Author + information
- Received July 29, 2004
- Revision received October 16, 2004
- Accepted October 25, 2004
- Published online February 15, 2005.
- Omid Salehian, MD, MSc, FRCPC*,
- Eric Horlick, MD, FRCPC*,
- Markus Schwerzmann, MD*,
- Kim Haberer, BArt, Sc, MA*,
- Peter McLaughlin, MD, FRCPC*,
- Samuel C. Siu, MD, SM, FRCPC, FACC*,†,
- Gary Webb, MD, FRCPC, FACC* and
- Judith Therrien, MD, FRCPC,*,‡,* ()
- ↵*Reprint requests and correspondence:
Dr. Judith Therrien, Sir MB Davis Jewish General Hospital, 3755 Cote Ste. Catherine, Montreal, Quebec, Canada, H3T 1E2
Objectives We set out to study the effect of transcatheter closure of atrial septal defect (ASD) on right ventricular (RV) and left ventricular (LV) function assessed by myocardial performance index (MPI), as well as left atrial (LA) volumes.
Background The hemodynamic response to the closure of ASD is well-documented in surgically treated patients. However, few studies have documented echocardiographic evaluation of ventricular function in patients undergoing transcatheter closure of ASDs.
Methods Pre- and post-ASD device closure echocardiograms of 25 consecutive patients were retrospectively reviewed. Measurements of RV and LV MPI and LA volumes were made.
Results Twenty-five patients with an average age of 45.5 ± 16.3 years underwent transcatheter closure of ASD. There was statistically significant improvement in RV MPI (0.35 to 0.28, p = 0.004), LV MPI (0.37 to 0.31, p = 0.04), and LA volume index (25.7 to 21.8 ml/m2, p < 0.001) after closure of ASD.
Conclusions Device closure of ASDs leads to improvement of both RV and LV function as well as reduction in LA volume. These hemodynamic improvements provide insights into the symptomatic benefits gained in closure of ASDs using the transcatheter approach.
Atrial septal defects (ASDs) account for approximately 10% of all congenital heart lesions (1) and can sometimes result in the development of pulmonary hypertension, atrial arrhythmias, and right heart decompensation (2). Although surgical closure of ASDs has been the mainstay of treatment, and although both medium- and long-term studies show excellent results (3), there remains significant morbidity and mortality associated with surgical repair (4). Transcatheter closure of isolated secundum ASDs has been established as a treatment alternative to surgical closure (5). Although there are a number of short-term studies reporting its utility (6), no data are currently available on its long-term benefit.
Studies have shown hemodynamic and functional improvement in right ventricular (RV) function after both surgical and device closure of ASDs (7,8). However, there is a paucity of data assessing left ventricular (LV) function after closure of ASDs. Myocardial performance index (MPI) was initially described by Tei et al. (9) as a measure of combined LV systolic and diastolic function. Although MPI has been used to study patients with congenital heart disease, the effect of transcatheter closure of ASD on LV MPI and RV MPI has not been studied to date.
It is generally agreed that surgical closure of ASDs in adulthood is associated with significant mortality benefit; however, there is limited benefit when it comes to prevention of atrial arrhythmias (2). An increase in left atrial (LA) diameters predicts the development of atrial fibrillation, stroke, and death (10). Currently, no data exist on LA volumes in patients with ASD and the effect of closure on these volumes.
The goals of this study were: 1) to assess the effect of transcatheter closure of ASDs on RV and LV function; and 2) to study the effect of ASD closure on LA and LV volumes.
After approval from the institutional research ethics board, adult patients who had undergone transcatheter closure of secundum ASDs between June 1, 2002, and June 1, 2003, were identified from the Toronto Congenital Cardiac Center for Adults database. Patients were excluded if they were clinically unstable, were not in sinus rhythm at the time of either echocardiographic examination, or had inadequate echocardiographic assessment. Demographic information, clinical status, and the size of device implanted were collected from the medical records.
The Amplazer septal occluder (AGA Corp., Golden Valley, Minnesota) was used for ASD closure as previously described (11). The device size chosen was either 2 mm or 10% greater than the stretched diameter size, whichever was larger.
Echocardiographic studies were performed in the standard manner using commercially available systems. Doppler recordings of LV and RV inflows (interval a) and outflows (interval b) were used for the calculation of MPI for each ventricle (9) (Fig. 1).Measurements were made on four consecutive beats, and MPI was calculated from the average values as per the formula: MPI = (a − b)/b.
Left ventricular volumes were determined using the Teicholz method (12). Left ventricular ejection fraction (EF) was estimated using the Quinones method (13); LA volumes were estimated using a length-diameter ellipsoid method (14) (Fig. 2)and were indexed to body surface area, and left atrial volume index (LAVI) was reported for each patient. Echocardiograms were examined by the primary reader (O.S.) who was blinded to the clinical information.
Intraobserver variability was assessed in a randomly selected subset of 10 patients by repeating all measurements on a separate occasion. To test the interobserver variability, all measurements were performed offline by a second observer (M.S.) who was blinded to the clinical data as well as the results of the initial echocardiographic examination.
All statistical analyses were done using SPSS 12.0 (SPSS Inc., Chicago, Illinois). Data are expressed as mean values ± SD. Comparisons between pre- and post-device closure data were made using Wilcoxon signed-rank test. A p value of <0.05 was considered to be statistically significant. A Spearman correlation was used to determine the relationship between RV MPI and LV MPI. Intra- and interobserver variability were calculated as mean percentage error, derived from the difference between the two sets of measurements, divided by the mean observation.
Demographic data are shown in Table 1.A total of 30 patients were identified. Five patients were excluded (four had inadequate echocardiograms, and one patient was in atrial fibrillation). The remaining 25 patients (28% male) had an average age of 45.5 ± 16.3 years (range 20 to 79 years) at the time of ASD closure. All patients had successful deployment of their device, and no significant residual leak immediately after closure.
The average time interval between transcatheter closure and echocardiographic examination was 94.8 ± 73.5 days (range 8 to 270 days). Echocardiographic data are presented in Table 2.The average RV systolic pressure decreased from 40.3 ± 6.2 mm Hg to 31.0 ± 5.3 mm Hg after device closure, a decrease of 23.1% (p = 0.0001). The LV end-diastolic diameter (LVEDD) increased from 41.5 ± 6.7 mm to 44.6 ± 5.1 mm, an increase of 7.5% (p = 0.004). There was a non-significant increase in left ventricular end-systolic diameter (LVESD) (27.2 ± 6.5 mm to 28.7 ± 5.2 mm, p = 0.15). The LV end-diastolic volume increased from 79.2 ± 31.4 ml to 92.1 ± 24.7 ml, an increase of 16.3% (p = 0.007). The LV end-systolic volume showed a non-significant increase from 30.1 ± 18.8 ml to 31.5 ± 13 ml (p = 0.1). The LV EF showed an increase from 62.9 ± 11.1% to 65.8 ± 8.0%, which was not statistically significant (p = 0.38).
The RV MPI showed an improvement from 0.35 ± 0.14 to 0.28 ± 0.09 (Fig. 3),a 20% change (p = 0.004) after closure. The LV MPI also showed an improvement from 0.37 ± 0.12 to 0.31 ± 0.11, a 16.2% change (p = 0.04) after closure (Fig. 3). There was a positive correlation between the RV and LV MPIs before device closure (r = 0.47, p = 0.02). After closure there was no significant correlation between LV and RV MPIs (r = 0.12, p = 0.6).
Left atrial volume index decreased from 25.7 ± 8.0 ml/m2to 21.8 ± 6.6 ml/m2after transcatheter closure (Fig. 4),a decrease of 15.2% (p = 0.0001). All three parameters used to calculate the LA volume by the length-diameter ellipsoid method showed statistically significant reductions of 3.9% to 6.5% after device closure (Table 2).
Intra- and interobserver variability
For RV MPI measurements, the intraobserver variability was 4.1 ± 3.3%, and the interobserver variability was 5.9 ± 3.3%. For LV MPI, the intraobserver variability was 4.8 ± 2.8%, and the interobserver variability was 7.3 ± 4.6%. For measurements of LA volume, the intraobserver variability was 10.1 ± 5.0%, and the interobserver variability was 13.6 ± 4.0%.
Our study is the first to demonstrate that transcatheter closure of secundum ASD is associated with improvement of both RV and LV functions as assessed by MPI. We also showed a decrease in LA volumes after transcatheter closure of ASD.
Studies have reported an average value for RV MPI in the normal adult population between 0.26 to 0.28 (15). Our population had an abnormally high pre-closure RV MPI of 0.35 ± 0.14, reflecting a worse RV function. However, after closure there was a 20% improvement in the RV MPI into the normal range with a value of 0.28 ± 0.09. Our intra- and interobserver variability for RV MPI was 4.1 ± 3.3% and 5.9 ± 3.3%, respectively, in keeping with those reported previously (15).
Eidem et al. (16) reported an average RV MPI of 0.38 ± 0.04 in 40 patients with ASD; 16 of these patients underwent surgical closure of their ASD. Postoperative assessment revealed an RV MPI of 0.35 ± 0.03, which was not a statistically significant change from the preoperative state. The lack of reduction in RV MPI could reflect that surgical closure of ASD required cardiopulmonary bypass. Some investigators have indicated a potential role of cardiopulmonary bypass in post-surgical RV impairment (7,8). Perhaps postoperative RV impairment negates any gain in the RV function from ASD closure. Transcatheter closure of ASD avoids this potential problem.
This is the first study to report on the effect of ASD closure on LV MPI. The average value for LV MPI in the normal adult population is in the range of 0.34 to 0.40 (17). In this study we show that device closure of ASD was associated with a 16% improvement in the LV MPI from 0.37 ± 0.12 to 0.31 ± 0.11 (p = 0.04). Although we saw an improvement in LV MPI after closure, pre-closure values were well within the normal range. Our intra- and interobserver variability for LV MPI was 4.8 ± 2.8% and 7.3 ± 4.6%, respectively, in keeping with those reported in published data (18).
Two other studies have reported LV MPI in patients with ASD. Kim et al. (19) examined LV MPI in RV volume and pressure overloaded states. The 22 patients with ASD in this study had an average LV MPI value of 0.36 ± 0.03. Cheung et al. (20) in a pediatric population using tissue Doppler imaging, compared LV MPI in 17 normal subjects (0.42 ± 0.07), 17 with surgically closed ASD (0.49 ± 0.12) and 17 patients with percutaneously closed ASD (0.40 ± 0.13). Although there was no statistical difference between these groups, there was a trend toward benefit (lower LV MPI) with percutaneous closure.
One potential explanation for the improvement in LV MPI in our study is the influence of ventricular interdependence. We saw significant positive correlation between the RV and LV MPI values before device closure. A number of studies have indirectly alluded to this interaction leading to abnormal LV diastolic function in patients with ASDs (21). Using invasive hemodynamic assessment in patients with ASD, Satoh et al. (22) showed a 30% prolongation of LV isovolumic relaxation times (IVRT) compared with normal controls. Their data also showed that impaired LV filling shortened LV ejection time in this population. The combination of these two processes will give an LV MPI value, which would be higher (worse function) than normal. After closure of ASD, the volume load to the RV will decrease and will lead to shortening of LV IVRT, yielding a more normal LV filling. This will result in an LVMPI value that is lower than the pre-closure state.
Age at time of ASD closure appears to be the most important predictor of development of atrial arrhythmias (2,23). Oliver et al. (24) showed an association between LA diameter and development of atrial fibrillation in patients with ASD whether or not they had undergone surgical closure. Du et al. (25) showed a significant decrease in LA diameter measured by echocardiography from 37 ± 8 mm to 30 ± 8 mm at 24 h after device closure, and 27 ± 6 mm at 6-month follow-up.
To our knowledge ours is the first study showing a significant reduction in LAVI after transcatheter closure of ASD. The intra- and interobserver variability in the estimates of LA volumes in our study was 10.1 ± 5.0% and 13.6 ± 4.0%, respectively, similar to those reported in published reports in other conditions (26). The decrease in LA volume seen in our study is likely secondary to improved LV filling (22) and, perhaps more importantly, to the decrease in LA preload after elimination of the interatrial shunt. Potential benefit of this reduction in volume is the decrease in the risk of development of atrial fibrillation and its associated risk of vascular events.
LV dimensions and EF
The increase in LVEDD after closure of ASD seen here is similar to results reported for both catheter-based as well as surgical closure of ASDs seen in other studies (7,25). A recently published study by Giardini et al. (27) showed a similar increase in the LVEDD in patients after device closure of ASD. Furthermore, their study showed a statistically significant increase in LV EF and no change in the LVESD. In their study comparison of pre- and post-closure (six-month) cardiopulmonary testing showed that the echocardiographic changes after ASD closure was associated with a 17% increase in peak oxygen consumption, which was highly significant. They also showed a 15% increase in peak oxygen pulse, a surrogate for stroke volume. In our study we saw an increase of 23% in stroke volume after device closure. Another recent study by Walker et al. (21) further confirms the previously stated findings. They found a 17% increase in LV end-diastolic volumes immediately after device closure of ASDs. There was no significant change in end-systolic volumes after closure. This is very similar to our findings of a 16% increase in LV end-diastolic volume after device closure of ASD. Hence, the clinical improvements seen in patients after ASD closure can be explained by the augmentation in LV filling and, consequently, the stroke volume.
Several potential limitations of this study should be examined. There are variable lengths of time between transcatheter closure of ASD and follow-up echocardiographic examination in our population. Studies have shown that the majority of hemodynamic changes have taken place within three months of ASD closure (6,25), so our echocardiographic studies after an average of three months after closure likely reflect a fairly complete remodeling process. Unfortunately our study did not include enough patients to perform a meaningful comparison between an early and late device closure group. Secondly, although MPI is considered to be relatively load-independent, this has been challenged by recent studies (28). The effect of varying loading conditions on MPI in patients with ASDs is currently not completely understood. Hence, it is possible that different loading conditions in our patients might have affected the MPI values. Thirdly, the 15% reduction seen in the LAVI, although statistically significant, should be seen in the context of 13.6% interobserver variability. However, we believe that the change seen here is real because every patient showed a reduction in LA volume. Finally, this is a retrospective study with a small number of patients; a larger prospective study would be welcome to further assess results seen in our study.
Device closure of ASDs leads to improvement of both RV and LV function as well as reduction in LA volume. These hemodynamic improvements provide insights into the symptomatic benefits gained in closure of ASDs using the transcatheter approach. Whether a reduction in LA volume predicts a reduction in probability of later arrhythmias in a given patient remains to be determined.
Dr. Schwerzmann was supported by a grant from the Swiss National Science Foundation.
- Abbreviations and acronyms
- atrial septal defect
- ejection fraction
- isovolumic relaxation time
- left atrial/atrium
- left atrial volume index
- left ventricle/ventricular
- left ventricular end-diastolic diameter
- left ventricular end-systolic diameter
- myocardial performance index
- right ventricle/ventricular
- Received July 29, 2004.
- Revision received October 16, 2004.
- Accepted October 25, 2004.
- American College of Cardiology Foundation
- Dickinson D.F.,
- Arnold R.,
- Wilkinson J.L.
- Roos-Hesselink J.W.,
- Meijboom F.J.,
- Spitaels S.E.,
- et al.
- Galal M.O.,
- Wobst A.,
- Halees Z.,
- et al.
- Chan K.C.,
- Godman M.J.,
- Walsh K.,
- Wilson N.,
- Redington A.,
- Gibbs J.L.
- Veldtman G.R.,
- Razack V.,
- Siu S.,
- et al.
- Dhillon R.,
- Josen M.,
- Henein M.,
- Redington A.
- Benjamin E.J.,
- D'Agostino R.B.,
- Belanger A.J.,
- Wolf P.A.,
- Levy D.
- Thanopoulos B.D.,
- Laskari C.V.,
- Tsaousis G.S.,
- Zarayelyan A.,
- Vekiou A.,
- Papadopoulos G.S.
- Quinones M.A.,
- Waggoner A.D.,
- Reduto L.A.,
- et al.
- Pritchett A.M.,
- Jacobsen S.J.,
- Mahoney D.W.,
- Rodeheffer R.J.,
- Bailey K.R.,
- Redfield M.M.
- Giardini A.,
- Donti A.,
- Formigari R.,
- et al.