The Impact of Patch Augmentation on Left Atrioventricular Valve Dynamics in Patients with Atrioventricular Septal Defects: Early and Midterm Follow-up

The Impact of Patch Augmentation on LeftAtrioventricular Valve Dynamics in Patientswith Atrioventricular Septal Defects: Early

and Midterm Follow-upKevin S. Roman, MD, FCP (SA), MRCP (UK), Masaki Nii, MD, PhD,

Christopher K. Macgowan, PhD, Catherine Barrea, MD, John Coles, MD,
and Jeffrey F. Smallhorn, MBBS, FRCPC, Toronto, Ontario, Canada
Objective: Left atrioventricular valve pericardialpatch may prevent valve replacement. We assessedpatch annular dynamics compared with conven-tional repair and normal annuli.Methods: Transesophageal 3-dimensional echocar-diography was acquired preoperatively and postop-eratively in atrioventricular septal defects (n � 10, 5patch, 5 conventional repair). Real-time 3-dimen-sional annular motion at midterm was comparedwith that of healthy children (n � 10). Parameterswere: annular area, perimeter, segmental diameter,bending angle, stenosis, and regurgitation.Results: Regurgitant jet area ratio decreased in both
patient groups. Conventional repair reduced annu-
doi:10.1016/j.echo.2006.05.019

1382

lar area (P � .02). Patch repair showed an annulararea larger than normal (P � .01). Control subjectshad increased systolic area whereas operativegroups showed a reduction. Patch repair had seg-mental diameters similar to normal whereas con-ventional repair was inhomogeneous. Annularbending angle was maintained after operation.Conclusion: Patch repair in pediatrics shows dura-bility without shrinkage or expansion. Improvedstenosis and regurgitation does not change by mid-term. Operation causes increased annular stiffnessand diminished compliance. Neither technique es-tablishes normal annular eccentricity. (J Am Soc
Echocardiogr 2006;19:1382-1392.)
Left atrioventricular valve (LAVV) regurgitation isthe major indication for surgical reintervention inatrioventricular septal defect (AVSD) with reportedrates of 3% to 18%1-5 for partial and 6% to 14%2,6-9 forcomplete forms. Although valve replacement oper-ation in young children has an impressive 5-yearfreedom from reoperation rate of 81% to 85.7%,10,11

valve repair remains the most attractive pediatricsolution. If adequate valve leaflet tissue is present topermit good coaptation of the free edges, theninvariably a successful primary repair or reoperationcan usually be achieved by conventional techniques,

From the Division of Cardiology, Department of Diagnostic Im-aging (C.K.M.), and Cardiothoracic Surgery (J.C.), The Hospitalfor Sick Children, The University of Toronto, Faculty of Medicine.Supported by a grant from the Ontario Consortium of Imaging(Drs Nii and Roman).Philips Canada is a partner in the grant from the Ontario Consor-tium of Imaging.Reprint requests: Jeffrey F. Smallhorn, MBBS, FRCPC, Divi-sion of Cardiology, Room 4C2, Walter C. Mackenzie HealthSciences Centre, Edmonton, Alberta, Canada T6G 2B7 (E-mail:[email protected]).0894-7317/$32.00Copyright 2006 by the American Society of Echocardiography.

involving cleft repair plus or minus annuloplasty.However, in some cases where leaflet dysplasia is amajor contributor to valve failure, traditional surgi-cal approaches may lead to increased leaflet tensionwith resulting persistent regurgitation, stenosis, orboth.12,13 In these cases the leaflet tissue is oftendeficient, with thickened rolled edges that do notpermit good coaptation of the free edges.

To avoid valve replacement in such cases, ourinstitution developed a novel surgical repair of theatrioventricular valve for patients with an AVSD andassociated leaflet dysplasia.14 This technique usesglutaraldehyde-treated autologous pericardium topatch augment the leaflet surface. In these cases theaffected part of the valve is removed from thesupporting annulus and the pericardial patch sewninto position. This results in less tension and bettercoaptation of the free edges of the leaflets. Thisappears to have a reasonable impact on early LAVVfunction14,15; however, questions remain regardingthe durability of pericardial leaflet augmentation inpediatric patients. Of importance, this process maypotentially lead to splinting of the LAVV annulus withresultant abolition of “normal” annular dynamics.

The aims of this study were firstly to assess themidterm effects of pericardial patch augmentation
on LAVV function in AVSD. Secondly, we addressed

Journal of the American Society of EchocardiographyVolume 19 Number 11 Roman et al 1383

the impact of this technique on LAVV annulardynamics by comparing this population with a sim-ilar group undergoing conventional repair, and withhealthy pediatric control subjects.

METHODS

Study Population

The comparison control group for normal mitral annularmotion consisted of a cohort of children from our institu-

Table 1 Clinical and demographic data of patients Patch a

Patient 1 Patient 2

Diagnosis Primum ASD, doubleorifice LAVV (noprior surgery)

Primum ASD (noprior surgery)

Age at priorsurgery, mo

Time toreoperation, y

Indication forsurgery

Primum ASDLVOTO, moderateLAVVregurgitation

Primum ASDModerateLAVVregurgitation

Weight atsurgery, kg

28 19.9

Age at surgery, y 9.2 4.7Surgical

procedurePatch repair of SBL

and IBL, cleftclosure, resectchordae in LVOT

Patch repair ofSBL and IBL,closure of smallresidual cleft

Follow-up periodto RT3D, mo

24 41

Conventional repairDiagnosis Primum ASD (no

prior surgery)Primum ASD (no

prior surgery)

Age at priorsurgery, mo

Time toreoperation, y

Indication forsurgery

Primum ASDModerate LAVVregurgitation

Primum ASDmild LAVVregurgitation

Weight atsurgery, kg

30.6 15.5

Age at surgery, y 7.7 2Surgical

procedureClose primum defect

and cleft, LAVVannuloplasty

Close primumdefect andcleft, LAVVannuloplasty

Follow-up periodto RT3D, mo

10 12

ASD, Atrial septal defect; AVSD, atrioventricular septal defect; IBL, inferiotract; LVOTO, left ventricular outflow tract obstruction; ML, mural leaflet; Rdefect.

tion who were the subject of a recent publication.16 In

that study, normal cardiac anatomy and function wasassessed in all participants by 2-dimensional (2D) echocar-diography. The patient group consisted of 10 patientswith an AVSD and a normal ventriculoarterial connectionwho were prospectively studied between June 2000 andSeptember 2003. Of note, these cases were not part of theoriginal surgical study reported in 2000.14 These casescomprised 5 patients undergoing operation using pericar-dial patch augmentation (mean age 7.4 years) and 5 havinga conventional repair or rerepair of an AVSD (mean age4.2 years) (Table 1). To be included in the study all

tation operation

Patient 3 Patient 4 Patient 5

um ASD (previousair)

Complete AVSD(previousrepair)

Primum ASD, doubleorifice LAVV(previous repair)

15 5.5 5

6 7.6 6

erate LAVVurgitation

LVOTO, mild-moderateLAVVregurgitation

Moderate LAVVregurgitation

28 22 18

7.25 8.1 6.8repair of the

VV between IBLd ML

Patch repair ofSBL and IBL,resectsubaorticridge andchordae

Patch repair of SBL/IBL, resectaccessory chords

39 31 3

um ASD (previousair)

Complete AVSD(previousrepair)

Complete AVSD(previous repair)

8 11 5

2.1 2 5.5

erate LAVVurgitation

Moderate LAVVregurgitation

Severe LAVVregurgitation

13.7 12.6 17

2.8 2.9 5.9e residual LAVVft and small VSD

Close residualLAVV cleft

Close residual LAVVcleft andannuloplasty

14 20 7

g leaflet; LAVV, left atrioventricular valve; LVOT, left ventricular outflowl-time 3-dimensional; SBL, superior bridging leaflet; VSD, ventricular septal

ugmen

Primrep

Modreg

PatchLAan

Primrep

Modreg

Closcle

r bridginT3D, rea

patients had to have two separate atrioventricular valves,

Journal of the American Society of Echocardiography1384 Roman et al November 2006

and an intact ventricular septum. During the study periodthere were no other patients who underwent patchaugmentation. Of note, in all but two (one from the patchgroup with left ventricular [LV] outflow tract obstructionand one from conventional repair cohort) there wasassociated significant LAVV regurgitation. Patch augmen-tation was chosen in 4 cases, as there was insufficientnative LAVV tissue to permit repair of the regurgitantvalve without creating stenosis. In the fifth case it wasperformed to aid in the relief of the LV outflow tractobstruction. Clinical and demographic details are given inTable 1.

All surgical cases were studied the same way. Thesepatients underwent routine 2D transthoracic echocardiog-raphy preoperatively, postoperatively, and at follow-up.Transthoracic real-time 3-dimensional (RT3D) echocardi-ography was used to assess LAVV annular motion duringmidterm follow-up of all surgical patients and comparedwith healthy, age-matched volunteers (n � 10). Midtermdata were collected at a median of 31 months (range: 3-41)after patch augmentation and 12 months (range: 7-20)after conventional repair.

Three-dimensional (3D) transesophageal echocardio-graphic (TEE) data were acquired preoperatively and earlypostoperatively in all surgical patients. The use of 3D TEEin the preoperative and postoperative assessment of pa-tients with AVSD was approved by our ethics committee.Of note, in all cases RT3D echocardiography was notavailable at the time of the initial surgical repair.

3D TEE

TEE 3D data sets were acquired (Sonos 5500, PhilipsMedical Systems, Bothwell Wash) with an omniplane 6.2-to 5-MHz probe that is interfaced with online software. Arotational device built into the probe collects data at3-degree increments around a 180-degree arc, yielding 61sequential images. Respiratory and electrocardiographicgating was used to obtain optimal temporal and spatialregistration. A complete cardiac cycle was recorded at theend of expiration. Raw data were stored on a magneto-optical disk for subsequent offline reconstruction. Preop-erative and postoperative TEE data sets were collected inthe operating department under general anesthesia beforeand after chest closure. Anesthesia allowed for a con-trolled heart and respiratory rate during data acquisition.

RT3D Echocardiography

Imaging was performed with a 2- to 4-MHz 3D matrix-arraytransthoracic probe (Sonos 7500, Philips Medical Sys-tems). Data sets were acquired during midterm follow-upof all surgical patients and healthy volunteers. Transtho-racic volumetric images of the LAVV were obtained fromthe standard apical 4-chamber view. A volumetric framerate of 16 to 24 frames/s at an imaging depth of 6 to 16cm was obtained. Between 4 and 10 scans were per-formed in each participant, and the best data setsselected. All volumetric images were digitally stored for
offline analysis.
Annular Analysis from 3D TEE and RT3D Data Sets

A commercially available computer workstation was used forreconstruction of the LAVV annulus (4D LV-Analysis,TomTec, Munich, Germany). Data sets were sliced along 8planes around a central axis that joins the center of theannulus and the cardiac apex (ie, every 22.5 degrees). Theannular margins were placed at the hinge points ofthe LAVV and were manually traced for each frame throughthe cardiac cycle. Reconstructed 3D volumetric images wereconverted into spatial coordinates. Sixty points (x, y, zcoordinates) covering the annulus were then exported tocustom-designed software (developed in Mathlab, Math-works Inc, Natick, Mass) and reconstructed as a 3D graphfor 3D measurements (Figure 1, A). All data were ac-quired, processed, and analyzed by the same two observ-ers.

LAVV Annular Measurements

Annular area and perimeter. The annular area wascalculated as the sum of 60 triangles composed of twoadjacent annular points with the mathematically calcu-lated gravity center of the annulus forming the apex of thetriangles. The perimeter was calculated as the sum of thepoint-to-point distance traversing 60 points on the annulus(Figure 1, A). The annular area and perimeter wereindexed to body surface area (mm2/m2 for area, mm/m2

for perimeter).Segmental diameter contraction. A segmental diameter

was calculated as the distance between two oppositepoints on the annulus (Figure 1, A). Thirty segmentaldiameters around the annulus were measured and num-bered starting with number 1 closest to the aorta follow-ing a counterclockwise direction around the LAVV (Figure1, B). Segmental contraction of the annulus was used toassess annular eccentricity.

Annular bending angle. The annulus was divided intotwo planes along the anterior-posterior commissure-to-commissure line and the bending angle was calculated asthe angle between the least square planes fitted to the twoparts of the divided annulus (Figure 1, C ). This bendingangle is a measure of annular plasticity and saddle shapeduring systole and diastole.

LAVV Regurgitation and Stenosis

Preoperative and immediate postoperative 3D TEE datasets were used to assess the degree and location of LAVVregurgitation. This was achieved through the use of theoffline analysis system (TomTec). Each reconstructed 3Dimage included anatomic landmarks and color Dopplerregurgitant jets. Midsystolic color jets were optimized andmanually transected in a plane that was at right angles tothe vena contracta. The area of the enface view of thevena contracta was then measured preoperatively andimmediately postvalve repair. A summation of the areaswas undertaken if more than one regurgitant jet was
present. A regurgitant jet area ratio was calculated using


the midsystolic vena contracta and valve area from the 3Dimage. The location of the regurgitant jet was also notedfor each time period.17

Follow-up real-time color flow regurgitant jets are notcurrently comparable with those obtained by 3D TEE and,hence, were not used to quantify the degree of regurgita-tion. Instead, 2D color Doppler data sets were used forcomparison.

LAVV inflow stenosis was defined as a continuous wavemean Doppler gradient of greater than 5 mm Hg. Preop-erative, postoperative, and follow-up data were compared

Figure 1 A, Reconstructed 3-dimensional imageB, Corresponding anatomic points on valve anseptolateral regions of annulus. C, Left atrioventrangle (curved white line) between anterior (A) an(black) line. AL, Anterolateral; AO, aorta; MV, mstar, mathematic center of gravity.

within each group.

Timing of Cardiac Events

Because there is no electrocardiography on 3D data sets,the timing of cardiac events were defined visually from thedata set (TomTec). Systole was identified from aortic valveopening, with isovolumic contraction being the periodbetween mitral valve closure and aortic valve opening.The cessation of systole was identified by aortic valveclosure, with the onset of diastolic ventricular fillingoccurring at mitral valve opening. Isovolumic relaxationwas identified between aortic valve closure and mitral

ng 60 points on annulus at 6-degree increments.For example, points 10 to 20 correspond to

valve saddle shape as assessed by annular bendingterior (P) leaflets along commissure-commissurealve; PA, pulmonary artery; PM, posteromedial;

showinulus.iculard positral v

valve opening.


For each control subject and patient annular area atfollow-up was calculated as a function of time throughoutthe cardiac cycle. Because the duration and number offrames for systole and diastole varied between partici-pants, it was necessary to align corresponding segments ofeach curve before averaging. Alignment was based on theopening and closing of the mitral and aortic valves. Thealigned curve was then cubic interpolated, resampled to acommon temporal resolution, and averaged across allparticipants.

The 3D TEE data acquired preoperatively and immedi-ately postoperatively were used to compare the annularfindings of patch augmentation to conventional repair.This comparison only included maximum, minimum, andpercentage change for annular area, and perimeter, asanesthesia can have an impact on annular dynamics.

However, for RT3D data in addition to the abovechanges we also analyzed the percentage change insegmental diameters around the annulus, and the maxi-mum, minimum, and change in bending angle. At midtermfollow-up, RT3D data were also used to track LAVVannular area throughout the cardiac cycle for all 3 groups.In addition, the effect of the two surgical techniques onLAVV stenosis and regurgitation were assessed as de-scribed above.

Statistical Analysis

Data are expressed as mean � SD unless otherwise stated.A nonparametric repeated measures analysis of variance(ANOVA) was used to assess data within a group atpreoperative, postoperative, and follow-up time points.Data between groups were compared using an ordinarynonparametric 1-way ANOVA with posttesting of signifi-cant P values. When two variables were compared be-tween groups a 2-way ANOVA without repeated measureswas used (segmental diameter). A value of P less than .05was considered significant. All analyses were performedusing software (GraphPad Prism, Version 4; GraphPadSoftware, San Diego, Calif) using default options.

RESULTS

Prospective 3D data collection (preoperative, post-operative, and follow-up) was complete and success-ful in all cases. Our technique using both 3D TEEand RT3D echocardiography demonstrated goodreproducibility with an intraobserver variability of1.7% and an interobserver variability of 4% (area andperimeter). All patients had normal global and re-gional LV function at the time of the initial andfollow-up study with a mean fractional shortening of36.3% (range: 29%-44%). The mean age at study wassimilar (4.2 years for conventional repair vs 7.4 yearsfor patch group). The mean follow-up period was 27months for the patch group and 12 months for those
with a conventional repair.
LAVV Regurgitation and Stenosis

A reduction in LAVV regurgitant jet area ratio wasobserved in both patient groups (Figure 2). Of note,at follow-up the degree or location of the earlypostoperative LAVV regurgitation did not change asseen from 2D color Doppler images. Mild LAVVstenosis was present in two cases in each surgicalgroup at follow-up, whereas one patient with apreoperative mean Doppler gradient of 12 mm Hgshowed a reduction to 4 mm Hg at follow-up afterpatch augmentation (Figure 2, A).

LAVV Maximum and Minimum Annular Areaand Perimeter

As shown in Figure 3 and Table 2, after conventionalrepair, both maximum and minimum annular areawere significantly reduced (P � .02 and P � .03,respectively) when compared with preoperativevalues. Of note, annular perimeter measurementswere not significantly affected by those undergoingconventional repair.

Analysis at the time points of preoperative, post-operative, and follow-up revealed no significantdifferences for maximum, minimum, and percentchange for annular area and perimeter within thepatch repair group.

When annular data at follow-up were comparedwith control subjects, both surgical groups hadlarger measurements for maximum and minimumannular area indices. Conventional repair at fol-low-up showed a minimum annular area of 742.4 �130.4 mm2/m2, which was significantly differentfrom normal values (P � .01). Patch repair atfollow-up showed a minimum annular area of 925.7� 421.6 mm2/m2 and a maximum of 1081 � 477.3mm2/m2, which were both different to age-matchednormal values (P � .002 and P � .01, respectively).Percentage changes in annular area and perimeterwere not different from control values. At follow-up,comparisons between the two surgical groupsshowed no significant values.

Changes in LAVV Annular Area TrackedThrough the Cardiac Cycle

As shown in Figure 4, our control group showedan increase in mitral annular area during systole,reaching a maximum before mitral valve openingfollowed by an abrupt decrease during early dia-stolic filling. Both surgical groups showed theopposite in that LAVV annular area reduced dur-ing systole. Furthermore, the abrupt decrease inannular area as seen in control subjects duringearly diastolic filling was attenuated after both
surgical techniques.

ch ope


Segmental Diameter Change and AnnularBending Angle

Control subjects had the largest changes in segmen-tal diameter during the cardiac cycle occurringbetween segments 8 and 18, ie, septolateral segments(Figures 1 and 5). Results show that both surgicalgroups had less dynamic changes in segmental diame-ter when compared with the curve for control sub-jects. Patients with patch repair retained a curvesimilar to normal whereas conventional repair showeda markedly different curve. Two-way ANOVA compar-ing data between the surgical groups showed thatdifferences were a result of the type of operation(P � .0001). Maximum, minimum, and percent changein bending angle were not different between surgi-cal groups or control subjects (Table 2).

DISCUSSION

Adult mitral valve pericardial leaflet augmentationoperation was previously described by the Chau-vaud group and they have shown normal compli-ance of pericardial tissue over a 3-year period.18 Theuse of flexible material like pericardial tissue is wellestablished as being superior to rigid materials inpreserving annular dynamics.19,20 This technique isideally suited to repair dysplastic valves but ques-

Figure 2 Left atrioventricular valve (LAVV) raugmentation (A) and conventional (B) operatio(post-op) in all cases. Two patients have inflow mpatient shows good relief of obstruction after pat

tions remain regarding its effect on leaflet calcifica-

tion, fibrosis, shrinkage in children, and LAVV annu-lar dynamics. LAVV annular dynamics afterpericardial leaflet patch augmentation has not beenstudied before.

Mitral valve coaptation relies on annular plasticityduring the cardiac cycle. Area reduction is achievedby conformational changes to the 3D shape of theannulus.21 Two-dimensional perimeter shortening issupported by modulation of the annular saddleshape, allowing anterior and posterior segments tomove closer together. It is known that leaflet stressis determined by leaflet curvature and that thesestresses are reduced by leaflet billowing and thesaddle shape of the annulus.22 In the current studyannular dynamics were measured by changes inannular area and perimeter, eccentricity was as-sessed by segmental annular diameter changes, andthe saddle shape by the anterior-posterior bendingangle of the annulus.

The current study shows that both conventionaland patch repair produce satisfactory results regard-ing LAVV regurgitation and stenosis with most pa-tients showing an improvement; and at midtermfollow-up, no worsening of these parameters wereobserved. Both surgical groups had annular dimen-sions above normal values owing to chronic LAVVregurgitation. As expected, conventional repair (in-cluding a DeVega annuloplasty in most cases) led toa significant decrease in indexed maximum and

itation (top) and stenosis (bottom) after patchurgitant jet area ratio is improved after operationoppler gradient greater than 5 mm Hg. A, Oneration (*). Pre-op, Preoperative.

egurgn. Regean D

minimum annular area compared with preoperative


values. Patch repair does not diminish annular areaand, notably, no shrinkage in annular area wasobserved. At follow-up (both surgical groups), per-centage change in annular area and perimeter werenot different from control subjects proving that

Figure 3 Maximum and minimum indexed annsubjects. No shrinkage or dilatation is seen atpostoperative data of surgical groups. *P less thaConventional operation showed significant decregroups and normal control values. *P less than .0values.

Table 2 Comparison of mean annular parameters at follow

Patch

(n

Maximal area index, mm2/m2 1081Minimal area index, mm2/m2 925.7Area change ratio, (max-min)/max (%) 14.4Maximal perimeter index, mm/m2 118Minimal perimeter index, mm/m2 109.4Perimeter change ratio, (max-min)/max (%) 7.7Maximum bending angle, degree 167.9Minimum bending angle, degree 150Bending angle change, (max-min) degree 17.9

*P � .01.†P � .002.

these functional parameters are preserved.

An important difference was noted when annulararea was examined through the entire cardiac cycle(Figure 4). Control subjects showed an increase inmitral annular area during systole followed by anabrupt decrease during early diastolic filling. These

ea for surgical groups and age-matched controlerm follow-up. Top, Preoperative (pre-op) andwhen comparing pre-op and postoperative area.annular area. Bottom, Follow-up data of surgicaln comparing follow-up measurements to control

Conventional repair

(n � 5)

Control

(n � 10)

3* 886.8 � 177.4 665.4 � 70.56† 742.4 � 130.4* 527.5 � 55.5

15.9 � 4.6 19.4 � 6.1122.7 � 22.8 98.4 � 13.9111.1 � 20.7 89.2 � 13.3

9.4 � 1.6 9.3 � 3.1159.3 � 20.9 163.5 � 6.8138.1 � 21.5 146.3 � 8.4

21.2 � 2.4 17.2 � 5.9

ular armidt

n .05ase in5 whe

-up

repair

� 5)

� 477.� 421.� 4.7� 36.5� 34.8� 2.5� 7.1� 13.2� 7.9

findings were recently published from our institu-

ening.


tion16 and differ from published adult data23 wherea reduction in annular size has been demonstratedduring systole. Interestingly, after both patch andconventional repair, annular area through the car-diac cycle closely resembles an adultlike pattern inthat a reduction in annular size is seen duringsystole. We speculate that this adultlike pattern iscaused by diminished annular compliance and in-creased annular stiffness probably as a result ofextensive suture lines.

Contraction of septolateral segments is a promi-nent feature of the normal mitral annulus and, bycomparison, both surgical groups revealed lesschange within these segmental diameters. This ismost likely related to the fundamental difference ina normal atrioventricular junction, versus all hearts

Figure 4 Indexed left atrioventricular valve annuincrease in annular area during systole, whereasControl group mitral valve annular area indexconventional (right) repair. Mean and SEM vmaximum and minimum values. AVC, Aortic vafilling period; EP, ejection period; IC, isovolumiMVC, mitral valve closure; MVO, mitral valve op

with an AVSD. In the latter the atrioventricular

junction is sprung resulting in a different annularseptal configuration (Figure 5). This segment com-prises the atrioventricular valve bridging leaflets,patch material closing an atrial or ventricular septaldefect (VSD) and the interventricular septum, inthose without a ventricular component to theirdefect. As well, extensive suture annuloplasty mayresult in less physiologic changes in annular eccen-tricity as shown by a relatively inhomogeneoussegmental diameter contraction curve. Of interest,although both groups were different than controlsubjects, those with a conventional repair had amore homogeneous pattern compared with thepatch augmentation group. This may in part berelated to the observation that two patients had aVSD patch, whereas only one did in the patch

a during cardiac cycle. Control subjects showedl groups showed reduction during systole. Top,om, Annular area index after patch (left) andGraphs have different scales to accommodate

osure; AVO, aortic valve opening; DF, diastolicaction period; IR, isovolumic relaxation period;

lar aresurgica. Bottalues.lve cl

c contr

augmentation group. The presence of a VSD patch

0) whe


may remove any effect of the interventricular sep-tum on annular function.

Analysis of the saddle shape by the anterior-posterior bending angle showed that this parameterwas preserved after operation. There appears to bea close relationship between bending angle andventricular function that is primarily caused bygeometric changes as a result of torsional deforma-tion of the ventricle from base to apex. This angle isalso probably important to reduce leaflet stress

Figure 5 Annular segmental diameter change raBoth surgical groups have less dynamic changescontrol subjects. Patch operation retains more phloss of septolateral dynamics (arrows). Left, X axichange ratio (%) during cardiac cycle in eachFundamental difference in normal atrioventriculaseptal defect (bottom). In latter, atrioventricularconfiguration. AL, Anterior leaflet; AO, aorta; IBsuperior bridging leaflet; arrows, segments (10-2

during the cardiac cycle.

Limitations

Small sample size is the most important limitation ofthis investigation making statistical analysis less reli-able. TEE 3D data (annular area and perimeter) wereused to assess the annulus preoperatively and imme-diately postoperatively and probably cannot be com-pared with RT3D data at follow-up. As well, the TEE3D data sets were collected during general anesthe-sia that, in itself, may affect cardiac function. Like-

er cardiac cycle for control and surgical groups.mental diameter when compared with curve forgic changes whereas conventional repair leads toegmental diameters on annulus. Y axis: diameternt. Mean and SEM values are shown. Right,tion (top), versus all hearts with atrioventricularn is sprung resulting in different annular septalerior bridging leaflet; PL, posterior leaflet; SBL,re diameter changes are most dynamic.

tio ovin segysiolos: 30 ssegmer juncjunctioL, inf

wise the immediate postoperative study was after


cardiopulmonary bypass that, in itself, has an inher-ent impact on ventricular function. Although thesemay have caused inaccuracies, it should not affectthe follow-up comparisons between control sub-jects and patients as only RT3D data were used. Thetwo groups were not perfectly matched with regardto their morphology, however, all had two separateatrioventricular valve orifices, no VSD, significantLAVV regurgitation in all but two, and normal LVfunction. As well, our institution has a standardrepair for AVSD with or without a ventricular com-ponent, which includes closing the atrial and/orventricular communication with a patch and sutureclosure of the cleft. Although the follow-up periodwas different, being 12 months for the conventionalgroup and 27 months for the patch group, it is likelythat any immediate effect of the operation wouldhave been resolved by that time.

Although 3D ultrasound techniques do not allowa precise fixed point on the annulus to be tracked,the aorta does provide a fixed reference point,which can be used in the analysis. This permits theannulus to be divided into segments that can becompared throughout the cardiac cycle. Althoughsome of the patients had a moderate mean gradientthrough the LAVV at the preoperative assessment,this decreased after surgical intervention that indi-cated that the major lesion was regurgitation and notstenosis.

Conclusion

Pericardial leaflet augmentation in pediatric patientsshowed good durability with no evidence of shrink-age, calcification, or expansion in this group of pa-tients studied. LAVV stenosis and regurgitation appearto remain stable at midterm. Both patch augmentationand conventional repair lead to increased annularstiffness and diminished annular compliance. Annularbending angle is maintained by both surgical tech-niques but neither technique establishes normal annu-lar eccentricity, although patch repair allows for morephysiologic shape changes.

REFERENCES

1. Najm HK, Williams WG, Chuaratanaphong S, Watzka SB,Coles JG, Freedom RM. Primum atrial septal defect in chil-dren: early results, risk factors, and freedom from reoperation.Ann Thorac Surg 1998;66:829-35.

2. Permut LC, Mehta V. Late results and reoperation after repairof complete and partial atrioventricular canal defect. SeminThorac Cardiovasc Surg 1997;9:44-54.

3. Sadeghi AM, Laks H, Pearl JM. Primum atrial septal defect.Semin Thorac Cardiovasc Surg 1997;9:2-7.

4. Al-Hay AA, Lincoln CR, Shore DF, Shinebourne EA. The leftatrioventricular valve in partial atrioventricular septal defect:management strategy and surgical outcome. Eur J Cardiotho-
rac Surg 2004;26:754-61.
5. Murashita T, Kubota T, Oba J, Aoki T, Matano J, Yasuda K.Left atrioventricular valve regurgitation after repair of incom-plete atrioventricular septal defect. Ann Thorac Surg 2004;77:2157-62.

6. Bando K, Turrentine MW, Sun K, Sharp TG, Ensing GJ,Miller AP, et al. Surgical management of complete atrioven-tricular septal defects: a twenty-year experience. J ThoracCardiovasc Surg 1995;110:1543-54.

7. Hanley FL, Fenton KN, Jonas RA, Mayer JE, Cook NR,Wernovsky G, et al. Surgical repair of complete atrioventricu-lar canal defects in infancy: twenty-year trends. J ThoracCardiovasc Surg 1993;106:387-97.

8. Najm HK, Coles JG, Endo M, Stephens D, Rebeyka IM,Williams WG, et al. Complete atrioventricular septal defects:results of repair, risk factors, and freedom from reoperation.Circulation 1997;96:II311-5.

9. Tweddell JS, Litwin SB, Berger S, Friedberg DZ, Thomas JP,Frommelt PC, et al. Twenty-year experience with repair ofcomplete atrioventricular septal defects. Ann Thorac Surg1996;62:419-24.

10. Caldarone CA, Raghuveer G, Hills CB, Atkins DL, Burns TL,Behrendt DM, et al. Long-term survival after mitral valvereplacement in children aged �5 years: a multi-institutionalstudy. Circulation 2001;104:I143-7.

11. Moran AM, Daebritz S, Keane JF, Mayer JE. Surgical man-agement of mitral regurgitation after repair of endocardialcushion defects: early and midterm results. Circulation 2000;102:III160-5.

12. Alexi-Meskishvili V, Hetzer R, Dahnert I, Weng Y, Lange PE.Results of left atrioventricular valve reconstruction after pre-vious correction of atrioventricular septal defects. Eur J Car-diothorac Surg 1997;12:460-5.

13. Abbruzzese PA, Napoleone A, Bini RM, Annecchino FP,Merlo M, Parenzan L. Late left atrioventricular valve insuffi-ciency after repair of partial atrioventricular septal defects:anatomical and surgical determinants. Ann Thorac Surg 1990;49:111-4.

14. Poirier NC, Williams WG, Van Arsdell GS, Coles JG, Small-horn JF, Omran A, et al. A novel repair for patients withatrioventricular septal defect requiring reoperation for leftatrioventricular valve regurgitation. Eur J Cardiothorac Surg2000;18:54-61.

15. Barrea C, Levasseur S, Roman KS, Nii M, Coles JG, WilliamsWG, et al. Three-dimensional echocardiography improves theunderstanding of left atrioventricular valve morphology andfunction in atrioventricular septal defects undergoing patchaugmentation. J Thorac Cardiovasc Surg 2005;129:746-53.

16. Nii M, Roman KS, Macgowan CK, Smallhorn JF. Insight intonormal mitral and tricuspid annular dynamics in pediatrics: areal-time three-dimensional echocardiographic study. J AmSoc Echocardiogr 2005;18:805-14.

17. Shandas R, DeGroff CG, Kwon J, Trujillo N, Gill EA, Valdes-Cruz L. Utility of three-dimensional ultrasound Doppler flowreconstruction of the proximal jet to quantify effective orificearea: in vitro steady and pulsatile flow studies. J Am SocEchocardiogr 1998;11:313-21.

18. Chauvaud S, Fuzellier JF, Houel R, Berrebi A, Mihaileanu S,Carpentier A. Reconstructive surgery in congenital mitralvalve insufficiency (Carpentier’s techniques): long-term re-sults. J Thorac Cardiovasc Surg 1998;115:84-93.

19. Timek TA, Miller DC. Experimental and clinical assessment ofmitral annular area and dynamics: what are we actually mea-
suring? Ann Thorac Surg 2001;72:966-74.


20. Sharony R, Saunders PC, Nayar A, McAleer E, Galloway AC,Delianides J, et al. Semirigid partial annuloplasty band allowsdynamic mitral annular motion and minimizes valvular gradients:an echocardiographic study. Ann Thorac Surg 2004;77:518-22.

21. Flachskampf FA, Chandra S, Gaddipatti A, Levine RA, Wey-man AE, Ameling W, et al. Analysis of shape and motion of themitral annulus in subjects with and without cardiomyopathyby echocardiographic 3-dimensional reconstruction. J Am Soc
Echocardiogr 2000;13:277-87.
22. Salgo IS, Gorman JH III, Gorman RC, Jackson BM, BowenFW, Plappert T, et al. Effect of annular shape on leafletcurvature in reducing mitral leaflet stress. Circulation 2002;106:711-7.

23. Kaplan SR, Bashein G, Sheehan FH, Legget ME, MuntB, Li XN, et al. Three-dimensional echocardiographic as-sessment of annular shape changes in the normal and re-gurgitant mitral valve. Am Heart J 2000;139:
378-87.

Documents

The Impact of Patch Augmentation on Left Atrioventricular Valve Dynamics in Patients with Atrioventricular Septal Defects: Early and Midterm Follow-up