CONGENITAL MALFORMATIONS IN ONE OFMONOZYGOTIC TWINS

BY

J. EDGAR MORISON, M.D., B.Sc.(From the Institute of Pathology, Queen's University, Belfast)

The purpose of this communication is to describetwo cases of gross foetal abnormaity and to discussthe influence of intra-uterine environment on thedevelopment of congenital abnormalities when thecontribution of inherited or genetic factors is madeknown by the co-existence of a genetically similarbut normal twin. The anomalies will not bediscussed in detail, but some explanation of themechanismn of their occurrnce will be attempted,since, even if this proves incorrect, it may stimulateuseful reonsideration of the problems presented bycongenital malformations which now assume sucha great importance in practice.

like Twins and Factors Inth-mning li XnceIf the outcome of intra-uterine development is

determined sokly by genetic inheritance, mono-zygotic or like twins should be completelyindistinguishable from one another at birth. Rightand left asymmetry with a mirror image differenceoften does exist, but this is not a valid difference andneed not be discussed now. It is widely admitted,however, that even like twins may differ in size atbirth. Moreover. Schatz (quoted by Newman,Freeman, and Holzinger, 1937) found that at aboutthe middle period of pregnancy the size differencein monozygotic twins was greater than at term and,on the average, much greater than for dizygotic orunlike twins. The careful survey of the literatureon twin pregnancies with one twin blighted madeby Kindred (1944) is also relevant Here one twindies, usually between the third and the fifth month,and is retained in utero until the birth of its livingpartner at, or near, term. This condition has aroughly similar incidence in both monozygotic anddizygotic twins. The occurrence of this conditionmust suggest that the development of twins, evenwhen they are derived from the division of the sameferdlizd ovum and possess an identical geneticinheritance, can be very significantly influenced bydifferences in the environment they establish forthemselves in the maternal uterus. Joined orSiamese twins and double monsters are often closelyidentical (Szendi, 1939). Parasitic twins andteratomata, if it is insisted that these latter arisefrom cells laid aside so early in development, areboth examples of much greater inequality of growthand even of organ differentiation. Here, however,

division of the initial cell mass between the twoindividuals is relatively late, and may be unequal,and the juxtaposition of the growing tissues and theorganizers of the two individuals may disturbdevelopment These more extreme abnormalitiesare, therefore, unsuitable material for a study ongenetically similar material of the role of environ-mental differences in utero and especially for thestudy of the effect of nutritional differences.

Despite the value of the study of like and unliketwins in the assessment of the relative contributionof the environment and of heredity in post-natal lifein health (Newman, Freeman, and Holzinger, 1937,and Newman, 1942) and in disease (Ka1mann andReisner, 1943), twins have been little studied for dataon the relative importance of environment andheredity in intra-uterine development. The Mateerembryo of the pre-somite period (Streeter, 1920)showed a normal embryo and probably also anabnormal and yet monozygotic twin. Such casesare of interest, but have not been discussed in thisconnexion. Macklin (1936) collected from theliterature cases of malformation in twins. Therecords rarely indicated the nature of these twins.She found, however, that twins of like sex wereconcentrated in the group where both twins weremalformed. Where but one twin was malformed,the twins were presumably binovular (dizygotic),becausw they were as often of different sex asordinary siblings. She argued that this was goodevidence that malformations were not determinedby environmental causes, but were due to inheriteddefects of the genetic inheritance. To-day there isconvincng evidence that some congenital mal-formations can, and do, arise from environmentalcauses, such as rubella. It is not necessary to thinkthat even identical abnormalities always have thesame causes. Environmental causes may createcopies (phenocopies) of conditions which are moreoften determined by genetic factors (Goldschmidt,1938). Information on any environmental condi-tions which may operate in this way is mostdesirable.The probable nature of this environmental

disturbance must be considered. Studies of theoccurrence of malformation when the maternal dietis deficient, and where it is presumed in consequencethat embryonic nutrition is also deficient, providesome evidence in animals, but no satisfactory

214

Protected by copyright.

on Septem

ber 27, 2020 by guest.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.24.119.214 on 1 S

eptember 1949. D

ownloaded from

CONGENITAL MALFORMATIOANS IN ONE OF MONOZYGOTIC TWINSevidence that in man nutritional requirements candetermine malformations. The blighting or deathof one ovum of a monozygotic pregnancy is probablyto be regarded as the result of unequal sharing ofnutrient. The deficiency usually leads to death and,if abnormalities are present in the dead and blightedovum, it will very likely be impossible to establishthis months later from its macerated or papyraceousremains. The uterus otherwise must present a verysimilar environment to both twins and such maternalinfluences as general nutrition, infection, andtoxaemia should affect two genetically similarembryos equally.

Criteria for Diagnosis of Monozygotic TwinsIt is necessary to decide on what basis a diagnosis

of monozygotic or like twins can be made. Diagnosisby the study of the foetal membranes has fallen intosome disrepute among those who study twins in latepost-natal life. A casual examination of thesemembranes is useless and they must be properlyexamined grossly and microscopically and withadequate tissue sections. Very rarely monozygotictwins are enclosed in a common amnion; usuallythey are separated only by the walls of their amnioticcavities and the chorion should continue as acommon membrane around both. In dizygotictwins there are four membranes between thefoetuses, two amniotic and two chorionic layers(dichorionic placentae), and, though the twoplacentae may appear fused, there is no continuouslayer of chorion from one placenta to the other.Some confusion has been caused because someworkers have failed to recognize that a fairly highproportion, probably about 30 per cent., of mono-zygotic or like twins are born in dichorial placentae.In these cases the ovum cleavage has occurred early;the separated cells are not enclosed in a commonmembrane and each forms its own chorion.

It is important to decide if dichorionic, usuallydizygotic, twins can come to possess a singlechorion. Arey (1922) has suggested on insufficientevidence that occasionally the two:separate chorioniccavities may become continu-ous witn one anotner tnrougnthe breakdown of the adjacentchorionic walls. If this occursthe retention of a partitionformed by two thin and fusedamniotic walls would seemsomewhat unlikely and suchtwins would probably bedescribed as monochorionictwins with a common amnion.It is highly improbable thatthose monozygotic twins ofidentical inheritance whopossess a monochorionicplacenta and two amnions willre mnstaken ior auzygouc twmsof unlike inheritance, if the FIG. 1.-Tme circulamembranes are carefully

studied. Additional support for a monozygoticorigin may be provided by a study of the placentalblood vessels. There is some agreement that inman readily demonstrable blood vessel anastomosesare associated only with monozygotic twins. Thusvon Verschuer (1939) injected placentae and foundthirty-two cases with circulatory anastomoses andthese were all monozygotic twins on complete testsof resemblance. In one hundred cases there weretwo chorions and no anastomoses, and of theseseventy-six were dizygotic and twenty-four weremonozygotic after similar studies. It would seemthat only some monozygotic twins possess suchanastomoses and that in man, as opposed to cattle,unlike twins do not mix their blood.The only alternative to such a study of foetal

membranes is a meticulous comparison of bloodgroups, eye colour, iris pattern, hair and skin colour,form and texture, finger and palm prints, and generalfeatures. This is largely impossible in malformedinfants, one or both of whom may die at birth.It prejudges the issue since only if there is closephysical correlation will the diagnosis of monozy-gotic twinning be made. It was not employed inthe cases to be described. Blood grouping isdifficult in the newborn and could scarcely be usefulwhen there are communications between thecirculation of the two twins. Some companrsonsmight, however, give useful collaborative evidenceof genetic identity in future cases.

Case 1. Mrs. A, a primipara aged 45 years, wasadmitted at twenty-five weeks with severe pre-eclamptic toxaemia, oedema, urine loaded withalbumen, and a blood pressure of 160/80. Atthirty weeks spontaneous labour occurred and twostillborn female infants were delivered.The placentae were fused, the partitions between

the foetuses had been accidently torn in some places,but multiple sections of this and of the junctionbetween the placentae of each foetus showed onlyamnion in the partition and the chorionic membranewas continuous from one placenta to the other.Several large blood channels (fig. 1) passed

tions of the twins communicate by blood vessels, one ofwhich is clearly seen. (Case 1.)

215

Protected by copyright.

on Septem

ber 27, 2020 by guest.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.24.119.214 on 1 S

eptember 1949. D

ownloaded from

ARCHIVES OF DISEASE IN CHILDHOOD

from the circulation of one foetus to theother.FoETus B. Intraparum stillbirth. There was a

history of maternal toxaemia, breech presentation,prolapse of the cord through the incompletelydilated os, and cessation of cord pulsation duringattemts at replacement In the foetus there wasasphyxial petechial haemorrhage in the heart, mugs,thymus, and subserosal tissues, and gross congestionof all organs.

Prematurity was 30 weeks; weight 1,800 g. Crownto heel measurement, 41 cm. There was evidence ofearly intra-uterine maceration.FoErus C. Intrapartum stillbirth. The con-

genital anomalies of the heart were persistent ostiumatrio-ventriculare communis, with an incompleteseptum primum and absence of the septumsecundum, a sngle atrio-ventricular valve, and anincomplete interventriular septum; dextro-positionof the aorta; hypoplasia of left ventricle; and slightcoarctation of aorta (infantile type). There was con-genitalatresiaoftheoesophaguswithouttracheo-oeso-phageal communication (variant of Type I of Ladd,1944). There was gross dilation of the vagina anddilation of the cervical canal and corpus uteri witha shelf-like diaphragm at the lower end of thevagina (fig. 2) and retrogade lakage of vaginalsquamous epithelial debris into the peritoneal sacwith extensive vernix peritonitis (fig. 3). Prematurity

FIG. 2.-Te large vagina, the dilated endocervical

canal and uterus, and the complex shelf-likefolds in the lower vagina which constituted theonlyobstructo,areshown. (Case 1, FoetusC.)

FiG. 3.-Vernix squames have stimulated some peritonealreaction by mononuclear cells and by multinucleatedgiant cells. (Case 1, Foetus C.)

30 weeks; weight 800 g. The crown to heel measure-ment was 33 cm.

In foetus C the cardiac anomalies are those whichwould arise from disurbances of those growthprocesses which are normally most active at the endof the fifth and beginning of the sixth week ofembryonic life. The anlag of the septum primumhas then appeared, the antero-superior and postero-inferior endocardial cushions between the futureauricls and ventricles are present, the interventri-cular septum is incomplete, and the spiral subendo-cardial bulbar ridges have formed but have separatedthe aortic and pulmonary channels only in thedistal part of the bulbus. The actively growing freeedges of the septum pimum have not united with theendocardial cushions, nor have these fused todivide the single atrioventricular channel into rightand left c s The interventicular septum isgrowmg upwards, but will not complete the separa-tion of the ventricles until about the eighth week.The growth of the bulbar ridges is active, but notcomplete, in the proximal portion of the bulbus,and the positionig of the aortic and pulmonaryorific is therefore not determined. Most of thesechanges will occur in a few days and, whatevermight subsequently happen, the heart found in thisfoetus could not then be produced. In the fifthand sixth weeks there is an almost solid mass ofepithelial cells reptsenting the future lining of theoesophagus. The cells must multiply rapidly atthis time as the oesophagus is drawn out by therapid growth of this region with the development ofthe heart and lungs. Disturbed or incoordinated

216P

rotected by copyright. on S

eptember 27, 2020 by guest.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.24.119.214 on 1 Septem

ber 1949. Dow

nloaded from

CONGENITAL MALFORMATIONS IN ONE OF MONOZYGOTIC TWINS

growth may interrupt the continuity of theepithelium and stenosis or atresia may result. Theuterine anomaly is unusual. It can perhaps beregarded as a disturbance of the lower end of theMullerian duct system. It is difficult to explain itsformation at any stage of development. In thesixth week the right and left paramesonephric ductshave appeared in the mesenchyme lateral to thecranial extremities of the mesonephric ducts. Theyare growing downwards and in each a lumen isextending towards the growing tip. Disturbance ofgrowth activity at the lower end of these ducts maycreate abnormalities which later affect their complexdevelopment even after fusion. There are many

other active growth changes occurring at this period.The future ureter and pelvis is growing upwards andthe primordia of the metanephros is appearing.The kidneys and ureters in this case were notaffected but the critical growth period for these andother structures is uncertain. The urorectal septummay not have completed the separation of the rectumand urogenital sinus, but here growth activity at thisperiod is relatively slight.

Case 2. Mrs. X was a primigravida aged 23years. She had mitral stenosis. Twins were notexpected but at thirty-five weeks gestation she was

delivered first of a normal live female child and thiswas followed by a dead-born monster.The placentae were complete. There were two

amniotic cavities enclosed in a single chorion, andthe relationship of the membranes was confirmedby histological examination. The umbilical cordsarose within 1i in. of each other. Large com-

municating channels passed between the arteries ofeach twin and between the veins.

Child Y is alive and well and shows no congenitaldefect.FoETUs Z. Intrapartum stillbirth. There was

occipital encephalocoele, complete spina bifida,Arnold-Chiari malformation; gross dysgenesis ofthe larynx, complete atresia of the trachea andoesophagus, and isolated lung rudiment; cor

biloculare with imperfect incorporation of the sinusvenosus and bulbus cordis, and diffuse endocardialfibro-elastic thickening. Scoliosis and kyphosis andimperfect development of all limbs was noted.There was gross dysgenesis and extreme hypoplasiaof the liver, no recognizable stomach dilatation or

pancreatic outgrowth, also absence of the urorectalseptum and an imperforate anal membrane. Therewas a single hydronephrotic kidney with imperforateureter, and ovaries without a demonstrable ductsystem.The lungs were represented by a nodule of tissue

2 mm. in diameter in the neck, and recognizableonly by histological examination. No trachea or

oesophagus could be found and the larynx was a

flat groove in the floor of the pharynx with twocartilage plates lying in front of this. An ovary

was identified histologically.The disturbance of development affected almost

every structure, and the primary disturbance may

well have been one operating for a period about thefourth week of intra-uterine life. About this periodembryonic growth activity, which is slightly moreadvanced in the head region, is directed to theclosure of the neural-groove in the region of thespinal cord and to the organization of the overlyingmesoderm. The primitive pharynx is forming andthe laryngo-tracheal groove on its floor is the siteof active growth which will later form the tracheaand the paired lung buds. The four divisions of theheart are present and still separate (sinus venosus,atrium, ventricle, and bulbus cordis), but soon theseptum primum and the bulbar ridges are to appearand the heart should assume its more adult shapeand the auricular and ventricular chambers bedivided by their septa. The differentiation of theskeleton of the limb buds is at a critical stage. Theanlage of the liver has appeared but the stomachenlargement and the dorsal and ventral outgrowthsof the pancreas are scarcely to be recognized. Theexpanding cloaca has scarcely yet begun to bedivided by the growth of the urorectal septum.Development of a part of the genito-urinary systemis also active, but even if this were disturbed, renalstructures might still be derived at a later date fromthe caudal end of the Wolffian duct system wheredevelopmental activity is somewhat later.

Suggested Mechanism for Production ofAbnormalities

In both cases multiple sections of the partitionbetween the foetuses showed no extension of chorioninto the partition between the two foetuses. Both,and especially case 2, showed anastomoses of largeblood vessels of the two circulations. In neithercase was one foetus parasitic on the other. Com-petition between the circulations of two foetuses forthe utero-placental site is likely to be keenest in theearly growth period when villi are growing out fromthe entire surface of the chorion. Later, when thesite of villous attachment normally comes to coveronly a part of the larger chorion, there is a moreample area for each circulation to develop in someindependence from the other and from localizeddefects in the maternal decidua.

It is perhaps reasonable to suggest that one twinin both of these pregnancies experienced inadequatenutrition early in gestation, but, unlike mostblighted ova, survived to experience an amplenutrition which permitted continued growth asgestation continued. Any extraneous influence,such as maternal toxaemia or infection, should haveaffected genetically equal twins equally. Thisdeficiency of nutrition is not likely to have been acomplete deprivation of one or more specificsubstances, though those substances whose transferat that period of gestation was most difficult wouldbe most seriously reduced. It is most useful toregard the disturbance as a somewhat non-specificinfluence causing an ' arrest of development.'Ingalls (1947) and Ingalls and Gordon (1947) haveelaborated the earlier work of Stockard on

217P

rotected by copyright. on S

eptember 27, 2020 by guest.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.24.119.214 on 1 Septem

ber 1949. Dow

nloaded from

218 ARCHIVES OF DISEASE IN CHILDHOODnon-mammalian embryos and have suggested thatdevelopment can be arrested in a variety of waysand that the result is not specific to the agent, butdepends on the time of development at which theagent acts. The disturbance operates not on somespecific organ or stage of differentiation, but onwhatever organ or organs happen to be undergoingthe most active growth changes at that time.A profound knowledge of normal development

might enable the time of a developmental arrest tobe established and perhaps related sometimes to theoperation of a known agent such as German measles.Developmental changes are crowded together,especially in the second month, but, if the arrest ofdevelopment has affected several organs whosedevelopment has been at a recognizable stage, thecoincidental occurrence of abnormalities in differentorgans might be significant and suggest a non-specific environmental agent operating at a periodcritical for the development of all of them. Adisturbance in growth activity and in the productionof those chemical substances which are concernedwith organ differentiation will cause retardation anddisorganization in those organs which are at thatmoment developing most actively. Other parts ofthe body, which are also growing, but at a slowerrate, and whose demands are less exacting, may beless disturbed. When conditions improve these maycontinue their growth with little disturbance andestablish their normal connexions. A stage ofdevelopment once omitted cannot take place at alater period, but its omission may disturb develop-ment at a later stage or in another organ.Embryological data are much concerned with whendevelopmental processes start and finish, but neitherof these events may be of essential importance anddata on the critical periods of growth in differentorgans are most desirable. There are many diffi-culties in the detailed study of multiple congenitalanomalies, and embryology has contributed little toan understanding of the mechanism of theiroccurrence. The coincidence in different organsystems of what may be regarded as arrests ofdevelopment is often surprising in cases, such asthose presented, where there are multiple abnor-malities. Especially when the genetic contributionis revealed by the existence of a monozygotic andnormal twin, this seems to favour an environmentalcausation as opposed to a genetic basis.Though these two cases are presented to suggest

that environment sometimes determines congenital

malformation, it is not intended to suggest that theenvironment is the only cause of congenital malfor-mation. If this viewpoint should receive furthersupport efforts to improve the supply of the meta-bolic constituent or constituents limiting embryonicgrowth at the critical period, perhaps by increasingthe level in the maternal blood, are relevant andmight result in some reduction in the incidence ofcongenital malformation. A belief in the exclusiveoperation of an unalterable genetic inheritance mustbe avoided.

SmxmaryTwo sets of twins with monochorionic placentae

and placental vessel anastomoses are regarded asmonozygotic twins. One of each set showedmalformations affecting many different organs.Since the genetic inheritance of the malformed twinshould be similar to that of its normal partner, it issuggested that the malformations arose from anarrest of normal development when the placenta ofone twin was at some environmental disadvantagein obtaining nutrition from the utero-placental sitefor a short period early in intra-uterine life. Theabnormalities found were consistent with some suchnon-specific influence causing an arrest of develop-ment of those structures whose formation might bein a critical stage at about the sixth and fourth weekrespectively.

REFERENCESArey, L. B. (1922). Anat. Rec., 23, 253.Goldschmidt, R. B. (1938). 'Physiological Genetics.'

New York.Ingalls, T. H. (1947). Amer. J. Dis. Child., 73, 279.

and Gordon, J. E. (1947). Amer. J. med. Sci., 214,322.

Kallmann, F. J., and Reisner, D. (1943). Amer. Rei.Tuber., 47, 549.

Kindred, J. E. (1944). Amer. J. Obst. Gvnec., 48, 642.Ladd, W. E. (1944). NVew Eng. J. Med., 230, 625.Macklin, M. T. (1936). Amer. J. Obst. GVnec., 32, 258.Newman, H. H. (1942). 'Twins and Super Twins.'

London.Freeman, F. N., and Holzinger, K. J. (1937).'Twins: A study of Heredity and Environment.'Chicago.

Streeter, G. L. (1920). Contr. Embryol. Carneg. Inst.,9, 389.

Szendi, B. (1939). J. Obstet. Gynaec. Brit. Emp., 46, 836.von Verschuer, 0. (1939). Proc. roy. Soc. B., 128, 62.

Protected by copyright.

on Septem

ber 27, 2020 by guest.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.24.119.214 on 1 S

eptember 1949. D

ownloaded from