European Heart Journal (1999) 20, 221231 Article No. euhj.1998.1273, available online at http://www.idealibrary.

com on

Echocardiographic and morphological correlations in tetralogy of Fallot

M. A. Gatzoulis, N. Soukias, S. Y. Ho, M. Josen and R. H. Anderson
Royal Brompton Hospital and Imperial College School of Medicine at National Heart and Lung Institute, London, U.K.

Aims Our aim was to clarify the location and structure of the outlet septum relative to the free-standing subpulmonary infundibulum in the setting of tetralogy of Fallot and to examine its relationship to the other components of the subpulmonary outow tract, determining their potential inuence on clinical outcome. Methods and Results We studied prospectively 41 patients with tetralogy of Fallot (mean age 14 109 months) prior to surgical repair, and compared them with 15 patients undergoing closure of a ventricular septal defect associated with malalignment of the outlet septum but no subpulmonary infundibular stenosis (Eisenmenger ventricular septal defect), and 20 healthy controls. We also examined available autopsied hearts from cases with uncorrected tetralogy of Fallot (8) and Eisenmenger ventricular septal defect (13). Data were indexed for body surface area, and diameter of the tricuspid valve, respectively. The overall length of the subpulmonary infundibulum, including the extent of the muscular outlet septum, was signicantly greater for patients with tetralogy of Fallot compared to normals (234 06 vs 146 034 cm/BSA05, P <0001), whereas the dierence between those with tetralogy of Fallot and an Eisenmenger ventricular septal defect was

conned to the degree of narrowing of the subpulmonary outlet (043 022 vs 217 064 cm/BSA05, P <0001). Within the tetralogy of Fallot group, there were linear relationships between deviation of the outlet septum (r= 061, P <0005) and the diameter of the pulmonary valvar orice (r=075, P <0001), suggesting that growth of the pulmonary arteries may be related to this feature. When patients requiring a transannular patch as part of their surgical repair were compared with those not needing this procedure, dierences were found in the diameter of the pulmonary valvar orice and the pulmonary trunk, but not in the dimensions of the outlet septum. Conclusion The position of the outlet septum in relationship to the remainder of the muscular subpulmonary infundibulum represents a hallmark of tetralogy of Fallot, permitting its dierentiation from Eisenmenger ventricular septal defects and normal hearts. (Eur Heart J 1999; 20: 221231) Key Words: Tetralogy of Fallot, echocardiography, pathology, ventricular, outow obstruction, pulmonary trunk.

Introduction
Although the anatomical features of tetralogy of Fallot are well recognized, disagreement remains as to the basis morphology of this congenital lesion. At the centre of the continuing argument has been the dimensions of the subpulmonary infundibulum, and the potential contribution of infundibular morphology to morphogenesis. Van Praagh and colleagues[1] have long claimed that the principal anatomical feature of tetralogy of Fallot is overall hypoplasia of the subpulmonary infundibulum,
Revision submitted 23 July 1998, and accepted 29 July 1998. Correspondence: Professor Robert H. Anderson, Imperial College School of Medicine at National Heart & Lung Institute, Dovehouse Street, London SW3 6LY, U.K. 0195-668X/99/030221+11 $18.00/0

including its length. Previous morphological studies by Becker and colleagues[2], subsequently conrmed by Howell et al.[3], whilst conrming that the subpulmonary infundibulum was narrower in tetralogy of Fallot, showed its length to be increased in comparison with the normal heart. Geva et al.[4], in a recent echocardiographic study, produced ndings which they claimed supported the earlier concept[1], namely that the subpulmonary infundibulum in Fallots tetralogy was shorter compared with normals. One reason for these continuing disagreements is failure to distinguish between the two distinct components of the subpulmonary infundibulum in tetralogy of Fallot, namely the muscular outlet septum which separates the subpulmonary from the subaortic outow tracts within the ventricular mass, and the free-standing subpulmonary
 1999 The European Society of Cardiology

222

M. A. Gatzoulis et al. 35 MHz transducer. Intra-nasal midazolam was used for sedation in two patients. Following routine diagnostic cross-sectional and Doppler examination, particular attention was paid in obtaining satisfactory subcostal right anterior oblique and parasternal short axis views. Studies were recorded on super VHS videocassette tapes. Selected still frames were employed for subsequent measurements, using the built-in digitizing equipment. The right anterior oblique view was used for all measurements of the musculatures surrounding the right ventricular outow tract. The apical four-chamber view was used for determining the diameter of the tricuspid valvar orice and the right ventricular length (measured from the atrioventricular junction to the right ventricular apex). The views obtained in the parasternal short axis, and from the suprasternal notch, were used for measurements of the diameters of the pulmonary valve and the pulmonary arteries. The subpulmonary outow tract was divided into its two components (outlet septum and free-standing infundibulum) according to the location of the nadir of the hingepoint of the aortic valvar leaets (Fig. 1). Maximum thickness was measured at any level of these two structures, and planimetry was used to assess cross-sectional area. The superoinferior, and left-to-right, deviation of the outlet septum were taken in reference to the angle of its long axis relative to the long-axis of the right ventricular outow tract. Measurements were expressed in degrees of rotation, with 0 degrees being in the long axis of the right ventricular outow tract (Fig. 2). Positive values represented anti-clockwise rotation of the outlet septum as seen in the right anterior oblique view. Because of the inability to recognize any discrete muscular structure specically separating the subpulmonary infundibulum from the cavity of the left ventricle in the normal heart, as opposed to the cavity of the aortic valvar sinuses, it was not possible to make these measurements in the normal hearts. In the abnormal hearts we noted the narrowest point of the subpulmonary outow tract, allowing for assessment of the maximum thickness of the septoparietal trabeculations. Measurements of the oricial diameter of the pulmonary valve were made just below the lowest part of the hingepoints of the valvar leaets from the subpulmonary infundibulum, and at the level of the sinutubular junction. The maximum diameter of the pulmonary trunk (at any level from the sinutubular junction to the bifurcation), and the oricial diameters of the right and left pulmonary arteries, were measured in the standard manner. Systolic frames were used (immediately after opening of the aortic valve for the right anterior oblique and parasternal short axis views). Three measurements were made, and the average value was employed for analysis. All echocardiographic measurements were indexed for body surface area (linear measurements to the square root of body surface area and cross-sectional area measurements to body surface area). They are presented in the indexed form in the results and tables. Intra- and inter-observer variability with regard to measured length, thickness, cross-sectional area and angle of the outlet septum

infundibulum. A distinction between these two components has not yet been made in any anatomical study[13]. The free-standing infundibulum is a tube of ventricular muscle which is related to the aortic sinuses, but separated from them by extracardiac space. Only this free-standing tube can readily be distinguished as a discrete entity in the normal heart, making dicult meaningful comparisons with abnormal hearts where the outlet septum is recognized in its own right. These distinctions are also pertinent to analysis of patients with the so-called Eisenmenger type of ventricular septal defect. It is known that such patients have comparable morphological features of the outlet septum relative to the ventricular septal defect, but that they do not develop obstruction within the right ventricular infundibulum[5]. The aim of our study, therefore, was to ascertain the arrangement of the outlet septum and subpulmonary infundibulum in the setting of Fallots tetralogy, and to compare these features with those seen in patients with the Eisenmenger type of ventricular septal defect, and with the infundibular arrangement found in the normal heart.

Patients
We studied prospectively with cross-sectional echocardiography 41 consecutive patients with tetralogy of Fallot seen at the Royal Brompton Hospital prior to complete surgical repair, and 15 patients with the Eisenmenger type of ventricular septal defect who underwent surgical closure during the same period (January 1995December 1996). Patients with coexisting pulmonary atresia, double outlet ventriculo-arterial connection, and atrioventricular septal defects with common atrioventricular junction, were all excluded. We also studied 20 normal controls, matched for age and sex with the patients having Fallots tetralogy, chosen randomly from patients with innocent heart murmurs attending our echocardiographic clinic. In addition, we examined eight autopsied hearts with uncorrected tetralogy of Fallot, and 13 specimens of the Eisenmenger type of ventricular septal defect. These hearts were available in our cardiopathological collection at the National Heart & Lung Institute, and were compared with 20 dissected normal hearts also lodged in the collection.

Methods
Background clinical and surgical information was obtained from the medical records.

Cross-sectional echocardiography
Transthoracic echocardiographic examination was performed (prior to surgical repair) with a Hewlett-Packard Sonos 1500 Ultrasound System, using a 75, 50 or
Eur Heart J, Vol. 20, issue 3, February 1999

Echocardiographicalmorphological correlations in tetralogy of Fallot was examined by two independent observers blinded from the clinical data. Each observer analysed eight echocardiographic data sets randomly chosen from the group of patients with Fallots tetralogy.

223

Morphology
Autopsied hearts were examined in planes resembling the standard echocardiographic views. The same

measurements, apart from estimation of cross-sectional area, were made as far as possible in comparable manner (Fig. 3). Data for body surface area of the patients from whom the hearts were obtained was incomplete, so morphological measurements were indexed for tricuspid valvar diameter.

Statistics
Data analysis was performed using the SPSS for Windows software (Version 70). Descriptive data for continuous variables are presented as mean SD. Students t-tests were used to compare normally distributed variables. Regression analysis was performed using the method of least squares. The level of signicance was set at 005.

Results
Basic clinical characteristics of the two groups of patients, the healthy controls and the autopsied hearts are presented in Table 1. There were no peri-operative deaths.

Cross-sectional echocardiographic measurements

The percentage dierence in measurements of the length, thickness, cross-sectional area, and degree of deviation of the outlet septum for the random sample of eight

Figure 1 (a) Subcostal right anterior oblique view from a patient with Fallots tetralogy. Arrow indicates the level of the hingepoint of the aortic valvar leaets, marking the division of the subpulmonary infundibulum into its freestanding (upper) part and the outlet septum (lower part). Note the anterocephaled deviation of the hypertrophied outlet septum which, in combination with the hypertrophied septoparietal trabeculations, creates subpulmonary stenosis. (b) Subcostal right anterior oblique view from a patient with Eisenmenger ventricular septal defect. The outlet septum (below the arrow) is elongated and hypertrophied, as in patients with Fallots tetralogy. In contrast to patients with tetralogy of Fallot, the right ventricular outow tract is wide open. There is no anterocephaled deviation of the outlet septum, nor any signicant hypertrophy of the septoparietal trabeculations. Note the dilated pulmonary trunk and arteries. (c) Subcostal right anterior oblique view from a normal control. Note the o-setting of the leaets of the arterial valves (pulmonary valve superiorly), and the angulations of the two outow tracts between each other (in this view, long axis for the right and short axis for the left ventricular outow tract, respectively). AO=aorta; AV=aortic valve; FW=free wall of right ventricle; OS=outlet septum; PT=pulmonary trunk; PV=pulmonary valve; RA=right atrium; RV=right ventricle.
Eur Heart J, Vol. 20, issue 3, February 1999

224

M. A. Gatzoulis et al.

Anticlockwise (positive) 0 Freewall Outlet septum

Freewall Outlet septum 0 Clockwise (negative)

Figure 2 Diagram showing measurements for the angle of deviation between the long axis of the outlet septum and the right ventricular freewall. patients with Fallots tetralogy ranged from 314% for within-observer, and from 215% for between-observer variability. Detailed echocardiographic data from the three groups are presented in Table 2. When patients with Fallots tetralogy were compared to healthy controls, the outlet septum was recognizable as a structure with length and thickness within the hearts of patients with tetralogy. Such a structure could not be distinguished from the remainder of the muscular septum in the normal population (Fig. 3). When assessing only the length of the free-standing subpulmonary infundibulum, this was signicantly longer in the normal patients (146 034 vs 089 035, P <0001). Maximal thickness of the infundibulum, however, was greater in the patients with tetralogy of Fallot. This feature, in combination with the hypertrophied septoparietal trabeculations, produced the muscular subpulmonary stenosis. When the length of the outlet septum was combined with that of the free-standing infundibulum, the value of the overall length of the subpulmonary outlet was signicantly longer in the patients with tetralogy of Fallot as compared to the length of the free-standing subpulmonary infundibulum in normals (234 06 vs 146 034, P <0001). The diameters of the pulmonary valve and pulmonary trunk, as expected, were smaller in the setting of Fallots tetralogy, whereas the diameters of the left and right pulmonary arteries were closer to normal values. The presence of a left modied Blalock Taussig shunt in 12 patients with tetralogy of Fallot may have aected growth, explaining why, on average, there was no dierence in the size of the left pulmonary artery between patients with Fallots tetralogy and healthy controls. There were signicant correlations within the group of patients with Fallots tetralogy between the degree of deviation of the outlet septum (r= 061, P =0002) and the diameter of the pulmonary valve (r=075, P =0001), and the dimensions of the pulmonary trunk, respectively (Table 3). When the 20 patients who required a transannular patch as part of the repair
Eur Heart J, Vol. 20, issue 3, February 1999

were compared to the 21 patients who did not, those needing a transannular patch had a signicantly smaller pulmonary trunk (11 038 vs 145 03, P <002). In addition, there was a tendency toward a smaller diameter of the pulmonary valve in those requiring a transannular patch (078 021 vs 099 044, P =009). No signicant dierences were found between the two subgroups with regard to the overall dimensions of the outlet septum. The main dierence between patients with the Eisenmenger type of ventricular septal defect and those with tetralogy of Fallot was the absence of narrowing of the subpulmonary infundibulum in those with the Eisenmenger ventricular septal defect (Fig. 1). The malaligned outlet septum in the Eisenmenger type of ventricular septal defect was relatively long and thick, as in the patients with tetralogy. There was, however, no anterocephalad deviation in terms of counterclockwise rotation of the outlet septum relative to the long axis of the right ventricular outow tract nor was there as marked hypertrophy of the septoparietal trabeculations. As a result, the substrate for development of muscular subpulmonary stenosis was lacking in the patients with Eisenmenger ventricular septal defect. The pulmonary valve and pulmonary arteries were dilated, compared to patients with Fallots tetralogy and healthy controls, due to the excessive pulmonary arterial ow.

Morphological measurements
The values obtained from the specimens conrmed those of cross-sectional echocardiography (Table 4). The main dierence between the hearts showing Fallots tetralogy and those with an Eisenmenger type of ventricular septal defect was the lack of narrowing of the subpulmonary infundibulum in the latter group (Fig. 4). There were no other discernible dierences in the structure of the right

Echocardiographicalmorphological correlations in tetralogy of Fallot

225

Figure 3 (a) The right ventricle of a normal heart is displayed to show the septal insertion of the ventriculo-infundibular fold between the anterior (A) and posterior (P) limbs of the septomarginal trabeculation. (b) When the ventriculo-infundibular fold in the normal heart specimen is deected, the level of the pulmonary valve superior to that of the aortic valve can be seen. The free-standing subpulmonary infundibulum is continuous postero-inferiorly with the ventriculo-infundibular fold. (c) A dissection into the ventriculo-infundibular fold shows its close relationship with the right coronary aortic sinus and the septomarginal trabeculation. A discrete outlet septum cannot be distinguished.

Eur Heart J, Vol. 20, issue 3, February 1999

226

M. A. Gatzoulis et al.

Table 1

Clinical characteristics of cases studied

N Age at study (months) 14 58 91 122 1046 109 33 128 125 1669

Discussion
Our echocardiographic and morphological data support the view that the morphological substrate for tetralogy of Fallot is anterocephalad deviation of the malaligned outlet septum which, in combination with hypertrophy of septoparietal trabeculations, leads to narrowing of the subpulmonary infundibulum. In our investigation, this was tested not only against healthy controls, but also against patients with the Eisenmenger type of ventricular septal defect. Our ndings conrm earlier studies[5] showing that the infundibular morphology of Eisenmenger ventricular septal defect is indistinguishable from Fallots tetralogy apart from the narrowing of the subpulmonary infundibulum. In contrast to the ndings described in a recent study by Geva et al.[4], and previous suggestions from Van Praagh and his colleagues[1], we were unable to detect any shortening in the overall length of the narrowed subpulmonary outow tract in our patients with Fallots tetralogy compared to normals. Although the extent of the free-standing infundibulum was signicantly shorter in patients with Fallots tetralogy, the total length of the subpulmonary outlet, representing the extent of the infundibulum combined with the length of

Patients with tetralogy of Fallot Patients with Eisenmenger VSD Normals Specimens of tetralogy of Fallot Specimens of Eisenmenger VSD

41 15 20 8 13

Values are mean SD. Patients were studied echocardiographically immediately prior to repair. VSD=ventricular septal defect.

ventricular outow tract between the two groups. The pulmonary valve was signicantly smaller, however, and there was a tendency for a smaller pulmonary trunk, in those with tetralogy of Fallot. There were no statistically signicant dierences between the two groups with respect to the dimensions of the left and right pulmonary arteries. This reects, in part, the small numbers suitable for measurement (4 and 7 respectively). It also shows that, in simple Fallots tetralogy, the branches of the pulmonary trunk are not signicantly hypoplastic. Table 2 Echocardiographic ndings
Fallots tetralogy n=41 Os length (cm/BSA05) Os thickness (cm/BSA05) Os C/S area (cm/BSA) Deviation of Os (degree) F-S infundibular length (cm/BSA05) F-S infundibular thickness (cm/BSA05) Total infundibular length (cm/BSA05) Narrowest diameter (cm/BSA05) SP trabeculations thickness (cm/BSA05) Pulmonary valve/below (cm/BSA05) Pulmonary valve/above (cm/BSA05) Pulmonary trunk (cm/BSA05) Right pulmonary artery (cm/BSA05) Left pulmonary artery (cm/BSA05) 138 078 097 118 089 051 234 043 084 111 09 133 097 098 059 028 06 98 035 017 06 022 097 036 035 067 018 019

Eisenmenger VSD n=15 155 039 =002 102 036 <003 132 057 <007 81 8 <0001 099 044 ns 048 01 ns 245 06 =002 217 064 <0001 06 023 ns 243 039 <0001 263 041 <0001 258 066 <0001 161 034 <0001 159 025 <0001

Normals n=20 146 034 <0001 025 023 <0001 146 034 <0001 201 04 <0001 032 011 =0002 198 049 <0001 169 026 <0001 183 029 <0001 12 019 <0001 105 015 ns

BSA=body surface area; C/S=cross-sectional; F-S=free standing; ns=not signicant; Total=total subpulmonary infundibulum; i.e., free-standing infundibulum+outlet septum (for abnormal hearts); narrowest diameter=in the subpulmonary area; SP=septoparietal; VSD=ventricular septal defect; decimal values= P values compared to the rst column; Os=outlet septum. Eur Heart J, Vol. 20, issue 3, February 1999

Echocardiographicalmorphological correlations in tetralogy of Fallot

227

Table 3 Echocardiographic correlations in patients with Fallots tetralogy

PVb Osa PVb PVa LPA ns PVa ns P <0001 r=061 PT P =0002 r= 061 P =0001 r=060 P =0001 r=075 RPA ns P =0006 r=052 ns P =0003 r=05 LPA ns ns

Osa=os/free right ventricular wall angle; PVb=diameter below the pulmonary valve; PVa=diameter at the sinutubular junction; RPA=right pulmonary artery; LPA=left pulmonary artery; ns=not signicant; Os=outlet septum.

Table 4

Morphological ndings
Fallots tetralogy Eisenmenger VSD n=8 n=13

Os length (cm/TVd) Os thickness (cm/TVd) Deviation of Os (degree) F-S infundibular length (cm/TVd) Infundibular thickness (cm/TVd) Total infundibular length (cm/BSA05) Narrowest diameter (cm/TVd) SP trabeculations thickness (cm/TVd) Pulmonary valve/below (cm/TVd) Pulmonary valve/above (cm/TVd) Pulmonary trunk (cm/TVd) Right pulmonary artery (cm/TVd) Left pulmonary artery (cm/TVd)

054 050 17 021 05 075 034 06 047 051 093 057 072

024 021 31 038 018 038 015 01 02 019 056 031 026

038 014 =007 052 017 ns 55 18 <0001 038 044 ns 073 028 =006 076 041 ns 064 017 =0001 05 016 ns 062 009 <004 08 012 =0001 116 026 ns 083 015 =009 08 011 ns

F-S=free standing; Total=total subpulmonary infundibulum; i.e., free-standing infundibulum+outlet septum; narrowest diameter=in the subpulmonary area; SP=septoparietal; VSD= ventricular septal defect; decimal values P: values compared to the rst column; ns=not signicant; Os=outlet septum; TVd= measurements indexed to diameter of tricuspid valvar orice.

the outlet (infundibular or conal) septum within the cavity of the right ventricle, was signicantly longer than in normals. In the studies of Geva et al.[4], and of Van Praagh and associates[1], no distinction has been made between the outlet septum and the free-standing subpulmonary infundibulum. Yet, without distinguishing between the two, it is not possible to compare the

abnormal with the normal. The outlet septum, dened as the muscular structure which, within the ventricular cavity, separates the subaortic from the subpulmonary outlets, can be dened only in the presence of a ventricular septal defect. In the setting of tetralogy of Fallot, and also in the Eisenmenger type of ventricular septal defect, this septum is an extensive structure, albeit within the cavity of the right ventricle rather than between the ventricles (Figs 1 and 4). The free-standing subpulmonary infundibulum, which forms the distal component of the subpulmonary outlet in tetralogy, is the sleeve of muscle located between the cavity of the subpulmonary outlet and the extracardiac space, being related posteriorly to the aortic valvar sinuses rather than the cavity of the left ventricle (Fig. 4). This free-standing infundibular sleeve exhibited considerable length in Fallots tetralogy, but was longer in healthy controls because it represented the full extent of the normal subpulmonary outlet. But no part of the infundibular tube can be likened to a true septum. This distinction, between a discrete intracardiac muscular septum and the free-standing sleeve of subpulmonary infundibulum, which we failed to make in our earlier studies[2,3], has more than academic implications. It has obvious surgical signicance. Excessive resection of the hypertrophied body of the outlet septum can place at risks the aortic valvar leaets which are supported on its left ventricular surface. In contrast, should incisions be extended towards the pulmonary valvar leaets, the risk is then courted of exiting the heart and damaging the adjacent aortic sinus. This distinction of the proximal and distal components of the subpulmonary infundibulum has further implications for previous disagreements. In essence, the arrangement of the right ventricular outlet musculature seen in tetralogy of Fallot represents failure of the various building blocks of the developing outow tracts to coalesce and form the normal supraventricular crest. It is an easy matter to distinguish the free-standing subpulmonary infundibulum in the normal heart, and to identify its area of fusion with the muscular ventricular septum, the latter being reinforced by the septomarginals trabeculation (septal band) at the septal aspect of the union (Fig. 3(a)). This produces the well recognized arrangement of the supraventricular crest inserted between the two limbs of the septomarginal trabeculation. Virtually the entirety of the supraventricular crest, however, is formed of parietal rather than septal ventricular musculature. The larger part is the structure which we have previously dened as the ventriculo-infundibular fold[7] (Figs 1(c) and 3(b),(c)). This muscular crest is bordered on its epicardial aspect in part by the transverse sinus, containing the right coronary artery, but also by a virtual tissue space separating the subpulmonary infundibulum from the aortic sinuses. It is not possible, in the normal heart, to distinguish separate components within the muscular ventricular septum itself. Recognition of these anatomical relationships, well seen in echocardiographic images, brings into focus several important features (Fig. 1). First, there is just as
Eur Heart J, Vol. 20, issue 3, February 1999

228

M. A. Gatzoulis et al.

Figure 4 (a) The right ventricle of a heart with tetralogy of Fallot resembling subcostal right anterior oblique echocardiographic views, as shown in Fig. 1(a). The overriding of the aortic valve through the ventricular septal defect which is located between the anterior (A) and (P) limbs of the septomarginal trabeculation is evident. The septal insertion of the ventriculo-infundibular fold is attenuated (*), leaving an area of brous continuity between aortic, triscupid and mitral valves. The deviated outlet septum ( ), cut in prole, narrows the pulmonary outow tract. (b) A specimen of Eisenmenger ventricular septal defect showing morphological features corresponding to the echocardiogram shown in Fig. 1(b). The hatched areas indicate the free-standing subpulmonary infundibulum. much o-setting of the leaets of the arterial valves at the ventriculo-arterial junction as there is of the hinges of the atrioventricular valves at the septal atrioventricular junction. Second, the long axes of the two ventricular outow tracts are almost at right angles to each other. Third, the aortic sinuses are related spatially, in part, to the cavity of the right ventricular infundibulum, but the leaets of the normal aortic valve are, of necessity, attached exclusively within the left ventricle. It is potenEur Heart J, Vol. 20, issue 3, February 1999

tially misleading, therefore, to suggest that this usual arrangement represents overriding of the aortic valve[9]. In our opinion, overriding is better dened on the basis of biventricular connection of the aortic valvar leaflets[10]. This arrangement, self-evidently, can be seen only in abnormal hearts. In Fallots tetralogy, due to the abnormal development, each of these various building blocks can be recognized in its own right[7,11]. The outlet septum is malaligned relative to the muscular ventricular

Echocardiographicalmorphological correlations in tetralogy of Fallot

229

Figure 5 These right ventricular views show a specimen with Eisenmenger type of ventricular septal defect (a) and a specimen with tetralogy of Fallot (b). The outlet septum ( ) is an exclusively right ventricular structure in both cases. The pulmonary outow tract is unobstructed in (a). In contrast, there is severe restriction exacerbated by hypertrophy of septoparietal trabeculations in (b). A=anterior limb; P=posterior limb of septomarginal trabeculation. septum, and is inserted to the rest of the muscular septum in relation to the anterior limb of the septomarginal trabeculation (septal band). The inferior border of the ventricular septal defect, therefore, formed by the crest of the muscular ventricular septum, is cradled between the limbs of the septomarginal trabeculation. The aortic valvar leaets connected within the right ventricle are separated from the tricuspid valve by the ventriculo-infundibular fold, this musculature forming the ventricular roof. The extent of this fold relative to
Eur Heart J, Vol. 20, issue 3, February 1999

230

M. A. Gatzoulis et al. junction remains unclear. It may be that other factors, such as timing of repair, the dimensions of the pulmonary arterial tree, and the preference of dierent surgeons, also play a role. Further research is clearly needed in this area. It is also the case that some individual patients, or hearts, can show hypoplasia of the outlet septum and subpulmonary infundibulum, indicating that Fallots tetralogy represents a spectrum of anatomical malformations. Our measurements show unequivocally, nonetheless, that the subpulmonary infundibulum, when measured from the tip of the outlet septum, is elongated rather than shortened within the overall group. We conclude that it is the position of the outlet septum, combined with its relationship to the remainder of the muscular subpulmonary infundibulum, that represents a hallmark of Fallots tetralogy, allowing its dierentiation from both Eisenmenger ventricular septal defects and the normal heart.

the septomarginal trabeculation determines whether the right ventricular margin of the ventricular septal defect is, in part, brous or entirely muscular[12]. Sectional studies, both echocardiographic and morphological, then show that it is possible to distinguish the outlet septum from the rest of the muscular ventricular septum in tetralogy of Fallot. It is seen as a discrete muscular structure located exclusively within the cavity of the right ventricle. The sections also show that the septum is continuous distally, beyond the hinge of the leaets of the aortic valve from the aortic sinuses, with the posterior margin of the free-standing sleeve of subpulmonary infundibulum. This arrangement is seen in both tetralogy of Fallot and in the Eisenmenger type of ventricular septal defect. The insertion of the outlet septum relative to the anterior limb of the septomarginal trabeculation is comparable in the two lesions, and is responsible for the so-called malaligned ventricular septal defect[12]. The connection of the outlet septum in ring-like fashion with the hypertrophied and septoparietal trabeculations produces the os of the infundibulum, and distinguishes between Fallots tetralogy and Eisenmenger variations (Fig. 5). We should also emphasize that, although the overall extent of the subpulmonary infundibulum, as measured from the os, is unequivocally longer in hearts with Fallots tetralogy than in the normal heart, its volume is signicantly less than normal because of its narrowness. The inverse relationship between the degree of anterocephalad deviation of the outlet septum and the size of the pulmonary trunk in patients with Fallots tetralogy suggests that this deviation is not only the morphogenetic essence of the tetrad, but may also aect the growth of the pulmonary arteries. Linear correlations were identied between the dimensions of the pulmonary valve, pulmonary trunk and right pulmonary artery, suggesting a possible interrelationship between the size of these structures and the development of the pulmonary arterial tree. In this respect, Hornberger et al.[13], in a recent serial echocardiographic study of pulmonary arterial growth in Fallots tetralogy, found hypoplasia of the pulmonary trunk at the initial fetal examination to be a consistent feature in fetuses with more severe post-natal obstruction of the right ventricular outow tract. The relationship between the type of surgical reconstruction of the right ventricular outow tract and pulmonary trunk is also of interest. In our Institution, the preferred surgical approach is construction of a modied BlalockTaussig shunt in symptomatic patients seen before the age of 6 months, and elective primary repair between 6 and 18 months of age[14]. Transannular patching, as encountered in this current study, bore no direct relationship to the dimensions of structures within the right ventricle, such as the outlet septum or septoparietal trabeculations. In contrast, hypoplasia of the pulmonary trunk, and perhaps more severe degrees of pulmonary valvar stenosis, seemed to predispose to the need of augmentation with a patch. The exact substrate, morphological or other, necessitating incisions across the ventriculo-pulmonary
Eur Heart J, Vol. 20, issue 3, February 1999

Limitations of the study

In the interest of simplicity, we chose standard echocardiographic views with which to make our measurements. In addition, we employed comparable sections for our morphological measurements, thus allowing for relative comparisons of the ndings from the two parts of our study. No attempts were made for direct comparisons between the echocardiographic and morphological data, however, as they represent two dierent modes, and have been indexed in dierent ways. There are potential problems in making accurate echocardiographic measurement of the structures forming the right ventricular outow tract and pulmonary arterial tree. Intraand inter-observer variability in this study, however, was acceptable. It should also be noted that, when problems of nomenclature are excluded, our ndings were remarkably similar to those of Geva and his colleagues[4]. The major point which emerges from our study is the need to distinguish between septal and free-standing infundibular structures.
We thank our colleagues at the Royal Brompton Hospital for allowing us to study their patients. Drs Ho and Anderson are supported by the British Heart Foundation, together with the Joseph Levy Foundation.

References
[1] Van Praagh R, Van Praagh S, Nebesar RA et al. Tetralogy of Fallot: Underdevelopment of the pulmonary infundibulum and its sequel. Am J Cardiol 1970; 26: 2433. [2] Becker AE, Connor M, Anderson RH. Tetralogy of Fallot: a morphometric and geometric study. Am J Cardiol 1975; 35: 40212. [3] Howell CE, Ho SY, Anderson RH, Elliot MJ. Variables within the brous skeleton and ventricular outow tracts in tetralogy of Fallot. Ann Thorac Surg 1990; 50: 4507. [4] Geva T, Ayres NA, Pac FA, Pignatelli R. Quantitative morphometric analysis of progressive infundibular obstruction in tetralogy of Fallot: A prospective longitudinal echocardiographic study. Circulation 1995; 92: 88692.

Echocardiographicalmorphological correlations in tetralogy of Fallot

231

[5] Oppenheimer-Dekker A, Gittenberger-de Groot AC, Bartelings MM, Wenink ACG, Moene RJ, van der Harten JJ. Abnormal architecture of the ventricles in hearts with an overriding aortic valve and a perimembranous ventricular septal defect (Eisenmenger VSD). Int J Cardiol 1985; 9: 34155. [6] Al-Marsafawy AHF, Ho SY, Redington AN, Anderson RH. The relationship of the outlet septum to the aortic outow tract in hearts with interruption of the aortic arch. J Thorac Cardiovasc Surg 1995; 109: 122536. [7] Anderson RH, Becker AE, Van Mierop LHS. What should we call the crista? Br Heart J 1977; 39: 8569. [8] Goor DA, Lillehei CW. Congenital malformations of the heart. Embryology, anatomy, and operative considerations. New York: Grune & Stratton, 1975: 145. [9] Kleinert S, Geva T. Echocardiographic morphometry and geometry of the left ventricular outow tract in xed subaortic stenosis. J Am Coll Cardiol 1993; 22: 15018.

[10] Alva C, Rigby ML, Ho SY, Anderson RH. Overriding and biventricular connection of arterial valves. Cardiol Young 1998; 18: 15064. [11] Capuani A, Uemura H, Ho SY, Anderson RH. Anatomic spectrum of abnormal ventriculo-arterial connections: surgical implications. Ann Thorac Surg 1995; 59: 35260. [12] Baker EJ, Leung MP, Anderson RH, Fischer DR, Zuberbuhler JR. The cross-sectional anatomy of VSDs: a reappraisal. Br Heart J 1988; 59: 33951. [13] Hornberger LK, Sanders SP, Sahn DJ et al. In utero pulmonary artery and aortic growth and potential for progression of pulmonary outow tract obstruction in tetralogy of Fallot. J Am Coll Cardiol 1995; 25: 73945. [14] Norgard N, Gatzoulis MA, Moraes F et al. The relationship between type of outow tract repair and postoperative right ventricular diastolic physiology in tetralogy of Fallot: Implications for long-term outcome. Circulation 1996; 94: 327680.

Eur Heart J, Vol. 20, issue 3, February 1999