West Nile Virus Seroprevalence in Blood Donors from Central Anatolia,Turkey

Abstract

Introduction:

West Nile virus (WNV) is a reemerging flavivirus that has displayed a drastic change in epidemiology in the last decade. Data on WNV activity in Turkey are currently limited. This study investigated WNV exposure in blood donors from Central Anatolia, Turkey.

Materials and Methods:

A total of 2516 sera, collected from blood donors at four major branches of the Turkish Red Crescent Middle Anatolia Regional Blood Center, were evaluated by a commercial WNV immunoglobulin G (IgG) enzyme-linked immunosorbent assay (ELISA). Positive and borderline samples were investigated further by a WNV IgG indirect immunofluorescence test (IIFT), IgG ELISAs for tick-borne encephalitis virus and dengue virus, an IgG IIFT for yellow fever virus, and a multi-Flavivirus biochip IgG IIFT. WNV antibody specificity and titer values were determined by plaque reduction neutralization assay. IgG avidity and IgM were determined for confirmed samples. IgM-positive samples were also evaluated by a real-time reverse transcription polymerase chain reaction assay.

Results:

Twenty-five samples (25/2516; 0.99%) were found reactive in the WNV ELISA/IIFT assays, and 14 could be confirmed by the plaque reduction neutralization assay (14/2516; 0.56%). All IgGs were of high avidity, and four samples (4/14; 28.6%), which were negative for viral RNA, were IgM positive. Although samples with neutralizing WNV IgGs had strong fluorescence intensity in IIFTs, no correlation between antibody titer values and IIFT intensity or quantitative ELISA results could be found. Three WNV nonreactive samples were positive in the dengue IgG ELISA test; one of these also displayed positive results for dengue virus in the mosaic biochip IIFT and reactivity in yellow fever virus IIFT.

Discussion:

WNV exposure is confirmed in 0.56% of the tested healthy blood donors in Central Anatolia, with evidence for dengue/yellow fever and/or other flaviviral infections. This study is the first to document WNV exposure in individuals from Konya, Yozgat, and Sivas provinces in Central Anatolia, and it also establishes viral activity in Ankara, the capital of Turkey.

Introduction

W est Nile virus (WNV) is an enveloped positive-sense RNA virus classified in the Japanese encephalitis serocomplex of Flavivirus genus in the family Flaviviridae (Monath 1990). The natural WNV replication cycle involves culicine mosquitoes and various bird species. Humans, horses, and other mammalian species are “dead-end” hosts, characterized by WNV infections with potential clinical symptoms but transient and low virus levels insufficient to establish a mosquito–mammalian replication cycle (Hayes et al. 2005a). The majority of human WNV exposures follow an asymptomatic course with approximately 20% of infected individuals developing a febrile illness of sudden onset called the West Nile fever, often associated with a long recovery period, and less than 1% developing encephalitis or meningitis (Petersen and Marfin 2002, Hayes et al. 2005b).

Geographical distribution of WNV before 1999 was limited to the Old World, namely, Africa, the Middle East, India, and western and central Europe. WNV was first isolated in the United States in 1999 and has since spread rapidly in the continent, becoming a significant public health problem (Gubler 1996, Hayes et al. 2005a). The North American experience with the virus has identified the elderly and/or immunocompromised patients, including solid organ transplant recipients, as having increased risk for more severe manifestations of the illness. Significant morbidity/mortality rates are also found in wild and domestic animals (Ravindra et al. 2004, Briese and Bernard 2005). The incidence of WNV in Europe has been comparatively poorly studied, and the risk for a similar epidemic, although low, cannot be precisely estimated (Kallio-Kokko et al. 2005). Data on WNV activity in Turkey, an endemic region for WNV, are also limited (Serter 1980). WNV seroreactivity in Anatolia was first reported by Ari (1972), followed by Meco (1977), based on the detection of group-specific antibodies by hemagglutination inhibition (HI) assay. The HI antibodies were prevalent in 6% of humans and 1.5% of sheep in western Anatolia. High reactivities were found in 937 human sera randomly collected from southeast Anatolia: 40.5% in Diyarbakir, 47.8% in Mardin, 44.8% in Siirt, 38% in Sanliurfa, and 41.2% in Elazig (Ari 1972, Meco 1977). These results probably have been affected by the well-known cross-reactions among flaviviral antigens in serological tests (Allwinn et al. 2002). Serter (1980) reported a 29.1% prevalence of HI antibodies in sera collected from the Aegean region of western Anatolia. Seventy-four percent of the positive samples in this case were reactive in neutralization assay, providing the first confirmed seroprevalence data. Ozkul et al. (2005) identified WNV exposure in a wide range of mammals and a human seroprevalence of 20.4% from 88 sera. In another survey, WNV neutralizing antibodies were detected in 9.4% of 181 subjects from Sanliurfa and Siverek in southeast Anatolia (Ergunay et al. 2007). All these confirmed data for WNV activity originate from different regions with diverse ecological properties and/or from limited number of subjects. Thus, we aimed to study WNV exposure in healthy blood donors from Central Anatolia, Turkey.

Materials and Methods

Study population

A total of 2516 sera, collected after informed consent from volunteer blood donors at four major branches (Ankara, Konya, Eskişehir, and Zonguldak provinces) of the Turkish Red Crescent Middle Anatolia Regional Blood Center between January and April 2009, were included in this study. The study was approved by the Hacettepe University Medical Ethics Committee and was conducted according to the offical guidelines for blood donations of the Turkish Red Crescent Society as approved by the Turkish Ministry of Health. The distribution of sera according to the sampling locations, determined based on the population density of each area, was as follows: Ankara (39°56′N–32°52′E), 1395; Konya (37°52′N–32°31′E), 680; Eskişehir (39°46′N–30°32′E), 240; and Zonguldak (41°27′N–31°49′E), 201 (Table 1). Ankara is the capital and the second most densely populated city of Turkey. Mean age of the study group, consisting of 12.9% women, was 38.1 years (range: 19–72; standard deviation: 14.91). All participants filled out a survey questionnaire to reveal the presence of risk factors for vector-borne infections. All sera were transported on dry ice and stored in aliquots at −20°C and −80°C.

Table 1.

Distribution of Samples According to the Region of Blood Donation, and Positive West Nile Virus Immunoglobulin G Assay Results in the Study

Sampling location	Number of samples	ELISA and IIFT a positives (IgG)	PRNA confirmed
Ankara (39°56′N–32°52′E)	1395	20	12
Konya (37°52′N–32°31′E)	680	4	2
Eskişehir (39°46′N–30°32′E)	240	1	—
Zonguldak (41°27′N–31°49′E)	201	—	—
Total	2516 (100%)	25 (0.99%)	14 (0.56%)

These assays were performed only for ELISA positive/borderline samples and include results from single and mosaic biochip assays.

ELISA, enzyme-linked immunosorbent assay; IIFT, indirect immunofluorescence test; IgG, immunoglobulin G; PRNA, plaque reduction neutralization assay.

Primary surveillence testing

All samples were analyzed quantitatively for WNV immunoglobulin G (IgG) by a commercial ELISA test (Anti-WNV IgG ELISA; Euroimmun, Lübeck, Germany) according to the manufacturer's instructions. Photometric measurement of color intensity was performed at a wavelength of 450 nm and a reference wavelength of 620–650 nm. Test results were expressed in relative units per milliliter (RU/mL), calculated according to the standard curve formed by including three calibrator sera with known antibody titers at each run. Samples with antibody levels >20 RU/mL were considered positive and those with 15–20 RU/mL were considered borderline. Specificity and sensitivity of the assay were 96.9% and 99.5%, respectively, as indicated by the manufacturer.

Secondary and confirmatory testing

WNV and mosaic indirect immunofluorescence tests

Samples with positive/borderline WNV IgG ELISA results were evaluated by a commercial indirect immunofluorescence test (IIFT) (Anti-WNV IgG IIFT; Euroimmun) and a mosaic Flavivirus biochip IIFT (Flavivirus Profile 2 IgG IIFT; Euroimmun) according to the manufacturer's instructions. The mosaic biochip IIFT detects IgG antibodies against tick-borne encephalitis virus (TBEV), WNV, Japanese encephalitis virus, yellow fever virus (YFV), and dengue viruses (types 1–4). The results were interpreted by fluorescence microscopy. Antibody levels were qualitatively determined at serum dilutions of 1:10 and 1:100 (for positive samples). The positive results were evaluated as weak (+), moderate (++), or strong (≥+++) according to visual intensity of fluorescence compared to control sera.

TBEV/dengue ELISA and YFV IIFT tests

Samples with positive/borderline WNV IgG ELISA results were also tested using commercial TBEV and dengue IgG ELISA assays (Anti-TBE and Anti-Dengue Virus ELISA; Euroimmun) and for YFV IgG by a commercial IIFT (Anti-YFV IIFT; Euroimmun) to identify cross-reactions and/or coinfections. All assays were performed and evaluated according to the manufacturer's instructions.

WNV plaque reduction neutralization assay, IgM detection, IgG avidity, and real-time reverse transcription polymerase chain reaction assay

Sera positive for WNV IgG by ELISA and IIFT assays were tested for neutralizing antibodies by plaque reduction neutralization assay (PRNA) to confirm specificity. WNV strain NY99-4132 and Vero cells were employed as previously described (Bunning et al. 2002). Briefly, diluted serum (1:5) was inactivated at 56°C for 30 min and mixed with an equal volume (0.1 mL) of virus in Dulbecco's modified Eagle's medium containing 5% fetal calf serum to produce an estimated 100 plaque-forming units of virus per 0.2 mL. Virus–antibody mixtures were then inoculated as 0.2 mL volumes onto Vero cell monolayers in six-well plates and overlaid with 3.2% carboxymethyl cellulose in 2 × Dulbecco's modified Eagle's medium. Plaques were scored on approximately day 4 following incubation at 37°C. Sera that neutralized >70% of the challenge virus were considered reactive, and further dilutions (1:10 to 1:80) were tested. All experiments were performed in duplicate.

PRNA-confirmed samples were further evaluated for IgM and IgG avidity by commercial tests (Anti-WNV avidity ELISA, WNV IgM IIFT; Euroimmun) according to the manufacturer's instructions. All IgM-positive samples were also evaluated for the presence of viral RNA. For this purpose, a set of primers and a probe were selected from the WNV FJ483549 sequence for the in-house real-time reverse transcription polymerase chain reaction (RT-PCR) assay: forward primer, WNV/F/9514 (5′-CTAAACACTTTCACCAACC-3′); reverse primer, WNV/R/9640 (5′-TCTCTCTTCCCCTTCTC-3′); and TaqMan probe, WNV/Pr/9549 (FAM 5′-CCTTCCCCTTCCATCATCCTC-3′ BHQ1). High Pure Viral Nucleic Acid Kit (Roche Diagnostics, Mannheim, Germany) was used for nucleic acid purification. The assay was carried out using QuantiTect Probe RT-PCR Kit (Qiagen, Hilden, Germany) with 5 μL of RNA, 0.6 μM of each primer, and 0.2 μM of the TaqMan probe. Rotor-Gene 6000 instrument (Corbett Research, Sydney, Australia) was used with the following steps: reverse transcription at 42°C for 30 min; initial denaturation at 95°C for 8 min, followed by 45 cycles of denaturation at 95°C for 15 s; and annealing at 60°C for 60 s. Dilutions of RNA extracted from culture-grown virus were employed as positive controls.

Results

WNV ELISA/IIFT assays, PRNA, and avidity

A total of 20 samples (20/2516; 0.79%) were positive and 5 (5/2516; 0.19%) were borderline in the WNV ELISA assay. Twenty-four (24/25; 96%) of these sera were also reactive for WNV in single and/or mosaic IIFT assays. One ELISA borderline sample (16.025 RU/mL) was nonreactive in both IIFT assays and was excluded from PRNA. All samples originated from subjects with no TBEV/YFV vaccination history.

PRNA idenitified 14 samples (14/24; 58.3%) with neutralizing antibodies for WNV. Thus, the confirmed IgG seroprevalence was 14/2516 (0.56%) for all samples (Table 1). All IgG antibodies were of high avidity, and IgM was identified in four (4/14, 28.6%) persons with specific IgGs (Table 2). All IgM-reactive samples were negative for viral RNA.

Table 2.

Demographic Characteristics and West Nile Virus Assay Results for Individuals with West Nile Virus Neutralizing Immunoglobulin G Antibodies

No.	Age/sex/place of residence	ELISA (IgG) (RU/mL)	IIFT ^a (IgG)	IIFT (IgM)	PRNA titer	IgG avidity
1	56/male/Sivas	17.05	+++	−	1:5	High
2	41/male/Ankara	152.86	++++	−	1:80	High
3	25/Female/Ankara	15.72	+++	−	>1:80	High
4	54/male/Konya	>200	++++	+	>1:80	High
5	30/male/Konya	141.17	+++++	−	1:80	High
6	60/male/Ankara	68.36	++++	−	1:80	High
7	48/male/Ankara	111.42	+++++	−	>1:80	High
8	31/male/Yozgat	129.26	++++	−	1:40	High
9	59/male/Ankara	186.88	+++++	−	>1:80	High
10	21/male/Ankara	>200	++++	+	>1:80	High
11	61/male/Ankara	102.73	+++	−	1:80	High
12	57/male/Ankara	181.46	++++	+	>1:80	High
13	35/male/Ankara	155.91	+++	−	>1:80	High
14	25/male/Ankara	159.78	++++	+	>1:80	High

This includes results from single and mosaic biochip assays.

For persons with confirmed WNV exposure, mean age was 43.07 years (range: 21–61; standard deviation: 14.71) with a male:female ratio of 13:1. No history of apparent risk factors or probable virus activity (dead bird sighting, etc.) in the proximity was mentioned by these persons in the survey. Of the two persons with WNV-specific antibodies who donated blood in Ankara one resided in Sivas (39°45′N–37°02′E) and the other in Yozgat (39°50′N–34°48′E), two other provinces in Central Anatolia. Demographic features, detailed assay results, and WNV antibody titers in positive sera are given in Table 2.

Evaluation of confirmed antibody positivity and IIFT results showed that all samples with specific WNV IgG had strong intensity at 1:10 dilution and moderate/strong intensity at 1:100 dilution in both IIFT assays. Although cross-reactions with other flaviviruses were observed on slides for some samples in the mosaic biochip IIFT, fluorescence intensity was clearly stronger for the WNV IgG specific samples at 1:10 and 1:100 dilutions. No correlation was observed between WNV neutralizing antibody titer and IIFT intensity or quantitative ELISA results (Table 2).

Evaluation of WNV nonreactive samples

Seven positive and three borderline samples in WNV ELISA could not be confirmed by PRNA (total: 10/24; 41.7%). All of these sera displayed low-to-moderate positivity for WNV in IIFT assays regardless of the quantitative ELISA results. Three samples (originating from Eskişehir, Konya, and Ankara) were positive in dengue IgG ELISA, and one of these was also positive for dengue virus in the mosaic biochip IIFT. This sample was positive in YFV IgG IIFT and for YFV, TBEV, and Japanese encephalitis virus in the mosaic biochip at 1:10 and 1:100 dilutions. All samples were negative in TBEV IgG ELISA.

Discussion

WNV is a reemerging flavivirus that has displayed a drastic change in epidemiology in the last decade and has since become a public health problem in a number of countries (Hayes et al. 2005a). Information about the epidemiology of WNV infections in Turkey is scarce. Although vector activity is present in certain regions of Turkey for most of the medically important flaviviruses including WNV, no established vector surveillance programs exist and risk areas for WNV infections cannot be predicted (Alten et al. 2000). Because there are no documented cases of WNV from Turkey, contribution of WNV-induced syndromes to fever of unknown origin and/or central nervous system infections are not known. Previously published WNV seroprevalence rates were based on assays with low specificity in differentiating WNV from other flaviviral exposures, or performed in regions/provinces with different climatic/ecological conditions, or performed with a limited number of subjects (Ari 1972, Meco 1977, Serter 1980, Ozkul et al. 2005, Ergunay et al. 2007). Therefore, WNV activity in Anatolia is still not established, and confirmed seroprevalence data is lacking. This study was designed to investigate WNV seroprevalence in persons who donated blood at four branches of the Turkish Red Crescent Middle Anatolia Regional Blood Center (in the Ankara, Konya, Eskişehir, and Zonguldak provinces) in Ankara. A commercial WNV IgG ELISA was employed for primary surveillance, and all borderline/positive results were further evaluated by IIFT assays. Reactive sera in both assays were confirmed for neutralizing antibodies by PRNA, and presence of IgM/viral RNA along with IgG avidity was further determined. To our knowledge, this is the largest study investigating human WNV exposure in Turkey.

WNV neutralizing antibodies were identified in 14 (0.56%) of 2516 blood donors. Ten of the exposed individuals were residents of Ankara while the rest lived in Konya, Yozgat, and Sivas provinces in Central Anatolia (Table 2). In previous studies on WNV seroprevalence, Serter (1980) reported a 21.5% confirmed IgG rate in 1074 sera from Aegean, a region that forms the western coastal border of Turkey. Ozkul et al. (2005) demonstrated neutralizing IgG antibodies in 14 of 48 (29.1%) sera from Ankara and in 4 of 40 sera (10%) from Antalya, which lies in the Mediterranean region of Turkey. We have previously identified and confirmed WNV exposure in 9.4% of 181 individuals from the Sanliurfa and Siverek provinces in southeastern Anatolia (Ergunay et al. 2007). The basis of these apparent variations and their impact on seroprevalence rates need to be identified. Because results from various surveillance tests were confirmed by PRNA in all of these studies, differences in seroprevalence rates possibly indicate actual variances in human WNV exposure in different regions. Implementation of vector surveillance programs and identification of high-risk areas will enable better understanding of the epidemiology of WNV and/or other vector-borne viral infections in Turkey.

Although we have demonstrated in this study that sera with specific WNV IgG have strong fluorescence intensity in IIFT assays, no correlation between neutralizing antibody titers and quantitative ELISA or IIFT results could be identified (see Results). This fact was also reported by Niedrig et al. (2007), who suggested that ELISA and IIFT identified antibodies to structural proteins whereas PRNA detected neutralizing antibodies to the envelope protein of the virus. Although commercial ELISA and IIFT assays are based on cell culture–grown virus antigens, which presumably contain the E-protein, their relatively low correlation with PRNA indicates that ELISA and IIFT are detecting predominantly antibodies to the capsid protein, the most abundant antigen found in cells infected with WNV (Niedrig et al. 2007).

All anti-WNV IgG antibodies detected in the current study were of high avidity, and 28.6% of the individuals with WNV IgG antibodies were simultaneously positive for IgM (Table 2). All samples with WNV IgM were negative for viral RNA in the real-time RT-PCR assay. WNV-specific IgM has been shown to persist for up to 500 days in some patients (Roehrig et al. 2003, Prince et al. 2007). To overcome the problem of differentiating current WNV infections from persistent seropositivity derived from previous exposures, IgG avidity determination has been suggested. Sera tested 6 months or more after WNV exposure contain high-avidity IgG antibodies (Levett et al. 2005). Further, recent and past exposure to the virus can be differentiated using avidity assays, where a high avidity index suggests infection of the order of 40 days or more in a patient with a primary WNV infection or an anamnestic response resulting from a previous infection (Fox et al. 2006). Thus, the presence of IgM with high-avidity IgG antibodies and without viral RNA in the donors in the current study indicate residual/persistent antibody response or previous WNV exposure, which presumably occurred during the preceding mosquito seasons.

Interestingly, three persons whose reactivity in the WNV ELISA/IIFT assays could not be confirmed by PRNA yielded positive dengue IgG results in ELISA and one of them also had detectable YFV IgG in single IIFT and a complex pattern of reactivity in mosaic biochip IIFT (see Results). All displayed negative results in TBEV ELISA. Because no neutralization assays were performed for these viruses, antibody cross-reactions with other flaviviruses cannot be ruled out entirely. Only limited information is available on dengue and yellow fever epidemiology in Turkey. Human exposure to dengue serotypes 1–3 was previously identified and confirmed in 53%, 2.1%, and 0.9% of sera, respectively, in one study, which also identified YFV antibodies in 9.7% of the samples though none could be confirmed by neutralization (Serter 1980). TBEV vaccination is not performed in Turkey, and all individuals who participated in our study reported not to have received YFV vaccine. Hence detection of these antibodies in the current study may indicate exposure to dengue virus/YFV or an antigenically similar flavivirus. Niedrig et al. (2007) found that sera positive for YFV, dengue virus, and TBEV have individual cross-reaction patterns that could differ significantly from one serum to the other. This can be more pronounced in a multiantigen biochip assay as used in this study. More data are required to fully assess YFV and dengue virus activity in various regions of Turkey.

To conclude, WNV exposure is confirmed in 0.56% of healthy blood donors in Central Anatolia, with evidence for dengue/yellow fever and/or other flaviviral infections. This study documents for the first time WNV exposure in individuals from Konya, Yozgat, and Sivas provinces, and establishes viral activity in Ankara. The clinical impact of WNV in Turkey still needs to be determined, and WNV should be considered as a possible etiologic agent in cases of fever of unknown origin and/or viral meningoencephalitis.

Footnotes

Acknowledgments

This project was supported by the Turkish Red Crescent Society and the Hacettepe University Research Fund (project no. 08 O 101 001).

Disclosure Statement

All authors announce that no competing financial interests exist.

References

Allwinn

, Doerr

, Emmerich

, Schmitz

. et al.

Cross-reactivity in flavivirus serology: new implications of an old finding?

Med Microbiol Immunol, 2002; 190:199–202.

Alten

, Bellini

, Caglar

, Simsek

. et al. Species composition and seasonal dynamics of mosquitoes in the Belek region of Turkey. J Vector Ecol, 2000; 25:146–154.

Ari

Studies on activity and ecology of arboviruses in Turkey. Turk Hij Tecr Biyol Derg, 1972; 32:134–143.

Briese

, Bernard

. West Nile virus—an old virus learning new tricks? J Neurovirol, 2005; 11:469–475.

Bunning

, Bowen

, Cropp

, Sullivan

. et al. Experimental infection of horses with West Nile virus. Emerg Infect Dis, 2002; 8:380–386.

Ergunay

, Ozer

, Us

, Ozkul

. et al. Seroprevalence of West Nile Virus and tick-borne encephalitis virus in southeastern Turkey: first evidence for tick-borne encephalitis virus infections. Vector Borne Zoonot Dis, 2007; 7:157–161.

Fox

, Hazell

, Tobler

, Busch

. Immunoglobulin G avidity in differentiation between early and late antibody responses to West Nile virus. Clin Vaccine Immunol, 2006; 13:33–36.

Gubler

. The global resurgence of arboviral diseases. Trans R Soc Trop Med Hyg, 1996; 90:449–451.

Hayes

, Komar

, Nasci

, Montgomery

. et al. Epidemiology and transmission dynamics of West Nile Virus disease. Emerg Infect Dis, 2005a. 11:1167–1173.

10.

Hayes

, Sejvar

, Zaki

, Lanciotti

. et al. Virology, pathology and clinical manifestations of West Nile Virus disease. Emerg Infect Dis, 2005b. 11:1174–1179.

11.

Kallio-Kokko

, Uzcategui

, Vapalahti

, Vaheri

. Viral zoonoses in Europe. FEMS Microbiol Rev, 2005; 29:1051–1077.

12.

Levett

, Sonnenberg

, Sidaway

, Shead

. et al. Use of Immunoglobulin G avidity assays for differentiation of primary from previous infections with West Nile Virus. J Clin Microbiol, 2005; 43:5873–5875.

13.

Meco

. West Nile arbovirus antibodies with hemagglutination inhibition (HI) in residents of Southeast Anatolia. Mikrobiyol Bul, 1977; 11:3–17.

14.

Monath

. Flaviviruses. Fields

, Knipe

. Virology, 2. New York: Raven Press, 1990; 763–814.

15.

Niedrig

, Sonnenberg

, Steinhagen

, Paweska

. Comparison of ELISA and immunoassays for measurement of IgG and IgM antibody to West Nile virus in human sera against virus neutralisation. J Virol Methods, 2007; 139:103–105.

16.

Ozkul

, Yildirim

, Pinar

, Akcali

. et al. Serological evidence of West Nile Virus (WNV) in mammalian species in Turkey. Epidemiol Infect, 2005; 29:1–4.

17.

Petersen

, Marfin

. West Nile virus: a primer for clinician. Ann Intern Med, 2002; 137:173–179.

18.

Prince

, Tobler

, Yeh

, Gefter

. et al. Persistence of West Nile Virus-specific antibodies in viremic blood donors. Clin Vaccine Immunol, 2007; 14:1228–1230.

19.

Ravindra

, Freifeld

, Kalil

, Mercer

. et al. West Nile virus associated encephalitis on recipients of renal and pancreas transplants: case series and literature review. Clin Infect Dis, 2004; 38:1257–1260.

20.

Roehrig

, Nash

, Maldin

, Labowitz

. et al. Persistence of virus-reactive serum immunoglobulin M antibody in confirmed West Nile virus encephalitis cases. Emerg Infect Dis, 2003; 9:376–379.

21.

Serter

. Present status of arbovirus sero-epidemiology in the Aegean region of Turkey. Vesenjak-Hirjan

, Caliserh

. Arboviruses in the Mediterranean countries, (Zentralbl Bakteriol Suppl. 9) Stuttgart: Gustav Fischer Verlag, 1980; 155–161.