Journal of Viral Hepatitis, 2015, 22, 957–963
doi:10.1111/jvh.12397
Evaluation of an antigen-capture EIA for the diagnosis of hepatitis E virus infection C. Zhao,1,2 Y. Geng,3 T. J. Harrison,4 W. Huang,1 A. Song1 and Y. Wang1,2
1
College of Life Sciences,
2
Jilin University, Changchun, China; Division of HIV/AIDS and Sex-Transmitted Virus Vaccines, National Institutes for Food and Drug Control, Beijing, China; 3Health Science Center, Hebei University, Baoding, China; and 4Division of Medicine, University College London Medical School, London, UK Received September 2014; accepted for publication January 2015
SUMMARY. An enzyme immunoassay (EIA) has been devel-
oped for hepatitis E virus (HEV) antigen (HEV-Ag) detection and marketed in China. This study aimed to evaluate the sensitivity of the assay and assess the value of HEV-Ag detection in the diagnosis of HEV infection in comparison with HEV RNA detection. Using serial dilutions of a genotype 4 HEV strain, significant correlation was found between the EIA (S/CO) and HEV RNA (IU/mL) concentration in the range 103.5 to 100.5 IU/mL HEV RNA, the Pearson correlation coefficient r approached 0.97. The EIA detection limit was 54.6 IU/mL, compared to 24 IU/mL for HEV RNA using real-time RT-PCR. In clinical samples from hepatitis E patients, the HEV-Ag and HEV RNA positivity rates were 55.6% (65/117) and 60.7% (71/117) in sera and 76.7% (56/73) and 84.9% (62/73) in stools, and the concordance of these two markers was 77.8% in sera and
INTRODUCTION Hepatitis E virus (HEV), the causative agent of hepatitis E, is a nonenveloped, single stranded, positive sense RNA virus. The genome of HEV contains three open reading frames (ORF1-3): ORF1 encodes a nonstructural polyprotein, ORF2 encodes the major viral capsid protein, which contains potential glycosylation, dimerization and host interacting sites and ORF3 encodes a small protein that may be involved in virus replication [1]. Four major genotypes of HEV are known to infect humans. Genotypes 1 and 2 are endemic only in developing countries and can cause outbreaks through contaminated water. Genotypes 3 and 4 HEV are zoonotic and mainly responsible for the Abbreviations: BSA, bovine serum albumin; EIA, enzyme immunoassay; ELISA, enzyme-linked immunosorbent assay; HEV, hepatitis E virus; ORF, open reading frame. Correspondence: Youchun Wang, PhD, MD, Division of HIV/AIDS and Sex-Transmitted Virus Vaccines, National Institutes for Food and Drug Control, No 2 Tiantianxili, Beijing 100050, China. E-mail: [email protected]
© 2015 John Wiley & Sons Ltd
80.8% in stools. In serum samples, the HEV-Ag positivity rate and the concordance between HEV-Ag and HEV RNA were inversely proportional to the presence of anti-HEV antibody. The presence of anti-HEV IgG could reduce the S/CO of the HEV-Ag EIA. These results reveal a significant correlation between the detection of HEV-Ag and HEV RNA. The sensitivity of the HEV-Ag EIA was lower than real-time RT-PCR but could be higher than conventional nested RT-PCR. Therefore, the detection of HEV-Ag in serum and faeces is valuable for the diagnosis and prognosis of HEV infection in developing regions where real-time RT-PCR is not available. Keywords: diagnostic marker, enzyme immunoassay, hepatitis E virus, HEV antigen, HEV ORF2 protein.
sporadic cases in both developing and developed countries [2]. In recent years in China, most of the HEV isolations from hepatitis E patients were genotype 4 HEV, only a few strains were genotype 3 or genotype 1 [3,4]. Hepatitis E manifests generally as a self-limiting acute disease. However, the severity of the disease can be high under certain circumstances. Acute hepatitis E can progress to fulminant hepatitis with encephalopathy and coagulation disorders. In HEV-infected pregnant women, the mortality can be as high as 20%–30% in endemic regions [5]. HEV infections in immunocompromised patients may evolve to chronic hepatitis [6]. Chronic HEV infection also has been reported recently in immunocompetent patients [7]. Four biomarkers, viral RNA, anti-HEV IgM, anti-HEV IgG (at least fourfold rise) and low avidity of anti-HEV IgG, have been found to be valuable for the diagnosis of acute HEV infection [8]. Currently, the clinical diagnosis of acute hepatitis E is determined primarily by the serological detection of anti-HEV antibodies [9]. For the early diagnosis of acute hepatitis E cases in the window period and chronic HEV infection in immunocompromised patients whose HEV
958
C. Zhao et al.
seroconversion may be delayed or in whom antibodies fail to develop, PCR assays are recommended to detect viral RNA [10,11]. Various in-house assays have been described for HEV RNA detection. In general, HEV RNA detection provides a promising sensitivity rate and rapid diagnosis. However, it requires specialized laboratory equipment, experienced technicians and, more importantly, high cost. These notable limitations preclude the use of PCR detection in most hospitals and clinical centres in developing countries where HEV is hyperendemic. Viral antigen detection methods that are convenient and cost-effective have been used successfully for the diagnosis of various infectious diseases and monitoring of disease activity, for example hepatitis B virus (HBsAg), hepatitis C virus (core antigen) and HIV-1 (p24 antigen). HEV ORF2 encodes a single capsid protein of 660 amino acids. During viral infection, the immune system responds predominantly to the capsid protein, which has been shown to be highly immunogenic [8]. An assay has been developed for HEV antigen detection, using a combination of monoclonal antibodies to capture the HEV capsid protein [12]. The aim of this study was to evaluate the sensitivity of HEV ORF2 antigen detection using a commercial EIA manufactured by Wantai, Inc., (Beijing, China) and assess the potential use of ORF2 antigen as a marker for the diagnosis of HEV infection.
vol) suspensions. The suspensions were clarified by centrifugation at 1000 g for 15 min. Approval for this study was obtained from the Health Research Ethics Committee of the Health Science Center, Hebei University. The patients gave written or oral informed consent for the use of their samples.
HEV-Ag EIA The test is a Two-Antibody Sandwich Microplate Enzyme Immunoassay (Wantai Biopharmaceutical Inc., Beijing, China). In this system, goat polyclonal anti-ORF2 antibodies are used for antigen capture and enzyme-linked monoclonal antibodies against the ORF2 protein are used for detection. The procedures for detecting HEV-Ag in the samples were in accordance with the manufacturer’s instructions. One hundred microlitres of each sample was used per reaction.
Anti-HEV IgG and IgM ELISA Anti-HEV IgM and anti-HEV IgG were detected in sera using commercial enzyme-linked immunosorbent assay (ELISA) kits (Beijing Wantai Biological Pharmacy Enterprise Co., Ltd., Beijing, China) according to the manufacturer’s instructions.
MATERIALS AND METHODS RNA extraction and real-time RT-PCR Viruses and clinical samples Cell culture supernatants from a genotype 4 HEV strain (GenBank Accession Number: AJ272108) propagated in PLC/PRF/5 cells [13] were used to evaluate the correlation between the HEV-Ag EIA (S/CO) and HEV RNA concentration (log10 IU/mL) and to determine the minimum detection limits of the two assays. The harvested cell culture supernatant was clarified by centrifugation at 1000 g for 15 min and diluted with PBS containing 2% BSA. The WHO International Standard (WHO/BS/2011.2175; 250 000 IU/Vial) [14] was used to calibrate the HEV RNA concentration in the real-time RT-PCR detection system. The dry powdered WHO Standard was reconstituted at a concentration of 5 9 105 IU/mL in PBS containing 2% BSA. Half log10 fold serial dilutions were prepared and used to generate a standard RNA curve by analysing each dilution in triplicate by real-time RT-PCR. Sera and stools from 117 patients with a clinical diagnosis of acute hepatitis were collected from Baoding Hospital for Infectious Diseases in Hebei province, China from October 2011 to August 2013. In these patients, infections by hepatitis A, B, C viruses and cytomegalovirus were excluded by testing for specific antibodies or antigens and drug- and alcohol-induced hepatitis was excluded by interrogation of the patients. The faecal samples were suspended in PBS containing 2% BSA to generate 10% (wt/
Total RNA was extracted using a QIAamp Viral RNA Mini kit (QIAGEN GmbH, Hilden, Germany) according to the manufacturer’s instructions. A 140 lL volume of each sample was used for RNA extraction, and the RNA was recovered in 50 lL of elution buffer. The real-time RT-PCRs were performed on an ABI 7500 Fast Sequence Detection System (Applied Biosystems, Foster City, CA, USA) using a Diagnostic kit for hepatitis E virus RNA (FQ-PCR) (Beijing Jinhao Biological Pharmacy Enterprise Co., Ltd., Beijing, China), according to the manufacturer’s instructions. In each reaction, 10 lL aliquots of extracted RNA were used. The forward and reverse primers were 50 -CGG TGG TTT CTG GGG TGA-30 and 50 -GCG AAG GGG TTG GTT GGA-30 , respectively, and the probe was 50 -FAM-TGA TTC TCA GCC CTT CGC-TAMRA-30 .
Correlation between the antigen EIA and concentration of HEV RNA The cell culture supernatant of the genotype 4 HEV strain (AJ272108) was assayed quantitatively for HEV RNA concentration using the HEV WHO International Standard for NAT (WHO/BS/2011.2175; 250 000 IU/mL) [14] as a reference. The supernatant was diluted to a concentration of 104 IU/mL and further serially half log10 fold diluted with PBS containing 2% BSA. Real-time RT-PCR and HEV-Ag © 2015 John Wiley & Sons Ltd
HEV-Ag detection for the diagnosis of HEV infection
Fig. 2 Correlation between HEV-Ag (S/ CO) and HEV RNA concentrations (Log10 IU/mL) in clinical samples, sera (a) and faeces (b) from hepatitis E patients.
HEV Ag (S/CO)
25
30
(b) 30
25
25
HEV-Ag (S/CO)
HEV-Ag (S/CO)
(a)
20 15 10
15
0
0.5
1
1.5
8
6 25
12
64
32
16
8
2
2.5
3
3.5
4
HEV RNA (Log10 IU/mL)
0
4
10
0 0
5
2
15
5
10
2 fold serial dilutions of HEV samples
Fig. 3 Antigen EIA reactivity affected by spiking in Ab+Ag sera. The HEV-Ag EIA results (S/CO) are shown in Fig. 2. EIA was conducted after two-fold serial dilutions of the two HEV samples with AbAg or Ab+Ag which were mixed and incubated at 37 °C for 15 min.
coinfection in HIV-positive and other immunocompromised patients because of delayed or failed seroconversion. In this case, HEV RNA detection is also necessary for the diagnosis of chronic hepatitis E. For assay simplicity, cost reduction and applicability in developing countries, the use of ORF2 antigen detection is being evaluated as a surrogate of HEV RNA detection for the diagnosis of HEV infection. The presence of HEV ORF2 antigen in serum was found to be in concordance with HEV RNA [12,17,18]. However, the relatively low sensitivity of ORF2 assays compared to molecular methods has been reported: HEV antigen was not detected in some blood donors, while the corresponding viremias were greater than 103 IU/mL [19]. This disparity prompted questions about the detection limit of the antigen EIA. In this study, standard curves of S/CO of the antigen EIA vs the HEV RNA concentrations were established by detecting serial dilutions of cell culture supernatant and a faecal sample from a patient (Fig. 1). In the standard curve of the genotype 4 HEV, a high correlation between S/CO and HEV RNA concentrations was found in the range of 103.5 to 100.5 IU/mL of HEV RNA, and the Pearson correlation coefficient r was up to 0.97, indicating a good corre© 2015 John Wiley & Sons Ltd
20
5
Sample1:Ab– Sample1:Ab+ Sample2:Ab– Sample2:Ab+
20
961
4.5
5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
HEV RNA (Log10 IU/mL)
lation between the two assays (Fig. 1). The minimum detection limits of the HEV-Ag EIA and real-time RT-PCR were determined to be 54.2 IU/mL and 24 IU/mL, respectively, by measuring the end-point of dilutions. The statistical difference of the detection limits of these two assays indicates that the HEV-Ag EIA is less sensitive than realtime RT-PCR. Real-time RT-PCR methods generally have proven to be more sensitive than conventional RT-PCR assays [20]. Our previous study showed that real-time RTPCR could detect a 10- to 100-fold higher dilution of genotype 1 and 4 isolates than conventional RT-PCR [17]. Thus, it could be inferred that the sensitivity of the antigen EIA is at least not less than that of the conventional RTPCR. The result of the sensitivity of real-time RT-PCR in this study is similar to previous reports, with a detection limit of 5 IU/mL–180 IU/mL in genotype 3 HEV [19]. A high variability of 100- to 1000-fold difference in the sensitivity of RT-PCR is seen between different laboratories [20]. The reproducibility of antigen detection by EIA was better than that of the real-time RT-PCR in this study. Clinical serum samples from acute hepatitis E patients were divided into four groups according to the presence of anti-HEV antibodies. Analysis showed that the concordance of HEV antigen and HEV RNA detection was highest in the group in which anti-HEV antibodies were not present (Table 2). All three HEV RNA-positive cases in this group also tested positive for ORF2 antigen (Table 2). This result indicates that HEV RNA and ORF2 antigen could be detected before the presence of detectable IgM. Antigen detection is useful for the early diagnosis of acute infection. The concordance of these two markers dropped dramatically as the illness proceeded, with the appearance of antibodies. In the IgM+/IgG+ group comprising 76 samples, the concordance rate of these two markers was 69.7%, which was significantly lower than that of the 77.8% in the total 117 samples. In addition, the concordance rate of these two markers was lower than that in the faecal samples (Fig. 2). To confirm the influence of anti-HEV IgG on the detection of HEV ORF2 by EIA, we spiked IgG+ Ag sera into IgG-Ag+ sera and incubated the mixture at 37 °C for 15 min. As expected, the OD value of the antigen reduced significantly compared to the
960
C. Zhao et al.
Table 1 Detection limit comparison between the antigen EIA and real-time RT-PCR using the WHO standard (250 000 IU/mL)
Table 3 HEV-Ag and RNA detection in faeces from acute hepatitis E patients HEV-Ag
Positive tests Concentration (IU/mL) Replicates Real-time PCR ELISA 100 50 25 12.5 6.3 3.2 1.6 0.8
12 12 12 12 12 12 12 12
12 12 11 10 7 3 1 0
12 12 4 2 0 0 0 0
groups: IgM/IgG, IgM+/IgG, IgM+/IgG+ and IgM/ IgG+. The positivity rates of HEV-Ag in the HEV RNA-positive samples were 100% (3/3), 93.8% (15/16) and 71.2% (37/52) in the IgM/IgG, IgM+/IgG and IgM+/IgG+ samples, respectively, showing a gradual decrease in these three groups (Table 2). The concordance rate of these two markers in each group varied and was inversely proportional to the titre of the antibodies. The correlation between HEV-Ag EIA activities (S/CO) and the corresponding HEV RNA concentrations (Log10 IU/mL) was determined. The Pearson correlation coefficient r was 0.55 (P < 0.01), and thus, the correlation was significant (Fig. 2a).
Detection of HEV antigen in faecal samples from acute hepatitis E patients Faecal samples from 73 acute hepatitis patients, tested previously for anti-HEV IgM or HEV RNA in serum, were available for detecting the presence of ORF2 antigen and viral RNA. Among the 73 samples, 76.7% (56/73) samples were HEV-Ag positive and 84.9% (62/73) were HEV RNA positive (Table 3). The concordance rate of HEV RNA detection by real-time RT-PCR and HEV-Ag detection by EIA was 80.8%. The correlation between HEV-Ag EIA (S/CO) and the corresponding HEV RNA concentrations (Log10 IU/mL) in
HEV RNA
Positive
Negative
Total
Positive Negative Total
52 4 56
10 7 17
62 11 73
faecal samples also was significant (Fig. 2b). The Pearson correlation coefficient r was 0.93 (P < 0.01). The concordance of HEV-Ag and HEV RNA in faecal samples seemed to be better than that in serum samples.
The presence of anti-HEV antibody affects the detection of HEV-Ag To determine the effect of anti-HEV antibodies on the detection of HEV-Ag, serial dilutions of the two HEV samples were spiked into anti-HEV IgG-positive or IgG-negative sera in vitro and tested by the EIA. By statistical analysis, the HEV-Ag EIA S/CO values of the serum sample and cell culture supernatant mixed with anti-HEV IgG-positive sera were significantly reduced compared to the control (Fig. 3).
DISCUSSION Hepatitis E is endemic in industrialized and developing countries worldwide. Its prevalence is significantly higher in developing countries, such as India, Mexico and Uganda, where outbreaks associated with faecal contamination of drinking water occur frequently [15]. Diagnosis of HEV infection mostly relies on the serological detection of anti-HEV antibodies. Several diagnostic kits are available for HEV serology in humans and have been validated. Commercial kits for the detection of anti-HEV IgM seem to be specific and reliable for the diagnosis of acute hepatitis E [16], but anti-HEV IgM may not be present at the onset of the illness. This is when HEV RNA should be used as an early diagnostic marker of acute hepatitis E. In addition, serology alone may be insufficient to diagnose HEV
Table 2 HEV-Ag and HEV RNA detections in sera from acute hepatitis E patients Antibody detection
Case
RNA+/Ag+
RNA+/Ag
RNA/Ag+
RNA/Ag
IgM/IgG IgM+/IgG IgM+/IgG+ IgM/IgG+ Total
3 17 76 21 117
3 15 37 0 55
0 1 15 0 16
0 1 8 1 10
0 0 16 20 36
© 2015 John Wiley & Sons Ltd
HEV-Ag detection for the diagnosis of HEV infection
Fig. 2 Correlation between HEV-Ag (S/ CO) and HEV RNA concentrations (Log10 IU/mL) in clinical samples, sera (a) and faeces (b) from hepatitis E patients.
HEV Ag (S/CO)
25
30
(b) 30
25
25
HEV-Ag (S/CO)
HEV-Ag (S/CO)
(a)
20 15 10
15
0
0.5
1
1.5
8
6 25
12
64
32
16
8
2
2.5
3
3.5
4
HEV RNA (Log10 IU/mL)
0
4
10
0 0
5
2
15
5
10
2 fold serial dilutions of HEV samples
Fig. 3 Antigen EIA reactivity affected by spiking in Ab+Ag sera. The HEV-Ag EIA results (S/CO) are shown in Fig. 2. EIA was conducted after two-fold serial dilutions of the two HEV samples with AbAg or Ab+Ag which were mixed and incubated at 37 °C for 15 min.
coinfection in HIV-positive and other immunocompromised patients because of delayed or failed seroconversion. In this case, HEV RNA detection is also necessary for the diagnosis of chronic hepatitis E. For assay simplicity, cost reduction and applicability in developing countries, the use of ORF2 antigen detection is being evaluated as a surrogate of HEV RNA detection for the diagnosis of HEV infection. The presence of HEV ORF2 antigen in serum was found to be in concordance with HEV RNA [12,17,18]. However, the relatively low sensitivity of ORF2 assays compared to molecular methods has been reported: HEV antigen was not detected in some blood donors, while the corresponding viremias were greater than 103 IU/mL [19]. This disparity prompted questions about the detection limit of the antigen EIA. In this study, standard curves of S/CO of the antigen EIA vs the HEV RNA concentrations were established by detecting serial dilutions of cell culture supernatant and a faecal sample from a patient (Fig. 1). In the standard curve of the genotype 4 HEV, a high correlation between S/CO and HEV RNA concentrations was found in the range of 103.5 to 100.5 IU/mL of HEV RNA, and the Pearson correlation coefficient r was up to 0.97, indicating a good corre© 2015 John Wiley & Sons Ltd
20
5
Sample1:Ab– Sample1:Ab+ Sample2:Ab– Sample2:Ab+
20
961
4.5
5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
HEV RNA (Log10 IU/mL)
lation between the two assays (Fig. 1). The minimum detection limits of the HEV-Ag EIA and real-time RT-PCR were determined to be 54.2 IU/mL and 24 IU/mL, respectively, by measuring the end-point of dilutions. The statistical difference of the detection limits of these two assays indicates that the HEV-Ag EIA is less sensitive than realtime RT-PCR. Real-time RT-PCR methods generally have proven to be more sensitive than conventional RT-PCR assays [20]. Our previous study showed that real-time RTPCR could detect a 10- to 100-fold higher dilution of genotype 1 and 4 isolates than conventional RT-PCR [17]. Thus, it could be inferred that the sensitivity of the antigen EIA is at least not less than that of the conventional RTPCR. The result of the sensitivity of real-time RT-PCR in this study is similar to previous reports, with a detection limit of 5 IU/mL–180 IU/mL in genotype 3 HEV [19]. A high variability of 100- to 1000-fold difference in the sensitivity of RT-PCR is seen between different laboratories [20]. The reproducibility of antigen detection by EIA was better than that of the real-time RT-PCR in this study. Clinical serum samples from acute hepatitis E patients were divided into four groups according to the presence of anti-HEV antibodies. Analysis showed that the concordance of HEV antigen and HEV RNA detection was highest in the group in which anti-HEV antibodies were not present (Table 2). All three HEV RNA-positive cases in this group also tested positive for ORF2 antigen (Table 2). This result indicates that HEV RNA and ORF2 antigen could be detected before the presence of detectable IgM. Antigen detection is useful for the early diagnosis of acute infection. The concordance of these two markers dropped dramatically as the illness proceeded, with the appearance of antibodies. In the IgM+/IgG+ group comprising 76 samples, the concordance rate of these two markers was 69.7%, which was significantly lower than that of the 77.8% in the total 117 samples. In addition, the concordance rate of these two markers was lower than that in the faecal samples (Fig. 2). To confirm the influence of anti-HEV IgG on the detection of HEV ORF2 by EIA, we spiked IgG+ Ag sera into IgG-Ag+ sera and incubated the mixture at 37 °C for 15 min. As expected, the OD value of the antigen reduced significantly compared to the
962
C. Zhao et al.
control group (Fig. 3). Thus, it was seen that the presence of antibodies affects the detection of ORF2 antigen. This result may help to explain why HEV antigen was not detected in some blood donors when the corresponding viremias were greater than 103 IU/mL [19]. Therefore, in clinical diagnosis of HEV infection, HEV antigen detection in serum should be combined with antibody detection, followed by less frequent confirmatory testing with molecular assays. HEV is transmitted primarily by the faecal–oral route. Faeces from hepatitis E patients might contain infectious viruses. Of the 73 faecal samples from hepatitis E patients confirmed by detecting anti-IgM or HEV RNA in sera, the ORF2 antigen and HEV RNA positivities were 56 (76.7%) and 62 (84.9%), respectively. That the concordance rate of these two assays (80.8%) was better than those of the serum samples may be attributable to the absence of antiHEV antibodies in the faeces. Animal experiments have shown that HEV RNA was detectable early and disappeared later from faecal samples than from the sera of infected animals [21]. However, because anorexia is a presenting symptom of acute hepatitis E patients, the collection of faecal samples might not be straightforward in the early period of the infection. Therefore, detecting HEV RNA or HEV-Ag in faecal samples for early diagnosis of
acute hepatitis E may be less practical; nevertheless, HEVAg in faeces should be valuable to monitor the prognosis of HEV-related diseases, such as chronic HEV infection, and will provide clinical researchers with an alternative method of low cost and easy manipulation to evaluate the infectious status of HEV-infected patients. In summary, the S/CO of the HEV-Ag EIA shows a good relationship to HEV RNA concentrations. Although the HEV-Ag EIA was less sensitive than real-time RT-PCR, it might be more sensitive than conventional RT-nested PCR. HEV-Ag detection in serum is valuable for the early diagnosis of acute hepatitis E. HEV-Ag detection in stool is likely to be useful for the diagnosis of chronic HEV infection and for monitoring the disease progression in both acute and chronic hepatitis E.
ACKNOWLEDGEMENTS Funding: The study were supported by the National Natural Science Foundation of China (grant: 81171549) and by the Science and Technology Program of Hebei Province (grant: 12277726). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
REFERENCES 1 Ahmad I, Holla RP, Jameel S. Molecular virology of hepatitis E virus. Virus Res 2011; 161: 47–58. 2 Meng XJ. Recent advances in hepatitis E virus. J Viral Hepat 2010; 17: 153–161. 3 Zhang W, He Y, Wang H et al. Hepatitis E virus genotype diversity in eastern China. Emerg Infect Dis 2010; 16: 1630–1632. 4 Geng Y, Zhao C, Fan J et al. Genotype analysis of hepatitis E virus from sporadic hepatitis E cases in northern China. Infect Genet Evol 2013; 20: 413–417. 5 Navaneethan U, Al Mohajer M, Shata MT. Hepatitis E and pregnancy: understanding the pathogenesis. Liver Int 2008; 28: 1190– 1199. 6 Kamar N, Izopet J, Rostaing L. Hepatitis E virus infection. Curr Opin Gastroenterol 2013; 29: 271–278. 7 Grewal P, Kamili S, Motamed D. Chronic hepatitis E in an immunocompetent patient: a case report. Hepatology 2014; 59: 347–348. 8 Zhao Q, Zhang J, Wu T et al. Antigenic determinants of hepatitis E
9
10
11
12
13
virus and vaccine-induced immunogenicity and efficacy. J Gastroenterol 2013; 48: 159–168. Dreier J, Juhl D. Autochthonous hepatitis E virus infections: a new transfusion-associated risk? Transfus Med Hemother 2014; 41: 29–39. Kamar N, Selves J, Mansuy JM et al. Hepatitis E virus and chronic hepatitis in organ-transplant recipients. N Engl J Med 2008; 358: 811–817. Haagsma EB, van den Berg AP, Porte RJ et al. Chronic hepatitis E virus infection in liver transplant recipients. Liver Transpl 2008; 14: 547–553. Zhang F, Li X, Li Z et al. Detection of HEV antigen as a novel marker for the diagnosis of hepatitis E. J Med Virol 2006; 78: 1441–1448. Zhang F, Qi Y, Harrison TJ et al. Hepatitis E genotype 4 virus from feces of monkeys infected experimentally can be cultured in PLC/ PRF/5 cells and upregulate host interferon-inducible genes. J Med Virol 2014; 86: 1736–1744.
14 Baylis SA, Bl€ umel J, Mizusawa S et al. World Health Organization International Standard to harmonize assays for detection of hepatitis E virus RNA. Emerg Infect Dis 2013; 19: 729–735. 15 Aggarwal R, Jameel S. Hepatitis E. Hepatology 2011; 54: 2218–2226. 16 Legrand-Abravanel F, Thevenet I, Mansuy JM et al. Good performance of immunoglobulin M assays in diagnosing genotype 3 hepatitis E virus infections. Clin Vaccine Immunol 2009; 16: 772–774. 17 Zhao C, Ma Z, Harrison TJ et al. A novel genotype of hepatitis E virus prevalent among farmed rabbits in China. J Med Virol 2009; 81: 1371–1379. 18 Majumdar M, Singh MP, Pujhari SK, Bhatia D, Chawla Y, Ratho RK. Hepatitis E virus antigen detection as an early diagnostic marker: report from India. J Med Virol 2013; 85: 823–827. 19 Vollmer T, Knabbe C, Dreier J. Comparison of real-time PCR and antigen assays for detection of hepatitis E virus in blood donors. J Clin Microbiol 2014; 52: 2150–2156.
© 2015 John Wiley & Sons Ltd
HEV-Ag detection for the diagnosis of HEV infection 20 Baylis SA, Hanschmann KM, Bl€ umel J, N€ ubling CM; HEV Collaborative Study Group. Standardization of hepatitis E virus (HEV) nucleic acid amplification tech-
© 2015 John Wiley & Sons Ltd
nique-based assays: an initial study to evaluate a panel of HEV strains and investigate laboratory performance. J Clin Microbiol 2011; 49: 1234–1239.
963
21 Liu P, Bu QN, Wang L et al. Transmission of hepatitis E virus from rabbits to cynomolgus macaques. Emerg Infect Dis 2013; 19: 559–565.
doi:10.1111/jvh.12397
Evaluation of an antigen-capture EIA for the diagnosis of hepatitis E virus infection C. Zhao,1,2 Y. Geng,3 T. J. Harrison,4 W. Huang,1 A. Song1 and Y. Wang1,2
1
College of Life Sciences,
2
Jilin University, Changchun, China; Division of HIV/AIDS and Sex-Transmitted Virus Vaccines, National Institutes for Food and Drug Control, Beijing, China; 3Health Science Center, Hebei University, Baoding, China; and 4Division of Medicine, University College London Medical School, London, UK Received September 2014; accepted for publication January 2015
SUMMARY. An enzyme immunoassay (EIA) has been devel-
oped for hepatitis E virus (HEV) antigen (HEV-Ag) detection and marketed in China. This study aimed to evaluate the sensitivity of the assay and assess the value of HEV-Ag detection in the diagnosis of HEV infection in comparison with HEV RNA detection. Using serial dilutions of a genotype 4 HEV strain, significant correlation was found between the EIA (S/CO) and HEV RNA (IU/mL) concentration in the range 103.5 to 100.5 IU/mL HEV RNA, the Pearson correlation coefficient r approached 0.97. The EIA detection limit was 54.6 IU/mL, compared to 24 IU/mL for HEV RNA using real-time RT-PCR. In clinical samples from hepatitis E patients, the HEV-Ag and HEV RNA positivity rates were 55.6% (65/117) and 60.7% (71/117) in sera and 76.7% (56/73) and 84.9% (62/73) in stools, and the concordance of these two markers was 77.8% in sera and
INTRODUCTION Hepatitis E virus (HEV), the causative agent of hepatitis E, is a nonenveloped, single stranded, positive sense RNA virus. The genome of HEV contains three open reading frames (ORF1-3): ORF1 encodes a nonstructural polyprotein, ORF2 encodes the major viral capsid protein, which contains potential glycosylation, dimerization and host interacting sites and ORF3 encodes a small protein that may be involved in virus replication [1]. Four major genotypes of HEV are known to infect humans. Genotypes 1 and 2 are endemic only in developing countries and can cause outbreaks through contaminated water. Genotypes 3 and 4 HEV are zoonotic and mainly responsible for the Abbreviations: BSA, bovine serum albumin; EIA, enzyme immunoassay; ELISA, enzyme-linked immunosorbent assay; HEV, hepatitis E virus; ORF, open reading frame. Correspondence: Youchun Wang, PhD, MD, Division of HIV/AIDS and Sex-Transmitted Virus Vaccines, National Institutes for Food and Drug Control, No 2 Tiantianxili, Beijing 100050, China. E-mail: [email protected]
© 2015 John Wiley & Sons Ltd
80.8% in stools. In serum samples, the HEV-Ag positivity rate and the concordance between HEV-Ag and HEV RNA were inversely proportional to the presence of anti-HEV antibody. The presence of anti-HEV IgG could reduce the S/CO of the HEV-Ag EIA. These results reveal a significant correlation between the detection of HEV-Ag and HEV RNA. The sensitivity of the HEV-Ag EIA was lower than real-time RT-PCR but could be higher than conventional nested RT-PCR. Therefore, the detection of HEV-Ag in serum and faeces is valuable for the diagnosis and prognosis of HEV infection in developing regions where real-time RT-PCR is not available. Keywords: diagnostic marker, enzyme immunoassay, hepatitis E virus, HEV antigen, HEV ORF2 protein.
sporadic cases in both developing and developed countries [2]. In recent years in China, most of the HEV isolations from hepatitis E patients were genotype 4 HEV, only a few strains were genotype 3 or genotype 1 [3,4]. Hepatitis E manifests generally as a self-limiting acute disease. However, the severity of the disease can be high under certain circumstances. Acute hepatitis E can progress to fulminant hepatitis with encephalopathy and coagulation disorders. In HEV-infected pregnant women, the mortality can be as high as 20%–30% in endemic regions [5]. HEV infections in immunocompromised patients may evolve to chronic hepatitis [6]. Chronic HEV infection also has been reported recently in immunocompetent patients [7]. Four biomarkers, viral RNA, anti-HEV IgM, anti-HEV IgG (at least fourfold rise) and low avidity of anti-HEV IgG, have been found to be valuable for the diagnosis of acute HEV infection [8]. Currently, the clinical diagnosis of acute hepatitis E is determined primarily by the serological detection of anti-HEV antibodies [9]. For the early diagnosis of acute hepatitis E cases in the window period and chronic HEV infection in immunocompromised patients whose HEV
958
C. Zhao et al.
seroconversion may be delayed or in whom antibodies fail to develop, PCR assays are recommended to detect viral RNA [10,11]. Various in-house assays have been described for HEV RNA detection. In general, HEV RNA detection provides a promising sensitivity rate and rapid diagnosis. However, it requires specialized laboratory equipment, experienced technicians and, more importantly, high cost. These notable limitations preclude the use of PCR detection in most hospitals and clinical centres in developing countries where HEV is hyperendemic. Viral antigen detection methods that are convenient and cost-effective have been used successfully for the diagnosis of various infectious diseases and monitoring of disease activity, for example hepatitis B virus (HBsAg), hepatitis C virus (core antigen) and HIV-1 (p24 antigen). HEV ORF2 encodes a single capsid protein of 660 amino acids. During viral infection, the immune system responds predominantly to the capsid protein, which has been shown to be highly immunogenic [8]. An assay has been developed for HEV antigen detection, using a combination of monoclonal antibodies to capture the HEV capsid protein [12]. The aim of this study was to evaluate the sensitivity of HEV ORF2 antigen detection using a commercial EIA manufactured by Wantai, Inc., (Beijing, China) and assess the potential use of ORF2 antigen as a marker for the diagnosis of HEV infection.
vol) suspensions. The suspensions were clarified by centrifugation at 1000 g for 15 min. Approval for this study was obtained from the Health Research Ethics Committee of the Health Science Center, Hebei University. The patients gave written or oral informed consent for the use of their samples.
HEV-Ag EIA The test is a Two-Antibody Sandwich Microplate Enzyme Immunoassay (Wantai Biopharmaceutical Inc., Beijing, China). In this system, goat polyclonal anti-ORF2 antibodies are used for antigen capture and enzyme-linked monoclonal antibodies against the ORF2 protein are used for detection. The procedures for detecting HEV-Ag in the samples were in accordance with the manufacturer’s instructions. One hundred microlitres of each sample was used per reaction.
Anti-HEV IgG and IgM ELISA Anti-HEV IgM and anti-HEV IgG were detected in sera using commercial enzyme-linked immunosorbent assay (ELISA) kits (Beijing Wantai Biological Pharmacy Enterprise Co., Ltd., Beijing, China) according to the manufacturer’s instructions.
MATERIALS AND METHODS RNA extraction and real-time RT-PCR Viruses and clinical samples Cell culture supernatants from a genotype 4 HEV strain (GenBank Accession Number: AJ272108) propagated in PLC/PRF/5 cells [13] were used to evaluate the correlation between the HEV-Ag EIA (S/CO) and HEV RNA concentration (log10 IU/mL) and to determine the minimum detection limits of the two assays. The harvested cell culture supernatant was clarified by centrifugation at 1000 g for 15 min and diluted with PBS containing 2% BSA. The WHO International Standard (WHO/BS/2011.2175; 250 000 IU/Vial) [14] was used to calibrate the HEV RNA concentration in the real-time RT-PCR detection system. The dry powdered WHO Standard was reconstituted at a concentration of 5 9 105 IU/mL in PBS containing 2% BSA. Half log10 fold serial dilutions were prepared and used to generate a standard RNA curve by analysing each dilution in triplicate by real-time RT-PCR. Sera and stools from 117 patients with a clinical diagnosis of acute hepatitis were collected from Baoding Hospital for Infectious Diseases in Hebei province, China from October 2011 to August 2013. In these patients, infections by hepatitis A, B, C viruses and cytomegalovirus were excluded by testing for specific antibodies or antigens and drug- and alcohol-induced hepatitis was excluded by interrogation of the patients. The faecal samples were suspended in PBS containing 2% BSA to generate 10% (wt/
Total RNA was extracted using a QIAamp Viral RNA Mini kit (QIAGEN GmbH, Hilden, Germany) according to the manufacturer’s instructions. A 140 lL volume of each sample was used for RNA extraction, and the RNA was recovered in 50 lL of elution buffer. The real-time RT-PCRs were performed on an ABI 7500 Fast Sequence Detection System (Applied Biosystems, Foster City, CA, USA) using a Diagnostic kit for hepatitis E virus RNA (FQ-PCR) (Beijing Jinhao Biological Pharmacy Enterprise Co., Ltd., Beijing, China), according to the manufacturer’s instructions. In each reaction, 10 lL aliquots of extracted RNA were used. The forward and reverse primers were 50 -CGG TGG TTT CTG GGG TGA-30 and 50 -GCG AAG GGG TTG GTT GGA-30 , respectively, and the probe was 50 -FAM-TGA TTC TCA GCC CTT CGC-TAMRA-30 .
Correlation between the antigen EIA and concentration of HEV RNA The cell culture supernatant of the genotype 4 HEV strain (AJ272108) was assayed quantitatively for HEV RNA concentration using the HEV WHO International Standard for NAT (WHO/BS/2011.2175; 250 000 IU/mL) [14] as a reference. The supernatant was diluted to a concentration of 104 IU/mL and further serially half log10 fold diluted with PBS containing 2% BSA. Real-time RT-PCR and HEV-Ag © 2015 John Wiley & Sons Ltd
HEV-Ag detection for the diagnosis of HEV infection
Fig. 2 Correlation between HEV-Ag (S/ CO) and HEV RNA concentrations (Log10 IU/mL) in clinical samples, sera (a) and faeces (b) from hepatitis E patients.
HEV Ag (S/CO)
25
30
(b) 30
25
25
HEV-Ag (S/CO)
HEV-Ag (S/CO)
(a)
20 15 10
15
0
0.5
1
1.5
8
6 25
12
64
32
16
8
2
2.5
3
3.5
4
HEV RNA (Log10 IU/mL)
0
4
10
0 0
5
2
15
5
10
2 fold serial dilutions of HEV samples
Fig. 3 Antigen EIA reactivity affected by spiking in Ab+Ag sera. The HEV-Ag EIA results (S/CO) are shown in Fig. 2. EIA was conducted after two-fold serial dilutions of the two HEV samples with AbAg or Ab+Ag which were mixed and incubated at 37 °C for 15 min.
coinfection in HIV-positive and other immunocompromised patients because of delayed or failed seroconversion. In this case, HEV RNA detection is also necessary for the diagnosis of chronic hepatitis E. For assay simplicity, cost reduction and applicability in developing countries, the use of ORF2 antigen detection is being evaluated as a surrogate of HEV RNA detection for the diagnosis of HEV infection. The presence of HEV ORF2 antigen in serum was found to be in concordance with HEV RNA [12,17,18]. However, the relatively low sensitivity of ORF2 assays compared to molecular methods has been reported: HEV antigen was not detected in some blood donors, while the corresponding viremias were greater than 103 IU/mL [19]. This disparity prompted questions about the detection limit of the antigen EIA. In this study, standard curves of S/CO of the antigen EIA vs the HEV RNA concentrations were established by detecting serial dilutions of cell culture supernatant and a faecal sample from a patient (Fig. 1). In the standard curve of the genotype 4 HEV, a high correlation between S/CO and HEV RNA concentrations was found in the range of 103.5 to 100.5 IU/mL of HEV RNA, and the Pearson correlation coefficient r was up to 0.97, indicating a good corre© 2015 John Wiley & Sons Ltd
20
5
Sample1:Ab– Sample1:Ab+ Sample2:Ab– Sample2:Ab+
20
961
4.5
5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
HEV RNA (Log10 IU/mL)
lation between the two assays (Fig. 1). The minimum detection limits of the HEV-Ag EIA and real-time RT-PCR were determined to be 54.2 IU/mL and 24 IU/mL, respectively, by measuring the end-point of dilutions. The statistical difference of the detection limits of these two assays indicates that the HEV-Ag EIA is less sensitive than realtime RT-PCR. Real-time RT-PCR methods generally have proven to be more sensitive than conventional RT-PCR assays [20]. Our previous study showed that real-time RTPCR could detect a 10- to 100-fold higher dilution of genotype 1 and 4 isolates than conventional RT-PCR [17]. Thus, it could be inferred that the sensitivity of the antigen EIA is at least not less than that of the conventional RTPCR. The result of the sensitivity of real-time RT-PCR in this study is similar to previous reports, with a detection limit of 5 IU/mL–180 IU/mL in genotype 3 HEV [19]. A high variability of 100- to 1000-fold difference in the sensitivity of RT-PCR is seen between different laboratories [20]. The reproducibility of antigen detection by EIA was better than that of the real-time RT-PCR in this study. Clinical serum samples from acute hepatitis E patients were divided into four groups according to the presence of anti-HEV antibodies. Analysis showed that the concordance of HEV antigen and HEV RNA detection was highest in the group in which anti-HEV antibodies were not present (Table 2). All three HEV RNA-positive cases in this group also tested positive for ORF2 antigen (Table 2). This result indicates that HEV RNA and ORF2 antigen could be detected before the presence of detectable IgM. Antigen detection is useful for the early diagnosis of acute infection. The concordance of these two markers dropped dramatically as the illness proceeded, with the appearance of antibodies. In the IgM+/IgG+ group comprising 76 samples, the concordance rate of these two markers was 69.7%, which was significantly lower than that of the 77.8% in the total 117 samples. In addition, the concordance rate of these two markers was lower than that in the faecal samples (Fig. 2). To confirm the influence of anti-HEV IgG on the detection of HEV ORF2 by EIA, we spiked IgG+ Ag sera into IgG-Ag+ sera and incubated the mixture at 37 °C for 15 min. As expected, the OD value of the antigen reduced significantly compared to the
960
C. Zhao et al.
Table 1 Detection limit comparison between the antigen EIA and real-time RT-PCR using the WHO standard (250 000 IU/mL)
Table 3 HEV-Ag and RNA detection in faeces from acute hepatitis E patients HEV-Ag
Positive tests Concentration (IU/mL) Replicates Real-time PCR ELISA 100 50 25 12.5 6.3 3.2 1.6 0.8
12 12 12 12 12 12 12 12
12 12 11 10 7 3 1 0
12 12 4 2 0 0 0 0
groups: IgM/IgG, IgM+/IgG, IgM+/IgG+ and IgM/ IgG+. The positivity rates of HEV-Ag in the HEV RNA-positive samples were 100% (3/3), 93.8% (15/16) and 71.2% (37/52) in the IgM/IgG, IgM+/IgG and IgM+/IgG+ samples, respectively, showing a gradual decrease in these three groups (Table 2). The concordance rate of these two markers in each group varied and was inversely proportional to the titre of the antibodies. The correlation between HEV-Ag EIA activities (S/CO) and the corresponding HEV RNA concentrations (Log10 IU/mL) was determined. The Pearson correlation coefficient r was 0.55 (P < 0.01), and thus, the correlation was significant (Fig. 2a).
Detection of HEV antigen in faecal samples from acute hepatitis E patients Faecal samples from 73 acute hepatitis patients, tested previously for anti-HEV IgM or HEV RNA in serum, were available for detecting the presence of ORF2 antigen and viral RNA. Among the 73 samples, 76.7% (56/73) samples were HEV-Ag positive and 84.9% (62/73) were HEV RNA positive (Table 3). The concordance rate of HEV RNA detection by real-time RT-PCR and HEV-Ag detection by EIA was 80.8%. The correlation between HEV-Ag EIA (S/CO) and the corresponding HEV RNA concentrations (Log10 IU/mL) in
HEV RNA
Positive
Negative
Total
Positive Negative Total
52 4 56
10 7 17
62 11 73
faecal samples also was significant (Fig. 2b). The Pearson correlation coefficient r was 0.93 (P < 0.01). The concordance of HEV-Ag and HEV RNA in faecal samples seemed to be better than that in serum samples.
The presence of anti-HEV antibody affects the detection of HEV-Ag To determine the effect of anti-HEV antibodies on the detection of HEV-Ag, serial dilutions of the two HEV samples were spiked into anti-HEV IgG-positive or IgG-negative sera in vitro and tested by the EIA. By statistical analysis, the HEV-Ag EIA S/CO values of the serum sample and cell culture supernatant mixed with anti-HEV IgG-positive sera were significantly reduced compared to the control (Fig. 3).
DISCUSSION Hepatitis E is endemic in industrialized and developing countries worldwide. Its prevalence is significantly higher in developing countries, such as India, Mexico and Uganda, where outbreaks associated with faecal contamination of drinking water occur frequently [15]. Diagnosis of HEV infection mostly relies on the serological detection of anti-HEV antibodies. Several diagnostic kits are available for HEV serology in humans and have been validated. Commercial kits for the detection of anti-HEV IgM seem to be specific and reliable for the diagnosis of acute hepatitis E [16], but anti-HEV IgM may not be present at the onset of the illness. This is when HEV RNA should be used as an early diagnostic marker of acute hepatitis E. In addition, serology alone may be insufficient to diagnose HEV
Table 2 HEV-Ag and HEV RNA detections in sera from acute hepatitis E patients Antibody detection
Case
RNA+/Ag+
RNA+/Ag
RNA/Ag+
RNA/Ag
IgM/IgG IgM+/IgG IgM+/IgG+ IgM/IgG+ Total
3 17 76 21 117
3 15 37 0 55
0 1 15 0 16
0 1 8 1 10
0 0 16 20 36
© 2015 John Wiley & Sons Ltd
HEV-Ag detection for the diagnosis of HEV infection
Fig. 2 Correlation between HEV-Ag (S/ CO) and HEV RNA concentrations (Log10 IU/mL) in clinical samples, sera (a) and faeces (b) from hepatitis E patients.
HEV Ag (S/CO)
25
30
(b) 30
25
25
HEV-Ag (S/CO)
HEV-Ag (S/CO)
(a)
20 15 10
15
0
0.5
1
1.5
8
6 25
12
64
32
16
8
2
2.5
3
3.5
4
HEV RNA (Log10 IU/mL)
0
4
10
0 0
5
2
15
5
10
2 fold serial dilutions of HEV samples
Fig. 3 Antigen EIA reactivity affected by spiking in Ab+Ag sera. The HEV-Ag EIA results (S/CO) are shown in Fig. 2. EIA was conducted after two-fold serial dilutions of the two HEV samples with AbAg or Ab+Ag which were mixed and incubated at 37 °C for 15 min.
coinfection in HIV-positive and other immunocompromised patients because of delayed or failed seroconversion. In this case, HEV RNA detection is also necessary for the diagnosis of chronic hepatitis E. For assay simplicity, cost reduction and applicability in developing countries, the use of ORF2 antigen detection is being evaluated as a surrogate of HEV RNA detection for the diagnosis of HEV infection. The presence of HEV ORF2 antigen in serum was found to be in concordance with HEV RNA [12,17,18]. However, the relatively low sensitivity of ORF2 assays compared to molecular methods has been reported: HEV antigen was not detected in some blood donors, while the corresponding viremias were greater than 103 IU/mL [19]. This disparity prompted questions about the detection limit of the antigen EIA. In this study, standard curves of S/CO of the antigen EIA vs the HEV RNA concentrations were established by detecting serial dilutions of cell culture supernatant and a faecal sample from a patient (Fig. 1). In the standard curve of the genotype 4 HEV, a high correlation between S/CO and HEV RNA concentrations was found in the range of 103.5 to 100.5 IU/mL of HEV RNA, and the Pearson correlation coefficient r was up to 0.97, indicating a good corre© 2015 John Wiley & Sons Ltd
20
5
Sample1:Ab– Sample1:Ab+ Sample2:Ab– Sample2:Ab+
20
961
4.5
5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
HEV RNA (Log10 IU/mL)
lation between the two assays (Fig. 1). The minimum detection limits of the HEV-Ag EIA and real-time RT-PCR were determined to be 54.2 IU/mL and 24 IU/mL, respectively, by measuring the end-point of dilutions. The statistical difference of the detection limits of these two assays indicates that the HEV-Ag EIA is less sensitive than realtime RT-PCR. Real-time RT-PCR methods generally have proven to be more sensitive than conventional RT-PCR assays [20]. Our previous study showed that real-time RTPCR could detect a 10- to 100-fold higher dilution of genotype 1 and 4 isolates than conventional RT-PCR [17]. Thus, it could be inferred that the sensitivity of the antigen EIA is at least not less than that of the conventional RTPCR. The result of the sensitivity of real-time RT-PCR in this study is similar to previous reports, with a detection limit of 5 IU/mL–180 IU/mL in genotype 3 HEV [19]. A high variability of 100- to 1000-fold difference in the sensitivity of RT-PCR is seen between different laboratories [20]. The reproducibility of antigen detection by EIA was better than that of the real-time RT-PCR in this study. Clinical serum samples from acute hepatitis E patients were divided into four groups according to the presence of anti-HEV antibodies. Analysis showed that the concordance of HEV antigen and HEV RNA detection was highest in the group in which anti-HEV antibodies were not present (Table 2). All three HEV RNA-positive cases in this group also tested positive for ORF2 antigen (Table 2). This result indicates that HEV RNA and ORF2 antigen could be detected before the presence of detectable IgM. Antigen detection is useful for the early diagnosis of acute infection. The concordance of these two markers dropped dramatically as the illness proceeded, with the appearance of antibodies. In the IgM+/IgG+ group comprising 76 samples, the concordance rate of these two markers was 69.7%, which was significantly lower than that of the 77.8% in the total 117 samples. In addition, the concordance rate of these two markers was lower than that in the faecal samples (Fig. 2). To confirm the influence of anti-HEV IgG on the detection of HEV ORF2 by EIA, we spiked IgG+ Ag sera into IgG-Ag+ sera and incubated the mixture at 37 °C for 15 min. As expected, the OD value of the antigen reduced significantly compared to the
962
C. Zhao et al.
control group (Fig. 3). Thus, it was seen that the presence of antibodies affects the detection of ORF2 antigen. This result may help to explain why HEV antigen was not detected in some blood donors when the corresponding viremias were greater than 103 IU/mL [19]. Therefore, in clinical diagnosis of HEV infection, HEV antigen detection in serum should be combined with antibody detection, followed by less frequent confirmatory testing with molecular assays. HEV is transmitted primarily by the faecal–oral route. Faeces from hepatitis E patients might contain infectious viruses. Of the 73 faecal samples from hepatitis E patients confirmed by detecting anti-IgM or HEV RNA in sera, the ORF2 antigen and HEV RNA positivities were 56 (76.7%) and 62 (84.9%), respectively. That the concordance rate of these two assays (80.8%) was better than those of the serum samples may be attributable to the absence of antiHEV antibodies in the faeces. Animal experiments have shown that HEV RNA was detectable early and disappeared later from faecal samples than from the sera of infected animals [21]. However, because anorexia is a presenting symptom of acute hepatitis E patients, the collection of faecal samples might not be straightforward in the early period of the infection. Therefore, detecting HEV RNA or HEV-Ag in faecal samples for early diagnosis of
acute hepatitis E may be less practical; nevertheless, HEVAg in faeces should be valuable to monitor the prognosis of HEV-related diseases, such as chronic HEV infection, and will provide clinical researchers with an alternative method of low cost and easy manipulation to evaluate the infectious status of HEV-infected patients. In summary, the S/CO of the HEV-Ag EIA shows a good relationship to HEV RNA concentrations. Although the HEV-Ag EIA was less sensitive than real-time RT-PCR, it might be more sensitive than conventional RT-nested PCR. HEV-Ag detection in serum is valuable for the early diagnosis of acute hepatitis E. HEV-Ag detection in stool is likely to be useful for the diagnosis of chronic HEV infection and for monitoring the disease progression in both acute and chronic hepatitis E.
ACKNOWLEDGEMENTS Funding: The study were supported by the National Natural Science Foundation of China (grant: 81171549) and by the Science and Technology Program of Hebei Province (grant: 12277726). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
REFERENCES 1 Ahmad I, Holla RP, Jameel S. Molecular virology of hepatitis E virus. Virus Res 2011; 161: 47–58. 2 Meng XJ. Recent advances in hepatitis E virus. J Viral Hepat 2010; 17: 153–161. 3 Zhang W, He Y, Wang H et al. Hepatitis E virus genotype diversity in eastern China. Emerg Infect Dis 2010; 16: 1630–1632. 4 Geng Y, Zhao C, Fan J et al. Genotype analysis of hepatitis E virus from sporadic hepatitis E cases in northern China. Infect Genet Evol 2013; 20: 413–417. 5 Navaneethan U, Al Mohajer M, Shata MT. Hepatitis E and pregnancy: understanding the pathogenesis. Liver Int 2008; 28: 1190– 1199. 6 Kamar N, Izopet J, Rostaing L. Hepatitis E virus infection. Curr Opin Gastroenterol 2013; 29: 271–278. 7 Grewal P, Kamili S, Motamed D. Chronic hepatitis E in an immunocompetent patient: a case report. Hepatology 2014; 59: 347–348. 8 Zhao Q, Zhang J, Wu T et al. Antigenic determinants of hepatitis E
9
10
11
12
13
virus and vaccine-induced immunogenicity and efficacy. J Gastroenterol 2013; 48: 159–168. Dreier J, Juhl D. Autochthonous hepatitis E virus infections: a new transfusion-associated risk? Transfus Med Hemother 2014; 41: 29–39. Kamar N, Selves J, Mansuy JM et al. Hepatitis E virus and chronic hepatitis in organ-transplant recipients. N Engl J Med 2008; 358: 811–817. Haagsma EB, van den Berg AP, Porte RJ et al. Chronic hepatitis E virus infection in liver transplant recipients. Liver Transpl 2008; 14: 547–553. Zhang F, Li X, Li Z et al. Detection of HEV antigen as a novel marker for the diagnosis of hepatitis E. J Med Virol 2006; 78: 1441–1448. Zhang F, Qi Y, Harrison TJ et al. Hepatitis E genotype 4 virus from feces of monkeys infected experimentally can be cultured in PLC/ PRF/5 cells and upregulate host interferon-inducible genes. J Med Virol 2014; 86: 1736–1744.
14 Baylis SA, Bl€ umel J, Mizusawa S et al. World Health Organization International Standard to harmonize assays for detection of hepatitis E virus RNA. Emerg Infect Dis 2013; 19: 729–735. 15 Aggarwal R, Jameel S. Hepatitis E. Hepatology 2011; 54: 2218–2226. 16 Legrand-Abravanel F, Thevenet I, Mansuy JM et al. Good performance of immunoglobulin M assays in diagnosing genotype 3 hepatitis E virus infections. Clin Vaccine Immunol 2009; 16: 772–774. 17 Zhao C, Ma Z, Harrison TJ et al. A novel genotype of hepatitis E virus prevalent among farmed rabbits in China. J Med Virol 2009; 81: 1371–1379. 18 Majumdar M, Singh MP, Pujhari SK, Bhatia D, Chawla Y, Ratho RK. Hepatitis E virus antigen detection as an early diagnostic marker: report from India. J Med Virol 2013; 85: 823–827. 19 Vollmer T, Knabbe C, Dreier J. Comparison of real-time PCR and antigen assays for detection of hepatitis E virus in blood donors. J Clin Microbiol 2014; 52: 2150–2156.
© 2015 John Wiley & Sons Ltd
HEV-Ag detection for the diagnosis of HEV infection 20 Baylis SA, Hanschmann KM, Bl€ umel J, N€ ubling CM; HEV Collaborative Study Group. Standardization of hepatitis E virus (HEV) nucleic acid amplification tech-
© 2015 John Wiley & Sons Ltd
nique-based assays: an initial study to evaluate a panel of HEV strains and investigate laboratory performance. J Clin Microbiol 2011; 49: 1234–1239.
963
21 Liu P, Bu QN, Wang L et al. Transmission of hepatitis E virus from rabbits to cynomolgus macaques. Emerg Infect Dis 2013; 19: 559–565.
