IMPERFORATE VAGINA IN THE

IMPERFORATE VAGINA IN THE NEWBORN*

NEONATAL HYDROCOLPOS

BY

WALLACE M. DENNISON and PAUL BACSICHFrom the Royal Hospital for Sick Children and the Department of Anatomy,

the University of Glasgow

The purpose of the present communication is todiscuss hydrocolpos and hydrometrocolpos, a notuncommon complication of imperforate vagina inthe newborn. We do not propose to discuss therelatively common problem of vaginal occlusiondue to labial adhesions (Fig. 1), which can beeasily separated with little or no bleeding.

Hydrocolpos in the newborn female due toimperforate hymen or to occlusion by a membranein the lower third of the vagina is not excessivelyrare. At or shortly after birth the infant presentswith lower abdominal swelling and a bulging mem-brane is seen at the vaginal orifice (Fig. 2). Thisbulging membrane looks like a cystocele and whengently reduced the urethral orifice can be seenanteriorly. Occasionally there may be a tinyopening in the occluding membrane through whicha white mucoid material can be extruded. Theanomaly has been fully described by Mahoneyand Chamberlain (1940), by Gross (1953), by Caffey(1956), and by many others and for the purposesof the present communication a brief descriptionof their findings will be sufficient.

In this condition the vagina is invariably greatlydistended-hydrocolpos-and on occasions the accu-mulated fluid can stretch the cervical canal and thebody of the uterus as well-hydrometrocolpos-though we have seen the uterine cavity distendedonly on one occasion. The ballooned genital tractarises from the pelvic floor into the abdominalcavity. The cyst is filled with clear or mucoid fluidwhich tends to become purulent from bacterialinvasion, giving rise to pyocolpos. The uterinetubes have not been involved in any of the reportedcases. The distended genital tract may causeurinary obstruction and the enlargement of thebladder may hinder palpation of the vaginal cyst.Megaloureter and hydronephrosis may follow andpyuria will be a further complication. Posteriorpressure may cause rectal obstruction. There may

* A paper read at a meeting of the British Association of PaediatricSurgeons held in London in July 1960.

be respiratory embarrassment through the upwarddisplacement of the abdominal contents. Ondigital rectal examination the pelvis is filled by amass projecting against the hollow of the sacrum.On inspection of the vagina a bulging membraneis seen at or above the hymen. In such patients aneedle is inserted through the membrane and fluidis aspirated to confirm the diagnosis. Radio-opaque fluid may be instilled to outline theobstructed vagina (Fig. 3). Having made thediagnosis the occluding membrane is slit with aknife and the cyst contents are evacuated.

Less commonly the hydrocolpos is due to atresiaof the lower vagina (White and Dennison, 1958) and

FIG. 1.-Occlusion of vagina due to labial adhesions.

156

copyright. on M

ay 30, 2022 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.36.186.156 on 1 April 1961. D

ownloaded from

http://adc.bmj.com/

IMPERFORATE VAGINA IN THE NEWBORN

FIG. 4.-Hydrocolpos due to membrane occlusion.

............ ...... :...... - I.-.

FIG. 2.-Hydrocolpos, showing bulging membrane at the vaginalorifice.

in our limited experience of five cases the diagnosiswas finally confirmed only at laparotomy. One ofus (W.M.D.) first encountered this condition some10 years ago and the account of his misdemeanoursmay serve as an awful warning to any juniorcolleague. A very ill baby, 1 day old, was admittedsuffering from peritonitis and intestinal obstruction.At laparotomy dense adhesions were found and alarge supravaginal (? urachal) cyst was demon-strated. This was freed, the ureters dissected offand the cyst excised in toto. When the specimenwas examined later, it became evident that theoperator had performed what is probably the earliestpanhysterectomy on record! The so-called urachalcyst consisted of a grossly distended vagina on topof which was a tiny uterus complete with tubesand ovaries. The patient did not survive surgicalintervention.

Figs. 4 and 5 show in a schematic fashion theessential anatomical differences between vaginalocclusion by a membrane, a simple well-knownproblem and the relatively undocumented hydro-colpos due to vaginal atresia. In spite of theanatomical differences the clinical features and the

FIG. 5.-Hydrocolpos associated with vaginal atresia.

157

FIG. 3.-Vaginogram in infant with hydrocolpos.

copyright. on M


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

ARCHIVES OF DISEASE IN CHILDHOODcomplications are similar when the hydrocolposis due to atresia of the vagina. During the pastfew years we have been particularly unlucky toencounter five infants with atresia of the lowervagina. In all these cases there was a solid blockof tissues (at least 2 cm.) separating the distendedvagina from the perineum and there was no bulgingmembrane present to aid diagnosis. In spite ofthis we have confidently diagnosed hydrocolpos dueto vaginal atresia in all these infants, four babiesand one girl aged 7 months. Unfortunately onlyone of our five patients is alive and well and she haslost both her vagina and her uterus.

In the newborn babies dense adhesions makeexploration of the abdomen difficult. The cystsvary from 10 to 15 cm. in diameter and in two of thefour babies the 'cyst' contents were grossly infected(pyocolpos). In only one case was the uterusdistended (hydrometrocolpos). On three occasionsthe cyst was evacuated through the abdomen andno attempt was made to drain the vagina via theperineum. In the uninfected case it was theoreticallypossible to evacuate the cyst through the abdomenand after dissecting down towards the lower poleof the cyst to pass a probe through the inferior poleof the cyst until it presented behind the urethra inthe region of the posterior fornix. A perinealincision over the probe allows drainage of thedistended vagina through an indwelling Malecotcatheter. It was difficult to visualize the next stage,as it was assumed that the resulting fibrosis followingrepeated dilatation would make a later Mclndoe andBanister (1938) operation almost impossible. Inanticipation it was thought that it might be possibleto maintain the passage by using a split-skin graftover a mould until the patient was old enough forconstruction of an artificial vagina. Having learnedfrom our bitter experience, Mr. John Grant, of ourStaff, recently dealt with a non-infected hydrocolposin an older infant. Before evacuating the cyst hecarefully dissected down into the pelvis only tofind that suddenly an intact cyst was deliveredinto the abdomen (Fig. 6). The only reasonablesolution to the problem then was to separate thetubes and the ovaries, to remove the vaginal cystand the uterus and to leave the ovaries behind.This infant made an uneventful recovery and remainswell.From the aforesaid it will be evident that we in

Glasgow, as yet, do not know how to deal success-fully with hydrocolpos or hydrometrocolpos due tovaginal atresia. On the other hand, to the best ofour knowledge, there is no evidence of any moreencouraging results from other centres.One could therefore legitimately ask: Why is the

anomaly so rare ? Haematocolpos at or shortlyafter puberty is well known. Due to an imperforatehymen the uterine discharge collects within thevagina and the girl presents with abdominal orpelvic pain. There is tenderness on suprapubicpressure and the ballooned vagina can be felt onrectal examination. This condition should besuspected in a girl with pubic hair and enlargementof the breasts who has not yet menstruated. Whydo these girls not present with hydrocolpos in theneonatal period? Obstruction of the vagina alonedoes not normally give rise to symptoms untilpuberty. One must therefore postulate that apartfrom the vaginal obstruction, which provides thenecessary anatomical conditions, the real cause liesin the over-abundant secretory activity of the neo-natal uterine or cervical glands.

FIG. 6.-Intact vaginal 'cyst' and uterus.

In order to find some support in favour of thispostulate we have been reinvestigating during thepast four years certain aspects of the developmentof the vagina and also some features of the physiologyof the neonatal reproductive organs. Our findingsand conclusions, together with the discussion of therelevant literature, will be presented elsewhere andfor the time being we shall confine ourselves to thefollowing brief statements.

Contrary to the most generally accepted view in

158

copyright. on M


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

IMPERFORATE VAGINA IN THE NEWBORN 159

the Anglo-American textbooks, according to whichonly the lower fifth of the vagina is formed by thefused 'sino-vaginal bulbs' growing out from thedorsal wall of the urogenital sinus, we believe withVilas (1932), Politzer (1955) and Bulmer (1957) thatthe boundary between the sinus and Mullerianderivatives of the human vagina lies either near tothe portio vaginalis or in the cervical canal. Wecan also confirm Politzer's (1955) observations onthe surprisingly great variability of the grade ofdevelopment of the vagina in foetuses of similarages and we believe that sometimes the neonatalvagina and uterus can be too immature histologicallyto be able to respond to hormonal stimuli in thenormal fashion.

It is accepted that in general the uterus of thenewborn shows a considerable degree of precociousmaturity and there is a marked hyperplasia of thelining epithelium and of the glands (Rosenthal andHellman, 1952). In fact these glands produce acopious mucus secretion, presumably, under theimpact of oestrogens and progesterone derived bothfrom the placenta (Brown, 1959) and from the foetalovaries (Ober and Bernstein, 1955). While there isa tacit assumption that this mucus secretion islargely the product of the glands of the body of theuterus, our own observations revealed a surprisinglygreater maturity of the glands of the cervical canaland we believe that these glands are the main, if notthe exclusive source of the mucus discharge. Withregard to the hormonal control of the secretion ofthe cervical glands of the uterus there is accumulatingevidence to indicate the cardinal role played byoestrogens (Moricard, 1936; Watson, 1939). Laterit is shown (Aberbanel, 1946; Zondek and Rozin,1954; Macdonald and Sharman, 1959; Stern, 1960)that administration of progesterone inhibits cervicalmucus secretion.Most authorities remark on the great individual

variations that exist in the urinary output ofoestrogens during pregnancy and we believe that thenon-appearance of hydrocolpos in some cases ofimperforate vagina of the newborn may be dueto the low output of oestrogens or in some instancesto the blocking effect of increased discharge ofprogesterone or progesterone-like substances fromthe placenta or more likely from the foetal adrenalcortex.

While the above two hypotheses could accountfor the dormancy in an odd case of imperforatevagina, it would stretch the imagination too far toexpect them to account for the large number ofcases observed in the adult by Bryan, Nigro andCounseller (1949), by Mclndoe (1959) and bySimmons (1959). Clearly we must look for some

other explanation. In this respect it seems to ushighly significant that in the majority of the adultcases of congenital absence of the vagina or ofvaginal atresia, palpation or laparotomy revealsovarian agenesis or some other malformation ofthe internal reproductive organs. Thus, while inthe cases occurring among newborn babies and smallinfants the abnormality is strictly and exclusivelyconfined to the vagina, a derivative of the urogenitalsinus, in the adult cases both urogenital sinus andMullerian structures are involved and create ananatomical situation with entirely different physio-logical behaviour and clinical manifestations. Con-sequently we believe that the two groups should betreated as separate clinical entities which can beusefully compared but should not be confused witheach other.Our main objective in putting forward these

somewhat provocative ideas is two-fold: partlywe seek advice from our learned audience concerningthe acceptability of these views and partly we hopethat by reaching some agreement we may gain acloser understanding of the aetiology and also obtaina necessary theoretical basis for the correct treatmentof neonatal hydrocolpos and hydrometrocolpos dueto vaginal atresia.

We thank Mr. J. R. Devlin for preparing the illus-trations and Dr. A. M. MacDonald, Dr. S. P. Rawsonand Mr. John Grant for their co-operation in ourinvestigations.

REFERENCES

Aberbanel, A. R. (1946). Artificial reproduction of cyclic changesin cervical mucus secretion throughout the cycle and correlationwith basal temperature. Trans. Amer. Soc. Study Steril., 46, 62.

Brown, J. B. (1959). Estrogen excretion of the pregnant woman.In Recent Progress in the Endocrinology of Reproduction,ed. C. W. Lloyd, p. 335. Academic Press, New York and London.

Bryan, A. L., Nigro, J. A. and Counseller, V. S. (1949). One hundredcases of congenital absence of the vagina. Surg. Gynec. Obstet.,88, 79.

Bulmer, D. (1957). The development of the human vagina. J. Anat.(Lond.), 91, 490.

Caffey, J. (1956). Pediatric X-ray Diagnosis, 3rd ed. The YearBook Publishers, Chicago.

Gross, R. E. (1953). The Surgery of Infancy and Childhood. W. B.Saunders, Philadelphia and London.

Macdonald, R. R. and Sharman, A. (1959). Cervical mucus in earlypregnancy. Int. J. Fertil., 4, 338.

McIndoe, A. (1959). Discussion on treatment of congenital absenceof vagina with emphasis on long-term results. Proc. roy. Soc.Med., 52, 952.and Banister, J. B. (1938). An operation for the cure of con-

genital absence of the vagina. J. Obstet. Gynaec. Brit. Emp.,45, 490.

Mahoney, P. J. and Chamberlain, J. W. (1940). Hydrometrocolpos ininfancy; congenital atresia of the vagina with abnormallyabundant cervical secretions. J. Pediat., 17, 772.

Moricard, R. (1936). D6veloppement du tractus g6nital et men-struation folliculinique par injection de benzoate de dihydro-folliculine. Bull. Soc. Obste't. Gynec. Paris, 25, 426.

Ober, W. B. and Bernstein, J. (1955). Observations on the endo-metrium and ovary in the newborn. Pediatrics, 16, 445.

Politzer, G. (1955). Zur normalen and abnormen Entwicklung dermenschlichen Scheide. Anat. Anz., 102, 271.

Rosenthal, A. H. and Hellman, L. M. (1952). The epithelial changesin the fetal cervix including the role of the 'reserve cell'. Anmer.J. Obstet. Gynec., 64, 260.

copyright. on M


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

160 ARCHIVES OF DISEASE IN CHILDHOODSimmons, C. A. (1959). Discussion on treatment of congenital

absence of vagina, with emphasis on long-term results. Proc.roy. Soc. Med., 52, 953.

Stem, A. (1960). Arborization of the cervical mucus: its assessmentby the fern test. J. int. Coll. Surg., 33, 163.

Vilas, E. (1932). Olber die Entwicklung der menschlichen Scheide.Z. Anat. Entwickl. Gesch., 98, 263.

Watson, M. C. (1939). The effect of cervical secretions on thevitality of spermatozoa. Canad. med. Ass. J., 40, 542.

White, M. and Dennison, W. M. (1958). Surgery in Infancy andChildhood. Livingstone, Edinburgh and London.

Zondek, B. and Rozin, S. (1954). Cervical mucus arborization:its use in the determination of corpus luteum function. Obstel.and Gynec., 3, 463.

copyright. on M


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

Documents

IMPERFORATE VAGINA IN THE