HIV Testing

in the Era of PrEP:

When the Tests are Discordant

Raphael J. Landovitz, MD MSc

Associate Professor of Medicine
UCLA Center for Clinical AIDS Research & Education
HPTN Regional Meeting, Lima Peru, March 26, 2018
Scope of the Challenge

• In HPTN 083, a full HIV testing algorithm is

used at 57 study visits
– Each includes a rapid FDA-cleared HIV EIA and
an instrumented (4th or 5th generation) Ag/Ab
test
– Also an HIV RNA at screening and as clinically
indicated during the study
– 4,500 participants will be enrolled
– ~256,500 HIV testing events
256,500 HIV Testing Events

• False Positive Rate (FPR) = 1-Specificity

• Each Instrumented Ag/Ab platform has
unique performance characteristics
– Reported Specificity of Abbott Architecht Ag/Ab
test is 100% with 95% CI 99.18-100%
– FPR 0 (0-0.82%)
– 0.82% x 256,500 = 2,103 false positive tests
2,103 False Positive Tests
• Likely overestimate, as does not account for
those screened out for “false positive” testing
prior to study entry
• Multiple testing methods at each visit also
provide additional opportunity for false positive
results
• Must be balanced with high probability that
a newly detected reactive/positive test in
high-risk population more likely to
represent real infection
TDF/FTC PrEP Delays Seroconversion

• 25 days vs. 17 days

to Feibig V
• 7-fold odds of >100
day delay in site
detection of
seroconversion
• 0.75 log decrease in
viral load
MARKOWITZ, JAIDS, 2017
HOORNENBORG, LANCET HIV, 2017
We have no information on delay of
Seroconversion process with CAB
• Oral CAB may (or may not) be similar to oral
TDF/FTC
• Limited information available for “time-to-
viremia” during CAB LA decay
– Few failures thus far in CAB LA treatment studies
– 4x PA-IC90 may be reasonable estimate
• CAB LA, if approved, will be deployed in
varying-resources areas, thus a pragmatic
approach must be balanced with caution
Approaches Considered

• Maximally conservative
– Any reactive HIV testing = No further study
products
– Impractical: Impugns primary study outcome
– Not consistent with clinical practice
• Minimally conservative
– Absent RNA viremia, low risk for resistant
quasispecies selection, continue
Consensus Panel Convened
Grace Aldrovandi MD PhD Mina Hosseinipour MD MPH
Jared Baeten MD PhD James P. Hughes PhD
Bernard Branson MD Sinead Delaney-Moretlwe
David Burns MD Daniel R. Kuritzkes
Connie Celum MD MPH David A. Margolis, MD MPH
Robert Coombs, PhD Marybeth McCauley
Wafaa El-Sadr MD MPH Jean-Michel Molina MD PhD
Sue Eshleman MD PhD Deborah Persaud MD PhD
Jennifer Farrior Estelle Piwowar-Manning
Myron S. Cohen MD James Rooney MD
Deborah Donnell, PhD Paul E. Sax MD
Joseph J. Eron MD Scott Rose
Kailazarid Gomez-Feliciano Nirupama Deshamane Sista
Robert Grant MD PhD PhD
Beatriz Grinsztejn MD PhD Sheryl Zwerski RN PhD
 ACKNOWLEDGEMENTS

The HIV Prevention Trials Network is funded by the

National Institute of Allergy and Infectious Diseases
(UM1AI068619, UM1AI068613, UM1AI1068617),
with co-funding from the National Institute of Mental
Health, and the National Institute on Drug Abuse, all
components of the U.S. National Institutes of Health.
 Clinical Scenario #1
Index Sample: Reactive HIV rapid test, Negative instrumented HIV Ag/Ab test

Rationale:
Clinical Management:
This scenarioHold is consistent
study productswith false-positive
when the instrumented
HIV rapid testing.
HIV Ag/Ab
In thetest
absence
result isof
received.
a positive Samples
instrumented will be
Ag/Ab
sent test,
for testing
an indeterminate
using a discriminatory
or positive discriminatory
test (if available,
test,
index
or aand
confirmatory
positive HIV RNAvisits),test,
an itHIV
is very
RNAunlikely
test (index
that and
a reactive
confirmatory
point-of-care
visits) and
HIV an
rapid
HIVtest
DNA test
(confirmatory
represents truevisit only). Additional
infection. The goal of HIV testing
this will also be
management performed
strategy after a 4-week
is to minimize time off product
study-
hold; thisinwill
product include an
a high-risk HIV rapidThe
population. test,risk
a discriminatory
of selecting fortest (if resistant
drug available), an HIV RNA test,
and an
HIV HIV DNA test.
by continuing PrEP Blinded
with a study
singleproduct may be restarted
ARV (cabotegravir) after the 4-week
or a dual-agent regimenproduct hold if
(TDF/FTC)
thevery
is Architect
low. IfAg/Ab test HIV
a positive fromDNA
the confirmatory
test result is visit is ≤10
obtained (or the
from if this value is notvisit
confirmatory reported by the
after study
laboratory),
product and if thestudy
is resumed, HIV rapid tests,
products willthe
bediscriminatory tests, HIV RNA
permanently discontinued andtests,
SOCandARTHIV willDNA
be
tests fromIf the
initiated. HIVindex
rapidvisit,
teststhe
areconfirmatory visit, and the
reactive at subsequent post-product
study hold visit arewho
visits for participants all non-
reactive/negative.
resume study product, please refer to additional considerations below (“Persistently
Positive/Reactive HIV Screening Tests”).
 Clinical Scenario #2
Index Sample: Reactive HIV rapid test, Reactive instrumented HIV Ag/Ab test (as
defined by the manufacturer)

Rationale:
Clinical Management:
Participants Hold
are presumed
study products
to be HIV
afterinfected
the first when
reactive
reactive/positive
HIV rapid testresults
result isare
received. for
obtained When
two the
different
reactive
HIVresult
screening
is obtained
assays,foreven
the instrumented
if discriminatoryAg/Ab
tests,
test,
HIVstudy
RNAproduct
tests,
will be
and thepermanently
HIV DNA test discontinued
are negative.andInSOC
theseART
cases,
will viral
be initiated.
replication
Additional
and anti-
HIV
HIVtesting
antibody
will be
performed at
production thebe
may index and confirmatory
suppressed by PrEP. visits. However,
The guiding results
principle forofthis
thatmanagement
testing will not change
strategy
clinical
is management,
to minimize the timeunless there is clearincident
from presumptive evidence that the
infection toreactive/positive HIV screening
initiation of fully suppressive
test results
ART, were the
to minimize result of a sample/data
establishment of an HIV and error.
maximize the potential for future
elimination/cure of HIV infection.
 Clinical Scenario #3
Index Sample: Reactive HIV rapid test, Reactive instrumented HIV Ag/Ab test (as
defined by the manufacturer)

Rationale:
Clinical Management:
Participants Permanently
are presumeddiscontinue
to be HIV infected
study product
if the Architect
and initiate
HIVSOCAg/Ab
ART,
Combo
regardless
test is reactive
of the
with
results
a S/CO
of other
ratio >10.
testsInfrom
these
thecases,
index and
viralconfirmatory
replication andvisits.
anti-HIV
Additional
antibody
HIV
testing will be
production may performed
be suppressed
at the index
by PrEP.andThe
confirmatory
guiding principle
visits. However,
for this management
results of thatstrategy
testing
will
is tonot changethe
minimize clinical
time management,
from presumptive unless thereinfection
incident is clear evidence that
to initiation of the
fullyreactive/positive
suppressive
Ag/Ab
ART, totest result was
minimize the result of
establishment of aansample/data error. the potential for future
HIV and maximize
elimination/cure of HIV infection.
 Clinical Scenario #4
Index Sample: Negative rapid HIV test, Reactive Architect HIV Ag/Ab test with a
signal-to-cutoff ratio (S/CO) ≤10 or S/CO ratio not reported by the laboratory

Rationale:
Clinical Management:
There is some Hold
evidence
study products
that HIV when
RNA should
the instrumented
be detected HIV4 weeks
Ag/Ab aftertest result is
received. Samples
TDF/FTC discontinuation.
will be sent
Negative
for testing
test results
using for
a discriminatory
a discriminatory testtest
(if available,
(if performed),
indexan and
confirmatory
HIV RNA test,visits),
and an anHIV
HIVDNARNAtest,
test all
(index
performed
and confirmatory
4 weeks after visits)
TDF/FTC
and andiscontinuation,
HIV DNA test
(confirmatory
provide reasonable
visit only).
assurance
Additional
that HIV
a participant
testing will
is also
not infected.
be performed
Thereafter
are no
a 4-week
data onproduct
the
hold; thisofwill
amount timeinclude
required
an HIV
afterrapid
injections
test, astop
discriminatory
before HIV test
antibody,
(if available),
HIV RNA anorHIVHIVRNADNAtest,
will
anddetected
be an HIV DNAin individuals
test. Blinded
who study
become product
infected
maywhile
be restarted
receivingafter
CAB-LA
the 4-week
PrEP. One product
couldhold if
the
argueArchitect Ag/Ab test
that 4 weeks afterfrom the confirmatory
discontinuation visit isinjections
of CAB-LA ≤10 (or ifwould
this value
not beis not
longreported
enoughby forthe
laboratory),
HIV infectionand if the
to be HIV rapidHowever,
unmasked. tests, thewithholding
discriminatory tests,from
of PrEP HIVparticipants
RNA tests, and HIV DNA
at high-risk for
tests from the index
HIV acquisition for 8 visit,
or 12the confirmatory
weeks (the timevisit, and thefor
anticipated post-product
CAB-LA to hold“decay”visittoare
4 xallPA-IC90)
non-
reactive/negative.
is not prudent. In addition, in the absence of detectable HIV RNA and HIV DNA, the risk of
selecting for drug-resistant HIV in those with true infection should be very low. Further,
continued use of single- or dual-drug PrEP, even in cases with low-level reservoir seeding,
could theoretically lead to reservoir eradication (cure of infection). Therefore, a 4-week
interval after the last CAB-LA injection seems to provide a reasonable balance of risks and
benefits in this setting. If instrumented HIV Ag/Ab tests are reactive at subsequent study
visits for participants who resume study product, please refer to additional considerations
below (“Persistently Positive/Reactive HIV Screening Tests”).