Evolution of HIV Diagnostics andand
Goals for the 2010 Conference
Bernard M. Branson, M.D.
Associate Director for Laboratory DiagnosticsAssociate Director for Laboratory Diagnostics
CDC Division of HIV/AIDS Prevention
or other influences – like Creationism?…or other influences like Creationism?
FDAAPHL
C
CLSIIndustry
CLIA CDC
Consumer
CLIA CDC
Medical Practitioners
1987 1992 1987 Vironostika HIV-1 EIA
1992 Fluorognost
IFA
1985 Abbott
HIV 1 EIA
1991 Cambridge
Western blotHIV-1 EIA Western blot
i i Al i h Diagnostic Algorithm: 1989
The Public Health Service recommends that no positive test results be given to clients/patients until a screening test has been repeatedly reactive(i.e., greater than or equal to two tests) on the same specimen and a supplemental more specific testspecimen and a supplemental, more specific testsuch as the Western blot has been used to validate those results validate those results
1987 1994 1992
1987 Vironostika HIV-1 EIA
Vironostika Oral fluid
EIA
Abbott HIV-1/HIV-2
EIA
1985 Abbott
HIV 1 EIA
1992 Murex SUDS
1994 Orasure
Specimen HIV-1 EIA SUDS Specimen Collection
Device
1995 Human Retrovirus Testing Conference:1995 Human Retrovirus Testing Conference:Provide results from rapid tests?
1987 1994 1996 1996 1992
Abb tt 1987 Vironostika
EIA
1994 Vironostika
Oral fluid EIA
Abbott p24 Donor Screen
Calypte HIV-1
Urine EIA
Abbott HIV-1/HIV-2
EIA
1985 Abbott
HIV-1 EIA
1992 Murex SUDS
1994 Orasure
Specimen
1996 Coulter p24
D
1996 HomeAccess HIV 1 Home Collection
DeviceDonor Screen
HIV-1 Home Test System
Alternative specimens Alternative specimens
p24 Antigen testing p24 Antigen testing
Clinical applications Clinical applications of PCR
Home collection Home collection: impact on public healthhealth
HIV-1 viral load HIV-1 viral load assays
Mfg demos: RT PCR RT PCRUrine kitOral fluid deviceHome collection
d i d iNew Recommendation …and a Promise Health-care providers should provide preliminary Health care providers should provide preliminary
positive test results before confirmatory results are available in situations where tested persons benefit.
When additional rapid tests become available for puse in the United States, the PHS will re-evaluate algorithms using specific combinations of two or
id f i d fi i HIV more rapid tests for screening and confirming HIV infection.
Rapid Testing E l ti Evaluation Applications
Challenges to the Testing Algorithm:Testing Algorithm: Repeat sampleOral fluidO u Nucleic Acid
Testing Strategies R id EIA Rapid EIADouble EIAWestern blotWestern blot
New Testing Strategy: New Testing Strategy: detecting early HIV infection for prevention and for estimating incidence
2000 Genetic Systems
1996 1992
Genetic Systems HIV-1/HIV-2 Peptide EIA
1987 Vironostika
EIA
1994 Vironostika
Oral fluid EIA
1996 Abbott p24
Donor Screen
1992 Abbott
HIV-1/HIV-2 EIA
1985 Abbott
HIV-1 EIA
1992 Murex SUDS
1994 Orasure
Specimen
1996 Coulter p24 Donor
SSUDS Specimen Collection
Device
Screen
1999 Roche
Amplicor Amplicor HIV-1
Monitor
Preliminary Preliminary evaluation of avidity index to distinguish index to distinguish incident vs prevalent infection
Evaluation of the Inno-LIA in a testing algorithm for li i ti f th elimination of the
Western blot
2000 2000
COBAS 2002
P l i 2003
GS HIV 1
1992
Genetic Systems HIV-1/HIV-2 Peptide EIA
COBAS Ampliscreen
HIV-1
Procleix HIV-1/HCV
NAT
GS HIV-1 HIV-2 Plus O
EIA
1987 Vironostika
EIA
1992 Abbott
HIV-1/HIV-2 EIA
1985 Abbott
HIV-1 EIA
1992 Murex SUDSSUDS
1999 Roche
Amplicor
2002 OraQuick
HIV 1/HIV 2
2004 Multispot
HIV 1/HIV 2 HIV-1 Monitor
HIV-1/HIV-2 Rapid Test
HIV-1/HIV-2 Rapid Test
Approved package insert:
Intended Use Biological Principles Restrictions Restrictions Warnings Directions for Use I t t ti Interpretation Limitations Performance Characteristics
Ch i Change in Language
“This test is suitable for use in multi-test algorithms designed for statistical validation of rapid HIV test results. When multiple rapid HIV tests are available, this test should be used in appropriate multi test algorithms ”multi-test algorithms.
Comparing new EIAs Comparing new EIAs with old stand-bys
NAAT testing for acute infectionacute infection
Roll-out of rapid Roll out of rapid testing
Options for confirmatory testingy g
2000 2002
Procleix 2003 2006
1992
Genetic Systems HIV-1/HIV-2 Peptide EIA
ProcleixHIV-1/ HCV NAT
GS HIV-1 HIV-2 Plus
O EIA
Procleix Ultrio HIV/HCV/HBV
1987 Vironostika
EIA
1992 Abbott
HIV-1/HIV-2 EIA
2006 Advia
Centaur
1985 Abbott
HIV-1 EIA
1992 Murex SUDS
1/O/2 CIA
SUDS
1999 Roche
Amplicor
2002 OraQuick
HIV-1/HIV-2 id
2004 Multispot
HIV-1/HIV-2 id
2006 Aptima
Qualitative
2008 Ortho Vitros
HIV 1+2 HIV-1 Monitor
Rapid Test Rapid Test Qualitative RNA
HIV 1+2 CIA
Random Access EIAs Random Access EIAs
Acute HIV screening Acute HIV screening
Ch ll ith th Challenges with the current algorithm
Alternatives to the current algorithmcurrent algorithm
Sequence of Test Positivity Relative to WB50 % Positive Cumulative Frequency
natio
n*
PTIM
A
A+O
bC
ombi
n
HIV
-1/2
PA
K
pos
itive
ocle
ix/ A
P
ott r
DN
ARa
d 1/
2+
vlti-S
pot
veal
G2
onos
tka
aQui
ckig
old
Ind
ott A
g/Ab
CO
MPL
ETE
V-1/
2 ST
AT-P
WB
Pro
Abb
Bio- rLav
MulRev
Viro
Ora
Uni
WB
Abb C
HIV
Days before WB positive
25 20 10 5 -2015
Modified from Owen et al J Clin Micro 2008
* = not currently FDA approved
Al i h fi i iAlgorithm: Definition
Overall sensitivity or specificity may be improved by using test combinations under one of two decision rules for resolving discordant results. Assumes that errors in the tests are random and
independent but errors can be systematic as with falseindependent, but errors can be systematic, as with false-negative results in early infection.
Few testing strategies involve only a single test, but the g g y gcosts of “confirmation” can be prohibitive.
i i i f h C Al i hLimitations of the Current Algorithm
Antibody tests do not detect infection in ~ 10% of Antibody tests do not detect infection in ~ 10% of infected persons at highest risk of transmission
Western blot confirmation is less sensitive during early infection than many widely used screening early infection than many widely used screening tests
Delays inherent with centralized screening reduce the “effective sensitivity” because infected persons y pdo not learn their test results
i i i f d Al i hLimitations of Proposed Algorithms
Most look like the traditional algorithm Most look like the traditional algorithm Include Western blot Additional (more expensive) tests if WB is negative Additional (more expensive) tests if WB is negative
Logistics of multiple POC tests can be daunting Cost and space requirements of multiple Cost and space requirements of multiple
laboratory platforms is prohibitive Tests are usually repeated at care site anyway, with ests a e usua y epeated at ca e s te a y ay, t
additional tests (e.g., viral load) for clinical staging and management
Problems with Regulatory ConstraintsProblems with Regulatory Constraints
Not responsive to clinical needs:p Preliminary results from rapid tests but not EIAsUse of rapid tests on potentially exposed infants is
drestricted Not feasible to accumulate sufficient numbers for
required validationsrequired validations
Delays the availability (and improvement) of state-of-the-art technologygy
Cost discourages innovation
Goals for 2010 ConferenceGoals for 2010 Conference
Review the available evidence
Identify the needs of different stakeholdersy
Develop a menu for specific applicationsp p pp
Begin consensus process for updating testing
Agree on a new Algorithm
Begin consensus process for updating testing recommendations