....viral
blips were not associated with viral failure....
AIDS, July 2005
Martinez,
Valeriea; Marcelin, Anne-Genevièveb; Morini, Jean-Pierrea; Deleuze, Jeana;
Krivine, Annec; Gorin, Isabellea; Yerly, Sabined; Perrin, Lucd; Peytavin,
Gillese; Calvez, Vincentb; Dupin, Nicolasa
From the
aService de Dermatologie, Hopital Tarnier-Cochin, 89, rue d'Assas, 75006 Paris
bLaboratoire de
Virologie, Hopital Pitie-Salpêtrière, 45-83, boulevard de l'hopital, 75013
Paris
cLaboratoire de
Virologie, Hopital Saint-Vincent-de-Paul, 74-82, avenue Denfert Rochereau,
75674 PARIS Cedex 14, France
dLaboratory of
Virology, Geneva University Hospital, Switzerland
ePharmacie,
Hopital Bichat-Claude-Bernard, 46, rue Henri Huchard, 75018 Paris, France.
This work was
supported by the French National AIDS Research Agency (Agence Nationale de
Recherche sur le SIDA, ANRS) and by Sidaction.
This work was
presented in part at the XIII International HIV Drug Resistance Workshop, 8-12
June 2004, Tenerife, Spain (abstract 63).
".....Despite
long term viral suppression and apparently maintained CD4 cell counts, eight
patients under NNRTI therapy showed intermittent viraemia with a median level
at 350 copies/ml but again achieved plasma viral load below the limits of
quantification during the follow-up (< 50 copies/ml). Patients harbouring
blips had a higher HIV-1 RNA load using a more sensitive assay with a detection
limit of 3 copies/ml, in comparison with those who remained always
undetectable. Moreover, the occurrence of intermittent viraemia seems to have a
deleterious effect on the evolution of CD4 cell counts during the follow-up.
These results suggest that, among virologically well-controlled patients who
were treated with a NNRTI-based regimen after switching from PI-containing
regimen, intermittent viraemia was a frequent event, but one which did not
predict subsequent virological failure...the prevalence of intermittent viraemia
was 18% and was comparable with the prevalence of intermittent viraemia
reported in previous studies....in our study evaluating patients under
NNRTI-based regimen, blips were not associated with suboptimal plasma
concentrations of NNRTI drugs....."
Introduction
Therapy based on
efficient combination of antiretroviral drugs or highly active antiretroviral
therapy (HAART) changed the natural course of HIV infection, decreasing
morbidity, mortality, occurrence of opportunistic infection and increasing
survival [1,2]. Most of patients treated by HAART achieved and maintained HIV-1
plasma viral load levels below the limit of detection of currently approved
assays. Using a more sensitive assay, several studies have shown that residual
viraemia between 3 and 50 copies/ml was detectable in 50 to 70% of treated
patients [3-9]. Moreover, intermittent viraemia occurred frequently and was
associated with higher levels of replication but was not associated with
virological failure in patients treated by protease inhibitor (PI)-containing
regimen [10]. In this population, treatment changes may not be necessary to
maintain long-term virological suppression with low-level or intermittent
viraemia. Regimen that include non-nucleoside reverse transcriptase inhibitors
(NNRTIs) are potent inhibitors of HIV-1 replication, with rapid emergence of
resistant viruses both in vitro and in vivo in cases of therapeutic failure
[11-14]. Moreover, certain cross-class mutations confer significant resistance
to all NNRTI drugs and additional NNRTI resistance mutations observed with
K103N mutation enhanced the degree of phenotypic resistance to NNRTI [15-17].
The effects of intermittent viraemia need to be evaluated in patients treated
by NNRTI-based regimen.
The aims of this
study were to measure the prevalence of HIV-1 intermittent viraemia or 'blips',
to determine their impact on CD4 cell counts, the evolution of plasma viral
load, and the occurrence of virological failure in adherent patients with
long-term suppressed plasma viral load under NNRTI-based therapy.
Abstract
Background: It
has been demonstrated that, in patients treated by protease-inhibitor-based
regimen, intermittent viraemia occurred frequently and was associated with
higher concentrations of residual replication but not with virological failure.
Risk factors associated with intermittent viraemia and its impact in patients
treated by non-nucleoside-reverse-transcriptase-inhibitor-based (NNRTI) regimen
need to be evaluated.
Methods: We
analyzed the occurrence of blips (one HIV-1 RNA > 50 copies/ml with a
subsequent value < 50 copies/ml), the level of these blips (between 3 and 50
copies/ml) and their effect on CD4 cell count and the occurrence of virological
failure in 43 patients with stable suppression of HIV-1 plasma viraemia (<
50 copies/ml) under NNRTI-based therapy.
Results: Eight
out of 43 patients had one episode of blips during the follow-up (median = 350
copies/ml). Comparing patients with and without blips, the median level of
HIV-1 RNA at baseline was 7.5 versus 3 copies/ml (P = 0.008), respectively.
Patients with blips had a significantly lower CD4 cell count after 12 and 18
months than the others. Plasma concentrations of NNRTI before, during, and
after the blips were adequate. In addition, the occurrence of blips was not
associated with virological failure.
Conclusion:
These results suggest that blips may reflect ongoing viraemia of below 50
copies/ml and can impair the CD4 cell count recovery under an NNRTI regimen.
The impairment of CD4 cell count recovery seems to be affected more by the
occurrence of blips than by the level of viraemia (< 50 copies/ml) itself.
Nevertheless, despite a tight genetic barrier for resistance with NNRTI drugs,
no virologic failure occurred during the follow-up.
Results
Forty-three
HIV-1 infected patients were enrolled in this study. Median follow-up was 18
months (range, 6-24 months). The median age was 45 years (range, 29-75 years)
and 97.6% (42 of 43) were men. All patients have been infected by sexual
contacts: homosexual (n = 36) or heterosexual contacts (n = 7). PI-containing
regimen was the first-line of antiretroviral therapy for four patients, the
others were pre-treated by various combinations of
nucleoside-reverse-transcriptase-inhibitors (NRTI) and PI antiretroviral drugs
but were naive of NNRTI drugs. PI-containing regimen was used for a median time
of 41.5 months (range, 14-72 months). The reasons for modification from PI- to
NNRTI-containing regimen were to limit the occurrence of adverse effects (n =
24), to improve adherence (n = 5) by reducing pill burden or to simplify drugs
regimen (n = 14). Adverse effects with PI were as follows: lipodystrophy (n =
10), elevated triglyceride, cholesterol and/or glycemia levels (n = 5),
diarrhoea (n = 6) and kidney stones (n = 3). In PI-containing regimen, the
drugs used were: nelfinavir 750 mg three times a day (tid) or 1250 mg twice
daily (bid) (n = 25), indinavir 800 mg tid (n = 13) and ritonavir 600 mg bid (n
= 5). Comparison between patients treated by nelfinavir, indinavir or ritonavir
revealed no difference in plasma viral load, using a detection limit at 3
copies/ml (P = 0.86) at the time of the modification of the regimen. The median
of CD4 cell counts before the switch was 524 (range, 211-1026).
For combination
of therapy including NNRTI drugs, nevirapine was prescribed for 38 patients and
efavirenz for five patients. After the introduction of NNRTI-containing
regimen, the median CD4 cell counts was 543 (range, 229-1004) after 6 months,
561 (range, 137-1023) after 12 months and 564 (range, 166-954) after 18 months.
There was no statistical difference between CD4 cell counts before and after
the switch at any time (6, 12 or 18 months). With the standard determination of
HIV-1 plasma viral load (limit of detection 50 copies/ml), no significant variation
of the viral load was observed between baseline and months 6, 12 or 18.
Nevertheless, despite no significant differences, a gain of CD4 cell count of
40 in median was observed after 18 months under NNRTI therapy.
We observed an
intermittent viraemia or blips for eight patients during the follow-up. The
median level of blips was 350 copies/ml (range, 95-1750 copies/ml). The
duration of NRTI-exposure was similar between patients with blips [median 40.5
months (range, 7-120)] and the others [median 48 months (range, 17-114 months)]
(P = 0.5). Patients with intermittent viraemia had significantly lower CD4 cell
counts at 12 and 18 months than those with persistent suppressed viral load (P
= 0.04 and P = 0.02, respectively) (avg CD4 count was about 630 for patients
without blips & 500 for patients with blips at both time points). We
observed no difference in the level of plasma viral load at the time of the
switch between patients with and without blips using a limit of detection of 50
copies/ml. However, using a more sensitive assay with a detection limit at 3
copies/ml, patients with intermittent viraemia harboured higher plasma viral
load than patients with apparent suppressed viral load under PI before the
change for NNRTI-containing regimen, with a median of 7.5 versus 3 copies/ml (P
= 0.008). Eighteen months after the switch, plasma viral load was different
between the two groups using both HIV-1 RNA assays (P = 0.003 and P = 0.01,
using a limit of detection of 50 and 3 copies/ml, respectively). Patients with
a plasma viral load greater than 50 copies/ml at 18 months (n = 3) harboured a
subsequent value of less than 50 copies/ml at 24 months, confirming the nature
of the blips and thus excluding virologic failure. Plasma concentrations of
NNRTI-drugs were measured in two patients treated by NVP and two patients
treated by EFV. Plasma concentrations of NVP and EFV measured before, during,
and after the blips showed that blips occurred despite efficient plasma levels
of NNRTI drugs at these different times. At the time of blips, NVP plasma
concentrations were 5938 and 7951 ng/ml and EFV plasma concentrations were 2695
and 3064 ng/ml.
Author
Discussion
Despite long
term viral suppression and apparently maintained CD4 cell counts, eight
patients under NNRTI therapy showed intermittent viraemia with a median level
at 350 copies/ml but again achieved plasma viral load below the limits of
quantification during the follow-up (< 50 copies/ml). Patients harbouring
blips had a higher HIV-1 RNA load using a more sensitive assay with a detection
limit of 3 copies/ml, in comparison with those who remained always
undetectable. Moreover, the occurrence of intermittent viraemia seems to have a
deleterious effect on the evolution of CD4 cell counts during the follow-up.
These results suggest that, among virologically well-controlled patients who
were treated with a NNRTI-based regimen after switching from PI-containing
regimen, intermittent viraemia was a frequent event, but one which did not
predict subsequent virological failure.
In our study,
the prevalence of intermittent viraemia was 18% and was comparable with the
prevalence of intermittent viraemia reported in previous studies [3,10,22,23].
Sklar et al., analysed the impact of intermittent viraemia in patients with
previous virologic suppression treated either with PI- or non-PI-containing
regimen and showed no subsequent virologic rebound whatever the HAART regimen
used or the past history of antiretroviral therapy [23]. In the APROVIR study
based on naive patients, the occurrence of blips was correlated only to
suboptimal levels of PI drugs [22]. In contrast, in our study evaluating
patients under NNRTI-based regimen, blips were not associated with suboptimal
plasma concentrations of NNRTI drugs. In the ACTG 343 and in Merck 035 trials,
intermittent viraemia was a frequent event under PI-based regimen and was
associated with higher steady-state level of viral replication (Merck 035
trial), but not with virologic failure (ACTG 343 and Merck 035) after 4.5 years
of follow up [10]. In these two last trials, no pharmacologic data were
available. Similarly, in our study evaluating patients under NNRTI-containing
therapy, the occurrence of intermittent viraemia was a frequent event, but was
not associated with virologic failure.
Moreover,
previous studies about switching from PI to NNRTI based regimen used a
threshold above 50 copies/ml for measurement of HIV-1 plasma viral load. These
studies have shown clinical, virological and immunological safety of the switch
strategy [24]. Our study reports the occurrence of intermittent viraemia under
NNRTI therapy (35 of 43) in the context of PI pre-treated patients, despite
adequate concentrations of NNRTI drugs. This result suggests the interest in a
more sensitive assay to quantify HIV-1 plasma viral load before switching from
PI to NNRTI-based regimen to explain the occurrence of intermittent viraemia
during the follow up. However, one limitation of our study is the retrospective
nature of the analysis and a prospective followed cohort could be useful to
determine the clinical impact of such ongoing viraemia.
In conclusion,
these results suggest that intermittent viraemic episodes may be markers of
ongoing viraemia below 50 copies/ml and can impair the CD4 cell count recovery
under NNRTI therapy, in the context of PI-pretreated patients. The impairment
of CD4 cell count recovery seems to be affected more by the occurrence of blips
than by the level of viraemia below 50 copies/ml itself. Nevertheless, despite
the tight genetic barrier for resistance with NNRTI drugs, no virologic failure
occurred during the follow-up.
Patients and
methods
Patients
This
retrospective study was conducted in one institution. Subjects were eligible
for inclusion on the following criteria: (1) HIV-1 infection, (2) previous
HAART with PI-containing regimen at baseline, (3) three consecutive HIV-1 RNA
loads below the limit of quantification (< 50 copies/ml) with PI-containing
regimen, and (4) switching from PI- to NNRTI-containing regimen for at least
the last 6 months. All patients were more than 18 years old.
Methods
Baseline data
Baseline was
defined as the day of the switch from PI- to NNRTI-containing regimen. At
baseline, the following data were collected: age, sex, risk group,
antiretroviral therapy, CD4 cell counts and HIV-1 plasma viral load. Follow-up
data were collected 6, 12 and 18 months after NNRTI substitution: CD4 cell
counts, HIV-1 plasma viral load, combination of HAART used, reasons for
modification of HAART and clinical AIDS events.
HIV-1 RNA
assays
The standard
determination of plasma HIV-1 RNA levels was performed according to the
manufacturer's instructions (Amplicor HIV-1 Monitor Assay 1.5 ultrasensitive;
Roche, Basel, Switzerland). This assay had a detection limit of 50 copies/ml.
Intermittent
viraemia or blips were defined as a plasma HIV-1 RNA level greater than 50
copies/ml with a subsequent value of less than 50 copies/ml without evidence of
virological failure. Virological failure was defined as two consecutive plasma
HIV-1 RNA levels greater than 200 copies/ml. We compared plasma HIV-1 RNA
levels from patients with blips to those from patients with persistent
suppressed plasma viral load using a more sensitive version of the Amplicor
assay adapted to permit detection of 3 copies/ml of plasma HIV-1 RNA [8].
Determination
and interpretation of NNRTI plasma concentrations NNRTI plasma trough concentrations at
steady state were measured in a subset of patients with intermittent viraemia
at three different times: before, during and after the blips. Efavirenz (EFV)
and nevirapine (NVP) plasma trough concentrations were determined by
reverse-phase high performance liquid chromatography with ultraviolet
detection, after liquid-liquid extraction. The EFV and NVP assays were
validated over concentration ranges of 10 to 10 000 ng/ml and 50 to 10 000
ng/ml and limits of quantification were 10 and 50 ng/ml, respectively.
According to the
literature, NNRTI plasma trough concentrations were considered adequate if they
reached the following concentrations: 1100 ng/ml for EFV and 3400 ng/ml for
NVP, respectively [18-21].
Statistical
analysis
Data analysis
was performed with StatView F-4.5 software (1992-1995; Abacus Concepts, Inc,
Berkeley, California, USA). CD4 cell counts and the HIV-1 plasma viral load
changes from baseline to those at 6, 12 and 18 months after the NNRTI
substitution were compared using Wilcoxon rank test. Non-parametric
Mann-Whitney or Kruskal-Wallis tests were used to compare quantitative values.
