S

Mi

MAa

b

c

d

a

ARRA

KURHDOS

1

ctdso

CT

fG

t

1h

Journal of Clinical Virology 58 (2013) 718– 721

Contents lists available at ScienceDirect

Journal of Clinical Virology

jo u r n al hom epage: www.elsev ier .com/ locate / j cv

hort Communication

easure of herpesvirus-specific ocular antibody productionn patients with uveitis

agdalena Bojanovaa,1, Bahram Bodaghib,c, Naila Hannachia,2, Thomas Jouffroyb,udrey Felb, Phuc Le Hoangb,c, Flore Rozenberga,d,∗

Assistance Publique – Hôpitaux de Paris, Hôpital Cochin, Service de Virologie, Paris, FranceAssistance Publique – Hôpitaux de Paris, Hôpital Pitié-Salpétrière, Service d’Ophtalmologie, Paris, FranceUPMC, DHU ViewMaintain, Paris, FranceUniversité Paris Descartes, Paris, France

r t i c l e i n f o

rticle history:eceived 10 June 2013eceived in revised form 27 August 2013ccepted 9 October 2013

eywords:veitisetinitiserpesvirusiagnosiscular antibody productionerology

a b s t r a c t

Background: The amount of specific antiviral IgG in aqueous humour (AH) provides a major contributionto the diagnosis of herpesvirus uveitis. Ocular antibody production is often evaluated by comparing levelsof specific and total IgG in serum and AH. The small volume of AH is a major limit for diagnosis.Objectives: To simplify the measure of ocular antibody production, we tested the quotient of serum/AHratios of specific and control antiviral IgG, using automated quantitative serology methods on minimalvolumes of AH, in confirmed and suspected herpesvirus uveitis.Study design: Serum and AH samples from herpesvirus PCR-positive uveitis patients, and from PCR-negative cases who were highly suspected to have viral uveitis were retrospectively analysed for ocularproduction of specific antiviral IgG using 40 �l of AH, and quantitative Enzygnost ELISA-based methods.Cataract and Fuchs cyclitis cases were used as controls.Results: Ocular production of specific antiviral IgG was demonstrated in 32 (51.6%) of 62 herpesvirus

PCR-positive uveitis cases, in none of 42 controls, and in 21 (55.2%) of 38 PCR-negative cases clinicallysuspected to have herpesvirus uveitis. The test had absolute specificity, and its sensitivity depended onthe virus, pathology and timing of sampling.Conclusion: Ocular antibody production can be measured by simple quantitative ELISA-based methods onserum and minimal volumes of AH. This specific and sensitive test, implemented in the routine virologylaboratory should help the diagnosis and specific antiviral therapy management of herpesvirus uveitis.

. Background

Herpes simplex virus (HSV), varicella zoster virus (VZV) andytomegalovirus (CMV) are major causes of uveitis [1–9]. Diagnos-ic confirmation by early PCR on aqueous humour (AH) or lateretection of ocular antibody production (OAP) [10–13] is crucial for

pecific antiviral therapy management [2,14–16]. Limited volumesf AH hamper exhaustive investigation [17–19].

∗ Corresponding author at: Service de Virologie, Bâtiment Jean Dausset, Hôpitalochin, 27 rue du Faubourg Saint Jacques, 75679 Paris cedex 14, France.el.: +33 01 58 41 33 19; fax: +33 01 58 41 33 18.

E-mail address: flore.rozenberg@cch.aphp.fr (F. Rozenberg).1 Present address: Universitätsklinikum Giessen und Marburg (UKGM), Institut

ür Medizinische Virologie, Labor für Virusdiagnostik, Schubertstrasse 81, 35392iessen, Germany.2 Present address: Laboratoire de Microbiologie-Immunologie, UR02SP13, Hôpi-

al Farhat Hached, Sousse, Tunisia.

386-6532/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jcv.2013.10.013

© 2013 Elsevier B.V. All rights reserved.

2. Objectives

We reassessed the measure of OAP by establishing the quotientof serum/AH ratios of different antiviral IgGs using quantitativeELISA immunoassays.

3. Study design

3.1. Patients and samples

From 2003 to 2009, serum (1 ml) and AH (100 �l) from patientssuspected with HSV, VZV or CMV uveitis was drawn simultaneouslyfor serology and AH PCR, then stored frozen at −20 ◦C. Clinical andvirological diagnoses were confronted. Samples from 100 patients(65 males and 35 females from 13 to 82 years) were retrieved, i.e.

55 PCR-positive anterior uveitis (AU), 7 PCR-positive acute retinalnecrosis (ARN), and 38 PCR-negative cases, plus 12 cataract and30 Fuchs iridocyclitis controls [20]. Patients gave their informedconsent about investigations.

Clinica

3

fcGCfac

3

aEBUlSdfFosfC

fu

3

w

FIa

M. Bojanova et al. / Journal of

.2. PCR methods

Classical or real time HSV, VZV and CMV PCR was per-ormed as described [3,4,21–23] and since 2008 usingonsensus 5′TTTGTGTACATGTCCCCGTTTTAC3′ and 5′AGGTCGC-CGCGTAGAAG3′ primers and specific 5′FAMCACACCGAA-ACACCA-MGB3′ and 5′FAM-CACACCGAGCACACCA-MGB3′ probesor HSV-1 and HSV-2, respectively. The sensitivity of real time PCRssays evaluated using European quality control panels (qualityontrol for molecular diagnostic QCMD), was 10 copies/sample.

.3. Measurement of ocular antibody production

ELISA immunoassays for HSV, VZV, CMV, and rubella or measless controls (depending on serum IgG levels) were performed usingnzygnost® anti-HSV, -VZV, -CMV, -rubella and -measles IgG (Dadeehring or Siemens) and the automated ETI-Max 3000 (DiaSorin).sing the � method, IgG levels quantified in international Units are

inear for optical density (OD) values between 0.195 and 2.195 [24].erial serum (1/231, 1/1155, 1/2310) and AH (1/23, 1/115, 1/230)ilutions were tested in same analytical runs where three-fold dif-erences are significant. Serum/AH coefficients (C) were calculated.or undetectable AH IgG, C was assigned the mean ratio value (300)f the physiological blood-aqueous barrier [25]. The C′ quotient oferum/AH C ratios differentiated passive serum-AH diffusion trans-er or non-specific AH IgG from specific OAP, defined by C < 80 and′ ≥ 3, as reported for Goldmann–Witmer coefficient (GWC) [10].

In addition to cataract and Fuchs samples, PCR-positive samplesor any of the tested viruses with or without confirmed OAP weresed as negative controls for the others [17].

.4. Statistical analysis

The t and Mann–Whitney test were used for statistical analysis,ith significant p value < 0.05.

A.

evitisopRCPVSH

(n=18)

Median C 32

Median C’ 1.7

B. C.

HSV P

CR +

HSV P

CR -

0.1

1

10

100

1000 ns

C R

ati

o H

SV

ig. 1. Ocular production of HSV IgG in HSV anterior uveitis cases. The C ratio of serum/gG were compared in PCR-positive versus PCR-negative HSV anterior uveitis cases (A). Dnterior uveitis.

l Virology 58 (2013) 718– 721 719

4. Results

4.1. OAP in PCR-positive cases

Among 18 HSV AU cases, HSV IgG were compared to VZV(n = 17) or CMV (n = 1) IgG. HSV-specific OAP was confirmed in6 cases by decreased C (median = 29.5, range 1–77) and elevatedC′ (median = 6.3, range 3.9–27.3). Nine cases had dual decreaseof HSV/VZV IgG C ratios, and 3 had undetectable ocular HSV IgG(Fig. 1). None of 25 controls (1 Fuchs syndrome, 7 CMV and 9 VZVuveitis, and 8 VZV retinitis cases) had HSV OAP.

Among 11 VZV AU cases, VZV IgG were compared to HSV (n = 9),CMV (n = 1) and measles (n = 1) IgG. VZV-specific OAP was con-firmed in 4 cases by decreased C (median = 15, range 1–23) andelevated C′ (median = 17.2, range 3.4–60), and 7 cases had unde-tectable ocular VZV IgG. Among 54 controls (12 cataract, 17 CMVand 17 HSV uveitis, 8 Fuchs cyclitis cases), none had VZV OAP.

In 7 VZV ARN cases, VZV and HSV IgG were compared. VZV-specific OAP was confirmed in all by low C (median = 2, range 1–35)and elevated C′ (median = 13, range 6–16), despite presence of ocu-lar HSV IgG (median C ratio = 30) possibly reflecting inflammationor cross stimulation of VZV and HSV immunity [26] (Fig. 2).

Among 26 CMV uveitis patients, CMV IgG were compared toHSV (n = 8), VZV (n = 17) or Rubella (n = 1) IgG. CMV-specific OAPwas confirmed in 15 cases (58%) by low C (median = 13, range 3–74)and elevated C′ (mean = 13, range 3.1–71). Four patients had equiv-ocal results. Seven others had undetectable ocular CMV IgG, as 32controls (9 cataract, 21 Fuchs syndrome, 1 VZV and 1 HSV uveitiscases).

Overall, specific OAP was unambiguously demonstrated in25/55 (45%) PCR-positive AU and 7/7 VZV ARN cases.

4.2. OAP in PCR-negative cases

In 13/24 cases (10 AU and 3 ARN) suggesting HSV or VZV infec-tion, HSV-specific OAP was demonstrated by low C (median = 17,

evitagenRCPVSH p

(n=10)

18.5 ns

10.5 0.001

HSV P

CR +

HSV P

CR-

0.1

1

10

100 p=0.001

C' q

uo

tien

t

AH levels of HSV IgG and the C′ quotient of serum/AH ratios of HSV versus controlistribution of C (B) and C′ quotient (C) values of PCR positive versus negative HSV

720 M. Bojanova et al. / Journal of Clinical Virology 58 (2013) 718– 721

A.

VZV uve itis VZV ARN p

(n=11) (n=7)

Median C 117 2 0.009

Median C’ 2 13 ns

B. C.

VZV

ante

rior

uveiti

s

VZV

ARN

0

100

200

300

400p=0,00 9

C r

ati

o V

ZV

VZV

ante

rior

uveiti

s

VZV

ARN

0

20

40

60

80ns

C' quotient

Fig. 2. Ocular production of VZV IgG in VZV anterior uveitis versus VZV ARN. OAP was comC (B) and C′ quotient (C) values in VZV uveitis versus retinitis cases.

Table 1Contribution of PCR and ocular antibody production to diagnosis of herpesvirus-associated uveitis.

Virus PCR+ PCR+ OAP+ PCR–OAP+ Total

HSV 12 (39%) 6 (19%) 13 (42%) 31VZV 7 (30%) 11 (47%) 5 (23%) 23CMV 11 (38%) 15 (52%) 3 (10%) 29

O

rpwCoh(Ic

4

hfipvr

5

rr

Total 30 (36%) 32 (38%) 21 (25%) 83

AP: ocular antibody production.

ange 7–73) and elevated C′ (median = 9.2, range 3.9–27.3) com-ared to VZV. In 5 cases (4 AU and 1 ARN), VZV-specific OAPas evidenced by low C (median = 22, range 1–44) and elevated

′ (median = 6.8, range 5.3–63) relative to HSV (n = 3), CMV (n = 1)r measles (n = 1). Only 3/14 patients suspected with CMV uveitisad CMV-specific OAP. Overall, OAP was demonstrated in 21/3855%) of PCR negative cases. None of 30 Fuchs controls had oculargG to CMV (n = 21), VZV (n = 8) or HSV (n = 1). None of 12 cataractontrols had detectable IgG towards any virus.

.3. Time-course evolution of OAP

OAP contributed to 63% of all diagnoses, 25% of which wouldave been missed without this analysis (Table 1). OAP was more

requent (75% versus 33% of cases) and significantly higher (Fig. 1)n suspected PCR-negative than PCR-positive HSV AU cases. In twoatients with CMV AU, OAP was detected twice at 1 or 2-year inter-als. Levels of IgG were higher in VZV retinitis than VZV AU, withouteaching significance (Fig. 2).

. Discussion

The serological diagnosis of ocular and neurological infectionsaises similar methodological concerns related to low volumes ofelevant fluids, multiple causal pathogens, and interpretation of

pared in VZV uveitis and retinitis cases. C and C′ same as Fig. 1 (A). Distribution of

results [27–29]. Forty �l AH sufficed to measure the C′ quotientof serum/AH ratios of IgG directed to two viruses. Given OAP per-sistence, C′ can be measured retrospectively. In the absence of goldstandard, the sensitivity of C′ is hardly calculated [30]. The delayfrom onset of disease to AH sampling was unknown based on latereferral of patients to our tertiary centre. OAP possibly dependedon timing [11] because OAP was more frequent and higher inHSV/VZV-suspected PCR-negative than HSV/VZV PCR-positive AUcases, suggesting the latter were sampled earlier in disease than theformer. OAP was detected in 57.7% of CMV PCR-positive and only21% of CMV suspected PCR-negative cases, may be due to atypicalclinical presentation of CMV uveitis. Overall, the sensitivity of C′

was equivalent to the reported sensitivity (47–85%) of GWC [16,30],which measures the quotient between specific antibody IgG andtotal immunoglobulin AH/serum ratios. In 175/185 published cases,C′ was concordant to GWC which gave false-negative and false-positive results due to high serum IgG values and polyclonal IgGproduction, respectively [31].

OAP was directed to the virus diagnosed by PCR. In PCR-negative AU, OAP was specific to the virus clinically suspected.Specificity was absolute, as shown using inflammatory controls.C′ accurately differentiated specific OAP from blood-aqueous leak-age despite the presence of dual IgG as reported [2,3,31], and as forblood–brain barrier [21,27,32,33]. Dual concentration of HSV/VZVIgG observed in some cases could be due to inflammation andblood-aqueous leakage, polyclonal stimulation, specific cross acti-vation of VZV/HSV immune responses, or dual viral reactivation asin neurological infections [34]. Non cross reactive serological assayscould discriminate these hypotheses [35]. However, OAP towardsmultiple pathogens was described using GWC, the highest valueindicating the causative pathogen [11].

In conclusion, the simple, sensitive and specific measure of OAPshould help the specific antiviral therapy management of uveitis,to avoid major complications such as secondary glaucoma and per-manent visual loss.

Clinica

F

C

E

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

M. Bojanova et al. / Journal of

unding

None.

ompeting interests

None declared.

thical approval

Not required.

eferences

[1] Jakob E, Reuland MS, Mackensen F, Harsch N, Fleckenstein M, Lorenz HM, et al.Uveitis subtypes in a German Interdisciplinary Uveitis Center – analysis of 1916patients. J Rheumatol 2009;36:127–36.

[2] de Schryver I, Rozenberg F, Cassoux N, Michelson S, Kestelyn P, Lehoang P,et al. Diagnosis and treatment of cytomegalovirus iridocyclitis without retinalnecrosis. Br J Ophthalmol 2006;90:852–5.

[3] Bodaghi B, Rozenberg F, Cassoux N, Fardeau C, LeHoang P. Nonnecrotizing her-petic retinopathies masquerading as severe posterior uveitis. Ophthalmology2003;110:1737–43.

[4] Tran TH, Rozenberg F, Cassoux N, Rao NA, LeHoang P, Bodaghi B. Polymerasechain reaction analysis of aqueous humour samples in necrotising retinitis. BrJ Ophthalmol 2003;87:79–83.

[5] Dworkin LL, Gibler TM, Van Gelder RN. Real-time quantitative polymerasechain reaction diagnosis of infectious posterior uveitis. Arch Ophthalmol2002;120:1534–9.

[6] Van Gelder RN, Willig JL, Holland GN, Kaplan HJ. Herpes simplex virus type 2 asa cause of acute retinal necrosis syndrome in young patients. Ophthalmology2001;108:869–76.

[7] Ganatra JB, Chandler D, Santos C, Kuppermann B, Margolis TP. Viralcauses of the acute retinal necrosis syndrome. Am J Ophthalmol 2000;129:166–72.

[8] Knox CM, Chandler D, Short GA, Margolis TP. Polymerase chain reaction-basedassays of vitreous samples for the diagnosis of viral retinitis. Use in diagnosticdilemmas. Ophthalmology 1998;105:37–44 [discussion 44-5].

[9] Yamamoto S, Pavan-Langston D, Kinoshita S, Nishida K, Shimomura Y, Tano Y.Detecting herpesvirus DNA in uveitis using the polymerase chain reaction. Br JOphthalmol 1996;80:465–8.

10] de Boer JH, Verhagen C, Bruinenberg M, Rothova A, de Jong PT, BaarsmaGS, et al. Serologic and polymerase chain reaction analysis of intraocularfluids in the diagnosis of infectious uveitis. Am J Ophthalmol 1996;121:650–8.

11] De Groot-Mijnes JD, Rothova A, Van Loon AM, Schuller M, Ten Dam-Van LoonNH, De Boer JH, et al. Polymerase chain reaction and Goldmann–Witmer coef-ficient analysis are complimentary for the diagnosis of infectious uveitis. Am JOphthalmol 2006;141:313–8.

12] Westeneng AC, Rothova A, de Boer JH, de Groot-Mijnes JD. Infectious uveitis inimmunocompromised patients and the diagnostic value of polymerase chainreaction and Goldmann–Witmer coefficient in aqueous analysis. Am J Ophthal-mol 2007;144:781–5.

13] Errera MH, Goldschmidt P, Batellier L, Degorge S, Heron E, Laroche L, et al.Real-time polymerase chain reaction and intraocular antibody production forthe diagnosis of viral versus toxoplasmic infectious posterior uveitis. Graefes

Arch Clin Exp Ophthalmol 2011;249:1837–46.

14] Wensing B, Relvas LM, Caspers LE, Valentincic NV, Stunf S, de Groot-MijnesJD, et al. Comparison of rubella virus- and herpes virus-associated ante-rior uveitis: clinical manifestations and visual prognosis. Ophthalmology2011;118:1905–10.

[

l Virology 58 (2013) 718– 721 721

15] Rothova A, de Boer JH, Ten Dam-van Loon NH, Postma G, de Visser L, ZuurveenSJ, et al. Usefulness of aqueous humor analysis for the diagnosis of posterioruveitis. Ophthalmology 2008;115:306–11.

16] Errera MH, Goldschmidt P, Batellier L, Degorge S, Heron E, Laroche L, et al. Find-ings in detection of herpesviridae by polymerase chain reaction and intraocularantibody production in a case series of anterior uveitis. Ocul Immunol Inflamm2013;21:61–8.

17] de Groot-Mijnes JD, de Visser L, Zuurveen S, Martinus RA, Volker R, ten Dam-van Loon NH, et al. Identification of new pathogens in the intraocular fluid ofpatients with uveitis. Am J Ophthalmol 2010;150:628–36.

18] Drancourt M, Berger P, Terrada C, Bodaghi B, Conrath J, Raoult D, et al. Highprevalence of fastidious bacteria in 1520 cases of uveitis of unknown etiology.Medicine (Baltimore) 2008;87:167–76.

19] Kuchtey RW, Kosmorsky GS, Martin D, Lee MS. Uveitis associated with WestNile virus infection. Arch Ophthalmol 2003;121:1648–9.

20] de Groot-Mijnes JD, de Visser L, Rothova A, Schuller M, van Loon AM, WeersinkAJ. Rubella virus is associated with fuchs heterochromic iridocyclitis. Am JOphthalmol 2006;141:212–4.

21] Rozenberg F, Lebon P. Amplification and characterization of herpesvirus DNAin cerebrospinal fluid from patients with acute encephalitis. J Clin Microbiol1991;29:2412–7.

22] Campsall PA, Au NH, Prendiville JS, Speert DP, Tan R, Thomas EE. Detection andgenotyping of varicella-zoster virus by TaqMan allelic discrimination real-timePCR. J Clin Microbiol 2004;42:1409–13.

23] Leruez-Ville M, Ouachee M, Delarue R, Sauget AS, Blanche S, Buzyn A, et al.Monitoring cytomegalovirus infection in adult and pediatric bone marrowtransplant recipients by a real-time PCR assay performed with blood plasma. JClin Microbiol 2003;41:2040–6.

24] Dopatka HD, Giesendorf B. Single point quantification of antibody by ELISAwithout need of a reference curve. J Clin Lab Anal 1992;6:417–22.

25] Mondino BJ, Rao H. Complement levels in normal and inflamed aqueous humor.Invest Ophthalmol Vis Sci 1983;24:380–4.

26] Kuhn JE, Klaffke K, Munk K, Braun RW. HSV-1 gB and VZV gp-II crossreactiveantibodies in human sera. Arch Virol 1990;112:203–13.

27] Linde A, Klapper PE, Monteyne P, Echevarria JM, Cinque P, Rozenberg F, et al.Specific diagnostic methods for herpesvirus infections of the central nervoussystem: a consensus review by the European Union Concerted Action on VirusMeningitis and Encephalitis. Clin Diagn Virol 1997;8:83–104.

28] Monteyne P, Albert F, Weissbrich B, Zardini E, Ciardi M, Cleator GM, et al.The detection of intrathecal synthesis of anti-herpes simplex IgG antibod-ies: comparison between an antigen-mediated immunoblotting technique andantibody index calculations. European Union Concerted Action on Virus Menin-gitis and Encephalitis. J Med Virol 1997;53:324–31.

29] Reiber H, Lange P. Quantification of virus-specific antibodies in cerebrospinalfluid and serum: sensitive and specific detection of antibody synthesis in brain.Clin Chem 1991;37:1153–60.

30] Kongyai N, Sirirungsi W, Pathanapitoon K, Tananuvat N, Kunavisarut P,Leechanachai P, et al. Viral causes of unexplained anterior uveitis in Thailand.Eye (London) 2012;26:529–34.

31] Dussaix E, Cerqueti PM, Pontet F, Bloch-Michel E. New approaches to the detec-tion of locally produced antiviral antibodies in the aqueous of patients withendogenous uveitis. Ophthalmologica 1987;194:145–9.

32] Mathiesen T, Fridell E, Fredrikson S, Linde A, Sundqvist VA, Edler D, et al.Combination ELISAs for antiviral antibodies in CSF and serum in patientswith neurological symptoms and in healthy controls. J Virol Methods1988;19:169–79.

33] De Tiege X, Rozenberg F, Des Portes V, Lobut JB, Lebon P, Ponsot G, et al. Her-pes simplex encephalitis relapses in children: differentiation of two neurologicentities. Neurology 2003;61:241–3.

34] Casas I, Tenorio A, de Ory F, Lozano A, Echevarria JM. Detection of both herpessimplex and varicella-zoster viruses in cerebrospinal fluid from patients withencephalitis. J Med Virol 1996;50:82–92.

35] Grahn A, Studahl M, Nilsson S, Thomsson E, Backstrom M, Bergstrom T.Varicella-zoster virus (VZV) glycoprotein E is a serological antigen for detec-tion of intrathecal antibodies to VZV in central nervous system infections,without cross-reaction to herpes simplex virus 1. Clin Vaccine Immunol2011;18:1336–42.