A Clinical Study of Antigen-Antibody Rapid Testing for Acute HIV Infection Christopher D. Pilcher1, Brian Louie2, Sheila Keating3, Mark Pandori2, Falk Fish4, Tomer Keren4, Sally Liska2, Michael P. Busch3, Frederick M. Hecht1, Jeffrey Klausner1,2 and

1
A Clinical Study of Antigen-Antibody Rapid
Testing for Acute HIV Infection Christopher D.
Pilcher1, Brian Louie2, Sheila Keating3, Mark
Pandori2, Falk Fish4, Tomer Keren4, Sally Liska2,
Michael P. Busch3, Frederick M. Hecht1, Jeffrey
Klausner1,2 and Grant Colfax1,2 from 1UCSF
HIV/AIDS Division 2SF Department of Public
Health 3Blood Systems Research Institute
4Inverness Medical Innovations
992
Acute HIV infections, as a proportion of all HIV
cases detected
Results
Methods - Testing Antibody
(Ab)-plus-RNA testing reference standard The
complete algorithm, including Ab and RNA tests,
was used to define HIV infection. The combined
reference standard was used to determine Sensitivi
ty, Specificity and Positive/Negative Predictive
Values for each screening assay. Western blot was
used to categorize samples as representing acute
(WB negative/indeterminate) or established
(WB positive) HIV infection.
Background A key factor influencing HIV test
performance, particularly in higher-risk testing
sites, is the ability to detect acute HIV cases. 
These cases are often negative on tests designed
to detect only anti-HIV IgG 4th generation
immunoassays (e.g., the ARCHITECT HIV Ag/Ab
Combo) can detect both HIV p24 antigen and
anti-HIV antibodies for earlier detection. A new
rapid test (the Determine HIV-1/2 Ag/Ab Combo) is
also designed to detect HIV p24 antigen and
anti-HIV antibodies. We sought to measure the
potential impact of these new screening tests on
the performance of real-world, public health and
research testing programs in San Francisco.
by each testing program.
The of all HIV infected cases
identified that were Western blot
-
negative or indeterminate is shown.
The majority of RNA patients found in high risk
testing populations (high risk MSM, non-targeted
STD clinic VCT, partner services and nPEP) were
detectable by the 4th generation EIA or by the
new rapid Ag/Ab combo assay, but not by 1st
generation or even 3rd generation EIAs.
50
45
40
40
100
1st Gen EIA
90
35
Programs with the highest yield of acute HIV
infections were those targeting patients with
high risk exposures
80
Stat
-
Pak Rapid
Screening HIV
30
-

70
Ab
Test

Genetic Systems HIV
-
1/2
60
25
Acute Infections Positive on Each Test
Plus 0
50
Determine HIV
-
1/2 Ag/
Ab
20
40
Combo Rapid
30
15
ARCHITECT HIV Ag/
Ab
10
Ab
Status
Pooled
RNA
-
20

Combo
7
10
10
6
Confirmation
0
HIV RNA
5
nPEP
WB
Realtime
PCR
0
STD Clinic
nPEP
High Risk MSM
Partner Services
High Risk MSM
STD Clinic
Partner Services
Acute HIV Diagnostic
Testing Population
-
San Francisco
-

Sensitivity was compared head-to-head in the high
risk MSM testing population, The sensitivity
estimates associated with each rapid test (light
blue) and central lab immunoassay (dark blue) are
shown. Despite differences in
clinical sensitivity, negative predictive values
for each test to were gt99.5. Specificity was
estimated to be 100.00 for all immunoassays.
Oraquick Advance specificity was lower on oral
fluid (Sp 99.86) vs. blood (99.96). The
Determine HIV Ag/Ab Combo correctly identified 80
of 81 HIV negative Options specimens, with 1
false positive result (Sp 98.9).

• Summary
• In practice in San Francisco testing programs,
  HIV screening tests differed significantly in
  their ability to detect HIV infections. Assays
  that detected acute HIV substantially increased
  case identification.
• The Determine HIV-1/2 Ag/Ab Combo, a novel point
  of care assay, increased detection of HIV
  infection compared to all existing antibody-only
  rapid tests. Separate signals for HIV-1 p24
  antigen and HIV antibodies make immediate, point
  of care identification of acute cases possible.
  Speed and simplicity make the test appropriate
  for both developed and developing world settings.
• The ARCHITECT HIV Ag/Ab Combo, a 4th
  generation immunoassay, showed even greater
  sensitivity for acute HIV infections. This assay
  could eliminate the need for pooled HIV RNA for
  central laboratory screening of acute HIV
  infections.
• Confirmatory algorithms must be modified (e.g.,
  to incorporate HIV RNA), as these screening tests
  are more sensitive than established Western blot
  or immuno-fluorescence assays, and have imperfect
  specificity.
• The Oraquick oral fluid test device (with an
  estimated sensitivity of 86) may not be
  appropriate for testing of high risk individuals
  in San Francisco.

HIV
Ab
()
HIV
Ab
(
-
) RNA (
-
)
HIV
Ab
(
-
) RNA ()
Established HIV
HIV Negative
Acute HIV
Rapid Test Panel
Methods - Specimens Testing programs were
included that used both HIV antibody and HIV
RNA amplification tests in their diagnostic
algorithm High Risk MSM, Targeted Testing
Population (n6661) MSM at public testing sites
meeting high risk criteria (sex with HIV,
unprotected anal intercourse, or STD) STD Clinic
Population (n14573) attendees to the muncipal
STD clinic, not meeting above criteria Non-occupa
tional Post Exposure Prophylaxis (nPEP)
Population (n989) patients evaluated for, or
receiving, nPEP Partner Services (PS) Testing
Population (n173) sexual or needlesharing
partners to newly diagnosed HIV Acute HIV
Diagnostic Testing Population (n1114
1998-2008) patients screened for having acute
HIV by the UCSF Options study based on
acute retroviral symptoms and/or high risk
exposure.
Central Lab Test Panel
Oraquick
Advance
3
rd
Gen EIA (
Gen Systems
HIV
½
Plus O)
Clearview
Stat
-
Pak
4
th
Gen IA (Abbott ARCHITECT HIV Ag/
Ab
Combo)
Unigold
Recombigen
Western Blot (
Biorad
Genetic Systems)
Multispot
HIV 1/2
Determine HIV
1/2 Ag/
Ab
Combo
Head to head comparative performance evaluation
Acute specimens (RNA, WB- or indeterminate)
were submitted to Rapid Test and Central Lab
panel tests. Sensitivity and Negative Predictive
Values (NPV) were calculated for each panel test,
assuming 100 detection of WB positive specimens.
Determine HIV-1/2 Ag/Ab Combo Rapid Test
Readout
Acknowledgements The work was supported by R01-
MH068686 we are grateful to Shelley Facente for
analysis and to John Hackett, Jr. for providing
4th generation IA data, previously submitted for
publication (Pandori MW, Hackett Jr., Louie B, et
al., as Assessment of the Ability of a Fourth
Generation Immunoassay for HIV Antibody and p24
Antigen to Detect both Acute and Recent HIV
Infection in a High-Risk Setting.)