Author + information
- Received October 28, 1997
- Revision received May 6, 1998
- Accepted May 20, 1998
- Published online September 1, 1998.
- Nigel J Wilson, MB∗,* (, )
- Patricia M Clarkson, MB∗,
- Brian G Barratt-Boyes, ChM†,
- A.Louise Calder, MD∗,
- Ralph M.L Whitlock, MB‡,
- Ronald N Easthope, MB§ and
- John M Neutze, MD∗
- ↵*Address for correspondence: Dr. Nigel Wilson, Cardiology Department, Green Lane Hospital, Private Bag 92 189, Auckland 1030, New Zealand
Objectives. This study examines the late outcome in patients with simple transposition of the great arteries (TGA) after a Mustard operation.
Background. Continuing medical follow-up for patients after the Mustard procedure, now extending to three decades, is required. The quality of life of adult survivors has not been well documented.
Methods. Survival and quality of life among 113 hospital survivors of the Mustard operation performed for simple TGA between 1964 and 1982 were assessed by medical review and a lifestyle questionnaire. The incidence of right ventricular failure and echocardiographic right ventricular dysfunction (RVD) were determined. A measure of lifestyle, the ability index, was determined.
Results. Actuarial survival was 90%, 80%, and 80% at 10, 20, and 28 years, respectively, with 76% of survivors being New York Heart Association class 1. Sudden death, with an incidence of 7% without identifiable risk factors, was the most common cause of late demise. RVD was identified in 18% of patients who had echocardiography, but there was right ventricular failure in only two patients. Seventy-five percent of current survivors lead a normal life, 20% have some symptoms or lifestyle modification, and 5% are unable to work.
Conclusions. The survival of patients to 28 years with the Mustard repair has been good. Late sudden death is the most worrisome feature. There is a 97% freedom from right ventricular failure to date. The quality of life of late survivors is good, most achieving a normal level of education and employment.
Sequelae after the Mustard operation previously identified include sudden death, arrhythmia, and right ventricular failure (1). Follow-up now extends to 30 years in some patients (2). However, the frequency of right ventricular dysfunction (RVD) and right ventricular failure are poorly defined. There is little information on the quality of life of survivors reaching adulthood. This study reviews the outcome, medical status, and quality of life of late survivors who are now young adults.
All 113 consecutive hospital (30-d) survivors of the Mustard operation for simple transposition of the great arteries (TGA) at Green Lane Hospital between 1964 and 1982 were reviewed. Simple TGA is defined as a congenital cardiac defect with situs solitus, atrioventricular concordant, and ventriculoarterial discordant connections and essentially intact ventricular septum. Patients with TGA and a small (<4-mm diameter) ventricular septal defect (VSD) or left ventricular outflow tract obstruction (LVOTO) or a patent ductus arteriosus (PDA) were included (3). Those with a moderate or large VSD were excluded.
The Mustard atrial repair was performed using pericardium as the baffle. In the first 60 patients a conventional right oblique atriotomy was used, but from 1975 onwards this was replaced by a “VY” plasty designed to enlarge the pulmonary venous compartment without the use of patch enlargement of the right atrial wall (3). Specifically, the right atrium was opened using a V-shaped incision that, after baffle insertion was completed, was enlarged backwards between the ostia of the right upper and lower pulmonary veins, converting it into a Y. The atriotomy was closed horizontally. Initially, cardiopulmonary bypass with mild hypothermia and bicaval cannulation was used as well. Since 1970, profound hypothermia circulatory arrest techniques were adopted and a single venous cannula was inserted in the right atrial appendage. The baffle suture line was placed into the right atrial wall well clear of the superior vena caval ostium and the sinus node (4).
All 113 patients have had serial cardiac review, which included cardiac catheterization in 65% at an average age of 4.3 years postrepair (3–6). A review of current survivors was undertaken between 1992 and 1994 by cardiologists in 91%, other physicians in 5%, and by phone call only in 4% of patients. Causes of late death were assessed from the clinical history and autopsy data. Deaths that were sudden and unexpected were presumed to be due to arrhythmia.
Assessment of right ventricular function
RVD was diagnosed when there was abnormal echocardiographic dilatation of the right ventricle with unequivocally decreased systolic function. Videotapes were reviewed by one investigator (NJW) in 76% of studies and by outside cardiologists in 24%. In patients without echocardiography, assessment of RVD was based on symptomatology, serial radiographic heart size, and cineangiography. Right ventricular failure was diagnosed when there was symptomatic (New York Heart Association [NYHA] class 3 or 4) RVD with pulmonary venous congestion; this definition is essentially the same as that used by Senning’s group (7).
Serial standard 12-lead electrocardiograms were used to assess rhythm and conduction after hospital discharge. Patients in slow junctional rhythm were classified as having junctional rhythm even if at a later date, or during exercise, they exhibited sinus rhythm. Episodes of paroxysmal supraventricular tachycardia and atrial flutter occurring late postoperatively were noted, but when the usual rhythm was sinus, patients were not considered to have lost sinus rhythm.
A questionnaire elicited information about symptoms, medication, recreational activities, education, employment, personal relationships, parenthood, life insurance, and motor driving license. These factors allowed assessment of an ability index (8). Ability index grade 1 is defined as leading a normal life with full-time work or education. Women are able to manage pregnancy. Grade 2 indicates those able to do part-time work but life is modified by symptoms. Grade 3 identifies those unable to work with noticeable limitation of activities, and grade 4 those with extreme limitation, dependent on others for daily living.
The probability of survival was estimated by the Kaplan-Meier method (9). Variables for values that were statistically significant by univariate analysis were evaluated by multivariate analysis using the Cox proportional hazards model or multiple logistic regression as appropriate (10). Variables for sudden death were age at operation, type of operation (Mustard or VY plasty), reoperation, last documented rhythm, and RVD. Variables for elevated pulmonary vascular resistance index (PVRI) were presence or absence of associated defects (intact ventricular septum and/or LVOTO vs. PDA and small VSD), previous atrial septectomy or balloon septostomy, age at operation, type of operation, and gender.
There were 113 hospital survivors. Prior neonatal creation of an atrial defect had been undertaken in 112 patients (20 surgical septectomy, 92 balloon septostomy). Forty patients had one or more associated anatomical defects before the Mustard repair: 21 small VSDs (four had suture closure at Mustard repair), 13 LVOTO, one coarctation of the aorta, two partial anomalous pulmonary venous drainage, and three large aortopulmonary collaterals. The mean age at repair is detailed in Table 1. Nine patients underwent reoperation 1–7 years post-Mustard repair: six for superior vena caval obstruction, two for a significant atrial baffle leak, and one for coarctation of the aorta.
Patients were followed for up to 28 years postrepair. Actuarial survival was 93%, 90%, 84%, 80%, 80% at 5, 10, 15, 20, 28 years, respectively (Fig. 1). Forty patients were alive over 20 years postrepair. The age of the 94 current survivors was 12–32 (mean 20.2) years. Survival for the standard Mustard and for VY plasty are compared in Figure 1. There were 19 deaths with 14 autopsies (Table 2). Seven deaths predominantly related to baffle obstruction occurred within 2 years of repair. Twelve patients died in a later phase between 6 and 19 years postrepair, seven suddenly. The incidence of sudden unexpected death was 7% (95% confidence limits 4–14%). There was a single phase hazard function for sudden death since repair (1). Freedom from sudden death is 92% at 19 years when the last event occurred (Fig. 2). The last recorded rhythm prior to death in these patients was junctional in six and sinus in one; one had complete heart block. No preoperative or late variables entered were significantly associated with late sudden death (10). The coronary artery pattern was normal for TGA in seven of the eight patients in whom this was known from autopsy (three patients) or cineangiography (four patients), without evidence of an interarterial course as a possible etiology of sudden death.
Six patients, known to be alive 10–15 years postrepair, were lost to current follow-up. Seventy-six percent of survivors were NYHA class 1, 22% class 2, 1% class 3, and 1% unknown. No individual was class 4.
Twenty-seven local patients, at a mean age of 18.3 years, underwent an exercise test using the Bruce protocol. The mean time exercised was 11.3 ± 2.3 minutes, range 6–15 minutes, median time 12 minutes. Five patients did not reach the tenth centile for exercise time for age- and sex-matched normal individuals (11). The mean maximal heart rate was 160 ± 26 bpm, range 94–205 bpm, with 12 patients reaching 85% of the predicted maximal heart rate for age. The mean systolic blood pressure rise with exercise was 42 ± 18 mm Hg; in only three did it fail to increase as expected (12). Exercise parameters had low correlation coefficients with radionuclide right ventricular ejection fraction (RVEF) (exercise time r = 0.21, p > 0.1; maximal heart rate r = 0.22, p > 0.1; maximal systolic blood pressure rise r = 0.25, p > 0.1).
Of the 78 patients with current echocardiography, 60% had no or trivial tricuspid regurgitation, 34% mild, and 6% moderate tricuspid regurgitation. Three patients had mild aortic regurgitation. One patient had mild pulmonary venous obstruction. Usually the pulmonary venous atrium was larger than the systemic venous atrium, but significant systemic venous obstruction was not currently diagnosed in any patient. Left (pulmonary) ventricular function was well maintained in all but one patient (case 18, Table 2).
Radionuclide angiography (13)was performed in 26 patients, mean age 18.7 years, by first pass and gated scans. The average RVEF was 44 ± 9%, range 23–63%. Neither later age at repair nor longer duration of follow-up predicted decreasing RVEF. The average left ventricular ejection fraction (LVEF) was 59 ± 7%, range 39–76%.
Right ventricular dysfunction
RVD was present in 14 of 78 patients (18%) who had echocardiography. The incidence of RVD diagnosed by the investigator (8/49) as compared with outside cardiologists (6/19) was similar (p = 0.09). Their findings are detailed in Table 3, illustrating poor correlation of variables. There remain 35 patients without echocardiography, including 18 of the 19 patients who died. Only one of these latter patients had RVD with impaired right ventricular contractility on cineangiography postrepair (case 4, Table 2). Thus RVD without failure was identified in 13% of the entire cohort.
Right ventricular failure
Right ventricular failure occurred in two patients. There was a 97% (95% confidence limits 95–99%) freedom from right ventricular failure by 18 years, with no further occurrence in patients with 18–28-year follow-up. The first patient (case 15, Tables 2 and 4) ⇓was the only one to have a right ventriculotomy for VSD closure. At 20 months of age there was a high PVRI, which fell with isoproterenol (14). The right ventricle was dilated with moderate tricuspid regurgitation. Right ventricular failure was diagnosed from that time. Progress was poor, with persistent congestive cardiac failure and the development of a fixed high PVRI. He died at 14 years of age. The second patient (case 18, Table 2) remained well until the development of pulmonary edema and right ventricular dilatation at an age of 19 years. There was moderate tricuspid regurgitation, and radionuclide angiography showed an RVEF of 19% and an LVEF of 39%. Pulmonary artery banding was undertaken, but left ventricular function deteriorated further precluding an arterial switch procedure. At autopsy there was grade 3 pulmonary vascular disease (15).
Elevated pulmonary vascular resistance
Fourteen (19%) of the 74 patients who underwent postrepair cardiac catheterization had a measured resting PVRI of ≥4 u.m2, 1.5–7.5 years postrepair (Table 4). An anatomical cause was associated with elevated PVRI in cases 16, 34, 35, and 39 (Table 4). Eight of the other 10 patients underwent repair after 2 years of age. Older age at operation and palliation by surgical septectomy were risk factors for postrepair elevated PVRI on univariate analysis, but on multivariate analysis only older age at operation remained significant. Four of the eight patients with a resting PVRI >8 u.m2have died (Table 4). The four survivors are stable 8–24 years later with mild desaturation and exercise limitation. The remaining six patients who had a resting PVRI of 4.0–7.7 u.m2are NYHA class 1 or 2, 10–23 years after postrepair catheterization. In addition, five patients who died (Table 2)without a calculated PVRI had grade II–IV pulmonary vascular disease at autopsy. Overall, nine patients (8%) were considered to have severe pulmonary vascular changes as judged by PVRI >8 u.m2despite vasodilators and/or grade III–IV changes at autopsy (Table 2) (3).
Left ventricular outflow tract (pulmonary) obstruction
LVOTO in 13 patients at the time of Mustard repair did not influence late outcome. In seven of 11 patients with angiographic subvalvar obstruction, resection abolished the gradient in six, and left a mild gradient in one. In four, the unresected LVOTO was mild and remained mild at late follow-up. Two patients had pulmonary valve stenosis. One was left with mild obstruction after valvotomy; the other, whose lesion was not appreciated at the time of repair, had a gradient of 120 mm Hg. Two late deaths occurred in patients with LVOTO from other causes (Table 2, cases 11 and 15). All survivors are NYHA class 1. Two other patients without LVOTO at the time of the Mustard repair have acquired a hypertrophic cardiomyopathy type obstruction with gradients of 50–60 mm Hg.
There has been a progressive loss of sinus rhythm as shown by the resting 12-lead electrocardiogram (Figure 2). Patients in junctional rhythm at rest usually developed an atrial or sinus rhythm with exercise. No patient developed a supraventricular or ventricular arrhythmia during exercise testing. Episodes of supraventricular tachycardia occurred in 12 patients. In only one of the five patients with atrial flutter did this become persistent. Atrial pacemakers were inserted in four patients because of bradycardia associated with decreasing exercise tolerance in two and sinus pauses >3.5 s on Holter monitoring in two. A patient with a functioning ventricular pacemaker inserted 2 years earlier because of persistent atrial flutter, variable atrioventricular block, and syncope died suddenly (case 14, Table 2).
Quality of life of current survivors
Lifestyle information was available in 88 of the 94 current survivors. Forty-nine percent of patients reported nonischemic chest pains, and in two the pains were frequent and interfered with lifestyle. Thirty-one percent experienced palpitations and 6% an episode of syncope or presyncope. Overall, 67% of patients had at least one of these symptoms, albeit often mild, with little influence on lifestyle. Eleven percent of patients at current follow-up were receiving cardiac and 11% noncardiac medication. Six patients had a cerebrovascular accident with late sequelae, five prior to the Mustard repair and one 21 years postoperatively; four have motor and two cognitive impairment. Five patients, including one of the above and two with neonatal preoperative seizures, had epilepsy late postoperatively and receive antiepileptic medication. One other patient with neonatal convulsions has a movement disorder.
All patients surviving beyond age 5 years have attended school. The educational level at high school was age-appropriate in 88%, although 6% reported speech or learning difficulties that required special tuition. The educational level for 12% of patients was 1–2 years below age-appropriate level. Thus, 18% had some educational difficulties. Twenty-eight percent of school leavers (16/57) had at least 1 year of tertiary education. This compares with 27% of the New Zealand population between 18 and 24 years enrolled in a tertiary institution in 1993 (16). Forty-one (72%) of the 57 school leavers were in full-time employment, including 10 mothers, all of whom were previously in paid employment, four (7%) were in part-time employment, three (5%) were full-time students, and nine (16%) were unemployed, seven of these having had previous employment. Two patients (4%) were regarded as unemployable. The percentage of the school leavers who were unemployed (16%) compares with the New Zealand average of 9% in the early- to mid-1990s.
Eighty percent of survivors reported participating in sporting or recreational activities, with 91% participation in the 12–16-year age group and 71% participation in those ≥17 years. Thirteen of 25 applications for life insurance were accepted, albeit at an increased rate in eight. The usual local acceptance rate is 99%. Eighty-two percent of those ≥16 years have a driving license, one being declined because of epilepsy.
Of those ≥20 years, 56% (23/41) were in a married or stable relationship. Eleven women had 20 pregnancies resulting in 18 live births. There was no overt cardiac deterioration during or after pregnancy. The outcome in nine women has been detailed (17). Four men have fathered four children. None of the 22 offspring has a congenital heart defect.
The above aspects of lifestyle, categorized by the ability index (8), show that 75% of current survivors have ability index 1, 20% ability index 2, and 5% ability index 3. None had ability index 4. Causes for classification as ability index class 2 or 3 are detailed in Table 5.
Survival and quality of life are good in young adults after the Mustard repair for simple TGA. The most important complication is an ongoing risk of late unexpected death, with a reported incidence of 2–8% (2,18–20). In our series there were eight late sudden deaths without identifiable risk factors and a freedom from this event of 91% at 28 years. Atrial flutter has been suggested as a risk factor for late sudden death (21). However, in a multi-institutional study, supraventricular tachycardia, including atrial flutter, was not a risk factor unless it was the dominant rhythm (18). Atrial flutter was recognized in 4% of patients in the present study, lower than in most other series (2,18,21). It appears that risk of sudden death is an immutable part of atrial baffle repair. Pacing was seldom used and did not prevent sudden death in our series or others (18).
This study indicates that the right ventricle is capable of performing as the systemic pumping chamber for at least three decades. Only two patients have developed right ventricular failure and in each coexistent factors may have been contributory: both had pulmonary vascular disease along with a right ventriculotomy in one and impaired left ventricular function in the other. The incidence of late right ventricular failure is low, with no time interval since repair identified as influencing its onset. These data are almost identical to those for the Senning repair, with a 4% incidence of right ventricular failure for those with an intact ventricular septum (7). An associated large VSD requiring patch closure has been shown to be a risk factor for atrial baffle repair (3), late death (2,7,22), and right ventricular failure (7). Of the 33 such patients who underwent a Mustard repair from our institution there are only 10 current survivors. An incision in the right ventricle and placement of a VSD patch are potential confounding variables for long-term right ventricular function. For these reasons we excluded those with a large VSD from this study.
There are no uniform definitions of RVD. We judged this by qualitative review of echocardiograms. An echocardiographic and radionuclide comparative study detailed right ventricular function in a subset of 22 of these patients (13), but we were unable to use planimetered right ventricular volume or RVEF routinely as many examinations we reviewed from outside institutions lacked sufficient attention to endocardial border definition. There is clearly a spectrum in these patients between the “normal” systemic right ventricle exhibiting physiologically mild dilatation and more significant degrees of dilatation with abnormal systolic function. The majority of patients with echocardiographic RVD were NYHA class 1, and in four of 14 both dilatation and dyskinesia were mild. It is uncertain whether those right ventricles we have labeled dysfunctional will ultimately fail or whether, despite appearances, they too represent a physiological response to functioning as the systemic ventricle. RVD (1,13,23,24)is not synonymous with right ventricular failure (7). In the only reported longitudinal study, patients showed no decline in right ventricular function over a mean follow-up time of 4.3 years (24). The Toronto data for simple TGA showed 10% of patients had angiographically significantly impaired right ventricular contractility (1). In the present study, radionuclide angiography did not show lower RVEF with longer follow-up. These data suggest that undue pessimism is inappropriate regarding the long-term fate of the right ventricle after atrial baffle repair for simple TGA.
The relationship between right ventricular myocardial function and tricuspid regurgitation is ill-defined. On echocardiography moderate tricuspid regurgitation was associated with RVD in four, and with good right ventricular function in one patient. Mild nonprogressive tricuspid regurgitation was frequently seen. No patient had severe tricuspid regurgitation. A minority of our patients had abnormal exercise parameters, similar to other series (24), but the reduction of exercise capacity was insufficient to affect daily living, the majority of patients being NYHA class 1.
Elevated postrepair PVRI or pulmonary vascular disease was identified in 17% of patients, being severe in 8%, and has been encountered in other series (23). Most of our patients have remained remarkably well. While neonatal repair would be expected to reduce important elevation of PVRI, values ≥4 u.m2were found in over 20% of TGA patients following an arterial switch operation, predominantly performed before 2 weeks of age (25).
Survival between neonatal palliation and Mustard repair, and risk factors for repair, have been documented for our and other series (1,3,20). For the entire study period, the hospital mortality was 11% for the Mustard operation, and 4% for the VY plasty from 1975 to 1982 (3). The surgical techniques of placement of the baffle suture line away from the superior vena caval ostium and avoidance of caval purse string with deep hypothermia circulatory arrest may have resulted in more patients in sinus rhythm at late follow-up than other series (1).
The quality of life of survivors has been very good, most achieving normal levels of education, employment, and activity. Women have had successful pregnancies and manage motherhood without problems. These young adults participate in a wide range of sports and recreational activities. The overall quality of life is quite remarkable considering that many are survivors of complex heart surgery performed in the 1960s and 1970s. However, nearly 20% of survivors have experienced educational difficulties. Preoperative neurological events contributed to an excess of motor and learning difficulties at late follow-up. Neonatal repair would be expected to obviate most such events and their long-term sequelae.
This study reviewed all hospital survivors from two decades of surgery commencing in the 1960s, with a longer follow-up than reported by most (1,2,19,22). The incidence of right ventricular failure has been low. Sudden unexpected death, without identifiable risk factors, was the commonest cause of late death. Institutions performing atrial baffle repairs in the 1980s might expect their survivors to fare as well as, if not better than, this study group.
- left ventricular ejection fraction
- left ventricular outflow trace obstruction
- New York Heart Association
- patent ductus arteriosus
- pulmonary vascular resistance index
- right ventricular dysfunction
- right ventricular ejection fraction
- transposition of the great arteries
- ventricular septal defect
- Received October 28, 1997.
- Revision received May 6, 1998.
- Accepted May 20, 1998.
- American College of Cardiology
- Gelatt M,
- Hamilton R.M,
- McCrindle B.W,
- et al.
- Kirklin J.W,
- Barratt-Boyes B.G
- Clarkson P.M,
- Barratt-Boyes B.G,
- Neutze J.M
- Clarkson P.M,
- Neutze J.M,
- Barratt-Boyes B.G,
- Brandt P.W.T
- Clarkson P.M,
- Neutze J.M,
- Wardill J.C,
- Barratt-Boyes B.G
- Siebenmann R,
- von Segesser L,
- Schneider K,
- Schneider J,
- Senning A,
- Turina M
- Warnes C.A,
- Somerville J
- Cox D.R
- Heath D,
- Edwards J.E
- ↵Education statistics of New Zealand 1994. Wellington: Ministry of Education, 1994.
- Clarkson P.M,
- Wilson N.J,
- Neutze J.M,
- North R.A,
- Calder A.L,
- Barratt-Boyes B.G
- Champsaur G.L,
- Sokol D.M,
- Trusler G.A,
- Mustard W.T
- Vetter V.L,
- Tanner C.S,
- Horowitz L.N
- Wong K.Y,
- Venables A.W,
- Kelly M.J,
- Kalff V
- Mantel N,
- Haenszel W