Video-Assisted Cardioscopy for IntraventricularRepair in Congenital Heart DiseaseKagami Miyaji, MD, Robert L. Hannan, MD, Jorge Ojito, BS, James M. Dygert, MD,Jeffrey A. White, MS, and Redmond P. Burke, MDDepartment of Cardiovascular Surgery, Miami Children’s Hospital, Miami, Florida

Background. Video-assisted thoracoscopic surgicaltechniques have been widely adopted as a means toreduce surgical trauma. By adapting pediatric thoraco-scopic instrumentation, we have developed a techniquefor video-assisted cardioscopy (VAC). We report ourexperience and describe the technical feasibility of VAC.

Methods. Since June 1995, 409 consecutive patientsunderwent 431 intracardiac procedures (ventricular sep-tal defect, 150; tetralogy of Fallot or double outlet rightventricle, 101; atrioventricular canal, 52; subaortic steno-sis, 43; valve repair, 50; Rastelli procedure, 12; Konno orRoss Konno operation, 11; and miscellaneous, 12) usingVAC at Miami Children’s Hospital. Using a prospectivedatabase, we tracked outcomes and operative events todelineate the usefulness and efficacy of this technique.

Results. VAC provided clear and precise imaging ofsmall or remote intracardiac structures during repair of

congenital heart defects without technical complications.Procedure times and aortic cross-clamp times using VACwere not prolonged. Intraoperative images were col-lected for every operation, documenting each patient’scardiac anatomy before and after repair. Surgery throughsmall incisions was facilitated. Operative mortality was1.2% (5 of 409), and no patient required reoperationbefore discharge. At a mean follow-up interval of 22months, the incidence of reoperation for residual orrecurrent lesions was 1.2% (5 of 404).

Conclusions. Our experience demonstrates the techni-cal feasibility and clinical utility of routine endoscopicimaging during open heart surgery for congenital heartrepair.

(Ann Thorac Surg 2000;70:730–7)© 2000 by The Society of Thoracic Surgeons

Video-assisted thoracoscopic surgical techniques arenow commonly used to improve anatomic visual-

ization and reduce surgical trauma in general thoracic [1]and cardiovascular surgery [2, 3]. Video-assisted endo-scopic techniques in congenital heart surgery have beendescribed for patent ductus arteriosus interruption [4–7]and vascular ring division [8–10]. These procedures dem-onstrate the efficacy of video-assisted thoracoscopy inproviding pediatric thoracic access and exposure withinconfined and delicate anatomic spaces, ie, patent ductusinterruption in premature newborns.

Open heart operations for congenital heart disease inneonates and infants also require clear visualization ofsmall structures within confined spaces. By adaptingpediatric thoracoscopic instrumentation, we developed atechnique for routine video-assisted cardioscopy (VAC)during open heart repair of congenital heart disease [11].This technique was initially used in selected operationsto visualize remote intracardiac structures and to facili-tate repairs while avoiding the need for extended cardiacincisions or vigorous cardiac retraction. We subsequentlyfound that routine use of these techniques could be usedto enhance operative visualization and precision.

Patient demand for less traumatic, more cosmetic sur-

gery has increased, particularly for congenital heart sur-gery. Several authors have reported less-invasive proce-dures for repair of atrial septal defect (ASD) [12, 13] andventricular septal defect (VSD) [14]. Many of these oper-ations have come under scrutiny for violating key surgi-cal tenets of exposure and precision. The VAC techniquecan provide clear and precise visualization of smallintracardiac structures within a limited space. This tech-nique is effective and necessary to achieve completerepair of more complex intraventricular lesions in mini-mally or less invasive congenital heart surgery.

Since June 1995, 409 consecutive patients underwentintracardiac repair using a VAC technique to exposeremote or small intracardiac structures and facilitatesurgical repair in Miami Children’s Hospital. Here, wereport our experience and clarify the usefulness andefficacy of this technique.

Material and Methods

Cardioscope EquipmentPediatric VAC was performed with the same equipmentused in pediatric video-assisted thoracic operations [5–7].Videoscopes (Smith and Nephew, Dyonics, Inc, Andover,MA) were chosen based on size (4-mm diameter with a7-cm working length) and the angle at the camera face (30degrees) and produced 43 magnification. Endoscopicinstruments were not required. In all cases, the video

Accepted for publication Mar 8, 2000.

Address reprint requests to Dr Burke, Department of CardiovascularSurgery, Miami Children’s Hospital, 3200 SW 60 Ct, Suite 102, Miami, FL33155-4069; e-mail:redmondlll@aol.com.

© 2000 by The Society of Thoracic Surgeons 0003-4975/00/$20.00Published by Elsevier Science Inc PII S0003-4975(00)01497-1

equipment was on the operative field before cardiopul-monary bypass was initiated.

Patient CharacteristicsBetween June 1995 and April 1999, 1,056 congenital car-diovascular operations were performed at Miami Chil-dren’s Hospital, Miami, Florida. The VAC technique wasused for 431 consecutive intraventricular procedures in409 patients (Table 1). The patients’ ages ranged from 2days to 38 years (mean 3.4 years) and their weight rangedfrom 2.2 to 80.0 kg (mean 14.1 kg). All intraventricularrepairs were classified into one of the following eightcategories: (1) ventricular septal defect (VSD) closure; (2)tetralogy of Fallot (TOF) and pulmonary atresia with VSD(PA/VSD) repair including double outlet of right ventri-cle (DORV) repair; (3) atrioventricular canal (AVC) re-pair; (4) subaortic stenosis (SAS) repair; (5) valvuloplasty;(6) Rastelli and reparation a l’etage ventriculaire (REV)procedure; (7) Ross and Ross/Konno procedure; and (8)miscellaneous procedures.

VSD ClosureVSD closure was performed in 150 patients. Patient ageranged from 4 days to 18 years (mean 2.9 years) and bodyweight ranged from 2.3 kg to 76.8 kg (mean 13.0 kg). Thetypes of VSD and associated procedures are shown inTable 2. Once the heart was arrested, the cardioscope was

inserted into the right ventricle through the right atri-otomy, pulmonary artery, or right ventriculotomy. VACwas then used to expose and clearly define the bounds ofthe VSD and suture line for closure.

TOF and PA/VSD RepairTOF and PA/VSD repair was performed in 101 patients.Patient age ranged from 2 days to 22 years (mean 1.6years) and body weight ranged from 2.7 kg to 46 kg (mean9.4 kg). Types of TOF and associated procedures areshown in Table 3. The cardioscope was inserted into theright ventricle cavity through a ventriculotomy afterheart arrest was achieved. The VAC was used to exposeboth the right ventricular outflow tract (RVOT) and theVSD, and to decide the resection of RVOT obstructedmuscle and the suture line for VSD closure.

AVC RepairAVC repair was performed in 52 patients. Patient ageranged from 40 days to 13 years (mean 1.4 years) andbody weight ranged from 2.7 kg to 52 kg (mean 8.2 kg).The types of AVC include complete AVC (36) and tran-sitional AVC (16). Associated procedures were PDA liga-

Table 1. Patient Data

VSD repair 150TOF repair 88DORV repair 7TOF/AVC repair 6Complete AVC repair 52SAS repair 43Valve repair 50Rastelli/REV operation 12Konno, Ross/Konno 11Miscellaneous 12Total 431

AVC 5 atrioventricular canal; DORV 5 double outlet right ventricle;SAS 5 subaortic stenosis; TOF 5 tetralogy of Fallot; VSD 5ventricular septal defect.

Table 2. VSD Repair

Number ASO Arch IAA/APW PTA ASD PDA DCRV PAP CoA SAS VP RVOTR

VSD I 13 1 1 3 2 1 1 1VSD II 113 5 4 6 32 24 17 8 2 4 3VSD III 4 1 2 1VSD IV 3 1 1 2 1Multiple VSD 9 4 2 3 1Residual VSD 5 1 1 3L-TGA VSD 3 1 2 1 1Total 150 6 6 6 1 43 31 18 14 6 5 4 4

Arch 5 aortic arch repair; ASD 5 atrial septal defect closure; ASO 5 arterial switch operation; CoA 5 coarctation of aorta repair; DCRV5 double chamber of right ventricle repair; IAA/APW 5 interruption of aortic arch or aortopulmonary window repair; PAP 5 pulmonary arteryplasty; PDA 5 patent ductus arteriosus closure; PTA 5 truncus arteriosus repair; RVOTR 5 right ventricular outflow tract reconstruction;SAS 5 subaortic stenosis repair; VP 5 valve plasty; VSD 5 ventricular septal defect.

Table 3. TOF and PA/VSD Repair

NumberPA

PlastyPV

Plasty ASD PDA UF

Non-TAP 25 4 3 4 1TAP 41 4 6 5PA/VSD (TAP) 4 1PA/VSD (homograft) 7 1 1PA/VSD (staged) 8 5 2TOF/AVC 6 1TOF absent PA valve 3 2 2DORV 7 1 1 2Total 101 18 4 14 9 1

ACV 5 atrioventricular canal; ASD 5 atrial septal defect closure;DORV 5 double outlet of right ventricle; PA plasty 5 pulmonaryartery plasty; PA valve 5 pulmonary valve; PA/VSD 5 pulmonaryatresia and ventricular septal defect; PDA 5 patent ductus arteriosusclosure; PV plasty 5 pulmonary valve plasty; TAP 5 transannularpatch; TOF 5 tetralogy of Fallot; UF 5 unifocalization of pulmo-nary arteries.

731Ann Thorac Surg MIYAJI ET AL2000;70:730–7 VAC IN CONGENITAL HEART SURGERY

tion or division (16), atrioventricular valve repair (12),ASD secumdum closure (9), pulmonary artery plasty (5),RVOTR (4), and aortic arch repair (1). The cardioscopewas inserted into the right ventricle through a rightatriotomy after heart arrest was achieved. The VAC wasused to expose the VSD component of AVC and AV valveapparatus including chordal structures in order to deter-mine the dividing line of the anterior and posteriorleaflets of the common AV valve and the suture line forVSD patch closure. Before closing the ASD component,we inserted the cardioscope into the left atrium throughthe ASD and a left-sided AV valve competence test usingcold saline installation without any retraction was per-formed and used in all cases.

SAS RepairFor SAS, resection of transaortic subaortic membranewas performed in 43 patients. A minimally invasiveprocedure using a partial upper sternotomy was per-formed in 15 patients [15], and a conventional approachusing full median sternotomy was used in 28. Amongthese 28 patients, 14 underwent the following associatedprocedures: mitral valvuloplasty (6), VSD closure (5),aortic arch repair (2), aortic valvuloplasty (2), ASD clo-sure (1), and PDA ligation (1). Patient age ranged from 2months to 26 years (mean 6.3 years) and body weightranged from 5.4 kg to 68 kg (mean 24.2 kg). After goodexposure of the aortic valve leaflet, the VAC was appliedto visualize the subaortic lesion. Inspection of the sub-aortic area using the VAC technique was performed toconfirm the absence of any obstructive lesion.

ValvuloplastyValvuloplasty was performed in 50 patients (Table 4).Patient age ranged from 1 month to 28 years (mean 6.2years) and body weight ranged from 3.5 kg to 80 kg (mean21.2 kg). Before and after the valve repair procedure, avalve competence test using cold saline installation was

performed without any retraction in all cases and theVAC technique was applied to check valve competence.

Rastelli and REV ProcedureRastelli and REV procedures were performed in 12 pa-tients. Patient age ranged from 5 days to 8 years (mean 2.6years) and body weight ranged from 2.2 kg to 31.2 kg(mean 12.0 kg). The Rastelli operation was performed in9 patients and the REV procedure in 3 patients. Associ-ated procedures were VSD closure (12), ASD closure (9),VSD enlargement (5), Damus-Kay-Stansel (DKS) anasto-mosis (1), RVOTR (1), PA plasty (1), and PDA ligation ordivision (1). The cardioscope was inserted into the rightventricle through a ventriculotomy after heart arrest wasachieved. The VAC was then used to expose the anatomicrelationships in the aortic valve, VSD and tricuspid valveapparatus including chordal structures. We determinedthe suture line for VSD closure and rerouting to the aorticvalve, avoiding injuries to the aortic valve and conduc-tion system. The VAC confirmed that stitches were in theproper position during the procedure and that the VSDwas completely closed and the LVOT wide open after theprocedure.

Konno and Ross/Konno ProceduresThe Konno aorticoventriculoplasty or Ross/Konno proce-dure was performed in 11 patients. Patient age rangedfrom 3 months to 38 years (mean 10.0 years) and bodyweight ranged from 5.4 kg to 64.8 kg (mean 29.8 kg). TheKonno aorticoventriculoplasty was performed in 3 pa-tients with subaortic stenosis, and 2 of them underwentaortic valve replacement. The Ross/Konno procedurewas performed in 8 patients. The aorticoventriculoplastywas performed using a Dacron graft patch. The cardio-scope was inserted into the left ventricle through theaortotomy, after heart arrest had been achieved. VACwas used to expose the subaortic lesion and define theincision line of the ventricular septum for ventriculo-plasty. Inspection of the subaortic area was also per-formed to confirm the absence of any obstructive lesionafter the procedure.

Miscellaneous ProceduresA thrombectomy was performed in 4 patients (leftatrium, 2; left ventricle, 1; and right atrium, 1). After aorticclamping, the cardioscope was inserted into the leftventricle through a transverse incision of the ascendingaorta for left ventricle thrombectomy [16]. VAC was usedto expose the intracardiac thrombus and to make surethat there was no remnant of thrombus. Video-assistedintraoperative stenting for branch pulmonary artery ste-nosis was performed in 6 patients and a pulmonaryartery stent removal in 1. Associated procedures were PAplasty (7), RVOTR (6), VSD closure (5), ASD closure (1).VAC was used to expose the pulmonary stenosis andmake sure that the stent was positioned properly duringand after the procedure. Coronary artery fistula repairwas performed in 1 patient.

Table 4. Valvuloplasty

NumberASD(2) PDA CoA VSD SAS MVR

Partial AVC with MR 17 2 2 1MR s/p AVC repair 11 (1) 1 1

MR 7 1 1 5MS 3 1

Ebstein’s anomaly(TR)

3 2

CTGA (Lt. AVVR) 2AR 5 3 1AS 2Total 50 5 2 1 4 8 1

AR 5 aortic valve regurgitation; AS 5 aortic stenosis; ASD (2) 5atrial septal defect closure; AVC 5 atrioventricular canal; CoA 5coarctation of aorta repair; CTGA 5 corrected transposition of greatartery; Lt. AVVR 5 left atrioventricular valve regurgitation; MR 5mitral valve regurgitation; MS 5 mitral stenosis; MVR 5 mitralvalve replacement; PDA 5 patent ductus arteriosus closure; SAS5 subaortic stenosis TR 5 tricuspid regurgitation; VSD 5 ventric-uar septal defect.

732 MIYAJI ET AL Ann Thorac SurgVAC IN CONGENITAL HEART SURGERY 2000;70:730–7

Study ProtocolWe reviewed the postoperative transesophageal echocar-diograms (TEE) and transthoracic echocardiograms(TTE) and mortality and morbidity related to the above-mentioned procedures. Procedure time and aortic crossclamp time were also reviewed to evaluate the influenceof VAC upon the procedure itself. All data are expressedas mean 6 standard deviation (SD).

Results

VSD ClosureIntraoperative VAC provided clear vision of all VSDs in150 patients (Fig 1). VAC was also used to make sure thatstitches were in proper position during procedures, andthe procedure made it easy to close the VSD withoutinjury to the aortic valve or conduction system (Fig 1).After the procedure, VAC was used to make sure that the

VSD was completely closed. Intraoperative TEE andpostoperative TTE findings are shown in Table 5. Smallresidual VSDs were found in 12 patients (8%; 5 VSD typeII and 7 multiple VSD). There were 2 hospital deaths(1.3%); 1 patient died of postoperative vascular thrombo-sis after ASO with VSD closure and the other of pulmo-nary hypertension. The procedure times and aortic crossclamp times are shown in Table 6 for 73 patients withsimple VSD closure. The mean follow-up period was 22months, and during this period no patient neededreoperation.

TOF and PA/VSD RepairIntraoperative VAC provided a clear view of malalign-ment VSD in all 101 patients. The VAC technique madeit clear that stitches were in proper position during theprocedure (Fig 2A), and made it easy to close the VSDwithout injuring aortic valve and conduction system.After the procedure the VAC was used to confirm thatthere was no right ventricular outflow tract obstruction(RVOTO) and that the VSD was closed completely. ForDORV, in order to determine the absence of left ventric-ular outflow tract obstruction (LVOTO) and suture lineplacement of the left ventricle to aorta graft patch, thecardioscope was inserted inside of the Dacron grafttunnel (Fig 2B). Intraoperative TEE and postoperative

Fig 1. Representative picture of perimembranous ventricular septaldefect taken using video-assisted cardioscopy (VAC). The aorticvalve can be seen clearly by inducing cardioplegia. The VAC con-firmed that stitches were in the proper position during the procedureand made it easy to close the defect without injuring the aortic valveand conduction system.

Table 5. Results

No VSDLeakage

PermanentCAVB

Aortic ValveInjury

No RVOTOor LVOTO

Less ThanMild VR

EarlyReoperation

VSD (n 5 150) 138 (92.0%) 1 (0.7%)a 0 — — 0TOF (n 5 101) 97 (96.0%) 0 0 90 (89.1%) — 0AVC (n 5 52) 48 (92.3%) 1 (1.9%) 0 — 51 (98.1%) 1 (1.9%)SAS (n 5 43) — 0 0 41 (95.3%) — 0Valvuloplasty (n 5 50) — — — — 49 (98%) 1 (2.0%)Rastelli/REV (n 5 12) 12 (100%) 1 (8.3%)b 0 10 (83.3%) — 0Konno/Ross-Konno

(n 5 11)10 (90.9%) 0 — 0 — 0

a One patient with multiple VSDs had permanent heart block needed pace maker implantation. b Patient who underwent a Rastelli operation forcorrected transposition of great arteries with pulmonary stenosis and VSD.

AVC 5 atrioventricular canal; CAVB 5 complete atrioventricular block; LVOTO 5 left ventricular outflow tract obstruction; RVOTO 5 rightventricular outflow tract obstruction; SAS 5 subaortic stenosis; TOF 5 tetralogy of Fallot; VR 5 valve regurgitation; VSD 5 ventricularseptal defect.

Table 6. Procedure and Aortic Cross-Clamp Time

Procedure Time(min)

ACC Time(min)

VSD (n 5 76) 135 6 33 37 6 13TOF (n 5 66) 171 6 32 60 6 19AVC 181 6 33 74 6 22SAS 167 6 53 41 6 14Valvuloplasty 182 6 49 61 6 22Rastelli/REV 297 6 65 108 6 24Konno/Ross-Konno 364 6 100 154 6 21

ACC 5 aortic cross clamp; AVC 5 atrioventricular canal; REV 5reparation a l’etage ventriculaire; SAS 5 subaortic stenosis; TOF5 tetralogy of Fallot; VSD 5 ventricular septal defect.

733Ann Thorac Surg MIYAJI ET AL2000;70:730–7 VAC IN CONGENITAL HEART SURGERY

TTE findings are shown in Table 5. Small residual VSDswere found in 4 patients. Mild RVOTO less than25 mm Hg was found in 11 patients. There were nohospital deaths. The procedure times and aortic crossclamp times are shown in Table 6 for 66 patients withTOF including TAP and non-TAP repair. During a meanfollow-up period of 21 months, 1 patient required reop-eration. This patient underwent PA/VSD repair withRVOTR using a homograft. Postoperative echocardiogra-phy revealed no residual VSD and RVOTO, and thepatient was discharged home. One month later, a resid-

ual VSD was found on the follow-up echocardiographyand required VSD closure.

AVC RepairIntraoperative VAC was utilized in all 52 AVC patientsand clearly exposed the VSD component of the AVC andAV valve apparatus including the chordal structures.This technique was very helpful in determining thedividing line of the anterior and posterior leaflets of thecommon AV valve and the suture line for VSD patchclosure, preserving AV valve function and avoiding in-jury of the conduction system. The VAC also made surethat stitches were in the proper position during theprocedure and that the AVC was completely repairedafter the procedure. Before closure of the ASD compo-nent, a left-sided AV valve competence test was per-formed and visualized using the VAC technique. TheVAC provided a clear view of valve competence withoutretraction of the right atrium (Fig 3). Valve incompetencewas found in 12 patients, requiring valve repair. Onepatient needed a valve replacement because the VACrevealed severe left-sided valve regurgitation after valverepair. Intraoperative TEE and postoperative TTE find-ings are shown in Table 5. Minor residual VSDs werefound in 3 patients, with 1 patient needing reoperationfor residual VSD closure. One patient who had unbal-anced AVC with a hypoplastic left ventricle and neededaortic arch repair died because of heart failure andinfection. The procedure times and aortic cross clamptimes are shown in Table 6. The mean follow-up periodwas 25 months, and during that period 1 patient neededreoperation (valve repair).

SAS RepairIntraoperative VAC provided a clear view of the SAS inall 43 patients. After exposing the aortic valve, the car-dioscope clearly revealed the subaortic lesion. This tech-

Fig 2. (A) Representative picture of a malalignment ventricular sep-tal defect (TOF type) taken using video-assisted cardioscopy (VAC).The overriding aortic valve can be seen clearly by inducing cardio-plegia. The VAC confirmed that stitches were properly positionedduring the procedure and made it easy to close the defect withoutinjuring the aortic valve and conduction system. (B) Representativepicture of the inside of the Dacron graft tunnel in double-outlet rightventricle repair. There was no left ventricular outflow tract obstruc-tion, and the suture line of left ventricle to aorta graft patch wasperfect.

Fig 3. Before closure of an atrial septal defect component, a left-sided atrioventricular valve competence test was performed, and vid-eo-assisted cardioscopy provided a clear view of valve competencewithout retraction of the right atrium.

734 MIYAJI ET AL Ann Thorac SurgVAC IN CONGENITAL HEART SURGERY 2000;70:730–7

nique was very useful to resect the subaortic fibromem-brane using blunt and sharp dissection, while avoidinginjury of the aortic valve and conduction system [15].Inspection of the subaortic area using the VAC techniqueconfirmed the absence of any obstructive lesion. Postop-erative echocardiographic findings are shown in Table 5.Two patients had moderate residual SAS (35 mm Hg and40 mm Hg), but there were no deaths. The proceduretimes and aortic cross clamp times are shown in Table 6.The mean follow-up period was 21 months, and duringthat period no patient required reoperation.

ValvuloplastyThe VAC provided an excellent view of the valve appa-ratus in 50 patients during the repair. After completion ofthe procedure, a valve competence test was performed,and the VAC technique was used to provide confirmationof valve competence without any retraction of the left andright atrium or aortic wall. Severe valve incompetencewas found in 1 patient postoperatively, and a mitral valvereplacement was performed. Intraoperative TEE andpostoperative TTE findings are shown in Table 5. Forvalve stenosis, all patients had less than mild valvarstenosis. There were 2 hospital deaths. One patient whohad critical aortic stenosis died of postoperative neuro-logic complication and another with severe Ebstein’sanomaly died of postoperative cardiac failure. The pro-cedure times and aortic cross clamp times are shown inTable 6. The mean follow-up period was 22 months,during which no patient underwent reoperation.

Rastelli and REV ProcedureThe VAC was used to view the anatomical relationship ofthe aortic valve, VSD and tricuspid valve apparatusincluding chordal structures in 12 patients. In 5 patients,a VSD enlargement was needed due to restrictive VSD.This technique was useful in deciding on the suture linefor the VSD closure and rerouting to aortic valve, whileavoiding injuries to the aortic valve and conductionsystem. The VAC also confirmed that stitches were in theproper position during the procedure, and made it easyto close the VSD completely. After the procedure, thecardioscope was inserted into the buffle graft in order tocheck for LVOTO and the suture line of the buffle graft(Fig 4). Intraoperative TEE and postoperative TTE find-ings are shown in Table 5. There were no hospital deaths.Procedure times and aortic cross clamp times ae shownin Table 6. The mean follow-up period was 21 months,and during this period no patient needed reoperation.

Konno and Ross/Konno ProcedureThe VAC provided a clear view of subaortic lesions andwas helpful while determining the incision line of theventricular septum for ventriculoplasty, avoiding theconduction system (Fig 5). The VAC also made sure thatstitches were in the proper position during these proce-dures and that the ventriculopalsty patch was completelyclosed. Inspection of the subaortic area using VAC pro-vided a picture clear and precise enough to verify thatthere was no obstructive lesion in the LVOT after the

procedure. Intraoperative TEE and postoperative TTEfindings are shown in Table 5. There were no hospitaldeaths. Procedure times and aortic cross clamp times areshown in Table 6. The mean follow-up period was 20months, during which no patient needed reoperation.

Miscellaneous ProceduresVAC provided a clear intracardiac view and was useful inremoving the intracardiac thrombus. After the proce-dure, the VAC confirmed that there was no residualthrombus remnant [16]. Postoperative TTE also revealedthat there was no intracardiac thrombus in these pa-tients. VAC was also useful in removing the previousstent device from the left distal pulmonary artery. The

Fig 4. Representative picture of the inside of Dacron buffle graftpatch in a reparation a l’etage ventriculaire procedure with ven-tricular septal defect enlargement. There was no left ventricular out-flow tract obstruction.

Fig 5. Representative picture of Konno ventriculoplasty. Video-as-sisted cardioscopy provided a clear view of the subaortic lesion andwas used to determine the incision line of the ventricular septum,avoiding the conduction system, and mitral and tricuspid papillarymuscles.

735Ann Thorac Surg MIYAJI ET AL2000;70:730–7 VAC IN CONGENITAL HEART SURGERY

cardioscope was also used during stenting of the pulmo-nary artery, providing a clear picture of the pulmonarystenosis and assuring proper position of the stent device.After the procedure, the cardioscope was inserted intothe pulmonary artery distal to the stent and used toconfirm that the device was positioned in the properposition and that the pulmonary stenosis lesion wasdilated. The intraoperative TEE or postoperative TTEshowed there was no residual pulmonary stenosispresent in these patients. For coronary artery fistularepair, VAC provided a clear vision of the fistula andmade it possible to find the discrete opening of the fistulainto the right ventricle. After closure of the fistula, VACwas used to confirm the complete closure by means ofcardioplegia through the coronary artery. The intraoper-ative TEE revealed no residual coronary fistula.

Overall ResultsThe VAC technique was applied to a total of 409 patientsand 431 procedures. There were 5 hospital deaths (1.2%).The mean follow-up period is 22 months, and during thisperiod there were 6 late deaths (1.5%) and only 5 patientswho needed reoperation (1.2%).

Comment

Experience with endoscopic techniques for PDA inter-ruption and vascular ring division in neonates led us toconsider using video assistance during congenital openheart operations when difficult anatomic situations wereanticipated preoperatively or encountered intraopera-tively. In 1994, we reported that video-assisted cardios-copy is feasible for imaging small, inaccessible structuresduring repair of complex congenital heart defects [11].VAC allows atraumatic visualization and magnificationof inaccessible structures while avoiding vigorous cardiacmanipulation and extended incisions.

Since June 1995, the VAC technique has been appliedto intracardiac procedures at Miami Children’s Hospital.VAC provided clear and precise imaging of intracardiacstructures in all cases. This technique was also a veryuseful aid to understanding the anatomy of intracardiaclesions such as malalignment VSD and subaortic steno-sis. The surgical results using VAC were excellent, and nopatients needed reoperation before being dischargedhome, although 5 patients needed reoperation during thefollow-up period (22 months). Potential problems withVAC include prolonged procedure time, prolonged car-diac arrest time, valve laceration, and ventricular or atrialwall perforation. To prevent trauma, the rigid videoscopemust be advanced along a straight anatomic path avoid-ing cardiac distortion. The procedure time and aorticcross-clamping time using the VAC technique were ac-ceptable in all cases. There were no valve lacerations andno ventricular or atrial perforations detected in ourseries.

Video-assisted minimally invasive cardiac surgery isbecoming more and more common in adult cardiacsurgery. This surgical technique has been applied in themanagement of intracardiac lesions, including mitral

valve procedures, as well as coronary artery bypassgrafting [2, 3, 17]. In congenital cardiac surgery, video-assisted thoracoscopic PDA interruption [4–7] and vas-cular ring division [8, 10] were reported as producingexcellent results. Recently, since improved outcomes andless morbidity have been experienced, minimally inva-sive procedures for ASD [12, 13] and VSD [14] have beenreported. As mentioned, the VAC technique allowsatraumatic visualization and magnification of inaccessi-ble structures without vigorous cardiac manipulation andextended incisions.

Minimally invasive surgery with this technique can bea sophisticated and useful procedure, especially in com-plex congenital cardiac surgery. Since March 1997 wehave performed resection of the subaortic membrane viaan upper sternotomy as a minimally invasive procedure.This approach is basically the same as minimally invasiveaortic valve surgery in adults. However, as the operativefield is limited and smaller than that of an adult, the VACtechnique was used to expose the subaortic lesion. Re-cently, video-assisted robotic surgery was reported for amitral valve procedure and coronary artery bypass graft-ing [18–20]. This new technology may be applied tocongenital cardiac surgery in the near future, althoughthere are several problems to be solved such as access forcardiopulmonary bypass in small children. We believethat VAC will play an important role in robot-assistedminimally invasive adult and pediatric cardiac surgery inthe future.

In conclusion, 409 consecutive underwent intracardiacrepair using the VAC technique. Our experience showedits technical feasibility for imaging small, inaccessibleintracardiac structures and for obtaining excellent surgi-cal results without any valve laceration or perforation.Long-term follow-up will be necessary to ensure that theexcellent short-term results continue.

References

1. Kohno T, Murakami T, Wakabayashi A. Anatomic lobectomyof the lung by means of thoracoscopy. An experimentalstudy. J Thorac Cardiovasc Surg 1993;105:729–31.

2. Mack MJ, Acuff TE, Casimir-Ahn H, et al. Video-assistedcoronary bypass grafting on the beating heart. Ann ThoracSurg 1997;63:S100–3.

3. Miyaji K, Wolf RK, Flege JB Jr. Surgical results of video-assisted minimally invasive direct coronary artery bypass.Ann Thorac Surg 1999;67:1018–21.

4. Laborde F, Noirhomme P, Karam J, et al. A new video-assisted thoracoscopic surgical technique for interruption ofpatient ductus arteriosus in infants and children. J ThoracCardiovasc Surg 1993;105:278–80.

5. Burke RP, Video-assisted thoracoscopic surgery for patentductus arteriosus. Pediatrics 1994;93:823–5.

6. Laborde F, Folliguet T, Batisse A, et al. Video-assistedthoracoscopic surgical interruption: the technique of choicefor patent ductus arteriosus. Routine experience in 230pediatric cases. J Thorac Cardiovasc Surg 1995;110:1681–5.

7. Burke RP, Wernovsky G. Images in clinical medicine. Tho-racoscopic clipping of patent ductus arteriosus. N EnglJ Med 1997;336:185.

8. Burke RP, Chang AC. Video-assisted thoracoscopic divisionof a vascular ring in an infant: a new operative technique.J Card Surg 1993;8: 537–40.

736 MIYAJI ET AL Ann Thorac SurgVAC IN CONGENITAL HEART SURGERY 2000;70:730–7

9. Burke RP, Rosenfeld HM, Wernovsky G, Jonas RA. Video-assisted thoracoscopic vascular ring division in infants andchildren. J Am Coll Cardiol 1995;25:4, 943–7.

10. Burke RP, Wernovsky G, van der Velde M, et al. Video-assisted thoracoscopic surgery for congenital heart disease.J Thorac Cardiovasc Surg 1995;109:499–508.

11. Burke RP, Michielon G, Wernovsky G. Video-assisted car-dioscopy in congenital heart operations. Ann Thorac Surg1994;58:864–8.

12. Black MD, Freedom RM. Minimally invasive repair of atrialseptal defects. Ann Thorac Surg 1998;65:765–7.

13. Barbero-Marcial M, Tanamati C, Jatene MB, et al. Transxi-phoid approach without median sternotomy for the repair ofatrial septal defects. Ann Thorac Surg 1998;65:771–4.

14. Lin PJ, Chang CH, Chu JJ, et al. Minimally invasive cardiacsurgical techniques in the closure of ventricular septal defect:an alternative approach. Ann Thorac Surg 1998;65:165–70.

15. Miyaji K, Ojito J, White JA, et al. Minimally invasive resec-

tion of congenital subaortic stenosis. Ann Thorac Surg 2000;69:1273–5.

16. Mazza IL, Jacobs JP, Aldousany A, et al. Video-assistedcardioscopy of left ventricular thrombectomy in a child. AnnThorac Surg 1998;66:248–50.

17. Chitwood WR Jr, Wixon CL, Elbeery JR, et al. Video-assistedminimally invasive mitral valve surgery. J Thorac CardiovascSurg 1997;114:773–82.

18. Stephenson ER Jr, Sankholkar S, Ducko CT, Damiano RJ Jr.Robotically assisted microsurgery for endoscopic coronaryartery bypass grafting. Ann Thorac Surg 1998;66:1064–7.

19. Falk V, Walther T, Autschbach R, et al. Robot-assistedminimally invasive solo mitral valve operation. J ThoracCardiovasc Surg 1998;115:470–1.

20. Shennib H, Bastawisy A, McLoughlin J, Moll F. Roboticcomputer-assisted telemanipulation enhances coronary ar-tery bypass. J Thorac Cardiovasc Surg 1999;117:310–3.

INVITED COMMENTARY

In the year 1927 Sir Russel Brock attempted thrice to usea cardioscope designed for him by Schrantz of the Gen-itourinary manufacturing company for intraoperative vi-sualization and management of pulmonic valve stenosisin children. All three children died, clearly not related tothe use of the cardioscope but to the general conditionsin which cardiac surgery was performed during that era.Since then, cardiac surgery evolved to become a safe andeffective practice. This primarily is attributed to thedevelopment of cardiopulmonary bypass and good myo-cardial protection.

Then came minimally invasive cardiac surgery. Ingeneral, three factors stimulate the adoption of newtechnology: market needs, threat of distinction whether itbe war or take over by other specialty and last but notleast; just simply being there (Wright time and Wrightplace) when technology is developed for other purposes.Minimally invasive cardiac surgery (MICS) is so far theproduct of all three. Its technology, except for coronaryartery stabilizers has yet to be proven of added value tomost patients, hospitals or medical industry.

However, the failure to demonstrate benefit from theuse of MICS technology may reflect true lack of benefit orsimply the lack of sensitive methods to demonstrate anadvantage.

In this article, Drs Burke and coworkers at MiamiChildren’s hospital systematically applied cardioscopy inpatients requiring intraventricular repairs. In a seriesexceeding four hundred children they demonstrated the

safety of exposing defects using indirect visualization ormagnification with a cardioscope. Even though it was notpossible to compare this experience with that of a con-ventional direct approach it is intuitively likely that bettervisualization may in fact lead to better quality repair.Only time and careful follow-up will tell. Has this ap-proach led to more cosmetic less lengthy incisions? Thisdoes not appear in this article, yet it is likely that betterindirect exposure using cardioscopes will reduce theextent of surgical incisions. Totally endoscopic repairhowever will only happen when other adjunct enablingtechnology such as closure devices, one shot endoscopicautomated patching and others become available. Cath-eter based technology will most likely challenge surgeonsto further evolve their minimally invasive techniques. DrBurke and his group are to be commended for pushing thefrontier and leading the application of innovative minimallyinvasive cardiac surgery technology in children.

Hani Shennib, MD

Center for InnovativeCardiovascular TherapyHeart InstituteBeth Israel Medical CenterFirst Ave at 16th StNew York, NY 10003e-mail:hshennib@bethisraelny.org.

737Ann Thorac Surg MIYAJI ET AL2000;70:730–7 VAC IN CONGENITAL HEART SURGERY

© 2000 by The Society of Thoracic Surgeons 0003-4975/00/$20.00Published by Elsevier Science Inc PII S0003-4975(00)01626-X