Prospective BK JC recipients · Study ofhumanpolyomaviruses BKandJCandcytomegalovirus in renal transplant recipients a later investigation, this infection was associated withureteric

J Clin Pathol 1984;37:578-586

Prospective study of the human polyomaviruses BKand JC and cytomegalovirus in renal transplantrecipientsSD GARDNER,* EFD MACKENZIE,t C SMITH,t AA PORTER*

From the *Virus Reference Laboratory, Central Public Health Laboratory, Colindale Avenue, LondonNW9 SHT, and the tDepartment ofPathology, Southmead General Hospital, Westbury on Trym, BristolBSIO 5NB

SUMMARY Forty eight renal transplant recipients were investigated prospectively for evidence ofinfection with the polyomaviruses BK and JC and cytomegalovirus. An active polyomavirusinfection was shown in 31 patients (65%) and cytomegalovirus in 30 (62.5%). Half of the BKand JC virus infections occurred within the first three months after transplantation compared with93% of the cytomegalovirus infections. Very late polyomavirus infections two or more years afterthe transplant were also shown.

Cytology was useful in identifying polyomavirus but not cytomegalovirus infections, and 21(68%) of the 31 polyomavirus infected patients excreted inclusion-bearing cells.Only three patients had symptoms possibly associated with the polyomavirus infection. One

patient with BK virus infection developed ureteric stenosis and a second patient had malaise andvomiting. One patient with JC virus infection developed pericarditis and effusion. Renal functionbecame impaired at the time of the polyomavius infection in eight patients (26%) and uretericobstruction and pericarditis developed in two patients treated with methyl prednisolone forpossible rejection. At the end of the study 25 of the 31 polyomavirus infected patients (81%) hadfunctioning renal grafts.The detection of polyomavirus infection is important as increased immunosuppression needs to

be avoided to prevent possible complications such as ureteric stenosis in transplant recipients.

Viral infections in renal transplant recipients havebeen well documented as occurring in the post-transplant period. In particular, infections withcytomegalovirus and herpes simplex and varicellazoster viruses can cause serious problems. It hasbeen unclear whether these infections initiate rejec-tion episodes or have an adverse effect on renalfunction and graft survival. Recent studies have sug-gested that patients with cytomegalovirus infectionand an associated viraemia may have circulatingimmune complexes which deposit in the kidneycausing glomerulitis and deterioration of renal func-tion.'

Investigations have been done by various workersin an attempt to relate the time of onset of viralinfection to the transplant operation and also todetermine whether the infection is due to reacti-

Accepted for publication 25 January 1984

vated latent virus or arises from exogenous infec-tions introduced via the donor kidney or blood.Warrell and her colleagues2 showed in a prospec-tive study that most primary infections withcytomegalovirus occurred within the first twomonths after grafting and that the likely source ofthese infections was the donor kidney. Secondaryinfection in the seropositive patients occurredindependently of the antibody status of the kidneydonor.

In 1971 a new virus, a human polyomavirus, wasisolated from the urine of a renal transplant reci-pient and named BK virus (the initials of thepatient).3 Subsequent work in England with BKvirus showed that 70% of renal transplant recipientswere seropositive at the time of the transplant oper-ation and that just over half of them had serologicalevidence of an active infection after transplant.4 Inthe original patient BK and in four other patients in

578

on February 20, 2021 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.37.5.578 on 1 May 1984. D

ownloaded from

http://jcp.bmj.com/

Study ofhuman polyomaviruses BK and JC and cytomegalovirus in renal transplant recipients

a later investigation, this infection was associatedwith ureteric stenosis and recognised because of thelarge number of inclusion-bearing cells seen in theurine.5 6An antigenically distinct but similar

polyomavirus, JC virus, was also isolated in 1971from the brain of a patient with Hodgkin's diseaseand the rare demyelinating disease progressive mul-tifocal leucoencephalopathy (PML).7 PML isuncommon and is seen as a late complication inpatients with altered immunity subsequent to dis-eases such as chronic lymphocytic leukaemia,Hodgkin's disease, lymphosarcoma and sarcoidosis.Ten cases of PML have been reported in renaltransplant recipients.8'0 JC virus has beenidentified in 44 cases of PML and is considered themain agent of this disease."I Silent infection with JCvirus is common in children and adults, unassociatedwith any neurological disease, and the virus has beenfound in the urine of renal transplant recipients'2 13and also pregnant women.'4A prospective study of renal transplant recipients

was started in 1977 in an attempt to determine: (a)whether primary or secondary infections with BKand JC viruses were associated with any clinical ill-ness or effect on renal function; and (b) the time ofthese infections in relation to the transplant opera-tion. The patients were also studied for evidence ofcytomegalovirus infection in order to be able toeliminate any possible effects due to this virus.

Material and methods

STUDY GROUPDuring the three year period from December 1977until September 1980, 65 patients received kidneytransplants and initially entered the study. Thirteenof them, however, suffered acute rejection andtransplant nephrectomy within four weeks of thetransplant operation; three patients died within thesame period; and one patient returned to Greeceand was not followed further. These 17 patientswere excluded from the investigation.The study group therefore consisted of 48

patients, with a mean age of 37 years (range 11-55years). Twenty nine patients were male and 19female. Forty four patients were recipients of a firstkidney transplant and four were recipients of a sec-ond graft. Only two patients were given kidneysfrom living related donors, whereas 46 receivedcadaver kidneys. Serial serum samples wereobtained from all patients. These were takenimmediately before the transplant operation, twiceweekly during the first few weeks, then at increasingintervals of two weeks, one month, and three

months, depending on when they attended the out-patients department or were readmitted to hospital.Urine samples for cytological examination wereobtained as described below. The patients were fol-lowed from two to 46 months (mean period 23months).A flow sheet was kept on each patient by the renal

physicians. Information relating to temperature,white cell count, renal function, blood glucose con-centration, immunosuppressive therapy, and com-plications was extracted from .the data available tothe authors, and this was correlated with antibodyresponse and virus excretion.

IMMUNOSUPPRESSIONThe transplant operations were done by twosurgeons who used different regimens forimmunosuppression. All the patients received2 mg/kg azathioprine daily. Patients operated on bysurgeon A were given 300 mg of hydrocortisone and60 mg of prednisolone on day 1 after the operationand 100 mg of hydrocortisone and 60 mg pred-nisolone on day 2. Thereafter they were given 30-60 mg of prednisolone daily with a total dose of630 mg given during the first 14 postoperative days.Surgeon B's patients received 30-200 mg pred-nisolone daily with a total dose of 1140 mg pred-nisolone given during the first two weeks. Methylprednisolone (Solu-Medrone) was used by bothsurgeons to treat rejection episodes.

CYTOLOGICAL INVESTIGATIONSSpecimens of urine were collected from patientsafter transplantation on a daily basis in the post-operative period and thereafter at every readmissionto hospital and outpatient visit. They were screenedfor evidence of rejection and the presence ofinclusion-bearing cells associated with virus infec-tion. The specimens were processed in the labora-tory using a cytocentrifuge and three slides wereobtained, two stained with Papanicolaou and onewith methyl green pyronine, the latter being used toidentify pyronine positive lymphocytes. Whereinclusion-bearing cells were identified furtherspecimens of urine and serum were sent to Colindalefor virus isolation and antibody studies.

VIROLOGICAL INVESTIGATIONSSerologyAntibody to BK and JC viruses was studied bymeans of a haemagglutination inhibition test. A mic-rotechnique was employed using 16 units of eachviral antigen.'4 Serum samples were considered posi-tive if a haemagglutination inhibition titre of 20 ormore was found. IgM antibodies against BK and JCviruses were measured by haemagglutination inhibi-

579



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

580

tion after serum samples were fractionated on suc-

rose density gradients.An eight fold or greater rise in IgG antibody was

considered evidence of infection and this was

confirmed when possible by the demonstration ofpolyomavirus particles in the urine by electronmicroscopy or isolation of virus, or both.Antibody to cytomegalovirus was determined by a

complement fixation test using the Rawles strain ofcytomegalovirus. Only patients with antibody titresof <2 were considered seronegative and susceptibleto a primary infection.

Electron microscopySelected urine samples were examined by electronmicroscopy. Three millilitre volumes were cen-

trifuged at 20 000 rpm for 1 h and the resultingpellets were diluted in distilled water, mixed with0-05% bovine plasma albumin, and negativelystained with 3% phosphotungstic acid at pH 6.3. Adrop was applied to a 400 mesh formvar-carbon-coated grid and examined at a magnification ofX60 000 in an AEI EM 6B electron microscope.

Virus isolationThe urine samples were inoculated into human fetalbrain cells, human embryo kidney cells, and humanembryo lung fibroblast cells for the isolation of JCvirus, BK virus, and cytomegalovirus respectively as

described previously.'4 A minimum total incubationtime of three months was achieved before cultureswere considered negative for BK and JC viruses.The polyomaviruses were identified by the haemag-glutination inhibition method using specific antisera.

Results

CYTOLOGYDuring the study 3497 urine specimens were

examined cytologically and inclusion-bearing cellswere seen in 327 (9%). Twenty one (44%) of the 48patients were identified as excretors.The inclusion-bearing cells were easily identified

in the cytological preparations. Invariably they werelarger than normal urothelial cells. The nuclei were

mainly of three types; those with an opaque groundglass appearance; those with only the nuclear borderof condensed chromatin remaining; and, rarely,those with the classic "owl's eye" appearance withthe hyperchromatic inclusion centrally placed.

VIROLOGYAn active polyomavirus infection, diagnosedserologically or by electron microscopy or virus iso-lation, or both, was present in 31 (65%) of the 48renal transplant recipients. Twenty one patients

Gardner, MacKenzie, Smith, Porter

were infected with BK virus, 17 with JC virus, andseven of them apparently had infections with both.

BK virusNine hundred and six serum samples from the 48patients were tested for BK virus antibody. Only sixof the 48 patients (12%) were seronegative beforethe transplant operation (Table 1). Two of the sixdeveloped BK virus antibody after the operation(titres 1280 and 20 480), but only one wasconfirmed as a primary infection. BK virus was iso-lated from the urine of this patient six weeks aftertransplantation, when the antibody titre was lessthan 10. Two weeks later the antibody level was20 480. BK virus was not isolated from subsequenturines. The second patient may have been respond-ing anamnestically to a concurrent primary infectionwith JC virus or could have had a double infection(Fig. 1). Details of this patient (patient B) are givenlater.Nineteen of the 42 seropositive patients (45%)

had eight fold or greater increases in BK virus anti-body titre and experienced secondary infection(Table 1). The antibody rise was substantial (32 foldor more) in 17 patients. Seven of the 19 patients alsohad rises in JC virus antibody but in three this mayhave been an anamnestic response.

Urine samples were examined from 14 of the 19patients and infection was confirmed in eight.Polyomavirus particles were seen in the urine ofseven patients and BK virus was isolated from fiveand JC virus from one. One other patient wasexcreting BK virus.

Thirteen urine samples were examined from thesix patients excreting BK virus. The virus was pres-ent in the urine between five and eight weeks afterthe transplant operation in four of the six excretors;between 12 and 17 weeks in one; and at 34 weeks inanother. In all six patients BK virus was first isolatedat a time when the BK virus antibody titre was low.Subsequently there was a greater than 100 fold

Table 1 Serological findings in 48 renal transplantrecipients

BK virus JC virus CMV

Number of patientsseronegative beforetransplantation 6 (12%) 22 (46%) 19 (40%)Number of seronegativepatients withpnmary infection 2 (33%) 5 (23%) 6 (32%)Number of seropositivepatients with secondaryinfection 19 (45%) 12 (46%) 24 (83%)Number of seropositivepatients with stable titres or afall in antibody 23 (55%) 14 (54%) 5 (17%)

CMV = cytomegalovirus.



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Study ofhuman polyomaviruses BK and JC and cytomegalovirus in renal transplant recipientsTransplant 2 1 InclusionsTN IMP mJcv

InclusionsMP ' I

4 10 36 52 76 100 108 116. 124 132 144 152Weeks after transplant operation

Fig. 1 Patient B. Serological response to polyomavirus BK and JC (HI) and cytomegalovirus (CF) after two renaltransplant operatons and treatment with methyl prednisolone (MP). HI = haemagglutination inhibition; CF = complementfixation. TN = transplant nephrectomy.

Table 2 BK virus infection related to time of the renal transplant operation and clinical responsef~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Type of Time relatve No of Clinical features Other virus infectioninfection to transplant cases (no. ofpatients) during same period

Primary (2 patients) 10 days 1 Pyrexia JC virus6 weeks 1 None

Secondary (19 patients) 5-8 weeks 9 (47%) Pyrexia (2) CMVPyrexia, respiratory 1 CMVUreteric obstruction 1 CMVPyrexia, malaise, vomiting (1)None (4)

3-5 months 6 (32%) None (6)8 months - 4 (21%) None (3)

11 months t24 months r17 months J Pyrexia, influenza-like

illness, respiratory (1) CMV


increase and this high level of antibody was main-tained thereafter.The patient infected with JC virus excreted virus

intermittently from 14 months to three years afterthe transplant operation. An increase in antibodywas not shown before two years, however, whenthere was a gradual slight increase in antibody (eightfold) to both JC and BK viruses.The secondary infection was not confirmed in six

patients either by electron microscopy or virus isola-tion: the urines were not collected at the optimumtime in four of them.

Just under half (42%) of the study group hadactive infections with BK virus after transplantationand 16 of the 21 patients had inclusion-bearing cellsin the urine.

Time of the BK virus infection related to thetransplant operation and the clinical responseThe definite primary BK virus infection occurred sixweeks after the transplant operation and nearly half

of the 19 secondary infections occurred between fiveand eight weeks. Of the remaining 10 secondary

Trarnplant

I Inclusxons

- 2IX)

2 (512)tLU

0 (128)

(32)

< (8)

(2)ic

8 16 24 32 40 48 56 64 72 80 88 96 104Weeks after trcnsplant operation

Fig. 2 Patient F. Serological response to polyomavirus BKand JC (HI) and cytomegalovirus (CF) after a renaltransplant operation and treatment with methylprednisolone (MP). HI = haemagglutination inhibition; CF= complement firation.

I Transplant 1

I MP

I

IL(-0)

a

581



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

582

infections six were recognised between three andfive months and four were late infections at 8, 11,17, and 24 months respectively (Table 2). With theexception of one patient (patient F) who developedureteric stenosis, the clinical features present at thetime of the BK virus infection were minimal. Norecognisable illness was associated with the primaryinfection and no symptoms were apparent in 13 ofthe'19 patients with secondary infections.

Six patients had symptoms temporally related tothe BK virus infection. Five of them experiencedfever, which in one patient was accompanied by aninfluenza-like illness and in another by respiratorysymtoms. Four of the five patients also had concur-rent cytomegalovirus infection, however, which mayhave accounted for most if not all of the symptoms.Only two patients therefore may have had a clini-

cal response to the secondary BK virus infection.One of them experienced fever, malaise, and vomit-ing in the second month after transplantation andthe other, patient F, developed ureteric stenosis.Details of the laboratory findings in the latter caseare shown in Fig. 2. In 1977 patient F received a firstrenal transplant and was treated for early rejectionat eight days with methyl prednisolone. Five weekslater virus inclusion-bearing cells were seen in theurine. At this time his renal function deterioratedand he was again treated with methyl prednisolonefor possible rejection. Excretory urography showedthat the donor ureter was partially obstructed and a

renal biopsy showed no evidence of acute rejection.In view of this finding and because of the concurrentinfection with BK virus, further acute rejectiontherapy was withheld. The BK virus antibody con-centrations rose sharply, the inclusion-bearing cellsdisappeared from the urine, and his renal functionrecovered. The patient remained well with a func-tioning graft four years after the operation. At thesame time as the BK virus infection, hiscytomegalovirus antibody titre rose from 64 to 512;four weeks later he developed JC virus antibody(Fig. 2), having been previously seronegative. The


JC virus antibody was transient and he eventuallyreverted to being seronegative, but IgM antibody toboth BK and JC viruses was detected after the BKvirus infection.

JC virusSix hundred and fifty one serum samples from the 48patients were tested for JC virus antibody. Twentytwo of the 48 patients (46%) had no antibodybefore the transplant operation and five of theseronegative cases (23%) subsequently had primaryJC virus infections (Table 1). The increase in titrewas 16 fold in two cases and more than 100 fold inthree. Two further patients also seroconverted, butthe JC virus antibody was transient and most likelyto have been an anamnestic response to BK infec-tion. Urines were examined from four of the fivepatients with primary infections and polyomavirusparticles were seen in two of them by electron mic-roscopy 16 weeks and 10 months after transplanta-tion. JC virus was isolated from the former urine butnot from the latter, probably because the virus parti-cles had been coated with antibody.Twelve of the 26 seropositive patients (46%) had

eight fold or greater increases in JC virus antibody,suggesting a secondary infection, but six of themalso had rises in antibody to BK virus and in mostcases this was the major response. Nevertheless, intwo patients known to be excreting BK virus bothBK and JC virus specific IgM was present in theirserum. It was not possible to be sure in thosepatients with serological evidence of infection withboth viruses whether these were dual infections oranamnestic responses.

Seventeen urine samples were examined fromseven of the 12 patients. Polyomavirus particleswere seen by electron microscopy in the urine of fivepatients. JC virus was isolated on two occasionsfrom one of them at 14 months and 31 months aftertransplantation and BK virus from three. Anunidentified polyomavirus was isolated 16 monthsafter transplantation from the other patient; this

Table 3 JC virus infection related to time ofthe renal transplant operation and clinical response

Type of Time relative No of Clinical features Other virus infectioninfec on to transplant cases (no ofpatients) during same period

Primary (5 patients) 10 days 1 Pyrexia BK virus15 and 16 weeks 2 None11 and 12 months 2 None

Secondary (12 patients) 4-7 weeks 6 (50%) Malaise, vomiting 1 CMVPyrexia, ?viral pericarditis 1)None ( M

10-15 weeks 3 None 311 months 3 None 314 months30 months




j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


virus was probably JC virus from its restrictedgrowth pattern in cell culture.About one third of the transplant recipients

studied were infected with JC virus after transplan-tation and inclusion-bearing cells were found in 13of these 17 patients.

Time of JC virus infection related to the transplantoperation and the clinical responseOne of the patients (patient B) with a primary infec-tion with JC virus had evidence of infection 10 daysafter a second transplant operation (Table 3). PatientB, a child of 12 years, had not seroconverted afterthe first transplant operation and he remainedseronegative to both JC and BK viruses for twoyears (Fig. 1). During this period the original renaldisease recurred and renal function deteriorated andhe was managed by haemodialysis. He received anincreased dose of azathioprine before the secondtransplant operation and six days after transplanta-tion was given methyl prednisolone because ofreduced renal function. At 10 days he had a mildfever and developed antibody to both JC and BKviruses. After two months, antibody to both virusesdeclined and remained low for the next 10 weeks.During this interval inclusion-bearing cells werepresent in the urine and JC virus isolated. Eighteenweeks after the operation there was a second sub-stantial rise in antibody titre to both viruses (Fig. 1)and the JC virus antibody concentrations remainedraised until his death 18 months after the secondrenal transplant. Specific IgM antibodies to both BKand JC viruses were present and a rise followed by afall and then subsequent rise was shown similar tothe IgG response.The other four primary infections occurred be-

tween 15 and 16 weeks after the operation (twocases) and at 11 and 12 months (Table 3). Apartfrom a mild fever in one patient there was no illnessor symptoms associated with the primary infection.

Six of the 12 secondary JC virus infections occur-red between four and seven weeks, three between10 and 15 weeks, and three were late infections at11, 14, and 30 months. No symptoms were seen in10 patients (Table 3). One patient experiencedmalaise and vomiting but was also infected withcytomegalovirus. Only one patient had symptomspossibly associated with the JC virus infection. Thispatient had no antibody at the time of a first trans-plant operation and transplant nephrectomy threemonths later. When he was given his second trans-plant 12 months afterwards, however, he hadseroconverted and had low levels of JC virus anti-body. Twelve weeks after the second transplant hewas given methyl prednisolone for deterioratingrenal function and subsequently his JC virus anti-

body levels increased from 40 to 640. At this timehe developed palpitations and severe chest pain andwas admitted as an emergency case to the intensivecare unit, where he was found to have a pericardialrub and diagnosed as possibly having a viral pericar-ditis. Six months later he had a recurrence of symp-toms and developed a pericardial effusion. His con-dition improved and his renal function has remainedstable.

CytomegalovirusThirty (62.5%) of the 48 renal transplant recipientsexperienced an infection with cytomegalovirus aftertransplantation. Inclusion-bearing cells were notseen in the urine of 17 of these patients. Thirteenpatients were excreting such cells but all of themalso had concurrent infection with BK and/or JCvirus.

Primary infectionsNineteen of the 48 patients (40%) had nocytomegalovirus antibody before the transplantoperation. Six (32%) of these 19 seronegative casesexperienced a primary infection (Table 1). Five ofthe infections occurred between four and sevenweeks after transplantation and one was at 10weeks. Unlike the primary infection with BK and JCviruses all six patients had symptoms related to thecytomegalovirus virus infection and none of themwas experiencing a concurrent polyomavirus infec-tion. Three of the six patients were seriously ill withfever and pneumonia, two also had a pancreatitis,and one patient died from the infection. Neverthe-less, four of the five surviving patients made a goodrecovery and had functioning grafts at the end of thestudy period.

Secondary infectionsTwenty four (83%) of the 29 seropositive patientshad secondary infections with cytomegalovirus(Table 1). In 17 of the 24 patients (71%) the infec-tion occurred betweeen four and eight weeks aftertransplantation. In five it was between nine and 12weeks, and in only two patients (8%) was there alate infection at eight and 17 months.

Thirteen of the patients had symptoms possiblyrelated to the cytomegalovirus infection. Twelvedeveloped fever and in five this was accompanied byrespiratory symptoms. One patient had malaise andvomiting. Although six of the 13 patients also hadconcurrent infection with a polyomavirus it wasthought unlikely that this virus was responsible forthe symptoms encountered.

583



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

584

BK and JC virus infection related to renal functionand graft survivalRenal function was reduced at the time of infectionin eight of the 21 patients known to have had a BKvirus infection after transplantation. In three of thepatients the impaired renal function was the result ofacute or acute on chronic rejection unrelated to theBK virus infection and these three patients eventu-ally had transplant nephrectomies. One otherpatient had progressive deterioration of renal func-tion and transplant nephrectomy because of therecurrence of the original renal disease. In fourpatients, however, the reduced renal function coin-cided with the BK virus infection and in one patientwas associated with ureteric stenosis. Renal functionimproved without treatment (except decreasingimmunosuppression in one case) in the four patientsand all had functioning renal graft at the end of thestudy.Only four of the 19 patients with either a primary

or secondary JC virus infection after transplantationalso had impaired renal function at the time of theinfection. In one this occurred eight days after thetransplant operation and was transient. Afterwardsthe renal output in this patient was satisfactory foreight months until the original disease recurred. Inthe second patient there were no rejection episodesand renal function was good until the JC virus infec-tion four weeks after transplant. Subsequently thefunction did not recover and continued to deterio-rate.Two other patients had impaired renal function

five weeks and 12 weeks after the transplant opera-tion. Both were treated with methyl prednisoloneand one developed viral pericarditis with pericardialeffusion. Retrospectively they were shown to havehad secondary JC virus infection at this time. Renalfunction recovered quickly in both patients and thegrafted kidneys continued to function well.

Discussion

Nearly two thirds of the patients in this study grouphad either a primary or a secondary infection withone or both polyomaviruses BK and JC. Thepolyomavirus infections were as frequent as thoseassociated with cytomegalovirus, which are knownto be common after transplantation.'5 The incidenceof polyomavirus infection after renal transplant wasgreater than that reported in 1980 by Hogan et al inthe USA, where just half of the 61 patients investi-gated had evidence of an active infection.'3

Cytological examination of the urine proved to bea useful technique in identifying polyomavirusexcretors. Twenty one of the 31 patients (68%)identified serologically as infected with BK or JC


virus, or both, were found to be excretinginclusion-bearing cells, whereas none of the 11patients with cytomegalovirus infection only wereexcreting these cells.Most of the transplant patients in our study had

BK virus antibodies before the transplant operation.Therefore, few were at risk of a primary infection.Nearly half of the patients, however, had no anti-body to either JC virus or cytomegalovirus, and ofthese five (23%) had primary infections with JCvirus and six (32%) with cytomegalovirus. Of theseropositive patients in the present study the rate ofsecondary infections with BK and JC viruses were45% and 46% respectively. The rate of secondarycytomegalovirus infections in the patients was 83%,an incidence similar to that found by other work-ers.'5An increase in antibody titre to both BK and JC

viruses occurred in seven patients. In four patientsthe rise was observed at different times after thetransplant operation, but in three patients the anti-body responses were identical in timing and con-centrations as can be seen in patient B (Fig. 1). Simi-lar responses were seen in two patients reportedpreviously.'2 When the antibody responses occurredsimultaneously it was difficult to assess whether bothviruses were active at the same time or one responsewas anamnestic. One of the authors (SDG) has iso-lated both BK and JC viruses from the urine of abone marrow transplant recipient,'6 but usually onlyone virus is isolated. Hogan and his colleagues,'7using immunofluoresence to stain exfoliatedinclusion-bearing cells, have shown both BK and JCvirus antigens in the same urine sample in fivepatients. This technique is probably the most sensi-tive one for confirming mixed infections.Most of the previous reports of polyomavirus

infections in kidney transplant recipients have con-sidered only BK virus infections and have notincluded serological studies with JC virus. Thereported incidence of BK virus infections aftertransplantation deduced from serological observa-tions has ranged from 12% to 38%.4 3 118-20 Evi-dence of polyomavirus excretion confirmed by elec-tron microscopy or by virus isolation has varied from0% to 44%.4 12 18 21 The variation in these observa-tions may be because serial sera were not examinedand many of the patients studied were not investi-gated from before the transplant operation. Also inmany instances urines may not have been taken atthe peak time for virus excretion.

In the present investigation about half of the sec-ondary polyomavirus infections occurred betweenfour and eight weeks after transplantation comparedwith 71% of the cytomegalovirus infections. Latepolyomavirus infections occurring eight months or



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


more after operation were seen in seven of the 31patients (22-5%) whereas only two of the 24 (8%)patients had late cytomegalovirus infections. Simi-larly, four of the seven primary polyomavirus infec-tions were recognised later than three months. Theprimary cytomegalovirus infections occurred earlier,as reported by other workers.2

Clinically, there was an appreciable difference inthe patients' responses to polyomavirus infectionscompared with cytomegalovirus. Only one of the sixpatients with primary polyomavirus infection hadsymptoms. This patient (patient B) was probablyinfected with both JC and BK viruses and developedonly a mild pyrexia. On the other hand all sixpatients with primary cytomegalovirus infectionswere ill, three seriously, one dying of disseminateddisease including pancreatitis. Most of the secondarypolyomavirus infections were asymptomatic, but inthree patients there were complications. One patientinfected with BK virus developed ureteric obstruc-tion and another had malaise and vomiting. A thirdpatient infected with JC virus had pericarditis andeffusion. Ureteric stenosis associated withpolyomavirus infections has been reported in ninerenal transplant recipients.3 56 1322 The most recentreport of three cases was from the detailed study byHogan et al.'3 The three patients reported in theirpaper developed ureteric stricture two, five and 10months after transplantation. Only one patientdeveloped the stricture during the survey period,and this patient had been infected with both BK andJC viruses. The other two patients were found to beexcreting JC virus some months after the uretericstricture was diagnosed. It is of great importancethat polyomavirus associated ureteric stenosisshould not be mistaken for acute rejection as anyincrease of immunosuppressive drugs during activepolyomavirus infection may impede host response tothe infection and result in an unnecessary transplantnephrectomy. The examination of urine deposits forinclusion-bearing cells would be the quickestmethod to detect these infections.The only disease so far associated with JC virus

has been PML. No neurological symptoms werenoted in any of our patients with active JC virusinfection. Pericarditis has not been reported inassociation with JC virus infection before. AlthoughJC virus was not actually isolated from the pericar-dial fluid (S Clarke, personal communication) asspecial procedures are needed, it is interesting tospeculate that JC virus infection in this patient mayhave affected renal function, and becauseimmunosuppressive drugs were then increased viralpericarditis developed.The polyomavirus infections had only a moderate

effect on renal function and the outcome of the

transplant operations. Reduction in renal functioncoincided with the polyomavirus infection in eight(26%) of the 31 infected patients; four of thesepatients were infected with BK virus and four withJC virus. Renal function continued to deteriorate inonly one of them. At the end of the study period 25of the 31 polyomavirus infected patients (81%) hadfunctioning renal grafts.

We thank Mr B Pentlow and Mr H White for per-mission to study their patients and gratefully ack-nowledge the help of the staff in the renal transplantunit.

References

'Richardson WP, Colvin BB, Cheeseman SH, etal.Glomerulopathy associated with cytomegolavirus viraemia inrenal allografts. N Engl J Med 1981;305:57-63.

2Warrell MJ, Chinn I, Morris PJ, Tobin JCYH. The effects of viralinfections on renal transplants and their recipients. Q J Med(NS) 1980;49:219-31.

3 Gardner SD, Field AM, Coleman DV, Hulme B. New humanpapovavirus (B.K.) isolated from urine after renal trans-plantation. Lancet 1971;i: 1253-7.

4Coleman DV, Gardner SD, Field AM. Human polyomavirusinfection in renal allograft recipients. BrMedJ 1973;iii:371-5.

5Coleman DV, MacKenzie EFD, Gardner SD, Poulding JM,Amer B, Russeil WJI. Human polyomavirus (B.K.) infectionand ureteric stenosis in renal allograft recipients. J Clin Pathol1978;31:338-47.

6 MacKenzie EFD, Poulding JM, Harrison PR, Amer B. Humanpolyomavirus (HPV)-a significant pathogen in renal trans-plantation. Proc Europ Dial Trans Ass 1978;15:352-60.

7Padgett BL, Walker DL, Zu Rhein GM, Eckroade RJ, DesselBH. Cultivation of papova-like virus from human brain withprogressive multifocal leucoencephalopathy. Lancet 1971;i:1257-60.

8 Walker DL. Progressive multifocal leucoencephalopathy: anopportunistic viral infection of the central nervous system. In:Vinken PJ, Bryum GW, eds. Handbook of clinical neurology.Amsterdam: North-Holland Publishing Company, 1978:307.

9 Janisch von W, Gerlach H, Holzhausen H-J. Die multifokaleprogressive leukenzephalopathie-eine papovavirus-enzephalitis. Zentralb AUg Pathol 1977;121:33-9.

'° Reznik M, Halleux J, Urbain E, Mouchette R, Castermans R,Beaujean M. Leucoencephalopathie multifocale progressiveapres transplantation renale. Acta Neurol Belg 1981;81:205-14.

" Padgett BL, Walker DL. Human papovavirus JCV: Naturalhistory, tumorigenicity and interaction with human cells inculture. In: Essex M, Todaro G, zur Hausen H, eds. Viruses innaturally occurring cancers. Cold Spring Harbor Laboratory,1980:319-28.

2 Gardner SD. Implication of papovaviruses in human diseases. In:Kurstak E, ed. Human and related viruses, part A. Compara-tive diagnosis of viral disease. New York: Academic Press,1977:41-84.

"Hogan TF, Borden EC, McBain JA, Padgett BL, Walker DL.Human polyomavirus infections with JC virus and BK virus inrenal transplant patients. Ann Intern Med 1980;92:373-8.

"Coleman DV, Wolfendale MR, Daniel RA, et al. A prospectivestudy of human polyomavirus infection in pregnancy. J InfeciDis 1980;1:142-9.

585



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

586"5Ho M. Virus infections after transplantation in man. Arch Virol

1977;55: 1-24.16 Gibson PE, Gardner SD. Strain differences and some serological

observations on several isolates of human polyomaviruses. In:Sever JL, Madden DL, eds. Polyomaviruses and humanneurological diseases. Progress in clinical and biologicalresearch. New York: Alan R Liss Inc, 1983;105: 119-32.

Hogan TF, Padgett BL, Walker DL, Borden EC, McBain JA.Rapid detection and identification of JC virus and BK virus inhuman urine by using immunofluorescence microscopy. J ClinMicrobiol 1980;11:178-83.

8 Shah KV, Daniel RW, Ziegel RF, Murphy GP. Search for BKand SV40 virus reactivation in renal transplant recipients.Transplantation 1974;17:131-4.

Krech U, Jung M, Price PC, Thiel G, Sege S, Reutter F. Virusinfections in renal transplant recipients. Z Immunforsch1975;148:341-55.

Gardner, MacKenzie, Smith, Porter20 Billaudel S, Le Bris JM, Soulillou JP, Chippaux-Hyppolite C,

Courtieu AL. Anticorps inhibant rhaemagglutination du virusBK: Breve surveillance de 52 transplantes r6naux et pr6val-ence dans diff6rents groupes d'&ge de rouest de la France. AnnVirol (Paris) 1981;132:337-45.

21 Lecatsas G, Prozesky OW, wan Wyk J, Els HJ. Papovavirus inurine after renal transplantation. Nature 1973;241: 343-4.

22 Traystman MD, Gupta PK, Shah KV, et al. Identification ofviruses in the urine of renal transplant recipients by cytomor-phology. Acta Cytol 1980;24:501-10.

Requests for reprints to: Dr Sylvia D Gardner, VirusReference Laboratory, Central Public Health Laboratory,175 Colindale Avenue, London NW9 5HT, England.



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Documents

Prospective BK JC recipients · Study ofhumanpolyomaviruses BKandJCandcytomegalovirus in renal transplant recipients a later investigation, this infection was associated withureteric