GM, Nardi P, Scaffa R, Bassano C, Chiariello L. Analysis of valve motion after the reimplantation type of valve-sparing procedure (David I) with a new aortic root conduit. Ann Thorac Surg 2002;74: Vollebergh FE, Becker AE. Minor congenital variations of cusp size in tricuspid aortic valves. Possible link with isolated aortic stenosis. Br Heart J 1977;39: Silver MA, Roberts WC. Detailed anatomy of the normally functioning aortic valve in hearts of normal and increased weight. Am J Cardiol 1985;55: Choo SJ, McRae G, Olomon JP, et al. Aortic root geometry: pattern of differences between leaflets and sinuses of Valsalva. J Heart Valve Dis 1999;8: Berdajs D, Lajos P, Turina M. The anatomy of the aortic root. Cardiovasc Surg 2002;10: Friedman T, Mani A, Elefteriades JA. Bicuspid aortic valve: clinical approach and scientific review of a common clinical entity. Expert Rev Cardiovasc Ther 2008;6: David TE, Armstrong S, Ivanov J, Webb GD. Aortic valve sparing operations: an update. Ann Thorac Surg 1999;67: David TE, Feindel CM, Webb GD, Colman JM, Armstrong S, Maganti M. Aortic valve preservation in patient with aortic root aneurysm: result of the reimplantation technique. Ann Thorac Surg 2007;83(Suppl):S Southern Thoracic Surgical Association: Fifty-Sixth Annual Meeting The Fifty-Sixth Annual Meeting of the Southern Thoracic Surgical Association (STSA) will be held November 4 7, 2009, in Marco Island, Florida. Manuscripts accepted for the Resident Competition must be submitted to the STSA headquarters office no later than October 12, The Resident Award will be based on abstract, presentation, and manuscript. Please visit for more information on the meeting and membership by The Society of Thoracic Surgeons Ann Thorac Surg 2009;87: /09/$36.00 Published by Elsevier Inc