Aortic valve replacement with a pulmonary autograft

Normalization of Left Ventricular Dimensions After Ross Operation With Aortic Annular Reduction Kazuo Niwaya, MD, Ronald C. Elkins, MD, Christopher J. Knott-Craig, MD, KathyLee Santangelo, MD, M. Bruce Cannon, MD, and Mary M. Lane, PhD Section of Thoracic and Cardiovascular Surgery, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma Background. Fifty-seven patients (August 1995 to November 1998) with a dysplastic dilated aortic root, a relative contraindication to the Ross operation, received an extended Ross operation with aortic annulus reduction and external cuff fixation (age 14 54 years). To assess the efficacy of these operations, echocardiographic assessment of autograft valve function and left ventricular function and dimensions were reviewed. Methods. Preoperative and postoperative assessment of 27 patients with aortic insufficiency (AI group) and 30 patients with aortic stenosis (> 20 mm Hg peak gradient) and aortic insufficiency (AS group) were compared. Aortic annulus size, valvular gradient, valve insufficiency, left ventricular dimensions at end-systole and enddiastole, left ventricular fractional shortening, and left ventricular mass were assessed. Results. There was one late death. Aortic annulus size, degree of AI, left ventricular internal dimensions, and left ventricular mass were all significantly reduced (p < 0.05) postoperatively in the AI group. Mean peak pressure gradients for this group were 6.8 6.7 mm Hg before operation and 8.7 6.4 mm Hg at 1 year after operation. Peak pressure gradient, aortic annulus size, degree of AI, left ventricular internal dimensions, and left ventricular mass were significantly reduced (p < 0.05) in the AS group. Mean fractional shortening was within normal limits pre- and postoperatively for both groups. Conclusions. Regression of left ventricular dilatation and hypertrophy, excellent autograft valve function, and survival suggest that this modification of the Ross operation may be offered to patients with a dysplastic aortic root requiring aortic valve replacement. (Ann Thorac Surg 1999;68:812 9) 1999 by The Society of Thoracic Surgeons Aortic valve replacement with a pulmonary autograft (Ross operation) is considered by many surgeons as the operative procedure of choice in children and young adults with aortic valve disease requiring valve replacement [1 3]. Late autograft valve insufficiency requiring reoperation with repair or replacement of the autograft valve has occurred more frequently in patients with significant size discrepancy between the pulmonary valve annulus and the aortic annulus and in patients with significant aortic valve insufficiency [1, 4, 5]. Replacement of the abnormal aortic root with a pulmonary autograft root replacement has been associated with early and late autograft failure, which has led many surgeons to question the use of the Ross operation in these patients and therefore to deny many young patients the benefits of this operation. To extend the indications for the Ross operation, we have performed concomitant aortic annulus reduction and fixation as well as reduction aortoplasty or ascending aortic resection and replacement with a Dacron graft when indicated in patients with significant aortic root pathology [6]. The dilated aortic annulus is reduced by a modification of Chauvaud s technique [7] and fixed with Presented at the Thirty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 25 27, 1999. Address reprint requests to Dr Elkins, Section of Thoracic and Cardiovascular Surgery, University of Oklahoma Health Sciences Center, PO Box 26901, Oklahoma City, OK 73190. an external cuff of Dacron. We have elected to adjust the aortic annulus size to a size appropriate for the patient s body surface area, recognizing the distensibility of the pulmonary annulus and the associated difficulty of assessing its normal size by echocardiography or by direct measurement. Previous reports of normalization of left ventricular function after a Ross operation have been primarily in children and have included intraaortic autograft implants as well as root replacements [8 10], but have not included patients with aortic annulus reduction. To assess the efficacy of this extended Ross operation on autograft valve function and on left ventricular function, we have reviewed and compared the preoperative and postoperative echocardiographic results. Patients and Methods Three hundred and twenty-nine patients have had a Ross operation at the University of Oklahoma Health Sciences Center between August 1986 and November 1998. Beginning in August 1995, we introduced concomitant aortic annulus suture reduction and fixation with external synthetic cuff fixation in patients with an enlarged aortic annulus who were otherwise candidates for a Ross operation. The patient s aortic annulus was measured by preoperative transthoracic echocardiogram and confirmed by intraoperative transesophageal echocardiogram and by direct measurement with Hegar (uterine) 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00 Published by Elsevier Science Inc PII S0003-4975(99)00765-1

Ann Thorac Surg NIWAYA ET AL 1999;68:812 9 ANNULUS REDUCTION WITH ROSS OPERATION 813 Table 1. Preoperative Patient Characteristics Variable Total (n 57) AI Group (n 27) AS Group (n 30) Age (years) 31 (13 54) 27 (14 49) 34 (13 54) Sex (male/female) 49/8 25/2 24/6 Body surface area (m 2 ) 2.0 (1.4 2.5) 2.0 (1.5 2.5) 1.9 (1.4 2.5) Aortic annulus (mm) 29 (23 43) 31 (25 43) 29 (23 35) Pulmonary annulus 24 (17 29) 24 (17 29) 24 (19 28) (mm) Aortic valve morphology Tricusp 5 5 0 Bicusp 46 22 24 Unicusp 6 0 6 Aortic root pathology Aneurysm 16 7 9 Dilated 17 4 13 Previous cardiac surgery Co-A 2 AVP 4 AV tunnel 1 AVR 1 CABG 1 PDA 1 Age, body surface area, aortic annulus, and median and range are displayed for pulmonary annulus. AVP aortic valvuloplasty; AVR aortic valve replacement; AV tunnel aorto-ventricular tunnel; CABG coronary artery bypass grafting; Co-A coarctation of aorta; PDA patent ductus arteriosus. dilators. When the aortic annulus size was at least 2 mm larger than the predicted aortic annulus size based on body surface area [11], and there was moderate to severe aortic insufficiency on the preoperative echocardiogram, the aortic annulus was reduced and fixed at the predicted size for the patient s body surface area. Fifty-seven patients have met these criteria and have been divided into two groups: the AI group (27 patients with predominant aortic insufficiency and less than 20 mm Hg peak gradient across their aortic valve) and the AS group (30 patients with 20 mm Hg gradient or more, irrespective of their degree of aortic insufficiency). In the AS group, 6 patients had trace to mild aortic insufficiency and the remainder had moderate to severe aortic insufficiency. The patient demographics are listed in Table 1. Our operative techniques for the Ross operation and aortic annulus reduction and fixation have been previously described [12]. All operations were performed with the use of mild systemic hypothermia (26 C 28 C) and myocardial protection was provided by intermittent retrograde cold blood cardioplegia with topical cooling. The dilated aortic annulus was decreased in size with two purse-strings of 2-0 or 3-0 polypropylene placed at the nadir of the coronary sinuses and continuing below the aortic annulus in the interleaflet triangle. The sutures were brought external to the aorta in the nadir of the noncoronary sinus and tied over a felt pledget with a Hegar dilator of appropriate size for the patient s body surface area in the aortic annulus. The proximal suture line between the pulmonary autograft and the narrowed aortic annulus was interrupted 4-0 polypropylene. These sutures were placed to include the annular reduction Table 2. Operative Characteristics Characteristic Total (n 57) AI Group (n 27) AS Group (n 30) Cardiopulmonary bypass 159 34 152 26 165 38 time (min) Aortic cross-clamp time 111 24 108 24 114 28 (min) Postop aortic annulus 23.4 1.2 23.5 1.1 23.2 1.3 (mm) Modification of ascending aorta Replacement 16 7 9 Aortoplasty 15 4 11 Resection 2 0 2 Concomitant surgery CABG 4 1 3 PFO 2 2 0 CABG coronary artery bypass grafting; foramen ovale. PFO closure of patent sutures, and those sutures in the nadir of the autograft sinuses include the autograft valve annulus. The sutures were tied over a 3-mm strip of woven Dacron in all patients except 2, in whom autologous pericardium was used, as these patients were operated on for active endocarditis. The two ends of the external strip were tied together with the last two sutures of the proximal suture line. After implantation of the left coronary artery, the distal suture line between the ascending aorta and the pulmonary artery was constructed just distal to the sinotubular junction of the pulmonary autograft. The ascending aorta with mild dilatation was reduced in size with a vertical aortoplasty to the size of the sinotubular dimension of the pulmonary autograft. In patients with aneurysmal dilatation of the ascending aorta, the ascending aorta was replaced with a Dacron graft (Hemashield; Meadox Medical Inc, Oakland, NJ) similar in size or 2-mm smaller than the size of the reduced aortic annulus. Vertical aortoplasty was required in 15 patients, replacement of the ascending aorta in 16 patients, and resection of localized dilatation in 2. In all patients, the right ventricular outflow tract was reconstructed with a cryopreserved pulmonary homograft (CryoLife, Inc, Kennesaw, GA), using the largest pulmonary homograft that could easily be inserted and attempting to match the homograft donor s age to the age of the recipient. Intraoperative direct measurement of the prereduction aortic annulus size was 29.6 3.2 mm (range 23 to 43 mm, median 29.0 mm), and postreduction aortic annulus size was 23.4 1.2 mm (range 20 to 25 mm, median 24.0 mm). Pulmonary annulus size was not measured directly. Intraoperative transesophageal pulmonary artery annulus size was 24.1 2.1 mm (range 19 to 29 mm, median 24.0 mm). Operative data are summarized in Table 2. Patient Follow-up Follow-up was complete within 1 year of completion of the study (November 1998), with a mean follow-up of 2

814 NIWAYA ET AL Ann Thorac Surg ANNULUS REDUCTION WITH ROSS OPERATION 1999;68:812 9 Table 3. Echocardiographic Data of Aortic Insufficiency Group Variable Preoperative (1 week) (6 months) (1 year) (2 years) ANN (mm) 30.7 3.5 23.5 1.1 a 22.9 3.1 a 23.7 1.4 a 24.3 1.5 a Z value 3.1 1.6 0.7 0.8 a 1.3 2.1 a 0.4 0.9 a 0.2 0.9 a AI degree 3.4 0.7 1.0 0.2 a 1.0 0.4 a 1.0 0.4 a 1.1 0.8 a dp (mm Hg) 6.8 6.7 7.5 5.6 6.8 5.9 8.7 6.4 4.5 5.9 LVDD (mm) 64.2 7.1 50.4 5.4 a 53.5 4.5 a 54.3 4.0 a 54.0 4.3 a Z value 2.6 1.3 0.2 0.9 a 0.6 0.8 a 0.9 0.8 a 0.8 0.9 a LVDS (mm) 42.6 9.4 35.5 6.6 a 36.4 4.9 a 37.5 4.6 37.0 4.7 Z value 2.0 1.8 0.7 1.3 a 0.8 1.0 a 1.2 0.9 1.0 0.9 IVS (mm) 11.3 2.2 11.7 1.2 11.0 1.6 10.6 3.1 10.7 1.8 Z value 0.6 1.0 0.9 0.7 0.3 0.9 0.2 1.5 0.3 1.1 LVPW (mm) 11.7 2.7 12.1 2.6 11.3 2.0 10.5 3.0 9.5 1.7 a Z value 1.1 1.6 1.5 1.7 0.7 1.2 0.4 1.8 0.3 1.1 a FS (%) 34 9 31 9 32 6 30 7 31 5 LVM (g) 410 156 294 92 a 277 65 a 274 105 a 251 73 a LVM/BSA (g/m 2 ) 201 72 148 49 a 132 33 a 135 48 a 125 37 a Data are mean SD. a p 0.05 compared with preop. AI aortic insufficiency; ANN aortic annulus size; BSA body surface area; dp left ventricular outflow tract peak pressure gradient; FS fractional shortening; IVS interventricular septal thickness; LVDD left ventricular dimension at end-diastole; LVDS left ventricular dimension at end-systole; LVM left ventricular muscle mass; LVPW left ventricular posterior wall thickness. years after the Ross Operation. Follow-up assessment included late death, all complications, and all valverelated complications [13]. echocardiograms were obtained within 1 week of operation, at 6 months, 1 year, and annually thereafter. End-systolic and end-diastolic left ventricular internal dimensions, left ventricular fractional shortening, left ventricular posterior wall thickness, and interventricular septal thickness were measured by M-mode echocardiography. Aortic annulus dimension was assessed by two-dimensional Doppler echocardiography. Pulse- and continuous-wave Doppler were used to measure left ventricular outflow tract obstruction, and color-flow mapping was used to assess autograft insufficiency. Left ventricular mass was calculated by means of the modified formulae of Devereux and Reichek [14]. Autograft valve insufficiency was graded: 0 (none), 1 (trace), 2 (mild), 3 (moderate), and 4 (severe). Statistical Analysis Data are expressed as the mean one standard deviation of the mean unless otherwise specified. All analyses were performed using SAS System software, version 6.12 (SAS Institute, Cary, NC). Between-group differences of continuous variables were analyzed using analysis of variance methods, and 2 or Fisher exact methods were used to test differences between proportions. All tests were two-tailed, and p 0.05 was considered to indicate statistical significance. To assess time-related changes in left ventricular dimensions and function and aortic annulus size, measurements of the aortic annulus, left ventricular dimension at end-diastole (LVDD) and endsystole (LVDS), ventricular septal thickness at enddiastole (IVS), and left ventricular posterior wall thickness at end-diastole (LVPW) were expressed as the number of standard deviations away from the expected mean value of a normal population with given body surface area (Z value, normal deviate [15]). Normal value equations for aortic annulus measurements [16] and LVDD, LVDS, IVS, and LVPW [17] were obtained from echocardiographic studies. The Z value or number of standard deviations from the normal mean, where the mean and standard deviations are dependent on body surface area, of each measurement was compared, with 0 (the expected normal mean for a given body surface area has a Z value of 0) using a single-sample t test. Results Mortality and Morbidity There were no operative deaths and one late death. This patient died 3 months postoperatively due to respiratory failure after a severe neurologic injury associated with a hypertensive crisis in the early postoperative period. Perioperative morbidity occurred in 2 patients. Both required pacemaker implantation, one related to the annular reduction suture placement in a patient with a unicommissural valve, and the other related to calcium debridement of the membranous and perimembranous septum. One patient required reoperation for pulmonary homograft stenosis at 13 months post-ross operation. Autograft Valve Function Preoperative and postoperative echocardiographic data are shown in Tables 3 and 4. The preoperative aortic annulus Z value was 3.1 in the AI group and 2.6 in the AS group. Annulus reduction and fixation reduced the aortic annulus Z value to 0.7 in the AI group and to 0.2 in the AS group at 1 week after operation. The aortic annulus

Ann Thorac Surg NIWAYA ET AL 1999;68:812 9 ANNULUS REDUCTION WITH ROSS OPERATION 815 Table 4. Echocardiographic Data of Aortic Stenosis Group Variable Preoperative (1 week) (6 months) (1 year) (2 years) ANN (mm) 28.6 2.6 23.2 1.3 a 23.1 3.0 a 23.7 2.9 a... Z value 2.6 1.5 0.2 0.8 a 0.2 1.6 a 0.2 1.4 a... AI degree 2.8 0.9 0.9 0.3 a 0.7 0.5 a 0.8 0.4 a 0.9 0.4 a dp (mm Hg) 51.8 28.6 6.9 6.4 a 4.0 3.8 a 5.7 7.7 a 4.0 4.1 a LVDD (mm) 57.3 8.8 49.8 7.5 a 48.8 5.4 a 49.7 4.8 a 50.4 2.3 a Z value 1.8 1.8 0.4 1.2 a 0.3 0.9 a 0.4 0.9 a 0.4 0.6 a LVDS (mm) 35.1 6.8 32.3 6.2 30.8 3.8 a 31.0 4.2 a 31.9 2.8 Z value 0.9 1.4 0.3 1.0 0.1 0.7 a 0.0 0.9 a 0.1 0.4 IVS (mm) 12.2 4.4 12.2 3.0 11.4 3.0 a 10.7 3.5 10.3 2.6 Z value 1.2 2.0 1.3 1.7 0.8 1.4 0.4 1.7 0.4 1.4 LVPW (mm) 11.8 3.7 12.0 2.3 10.7 2.0 9.8 2.2 a 10.0 2.7 Z value 1.3 1.9 1.6 1.6 0.6 1.1 0.1 1.4 a 0.4 1.5 FS (%) 39.2 8.2 35.2 6.7 36.8 8.0 40.8 6.4 36.6 5.4 LVM (g) 354 186 285 102 a 239 91 a 221 73 a 225 78 a LVM/BSA (g/m 2 ) 187 92 151 50 a 126 42 a 115 35 a 116 31 a Data are mean SD. a p 0.05 compared with preop. AI aortic insufficiency; ANN aortic annulus size; BSA body surface area; dp left ventricular outflow tract peak pressure gradient; FS fractional shortening; IVS interventricular septal thickness; LVDD left ventricular dimension at end-diastole; LVDS left ventricular dimension at end-systole; LVM left ventricular muscle mass; LVPW left ventricular posterior wall thickness. size has remained significantly reduced postoperatively in both groups ( 0.2 at 2 years in the AI group, 0.2 at 1 year in the AS group). In the AI group, the mean preoperative peak left ventricular outflow tract gradient was 6.8 6.7 mm Hg and has remained unchanged, with a mean gradient of 4.5 5.9 mm Hg at 2 years after operation. In the AS group, the mean preoperative peak left ventricular gradient was 51.8 28.6 mm Hg, and at 1 week after operation, the gradient had been reduced to 6.9 6.4 mm Hg. The gradients have remained constant with a mean value of 4.0 4.1 mm Hg at 2 years after operation. None of the patients in the two groups currently has a gradient greater than 17 mm Hg. Autograft valve insufficiency has remained stable in 56 of the 57 patients, with 0 or 1 insufficiency on their most recent echocardiogram. One patient, a competitive swimmer, had stable 1 to 2 autograft insufficiency until the patient was involved in an automobile accident with significant chest trauma and was noted to have 2 to 3 insufficiency on his postinjury echocardiogram. The significance of the trauma in the progression of his autograft valve dysfunction is unknown. The mean grade of autograft insufficiency in all patients was 1.0 0.6 at 2 years. Homograft Valve Function One patient developed homograft valve dysfunction at 13 months after operation and required replacement of the homograft. One additional patient, who is 3 years postoperative, has a mean gradient across his homograft valve of 40 mm Hg but is asymptomatic at this time and is being followed. Left Ventricular Dimensions and Function The preoperative and postoperative echocardiographic data on ventricular size, wall thickness, and function are shown for the AI group in Table 3 and for the AS group in Table 4. In the AI group, the before operation LVDD was 64.2 7.1 (Z value 2.6 1.3), and by 1 week after operation, this had decreased to 50.4 5.4 (Z value 0.2 0.9 [ p 0.05]). During the 2 years of follow-up, the mean Z value has increased, but remained less than 1. In the AS group, the before operation Z value was 1.8 1.8, which decreased to 0.4 1.2 ( p 0.05) at 1 week after operation. The mean Z value in this group has remained constant during the 2 years of follow-up (Fig 1). Left ventricular dimension in end systole was 42.6 9.4 mm (Z value 2.0 1.8) by preoperative echocardiogram in the AI group. The mean LVDS remained in the normal range without a significant change, although the Fig 1. The mean Z value ( standard deviation) of left ventricular dimension at end-diastole based on body surface area and age for the AI and the AS groups. Significant reduction from preoperative value (p 0.05) was present at all postoperative intervals.

816 NIWAYA ET AL Ann Thorac Surg ANNULUS REDUCTION WITH ROSS OPERATION 1999;68:812 9 Fig 2. The mean left ventricular mass indexed for body surface area ( standard deviation). Significant reduction from preoperative value was seen at all postoperative measurements in the two groups (p 0.05). Z value decreased to 1 0.9 by 2 years. In the AS group, the mean preoperative LVDS was 35.1 6.8 (Z value, 0.9 1.4) and decreased to 31.9 2.8 (Z value, 0.1 0.4) at 2 years. The mean preoperative Z-value for IVS in the AI group was 0.6 1.0, and at 1 week had increased slightly, probably due to edema. A nonsignificant steady decline was seen in the Z value to 0.3 1.1 at 2 years. In the AS group, the mean preoperative Z value was 1.2 2, and the Z values in this group had a mild increase at 1 week with a nonsignificant decrease to 2 years. Mean preoperative left ventricular posterior wall thickness Z value was 1.1 1.6 in the AI group. During the 2 years of follow-up, there was a decrease in the Z value to 0.3 1.1, a nonsignificant change. The mean Z value in the AS group had a similar decrease from 1.3 1.9 to 0.4 1.5. Preoperative data were available to calculate the left ventricular mass index in 20 of the AI group and in 19 of the AS group. In 17 of the 20 AI patients, the left ventricular mass index was greater than the 95th percentile of 129.7 g/m 2 [18] before operation, and in the AS group, 16 of 19 patients had a preoperative left ventricular mass index greater than normal. The mean preoperative value of the left ventricular mass index in the AI group was 201 72 g/m 2, but by 1 week, had decreased to 148 49 g/m 2 ( p 0.05). At 2 years, the mean value of left ventricular mass index is 125 37 g/m 2, a value within the normal range. In the AS group, the mean preoperative left ventricular mass index was 187 92 g/m 2, and by 1 week after operation, had decreased to 151 50 g/m 2 ( p 0.05). By 6 months after operation, the mean value of left ventricular mass index in this group had become normal and has remained in the normal range for 2 years (Fig 2). The mean values for left ventricular fractional shortening (FS) have remained in the normal range for both Fig 3. Mean left ventricular fractional shortening ( standard deviation) for AI and AS groups. All mean values are within normal range 95% confidence limits of 28% 44%. groups throughout the study (Fig 3). In the AS group, 7 patients initially had a supranormal FS and all but one of these patients had 3 or 4 AI associated with their stenosis. By their 6-month postoperative study, all but 2 of these had returned to normal. In the AI group, 4 patients had a supranormal preoperative FS but had tended toward the normal range by 6 months after operation. One patient (age 49 years) with a very dilated left ventricle (LVDD 7.2 cm) and a very low FS (7.3) has had a gratifying decrease in his LVDD to 5.6 cm and an increase in his FS to 23.2. Comment The introduction of the pulmonary autograft replacement of the aortic valve by Ross in 1967 [19] and the subsequent excellent event-free survival at over 20 years [1] led to the introduction of this procedure for children and young adults in this country. Recognizing the technical difficulty of implanting an anatomically perfect pulmonary valve using the scalloped subcoronary technique of Ross, surgeons began using the autograft root replacement, hoping to decrease the incidence of autograft valve dysfunction and failure [2, 20]. Routine echocardiographic surveillance identified early autograft dysfunction associated with aortic annulus dilatation and sino-tubular dilatation of the autograft root, particularly in patients with aortic insufficiency as the predominant preoperative valvular abnormality [4, 6]. Reoperative restoration of autograft function was frequently possible, if the aortic annulus could be reduced to a more normal size and, if necessary, the sinotubular dimension was restored to its normal size [4]. Careful longitudinal follow-up of our clinical series suggested that patients with aortic insufficiency and mismatch between the aortic annulus size and the pulmonary artery annulus had a less satisfactory result after the Ross operation, whether

Ann Thorac Surg NIWAYA ET AL 1999;68:812 9 ANNULUS REDUCTION WITH ROSS OPERATION 817 it was done as an intraaortic implant or as a root replacement. This information has led to our present policy of elective reduction of the aortic annulus and external fixation in patients with an aortic annulus that is greater than one would anticipate for their body surface area. If the aortic valve annulus is considered as a circle, reduction of the valve diameter from 25 to 23 mm will produce a 15% decrease in the valve area; therefore, we have reduced and fixed the valve annulus size when the measured valve diameter is more than 2 mm greater than the expected mean diameter for the patient s body surface area. The postoperative reduced valve diameters measure 20 to 25 mm, with a mean reduction of 7 mm in the AI group and 5.3 mm in the AS group. Follow-up echocardiograms at 6 months identified 5 patients with peak aortic gradients greater than 10 mm Hg, with the highest gradient being 18 mm Hg. The most recent echocardiograms on these patients showed only 1 patient with a left ventricular outflow gradient greater than 10 mm Hg, and that was 17 mm Hg. This patient, a competitive swimmer, has a left ventricular mass index of 120.6 g/m 2, suggesting a normal left ventricular mass at his most recent echocardiogram, 2 years after operation. Progression of left ventricular outflow tract obstruction has not been seen after a Ross operation with annular reduction and external fixation. The postoperative gradients measured in these patients are similar to those recently reported in patients with tissue valve replacement of their aortic valve [21 23]. The short follow-up available in this subgroup of patients does not allow a solid response to the wisdom of managing patients with aortic valve disease and associated significant ascending aortic disease with an extended Ross operation and replacement or repair of their ascending aorta. Thirty-three of the 57 patients had significant dilation of their ascending aorta, and in 17, a knitted Dacron graft (Hemashield) was used to replace their ascending aorta. In an additional 15 patients, the ascending aortic dilatation was managed with a vertical aortoplasty, and 2 had a resection of their localized dilatation with anastomosis of the autograft root to their remaining ascending aorta. Because of the relative young age of these patients, the aortoplasty was unsupported with an external wrap. None of the patients had evidence to suggest an inherited defect such as Marfan disease, and all but 1 patient had a bicuspid or a unicuspid aortic valve. Pathologic evaluation of the resected aortic tissue showed aneurysmal changes in some, but most had only atherosclerotic changes associated with thinning of the aortic wall. There have been no operative or postoperative complications associated with concomitant management of their ascending aortic disease, and these patients have avoided the risks and complications of a prosthetic valved conduit. Whether the present philosophy will provide a more permanent solution than aortic homograft replacement of the aortic valve and ascending aorta using the homograft as a root replacement conduit will only be demonstrated with long-term follow-up of this patient series. David [5], Eishi [24], and Durham [25] and associates have reported experiences with annular reduction in patients with a mismatch between the aortic annulus size and the pulmonary valve annulus size. Sizing for the aortic annulus has been based on direct measurement of the pulmonary valve annulus or calculation of the pulmonary valve annulus size based on measurement of the sinotubular dimension of the autograft valve. At our institution, we have elected to modify the technique of Chauvaud and associates [7] by employing an additional purse-string suture and to reinforce the annuloplasty with an external ring of woven Dacron. This has allowed us to reduce the aortic annulus with a reproducible technique that can be used in patients requiring a limited reduction or those who require a reduction of more than 1 cm in their aortic annulus diameter. The reduction annuloplasty was probably involved in one operative complication, an episode of complete heart block. There have been no other operative or postoperative complications associated with the annuloplasty technique or the technique of external fixation. The echocardiographic assessment in these patients was part of the routine postoperative care and is not complete in all patients. We do not have adequate data in some to assess all the determinants of left ventricular dimensions or function; however, adequate data are available to strongly suggest that the use of annular reduction has not been associated with left ventricular outflow tract obstruction. The changes in left ventricular end-diastolic dimension index and the changes in left ventricular mass index are not dissimilar to those previously reported in patients having a Ross operation [8, 9] or an aortic valve replacement with an unstented tissue valve [21 23]. The present early and limited midterm results suggest that extension of the Ross operation as a root replacement with annular reduction and external Dacron cuff fixation may be employed in the young patient with aortic annular dilatation and aortic root disease who is otherwise a candidate for a Ross operation. In young patients with anticipated somatic growth, the annuloplasty and fixation technique should be modified to allow for anticipated growth, or the method of Durham and associates [25] should be used. The excellent early survival, limited operative complications, excellent postoperative autograft valve function, and normalization of left ventricular function have encouraged us to continue to use this approach in these patients. Whether this operative approach will provide a more durable aortic valve replacement and be associated with a decreased incidence of aortic valve reoperation and decreased incidence of late valve-related complications than an aortic homograft root replacement will only be determined by continued close surveillance of these patients for several years. The decrease in left ventricular dimensions and the normalization of left ventricular mass index strongly suggest resolution of left ventricular dilatation and hypertrophy that has been maintained over the 2 years of follow-up.

818 NIWAYA ET AL Ann Thorac Surg ANNULUS REDUCTION WITH ROSS OPERATION 1999;68:812 9 References 1. Ross D, Jackson M, Davies J. The pulmonary autograft: a permanent aortic valve. Eur J Cardiothorac Surg 1992;6: 113 7. 2. Kouchoukos NT, Davila-Roman VG, Spray TL, Murphy SF, Perrillo JB. Replacement of the aortic root with a pulmonary autograft in children and young adults with aortic valve disease. N Engl J Med 1994;330:1 6. 3. Elkins RC, Knott-Craig CJ, Ward KE, McCue C, Lane MM. Pulmonary autograft in children: realized growth potential. Ann Thorac Surg 1994;57:1387 94. 4. Elkins RC, Lane MM, McCue C. Pulmonary autograft reoperation: incidence and management. Ann Thorac Surg 1996; 62:450 5. 5. David TE, Omran A, Webb G, et al. Geometric mismatch of the aortic and pulmonary roots causes aortic insufficiency after the Ross procedure. J Thorac Cardiovasc Surg 1996;112: 1231 9. 6. Elkins RC, Knott-Craig CJ, Howell CE. Pulmonary autografts in patients with aortic annulus dysplasia. Ann Thorac Surg 1996;61:1141 5. 7. Chauvaud S, Serraf A, Mihaileanu S, et al. Ventricular septal defect associated with aortic valve incompetence: results of two surgical managements. Ann Thorac Surg 1990;49: 875 80. 8. Rubay JE, Shango P, Clement S, et al. Ross procedure in congenital patients: results and left ventricular function. Eur J Cardiothorac Surg 1997;11:92 9. 9. Hokken RB, Cromme-Dijkhuis AH, Bogers AJ, et al. Clinical outcome and left ventricular function after pulmonary autograft implantation in children. Ann Thorac Surg 1997;63: 1713 7. 10. Santangelo K, Elkins RC, Stelzer P, et al. Normal left ventricular function following pulmonary autograft replacement of the aortic valve in children. J Card Surg 1991;6:633 7. 11. Kirklin JW, Barratt-Boyes BG. Cardiac surgery. New York: Churchill Livingstone, 1993:40 3. 12. Elkins RC. The Ross operation in patients with dilation of the aortic annulus and of the ascending aorta. In Cox JL, Sundt TM III, eds. Operative techniques in cardiac and thoracic surgery: a comparative atlas. Philadelphia: W.B. Saunders Co, 1997:331 41. 13. Edmunds LH Jr, Clark RE, Cohn LH, et al. Guidelines for reporting morbidity and mortality after cardiac valvular operations. Ann Thorac Surg 1996;62:932 5. 14. Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man. Anatomic validation of the method. Circulation 1977;55:613 8. 15. Zar JH. Biostatistical Analysis. Englewood Cliffs, NJ: Prentice-Hall, 1974:83. 16. El Habbal M, Somerville J. Size of the normal aortic root in normal subjects and in those with left ventricular outflow obstruction. Am J Cardiol 1989;63:322 6. 17. Henry WL, Gardin JM, Ware JH. Echocardiographic measurements in normal subjects from infancy to old age. Circulation 1980;62:1054 61. 18. Lauer MS, Anderson KM, Larson MG, Levy D. A new method for indexing left ventricular mass for differences in body size. Am J Cardiol 1994;74:487 91. 19. Ross DN. Replacement of aortic and mitral valves with a pulmonary autograft. Lancet 1967;2:956 8. 20. Stelzer P, Jones DJ, Elkins RC. Aortic root replacement with pulmonary autograft. Circulation 1989;80:III209 13. 21. Jin XY, Zhang ZM, Gibson DG, Yacoub MH, Pepper JR. Effects of valve substitute on changes in left ventricular function and hypertrophy after aortic valve replacement. Ann Thorac Surg 1996;62:683 90. 22. Westaby S, Jin XY, Katsumata T, Arifi A, Braidley P. Valve replacement with a stentless bioprosthesis: versatility of the porcine aortic root. J Thorac Cardiovasc Surg 1998;116: 477 84. 23. Kon ND, Cordell AR, Adair SM, Dobbins JE, Kitzman DW. Aortic root replacement with the Freestyle stentless porcine aortic root bioprosthesis. Ann Thorac Surg 1999;67:1609 16. 24. Eishi K, Nakajima S, Nakano K, et al. Pulmonary autograft implantation in the dilated aortic annulus. Ann Thorac Surg 1997;63:1155 8. 25. Durham LA III, desjardins SE, Mosca RS, Bove EL. Ross procedure with aortic root tailoring for aortic valve replacement in the pediatric population. Ann Thorac Surg 1997;64: 482 6. DISCUSSION DR EDWARD D. VERRIER (Seattle, WA): I would like to thank the Society for the opportunity to discuss this excellent paper from Dr Ron Elkins group at Oklahoma. I also appreciate receiving the manuscript, as some of my questions may relate to the manuscript. The presentation is obviously precise and clear. The contributions of this group towards our understanding of the technical considerations, the growth potential of the autograft, and both short-term and intermediate-term results with the Ross procedure for children and young adults with aortic valve and now aortic root disease is significant. Their comprehensive series of over 330 patients treated with the Ross procedure is among the largest in the literature. From that group of patients, the authors have technically reduced the aortic annulus in 27 patients with predominantly AI and 30 patients with AS. The mean reduction in annular diameter was 7 mm in the AI group and 5.3 mm in the AS group. The presentation details not only the echocardiographic follow-up but the significant refinements and extensions of their surgical techniques in this complex group of patients with remarkably low morbidity and mortality. Two critical components of their technique are highlighted. First, the authors precisely size the annulus based on body surface area estimates rather than visual or pulmonary valve annular estimates. They then fix the annulus with a 3-mm strip of Dacron incorporated into the autograft anastomosis to the left ventricular outflow tract. I believe the observed reductions in left ventricular systolic and diastolic dimensions, left ventricular septal and free-wall thickness, left ventricular mass, and outflow tract gradients in both the AS and AI groups are to be expected and are a testimony to the short-term technical expertise of this group. I have a few questions. I think that many of us have relied on preoperative transthoracic echocardiography, intraoperative transesophageal echocardiography, and direct visualization of the pulmonary valve, even to the point of opening the right ventricular pulmonary artery to assess the discrepancy in size of the aortic and pulmonary autograft annular dimensions. Your approach completely eliminates any evaluation of the pulmonary autograft diameter in the decision to proceed. Have you had an instance where this presented a problem after the annulus is fixed when you have transferred the pulmonary autograft that is either larger or smaller than expected? Secondly, based on the very tight mean and standard deviation of the completed annular size, which is about 23 mm Hg, were there any gender differences, and has your sizing protocol shortened based on this observation? Thirdly, I noted that 4 patients required concomitant CABG in this relatively young patient population. Was this due to tech-