Partial atrioventricular septal defect (AVSD), the socalled

Transcription

1 Surgery for Partial Atrioventricular Septal Defect in the Adult Michael A. Gatzoulis, MD, Sloane Hechter, Gary D. Webb, MD, and William G. Williams, MD The Toronto Congenital Cardiac Centre for Adults, The Toronto Hospital, and The Hospital for Sick Children, Toronto, Ontario, Canada Background. We sought to determine the clinical profile, operative results, and long-term outcome for adult patients undergoing operations for partial atrioventricular septal defects. Methods. Between 1976 and 1996, 50 adults (mean age, years) underwent surgery for partial atrioventricular septal defects. Thirty-nine of them underwent primary repair for a substantial left-to-right shunt (> 1.8), associated with symptoms in 29. The remaining 11 patients had previous atrioventricular septal defect repair in childhood, but required reoperation as adults for severe left atrioventricular valve regurgitation (6), subaortic (3) or mitral (1) stenosis, and a residual atrial shunt (1). Results. No patients died in hospital. Of the 39 patients first repaired in adulthood, left atrioventricular valve repair was performed in 37, valve replacement in 1, and no repair in 1. In contrast, left atrioventricular valve replacement was necessary in 2 of the 6 adults undergoing reoperation for left atrioventricular valve regurgitation. At 7 years median follow-up, 8 patients have died (2 from noncardiac causes). Of 42 patients alive in 1997, 39 are New York Heart Association class I or II, and 3 were class III (class improved in 81%). Two patients required left atrioventricular valve replacement (1 week and 5 years after repair, respectively) for valvar failure. Conclusions. Low operative risk and excellent longterm results support repair of partial atrioventricular septal defect in adults. (Ann Thorac Surg 1999;67:504 10) 1999 by The Society of Thoracic Surgeons Partial atrioventricular septal defect (AVSD), the socalled ostium primum atrial septal defect (ASD), is part of the spectrum of AVSDs. These lesions are characterized by a common atrioventricular (AV) junction. Associated morphologic features are (1) a trileaflet (cleft) left atrioventricular valve (LAVV) guarding the left ventricular component of the common AV junction, comprised of the mural leaflet and the superior and inferior bridging leaflets; (2) a defect in the atrial component of the AV septum; and (3) an unwedged aorta caused by the common AV orifice displacing anteriorly the aortic root [1 3]. In contrast to the complete form, patients with the partial form of AVSD have two separate AV valves, resulting from the fusion of the superior and inferior bridging leaflets, and no large intraventricular communication. Intermediate forms include a partial AVSD with a coexisting small and restrictive ventricular septal defect. Improved understanding of the anatomy, coupled with advances in surgical techniques and perioperative care, have improved the outcome in children with partial AVSD undergoing repair [4 6]. Medium- and long-term follow-up, however, has revealed ongoing morbidity and need for reoperation, often related to residual problems of the systemic AV valve [7 9]. Accepted for publication July 2, Address reprint requests to Dr Williams, Cardiovascular Surgery, The Hospital for Sick Children, Suite 1525, 555 University Ave, Toronto, ON M5G 1X8, Canada ( Bill.Williams@mailhub.sickkids.on.ca). Complete AVSDs present early in life, and unless treated expediently develop into irreversible pulmonary vascular disease. Presentation of partial AVSDs is more variable. Elective age for repair of partial AVSD is 3 to 5 years, but some patients will not present until later in life. Two recent reports by Bergin and associates [10] and Burke and colleagues [11] suggest that operations for adults with partial AVSDs can be carried out at low risk and have good long-term results. The purpose of our study was to (1) review our experience with operations for partial AVSD at the Toronto Congenital Cardiac Centre for Adults (TCCCA), with a particular focus on clinical profile, surgical approach, and symptomatic relief, and (2) establish further outcome data of adults undergoing repair of partial AVSD. Patients and Methods The study group comprised all 50 adult patients (18 years and older) who underwent surgery for partial AVSD at the TCCCA between 1976 and This 20-year period was selected both to reflect a reasonably contemporary surgical approach and to allow a minimum follow-up time of 6 months. Thirty-nine of these patients underwent primary repair for symptoms or a substantial leftto-right shunt. The remaining 11 patients had previous AVSD repair in childhood, and required reoperation as adults for residual or new and progressive hemodynamic problems. We include this subgroup in our study to 1999 by The Society of Thoracic Surgeons /99/$20.00 Published by Elsevier Science Inc PII S (98)

2 Ann Thorac Surg GATZOULIS ET AL 1999;67: PARTIAL AVSD IN THE ADULT 505 Table 1. Demographic, Hemodynamic, and Operative Profile (n 50) Characteristic Mean SD Range Age (years) Sex (male:female) 22/28... Shunt size (Q p:q s) Systolic PAp (mm Hg) Diastolic PAp (mm Hg) Mean PAp (mm Hg) Weight at repair (kg) BSA at repair (m 2 ) Bypass time (min) Cross-clamp time (min) BSA body surface area; SD standard deviation. address common complications leading to reoperation after AVSD repair, seen in the adult. Hospital records were reviewed for preoperative and postoperative clinical, electrocardiographic, and hemodynamic details. Detailed operative notes were obtained from the surgical database. Postoperative cardiac catheterization data were not always available, as there were no clinical indications for repeat invasive studies in the majority of these patients. In addition to determinations from death certificates, the preoperative and current follow-up symptomatic status of the patients was noted from detailed outpatient clinic assessment (for patients reviewed periodically at the TCCCA), records of the patient s referring physician, and patient questionnaires; information was obtained from one or more of these sources for all patients alive in Particular emphasis was placed on exercise performance, arrhythmias and other cardiovascular events (thromboembolism, stroke or transient ischemic attack, infective endocarditis), need for reoperation, and cause of death. Patients were asked to grade any changes (worse, no change, better, or much better) in exertional dyspnea and overall exercise performance compared with their preoperative condition. Data are presented as the mean 1 standard deviation, unless indicated otherwise. Student s t test was used for continuous variables. The 2 test was used to compare discrete variables. Cox regression analysis was performed to identify determinants or predictors of clinical outcome. Survival was expressed in a Kaplan-Meier actuarial analysis curve. In all tests, p less than 0.05 was considered significant. Results PAp pulmonary arterial pressure; The overall profile of the 50 patients enrolled in the study is shown in Table 1. Diagnoses additional to partial AVSD are listed in Table 2. Preoperative Data PRIMARY REPAIR. During the study period, 39 patients (17 men, 22 women) underwent primary surgical closure of a partial AVSD. Age ranged from 18 to 72 years (median age, 35.5 years; mean, years). The diagnosis of a partial AVSD was established by means of echocardiography in 27 patients (transesophageal echocardiography in 6) and during cardiac catheterization in 12. Nineteen additional patients underwent cardiac catheterization, mainly to assess pulmonary vascular resistance and status of the left AV valve and, in patients older than 40 years, to exclude coexisting coronary arterial disease. Thirteen patients had elevated pulmonary arterial pressure (systolic 45 mm Hg), but none had a reversed atrial shunt. Left atrioventricular valve regurgitation was moderate to severe in 6 (15%) of 39 patients, as established by echocardiography or cardiac catheterization. Only one of them had moderately severe left ventricular dysfunction. Two patients had associated coronary arteriosclerosis severe enough to require coronary arterial bypass grafting at the time of partial AVSD repair. Before the operation, 29 patients were symptomatic with one or more of the following symptoms: exertional dyspnea (27), palpitations (10), presyncope or syncope (4), angina (2), transient ischemic attack (1), and bacterial endocarditis (1). The 10 asymptomatic patients were referred for AVSD repair because of a substantial left-toright atrial shunt (pulmonary to systemic flow ratio 1.8:1) and echocardiographic or angiographic evidence of right heart volume overload. Sustained atrial arrhythmias were documented preoperatively in 8 (21%) of 39 patients. Five of them had atrial fibrillation, whereas 3 had atrial flutter. Patients older than 50 years at the time of repair were more likely to have previous history of atrial tachyarrhythmia, compared with patients aged 50 years or younger (5 of 8 versus 3 of 31; p 0.02). PREVIOUS AVSD REPAIR. Eleven patients (5 men, 6 women) underwent reoperations during the study period: 6 for severe LAVV regurgitation (none of whom had previous LAVV repair), 3 for left ventricular outflow tract (LVOT) obstruction, 1 for LAVV stenosis, and 1 for residual atrial shunt (Table 3). Their age at reoperation ranged from 21 to 51 years (median, 33 years; mean, years). Time Table 2. Diagnoses Additional to Partial Atrioventricular Septal Defect (n 50) Diagnosis Number of Patients Down syndrome 4 Secundum atrial septal defect 5 Unroofed coronary sinus 2 Small ventricular septal defect a 2 RAVV abnormalities b 6 Subaortic stenosis 4 Pulmonary valve stenosis 1 Coronary artery disease b 2 b Re- a Intermediate form of atrioventricular septal defect (AVSD). quiring intervention at the time of surgery for partial AVSD. RAVV right atrioventricular valve.

3 506 GATZOULIS ET AL Ann Thorac Surg PARTIAL AVSD IN THE ADULT 1999;67: Table 3. Clinical, Hemodynamic, and Surgical Profile of 11 Patients With Previous Repair of Atrioventricular Septal Defect Sex Age a (y) TFFO (y) Reason for Reoperation PAps (mm Hg) LAVV Follow-up Outcome b Years F Moderate/severe LAVVI... Repaired Much better 5 F Residual ASD (3.7:1) 70 Repaired Died (CHF) 6 F Severe LVOTO... Repaired Better 2 M LVOTO and LAVVI... Replaced Better 1 M Severe LAVVI 65 Repaired Same week later Replaced Better 6 F Severe LAVVI 63 Repaired Better 3 M Severe LAVVI 35 Replaced Better 17 M Severe LAVVI... Repaired Died (cancer) 16 F Moderate LVOTO 52 Repaired Better 2 M Moderate/severe LAVVS/I 60 Replaced Same 12 F Severe LAVVI 66 Replaced Better 7 a Age at left atrioventricular valve operation. b Same/better/much better outcomes are reported according to the written questionnaire (see Methods). ASD atrial septal defect; CHF congestive heart failure; LAVV left atrioventricular valve; LAVVI left atrioventricular valve incompetence; LAVVS/I left atrioventricular valve stenosis and incompetence; LVOTO left ventricular outflow tract obstruction; PAps systolic pulmonary arterial pressure; TFFO time from first operation. from initial AVSD repair (performed during childhood) was 11 to 35 years (median, 20 years; mean, years). Nine patients had one or more symptoms before reoperation (exertional dyspnea in 8, palpitations in 5, and syncope in 1). Three patients had sustained atrial arrhythmia (atrial flutter in 2, atrial fibrillation in 1). Perioperative Data At the TCCCA, all patients (Table 1) underwent repair during cardiopulmonary bypass with moderate hypothermia. For the past 12 years myocardial protection included potassium-enriched warm-induction blood cardioplegia, with cardioplegic cooling to a myocardial temperature of 15 C. A supplemental topical cooling device has been used to maintain a consistent temperature. PRIMARY REPAIR. Through a right atriotomy, the primum ASD was closed with an appropriately shaped patch in all 39 patients (autologous pericardium in 28, prosthetic patch material in 11), whereas primary suture was used for closing additional secundum defects in 3 of 5 cases. The coronary sinus was incised and left draining into the left atrium in 5 cases to avoid damage to the AV node during placement of suture lines for AVSD closure. In the other 34 patients, the suture line was kept on the right and superior aspect of the AV node area, placing the sutures at the base of the AV valve tissue adjacent to the node. The type of operation done on the LAVV is indicated in Table 4. All but 2 patients underwent suture repair between the adjacent edges of the inferior and superior bridging leaflets to improve valvar competence. Left atrioventricular valve competence was assessed routinely with saline injection and more recently with intraoperative transesophageal echocardiography. Additional annuloplasty to reduce the valve annulus further was required in 5 patients: 3 of them had insertion of a flexible annuloplasty ring, and 2 had suture plication of the valve annulus at the commissure between bridging leaflets and the mural leaflet. Only 1 (2.6%) primary repair patient, who had extensive reconstruction of the LVOT for moderately severe LVOT obstruction, required LAVV replacement. Finally, there was one patient with competent LAVV who did not require repair of the valve. Right AV valve repair was performed in 5 patients, with insertion of a flexible annuloplasty ring (2) or suture plication of the valve annulus (4 [circumferential in 1 and between valve commissures in 3]). Table 4. Left Atrioventricular Valve Operation: Demographics and Surgical Characteristics of Patients With Primary and Previous Repair of Partial Atrioventricular Septal Defect Characteristic Primary AVSD Repair (n 39) Previous AVSD Repair (n 11) Age (years) Sex (male:female) 17/22 5/6 Previous LAVV repair NA 4 LAVV incompetence a NA 6 LAVV stenosis 0 1 CS draining to left atrium 6 1 Cleft repair performed 37 4 b Additional annuloplasty g 5 1 LAVV replacement 1 c 4 cde RAVV repair 5 2 Elapsed follow-up (years) LAVV replacement f 1 (5 y) 1 (1 wk) a The primary reason for re-repair. b One required left atrioventricular valve replacement (see LAVV replacement f ). c One patient with coexisting moderately severe left ventricular outflow tract obstruction. d Two for isolated severe left atrioventricular valve regurgitation. e One for left atrioventricular valve stenosis. f During follow-up. g With or without placement of a flexible ring. AVSD atrioventricular septal defect; CS coronary sinus; LAVV left atrioventricular valve; NA not applicable; RAVV right atrioventricular valve.

4 Ann Thorac Surg GATZOULIS ET AL 1999;67: PARTIAL AVSD IN THE ADULT 507 Table 5. Mortality (8 of 50 Patients) Sex Age at Operation (y) Type of Operation Previous Operation Postoperative (y) Rhythm Mode of Death M 26 AVSD repair Sinus Sudden cardiac. Went to a party 20 days postoperatively (11 days after discharge); died next day. Post mortem showed excellent repair. M 37 AVSD repair AFib Sudden cardiac; post mortem showed mild LVOT and LAVV stenoses. M 39 AVSD repair... 1 Atrial flutter Acute pulmonary embolus. Good leftside hemodynamics, but previous noncompliance with anticoagulants. M 59 AVSD repair, CABG... 3 Sinus Bacterial endocarditis. F 22 Residual ASD (3.7:1) closure Yes 6 AFib Congestive heart failure; LAVVI moderate/severe. F 60 AVSD repair Sinus Congestive heart failure, poor LV function. M 35 LAVVI, repair Yes 16 Sinus Noncardiac malignancy. M 41 AVSD repair Sinus Traffic accident Afib atrial fibrillation; ASD atrial septal defect; AVSD atrioventricular septal defect; CABG coronary artery bypass graft; LAVV left atrioventricular valve; LAVVI left atrioventricular valve incompetence; LV left ventricular; LVOT left ventricular outflow tract. There were no intraoperative deaths. The postoperative stay ranged from 4 to 37 days (mean, days). Transient postoperative atrial arrhythmias developed in 8 patients who had been in sinus rhythm preoperatively. Pacemaker insertion for complete AV block was required in 2 patients, 1 week and 1 month after operation, respectively. One patient with postoperative atrial fibrillation had a transient hemiparesis 2 weeks after repair. PREVIOUS AVSD REPAIR. Demographics, indications, and type of operation performed in this subgroup of patients with previous AVSD repair in childhood are shown in Table 3. There were no hospital deaths. Four (36%) patients required LAVV replacement. Postoperatively, 4 of the 11 patients had new atrial arrhythmias, which were transient in 3 cases. Follow-Up (All Patients) Six patients have died of cardiac causes (Table 5): 2 of sudden death, 2 of congestive heart failure, 1 of pulmonary embolus, and 1 of endocarditis. One patient (the first patient listed in Table 5) suffered sudden cardiac death at home on the 21st postoperative day, giving an early mortality (within 1 month after operation) of 2%. Preoperative pulmonary arterial pressure (systolic) was greater than 50 mm Hg in 4 of the 6 patients who died. Four other patients with pulmonary arterial systolic pressure greater than 50 mm Hg remain alive. Overall, preoperative systolic pulmonary arterial pressure was significantly higher in those who died ( mm Hg) than in survivors ( mm Hg; p 0.03). Two patients died from noncardiac causes (Table 5). Actuarial survival for the 39 patients who underwent primary repair is 89% and 86% at 5 and 10 years, respectively (Fig 1). Forty-two patients were alive in 1997 when this study was conducted. Median follow-up is 7 years (range, 6 months to 20 years, mean years). Thirty-two of them were in New York Heart Association (NYHA) class I, 7 in class II, and 3 in class III. NYHA class improved in 26 of the 32 patients with symptoms before operation ( p 0.004). Of the 42 patients, 36 replied to the written questionnaire; the remaining 6 were assessed at recent TCCCA clinic visits. Six of the 10 patients who were asymptomatic before primary repair reported improved functional capacity after the operation. Recent echocardiographic data are available for 22 patients: LAVV regurgitation is moderate to severe in 3 (with moderate left ventricular dysfunction in 2). All 3 had severe preoperative LAVV regurgitation. No patient has LAVV stenosis. Two patients have mild to moderate LVOT obstruction. ARRHYTHMIA. Only 1 of the 11 patients with preoperative arrhythmia converted to sinus rhythm; this occurred 2 months after repair, and there has been no further Fig 1. Kaplan-Meier actuarial curve for 39 adults who underwent primary repair of partial atrioventricular septal defect (AVSD). Calculated survival is 89% at 5 years and 86% at 10 years.

5 508 GATZOULIS ET AL Ann Thorac Surg PARTIAL AVSD IN THE ADULT 1999;67: evidence of arrhythmia during 5 years of follow-up. Five patients experienced new atrial arrhythmias (3 atrial flutter and 2 atrial fibrillation) at a mean of years after AVSD repair. Overall, atrial arrhythmias were noted in 22% of the patients preoperatively, and in 30% at the latest follow-up. One patient required pacemaker insertion for complete AV block, which developed spontaneously 5 years after repair. No patient had cardiovascular events such as thromboembolism, stroke, or transient ischemic attacks late after repair. REOPERATION. Two patients required reoperation for severe LAVV regurgitation. The first, who had undergone previous AVSD repair, required LAVV replacement 1 week after a failed LAVV repair; the second underwent elective LAVV replacement 5 years after primary AVSD repair. In addition, 1 patient elected for replacement of a mechanical LAVV prosthesis 9 years after initial implantation, as a result of a manufacturer s recall for risk of strut fracture. Cox regression analysis failed to identify any clinical, surgical, or hemodynamic variables predictive of outcome (death, need for reoperation, or symptomatic improvement). Comment In agreement with recent reports by Bergin and associates [10] and Burke and colleagues [11], perioperative mortality and morbidity in our series of 50 adults undergoing operation for partial AVSD is low. Although the feasibility and safety of surgical repair are now well established, controversy continues to exist as to the optimal management of the adult with an ASD. Partial AVSDs share the same hemodynamic burden of chronic right heart volume overload with secundum ASDs, allowing for certain parallels to be drawn [12 14]. It has been suggested that medical management may be preferable to operation for adults with secundum ASDs [15, 16]. We [17] and others [18] do not share this view. Case reports describing occasional patients with secundum or primum ASDs who remain asymptomatic, surviving into old age without an operation, should be seen as the exception and not the rule. Somerville, in an early report [19], recognized that patients with partial AVSDs experience symptoms with increasing frequency after the third decade of life. All large surgical series [10, 11], including ours, clearly demonstrate that these patients are symptomatic by early to mid-adulthood. Indeed, three quarters of our patients who underwent primary repair of partial AVSD were symptomatic at presentation (mean age, 37.7 years), with impaired exercise capacity and exertional dyspnea being by far the most common manifestations. Furthermore, in 6 of the 10 patients thought to be asymptomatic before operation, functional capacity improved postoperatively. It is evident that patients with ASD who have not had an operation may consider themselves asymptomatic because of chronic adaptation of their lifestyle to subtle disability. The majority of our patients improved their exercise capacity after surgical intervention, regardless of age. Ideally, formal studies of cardiopulmonary status should be performed before and after operation and this is our current policy. Long-term survival after repair of partial AVSD is good [10, 11]. Actuarial survival in our series was 89% and 86%, 5 and 10 years after repair, respectively. No direct relationship to type of repair was found. These findings resemble those from pediatric series with repaired partial AVSDs [14, 20], suggesting a relationship between late death (and its low incidence) to the modified history, rather than to the surgical repair per se. The absence of associated pulmonary vascular disease in patients with partial AVSD as a whole may also account for the improved late survival over that of patients having the complete form. However, when pulmonary arterial pressure was elevated (for whatever reason), there was a trend toward an increased incidence of late cardiac death. This is in agreement with the recent report from the Mayo Clinic [10] and a much earlier report by Mattila and coworkers [21], both of which showed a relationship between preoperative pulmonary arterial hypertension and postoperative mortality. As with complete AVSDs [22], LAVV regurgitation is the most common cause of late morbidity after primary repair. Reoperation rates for LAVV regurgitation have varied between 6% and 18% in adult and pediatric series respectively [7, 10, 14]. Left atrioventricular valve replacement is often required under these circumstances, as was the case in our study. Whether routine LAVV valvuloplasty (cleft closure) during primary repair in adult patients with partial AVSD will decrease the need for reoperation remains unclear. It has been suggested [23] that routine valvuloplasty may not be required in such patients because of the tendency of the LAVV leaflets to thicken, which in turn leads to spontaneous closure of the cleft or indeed may not even be feasible because of calcification of the LAVV annulus. In our series, this did not appear to be so. Elective cleft repair of the LAVV was performed in 37 of the 39 patients (during primary repair) to improve valve competence, which (at 95%) is a more frequent rate than in the two recent reports from Burke and associates [11] and Bergin and colleagues [10] (30% and 74%, respectively). Surgical preference has almost certainly played a role. This may also be, at least in part, the result of the relatively younger age of our patients (mean age at primary repair, 37.7 years), compared with those in the series by Burke and coworkers [11] (42 years) and Bergin and associates [10] (51 years). Earlier age at operation for partial AVSDs in adults may facilitate LAVV repair over replacement, with all the potential benefits. Standard interrupted suture technique between the bridging leaflets was used to close the so-called LAVV cleft. In some cases, additional annuloplasty, with or without insertion of a flexible ring, is required to reduce the functional orifice of the LAVV to normal levels. Right atrioventricular valve annuloplasty was also required in 5 patients. Two recent reports of pediatric series [22, 24] have suggested that incomplete closure of the cleft is

6 Ann Thorac Surg GATZOULIS ET AL 1999;67: PARTIAL AVSD IN THE ADULT 509 associated with a higher risk of reoperation for LAVV regurgitation. In our adult series, only 1 patient from the primary repair group needed LAVV replacement; this patient required extensive reconstruction of the LVOT for moderately severe LVOT obstruction. In contrast, 5 patients from the reoperation group required valve replacement. Degeneration of the leaflets, annular dilatation with thickening, and disruption of the chordae led to severe valvular dysfunction that was deemed unsuitable for repair in 2 of the 6 patients in whom the primary indication for reoperation was LAVV regurgitation. Elective LAVV replacement as opposed to repair was performed in 2 other patients undergoing reoperations for LVOT obstruction or LAVV stenosis. The fifth patient required LAVV replacement 1 week after initial repair, for acute valvar failure. With newer reparative techniques incorporating augmentation of the leaflets with autologous pericardium and further plasty of the chordae [6], some of these valves may be amenable to repair. This is speculative, however. In contrast to our usual surgical practice of repairing, whenever possible, the LAVV [6], none of the 5 patients from the reoperation group who required LAVV replacement had had LAVV repair during their initial, childhood operations. These patients may have selected themselves out of a pediatric cohort and presented to the TCCCA for LAVV reoperation for this reason. However, any common underlying factors in the subgroup with previous repair in childhood are unknown, as not all of them had their initial operation at The Hospital for Sick Children, Toronto. Comparisons between these two groups within our study are therefore inappropriate. We have included patients with previous, childhood operations in our study to report our entire experience with partial AVSD operation in the adult, and to illustrate common problems such as LAVV failure, LVOT obstruction, and residual intracardiac shunts [6 9, 14, 25] encountered in our adult congenital heart program. These late complications required surgical attention at a mean of 22 years after initial repair, emphasizing the need for lifelong and continuous follow-up of patients with partial AVSD. Only 1 (2.6%) of the 39 patients who underwent primary repair required reoperation (LAVV replacement, 5 years later) during a mean follow-up of 9 years. Reoperations for LAVV problems in the series reported by Burke and coworkers [11] and Bergin and associates [10] ranged between 6% and 9.7% respectively (mean follow-up of 5.3 and 14 years). Whether reoperation rates in the adult with a partial AVSD relate to age at primary repair or length of follow-up and surgical preferences remains to be seen. Clearly, further longitudinal data from these three and other studies are required before any conclusions about the optimal management of LAVV in adults with partial AVSD can be drawn. Complete AV block and atrial arrhythmias remain a cause of morbidity for these patients. The former is part of the natural history of AVSDs; indeed, it occurred spontaneously in 1 of our patients 5 years after repair. The patient, who required a pacemaker 1 week after operation, had a calcified LAVV requiring replacement, and underwent extensive reconstruction of the LVOT for subaortic stenosis. Better understanding of the surgical anatomy of partial AVSD [1 3] has already reduced the incidence of postoperative complete AV block [6], and the advent of new-generation pacemakers is expected to reduce morbidity further in this respect. Atrial arrhythmias remain problematic. One fifth of our patients had established atrial flutter or fibrillation at presentation. Only 1 of them converted to sinus rhythm after primary repair. Furthermore, 5 patients experienced sustained atrial tachyarrhythmia beyond the immediate postoperative period. Older patients are more likely to present with arrhythmia, suggesting that earlier AVSD repair may reduce this risk. However, even younger patients may have atrial flutter or fibrillation after primary AVSD repair, raising the question whether adults with partial AVSD should be considered for additional surgical procedures targeting arrhythmias at the time of initial repair. We were unable to identify predictors of clinical outcome, perhaps because of the relatively small number of cases and the limitations of a retrospective study. Nevertheless, we have shown that operation for partial AVSDs in adults with substantial left-to-right atrial shunts can be performed with low mortality and morbidity. Long-term survival was good and the probability of sustained symptomatic improvement, high. Late complications, although uncommon, necessitate continuous, lifelong follow-up. Elective LAVV valvuloplasty during primary repair may confer a better outcome by reducing the need for reoperation; long-term data are required, however, before definitive recommendations can be made. In conclusion, low operative risk and excellent long-term results support our policy of elective repair of partial AVSDs in adults. We acknowledge Drs Rebeyka and Van Arsdell, who performed AVSD repair in some of the patients from this series. We also thank our colleagues from The Hospital for Sick Children, Toronto, for their continuing support of the TCCCA program. This paper was prepared with the assistance of Editorial Services, The Hospital for Sick Children. Doctor Gatzoulis was supported in part by a 1997 University of Toronto, Department of Medicine, Postgraduate Fellowship Award. References 1. Anderson RH, Zuberbuhler JR, Penkoske PA, Neches WH. Of clefts, commissures, and things. J Thorac Cardiovasc Surg 1985;90: Penkoske PA, Neches WH, Anderson RH, Zuberbuhler JR. Further observations on the morphology of atrioventricular septal defects. J Thorac Cardiovasc Surg 1985;90: Ebels T, Anderson RH, Devine WA, Debich DE, Penkoske PA, Zuberbuhler JR. Anomalies of the left atrioventricular valve and related ventricular septal morphology in atrioventricular septal defects. J Thorac Cardiovasc Surg 1990;99: Kirklin JW, Barratt-Boyes BG. Cardiac surgery. New York: Churchill Livingstone, 1993: Hanley FL, Fenton CN, Jonas RA, et al. Surgical repair of complete atrioventricular septal defects in infants. Twentyyear trends. J Thorac Cardiovasc Surg 1993;106:

7 510 GATZOULIS ET AL Ann Thorac Surg PARTIAL AVSD IN THE ADULT 1999;67: Najm HK, Williams WG, Chuaratanaphong S, Watzka SB, Coles JG, Freedom RM. Primum atrial septal defect in children: early results, risk factors, and freedom from reoperation. Ann Thorac Surg 1998;66: Abbruzzese PA, Napoleone A, Bini RM, Annechino FP, Merlo M, Parenzan L. Late left atrioventricular valve insufficiency after repair of partial atrioventricular septal defects: anatomical and surgical determinates. Ann Thorac Surg 1990;49: Najm HK, Coles JG, Endo M, et al. Complete atrioventricular septal defects: results of repair, risk factors, and freedom from reoperation. Circulation 1997;96(Suppl 2):II Alexi-Mesjishvili V, Hetzer R, Dahnert I, Weng Y, Lange PE. Results of left atrioventricular valve reconstruction after previous correction of atrioventricular septal defects. Eur J Cardiovasc Surg 1997;12: Bergin ML, Warnes CA, Tajik AJ, Danielson GK. Partial atrioventricular canal defect: long-term follow-up after initial repair in patients 40 years old. J Am Coll Cardiol 1995; 25: Burke RP, Horvath K, Landzberg M, Hyde P, Collins JJ, Cohn LH. Long-term follow-up after surgical repair of ostium primum atrial septal defects in adults. J Am Coll Cardiol 1996;27: Murphy JG, Gersh BJ, McGoon MD, et al. Long-term outcome after surgical repair of isolated atrial septal defect: follow-up at 27 to 32 years. N Engl J Med 1990;323: Barnett MG, Chopra PS, Young WP. Long-term follow-up of partial atrioventricular septal defect repair in adults. Chest 1988;94: Permut LC, Mehta V. Late results and reoperation after repair of complete and partial atrioventricular canal defect. Semin Thorac Cardiovasc Surg 1997;9: Ward C. Secundum atrial septal defect: routine surgical treatment is not of proven benefit. Br Heart J 1994;71: Shah D, Azhar M, Oakley CM, Cleland JGF, Nihoyannopoulos P. Natural history of secundum atrial septal defect in adults after medical or surgical treatment: a historical prospective study. Br Heart J 1994;71: Gatzoulis MA, Redington AN, Somerville J, Shore DF. Should atrial septal defects in adults be closed? Ann Thorac Surg 1996;61: Konstantinides S, Geibel A, Olschewski M, et al. A comparison of surgical and medical therapy for atrial septal defect in adults. N Engl J Med 1995;333: Somerville J. Ostium primum defect: factors causing deterioration in the history. Br Heart J 1965;27: Pillai R, Ho SY, Anderson RH, Lincoln C. Ostium primum atrioventricular septal defect: an anatomical and surgical review. Ann Thorac Surg 1986;41: Mattila S, Merikallio E, Tala P. ASD in patients over 40 years of age. Scand J Thorac Cardiovasc Surg 1979;13: Bando K, Turrentine MW, Sun K, et al. Surgical management of complete atrioventricular septal defects. J Thorac Cardiovasc Surg 1995;110: Warnes CA, Shugoll GI, Wallace RB, Roberts WC. Atrioventricular septal defect (primum atrial septal defect) with prolonged survival (despite severe mitral regurgitation and pulmonary hypertension) and associated cardiac calcification (mitral annulus, coronary artery, and pulmonary trunk). Am J Cardiol 1984;54: Nacker CL, Mavroudis C, Alboliras ET, Zales VR. Repair of complete atrioventricular canal defects: results with the two-patched technique. Ann Thorac Surg 1995;60: Van Arsdell GS, Williams WG, Boutin C, et al. Subaortic stenosis in the spectrum of atrioventricular septal defects: solutions may be complex and palliative. J Thorac Cardiovasc Surg 1995;110: