Left ventricular remodeling and change of systolic function after closure of patent ductus arteriosus in adults: Device and surgical closure

Clinical Investigations

Valvular and Congenital Heart Disease

Left ventricular remodeling and change of systolic function after closure of patent ductus arteriosus in adults: Device and surgical closure Young-Hoon Jeong, MD,a Tae-Jin Yun, MD,b Jong-Min Song, MD,a Jung-Jun Park, MD,b Dong-Man Seo, MD,b Jae-Kon Koh, MD,c Se-Whan Lee, MD,a Mi-Jeong Kim, MD,a Duk-Hyun Kang, MD,a and Jae-Kwan Song, MDa Seoul, South Korea

Background

Left ventricular (LV) remodeling and predictors of LV systolic function late after closure of patent ductus arteriosus (PDA) in adults remain to be clearly demonstrated.

Methods In 45 patients with PDA, including 28 patients who received successful occlusion using the Amplatzer device (AD group) (AGA, Golden Valley, MN) and 17 patients who received surgical closure (OP group), echocardiography studies were performed before closure and 1 day (AD group) or within 7 days (OP group) after closure, and then were repeated at z6 months (17 F 13 months). Results In both groups, LV ejection fraction (EF) and end-diastolic volume index were significantly decreased immediately after closure, whereas end-systolic volume index did not change. During the long-term follow-up period, end-systolic as well as end-diastolic volume indices decreased significantly in both groups and LV EF recovered compared to the immediate postclosure state. However, LV EF remained low compared to the preclosure state. Five patients (11.1%) including 3 patients in the AD group and 2 patients in the OP group showed persistent late LV systolic dysfunction (EF b50%). In stepwise, multiple logistic regression analysis, preclosure EF was the only independent predictor of late normal postclosure EF (odds ratio, 1.230; 95% CI, 1.054-1.434; P = .008). Receiver operating characteristic curve analysis showed that preclosure EF z62% had a sensitivity of 72% and a specificity of 83% for predicting late normal LV EF after closure. Conclusions

Left ventricular EF remains low late after PDA closure compared with preclosure state in adults. Preclosure LV EF is the best index to predict late postclosure LV EF. (Am Heart J 2007;154:436240.)

Patent ductus arteriosus (PDA) is a congenital anomaly causing left ventricular (LV) volume overload, which may be a compensatory phenomenon to maintain systemic cardiac output.1,2 Previous reports demonstrated that PDA closure led to immediate deterioration of LV systolic function, which recovered within 6 months in children.2,3 On the other hand, LV systolic dysfunction was reported to persist for 4 months after PDA closure in a 12-year-old child.4 However, preclosure predictors of the late persistent LV systolic dysfunction

From the Divisions of aCardiology, bPediatric Cardiac Surgery, and cPeiatric Cardiology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea. Submitted March 8, 2007; accepted April 25, 2007. Reprint requests: Jong-Min Song, MD, PhD, Division of Cardiology, Asan Medical Center, University of Ulsan College of Medicine, 388-1 Poongnap2-dong, Songpa-gu, Seoul 138-736, South Korea. Tae-Jin Yun, MD, PhD, Division of Pediatric Cardiac Surgery, Asan Medical Center, University of Ulsan College of Medicine, 388-1 Poongnap2-dong, Songpa-gu, Seoul 138-736, South Korea. E-mails: [email protected], [email protected] 0002-8703/$ - see front matter n 2007, Mosby, Inc. Allrights reserved. doi:10.1016/j.ahj.2007.04.045

and volume overload have not yet been clearly demonstrated. Furthermore, LV remodeling and changes in LV systolic function after PDA closure in adult patients have not been reported. Recently, PDA closure using the Amplatzer duct occluder device (AD) has been proven to be a safe and rational therapeutic option, especially in adults.5-12 However, the difference between surgical closure and the closure using AD in terms of LV remodeling and LV systolic functional changes after PDA closure has not been demonstrated. Therefore, we sought to investigate LV remodeling and changes in systolic function after PDA closure using surgery and AD, and to identify preclosure predictors of long-term LV systolic function after PDA closure in adults.

Methods Study population In a total of 45 adult patients with PDA, 2-dimensional echocardiographies were repeated at more than 6 months after successful PDA closure. Percutaneous PDA closure using AD

American Heart Journal Volume 154, Number 3

Jeong et al 437

Figure 1

Changes in LV EF, ESVI, and EDVI after PDA closure using surgery (OP group) and Amplatzer duct occluder device (AD group). *P b .05 versus preclosure; yP b .05 versus corresponding value of the OP group. BSA, body surface area.

Table I. Changes in LV volumes and EF after closure of PDA in adults AD group Preclosure EF (%) EDVI (mL/m2 BSA) ESVI (mL/m2 BSA)

60 F 10 93 F 34y 39 F 24

OP group

Postclosure

Follow-up

Preclosure

Postclosure

54 F 104y 79 F 264 39 F 23

58 F 84 62 F 134 26 F 114

64 F 6 139 F 55 51 F 24

44 F 124 90 F 314 54 F 27

Follow-up 56 F 64 63 F 164 27 F 84

4P b .05 versus preclosure. yPb .05 versus corresponding value of OP group.

was performed in 28 patients (age, 40 F 13 years, 26 women; AD group) and surgical PDA closure was performed in 17 patients (age, 34 F 11 years, 16 women; OP group). Inclusion criteria were age z18 years, documented PDA, and presence of audible continuous cardiac murmur. Exclusion criteria were markedly elevated pulmonary vascular resistance-systemic vascular resistance ratio N0.4 or combined other congenital heart diseases. PDA sizes were evaluated by transthoracic and transesophageal echocardiography. Percutaneous PDA occlusion procedures were performed by standard techniques.11 Immediately after the closure using AD, a descending aortogram was performed to assess the presence of a residual shunt. In the OP group, PDA division with primary closure was performed in 13, PDA ligation in 1, and division with patch in 3 patients. Eleven patients (65%) underwent descending thoracic aorta clamping, and 3 patients (18%) needed cardiopulmonary bypass assist. The cardiopulmonary pump time was 57 to 88 minutes.

Echocardiographic evaluation and follow-up Transthoracic Doppler echocardiography was performed before and after PDA closure (the next day of the procedure in the AD group, and within 1 week in the OP group) to measure LV end-systolic (ESVI) and end-diastolic volume

indices (EDVI) using modified Simpson’s method and standardization by body surface area of each patient, and to calculate LV ejection fraction (EF). LV end-diastolic and endsystolic internal dimensions were measured and LV mass was calculated using LV dimensions and myocardial wall thicknesses on M-mode echocardiography.13 The presence of residual shunt was assessed by color Doppler echocardiography. Follow-up echocardiography was repeated at z6 months (17 F 13 months) in all patients, and echocardiographic follow-up duration of the OP group was not significantly different from that of the AD group (21 F 19 vs 14 F 8 months, P = .056).

Statistical analysis Statistical analysis was performed using SPSS 13 software program (SPSS, Chicago, IL). Categorical data were presented as frequencies (%) and compared using the Fisher exact test. Continuous variables were presented as mean F SD and compared using the Student t test. Changes in LV volume indices and EF were evaluated by the paired t test. Correlation between 2 continuous variables was evaluated using linear regression analysis. Multiple stepwise linear regression analysis was performed to identify preclosure determinant of late postclosure LV EF. Multivariate stepwise logistic analysis using

American Heart Journal September 2007

438 Jeong et al

Table II. Multiple linear regression analysis for the LV EF late after closure of PDA Preclosure variables Age (y) BSA (m2) PDA size (mm)4 EF (%) LVMI (g/m2 BSA) ESVI (mL/m2 BSA) EDVI (mL/m2 BSA) LVIDs (mm) LVIDd (mm)

r 0.107 0.110 0.913 0.647 0.170 0.570 0.302 0.496 0.349

Figure 2

Univariate P Multivariate P .483 .472 .913 b.001 .294 b.001 .049 .001 .019

b.001

LVIDd, Left ventricular end-diastolic internal dimension; LVIDs, left ventricular endsystolic internal dimension; LVMI, left ventricular mass index. 4Size measured by transthoracic echocardiography.

backward stepwise elimination was performed to evaluate independent preclosure predictors for the late normal LV systolic function. Variables with P b .3 were considered significant in univariate analyses and included in multivariate analysis. Receiver operating characteristic curve analysis was performed to identify the best cutoff value for predicting late normal LV EF. P b .05 was considered significant.

Results In the AD group, PDA diameters were 5.9 F 2.2 mm (range, 3-11 mm) and 6.5 F 2.6 mm (range, 2-13 mm) on the transthoracic and transesophageal echocardiographies, respectively. There were significant correlations between successfully implanted AD size and measured PDA sizes using transthoracic (r = 0.792, P b .001) and transesophageal echocardiographies (r = 0.544, P = .005). There was no death or major complications such as device migration and embolization. Immediate angiogram showed residual shunts in 2 patients. Of these 2 patients, 1 patient showed delayed complete PDA occlusion and only 1 patient showed a minimal residual shunt on the long-term follow-up echocardiography. In the OP group, no immediate and long-term major complications developed after surgery. Postclosure hospital stay duration in the OP group was significantly longer than that in the AD group (8.3 F 3.2 vs 2.0 F 0.2 days, P b .001). Before PDA closure, ESVI and LV EF were not significantly different between the AD group and the OP group, whereas EDVI in the OP group was significantly larger than that in the AD group (Figure 1, Table I). In both groups, EDVI was significantly decreased immediately after PDA closure, whereas ESVI did not change (OP group, P = .633; AD group, P = .442). Consequently, LV EF significantly decreased immediately after closure in both groups. The immediate postclosure deterioration of LV EF in the AD group was smaller than that in the OP group (6.5% F 6.9% vs

Correlation between LV EFs (EF) at preclosure and late postclosure of PDA.

20.7% F 9.6%, P b .001). In the OP group, postoperative LV EF deterioration was not significantly different between the patients with and without descending thoracic aorta clamping (20.8% F 10.7% vs F20.3% F 8.0%, P = .924) or between the patients with and without cardiopulmonary bypass (25.0% F 6.2% vs 19.7% F 10.1%, P = .403). During the long-term follow-up period, ESVI as well as EDVI decreased significantly in both groups and LV EF recovered compared to the immediate postclosure state. However, LV EF remained low compared to the preclosure state in both groups (OP group, P = .001; AD group, P = .049). Late postclosure LV EF was significantly correlated with preclosure values of LV EF, ESVI, EDVI, and LV end-systolic and end-diastolic dimensions on M-mode (Table II, Figure 2). Multivariate stepwise linear regression analysis revealed that preclosure LV EF was the only determinant of late postclosure LV EF ( P b .001). Five patients (11.1%) including 3 patients in the AD group and 2 patients in the OP group showed persistent late LV systolic dysfunction (EF b50%) until 18 F 16 months (range, 6-45 months) after PDA closure. To determine the preclosure predictors of late normal LV systolic function (EF z50%), stepwise multiple logistic regression analysis was performed (Table III). Preclosure LV EF was the only independent predictor of late normal LV systolic function (odds ratio [OR], 1.230; 95% CI, 1.054-1.434; P = .008). By receiver operating characteristic curve analysis, the area under the curve of preclosure LV EF for predicting long-term normal LV systolic function was 0.818 (95% CI, 0.6231.014), and the preclosure EF z 62% showed a sensitivity of 72% and a specificity of 83% (Figure 3).

American Heart Journal Volume 154, Number 3

Jeong et al 439

Table III. Preclosure predictors of late postclosure normal EF (z50%) in adult patients with PDA Univariate analysis Preclosure variables Age (y) Sex BSA (m2) PDA size (mm)4 EF (%) LVMI (g/m2 BSA) ESVI (mL/m2 BSA) EDVI (mL/m2 BSA) LVIDs (mm) LVIDd (mm) AF

Multivariate analysis

OR

95% CI

P

0.966 0.211 21.228 0.996 1.189 0.997 0.944 0.985 0.863 0.905 26.000

0.896-1.041 0.015-2.869 0.012-38156 0.540-1.838 1.049-1.347 0.985-1.009 0.905-0.985 0.967-1.003 0.757-0.984 0.800-1.023 1.796-376

.363 .242 .424 .990 .007 .641 .008 .102 .028 .111 .017

OR

95% CI

P

1.230

1.054-1.434

.008

AF, Atrial fibrillation. 4Size measured by transthoracic echocardiography.

Discussion Left ventricular volume overload is frequently associated with PDA and seems to be required to increase cardiac output by Frank-Starling response to overcome significant left-to-right shunt and maintain systemic circulation.1,14,15 Because the LV remodeling is caused by significant left-to-right shunt through PDA, it is conceivable that LV reverse remodeling occurs after PDA closure. Recently, Eerola et al3 demonstrated, using 2- and 3-dimensional echocardiography, that changes in LV volume and function caused by PDA closure disappeared by 6 months after percutaneous closure in children. Galal et al2 reported that closure of relatively large PDA in children led to significant immediate deterioration of LV systolic performance, which appeared to recover within a few months. Consistent with these previous reports, our current study has demonstrated that LV EF was depressed immediately after PDA closure and recovered during the long-term follow-up period in adult PDA patients. Immediate postclosure deterioration of LV EF was attributed to immediate decrease in LV EDVI and stationary ESVI, whereas late recovery of LV EF was mainly due to late decrease in ESVI. In contrast to previous studies performed with children with PDA, our study population was composed of adults with relatively large PDA size. A unique finding in this study performed with adult patients with PDA was that LV systolic function was not completely recovered until z6 months after PDA closure compared with the preclosure state. About 11% of patients showed persistent long-term deterioration of LV systolic function in our study population. These inconsistent findings between this study and previous studies suggest that reverse LV remodeling after PDA closure in adults might be different from that in children. A more prolonged LV remodeling induced by PDA in adults compared to children might cause

Figure 3

Receiver operating characteristic curve of preclosure LV EF (EF) for predicting late normal postclosure EF (z50%). The number on the curve indicates the cutoff value of the point.

some structural changes in the LV myocardium and irreversible deterioration of LV contractility. This study also demonstrated for the first time that preclosure LV EF is the sole predictor of normal LV EF late after PDA closure in adults. These results suggest that PDA closure should be performed before LV EF decreases (b62%) to achieve a normal LV EF after PDA closure. There were no differences in basic patterns of LV remodeling and changes of systolic function between AD and OP groups, but immediate deterioration of LV systolic function was less prominent in the AD group than in the OP group. The cause of this phenomenon could not be clearly demonstrated in this study, but might result from perioperative intravascular volume

440 Jeong et al

change, bleeding, physical and emotional stress, and/or myocardial depressive effect of general anesthesia and inflammatory reaction in the surgical wound. The present study had some limitations. The study was retrospectively analyzed in a single center. Echocardiographic follow-up was dependent on the attending physician’s discretion and was not uniform. Finally, we did not demonstrate the effect of medication on the LV remodeling and systolic functional change. In conclusion, LV EF remains low late after PDA closure compared with the preclosure state in adult patients with PDA, although it is recovered from immediate postclosure deterioration. Preclosure LV EF is the best index to predict late postclosure LV EF.

References 1. Takahashi Y, Harada K, Ishida A, et al. Changes in left ventricular volume and systolic function before and after the closure of ductus arteriosus in full-term infants. Early Hum Dev 1996;44:77 - 85. 2. Galal MO, Amin M, Hussein A, et al. Left ventricular dysfunction after closure of large patent ductus arteriosus. Asian Cardiovasc Thorac Ann 2005;13:24 - 9. 3. Eerola A, Jokinen E, Boldt T, et al. The influence of percutaneous closure of patent ductus arteriosus on left ventricular size and function: a prospective study using two- and three-dimensional echocardiography and measurements of serum natriuretic peptides. J Am Coll Cardiol 2006;47:1060 - 6. 4. Galal MO, Arfi MA, Nicole S, et al. Left ventricular systolic dysfunction after transcatheter closure of a large patent ductus arteriosus. J Coll Physicians Surg Pak 2005;15:723 - 5.

American Heart Journal September 2007

5. Masura J, Tittel P, Gavora P, et al. Long-term outcome of transcatheter patent ductus arteriosus closure using Amplatzer duct occluders. Am Heart J (2006);151:755 e7-e10. 6. Thanopoulos BD, Hakim FA, Hiari A, et al. Further experience with transcatheter closure of the patent ductus arteriosus using the Amplatzer duct occluder. J Am Coll Cardiol 2000;35:1016 - 21. 7. Faella HJ, Hijazi ZM. Closure of the patent ductus arteriosus with the Amplatzer PDA device: immediate results of the international clinical trial. Catheter Cardiovasc Interv 2000;51:50 - 4. 8. Bilkis AA, Alwi M, Hasri S, et al. The Amplatzer duct occluder: experience in 209 patients. J Am Coll Cardiol 2001;37:258 - 61. 9. Fischer G, Stieh J, Uebing A, et al. Transcatheter closure of persistent ductus arteriosus in infants using the Amplatzer duct occluder. Heart 2001;86:444 - 7. 10. Kong H, Gu X, Bass JL, et al. Experimental evaluation of a modified Amplatzer duct occluder. Catheter Cardiovasc Interv 2001;53:571 - 6. 11. Masura J, Walsh KP, Thanopoulous B, et al. Catheter closure of moderate- to large-sized patent ductus arteriosus using the new Amplatzer duct occluder: immediate and short-term results. J Am Coll Cardiol 1998;31:878 - 82. 12. Pass RH, Hijazi Z, Hsu DT, et al. Multicenter USA Amplatzer patent ductus arteriosus occlusion device trial: initial and one-year results. J Am Coll Cardiol 2004;44:513 - 9. 13. Devereux RB, Alonso DR, Lutas EM, et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450 - 8. 14. Mathew R, Thilenius OG, Arcilla RA. Comparative response of right and left ventricles to volume overload. Am J Cardiol 1976; 38:209 - 17. 15. Gittenberger-deGroot AC. Persistent ductus arteriosus: most probably a primary congenital malformation. Br Heart J 1977; 39:610 - 8.