infectious complicationsafter transplantation › content › thoraxjnl › 38 › 11 ›...

Thorax 1983;38:822-828

Infectious complications after heart transplantationDKC COOPER, RP LANZA, S OLIVER, AA FORDER, AG ROSE, CJ UYS, D NOVITZKY,CN BARNARD

From the Departments of Cardiac Surgery, Medical Microbiology, and Pathology, Groote Schuur Hospitaland University of Cape Town Medical School, Cape Town, South Africa

ABsTRAcr Infection has been the major cause of death and morbidity in patients undergoingcardiac transplantation at Groote Schuur Hospital. Twenty-two (55%) patients suffered at leastone major episode of infection, which accounted for 10 (59%) of the deaths in the first year. Themajor site of origin of infection was the lung, though dissemination was not infrequent. Bacteriaaccounted for 22 (59%) infections; but viral, fungal and protozoal infections were not uncommonand in fact accounted for seven (64%) of the fatal infections. Several unusual causative micro-organisms have been isolated in this group of immunocompromised subjects. There is a higherincidence of infection in patients over the age of 35 years and in patients who did not comply withinstructions and advice.

Infection is the major cause of death in patientsundergoing cardiac transplantation, especially in thefirst few months when immunosuppression isheaviest.' 2 Patients are susceptible to infection by awide range of pathogens, bacterial infection beingby far the most common; other organisms primarilyresponsible include viruses, fungi, and protozoa.3We report our experience with such infections inrecipients of heterotopic heart transplants over thepast seven years.

The patients

From 25 November 1974 to 29 October 1981 44heterotopic heart transplants were performed in 40patients, four patients undergoing retransplantationafter irreversible rejection; all have been followedup for at least five months. Details of the selectionand management of patients have been describedpreviously.2The basic immunosuppressive treatment through-

out this period has consisted of: (1) azathioprine atthe highest tolerated level as judged by the absenceof bone marrow and hepatic toxicity, the mainte-nance dose for adults being 1-5-4-0 mg/kg/day; (2)

Address for reprint requests: Mr DKC Cooper FRCS, Departmentof Cardiac Surgery, University of Cape Town Medical School,Observatory 7925, Republic of South Africa.

Accepted 20 July 1983

methylprednisolone 600 mg given intravenously onthe day of operation and reduced by daily incre-ments of 100 mg until discontinued on the sixthpostoperative day, after which time oral doses of 64mg/day have been administered, reducing to about32 mg/day at three months if the progress of thepatient permits; (3) from 1970 to 1979 equine anti-lymphocyte globulin was given intravenously duringthe first month, in doses related to its effect on T cellrosetting; rabbit antithymocyte globulin has beenused since mid-1979. Acute rejection episodes havebeen treated with three to five daily intravenouspulses of methylprednisolone and short courses ofRATG and/or actinomycin D (200 ,g/day for 3 to 4days).

Attention was paid to the location and eliminationof all sources of infection before transplantation.Patients with active infection or conditions whichmight predispose to infection, such as diabetes mel-litus requiring insulin, were not selected for trans-plantation. Perioperative antibiotic prophylaxis atthe time of transplantation consisted of cephaman-dole nafate 1 g intravenously on induction of anaes-thesia, followed by 1 g intravenously six hourly forfour days. After transplantation observation for clin-ical symptoms and signs of infection was continuous,laboratory investigation urgent and intense, andtreatment aggressive. Specific treatment was with-held until the infecting organism had been positivelyidentified, unless identification proved impossible inlife threatening conditions.

822

on July 23, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.38.11.822 on 1 Novem

ber 1983. Dow

nloaded from

http://thorax.bmj.com/

Infectious complications after heart transplantation

Table 1 Primary cause ofdeath after heterotopic hearttransplantation in 28 patients

Cause ofdeath Months after transplantation Total'3 3-12 > 12 No (%b)

Acute rejection 2 0 1 3 1Infection 7 3 1 11 39Graft atherosclerosis 0 1 6 7 25Other 3 1 3 7 25

Total No(%) 12 (43) 5 (18) 11 (39) 28 (100)

Results

CAUSES OF DEATHOf 40 patients in this series, 28 have subsequentlydied. Survival ranged from one day to over fouryears. Survival of those who remain alive extendsfrom 5 months to over 7 years. The primary causesand timing of death are given in table 1. Infectionaccounted for 11 (39%) of the deaths and was themajor cause of death in the first year after transplan-tation, when it accounted for 10 (59%) of the 17deaths.

OVERALL INFECTIONTwenty two patients (55%) suffered at least onemajor episode of infection. The cause was bacterialin 22 (59%), viral in 8 (22%), fungal in 5 (14%),and protozoal in 2 (5%). Disseminated and pulmo-nary infections accounted for over half of allepisodes.

823

FATAL INFECTIONSThe number and type of fatal infections and theorganisms responsible are shown in table 2. Withtwo exceptions, all were or became disseminated. Atleast seven fatal infections originated in the lungs,two remaining confined there. Only one fatal infec-tion occurred after the first year. Several of theorganisms cultured from these patients were ofdoubtful significance, though determination ofsignificance is a recurring problem in immunosup-pressed patients.

All four fatal bacterial infections occurred withinthree months of transplantation and two occurred inpatients who had undergone retransplantation.Fungal infections in general developed later aftertransplantation. Fatal viral infection was seen bothearly and late.Though bacteria accounted for 59% of all fatal

and non-fatal infections, fungi and virusesaccounted for 64% of all fatal infections (seven of11). Indeed, three of four fungal infections were

fatal, reflecting the great difficulty experienced ineradicating these organisms in immunosuppressedpatients.Gas gangrene of the leg occurred in one patient,

amputation being unsuccessful in delaying death.The source of the Clostridium perfringens was

uncertain, but possibly followed an endomyocardialbiopsy performed through the femoral vein a fewdays previously. At necropsy the organism was alsogrown from the lungs.

Table 2 Fatal infections in 11 patients after heterotopic cardiac transplantation

Site of Organisms cultured or Clinical course Time ofdeathorigin isolated* (days after

transplantation)Bacterial ? Leg Leg: ?Followed endomyocardial biopsy performed 76(4 patients) Clostridiun perfringens through femoral vein; gas gangrene right leg;

amputationLungs Sputum: (Retransplant) Adult respiratory distress syndrome; 24

Staphylococcus aureus tracheostomy; pneumonia; septicaemiaKlebsiella pneumoniaeBlood:Pseudomonas aeruginosaEscherichia coli

Gastro- Stools and blood: Diarrhoea; lung abscess; bronchopleural fistula; 64intestinal tract Salmonella group B empyema

Blood:(Serratia marcescens)(E coli)(Enterobacter sp)

Lungs E coli (Retransplant) Pneumonia 93Enterobacter spPeptococcus

Fungal(4 patients) Lungs Aspergillus famigatus Pneumonia; dissemination 74

Lungs Aspergillus sp Pneumonia; dissemination 234Lungs Petrieuidium boydii Pneumonia; dissemination 233Lungs AspergiUlus fumigatus Pneumonia; dissemination 527

Viral ?Blood Cytomegalovirus Cytomegaloviraemia 61(3 patients) ?Blood ?Herpes simplex virus Infection of oesophagus, tongue, and skin; encephalitis 220

Lungs Unidentified Pneumonia 35

*The significance of organisms in parentheses remains uncertain.



/T


ber 1983. Dow

nloaded from


Cooper, Lanza, Oliver, Forder, Rose, Uys, Novitzky, Barnard

Salmonella infection was responsible for onedeath. The patient developed diarrhoea andpneumonia and had positive blood and stool cul-tures. A lung abscess formed and ruptured into thepleural cavity, resulting in an empyema.

Fungal pneumonia has been a considerable prob-lem, particularly with Aspergillus sp. Petriellidiumboydii was found at necropsy in one patient, thefindings suggesting primary infection of the lungswith haematogenous dissemination to the brain (fig1), kidneys, and skin.4

Fatal viral infections were thought to have occur-red in three patients, though the evidence wasinconclusive in two. In one there were patho-gnomonic inclusion bodies of cytomegalovirus inmany organs at necropsy. In a second patient inclu-sion bodies of a herpetic type were seen in severalorgans, but although the patient died with clinicaland histological encephalitis no virus was isolatedfrom the cerebrospinal fluid or from the brain. Thethird patient developed a severe pneumonia, thoughno micro-organisms could be cultured; at necropsythere were histological features which supported adiagnosis of viral pneumonia. Evidence of herpes-virus infection and of cytomegalovirus infectionwere present in about 40% of patients dying ofother causes.4NON-FATAL INFECTIONSNon-fatal infections are listed in table 3. Again, thelung was the most common site of infection, being

affected in a quarter of the cases. Certain cases areworthy of comment.

Eighteen months after heterotopic heart trans-plantation endocarditis caused by Staphylococcusaureus developed on an aortic valve prosthesis in thepatient's own heart; antibiotic treatment was unsuc-cessful. The prosthesis was removed, the recipient'sleft ventricle being partially resected and excludedfrom the circulation. The patient made an excellentrecovery, since when he has been symptom free;over seven years later he is our longest survivor.5Two patients developed severe postoperative

sternal wound infections, both of which wereextremely difficult to eradicate. In one the originalinfecting organism was Staphylococcus aureus, butsubsequently seven further organisms were grownfrom discharging sinuses in this tegion. During thecourse of two years the patient underwent nineoperative procedures in an attempt to eradicate thisinfection. Chronic discharging sinuses remained forsome 36 months but then spontaneously healed, thepatient dying of chronic rejection in his fifth post-operative year. In the second patient the originalinfecting organism remains uncertain, thoughEnterobacter sp and Staphylococcus epidermidiswere the first organisms cultured from the pus exud-ing from the wound. Eight further organisms, includ-ing Aspergillus sp and Clostridium perfringens, werecultured subsequently. The patient underwent threesurgical drainage procedures before the infection

Fig 1 Area of necrosis in the brain containing septate hyphae of Petriellidium boydii; the lattershow multiple vesicular expansions. (Methenamine silver, x 500.)

824



/T


ber 1983. Dow

nloaded from


825Infectious complications after heart transplantation

Table 3 Non-fatal infections after heterotopic cardiac transplantation

Site of Organisms cultured or isolated* Months Clinical courseorigin after

trans-plantation

Bacterial Blood(15 patients) Endocardium

Endocardium

LungsLungsLungsLungsSternum

SternumGastrointestinaltractGastrointestinaltractGastrointestinaltractSubcutaneoustissue (leg)Subcutaneoustissue(perineum)Subcutaneoustissue(perineum)

Mycobacterial Lungs(3 patients)

Lungs

Bone

Fungal(1 patient)Viral(5 patients)

Protozoal(2 patients)

Meninges

Face and scalpChest wallChest wallEar

Cervix

Myocardium

Lungs

Acinetobacter spStaphylococcus aureus

Salmonella dublinBaciUlus cereusHaemophilus influenzaeHaemophilus influenzaeMtreptococcus pneumoniaeUnidentifiedStaphylococcus aureus(+ 7 other organisms)Multiple organisms (10)Shigefla dysenteriae

Salmonella sp

Salmonella sp

Citrobacter sp

Escherichia coli

Anhaemolytic streptococcus

Mycobacterium kansasii

Myco tuberculosis

Myco haemophilum

Cryptococcus neoformans

Herpes zosterHerpes zoster t clinical diagnosisHerpes zosterJHerpes simplex (virusidentified in isolate)Herpes, virus unidentified(clinical diagnosis)roxoplasma gondii

Pneumocystis carinii

3 Septicaemia*18 Infection on prosthetic valve; prosthesis

removed; resolved25 Septicaemia; embolic brain abscess; infected

saddle embolus*6 Pneumonia*

1 2 Pneumonia*3 Pneumonia*

37 Pneumonia*1 Nine surgical drainage procedures; chronic

discharging sinus for 36 months; resolved1 Three surgical drainage procedures; resolved1 *

8 *

5

9

Extensive abscesses of both legs; surgicallydrained; resolvedPerineal abscess following sigmoidoscopy andbarium enema; surgically drained; resolved

54 Perineal abscess; surgically drained; resolved

9 Chronic low grade pneumonia, treated butremained unresolved; died at 14 months fromdisseminated Kaposi's sarcoma

20 Chronic low grade pneumonia; refused treatment;died at 24 months from cerebral embolus

48 Chronic discharging sinuses at left wrist, leftthigh, and left lower leg; low grade purulentinfective arthritis left knee; not treatedsystemically; died at 54 months from chronicrejection

56 *

369

3560

18

ResolvedResolvedResolvedResolved

Resolved

3 Seen on endomyocardial biopsy specimen;?transferred from donor; clinicallyasymptomatic*

51 *

*Resolved with treatment.was eradicated; he remains alive and well almostseven years later.

Myobacterial infection has become an importantproblem and has proved difficult to treat adequatelyin these immunosuppressed patients. One patientdeveloped a cavity in the right lung, seen on a chestradiograph 20 months after transplantation, butrefused further investigation and treatment, dyingfour months later from a cerebral embolus. Atnecropsy Mycobacterium tuberculosis was isolatedfrom the cavity. In a second patient chronicMyco haemophilum infection of limb bones occur-red and persisted for several months before thepatient died of chronic rejection almost four and ahalf years after transplantation. Myco kansasii was

responsible for a chronic lung infection withempyema in a third patient, who subsequently died14 months after transplantation from disseminatedKaposi's sarcoma.Minor herpes simplex infections are very common

in immunosuppressed patients, but herpes zoster hasalso occurred on three occasions, though it has notbecome disseminated.Toxoplasmosis was found on endomyocardial

biopsy of both donor and recipient hearts in onepatient (fig 2), though at no time did the patientdevelop clinical signs of disseminated infection.6While the infection could have been opportunistic,two other possibilities must be considered. Firstly,the patient's idiopathic cardiomyopathy may have



/T


ber 1983. Dow

nloaded from


Cooper, Lanza, Oliver, Forder, Rose, Uys, Novitzky, Barnard

Fig 2 Donor heart biopsy specimen six weeks after transplantation; the swollen myofibre in thecentre ofthe picture contains many dark staining toxoplasmas. (Haematoxylin and eosin, x 500.)

been due to toxoplasmosis; an absence of toxop-lasma antibodies before transplantation and a raisedtitre after transplantation (1/1024), however, argueagainst prior infection in the recipient. Secondly, thedonor could have had toxoplasmosis, the organismbeing transferred with the donor heart. The donordid have a mildly raised antibody titre (1I256)7 totoxoplasma so the recipient could have beeninfected from the donor, although the organism wasnot seen in endomyocardial biopsy specimens takenin the first month. After treatment, the organismwas not seen again on biopsy and the recipient'santibody titres fell to lower levels.That immunosuppressed patients remain at risk

from serious infection even when receiving lowdoses of drugs is illustrated by one of our patients,who during the fifth year after transplantationdeveloped Pneumocystis carinii pneumonia followedshortly by cryptococcal meningitis; his local physi-cians successfully treated both episodes.

Discussion

In patients undergoing heterotopic heart transplan-tation at Groote Schuur Hospital infection is themajor cause of death during the first year aftertransplantation. The Stanford group has reported aneven higher proportion of deaths from infection(58%). Whether the greater incidence of fatal

infection at Stanford is related to differences in theimmunosuppressive regimen used at the two centresis uncertain, though possibly a rather greateremphasis has been placed on rabbit antithymocyteglobulin at Stanford than in Cape Town.' 2Our impression has been that during the first

three months after transplantation infection notinfrequently follows increased immunosuppressivetreatment in patients undergoing acute rejectionepisodes, particularly if such episodes are repeatedor prolonged. After three months infectiousepisodes would appear to be less clearly related tothe level of immunosuppressive treatment.The lung was clearly the most common site for

infection in our series, though not infrequently dis-semination took place (tables 2 and 4). A higherincidence of purely pulmonary infections has beenreported from Stanford with a lower incidence ofdisseininated infection (table 5).3 Minor differencesbetween the two series include a greater number ofurinary tract infections at Stanford and of boneinfections at Groote Schuur. None of these differ-ences would appear to be related to the differenttype of operation performed at Stanford (orthotopicas opposed to heterotopic transplantation).The causal organisms responsible for infectious

episodes at the two centres are very similar, bacter-ial infection accounting for about 60% in bothseries.3

826



/T


ber 1983. Dow

nloaded from


Infectious complications after heart transplantation

Table 4 Site ofpresentation ofnon-fatal infections afterheterotopic cardiac transplantationSite No (%) of No of

episodes patients

Blood 1 4) 1Lung 7 26 6Skin 4 15 4Bone 3 12 2Gastrointestinal tract 3 12 3Subcutaneous tissue 3 12 2Endocardium 2 7 2Myocardium 1 4 1Brain or meninges 1 4 1Cervix 1 4 1

Total 26 (100)

Table 5 Comparison ofsites ofinfection at Groote SchuurHospital (GSH) and Stanford Medical Center (SMC)

Site ofinfection GSH (%o) SMH (%o)

Disseminated 29 12Lungs 24 45Bone 8 1Central nervous system 3 4Urinary tract 0 7Other 36 32

Total 100 100

Several unusual organisms were seen in the pres-

ent series. Petriellidium boydii is a rare fungus whichis usually responsible for localised soft tissuemycetomas. When our patient died of this infectiononly three previous reports of disseminated petriel-lidosis had been published. Steroids had beenadministered to all three patients, in one as

immunosuppression for a renal allograft and in twofor lupus nephritis.8Three patients developed mycobacterial infec-

tions, none of which were fatal, though all threepatients died while still infected. The susceptibilityof immunocompromised patients to infection withMyco tuberculosis has long been recognised9 and thenumber of reports of infection caused by mycobac-teria other than Myco tuberculosis is growing.9-12Myco haemophilum was first isolated relativelyrecently and has subsequently been described as an

infective agent in a small number of immunosup-pressed patients. 12 Myco kansasii is a relativelycommon cause of pulmonary disease in certaingeographical areas;'3 it has been described previ-ously in immunosuppressed patients and in thosewith malignancy and haematological disorders,'0 inwhom it is associated with a high death rate.

In the present series the incidence of death due toinfection was five times greater in patients over theage of 35 years than in the younger patients (34% v

7%)(p < 0-05). There were also significantly more

827

infectious episodes in patients with ischaemic heartdisease as their underlying cardiac condition (22episodes in 29 patients) than in patients with car-diomyopathy (five in 17 patients)(p < 0.01).'14Eighty per cent of the episodes of acute pneuinoniawere in patients with ischaemic heart disease. Thehigher incidence of infection in ischaemic heart dis-ease patients may well be related to their greatermean age (41 years compared with 30 years forthose with cardiomyopathy).We found a high incidence of infectious complica-

tions in patients who for one reason or another didnot comply with medical advice and instructions. Onoccasion this was due to a mild depressive disorder,but more frequently it was due to a casual approachor even irresponsibility on the part of the patient.The possible reasons for non-compliance will be dis-cussed elsewhere.'5The operation of heterotopic heart transplanta-

tion requires insertion of a graft of prosthetic mater-ial, such as Dacron, between donor and recipientpulmonary arteries. The presence of prosthetic mat-erial in an immunosuppressed patient might beassociated with a risk of infection but so far no suchcomplication has occurred.

Infection remains the major cause of death and amajor cause of morbidity during the first year aftertransplantation. If the patient is to be given the bestchance of long term survival, the clinician must beobservant for symptoms and signs of infection, clini-cal investigation in suspicious cases must be urgentand thorough, laboratory investigation must includea search for unusual micro-organisms not frequentlymet in non-immunocompromised patients, andtreatment must be early and aggressive.

The authors thank the many members of the medi-cal, nursing, and paramedical staff of Groote SchuurHospital and the University of Cape Town MedicalSchool who have contributed towards the care ofpatients undergoing transplantation.

References

'Baumgartner WA, Reitz BA, Oyer PE, et al. Cardiachomotransplantation. Curr Probi Surg 1979;16:3-16.

2 Barnard CN, Barnard MS, Cooper DKC, et al. The pres-ent status of heterotopic cardiac transplantation. JThorac Cardiovasc Surg 1981;81:433-9.

'Jamieson SW, Oyer PE, Reitz BA, et al. Cardiac trans-plantation at Stanford. Heart Transplantation1981 ;1:86-92.

4 Uys CJ, Rose AG, Barnard CN. The pathology ofhuman cardiac transplantation. S Afr Med J1979;56:887-96.

5Losman JG, Curcio A, Barnard CN. Normal cardiacfunction with a hybrid heart. Ann Thorac Surg1978;26: 177-84.



/T


ber 1983. Dow

nloaded from


Cooper, Laniza, Oliver, Forder, Rose, Uvs, Novitzkv, Barniardl6 Rose AG, Uys CJ, Novitzky D, et al. Toxoplasmosis of

donor and recipient hearts following heterotopic car-diac transplantation. Arch Pathol Lab Med1 983;107:368-73.

7Beverley JKA, Beattie CP. Standardisation of the dyetest for toxoplasmosis. J Clin Pathol 1952;5:350-3.

8 Walker DH, Adamec T, Krigman M. Disseminated pet-riellidosis (allescheriosis). Arch Pathol Lab Med1 978;102: 158-60.

9 King EQ, Johnson JB. Batten GS, et al. Tuberculosisfollowing cortizone therapy. JAMA 1951:147:238-41.

"1 Fraser DW, Buxton AE, Naji A, etal DisseminatedMycobacterium kansasii infection presenting as cel-lulitis in a recipient of renal homografts. Am RevRespir Dis 1975; 112: 125-9.

Graybill JR, Silva J, Fraser DW, et al. Disseminated

myobacteriosis due to Mycobacterium abscessus intwo recipients of renal homografts. Am Rev Respir Dis1974,109:4-10.

2 Mezo A, Jennis F, McCarthv SW, et al. Unusualmycobacteria in five cases of opportunistic infections.Pathology 1979;1 1:377-84.

13Resnikov M. Atypical mycobacteria, their classification,identification and etiological significance. Med J Aust1970;i:553-7.

Lanza RP, Cooper DKC, Boyd ST, et al. A comparisonof patients with ischemic, myopathic and rheumaticheart disease as cardiac transplant recipients. AmHeart J (in press).

Cooper DKC, Lanza RP, Nash ES, et al. The problem ofnon-compliance in recipients of cardiac transplants.Heart Transplantation (in press).

828



/T


ber 1983. Dow

nloaded from


infectious complicationsafter transplantation › content › thoraxjnl › 38 › 11 ›...

Documents