Development and characterization of secondary standards for nucleic acid amplification technology (NAAT) assays for detection of hepatitis E virus

Abstract

Background:

To harmonize assays for detection of HEV RNA, a World Health Organization International Standard (WHO IS) was established. The WHO IS represents the highest order standard for HEV RNA but is limited in quantity. Secondary standards are needed to limit the use of WHO IS and minimize the need to replace it.

Objective:

Establish secondary standards for HEV NAAT assays and to calibrate these against the WHO IS.

Methods:

Stocks of genotype 3 HEV were prepared using both cell lysates and cell culture supernatants to produce non-enveloped and quasi-enveloped virus stocks, respectively. Both stocks were heat-inactivated, diluted in negative human plasma, and lyophilized to produce two candidate secondary standards: HEV-RR (non-enveloped virus) and HEV-RR.1 (quasi-enveloped virus). Both candidate standards were characterized and calibrated against the WHO IS for HEV RNA in an international collaborative study.

Results:

The collaborative study returned a total of 15 data sets, with different RNA extraction and amplification methods. The estimated mean values relative to the WHO IS (250,000 IU/ml) are 229,000 IU/ml and 355,000 IU/ml for HEV-RR and HEV-RR.1, respectively.

Conclusion:

We have established two secondary standards for HEV RNA calibrated against the WHO IS. These standards are non-infectious and stable under different storage temperatures.

INTRODUCTION

Hepatitis E virus (HEV) is an important cause of acute hepatitis worldwide [1]. In developing countries with poor sanitation infrastructure, HEV is spread through contaminated water and causes large disease outbreaks associated with HEV genotypes 1 and 2 [2]. Most persons infected with HEV genotypes 1 and 2 experience acute, self-resolving hepatitis, but severe disease and high mortality are seen in pregnant women [2]. In industrialized countries, sporadic autochthonous infections caused by HEV genotypes 3 and 4 present in undercooked meat from infected animals are more common [3]. These infections are mostly asymptomatic [4], but the virus can establish chronic infection and cause progressive liver disease in immune suppressed persons (e.g. organ transplant recipients) [5].

HEV is transmitted by the fecal-oral route as a non-enveloped virus. In infected persons, HEV circulates in the blood as a quasi-enveloped virus, cloaked in host membranes with no virus encoded glycoproteins on the outside [6]. HEV viremia has been observed at high levels in blood donors in several European countries. In England between March 2016 and December 2017, HEV RNA was detected in 480 out of 1,838,747 blood donations or 0.026% [7]. A study of plasma donations in Germany and Portugal from 2015 to 2018 detected HEV RNA in 171 out of 187,515 donations, or 0.09% [8]. A more recent study of German blood donors between 2019 and 2020 detected HEV in 31 out of 16,236 donations or 0.19% [9]. Studies in the US have shown lower rates of viremia with RNA detectable in 3 out of 50,724 donations [10]. Nonetheless, HEV is present in the US at high rates in pigs [11] and has been detected in commercial offal products [12]. Thus, HEV represents a potential threat to both food and blood safety globally.

Diagnosis of acute HEV infection relies on detection of anti-HEV IgM in the serum or detection of viral RNA in the serum or feces. Screening of blood donations for HEV RNA is the only effective technique currently available to mitigate transfusion-transmission. Methods for detection of HEV RNA (either for the purpose of diagnosis or for screening blood donations) rely on Nucleic Acid Amplification Technology- (NAAT-) based assays. NAAT assays can differ according to the sample type, sample volume, nucleic acid isolation method, and the assay itself. To harmonize assays for detection of HEV RNA, a World Health Organization International Standard (WHO IS) was established in 2011 using a genotype 3a HEV strain from a blood donation [13]. The WHO IS represents the highest order standard for HEV RNA but is limited in quantity. WHO recommends that use of a WHO IS should be limited to the calibration of secondary biological reference materials [14]. An example of such a secondary standard is the European Pharmacopeia Biological Reference Preparation for HEV nucleic acid testing, established by the European Directorate for the Quality of Medicines and Healthcare for use as a positive control when screening solvent/detergent-treated plasma [15]. Secondary standards are needed to limit the use of WHO IS and minimize the need for replacement.

The objective of this study was to establish secondary standards for HEV NAAT assays calibrated against the WHO IS.

MATERIALS AND METHODS

Candidate Standard Preparation

Methods for preparation of two candidate secondary standards, heat inactivation and lyophilization are described in the Supplemental Information.

International Collaborative Study

The purpose of the Collaborative Study was to determine the potency of the candidate secondary standards (i.e. viral RNA concentration in International Units - IU) by calibration against the WHO IS for HEV RNA NAAT assays in independent laboratories. HEV-RR and HEV-RR.1 lyophilized reagents and WHO IS for HEV RNA were sent to ten laboratories from six different countries, including regulatory agencies, national control laboratories, clinical diagnostic laboratories, and in vitro diagnostic manufacturers. Our laboratory at CBER/FDA was also one of the testing laboratories, making a total of 11 participating laboratories. Each laboratory received sufficient material to perform three independent test runs, each performed in triplicate on a different day using their routine assay for HEV RNA. Participating laboratories with qualitative assays were asked to perform, on the first day of testing, an initial ten-fold dilution series in HEV negative human plasma to determine the endpoint of detection. For the next two testing runs, laboratories tested half-log dilutions, including 2 dilutions above and 2 dilutions below the endpoint. Participating laboratories with quantitative assays were asked to create a standard curve using the WHO IS for HEV RNA and run in parallel with our candidate secondary standards with the same number of dilutions and replicates per dilution, in three independent test runs.

Statistical analysis

For qualitative assays, the relative potency of the two candidate secondary standards (i.e., HEV-RR and HEV-RR.1) against the 1st WHO IS for HEV RNA (code 6329/10; 250,000 IU/ml) was calculated using Poisson regression analysis, which assumes a single NAAT-detectable unit (NDU) will result in a positive test. For each laboratory, data from all runs at each dilution were pooled to obtain the proportion of positive results and estimate the log₁₀ NDU/ml. The overall mean estimate for qualitative assays was calculated as the mean of the estimates from all laboratories. Variation between laboratories was assessed with standard deviation (SD) and geometric coefficient of variation (%GCV).

For quantitative assays, the relative potency of the two candidate secondary standards (i.e., HEV-RR, HEV-RR.1) against the 1st WHO IS for HEV RNA (250,000 IU/ml) was calculated using the parallel-line regression analysis. To check parallelism, we fitted separate linear regressions for WHO IS, HEV-RR, and HEV-RR.1 (expressed in log₁₀ IU/mL) against log₁₀ dilution level for each run in each laboratory. The ratios of slopes and associated 95% confidence intervals (CIs) were calculated. We considered no obvious deviation from parallelism when ratios of slopes for each run fell within 0.7–1.3 with associated 95% CIs between 0.5–1.5. For all assays, the mean ratio of slopes across runs fell within 0.8–1.2. For each laboratory, a single estimate was calculated for each run and a mean estimate was calculated across three runs. The overall mean estimate for quantitative assays was calculated as the mean of the estimates from all laboratories. Variation between laboratories was assessed with SD and %GCV.

Qualitative and quantitative results were combined to calculate the overall means and associated 95% CIs for HEV-RR and HEV-RR.1 using linear mixed model.

Stability and accelerated thermal degradation studies

After lyophilization, twelve vials of HEV-RR and HEV-RR.1 were stored at +4°C, +25°C, +37°C and +45°C for accelerated thermal degradation studies, and at −20°C and −80°C for stability assessment. Accelerated thermal degradation was evaluated weekly for 8 weeks of storage. Stability studies have been performed monthly up to 3 months, then every 3 months up to one year, and will continue to be performed annually after the first year of storage. At each time interval, the contents of the vials were resuspended in 0.5mL of nuclease-free water, RNA extracted and analyzed by RT-qPCR.

RESULTS

Two candidate secondary standards were prepared from HEV-infected hepatoma cell lysates (HEV-RR) and cell culture medium from HEV-infected human hepatocytes (HEV-RR.1). Both secondary standards were heat-inactivated, and the complete loss of infectivity was confirmed by infectivity assays (Supplemental Information and Figure S1). Sequencing showed that both preparations were identical to the molecular clone used to prepare the source material (Genbank JQ679013). Isopycnic gradient analyses indicated that the virions in the stocks used to prepare HEV-RR were mostly non-enveloped, like virus shed in feces of infected persons. In contrast, the viral particles in HEV-RR.1 possessed a buoyant density consistent with membrane-associated or quasi-enveloped HEV, like virus circulating in the blood of infected persons (Figure S1).

The two candidate secondary standards, HEV-RR and HEV-RR.1, were calibrated against the WHO IS for HEV RNA in an international Collaborative Study resulting in six qualitative and nine quantitative data sets with different RNA extraction and amplification methods (Table 1). All laboratories prepared RNA using commercially available extraction kits: five laboratories used a methodology based on silica membrane, three laboratories used silica-coated magnetic-particle technology, two laboratories used a magnetic beads-based methodology and one laboratory used target capture oligonucleotides (Table 1). Nine out of eleven laboratories used one step RT-PCR TaqMan assays for the detection and quantification of HEV RNA targeting mainly a conserved region of the genome where ORF2 and ORF3 overlap. One laboratory used primers and probe targeting RdRP. Four laboratories did not provide the targeted genome region because this was proprietary information (Table 1). Three laboratories (Lab codes 1, 5 and 10) provided qualitative and quantitative data. One laboratory (Lab 6) used two different quantitative assays and reported the results separately (6a and 6b). Therefore, overall means for HEV-RR and HEV-RR.1 were calculated using a total of 15 data sets.

Table 1 –

Qualitative and Quantitative assays used in the collaborative studies

RNA extraction method	Amplification assay method	Region of genome amplified	Assay Reference
QIAamp MinElute Virus Spin kit (Qiagen)	One-step RT-qPCR using TaqMan® Fast Virus 1 step (Applied Biosystems) on Roche LightCycler® 480 instrument (Roche)	ORF3	[4, 21]
QIAamp Viral RNA mini kit (Qiagen)	One-step RT-qPCR using RNA to CT Kit on ABI 7300 PCR System (Applied Biosystems)	ORF3	[21, 22]
NUCLISENS easyMAG (BioMérieux)	Hemi-nested PCR on GeneAmp PCR 9700 (Applied Biosystems) followed by QIAxcel capillary electrophoresis (Qiagen) and sequencing.	RdRp	[23, 24]
HEV ELITe MGB® Kit (ELITechGroup)	One-step RT-qPCR on ELITe InGenius® instrument (ELITechGroup)	ORF2	N/A
Procleix UltrioPlexE (UPE) (Grifols Diagnostic Solutions Inc.)	TMA	ORF3	N/A
QIAamp MinElute Virus Spin kit (Qiagen)	Real-Star HEV RT PCR Kit v 2.0 (Altona Diagnostics) on LightCycler 480 II instrument (Roche)	Proprietary information	N/A
cobas® HEV test	RT-PCR on cobas® 6800 and cobas® 8800 Systems (Roche)	Proprietary information	N/A
NUCLISENS easyMAG (BioMérieux)	Real-Star HEV RT-qPCR RUO kit 2.0 (Altona Diagnostics) on QuantStudio5 instrument (Applied Biosystems)	Proprietary information	N/A
EZ1 Virus Mini Kit v2.0 (Qiagen)	One-step RT-qPCR using TaqPath 1-Step kit (Applied Biosystems) on StepOnePlus RT-PCR system (Thermo Fisher)	ORF3	[21, 22]
QIAamp MinElute Virus Spin kit (Qiagen)	One-step RT-PCR using EurobioPlex HEV Real Time RT-PCR kit (Eurobio Scientific)	Proprietary information	N/A
QIAamp MinElute virus spin kit (Qiagen)	One-step RT-qPCR using QuantiTect Probe RT-PCR kit (Qiagen) on ABI 7500 PCR System (Applied Biosystems)	ORF2/3	[25] (modified)
eMAG/easyMAG (bioMerieux)	One-step RT-qPCR using 1xTaqMan® Fast Virus 1-Step Master on ABI 7500 PCR System (Applied Biosystems)	ORF3	[26]
Total Nucleic Acid Extraction Roche MagNAPure (Roche)	One-step RT-qPCR on Roche LightCycler® 480 2.0 instrument (Roche)	ORF2/3	N/A

RT-PCR: Real-Time Reverse Transcription Polymerase Chain Reaction

RT-qPCR: Real-Time Reverse Transcription quantitative Polymerase Chain Reaction

N/A: Not Applicable

Results from qualitative assays were analyzed assuming a Poisson distribution and a single estimate expressed as log₁₀ NDU/mL was calculated for each laboratory (Table 2, Figure 1A). Overall mean estimates for qualitative assays were 5.28 log₁₀ NDU/ml and 5.6 log₁₀ NDU/ml for HEV-RR and HEV-RR.1, respectively (Table 2).

Table 2 –

The mean estimates of HEV-RR and HEV-RR.1 calibrated against WHO IS for HEV RNA based on qualitative (Log10 NDU/ml) and quantitative assays (Log₁₀ IU/ml)

Laboratory Code	HEV-RR	HEV-RR.1
Qualitative assays
1	5.70	6.29
4	5.47	5.47
5	5.77	5.86
7	5.21	5.49
10	4.84	5.07
11	4.69	5.41
Mean	5.28	5.60
SD	0.48	0.42
%GCV	155	124
Quantitative assays
1	5.35	5.74
2	5.60	5.68
3	5.67	5.79
5	5.57	5.62
6a	4.99	4.71
6b	5.43	5.71
8	5.70	5.40
9	5.32	5.69
10	4.98	5.37
Mean	5.40	5.52
SD	0.27	0.34
%GCV	69	92

SD – Standard Deviation

%GCV – % Geometric Coefficient of Variation = $\sqrt{e^{SD*ln{(10)}^2-1}}$

Figure 1 –

Estimated potency values of candidate secondary standards HEV-RR and HEV-RR.1 based on qualitative (A), and quantitative (B) assays as determined by 11 independent laboratories. Qualitative results are reported in Log₁₀ NDU/ml and quantitative results are reported in Log₁₀ IU/ml. Blue horizontal dashed line represents the overall mean. Solid, horizontal lines represent mean +/− two standard deviations (2SD).

Results from quantitative assays were analyzed using parallel-line regression. For each laboratory, a single estimate was calculated for each run and a mean estimate was calculated across three runs (Table 2, Figure 1B). The overall mean estimates for quantitative assays were 5.40 log₁₀ IU/ml and 5.52 log₁₀ IU/ml for HEV-RR and HEV-RR.1, respectively.

The mean estimates of HEV-RR and HEV-RR.1 from each laboratory were in good agreement with each other for both qualitative and quantitative assays (Figure 2). For HEV-RR, most estimates (11 out of 15) were within a range of 5.2 to 5.8 log₁₀ IU/ml while the remaining estimates were ≤ 5 log₁₀ IU/ml, including two qualitative and two quantitative assays. For HEV-RR.1, most estimates (12 out of 15) were within a range of 5.4 to 5.9 log₁₀ IU/ml. One laboratory presented an estimate greater than 6 log₁₀ IU/ml using a qualitative assay, while two laboratories presented estimates ≤ 5.1 log₁₀ IU/ml with one qualitative and one quantitative assay. Qualitative and quantitative results were combined to calculate the overall mean estimates and associated 95% CIs for HEV-RR and HEV-RR.1 (Table 3). The overall means were 5.36 log₁₀ IU/ml for HEV-RR and 5.55 log₁₀ IU/ml for HEV-RR.1. Therefore, the estimated mean values relative to the WHO IS (250,000 IU/ml) are 229,000 IU/ml and 355,000 IU/ml for HEV-RR and HEV-RR.1, respectively (Table 3).

Figure 2 –

Histograms of estimated potency values (Log₁₀ IU/ml) of candidate secondary standards HEV-RR and HEV-RR.1. Results for qualitative assays are represented by white squares, and results for quantitative assays are represented by gray squares. The number listed in each box represents the laboratory code number.

Table 3 –

Overall mean estimates of HEV-RR and HEV-RR.1

Sample	Mean		95% confidence interval (95% CI)
	Log IU/mL	IU/mL	Log IU/mL	IU/mL
HEV-RR	5.36	2.29E+05	5.16 – 5.55	1.45E+05 – 3.55E+05
HEV-RR.1	5.55	3.55E+05	5.37 – 5.74	2.34E+05 – 5.50E+05

Stability testing was conducted to monitor the potency of the secondary standards over time. There was no decrease of HEV RNA titers in HEV-RR and HEV-RR.1 stored at −20°C and −80°C up to 24 weeks compared to the original value at week 0, demonstrating good stability of the secondary standards (Figure 3). An accelerated thermal degradation study was performed to characterize the robustness of lyophilized standard materials and to assess the stability of the material after extreme shipping conditions. RNA degradation was observed only at 45°C for HEV-RR.1 after two weeks, with a less than 10-fold decrease compared to the original value at week 0 (Figure 4).

Figure 3 –

Real-time stability analysis of candidate secondary standards HEV-RR and HEV-RR.1. Vials of HEV-RR and HEV-RR.1 were stored at −20°C or −80°C and analyzed over time. At each timepoint the contents of the vials were resuspended in 0.5mL of nuclease-free water, viral RNA extracted, and analyzed by RT-qPCR. Results are shown in log₁₀ IU/ml. The dashed line represents the original value at week 0 for each candidate.

Figure 4 -.

Accelerated thermal degradation analysis of candidate secondary standards HEV-RR and HEV-RR.1. Vials of HEV-RR and HEV-RR.1 were stored at 4°C, 24°C (room temperature), 37°C, 45°C and analyzed over time. At each timepoint the contents of the vials were resuspended in 0.5mL of nuclease-free water, viral RNA extracted, and analyzed by RT-qPCR. Results are shown in log₁₀ IU/ml. The dashed line represents the original value at week 0 for each candidate.

DISCUSSION

Biological reference materials are used for standardization of blood screening and diagnostic assays for infectious diseases. The WHO IS is the highest order biological reference material. However, WHO recommends that use of a WHO IS should be limited to the calibration of secondary biological reference materials, i.e. secondary standards, to preserve stocks of the IS [14]. Secondary standards are defined by WHO as “reference standards established by regional or national authorities, or by other laboratories, that are calibrated against and traceable to WHO reference materials” [14]. The objective of this project was to establish secondary standards for HEV NAAT assays calibrated against the WHO IS for HEV RNA.

The first step in preparation of secondary standard is the identification of a suitable stock material, such as pooled viremic plasma or cell culture-propagated virus [16]. High titer viremic plasma can be difficult to source for HEV because of the short duration of viremia in acute infection. HEV-positive stool specimens offer an alternative starting material but can also be challenging to source and may show variable stability over time [17]. To overcome this challenge, high titer virus stocks were prepared using HEV propagated in cell culture. However, cell-culture systems for HEV propagation present limitations and the ideal infection model is yet to be established. For the present study, two methods were tested for production of HEV virus stocks. The first method involved purification of HEV particles from lysates of Huh7.S10–3 cells that had been transfected with viral genomes to launch replication. The second method involved pooling cell culture medium collected from HEV-infected HLCM-HH cultures. Virus stocks were heat-inactivated and lyophilized to produce two candidate secondary standards, HEV-RR and HEV-RR.1, that were further characterized and calibrated against WHO IS for HEV RNA [13] in an international collaborative study.

During infection, HEV is shed in feces as non-enveloped virus but circulates in blood as a quasi-enveloped virus, cloaked in host membranes [18, 19]. Although the quasi-enveloped virus is the predominant form of HEV in blood, non-enveloped HEV has also been detected in plasma from a subset of blood donors at lower levels. A recent study of HEV-infected blood donors detected non-enveloped HEV in 8 out of 23 plasma samples [20]. The virus stock produced from cell lysates, used to prepare HEV-RR, was mostly non-enveloped like HEV shed in feces. The virus stock produced from cell culture medium, used to prepare HEV-RR.1, was predominantly quasi-enveloped like HEV circulating in blood. The two different standards provide end-users with the opportunity to evaluate the capacity of different assays to accurately detect and quantify the different forms of HEV (non-enveloped vs quasi-enveloped). Real time stability and accelerated thermal degradation studies showed little difference between the two preparations post lyophilization.

Our secondary standards are heat-inactivated, non-infectious, and lyophilized to have a longer shelf-life. Furthermore, both secondary standards were prepared in negative human plasma to mimic relevant clinical sample matrices.

Eleven independent laboratories participated in the study using their routine assays to calibrate HEV-RR and HEV-RR.1 against the WHO IS. Since some laboratories returned data from more than one assay, a total of 15 data sets were returned to us, including both qualitative and quantitative methods. Data from all laboratories were used to calibrate HEV-RR and HEV-RR.1 against the WHO IS, providing a potency estimate in IU for each of the secondary standards. Most potency estimates (12 out of 15 data sets) for the secondary standards were within a range of 5.2 to 5.9 log IU/ml, indicating good agreement among all the values. Stability testing demonstrated no significant loss of potency of the secondary standards overtime.

CONCLUSION

In conclusion, we have established two secondary standards for HEV, which were well-characterized and calibrated against the current WHO IS. The estimated mean values relative to the WHO IS (250,000 IU/ml) are 229,000 IU/ml (HEV-RR) and 355,000 IU/ml (HEV-RR.1). These secondary standards are non-infectious, stable over the range of storage temperatures from −20°C to −80°C and represent a useful tool for interlaboratory harmonization of HEV NAAT assays. HEV secondary standards are available to researchers, clinical diagnostic laboratories, and developers of NAAT-based assays upon request.

Highlights:

• High titer stocks of Hepatitis E virus were produced using cell culture systems

• Two secondary standards were calibrated against the WHO IS in a Collaborative Study

• The secondary standards are non-infectious and stable

• These standards are useful for interlaboratory harmonization of HEV NAAT assays