canmedaj00797-0002

The (rnrn Ctrnt lIl Itle

t

JANUARY 15. 1959 * VOL. 80 NO. 2

POLYHYDRAMNIOS ANDOLIGOHYDRAMNIOS*

T. N. A. JEFFCOATE,t M.D., F.R.C.S.E.,F.R.C.O.G. and J. S. SCOTT,t M.B., Ch.B.,

M.R.C.O.G., Liverpool, Eng.

THEORIES CONCERNING the origin of liquor amniiwere well reviewed by Shaw and Marriott36 in1949 and do not require repetition. Suffice it tosay that most recent opinion has been coloured byobservati;ons on the transfer of radioactive isotopesto and from the amniotic cavity, and especially bythe studies reported by Vosburgh et al.41 Theseworkers concluded that the water content ofliquor amnii is completely exchanged every threehours. This conclusion, however, does n.ot tallywvith the clinical behaviour of women who sustaina rupture of the membranes during pregnancy andwho do not immediately go into labour. Suchwomen never lose fluid at the rate of one to twopints each three hours, as might be expected ifliquor is formed so quickly.

It is not always realized that Vosburgh's experi-ments on the rate of water exchange were limitedto five women whose pregnancies varied in dura-tion from 14 to 40 weeks. The results differedvidely from case to case, without any correlationaccording to the period of pregnancy, and thefigure of 34.5% renewal of fluid per hour wasobtained by taking an average. The sodium ex-change was studied in 20 pregnancies, the durationof which varied from 10 to 40 weeks; the findingswere again reduced to an average to suggest acomplete turnover once in 14.5 hours, a rate of re-newal five times less than that of the water.

This work received a measure of confirmation byPlentl and Hutchinson,'9 but the evidence from thistype of experiment is as yet so limited that theabove figures should not be allowed to prejudiceother observations. It would seem especially im-portant that all fututre investigations in this field,whether they be by following the behaviour oftracer substances or by physico-chemical analysis,should be carried out for each phase of pregnancy

*An elaboration of a, paper piresented by T. N. A. Jeffcoateto the Section of Obstetrics and Gynsecology of the OntarioMedical Association at Toronto, May 15, 1958.tFrom the Department of Obstetrics and Gynecology, Univer-sity of Liverpool.

77

separately. This is because there is much to suggestthat the rates of formation and removal, as well asthe origin and destination, of liquor amnii changewith the progress of pregnancy. For example, theliquor is isotonic with plasma during the earlvmonths but becomes increasingly hypotonic in thelater stages of pregnancy.24 Again, its urea, uricacid and creatinine contents are similar to thoseof plasma up to the fifth month; thereafter they aregreater."For the present, the most important contribution

to knowledge made by radioactive isotope studiesis the clear demonstration that the pool of liquoris not a static reservoir. Both the fluid and itschemical contents are in a state of continual flux,passing to and from their surroundings. The volumeof liquor, however, remains fairly constant for eachparticular stage of pregnancy. The normal volume,and pathological variations constituting poly-hydramnios and oligohydramnios, are likely to bedetermined by the relative efficiency of themechanisms for production and disposal.

ORIGIN OF LIQuoR AMNIIFrom or Through the MembranesLiquor appears early in the pregnancy, when the

fetal tissues are underdeveloped and non-func-tional. Moreover, it is found within the sac ofblighted ova in which the fetus is absent or rudi-mentary. Such observations strongly suggest thatfluid enters via the membranes. Moreover, in earlypregnancy, liquor is so similar to plasma in com-position that it could well represent a dialysate.Amniotic epithelium is so constituted as to suggestthat it could have a secretory function; indeed,such a function has been demonstrated histo-logically.40 It is therefore possible that the amnionplays an active part in the formation of liquor.If liquor is regarded as a transudate in early preg-nancy and as the product of a selective secretoryaction of the amnion in the later months, the pro-gressive changes in its chemical and physicalproperties might be explained. It is, however, dif-ficult to conceive of the amnion producing or trans-mitting fluid in large amounts, because most of itis not in direct contact with a rich blood supply.The exception is that portion of the membranewhich covers the placenta, and it has been stig-

78 JEFFCOATE AND SCOTr: POLYHYDRAMNIOS AND OLIGOHYDRAMNIOS

gested that this specialized area, with its extensivevascular connections, plays some part in liquorproduction. The common association of poly-hydramnios with hemangiomatous tumours of thechorion is quoted in support. This finding is capableof different interpretations. Thus, McInroy andKelsey23 concluded that, even in these cases, it isthe fetus which is responsible for the hydramnios.They argued that the placental tumour occupies somuch of the intervillous space that it leads toplacental insufficiency with retention of metaboliteswithin the fetus. These in turn stimulate the fetalkidneys to excrete abnormally large quantities offluid into the amniotic cavity (see later).

From the FetusThe parts of the fetus which might secrete fluid

into the amniotic cavity are the skin and its ap-pendages, the respiratory tree and the kidneys.The main argument against a purely fetal originis that liquor is present early in pregnancy beforemost fetal organs are formed. This objection isvalid except in relation to fetal ectoderm which ispresent at a very early stage, which has the sameorigin as amnion, and which might conceivablyhave a similar function. Moreover, the volume ofliquor in early pregnancy varies directly with thesurface area of the fetus.15Once fetal organs have developed, the production

of liquor by the respiratory tract and kidneys, ifnot by the skin, becomes feasible. The lungs aregenerally believed to be collapsed and in receiptof a relatively poor blood supply during the ante-natal period. Alveolar secretion of fluid is there-fore presumably small, and a contribution from thetrachea and bronchi, although possible, is unlikelyto be a major one. Nevertheless, Whitehead andothers,"4 Macafee25 and Morison26 consider that, inpart at least, the liquor amnii is derived from therespiratory tract of the fetus.The fetal kidneys have attracted attention ever

since Hippocrates postulated that liquor amnii isfetal urine. That the kidneys can function beforebirth is made clear by the fact that a baby oftenvoids urine immediately after birth, and by thefinding of fluid in the bladders of approximatelyone-third of stillborn babies which come toautopsy.20 Nevertheless, Holtermann'8 concludedthat fetal urine does not contribute to the liquoramnii. He administered methylene blue to themother and recovered colourless liquor at delivery,yet found the fetal urine stained with dye in theneonatal period. The test agent, however, wasgiven only a short time before delivery and it nowseems probable that the fetus did not have suffi-cient time to void dye-stained urine before itsbirth.Because of such experiments there arose, in the

earlier part of this century, the concept that thekidneys do not normally secrete while the fetus is

Canad. M. A. J.Jan. 15, 1959, vol. 80

function is deficient-as it sometimes is duringpregnancy and as it often is during labour. Thisidea, reviewed by one of us (T.N.A.J.) in 1932,is supported by little more than conjecture. Thenewly born baby does not have to withstand' labourbefore voiding urine; it does so even when de-livered by elective Caesarean section. When thefetal urethra is obstructed, the whole urinary tractbecomes distended with quite large volumes offluid before labour, and that without there beingany evidence of placental failure. Nevertheless,even if the fetal kidneys normally secrete beforebirth, it still remains possible that they pour outunusually large amounts of fluid when the func-tion of the placenta23 or of the maternal kidneys45is grossly defective.

Urine is reported to have been found in thefetal bladder as early as the 14th week of preg-nancy,24 and it is now generally accepted that thekidneys are invariably capable of some functionby the 20th week. This timing may be of consider-able significance in explaining the reported pro-gressive change in the chemical composition ofliquor which commences about mid-pregnancy.Summation of the evidence suggests that, up to thetime the fetal organs are developed, liquor amniiis derived almost entirely from the amnion, or fromfetal skin and amnion. After mid-pregnancy, anadditional and perhaps even more importantsource of supply is the fetal kidneys, with possiblya contribution from the fetal respiratory tract. Ithas to be recognized, however, that the composi-tion of the excretion of the fetal kidneys differsfrom that of urine excreted after birth. The renalfiltration and concentration functions vary withthe period of intrauterine development"' and alsochange progressively during infancy. Moreover,certain components of the fetal renal excretion maybe reabsorbed by the amnion. Such factors needto be considered when the results of chemicalanalysis of liquor amnii are interpreted.The importance of the renal contribution to the

liquor during late pregnancy is further borne outby study of cases of oligohydramniQs (see later).

Tim DISPOSAL OF LIQUOR AMNIIThrough the MembranesThe amnion might absorb as well as secrete

fluid or, depending on osmotic pressures, at leastallow a transudation back into the maternal circula-tion. Whether production is a more importantamniotic function than is disposal, and whethertheir relative importance varies with the stage ofpregnancy, are unknown. Nevertheless, the possi-bility of the amnion playing some part in the returnof the various components of liquor to the maternalcirculation must be conceded. It is, for example,the only obvious route to account for the dis-appearance of fluid from the pregnancy sac afterfetal death in utero; this at least shows that the

in utero and that they only do so when placental membranous lining is not waterproof.

JEFFCOATE AND Scolr POLYHYDRAMNIOS AND OLIGOHYDRAMNIOS

Absorption by the FetusFetal skin, like amnion, might absorb water, but

this is more likely when it is represented by primi-tive ectoderm than when it is a well-developedstratified epithelium covered with vernix caseosa.The fetal respiratory tract might also form adisposal route, at least after mid-pregnancy. Thepathway which seems most likely to offer an exitfor substantial amounts of fluid is, however, thefetal alimentary canal. The finding of epithelialdebris and lanugo hairs in the fetal stomach sug-gested to obstetricians of the past that the fetusin utero does swallow. Dyes introduced into theamniotic sac were also found to stain the fetalalimentary tract. Later, at a time when amni-ography was considered a reasonable diagnosticprocedure, radiographs often showed incidentallythat the fetus had imbibed radio-opaque materialwhich, having been introduced into the amnioticsac, could be followed through the stomach and theintestines (Fig. 1). Clinical studies of cases ofpolyhydramnios (see later) go to show that thisdisposal route is an important one, if not in termsof the total quantity of fluid absorbed, certainly inthe control of the volume of liquor contained bythe amnion. Of the liquor amnii swallowed, thefluid component presumably enters the fetal circula-tion, with or without subsequent passage acrossthe placenta to the maternal circulation.

Rosa,3' using inulin as a tracer substance, calcu-lated that the fetus swallows 500 ml. of liquoramnii every 24 hours. Of this 40 ml. is excretedback into the amniotic cavity by the fetal kidneys,*and 435 ml. passes into the maternal circulation viathe placenta. These conclusions, interesting thoughthey be, cannot yet be accepted as final becausethe experimental methods employed depended ona number of assumptions. Gray, Neslen and Plentl,s3working with an isotope tracer technique andcomparing their findings in the human being withthose obtained on hydrodynamic models, calculatedthat at least 25%, and probably more than 50%,of the water transfer from amniotic fluid to motheris accomplished through the intermedium of thefetus. Plentl and Grayn8 later gave figures for thevolumes of fluid passing each way between mother,fetus and liquor, and concluded that fluid normallycirculates from mother to fetus to liquor to fetusand back to mother. In hydramnios they foundthat this. circulation, although still present and inthe same direction, is reduced. In this conditionthey found an additional direct transfer of waterfrom mother to liquor without passing through thefetus; this, they said, does not occur normally.

These conclusions must again be accepted withreserve because some of Plentl and Gray's argu-ments were based on statements about clinical

*This work, if confirmed, does not mean that the total out-put of the fetal kidneys is only 40 ml. in day. An additional,and probably much greater, amount of urine is likely tobe derived from fluid entering the fetal circulation throughthe placenta.

79

Fig. 1.-An amniograph showing that the fetus hasswallowed radio-opaque material which had been introducedInto the amniotic cavity. (Reproduced by permission of theEditor of The Laincet.)

associations of hydramnios for which the evidenceis slender. Thus they stated that placental infarc-tion and "especially the circumvallate placenta aresignificant and consistent findings" in cases ofhydramnios. In a series of 578 placentae of circum-vallate type studied by one of us (J.S.S.) poly-hydramnios had been an antenatal association ofonly 3 (0.5%). This compares with an 0.8% over-all incidence of hydramnios amongst all patientsdelivered in the same hospital during the periodof the survey. Plentl and Gray28 also assumed thatmongolism, achondroplasia and cardiac malforma-tions are significant causes of hydramnios. We donot know of any evidence to substantiate this, andour own experience is to the contrary.Although details remain doubtful there emerges,

from the foregoing considerations, a working hypo-thesis. This envisages that liquor amnii is in a stateof coming and going throughout pregnancy; itsrate of exchange may be rapid but it almostcertainly varies with different stages of pregnancy.During its continual circulation, certain constitu-ents of the liquor are added and subtracted as inthe case of the circulation of blood. The origin ofliquor amnii is at least twofold: (1) from theamnion and possibly the fetal ectoderm, (2) fromthe fetal kidneys and possibly the fetal respiratorytract. There are at least two pathways for itsdisposal: (1) by the amnion, (2) by the fetal ali-mentary tract, and possibly by the fetal lungs or

skin. The amount of liquor present at any one timeis determined by a balance between flow and ebb.If the supply is relatively greater than the disposalmechanism can handle, polyhydramnios results.

Canad. M. A. J.Jan. 15, 1959, vol. 80

80 JEFFCOATE AND SCOTT: POLYHYDRAMNIOS AND OLIGOHYDRAMNIOS

When disposal outruns supply, oligohydramniosoccurs.

This concept is strongly supported by clinicalobservation of cases in which the amount ofamniotic fluid is increased or reduced. Gross devia-tions from the normal volume of 500-1500 ml. ofliquor in the later weeks of pregnancy are notcommon. In many cases where some abnormalityin this respect is noted, an obvious cause is notfound. Nevertheless, in a considerable proportionsome factor is present which appears to be re-sponsible for the unusual amount of liquor.

POLYHYDRAMNIOSA review of 169 consecutive cases of hydramnios

encountered in the practice of one hospital withwhich we are associated revealed that in 54 (32%)the excessive amount of liquor was associated witha malformation which would probably interferewith the fetus's swallowing or absorbing liquoramnitifom the intestine.3" These conditions wereas follows:

Anencephaly .................................

Iniencephaly ..................................Hydrops fztalis with gloss (deema of the lips and

fauces ...................................(Esophageal atresia .........................Duodenal atiesia.............................Diaphragznatic hernia

32 cases2 "4 "123 "1 case

1. Anencephaly and iniencephaly.-These mal-formations could interfere with swallowing, eitherbecause the deglutition centre is not developed inthe brain or because of the local deformity affect-ing the mouth and neck. In order to establish thatfailure of the fetus to swallow is the mechanismwhereby these lesions cause hydramnios, amni-ography was carried out in four cases in whichanencephaly was diagnosed in pregnancy. It was

expected that, unlike the normal baby, the an-

encephalic fetus would not show any radio-opaquematerial in its alimentary tract. This happened inthree of the cases but the fourth fetus did swallow.This exceptional case only goes to prove the rule,because delivery showed that the antenatal radio-graphic diagnosis of anencephaly was incorrect.The vault of the fetal skull, together with normalcerebral tissue, was present and the actual de-formity was a severe degree of occipital encephalo-ccele. This case, incidentally, was the only one ofthe four which did not show clinical evidence ofhydramnios. Anencephaly is sometimes seen with-out polyhydramnios, but the deformity is thenusually incomplete and of a degree which mightstill allow fetal swallowing.

2. Hydrops fetalis.-This interesting condition ofthe fetus produces hydramnios only occasionally.When it does, the fetus usually suffers from grosscedema of the lips, mouth and fauces; such swellingcould prevent swallowing.

3. CEsophageal and duodenal atresia.-The oc-

currence of polyhydramnios in cases in which there

Canad. M. A. J.Jan. 15, 1959, vol SO)

is an obstruction of the upper part of the fetalalimentary tract provides the strongest evidencethat swallowing and intestinal absorption play animportant role in the disposal of liquor amnii. Dur-ing the period under review there were in all 13cases of cesophageal atresia but polyhydramnioswas present in only 12. The one exception againgoes to establish the importance of swallowing illthe removal of liquor amnii. In this case the con-genital defect was not of the-common variety butone in which the upper and lower parts of thecesophagus had an indirect communication via thetrachea.33 This made it possible for liquor to reachthe intestine to be absorbed.

It seems probable that, in the one case ofdiaphragmatic hernia, the swallowed fluid wasprevented from reaching the lower absorptiveregions of the intestine by obstruction due tokinking.Polyhydramnios associated with fetal ingestion

defects is rarely clinically manifest until after the20th week. This may mean that fetal swallowingdoes not become the main method for disposal ofliquor until relatively late in pregnancy. Alterna-tively, it might signify that, after mid-pregnancy,there is an additional amount of liquor to be re-moved and that this is more than can be clearedby routes other than the fetal alimentary tract.

It is not to be overlooked that in two-thirds ofour cases of polyhydramnios there was no evidencethat the fetus was unable to dispose of liquor. Inthese a definite causal mechanism was not found,but it can be postulated that either another outletfor liquor was not operating or, even more likely,that fluid was being produced in excessive amounts.Polyhydramnios could in theory be caused by anoveractive secretory function of the amnion, apossibility which Plentl and Gray's28 work mightbe construed to support. It is tempting to questionwhether the finding of an excessive amouint of fluidlin the pregnancy sac in association with maternaldiabetes is related to the reported finding of cortisolin the liquor in such cases.17

Polyhydramnios might even be the result offetal polyuria, a possible example of this being theexcessive amount of liquor sometimes found in onesac of uniovular twins. In such case it is sug-gested9 23 that there is a placental communicatioinbetween the circulations of the two fetuses and thatthe heart of one over-rides that of the other. Theresult is unequal renal circulations and functions,the dominant fetus secreting urine for both andthus surrouncding itself with more liquor thainnormal. This largely theoretical concept may besupported by the following recently observed case.

When a primigravid womani w,vas nioted to be suffer-ing from hydramnios, rcadiography revealed a twillpregnaincy, one fetuLs showing evidence of gross mal-formation. Because of incoordinate uiterine action inlabour, delivery was effected byCI.%sarean section, andthis allowed stuidy of the exact state of affairs. The

JEFFCOATE AND SCOrr: POLYHYDRAMNIOS AND OLIGOHYDRAMNIOS 81;

amnion of the normal fetus contained more than a

measured 5 pints (3 litres) of liquor; the other sac

was almost dry. The fetus without liquor was an

acardiac acephalic monster, with only a small amountof rudimentary renal cortex on one side and that notserved by a patent ureter. The twins were uniovularand there was an obvious anastomosis between the twoplacental circulations. Here would at first appear tobe a clear example of polyhydramnios associated withcardiac dominance. A more acceptable explanation,however, is the absence of functional renal tissuefrom one fetus. This could account for its owni oligo-hydramnios aind also for induced polyuria in its brother.

That cardiac dominance alone is not alwaysan adequate explanation of polyhydramnios in one

sac of uniovular twins is suggested by a secondcase.*

Because of the large size of the uterus at the 20thweek of pregnancy, a primigravida was radiographedand found to have twins. By the 27th week so muchliquor was present that the abdominal girth measured47 inches (118 cm.). Spontaneous labour two weekslater resulted in the delivery of a stillborn hydropicfetus (weight 2350 g.) from a sac which containeda massive but unmeasured amount of liquor amnii,and of a normal fetus (weight 790 g.) from a sac con-

taining relatively little fluid. The twins were uniovular,and the placenta was single (weight 910 g.) with a

definite although not prominent vascular communica-tion between its two halves. The exclusion of iso-immunization to any of the blood factors made itnecessary to look for another cause of the hydropsfetalis. The most likely one is that the cedema was theresult of, relative cardiac inefficiency. A mechanicalcirculatory basis for hydrops fetalis is an old conceptwhich has tended to be overlooked as a result of moderninterest in rhesus incompatibility. It is supported bythe fact that acardiac monsters (including the one

mentioned above) nearly always show generalizedcedema of the tissues. If this is the explanation of theextraordinary finding of hydrops fetalis affecting onlyone of uniovular twins, then polyhydramnios did niotaccompany cardiac dominance; it was found in thesac-of the fetus with presumed cardiac inefficiency. Inthis case, as in the first, that is another possible ex-

planation of the polyhydramnios, namely, hydropsfetalis alone. This condition, however, rarely operatesso early in pregnancy.

OLIGOHYDRAMNIOSOligohydramnios is recorded far less frequently

than is polyhydramnios, possibly because it tendsto pass unnoticed.

In the first half of pregnancy oligohydramniosis not seen as a recognizable clinical -entity. If itdoes occur, it is presumably the result of a faultin the amnion, or in the chemical and othermechanisms which control the exchanges, to andfrom the maternal tissue fluids through the mem-

branes. In fact, its occurrence in early pregnancy

would almost inevitably cause abortion becausethe accumulation of fluid is essential to stimulate

*The details of this case are described by Hibbard.16

the uterus to grow to accommodate the youngfetus.15In late pregnancy the finding of anhydramnios.or

a severe degree of oligohydramnios often indicatesthat the fetus in utero is suffering from a gross.renal abnormality or from an obstructive lesion inthe lower urinary tract. In other words, the fetusis unable to void urine into the amniotic sac.

Bardram5 reported concomitant oligohydramnios in10 out of 13 cases of gross defects in the fetalurinary system, and Bates7 noted a deficiency ofliquor in 13 out of 14 cases of renal agenesis anddysplasia. Shaw and Marriott,36 as well as others,also reported isolated examples. Potter30 reviewed30 cases in which the fetus suffered some conditionrendering it unable to void urine, and said that innone had the obstetrician seen liquor escape duringlabour. Selby and Parmelee35 recorded a case inwhich a baby without kidneys was born with theamnion intact. There was definitely no liquor in thesac. An equally convincing account of anhy-dramnios discovered at elective Caesarean sectionwas given by Bates.7One of us (J.S.S.), together with A. D. Bain,3

recently collected 50 consecutive cases of renalagenesis and dysplasia occurring in the maternitydepartments of certain hospitals in Liverpool andEdinburgh. In going into the clinical histories ofthe mothers of these babies retrospectively, it was

found that liquor amnii had been definitely seen inonly one case. In 14 the records contained no com-

ment about the -presence or absence of liquor; in28 there were various -entries which stronglysuggested that the liquor had been deficient; inthe remaining seven there was a specific note stat-ing that liquor amnii was negligible in amount or

absent. We now have four further cases to addto this reported series; in none was a significantamount of liquor amnii present. Davidson and

Ross,10 writing in 1954, collected 232 cases ofrenal agenesis and severe dysplasia from theliterature. The addition of cases recorded byBaron,6 Osmond,27 Bound and others,8 Selby andParmelee,35 Welch,43 and our own 54 cases, givesa total of 295 recorded cases. Among these therewas firm or presumptive evidence of oligo-hydramnios or anhydramnios in 100, while in thevast majority of the others there was no statementas to the amount of liquor amnii. Amongst thevery few cases in which liquor was present innormal or excessive amounts are those reportedby Wagner,42 Balard,4 Schiller and Toll,32 Gowar,12Allen and Orchard' and Silvester and Hughes.37There is, in addition, the single case in our seriesof renal agenesis in which liquor was definitelypresent: indeed this patient suffered from poly-hydramnios.

Consideration of this last exception offers a clueto the explanation of the other isolated cases in

which renal non-function in the fetus was notassociated with oligohydramnios. In our case thefetus not only had renal agenesis but also had

Canad. M. A. J.Jan. 15, 1959, vol. 80

82 JEFFCOATE AND SCOrr: POLYHYDRAMNIOS AND OLIGOHYDRAMNIOS

iniencephaly. This anomaly, as previously notedshares with anencephaly a predisposition to causehydramnios, apparently because it interferes withfetal swallowing. It is therefore presumed that, inthis case, the amount of amniotic fluid of non-renal origin exceeded that which could be takenaway by routes other than the fetal alimentarytract.

This example serves to illustrate well that liquoramnii has multiple sources of origin and multipleroutes of disposal. In the other recorded cases ofrenal agenesis in which liquor volume was normalor excessive it is almost certain that additionalunrecognized factors either interfered with someroute of disposal or caused an over-production offluid from a non-renal source.

Babies without functional kidneys are born deador die within a few hours of birth. Their bloodurea is not raised, and they do not succumb fromrenal failure. Their death is usually an asphyxialone and is the result of pulmonary hypoplasiawhich invariably accompanies renal agenesis. It is,therefore, necessary to consider the possibility thatthe oligohydramnios found in these cases is theresult of a failure of the lungs to contribute fluidto liquor amnii. This is unlikely because whenthe fetal kidneys and lungs are normal, but theurethra is obstructed, oligohydramnios still occurs.Indeed, pulmonary hypoplasia coincident withrenal agenesis may well be the result rather thanthe cause of oligohydramnios.3 Be this as it may,the evidence strongly suggests that the urinarytract is more important than the respiratory systemin the normal production of liquor.Although Schiller and Toll32 noted a gross ab-

normality in the fetal renal tract in only 15 outof 57 reported cases of oligohydramnios, Potter30considered that a deficiency of liquor is rareexcept in association with bilateral renal agenesisor obstruction of the urethra. Excluding cases ofpostmaturity and of retention of a dead fetusin utero, in which the amount of liquor doesbecome reduced, we did, until recently, subscribeto Potter's view. During the past 18 months,however, among approximately 5000 deliveries inone of the hospitals in which we work, therewere six proved cases of oligohydramnios. In threeof these the fetus suffered from renal agenesis butin the remaining three the fetal urinary tract wasnormal. In each of the latter group there was ahistory of threatened abortion in the early monthsof pregnancy. Moreover, at delivery, large areasof the membranes were found to be covered ontheir maternal aspect with partially organized bloodclot. These lesions were such that they could havehindered the transfer of fluid through the mem-branes, so it seems likely that, in these particularcases, the cause of the oligohydramnios was afailure of the amnion to contribute its quota tothe total volume of liquor.

Canad. M. A. J.Jan. 15, 1959, vol. 80

Fig. 2.-Radiograph in the case of oligohydramnios showingcompaction of the fetus, extreme flexion of the spine andcrowding of the skeleton towards the pelvis. The durationof the pregnancy was 34 weeks. The breech presentation isalso a feature of oligohydramnios.

The low incidence of oligohydramnios in thislast series again emphasizes the apparent rarityof the condition. Additional cases may, however,have passed unrecognized because the diagnosisis so difficult. It generally rests on noting duringabdominal palpation a deficiency of fluid aroundthe baby, and on observing that liquor does notescape vaginally before or during labour. Radio-graphs of the uterus and its contents are said toshow typical signs of oligohydramnios.21 Theseresult from compaction of the fetus and consist ofextreme flexion of the spine and crowding of theskeleton towards the pelvic cavity (Fig. 2). Theappearance is similar to that described by Tager(1954) for fetal death in utero, except thatSpalding's sign is absent. Radiographs were ob-tained in seven recent cases of oligohydramnios.Extreme flexion of the spine was a feature in four;in the remaining three this sign was absent but thelimbs appeared cramped and the soft tissue outlineof the uterus was closely applied to the fetus. Theradiological signs of oligohydramnios are ofinterest but do not contribute much to the practicalrecognition of the condition, since some abnormalityhas to be noted clinically before radiographicexamination is justified. They are therefore rarelymore than confirmatory.As a more certain means of diagnosing oligo-

hydramnios, we draw attention to changes in theamnion which are almost specific and which, ifsearched for diligently during routine examinationof the secundines after every delivery, shouldensure that paucity of liquor is rarely overlooked.The diagnostic lesion is "amnion nodosum".22, 34

Canad. M. A. J.Jan. 15, 1959, vol. 80 JEFFCOATE AND SCOTrr POLYHYDRAMNIOS AND OLIGOHYDRAMNIOS 83

transition from cuboidal cells in, the deep layerto flattened and keratinized cells on the surfaceof a stratified epithelium. In amnion nodosumthere is merely a disorderly mass of keratinizedsquames loosely embedded on the amnion. Theirattachment is so unstable that it is almost certainthat they are of fetal origin, and that they representdeposits of squames shed from the fetal skin.We have so far seen 14 cases of amnion nodosum,

all collected within two years; nine of these werereported by Scott and Bain.34 The accompanyingconditions were as follows:

Oligohydramnios and renal agenesis.............Oligohydramnios with a normal fetal urinary tract.Intrauterine death of the fetus followed by its

retention in utero for three weeks............

10 cases3 "

1 case

Fig. 3.-Amnion nodosum. A piece of amnion has beendetached from the chorion and shows plaques and noduleson its fetal aspect. These appear quite dark because, in thisparticular case, they had a reddish colour; usually they are

paler.

AMNION NODOSUMThis condition takes the form of small nodules

or plaques on the fetal aspect of the amnion(Fig. 3). They are usually a dull grey-yellow incolour, but in two of our cases they were brick-red. The nodules can be found anywhere on thefetal surface of the amnion but are generally mostnumerous and obvious on the part of the mem-

brane covering the placenta. They move with theamnion when it slides on the chorion but can bepicked off the underlying membrane without dis-rupting its continuity.

Histologically, the nodules consist of masses -ofkeratinized squames embedded in an acidophilmatrix (Fig. 4). The cuboidal cells of the amnioticepithelium are generally absent in the region ofthe nodules but there is no sign of squamous meta-

plasia. In the latter condition there is a clear

Fig. 4.-Amnion nodosum. Section of a nodule showingthat it consists of degenerated squames clustered on thesurface of the amnion. (Photomicrograph X 45.)

The early cases of amnion nodosum were allfound in association with renal agenesis and itwas at first considered possible that the lesion wasan associated developmental error, possibly ofgenetic origin. However, among 11 cases of renalagenesis in which the membranes were studied,there was one in which the lesion was not present.This was the one, mentioned earlier, in which con-comitant iniencephaly caused polyhydramnios. Inother words, only those cases showing oligohy-dramnios had nodules on the amnion. The con-clusion that the lesion is associated with a shortageof liquor, and not with the renal defect directly,was confirmed by the later finding of three casesof amnion nodosum when the fetal urinary tractwas normal. In these, as mentioned earlier, therewas an extensive deposit of fibrin between thechorionic membrane and the decidua vera. Finally,there was one case in which a shortage of liquorresulted from prolonged retention of a dead fetusin utero.

In all pregnancies it is normal for superficialsquames to be cast from the fetal skin into theliquor amnii. These remain floating freely. When,however, the amount of liquor is deficient, itwould appear that the amniotic epithelium comesinto continuous close apposition with the prom-inences of the fetal ovoid. This results in unusualfriction which leads to excessive desquamationfrom the fetal skin. The squames, having littleif any fluid in which to circulate, become depositedon the amnion. The amniotic cuboidal cells at thesite of contact become destroyed, either by directfriction with the fetus or by the influence of thedeposit of squames.

LIQUOR VOLUME AND FETAL WEIGHT

Study of the weights of babies born with renalagenesis or severe dysplasia raises another matterof some importance. Among the 50 cases recordedby Scott and Bain34 it is known with reasonablecertainty that oligohydramnios was present in 35and it can be assumed that it was a feature ofthe majority of the others. Fig. 5 shows the weights

84 JEFFCOATE AND SCOrr: POLYHYDRAMNIOS AND OLIGOHYDRAMNIOS

RENAL DYSPLASIA

Birth weight in relation to maturity

7.

6

5

oD 4 -

KX

42Ix,

qz

S Normal average birthweight

Average birth weightof renal agenesis babies

5 9 17 19Cases Cases Cases Cases

26 - 29 30 - 33 34 -37 38 and over

MATURITY IN WEEKS

Fig. 5.-Average birth weights of 50 babies born withrenal dysplasia, grouped according to maturity, comparedwith the average birth weights of normal babies of thesame maturity.

of these abnormal babies, grouped according tomaturity, compared with the average weights ofnormal babies of similar maturity. From this it isclear that whereas growth continued as expectedup to the 34th week of pregnancy, the affectedbabies thereafter failed to gain weight. Indeed theheaviest birth weight recorded was 5 lb. 1 oz.(2300 g.). The findings are too consistent for themto be dismissed as a chance happening. Nor canthey be explained simply on the basis of absenceof certain organs, because: (1) the weight of thekidneys is trivial compared with the overall bodyweight; (2) 22 babies suffered cystic renal dys-plasia and their kidneys were heavier than normal;(3) it is only in late pregnancy that the fetalweight is less than expected.

It has previously been assumed, even by thosewho accept that urine is normally secreted beforebirth, that this function is not essential to thebaby's health. Our findings, however, suggest that,without it, there is retardation of growth. For thisthere are three possible explanations.

1. That the fetal kidneys normally remove fromthe fetal blood stream some noxious metabolite andexcrete it into the liquor amnii where it exerts noill effect. This concept is contrary to the commonview that the placenta exerts a comprehensiveexcretory function for the fetus. Moreover, theurea and electrolyte levels in the blood of babieswithout kidneys are normal at birth.10' 2

2. That the fetus derives some of the nutiimentnecessary for its growth by imbibing liquor amnii,

Canad. M. A. J.Jan. 15, 1959, vol. 80

and that deprival of this impairs its weight gain.It is unlikely that the fetus thrives on what it hasalready excreted, which is the logical implicationof this idea. Even if it is postulated that the lowbirth weight of the fetus is merely the resultof its failing to drink sufficient water, it has tobe recognized that, except for some dryness ofthe skin, these babies do not show signs ofdehydration at birth.

3. That normal growth of the fetus depends onits fluid environment. An adequate amount ofliquor amnii could exert its effect by-allowing thefetus freedom of movement, by stimulating growthof the uterus, and possibly by encouraging a goodplacental circulation. This view, which appears themost reasonable, is in accord with the findings ofHarrison and Malpas.15 They studied liquorvolumes and embryo sizes during the early monthsof pregnancy and concluded that the functionof the liquor, even at that time, was so to distendand stimulate the uterus as to allow the properdevelopment of the fetus. They showed, too, adefinite correlation between liquor volume andfetal size.

If it be accepted that the fetus which fails tourinate becomes stunted because of the associatedoligohydramnios, it is relevant to emphasize thatits growth is slowed only from the 34th week ofpregnancy onwards. This again suggests that therenal contribution to liquor amnii is relatively

9

87

7.

Q.

a.I

HYDRAMN IOS

Birth weight in relation to maturity

Normal average birth

weight.

Average birth weight of

babies from hydramniotic

SsacS.17

Cases Cases Cases Cases26-29 30 - 33 34 -37 38 and over

MATURITY IN WEEKS

Fig. 6.-Average birth weights of babies born from poly-hydramnniotic sacs, grouped according to maturity, comparedwith the average birth weight for normal babies of thesame maturity. Hydropic fetuses and those born to diabeticmothers were excluded because their excessive weight couldhave been caused by factors other than polyhydramnios.Anencephalic fetuses were also excluded because it wasimpossible to allow for their structural deflciency. The figuresfor both columns apply to single pregnancies only.

JEPFCOATE AND SCOTrT: POLYHYDRAMNIOS AND OLicoHYDRAmNNIos 85

important only in late pregnancy. Earlier, thefluid is derived mainly from other sources.

It may be added that an association betweenbirth weight and liquor volume has long beennoted by clinicians. Thus it is commonly statedthat a large fetus is a cause of polyhydramnios.Excluding those whose weights could have beeneither increased or decreased by another con-

dition, we studied 119 babies born from motherswho suffered from polyhydramnios, comparingtheir birth weights with those of babies of similarmaturity born from amniotic sacs containing a

normal amount of liquor. The results, expressed asaverages, are shown in Fig. 6, and confirm the

general impression that a large baby and a largeamount of liquor often go together. The natureof the association, however, is not necessarily theaccepted one, and our observations lead to thesuggestion that whereas oligohydramnios causesdwarfing, polyhydramnios encourages excessivefetal development. It follows that it may also benecessary to revise the outlook in regard to thesmall fetus with its small volume of liquor des-cribed as part of the ill-defined syndrome ofplacental insufficiency. What is the sequence ofevents? Is it small placenta > small baby

>* small volume of liquor? Or is it small volumeof liquor > small placenta -> small baby?

PLACENTAL SIZE AND LIQUOR VOLUME

Evidence on the growth of ihe placenta withthe advance of pregnancy is conflicting, but itdoes seem that the chorionic plate conrtinues toincrease in area beyond mid-pregnancy.4' 39 It isdifficult to conceive that th-e placenta has a grow-ing edge or that it shifts its connections to accom-

'modate its increasing area. It must, therefore, bepresumed that placental growth occurs pari passu

with that' of the portion of the uterine wall towhich it is attached. This means that the greaterthe uterine distension the larger the area of theplacenta, and so may be explained the rough cor-

relation (for each phase of the pregnancy)between a large baby and polyhydramnios, and a

small baby and oligohydramnios.A previous interpretation of these relationships

was that the unusual placental size accounts for theabnormal liquor volume;40 this may still hold goodin certain circumstances. However, in the twogroups of cases with which we are mainly con-

cerned, viz., polyhydramnios associated with failureof the fetus to imbibe fluid, and oligohydramniosassociated with failure of the fetus to void urine,it is almost certain that it is the fetal abnormalitywhich initiates a whole chain of events. The firstresult is an alteration in the volume of liquor; this

then affects the uterine dimensions which control'the placental size; the last is mainly responsiblefor determining the fetal bulk.

CONCLUSION

There is clearly no simple solution to theproblems of the origin and fate of liquor amnii.

From the evidence available it seems certain thateach function' is covered by more than one

mechanism, and that the relative importance ofdifferent mechanisms changes with the progressof pregnancy.

Our clinical observations lead us, so far as latepregnancy is concerned, to disagree with Plentland Gray's28 statement that "Fetal micturition anddeglutition play, at most, a very subordinate rolein the physiology and physiopathology of theamniotic fluid." The exact place of 'thesemechanisms in normal physiology remains to bedecided, but failure of one or other of them ispractically always associated with a disturbance ofliquor volume. Provided the other mechanisms forthe production and absorption of liquor are oper-ating normally, pathological conditions of the fetuswhich are incompatible with fetal micturitioncause oligohydramnios while lesions which preventthe fetus swallowing and absorbing liquor fromthe intestine result in polyhydramnios. Moreover,irrespective of their cause, oligohydramnios retardsfetal growth and polyhydramnios encourages it.The conclusion of Hutchinson et al.,19 from their

isotope experiments, that the rate of waterexchange is independent of liquor volume mnayoffer a basis of explanation for some of the apparentcontradictions between the findings of cliniciansand' those of experimentalists. The actual volumeof liquor at one'time may not be determined byits rate of renewal and removal. Fetal swallowingand voiding could hardly be responsible for theenormous turnover of water which is suggestedby those working with tracer substances, yet thesetwo mechanisms do seem to be of vital importanceto the control of the volume of liquor amnii foundin late pregnancy. This raises the idea of a poolof fluid, the volume of which is largely though notentirely controlled by a relatively simple "tap"inflow (fetal micturition) and a "drain" outflow(fetal deglutition), with its molecules meanwhileconstantly diffusing in and out of the maternal'andfetal circulations. It is a concept which, althoughhighly conjectural, might reconcile apparentlyvalid clinical observations wlth the results of theradioactive isotope experiments so far reported.Failing such a reconciliation, however, well-documented clinical happenings cannot be ignoredmerely because they do not tally with experimentalfindings.

ADDENDUM

According to D. C. McKay, C. C. Roby, A. T. Hertigand Margaret V. Richardson (Am. J. Obst. &. Gynec., 69:722, 1955), the liquor amnii in the early months of preg-nancy is hypotonic to maternal plasma and not isotonic.This is contrary to the findings of Makepeace et al.,24 whichare quoted in the introduction, but it does not affect ourmain arguments and conclusions.

Canad. M. A. J.Jan. 15, 1959, vol. 80

3 HANDFORTH AND WOODBURY: CVS IN RUMATOID ARTERITI Canad. M. A. J.Jan. 15, 1959, vol. 80

REFERENCES

1. ALLEN, G. AND ORCHARD, N. P.: J. Obst. 6 ynao. Brit.Emp., 59: 810, 1952.

2. BAIN, A. D.: Personal communication, 1958.3. BAIN, A. D. AND SCOTT, J. S.: In press, 1958.4. BALARD: Quoted by SCHILLER, W. AND TOLL, R. M.: Am.

J. Obst. d Gynec., 13: 689, 1927.5. BARDRAM, E.: Acta obst. et gynec. scandinav., 10: 134,

1930.6. BARON, C.: Am. J. Obst. & Gynec., 67: 667, 1954.7. BATES, G. S.: Ibid., 25: 41, 1933.8. BOUND, J. P., BUTLER, N. R. AND SPECTOR, W. G.: Brit.

M. J., 2: 1191 and 1260, 1956, (and in personal com-munication quoted by WELCH, R. G.: Brit. M. J., 1:1102, 1958).

9. BROWNE, F. J. AND BROWNE, J. C. M.: Antenatal andpostnatal care, 8th ed., J. & A. Churchill Ltd., London,1955.

10. DAVIDSON, W. M. AND Ross, G. I. M.: J. Path. & Bact.,68: 459, 1954.

11. FRRIEDBERG, V.: Gynecologia, 140: 34, 1955.12. GOWAR, F. J. S.: J. Obst. & GynxeC. Brit. Emp., 42:

871, 1935.13. GRAY, M. J., NESLEN, E. D. AND PLENTL, A. A.: Proc.

Soc. Exper. Biol. d Med., 92: 463, 1956.14. HAMILTON, W. J. AND BOYD, J. D.: Proc. Roy. Soc. Med.,

44: 489, 1951.15. HARRISON, R. G. AND MALPAS, P.: J. Obst. & Gynac. Brit.

Emp., 60: 632, 1953.16. HIBBARD, B. M.: Proc. North of England Obst. & Gyn.ec.

Soc. (In Press), 1958.17. HOET, J. P.: In: Symposium on quantitative biology Vol.

19: The mammalian fetus: physiological aspects ofdevelopment. Biological Laboratory, Cold SpringsHarbor, New York, 1954, p. 187.

18. HOLTERMANN, C.: Zentralbl. f. Gyndk., 48: 2536, 1924.19. HUTCHINSON, D. L et al.: Surg. Gynec. & Ob8t., 100:

391, 1955.20. JEFFCOATE, T. N. A.: J. Obst. & Gynae. Brit. Emp., 38:

814, 1932.

21. KENT, B. K., RUIBIN, S. AND DANN, D. S.: Am. J. Ob8t.& Gynee., 74: 1106. 1957.

22. LANDING, B. H.: Ibid., 60: 1339, 1950.23. MCINROY, R. A. AND KELSEY, H. A.: J. Path. - Bact., 68:

519, 1954.24. MAKEPEACE, A. W. et al.: Surg. Gynee. d Obst., 53:

635, 1931.25. MACAFEE, C. H. G.: J. Obst. & Gynec. Brit. Emp., 57:

171. 1950.26. MORISON, J. E.: Fetal and neonatal pathology, Butter-

worth & Co. Ltd., London, 1952.27. OSMOND, R.: M. J. Australia, 1: 42, 1955.28. PLENTL, A. A. AND GRAY, M. J.: S. Clin. North America,

37: 405, 1957.29. PLENTL, A. A. AND HUTCHINSON, D. L.: Proc. Soc. Exper.

Biol. & Med., 82: 681, 1953.30. POTTER, EDITH, L.: Pathology of the fetus and the new-

born. Year Book Publishers, Inc., Chicago, 1952.31. RoSA, P.: Gynec. et obst., 50: 463, 1951.32. SCHILLER, W. AND TOLL, R. M.: Am. J. Obst. & Gynec.,

13: 689, 1927.33. SCOTT, J. S. AND WILSON, J. K.: Lancet, 2: 569, 1957.34. ScoTT, J. S. AND BAIN, A. D.: Proc. Roy. Soc. Med., 51:

512, 1958.35. SELBY, G. W. AND PARMELEE, A. H., JR.: J. Pediat., 48:

70, 1956.36. SHAW, R. E. AND MARRIOTT, H. J.: J. Obst. & Gynae.

Brit. Emp., 56: 1004, 1949.37. SILVESTER, P. E. AND HUGHES, D. R.: Brit. M. J., 1:

77, 1954.38. STIEVE, H.: Anat. Anz., 90: 225, 1940.39. TAGER, S. N.: Am. J. Roentgenol., 67: 106, 1952.40. TAUSSIG, F. J.: Am. J. Obst. &- Gynee., 14: 505, 1927.41. VOSBURGH, G. J. et al.: Ibid., 56: 1156, 1948.42. WAGNER, G. A.: Quoted by SCHILLER, W. AND TOLL, R. M.:

Am. J. Ob8t. d Gynec., 13: 689, 1927.43. WELCH, R. G.: Brit. M. J., 1: 1102, 1958.44. WHITHEAD, W. H., WINDLE, W. F. AND BECKER, R. F.:

Anat. Rec., 83: 255, 1942.45. WOLFF, B.: Arch. f. Gynak., 71: 224, 1904.

CARDIOVASCULAR MANI-FESTATIONS OFRIEUMATOID ARTHRITIS*

C. P. HANDFORTH, B.SC., M.B., B.S. andJ. F. L. WOODBURY, B.Sc., M.D., C.M.,

Halifax, N.S.

RHEUMATOID ARTHRITIS is a chronic inflammatorydisease of unknown etiology, the principal mani-festations of which are polyarthritis and sub-cutaneous nodules. Rheumatoid disease of thecardiovascular system is known to occur but israre by contrast with rheumatic heart disease." 2The pathological changes which may occur in theheart in rheumatoid disease are distinct from thosecaused by rheumatic fever. The object of tlhis paperis to record three examples of fatal rheumatoidcardiovascular disease.

MATERIALCASE 1.-A 16-year-old white girl was admitted

to hospital in 1957 because of fever and pain inmultiple joints. These symptoms had been presentfor two weeks and followed a mild sore throat forwhich she had not sought treatment.A maternal uncle had died at the age of 48 after

having rheumatoid arthritis for 11 years. The cause ofhis death is unknown.

In 1951, at the age of 10, the patient had hadpain and swelling of multiple joints lasting for three

*From the Departments of Pathology and Medicine, DalhousieUniversity and the Victoria General Hospital, Halifax, NovaScotia.

months. During this illness her heart was apparentlynormal and she had no residual joint deformitiesafterwards. In 19K4, at the age of 13, she had afurther attack of polyarticular pain and swelling asso-ciated with fever and anorexia. There had been somepreceding lassitude but no apparent respiratory in-fection. She became bedridden and remained in bedfor about nine months, during which time her weightfell to 66 lb. The wrists, cervical spine, hips and kneeswere the most prominently affected joints. Flexion de-formities developed in her hips and knees (Fig. 1).i l l >,'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ..........' :':'. :: :.0

Fig. I. Case 1. Photograph showing emaciation and Jointdeformities.

There was no enlargement of the spleen or superficiallymph nodes, no psoriasis and no clinical evidence ofheart disease. Radiography showed partial fusion ofthe sacroiliac joints, narowing of the joint spaces ofthe hips and skeletal rarefaction. She was treated inhospital for one year with an active physiotherapeuticregimen and some cortisone. When she returned homeshe weighed 80 lb. She was able to walk with theaid of canes but her wrist movements were limitedto a few degrees.