Accuracy of Intraoperative Epicardial Echocardiography in the Assessment of Surgical Repair of Congenital Heart Defects confirmed.

OBJECTIVE: To determine the accuracy of epicardial echocardiography in detecting residual lesions after surgical repair of congenital heart defects. To determine the sensitivity, specificity, positive predictive valve, negative predictive valve, and false negative and false positive percentage of the same.
MATERIALS AND METHODS: One year hospital-based prospective study of epicardial echocardiography in patients undergoing cardiopulmonary bypass for surgical correction of congenital heart defects in children and adults.
RESULTS: Epicardial echocardiography was done in 158 patients. Residual lesions were detected in 38 patients by epicardial echocardiography. In 28 of these cases the residual lesions were significant. In 24 of them immediate reoperation was done with good outcome. Epicardial echo has high sensitivity, specificity, positive and negative predictive valve, in detecting residual lesion with congenital heart defects.
CONCLUSION: Routine use of intraoperative epicardial echocardiography allows detection of majority of residual defects. It has high accuracy. There were no complications associated with use of epicardial echocardiography. The use of transesophageal echocardiography (TEE) can be avoided in small children and neonates undergoing cardiac surgery.

INTRODUCTION

Surgical repair of congenital heart defects has undergone a remarkable evolution in a developing country like India.[1] A significant proportion of major defects are now often completely corrected in early infancy with improving rates of survival. Residual cardiovascular lesions, anatomical or functional are a major cause of morbidity and mortality after surgical repair.[23456] The increasing availability and use of epicardial echocardiography have made the pediatric cardiac surgeons aware that intraoperative epicardial echocardiography can provide valuable information during all types of procedures. Epicardial echocardiography was introduced 3 decades ago and resulted in prompt recognition of significant residual defects.[78910] Later, transesophageal echocardiography (TEE) was introduced and largely replaced it in adults undergoing cardiac surgery.[1112]

In our center, transesophageal echocardiography is not performed in neonates, infants, and small children undergoing cardiac surgery for congenital heart defects. In the above mentioned pediatric population epicardial echocardiography was introduced since 2010. Hence, a prospective study was designed to determine the accuracy of epicardial echocardiography to detect residual lesions and its impact on surgical outcome. To determine this, intraoperative epicardial echocardiography was compared with surgical findings and immediate postoperative transthoracic echocardiography done in intensive therapy unit.

MATERIALS AND METHODS

Patient population

Pediatric and adult patients with congenital heart disease undergoing cardiac surgery involving cardiopulmonary bypass between January 2012 and December 2012 were studied. Patients with simple fossa ovalis atrial septal defect (ASD) were not included. The study was approved by the hospital research ethics board. Data collection included cardiac diagnosis, surgical procedure, epicardial echocardiography findings, surgical revision, surgical outcome, and postoperative transthoracic echocardiography findings.

Intraoperative epicardial echocardiography

All the cases of epicardial echocardiography were done using SONOS 7500 echo machine with S12 probe in infants, S8 probe in children, and S3 probe in adults. All cases were done by a single pediatric cardiologist. The probe was inserted in sterile gloves of adequate size. A small quantity of ultrasound gel was put in gloves in addition to applying on the probe head. It was ensured not to trap any air in between the probe and gloves. Multiple epicardial views were performed for comprehensive analysis. Parasternal long axis view was useful for assessment of ventricular septal defect (VSD), left ventricular outflow tract, aortic valve, mitral valve, and pulmonary venous return to left atrium. With modification of long axis view it was possible to visualize the right ventricular inflow and outflow. Short axis view of great vessels was useful to assess flow across the pulmonary valve annulus, branch pulmonary artery, and residual patent ductus arteriosus. Short axis view at the level of ventricle was used to assess the muscular VSD, shape of the ventricle and systolic ventricular function. In those patients with residual lesions detected by epicardial echocardiography, additional hemodynamic data was obtained. Parameters measured included saturation and pressure data in various chambers of the heart.

Postoperative transthoracic echocardiography

It was performed as a routine in all patients within 24 h of surgery, usually before extubation in the intensive therapy unit. Subsequent echo studies were done according to the need of the patient. As a routine pre discharge echo was done in all the patients.

RESULTS

A total of 290 pediatric and adult patients with congenital heart disease were operated between January 2012 and December 2012. In 224 of them open heart surgery was done. Epicardial echocardiography was indicated in 180 cases. However, epicardial echocardiography could be done in 158 patients. The ages ranged from 3 days to 44 years (median 5 years). The weight ranged from 2.5 to 65 kg (median 9 kg). The overall mortality was 12 cases (4.1%). The primary diagnosis of patients in whom epicardial echo was done is given in Table 1.

Table 1

Distribution of various surgical cases for whom epicardial echocardiography was done

Primary diagnosis	Number
ASD repair	8
Sinus venousus ASD	4
Ostium primum ASD	4
VSD repair	53
Perimembranous VSD	29
Muscular VSD	4
Inlet VSD	6
Swiss cheese VSD	7
Complete AV canal	5
Redo VSD patch dehiscence	2
Tetralogy of Fallot (intracardiac repair)	38
Total anomalous pulmonary venous connection repair	10
Mitral valve repair	14
Bidirectional Glenn shunt	5
Arterial switch operation	4
Partial anomalous pulmonary venous connection repair	4
Fontan operation	3
Aortopulmonary window	3
Supramitral stenosing ring	3
Rastelli repair	3
Truncus arteriosus repair	3
Subaortic membrane	2
Senning’s repair	2
Supravalvar aortic stenosis (William’s syndrome)	1
Right atrial myxoma repair	1
ALCAPA (translocation of LMCA to aorta)	1

ASD = Atrial septal defect; VSD = Ventricular septal defect; AV = Arterioventricular; ALCAPA = Anomalous origin of left coronary artery from pulmonary artery; LMCA = Left main coronary artery

Based on the epicardial echo findings, a total of 38 out of 158 cases were having residual lesions. Significant residual lesions were found in 28 cases. In 24 out of these 28 cases, immediate reoperation was done to correct the residual lesions. Table 2 describes these 24 cases. In remaining four cases who had significant residual lesion, it was decided to observe the patients.

Table 2

Residual postoperative lesions requiring immediate reoperation (n = 24)

Primary surgical procedure	Epicardial echocardiography findings (number of cases)	Reoperation
VSD patch closure (Swiss cheese VSD)	Significant residual shunt (3)	Patch closure of additional VSD
VSD patch closure (single VSD)	Significant residual shunt from patch edge (2)	Pledget suture closure of VSD shunt
VSD patch closure	Newly diagnosed additional VSD (3)	VSD patch closure
Intracardiac repair with trans annular patch	Significant RVOT stenosis caused by muscle bundle (6)	Excision of muscle bundle
Intracardiac repair without trans annular patch	Significant RVOT stenosis at the level of pulmonary annulus (2)	Trans annular patch
Intracardiac repair with LPA plasty	Significant LPA origin stenosis (1)	LPA patch augmentation
VSD closure, ICR, Rastelli	New onset severe tricuspid regurgitation (8)	Tricuspid valve repair (4)
Arterial switch operation	Branch pulmonary artery stenosis due to kincking at anastamosis site (2)	Surgical relief of pulmonary stenosis
IVC type of sinus venous ASD patch closure	Routing of IVC to left atrium (1)	Rerouting of IVC to right atrium

RVOT = Right ventricular outflow tract; VSD = Ventricular septal defect; ICR = Intracardiac repair; LPA = Left pulmonary artery; IVC = Inferior vena cava; ASD = Atrial septal defect

VSD

Out of 53 cases of VSD closure, there were significant residual VSD seen in 10 cases. In three of them had residual shunt across Swiss cheese VSDs, four were newly diagnosed additional VSD, and three had significant shunting through the edge of the patch. In four cases of VSD patch closure, there was tiny residual leak detected in the postoperative period which were not mentioned in the epicardial echocardiography report [Figures 1 and 2]. However, the shunt was nonsignificant and postoperative course was uneventful. There were two cases with significant residual VSD shunt detected by postoperative echocardiography. In both these cases epicardial echo was not done. In both these cases reoperation was done to close the residual VSD.

Figure 1

Epicardial echocardiography showing ventricular septal defect (VSD) patch with no residual shunt

Figure 2

Epicardial echocardiography showing VSD patch with residual VSD

New onset severe tricuspid regurgitation (TR)

It was seen in eight cases. In five out of these eight cases had undergone VSD patch closure and in remaining three cases ICR was done. The chordae of septal leaflet of the tricuspid valve had got trapped in VSD patch closure, resulting in non-coaptation between anterior and septal leaflet of tricuspid valve. Tricuspid valve repair to correct TR was done in four out of these eight cases. In remaining cases of new onset severe TR, there was gradual reduction in severity from severe to moderate.

New onset ventricular dysfunction

Significant left ventricular systolic dysfunction was observed in 14 cases on postoperative echocardiography. However, in none of these cases the epicardial echo showed left ventricular dysfunction. In 10 out of 14 patients there was left ventricular dysfunction and rest had biventricular dysfunction. It was an observation seen in adolescent and adults with large VSD who had undergone surgical patch closure. Left ventricular systolic dysfunction was noticed in 1st postoperative day which subsequently worsened. In an infant with large VSD who had undergone patch closure developed new onset severe ventricular dysfunction with mitral regurgitation which was not seen in the epicardial echo. On a detailed echocardiography evaluation it was diagnosed as a case of anomalous origin of left coronary artery from pulmonary artery (ALCAPA), which was not reported in preoperative and epicardial echocardiography. Preoperatively the child had normal left ventricle systolic function, large VSD with severe pulmonary arterial hypertension (PAH). After surgical patch closure of the VSD, the pulmonary arterial pressures significantly dropped resulting in coronary steal phenomena and new onset left ventricle dysfunction. In four cases, significant right ventricular dysfunction was noticed. Two had undergone repair for tetralogy of Fallot, one each for Rastelli and truncus repair.

Intracardiac repair (ICR)

A total of 38 cases of the ICR for tetralogy of Fallot were done. In four patients, there was a small residual shunt which was not considered significant. In eight cases, there was a significant right ventricular out flow tract obstruction. In 5/8 cases there was significant residual muscle bundle. In remaining 2/3, the pulmonary artery annulus was causing significant turbulence and there was significant gradient across the pulmonary valve. In remaining one case, there was left pulmonary artery origin stenosis which had persisted even after left pulmonary artery plasty. There was a significant right ventricular outflow tract obstruction noted in one case in postoperative echocardiography, which was not mentioned in the epicardial echo. Doppler echocardiography was not able to detect significant gradient. However, direct measurement of pressure in operation theatre showed near systemic right ventricular pressure. It was discovered that child had borderline sized pulmonary artery resulting in increased main pulmonary artery pressure and hence significant gradient was not detected in the right ventricular outflow tract.

ASD

In a total of eight cases of ASD (four ostium primum + four sinus venous defects) epicardial echo was done. Associated cleft repair was done. There was no significant residual mitral regurgitation seen. However, one case of inferior vena cava (IVC) type of sinus venosus ASD had postoperative desaturation. Epicardial echo confirmed direction of IVC flow to left atrium which was corrected.

Arterial switch operation (ASO)

In two out of five cases of ASO, there was severe stenosis of both branch pulmonary arteries. In one case, it was due to kinking of pulmonary artery anastamosis and in another there was narrowing of anastamosis site. In both these cases, the diagnosis was done by epicardial echocardiography and immediate repair was done.

Total anomalous pulmonary venous connection (TAPVC)

Out of 10 cases of TAPVC for whom epicardial echo was done, one case of infracardiac type of TAPVC had significant stenosis at the anastamosis site. Repair of stenosis was done immediately. In all the remaining cases the repair was adequate.

In one case of Senning, there was stenosis of systemic venous baffle that caused superior vena caval obstruction as detected by turbulent flow and mean gradient of 5 mmHg. In three out of seven cases reported to have mild mitral regurgitation after mitral valve repair by epicardial echocardiography, the severity increased to moderate. However, it did not need further repair.

In one patient of Glenn shunt, there was a significant obstruction at the anastamosis site. Two cases of VSD patch closure had significant residual VSD shunt. One case of patent ductus arteriosus (PDA) ligation had residual PDA flow seen in postoperative echocardiography. Epicardial echo was not done in all the above cases.

DISCUSSION

Overall, out of 158 cases for which epicardial echo was done, there was a significant residual lesion seen in 28 cases. In 24 out of these 28 cases, the lesions were operated upon. In the remaining four cases reported to have new onset severe TR in whom the residual lesion was not corrected, had poor outcome. In 10 cases in whom the residual lesions were reported to be nonsignificant and repair was not advised, all patients did well in postoperative period. In four out of 130 cases reported to have normal epicardial echo without residual shunt needed reoperation [Table 3]. In one case there was ALCAPA, detected in postoperative period. Second was a case of ICR with residual infundibular obstruction. Third was a case of mitral regurgitation which was reported as moderate, but progressed to severe mitral regurgitation. Fourth was a case of branch pulmonary artery stenosis in transposition of the great arteries (TGA) following ASO.

Table 3

New findings in postoperative (ICU) echocardiography which were not mentioned in epicardial echocardiography

Epicardial echocardiography	Postoperative echocardiography (number)
No residual shunt in VSD patch (35)	Mild residual shunt in VSD patch (4)
New onset moderate tricuspid regurgitation (5)	Reduced to mild regurgitation (3)
Normal left ventricular systolic function in large VSD closure (40)	New onset significant left ventricular systolic dysfunction (10)
ALCAPA not detected (1)	ALCAPA was diagnosed (1)
Well opened RVOT in ICR (29)	Significant RVOT stenosis (1)
Mild mitral regurgitation after mitral valve repair (7)	Increased to moderate mitral regurgitation (3)
*Epicardial echo not done (22)	*Significant residual VSD shunt (2)
	*Residual PDA shunt after PDA ligation (1)
	*Glenn shunt anastamosis site stenosis (1)

*Epicaridal echocardiography was not done; ICU = Intensive care unit; VSD = Ventricular septal defect; ICR = Intracardiac repair; RVOT = Right ventricular outflow tract; ALCAPA = Anomalous origin of left coronary artery from pulmonary artery; PDA = Patent ductus arteriosus

In eight out of 58 (14%) cases of VSD closure had either significant residual VSD or additional VSD diagnosed by epicardial echocardiography. In all these cases the diagnosis was confirmed and repair of residual lesion was done. In seven cases there was tiny to small residual shunt for which repair was not advised. In none of these cases there was increase in shunt severity or complications in postoperative period. Similarly, for ICR in tetralogy of Fallot, eight out of 38 cases had significant right ventricular outflow tract obstruction diagnosed by epicardial echocardiography. In all these cases the diagnosis was confirmed intraoperatively and operated upon. In four out of 38 cases there was small residual VSD and reoperation was not advised. In none of them there was increase in VSD shunting.

Hence, the overall sensitivity and specificity of epicardial echocardiography in detecting significant residual lesions is 86 and 100%, respectively. The positive predictive value and negative predictive value of epicardial echocardiography to detect significant residual lesion in immediate post cardiopulmonary bypass period was 100 and 97%, respectively. The percentage of false negative and false positive of epicardial echocardiography was 14 and 0%, respectively.

Intraoperative epicardial echocardiography is a well-established diagnostic modality to assess the adequacy of repair in surgery for congenital heart disease. However, in our center it is in use since 2010. Hence, it is important to document both the advantages and limitation of the technique before using it on a routine basis. Advantages were timely and accurate decision-making based on epicardial echocardiography in cases where there were significant residual lesions. As a result patient underwent reoperation in the same sitting and thus avoiding excessive complications and cost of surgery.

A recent study was done by Dragulescu et al., to evaluate visualization of branch pulmonary arteries, coronary arteries, Blalock Taussig shunt, Glenn/Fontan circuit, pulmonary veins/baffles, and residual VSDs by epicardial echocardiography and TEE.[13] It was concluded that epicardial echocardiography detects residual intraoperative lesion not visualized by TEE. In most of these cases, pulmonary arteries were not well-visualized by TEE. The use of epicardial echocardiography may decrease the need for early reintervention.

In a study done by Narayanswami et al., on comparative roles of intraoperative epicardial and early postoperative transthoracic echocardiography it was detected that epicardial echocardiogram correctly identified six out of seven patients with residual lesion that needed revision.[14] However, they noted that in three patients there was underestimation of left atrioventricular valve regurgitation.

Ungerleider et al., did a study to evaluate the exactness of surgical repair using epicardial echo in children with congenital heart disease it was found that the rate of reoperation and early death was significantly higher in those with unrepaired residual lesion detected by epicardial echo.[4] Accordingly in those subset of patients in whom the residual lesion was corrected based on epicardial echocardiography had no postoperative events with good long-term results. Another study done by the same author Ungerleider et al., it was found that in 18% of patients in whom pre-bypass epicardial echo was done, there were significant newly detected lesions.[10] After bypass 7% of patients had unsuspected residual lesion in whom immediate revision was done. Thus authors concluded that routine intraoperative epicardial echocardiography was useful in aiding, planning, conduct, and assessment of results in operations for congenital heart disease.

In a study done by Kaushal et al., to evaluate the role of routine use of intraoperative echocardiography to assess surgical repair, it was found that intra operative echocardiography was inexpensive, accurate, valuable and safe.[15] It should be made mandatory for all corrective surgical procedures for congenital heart disease.

There were no comparative studies between epicardial echo and micro TEE probe to assess the adequacy of surgical repair for congenital heart disease. Zyblewski et al., have reported that micro TEE provides high quality images without hemodynamic or ventilation compromise in small infants undergoing cardiac operations.[16]

CONCLUSION

Routine use of intraoperative epicardial echocardiography allows detection of majority of residual defects. Immediate surgical revision of such lesion results in excellent outcome, thus saving them from reoperation and postoperative complications. In our study we have demonstrated a very high sensitivity and specificity along with high positive and negative predictive value of epicardial echocardiography in detecting significant residual lesions. However, it is possible to miss out on certain unsuspected lesions such as ALCAPA. Rarely the epicardial echocardiography may underestimate the mitral valve regurgitation in cases of mitral valve repair. Similarly, residual gradient in right ventricular outflow tract (RVOT). In a significant number of cases of operated large VSD with PAH, it was observed that on 1st postoperative day there was significant left ventricular dysfunction which was not observed at the time of epicardial echocardiography. In our experience and that of others this method has not resulted in complications such has infection or serious arrhythmias requiring cardioversion.[17] We recommend that epicardial echocardiography is accurate, economical, and safe modality of investigation; it should be used routinely in all patients undergoing corrective surgical procedure. Thus, it possible to avoid complications and limitations in using TEE in neonates and children.

EVALUATION OF A SCREENING TEST

The following measures are used to evaluate a screening test:

Sensitivity = a/(a + c) × 100

Specificity = d/(b + d) × 100

Predictive value of a positive test = a/(a + b) × 100

Predictive value of a negative test = d (c + d) × 100

Percentage of false of negatives = c/(a + c) × 100

Percentage of false positive = b/(b + d) × 100

Statistics:
Screening test results	Diseased	Not diseased	Total
Positive	A (true positive)	B (false positive)	A + B
Negative	C (false negative)	D (true negative)	C + D
Total	A + C	B + D	A + B + C + D