Download HIV drug resistance acquired through superinfection

HIV drug resistance acquired through superinfection
Davey M. Smitha, Joseph K. Wonga,b,d, George K. Hightowera, Caroline
C. Ignacioa, Kersten K. Koelscha, Christos J. Petropoulosc, Douglas D.
Richmana,b and Susan J. Littlea
Objective: HIV interclade B superinfection has previously been described in individuals initially infected with drug resistant virus who then become superinfected by a drug
susceptible strain. We report an individual initially infected with a drug-sensitive clade
B strain of HIV who was superinfected with another clade B strain resistant to two
classes of antiretroviral drugs.
Methods and design: To differentiate superinfection from possible co-infection we
applied three independent molecular techniques: dye-primer sequencing of a pol
fragment, length polymorphism analysis of the V4–5 coding region of the env gene
and clonal sequencing of the V3 coding region of the env gene. To assess viral fitness we
performed replication capacity assays of the pol gene.
Results: These investigations supported the conclusion that this was a case of superinfection and not co-infection. Coincident with acquiring the new strain, the individual’s viral load increased by about 10 000 copies/ml with a decrease of 150 CD4 T
cells/ml over the next 6 months. The greater in vivo fitness of the second virus was not
supported by the replication capacity assay. Furthermore, superinfection negatively
impacted this individual’s treatment course. It was not known that he had acquired a
drug resistant strain before entering a treatment study, and he had an incomplete
response to therapy most likely because the superinfecting viral strain had a decreased
susceptibility to two of the prescribed medications.
Conclusion: HIV drug resistance acquired through superinfection significantly lowers
the likelihood of successful antiretroviral therapy and undermines the clinical value of a
patient’s prior drug resistance testing and lack of prior antiretroviral use.
ß 2005 Lippincott Williams & Wilkins
AIDS 2005, 19:1251–1256
Introduction
Superinfection with HIV is presumed to occur frequently
because of the prevalence of interclade recombinants
[1,2]. However, documented superinfection in humans
has only recently been reported, first with different clades
and subsequently, two instances of intraclade B superinfection [3–7]. Koelsch et al. reported a patient who was
initially infected with drug-resistant HIV and then
secondarily infected by drug-sensitive (‘wild-type’) virus.
The superinfecting wild-type HIV replicated at high
levels and masked from standard drug resistance tests the
underlying presence of the initial drug-resistant virus [4].
We now report a patient initially infected with a drugsensitive clade B strain of HIV who was superinfected
with another clade B strain resistant to two important
classes of antiretroviral drugs.
Case report
In September 2001, a 31-year-old man was enrolled into
the San Diego Acute HIV Infection and Early Disease
Research Program (AIEDRP) after recently testing
positive for HIV at a community clinic. He previously
From the aUniversity of California San Diego, San Diego, California, USA, the bVeterans Affairs San Diego Healthcare System, San
Diego, California, USA, the cViroLogic, Inc., South San Francisco, and the dVAMC, San Francisco, San Francisco, California, USA.
Correspondence to D. M. Smith, 9500 Gilman Drive 0679, La Jolla, CA 92093-0679, USA.
Tel: +1 858 552 8585 extn 2624; fax: +1 858 552 7445; e-mail: [email protected]
Received: 6 August 2004; revised: 20 September 2004; accepted: 17 November 2004.
ISSN 0269-9370 Q 2005 Lippincott Williams & Wilkins
1251
AIDS 2005, Vol 19 No 12
had a negative HIV test in October 2000, but reported
having had symptoms consistent with acute HIV
infection in April 2001. A repeat HIV ELISA in
September 2001 was positive, but the detuned assay
was negative, consistent with infection within 6 months.
His reported risk factor was unprotected receptive anal
intercourse with partners of unknown HIV serostatus. At
presentation, his HIV viral load (VL) was 1203 copies/ml
(AMPLICOR HIV-1 MONITOR Test, v.1.5, Roche
Diagnostics, Basel, Switzerland) and his CD4 cell count
was 571 /l (34% of total T cells). Genotypic and
phenotypic assays (Viroseq HIV-1 Genotyping System,
Celera Diagnostics, Alameda, California, USA and
PhenoSense HIV, Virologic, Inc., South San Francisco,
California, USA) detected no drug resistance and the
patient elected to defer antiretroviral therapy. Blood
samples were collected weekly for 4 weeks, monthly for
5 months and every 8 weeks thereafter.
Initially, the patient’s HIV VL remained low (140–2927
copies/ml) until 260 days after presentation when it
increased to 11 611 copies/ml. It continued to increase to
142 566 copies/ml 4 months later, which corresponded
to a drop in CD4 cell count to 331 /l (Fig. 1). A month
later, he entered a double-blind randomized treatment
trial, AIDS Clinical Trials Group (ACTG) 5095 [8]. After
the trial was closed and unblinded, his treatment regimen
was identified as lamivudine, zidovudine, and abacavir
(Trizivir, GlaxoSmithKline, Research Triangle Park,
North Carolina, USA). He reported that because of
nausea, he was only intermittently compliant to the
antiretroviral therapy. His VL remained detectable
throughout the period of follow-up (Fig. 1). After
voluntary withdrawal from the trial at 4 months, testing
revealed high-level drug resistance, including resistance
to protease inhibitors (L90M) and other new polymorphisms in protease (M36I, L63P), although protease
inhibitors were not used in the trial. When questioned
after resistance was detected, he denied taking any other
antiretroviral medications, but he did report frequent
unprotected anal receptive intercourse since his initial
HIV diagnosis, with partners of unknown HIV serostatus.
Methods
Clonal env sequencing
HIV RNA was isolated from the subject’s blood plasma
from two time-points 228 days apart (11 September 2001,
29 April 2002) using the QIAamp Viral RNA extraction
kit (QIAGEN Inc., Valencia, California, USA) according
to manufacturer’s instructions. Nested PCR of the V3
coding region was performed using previously described
primers and methods [4]. PCR products were cloned
using TOPO Cloning vectors (Invitrogen Corporation,
Carlsbad, California, USA) per the manufacturer’s
instructions. Fifteen clones were sequenced using dye
terminator sequencing and analyzed using the ABI
PRISM 3100 Genetic Analyzer (Applied Biosystems,
Foster City, California, USA).
Phylogenetic analysis of pol and env sequences
Sequences were initially compiled, aligned, and edited in
BioEdit using CLUSTAL W alignment tool. The
alignment was then manually edited to preserve frame
700
6
B
600
C
CD4 cells/µl
500
4
400
3
300
A
2
200
Plasma viremia (*1000)
5
CD4
VL
1
100
0
11
9/ /20
17 0
1
9/ /200
20 1
/
2
9/ 0
27 01
10 /20
/4 01
11 /20
/1 01
12 /20
/7 01
1/ /20
7/ 0
20 1
2/ 02
4/
2
3/ 002
4/
4/ 20
29 02
8/ /20
13 02
/2
9/ 00
4/ 2
9/ 20
26 0
/ 2
10 200
/9 2
11 /20
/6 02
12 /20
/5 02
/2
2/ 00
5/ 3
3/ 200
19 3
4/ /20
21 03
/2
00
3
0
9/
1252
Time course
Fig. 1. CD4 cell counts and VL for the subject’s first 19 months of enrollment in the primary infection cohort. VL rose from 395
copies/ml on 4 March 2002 to 11 611 copies/ml on 29 April 2002 (arrow A), which corresponded to the first appearance of highly
drug-resistant virus, which was identified retrospectively. On 9 October 2002 the subject started antiretroviral therapy (arrow B)
with a triple nucleoside regimen, which resulted in a transient decrease in VL to 304 copies/ml on 6 November 2002. He elected to
discontinue medication on 5 February 2003, and at the next sample collection on 19 March 2003 (arrow C), his VL was 35 972
copies/ml, which also corresponded to a reversion of the M184V mutation in reverse transcriptase.
HIV superinfection and drug resistance Smith et al.
Fig. 2. A Inferred maximum likelihood phylogeny of clonal env (a) and pol (b) sequences. (a) All 15 sequenced clones of the V3
coding region obtained from blood plasma collected on 11 September 2001 during acute infection clustered together (subject
time-point 1), and these sequences did not overlap with the cluster of 15 clones sequenced from virus collected on 29 April 2002
(subject time-point 2). The unlabeled branches are sequences from 18 other B clade viruses collected from epidemiologically
unrelated subjects in the San Diego area. The other labeled branches are clade B reference lab strains: NL43, LAI, NY-5, HXBR,
JRFL, YU-2, SF162, ADA, BAL, and JRCSF. (b) All pol sequences from viruses collected before 29 April 2002 harbored the wildtype 184M codon and clustered independently from the pol sequences from viruses collected on and after 29 April 2002 even
though the 13 August 2002 sample revealed a virus harboring the 184M codon. The unlabeled branches are sequences from 17
other B clade viruses collected from epidemiologically unrelated study participants in the San Diego area.
1253
1254
AIDS 2005, Vol 19 No 12
insertions and deletions. Phylogenetic analyses were
performed using Phylogeny Inference Package software (PHYLIP) and FastDNAml (J. Felsenstein, 1993.
PHYLIP version 3.5c. University of Washington, Seattle;
G. Olsen, University of Illinois) starting from a neighborjoining tree using the previously measured parameters
from the maximum likelihood tree. To evaluate for
possible sample contamination sequences from standard
lab strains were included in the phylogenetic analysis.
Length polymorphism analysis
The V4–5 coding region of the env gene from HIV RNA
was amplified by a nested PCR using previously
described primers and protocols [4]. These fluorescently
labeled PCR products were separated by capillary
electrophoresis using an automated sequencer (ABI
PRISM 3100) and analyzed by GeneScan Analysis
Software (Applied Biosystems, Foster City, California,
USA). These experiments were performed in triplicate.
Dye-primer sequencing
From HIV RNA isolated from blood plasma, a 500-base
region of the pol gene was sequenced using ABI PRISM
BigDye Primer v.3.0 kits with 20 M13 and M13
primers (Applied Biosystems, Foster City, California,
USA) according to the manufacturer’s protocols. Base
mixtures were determined from ABI trace files using
Vector NTI v.5.0 (InforMax, Frederick, Maryland,
USA). Values from forward and reverse strand sequencing
were averaged and correlated with those obtained from
control samples [9].
Results
HIV superinfection was suspected when retrospective
drug resistance testing (260 days after enrollment into the
AIEDRP cohort) on previously collected samples
detected acquisition of resistance to protease inhibitors,
to which the patient was never exposed. In retrospect, this
switch in drug susceptibility occurred 4 months before
the subject entered the treatment trial and in conjunction
with a rise in plasma HIV RNA and drop in CD4 cell
count. Both viral isolates were clade B. To discriminate
superinfection or coinfection from the selection of drug
resistance, cloning of the V3 region of HIV envelope was
performed on samples collected before and after the
emergence of drug resistance. Phylogenetic reconstruction revealed no intermingling of sequences between
the two time-points, and genetic distances between
the two clusters of sequences were greater then 10%
(Fig. 2a), indicating two epidemiologically distinct strains
[10–12].
To distinguish further between coinfection and superinfection, length polymorphism detection (GeneScan) of
the V4–5 region of HIV envelope from HIV RNA
extracted from blood samples collected at the two
time-points (September 2001 and April 2002) was also
performed. The assay can discern minority variants that
make up 1–5% of the viral population [4,13]. The viral
populations differed by a two-codon (six-base pair)
insertion in the region at the second time-point with
homogenous populations at both time-points (data not
shown). To further confirm the superinfection, dyeprimer sequencing of a 500-base pair fragment of the
HIV pol, which spanned the 184 region of reverse
transcriptase, was also performed. This method is
reported to detect a minor variant that makes up as little
as 5–10% of the total viral population [4,9]. In the two
samples obtained before the genotypic switch (11
September 20, 4 February 2002) only 184M (wild-type)
virus was detected. In the sample collected from the timepoint when drug resistance was first observed (29 April
2002), only 184V (lamivudine resistant) virus was
detected. In the sample collected 104 days later (13
August 2002) the 184M (wild-type) had re-emerged but
as a reversion in the context of the superinfecting virus.
There were no mixtures of drug-sensitive and drugresistant viruses at any time-point (data not shown).
Phylogenetic analysis of pol sequences, obtained through
population based sequencing (Viroseq), revealed that all
sequences collected before 29 April 2002 clustered
separately from all sequences obtained on and after 29
April 2002 (Fig. 2b).
Replication capacity (RC) of the viral isolates using a pol
sequence in a recombinant assay also changed between
time-points [14]. The first virus had an RC of 57%, while
the second virus had an RC of 18%, which is consistent
with the presence of the M184V mutation and protease
mutations [14,15]. At the next time-point (13 August
2002) when the 184V mutation had reverted to 184M,
the RC was 63% even though the protease mutations
remained. This change in RC did not explain the rise in
VL seen after the second HIV infection, suggesting viral
fitness based on HIV genes outside of the pol coding
region.
Discussion
Although unlikely, it is possible that the individual
described in this report was coinfected with both a drug
susceptible and drug resistant strain and that the drug
resistant strain only emerged at a later time. However,
four independent lines of molecular investigation: env
cloning, sequencing, and phylogenetic analysis; dyeprimer sequencing; length polymorphism detection;
and pol sequencing and phylogenetic analysis provide
evidence that the individual was initially infected with
wild-type HIV and subsequently superinfected with a
HIV superinfection and drug resistance Smith et al.
second strain resistant to two classes of antiretroviral
drugs, nucleoside reverse transcriptase inhibitors and
protease inhibitors. He maintained relatively good
control of the first virus, with low HIV VL and high
CD4 cell counts, without antiretroviral therapy. He
subsequently experienced a rise in VL and a drop in CD4
cell count coincident with the appearance of the second
virus, which is similar to previous reports of superinfection [4,16]. The greater fitness of the second virus in
this individual, demonstrated by its predominance in vivo,
was not supported by the replication capacity assay of the
viral isolates. The replication capacity assay only interrogates (or assesses) replication fitness contributed by the
pol coding region, while other genes such as env most
likely play an important role in determining in vivo fitness
of the virus [15].
This superinfection negatively affected this subject’s
disease and course of treatment. It was not known that
he had acquired a drug-resistant virus before he entered
the trial for treatment-naive patients. During the ACTG
5095 trial, he received the combination therapy of
zidovudine, lamivudine, and abacavir and his VL
decreased from 65 954 copies/ml to 263 copies/ml,
but never became undetectable (Fig. 1). This incomplete
response may have resulted from the patient’s intermittent
compliance with his medications secondary to nausea
during the trial or it may have been that the prescribed
treatment regimen was suboptimal since there was
unrecognized reduced susceptibility to two of the
medications used (M184V mutation in reverse transcriptase).
This first observation of superinfection with multidrugresistant virus was initially recognized as a failure of
antiretroviral therapy, which raises the possibility that
some patients who fail antiretroviral therapy might do so
because of superinfection rather than the de novo
evolution of drug resistance. At present, it appears
unlikely that superinfection contributes substantially to a
significant number of cases of drug resistance in clinical
practice since a recent retrospective study by Gonzalez
et al. comparing viral sequences from patients undergoing
serial resistance genotyping found no evidence for such
cases. However, it is unclear whether any baseline viral
sequences were compared in that study [12].
HIV drug resistance acquired through superinfection
significantly lowers the likelihood of successful antiretroviral therapy. It undermines the clinical value of a
patient’s prior drug resistance testing and lack of prior
antiretroviral use. Vigilant personal protection, through
safe sex practices or clean needle use for injection drugs,
must be counseled to patients already infected with HIV
even if their risk exposures are with other HIV infected
people.
Acknowledgements
We are grateful to the AIEDRP study participants for
their unwavering generosity and to the University of
California, San Diego AIEDRP staff: Heidi Aiem, Tari L.
Gilbert, Paula Potter, and Joanne Santangelo. We also
thank Eric Daar, Andrew J. Leigh-Brown, and Simon D.
W. Frost for their insightful comments, Nancy Keating,
Melissa Moore, Ruby Lam, and Theresa Russell for their
technical assistance, and Laureen Copfer for her administrative assistance.
Sponsorship: Supported by grants 5K23AI055276,
AI27670, AI38858, AI43638, AI43752, UCSD Centers
for AIDS Research (AI36214), AI29164, the General
Clinical Research Center for Research Resources
M01-RR00425 from the National Institutes of Health
and the Research Center for AIDS and HIV Infection of
the San Diego Veterans Affairs Healthcare System.
References
1. Blackard JT, Cohen DE, Mayer KH. Human immunodeficiency
virus superinfection and recombination: current state of
knowledge and potential clinical consequences. Clin Infect
Dis 2002; 34:1108–1114.
2. Goulder PJ, Rowland-Jones SL, McMichael AJ, Walker BD.
Anti-HIV cellular immunity: recent advances towards vaccine
design. AIDS 1999; 13 (Suppl A):S121–S136.
3. Altfeld M, Allen TM, Yu XG, Johnston MN, Agrawal D, Korber
BT, et al. HIV-1 superinfection despite broad CD8R T-cell
responses containing replication of the primary virus. Nature
2002; 420:434–439.
4. Koelsch KK, Smith DM, Little SJ, Ignacio CC, Macaranas TR,
Brown AJ, et al. Clade B HIV-1 superinfection with wild-type
virus after primary infection with drug-resistant clade B virus.
AIDS 2003; 17:F11–F16.
5. Jost S, Bernard MC, Kaiser L, Yerly S, Hirschel B, Samri A, et al. A
patient with HIV-1 superinfection. N Engl J Med 2002;
347:731–736.
6. Ramos A, Hu DJ, Nguyen L, Phan KO, Vanichseni S, Promadej
N, et al. Intersubtype human immunodeficiency virus type 1
superinfection following seroconversion to primary infection in two injection drug users. J Virol 2002; 76:7444–7452.
7. Yang O, Daar E, Jamieson B, Balamurugan A, Smith D, Pitt J,
Petropoulos C, Richman D, Little S, Leigh-Brown A. HIV-1
Clade B superinfection: evidence for differential immune containment of distinct clade b strains. J Virol 2005; 79:860–868.
8. Gulick RM, Ribaudo HJ, Shikuma CM, Lustgarten S, Squires
KE, Meyer WA 3rd, et al., AIDS Clinical Trials Group Study
A5095 Team. Triple-nucleoside regimens versus efavirenzcontaining regimens for the initial treatment of HIV-1 infection. N Engl J Med 2004; 350:1850–1861.
9. Strain MC, Gunthard HF, Havlir DV, Ignacio CC, Smith DM,
Leigh-Brown AJ, et al. Heterogeneous clearance rates of longlived lymphocytes infected with HIV: intrinsic stability predicts
lifelong persistence. Proc Natl Acad Sci USA 2003; 100:4819–
4824.
10. Albert J, Wahlberg J, Leitner T, Escanilla D, Uhlen M. Analysis of
a rape case by direct sequencing of the human immunodeficiency virus type 1 pol and gag genes. J Virol 1994; 68:5918–
5924.
11. Yirrell DL, Robertson P, Goldberg DJ, McMenamin J, Cameron
S, Leigh Brown AJ. Molecular investigation into outbreak of
HIV in a Scottish prison. BMJ 1997; 314:1446–1450.
12. Gonzales MJ, Delwart E, Rhee SY, Tsui R, Zolopa AR, Taylor J,
et al. Lack of detectable human immunodeficiency virus type 1
superinfection during 1072 person-years of observation.
J Infect Dis 2003; 188:397–405.
1255
1256
AIDS 2005, Vol 19 No 12
13. Zhang L, Chung C, Hu BS, He T, Guo Y, Kim AJ, et al. Genetic
characterization of rebounding HIV-1 after cessation of highly
active antiretroviral therapy. J Clin Invest 2000; 106:839–845.
14. Deeks SG, Wrin T, Liegler T, Hoh R, Hayden M, Barbour JD,
et al. Virologic and immunologic consequences of discontinuing combination antiretroviral-drug therapy in HIV-infected
patients with detectable viremia. N Engl J Med 2001; 344:472–
480.
15. Simon V, Padte N, Murray D, Vanderhoeven J, Wrin T, Parkin N,
et al. Infectivity and replication capacity of drug-resistant
human immunodeficiency virus type 1 variants isolated during
primary infection. J Virol 2003; 77:7736–7745.
16. Gottlieb GS, Nickle DC, Jensen MA, Wong KG, Grobler J, Li F,
Liu SL, et al. Dual HIV-1 infection associated with rapid disease
progression. Lancet 2004; 363:619–622.