Pregnancy andthe changing pattern rheumatic heart disease · 1294 Szekely, Turner, andSnaith TABLE...

British Heart Journal, I973, 35, I293-I303.

Pregnancy and the changing pattern of rheumaticheart disease

Paul Szekely, Richard Turner, and Linton SnaithFrom the Cardiovascular Department and Department of Obstetrics, Newcastle General Hospital, Newcastleupon Tyne; and the University Department ofMedicine, Western General Hospital, Edinburgh

The clinical aspects of rheumatic heart disease in relation to pregnancy are reviewed in a combined series ofpatients chiefly from the Newcastle General Hospital studied over a period of 28 years, and from theWestern General Hospital in Edinburgh studied over 25 years.In both centres, since about I960, there has been a progressive decrease in the number of patientsfound in

antenatal clinics to have rheumatic heart disease, and also in the severity of this condition. There has also beena significant fall in the incidence of major complications, such as pulmonary oedema, right heart failure, andatrialfibrillation which formerly were often encountered.

In addition to the changes in the natural course of the disease, improved medical management of the more

severe cases, together with the introduction of cardiac surgery, has been responsiblefor a much morefavourablecourse in the pregnant patient.

The disadvantages and potential dangers of the New York Heart Association classification are emphasizedand illustrated.

The risks of valvotomy during pregnancy are no greater than in cases of comparable severity in the non-

pregnant state. Subsequent close medical supervision is important because complications may still occur laterin pregnancy. Likewise should pregnancy occur after successful valvotomy, close observation throughout isessential.

Current concepts ofmedical, surgical, and obstetric management are outlined.Routine medical examinations at about IO, I5, and 20 years of age with appropriate management would

render pregnancy virtually safefor every patient with rheumatic heart disease.

As judged by the number of new patients found atpresent to have rheumatic heart disease when ex-amined in antenatal clinics in Newcastle, Edinburgh,and in three other regions of Scotland (Szekely,i968; Turner, I968b; R. W. D. Turner andN. M. B. Dean, i969, unpublished data), and by thedecreasing frequency of serious complications dur-ing pregnancy such as pulmonary oedema, rightheart failure, and atrial fibrillation, the prevalenceand the severity of rheumatic heart disease have de-clined in recent years. A similar trend has recentlybeen reported from London (Barnes, I970). A recentinquiry carried out by one of us regarding the num-ber of first mitral valvotomies performed anuallysince its introduction in i8 centres of cardiac sur-gery in Britain also confirms the clinical impressionof a continuing decline in the incidence and severityof rheumatic heart disease in this country (R. W. D.Received iI June 1973.

Tumer and N. M. B. Dean, i969, unpublisheddata). These facts reflect the favourable changewhich has taken place in the past two decades in theprevalence and in the severity of acute rheumaticfever in the Western hemisphere (Bland, I960;Massell, Amezcua, and Pelargonio, 1964; WorldHealth Organization, I966; Scottish Health ServiceCouncfl, I967; Szekely, I968; Perry, I969). How-ever, this favourable trend is not universal as it doesnot apply to countries with less satisfactory socio-economic conditions.

Despite these favourable changes rheumatic heartdisease is still sufficiently prevalent to constitute oneof the potentially most serious medical risks en-countered in pregnancy. Also because of the steepdecline in maternal deaths from other causes, andin particular from infection, haemorrhage, and tox-aemia, heart disease has proportionately come toassume a more prominent position as a cause of

copyright. on D

ecember 10, 2020 by guest. P

rotected byhttp://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.35.12.1293 on 1 Decem

ber 1973. Dow

nloaded from

http://heart.bmj.com/

1294 Szekely, Turner, and Snaith

TABLE I Incidence of rheumatic heart disease in pregnancy

Newcastle Edinburgh Total

Years No. of all Pregnancies No. of all Pregnancies No. of all Pregnanciespregnancies complicated by pregnancies complicated by pregnancies complicated

rheumatic heart rheumatic heart by rheumaticdisease disease heart disease

No. % No. % %

I942-I947 5617 I93 3.5 5617 3.5I948-I957 I6672 517 31I 49915 845 17 66587 2-0I958-I965 14912 235 i-6 45287 827 i-8 60I99 I.7I966-i969 9875 103 I-I 25551 I36 0-5 35426 0o7Total 47076 I048 2-2 I20753 i8o8 I-5 I67829 I-7

maternal morbidity and mortality. Every maternaldeath from rheumatic heart disease can be regardedas potentially preventable. Appropriate prophylac-tic and therapeutic management of rheumaticcarditis from the onset of this disease and the earlydiscovery of insidious asymptomatic but severemitral stenosis by routine medical examination atschool leaving age and again at the age of about 20would render pregnancy virtually safe in everypatient with rheumatic heart disease.

In view of this situation and of the changes in thecourse of severe valve lesion brought about by theintroduction of cardiac surgery, and in the light of abetter understanding of the haemodynamics ofrheumatic heart disease and of the circulatorychanges of pregnancy made possible by improvedlaboratory methods of investigation, a reappraisalof rheumatic heart disease in relation to pregnancyis warranted. It is the purpose of this paper to pre-sent data obtained in two medical centres over a28-year period from 1942 to i969 in a large numberof patients, many of whom have been also observedin the years preceding pregnancy. In this analysis,an attempt has been made to emphasize the chang-ing clinical pattern of rheumatic heart disease and toput the present situation in statistical perspective.

SubjectsThe data compiled at the Newcastle General Hospitalare based on 1048 consecutive pregnancies in 723patients with rheumatic heart disease studied betweenI942 and I969. These patients have been closely followedby the same obervers for 7456 patient-years with anaverage period of more than I0 years. The dominantvalve lesion was mitral stenosis in go per cent of thecases, mitral regurgitation in 6-6 per cent, aortic re-gurgitation in 2-5 per cent, and aortic stenosis in i percent.The Edinburgh material which is particularly con-

sidered in relation to surgical treatment carried outbefore or during pregnancy, and to matemal mortality,

comprises a total of i8o8 cardiac pregnancies (SimpsonMemorial Maternity Pavilion Reports, I962-69; Turnerand Marquis, I965; Turner, I968b; Marquis, I969).The distribution of valve defects was similar to that inthe Newcastle series. In all respects the trend of ex-perience has been the same in both centres.There has been a gradual fall in the incidence of

rheumatic heart disease as seen in Table i.

Clinical observations and resultsThe principal complications of rheumatic heartdisease encountered during pregnancy, all of whichhave become much less frequent in recent years,were pulmonary congestion, pulmonary oedema,right heart failure, dysrhythmias, and embolism.The incidence of these complications in the New-castle series is analysed in detail (Table 2).

Heart failureIt is appreciated that at the present time there is noagreed definition as to what constitutes heartfailure. In mitral stenosis, which is the most fre-quent valve defect found in pregnancy, pulmonarycongestion and oedema are not primarily caused bymyocardial failure, but by mechnical obstruction atthe mitral valve with consequent increase in leftatrial and pulmonary venous pressures. Also theconventional manifestations of right heart failure,namely raised jugular venous pressure, engorge-ment of the liver, and peripheral oedema, are closelyrelated to fluid retention which is also determinedby a fall in cardiac output, so-called 'forward' failurewith consequent renal retention of salt and water.However, in this paper it is not our purpose to dis-cuss the pathogenesis of the clinical manifestationsof heart failure in general, but it is necessary toensure clarity as to what is being described whenusing the familiar subdivision of heart failure intopulmonary congestion, pulmonary oedema, andright heart failure. Such a subdivision proved help-

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from


Pregnancy and rheumatic heart disease 1295

TABLE 2 Frequency of acute pulmonary oedema, right heart failure, atrial fibrillation, and systemic embolismin pregnancy complicated by rheumatic heart disease (Newcastle)

Years No. of Pulmonary oedema Right heartfailure Atrialfibrillation Systemic embolismpregnanciescomplicated by No. of % No. of % No. of % No. of %rheumatic heart pregnancies pregnancies pregnancies pregnanciesdisease

I942-I95I 413 I0 2.5 I4 3.4 27 6-5 7 I.7(pre-valvotomy era)

I952-196I 432 10 2-3 6 I.4 36 8.4 7 I6I962-I969 203 0 0 0 0 6 3-0 I 0-5Total I048 20 2-0 20 2-0 69 6-8 15 I.4

TABLE 3 Heart failure in relation to cardiac rhythm in pregnancy complicated by rheumatic heart disease(Newcastle)

Cardiac rhythm No. of Pulmonary Acute pulmonary Right heart All three types ofpregnancies congestion oedema failure heart failure

No. %

Sinus rhythm throughout 94I i58 I5 II I84 I9-5Paroxysmal atrial tachycardia 38 5 2 0 7 i8.5

Onset beforeParoxysmal or pregnancy 36 4 I 3 8 22-0

established Onset inatrial pregnancy or 33 13 2 6 21 63-ofibrillation puerperium - -

Total 69 I7 3 9 29 42-0

Total 1048 i8o 20 20 220 22-0

ful from the point of view of management and prog-nosis.

Pulmonary congestion Criteria for the diagnosisof pulmonary congestion were a history of unduedyspnoea with persistent basal crepitations and/orradiographic signs of interstitial oedema.Pulmonary congestion was observed in i8o (i8%

of all pregnancies and in 82% of those with one ofthe three types of heart failure (Table 3)). In about6 per cent of the cases pulmonary congestionoccurred in the puerperium. Pure or dominantmitral stenosis was present in I76 instances andmitral regurgitation was the dominant lesion in 4cases.The incidence of pulmonary congestion increased

with age from ii per cent in patients under 20 yearsto 26 per cent in those 30 years or older.

There was also a significant rise in its incidencewith increase in cardiac size from i6 per cent inpatients with little or no cardiac enlargement(cardiothoracic ratio < 55%) to 35 per cent in pat-ients with great cardiac enlargement (CTR > 6o%).

The incidence was I7 per cent in those in sinusrhythm throughout, I3 per cent in those who ex-perienced episodes of atrial tachycardia, and 25 percent in those with transient or established atrialfibrillation. In the atrial fibrillation group, pulmon-ary congestion was more than three times as frequentin those in whom the dysrhythmia developed duringpregnancy or in the puerperium than in those inwhom it had been previously present and the ven-tricular rate controlled by digitalis before the onsetof pregnancy (Table 3).

Pulmonary oedema Acute pulmonary oedemawas diagnosed by the rapid onset of dyspnoeanecessitating the upright position, associated withcough and often the production of frothy sputumwith or without blood staining, and by evidence ofdiffuse pulmonary congestion. The difference be-tween pulmonary congestion and acute pulmonaryoedema is only one of degree, but clinically the twoconditions are readily distinguishable.Pulmonary oedema occurred in 20 of the Io48

pregnancies (2%), and in 9 per cent of those with

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from


x2g6 Szehely, Turner, and Snaith

one of the three types of heart failure (Table 3).These 20 pregnancies occurred in i8 patients whobetween them experienced a total of 25 attacks.Twenty attacks occurred during pregnancy and 5 inthe puerperium. Sixteen of the i8 patients had puremitral stenosis. In 2, the dominant lesion was aorticregurgitation, but there were important associateddisorders in both cases: hypertensive toxaemiaand hemichorea, respectively.Twenty-two of the attacks occurred in patients

under the age of 30 and 3 in those over the age of30. This age distribution is in sharp contrast to thatfound in the two other types of heart failure whichoccurred more frequently in patients over 30.

In all patients the cardiothoracic ratio was under6o per cent and there was only a marginal increasein the incidence of pulmonary oedema with increasein heart size. The fact that no episode of pulmonaryoedema occurred in patients with a cardiothoracicratio over 6o per cent is again in sharp contrast topulmonary congestion and right heart failure.

In 4 patients pulmonary oedema developed withthe onset of atrial fibrillation or atrial tachycardia.Another patient had long-established atrial fibrilla-tion. The remaining 20 episodes occurred in sinusrhythm. The precipitating factors which wereassumed to be operative are listed in Table 4. In 6instances no particular event was observed immed-iately before the attack, but since emotional upsetis known to be often responsible this may havebeen a cause.The incidence of pulmonary oedema in successive

periods since I942 is shown in Table 2. It is note-worthy that no episodes were observed after I960, afact that can be attributed to the introduction of

TABLE 4 Acute pulmonary oedema in pregnancycomplicated by rheumatic heart disease (Newcastle).Precipitating factors in 25 attacks occuring in 29pregnancies in i8 patients

Conditions immediately preceding onset ofpulmonary No. ofoedema attacks

Ectopic tachycardia 4Respiratory infection 4Physical effort 3Cerebral embolism IPulmonary embolism IPlasma transfusion IToxaemia of pregnancy 2Hemichorea IEmotional upset IMiscarriage INothing noticeable 6

mitral valvotomy and also to the natural decrease inthe severity of rheumatic heart disease.

Right heart failure A diagnosis of right heartfailure was made when in addition to pulmonarycongestion the jugular venous pressure was raisedand peripheral oedema developed in the absence ofother obvious causes such as severe anaemia or tox-aemia of pregnancy. In these cases, pulmonary con-gestion was followed by what we have termed rightheart failure without the clinical picture of acutepulmonary oedema. In these patients the heart wasnearly always grossly enlarged but the valve defectswere not necessarily severe.

Right heart failure thus defined occurred in 20pregnancies, constituting 2 per cent of the wholeseries and 9 per cent of those with one of the threetypes of heart failure (Table 2).

All 20 patients had dominant mitral stenosis: purestenosis in i6, and slight mitral regurgitation as wellin 4.The incidence of right heart failure increased with

advancing age, reaching I0 per cent in patients of40 years or over. It was also more frequent whenthe heart was greatly enlarged, the incidence beingonly o025 per cent in patients with slight cardiacenlargement and if per cent in those with cardio-thoracic ratio over 60 per cent.

Right heart failure occurred in ii (I%) of thepregnancies in the sinus rhythm group, and in 9(I3%) of those complicated by transient or estab-lished atrial fibrillation. In the latter group itoccurred twice as often in patients in whom atrialfibrillation developed during pregnancy or in thepuerperium than in those in whom it had been pre-viously present and the ventricular rate controlled(Table 3).The incidence of right heart failure in successive

periods since i942 is shown in Table 2. It is note-worthy that this complication has not been seensince I960, which again reflects the decreasingseverity of myocardial damage as judged by thedegree of cardiac enlargement and/or atrial fibrilla-tion.

Distribution of heart failure throughout preg-nancy

The data summarized in Table 5 clearly show thatthe incidence of all three types of heart failure in-creases as pregnancy advances. No less than ii(44 per cent) of the 25 attacks of acute pulmonaryoedema observed in 20 pregnancies occurred duringthe last four weeks, and 5 of the attacks (20%) in thepuerperium. However, on only 2 occasions did theattacks appear first in the puerperium, the remain-ing patients having had previous attacks.

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from



TABLE 5 Onset of heart failure in pregnancy complicated by rheumatic heart disease (Newcastle)

Weeks of No. of pregnancies Percentage distributionpregnancy of heart failure

Pulmonary Acute pulmonary Right heartcongestion oedema failure

-15 I0 0 0 4.5I6-20 I8 I 0 8-52I-25 22 3 I I2*026-30 32 4 2 I7031-35 36 I 5 I8.536-40 51 II 9 3I10Puerperium I I 5 3 8.5

Three pregnancies were complicated by two and one pregnancy by three attacks of pulmonary oedema.

Cardiac dysrhythmiasAtrial and ventricular extrasystoles, sometimes as abigeminal rhythm, are commonly encountered inpregnancy and are as a rule of no significance.Paroxysmal or established atrial fibrillation andatrial tachycardia are not infrequent, but ventric-ular tachycardia is rare (Szekely and Snaith, I953,i96i; Pine, Fox, and Shook, i965).

Atrial fibrillation Transient or permanent atrialfibrillation complicated 69 pregnancies in 59 patientsbetween the ages of i8 and 44, with an average ageof 32 years. In 36 instances it was either known withcertainty or assumed on reasonable grounds that thedysrhythmia had been present before pregnancy. In23 instances it appeared during pregnancy and in ioin the puerperium.

Mitral valve disease was present in 56 patients:pure stenosis in 46, and slight regurgitation as wellin io. Three patients had dominant mitral regurgita-tion. Mitral valve disease was associated with aorticregurgitation in I7 cases but in no case was theaortic regurgitation the dominant lesion.

Thirteen patients had slight cardiac enlargement(CTR <55%), 36 moderate enlargement (CTR55-6o%), and Io patients considerable enlargement(CTR > 6o%).

Heart failure occurred in 29 of the 69 pregnancies(42%), though atrial fibrillation was not necessarilythe immediate precipitating factor. When estab-lished atrial fibrillation had been present beforepregnancy, the three types of heart failure occurredin 8 out of 36 pregnancies (22%). When the dysrhy-thmia first developed during pregnancy or in thepuerperium heart failure occurred in a much higherproportion (63%, Table 3).The incidence of atrial fibrillation in successive

periods since i942 is shown in Table 2. It is now-adays much less frequently encountered in preg-nancy.

Atrial flutter Atrial flutter is uncommon duringpregnancy and was only encountered twice. Its sig-nificance is similar to that of atrial fibrillation.

Atrial tachycardia A diagnosis of paroxysmalatrial tachycardia was made in 28 patients withrheumatic heart disease between the ages of 2I and41 who had 52 documented attacks in 38 pregnan-cies. Forty-four attacks occurred during pregnancyand 8 in the puerperium. Twenty-one patients hadattacks only in pregnancy, mainly during the lasttrimester, 3 only in the puerperium, and 4 in bothpregnancy and the puerperium.Moderate to severe pure or dominant mitral

stenosis was present in 25 patients and dom nantmitral regurgitation in 3 patients. Associated aorticregurgitation was found in 6 patients but in no casewas it the dominant lesion.

In I6 patients the first attack occurred duringpregnancy or in the puerperium and I2 of these didnot have attacks at any other time. The other 4 aswell as the remanng I2 patients continued to haveoccasional attacks after pregnancy. Eight attackslasted between 6 and 72 hours, and 5 of these pre-cipitated pulmonary congestion and 2 acute pul-monary congestion and 2 acute pulmonary oedema.All other attacks lasted from 30 to go minutes onlyand were not followed by heart failure.

Systemic embolismFifteen episodes of systemic embolism were ob-served in i5 patients (i-5%). Eleven episodesoccurred during pregnancy and 4 in the puerperium.Eight patients were in sinus rhythm and 7 hadproven paroxysmal or established atrial fibrillation.In 6 of the 7 patients with atrial fibrillation em-bolism occurred between less than 24 hours and 9days after the onset of the ectopic tachycardia. Inone patient atrial fibrillation had been presentbefore pregnancy.

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from



TABLE 6 Mitral valvotomy during pregnancy

Newcastle Edinburgh Total

No. of patients 25 44 69Age at operation< 30 years I4 28 42> 30years II I6 27

Time of operationISt trimester 3 7 IO2nd trimester 10 30 403rd trimester IO 7 I7Puerperium 2 0 2

Degree of stenosisSevere i8 41 59Moderate 7 3 IO

Operative deaths I I 2Postoperative complications

Heart failure 2 2 4Transient or permanent atrial fibrillation 8 0 8Systemic embolism 2 0 2

The dominant valve lesion was mitral stenosis inall cases: pure stenosis in 12 (80%) and slight re-gurgitation as well in 3 (20%).

Pulmonary embolismPulmonary embolism occurred in 20 patients (2%).Nine were in sinus rhythm and ii had paroxysmalor established atrial fibrillation. Four patients in thelatter group also had systolic embolism. Sixteenepisodes occurred during pregnancy and 4 in thepuerperium.

Pure or dominant mitral stenosis was present ini8 patients, and dominant mitral regurgitation in 2patients.Only 4 of the 20 patients had clinically overt deep

vein thrombosis. Haemoptysis was observed in 6 ofthe 20 cases. None of the patients was on anticoagu-lant treatment at the onset of pulmonary embolism.The 4 patients with overt phlebothrombosis wereobserved in the pre-anticoagulant era.

In one patient pulmonary embolism followedtermination of pregnancy and in 2 delivery bycaeserian section.

In I942, a 24-year-old woman who developedatrial fibrillation Io days after an uneventful de-livery, and a popliteal embolism 9 days later, diedon the 23rd postpartum day exhibiting the clinicalsigns of massive pulmonary embolism. There wasno evidence of overt phlebothrombosis. Permissionfor necropsy could not be obtained.

Cardiac surgery in relation to pregnancyMitral valvotomy during pregnancyIn the combined experience of the Newcastle Gen-eral Hospital, the Western General Hospital, and

the Royal Infirmary in Edinburgh mitral valvotomyhas been carried out in 69 patients since I95I(Table 6). Ten patients had their operation in thefirst, 40 in the second, and I7 in the third trimester,and 2 in the puerperium.The main postoperative complications were heart

failure and atrial fibrillation. Heart failure developedin 4 of the 69 patients: 2 of these had previous rightheart failure suggesting considerable myocardialdamage. Postoperative atrial fibrillation developedin 8 patients, but sinus rhythm was restored in 6 ofthem.There were 2 operative deaths: one from trau-

matic mitral regurgitation (before the introductionof open heart surgery) and the other after an emer-gency mitral valvotomy for uncontrolled pul-monary oedema.There were 2 spontaneous abortions and 6 peri-

natal deaths.In the first I5 years after the introduction of

mitral valvotomy, 64 operations were considerednecessary in pregnancy. Between I966 and I969,using the same criteria for operative treatment, therehave only been 5 such cases.

Pregnancy after mitral valvotomyIn the combined series from the three hospitals, i6ipatients experienced 210 pregnancies after mitralvalvotomy. Thirty-seven patients developed unduedyspnoea or cardiac failure at some stage of preg-nancy.There were 3 maternal deaths: i from acute pul-

monary oedema in a 27-year-old woman who had aclosed mitral valvotomy in a previous pregnancy 3years earlier, and 2 patients died from infectiveendocarditis.

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from



TABLE 7 Causes of maternal death in pregnancy complicated by rheumatic heart disease

Cause of death Newcastle Edinburgh Total

Pulmonary oedema 6 7 13Right heart failure I O IPulmonary embolism I 2 3Bacterial endocarditis 0 2 2Acute rheumatic carditis I I 2Acute circulatory failure I I 2Traumatic (operative) mitral regurgitation o I IObstruction of tricuspid orifice by aneurysm of atrial septum I O INon-cardiac death I 0 I

Total I2 I4 26

Pregnancy after valve repair or replacementAt the Edinburgh Western General Hospital 2patients had pregnancies after mitral valve repairfor mitral regurgitation, 3 after mitral, and i afteraortic valve replacement by Starr-Edwards pros-thesis. At the Newcastle General Hospital 2 patientsbecame pregnant after mitral valve repair, and ipregnancy was observed after mitral valve replace-ment by Starr-Edwards prosthesis. All 9 pregnan-cies were uncomplicated and there was no foetalloss.The 5 patients with valve replacement continued

with oral anticoagulant treatment until the onset oflabour and resumed it within 24 hours after delivery.

Maternal mortalityIn the Newcastle series there were I2 deaths, butnone since I960 (Table 7). In 6 patients the cause ofdeath was acute pulmonary oedema, 3 of whichoccurred in the prevalvotomy era. The other 3deaths caused by pulmonary oedema occurred, re-spectively, in a patient who was admitted to hospitalmoribund and died within an hour, in anotherpatient who had closed mitral valvotomy in a pre-vious pregnancy for pulmonary oedema and died ina subsequent pregnancy before arrangements for asecond mitral valvotomy could have been com-pleted, and in a further patient who was actuallyoperated on at 38 weeks, but died i8 hours aftermitral valvotomy without regaining consciousness.Two other deaths which occurred in the pre-valvotomy era and could be regarded today aspotentially preventable were due, respectively, toprogressive right heart failure and to peripheralcirculatory failure associated with profuse haemopty-sis. Rheumatic carditis and pulmonary embolismwere each responsible for one death. Another patientdied suddenly in the puerperium and the necropsyshowed obstruction of the tricuspid orifice by atumour-like structure which proved to be an

aneurysm of the atrial septum filled with blood clots.In the twelfth patient the cause of death was non-cardiac.

In the Edinburgh series there were I4 deathsfrom rheumatic heart disease, but none since I960.Before mitral valvotomy became established 5patients with severe mitral stenosis died from pul-monary oedema. On the other hand, between I952and I960 only 2 patients died from pulmonaryoedema and neither had been seen in an antenatalclinic. Two of the other deaths were from pulmonaryembolism, 2 from infective endocarditis, i fromsudden circulatory failure in a patient with a verylarge heart, i from acute rheumatic carditis, and ifrom traumatic (operative) mitral regurgitation.

Foetal mortalityThe perinatal mortality rate as observed at the New-castle General Hospital has been gradually fallingover the whole period from 8-8 per cent (I942-I95I) to under 5 per cent (I96I-I969). It is, how-ever, still higher than the current general perinatalmortality rate which is less than 3 per cent.

DiscussionThe data presented here cover a period of 28 yearsbetween i942 and I969 and provide a spectrum ofthe changing clinical pattem of rheumatic heartdisease as observed in the child-bearing age and inthe pregnant woman. They reflect a gradual declinein the incidence of rheumatic heart disease and alsoa favourable change in the severity of the disease.This trend has continued since i969.

Principles of antenatal careThe clinical course of rheumatic heart disease whichis in the majority of cases well established before thechild-bearing age may be temporarily influenced bypregnancy. The purpose of cardiac antenatal care isthe detection of the nature of the cardiac abnor-

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from



mality, the precise assessment of its severity, and theintroduction of appropriate measures aimed at theprevention of complications.Any form of purely functional grading, such as

that recommended by the New York Heart Associa-tion which is still widely used, can in our opinion bemisleading, potentially dangerous, and should there-fore be abandoned. We have already defended thispoint of view in several discussions (Turner andMarquis, I965; Tumer, I968a; Szekely and Tumer,I968). It is usually assumed that symptoms nec-

essarily reflect the severity of the underlying heartdisease. However, experience has shown that this isnot the case. It is true that when a functional classi-fication has been used, published figures have showna comparatively low maternal mortality rate ingrades I and II. However, this is so because in thegreat majority of cases the rhuematic valve defect isnot severe. It is equally true that the majority ofmaternal deaths during pregnancy occur in patientswho when first seen were placed in functional gradesI or II. In particular, a patient with severe mitralstenosis but good myocardium and little or no car-

diac enlargement can be practically symptom free atthe beginning of pregnancy and yet develop laterfatal pulmonary oedema. We believe, therefore,that cardiac antenatal care should be governedby an assessment based essentially on objectivefindings, with a view to establishing the severityof the lesion and thus the potential risk ofpregnancy and the need for special medical orsurgical treatment.

Complications The most important complica-tions encountered during pregnancy included acutepulmonary oedema, right heart failure, atrial fibrilla-tion, and embolism. Formerly these complicationswere responsible for serious maternal morbidity andmortality, but they have gradually become less fre-quent. The steady decrease in maternal mortalityfrom cardiac causes is well documented (Gilchrist,I963; Department of Health and Social Security,I968). Important factors which have been respons-ible for this decrease are improved medical care

during pregnancy and during the years precedingthe childbearing age, including antistreptococcalprophylaxis, surgical relief of severe mitral valveobstruction, and also the natural decline in theseverity of rheumatic heart disease.

Despite this favourable trend, serious and poten-tially dangerous complications may still occur. Thus,the report on confidential inquiries into maternaldeaths in England and Wales (Department ofHealthand Social Security, I968) lists 43 deaths fromacquired heart disease between I964 and i966, 23 ofwhich were rheumatic in origin. In the last report

covering the period I967 tO I969 there were I4deaths from rheumatic heart disease.The pattern of heart failure during pregnancy has

undergone a gradual change during the past 20years. Gilchrist (I963) reported that between I928and I947 right heart failure accounted for 6i percent and acute pulmonary oedema for 13 per cent ofcardiac deaths, whereas between I948 and I960 thecorresponding figures were 8 per cent and 67 percent, respectively. This trend has continued.Acute pulmonary oedema, though much less

common than previously, is today the most impor-tant single cardiac cause of maternal mortality. Awoman with even moderately severe mitral stenosismay develop pulmonary oedema during pregnancybecause of the increase in blood volume, in cardiacoutput, and in heart rate, accompanied by an in-crease in left atrial, pulmonary venous, and pul-monary capillary pressures (Burwell and Metcalfe,1958). In this way an anatomically moderatestenosis can be transformed into a functionallysevere one. It is helpful to recognize the type ofpatient who is especially susceptible to pulmonaryoedema. She is likely to be under the age of 30 withfew if any previous symptoms, with clinical signs ofmoderate to severe mitral stenosis, and, as a rule,with only slight cardiac enlargement and normalsinus rhythm. In assessing the risks of pregnancy inthis type of patient it is essential to appreciate thatthe functional capacity of the heart before pregnancyor at its onset is no measure of the risk of acute pul-monary oedema during pregnancy. It also appearsthat the combination of mitral stenosis or aorticregurgitation and of toxaemia of pregnancy is un-favourable as this association potentiates the ten-dency to pulmonary oedema.Right heart failure has in our experience virtually

disappeared in recent years. It used to be observedas a rule in older patients with larger hearts, con-siderable myocardial damage, and symptoms beforepregnancy. These observations are similar to thosereported earlier by Jones (I959).The available data show that heart failure can

occur at any stage of pregnancy and that its inci-dence increases as pregnancy progresses. This trendis maintained until term. These long-appreciatedclinical observations have now found their explana-tion in the recent findings that the cardiac outputdoes not fall as previously thought until after de-livery. When the cardiac output is measured in thelateral recumbent position a fall towards the end ofpregnancy is not observed (Lees et al., I967; Kerr,I968; Mulholland and Boyle, I968; Ueland et al.,I969; Hytten and Leitch, I97I). The decrease incardiac output previously described in late preg-nancy in the supine position is due to the reduction

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from


Pregnancy and rheunatic heart disease 1301

in venous return brought about by compression ofthe inferior vena cava by the gravid uterus. Theearly puerperium is also a critical period, probablybecause of the redistribution of the blood volumewhich occurs at this time and results in a furthertransient rise in cardiac output.

Uncontrolled atrial fibrillation carries the risk ofprecipitating pulmonary congestion or oedema andalso systemic embolism (Szekely and Snaith, I96I).In our experience, atrial fibrillation is a less seriouscomplication when established before pregnancyand the heart rate already well controlled with digi-talis than when it develops during pregnancy.Patients who first develop atrial fibrillation duringpregnancy do so as a rule after the fourth month at atime when the blood volume is already increasedalmost to its maximum. It is therefore possible thatthe increase in circulating blood volume in a patientwith severe rheumatic heart disease may be animportant factor in determining the onset of atrialfibrillation. Another possible factor might berheumatic activity with carditis, a raised anti-streptolysin titre having been found not uncom-monly, but the evidence for this is conflicting andin the individual patient the diagnosis of rheu-matic activity may be difficult or impossible to make.In any event, atrial fibrillation reflects significantmyocardial damage, especially if associated withmore than slight cardiac enlargement. Atrial tachy-cardia can also precipitate pulmonary congestion oroedema, especially if it lasts for several hours, andshould therefore be terminated as early as possible.

In our series, all IS patients with systemic em-bolism had dominant mitral stenosis: 8 were insinus rhythm; 7 had atrial fibrillation, and in 6 ofthese embolism occurred between i and 9 days afterthe onset of the dysrhythmia. This feature reflectsthe general tendency for embolism to occur shortlyafter the onset of atrial fibrillation.The fatal episode of massive pulmonary embolism

we observed occurred in the pre-anticoagulant era.However, in this case there were no clinical warningsigns which are often present preceding a fatal epi-sode (Marquis, I966), except that the patientdeveloped atrial fibrillation I3 days earlier andsystemic embolism 4 days earlier. Contrary to someother series, we observed more episodes of pulmon-ary embolism in the antepartum period than in thepuerperium. In a recent report, Henderson, Lund,and Creasman (I972) suggested that there might bea decreasing incidence of pulmonary embolism inthe puerperium and a relative increase in the num-ber of cases in the antepartum period.

Current concepts of management(i) Medical aspects Potentially, acute pulmon-

ary oedema is preventable. The recognition of earlypulmonary congestion and its appropriate manage-ment constitutes an important step towards its pre-vention. Respiratory infection is a frequent preced-ing event and requires early and energetic treatment.If pulmonary congestion appears, the patientshould be rested in bed and treated with digitalis anddiuretics. The medical management of acute pul-monary oedema consists of the administration ofmorphine together with a drug aimed at preventingvomiting, oxygen, a rapidly-acting diuretic, digi-talis, and assisted respiration if necessary. Thepatient should be nursed in a propped-up position.

Atrial fibrillation with a rapid ventricular rate re-quires immediate bed-rest and digitalis. Sometimesa beta-blocking agent such as practolol or oxprenololis helpful. Owing to the potential danger of systemicembolism early after the onset of atrial fibrillation,anticoagulant treatment should be considered inselected cases, especially if significant mitral stenosisis present.

Although in a given series of cases there may beno evidence that systemic embolism in the presenceof rheumatic heart disease occurs more frequentlyduring pregnancy than in non-pregnant patients ofthe same age group, in our experience the risk perpatient-year in the atrial fibrillation group provedto be higher in pregnancy than in the non-pregnantstate. It is also noteworthy that when systemicembolism occurred in atrial fibrillation, it did sowithin I0 days after its onset in 85 per cent of thecases. Intravascular thrombosis probably occursmore frequently in pregnancy than in the non-pregnant state (Barnes, 1970) and certain obstetricfactors, such as high parity, delivery by caesariansection, and hormone-induced suppression of lacta-tion may further increase the tendency to thrombo-embolism. The desirability of anticoagulant therapyin the individual patient should therefore be deter-mined in the light of these observations. Thoughanticoagulant treatment carries some risk to thefoetus, it is feasible and can be made safe understrict supervision and laboratory control (Turnerand Kitchin, I968; Szekely and Snaith, 1969;Barnes, 1970). Phenindione or warfarin have provedmore practicable in our experience than heparin,though heparin is theoretically safer because it doesnot cross the placental barrier. Macdonald (1970)suggested a reasonable compromise in giving oralanticoagulants during pregnancy and replacing it byheparin shortly before the onset of labour, as trans-placental transmission may be most dangerous atthe time of delivery. In a recent review, Hirsh, Cade,and O'Sullivan (I970) have also stated that anti-coagulant treatment can be safely carried out duringpregnancy. They observed only occasional minor

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from



haemorrhagic episodes and no foetal or neonatalcomplications. If anticoagulants are used until theonset of labour it is reasonable to give vitamin K tothe newborn (Mendelson, I960).The conventional antidysrhythmic agents for the

abolition of atrial fibrillation or atrial tachycardiaare not contraindicated in pregnancy and can beused with success (Szekely and Snaith, I953). Directcurrent shock has also been safely used in pregnancy(Vogel, Pryor, and Blount, I965; Schroeder andHarrison, I97I). The indications for these pro-cedures are similar to those in the non-pregnantpatient.

(2) Surgical aspects: (a) Closed mitral valvo-tomy. In general, the indications for closed mitralvalvotomy during pregnancy and the operativemortality do not differ from those outside pregnancy(Logan and Turner, I952; Szekely and Snaith, I963;Ueland, I965; Snaith and Szekely, I967; Metcalfe,I968; Turner, I968b; Metcalfe and Ueland, I970).However, in certain cases operative treatmentappears to be more urgent than if the patient werenot pregnant, and in other cases operation should bepostponed until some time after delivery. As closedmitral valvotomy carries today a low operative risk,it should be carried out in the presence of puremitral stenosis if pulmonary congestion developswith undue dyspnoea and there is no prompt re-sponse to medical management. Operative treat-ment should be considered during pregnancy whena reliable history of pulmonary oedema can beelicited, even if the patient is symptom freeat the time of assessment, because pulmonaryoedema is likely to recur. Profuse and uncontrollablehaemoptysis usually associated with pulmonaryhypertension and a severe degree of mitral stenosisalso requires urgent valvotomy. It should be againemphasized that an anatomically moderate stenosiscan be transformed into a functionally severe one bythe haemodynamic changes of pregnancy.

Maternal mortality from acute pulmonary oedemacan be reduced by the appropriate selection ofpatients for surgical treatment in whom the risk ofcontinuing with conservative management is likelyto be greater than that of operating during preg-nancy. Though the operative risk is not higherduring pregnancy than in the non-pregnant state,it is of course true that two lives are at stake. Theincidence of traumatic mitral regurgitation shouldnowadays be very low and with facilities for openheart surgery immediate repair or valve replace-ment can be undertaken.

If mitral valvotomy is considered desirable duringpregnancy, it can be performed at any stage. How-ever, it must be emphasized that mitral valvotomy

can only be regarded as a therapeutic episode in thecourse of rheumatic heart disease, and strict medicalsupervision thereafter continues to be imperative.The valve is not restored to normal, and even anapparently good valvotomy is not an absolute pro-tection against complications in the remainder ofpregnancy. Likewise, subsequent pregnancies re-quire careful assessment and strict supervision. Anincreasing number of patients require further car-diac surgery in later years.

(b) Open heart surgery Pregnancy is no absolutebar to valve replacement under extracorporeal circu-lation. At the time of writing 22 patients have beenreported who underwent open heart surgery duringpregnancy, and about half of these were undertakenfor mitral or aortic valve disease (Zitnik et al.,I969). The overall maternal mortality was 5 percent in these cases, which suggests that there is nosignificant additional hazard in open heart surgeryundertaken during pregnancy as compared withthat in non-pregnant patients. However, the foetalmortality amounted to 33 per cent. A large numberof patients were operated on early in pregnancy.Zitnik et al. (I969) state that open-heart surgerywith extracorporeal circulation should preferably beundertaken well after the first trimester whenorganogenesis is completed.

Pregnancy after successful valve replacement, in-cluding the continuation of anticoagulant prophy-laxis, is a manageable situation and does not con-stitute an unacceptable hazard (Turner and Kitchin,I968; Szekely and Snaith, I969; Macdonald, I970).However, in addition to the usual potential compli-cations, due consideration should be given to theincreased risk of thromboembolism and infectiveendocarditis as in cases after valve replacement ingeneral.

(3) Obstetric aspects The principles of obstetricmanagement of patients with rheumatic heartdisease do not differ greatly from those of generalobstetric practice. The patients should be suitablysedated at the onset of labour and during the firststage and kept in the propped-up position as far aspossible. The second stage should be shortened byforceps delivery when appropriate.

Prevailing opinion is that caesarian section shouldas a rule only be carried out for obstetric reasons.

Opinions are still divided on the use of ergome-trine. On the whole, it should be used sparingly asit can precipitate acute pulmonary oedema in thepresence of mitral stenosis.As there is evidence that the administration of

stilboestrol for the suppression of lactation afterdelivery is associated with an increased incidence of

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from



thrombophlebitis and pulmonary embolism (Daniel,Campbell, and Turnbull, i967), it should not beused in patients with rheumatic heart disease, if atall.

We are grateful to Professor R. J. Kellar for permissionto quote from the Annual Reports of the SimpsonMemorial Maternity Pavilion, Edinburgh.

ReferencesBarnes, C. G. (I970). Medical Disorders in Obstetric Practice,

3rd ed., pp. II, 26, and 214. Blackwell Scientific Publica-tions, Oxford and Edinburgh.

Bland, E. F. (I960). Declining severity of rheumatic fever:a comparative study of the past four decades. New EnglandJ'ournal of Medicine, 262, 597.

Burwell, C. S., and Metcalfe, J. (I958). Heart Disease andPregnancy, p. I25. Little, Brown, Boston.

Daniel, D. G., Campbell, H., and Turnbull, A. C. (1967).Puerperal thromboembolism and suppression of lactation.Lancet, 2, 287.

Department of Health and Social Security (1968). Report onConfidential Enquiries into Maternal Deaths in England andWales, I964-66. H. M. Stationery Office, London.

Gilchrist, A. R. (I963). Cardiological problems in youngerwomen including those of pregnancy and the puerperium.British Medical Journal, I, 209.

Henderson, S. R., Lund, C. J., and Creasman, W. T. (I972).Antepartun pulmonary embolism. American Journal ofObstetrics and Gynecology, 112, 476.

Hirsh, J., Cade, J. F., and O'Sullivan, E. F. (1970). Clinicalexperience with anticoagulant therapy during pregnancy.British Medical Journal, I, 270.

Hytten, F. E., and Leitch, I. (I97i). The Physiology ofHumanPregnancy, 2nd ed., p. 69. Blackwell Scientific Publications,Oxford.

Jones, A. M. (I959). Heart disease in pregnancy. Proceedingsof the Royal Society of Medicine, 52,767.

Kerr, M. G. (I968). Cardiovascular dynamics in pregnancyand labour. British Medical Bulletin, 24, I9.

Lees, M. M., Scott, D. B., Kerr, M. G., and Taylor, S. H.(I967). The circulatory effects of recumbent posturalchange in late pregnancy. Clinical Science, 32,453.

Logan, A., and Turner, R. W. D. (I952). Mitral valvotomy inpregnancy. Lancet, I, I286.

Macdonald, H. N. (I970). Pregnancy following insertion ofcardiac valve prostheses: a review and further case report.Journal of Obstetrics and Gynaecology of the British Com-monwealth, 77, 603.

Marquis, R. M. (I966). Pulmonary embolism and pregnancy.British Medical Journal, I, 1480.

Marquis, R. M. (I969). Mitral disease and pregnancy. Journalof the Royal College of Physicians of London, 3, I21.

Massell, B. F., Amezcua, F., and Pelargonio, S. (i964).Evolving picture of rheumatic fever. Journal of the Ameri-can Medical Association, I88, 287.

Mendelson, C. L. (I960). Cardiac Disease in Pregnancy, p. 67.F. A. Davis, Philadelphia.

Metcalfe, J. (I968). Rheumatic heart disease in pregnancy.Clinical Obstetrics and Gynecology, II, 1010.

Metcalfe, J., and Ueland, K. (I970). The heart and pregnancy.

In The Heart, 2nd ed., p. I393. Ed. by J. W. Hurst andR. B. Logue. McGraw-Hill, New York.

Mulholland, H. C., and Boyle, D. McC. (I968). The effect ofposture on the cardiac output during the last six weeks ofpregnancy. American Heart Journal, 76, 291.

Perry, C. B. (I969). The natural history of acute rheumatism.Annals of the Rheumatic Diseases, 28, 471.

Pine, H. L., Fox, L., and Shook, D. McK. (I965). Paroxysmalventricular tachycardia complicating pregnancy. AmericanJournal of Cardiology, 15, 732.

Schroeder, J. S., and Harrison, D. C. (I97i). Repeatedcardioversion during pregnancy. American Journal ofCardiology, 27, 445.

Scottish Health Service Council (I967). Rheumatic Fever inScotland. H.M. Stationery Office, Edinburgh.

Simpson Memorial Maternity Pavilion, Medical and ClinicalReports (I962-1969). Royal Infirmary, Edinburgh.

Snaith, L., and Szekely, P. (I967). Cardiovascular surgery inrelation to pregnancy. In Advances in Obstetrics and Gyne-cology, p. 220. Ed. by S. L. Marcus and C. C. Marcus.Williams and Wilkins, Baltimore.

Szekely, P. (I968). Current incidence of rheumatic fever andthe present status of prevention of rheumatic heartdisease. In Proceedings of the 5th International Congress ofHygiene and Preventive Medicine, Rome, Vol. I, p. II9.

Szekely, P., and Snaith, L. (I953). Paroxysmal tachycardiain pregnancy. British Heart Journal, I5, 195.

Szekely, P., and Snaith, L. (I96I). Atrial fibrillation andpregnancy. British Medical Journal, I, 1407.

Szekely, P., and Snaith, L. (I963). The place of cardiac sur-gery in the management of the pregnant woman with heartdisease. Journal of Obstetrics and Gynaecology of the BritishComnonwealth, 70, 69.

Szekely, P., and Snaith, L. (I969). Mitral-valve prosthesis,warfarin anticoagulation and pregnancy. Lancet, 2,598.

Szekely, P., and Turner, R. W. D. (i968). Indications fortermination of pregnancy. British Medical Journal, 2, 556.

Turner, R. W. D. (I968a). Controversial aspects of rheumaticheart disease. British MedicalJournal, 2, 383.

Turner, R. W. D. (I968b). Heart disease in pregnancy.Hospital Medicine, 2, 1294.

Turner, R. W. D., and Kitchin, A. H. (I968). Pregnancy aftermitral-valve prosthesis. Lancet, I, 862.

Turner, R. W. D., and Marquis, R. M. (i965). Symposium onheart disease and pregnancy. Scottish MedicalJournal, Io,47.

Ueland, K. (I965). Cardiac surgery and pregnancy. AmericanJournal of Obstetrics and Gynecology, 92, I48.

Ueland, K., Novy, M. J., Peterson, E. N., and Metcalfe, J.(I969). Maternal cardiovascular dynamics. AmericanJournal of Obstetrics and Gynecology, 104, 856.

Vogel, J. H. K., Pryor, R., and Blount, S. G. (I965). Direct-current defibrillation during pregnancy. Journal of theAmerican Medical Association, 193, 970.

World Health Organization (1966). Prevention of rheumaticfever. World Health Organization. Technical Report Series,No. 342.

Zitnik, R. S., Brandenburg, R. O., Sheldon, R., and Wallace,R. B. (I969). Pregnancy and open-heart surgery. Circula-tion, 39, Suppl. I, 257.

Requests for reprints to Dr. Paul Szekely, 34 EslingtonTerrace, Jesmond, Newcastle upon Tyne NE2 4RN.

copyright. on D



j.com/

Br H


ber 1973. Dow

nloaded from


Pregnancy andthe changing pattern rheumatic heart disease · 1294 Szekely, Turner, andSnaith TABLE...

Documents

Transcript of Pregnancy andthe changing pattern rheumatic heart disease · 1294 Szekely, Turner, andSnaith TABLE...