Evaluation of a new human immunodeficiency virus antigen and antibody test using light-initiated chemiluminescent assay

Abstract

Objectives
The goal of this study is to evaluate the analytical and clinical performance of a new human immunodeficiency virus antigen and antibody (HIV Ag/Ab) test using light-initiated chemiluminescent assay (LiCA®) and compare it with the well-established Architect® HIV Ag/Ab combo assay in a clinical setting.

Methods
We used banked samples and national reference controls to identify the ability to detect HIV Ag/Ab and different viral subtypes. Thirteen seroconversion panels were tested to evaluate early detection of HIV. A total of 21,042 patient samples were collected to compare the diagnostic performance of LiCA® with Architect®. Screening-reactive results were confirmed by Western blotting and nucleic acid testing.

Results
Total imprecision was within 2.49%–6.56%. The C5–C95 interval was within −10.20%–7.67% away from C50. The limit of detection for p24 antigen was <1.00 IU/mL. Using national reference panels and banked sample pools, LiCA® successfully detected all negative and positive controls in line with the criteria, and all HIV-positive specimens containing different viral subtypes. In 13 seroconversion panels, LiCA® detected reactive results on average 5.73 days (95% CI: 3.42–8.04) after the initial RNA test results were confirmed positive, which was 1.27 days earlier (−3.75 to 1.21) compared to Architect®. Paired comparisons in 21,042 clinical patient samples demonstrated that LiCA® detected HIV Ag/Ab with a slightly better performance in sensitivity (100.00% vs. 99.65%), specificity (99.85% vs. 99.81%), negative predictive value (NPV, 100.00% vs. 99.99%), and positive predictive value (PPV, 89.84% vs. 87.85%) than Architect®. Total agreement between two assays was 99.67% with a kappa value of 0.89.

Conclusion
LiCA® HIV Ag/Ab is a precise and highly sensitive assay for measuring HIV-1 p24 antigen and HIV-1/2 antibodies with differentiated S/Co values of Ag/Ab. The assay is appropriate for use in the clinical routine test for the early detection of HIV.

Full text

Evaluation of a new human
immunodeficiency virus antigen
and antibody test using light-
initiated chemiluminescent assay
Yijun Li
1
†
, Fangfang Jin
2
†
, Yunhui Li
2
†
, Yan Li
1
, Yajie Wang
2
*
and Ximing Yang
1
*
1
Clinical Laboratory, Dongzhimen Hospital, Beijing University of Chinese Medicine, Beijing, China,
2
Clinical Laboratory, Beijing Ditan Hospital, Capital Medical University, Beijing, China
Objectives: The goal of this study is to evaluate the analytical and clinical performance
of a new human immunodeficiency virus antigen and antibody (HIV Ag/Ab) test using
light-initiated chemiluminescent assay (LiCA®) and compare it with the well-
established Architect®HIVAg/Abcomboassayinaclinicalsetting.
Methods: We used banked samples and national reference controls to identify
the ability to detect HIV Ag/Ab and different viral subtypes. Thirteen
seroconversion panels were tested to evaluate early detection of HIV. A total
of 21,042 patient samples were collected to compare the diagnostic
performance of LiCA®with Architect®. Screening-reactive results were
confirmed by Western blotting and nucleic acid testing.
Results: Total imprecision was within 2.49%–6.56%. The C
5
–C
95
interval was
within −10.20%–7.67% away from C
50
. The limit of detection for p24 antigen was
<1.00 IU/mL. Using national reference panels and banked sample pools, LiCA®
successfully detected all negative and positive controls in line with the criteria, and
all HIV-positive specimens containing different viral subtypes. In 13 seroconversion
panels, LiCA®detected reactive results on average 5.73 days (95% CI: 3.42–8.04)
after the initial RNA testresults were confirmed positive, which was 1.27 days earlier
(−3.75 to 1.21) compared to Architect®. Paired comparisons in 21,042 clinical
patient samples demonstrated that LiCA®detected HIV Ag/Ab with a slightly better
performance in sensitivity (100.00% vs. 99.65%), specificity (99.85% vs. 99.81%),
negative predictive value (NPV, 100.00% vs. 99.99%), and positive predictive value
(PPV, 89.84% vs. 87.85%) than Architect®. Total agreement between two assays
was 99.67% with a kappa value of 0.89.
Conclusion: LiCA®HIV Ag/Ab is a precise and highly sensitive assay for
measuring HIV-1 p24 antigen and HIV-1/2 antibodies with differentiated S/Co
values of Ag/Ab. The assay is appropriate for use in the clinical routine test for the
early detection of HIV.
KEYWORDS
human immunodeficiency virus, light-initiated chemiluminescent assay, homogeneous
immunoassay, performance evaluation, LiCA
®
Frontiers in Cellular and Infection Microbiology frontiersin.org01
OPEN ACCESS
EDITED BY
Kamal El Bissati,
The University of Chicago, United States
REVIEWED BY
Haifei Jiang,
Mayo Clinic, United States
Victor Manuel Luna-Pineda,
Hospital Infantil de Me
´xico
Federico Go
´mez, Mexico
*CORRESPONDENCE
Ximing Yang
dzmmys@163.com
Yajie Wang
wangyajie@ccmu.edu.cn
†
These authors have contributed
equally to this work and share
first authorship
RECEIVED 01 August 2024
ACCEPTED 06 January 2025
PUBLISHED 31 January 2025
CITATION
Li Y, Jin F, Li Y, Li Y, Wang Y and Yang X
(2025) Evaluation of a new human
immunodeficiency virus antigen and
antibody test using light-initiated
chemiluminescent assay.
Front. Cell. Infect. Microbiol. 15:1474127.
doi: 10.3389/fcimb.2025.1474127
COPYRIGHT
© 2025 Li, Jin, Li, Li, Wang and Yang. This is an
open-access article distributed under the terms
of the Creative Commons Attribution License
(CC BY). The use, distribution or reproduction
in other forums is permitted, provided the
original author(s) and the copyright owner(s)
are credited and that the original publication
in this journal is cited, in accordance with
accepted academic practice. No use,
distribution or reproduction is permitted
which does not comply with these terms.
TYPE Original Research
PUBLISHED 31 January 2025
DOI 10.3389/fcimb.2025.1474127

1 Introduction
Acquired immune deficiency syndrome (AIDS), caused by the
human immunodeficiency virus (HIV), is a global infectious disease
that seriously endangers human health. HIV attacks the patient’s
immune system, which causes a variety of complications and even
death. In addition, the virus continues to be spread through blood
transfusions, sexual contact, and drug use (Volberding and Deeks,
2010). As reported by the US Centers for Disease Control and
Prevention (CDC), approximately 1.7 million people were newly
infected with HIV worldwide in 2018, and 770,000 people died
among those living with AIDS (UNAIDS, 2019). Of added concern
is that several African countries and Middle East nations are far
from controlling this epidemic (El-Sadr et al., 2019). Fortunately,
continued access to antiretroviral therapy (ART) in the early stages
of infection can have a major positive effect on reducing
transmission and death among those living with HIV (Cohen
et al., 2011;Rodger et al., 2019). Therefore, early diagnosis of the
infection and linking the patients to the proper medicinal therapy
are critical for the management of AIDS patients. This heavily relies
on an effective viral screening strategy.
During recent decades, various methods for detecting HIV
antibody (Ab), p24 antigen (Ag), and ribonucleic acid (RNA) in
serum or plasma have been developed for the diagnosis of HIV
infection (Alexander, 2016;Gray et al., 2018). The detection
capabilities of various methods exhibit certain variations. For
instance, the sensitivity of first-generation HIV-Ab reagents is
99%, with a specificity ranging from 95% to 98% (Alexander,
2016). Taking the Alere HIV Combo POCT test as an example
for HIV-1 P24 antigen detection, its sensitivity is 88%, and its
specificity is 100% (Fitzgerald et al., 2017). The detection capability
of HIV RNA is primarily reflected in its sensitivity. Currently, the
lower limit of detection for ordinary HIV RNA quantitative
reagents is 100–200 copies/mL, while high-sensitivity quantitative
reagents can achieve a lower limit of detection as low as 20 copies/
mL. In general, a suspected subject is initially detected with the
screening test for HIV Ag/Ab, and the screening-reactive assay is
further confirmed by Western blot (WB) and/or nucleic acid test
(NAT) to clarify the infection condition or even a false-positive
result (Centers for Disease Control and Prevention, 2018;Branson,
2019). The Ag/Ab screening test with higher sensitivity and
specificity can be more favorable in clinical practice due to its
better detection capability for viral infection from first- to fifth-
generation commercial kits (Alexander, 2016). Unlike the third-
generation assay that detects Ab alone, the fourth- and fifth-
generation assays detect both Ab and p24 Ag in combination,
reducing the test-negative window to 8–14 days (Alexander, 2016;
Qiu et al., 2017). Moreover, the fifth-generation assay can
differentiate Ab and Ag reactivity instead of a single result by the
fourth-generation kit, thus facilitating the confirmatory strategy for
early detection of HIV (Muhlbacher et al., 2019;Yang et al., 2022).
Here, we introduce a new fifth-generation HIV Ag/Ab combination
test that is based on the light-initiated chemiluminescent assay (LiCA®)
(Bian et al., 2018). LiCA®provides a fully automatic homogeneous
immunoassay platform, which has been widely developed for detection
of various analytes with high sensitivity and specificity, such as
hormones (Yu et al., 2021;Wang et al., 2022), cardiac proteins (Yang
et al., 2022;Li et al., 2024), and Ag and Ab (Li et al., 2023;Yu et al.,
2023). In this study, we aim to evaluate the performance of the LiCA®
HIV Ag/Ab assay in analytical and clinical perspectives and compare it
with the well-established Architect®HIV Ag/Ab combo test in
clinical setting.
2 Materials and methods
2.1 Sample collection and HIV Ag/Ab
serological assays
We recruited a total of 21,042 clinical serum specimens from
inpatients and outpatients in Dongzhimen Hospital. HIV Ag/Ab
screening tests were performed on the LiCA®500 platform
(Chemclin Diagnostics, Beijing, China) and the Architect®
i2000SR system (Abbott Laboratories, IL, USA) in parallel
(Figure 1). LiCA®HIV Ag/Ab is a one-step fully automatic
homogeneous immunoassay. Serum samples were dispensed into
two cuvettes for detecting antibodies to HIV-1/HIV-2 subtypes and
HIV-1 p24 antigen, respectively. The Ag/Ab reactivity can
be differentiated in results. A signal-to-cutoff (S/Co) ratio ≥1.0 in
any cuvette is considered to be screening-reactive. Time to the first
report is approximately 25 min. Architect®HIV Ag/Ab combo is a
two-step indirect immunoassay and reports combined Ag/Ab
reactivity in a single result. A ratio of S/Co ≥1.0 is regarded as
screening-reactive. Any one assay with a reactive S/Co was retested
in duplicate. The repeated screening-reactive assays were then
allocated for antibody identification with the WB test of
recomLine HIV-1/HIV-2 IgG (Mikrogen Diagnostics, Neuried,
Germany). Subjects with WB-indeterminate and WB-negative
results were further identified with the Cobas®AmpliPrep/
Cobas®TagMan®HIV-1 RNA test (Roche Diagnostics,
Mannheim, Germany). Finally, the HIV Ag/Ab true-positive
group included both WB-positive and RNA-positive results, and
the true-negative group included those with screening-negative
results in both assays and RNA-negative results. The testing
protocol was plotted as shown in Figure 2 (Alexander, 2016;
Fitzgerald et al., 2017).
2.2 Precision study
We performed precision analysis in S/Co ratios according to the
EP15-A3 protocol of the Clinical and Laboratory Standards
Institute (CLSI) (CLSI, 2014), using three levels of patient serum
samples, and two levels of controls for HIV antibodies and p24
antigen, respectively. An acceptable coefficient of variation (CV)
was ≤15%. In addition, we followed the guideline of EP12-A2 (CLSI,
2008) and prepared a series of dilutions with a positive sera to
determine the C
50
target and the C
5
–C
95
interval for HIV antibodies
and p24 antigen, respectively. An acceptable C
5
–C
95
interval was
within C
50
± 15%.
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org02

2.3 Detection capability
We evaluated the assay detection capability to HIV antibodies
and p24 antigen with the reference panels of the China National
Institutes for Food and Drug Control (NIFDC). The panel for HIV-
1 p24 antigen (Lot 220015-201906) was composed of 20 negative
controls, 10 positive controls, and 10 levels of serial dilutions
for study of the limit of detection (LoD). The panel for HIV
antibodies (Lot 370045-201901) contained different types of
negative and positive controls for the detection of HIV-1/HIV-2
subtypes and controls for the LoD study of B/B′, BC, and AE
genotypes. Furthermore, we prepared a series of doubling dilutions
FIGURE 1
HIVAg/Ab combo assays on LiCA and Architectusing patient serum samples.
FIGURE 2
The S/Co of LiCA HIV Ag/Ab with P24-Ag samples and anti-HlV samples. (A) S/Co of LiCA HIV Ag/Ab with P24-Ag samples. (B) S/Co of LiCA HIV
Ag/Ab with anti-HlV samples.
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org03

to quantify the assay LoD to HIV-1 p24 antigen using the NIFDC
reference panel (Lot 220015-201906, baseline p24 concentration 20
IU/mL) and the World Health Organization (WHO) international
standard from the National Institute for Biological Standards and
Control (NIBSC, code 90/636, baseline p24 concentration 1,000
IU/mL).
2.4 Detection of seroconversion panels
and HIV subtypes
We used 74 banked clinical samples with identified HIV
subtypes, 79 HIV-positive patient sera with a low level of S/Co in
1.0–35.0, and 13 commercial seroconversion panels for comparative
detection between LiCA®and Architect®HIV Ag/Ab assays.
Seroconversion panels were purchased from BioMex (n= 2, SCP-
HIV 005–006, Heidelberg, Germany), ZeptoMetrix (n= 6, PIHIV
9011–9077, Franklin, MA, USA), and SeraCare (n= 5, PRB 953–
977, Milford, MA, USA).
2.5 Potential interference and
cross-reactivity
To evaluate potential interferences from bilirubin, triglycerides,
hemoglobin, and biotin, two patient sera (baseline S/Co 0.72 and
5.68) were used as the diluent to prepare a pool of samples with a
serial concentration of each interferent, respectively. The cross-
reactivity study was performed using 169 serum specimens free of
HIV but positive for potential interferents, such as auto-antibodies
and other viral infection. All samples were tested in duplicate. A
significant interference was considered when the recovery change of
mean S/Co was ≥15% in the sample with a baseline S/Co >0.8 or a
reactive result was recorded in the sample with a baseline S/Co <0.8.
2.6 Statistics
Statistical analyses were conducted using MedCalc (MedCalc
Software, Mariakerke, Belgium) and Excel (Microsoft, WA, USA).
Agreement between LiCA®and Architect®was analyzed based on
the screening-reactive (S/Co ≥1.0) or -nonreactive (S/Co <1.0) assay
results. Specifically, using the HIV Ag/Ab test kit, HIV-negative
samples, HIV p24 antigen-positive samples, and HIV antibody-
positive samples were tested. Through the ROC curve, with the
maximum Youden index as the criterion, the optimal cutoff signal
values corresponding to HIV p24 antigen and HIV antibody were
obtained, respectively. The ratio of the sample detection signal value
to the cutoff signal value was defined as S/Co. Since the signal value
corresponding to CO represents the optimal threshold for
distinguishing between negative and positive results, the S/Co
value of 1 was used as the cutoff for this distinction.
Diagnostic performance parameters, such as sensitivity, specificity,
negative predictive value (NPV), and positive predictive value
(PPV) were determined with confirmatory results by WB and RNA
tests. The t-test was used to evaluate the significant difference
between paired samples. p-value <0.05 was considered as
statistically significant.
3 Results
3.1 Precision analysis
A precision study on S/Co ratios presented that the assay CVs
for repeatability and within-lab imprecision were 2.49%–5.11% on
patient sera, 2.93%–5.05% on p24 antigen controls, and 3.73%–
6.56% on anti-HIV controls (Table 1). The C
50
target and the C
5
–
C
95
interval away from C
50
were 0.99 S/Co and −10.20%–7.67% for
p24 antigen and 1.01 S/Co and −8.79%–5.64% for HIV antibodies,
respectively (Supplementary Figure 1).
3.2 Detection capability
Using China national reference control panels (Table 2), LiCA®
presented nonreactive results in all negative controls for both HIV-
1 p24 antigen and HIV antibodies. The mean S/Co ratios with 95%
confidence interval (95% CI) were 0.33 (0.31–0.34) for p24 and 0.23
(0.19–0.27) for antibodies, respectively. All positive controls for p24
TABLE 1 Precision study for the LiCA®HIV Ag/Ab assay.
Sample Mean Repeatability Within-lab imprecision
(S/Co
a
)SD
a
%CV
a
SD % CV
Serum 1 2.63 0.07 2.49 0.10 3.95
Serum 2 11.55 0.37 3.19 0.53 4.60
Serum 3 37.61 1.18 3.15 1.92 5.11
p24 antigen QC1 1.56 0.05 3.52 0.08 5.05
p24 antigen QC2 3.20 0.09 2.93 0.14 4.47
Anti-HIV QC1 1.14 0.04 3.73 0.07 6.56
Anti-HIV QC2 2.38 0.09 3.88 0.15 6.46
a
S/Co, signal-to-cutoff ratio; SD, standard deviation; CV, coefficient of variation.
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org04

were detected to be reactive with a mean S/Co of 31.45 (19.03–
43.89). One positive control for the antibody of HIV-1 group O (n=
3) was undetected (S/Co = 0.32). All other positive controls for HIV
antibodies were measured with reactive S/Co ratios at a mean of
55.30 (22.84–87.76). The LoD for p24 antigen was identified to be
≤1.25 IU/mL. All positive controls for LoDs to B/B′, BC, and AE
genotypes were recorded with reactive S/Co values at a mean of
11.66 (4.80–18.53).
To further clarify the assay LoD to p24 antigen, linear
regression analyses were performed between the low range (0–10
IU/mL) of p24 antigen concentrations (Y) and assay S/Co values
(X). For the NIFDC reference material, the regression equation was
Y= 1.091X−0.365 (R= 0.999) and LoD was calculated to be 0.73
IU/mL. For the WHO international standard, the equation was Y=
2.908X−2.008 (R= 0.999) and LoD was assessed to be 0.90 IU/mL
(Supplementary Table 1).
3.3 Detection of seroconversion panels
and HIV subtypes
Thirteen seroconversion panels were measured to evaluate early
detection of HIV (Table 3). Among them, LiCA®presented 2/13
with earlier, 1/13 with later, and 10/13 with equal detections in
comparison to Architect®. In general, LiCA®detected the panels
with an average of 5.73 (95% CI, 3.42–8.04) days at the first reactive
result since the RNA-positive detection, which had a mean of −1.27
(−3.75 to 1.21) days earlier than Architect®. The relative sensitivity
coefficient was −0.08 (−0.38 to 0.22).
Detection of various HIV subtypes in clinical patient sera was
evaluated with 64 specimens containing HIV antibodies to different
types of HIV subtypes, 10 p24 antigen single positive samples, and 79
weak positive cases that were collected from HIV-confirmed patients
and measured with low reactive S/Co values (1.0–35.0) by the
Architect®HIV Ag/Ab combo assay. Both LiCA®and Architect®
successfully recorded reactive results for all subjects studied (Table 4).
3.4 Cross-reactivity and interference
The assay recovery changes were determined to be −4.01%–
4.76%, −4.91%–4.74%, −2.62%–4.79%, and −2.15%–6.98% on
LiCA®by spiking potential interferents (up to 342.08 µmol/L
bilirubin, 33.90 mmol/L triglycerides, 5.00 g/L hemoglobin, and
102.25 nmol/L biotin) into both low and high levels of specimens
(baseline S/Co 0.72 and 5.68), respectively. In all 169 samples free of
HIV but positive for potential interfering factors such as auto-
antibodies, other viral infections, and multiple pregnancies, no test-
reactive results were observed on both LiCA®and Architect®
(Supplementary Table 2).
3.5 Comparison of diagnostic performance
between LiCA®and Architect®
Among 21,042 clinical patient sera recruited, 283 (1.34%) were
confirmed to be HIV-positive and 20,759 (98.66%) were HIV-
negative (Figure 2). Compared to Architect®, LiCA®presented a
TABLE 2 Detection of the national reference panels for the LiCA®HIV Ag/Ab assay.
Sample types Sample
no. nAcceptable criteria Test results S/Co
a
mean
(95% CI
a
)
National reference panel for HIV-1 p24 antigen
Negative control N1–N20 20 Nonreactive results = 20/20 Nonreactive results = 20/20 0.33 (0.31–0.34)
Positive control P1–P10 10 Reactive results = 10/10 Reactive results = 10/10 31.45 (19.03–43.89)
LoD
a
control L1–L10 10 L10 = nonreactive, and LoD ≤2.50
IU/mL
L10 = nonreactive, and LoD ≤1.25
IU/mL
Not applicable
National reference panel for HIV antibodies
Negative control N1–N13 13 Nonreactive results = 13/13 Nonreactive results = 13/13 0.23 (0.19–0.27)
Positive control for HIV-1 group
M subtype P1–P14 14 Reactive results = 14/14 Reactive results = 14/14 72.45 (31.35–113.55)
Positive control for HIV-1 group
O subtype P15–P17 3 Reactive results ≥1/3 Reactive results = 2/3 4.40 (0.32–11.48)
Positive control for HIV-2 subtype P18–P20 3 Reactive results ≥2/3 Reactive results = 3/3 7.83 (2.07–13.03)
LoD control for B/B′genotype BB1–BB5 5 Reactive results ≥3/5 Reactive results = 5/5 19.02 (5.17–35.61)
LoD control for BC
recombinant genotype BC1–BC5 5 Reactive results ≥3/5 Reactive results = 5/5 10.13 (2.63–19.75)
LoD control for AE
recombinant genotype AE1–AE5 5 Reactive results ≥3/5 Reactive results = 5/5 5.84 (1.78–10.95)
a
S/Co, signal-to-cutoff ratio; 95% CI, 95% confidence interval; LoD, limit of detection.
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org05

slightly better but not significantly different (p> 0.05) performance
in sensitivity (100.00% vs. 99.65%), specificity (99.85% vs. 99.81%),
NPV (100.00% vs. 99.99%), PPV (89.84% vs. 87.85%), and overall
accuracy (99.85% vs. 99.81%) for the diagnosis of HIV
infection (Table 5).
With further analysis of the segmented S/Co values (Table 6),
we found that LiCA®detected true-positive results with a portion of
10.34% (n= 29), 60.00% (n= 10), 88.89% (n= 9), 94.74% (n= 19)
and 100% (n= 248), and 89.84% (n= 315) in an S/Co range of 1.00–
4.99, 5.00–9.99, 10.00–29.99, 50.00–99.99 and ≥100.00, and in
overall reactive S/Co ratios (≥1.00), respectively. In contrast, the
corresponding true-positive detection portions on Architect®were
7.41% (n= 32), 44.44% (n= 9), 89.47% (n= 19), 94.29% (n= 35)
and 100% (n= 226), and 87.85% (n= 321), respectively. One case,
TABLE 3 Detection of seroconversion panels.
Panel no.
(n= 13)
Sample size Numbers of nonreactive bleeds Days at the first reactive result since
RNA (+)
LiCA®Architect®LiCA®
vs. Architect®
LiCA®Architect®LiCA®
vs. Architect®
SCP-HIV-005 25 4 4 0 / / /
SCP-HIV-006 17 2 2 0 / / /
PIHIV9011 11 8 9 −108−8
PIHIV9016 10 8 8 0 3 3 0
PIHIV9020 22 19 19 0 7 7 0
PIHIV9021 17 14 13 1 7 4 3
PIHIV9031 19 15 16 −1615−9
PIHIV9077 24 11 11 0 4 4 0
PRB953 4 2 2 0 7 7 0
PRB955 5 2 2 0 7 7 0
PRB971 4 2 2 0 7 7 0
PRB974 4 2 2 0 2 2 0
PRB977 4 2 2 0 13 13 0
Total 166 91 92 −16377−14
Mean 12.77 7.00 7.08 −0.08 5.73 7.00 −1.27
95% CI
a
7.79–17.74 3.37–10.63 3.41–10.74 −0.38–0.22 3.42–8.04 4.31–9.68 −3.75–1.21
a
95% CI, 95% confidence interval.
TABLE 4 Detection of different types of HIV-positive serum samples.
Sample types Sample
size
Reactive samples on LiCA®Reactive samples on Architect®
nS/Co
a
mean (95% CI
a
) n S/Co mean (95% CI)
B/B′genotype 10 10 321.62 (213.29–429.95) 10 547.62 (275.39–819.85)
BC recombinant genotype 13 13 457.98 (304.51–611.46) 13 411.28 (251.01–571.56)
AE recombinant genotype 38 38 372.83 (326.32–419.34) 38 269.91 (203.06–336.76)
HIV-2 subtype 2 2 156.30 (106.44–206.24) 2 535.45 (16.34–1,054.62)
HIV-1 group O subtype 1 1 20.09 (20.09–20.09) 1 6.70 (6.70–6.70)
HIV-1 p24 antigen single positive 10 10 21.81 (13.68–29.94) 10 30.85 (14.87–46.83)
Weak positive* 79 79 72.19 (60.90–85.49) 79 13.97 (10.05–17.92)
Total 153 153 189.23 (158.52–219.95) 153 153.45 (113.63–193.26)
a
S/Co, signal-to-cutoff ratio; 95% CI, 95% confidence interval.
*The weak positive specimens were collected from HIV-positive patients and measured with a low reactive signal-to-cutoff ratio between 1.0 and 35.0 by Architect®HIV Ag/Ab combo.
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org06

which was confirmed to be HIV-positive, was misdetected on
Architect®(S/Co = 0.59) but strongly reactive on LiCA®(S/Co
= 64.14).
3.6 Agreement between LiCA®
and Architect®
Paired comparisons demonstrated that the overall agreement
between LiCA®and Architect®was99.67%(95%CI,99.58–99.74%,
n= 21,042) and the Cohen’s kappa was 0.89 (0.86–0.91). Agreements
in nonreactive and reactive assays were 99.85% (99.78%–99.90%, n=
20,721) and 88.16% (84.11%–91.49%, n= 321), respectively (Table 6).
The S/Co segmentation analysis for reactive results revealed that
more detailed agreements in a range of 1.00–4.99, 5.00–9.99, 10.00–
29.99, 50.00–99.99, and ≥100.00 were 9.38% (n= 32), 44.44% (n=9),
89.47% (n= 19), 94.29% (n= 35), and 100% (n= 226), respectively.
There were 70 (0.33%, n= 21,042) discrepant results between these
two assays (Table 7). Architect®contributed 38 (54.29%) false-
positive subjects and 1 (1.43%)false-negativesubject.The
remaining 31 (44.28%) cases had false-positive reactivity on LiCA®.
Among these 70 discrepancies, 63 (90.00%) were identified as having
false-positive reactivity in a low range of S/Co values (<10.00) either
from Architect®(34/63, 53.97%) or from LiCA®(29/63, 46.03%).
4 Discussion
The screening test to HIV Ag/Ab is the initial step for the
diagnosis of HIV infection (Centers for Disease Control and
Prevention, 2018;Branson, 2019). The viral genetic diversity and
geographic distribution change of the variants remain a great
challenge to the detection of early HIV infection (Gaudy et al.,
2004;Pyne et al., 2013;Xiao et al., 2017). Therefore, higher
sensitivity and higher detection capability to various HIV subtypes
are essential to the HIV screening assay. The current study included
HIV-positive cohorts, with HIV-1 p24 antigen, group O and M,
genotypes B/B′, BC, and AE, HIV-2, and low S/Co reactivity (1.0–
35.0), to evaluate the sensitivityand detection capability of the LiCA®
HIV Ag/Ab assay. LiCA®successfully detected all subtypes of HIV
and weak positive samples were recruited. Using the national
reference material and WHO international standard for p24
antigen, the LoD of LiCA®was estimated to be 0.73 and 0.90 IU/
mL, respectively. These data are favorably comparable to other
counterpart assays such as Architect®HIV Ag/Ab combo (LoD =
0.94–1.03 IU/mL), Elecsys®HIV combi PT (LoD = 1.05–1.10 IU/
mL), and Centaur®HIV Ag/Ab combo (LoD = 1.89–1.90 IU/mL)
(Ly et al., 2012;Muhlbacher et al., 2013).
Furthermore, the high sensitivity of LiCA®is explained with
excellent detection of seroconversion panels in comparison to
Architect®. In 13 panels tested on both assays, LiCA®presented
an average of 5.73 (3.42–8.04) days at the first reactive detection
after positive RNA and detected more positive samples with a of
mean −1.27 (−3.75–1.21) days earlier than Architect®.
Paired comparisons in 21,042 clinical patient samples revealed
that LiCA®detected HIV Ag/Ab with a slightly better performance
in sensitivity (100.00% vs. 99.65%), specificity (99.85% vs. 99.81%),
NPV (100.00% vs. 99.99%), and PPV (89.84% vs. 87.85%) than
Architect®. Excellent assay concordance was observed in nonreactive
results (99.85%) but decreased agreement occurred in reactive
measurements (88.16%). Most discrepancies (90.00%) primarily
resulted from false-positive assays in a low level of S/Co reactivity
(<10.00). Previous studies have demonstrated that the Architect®
TABLE 6 Comparisons between LiCA®and Architect®HIV Ag/Ab assays in patient serum samples (n= 21,042).
S/Co
a
segmentation
LiCA®Architect®LiCA®vs. Architect®
No. (%)
of samples
False-negative
or
true-positive
True-negative
or
false-positive
No. (%)
of samples
False-negative
or
true-positive
True-negative
or
false-positive
Agreement
(95% CI
a
)
<1.00 (nonreactive) 20,727 (98.50%) 0 (0.00%) 20,727 (100.00%) 20,721 (98.47%) 1 (0.01%) 20,720 (99.99%) 99.85% (99.78–99.90%)
≥1.00 (reactive) 315 (1.50%) 283 (89.84%) 32 (10.16%) 321 (1.53%) 282 (87.85%) 39 (12.15%) 88.16% (84.11–91.49%)
1.00–4.99 29 (0.14%) 3 (10.34%) 26 (89.66%) 32 (0.15%) 2 (7.41%) 25 (92.59%) 9.38% (1.98–25.02%)
5.00–9.99 10 (0.05%) 6 (60.00%) 4 (40.00%) 9 (0.04%) 4 (44.44%) 5 (55.56%) 44.44% (13.70–78.80%)
10.00–29.99 9 (0.04%) 8 (88.89%) 1 (11.11%) 19 (0.09%) 17 (89.47%) 2 (10.53%) 89.47% (66.86–98.70%)
50.00–99.99 19 (0.09%) 18 (94.74%) 1 (5.26%) 35 (0.17%) 33 (94.29%) 2 (5.71%) 94.29% (80.84–99.30%)
≥100.00 248 (1.18%) 248 (100.00%) 0 (0.00%) 226 (1.07%) 226 (100.00%) 0 (0.00%) 100.00%
(98.38–100.00%)
Total 21,042 (100.00%) 283 (1.34%) 20,759 (98.66%) 21,042 (100.00%) 283 (1.34%) 20,759 (98.66%) 99.67% (99.58–99.74%)
a
S/Co, signal-to-cutoff; 95% CI, 95% confidence interval.
TABLE 5 Assay performance in patient serum samples (95%
confidence interval).
n= 21,042 LiCA®Architect®
Sensitivity, % 100.00% (98.71–100.00%) 99.65% (98.05–99.99%)
Specificity, % 99.85% (99.78–99.90%) 99.81% (99.74–99.87%)
Negative predictive
value, %
100.00% (98.71–100.00%) 99.99% (99.96–99.99%)
Positive predictive
value, %
89.84% (86.22–92.59%) 87.85% (84.09–90.82%)
Accuracy, % 99.85% (99.79–99.90%) 99.81% (99.74–99.86%)
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org07

HIV Ag/Ab combo assay yielded a high rate of false-positive results,
especially in S/Co values <30.00 (Alonso et al., 2018;Wang et al.,
2019). The false reactivity can be generated due to non-specific
binding to the immune complex in the Ag–Ab combination assay
(Mahajan et al., 2010). The testing discrepancies in the same cohort
most likely result from the different Ag/Ab configuration in different
assays (Stickle et al., 2002). Notably, there was one subject that was
misdetected on Architect®(S/Co = 0.59) but strongly reactive on
LiCA®(S/Co = 64.14). This case was from an AIDS inpatient who
was at the late stage of ART during our study. The missed detection of
Architect®could be explained by the less sensitivity to certain
subtypes of HIV (Ly et al., 2012)ortheinfluence on the assay due
to viral mutation after treatment (Zuo et al., 2020).
It has been reported that false-positive HIV Ag/Ab screening tests
can be caused by other viral infections, autoimmune diseases, and
multiple pregnancies (Mahajan et al., 2010;Liu et al., 2016;Adhikari
et al., 2018). Our study indicated that no significant cross-reactivity or
interference was observed from any of 15 potential interference factors
assessed for the LiCA®assay, including auto-antibodies, viral infections
such as Epstein–Barr virus and hepatitis viruses, multiple pregnancies
such as different stages of normal pregnancy and co-infection
pregnancies, and various endogenous interferents. The essence of
high specificity and high sensitivity can be attributed to the unique
methodology and light-initiated multi-amplification signaling
mechanism for the LiCA®assay (Li et al., 2023;Yu et al., 2023).
Combining HIV-1 p24 antigen with antibodies classifies the HIV
screening test from the third-generation to the fourth-generation
method, which enables the assay to achieve higher sensitivity,
reducing the test-negative window to 8–14 days from approximately
3weeks(Alexander, 2016;Qiu et al., 2017). However, the fourth-
generation assay integrates the Ag/Ab reactivity together and can only
report a single result in combination of the Ag/Ab S/Co values. In
contrast, LiCA®performs immunoassays for detecting HIV p24
antigen and antibodies in two independent cuvettes and separates
the S/Co value of p24 antigen from antibodies. Differentiation of the
Ag/Ab reactivity can easily identify the preclinical infectious patient
with the single-positive p24 antigen (Salmona et al., 2014)andthus
facilitate the subsequent confirmatory process for early detection of
HIV infection (Muhlbacher et al., 2019;Yang et al., 2022).
In this study, most of the HIV-positive samples were collected
from the inpatients with AIDS and the outpatients with highly
suspicious history. The positive detection rate was 1.34% (n=
21,042). This situation is different from the clinical conditions for
screening populations and blood donors, in which the viral
prevalence rate can be quite lower (Wang et al., 2019).
Another investigation is valuable for further characterization of the
assay performance in low prevalence of HIV.
5 Conclusion
LiCA®provides a precise and fully automatic platform for
measuring HIV-1 p24 antigen and HIV-1/2 antibodies with high
sensitivity and specificity. The assay performance is favorably
comparable to the well-established Architect®HIV Ag/Ab combo
assay in analytical and clinical perspectives. Additionally, LiCA®
HIV Ag/Ab can differentiate the reactivity of p24 antigen from
antibodies in a separate S/Co result. It is appropriate for use in the
clinical routine test for the early detection of HIV.
Data availability statement
The original contributions presented in the study are included
in the article/Supplementary Material. Further inquiries can be
directed to the corresponding authors.
Ethics statement
Research involving human subjects complied with all relevant
national regulation, institutional policies and is in accordance with
the tenets of the Helsinki Declaration (as revised in 2013), and has
been approved by Dongzhimen Hospital, Beijing University of
Chinese Medicine (No. 2024DZMEC-363-01).
Author contributions
XY: Writing –review & editing. YW: Writing –review &
editing. YJL: Writing –original draft, Writing –review & editing.
FJ: Writing –original draft, Writing –review & editing. YHL:
Writing –original draft, Writing –review & editing. YL: Writing –
review & editing.
Funding
The author(s) declare that no financial support was received for
the research, authorship, and/or publication of this article.
TABLE 7 Analysis of discrepant assays between LiCA®and Architect®HIV Ag/Ab combo in patient serum samples (n= 70).
S/Co
segmentation
No. (%)*
of samples
LiCA®
false-positive
LiCA®
false-negative
Architect®
false-positive
Architect®
false-negative
1.00–4.99 54 (77.14%) 25 (46.30%) 0 (0.00%) 29 (53.70%) 0 (0.00%)
5.00–9.99 9 (12.86%) 4 (44.44%) 0 (0.00%) 5 (55.56%) 0 (0.00%)
≥10.00 7 (10.00%) 2 (28.57%) 0 (0.00%) 4 (57.14%) 1 (14.29%)
Total 70 (100.00%) 31 (44.28%) 0 (0.00%) 38 (54.29%) 1 (1.43%)
*The sample amount included subjects with a candidate segment of reactive signal-to-cutoff (S/Co) ratios either on LiCA®or on Architect®among the discrepant cohort.
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org08

Conflict of interest
The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be
construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online
at: https://www.frontiersin.org/articles/10.3389/fcimb.2025.1474127/
full#supplementary-material
References
Adhikari, E. H., Macias, D., Gaffney, D., White, S., Rogers, V. L., McIntire, D. D., et al.
(2018). Diagnostic accuracy of fourth-generation ARCHITECT HIV Ag/Ab Combo
assay and utility of signal-to-cutoff ratio to predict false-positive HIV tests in
pregnancy. Am. J. Obstet Gynecol 219, 408.e1–408.e9. doi: 10.1016/j.ajog.2018.06.008
Alexander, T. S. (2016). Human immunodeficiency virus diagnostic testing: 30 years
of evolution. Clin. Vaccine Immunol. 23, 249–253. doi: 10.1128/CVI.00053-16
Alonso, R., Perez-Garcia, F., Gijon, P., Collazos, A., and Bouza, E. (2018). Evaluation
of the Architect HIV Ag/Ab Combo Assay in a low-prevalence setting: The role of
samples with a low S/CO ratio. J. Clin. Virol. 103, 43–47. doi: 10.1016/j.jcv.2018.04.002
Bian, Y., Liu, C., She, T., Wang, M., Yan, J., Wei, D., et al. (2018). Development of a
light-initiated chemiluminescent assay for the quantitation of sIgE against egg white
allergens based on component-resolved diagnosis. Anal. Bioanal Chem. 410, 1501–
1510. doi: 10.1007/s00216-017-0791-y
Branson, B. M. (2019). HIV diagnostics: current recommendations and opportunities
for improvement. Infect. Dis. Clin. North Am. 33, 611–628. doi: 10.1016/j.idc.2019.04.001
Centers for Disease Control and Prevention (2018). 2018 Quick reference guide:
recommended laboratory HIV testing algorithm for serum or plasma specimens.
Available online at: https://stacks.cdc.gov/view/cdc/50872 (Accessed October 17, 2018).
CLSI (2008). “User protocol for evaluation of qualitative test performance; approved
guideline-second edition,”in CLSI document EP12-A2 (Clinical and Laboratory
Standards Institute, Wayne, PA).
CLSI (2014). “User verification of precision and estimation of bias; approved guidline-third
edition,”in CLSI document EP15-A3 (Clinical and Laboratory Standards Institute, Wayne, PA).
Cohen, M. S., Chen, Y. Q., McCauley, M., Gamble, T., Hosseinipour, M. C.,
Kumarasamy, N., et al. (2011). Prevention of HIV-1 infection with early
antiretroviral therapy. N Engl. J. Med. 365, 493–505. doi: 10.1056/NEJMoa1105243
El-Sadr,W.M.,Mayer,K.H.,Rabkin,M.,andHodder,S.L.(2019).AIDSinAmerica-back
in the headlines at long last. NEngl.J.Med.380, 1985–1987. doi: 10.1056/NEJMp1904113
Fitzgerald, N., Cross, M., O’Shea, S., and Fox, J. (2017). Diagnosing acute HlV
infection at point of care:a retrospective analysis of the sensitivity andspecificity of a
fourth-generation point-of-care testfor detection of HlV core protein p24. Sexually
Transmitted Infections 93, 100–101. doi: 10.1136/sextrans-2015-052491
Gaudy, C., Moreau, A., Brunet, S., Descamps, J. M., Deleplanque, P., Brand, D., et al.
(2004). Subtype B human immunodeficiency virus (HIV) type 1 mutant that escapes
detection in a fourth-generation immunoassay for HIV infection. J. Clin. Microbiol. 42,
2847–2849. doi: 10.1128/JCM.42.6.2847-2849.2004
Gray, E. R., Bain, R., Varsaneux, O., Peeling, R. W., Stevens, M. M., and McKendry,
R. A. (2018). p24 revisited: a landscape review of antigen detection for early HIV
diagnosis. AIDS 32, 2089–2102. doi: 10.1097/QAD.0000000000001982
Li, H., Yang, S., Cao, D., Wang, Q., Zhang, S., Zhou, Y., et al. (2023). A new double-
antigen sandwich test based on the light-initiated chemiluminescent assay for detecting
anti-hepatitis C virus antibodies with high sensitivity and specificity. Front. Cell Infect.
Microbiol. 13, 1222778. doi: 10.3389/fcimb.2023.1222778
Li, Z., Yang, S., Qiao, J., Tan, Y., Liu, Q., Yang, B., et al. (2024). Performance
evaluation of a novel high-sensitivity cardiac troponin T assay: analytical and clinical
perspectives. Clin. Chem. Lab. Med. 62, 979–987. doi: 10.1515/cclm-2023-0789
Liu, P., Jackson, P., Shaw, N., and Heysell, S. (2016). Spectrum of false positivity for
the fourth generation human immunodeficiency virus diagnostic tests. AIDS Res. Ther.
13, 1. doi: 10.1186/s12981-015-0086-3
Ly, T. D., Plantier, J. C., Leballais, L., Gonzalo, S., Lemee, V., and Laperche, S. (2012).
The variable sensitivity of HIV Ag/Ab combination assays in the detection of p24Ag
according to genotype could compromise the diagnosis of early HIV infection. J. Clin.
Virol. 55, 121–127. doi: 10.1016/j.jcv.2012.06.012
Mahajan, V. S., Pace, C. A., and Jarolim, P. (2010). Interpretation of HIV serologic
testing results. Clin. Chem. 56, 1523–1526. doi: 10.1373/clinchem.2009.139535
Muhlbacher, A., Sauleda, S., Piron, M., Rietz, R., Permpikul, P., Klinkicht, M., et al.
(2019). A multicentre evaluation of the Elecsys HIV Duo assay. J. Clin. Virol. 112, 45–50.
doi: 10.1016/j.jcv.2018.11.005
Muhlbacher, A., Schennach, H., van Helden, J., Hebell, T., Pantaleo, G., Burgisser, P.,
et al. (2013). Performance evaluation of a new fourth-generation HIV combination
antigen-antibody assay. Med. Microbiol. Immunol. 202, 77–86.
Pyne, M. T., Hackett, J. Jr., Holzmayer, V., and Hillyard, D. R. (2013). Large-scale
analysis of the prevalence and geographic distribution of HIV-1 non-B variants in the
United States. J. Clin. Microbiol. 51, 2662–2669. doi: 10.1128/JCM.00880-13
Qiu,X.,Sokoll,L.,Yip,P.,Elliott,D.J.,Dua,R.,Mohr,P.,etal.(2017).Comparative
evaluation of three FDA-approved HIV Ag/Ab combination tests using a genetically diverse
HIV panel and diagnostic specimens. J. Clin. Virol. 92, 62–68. doi: 10.1016/j.jcv.2017.05.005
Rodger, A. J., Cambiano, V., Bruun, T., Vernazza, P., Collins, S., Degen, O., et al.
(2019). Risk of HIV transmission through condomless sex in serodifferent gay couples
with the HIV-positive partner taking suppressive antiretroviral therapy (PARTNER):
final results of a multicentre, prospective, observational study. Lancet 393, 2428–2438.
doi: 10.1016/S0140-6736(19)30418-0
Salmona, M., Delarue, S., Delaugerre, C., Simon, F., and Maylin, S. (2014). Clinical
evaluation of BioPlex 2200 HIV Ag-Ab, an automated screening method providing
discrete detection of HIV-1 p24 antigen, HIV-1 antibody, and HIV-2 antibody. J. Clin.
Microbiol. 52, 103–107. doi: 10.1128/JCM.02460-13
Stickle, D. F., Pirruccello, S. J., Swindells, S., and Hinrichs, S. H. (2002). Discrepant
results of 2 screening tests for anti-HIV antibody. Clin. Infect. Dis. 35, 773–4;author
reply 774-5. doi: 10.1086/cid.2002.35.issue-6
UNAIDS (2019). 2018 Global HIV & AIDS statistics-fact sheet. Available online at:
https://www.unaids.org/en/resources/fact-sheet (Accessed Febuary 13, 2019).
Volberding, P. A., and Deeks, S. G. (2010). Antiretroviral therapy and management
of HIV infection. Lancet 376, 49–62. doi: 10.1016/S0140-6736(10)60676-9
Wang, M., Li, J., Huang, Y., Chen, T., Dong, S., Zhang, R., et al. (2022). Analytical
validation of the LiCA high-sensitivity human thyroid stimulating hormone assay. Clin.
Biochem. 101, 42–49. doi: 10.1016/j.clinbiochem.2021.11.018
Wang, L., Xiao, Y., Tian, X. D., Ruan, J. X., Chen, W., and Yu, Y. (2019). HIV infection
in Xi’an, China: epidemic characterization, risk factors to false positives and potential
utility of the sample-to-cutoff index to identify true positives using Architect HIV Ag/Ab
combo. Antimicrob. Resist. Infect. Control 8, 9. doi: 10.1186/s13756-019-0463-0
Xiao, P., Li, J., Fu, G., Zhou, Y., Huan, X., and Yang, H. (2017). Geographic
distribution and temporal trends of HIV-1 subtypes through heterosexual
transmission in China: A systematic review and meta-analysis. Int. J. Environ. Res.
Public Health 14 (7), 830. doi: 10.3390/ijerph14070830
Yang, M., Yang, W., Shi, W., and Tao, C. (2022). Clinical application evaluation of
elecsys((R)) HIV duo assay in southwest China. Front. Cell Infect. Microbiol. 12,
877643. doi: 10.3389/fcimb.2022.877643
Yang, S., Zhang, Q., Yang, B., Li, Z., Sun, W., and Cui, L. (2022). Analytical and
clinical performance evaluation of a new high-sensitivity cardiac troponin I assay. Clin.
Chem. Lab. Med. 60, 1299–1307. doi: 10.1515/cclm-2021-1136
Yu, M., Chen, D., Tang, X., Zhang, Y., Liang, P., Xiong, Y., et al. (2023). Evaluation of a high-
sensitivity SARS-CoV-2 antigen test on the fully automated light-initiated chemiluminescent
immunoassay platform. Clin. Chem. Lab. Med. 61, 1123–1130. doi: 10.1515/cclm-2022-1039
Yu, Y., She, T., Huang, L., Xu, J., Yan, J., Jiang, Q., et al. (2021). Establishment of a
homogeneous immunoassay-light-initiated chemiluminescence assay for detecting
anti-Mullerian hormone in human serum. J. Immunol. Methods 494, 113059.
doi: 10.1016/j.jim.2021.113059
Zuo, L., Liu, K., Liu, H., Hu, Y., Zhang, Z., Qin, J., et al. (2020). Trend of HIV-1 drug
resistance in China: A systematic review and meta-analysis of data accumulated over 17
years (2001-2017). EClinicalMedicine 18, 100238. doi: 10.1016/j.eclinm.2019.100238
Li et al. 10.3389/fcimb.2025.1474127
Frontiers in Cellular and Infection Microbiology frontiersin.org09