Annals of the Rheumatic Diseases 1994; 53: 106-1 11

Does parvovirus B 1 9 have a role in rheumatoidarthritis?

Simo Nikkari, Reijo Luukkainen, Timo Mott6nen, Olli Meurman, Pekka Hannonen,Mikael Skumik, Paavo Toivanen

AbstractObjectives-To determine whether parvo-virus B19 (B19) infection is associatedwith rheumatoid arthritis (RA).Methods-The polymerase chain reactionwas applied to serum, cells isolated fromsynovial fluid, and synovial fluid. Enzymeimmunoassay technique was used todetect antibodies against B19.Results-Of 142 patients with early RA(onset of disease under one year) and 67control patients, serological evidence ofrecent parvoviral infection was found in4/135 and 2/62, respectively. However, noevidence for the presence of parvoviralDNA was observed in 18 synovial fluids, 21samples of synovial fluid granulocytes or40 sera, all obtained from 65 patientsdiagnosed with early RA.Conclusion-Although there is publishedevidence of chronic rheumatoid-likearthropathy following acute parvovirusinfection, our findings do not support theinvolvement of B19 in the aetiopatho-genesis ofRA.

(Ann Rheum Dis 1994; 53: 106-1 1 1)

Turku University,Turku, FinlandDepartment ofMedicalMicrobiologyS NikkariM SkurnikP ToivanenDepartment ofVirology0 Meurman

Satalinna Hospital,Hariavalta, FinlandR LuukkainenTurku UniversityCentral Hospital,Turku, FinlandDivision ofRheumatology,Department ofMedicineT MottonenCentral Hospital,JyvAskyli, FinlandDepartment ofMedicineP HannonenCorrespondence to:Dr Simo Nikkari,Department of MedicalMicrobiology,Turku University,Kiinamyllynkatu 13,SF-20520 Turku, Finland.

Accepted for publication25 October 1993

Parvovirus B 19 (B 19) is a small single strandedDNA virus. B 19 is the only known pathogenichuman parvovirus. It is known to cause

erythema infectiosum,' aplastic crisis inpatients with haematological disorders or

immunosuppression,2 transient erythroblasto-penia of childhood,3 hydrops fetalis,4 and maybe associated with other conditions such as

fibromyalgia.5 Especially in adults, arthro-pathies following primary infection with B19are common, and cases of chronic rheumatoid-like arthropathy following parvovirus infectionshave been reported.6 I B19 DNA has beendetected in the synovial fluid of a patient withserologically proven parvovirus B 19 infectionusing dot blot hybridisation8 and in synovialfluid cells of a patient with 'reactive arthritis'using in situ hybridisation.9 In a recent reportB 19 DNA could be amplified by PCR from thesynovial tissue, but not from the serum of twopatients with chronic B19 arthropathy.'l

Parvovirus laboratory diagnostics is basedprimarily on serology. Because parvovirusesare difficult to culture in vivo, lack of nativeantigen has limited the availability of the test.

Antigen purified from viraemic human sera,synthetic peptides and recombinant antigenshave been used. " Methods based on the

detection of B 19 DNA have been described inseveral reports.7 12-15 As a routine clinical test,dot-blot hybridisation and DNA amplificationby polymerase chain reaction (PCR) have beenreported to be useful in acute B 19 infection,before seroconversion, in association withhaematological disorders,'2 PCR in chronic B 19infection of immunocompromised patients,'4and in screening of blood products B 19.'3

In the present work, we have studied the roleof B 19 in the aetiopathogenesis of RA. For thispurpose, samples from patients with early RA(onset of disease under one year) were analysedby using serological assay of B 19 antibodiesand DNA amplification by PCR. As controls,patients with reactive arthritis (ReA) or otherarthropathies and healthy blood donors wereused.

Patients and methodsPATIENTS

Samples from 62 patients initially diagnosedwith RA were studied by PCR for the presenceof B 19 virus and for serological evidence ofrecent B 19 infection. The detailed char-acteristics of the patients are given intable 1. The final diagnosis was made accordingto the American College of Rheumatologycriteria.'6 Forty patients were examined in theTurku University Central Hospital, 15 inSatalinna Hospital (situated also in south westFinland) and seven in Jyvaskyla Central Hospitalin central Finland. The samples were collectedbetween April 1989 and December 1991.The patients were primarily examined by

general practitioners and sent to the rheuma-tologist for verification of the diagnosis andtreatment. The samples were collected duringthe rheumatological examination and takenwithin one year from onset of disease [mean(SD) time from onset of disease 5-7 (3 2)months]. The diagnosis was confirmed approxi-mately a year after the sample collection.An additional group of 80 patients

[mean (SD) age 51-2 (15-4) years, 51 females,29 males] diagnosed with RA, with diseaseduration of one year or less [mean (SD) timefrom onset of disease 5-2 (2-6) months],was included. Paired sera taken at 3 monthintervals were available from 54 of thesepatients, and only the first sample from 26. Allpatients of this group were examined inJyvaskyla Central Hospital during the years1988 and 1989. In this group, the serumsamples were analysed for B19 antibodies andonly samples of patients with serological

106

group.bmj.com on July 29, 2011 - Published by ard.bmj.comDownloaded from

Does parvovirus B19 have a role in rheumatoid arthritis?

Table 1 The clinical specimens studied

Patients No Female/Male Age Results (positive */tested)Mean (SD)(years) Serology PCR

Serum Synovial Synovialfluid fluid cells

Early rheumatoid arthritis 62** 42/20 54-3 (14-7) lt/55 0/37*** 0/18 0/21Reactive arthritistt 27 12/15 42-9 (14-5) 1/27 0/16 0/16 0/5Otherarthropathies§ 15 9/6 42-7 (16-8) 0/10 0/8 0/5 0/6Healthy blood donors 25 9/16 40 0 (10-1) 1/25 0/25§§ 0/0 0/0

*Evidence of recent parvovirus infection; see table 2.**48/62 were RF positive and 17/62 had erosive disease at the time when the first clinical specimens were taken.***From 12 patients both serum and synovial fluid specimens were studied. In addition serum samples of 3 patients from theadditional group of patients with RA were negative.tIn addition, three patients from the additional group of 80 patients with RA were positive.ttSeven triggered by Salmonella, 6 by Yersinia, 5 by Chlamydiae, 1 by Chlostridium difficile and 8 by non-defined micro-organisms.§7 non-defined arthritis, 1 hydrops of the knee, 1 Lyme arthritis, 1 osteoarthrosis, 1 polymyalgia rheumatica, 1 psoriatic arthritis,1 sarcoidosis, 1 ankylosing spondylitis, 1 Sj6gren's syndrome.§§Plasma instead of serum studied.

evidence of recent B19 infection (detectable anti-B 19 IgM antibodies or a rise in anti-B 19 IgGantibody titre) were studied by PCR (table 2).As control groups, 27 patients with reactive

arthritis (ReA) and 15 patients with otherarthropathies were studied (table 1). Thesamples were taken during March 1990-January 1992. Plasma samples from 25 healthyblood donors were also screened for B 19specific antibodies and DNA; the sampleswere taken during April-June 1991. Severalserum samples taken during 1984-1991 frompatients with serological evidence of acute B 19infection were screened by PCR as well.

CLINICAL SPECIMENS

Synovial fluid (SF) was aspirated from kneejoints and mixed immediately with heparin(50 IU/ml). The synovial fluid cells werepelleted by low speed centrifugation (400 g)and the supematant was stored at -40°C.Mononuclear and polymorphonuclear cellswere separated by Ficoll-Histopaque gradientcentrifugation. Red blood cells were lysedusing 0-83% NH4Cl, pH 7 0. After washingtwice with Hanks' buffered saline the cells werecounted and cryopreserved in fetal calf serumcontaining 10% DMS0 at -1 35°C.

DNA extraction from the polymorpho-nuclear cells was performed after proteinase Kdigestion.'7 A 5 [LI aliquot of DNA equivalentto 50 000 cells was used as a template in PCR.For PCR, serum, plasma and synovial fluid

supematant samples were treated as describedearlier,'5 with slight modifications; 500-1000 RIof sample was heated at 95°C for 10 minutesand centrifuged for 1 hour in 20 000 g at +4°C(Beckman TL-100 ultracentrifuge, Palo Alto,CA, rotor type TLA-100, 3). The supematantwas collected and 5 ,ul used as a template inPCR. Additionally, supernatants of plasmasamples and synovial fluid were further diluted1: 10 in PCR buffer (Boehringer Mannheim,Mannheim, Germany) before the analysis.Inhibition of the polymerase enzyme wasexcluded by the ability of parallel amplificationof a small amount of known B 19 positiveserum mixed with the clinical sample. Theserum samples from the additional RA groupwere diluted 1: 10 in PCR buffer and 5 ,ul wasused as such for PCR analysis.

PCR CONDITIONSThe standard PCR amplification was per-formed as described by Saiki et all' in 0-5 mlpolypropylene tubes with a Perkin Elmer Cetus

Table 2 Patients with serological evidence of recent parvovirus infection.

Patient Age at Sex RF Characterisation ofdisease Duration of Duration Anti-parvovirus B19 antibodies (ELISA) Samples analysed byNo onset follow up of disease PCR (result)

(years) (years) at drawing IgM (OD 405 nm) IgG (OD 405 nm)ofsample Serum SF SFC(months) Sample* Negat ctrt Result Sample Negat ctrt

1 35 F - ParvovirusBl9 arthropathy. 3 6 1-019 0-162 + 0-201 0-119 ND NDFully recovered in 3 years. Non-erosive.Treated as seronegative, oligoarticular RA.

2§ 38 F - Parvovirus B19 arthropathy. 4 3 0-187 0-063 - 0-061t 0-086 - ND NDSymptoms preceded by sore throat. 6 ND ND 0 430 0-086 - ND NDTotal remission in 6 months. 9 ND ND 0-878 0-086 ND ND NDTreated as seronegative, polyarticular RA.

3 32 F - Seronegative RA, polyarticular, non-erosive, 4-5 2 1-149 0-143 + 0-219 0-127 - ND NDmild disease. Reactivation of disease afterchild delivery at 4 years follow up.Possibly B19 arthropathy.

4 68 M + Rheumatoid arthritis. 3 4 0-213 0-144 - 0 555 0-025 - ND NDSeropositive, polyarticular, erosive disease. 7 ND ND 0-767 0-025 - ND NDAlmost total remission with gold treatment. 10 ND ND 0-641 0-025 ND ND NDSerological evidence of B 19 infection unclear.

5 38 F + Parvovirus B19 arthropathy. 1 3 weeks 0-673 0-054 + 0-560 0-035 - - NDInitially diagnosed as self-limiting 7 weeks ND ND 0-565 0 035 ND ND NDreactive arthritis.Symptomless during follow up.

6 54 M - Healthy blood donor. - - 0-776 0-144 + 0-231 0-136 - ND ND

*Mean value of duplicate samples.tMean value of negative controls.jDiagnostic rise in I, II and III sample.SPatients 2-4 are from the additional RA group of 80 patients; from these only serum samples were available.

SF = synovial fluid.SFC = synovial fluid cells.ND = not done.

107

group.bmj.com on July 29, 2011 - Published by ard.bmj.comDownloaded from

Nikkari, Luukkainen, Mottonen, Meurman, Hannonen, Skurnik, Toivanen

thermal cycler. The total incubation volumewas 50 ,u, containing 10 mM Tris-HCl pH 8.3,50 mM KCI, 1 5 mM MgCl2, 041 mg/mlgelatin, 200 ,uM each dNTP, 2-5 units TaqDNA polymerase (Boehringer Mannheim,Mannheim, Germany), 50 pmol each primerand 5 [lI template. The PCR steps 94°C for45 s, 63°C for 1 minute, 72°C for 2 minuteswere repeated 40 times. When the human beta-globin primers H 1 and H2 were used, the PCRsteps 94°C for 45 seconds, 52°C for 1 minute,72°C for 2 minutes were repeated 30 times andonly 1 25 units of polymerase enzyme perreaction were used.

Special care was taken to avoid contami-nation of samples with amplicons.'9 Clinicalmaterial was handled whenever possible inlaminar flow hoods. The preparation of PCRreaction mixture and pipetting of samples wereperformed in laminar flow hoods situated inseparate rooms with restricted passage ofpersonnel and decontamination barriers,including changing of protective clothing onentering and positive air pressure of the area.The rooms and hoods were illuminated by UVirradiation lamps, when not in use. Theamplification procedure and analysis of PCRproducts were performed in another laboratoryroom. Special emphasis was placed on pre-venting carryover of amplicons from post-PCRto pre-PCR facilities. Positive displacementpipettes (Tri-Continent Scientific, Grass Valley,CA) were used in pipetting of specimens.

PRIMERS AND CONTROLS USED IN PCRThe oligonucleotide primers used wereselected according to previously publishedsequences'5 20 and analysed using Oligo(Version 3 4, MedProbe, Oslo, Norway) primeranalysis software. Primer P1 (5'-GTA CGCCCA TCC CCG GGA CCA GTT CAG G-3')is situated in position 2060-2087 and P5(5'-CCC ACA TGG CAG CTA CAT CGCACC AAA T-3') in position 2369-2342 of thesequence published by Shade et al.20 Thelength of the amplified product is 310 bp. P2(5'-CAG GTA AAC CCC TTA CAC CGTCCC ACA C-3', position 2230-2203) wasused as a probe in hybridisation experiments.Five pl of a known B 19 positive serum samplediluted 1:1000 in water was used in all assaysas positive control. A negative control consistingof all reaction reagents necessary for PCR, withwater as template, was always included.To exclude possible false negative results

caused by inhibitory factors, the presence ofhuman DNA in the synovial fluid cell sampleswas verified by collateral amplification ofhuman beta-globin gene using primers H 1(PCO4, 5'-CAA CTT CAT CCA CGT TCACC-3', position 54-73) and H2 (GH20, 5'-GAAGAG CCA AGG ACA GGT AC-3', position-195- -176), which amplify a 268 bp segmentof human beta-globin encoding genome.2"

DETECTION OF AMPLIFIED PRODUCTTen ,ul of the amplified product was analysedon 1-5% agarose gel electrophoresis and

visualised by UV fluorescence after stainingwith ethidium bromide. After electrophoresisthe DNA was transferred overnight to a

Hybridisation Transfer Membrane (Gene-ScreenPlus, Du Pont, Boston) with 0 4 NNaOH using the capillary blot procedure.The filter was washed with 0X2 M Tris-HCl,pH 7*5, 2 X SSC (0 3 M NaCl and 0 03 Msodium citrate) before drying. The amplified268 bp human beta-globin gene product wasvisualised after agarose gel electrophoresis andstaining with ethidium bromide.The hybridisation experiments with B 19

specific oligonucleotide probe (P2) were per-formed as reported earlier.22

SEROLOGY

B19 specific IgG and IgM antibodies weredetermined by enzyme linked immunosorbentassay using Parvoscan-B19TM (Ferring diag-nostica, Malmo, Sweden). The synthetic-peptide based assay has been described byFridell et al.23 From most patients serum wasavailable, if not, synovial fluid was used (in 15patients with RA, 13 with ReA, 4 with otherarthropathies). From the healthy blood donorsplasma was tested. Nonspecific anti-B19 IgMantibody-assay results caused by rheumatoidfactor were eliminated by GullSORB (Gulllaboratories, Salt Lake City, Utah) inactivationreagent or RF-Absorbant (Behring, Marburg,France).

ResultsPCR

The sensitivity of the assay was about 70molecules per assay when purified PCRproduct was used as template (figure). With theknown B 19 positive serum sample used aspositive control, B 19 DNA was detectable bySouthern hybridisation of the PCR product upto a serum dilution of 1:109. Addition ofhuman DNA equivalent o 50 000 synovialfluid granulocytes per assay did not affectsensitivity of the PCR. No B 19 DNA wasdetectable from the patients with differentforms of arthritis, including 65 patients withearly RA (table 1). Conversely, B 19 DNAcould be amplified from 6/21 serum samplestaken from patients with serologically provenacute B 19 infection as well as in cases ofhydrops fetalis and transient erythroblasto-penia. Prolonged B 19 viraemia in a child withaplastic crisis was detectable by our assay.Human beta-globin gene was amplified fromall the synovial granulocyte samples, indicatingthat significant inhibitory agents were notpresent and that DNA isolation had beenproperly carried out.

SEROLOGYThe patients with serological evidence ofrecent B19 infection are shown in table 2. Fourpatients of 135 initially diagnosed with earlyRA had serological evidence of recent B19infection (patients 1-4). Three of thesepatients seem to be true B 19 arthropathies

108

group.bmj.com on July 29, 2011 - Published by ard.bmj.comDownloaded from

Does parvovirus B1 9 have a role in rheumatoid arthritis?

Ad2 3 . 5 6 7 8 9 1 V112 M

B

M 1 2 3 4 5 6 7 8 9 10 11 12 M

JhIEiiiEMhi.

Sensitivity of the assay using different amounts ofamplified parvovirus Bl 9 DNA fragment as template. The 310 bpamplified Bl 9 DNA fragment was purifiedfrom a preparative agarose gel using an Ultrafree-MC Filter Unit with0-45 ,um membrane (Millipore Corporation, Bedford, A4) as recommended by the manufacturer. The DNA concentrationwas measured by UV260 spectrophotometry. (A) Agarose gel electrophoresis of amplified product after 40 cycles ofpolymerasechain reaction (PCR) with primers Pl and P5. Lane M, molecular weight marker; lane 1, 7 X 109 molecules; lane 2,7 x 1o8 molecules; lane 3, 7 X 107 molecules; lane 4, 7 X 1 6 molecules; lane 5, 7 x 105 molecules; lane 6, 7 x 104molecules; lane 7, 7 X 103 molecules; lane 8, 7 x 1o' molecules, lane 9, 70 molecules; lane 10, 7 molecules; lane 1 1, knownparvovirus positive serum, dilution 1 :1000; lane 12, negative control; (B) Southern blot hybridisation of the same agarosegel, using parvovirus B19 specific probe (P2).

(patients 1-3), and one fulfils ACR criteria forRA (patient 3). The serological diagnosis ofpatient number 1 is based on a positive anti-B19 IgM antibody level, which together withthe low anti-B19 IgG suggests a possibility ofsubclinical infection after the beginning ofarthritic symptoms. Patient number 2 isclinically a typical B19 arthropathy patient.The joint manifestations were preceded bysymptoms of upper respiratory tract infectionin concordance with acute B 19 infection.The serological diagnosis is based on slowdevelopment of anti-B 19 IgG antibodiesduring several months, a phenomenon we havenot encountered before in the diagnosis ofacute B 19 infection in otherwise healthyindividuals. The anti-B 19 IgM antibodiesdetected in patient 3 are consistent with arecent infection within two months. Unfor-tunately, paired sera were not available fromthis patient, nor from patient number 1. Inpatient number 4 a slight rise in anti-B 19 IgGantibodies between the first two serum samplesis observed, but on the basis of the clinicalpicture, B 19 arthropathy seems unlikely.

In the control groups two possible cases ofrecent B 19 infection were found. Patientnumber 5 was initially diagnosed as having self-limiting reactive arthritis, the triggering agentbeing uncertain. The IgM and IgG antibodyresponses against B 19 are consistent withrecent infection; the patient most probably hadB 19 arthropathy. In addition, anti-B 19 IgMantibodies were observed in one healthy blooddonor (patient 6), from which no clinicalinformation is available.

DiscussionAcute B 19 infections are quite commonlyfollowed by arthropathies, especially in adults.Chronic conditions resembling RA associatedwith B19 infection have also been reported.6 7Parvovirus DNA has been observed in synovial

fluid, synovial fluid cells and synovium ofpatients with B 19 arthropathy.8 '0 In acute B 19infection virus can be detected in high concen-trations in serum and for a longer period incirculating polymorphonuclear cells.24 Forthese reasons we have studied the serum,synovial fluid and synovial fluid cells frompatients with RA and from controls.

In a recent prospective study the presence ofB19 DNA was reported in 75% of synovialbiopsy samples taken from patients with RAand 1 6.7% of biopsies taken from patients withosteoarthritis.25 In our study B 19 DNA was notdetectable in clinical samples taken from anyof the patients with chronic arthritis; parvoviralDNA, however, was amplified from severalcontrol sera of patients with serological evi-dence of acute B 19 infection, includingindividual cases of aplastic crisis in a child withan underlying haematological disorder andtransient erythroblastopenia of an otherwisehealthy child.3 B19 could also be detected inthe amniotic fluid of a case of hydrops fetalis,caused by intrauterine B 19 infection.4 Thesensitivity of our PCR assay was approximately70 B19 DNA particles, which is the same asthat reported by Saal et al.25 We do not havea good explanation for the discrepanciesbetween these two studies. In our studysynovial biopsies were not studied, althoughthis does not explain the different results, asSaal et al found B 19 in peripheral blood mono-nuclear cells of 15% of patients in all groups.The duration of RA of the patients describedby Saal et al was approximately 1 1 years,whereas our patients suffered from recentonset RA. In our opinion it is preferable tostudy patients with disease duration as short aspossible when the aetiological agent is beingsought. We also studied by PCR samples takenfrom 9 patients with advanced RA [8 sera,2 synovial fluids, 1 synovial fluid cell sample;mean (SD) from onset of disease 5.1 (4 5)years]; B 19 DNA was not present in these

109

group.bmj.com on July 29, 2011 - Published by ard.bmj.comDownloaded from

Nikkari, Luukkainen, Mdttonen, Meurman, Hannonen, Skurnik, Toivanen

samples either. One possible explanation in thereport by Saal et al is a cross-reaction of theprimer sequences with othl}r parvoviruses. In

1984 Simpson et al26 suggested that parvo-

viruses are associated with RA, but the originof their RA-1 virus has remained unclear.27With PCR, initially three samples of our

patients with arthritis (two synovial fluid andone serum sample, all from different patientsand patient groups) were weakly positive afterhybridisation experiments of PCR productswith B19 specific probe. They were negativewhen PCR was repeated from the originalsamples; the contamination source of thesefalse positive results was not found. The riskfor contamination is great when the amplifi-cation method is optimised to yield highsensitivity.Cohen et a128 observed evidence for recent

B19 infection more often in patients with RAthan in the controls, using antibody detectionbased on a radioimmunoassay with mono-

clonal antibodies. In their study, four patientsout of 69 had evidence of B19 infection at theonset of RA; the prevalence of anti-B19 IgGwas also significantly higher (92.3%) in RApatients than in the controls (609%). Lefrereet al,29 using counterimmuno-electrophoresis,did not find significant differences in B19 sero-

positivity between patients with RA (33.3%)and the adult healthy population (25.3%). Theassay applied by us for detection of B19 anti-bodies is based on the use of synthetic peptidesas the antigen. We could find no significantdifferences in the mean values for anti-B19IgM and IgG antibodies in the different patientgroups (data not shown).The clinical specimens from patients with

arthritis studied in this work were taken duringthe years 1988-1992. According to localpaediatricians, during the study period onlysporadic cases of erythema infectiosum were

seen in the Turku area (Jussi Mertsola, Depart-ment of Paediatrics, Turku University CentralHospital, personal communication). Serologicaldata of B 19 infections from the Turku area are

available from the beginning of year 1991.During our arthritis investigations, no B 19epidemics were observed in the community,despite some serologically proven infections,and a few B19 viraemic (PCR positive) patients(unpublished findings).The epidemiological situation may account

for the differences between our findings andthose reported by Cohen et al.28 The earliestserum samples analysed from the 69 patientswith RA studied by Cohen et al were takenwithin five months from the onset of arthriticsymptoms. The first sample taken from our

62 patients with RA was taken within one

year from the beginning of arthritic symptoms(average 5.7 months). From 18 of these 62

patients with RA the serum sample was

available within three months from the onset of

disease; this is a reasonable window for

diagnosis of acute B 19 infection based on B 19

anti-IgM serology. In chronic arthritis, how-

ever, a prolonged B 19 anti-IgM response could

occur, as has been shown in arthritis associatedwith rubella virus infection.30 In one of our B 19

arthropathy patients (number 1, table 2) B 19anti-IgM antibodies were detectable 6 monthsafter the beginning of the arthritic symptoms;the patient did not present any symptoms of aninfectious febrile illness, so the true timing ofher B 19 infection remains unsolved. Amongthe additional 80 patients with RA the firstserum sample was taken from 22 patientswithin three months from the beginning ofthe symptoms (average 5-2 months), and con-valescent sera were available from 54 patientsfor detection ofIgG anti-B 19 antibodies.

It has been proposed that a significantproportion of patients diagnosed as sero-negative RA and a lesser proportion of sero-positive RA patients actually have B 19arthropathy.7 In the present study we observedserological evidence of recent B19 infection inonly four of 135 patients with early RA. In twocases the finding was based only on high IgMantibody concentrations; paired sera were notavailable. Evidence of recent B 19 infection wasalso found in one patient initially diagnosed asself-limiting reactive arthritis and in onehealthy blood donor. The discrepancy betweenour findings and those published earlier7 28

might be explained by the low prevalence ofB 19 infection circulating in the communityduring our study.According to our PCR and serological data,

the association of B19 infection with RA is arare phenomenon. Nevertheless, chronic B19arthropathy should be kept in mind in thedifferential diagnosis of RA.7 Our data do notexclude the possibility of the virus persisting atthe site of inflammation after acute infection,due to the small number of patients in ourstudy who actually had chronic B19 arthro-pathy. The present findings do not support theinvolvement ofB 19 in the aetiopathogenesis ofRA. However, its role as a triggering factorin a genetically or otherwise predisposed hostshould not be ruled out.

We are grateful to Dr Timo Hyypia for providing the knownpositive patient serum. Paivi Alenius, Tiina Haarala, MerjaKleme and Marju Niskala are thanked for their excellenttechnical assistance. This work was supported by the SigridJuselius foundation and Finnish Medical Foundation.

1 Anderson M J, Jones S E, Fisher-Hoch S P, et al.Human parvovirus, the cause of erythema infectiosum(fifth disease)? Lancet 1983; i: 1378.

2 Young N. Hematologic and hematopoietic consequences ofB19 parvovirus infection. Seminars in Hematology 1988;25: 159-72.

3 Nikkari S, Meurman 0, Wanne 0. Parvovirus B19 andtransient erythroblastopenia of childhood (letter). Br JfHaematol 1993; 83: 679.

4 Anand A, Gray E S, Brown T, Clewley J P, Cohen B J.Human parvovirus infection in pregnancy and hydropsfetalis. N EnglJMed 1987; 316: 183-6.

5 Leventhal L J, Naides S J, Freundlich B. Fibromyalgia andparvovirus infection. Arthritis Rheum 199 1; 34: 13 19-24.

6 White D G, Woolf A D, Mortimer P P, Cohen B J,Blake D R, Bacon P A. Human parvovirus arthropathy.Lancet 1985; i: 419-21.

7 Naides S J, Scharosch L L, Foto F, Howard E J.Rheumatologic manifestations of human parvovirus B19infection in adults. Arthritis Rheum 1990; 33: 1297-309.

8 Dijkmans B A C, van Elsacker-Niele A M W, SalimansM M M, van Albada-Kuipers G A, de Vries E,Weiland H T. Human parvovirus B19 DNA in synovialfluid. Arthritis Rheum 1988; 31: 279.

9 Kandolf R, Kirschner P, Hofschneider P H, Visher T L.Detection of parvovirus in a patient with 'reactivearthritis' by in situ hybridization. Clinical Rheumatology1989; 8: 398-401.

10 Naides S J, Foto F, Marsh J L, Scharosch L L, Howard E J.Synovial tissue analysis in patients with chronic parvo-virus B 19 arthropathy. Clinical Research 1991; 39: 733A.

110

group.bmj.com on July 29, 2011 - Published by ard.bmj.comDownloaded from

Does parvovirus B19 have a role in rheumatoid arthritis?

11 Soderlund M, Brown K, Meurman 0, Hedman K.Prokaryotic expression of a VP1 polypeptide antigenfor diagnosis by a human parvovirus B 19 antibodyenzyme immunoassay. J Clin Microbiol 1992; 30:305-11.

12 Salimans M M M, Holsappel S, van de Rijke F M,Jiwa N M, Raap A K, Weiland H T. Rapid detection ofhuman parvovirus B 19 DNA by dot-hybridization and thepolymerase chain reaction. J Virol Methods 1989; 23:19-28.

13 Clewley J P. Polymerase chain reaction assay of parvovirusB19 DNA in clinical specimens. J Clin Microbiol 1989; 27:2647-51.

14 Koch W C, Adler S P. Detection of human parvovirus B 19DNA by using the polymerase chain reaction. J ClinMicrobiol 1990; 28: 65-69.

15 Sevall J S. Detection of parvovirus B19 by dot-blot andpolymerase chain reaction. Molecular and Cellular Probes1990; 4: 237-46.

16 Arnett F C, Edworthy S M, Bloch D A, et al. The AmericanRheumatism Association 1987 revised criteria for theclassification of rheumatoid arthritis. Arthritis Rheum1988; 31: 315-24.

17 Tamm A, Ziegler T, Lautenschlager I, et al. Detection ofcytomegalovirus DNA in cells from synovial fluid andperipheral blood of patients with early rheumatoidarthritis. J Rheumatol 1993; 20: 1489-93.

18 Saiki R K, Gelefand D H, Stoffel S, et al. Primer-directedenzymatic amplification of DNA with a thermostabileDNA polymerase. Science 1988; 239: 487-91.

19 Kwok S, Higuchi R. Avoiding false positives with PCR.Nature 1989; 339: 237-8.

20 Shade R 0, Blundell M C, Cotmore S F, Tattersall P,Astell C R. Nucleotide sequence and genomeorganization of human parvovirus B19 isolated from theserum of a child during aplastic crisis. J Virol 1986; 58:921-36.

21 Bauer H M, Ting Y, Greer C E, et al. Genital humanpapillomavirus infection in female university students asdetermined by a PCR-based method. JAMA 1991; 265:472-7.

22 Nikkari S, Merilahti-Palo R, Saario R, et al. Yersinia-triggered reactive arthritis. Use of polymerase chainreaction and immunocytochemical staining in thedetection of bacterial components from synovialspecimens. Arthritis Rheum 1992; 35: 682-7.

23 Fridell E, Cohen B J, Wahren B. Evaluation of a synthetic-peptide enzyme-linked immunosorbent assay forimmunoglobulin M to human parvovirus B 19. J ClinMicrobiol 1991; 29: 1376-81.

24 Kurzman G J, Gascon P, Caras M, Cohen B, Young N S.B19 parvovirus replicates in circulating cells of acutelyinfected patients. Blood 1988; 71: 1448-54.

25 Saal J G, Steidle M, Einsele H, Muller C A, Fritz P,Zacher J. Persistence of B 19 parvovirus in synovialmembranes of patients with rheumatoid arthritis.RheumatolInt 1992; 12: 147-51.

26 Simpson R W, McGinty L, Simon L, Smith C,Godzeski C W, Boyd RJ. Association of parvoviruses withrheumatoid arthritis of humans. Science 1984; 223: 1425-8.

27 Phillips P E. Evidence implicating infectious agents inrheumatoid arthritis and juvenile rheumatoid arthritis.Clin Exp Rheumatol 1988; 6: 87-94.

28 Cohen B J, Buckley M M, Clewley J P, Jones V E,Puttick A H, Jacoby R K. Human parvovirus infection inearly rheumatoid and inflammatory arthritis. Ann RheumDis 1986; 45: 832-8.

29 Lefrere J-J, Meyer 0, Menkes C-J, Beaulieu M-J,Courouce A-M. Human parvovirus and rheumatoidarthritis. Lancet 1985; i: 982.

30 Meurman 0 H. Persistence of immunoglobulin G andimmunoglobulin M antibodies after postnatal rubellainfection determined by solid-phase radioimmunoassay.JClinMicrobiol 1978; 7: 34-8.

illl

group.bmj.com on July 29, 2011 - Published by ard.bmj.comDownloaded from

doi: 10.1136/ard.53.2.106 1994 53: 106-111Ann Rheum Dis

 S Nikkari, R Luukkainen, T Möttönen, et al. rheumatoid arthritis?Does parvovirus B19 have a role in

http://ard.bmj.com/content/53/2/106Updated information and services can be found at:

These include:

References http://ard.bmj.com/content/53/2/106#related-urls

Article cited in:

serviceEmail alerting

the box at the top right corner of the online article.Receive free email alerts when new articles cite this article. Sign up in

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:

group.bmj.com on July 29, 2011 - Published by ard.bmj.comDownloaded from