236 • CID 2005:41 (15 July) • HIV/AIDS
H I V / A I D SM A J O R A R T I C L E
Antiviral Activity of Lamivudine in Salvage Therapy
for Multidrug-Resistant HIV-1 Infection
Thomas B. Campbell,1Nancy S. Shulman,2Steven C. Johnson,1Andrew R. Zolopa,2Russell K. Young,1
Lane Bushman,3Courtney V. Fletcher,3E. Randall Lanier,4Thomas C. Merigan,2and Daniel R. Kuritzkes5
Division of Infectious Diseases, Departments of
Research Triangle Park, North Carolina; and
Harvard Medical School, Boston, Massachusetts
2Clinical Pharmacy, University of Colorado Health Sciences Center, Denver,
4Department of Clinical Virology, GlaxoSmithKline,
5Section of Retroviral Therapeutics, Brigham and Women’s Hospital and Division of AIDS,
3Division of Infectious Disease, Stanford Medical School, Stanford, California;
multidrug-resistant human immunodeficiency virus type 1 (HIV-1). Available data suggest that lamivudine con-
tributes to partial viral suppression, despite the presence of M184V mutationsandhigh-levelphenotypiclamivudine
Selective lamivudine withdrawal was studied in 6 subjects who had incomplete viral suppression
during antiretroviral treatment for multidrug-resistant HIV-1 infection.
Plasma levels of HIV-1 RNA increased to 0.5 log10copies/mL above baseline 6 weeks after the
withdrawal of lamivudine treatment (), even though reversion of lamivudine resistancewasnotyetdetected.P p .04
Early increases in plasma levels of HIV-1 RNA after lamivudine withdrawal were associated with the presence of
the T215Y/F mutation and broad phenotypic resistance to nucleoside reverse-transcriptase inhibitors at baseline.
Genotypic and phenotypic reversion of lamivudine resistance was detected in 4 subjects 8–14weeksafterwithdrawal
of lamivudine therapy. The duration of lamivudine withdrawal ranged from 8 to 22 weeks; all subjects resumed
lamivudine treatment. Plasma levels of HIV-1 RNA were 0.6 log10copies/mL above baseline (
lamivudine therapy was resumed. After the resumption of lamivudine treatment, plasma HIV RNA levelsdecreased
to baseline levels in 3 subjects but remained elevated in 3 subjects who had evolution of increased antiretroviral
drug resistance during the period of lamivudine withdrawal. Safety concerns raised by this latter finding led to
permanent closure of the study.
In select cases of multidrug-resistant HIV-1 infection, lamivudine contributes to suppression of
HIV-1 replication, despite the presence of M184V mutations and lamivudine resistance.
Maximum suppression of virus replication is often not achievable for persons infected with
) whenP p .03
Treatment options for persons infectedwithmultidrug-
resistant HIV-1 are limited, and, in many cases, sup-
pression of plasma levels of HIV-1 RNA to less than
detectable levels is not possible. However, the accu-
mulation of drug-resistance mutations in reverse tran-
scriptase (RT) and protease (PR) is sometimes associ-
ated with plasma HIV-1 RNA levels that are less than
pretreatment levels. In addition, antiretroviral drugs
may have partial activity, even in the presence of sub-
stantial resistance. Although complete viralsuppression
Received 29 November 2004; accepted 3 March 2005; electronically published
7 June 2005.
Reprints or correspondence: Dr.ThomasCampbell,CampusBoxB-168,University
of Colorado Health Sciences Center, 4200 E. Ninth Ave., Denver, CO 80262
Clinical Infectious Diseases 2005;41:236–42
? 2005 by the Infectious Diseases Society of America. All rights reserved.
may not be achievable in such patients, maintenance
of even partial viral suppression provides a clinicalben-
High-level lamivudine resistance is conferred bymu-
tations at RT codon 184 that produce a M184V or
M184I substitution in the conserved YMDD motif of
the RT polymerase domain . Despite the rapid ap-
pearance of the 184V mutation duringlamivudinether-
apy, plasma HIV-1 RNA levels remain partially sup-
pressed when lamivudine therapy is continued [3–5].
Withdrawal of lamivudine monotherapy or discontin-
uation of treatment with multidrug regimens that in-
clude lamivudine in persons infected with HIV-1 with
the 184V mutation is associated with increased plasma
levels of HIV-1 RNA and reversion to wild-type 184M
virus [6–8]. Explanations for the apparent paradox be-
tween the development of high-level lamivudine resis-
tance and continued partial suppression of HIV-1 rep-
by guest on November 26, 2015
HIV/AIDS • CID 2005:41 (15 July) • 237
lication are that the 184V mutation causes decreased viral rep-
lication fitness and increased susceptibility to other nucleoside
analogues. Alternatively, lamivudine could have residual anti-
despite high-level phenotypic resistance . To assess the po-
tential antiviral activity of lamivudine against lamivudine-
resistant virus, we conducted a pilot study of withdrawal of
lamivudine therapy (hereafter, “lamivudine withdrawal”)
among patients infected with lamivudine-resistant virus.
PATIENTS AND METHODS
The results of 2 prospective studies of open-label withdrawal
of the lamivudine component of the antiretroviral regimens
conducted at the University of Colorado General Clinical Re-
search Center and Stanford University School of Medicinewere
combined. Both studies followed the US Department of Health
and Human Services Guidelines for human experimentation
and wereapproved bytherespectiveinstitutionalreviewboards.
The entry criteria for the University of Colorado study were
as follows: (1) receipt of an antiretroviral regimen of ?3 drugs
that included zidovudine (or stavudine), lamivudine, and at
least 1 protease inhibitor (PI) and/or nonnucleoside reverse-
transcriptase inhibitor (NNRTI); (2) plasma HIV-1 RNA level
of ?1000 copies/mL within 60 days before study entry; (3)
screening RT/PR genotype showed the M184V mutation, at
least 3 thymidine analogue resistance mutations (i.e., M41L,
D67N, K70R, L210W, T215Y/F, or K219Q), and evidence of
resistance to other drugs that were part of the current regimen;
(4) cumulative duration of antiretroviral treatment of ?24
months, with failure of at least 2 previous antiretroviral regi-
mens; (5) no satisfactory alternative regimens available, or the
subject did not wish to initiate a new regimen; (6) willingness
to continue the current regimen without lamivudine;(7)ability
and willingness to give written, signed, informed consent; and
(8) age ?18 years. Potential subjects were excluded on thebasis
of the following criteria: (1) current abacavir, didanosine, or
zalcitabine use, because these drugs also may selectforthe184V
genotype; (2) treatment with an immune modulating agent
within 30 days before study entry; (3) treatment for an op-
portunistic infection within 14 days before study entry or pres-
ence of a malignancy requiring systemic chemotherapy; (4)
pregnancy or breastfeeding; (5) active drug or alcohol abuse
that would interfere with the study requirements.
Entry criteria for the Stanford University study were as fol-
lows: (1) CD4 lymphocyte count of 1200 cells/mm3; (2) plasma
HIV-1 RNA level of 1500 copies/mL; (3) receipt of stable an-
tiretroviral therapy including lamivudine for ?3 months; (4)
stable plasma HIV-1 RNA levels, definedas2consecutivevalues
within 0.5 log10copies/mL within 3 months; (5) presence of
the M184V mutation; and (6) age ?18 years. Subjects were
excluded on the basis of the following criteria: (1) receipt of a
regimen that included abacavir or didanosine, (2) active op-
portunistic infections, (3) chronic hepatitis B virus infection,
or (4) an anticipated change in antiretroviral therapy within
the subsequent 3 months.
At the time of study entry (day 0), subjects
were asked to discontinue taking lamivudine but to continue
taking all other antiretrovirals in their pre-entry regimen. After
withdrawal of lamivudine therapy, CD4+lymphocyte counts
and plasma HIV-1 RNA levels were determined, and genotypic
resistance tests were performed at weeks 1, 2, and 3 (University
of Colorado study only) and 4, 6, 8 and then every 2 weeks
(Stanford University study) or 4 weeks (University of Colorado
study). At each visit, current antiviral therapy was documented
to verify thatpatientscontinuedthenonlamivudinecomponent
of their regimens, and subjects were counseled to continue
adherence with this therapy. Resumption of therapy with la-
mivudine was required if any of the following events occurred:
(1) subject had a reduction in CD4+cell counts of 200 cells/
mm3or to ?50% of baseline,(2)subjectsustainedplasmalevels
of HIV-1 RNA of 11.0 log10copies/mL above baseline (Uni-
versity of Colorado study only), or (3) subject missed ?2 study
visits (Stanford University study only). Subjects who resumed
lamivudine therapy were asked to continue study follow-up.
Antiretroviral drug susceptibility assays.
sequence of RT and PR was determined by consensus sequence
analysis with the TruGene HIV-1 Sequencing Kit and
OpenGene DNA Sequence Analysis System (BayerNucleicAcid
Diagnostics) or Applied Biosystems Big Dye v3.1 (Applied Bio-
systems). Clonal analysis of HIV-1 RT was conducted by RT-
PCR amplification, as described elsewhere . All RT-PCRs
included samples that contained no RNA and a sample that
contained 4000 median tissue culture infective doses of HIV-
1 reference strain NL4-3. Phylogenetic trees were constructed
to rule out the occurrence of PCR contamination .
Phenotypic susceptibility to antiretrovirals (i.e., 50% inhib-
itory concentration [IC50]) and the contribution of RT/PR to
replication capacity (RC) were measured by the PhenoSense
HIV Drug Resistance Assay (ViroLogic). Specimens collected
at baseline and during interruption of lamivudine treatment
from individual subjects were assayed concurrentlytoeliminate
potential interassay variation.
Lamivudine, stavudine and zidovudine concentrations in
plasma were determined by reverse-phase high-performance
liquid chromatography (HPLC) [12, 13]. A simultaneous re-
versed-phase HPLC assay for the determination of indinavir,
amprenavir, nelfinavir, AG 1402 (active metabolite of nelfina-
vir), saquinavir, ritonavir, efavirenz, and lopinavir was per-
formed as described elsewhere .
Statistical methods and data analysis.
transformed plasma HIV-1 RNA data. Comparisons used a
by guest on November 26, 2015
238 • CID 2005:41 (15 July) • HIV/AIDS
treatment for multidrug-resistant HIV-1 infection.
Characteristics of and laboratory values at baseline for 6 subjects who had incomplete viral suppression duringantiretroviral
Reverse transcriptase/protease phenotype
ZDV D4T 3TCABCddITDFEFV NFV LPVIDV
D4T, 3TC, NFV 41L, 67N, 70R, 118I,
70R, 118I, 184V,
215F , 219E
41L, 184V, 210W,
67N, 70R, 103N,
41L, 69D, 118I,
184V, 210W, 215Y
67N, 70R, 103N,
D4T, 3TC, LPV,
D4T, 3TC, LPV,
D4T, 3TC, IDV
6 728 10,965ZDV, 3TC,
99 5014 30
tenofovir; ZDV, zidovudine; 3TC, lamivudine. Subjects 1, 2, 3, and 4 were studied at the University of Colorado Health Sciences Center. Subjects 5 and 6 were
studied at Stanford Medical School.
aMutations in HIV-1 reverse transcriptase associated with antiretroviral drug resistance, according to International AIDS Society–USA criteria.
bHIV-1 protease and reverse transcriptase replication capacity relative to the HIV-1 NL4-3 control, measured by the PhenoSense assay.
cFold-change in the concentration of the drug that inhibits virus replication by 50% in the PhenoSense assay. Values in boldface indicate reduced phenotypic
susceptibility, defined as a fold-change that exceeded the defined cutoff for full susceptibility.
dMean of values determined at screening (i.e., week ?1) and study entry (i.e., week 0).
ABC, abacavir; ddI, dideoxyinosine; D4T, stavudine; EFV, efavirenz; IDV, indinavir; LPV, lopinavir; NFV, nelfinavir; RTV(LD), low-dose ritonavir; TDF ,
Wilcoxon signed-rank test for paired continuous variables, a
Mann-Whitney test for groups of continuous variables, and
Fisher’s exact test for categorical variables. Statistical compar-
isons used Statview (Abacus Concepts) and assumed a 2-sided
significance level of .05. Ninety-five percent CIs based on the
binomial distribution were computed with Splus, version 6.2
A planned safety review was conducted
by an independent monitor after enrollment of the fourth sub-
ject at the University of Colorado. The monitor noted that
plasma levels of HIV-1 RNA did not return to baseline levels
after resumption of lamivudine therapy in 3 subjects and that
further evolution of antiretroviraldrugresistanceoccurreddur-
ing the period of lamivudine withdrawal. The monitor rec-
ommended permanent closure of the study.
43–64 years) participated in the studies (table 1). All subjects
had virus with nucleoside reverse-transcriptase inhibitor resis-
tance, including the M184V mutation, as determined by con-
sensus sequence analysis, and high-levelphenotypiclamivudine
resistance. Resistance-associated mutations in PR (D30N in
subjects 1 and 2; M46I, I84V, and L90M in subject 3; K20R,
I54V, A71T, and V82A in subject 4; M46I and I84V in subject
5; and M46I and V82A in subject 6) and decreased phenotypic
susceptibility (30- to 231-fold increased IC50) to the PI com-
ponent of the antiretroviral regimen were also present. At the
Six men (median age, 48 years;range,
time of entry into the study, all subjects had plasma HIV-1
RNA levels of 110,000 copies/mL (median, 20,000 copies/mL),
and plasma HIV-1 RNA levels were within ? 0.5 log10copies
of baseline levels 8–12 weeks before lamivudine withdrawal for
5 of 6 subjects (figure 1).
Effects of lamivudine withdrawal.
ued treatment with lamivudine but continued treatment with
the other components of their antiretroviral regimens. All sub-
jects had quantifiable plasma lamivudine concentrations at the
time of entry into the study (range, 374–2026 ng/mL) and
undetectable plasma lamivudine concentrations at week 6 (!20
ng/mL). The expected thymidine analogues, PR inhibitors and
NNRTI listed for each subject in table 1 were detected at both
week 0 and week 6.
The duration of lamivudine withdrawal ranged from 8 weeks
(for subjects 2 and 5) to 22 weeks (for subject 6). During
lamivudine withdrawal, subject 1 had acute antibiotic-associ-
ated colitis at week 6 that responded to treatment for Clos-
tridium difficile infection. Viral load measurements were not
performed during the period of acute illness, and plasma HIV-
1 RNA levels determined 2 weeks before onset and 4 days after
mL, respectively). No other serious adverse events occurred.
The 184VrM reversion was detected at least once by con-
sensus sequence analysis of plasma virus in 4 subjects during
the period of lamivudine interruption (subjects 1, 3, 4, and 6;
figure 1). The median time to first detection of 184VrM was
12 weeks after lamivudine withdrawal (range, 8–14 weeks).The
All subjects discontin-
by guest on November 26, 2015
transcriptase codon 184 genotype. CD4+lymphocyte counts and plasma HIV-1 RNA levels at baseline are provided in table 1. Horizontal dashed lines,
baseline values; vertical dashed lines, time of lamivudine withdrawal at week 0; arrows, the time of lamivudine treatment resumption; open circles,
gray circles, and black circles, time points at which the plasma virus was had 184V, 184M/V, and 184M mutations, respectively, as determined by
consensus sequence analysis. Panels A and B, patient 1; C and D, patient 2; E and F, patient 3; G and H, patient 4; I and J, patient 5; K and L,
Effects of lamivudine (3TC) withdrawal and resumption of therapy on CD4+lymphocyte count, plasma HIV-1 RNA levels, and reverse-
by guest on November 26, 2015
240 • CID 2005:41 (15 July) • HIV/AIDS
184VrM reversion was first detected as a 184V/M mixture in
subjects 2, 3, and 4, followed by detection of only 184M at the
next time point (figure 1A, 1F, 1H, and 1L). Only the 184V/
M mixed mutation was detected in subject 1, who resumed
treatment with lamivudine after the week 12 visit (figure 1B).
In these 4 subjects, there was a trend toward increased plasma
HIV-1 RNA levels at the time the 184M mutationfirstappeared
(median, +0.3 log10copies/mL;
P p .07
signed rank test).
lamivudine susceptibility and 2.2-fold increased abacavir sus-
ceptibility. On the other hand, the 184VrM reversion was as-
sociated with 10-fold decreased zidovudine and 2-fold de-
creased tenofovir susceptibility. The IC50 for lamivudine,
abacavir, zidovudine, and tenofovir remained unchanged dur-
ing lamivudine withdrawal in the 2 subjects who did not have
the 184VrM reversion. Susceptibility to didanosine,stavudine,
efavirenz, and protease inhibitors was not affected by184VrM.
Median susceptibility to lamivudine, abacavir, zidovudine, and
tenofovir returned to baseline levels in association with the
reappearance of the 184V mutation after resumption of la-
mivudine. MedianRT/PR replicationcapacitywas41%atentry,
60% at the time of the 184VrM reversion, and 38% at the
time of the reappearance of the 184V mutation after treatment
with lamivudine was resumed (
Wilcoxon signed rank comparisons of the latter values with
values at entry, respectively).
During lamivudine withdrawal, CD4+lymphocyte countsre-
mained near baseline values (figure 1A, 1C, 1E, 1G, and 1K),
except in subject 5, who had a sustained 30% decreased lym-
phocyte count (figure 1I). The 184VrM reversion was not
associated with changes in CD4+lymphocyte counts.
Median plasma HIV-1 RNA levels increased 0.5 log10copies/
mL above baseline levels at6 weeksafterlamivudinewithdrawal
( , by paired Wilcoxon signed-rank test), even though
P p .04
the 184VrM reversion was not detected in any subject at or
before week 6 by consensus sequence analysis (figure 1). Two
patterns of plasma virus load trajectories were observed in in-
dividual subjects through week 6 of lamivudine interruption,
before the first detection of 184VrM reversion. Plasma HIV-
1 RNA levels were at or below baseline values in subjects 4 and
6 through week 6 (figure 1H and 1L). Subject 4 subsequently
had increased plasma virus load when the 184VrM reversion
occurred (+0.3 log10copies/mL at weeks 12 and 14). Although
the 184VrM reversion occurred at week 12 in subject6,plasma
HIV-1 RNA levels remained near baseline values.
Subjects 1, 2, 3, and 5 had early increased levels of plasma
HIV-1 RNA, beginning within several weeks after lamivudine
withdrawal (figure 1B, 1D, 1F, and 1J) even though 184M was
P p .07P p .9
The 184M mutation was not detected by sequence analysis of
12–14 molecular clones (95% CI for frequency of the 184M
mutation, 0–0.3 for 12 clones and 0–0.2 for 14 clones) derived
from the plasma of each of these 4 subjects at the last time
during lamivudine withdrawal that the 184M mutation was not
detected by consensus sequence analysis of the plasma virus.
Early increases in plasma HIV-1 RNA levels after lamivudine
interruption were associated with T215Y/F and M41L muta-
tions at study entry (subjects 1, 2, 3, and 5; table 1) and a
greater frequency of reduced phenotypic susceptibility to drugs
in the nucleoside reverse-transcriptaseinhibitorclass,excluding
lamivudine (75% vs. 10%;
P p .001
differences in plasma antiretroviral drug concentrations were
observed in subjects with early or delayed increases of plasma
HIV-1 RNA levels after lamivudine withdrawal.
All subjects resumed lamivudine
therapy, either at their own request or at the recommendation
of their primary care provider. No subject met the CD4+cell
count or plasma HIV-1 RNA level criteria for mandated re-
sumption. At the time that treatment with lamivudine was
resumed, median plasma HIV-1 RNA levels were 0.6 log10cop-
ies/mL above baseline levels (range, 0.1 to 0.6 log10copies/mL
above baseline levels; , by paired Wilcoxon signed-rank
P p .03
test). Eight weeks after treatmentwithlamivudinewasresumed,
plasma HIV-1 RNA levels were not significantly different from
levels at the time of study entry (
mutation was detected in plasma by consensus sequence
The response of plasma HIV-1 RNA to lamivudine resump-
tion in individual subjects was dichotomous and was related
to the evolution of increased antiretroviral drug resistancedur-
ing lamivudine interruption. Plasma HIV-1 RNA levels re-
mained at least 0.3 log10copies/mL above baseline levels at all
visits after lamivudine resumption for subjects 1, 2, and 3 (fig-
ure 1B, 1D, and 1F), who had increased phenotypic antiret-
roviral drug resistance during lamivudine withdrawal. Virus
from subject 1 acquired a K219Q mutation in RT during la-
mivudine withdrawal and had persistently increased stavudine
resistance (stavudine IC50increased 2-fold from baseline levels)
after treatment with lamivudine was resumed, despite the reap-
the L10F and G73S mutations in protease and a V118I/Vmixed
mutation in RT during lamivudine withdrawal; resistance to
stavudine and lopinavir increased 2-fold above baseline levels,
even after treatment with lamivudine was resumed and the
M184V mutation returned. Virus from subject 2 did not have
the 184VrM reversion or the new appearance of other mu-
tations in RT associated with drug resistance but had increased
zidovudine resistance during lamivudine withdrawal (zidovu-
dine IC50increased 2-fold from baseline levels).
, by Fisher’s exact test). No
), and only the 184V
P p .2
by guest on November 26, 2015
HIV/AIDS • CID 2005:41 (15 July) • 241
Lamivudine contributed a suppression of plasma HIV-1 RNA
levels of ∼0.5 log10, despite high-level phenotypic resistance to
lamivudine and thymidine analogues. This finding is consistent
with the findings of previous studies in which lamivudine
monotherapy, despite the presence of the M184V mutation,
provided a 0.5- and 0.6-log10suppressionofplasmaHIV-1RNA
levels in either antiretroviral-naive or treatment-experienced
persons, respectively [3, 15]. Given that a 0.5-log10decrease of
plasma HIV-1 RNA levels during partially suppressive antiret-
roviral therapy is associated with an ∼50% reduction of risk
of clinical disease progression [16–18], lamivudine could pro-
vide benefit in persons with the M184V mutation and limited
Even though the present study did not include a comparable
group of control subjects who continued to receive lamivudine
therapy during the follow-up period, several findings suggest
that the observed increased plasma virus load was a directeffect
of lamivudine withdrawal. First, the observed virus load in-
creases exceeded the expected interassay variation for plasma
HIV-1 RNA quantification [19, 20]. Second, plasma HIV-1
RNA levels were stable before lamivudine withdrawal, and in-
creased plasma HIV-1 RNA levels were temporally related to
lamivudine withdrawal and/or 184VrM reversion. Third, the
presence of all prescribed antiretroviraldrugsinplasmaatstudy
entry and week 6 and the absence of lamivudine at week 6
provides objective evidence of adherence to study procedures.
Finally, the reversibility of virus load increases afterresumption
of lamivudine therapy in the subjects who did not develop
increased antiretroviral drug resistance provides additionalevi-
dence for a direct effect of lamivudine on plasma levels of HIV-
Our finding that plasma HIV-1 RNA levels increased prior
to the 184VrM reversion in 4 subjects suggeststhatlamivudine
contributed to partial viral suppression in salvage therapy, de-
spite the presence of the 184V mutation. This finding is con-
sistent with the results of other studies in which early increases
in plasma virus load occurred after withdrawal of other nu-
cleoside reverse-transcriptase inhibitors, but not after with-
drawal of protease inhibitors or NNRTIs [21, 22]. Collectively,
these observations suggest direct residual antiviral activity of
nucleoside analogues against resistant virus in some patients.
Because lamivudine acts synergistically with thymidine ana-
logues against zidovudine-resistant HIV-1  and partially
inhibits lamivudine-resistant RT , there are established
mechanisms for direct effects of lamivudine on lamivudine-
and zidovudine-resistant HIV-1. Loss of either lamivudine or
thymidine analogue synergy and/or residual RT inhibition by
lamivudine withdrawal would be expected to result in prompt
increases in HIV-1 replication, as occurred in our study. The
trends toward increased plasma HIV-1 RNA levels and repli-
cation capacity at the time of 184VrM reversion in some sub-
jects suggest that continued lamivudine therapy also provided
partial viral suppression by maintenance of the 184V mutation.
The potential virologic benefits of maintaining the 184V mu-
tation include increased thymidine analogue and tenofovirsus-
ceptibility and decreased RT replication fitness.
Our findings differ with the results of the COLATE trial,
which found that inclusion of lamivudine in a new treatment
regimen after failure of a lamivudine-containing regimen does
not provide additional virologic benefit. It is importanttopoint
out that all subjects in our study had incomplete suppression
of virus replication while they were receiving a lamivudine-
containing regimen. In contrast, subjects in COLATE received
a new antiretroviral regimen of ?3 drugs with or without la-
mivudine, and the majority of subjects had suppression of
plasma HIV-1 RNA levels to ?400 copies/mL while they were
receiving the new regimen. It is possible that the effects of
adding 3 new drugs to the antiretroviral regimens in the COL-
ATE treatment arms overshadowed any benefit of continued
Generalization of the results of the present study is limited
for several reasons. First, the present study was designed to be
a pilot study of relatively homogenous subjects. All subjects
had multiple thymidine analogue resistance mutations at the
time of study entry and receivedathymidineanaloguethrough-
out the course of lamivudine withdrawal, and no subjects re-
ceived abacavir or didanosine during the study. Because la-
mivudine monotherapy provides suppression of plasma HIV-1
RNA levels of ∼0.5 log10in patients with a M184V mutation
without thymidine analogue mutations , it is unlikely that
the antiviral effects of lamivudine in salvage therapy require
concomitant administration of a thymidine analogue or the
presence of thymidine analogue resistance mutations. The in-
crease in plasma HIV RNA levels after lamivudine withdrawal
noted in the present study suggests that lamivudinecontributes
to partial viral suppression during salvageantiretroviraltherapy
in a select group of patients who have resistance to both la-
mivudine and other nucleoside reverse-transcriptaseinhibitors.
Automated DNA sequence analysis was performed at the University of
Colorado Cancer Center DNA Sequencing Core (Denver, CO).
GlaxoSmithKline (grant COL40192 to T.B.C.), the
National Institutes of Health (grants AI42567 and RR16482 to D.R.K;grant
AI51199 to N.S.S; and grant AI33835 to C.V.F.), the Colorado Center for
AIDS Research (grant AI054907) and the University of Colorado General
Clinical Research Center (grant RR00051).
Potential conflicts of interest.
T.B.C. has received grant support from,
was on the speakers’ bureau of, and was a consultant for GlaxoSmithKline.
N.S.S. has received research funding from Bristol-Myers Squibb; was a
consultant for GlaxoSmithKline, Bristol-Myers Squibb, and Abbott; and
was on the speakers’ bureau for GlaxoSmithKline, Bristol-Myers Squibb,
by guest on November 26, 2015
242 • CID 2005:41 (15 July) • HIV/AIDS
Virologic, Roche Pharmaceuticals, Abbott, and Gilead. E.R.L. is an em-
ployee of GlaxoSmithKline. T.C.M. is a consultant for Quest Diagnostics.
D.R.K. is a consultant for and has received honoraria and research grant
support from GlaxoSmithKline. S.C.J. has received research funding from
Bayer Corporation and Bristol-Myers Squibb Company and is a consultant
for AstraZeneca and Aventis Pharmaceuticals. A.R.Z. has received research
funding from Bristol-Myers Squibb, Abbott, and Johnson & Johnson; is a
consultant for Bristol-Myers Squibb; and is a scientific advisor for Bristol-
Myers Squibb, Gilead, Virologic, and Roche Diagnostics. R.K.Y., L.B., and
C.V.F.: no conflicts.
1. Deeks SG, Wrin T, Liegler T, et al. Virologic and immunologic con-
sequences of discontinuing combination antiretroviral-drug therapyin
HIV-infected patients with detectable viremia. N Engl J Med 2001;
2. Tisdale M, Kemp SD, Parry NR, Larder BA. Raqid in vitro selection
of human immunodeficiency virus type 1 resistant to 3?-thiacytidine
inhibitors due to a mutation in the YMDD region of reverse tran-
scriptase. Proc Natl Acad Sci USA 1993;90:5653–6.
3. Eron JJ, Benoit SL, Jemsek JJ, et al. Treatment with lamivudine, zi-
dovudine, or both in HIV-positive patients with 200 to 500 CD4+cells
per cubic millimeter. N Engl J Med 1995;333:1662–9.
4. Catucci M, Venturi G, Romano L, et al. Development and significance
of the HIV-1 reverse transcriptase M184V mutation during combi-
nation therapy with lamivudine, zidovudine, and protease inhibitors.
J Acquir Immune Defic Syndr 1999;21:203–8.
5. Schurman R, Nijhuis M, van Leeuwen R, et al. Rapidchangesinhuman
immunodeficiency virus type 1 RNA load and appearance of drug-
resistant virus populations in persons treated with lamivudine (3TC).
J Infect Dis 1995;171:1411–9.
6. Kuritzkes DR, Quinn JB, Benoit SL, et al. Drug resistance andvirologic
zidovudine (ZDV) versus ZDV plus 3TC in previously untreated pa-
tients. AIDS 1996;10:975–81.
7. Verhofstede C, van Wanzeele F, Van Der Gucht B, De Cabooter N,
Plum J. Interruption of reverse transcriptase inhibitors or a switch
from reverse transcriptase to protease inhibitors resulted in a fast reap-
pearance of virus strains with a reverse transcriptaseinhibitor–sensitive
genotype. AIDS 1999;13:2541–6.
8. Devereux H, Youle M, Johnson MA, Loveday C. Rapid decline in
9. Quan Y, Brenner BG, Oliveira M, Wainberg MA. Lamivudine canexert
a modest antiviral effect against human immunodeficiency virus type
1 containing the M184V mutation. Antimicrob Agents Chemother
10. Schooley RT, Campbell TB, Kuritzkes DR, et al. A phase 1 study of
combination therapy with L-697,661andzidovudine.JAcquirImmune
Defic Syndr 1996;12:363–70.
11. Kuiken C, Korber B. Sequence quality control. In: Korber B, Kuiken
C, Fuley B, et al., eds. Human retroviruses and AIDS. Los Alamos,
NM: Los Alamos National Laboratory, 1998:80–90.
12. Kakuda TN, Page LM, Anderson PL, et al. Pharmacologic basis for
concentration-controlled therapy with zidovudine, lamivudine,andin-
dinavir. Antimicrob Agents Chemother 2001;45:236–42.
13. Fletcher CV, Brundage RC, Remmel RP, et al. Pharmacologic char-
acteristics of indinavir, didanosine, and stavudine in human immu-
nodeficiency virus–infected children receiving combination therapy.
Antimicrob Agents Chemother 2000;44:1029–34.
14. Remmel RP, Kawle SP, Weller D, Fletcher CV. Simultaneous HPLC
assay for quantification of IDV, NVP, RTV and SQV in human plasma.
Clin Chem 2000;46:73–81.
15. Castagna A, Danise A, Carini E, et al. Pilot study to evaluate immu-
nologic responses to lamivudine monotherapy vs. treatment interrup-
[abstract WeOrB1286]. In: Program and abstracts of the 15th Inter-
national AIDS Conference (Bangkok). Geneva: International AIDS So-
16. O’Brien WA, Hartigan PM, Martin D, et al. Changes in plasma HIV-
1 RNA and CD4+lymphocyte counts and the risk of progression to
AIDS. N Engl J Med 1996;334:426–31.
17. Coombs RW, Welles SL, Hooper C, et al. Association of plasmahuman
immunodeficiency virus type 1 RNA level with risk of clinical pro-
gression in patients with advanced infection. J Infect Dis 1996;174:
18. Marschner IC, Collier AC, Coombs RW, et al. Use of changesin plasma
levels of human immunodeficiency virus type 1 RNA to assess the
clinical benefit of antiretroviral therapy. J Infect Dis 1998;177:40–7.
19. Lin HJ, Myers LE, Yen-Lieberman B, et al. Multicenter evaluation of
quantification methods for plasma human immunodeficiency virus
type 1 RNA. J Infect Dis 1994;170:553–62.
20. Raboud JM, Montaner JSG, Conway B, et al. Variation in plasma RNA
levels, CD4 cell counts, and p24 antigen levels in clinically stable men
with human immunodeficiency virus infection. J Infect Dis 1996;174:
21. Deeks SG, Paxinos EE, Wrin T, et al. Limited genotypicandphenotypic
evolution after interruptionofa singletherapeuticdrugclassinpatients
with multidrug-resistant HIV. Antiviral Therapy 2003;8:S73.
22. Maldarelli F, Palmer S, Kearney M, et al. Short duration, single drug
discontinuation to assess the activity of individual drugs in patients
failing antiretroviral therapy. Antiviral Therapy 2003;8:S149.
23. Merrill DP, Moonis M, Chou TC, Hirsch MS. Lamivudine orstavudine
in two- and three-drug combinations against human immunodefi-
ciency virus type 1 replication in vitro. J Infect Dis 1996;173:355–64.
24. Dragsted U, Fox Z, Mathiesen L, et al. Final 48 week analysis of a
phase 4, randomized, open-label, multi-center trial to evaluate safety
and efficacy of continuedlamivudinetwicedailyversusdiscontinuation
of lamivudine in HIV-1–infected adults with virological failure on on-
going combination treatments containing lamivudine: the COLATE
Trial [abstract 549]. In: Program and abstracts of the 11th Conference
on Retroviruses and Opportunistic Infections (San Francisco). Alex-
andria, VA: Foundation for Retrovirology and Human Health, 2004:
by guest on November 26, 2015