Alternative hepatitis B virus DNA confirmatory algorithm identified occult hepatitis B virus infection in Chinese blood donors with non-discriminatory nucleic acid testing

Abstract

Background

Multiplex viral nucleic acid testing (NAT) and a discriminatory testing algorithm have been used to detect viral infections in blood donors. Non-discriminated reactive (NDR) results may arise from low hepatitis B virus (HBV) DNA levels and are challenging for donor management by blood services. The aim of this study was to evaluate the performance and feasibility of alternative viral particle concentration methods to confirm and to characterise HBV infection status in NDR donors from Dalian, China, in order to improve routine donor management according to the potential residual risk estimate.

Materials and methods

Individual donations were tested with ULTRIO Plus, and discriminated when reactive. Virions were concentrated from 12 and 6 mL plasma samples by ultracentrifugation (UC) and polyethylene glycol (PEG) precipitation, respectively. HBV DNA was detected with four nested polymerase chain reactions (95% limit of detection: 5–25 IU/mL). Amplified products were sequenced for definitive confirmation. Anti-HBc and anti-HBs were tested.

Results

Of 77,556 donors, 79 (0.1%) were NAT NDR. After viral particle concentration by UC and PEG precipitation, HBV DNA was detected in 46 (58.2%) and 34 (43.0%) NDR donors, respectively, including 61.7% of samples that were repeatedly non-reactive with multiple NAT testing. Anti-HBc and anti-HBs (median titre: 37 mIU/mL) were detected in 87.3% and 46.8% of NDR donors, respectively. Sequencing confirmed HBV DNA in 65.8% of NDR donors, of whom 96.2% were occult HBV carriers with rare mutations in S and core proteins.

Discussion

A HBV DNA confirmatory procedure with limited technical constraints was implemented successfully. The majority of NDR donors had occult HBV infections with extremely low viral DNA levels, which may constitute a potential residual threat for blood safety. Only a minority of anti-HBc+ NDR donors had anti-HBs levels high enough to consider their reinstatement as donors. The data support the permanent deferral of NDR donors to ensure maximum blood safety in areas of high HBV endemicity.

INTRODUCTION

The risk of viral transfusion-transmitted infections has been steadily reduced through the recruitment of volunteer donors, the selection of donors based on risk-behaviour evaluation, the development of increasingly more sensitive screening assays for antibodies and/or viral antigens, and the critical implementation of viral nucleic acid testing (NAT). Recently, the implementation in some countries of pathogen-reduction procedures for platelet concentrates and fresh-frozen plasma added another layer of safety. NAT for the three major transfusion-transmitted viruses, hepatitis B virus (HBV), hepatitis C virus (HCV), and human immunodeficiency virus type 1 (HIV-1) made it possible to reduce the diagnostic pre-seroconversion window period significantly and to uncover occult HBV infections (OBI) which are defined by a lack of reactivity with the most sensitive HBV surface antigen (HBsAg) assays and the presence of very low levels of HBV DNA^1,2. In 2010, the Ministry of Health of the People’s Republic of China initiated a programme of simultaneous testing for HBV DNA, HCV RNA and HIV-1 RNA in several blood banks across China. Multiplex NAT was implemented nationwide in China by 2015 with some heterogeneity in the assays and the testing algorithms used by the different blood banks. In 2011, the Dalian blood centre began routine screening of blood donations for HBV/HCV/HIV-1 nucleic acids using the Cobas TaqScreen MPX 1.0 assay (Roche Diagnostics, Mannheim, Germany) on mini-pools of six plasma samples or the PROCLEIX ULTRIO assay (Novartis, San Diego, CA, USA) for individual donations. In 2016, individual donor testing with the PROCLEIX ULTRIO Plus assay (Grifols, Barcelona, Spain) was implemented with the prospect of increased sensitivity for the detection of HBV DNA (95% limit of detection [LoD]: 3.4 IU/mL for ULTRIO Plus vs 10.4 IU/mL for ULTRIO)^3,4. The ULTRIO Plus multiplex assay indicates the presence of viral genomes in a sample with a single consensual signal that does not discriminate between the three viruses. A second step using three separate virus-specific amplification discriminatory assays is therefore necessary to identify the origin of the initial test signal. These additional discriminatory assays do not qualify as confirmation assays since they use the same methodology and the same reagents as the initial screening assay. This two-step screening system introduced the difficult problem of non-discriminated reactive (NDR) samples that are reactive with the initial multiplex screening assay but non-reactive with all discriminatory assays⁵. NDR results may be false-positive due to non-specific reactions or cross-contamination, leading to blood wastage and unnecessary loss of potentially eligible donors. However, NDR results may also be related to extremely low and fluctuating levels of viral nucleic acids, as mainly observed in donors with OBI⁶. Different approaches have been proposed to clarify the infection status of NDR donors. In Dalian, a follow-up programme was developed successfully to define infection status of NAT yield donors and donors with inconclusive test results⁷. However, this process required major logistic and technical investments, and proved to be labour-intensive and obviously time-consuming with results only being available several months after the initial testing. Additional anti-HBc testing might be considered to identify OBI in areas of low/medium HBV endemicity but it does not appear to be suitable in areas of high endemicity, such as China^1,8. Multiple repeat testing from the plasma bag, using either the multiplex assay or a second independent commercial or in-house assay may be considered to rule out potential false-positive results due to cross-contamination in the primary test tube. It may also increase the chance of detecting viral nucleic acids present at levels around the assay’s LoD according to the Poisson distribution^9,10. This approach is costly and even the most sensitive assays may fail to detect low levels of viral nucleic acids consistently^9,11. The sensitivity of detection can be significantly improved by purifying viral nucleic acids from a larger volume of plasma¹². However, the maximum volume of plasma used in commercial assays remains relatively limited (≤5 mL). Ultimately, high-speed (ultra)centrifugation has been successfully used to concentrate HBV particles from large volumes of plasma (>10 mL) allowing extremely low levels of viral DNA to be characterised¹¹. However, ultracentrifugation is usually used in specific studies involving a limited number of samples and requires well-trained staff and sophisticated equipment that are generally not available at blood banks. In addition, the method is not applicable for all viruses. The aim of the present study was to develop and evaluate an alternative viral particle concentration method for HBV, coupled with optimised sensitive nested polymerase chain reactions (PCR) to confirm the viral status of NDR donations, as a surrogate of ultracentrifugation when this latter cannot be implemented for technical or economic reasons.

MATERIAL AND METHODS

Viral serological and molecular screening of blood donations

Blood donors were mandatorily screened prior to their donations with an HBsAg rapid test (HBsAg Rapid test, In Tec Products, Xiamen, China; LoD: 5 IU/mL). Blood donations were collected from non-reactive donors and were tested further with two enzymatic immunoassays for HBsAg, anti-HCV and anti-HIV alone or in combination with HIV antigen as previously described⁷.

Between 2016–2018, viral NAT for HBV DNA, HCV RNA, and HIV-1 RNA was performed in individual donations using the multiplex PROCLEIX ULTRIO Plus assay (95% LoD: HBV 3.4 IU/mL, HCV 5.4 IU/mL, and HIV-1 21.2 IU/mL). Initially reactive samples were discriminated using the corresponding discriminatory (dHXV) assays (95% LoD of dHBV: 4.1 IU/mL) according to the manufacturer’s procedures. Discriminated NAT reactive samples were considered reactive and were not investigated further in the present study. Serology non-reactive/NAT initially reactive samples were re-tested twice with the multiplex assay and once with the dHXV assay as recommended by the manufacturer. Samples non-reactive with either the multiplex or the discriminatory assays were considered non-repeat reactive/NDR and the corresponding blood units were not released.

Viral particle concentration and nucleic acid purification

Plasma samples were centrifuged at 3,000 g (Eppendorf 5810R centrifuge [Sigma-Aldrich, St. Louis, MO, USA]) for 15 min at 4°C to eliminate cell debris. For each sample investigated, viral particles were concentrated with two different methods: (i) 12 mL of plasma were centrifuged at 250,000 g for 3 hours at 4°C using an Optima L-100XP ultracentrifuge and a SW41Ti rotor (Beckman Instruments, Brea, CA, USA). Pellets were re-suspended in 600 μL of nuclease-free water. (ii) 6 mL of plasma were precipitated with 6 mL of a 12% polyethylene glycol (PEG) 8000 solution overnight at 4°C. The precipitate was pelleted by centrifugation at 3,000 g for 20 min at 4°C, re-suspended in 600 μL of 1X phosphate-buffered saline containing ≥600 mU/mL of proteinase K, and incubated overnight at 50°C.

Nucleic acids were extracted from the concentrated material using a hepatitis B virus DNA diagnostic kit A: nucleic acids extraction reagent (WANTAI Bio-Pharm, Beijing, China) according to the manufacturer’s instructions.

Hepatitis B virus DNA amplification and sequencing

HBV DNA was amplified using nested PCR targeting the basic core promoter/precore (BCP/PC; 276 bp), the precore/core (817 bp), the pre-S/S (1,427 bp), and the S (499 bp) regions, as previously described^6,13. The PCR were optimised to a 95% LoD of 5–25 IU/mL using the TAKARA Ex Taq polymerase (0.5 U/reaction) and the World Health Organization International Standard 10/266 (National Institute for Biological Standards and Control, Potters Bar, UK). Purified amplicons were directly sequenced using the Sanger method.

Phylogenetic analysis of HBV sequences was performed using the neighbour-joining method and alignment of HBV genotypes A–H (HBV_A–H) full genome sequences available in GenBank. Genetic distances were calculated with the Kimura two-parameter method and 1,000-replicate parametric bootstrap analysis was performed to assess the reliability of the branching order. Previously characterised sequences of HBsAg+ and OBI donors from Dalian and Eastern Asia were used for comparison⁶. HBsAg+ donor sequences included 109 BCP (61 HBV_B and 48 HBV_C), 100 precore/core (57 HBV_B and 43 HBV_C), and 127 S (59 HBV_B and 68 HBV_C) sequences. OBI sequences were 46 BCP (5 HBV_B and 41 HBV_C), 118 precore/core (35 HBV_B and 83 HBV_C), and 131 S (58 HBV_B and 73 HBV_C) sequences.

Complementary serological testing

Anti-HBc and anti-HBs were tested with Elecsys Anti-HBc and Anti-HBs II (Roche Diagnostics, Mannheim, Germany) according to the manufacturer’s instructions.

RESULTS

Nucleic acid testing of blood donations and occurrence of non-discriminated reactive results

Between January 2016 and February 2018, 77,556 blood donations collected from eligible donors in the Dalian Blood Centre were tested with the ULTRIO Plus assay and 202 (0.26%) donations were reactive. Of these, 116 (57.4%) were dHBV reactive of which 60 (29.7%) tested HBsAg positive (Figure 1). One (0.5%) donation was dHIV and anti-HIV reactive. Eighty-five (42.1%) initially reactive donations were non-discriminated. Seventy-nine of these donations were re-tested twice with the multiplex NAT and once with the dHXV assays. Results showed 32/79 (40.5%) samples repeated reactive at least once with the multiplex assay (1/2 and 2/2 reactive replicates in 17 [21.5%] and 15 [19.0%] samples, respectively). All samples remained non-discriminated, leading to an estimated NAT NDR rate of 0.10% (79/77,556).

Figure 1.

Blood donation molecular screening and supplemental testing algorithm in Dalian blood centre, China

NAT: nucleic acid testing; IR: initially reactive; HBV: hepatitis B virus; HCV; hepatitis C virus; HIV-1: human immunodeficiency virus type 1; dHIV-1, dHBC and dHCV: discriminatory tests; HBsAg: hepatitis B virus surface antigen; EIA: enzymatic immunoassay; HCVAb: hepatitis B virus antibody; HIVAb/Ag: human immunodeficiency virus antibody/antigen, R/R: repeat reactive; R/NR: reactive/non-reactive; NR/NR: repeat non-reactive.

All 79 NDR donations were non-reactive for HBsAg, anti-HCV, and anti-HIV. Complementary serological testing identified 33 (41.8%) anti-HBc and anti-HBs (median: 37 IU/L; range: 10–1,000 IU/L) reactive samples, 36 (45.6%) samples that were anti-HBc reactive only, four (5.0%) that were anti-HBs reactive only (median: 276.5 IU/L; range: 20–595 IU/L), and six (7.6%) with no reactivity (Table I). The overall observed prevalence of 87.3% (69/79) of anti-HBc suggested a past exposure to HBV infection in the majority of NDR donors.

Table I.

Serological and host markers of 79 donors with non-discriminated reactive nucleic acid test results

Markers	All donors	HBV DNA confirmed status
		Positive	Negative	Indeterminate
Anti-HBc+/anti-HBs+	33 (41.8%)	22 (27.8%)	9 (11.4%)	2 (2.5%)
Anti-HBc+/anti-HBs−	36 (45.6%)	25 (31.6%)	10 (12.6%)	1 (1.3%)
Anti-HBc−/anti-HBs+	4 (5.0%)	3 (3.8%)	1 (1.3%)	-
Anti-HBc−/anti-HBs−	6 (7.6%)	2 (2.5%)	4 (5.0%)	-
Anti-HBs titre (IU/L)
Median	37	24*	114.5*	225
Range	10–1,000	11–1,000	10–519	58–392
Donor age (years)
Mean	43	44	42	43
Range	18–59	20–59	18–54	38–48
Donor status
First time	52 (65.8%)	16 (30.8%)	10 (41.7%)	1 (33.3%)
Repeat	27 (34.2%)	36 (69.2%)	14 (58.3%)	2 (66.7%)

^*p<0.0001

Hepatitis B virus DNA confirmatory testing

Forty-six (58.2%) and 34 (43.0%) NDR samples were reactive with at least one PCR assay after viral particle concentration by UC and PEG precipitation, respectively (Table II). The four in-house PCR assays showed similar amplification yields irrespective of the concentration procedure used. HBV DNA was considered definitively confirmed in a given sample only when at least one viral sequence was identified. As shown in Table III, the rate of HBV DNA confirmation was higher after UC (54.4% [43/79]) than after PEG precipitation (39.2% [31/79]) but the difference was not quite statistically significant (p=0.079). HBV DNA was detected consistently with both procedures in 27.8% (22/79) of the samples. No clear sequence could be obtained for six samples showing a very weak positive signal with either BCP/PC or S PCR, and they were classified as indeterminate. Overall, by using the two viral concentration procedures, HBV DNA was detected by PCR and definitively confirmed by sequencing in the plasma of 69.6% (55/79) and 65.8% (52/79) of NAT NDR donors, respectively. In addition, of the 47 donors who re-tested consistently non-reactive twice with ULTRIO Plus and once with the discriminatory assays (Figure 1), 25 (53.2%) and 17 (36.2%) were confirmed to be HBV DNA positive when using UC and PEG, respectively. Multiple repeat testing (12 replicates) was also performed with the multiplex ULTRIO Elite assay (HBV 95% LoD: 3.4 IU/mL) in 15 randomly selected anti-HBc+ NDR samples, of which six (40%) were reactive at least once and nine (60%) showed no reactivity. Virion concentration by UC and PEG precipitation confirmed HBV DNA in five (55.6%) and four (44.4%) of the ULTRIO Elite non-reactive samples (data not shown).

Table II.

Hepatitis B virus DNA amplification results of 79 donations that were non-discriminated reactive on nucleic acid testing after concentration of viral particles

Viral particle concentration	Number of PCR reactive samples*	Number of samples confirmed HBV reactive by sequencing	Amplified products
			BCP/PC (276 bp)	Precore/Core (817 bp)	Pre-S/S (1,427 bp)	S (499 bp)
UC	46 (58.2%)	43 (54.4%)	35 (44.3%)	32 (40.5%)	25 (31.6%)	25 (31.6%)
PEG precipitation	34 (43.0%)	31 (39.2%)	20 (25.3%)	18 (22.8%)	15 (19.0%)	22 (27.8%)
UC+PEG precipitation	55 (69.6%)	52 (65.8%)	35 (44.3%)	33 (41.8%)	29 (36.7%)	29 (36.7%)

^*At least one reactive PCR per sample.

PCR: polymerase chain reaction; HBV: hepatitis B virus; BCP/PC: basic core promoter/precore; bp: base pairs; UC: ultracentrifugation; PEG: polyethylene glycol.

Table III.

Hepatitis B virus detection rates associated with two viral particle concentration procedures

PEG precipitation	Ultracentrifugation
	Positive	Negative	Ind.*	Total (%)
Positive	22	8	1	31 (39.2)
Negative	19	24	2	45 (57.0)
Ind. *	2	1	0	3 (3.8)
Total (%)	43 (54.4)	33 (41.8)	3 (3.8)	79 (100)

^*Indeterminate: no sequence confirmation available.

Relationship between serological markers and hepatitis B virus DNA status

Anti-HBc reactivity rates were not significantly different in confirmed and non-confirmed DNA-positive samples (90.4% [47/52] vs 79.2% [19/24]; p=0.178). Similar results were obtained for anti-HBs reactivity rates (48.1% [25/52] vs 4 1.7% [ 10/24]; p =0.602). However, median anti-HBs levels were significantly lower in confirmed DNA-positive samples (24 IU/L; range: 11–1,000 IU/L) than in non-confirmed DNA-positive samples (114.5 IU/L; range: 10–519 IU/L) (p<0.0001). There was no difference in age distribution or donor status according to the HBV DNA status.

Hepatitis B virus sequence analysis

Overall, data on partial S, precore/core, and BCP/PC sequences were available for 18 (22.8%), 35 (44.3%), and 11 (13.9%) NDR donors, respectively. Only short BCP/PC sequences were available for four donors. Phylogenetic analysis of S and precore/core sequences identified ten (21.3%) genotype B, 35 (74.5%) genotype C, and two (4.2%) genotype D strains. Analysis of 2,409-nucleotide long sequences (including the whole pre-S/S region) from 101 previously characterised HBV NAT repeat reactive blood donors (37 HBsAg+ and 64 OBI carriers) from Dalian showed a similar HBV genotype distribution: 21 (20.8%) HBV genotype B, 78 (77.2%) genotype C, and two (2.0%) genotype D (data not shown). Sample cross-contamination was ruled out as sequences obtained after UC and PEG precipitation from the same sample showed >99% nucleotide similarity. In addition, precore/core and S sequences from the same individual showed identical genotype clustering.

Partial S amino acid sequences of 18 NDR donors (4 HBV_B and 14 HBV_C) were compared with their respective counterparts obtained from 127 HBsAg+ (59 HBV_B and 68 HBV_C) and 131 (58 HBV_B and 73 HBV_C) OBI blood donors (Figure 2). Four HBVC NDR sequences were identical with the two corresponding consensus sequences derived separately from the alignment of HBsAg+ and OBI HBV_C strain sequences. Multiple amino acid substitutions were observed, essentially within the major hydrophilic region (MHR), but no specific mutations were more commonly present in NDR than in HBsAg+ or OBI strains. NDR sequences shared several substitutions with OBI sequences not found in HBsAg+ sequences although many were only present in one NDR strain (HBV_B: P127L, Q129P, and G145A; HBV_C: Q101H/K/R, M103L, S114P/T, S116N, P120T, K122R, I126L, P127S, Q129P, T131P, T140S, S143L, S154P, E164G, and S167L). Substitutions found only in individual NDR sequences were G112N/R and the double mutation T131N/M133T that introduced putative N-glycosylation sites (GSS-->NSS and TSM-->NST), insertion of a threonine residue at position 117, G119A/V, F134Y, and W165L/S.

Figure 2.

Alignment of amino acid sequences deduced from partial S sequences of genotypes B and C of non-discriminated reactive hepatitis B virus strains

Non-discriminated reactive (NDR) strain sequences were aligned with consensus sequences derived from 59 HBV_B and 68 HBV_C sequences from HBsAg+ donors, and 58 HBV_B and 73 HBV_C sequences from donors with occult hepatitis B virus infection (OBI). Residues identical to the reference consensus are indicated by dots. Residues in circles were present in NDR sequences only; residues in boxes were observed in both non-repeat reactive and OBI control sequences but not in HBsAg+ sequences.

The core protein (n=35) and the BCP (n=45) sequences of NDR donors were mainly conserved when compared to sequences derived from HBsAg+ (n=109) and OBI (n=46) donors (data not shown). Eleven NDR strains (2 HBV_B and 9 HBV_c) showed amino acid substitutions (1–3 mutations per sequence) in the core protein N-terminal and the linker domains at positions previously associated with virion production and/or maturation: E76D, I97L, P130T/Q, T146I, and T147A. In the BCP region, 18 (40%) NDR strains had the G1896A mutation impairing HBeAg production. The insertion of a C nucleotide at position 1827 and CT nucleotides at positions 1823–1834 was observed in the DR1 domain of two strains.

DISCUSSION

Developing NAT assays with enhanced analytical sensitivity is usually proposed to further improve blood safety¹. However, while the detection of increasingly low viral loads is desirable, but technically challenging, it may also be associated with an increased risk of non-reproducible reactive results difficult to confirm.

In Dalian, in northeast China, the implementation of the ULTRIO Plus assay resulted in a detection yield 1.5 times higher than that previously obtained with the ULTRIO assay¹⁴. However, a similar 0.10% individual donor-NAT NDR rate was observed with both ULTRIO Plus (79/77,550 NDR donations) and ULTRIO (180/169,348 NDR donations). The improved 95% LoD for HBV DNA of ULTRIO Plus (3.4 IU/mL vs 10.4 IU/mL for ULTRIO) most probably accounted for the increased NAT yield but seemed to have no significant effect on the NAT NDR rate. The NDR rate observed in Dalian blood donors was half that recently reported in Shenzhen (0.21%), South China, but it was still higher than in South Korea (0.05%) and New Zealand (0.03%) when using the same NAT assay^12,15,16. The differences observed between studies might relate to differences in the regional prevalence of HBV infection (i.e. HBV initial testing yield of 0.72% [890/123,280] in Shenzhen¹² compared to 0.15% [120/77,556] in Dalian).

Blood services are challenged to provide appropriate management for NDR donors as such donors might seek re-testing at hospitals and be informed of a confusing negative result. Eventually, this situation may result in loss of confidence and dispute between the blood services and NDR donors who are not understanding why they are deferred permanently, and even discourage others from donating^17,18. Besides the regulatory requirements in place, blood centres are committed to provide appropriate counselling to deferred donors. To do so, they need to be able to rapidly confirm, or not, viral infection in these donors. In this study, the presence of HCV and HIV-1 RNA was not investigated further. The risk of low viraemic HCV or HIV-1 window period donations could not be totally ruled out and may constitute a limitation of the study. However, this risk appears very limited as no window period HCV or HIV-1 was identified when 103 NAT NDR donors from Dalian were followed up⁷. Additional anti-HBc testing revealed that 87.3% (69/79) of HBsAg-/NAT NDR donors carried detectable anti-HBc, suggesting past exposure to HBV and potential OBI. This was in agreement with previous studies reporting anti-HBc detection in 68–91% of Chinese donors with NAT NDR results^12,19,20. Thus, the present study focused on confirming HBV infection by using a methodological approach that proved to be efficient for detecting low levels of HBV DNA by combining viral particle concentration from large volumes of plasma and several nested PCR amplifications⁶. Viral particle concentration by UC and the sequencing of amplified products confirmed the presence of HBV DNA in 54.4% of NDR samples. However, this technology requires sophisticated equipment and is difficult to implement in blood banks and transfusion medicine laboratories. PEG precipitation of viral particles from a relatively large volume of plasma constituted an alternative approach that was easy to implement, as it required no special equipment and only limited additional training of staff already familiar with genetic amplification techniques. PEG precipitation was less efficient than UC at confirming HBV DNA in NDR samples (39.2% vs 54.4%) most probably because half the volume of plasma was used. The technical feasibility and analytical performance of PEG precipitation using up to 10 mL of plasma is currently under validation in Dalian (unpublished data). Despite these limitations, PEG precipitation was more effective at confirming HBV DNA in NDR samples compared to the manufacturer’s recommended procedure (repeating the multiplex assay twice and the discriminatory assay once) or when performing up to 12 replicate tests, while being less labour-intensive and cheaper. A recent study reported a HBV DNA confirmation rate of 46.7% when purifying viral DNA from 2.5 mL plasma with commercial reagents¹². This may raise questions about the benefit of using complicated methods and large volumes of plasma. However, the authors stated that finally only 26.3% of NDR samples were confirmed HBV-positive by nested PCR with sequencing¹². The final 46.7% figure was obtained when including samples that tested positive only with a highly sensitive quantitative PCR assay (95% LoD: 2–5 IU/mL) but without sequencing. In the present study, HBV infection was confirmed only when a viral sequence was obtained. Both UC and PEG precipitation methods appeared significantly more effective at confirming viral infection definitively when applying this stringent sequencing criterion. In addition, amplifying different HBV genomic regions appeared essential to increase the probability of HBV DNA detection irrespectively of the viral DNA capture methods used. However, a multiple viral DNA amplification-based confirmatory strategy may increase the risk of test sample cross-contamination in routine donation screening, and should be conducted in separated dedicated areas with particular attention paid to sample handling and continuous staff training. Overall, viral particle concentration procedures confirmed HBV infection in 65.8% (52/79) of NAT NDR donors, which was a higher rate than in previous reports except when long-term follow-up studies were conducted^12,15,19,21. Supplemental serological testing suggested most likely false NAT reactivity in four (5.0%) NDR donors with unconfirmed DNA reactivity and no serological marker. However, the possibility of HCV or HIV window period infections could not be ruled out. Without follow-up, it was not possible to distinguish between a window period infection and primary OBI in two (2.5%) confirmed DNA-reactive donors without a HBV serological marker. Finally, 47 (59.5%) NDR donors were identified as HBV occult carriers based on confirmed HBV DNA and anti-HBc reactivity. OBI may be underestimated since 79.2% (19/24) of NDR donors with unconfirmed HBV DNA were anti-HBc reactive and some of these may have had a DNA load below the LoD of both ULTRIO Plus assay and confirmatory assays¹¹. In addition, 53.2% (25/47) and 52.6% (10/19) of anti-HBc reactive donors with confirmed and unconfirmed DNA, respectively, had no detectable anti-HBs and carried a potential HBV transmission risk¹¹. The detection of viral DNA after UC and/or PEG precipitation suggested the presence of associated viral particles but did not presume infectivity. The lack of viral load quantification and whole HBV genome characterisation constituted limitations of the study. The prevalence of anti-HBc was higher among NDR donors (83.5% [66/79]) than that observed in the general donor population from Dalian (35.4%; data not shown), as reported by others^{12,15,16,19,22}. Several countries have proposed anti-HBc testing to decide about possible NDR donor reinstatement if the donors had anti-HBs titres higher than 100 IU/mL^16,23. This strategy has some limitations. There is no available confirmatory testing for anti-HBc and the risk of false-positive results persists¹. Simultaneous detection of HBV DNA and anti-HBs in the absence of detectable anti-HBc was observed in 3.8% (3/79) of NDR donors. Likewise, a 4% frequency was reported in “NAT yield” donors worldwide but the associated infectious risk remains unknown²⁴. In addition, the level of protection associated with low anti-HBs levels remains a matter of debate^25–27. A recent study found that anti-HBc+ blood should contain ≥200 mIU/mL of anti-HBs to be considered safe for transfusion²⁸. In the present study, 78.8% (26/33) anti-HBc+/anti-HBs+ NDR donors had anti-HBs levels of <200 mIU/mL Similarly, studies from China and New Zealand found that about 64–82% of anti-HBc+ NDR donors had anti-HBs levels <100 mIU/mL^15,19.

Sequencing provided genetic evidence to rule out any sample cross-contamination and the opportunity to investigate, at least partially, the genetics of a subgroup of OBI strains with barely detectable viraemia. HBV strains infecting NDR donors showed a genetic diversity within the S gene similar to that reported previously in OBI irrespective of HBV genotype^6,21. Compared to consensus sequences derived from either HBsAg+ or OBI blood donor sequences, amino acid substitutions were mainly concentrated within the MHR of NDR sequences and the majority of them (25/37) were found only in NDR and OBI sequences. However, caution should be taken when assuming a given mutation is OBI-specific as it may depend largely on the number and the origin of HBsAg+ control sequences included in the analysis. Substitutions within the MHR of OBI have been associated with antigenic variations and defective HBsAg production^29–31. However, only a limited number of these mutations have been fully functionally characterised, and their negative effect, if any, on HBsAg phenotype was not clarified by the studies³¹. In contrast, the TSM131–133NST mutant present in one NDR strain has been consistently associated with decreased HBsAg secretion^32,33. Similarly, among the 35 NDR core sequences examined, 11 had one to three substitutions (E76D, I97L, P130T/Q, T146I, and T147A) at positions potentially involved in viral replication^34,35. Ultimately, the negative effect on viral replication and/or HBsAg phenotype of the mutations identified is only putative and requires further in vitro functional analysis.

CONCLUSIONS

A confirmatory procedure using viral particle concentration and multiple target amplifications with limited technical constraints was implemented successfully to determine HBV status in donors with NAT NDR results. HBV DNA was definitively confirmed in 65.8% of NDR donors, of whom 96.2% were OBI carriers with extremely low viraemia but constituted a potential residual threat for blood safety. However, it should be emphasised that the observed association between NAT NDR results and OBI may apply only to the epidemiological context of Dalian in which the prevalences of HIV and HCV in blood donors are very low. Mutations in S and core proteins that may be involved in OBI genesis were observed but further functional investigations are needed. Only a minority of anti-HBc+ NDR donors had anti-HBs levels high enough to consider their reinstatement as donors. These data support the currently implemented systematic and permanent deferral of NDR donors to ensure maximum blood safety. In addition, being able to establish definitively the viral status of deferred donors who initially tested non-reactive for serological markers and NAT NDR is essential for appropriate management of the donors.