|
|
||||||||
a Division of Congenital Cardiovascular Surgery, University Children's Hospital, University of Zurich, Steinwiesstrasse 75, CH-8032 Zurich, Switzerland
b Statistical Unit, Institute for Social and Preventive Medicine, University of Lausanne, Lausanne, Switzerland
c Division of Pediatric Cardiology, University Children's Hospital, University of Zurich, Zurich, Switzerland
d Division of Pediatric Intensive Care, University Children's Hospital, University of Zurich, Zurich, Switzerland
Received 12 September 2007; received in revised form 20 January 2008; accepted 30 January 2008.
* Corresponding author. Tel.: +41 44 2668020; fax: +41 44 2668021. (Email: ali.dodge-khatami{at}kispi.uzh.ch).
| Abstract |
|---|
|
|
|---|
Key Words: Congenital Subaortic stenosis Reoperation
| 1. Introduction |
|---|
|
|
|---|
| 2. Materials and methods |
|---|
|
|
|---|
|
2.1 Statistical analysis
The endpoint of the study was the time to first reoperation. As potential risk factors, we considered preoperative pressure gradient (continuous), immediate postoperative pressure gradient (continuous), age (continuous), aortic valve insufficiency (grade 0–I vs grade II–IV), mitral valve insufficiency (grade 0 vs grade larger than 0), chromosomal anomalies (absent vs present), surgical technique (isolated membrane resection vs concomitant myectomy), arteria lusoria (absent vs present), and subaortic stenosis (simple vs complex). Strength of association and statistical significance between a risk factor and the endpoint were assessed using a hazard ratio and Cox regression. For the binary risk factors (such as surgical technique or subaortic stenosis), Kaplan–Meier curves were calculated. A p-value smaller than 0.05 was considered statistically significant. Initially, the entire cohort was analyzed (n
= 58), and a Kaplan–Meier freedom from reoperation curve was calculated to distinguish between the complex and simple groups (Fig. 2). As the two groups are otherwise anatomically and physiologically different, the analysis as described above was repeated considering each group separately (simple; n
= 43 and complex; n
= 15). Note that in cases where no event occurred in small subgroups, it was technically not possible to estimate a hazard ratio using Cox regression (see Table 2
).
|
|
| 3. Results |
|---|
|
|
|---|
|
One major postoperative complication included complete heart block requiring a pacemaker in one patient (1.7%). Reoperation was required for recurrent stenosis in 11 patients (19%) at 2.6 years (range 0.3–7.5) after initial surgery. The gradients measured across the left outflow tract indicating the need for reoperation ranged from 32 to 99 mmHg (mean 69 mmHg). Five reoperations were necessary in the simple group (5/43; 12%) and 6 in the complex group (6/15; 40%); Kaplan–Meier estimates of freedom from reoperation at 5 years were 92% for the simple group, and only 32% for the complex group (p = 0.003; Fig. 2 ). Reoperation consisted of repeat membrane resection, with concomitant myectomy in all but two cases, and was achieved without mortality. In two patients, the reoperation consisted of a modified Konno procedure with aortic valve tricuspidization and enlargement annuloplasty. After reoperation, gradients dropped to a mean of 19 mmHg (range 0–80 mmHg). Two patients needed a second reoperation, one of which survived without further restenosis at last follow-up, and the other died as described above.
The statistical relationship between various risk factors and time to first reoperation were analyzed using Cox regression. For the entire cohort, negative findings included preoperative pressure gradient across the LVOT (hazard ratio 1.004 per mmHg; p = 0.73), aortic valve insufficiency (hazard ratio 0.4; p = 0.19), mitral valve insufficiency (hazard ratio 0.4; p = 0.34), chromosomal anomalies (hazard ratio 2.1; p = 0.26), and surgical technique (isolated membrane resection vs concomitant myectomy (hazard ratio 1.7; p = 0.40). Risk factors for reoperation included complex defects (hazard ratio 6.7; p = 0.003; Fig. 2), immediate postoperative gradient (per mmHg; hazard ratio 1.052; p = 0.019), younger age (hazard ratio 0.7 per year; p = 0.012), and the presence of an arteria lusoria (hazard ratio 5.7; p = 0.014). The results of Cox regression with hazard ratios and 95% confidence intervals after analyzing the two groups separately are shown in Table 2. In summary, no variable was an independent risk factor for recurrence of stenosis leading to reoperation, either in the simple lesion group or in the complex stenosis group (Table 2).
In cases of isolated membrane subaortic stenosis (n = 43), concomitant myectomy provided a slightly better freedom from reoperation rate compared to membrane resection only, although this was not significant statistically (hazard ratio = 0.621; p = 0.60) (Fig. 3 ).
|
| 4. Discussion |
|---|
|
|
|---|
When analyzing the group as a whole, we found complex forms of subaortic stenosis, including tunnel-like stenosis, hypoplastic aortic annulus, mitral stenosis, and hypoplastic arch with coarctation, associated with other intracardiac defects, to be a risk factor for reoperation (hazard ratio 6.7; p = 0.003). Similar to our findings, Serraf et al. found tunnel-like stenosis associated with Shone's complex to be a univariate risk factor for operative mortality and recurrence rate [2]. Similarly, the presence of an arteria lusoria (aberrant right subclavian artery) was a risk factor for reoperation (hazard ratio 5.7; p = 0.014). Not surprisingly, all four patients with this anomaly were in the complex group, three of which required a reoperation. In these patients with duct-dependent systemic circulations, perfusion to the descending aorta distributes flow to the left subclavian artery and the aberrant right subclavian artery, and therefore antegrade flow through the aortic valve feeds only the remaining two carotid vessels. In this situation, the steal from antegrade proximal left-sided flow, compared to that with normal upper extremity vascular anatomy, potentially leads to underdevelopment of left-sided structures and multi-level stenosis, including the mitral valve, the left ventricle itself, the subaortic area, the aortic valve, and the ascending aorta and arch, giving the physiological substrate to Shone's complex. The presence of an arteria lusoria is more a surrogate of hypoplastic left-sided structures and hindrance of left-sided forward flow. Probably due to the small sample size, the presence of an arteria lusoria was not a risk factor for reoperation in the complex lesions group, whereas there was only one patient with this anomaly in the simple group who had no reoperation, therefore it was technically not possible to perform the Cox regression in the simple group.
Considering the entire cohort, we found early age at operation to be an independent risk factor for reoperation (hazard ratio of 0.7 per year; p = 0.012). In patients with isolated membrane subaortic stenosis, age at initial surgery was not an independent risk factor for reoperation (hazard ratio = 0.76; p = 0.10). The relatively short time of median follow-up in our series may explain this finding, and longer follow-up could reveal progressive recurrence of the disease. Partially concurring with our findings are the results of Serraf et al. who found that an age younger than 5 years at initial surgery was a univariate risk factor for recurrence, although this was not confirmed in their multivariate analysis [2]. Brauner et al. also found a higher recurrence rate in younger patients [10]. Others advocate the benefits of early correction, stating that earlier removal of the subaortic membrane reduces recurrence and the risk of later aortic valve insufficiency [3,10–12].
Considered as a continuous variable, the preoperative gradient across the LVOT was not a predictive risk factor for later recurrence. Immediate residual postoperative gradient after resection of stenosis was a risk factor for reoperation when the entire cohort was considered (p = 0.019), but not when the complex lesions and isolated membrane resections were analyzed separately. Somewhat similar to our findings, others have found preoperative and early postoperative gradients to be significant predictors of late gradient [3,5,10]. Interestingly, they also found that early postoperative residual gradient was also the strongest predictor of mild or greater AI at mid-term follow-up, and conclude that a radical excision of all pathologic tissue including an aggressive myectomy may reduce the incidence of late AI. However, a higher incidence of complications with this approach is to be feared, namely complete heart block in 3.7% of those undergoing first time surgery for relief of subaortic stenosis, and a staggering 80% (4/5 patients) who underwent redo surgery for recurrent stenosis [3]. Postoperatively, it would seem intuitive that the lowest possible residual gradient or the absence thereof across the left outflow should be the goal of surgery, and would preclude recurrence.
The incidence of recurrence requiring reoperation is a source of controversy, with regards to the initial surgical technique employed. When considering the entire cohort we found no additional benefit from concomitant myectomy to membrane resection in terms of recurrence of stenosis or need for reoperation. When considering only the simple membrane lesions, concomitant myectomy to membrane resection showed only an improved trend with regards to freedom from reoperation, without achieving statistical significance (Fig. 3). Supporting our findings are the results of Serraf et al. [2], who agree that surgery does not alter the fundamental geometry of the left ventricular outflow, namely the anterior displacement of the interventricular septum into the left ventricle [4], or the more pronounced left ventricular-aorta junction [2,13]. Some authors have found that a more aggressive approach is necessary, with systematic deep septal myectomy, thereby reducing the risk of recurrence and reoperation in their experience [3,4,10,14,15].
Although a genetic predisposition to subaortic stenosis has been postulated by some [16,17], partly due to abnormal aorto-septal geometry with a wider angle leading to increased shear stress and cellular proliferation [6], genetically known syndromes in our series showed no increase in prevalence with regards to this lesion, compared to a larger cohort of syndromal patients with other associated cardiac defects. Equally, although there was a tendency towards more reoperations in patients with chromosomal syndromes, we found no statistically significant difference with regards to risk of recurrence, neither when considering the entire cohort together, nor when analyzing the subgroups simple and complex defects separately. This finding is difficult to interpret, as there were five different syndromes grouped together as one variable, and larger numbers of patients in each syndrome may have skewed the results otherwise.
Aortic valve insufficiency may already be present at the time of operation, and in infants and children constitutes a surgical indication per se even in the absence of a significant gradient across the left outflow tract [4,18,19]. It may also appear years after successful surgery for subaortic stenosis [2,3,5], and in adult patients, appears to be even more prominent after surgical intervention, when compared to unoperated patients [9,20]. The latter authors conclude that in adults, prevention of aortic regurgitation is not a criterion for surgery in patients without significant obstruction of the left outflow tract. We found no influence of gradient or preoperative AI on the occurrence of late aortic valve insufficiency, a finding also confirmed by other groups [2,10]. Also, Rohlicek et al. found that surgery had no influence on preexisting AI, or the occurrence of new AI after successful surgery, suggesting that surgery has little impact on the fate of the aortic valve [5]. Contrary to our findings, other authors report preoperative gradients to correlate with postoperative mild/moderate aortic valve insufficiency [3,5].
Limitations to the study are inherent to the retrospective nature of data retrieval. Also, the median follow-up time is relatively short, and stronger inferences could be made with longer follow-up, namely with regards to potential recurrence of subaortic stenosis and potentially a larger subgroup of patients eventually requiring reoperation.
In summary, excellent immediate surgical results are the rule with subaortic membrane resection, even with associated complex cardiac defects. Recurrence of LVOT obstruction is a frequent finding and gives reason for ongoing concern and close follow-up of patients with this lesion. Complex subaortic stenosis with associated cardiac defects is a risk factor for reoperation, compared to more simple forms of isolated membrane-type lesions. Systematic myectomy concomitant to simple membrane resection does not decrease the rate of stenosis recurrence, either in isolated membrane lesions or in complex forms of tunnel-like obstruction where it is indicated, and should be tailored to individual anatomical findings.
| Footnotes |
|---|
Presented at the 21st Annual Meeting of the European Association for Cardio-thoracic Surgery, Geneva, Switzerland, September 16–19, 2007. | References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
Y. Hirata, J. M. Chen, J. M. Quaegebeur, and R. S. Mosca The role of enucleation with or without septal myectomy for discrete subaortic stenosis. J. Thorac. Cardiovasc. Surg., May 1, 2009; 137(5): 1168 - 1172. [Abstract] [Full Text] [PDF] |
||||
![]() |
O. Metton, W. B. Ali, O. Raisky, and P. R. Vouhe Modified Konno operation for diffuse subaortic stenosis MMCTS, September 15, 2008; 2008(0915): 3426. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Talwar, S. K. Choudhary, and B. Airan Reoperation after relief of congenital subaortic stenosis Eur. J. Cardiothorac. Surg., September 1, 2008; 34(3): 700 - 700. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |