Journal of Virological Methods 130 (2005) 59–65

Assessment of IgM enzyme immunoassay and IgG avidity assay fordistinguishing between primary and secondary immune response

to rubella vaccine

Rasool Hamkara,∗, Somayeh Jalilvanda, Talat Mokhtari-Azada, Keramat Nouri Jelyania,Hosein Dahi-Farb, Hoorieh Soleimanjahic, Rakhshandeh Nategha

a School of Public Health, Tehran University of Medical Sciences, Tehran 14155, Iranb School of Medicine, Shahid Beheshti University of Medical Sciences, Tehran, Iran

c School of Medicine, Tarbeiat Modarres University, Tehran, Iran

Received 25 April 2005; received in revised form 6 June 2005; accepted 9 June 2005Available online 18 July 2005

Abstract

primaryered when

sing well-de Behring,

avidity% compare

infection

bothram)bellaendsiza-al.,atally

,

ursnced,ientgnif-

The primary test for the laboratory confirmation of rubella is IgM serology. It is important to distinguish IgM reactivity caused byinfection from that caused by reinfection or persistence, especially in pregnant women; as termination of pregnancy is considprimary rubella is diagnosed during the first trimester.

In this study, the performance of rubella IgM enzyme immunoassay (IgM-EIA) and rubella IgG avidity assay were compared udefined panels of sera from persons vaccinated against rubella and commercial rubella IgM and IgG enzyme immunoassay kits (DaMarburg, Germany).

The sensitivity and specificity of rubella IgM-EIA were found to be 77.4 and 97.9%, respectively, while the results for rubella IgGassay were 100 and 100%. IgG avidity assay showed higher positive and negative predictive values than the IgM-EIA (100 and 100to 96.9 and 82.9%).

In conclusion, the rubella IgG avidity assay is more sensitive and specific than IgM-EIA for differential detection of primary rubellafrom rubella reinfection.© 2005 Elsevier B.V. All rights reserved.

1. Introduction

Rubella is a viral disease that results in a transient macu-lopapular rash, lymphadenopathy and low-grade fever (Bestand Banatvala, 2000; CDC, 2001). It is generally a mild ill-ness, and serious complications are rare. However, if acquiredduring the first trimester of pregnancy, there is a 90% riskof congenital malformations in the fetus. Prevention of con-genital rubella syndrome is the main goal of rubella immu-nization and can be achieved by universal immunization ofchildren and immunization of susceptible women of child-bearing age using the live attenuated rubella vaccine (CDC,

∗ Corresponding author. Tel.: +98 21 896 2343; fax: +98 21 895 0595.E-mail address: rhamkar@sina.tums.ac.ir (R. Hamkar).

2001). There is a significant burden of disease globally indeveloped (where there is no rubella vaccination progand developing countries as a result of congenital rusyndrome, and the World Health Organization recommthat countries should include rubella into their immuntion programs if possible (Cutts et al., 1997; Robertson et1997; WHO, 2000a, 2000b). Economic studies indicate thrubella immunization programs are beneficial economicin both developed and developing countries (Hinman et al.2002).

Reinfection with rubella is usually subclinical and occmost commonly among pregnant women who experieclose and prolonged contact with rubella virus (Best et al.1998). Evidence of reinfection would be accepted if a patwho had pre-existing rubella antibodies showed a si

0166-0934/$ – see front matter © 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.jviromet.2005.06.003

60 R. Hamkar et al. / Journal of Virological Methods 130 (2005) 59–65

icant rise of IgG antibody titer or a rubella-specific IgMresponse, or both (Best et al., 1998). The IgM response istypically weak, but may sometimes be strong. Documentedrubella reinfection in which the patient has developed clinicalsymptoms, especially a rubelliform rash, has been reportedvery infrequently (Best et al., 1998; Korhonen et al., 1999).Despite the high occurrence of congenital rubella syndromein primary rubella infection among pregnant women in firsttrimester of pregnancy, the documented reports of congen-ital rubella syndrome by rubella reinfection are very rare(Aboudy et al., 1997).

The diagnosis of acute rubella infection, other viral andsome microbial infections rely often on the serological detec-tion of immunoglobulin M (IgM) antibodies, but the availabletechniques have serious pitfalls that may lead to erroneousinterpretation (Bodeus et al., 1998; Korhonen et al., 1999).Detection of rubella-specific IgM alone cannot be consideredabsolute proof of a recent primary infection. IgM responseafter primary infection may be prolonged, lasting up to sev-eral years. Furthermore, in some reinfections, rubella IgM isdetectable. False-positive IgM results may also occur dueto rheumatoid factor of the IgM class, heat treatment ofthe sera, heterophil antibodies, or interference by parvovi-ral infections (Hedman et al., 1989) as well as some otherviral agents such as Epstein–Barr virus (Bottiger and Jensen,1997; Gutierrez et al., 1999) and Measles virus (Thomas et al.,1 in-i ge,t nga Asa tingd sitiveru cifict butd 96;T c-t in-f sincet ellai n,1

owa or asa ay isg mento al.,1 ays,t bye thyla diest s tot pro-p imeoe t

of IgG avidity assay is calculation of cut-off point betweenlow and high avidity IgG responses. Different values of thecut-off points were reported; in some studies avidity index<30% were regarded as low avidity while avidity index >50%were interpreted as high avidity IgG responses (Thomas et al.,1992; Thomas and Morgan-Capner, 1991). In another studythe avidity index higher than 60% and lower than 40% wasconsidered as high avidity and low avidity IgG, respectively(Bottiger and Jensen, 1997), whereas in another study, thecut-off point of avidity was reported equal to 55% (Gutierrezet al., 1999). These differences may arise as the result ofusing small numbers of defined sera as low and high avid-ity panels to calculate cut-off point. In this study, the abilityof rubella IgM enzyme immunoassay and rubella IgG avid-ity assay to differentiate between primary immune response(infection) and secondary immune response (reinfection) torubella vaccine were evaluated, using well-defined panels ofsera from rubella vaccines with and without previous immu-nity against rubella virus. These panels consisted of relativelylarge numbers of sera for precise calculation of avidity cut-offpoint.

2. Materials and methods

2.1. Commercial enzyme immunoassays

mei andI ols,c dingt

2

riedo tioniO er ro-c urg,G

2

natedp mu-n cam-p sh sam-p conds anti-r unes d byt nti-b

999). In particular, since the incidence of rubella has dimshed in countries with high rubella vaccination coverahe proportion of non-specific rubella IgM reactivity amoll samples with rubella IgM reactivity has increased.result, the positive predictive value (PPV) of IgM tes

ecreases such that there is a significant risk of false poesults (Best et al., 2002; Cutts and Brown, 1995). Althoughsing capture IgM-EIA methods, which are more spe

han indirect IgM-EIA, reduced non-specific reactivity,id not remove it totally (Hudson and Morgan-Capner, 19ipples et al., 2004). It is important to distinguish IgM rea

ivity caused by primary infection from that caused by reection or persistence, especially in pregnant women,ermination of pregnancy is considered when primary rubs diagnosed during the first trimester (Bottiger and Jense997).

The differential assay of rubella high avidity and lvidity IgG antibodies can be used as an alternativecomplementary to the IgM antibody assay. This ass

aining popularity as a diagnostic method for the assessf the time of infection (Bodeus et al., 1998; Korhonen et999). In protein-denaturing avidity enzyme immunoass

he patient’s IgG is allowed to bind to its antigen, followedlution with or without a protein denaturant, such as diemine. Diethyl amine dissociates weekly bound antibo

o their specific antigens but strongly bound antibodieheir specific antigens remain nearly intact. Thus, theortion of IgG remaining antigen bound indicate the tf primary infection or occurrence of reinfection (Bodeust al., 1998; Korhonen et al., 1999). The most critical poin

The commercial rubella-specific IgM and IgG enzymmunoassays were the enzygnost anti-rubella virus IgMgG (Dade Behring, Marburg, Germany). All assay protocut-offs and result interpretations were carried out accoro the manufacturer’s instructions.

.2. Hemagglutination inhibition assay

The hemagglutination inhibition assays were carut according to the standard rubella hemagglutina

nhibition assay instruction described by Best (Best and’Shea, 1995) with kaolin absorption of sera for th

emoval of non-specific inhibitors, and pigeon erythytes, using four units of antigen (Dade Behring, Marbermany).

.3. Serum samples from vaccines

Five hundred paired sera from measles–rubella vaccieople were collected as part of the measles–rubella imity assay Program in Iran during measles–rubella massaign in December 2003, and stored at−80◦C. First samplead been collected before vaccination and the secondles were collected one month after vaccination. The seerum samples were classified to primary and secondaryubella antibody response panels according to their immtatus for rubella virus before vaccination that revealeesting of their first serum samples for rubella-specific aodies.

R. Hamkar et al. / Journal of Virological Methods 130 (2005) 59–65 61

2.4. Primary anti-rubella antibody response (primaryrubella infection) panel

Two hundred and sixteen serum samples were selectedfrom measles–rubella vaccinated people, which did not con-tain any anti-rubella antibodies before vaccination. Their firstserum samples were tested for anti-rubella IgM and IgG usingabove-mentioned EIA kits, which indicated that they did nothave any anti-rubella antibodies. First samples of all seraalso were tested by hemagglutination inhibition assay, andthey were all negative for rubella antibody.

2.5. Secondary anti-rubella antibody response (rubellareinfection) panel

Two hundred and eighty-four serum samples were selectedfrom measles–rubella vaccinated people, who had anti-rubella IgG before vaccination. Their first serum sampleswere tested for anti-rubella IgM and IgG using above-mentioned EIA kits, and they had anti-rubella IgG antibody.First samples of all sera also were tested by hemagglutina-tion inhibition assay, and they were all positive for rubellaantibody.

2.6. Rubella panel

s paro asesw avep inI aireds 5a test-i ectedt ; 23 ot bella( tedp

2

ellaI “in-h ardsw

2a

aredu odsd tions( eachd tes( ns;t

2.9. Anti-rubella IgG avidity measurement

All the above-mentioned sera were also tested by the anti-rubella IgG avidity assay. The avidity of IgG for rubella viruswas measured by a protein-denaturing enzyme immunoas-say where the antibodies were first allowed to bind to therubella virus antigen, followed by elution by buffer with andwithout 35 mM diethyl amine (Thomas and Morgan-Capner,1991; Gutierrez et al., 1999). Each sample was tested at tworeplicates and single serum dilution (1:200) was applied toeach of two replicates. After incubation for 1 h, test plateswere washed four times, and then one replicate of each tworeplicates was soaked for 5 min in washing buffer and otherremained replicate for 5 min in washing buffer containing35 mM diethyl amine. Fresh buffers were applied and this stepwas carried out twice more. The plates were then washed fourtimes with washing buffer. Then test was continued accord-ing kit’s procedure. Remaining specific antibody was thendetected, and an avidity index (AI) was calculated as follows:

AI (%) = ODwells soaked with 35 mM DEA

ODwells soaked with wash buffer× 100

2.10. Statistical methods

ri-m byC urves Ss itivity,s s ofr Me werec werec testw weent -r ntd

2

thp na-l RocC sifiedb upsb Seraw plesc ityi higha nym tionp anels ined

Paired sera from acute rubella cases were collected af the measles–rubella surveillance program in Iran. Cere confirmed as rubella when their rubella IgM-EIAs gositive results; or if IgG seroconversion or >4-fold rise

gG titer was detected between acute and convalescent pera by hemagglutination inhibition (Best and O’Shea, 199)nd other rash illness (measles and parvovirus B19) IgM

ng was negative. One hundred serum samples were selheir acute sera were collected 4–7 days post-rash onsethem had documented history of measles, mumps and ruMMR) vaccination and reminders were from non-vaccinaeople, whit age-range 5–20 years old.

.7. Anti-rubella IgM assay

All sera from above panels were carried out to anti-rubgM assay using the above-mentioned kits. Two local (ouse” preparations) weak and strong positive IgM standere included as external control in every EIA run.

.8. Determining optimal serum dilution for rubella IgGvidity assay

In preliminary experiment a set of 20 sera, was compsing different dilutions of sera for avidity assay (methescribed below). Each sample was tested at six dilu1:100, 1:200, 1:400, 1:800, 1:1600 and 1:3200), andilution at four replicates. The highest reproducibility ra99%) for nearly all cases were found at 1:200 dilutioherefore, all sera were tested at that dilution.

t

;f

All avidity index values obtained by testing of both pary infection and reinfection panels were analyzedlassification and Regression Tree (CART) and Roc Ctatistical methods (Lewis, 2000) using SPLUS and SPSoftwares. Then assessing parameters including senspecificity and predictive values in differential diagnosiubella primary infection and reinfection for both rubella Ignzyme immunoassay and rubella IgG avidity assayalculated. For each parameter, 95% confidence limitsalculated. To compare two proportions the Chi-squareas used to determine the statistical significance bet

he parameters of the two measurements. Thep-values coresponding top < 0.05 indicated a statistically significaifference between the two assays.

.11. Calculation of the avidity cut-off point

All avidity index values obtained from testing of borimary rubella infection and reinfection panels were a

yzed by Classification and Regression Tree (CART) andurve statistic methods. Both panels’ sera were reclasased on avidity index values to low and high avidity groy Classification and Regression Tree (CART) method.ith avidity index values <53% were classified as samontaining low avidity rubella IgG and sera with avidndex values >53% were classified as samples containingvidity rubella IgG. Classification carried out without aisclassification error rate (i.e. all cases of primary infecanel are regarded as low avidity and all reinfection pera are determined as high avidity). Similar results obta

62 R. Hamkar et al. / Journal of Virological Methods 130 (2005) 59–65

by Roc Curve analysis. This analysis revealed that the high-est sensitivity and specificity of rubella-specific IgG avidityassay were achieved only at cut-off point equal to 53%.

3. Results

3.1. Assessment of rubella IgG avidity assay and rubellaIgM-EIA sensitivity

Sensitivity of the rubella IgG avidity assay and rubellaIgM-EIA were determined using primary anti-rubella anti-body response (primary infection) panel and the results arepresented inTables 1 and 2. As shown inTable 1, rubella-specific IgM was detected in 167 (77.4%) serum samplesout of 216 serum samples from primary rubella infection,while rubella IgM-EIA failed to detect any rubella-specificIgM in 49 (22.6%) cases of rubella primary infection panel.However, at avidity cut-off point equal to 53%, rubella IgGavidity assay detected all primary rubella infection panel seraas low avidity rubella IgG response. The sensitivity of bothtests, ranges from 77.4% for rubella IgM-EIA to 100% forrubella IgG avidity assay, and the difference between thesetwo assays is statistically significant. The failure of IgM-EIAto detect rubella IgM in a proportion of cases may arise fromdiminishing rubella-specific IgM from patients serum samplew s pre-v si ofc ditya tiono

3I

ci-fi anti-

rubella antibody response (rubella reinfection) panel and theresults are shown inTables 1 and 2. As shown inTable 1,rubella-specific IgM was detected in six cases (2.1%) ofrubella reinfection panel sera, but 278 (97.9%) remainingcases were negative for rubella-specific IgM, while highavidity anti-rubella IgG was detected in all cases of rubellareinfection. The specificity of both tests, ranging from 97.8%for rubella IgM-EIA to 100% for rubella IgG avidity assay,and the difference between these two assays is not statisticallysignificant.

3.3. Primary infection and reinfection among naturallyacquired rubella

According to a calculated avidity cut-off point value,high avidity rubella IgG were detected in 26 cases out of100 rubella panel sera (minimum avidity index, AI = 71%),while in remaining 74 serum samples low avidity rubellaIgG were determined (maximum avidity index, AI = 28%).Among high avidity cases, which regarded as rubella reinfec-tion, 23 cases (88%) had a documented history of measles,mumps and rubella (MMR) vaccination at 15 months ofage and they were all above 14 years old. Their symp-tomatic rubella reinfection may have been resulted fromsecondary failure of the vaccine (i.e. diminishing of vaccine-i rs).I tions eirs ltedf is-r ges.A aryi erea ellaI pri-m llai

TD r the ru

P tal

%

R 22.6R 97.9

TR d rube

A pecifici

R 00R 7.9 (96,

hen their sera were collected, but these results confirmiously reported findings (Tipples et al., 2004). These resultndicated that, IgM-EIA has low sensitivity for detectionurrent or recent rubella infection and rubella IgG avissay is more sensitive than rubella IgM-EIA for detecf rubella primary infection.

.2. Assessment of rubella IgG avidity assay and rubellagM-EIA specificity

The rubella IgG avidity assay and rubella IgM-EIA specity assessments were carried out using the secondary

able 1istribution of rubella IgG avidity assay and rubella IgM-EIA results fo

anels Rubella IgM-EIA

Positive Negative

# % #

ubella primary infection 167 77.4 49ubella reinfection 6 2.1 278a Total number of both panels = 500.b Low avidity.c High avidity.

able 2elative sensitivity, specificity, predictive values of rubella IgM-EIA an

ssay Sensitivity S

ubella IgG avidity assay 100 1ubella IgM-EIA 77.4 (71.8, 83) 9a Numbers in parentheses are 95% confidence intervals.

nduced immunity during a period of at least 13 yean three other high avidity cases, the rubella vaccinatatus or history of rubella infection were unknown; thymptomatic rubella reinfection may have been resurom diminished immunity against rubella induced by demembered vaccine or rubella infection during early all low avidity cases that are regarded as rubella prim

nfection, were from non-vaccinated people and they wll <10 years old. These findings indicate that the rub

gG avidity assay is also competent to distinguishary infection from reinfection in naturally acquired rube

nfection.

bella primary infection and secondary reinfection panelsa

Rubella IgG avidity assay To

LAb HAc

# % # %

216 100 0 0 2160 0 284 100 284

lla IgG avidity assaya

ty PPV NPV

100 10099.6) 96.9 (94.3, 99.5) 82.9 (80.6, 85.2)

R. Hamkar et al. / Journal of Virological Methods 130 (2005) 59–65 63

4. Discussion

The detection of IgM antibodies in serum samples byenzyme immunoassay is used commonly for diagnosingacute rubella infections (Best et al., 2002). The rubella-specific IgM antibodies are usually detected within 4 daysafter the onset of rash and for 4–8 weeks thereafter (Bestand Banatvala, 2000). However, when IgM is not detected,a recent infection cannot be ruled out absolutely, since theextent of IgM antibody responses varies with each individ-ual and some patients may not produce sufficient amounts ofIgM antibodies to be detected, although they respond withdetectable IgG antibodies (Inouye et al., 1984). Furthermore,high concentration of specific IgM may also be found insera from proven cases of rubella reinfection (Thomas et al.,1992). Non-specific reactivity in IgM assays may also occurand persistent specific IgM reactivity has been detected inthe absence of primary rubella infection or rubella reinfection(Thomas et al., 1992). This issue of false positive and negativerubella IgM results is especially important when investigat-ing suspected rubella in pregnant women because of the riskof congenital rubella syndrome and the associated criticalclinical management decisions, so additional diagnostic testsshould be used in such situations (Best et al., 2002; Enderset al., 1985; Health Canada, 2002; Tipples et al., 2004).Regarding the above limitations of IgM-EIA, additional con-fi testss , IgGs tibodya iag-n sta althC et al.1

llow-i ta ism l riskf et n ise atu-r ions( talr ocu-m ni ireds fre-q ctedia aryr as fr thefi ;B d-i IgM

reactivity to confirm or exclude a recent rubella infection orreinfection.

The most critical point of IgG avidity assay is precise cal-culation of the cut-off point for differentiating low avidity IgGfrom high avidity IgG. Therefore, it was essential that we havehad panels with certain immune status before vaccination.In this study, all samples of both panels were composed ofpaired (before/post-measles–rubella vaccine) sera obtainedfrom ages between 5- and 25-year-old populations in the con-text of measles–rubella immunity assay in Iran. Therefore,the anti-rubella IgG status of all cases before measles–rubellavaccination, were known. In the present study, a large num-ber of these samples were tested by IgG avidity assay, andthe results were analyzed by two statistical methods (Classi-fication and Regression Tree and Roc-curve). Avidity indexequal to 53% was determined as cut-off point by either oftwo methods. At this cut-off point, the most desired sensi-tivity and specificity (100 and 100%) were achieved to IgGavidity assay.

Rubella IgM specific antibody was detected in 77.4% ofprimary infection panel sera, but in 22.6%, IgM-EIA pro-vided negative results. On the other hand, in these casesIgM-EIA could not demonstrate primary infection. In onestudy, it was shown that the percentage of rubella-specificIgM positive sera decreased from 100% at 15–28 days afterthe onset of infection through 71, 28 and 9% at 1–2, 2–3 and3 neg-a itys nths.A stills ,t ftera tudy,w cuter nseto ntil1f gGc to theI osticme

.1%o IgG.T train.A bed thes oft imaryr preg-n ormr dw itht eg-n in-f ,

rmatory tests are therefore required. Other laboratoryuch as reverse transcription polymerase chain reactionerology on paired acute and convalescent sera, and anvidity testing may also be used for rubella laboratory dosis and confirmatory testing (Akingbade et al., 2003; Bend Banatvala, 2000; Best et al., 2002; CDC, 2001; Heanada, 2002; Rousseau and Hedman, 1988; Thomas992; Tipples et al., 2004).

Rubella reinfection occurs and has been reported fong both natural and vaccine-induced immunity (Thomas el., 1995). Reinfection following immunization to rubellaostly without consequences and presents a minima

or the fetus (Aboudy et al., 1997). The risk of intrauterinransmission of virus associated with maternal reinfectioxtremely small, since maternal immunity (vaccine or nally induced) usually protects against intrauterine infectAboudy et al., 1997). Only very few cases of congeniubella syndrome after maternal reinfection have been dented (Thomas et al., 1995). Classically rubella reinfectio

s proved by increase in rubella-specific IgG titer in paera in the absence of rubella IgM, but it is documenteduently that specific rubella IgM antibody can be dete

n rubella reinfection cases (Bottiger and Jensen, 1997). Thevidity of IgG antibodies test seems to differentiate primubella infection from past infection and reinfection iningle serum sample (Inouye et al., 1984). An increase oubella IgG antibody avidity was seen to occur duringrst 3 months after the onset of rash (Aboudy et al., 1997ottiger and Jensen, 1997). Measurement of rubella IgG avi

ty is a good supplemental test for cases with rubella

,

–4 months, respectively. After four months, all becametive for rubella-specific IgM antibody. However, low avidpecific IgG was detected in all of sera taken at 3 mot 3–4 months 91% and at 5–7 months 21% of serahowed low avidity (Thomas et al., 1992). In another studyhe time course of maturation of rubella IgG avidity acute infection was demonstrated. In above-mentioned sell-characterized serial samples from 15 patients with a

ubella were tested and followed up to 5 months after the of rash. A high avidity index (>60%) was not observed u3 weeks after infection (Bottiger and Jensen, 1997). There-

ore, the differential assay of high avidity and low avidity Ian be used as an alternative or a complementary testgM antibody assay and is gaining popularity as a diagnethod for the assessment of the time of infection (Korhonent al., 1999).

Our results revealed that IgM could be detected in 2f rubella reinfection panel cases that have high avidityhese cases displayed a reinfection by rubella vaccine ss mentioned above rubella-specific IgM antibody canetected in both primary infection and reinfection, thuspecificity of IgM-EIA decreases to differential diagnosishese cases. It is essential that correct diagnosis of prubella can be achieved for the management of theant women with a recent rash or contact with a rubellifash illness (Inouye et al., 1984). With the introduction anidespread use of the rubella vaccine, it is likely that, w

ime, relatively fewer cases of rubella infection during prancy will be primary infection and more will be rubella re

ection (Aboudy et al., 1997; Thomas et al., 1995). Therefore

64 R. Hamkar et al. / Journal of Virological Methods 130 (2005) 59–65

detection of IgM alone cannot differentiate primary infectionfrom reinfection. For these cases, the measurement of IgGavidity is very useful, since in the case of recent primaryinfection, IgG is low avidity and in the case of reinfection,IgG is high avidity (Enders and Knotek, 1989; Gutierrezet al., 1999; Thomas and Morgan-Capner, 1988; Thomaset al., 1992).

In this study, rubella IgM-EIA was compared with rubellaIgG avidity assay. Results showed that sensitivity of IgM-EIA for diagnosing of primary rubella infection is 77.4% andIgM-EIA could not detect all the primary infection. How-ever, IgM-EIA has appropriate specificity (97.9%) and itcan detect non-primary rubella cases. Similar results havedemonstrated in other studies. Previously the sensitivity andspecificity of seven commercial rubella-specific IgM kitswere assessed and it was shown that sensitivity of the mostkits was within the ranges of 66.4–78.9% (median 73.9%).Specificity of these kits estimated to be between ranges of85.6 and 96.1% (median 92.6%). In above-mentioned study,sensitivity and specificity of IgM-EIA for detection of rubellainfection had shown by Behring’s indirect EIA kit to be 75.9and 98.7%, respectively (Tipples et al., 2004). Another studyhas evaluated the sensitivity and specificity of 15 commer-cial rubella-specific IgM kits. Specificity of these kits wasfound to be in the range of 82–98% and their sensitivity was63–92%. In above-mentioned study sensitivity and specificityo pec-t tedt 6

andn nd8 low,s ver-a asess ults,ae la-s fort andp ein-f

t, thea llai rallyi vid-i ocu-m tion,a s pri-m

ni-t meds nitalr hasb nths,i ntsa ldren

over 18 months of age (Best and Banatvala, 2000). Studieshad revealed that the maturation of the immune response torubella virus is abnormally slow in congenital rubella casesboth in terms of the isotype switching and especially thedevelopment of high avidity specific IgG1 (Thomas et al.,1993). Thus, avidity studies can be useful for serologicaldiagnosis of congenital rubella even after disappearance ofspecific IgM, because of low avidity IgG can be detected upto the age of 3 years (Best and Banatvala, 2000; Herne et al.,1997).

With the introduction and widespread use of the rubellavaccine, it is likely that with time, relatively fewer cases ofrubella infection during pregnancy will be primary infec-tions and more will be rubella reinfection. Therefore, it willbe more appropriate that rubella IgG avidity assay be usedas an alternative or complementary assay to testing rubellasuspected cases especially in pregnant women, because ofthis the test is likely more precise and sensitive than rubellaIgM-EIA and can distinguish primary rubella infection fromrubella reinfection.

References

Aboudy, Y., Fogel, A., Barnea, B., Mendelson, E., Yosef, L., Frank, T.,Shalev, E., 1997. Subclinical rubella reinfection during pregnancy fol-

76.A idity

ellaary),

B tvala,logy,

B Fetal

B nette,ick-alth

B .M.,n of325,

B diesn in

B over.

C an-, and12),

C to. Rev.

C ontroltries.Bull.

E IgGeren-

f Behring’s kit has been reported to be 79 and 96%, resively. In this study, a higher value of sensitivity is attribuo IgM capture EIA kits (Hudson and Morgan-Capner, 199).

Our study shows that positive predictive value (PPV)egative predictive value (NPV) of IgM-EIA are 96.9 a2.9%, respectively. When the prevalence of rubella isuch as in countries with high rubella vaccination coge, the positive predictive value of IgM testing decreuch that there is a significant risk of false positive resnd additional confirming tests are therefore required (Bestt al., 2002; Tipples et al., 2004). The measurement of rubelpecific IgG avidity is a specific and sensitive methodhe serological diagnosis of recent primary infectionrovides the distinction between primary infection and r

ection possible (Gutierrez et al., 1999).Testing the natural rubella panel sera revealed tha

vidity cut-off point, which obtained from primary rubenfection and reinfection panels are applicable to natunfected patients. Regarding calculated cut-off point of aty, IgG avidity assay can detect all patients who had d

ented history of rubella vaccination, as rubella reinfecnd illness of non-vaccinated people are diagnosed aary rubella infection.Rubella IgG avidity assay is also useful in conge

al rubella syndrome cases. Without appropriately tipecimen collection, serological confirmation of congeubella infection may be a problem. Rubella-specific IgMeen detected in all confirmed cases until the age of 3 mo

n 86% of infants aged 3–6 months, and in 62% of infaged between 1 year and 18 months, but rarely in chi

lowed by transmission of virus to the fetus. J. Infect. 34, 273–2kingbade, D., Cohen, B.J., Brown, D.W., 2003. Detection of low-av

immunoglobulin g in oral fluid samples: new approach for rubdiagnosis and surveillance. Clin. Diagn. Lab. Immunol. (Janu189–190.

est, J.M., Banatvala, J.E., 2000. Rubella. In: Zuckerman, A.J., BanaJ.E., Pattison, J.R. (Eds.), Principles and Practice of Clinical Viro4th ed. John Wiley and Sons, UK, pp. 387–415.

est, J.M., Banatvala, J.E., Morgan-Capper, P., Miller, E., 1998.infection after maternal reinfection. Br. Med. J. 299, 773–775.

est, J.M., O’Shea, S., 1995. Rubella Virus. In: Lennette, E.H., LenD.A., Lennette, E.T. (Eds.), Diagnostic Procedures for Viral, Rettsial and Clamydial Infections, 7th ed. American Public HeAssociation, Washington, DC, pp. 583–600.

est, J.M., O’Shea, S., Tipples, G., Davies, N., Al-Khusaiby, SKrause, A., Hesketh, L., Jin, L., Enders, G., 2002. Interpretatiorubella serology in pregnancy, pitfalls and problems. Br. Med. J.147–148.

odeus, M., Feyder, S., Goubau, P., 1998. Avidity of IgG antibodistinguishes primary from non-primary cytomegalovirus infectiopregnant women. Clin. Diagn. Virol. 9, 9–16.

ottiger, B., Jensen, P.I., 1997. Maturation of rubella IgG aviditytime after acute rubella infection. Clin. Diagn. Virol. 8, 105–111

DC, 2001. Control and prevention of rubella: evaluation and magement of suspected outbreaks, rubella in pregnant womensurveillance for congenital rubella syndrome. MMWR 50 (RR-1–23.

utts, F.T., Brown, D.W.G., 1995. The contribution of field testsmeasles surveillance and control: a review of available methodsMed. Virol. 5, 35–40.

utts, F.T., Robertson, S.E., Diaz-Ortega, J.L., Samuel, R., 1997. Cof rubella and congenital rubella syndrome in developing counPart 1. Burden of disease from congenital rubella syndrom.World Health Organ. 75, 55–68.

nders, G., Knotek, F., 1989. Rubella IgG total antibody avidity andsubclass-specific antibody avidity assay and their role in the diff

R. Hamkar et al. / Journal of Virological Methods 130 (2005) 59–65 65

tiation between primary rubella and rubella reinfection. Infection 17,218–226.

Enders, G., Knotek, F., Pacher, U., 1985. Comparison of various sero-logical methods and diagnostic kits for the detection of acute, recent,and previous rubella infection, vaccination, and congenital infections.J. Med. Virol. 16, 219–232.

Gutierrez, J., Rodriguez, M.J., Ory, F.De., Piedrola, G., Maroto, M.C.,1999. Reliability of low-avidity IgG and of IgA in the diagnosis ofprimary infection by rubella virus with adaptation of a commercialtest. J. Clin. Lab. Anal. 13, 1–4.

Health Canada, 2002. Proceeding of a Meeting of the Expert AdvisoryGroup on Rubella in Canada. Can. Commun. Dis. Rep. 28 (S4), 1–24.

Hedman, K., Hietala, J., Tiilikainen, A., Hartikainen-Sorri, A.-L., Raiha,K., Suni, J., Vaananen, P., Pietilainen, M., 1989. Maturation ofimmunoglobulin G avidity after rubella vaccination studied by anenzyme linked immunosorbent assay (avidity-ELISA) and by haemol-ysisi typing. J. Med. Virol. 27, 293–298.

Herne, V., Hedman, K., Reedik, P., 1997. Immunoglobulin G avidity inthe serodiagnosis of congenital rubella syndrome. Eur. J. Microbiol.Infect. Dis. 16 (10), 763–766.

Hinman, A.R., Irons, B., Lewis, M., Kandola, K., 2002. Economic anal-yses of rubella and rubella vaccines: a global review. Bull. WorldHealth Organ. 80, 264–270.

Hudson, P., Morgan-Capner, P., 1996. Evaluation of 15 commercialenzyme immunoassays for the detection of rubella-specific IgM. Clin.Diagn. Virol. 5, 21–26.

Inouye, S., Hasegawa, A., Matsuno, Sh., Katow, Sh., 1984. Changes inantibody avidity after virus infections: detection by an immunosorbentassay in which mild protein-denaturing agent is employed. J. Clin.Microbiol. 20 (3), 525–529.

Korhonen, M.H., Brunstein, J., Haario, H., Katnikov, A., Rescaldani, R.,tilityb.

L treeg of

the Society for Academic Emergency Medicine, San Francisco, CA.Department of Emergency Medicine, Harbor-UCLA Medical Center,Torrance, CA.

Robertson, S.E., Cutts, F.T., Samuel, R., Diaz-Ortega, J.L., 1997. Controlof rubella and congenital rubella syndrome in developing countries.Part 2. Vaccination against rubella. Bull. World Health Organ. 75,69–80.

Rousseau, S., Hedman, K., 1988. Rubella infection and reinfection dis-tinguish by avidity of IgG. Lancet 14, 1108–1109.

Thomas, H.I.J., Barrett, E., Hesketh, L.M., Wynne, A., Morgan-Capner,P., 1999. Simultaneous IgM reactive by EIA against more than onevirus in measles, parvovirus B19 and rubella infection. J. Clin. Virol.14, 107–118.

Thomas, H.I., Morgan-Capner, P., Cradock-Watson, J.E., Enders, G., Best,J.M., O’Shea, S., 1993. Slow maturation of IgG1 avidity and persis-tence of specific IgM in congenital rubella: implications for diagnosisand immunopathlogy. J. Med. Virol. 41 (3), 196–200.

Thomas, H.I.J., Morgan-Capner, P., Enders, G., O’Shea, S., Caldicott, D.,Best, J.M., 1992. Persistence of specific IgM and low avidity specificIgG1 following primary rubella. J. Virol. Meth. 39, 149–155.

Thomas, H.I.J., Morgan-Capner, P., 1988. Rubella-specific IgG subclassavidity ELISA and it’s in the differentiation between primary rubellareinfection. Epidemiol. Infect. 101, 591–598.

Thomas, H.I.J., Morgan-Capner, P., 1991. Rubella-specific IgG1 avidity:a comparison of methods. J. Virol. Meth. 31, 219–228.

Thomas, H.I.J., Charlett, A., Cubie, H.A., 1995. Specific IgG1 aviditymaturation after rubella vaccination: a comparison with avidity mat-uration after primary infection with wild rubella virus. Serodiagn.Immunother. Infect. Dis. 7, 75–80.

Tipples, G., Hamkar, R., Mohktari-Azad, T., Gray, M., Ball, J., Head, C.,Ratnam, S., 2004. Evaluation of rubella IgM enzyme immunoassays.

W miol.

W Epi-

Hedman, K., 1999. A new method with general diagnostic ufor the calculation of immunoglobulin G avidity. Clin. Diagn. LaImmuonol., 725–728.

ewis, R.J., 2000. An introduction to classification and regression(CART) analysis. In: Proceedings of the 2000 Annual Meetin

J. Clin. Virol. 30, 233–238.HO, 2000a. Rubella vaccines: WHO position paper. Weekly Epide

Rec. 20, 161–72.HO, 2000b. Preventing congenital rubella syndrome. Weekly

demiol. Rec. 36, 290–295.