Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

578 Abstracts 32288-32292

12th World AIDS Conference

32288 Transmission of protease inhibitor resistant HIV-1 to a
recently infected antiretroviral-naive man: The UCSF
options primary HIV project
Frederick Hecht1, J.O. Kahn2, B. Dillon3, M. Chesney2, R.M. Grant4. 1995
Potrero Avenue, San Francisco, California 94110; 2University of California, San
Francisco, CA; 3Centers for Disease Control and Prevention, San Francisco, CA;
4Gladstone Institute of Virology & Immunology, San Francisco, CA, USA
Objective: To determine the prevalence of antiretroviral drug resistance in recently infected drug naive subjects.
Methods: 48 recently HIV-1 infected persons in San Francisco were identified
between 6/96 and 1/98 by documented seroconversion in the previous 6 months
or history of recent seroconversion with confirmation by a detuned HIV-1 antibody
EIA test. Drug resistance was assessed by genotypic analysis of plasma HIV-1
RNA using Affymetrix probe arrays and D-rhodamine dye terminator cycle sequencing. Subjects who reported intake of antiretroviral drugs prior to enrollment
were excluded from analysis.
Results: Of the first 20 subjects analyzed, 3TC resistance mutations (rt 184V)
were identified in 3 subjects (15%). One of the subjects with the rt 184 mutation,
diagnosed with acute HIV in 9/97, had several mutations in reverse transcriptase
(rt 184V and 215Y) and protease (M461, V82A, L90M). This genotypic mutational
pattern likely represents resistance to AZT, 3TC, indinavir, ritonavir, saquinavir,
and nelfinavir. In this subject, plasma viremia has continued to be detectable
(3,906 HIV-1 RNA copies/ml) at 16 weeks of therapy with AZT/3TC/indinavir,
while all other participants receiving similar treatment in the same study have
been undetectable (HIV-1 RNA < 500) by week 12. We have initiated analysis of viral drug resistance in this subject's sexual partner who is extensively
antiretroviral experienced.
Conclusion: This is the first case we are aware of with apparent transmission
of protease inhibitor resistant HIV-1. 3TC resistant HIV-1 is the most commonly
transmitted drug resistant HIV-1 variant in San Francisco. The high prevalence of
drug resistance among currently HIV-1 infected persons may portend transmission of drug resistant variants and resistance assessment in recently infected drug
naive subjects may be warranted in some settings. Interventions to prevent drug
resistance in individuals may have epidemiologic as well as clinical importance.
231 */32289 Phenotypic HIV resistance in vitro correlates with
viral load response to abacavir (1592, ABC) in vivo
E.R. Lanier, M.L. Smiley, M.H. St Clair, G. Pearce, W. Spreen, A. Cutrell,
S.W. Lafon. Glaxo Wellcome Inc., 5 Moore Drive, Research Triangle Park, NC,
USA
Objectives: To determine if in vitro phenotypic HIV resistance testing is predictive
of clinical viral load response to abacavir (ABC) in a large number of patients with
diverse histories of anti-retroviral usage.
Design: As part of the clinical development plan for the new nucleoside, HIV
reverse transcriptase inhibitor (NRTI) ABC, viral genotype and resistance phenotype have been extensively correlated with viral load response. These parameters
have been examined in numerous clinical trials that enrolled antiretroviral therapy
naive and therapy experienced patients.
Results: Current results from over 100 patients indicate that plasma virus
phenotype may predict both short (4 weeks) and longer term (24 weeks) viral
load response to ABC. Three groups have been distinguished by comparing
fold resistance of clinical isolates to a sensitive laboratory isolate. Subjects with
plasma virus that is <4x resistant to ABC seem to have a very good chance of
responding well to therapy while those with plasma virus that is 4-8x resistant
may not respond as well or for as long a time. To date patients with isolates that
are >8x resistant to ABC do not benefit from therapy with this drug as defined by
viral load reduction. Given the small number of isolates in this group (n = 12) it
appears most patients should respond to ABC; however, a much larger data set
is being compiled to confirm these findings.
Conclusions: HIV resistance phenotype prior to therapy appears to be an
independent predictor of viral load response to ABC. Evaluation of plasma virus
resistance by in vitro phenotyping may prove valuable in selecting therapies to
maximally suppress viral replication for as long as possible in anti-retroviral naive
and experienced patients.
32290 Antiretroviral drug resistance testing to support the
clinical care of patients with HIV/AIDS
Clive Loveday, H. Devereux, L. Huckett, M. Johnson. Dept. Retrovirology, Royal
Free Hopspital, Med. Sch., Roland Hill St., London NW3 2PF, UK
Objectives: The high replication rate and mutability of the HIV-1 genome lends
itself to the rapid evolution of drug resistant quasispecies in the presence of
antiretroviral agents; these events are qne reason for drug failure. We have
undertaken the establishment and validation of an antiretroviral testing facility to
assist in patient care and describe laboratory and clinical requirements, problems
encountered, clinical findings to date and cost benefits within a London clinic.

Methods: Base changes associated with antiretroviral drug resistance were
determined using patient plasma HIV-1 RNA derived from viral load estimations,
this was reverse transcribed and RT (codons 1-245) and protease (codons 1-99)
were amplified by nested PCR. Bi-directional, consensus sequencing was performed using automated technology (ABI 377). The point mutation assay (PMA)
was performed to quantify proportions of wild to mutant at given codons to monitor
longitudinal resistance changes in certain cases.

The assays included internal and external controls to monitor quality, and strategies for translating complex sequence data into clinical reports were evaluated.
Evaluations included patients who were* drug-naive (n = 25), pregnant (n = 5),
drug experienced (n = 50)
Results: In patients with virological failure we identified genotypic changes
associated with resistance to: AZT (82%, M41L, D67N, L210W & T215Y/F)
and 3TC (70%: M184V), and multidrug resistance changes for AZT/3TC (50%:
M184V, R211K & L214F). No Q151M mutations were found. All patients had
some mutations associated with protease inhibitors and 25% had >5 mutations.
Combination therapies were designed on the basis of these results to include
drugs not implicated.
Conclusions: Establishment of molecular virological approaches for clinical
care require specific laboratory protocols, new controls and 'translational' approaches for the delivery of virological results and opinions. Cost benefits are
being determined on the basis of patient follow-up after 4 to 6 months on selected
new therapies.
32291 Viral load and CD4+ outcome following genotypic
antiretroviral resistance testing (GART) in primary care
settings in Atlanta, Georgia
Julia Gable, Melanie Thompson, L. Anisman, R. Hudson, R. Kauffman,
M. Patino, M. Tanner, R. Dudley. AIDS Research Consortium of Atlanta, 131
Ponce De Leon S. 130, Atlanta, GA, USA
Objective: To ascertain viral load and CD4+ outcomes in a continuously enrolling
cohort of patients who have used GART to assist with antiretroviral decision-making
Background: There are no clinical data yet to confirm that the use of GART results in better clinical, virologic, or immunologic outcome than clinical judgement,
yet GART is being used with increasing frequency as an aid to decisionmaking
regarding changes in antiretroviral therapy.
Methods: Patients who had used GART were referred for chart abstraction. Patients were selected only if their prior antiretroviral history, GART results, baseline
and followup CD4+ levels and plasma HIV-RNA quantitations were available.
Results: 24 patients were followed for a mean of 5 months after GART was
obtained. In these patients, 35 GART assays were obtained from 3 different commercial laboratories. Mean CD4+ count prior to GART was 336 cells//il. Mean
HIV-RNA level was 24,068 copies/ml (4.38 log). Mean prior antiretroviral therapy
was 51 months and patients had taken a mean of 6.5 AR drugs prior to GART
testing. The mean post-GART CD4+ count was 358 cells/til and HIV-RNA was
43,232 copies/ml (4.64 log). Overall, there was no significant difference between
CD4+ and viral load values pre- and post- GART. Seven patients (29%) achieved a
viral load below 500 copies/ml at any time following GART testing. These patients
are considered responders. The mean pre-GART CD4+ count for these patients
was 405 cells/pl, compared with 308 cells/il for non-responders (p = NS), and
HIV-RNA level was 23,660 copies/ml (4.37 log), compared with 24,247 copies/ml
(4.38 log) for non-responders (p = NS). For responders vs non-responders, prior
AR therapy was 49 vs 52 mos. (p = NS). There was no significant difference in
the use of nucleosides; however, patients had used 0.8 vs 2.3 Pi's (p < 0.05).
Conclusions: There was no significant difference between pre- and postGART CD4+ count or viral load in these experienced patients. The patients who
attained viral suppression below 500 copies did not differ from non-responders in
baseline CD4+ count or HIV-RNA level, or in total duration of prior AR therapy.
Responders had used fewer protease inhibitors than non-responders. GART may
be more useful in patients with less AR experience, particularly with protease
inhibitors. A randomized clinical trial, currently underway in the CPCRA, is needed
to fully assess the benefit of GART.
32292 Genotypic analysis for resistance testing (GART) as a
clinical tool: A retrospective analysis
James A. Newton12, A.S. Pemberton2, W.W. Emmons2, D.L. Mayers3. 13733
Farn Sworth Drive, Chesapeake Virginia; 2Naval Medical Center, Portsmouth,
VA; 3National Naval Medical Center, Bethesda, MD, USA
Objectives: To assess the clinical utility of GART in patients clinically judged as
failing HAART.
Design: Retrospective chart review.
Methods: We analyzed the charts of 32 HIV-infected patients clinically judged
as failing HAART (9.5% of our clinic). These 32 patients underwent 37 separate
GART analyses (RT-PCR of plasma HIV RNA with sequencing of protease and
first 250 amino acids of RT enzymes. Mean viral load (MVL) prior to GART was
compared with the MVL within 6 months after therapy was changed based on
GART data. The effect on MVL reduction of five specific mutations known to
reflect phenotypic resistance was also studied (PI codons 46, 82, & 90 and RT
codons 184 & 215). We developed a "GART score" assigning 1 point for each of
these mutations, resulting in a score ranging from 0-5, and assessed its effect
on MVL reduction.
Results: The mean reduction in VL after GART was 0.85 log. The MVL prior

to GART [(mean, 164 K copies/ml plasma, 95% CI, 110 K to 218 K; 4.95 log, CI,
4.77 to 5.13)] was significantly greater than the MVL within 6 months after therapy
was changed based on GART data [(mean 55 K copies/ml plasma, CI, 28 K to
82 K; 4.14 log, CI, 3.87 to 4.44)] (p = 0.028, CI, 0.077 to 1.274). ANOVA testing
showed a significant negative correlation, with higher GART scores resulting in
a smaller MVL reduction (p = 0.011, CI, -0.46 to -0.06). When each mutation
was analyzed independently, the presence of the PI codon 90 mutation had the