Journal of Medical Virology 77:23–28 (2005)
Impact of Drug Resistance Genotypes on
CD4þ Counts and Plasma Viremia in Heavily
Antiretroviral-Experienced HIV-Infected Patients
Berta Rodes,1Federico Garcı ´a,2Carolina Gutierrez,3Javier Martinez-Picado,4Antonio Aguilera,5
Maria Saumoy,6Alex Vallejo,7Pere Domingo,8David Dalmau,9Maria Angels Ribas,10
Jose ´ Luis Blanco,11Jose ´ Pedreira,12Maria Jesu ´s Perez-Elias,3Manuel Leal,7
Carmen de Mendoza,1and Vincent Soriano1* On behalf of the Red de Investigacio ´n en SIDA (RIS)
1Hospital Carlos III, Madrid, Spain
2Hospital San Cecilio, Granada, Spain
3Hospital Ramon y Cajal, Madrid, Spain
4IrsiCaixa Foundation, Hospital Germans Trias i Pujol, Universitat Auto `noma de Barcelona, Badalona,
5Hospital Universitario, Santiago de Compostela, A Corun ˜a, Spain
6Hospital Joan XXIII-Universitat Rovira i Virgili, Tarragona, Spain
7Hospital Virgen del Rocio, Sevilla, Spain
8Hospital de Sant Pau, Barcelona, Spain
9Mutua de Terrassa, Terrassa, Barcelona, Spain
10Hospital Son Dureta, Palma de Mallorca, Spain
11Hospital Clinic, Barcelona, Spain
12Hospital Juan Canalejo, A Corun ˜a, Spain
multiple treatment failures is increasing as time
passes by. The success of antiretroviral therapy
in these patients is often compromised by the
selection of drug-resistant viruses. Despite ini-
tial concerns, a rebound in AIDS cases among
heavily treatment-experienced patients failing
virologically their antiretroviral therapy has not
occurred yet. In a multicenter study conducted in
Spain, HIV-infected patients were assessed with
drug families who presented during the last
semester of the year 2003 with plasma HIV-RNA
values above 1,000 copies/ml, despite good
treatment adherence. The relationships between
genotypes were examined. A total of 273 pati-
ents were identified in 12 centers (78% male,
median age: 41 years). The mean viral load was
50,438 copies/ml and the mean CD4þ count was
328 cells/ml. Only 19.5% had less than 200 CD4þ
T cells/ml. Most patients (95%) were receiving nu-
their last antiretroviral regimen, while 63% were
treated with protease inhibitors (PI) and 27% on
non-nucleoside reverse transcriptase inhibitors
(NNRTI). Overall, 97.4% had at least one drug
resistance mutation (87.2% for NRTI, 68.5% for
type, resistance to three or more drugs within
each class was recognized in 45.8% for NRTI,
40.7% for NNRTI, and 44.7% for PI. Moreover,
cross-resistance to compounds from two or
three drug families was recognized in 41%
and 19.4% of patients, respectively. Nearly half
of the patients had plasma HIV-RNA below
10,000 copies/ml and they showed significantly
higher CD4þ counts than those with greater
viremia (408 versus 259 cells/ml; P<0.001). Pa-
tients with higher plasma viremia had signi-
ficantly more drug resistance mutations than
those with lower viremia. No favorable effect on
viral load could be recognized for individual drug
resistance mutations known to reduce viral fit-
ness in vitro (i.e., rtM184V, rtL74V, rtK65R,
proD30N, or proI50L). In summary, a large pro-
portion of treatment-experienced patients failing
Grant sponsor: Redes de Investigacio ´n en SIDA.
*Correspondence to: Dr. Vincent Soriano, Department of
Infectious Diseases, Hospital Carlos III, Calle Sinesio Delgado
10, 28010 Madrid, Spain. E-mail: email@example.com
Accepted 10 May 2005
Published online in Wiley InterScience
? 2005 WILEY-LISS, INC.
their current antiretroviral regimen carry viruses
with broad cross-resistant genotypes. Nearly
half of the patients with these multi-drug re-
sistant viruses had <10,000 HIV-RNA copies/ml
and80% have more than 200 CD4þ T cells/ml.
Thus, maintaining treatment HIV-infected indivi-
duals failing virologically and harboring drug-
resistant viruses might ameliorate immunolo-
gical deterioration until new drugs became
available. J. Med. Virol. 77:23–28, 2005.
? 2005 Wiley-Liss, Inc.
drug resistance; antiretroviral
therapy; HIV; viral fitness; pro-
The number of HIV-infected patients with prior
multiple treatment failures is increasing as time passes
by. The success of antiretroviral therapy in these
patients is often compromised by the selection of drug-
resistant viruses [Cheung et al., 2004; Richman et al.,
2004]. When resistance has developed to almost all
currently available antiretroviral agents, salvage inter-
ventions including mega-HAART or structured treat-
ment interruptions have been attempted, generally
without major benefits [Lawrence et al., 2003; Ruiz
et al., 2003]. In patients with high CD4þ T cell counts,
drug holidays may be advisable to spare drug toxicities
and give a chance to wild-type viruses to reappear
[Halfon et al., 2003]. However, in patients with low
CD4þ counts, continuation of antiretroviral therapy
despite virological failure has shown to provide immu-
nological benefit as long as plasma viremia is not in-
creasing beyond 10,000–20,000HIV-RNA copies/ml [Le
Moing et al., 2002; Raffanti et al., 2004; The PLATO
heavily pre-treated patients failing their current anti-
retroviral regimen. However, it is unclear what should
be expected from new agents when they belong to
the available drug families, since accumulation of mul-
tiple resistance mutations within the reverse transcrip-
tase (RT) or protease genes most likely will impact
deleteriously on virus susceptibility to these new drugs
to a more or less extent [Turner et al., 2004]. In this
setting, inhibitors of new targets may be the ultimate
unique option to regain the goal of complete virus sup-
pression. This has been clearly demonstrated recently
2003; Clotet et al., 2004].
In the present study, the main characteristics of a
large group of HIV-infected patients failing their cur-
rent antiretroviral regimen after exposure to agents
from all three main drug classes were evaluated.
The relationships between drug resistance genotypes,
CD4þ counts and plasma viremia were investigated
PATIENTS AND METHODS
prior virological failure under compounds from all three
the last semester of year 2003, and had plasma HIV-
RNA values above 1,000 copies/ml, despite good treat-
ment adherence, were assessed in a large multicenter,
in Spain. The main demographics and clinical charac-
teristics were recorded in a uniform case report form.
CD4þ T cell counts were measured using monoclonal
antibodies followed by cytofluorometry at each insti-
tution. Plasma HIV-RNA was quantified using one of
three commercially available viral load tests (Bayer
Population-based sequences of the HIV-1 pol gene were
obtained after RT-PCR amplification of HIV-RNA
extracted from plasma specimens using one of two com-
mercially available automatic sequencers (Celera Diag-
nostics and Visible Genetics).
The definitions of broad cross-resistance to a given
drug family were considered as follows. For NRTI, the
term was applied to genotypes known to confer reduced
susceptibility to at least three compounds within this
class [Gallego et al., 2004]. This definition applied to
samples with three or more thymidine analogue muta-
tions (TAMs), including T215Y/F, since this genotype
has been associated with resistance to zidovudine,
stavudine, and abacavir [Larder and Bloor, 2001]. If
TAMs included changes M41L or L210W, resistance to
tenofovir should further be expected [Barrios et al.,
2003; Miller et al., 2004].
For NNRTI, the presence of two or more mutations of
a list of nine changes (codons 100, 101, 103, 106, 108,
cross-resistance a given genotype. Although unique
single changes (i.e., K103N) may probably compromise
drug [Casado et al., 2002; Gonzalez de Requena et al.,
The definition of broad cross-resistance to PI was
based on the presence of a minimum of five resistance
mutations at the protease gene, following clinical data
reported in several studies [Barreiro et al., 2002; Valer
et al., 2002; De Mendoza and Soriano, 2004].
In order to confirm the reliability of the criteria used
for defining broad cross-resistance on the basis of geno-
types, when possible, the virtual phenotype (Tibotec-
Virco VA, Mechelen, Belgium) was obtained.
Descriptive statistics were expressed as mean and
standard deviations (CD4 counts and viral load) or
median and interquartile ranges (number of resistance
for categorical data. Median values were compared
using the Student’s t-test. The chi-square test was used
24Rodes et al.
were performed using the SPSS version 9.0 software
A total of 273 multitreated HIV-positive patients
failing theircurrent antiretroviralregimen were identi-
fied (78% male; median age, 41 years). Most individuals
were native Spaniards (96.1%), and the main risk
groups were injection drug users (39.2%), homosexual
men (31.9%), and heterosexuals (21.6%). The median
time on antiretroviral therapy was 8.5 years. The
median number of prior different antiretroviral regi-
mens was seven. At the time of the analysis, the mean
viral load was 50,438 copies/ml and the mean CD4þ
count was 328 cells/ml. Only 19.5% of patients had less
than 200 CD4þ cells/ml. Most patients (95%) were
receiving NRTIs in their last antiretroviral regimen,
while 63% were treated with PIs and 27% on NNRTIs.
Overall, 97.4% of patients showed at least one drug
resistance mutation recorded within the last IAS-USA
ing broad cross-resistance for each drug class were
recognized in 45.8% for NRTI, 40.7% for NNRTI, and
44.7% for PI (see Table I). Virtual phenotypic results
ed almost completely the interpretation based on geno-
types (data not shown). Phylogenetic analyses showed
that all patients were infected by HIV-1 subtype B
Overall, 87.2% of patients had mutations associated
with NRTI resistance. The genotypes causing more fre-
quently broad NRTI cross-resistance were: >3 TAMs
(37%), >3 TAMs þ M184V (14.7%), >3 TAMsþL74V
Besides, the Q151M complex and/or codon 69 inserts
were recognized in 5.1% of patients.
With NNRTIs, 68.5% of patients had at least one
NNRTI mutation, and 40.7% had ?2 mutations caus-
ing broad NNRTI cross-resistance. Finally, 92.7% of
patients had at least one protease resistance mutation,
and 44.7% carried ?5 protease resistance mutations,
which may cause multiple PI resistance. Broad cross-
resistance to compounds from two or three drug fami-
lies was recognized in 41% and 19.4% of patients,
Of note, mutations K65R (RT), whose rate has been
reported to be increasing following the approval of
tenofovir [Valer et al., 2004], and L33F/I/S (protease),
whose relevance on PI resistance has recently been
highlighted [Mayersetal.,2004],wereobserved in 6.2%
tested harbored the Q145M mutation, which has been
associated rarely with resistance to either NRTI and/or
NNRTI [Paolucci et al., 2003].
Overall, 59.7% of patients had less than 20,000 HIV-
RNA copies/ml and they had significantly higher CD4þ
counts than subjects with greater viral load values
(392 versus 234 cells/ml; P<0.001). This threshold of
of CD4þ T cell fall in patients with sustained virological
failure under antiretroviral therapy [Raffanti et al.,
2004], although similar studies have reduced this
critical threshold to 10,000 HIV-RNA copies/ml [Le
Moing et al., 2002; The PLATO collaboration, 2004].
CD4þ counts were significantly much higher in those
with lower viral loads (see Table II).
Patients with viruses harboring drug resistance
mutations known to reduce viral fitness in vitro (such
TABLE I. Resistance Mutations in 273 Heavily Antiretroviral-Experienced Patients
in Year 2003
NRTI, nucleoside reverse transcriptase inhibitors; NNRTI, non-nucleoside reverse transcriptase
inhibitors; PI, protease inhibitors.
 for interpreting resistance genotypes as well as the virtual phenotype for more than half of
Relationship Between Drug Resistance, Viral Load and CD4 Counts 25
as rtM184V, rtL74V, rtK65R, proD30N, or proI50L)
did not have plasma HIV-RNA values lower than the
rest (data not shown). Moreover, the number of resis-
tance mutations differed significantly when comparing
patients with low or high plasma viral load (Table III).
Accordingly, patients with less than 10,000 or 20,000
HIV-RNA copies/ml had viruses with less resistance
mutations than patients with higher plasma viremia.
This association was dependent mainly on a larger
number of drug-resistant genotypes against PI in
patients with elevated plasma viremia.
This study shows that a large proportion of heavily
treatment-experienced, HIV-infected patients failing
their current antiretroviral treatment in Spain carry
viruses with broad cross-resistant genotypes. Resis-
population, and 19.4% harbored viruses showing resis-
tance to multiple agents from all three main drug
patients. New agents which belong to the available
drug classes may have limited activity in this subset of
patients, given the large number of resistance muta-
tions found in their viruses. To overcome this problem,
to maximize the activity of new PIs such as atazanavir
and tipranavir in patients with prior PI exposure [De
Mendoza and Soriano, 2004; Turner et al., 2004]. Un-
fortunately, similar pharmacokinetic enhancement
cannot be done for RT inhibitors.
In this cross-sectional study, patients with less than
10,000 or 20,000 HIV-RNA copies/ml showed signifi-
cantly higher CD4þ counts than patients with greater
plasma viremia. This finding is in agreement with the
reported better immunological outcome of patients
failing antiretroviral therapy but showing viral load
values below a certain threshold [Le Moing et al., 2002;
Raffanti et al., 2004; The PLATO collaboration, 2004].
Although thereasonwhyCD4þ countsdonotdeclineas
expected in this subset of patients is not well under-
stood, an impaired virus replication capacity due to the
phenomenon [Barbour et al., 2002; Deeks et al., 2001;
Deval et al., 2004]. However, in our study patients with
viruses harboring classical resistance mutations known
rtK65R, proD30N, or proI50L [Martinez-Picado et al.,
1999; Menendez-Arias et al., 2003; De Mendoza et al.,
2004; Miller, 2004], did not show lower plasma HIV-
RNA values than the rest of patients. We believe that
the accumulation of compensatory mutations in these
heavily treatment-experienced patients may have pre-
cluded the recognition of any beneficial effect of such
resistance mutations on viral fitness [Nijhuis et al.,
1999; Barbour et al., 2002]. In support of this inter-
pretation is our finding of a much higher number of
resistance mutations in patients with viral loads above
10,000 or 20,000 HIV-RNA copies/ml with respect to
the accumulation of mutations in the protease gene was
the main force driving this association [Nijhuis et al.,
thymus by viruses carrying protease resistance muta-
tions could explain this finding [Stoddart et al., 2001;
Delgado et al., 2002].
The results of this study indicate that new inhibitors
targeting HIV elements other than the RT and pro-
tease enzymes are urgently needed. They may be the
only alternative to regain complete virus suppression in
heavily-treatment experienced patients. The recent
this concept [Lazzarin et al., 2003; Clotet et al., 2004],
although the low genetic barrier of this drug may has
precluded to maintain complete virus suppression for
[Poveda et al., 2004; Sista et al., 2004].
Finally, these data support indirectly the view that
antiretroviral therapy can provide immunological and
virological benefit even in patients with drug-resistant
viruses. It is noteworthy that only 20% of this heavily
TABLE II. Differences in the CD4þ T Cell Count
According to Plasma Viral Load Values in
TABLE III. Differences in the Median Number of Drug Resistance Mutations According to
Viral Load Values in 273 Heavily Treatment-Experienced Patients
Plasma HIV-RNA (copies/ml)No. NRTI NNRTI PITotal
NRTI, nucleoside reverse transcriptase inhibitors; NNRTI, non-nucleoside reverse transcriptase
inhibitors; PI, protease inhibitors.
26 Rodes et al.
less than 200 CD4þ T cells/ml. In a prior study, a shift
from drug-resistant to drug-sensitive viruses after dis-
continuing antiretroviral therapy was recognized in a
small group of chronically HIV-infected patients [Deeks
et al., 2002]. This shift was associated with an increase
in the viral load and a parallel sharp reduction in the
CD4þ count. Thus, keeping on treatment patients fail-
until new therapeutic options become available, as long
as drug toxicity issues are not relevant.
This work was supported in part by a grant from
Redes de Investigacio ´n en SIDA (project 173).
2002. Evolution of phenotypic drug susceptibility and viral re-
plication capacity during long-term virological failure of protease
inhibitor therapy in HIV-infected adults. J Virol 76:11104–11112.
Barreiro P, Camino N, de Mendoza C, Valer L, Nunez M, Martin-
efficacy, safety and predictive value of HIV genotyping using dis-
Barrios A, de Mendoza C, Martin-Carbonero L, Ribera E, Domingo P,
Galindo MJ, Galvez J, Estrada V, Dalmau D, Asensi V, Soriano V.
2003. Role of baseline HIV genotype as a predictor of viral response
to tenofovir in heavily treatment-experienced patients. J Clin
Casado JL, Moreno A, Hertogs K, Dronda F, Moreno S. 2002. Extent of
cross-resistance to efavirenz after nevirapine failure. AIDS Res
Human Retroviruses 18:771–775.
Cheung P, Wynhoven B, Harrigan PR. 2004. 2004: Which HIV-1 drug
resistance mutations are common in clinical practice? AIDS Rev
Clotet B, Raffi F, Cooper D, Delfraissy JF, Lazzarin A, Moyle G,
Rockstroh J, Soriano V, Schapiro J. 2004. Clinical management of
treatment-experienced, HIV-infected patients with the fusion in-
hibitor enfuvirtide: Consensus recommendations. AIDS 18:1137–
DeMendoza C,Soriano V.2004.Resistance to HIV proteaseinhibitors:
Mechanisms and clinical consequences. Curr Drug Metab 5:321–
de Mendoza C, Paxinos E, Barreiro P, Camino N, Nun ˜ez M, Soriano V.
2004. Different viral rebound following discontinuation of antire-
troviral therapy in cases of infection with viruses carrying L74V or
thymidine-associated mutations. J Clin Microbiol 42:862–866.
DeeksS, WrinT,Liegler T,Hoh R,HaydenM,BarbourJ,HellmannN,
Petropoulos C, McCune J, Hellerstein M, Grant R. 2001. Virologic
and immunologic consequences of discontinuing combination
antiretroviral drug therapy in HIV-infected patients with detect-
able viremia. N Engl J Med 344:472–480.
Deeks S, Hoh R, Grant R, Wrin T, Barbour J, Narvaez A, Cesar D, Abe
K, Hanley M, Hellmann N, Petropoulos C, McCune J, Hellerstein
who remain viremic despite long term treatment with protease
inhibitor-based therapy. J Infect Dis 185:315–323.
Delgado J, Leal M, Ruiz-Mateos E, Martinez-Moya M, Rubio A,
Merchante E, de la Rosa R, Sanchez-Quijano A, Lissen E. 2002.
Evidence of thymic function in heavily antiretroviral-treated HIV
type 1-infected adults with long-term virologic treatment failure.
J Infect Dis 186:410–414.
Deval J, White K, Miller M, Parkin N, Courcambeck J,Halfon P, Selmi
B, Boretto J, Canard B. 2004. Mechanistic basis for reduced viral
and enzymatic fitness of HIV-1 RT containing both K65R and
M184V. J Biol Chem 279:509–516.
resistance genotypes causing cross-resistance to nucleos(t)ide
analogues. AIDS 18:689–690.
Gonzalez de Requena D, Gallego O, Corral A, Jimenez-Nacher I,
Soriano V. 2004. Higher efavirenz concentrations determine the
response to viruses carrying NNRTI resistance mutations. AIDS
Halfon P, Durant J, Clevenbergh P, Carsenti H, Celis L, Khiri H,
De Smet K, De Brauwer A, Hulstaert F, Dellamonica P. 2003.
Kinetics of disappearance of resistance mutations and reappear-
ance of wild-type during structure treatment interruptions. AIDS
Schapiro J, Telenti A, Richman D. 2004. Update of the drug
Top HIV Med 12:119–124.
Larder B, Bloor S. 2001. Analysis of clinical isolates and site-directed
Antiviral Ther 6:38.
Lawrence J, Mayers D, Hullsiek K, Collins G, Abrams D, Reisler R,
Crane L, Schmetter B, Dionne T, Saldanha J, Jones M, Baxter J.
2003. Structured treatment interruption in patients with multi-
drug-resistant HIV. N Engl J Med 349:837–846.
Lazzarin A, Clotet B, Cooper D, Reynes J, Arasteh K, Nelson M,
Katlama C, Stellbrink HJ, Delfraissy JF, Lange J, Huson L,
DeMasiR, Wat C, Delehanty J,Drobnes C, Salgo M. TORO 2 Study
Group. 2003. Efficacy of enfuvirtide in patients infected with drug-
resistant HIV-1 in Europe and Australia. N Engl J Med 348:2186–
Le Moing V, Thiebaut R, Chene G, Leport C, Cailleton V, Michelet C,
Fleury H, Herson S, Raffi F, APROCO Study Group. 2002. Pre-
dictors of long-term increase in CD4þ cell counts in HIV-infected
patients receiving a protease inhibitor-containing antiretroviral
regimen. J Infect Dis 185:471–480.
Martinez-Picado J, Savara A, Sutton L, D’Aquila R. 1999. Replicative
Mayers D, Leith J, Valdez H, Boucher C, Schapiro J, Baxter J. 2004.
Impact of three or four protease mutations at codons 33, 82, 84 and
90 on 2 week virological responses to tipranavir, lopinavir, ampre-
navir and saquinavir all boosted by ritonavir in phase 2b trial BI
1182.51. Antiviral Ther 9:163.
Menendez-Arias L, Martı ´nez M, Quin ˜ones-Mateu M, Martinez-Picado
J. 2003. Fitness variations and their impact on the evolution of
antiretroviral drug resistance. Curr Drug Targets Infect Disord
Miller M. 2004. K65R, TAMs and tenofovir. AIDS Rev 6:22–33.
Miller M, Margot N, Lu B, Zhong L, Chen S, Cheng A, Wulfsohn M.
2004. Genotypic and phenotypic predictors of the magnitude of
response to tenofovir treatment in antiretroviral-experienced
patients. J Infect Dis 189:837–846.
Nijhuis M, Schuurman R, de Jong D, Erickson J, Gustchina E,
Albert J, Schipper P, Gulnik S, Boucher C. 1999. Increased fitness
of drug resistant HIV-1 protease as result of acquisition of com-
pensatory mutations during suboptimal therapy. AIDS 13:2349–
Paolucci S, Baldanti F, Tinelli M, Maga G, Gerna G. 2003. De-
tection of a new HIV-1 RT mutation (Q145M) conferring
resistance to nucleoside and non-nucleoside inhibitors in a
patient failing highly active antiretroviral therapy. AIDS 17:
Poveda E, Rodes B, Labernardiere JL, Benito JM, Toro C, Gonzalez-
Lahoz J, Faudon JL, Clavel F, Schapiro J, Soriano V. 2004.
Evolution of genotypic and phenotypic resistance to enfuvirtide in
HIV-infected patients experiencing prolonged virological failure.
J Med Virol 74:21–28.
Raffanti S, Fusco J, Sherrill B, Hansen N, Justice A, D’Aquila R,
Mangialardi W, Fusco G. 2004. Effect of persistent moderate
viremia on disease progression during HIV therapy. J Acquir
Immun Defic Syndr 37:1147–1154.
Richman D, Morton S, Wrin T, Hellmann N, Berry S, Shapiro M,
Bozzette S. 2004. The prevalence of antiretroviral drug resistance
in the United States. AIDS 18:1393–1401.
Ruiz L, Ribera E, Bonjoch A, Romeu J, Martinez-Picado J, Paredes R,
Diaz M, Marfil S, Negredo E, Garcia-Prado J, Tural C, Sirera G,
drug salvage antiretroviral regimen: The Retrogene study. J Infect
Sista P, Melby T, Davison D, Jin L, Mosier S, Mink M, Nelson E,
DeMasi R, Cammack N, Salgo M, Matthews T, Greenberg M. 2004.
Relationship Between Drug Resistance, Viral Load and CD4 Counts27
Characterization of determinants of genotypic and phenotypic
resistance to enfuvirtide in baseline and on treatment HIV-1
isolates. AIDS 18:1787–1794.
Stoddart C, Liegler T, Mammano F, Linquist-Stepps V, Hayden M,
of protease inhibitor-resistant HIV-1 in human thymus. Nat Med
The PLATO Collaboration. 2004. Predictors of trend in CD4þ T-cell
count and mortality among HIV-1-infected individuals with viro-
logical failure to all three antiretroviral drug classes. Lancet
Turner D, Schapiro J, Brenner B, Wainberg M. 2004. The influence of
PI resistance profiles on selection of HIV therapy in treatment-
naive patients. Antiviral Ther 9:301–314.
Study Group. 2002. Impact of HIV genotyping and drug levels on
the response to salvage therapy with saquinavir/ritonavir. AIDS
analogue mutations. AIDS 18:2094–2096.
28 Rodes et al.