bacterial endocarditis, mitral regurgitation, and intra...

Thorax (1968), 23, 11.

Bacterial endocarditis, mitral regurgitation, andintra-atrial thrombosis following mitral

valve replacementHARRY M. WINDSOR, PAUL FAGAN, AND

MARK X. SHANAHAN

From St. Vincent's Hospital, Sydney, Australia

Bacterial endocarditis with positive blood culture occurred on six occasions in a series of 140mitral valve replacements. In three of these, extensive detachment of the prosthesis with severemitral incompetence resulted. Re-operation was undertaken in two of these cases. Intra-atrialthrombosis occurred twice. In three other cases, in which intra-atrial thrombosis occurred,infection was strongly suspected to have been the responsible factor. Mitral regurgitationpresented no difficulty in diagnosis, as all patients had severe congestive cardiac failure andtypical physical signs. Confirmation was established by cineangiography. Intra-atrial thrombosispresented difficulty in diagnosis. A persistently positive blood culture, continued pyrexia, and ahistory of sepsis were the most consistent and significant findings in the cases reported. Theantibiotic regime described by Amoury and his colleagues has been used for the past year. Therehas not been a single case of bacterial endocarditis, intra-atrial thrombosis, or wound infectionin 120 consecutive valve replacements on this regime. Because of the high mortality associatedwith intra-atrial thrombosis, and the close association between endocarditis and thrombosis, thepresence of endocarditis with or without regurgitation might well be a compelling reason forre-operation. The mortality from these complications could, we believe, be reduced by the useof a more extensive antibiotic cover and a more vigorous attitude to surgical intervention. Wehave not had to consider re-operation since adopting the policy outlined, as these complicationshave not occurred.

Bacterial endocarditis with positive blood cultureoccurred six times in 140 patients who underwentmitral Valve replacement (see Table). In these six,mitral regurgitation and intra-atrial thrombosishad a significant incidence. In a further three cases,intra-atrial thrombosis was associated with infec-tion, but positive blood cultures were not obtained.The object of this paper is to correlate the

various aspects of these complications.

BACTERIAL ENDOCARDITIS ASSOCIATED WITHMITRAL REGURGITATION

CASE 1 (R.M. 859) (Windsor and Shanahan, 1967)A man aged 43 had a heavily calcified mitral valvereplaced on 12 August 1963. His post-operativecourse was uneventful. He was given sodium methi-cillin, 4 g. daily for one week, and sodium cloxacillin,1 g. daily for two months, following operation.

11

TABLEBACTERIAL ENDOCARDITIS AND ASSOCIATED

PHENOMENA

Case Intra-atrial MitralNo. Organism Aetiology Throm- Regurgi- Result

bosis tation

1 Staph. albus ? Cardiac + + Alive andcoagulase- catheteriza- wellnegative tion

2 Staph. aureus Empyema + + Diedcoagulase- thoracispositive

3 Staph. albus Calcific + + Diedcoagulase- remnantnegative; in mitralStaph. annulusaureuscoagulase-positive

4 Staph. albus Empyema + _ Diedcoagulase- thoracisnegative

5 Staph. aureus Unknown + - Diedcoagulase-positive

9 Staph. albus ? Calcific ? ? Died;coagulase- remnant in necropsynegative mitral refused

annulus

B

on 22 May 2018 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.23.1.11 on 1 January 1968. D

ownloaded from

http://thorax.bmj.com/

Harry M. Windsor, Paul Fagan, and Mark X. Shanahan

Twelve months after surgery at the time of routinecardiac catheterization he had a virtually normalexercise tolerance, no evidence of cardiac failure, andno abnormal cardiac murmurs. His pulmonary arterypressure was 38/12 mm. Hg at rest and 84/48 mm. Hgon exercise, compared with a resting pulmonary arterypressure of 90/50 mm. Hg prior to operation.Two months after this procedure he developed

fulminating bacterial endocarditis. Blood culturesgrew coagulase-negative Staphylococcus albus. Fol-lowing treatment with sodium methicillin, 12 g. dailyfor four weeks, he made a good recovery and wasdischarged on 10 December 1964.Three months later he presented in severe cardiac

failure with pulmonary oedema. A soft mitralsystolic murmur was heard. Blood cultures at thistime were negative, but deterioration was rapid, andanasarca developed. On 24 March 1965 he was againcatheterized: the pulmonary artery pressure was 160mm. Hg. The prosthesis was abnormally mobile, dis-playing a rocking movement on fluoroscopy. Grossmitral regurgitation was seen on cineangiography.

Operation was carried out on 29 March 1965 asan emergency measure. The left atrium was re-explored through a right antero-lateral thoracotomy.The prosthesis was found to be detached along theentire length of the mural aspect of the mitral annulus,and there was some soft thrombus present. All thesutures were intact around the Teflon rim but hadbecome detached from the annulus: the prosthesiswas replaced. He was given 2 million units sodiumpenicillin and 1 g. streptomycin daily for eight daysfollowing surgery.He had an uneventful post-operative course, re-

turned to work three months after operation, and hasremained well since.

CASE 2 (C.B. 12535) A man aged 51 had a heavilycalcified mitral valve replaced on 15 April 1965. Hesubsequently developed a right empyema from whichwas cultured a coagulase-positive Staphylococcusaureus. The empyema was drained but he continuedto have a moderate pyrexia with shive&ing and head-aches. Repeated blood cultures grew coagulase-positive Staph. aureus resistant to penicillin. Followingtreatment with sodium methicillin, 12 g. daily, thefever subsided. He was discharged apyrexial fiveweeks after admission, antibiotics being suspendedon discharge.

Five weeks later he was re-admitted in severecardiac failure, and now was noted to have a softmitral systolic murmur. On 4 August 1965 cardiaccatheterization and cindangiography revealed mitralregurgitation, the prosthesis displaying anabnormal movement. He was again pyrexial, and forthis reason operation, despite negative cultures, wasdeferred. When deterioration became rapid, a deci-sion was made to re-operate. He died suddenly on6 August 1965, the morning for which surgery hadbeen scheduled.

FIG. 1. Case 2. Detachment ofprosthesis from annulus ofmural leaflet.

At necropsy the prosthesis was found to bedetached around most of its mural circumference(Fig. 1). The sutures, with knots intact, were hangingloosely from the Teflon rim. Intra-atrial thrombus wasfound on the attached margin of the prosthesis.

CASE 3 (D.T. 10590) A man aged 39 had closedmitral and aortic valvotomies carried out in 1960. On30 November 1964 he underwent aortic and mitralvalve replacement through a median stemotomy.There was gross calcification of the mitral valve inthe region of the lateral commissure and adjacentleaflet tissue, extending into fused, contracted chor-dae. He was given 2 million units of sodium penicillindaily for 10 days following surgery.He was discharged three weeks after operation but

returned two weeks later when a diagnosis of bacterialendocarditis was made following culture of acoagulase-negative Staph. albus from the blood. Hewas treated with sodium methicillin, 12 g. daily,became febrile after four weeks, and remained sountil discharge seven weeks after admission. Threemonths later he had a recrudescence of endocarditis,the same organism being identified. A further courseof methicillin (12 g. daily) was given. During thenext two months severe cardiac failure supervened,infection remained uncontrolled, and at this stagecoagulase-positive Staph. aureus was isolated from theblood. Methicillin (12 g. daily) was continued withouteffect, and the response to massive penicillin dosageand later cephalosporin was equally disappointing.On 22 June 1965 cardiac catheterization and cin&-angiography revealed severe mitral regurgitation withan exaggerated rocking movement of the mitralprosthesis.At operation on 23 June 1965 the left atrium was

explored through a left postero-lateral thoracotomy.The sternal and pericardial adhesions were dense andextremely vascular. The mitral prosthesis was foundto be detached along two-thirds of the mural annulus.

12



j.com/


ownloaded from


Bacterial endocarditis following mitral valve replacement

There were numerous vegetations and soft thrombusmaterial attached to the annulus and adjacent Teflonrim. At the lateral commissure there was a projectingknob of calcium, 4 mm. wide, abutting against theTeflon rim. This calcific knob proved to be situatedin a remnant of contracted adherent chordae. It wasapparent that regurgitation began in this area due tofailure to seat the prosthesis accurately and thenextended due to detachment of the prosthesis. Theprosthesis was removed, and further debridement ofthe annulus was carried out before a new prosthesiswas inserted, but the post-bypass bleeding fromnumerous adhesions and many coronary veins couldnot be controlled and the patient died.

BACTERIAL ENDOCARDITIS WITH INTRA-ATRIALTHROMBOSIS

CASE 4 (R.L. 14906) A man aged 21 had a closedmitral valvotomy via a left thoracotomy in 1959. Hewas well for five years, after which he becameprogressively and severely disabled. Mitral valvereplacement was carried out on 6 September 1965 viaa right antero-lateral thoracotomy. The pericardiumwas densely adherent, and a heavily calcified valvewas excised. Considerable difficulty was experiencedin obtaining a satisfactory bed for the prosthesis inthe region of the lateral commissure, where calciumextended into the myocardium.

Convalescence was delayed by an empyema whichwas drained, and he left hospital with a wound sinusfive weeks after operation. The sinus closed rapidlyand he remained well for five months but thendeveloped recurrent joint pains. He was re-admittedthree hours after the onset of acute left hypochondrialpain, for which a diagnosis of systemic embolism wasmade. On admission he was pyrexial with a remittenttype chart reaching 39.50 C. The prothrombin indexwas 28%, haemoglobin 12 g./100 ml., and W.C.C.18,600/c.mm. Repeated blood cultures grew coagulase-negative Staph. albus sensitive to penicillin. He wastreated with intravenous penicillin, 20 million unitsdaily, but pyrexia continued. Four weeks after admis-sion he developed pulmonary oedema and congestivecardiac failure: repeated electrocardiograms showedatrial tachycardia and ventricular ectopic beats. Thechest radiographs following admission showed pro-gressive cardiomegaly and pulmonary venous con-gestion (Fig. 2).A diagnosis of intra-atrial thrombosis with obstruc-

tion of the prosthesis was made. Re-operation wasconsidered despite what appeared to be prohibitiverisks due to active endocarditis and the dense peri-cardial adhesions of previous post-operative infection.When temporary improvement occurred the idea wasabandoned, but death occurred suddenly on 11 March1966. At necropsy massive intra-atrial thrombosisobstructing the ball valve was found (Fig. 3). Septicinfarcts were present in the brain, liver, spleen, kidney,and mesentery. It was noted that during his lastillness the prothrombin index had been estimated on

alternate days and had fluctuated between 28%" and62%: the majority of the estimations were below50%.

CASE 5 (P.W. 12428) A youth aged 20 with coarcta-tion of the aorta had a grossly incompetent fenestratedmitral valve replaced on 2 April 1965.He was given soluble penicillin, 2 million units

daily, by intramuscular injection for five days afteroperation. On the sixth post-operative day he com-plained of dyspnoea and pleuritic pain in the left side.His temperature was 38.50 C., and the spleen waspalpable. Blood cultures initially were negative.Pyrexia and lassitude continued, and 16 days afteroperation further blood cultures grew coagulase-positive Staph. aureus. Intravenous therapy withmethicillin, 16 g. daily, and fusidic acid, 3 g. daily,was begun on 28 April 1965. Anticoagulants hadbeen started on 17 April 1965, and on 30 April 1965the prothrombin index was 29%. On 1 May 1965acute pulmonary oedema developed and rapidlyprogressed to a fatal termination in a few hours. Atnecropsy the mitral valve prosthesis was covered andloccluded by a mass of friable thrombus. Masses ofGram-positive cocci were seen on a smear of thisthrombus and culture grew Staph. aureus. There wasno evidence of mitral incompetence.

INTRA-ATRIAL THROMBOSIS ASSOCIATED WITH SEPSISWITHOUT POSITIVE BLOOD CULTURE

CASE 6 (G.F. 18181) A man aged 46 was admittedon 9 December 1965 with severe symptoms and signsof mitral incompetence. Two months previously hehad been admitted elsewhere with a diagnosis ofpulmonary embolus and had been on anticoagulantssince that time. On 16 December 1965 a thin in-competent valve was excised and replaced with a ball-valve prosthesis via a left thoracotomy. On the secondpost-operative day severe respiratory tract infectionnecessitated tracheostomy. Subsequently woundinfection and empyema necessitated drainage of thechest on 24 December 1965. Anticoagulants, whichhad been suspended pre-operatively, were begun on27 December 1965. Two days later the patientsuddenly became decerebrate. A diagnosis of cerebralembolus and intra-atrial thrombosis was made, andtreatment with heparin was begun, the prothrom-bin index at this time being 80%. Deterioration con-tinued and death occurred on 31 December 1965. Atnecropsy there was a large intra-atrial thrombusattached to the prosthesis, infarcts were present in thecerebrum, midbrain, and left kidney, and bilateralbronchopneumonia was present. No cultures weretaken.

CASE 7 (M.B. 9103) A woman aged 44 had a mitralvalve replacement on 7 September 1964. The valvewas moderately calcified but was excised in toto; thetricuspid valve was also explored at the time butno procedure was performed: post-operative con-

13



j.com/


ownloaded from


Harry M. Windsor, Paul Fagan, and Mark X. Shanahlan

(h)nGs 2a and b. Case 4. Progressive cardiomegaly and pulmonarycongestion.

14



j.com/


ownloaded from



_ll.', cvalescence was satisfactory. Ten weeks after opera-_ZSF tion she was re-admitted with a cerebral embolus and

transient hemiparesis following a road accident in|

_ 1** l i1which she was thrown against the dashboard. There

was no pyrexia during this three-week stay in hospital,and anticoagulant control was satisfactory. Tricuspidincompetence, however, was more severe. She was

1~Z~~K re-admitted on 2 January 1965 with severe tricuspidincompetence and cardiac failure, and a tricuspidvalve prosthesis was inserted on 28 January 1965,though the left atrium was not explored at this

~~~*,~~~~~operation. The patient was well until six hours post-operatively, when ventricular fibrillation suddenlyoccurred. Over the next five days 280 defibrillationswere necessary, and during this period she becameunconscious. A tracheostomy was performed as thetracheobronchial tree contained thick purulent secre-tions. Soluble penicillin, 2 million units daily, wascommenced pre-operatively and continued until 14February 1965. The state of consciousness slowlyimproved, but pulmonary infection remained severe.

\ Anticoagulants, suspended before operation, were notrecommenced because of the gross inflammation ofthe tracheobronchial tree. On 16 February 1965 a

remittent pyrexia to 39.50 C. began and continuedntil the patient died on 23 February 1965. No blood

cultures were taken. At necropsy the left atrium con-tained a thrombus occluding the mitral prostheticorifice. No cultures were taken at necropsy. The

(a)

ventricle, obstructing ball valve.

(b)

1 5



j.com/


ownloaded from



tricuspid valve prosthesis was normal and suturesappeared to be well clear of the conduction tissue.There was no obvious explanation for the numerousepisodes of ventricular fibrillation other than theatrial thrombus.

CASE 8 (A.S. 7323) A man aged 42 was admittedin uncontrollable cardiac failure on 31 May 1964. In1962 an open mitral annuloplasty had been performedelsewhere. This had been followed by a most com-plicated post-operative course with anuria requiringdialysis, chest and urinary tract infections, empyema,wound dehiscence, and pulmonary embolism. On 25June 1964 a fibrosed incompetent mitral valve wasexcised and replaced with a ball-valve prosthesis viaa right antero-lateral thoracotomy. His immediatepost-operative condition was good, but on the thirdday he became hypotensive and developed a con-solidated right upper lobe, for which tracheostomy wasnecessary. The sputum was thick and purulent andculture grew Pseudomonas pyocyanea. Soluble peni-cillin, 2 million units daily, started before operation,was continued. Oliguria occurred: peritoneal dialysiswas begun on the sixth day and resulted in a satis-factory diuresis on the eighth post-operative day.There was no pyrexia at this time and anticoagulantshad not been given. The wound became infected andseveral subcutaneous collections of pus were evacu-ated. Culture of this pus grew coagulase-positiveStaph. aureus. On the eleventh post-operative day thepatient, when apparently improving, suddenly col-lapsed and could not be resuscitated.At necropsy the left atrium contained a large

thrombus attached around the periphery of theprosthesis and obstructing the orifice. No cultureswere taken. The right atrial appendage also containedantemortem thrombus. There was a large embolusobstructing the right pulmonary artery with infarc-tion of the upper lobe.

DISCUSSION

The hospital mortality in 140 mitral replacementscarried out in this unit was 12%. This includedfour patients who underwent multiple valve re-placement, and two patients operated upon inextremis because of bacterial endocarditis.

Bacterial endocarditis was proved by positiveblood culture six times following replacement.Full information concerning the prosthesis was

available on five occasions either from re-opera-tion or necropsy examination. Mitral regurgita-tion, massive intra-atrial thrombosis, or both were

present in each of these cases. In a further threecases, in which massive intra-atrial thrombosis was

found at necropsy and appeared to be the majorfactor contributing to death, sepsis had been a

feature and endocarditis a possibility. These factsemphasize the association which exists between

infection following mitral valve replacement andthe development of lethal complications.We have attempted to develop' ,and assess a

routine of management which: will lessen theincidence of endocarditis and its associated com-plications, and which will encourage earlier andmore vigorous management of such complicationsshould they occur.

In, this group only two patients were re-operatedupon. One (case 1) is alive and well, the other(case 3) might have survived if an alternativeapproach to the valve, avoiding the adhesions ofprevious surgery, had been used. Case 2, in thelight of present experience (Windsor and Shana-han, 1967), should certainly have had a furtheroperation. The other patients (cases 4, 5, 6, 7, and8) all died, but the local lesion was amenable tocorrection either by removal of the thrombus orreplacement of the prosthesis.We have not removed thrombus from a pros-

thesis as a definitive procedure, but we believe thatto do so is feasible. In the two patients in whomre-operation was performed for regurgitation(cases 1 and 3) friable thrombus and vegetationshad to be removed in the process of replacing theprosthesis. This experience indicates that replace-ment of an infected prosthesis with or withoutthrombus and vegetations is a practical procedure.Our experience in operating in the presence of

bacterial endocarditis (Windsor and Shanahan,1967) where vegetations were present also indi-cates that infection can be controlled provided thenidus is removed and an adequate antibioticregime is used.

MITRAL REGURGITATION In cases 1 and 2 thedetachnment of the prosthesis was due to infection.In case I cardiac catheterization and cineangio-graphy one year after surgery and two monthsbefore the onset of endocarditis failed to revealregurgitation. In case 2 clinical evidence of endo-carditis preceded that of regurgitation.

Local factors likewise influence the associationby predisposing to infection. Failure to removea calcific remnant at the lateral commissureresulted in faulty seating of the prosthesis in case3. This undoubtedly acted both as a site for thecommencement of regurgitation and as a nidus forinfection. In case 9 (Table) calcific remnantswere strongly suspected as the basis for the endo-carditis (Windsor and Seldon, 1964).

Residual calcium in the annulus has constituteda hazard only if it prevented accurate seating ofthe prosthesis. It is, on occasion, impossible toremove all the calcium. To do so in the region ofthe lateral commissure can endanger the circum-

16



j.com/


ownloaded from



flex artery or lead to atrioventricular dehiscencewith resulting haemorrhage or false aneurysm.

In all three cases the prosthesis exhibited theabnormal rocking movement described by Bjorkand Bjork (1966).

INTRA-ATRIAL THROMBOSIS That intra-atrialthrombosis occurred to some degree in all patientswith bacterial endocarditis and has been associatedwith sepsis in three other patients in the seriesindicates that the principal predisposing factor isinfection. Blundell and McGoon (1964) drewattention to this association. They reportedmassive fatal thrombus formation on an aorticball valve in a calf. Blood cultures in this animalwere negative but the thrombus contained numer-ous colonies of Gram-positive cocci. Hughes(1965), in a review of the complications reportedfollowing mitral valve replacement, made refer-ence to fatal obstruction of the prosthesis bythrombosis in two patients, and suggested that, ifthis complication occurred, bacteriological exam-ination of the thrombus and adjacent tissues bemade prior to contamination at necropsy.

In our series of patients atrial thrombosisoccurred early in the post-operative period incases 5, 6, and 8, but in cases 4 and 7 it did notpresent for many months. We believe that in theselatter cases the basis for thrombosis was laid downin the early post-operative period. Anticoagulanttherapy appeared to be adequate in cases 4 and 5but did not prevent the complication. In the otherthree cases it was inadequate. In cases 6 and 7 anti-coagulants, which had been in use prior tooperation, were suspended afterwards, for reasonswhich seemed justified at the time. We nowbelieve that, when infection is present, everyeffort should be m-ade to institute early andadequate anticoagulant therapy and to maintainit, in spite of apparent contraindications.The diagnosis of intra-atrial thrombosis is

difficult, especially in the early post-operativeperiod. In cases 5 and 8 it was completely un-suspected and even in retrospect we believe it wasunrecognizable. In case 7 it was suspected, and incases 4 and 6 a diagnosis was made after theoccurrence of emboli. In case 4 additional clinicalfeatures made a diagnosis of prosthetic valvularobstruction possible. These were the symptomsand signs of pulmonary oedema and congestivecardiac failure, supra-ventricular tachycardia,ventricular ectopic beats, and a chest radiograph(Fig. 2) showing typical changes of pulmonaryvenous congestion.Cindangiography was not done in any of these

patients with massive intra-atrial thrombosis.

Retrospective study of accurate cineangiograms ofpatients from within whose left atria we haveremoved large amounts of clot at the time ofvalve replacement lead us to conclude that thisinvestigation often fails to detect intra-atrialthrombosis. For this reason we are not hopefulof it being of value in detecting post-operativeintra-atrial thrombosis. Furthermore, it is quitepossible that left heart catheterization is contra-indicated in the presence of intra-atrial thrombosis.

Diagnosis will largely hinge on a high degree ofsuspicion in a patient who has had a history ofsepsis or pyrexia and in whom abnormal symp-toms and signs are present.The unpredictability of the event is well illus-

trated by case 5, in which uncomplicated surgerywas followed by infection and massive throm-bosis. It can only be explained by the introductionof infection at operation.

ANTIBIOTICS Antibiotic therapy for patients under-going mitral or aortic valve replacement was notstandardized in this Unit until January 1966. Ourfirst three replacements in March 1963 were notgiven antibiotics; then a series of 30 were givensodium methicillin, 6 g. per day for 10 days,followed by oral sodium cloxacillin for an in-definite period. The next 100 cases were given 2million units of penicillin per day, combined, ifthe need appeared to have arisen, with variousother antibiotics.Over this period there were, in addition to the

cases reported here, four infections followingaortic valve replacement in a series of 100 cases.A nidus of infection was present in three. In thefourth unexplained infection led to death due tomassive thrombosis.The causative organism in each of the cases

reported was either coagulase-negative Staph.albus or coagulase-positive Staph. aureus (Table).Each occurred in three cases. These organisms,although occasionally insensitive in vitro, areundoubtedly on most occasions sensitive in vivo tothe bactericidal action of massive intravenousdosage of soluble penicillin and methicillin (cases1, 2, and 3).The cases presented in this paper all had

different post-operative antibiotics. When bacterialendocarditis developed, various combinations ofmethicillin, penicillin, fusidic acid, cephalosporin,and steroids were used.

In January 1966 a decision was made tostandardize antibiotic therapy. Since that time thefollowing regime has been used (Amoury,Bowman, and Malm, 1966);

17



j.com/


ownloaded from



1. Procaine penicillin, 1,200,000 units, sodiummethicillin, 4 g., and streptomycin, 0-5 g., aregiven 24 hours prior to surgery.

2. For two weeks following surgery, potassiumpenicillin, 20,000,000 units, and sodium methicil-lin, 12 g., are given daily through a right sub-clavian vein catheter. Streptomycin, 1 g. daily forsix days, is given unless there is concern regardingrenal function.

3. At the end of two weeks the cases aredivided into two groups. One is given oral sodiumcloxacillin, 2 g. daily, for three months. The otherreceives no antibiotics.

This regime was adopted for the followingreasons:

(a) to obviate haphazard antibiotic therapy ascan easily occur when many attendants are con-cerned with any one case;

(b) because of the tendency for haphazardtherapy to mask infection, to cause negative bloodcultures, and thereby to delay adequate therapy;

(c) because we believe that prophylactic anti-biotics in large doses are of value;

(d) because of the incidence of endocarditisfollowing wound infection (cases 2 and 4);

(e) because of the unpredictability of intra-atrialthrombosis and the difficulty of its recognition(case 5);

(I) because of the impossibility of maintainingsterile conditions in a modern operating room andthe ever-present risk of introducing infection(case 5).

Since adopting this regime there has not beenone case of bacterial endocarditis or intra-atrialthrombosis in 120 valve replacements so managed.During this period seven patients underwent valvereplacement who, because of drug sensitivity,were not given penicillin or methicillin. One ofthese, an aortic valve replacement for calcificaortic stenosis whose only antibiotic therapy waspost-operative erythromycin, 2 g. daily, developedStaph. albus endocarditis.

We suggest that this regime merely complementsaccurate surgical technique to avoid a nidus withinthe heart and the most stringent adherence toregulations designed to control infection in boththe hospital and the operating room.

RE-OPERATION The question of re-operation inthe patient with bacterial endocarditis and un-controlled mitral regurgitation is no longer debat-able. Surgery alone can rectify the haemodynamicstate. Removal of a thrombus or nidus can controlinfection (case 1).There can be little doubt that surgery is

indicated in the management of intra-atrialthrombosis. Diagnosis is the difficulty. Infection isno contraindication, as persistence of a positiveblood culture may be due to intra-atrial throm-bosis and the clue to its presence (cases 3, 4, and5). It could be that persistence of infection con-stitutes the single most important indication forre-operation.The timing of surgical intervention is a greater

difficulty. It is obvious that, if the infection canbe controlled before operation, then this is desir-able. However, if, after a reasonable period ofantibiotic therapy, infection or pyrexia is stillpresent, surgery should probably not be delayedas death can occur suddenly from regurgitation(case 2), from emboli (case 6), from dysrhythmia(case 7), or from obstruction to the prosthesis(case 4).

REFERENCESAmoury, R. A., Bowman, F. O., and Maim, J. R. (1966). Endo-

carditis associated with intracardiac prostbeses: diagnosis,management and prophylaxis. J. thorac. cardiovasc. Surg., 51,36.

Bjork, L., and Bj5rk, V. 0. (1966). Cineradiography in the post-operative control ofa Starr aortic valve prosthesis. Thorax, 21, 48.

Blundell, P. E., and McGoon, D. C. (1964). The fate of the Starr-Edwards ball valve in the aortic area of calves. J. thorac. cardio-vasc. Surg., 48, 704.

Hughes, R. K. (1965). Complications of Starr-Edwards mitral valvereplacement. Ibid., 49, 731.

Windsor, H. M., and Seldon, W. A. (1964). Initial experiences in thereplacement of the mitral valve. Med. J. Aust., 1, 112.- and Shanahan, M. X. (1967). Emergency valve replacement

in bacterial endocarditis. Thorax, 22, 25.

18



j.com/


ownloaded from


bacterial endocarditis, mitral regurgitation, and intra...

Documents