|
|
||||||||
J Thorac Cardiovasc Surg 2004;128:449-456
© 2004 The American Association for Thoracic Surgery
Surgery for acquired cardiovascular disease |
a Department of Thoracic and Cardiovascular Surgery, Rikshospitalet, Oslo, Norway
b Research Forum, Clinical Epidemiology Unit, Ullevaal University Hospital, Oslo, Norway
Received for publication October 28, 2003; revisions received April 7, 2004; accepted for publication April 23, 2004.
* Address for reprints: Runar Lundblad, MD, Rikshospitalet, N-0027 Oslo, Norway
runar.lundblad{at}rikshospitalet.no
| Abstract |
|---|
|
|
|---|
METHODS: We retrospectively reviewed 159 patients undergoing operations between 1989 and 2003. The epidemiologic design was of an exposed (simple linear repair, n = 74) versus nonexposed (endoventricular patch plasty, n = 85) cohort with 2 endpoints: early mortality and long-term survival. The crude effect of aneurysm repair technique versus endpoint was estimated by odds ratio, rate ratio, or relative risk and their 95% confidence intervals. Stratification analysis by using the Mantel-Haenszel method was done to quantify confounders and pinpoint effect modifiers. Adjustment for multiconfounders was performed by using logistic regression and Cox regression analysis. Survival curves were analyzed with the Breslow test and the log-rank test.
RESULTS: Early mortality was 8.2% for all patients, 13.5% after linear repair and 3.5% after endoventricular patch plasty. When adjusted for multiconfounders, the risk of early mortality was significantly higher after simple linear repair than after endoventricular patch plasty (odds ratio, 4.4; 95% confidence interval, 1.1-17.8). Mean follow-up was 5.8 ± 3.8 years (range, 0-14.0 years). Overall 5-year cumulative survival was 78%, 70.1% after linear repair and 91.4% after endoventricular patch plasty. The risk of total mortality was significantly higher after linear repair than after endoventricular patch plasty when controlled for multiconfounders (relative risk, 4.5; 95% confidence interval, 2.0-9.7). Linear repair dominated early in the series and patch plasty dominated later, giving a possible learning-curve bias in favor of patch plasty that could not be adjusted for in the regression analysis.
CONCLUSIONS: Postinfarction left ventricular aneurysm can be repaired with satisfactory early and late results. Surgical risk was lower and long-term survival was higher after endoventricular patch plasty than simple linear repair. Differences in outcome should be interpreted with care because of the retrospective study design and the chronology of the 2 repair methods.
Surgical treatment is indicated in established cases of CHF, angina pectoris, malignant ventricular arrhythmia, or recurrent embolization from the LV. The goal of surgical intervention is to correct the size and geometry of the LV, reduce wall tension and paradoxical movement, and improve systolic function. Intracavitary thrombi are removed, and coronary artery bypass grafting (CABG) is usually performed. Aneurysmectomy and linear repair of the LV was introduced by Cooley and colleagues in 19581 and remained the standard procedure until the late 1980s. Endoventricular patch plasty (EVPP) was then introduced as a more physiologic repair than the linear closure technique,2,3 especially when the aneurysm extends into the septum. However, there is still controversy whether EVPP is superior to simple linear resection on early and late clinical outcome. Mickleborough and associates4,5 have introduced a method combining septoplasty and linear resection. Simple linear closure reduces only the aneurysm of the free wall, while their newer technique targets also the septal part of the aneurysm. In the present retrospective study on patients with postinfarction left ventricular aneurysm, we compare simple linear repair and EVPP with respect to surgical risk and long-term survival.
| Materials and methods |
|---|
|
|
|---|
|
Operative technique
All procedures were performed with cardiopulmonary bypass (CPB) and moderate systemic hypothermia. In 153 of the 159 patients, the aorta was crossclamped, and cold crystalloid cardioplegic solution was infused through the aortic root as an initial bolus of 1000 mL, followed by 200 to 300 mL every 20 to 30 minutes. The operative sequence was as follows: the aneurysm was incised parallel to the interventricular septum, the intracavitary thrombus was removed, a ventricular vent was inserted, distal coronary anastomoses were performed, the aneurysm was repaired, the aortic crossclamp was removed, and, finally, proximal anastomoses were performed with the aorta partial excluded. The heart was cardioplegic during the whole procedure, except the proximal anastomosing in 67 of the 153 patients. In the remaining 86 patients, the crossclamp was removed after completion of the distal anastomoses, and therefore the aneurysm repair was performed partly or completely on beating heart. In 6 cases in which CABG was not performed, the aorta was not crossclamped, and the aneurysm repair was performed on a beating heart. For linear repair, the edges were readapted with 2 strips of felt or pericardium for reinforcement by using a combination of continuous sutures and interrupted mattress stitches. For EVPP, a patch of dacron (n = 76) or polytetrafluoroethylene (n = 9) was attached to the border zone between normal and scarred tissue eventually after a circular purse-string suture was passed to restore a new apical neck (n = 9). The patch was secured with continuous (n = 39) or interrupted mattress (n = 8) sutures or a combination of both techniques (n = 38). The aneurysmal sac was closed over the patch with 2 strips of felt or pericardium for reinforcement by using a combination of continuous sutures and interrupted mattress stitches. When performed, cryoablation was based on mapping or just visually directed.
During the 14-year period covered by the study, there was a substantial turnover of surgeons. Altogether, there were 18 surgeons involved in 159 operations, resulting in relatively little personal experience (median, 8 procedures). Five surgeons performed less than 5 procedures, and 9 surgeons performed 5 to 10 procedures. Four surgeons did more than 10 procedures, and 1 surgeon performed 47 repairs (linear repair, 26; EVPP, 21). Six surgeons participated only in the linear repair experience, 8 surgeons performed only EVPP, and 4 surgeons performed both procedures.
Statistical analysis and epidemiologic methods
Continuous numeric data are presented as means ± SD (range), and discrete numeric data and categoric data are presented as frequencies. Endpoints were early mortality (<30 days) and long-term survival. Univariate analysis was performed by using contingency tables with
2 or 2-tailed Fisher exact tests for discrete numeric data or categoric data, and the 2-tailed t test or the Mann-Whitney test were used for continuous numeric data. Total mortality was analyzed according to the Kaplan-Meier method, and differences between survival curves were estimated by using the Breslow test and the log-rank test.6
The epidemiologic design was of an exposed (linear repair, n = 74) versus nonexposed (EVPP, n = 85) cohort with the 2 endpoints of early mortality and long-term survival. First, we estimated the crude association between the aneurysmal repair technique and the endpoints. Then a comprehensive search for confounders and effect modifiers was performed by using stratification analysis with the Mantel-Haenszel method.7 The Breslow-Day test for heterogeneity of odds ratios (ORs), rate ratios, or relative risks between strata was used to pinpoint eventual effect modifiers. The confounding effect was quantified, comparing the adjusted Mantel-Haenszel ORs, rate ratios, or relative risks with the corresponding crude values:
![]() |
Logistic regression and Cox regression analyses were performed to adjust for confounders with a more than 5% numeric confounding effect.8 Manual backward elimination of variables was performed on the basis of the following criteria: clinical or pathophysiologic importance, correlation matrix between the variables, and statistical significance.
| Results |
|---|
|
|
|---|
|
Early mortality and long-term survival
Crude effect
Early mortality (<30 days) was 8.2% (13 patients) for the whole material. Early mortality was 15.4% in the first 39 patients (95% linear repair), 7.5% in the next 40 patients (77.5% linear repair), 2.5% in the following 40 patients (7.5% linear repair), and 7.5% in the last 40 patients (7.5% linear repair). Early mortality within each operative group did not change over time. We were not able to detect any effect of surgical experience (number of aneurysm repairs per surgeon) on early mortality. The crude risk of early mortality was significantly higher after linear repair than after EVPP (OR, 3.8; 95% CI, 1.1-13.4; Table 3).
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
Aneurysm repair and ventricular arrhythmia
Ventricular arrhythmia in the history indicates reduced long-term survival.10,15 Ventricular arrhythmia was present in 48 of the 159 patients and represented a significant risk factor for early and total mortality independent of surgical repair method (not shown). Ventricular arrhythmia arises in the border zone between viable and dead myocardium, usually on the interventricular septum. Simple linear resection with or without concomitant CABG often fails to control ventricular arrhythmia. EVPP and linear resection combined with septoplasty might reduce wall tension on the septal border zone, and it is hypothesized that this might have an inherent antiarrhythmic effect.16 However, this is not supported in other clinical studies, and no experimental data exist.
Aneurysm repair and CABG
Concomitant CABG is highly recommended for 2 reasons. First, it reduces or prevents angina pectoris. Second, although the LAD is occluded and the periphery on the free wall is thin or calcified, an ITA graft to the LAD might be particularly important to improve septal perfusion and control ventricular arrhythmia. In the present study revascularization of the LAD was significantly more frequent in the EVPP group (72%) than in the linear repair group (39%). The importance of an ITA graft to the LAD during complex operations was probably underestimated in the first period of the series, when linear repair dominated. Moreover, with this technique, the aneurysmal sac is partly removed, and the remnant is used for closure of the LV, which could render revascularization of the LAD more difficult. During EVPP, more of the aneurysmal sac is retained, and therefore the conditions for revascularization of the LAD might be better. CABG was performed in 128 of the 159 patients, and the LAD was revascularized in 78 of the 151 patients with LAD lesions. The number of distal anastomoses did not differ between groups and had no effect on early mortality (P = .30) and long-term survival (P = .59). Incomplete revascularization had different reasons: thin or calcified coronary peripheries, poor venuous or arterial conduits, or an LAD running into the aneurysmal sac to be removed. Lack of revascularization of the LAD had no statistically significant adverse effect on early mortality (P = .44) or long-term survival (P = .30). CABG is generally recognized as an important part of left ventricular aneurysm operations, irrespective of repair technique. Improved results might have been achieved with a more aggressive approach on revascularization, especially of the LAD.
Aneurysm repair and left ventricular function
Low LVEF is a predictor for reduced long-term survival after left ventricular aneurysm repair.10,13,15,17,18 Unless it is extremely low, a reduced LVEF per se should not contradict an operation because historical data show a poor lifespan without surgical intervention in symptomatic patients.19 Only global LVEF was calculated in this study. This represents both the contractile and the noncontractile aneurysmal parts of the LV. Because noncontractile tissue is removed during aneurysm repair, the state of the contractile portion of the LV is probably a better predictor for cardiac performance after surgical intervention.
EVPP was introduced as a more physiologic repair than the traditional simple linear resection technique. EVPP is supposed to provide a more physiologic apex, whereas the distal end of the LV is more flattened after simple linear repair. EVPP can improve left ventricular function17,20,21 and is probably more effective than traditional simple linear resection.9,16 Both methods clearly improve functional status.4,9,10,12,14-16,20 Some studies report superior functional improvement after EVPP than after simple linear repair,12,14-16 whereas others fail to demonstrate any functional difference between the 2 methods.9,10,13 The present study compares EVPP with the traditional simple linear resection. The more sophisticated technique of Mickleborough and associates,4,5 using septoplasty and linear repair, might provide equally good results as EVPP. They report improvement in left ventricular function associated with good symptomatic relief and long-term survival. In small aneurysms not involving the septum, the theoretic advantages of EVPP are less convincing.
Limitations of the study
Left ventricular aneurysm can be defined as an area of asynergy (either akinetic or dyskinetic) that is large enough to reduce left ventricular function. This is a very loose definition, and the lack of uniformity of aneurysm criteria complicates almost all discussions of left ventricular aneurysm repair. In the present report the diagnosis and criteria for aneurysmectomy were based on a dyskinetic aneurysm with systolic paradoxical movement. However, there is a continuum between akinesia and dyskinesia, and there are obvious border zones between these entities. Aneurysmectomy was originally introduced for dyskinetic aneurysms, and controversy has existed as to whether patients with akinetic aneurysms would benefit from surgical intervention. Recent reports suggest that both left ventricular dysfunction and the outcome of surgical intervention depend on the extent of asynergy rather than the type of asynergy (akinetic or dyskinetic).4,17,22,23 Wall thinning of the asynergic area was not systematically quantified in our study, either preoperatively or perioperatively.
We have performed an observational study with 2 cohorts exposed to different surgical methods and operated on at different time points. This is not the ideal design to evaluate the efficacy of a new surgical technique because of selection bias, residual confounding bias, and eventually some unknown confounders for which it is impossible to control. The ideal design is the randomized clinical trial. However, the promising results after EVPP, combined with the relatively small number of patients in each center, makes a randomized study difficult to carry out.
Our study covered a period of 14 years, and although we tried to control for multiconfounder bias by means of stratification analysis and multivariate methods, there could be hidden mechanisms of selection between the 2 surgical groups for which we did not manage to adjust. There was a continuous turnover of surgeons throughout the study, with 18 involved altogether. Only one surgeon gained significant experience (47 cases), whereas the other surgeons performed few procedures (median, 8), and therefore a benefit from surgical training with aneurysmectomy could not be demonstrated in the statistical analysis. Nevertheless, the general experience in cardiac surgery has improved over time. Linear resection dominated early in the series and EVPP dominated later, but with some overlapping. The effect of a learning curve on the institutional team could therefore represent a potential bias in favor of EVPP. However, this could not be corrected for in the regression analysis because of the problem of overadjustment. One cannot adjust for a variable (date of the operation) that is highly associated with exposure (repair technique) because it is a surrogate for the same phenomenon. Adjustment in this case will create difficulties in the analysis, and the results could be difficult to interpret.7 The lower incidence of reoperation for bleeding and use of an IABP late in the series could support a bias of learning curve favoring EVPP. On the other hand, for each particular surgical method, these 2 incidents did not change over time. Alternative explanations are that EVPP might be safer when it comes to bleeding because the LV is closed in 2 layers (patch and ventricular wall), and a more physiologic shape of the LV after EVPP might lead to reduced need for IABP. Altogether, differences in outcome should be interpreted with care because of the retrospective study design and because of the chronology of the 2 procedures, resulting in a possible learning-curve bias in favor of EVPP that could not be adjusted for in the regression analysis.
Conclusions
Postinfarction left ventricular aneurysm can be repaired with satisfactory early and late clinical outcomes. In the present study surgical risk was lower and long-term survival was higher after EVPP than after simple linear repair. The difference in outcome should be interpreted with care because of the retrospective study design and because of the chronology of the 2 methods, resulting in a possible learning-curve bias in favor of EVPP.
| References |
|---|
|
|
|---|
ener E, Özatik MA, Ta
demir O, Bayazit K. Left ventricular aneurysm repair: an assessment of surgical treatment modalities. Eur J Cardiothorac Surg. 1998;13:4956This article has been cited by other articles:
![]() |
S. Castelvecchio, M. Ranucci, and L. A. Menicanti The role of surgical technique in determining the outcome of left ventricular reconstruction: a difficult assessment Eur. J. Cardiothorac. Surg., June 1, 2009; 35(6): 1111 - 1111. [Full Text] [PDF] |
||||
![]() |
P. Klein, J. J. Bax, L. J. Shaw, H. H.H. Feringa, M. I.M. Versteegh, R. A.E. Dion, and R. J.M. Klautz Early and late outcome of left ventricular reconstruction surgery in ischemic heart disease Eur. J. Cardiothorac. Surg., December 1, 2008; 34(6): 1149 - 1157. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Mukaddirov, R. G. Demaria, L. P. Perrault, J.-M. Frapier, and B. Albat Reconstructive surgery of postinfarction left ventricular aneurysms: techniques and unsolved problems. Eur. J. Cardiothorac. Surg., August 1, 2008; 34(2): 256 - 261. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Mukaddirov, J.-M. Frapier, R. G. Demaria, and B. Albat Surgical treatment of postinfarction anterior left ventricular aneurysms: linear vs. patch plasty repair Interactive CardioVascular and Thoracic Surgery, April 1, 2008; 7(2): 256 - 261. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. D. Glower and J. E. Lowe Left Ventricular Aneurysm Card. Surg. Adult, January 1, 2008; 3(2008): 803 - 822. [Full Text] |
||||
![]() |
A. Parolari, M. Naliato, C. Loardi, P. Denti, M. Trezzi, M. Zanobini, M. Porqueddu, M. Roberto, S. Kassem, F. Alamanni, et al. Surgery of Left Ventricular Aneurysm: A Meta-Analysis of Early Outcomes Following Different Reconstruction Techniques Ann. Thorac. Surg., June 1, 2007; 83(6): 2009 - 2016. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Tonnessen and C. W. Knudsen Surgical left ventricular remodeling in heart failure Eur J Heart Fail, August 1, 2005; 7(5): 704 - 709. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Menicanti and M. Di Donato Left ventricular aneurysm/reshaping techniques MMCTS, April 25, 2005; 2005(0425): 596. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. E. Antunes, R. Silva, J. F. de Oliveira, and M. J. Antunes Left ventricular aneurysms: early and long-term results of two types of repair Eur. J. Cardiothorac. Surg., February 1, 2005; 27(2): 210 - 215. [Abstract] [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 |