anatomical variations conduction system with reference to … · fers according to the type...

British HeartJournal, I972, 34, I85-I90.

Anatomical variations in atrloventricularconduction system with reference toventricular septal defects

R. A. Latham' and R. H. AndersonFrom the Institute of Child Health, University of Liverpool;and Department of Anatomy, University of Manchester

The course of the atrioventricular conducting tissues was studied in IO hearts having ventricularseptal defects. Seven of the defects were in the membranous portion of the septum, and in thesehearts the common atrioventricular bundle passed down the posterior edge of the defect and bifur-cated on its inferior border. In 3 of these specimens the bundle deviated to the left side of the sep-tum before bifurcating. The 3 remaining hearts possessed defects of the muscular septum. In 2of these the common bundle bifurcated within the septum anterosuperiorly to the defect. The leftbranch descended to become intimately related to the anterior border of the defect. In the otherheart with low anterior intramuscular defects the right bundle-branch traversed the septum anddivided astride the defect. These results are discussed in relation to surgical repair of the defectsand to the reports of previous investigators. The added risk of damaging the conducting tissueis emphasized in cases which deviate from the usually accepted pattern.

Previous studies on hearts with intramem-branous defects of the ventricular septumhave emphasized the relation between theatrioventricular bundle and the postero-inferior border of the defect (Reemtsma andCopenhaver, I958; Reemtsma, Copenhaver,and Creech, I958; Truex and Bishof, I958;Lev, I960; Titus, Daugherty, and Edwards,I963a). While Lev (I960) suggested that thebundle may be related to intramuscular de-fects, the general inference was that thesewere unlikely to be intimately involved withthe conducting tissue. Our attention wasdrawn to this problem in the course of a

histological study of a heart in which an

intramuscular defect had been repaired. Wewere surprised to find that the right bundle-branch had been extensively traumatized, andthat the left branch passed down the anteriorborder of the defect. It was therefore decidedto reinvestigate the course of the atrioventri-cular specialized tissue in relation to varyingtypes of interventricular septal defects.

Materials and methodsThe hearts used in this investigation were takenfrom the cardiopathological collection of the

Received 7 May 1971.1 Present address: Dental Research Center, Universityof North Carolina, Chapel Hill, U.S.A.

Royal Liverpool Children's Hospital. Ten heartscontaining varying interventricular septal defectswere studied. In addition, three hearts withoutdefects were examined for comparative purposes.The type of defect present in the hearts is indi-cated in the Table.The hearts were first photographed, and then

a block of tissue was dissected out containing theatrioventricular septum from the coronary sinusto the defect. This block was then rephotographedfrom right and left aspects, before dehydrationand embedding in paraffin wax. Serial sectionswere cut at IosU and several sections in every 50

TABLE Defects in hearts studied

Specimen Block Age Type of defectNo. No.

Intra- Intra-mem- muscularbranous

782 io8 5 mth X73I 110 22dy X686 III 27 dy X6I2 112 i yr 5 mth X794 I13 12wk X798 II4 51 yr XI97 ii6 4 mth X X226 I17 io wk X647 ii8 2yr 8 mth X587 II9 2 yr I mth X

on May 8, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.34.2.185 on 1 February 1972. D

ownloaded from

http://heart.bmj.com/

i86 Latham and Anderson

were mounted and stained with Masson's tri-chrome technique. The intervening sections werestored and, after examination of the mountedsections, further sections were mounted andstained as required. In some cases adjacent sec-tions were stained by the PAS technique, afterblockade in 5 per cent dimedone, to show glycogen(Bulmer, I959).

ResultsIt was relatively easy to define histologicallythe atrioventricular bundle and its branches.The tracts of specialized cells were surroundedby connective tissue and were well demarcatedfrom the adjacent myocardium. The sectionsprocessed to show glycogen revealed littledifference between the specialized cells andmyocardial cells. In size, the specialized cellswere marginally larger than ventricular myo-cardial cells, and exhibited cross-striations.No cells were identified in the sections exam-ined which corresponded to the 'Purkinje'cells seen in artiodactyla except in poorlyfixed tissue.The course of the specialized tissue was

found to vary according to the type of defect.It was difficult to distinguish these defects onany strict macroscopical criteria, but 7 weredesignated as within the membranous portionof the septum while 3 were intramuscular.The intramembranous defects were generallymore anterior than the muscular type, closerto the atrioventricular ring, and higher inrelation to the crista supraventricularis.

a) Intramembranous defects Two groupswere identified according to the course of theatrioventricular bundle.

Five hearts were virtually identical andconstituted the first subgroup. These wereNo. 226, 6I2, 647, 686, and 73I in the RoyalLiverpool Children's Hospital collection (seethe Table). The atrioventricular bundle wasfound to pierce the fibrous annulus in a nor-mal fashion. As it reached the ventricularseptum the bundle came into contact withthe defect and descended along its posterioredge. Close to the inferior edge of the defectthe bundle began to bifurcate. The leftbundle-branch was an extensive structurewhich passed from the bifurcation along thelength of the inferior edge of the defect anddescended subendocardially on the left sideof the septum. The right branch was a thickcord given off the bifurcation near the middleof the inferior edge of the defect. This struc-ture passed intramyocardially to emerge in thesubendocardium near the anterior papillarymuscle of the tricuspid valve. The arrange-ment of specimen 686 is shown in Fig. i, 2,

ABCD

2 /BCI

3C

F I G. i Block Iii from specimen 686, showingintramembranous septal defect (VSD), openingof left superior vena cava (CS), and tricuspidvalve (TV).FIG. 2 Reconstruction of conducting tissueon tracing of Fig. i.FIG. 3 A-D Four representative sections atlevels A-D are shown. AVB, atrioventricularbundle; CB, common bundle; BIF, bifurcation;LBB and RBB, left and right bundle-bran-ches; TV, tricuspid valve. The arrow is in thedefect. All are at the same magnificationas shown by 0.55 mm line in 3D.

and 3 A-D. Specimen No. 647 varied slight-ly in that an interatrial septal defect was alsopresent. In this case the bundle deviated tothe leftafterpiercingthe annulusandcontinued



j.com/

Br H


ownloaded from


Conducting tissue in association with ventricular septal defects 187

along the left side of the bar of tissue betweenthe defects until it reached the ventricular de-fect. It then bifurcated on the posteroinferioredge in the manner described above.The second subgroup comprised specimens

794 and 798. The bundle descended onto theposterior edge of the defect as described, butthen deviated to the left. It then passed anappreciable distance down the left side of theseptum before bifurcating on the inferior bor-der of the defect. The left branch spread outsubendocardially while the right branch,again cord-like, ascended through the myo-cardium of the septum to emerge near theanterior papillary muscle (Fig. 4-7).

b) Intramuscular defects Three heartswere examiined with defects of this type. InNo. I97 and 782 the defect was close to theatrioventricular ring, simulating the intra-membranous type. In the third heart (No. 587)there were two defects, both remote from theAV ring. The conducting tissue followed dif-ferent routes in each case.

Specimen No. 197 was further complicated,since histological examination revealed thepresence of a small intramembranous defect.The bundle passed posteroinferiorly to this,as in the other fibrous defects, but did notbifurcate. Bifurcation occurred in the samefashion as that illustrated for No. 782 in Fig.8, 9, and IO A-D. This specimen was the firstheart examined in this series and it revealedwidespread trauma to the right bundle-branch in front of the defect. The atrioven-tricular bundle pierced the annulus in thenormal fashion before the bundle entered themuscular septum. In this position the righthalf was infiltrated with haemorrhagicexudate, though there was no evidence ofdirect trauma. The bundle passed throughthe septum to a point above the antero-superior edge of the defect and bifurcatedwithin the septum. The left branch immedi-ately fanned out to form a typical subendo-cardial ribbon. The posterior fascicles of thisbranch passed directly onto the anterior edgeof the defect, occupying its full thickness. Onreaching the inferior edge they curved aroundit and then descended on the left side of theseptum. The cord-like right branch wastraced into a large haematoma present in theseptum anterior to the defect and also involv-ing the anterior papillary muscle. The ter-minations of the bundle in heart No. 197showed identical relations to the musculardefect.The remaning heart in this series, No. 587,

had distal duplicated defects. The bundle in

FIG. 4-7 Representative sections of speci-men 798 with an intramembranous defect. Thearrow indicates the defect. Abbreviations asbefore.

this case bifurcated normally on the crest ofthe septum and the left branch occupied theexpected position. The cord-like right branchfollowed an extensive intramyocardial routebeneath the anterior papillary muscle andemerged on the posteroinferior edge of thesuperior defect (Fig. ii and I2). It then bifur-cated again in this position and the divisionspassed down on both right and left sides ofthe septum.

Fig. I3 illustrates the varying routes takenby the conducting tissue with reference tothe types of ventricular septal defectdescribed.

I MT" :,.Jps inall.

ors,



j.com/

Br H


ownloaded from


ix8 Latham and Anderson

Discussion

The results of this investigation indicate thatthe course of the atrioventricular bundle dif-fers according to the type of interventriculardefect. However, in each heart studied, somepart of the conducting tissue was intimatelyrelated to the defects. Thus, in any operationdesigned to repair a ventricular septal defectsome portion of the specialized conductingtissue is 'at risk'. It is important that surgeonsshould appreciate the probable course of theconducting tissue in relation to any particulardefect. Reemtsma and Copenhaver (I958),Reemtsma et al. (I958), Truex and Bishof(I958), Lev (I960), and Titus et al. (i963a)all studied high defects of the membranousportion of the interventricular septum. Wecan confirm their findings in this particulartype of defect. The bundle and its bifurcationare intimately related to the posteroinferiorquadrant of such a defect. In repairing sucha defect from the right the left bundle-branchis in a relatively safe position. Thus, if theposteroinferior quadrant is avoided the onlyother structure at risk is the right bundlebranch.Our investigations suggest that injury of

this intramyocardial fascicle is of importancenot only with regard to the production ofright bundle-block, as haemorrhagic exudatecan apparently track in a retrograde directionalong the conducting tissue. In the specimendescribed, injury of the right branch welldown the septum produced haemorrhagicexudate in the parent bundle as it pierced thefibrous annulus. Thus, injury of the distalright branch could possibly produce com-plete atrioventricular dissociation by retro-grade spread, producing oedema or fibrosis.Titus et al. (I963b) and McGoon, Ongley,and Kirklin (I964) pointed out that in patientssuffering from AV dissociation it was rare tofind evidence of direct trauma. This wouldalso suggest that the bundle is likely to besensitive to oedema or pericellular exudate.It therefore seems to be important to avoid theright bundle-branch if at all possible. In allthe specimens presently studied with intra-membranous defects this structure ran intra-myocardially from the inferior edge of thedefect towards the anterior papillary muscle.If this area is avoided during suturing, inaddition to the posteroinferior quadrant, thendamage to the conducting tissue should beavoided.The variations noted from this pattern

described in membranous defects were onthe left side of the septum. These should makethe bundle less at risk when repaired from the

9

ABC D

ABC D

R4JW T'^.U;F,.s mm

EF:^s$' d.gN;

IQE

FIG. 8 Block io8 from Specimen 782.Abbreviations as before; P, patch on defect;AP, anterior papillary muscle.FIG. 9 Reconstruction of conducting tissueon tracing of Fig. 8.FIG. I0 A-D Four representative sectionsat levels A-D. Abbreviations as before; he,haemorrhagic exudate; h, haematoma. Thetwo arrows indicate the anterior extent of thedefect.

right. The right bundle-branch followed thesame intramyocardial course in these heartsalso, so avoidance of the area above the an-terior papillary muscle should spare the con-ducting tissue in these patients.We found a relation between intramuscular

defects and the conducting tissue. This is



j.com/

Br H


ownloaded from


Conducting tissue in association with ventricular septal defects 189

quite different from the arrangement in mem-branous defects, and it is important that thedifference should be noted. This is particu-larly so in defects near the atrioventricularring which resemble membranous defects onmacroscopical examlination. Previous workersfound no relation between such defects andthe conducting tissue. Lev (I960), however,inferred that the conducting tissue might berelated to the anterior edge of a muscular de-fect. The present work presents evidence sub-stantiating this inference. In both heartsstudied where the muscular defect was nearthe ring, the bundle bifurcated above theanterosuperior quadrant of the defect. Theleft branch was intimately related to the an-terior edge of the defect. The right branchpassed intramyocardially towards the an-terior papillary muscle, again anterior to thedefect. Thus, in these hearts it is essentialto avoid the anterosuperior and anterior edgesof the defect, in contradistinction to avoidingthe posteroinferior quadrant in muscular de-fects. It is therefore of some importance todistinguish intramembranous and intramuscu-lar defects before starting a repair operation.In the present series the intramembranousdefects were in all cases closer to the AV ring,higher in relation to the crista supraventricu-laris, and more anterior in the septum.The other specimen studied with an intra-

myocardial defect was also instructive. Thisdefect was distant from the AV ring and thebifurcation, yet the right branch ran a lengthyintramyocardial route before bifurcating onthe inferior edge of the defect. Possibly thisrelation was coincidental, in that the defectwas close to the anterior papillary muscle.However, in the absence of contrary studies itmay be prudent to avoid the quadrant nearestthe anterior papillary muscle in repair opera-tions on such distal defects.The described relation of the specialized

conducting tissue to the differing types ofdefect is to be expected from its embryo-

II

FIG. II Photograph of Specimen 587. D,defects; apm, anterior papillary muscle.FIG. 12 Bifurcation of right bundle branch(RBB) on the lower edge of superior defect(indicated by arrow). IVS, interventricularseptum.

logical development. The bundle bifurcatesastride the developing muscular portion of theinterventricular septum, and below the por-tion formed from endocardial and bulbarmusculature (Boyd, I965). Thus, the special-ized tissue would be situated below mem-branous defects but above muscular defects.

This arrangement would be found irrespec-tive of the origin of the bundle. Walls (I947)considered that the bundle was formed byactive migration. On this principle, cases suchas that described by Huntingford (I960) areinterpreted as representing postjunctionalsevering ofthe bundle. However, James (I970)opined that these cases were better explainedby lack of union of the node and the bundle,

F I G. I3 Diagrams showing course of conducting tissue in relation to defects studied. A, B-simple intramembranous defects. C-intramembranous defect associated with ASD. D, E-intramuscular defects.

A B C E

/



j.com/

Br H


ownloaded from


igo Latham and Anderson

each developed in situ. Physiological findingssupport this contention (Lieberman, I970)while recent studies of embryological andcongenitally malformed specimens providemorphological evidence of its correctness(Anderson and Taylor, unpublished data).A further point arising from this investiga-

tion is of morphological interest. None of thehearts studied showed cells that correspondedto the cells identified in artiodactyla byPurkinje (I845). Indeed, the specialized cellsdiffered only slightly from ventricular myo-cardium, being recognized in collectionsrather than as single cells. Workers such asJames (I96I) used the presence of 'Purkinje'cells as criteria for recognizing conductingtissue in extranodal sites. On the other hand,Davies (I942) found no 'Purkinje' cells inthe upper parts of the interventricular sep-tum, a view endorsed by Copenhaver andTruex (I952) who were unable to identifysuch cells anywhere within the AV node.Truex (I96I) expanded this hypothesis andsuggested that the presence of 'Purkinje'characteristics might be a consequence offixation artefacts, since he noted segments ofnormal atrial cells possessing similar charac-teristics. He thought it probable that true'Purkinje' cells existed only in artiodactyla.The present investigation supports his andprevious workers' views with regard to thehuman heart. This point and others with re-gard to 'Purkinje' cells were discussed morefully in a paper relating to the architecture ofthe AV node in the hearts presently studied(Anderson and Latham, I97I).

The authors wish to thank Dr. R. S. Jones for hisinterest in this work, Mrs. A. Smith for histo-logical work, and Messrs. K. Walters and R.Wylie for photomicrography.

ReferencesAnderson, R. H., and Latham, R. A. (I97i). The

cellular architecture of the human atrioventricularnode, with a note on its morphology in the presenceof a left superior vena cava. Journal of Anatomy.In the press.

Boyd, J. D. (I965). Development of the heart. InHandbook of Physiology. Section 2: Circulation,Vol. 3, pp. 25II-2543. Ed. by W. F. Hamilton.American Physiological Society, Washington.

Bulmer, D. (1959). Dimedone as an aldehyde blockingreagent to facilitate the histochemical demonstra-tion of glycogen. Stain Technology, 34, 95.

Copenhaver, W. M., and Truex, R. C. (I952). His-tology of the atrial portion of the cardiac conduc-tion system in man and other mammals. Anatomi-cal Record, 114, 6oi.

Davies, F. (I942). The conducting system of the verte-brate heart. British Heart_Journal, 4, 66.

Huntingford, P. J. (I960). The aetiology and signifi-cance of congenital heart block. journal of Obstet-rics and Gynaecology of the British Empire, 67, 259.

James, T. N. (I96I). Morphology of the human atrio-ventricular node, with remarks pertinent to itselectrophysiology. American Heart_Journal, 62, 756.

James, T. N. (I970). Cardiac conduction system: fetaland postnatal development. American Journal ofCardiology, 25, 2I3.

Lev, M. (I960). The architecture of the conductionsystem in congenital heart disease. III. Ventricularseptal defect. Archives of Pathology, 70, 529.

Lieberman, M. (I970). Physiologic development ofimpulse conduction in embryonic cardiac tissue.American Journal of Cardiology, 25, 279.

McGoon, D. C., Ongley, P. A., and Kirklin, J. W.(I964). Surgical heart block. American journal ofMedicine, 37, 749.

Purkinje, J. E. (I845). Mikroskopisch-neurologischeBeobachtungen. Archiv fur Anatomie, Physiologieund wissenschaftliche Medicin, 12, 28I.

Reemtsma, K., and Copenhaver, W. M. (I958).Anatomic studies of the cardiac conduction systemin congenital malformation of the heart. Circula-tion, 17, 271.

Reemtsma, K., Copenhaver, W. M., and Creech, 0.(I958). The cardiac conduction system in congenitalanomalies of the heart: studies on its embryology,anatomy, and function. Surgery, 44, 99.

Titus, J. L., Daugherty, G. W., and Edwards, J. E.(I963a). Anatomy of the atrioventricular conduc-tion system in ventricular septal defect. Circulation,28, 72.

Titus, J. L., Daugherty, G. W., Kirklin, J. W., andEdwards, J. E. (I963b). Lesions of the atrioven-tricular conduction system after repair of ventricu-lar septal defect; relation to heart block. Circulation,28, 82.

Truex, R. C. (I96I). Comparative anatomy and func-tional considerations of the cardiac conductionsystem. In The Specialized Tissues of the Heart(Proc. ofa Symposium, I960, Rio de Janeiro), p. 22.Ed. by A. Paes de Carvalho, W. C. de Mello, andB. F. Hoffman. Elsevier, Amsterdam.

Truex, R. C., and Bishof, J. K. (I958). Conductionsystem in human hearts with interventricular septaldefects. Journal of Thoracic Surgery, 35, 421.

Walls, E. W. (I947). The development of the special-ized conducting tissue of the human heart. journalof Anatomy, 8I, 93.

Requests for reprints to Dr. R. H. Anderson,Anatomy Department, The University, Man-chester, MI3 9PL.



j.com/

Br H


ownloaded from


anatomical variations conduction system with reference to … · fers according to the type...

Documents