945
CONCISE COMMUNICATION
The Effect of Highly Active Antiretroviral Therapy on Binding
and Neutralizing Antibody Responses to Human Immunodeficiency Virus
Type 1 Infection
James M. Binley,1,a Alexandra Trkola,1,a Tom Ketas,1,a
Daryl Schiller,1,a Brian Clas,1 Susan Little,2
Douglas Richman,2 Arlene Hurley,1 Martin Markowitz,1
and John P. Moore,1,a,b
1
Aaron Diamond AIDS Research Center, Rockefeller University,
New York, New York; 2Departments of Medicine and Pathology,
University of California San Diego, and San Diego
VA Medical Center, San Diego, California
The effect on humoral immune responses of highly active antiretroviral therapy (HAART)
commenced during primary or chronic human immunodeficiency virus type 1 (HIV-1) infection was investigated. HAART inhibited the development of anti-gp120 antibodies when
initiated during primary infection and could sometimes reduce antibody titers in patients
treated within 2 years of HIV-1 infection. Conversely, antibody responses in patients infected
for several years were less sensitive to HAART. Administering HAART during primary infection usually did not substantially affect the development of weak neutralizing antibody
responses against autologous virus. However, 2 patients treated very early after infection did
not develop neutralizing responses. In contrast, 3 of 4 patients intermittently adherent to
therapy developed autologous neutralizing antibodies of unusually high titer, largely coincident with brief viremic periods. The induction of strong neutralizing antibody responses during
primary HIV-1 infection might require the suppression of virus replication by HAART, to
allow for the recovery of immune competency, followed by exposure to native envelope
glycoproteins.
Highly active antiretroviral therapy (HAART) has significantly reduced the morbidity and mortality of human immunodeficiency virus type 1 (HIV-1) infection in the developed
world, although eradication of the virus from infected patients
is unlikely [1]. How the viremia reductions caused by HAART
affect the immune system is being actively studied. Opposing
factors are involved; reducing HIV-1 replication limits the dam-
Received 14 March 2000; revised 15 May 2000; electronically published
1 August 2000.
Informed consent was obtained from patients, and human experimentation guidelines of the US Department of Health and Human Services were
followed. The studies were approved by the Rockefeller University Institutional Review Board.
Financial support: National Institutes of Health (AI-36082 to J.P.M.);
General Clinical Research Center (RR-00102 to M.M. and AI-41534 to
M.M. and J.P.M.); Center for AIDS Research (AI-36214 to S.L. and D.R.);
Research Center for AIDS and HIV Infection, San Diego VA Medical Center (S.L. and D.R.).
a
Present affiliations: Department of Microbiology and Immunology,
Weill Medical College of Cornell University, New York, New York (J.M.B.,
T.K., D.S., and J.P.M.); Division of Infectious Diseases, Department of
Internal Medicine, University Hospital, Zurich (A.T.).
b
J.P.M. is an Elizabeth Glaser Scientist of the Pediatric AIDS
Foundation.
Reprints or correspondence: Dr. John P. Moore, Weill Medical College
of Cornell University, Dept. of Microbiology and Immunology, 1300 York
Ave., W-805, New York, NY 10021 ([email protected]).
The Journal of Infectious Diseases 2000; 182:945–9
q 2000 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2000/18203-0041$02.00
age done by the virus to the immune system [1, 2], yet lowered
concentrations of viral antigens might restrict maturation or
maintenance of antiviral responses [3–6]. There have been many
reports of partial restoration of cellular immune functions in
HAART recipients, but there are few reports that address humoral immunity [1–7].
HIV-1 infection is associated with a vigorous antibody response to multiple viral antigens [8, 9]. The humoral immune
system is hyperstimulated, which leads to several abnormalities
[4, 8]. Immune destruction caused by primary infection affects
the development of effective, T helper (Th) cell–dependent immune responses [1, 2, 10]. Anti-gp120–binding antibody responses typically mature only after ∼1 year of infection; neutralizing antibody (NA) responses also develop slowly over a
similar timeframe and are usually moderate to low in titer [8, 9].
HIV-1 antigen–specific binding antibody titers and nonspecific B cell responses are both sensitive to HAART [3–5]. However, the effect of antiviral therapy on NA responses is unclear.
During zidovudine monotherapy, a modest decrease in NA titers was observed in one study [11], but another described an
increase [12]. In a third study, some patients receiving combination antiviral therapy had increased neutralization titers, but
in others, the titer levels decreased [13]. More recently, no effect
of HAART on NA titers was found [7]. The extent to which
HIV-1 replication is suppressed may influence what happens to
NA responses; the timing of therapy relative to the maturation
946
Binley et al.
of immune responses and the preservation of Th cell responses
are also probably relevant [2, 10].
Here we describe the effect of HAART initiation during primary infection on the development of autologous NA responses
and the effect of nonadherence to therapy. We also report on
how HAART affects binding antibody responses when it is first
administered during chronic infection.
Materials and Methods
Clinical specimens.
Archival plasma and peripheral blood
mononuclear cell (PBMC) samples were derived from cohorts that
have been described elsewhere [4–6, 10].
Patients acutely infected with HIV-1 and recruited within 90 days
of exposure received 1 of the following 3 drug regimens: (1) zidovudine, lamivudine, and ritonavir (patients 313#2, 313#5–
313#7, 313#8, 313#9, 313#11, and 313#12 [4–6, 10]; (2) zidovudine, lamivudine, and indinavir (patients 1304 and 1308); or (3)
zidovudine, lamivudine, abacavir, and amprenavir (patient 904).
Plasma samples from untreated acutely infected patients 1062 and
1071 were obtained from the University of California, San Diego,
Department of Medicine. Untreated, acutely infected patient AD13 has been described elsewhere [8, 9]. Patients 509#31–509#35
and 509#37–509#42, who were chronically infected with HIV-1,
were treated with zidovudine, lamivudine, and nelfinavir. Virus
loads for all patients except patients 1062 and 1071 were determined
by branched DNA assay; reverse transcription–polymerase chain
reaction was used to determine virus loads for patients 1062 and
1071
Antibody titers to p24 and gp120. ELISAs used to measure
plasma midpoint antibody titers against p24 and JR-FL gp120 are
described elsewhere [8].
Virus isolation and neutralization assay. Autologous viruses
were isolated from PBMC of infected patients on the following
days relative to the initiation of treatment (day 0): patient 313#2,
day 2157; patient 313#5, day 224; patient 313#6, day 0; patient
313#7, day 26; patient 313#8, day 21; patient 313#9, day 21;
patient 313#11, day 21; patient 313#12, day 210; patient 904,
day 218; patient 1304, day 612; and patient 1308, day 262. In the
case of the untreated control patients, virus was isolated as follows
(day 0 was the first sampling during acute infection): patient 1062,
day 0; patient 1071, day 0; and patient AD-13, day 270. The PBMC
blast assay was used to determine neutralization titers to autologous isolates [14].
Results
The humoral response to HIV-1 is suppressed when HAART
is administered during acute infection. We studied the development of humoral immunity in 7 patients who received
HAART during acute infection and whose plasma viremia was
effectively suppressed (figure 1A), compared with 3 control patients who did not receive HAART (figure 1B). Four additional
patients were intermittently adherent to therapy and underwent
detectable rebounds in plasma viremia (figure 1C). In each case,
we measured binding antibody titers to gp120 and autologous
JID 2000;182 (September)
neutralization titers (IC50 and IC90 values) against pretherapy
isolates.
All 7 patients who adhered to HAART during acute infection
seroconverted to HIV-1 antigens. Among 2 of the patients (patients 904 and 1308), anti-gp120 binding antibody titers were
extremely low after 1 year, compared with titers in untreated
control patients. These control patients were recruited significantly earlier after infection than the others, who had already
developed moderate anti-gp120 titers before receiving HAART.
However, little or no further increase in titer occurred after
HAART was commenced (figure 1A). Here we present binding
antibody data on only 3 control patients who seroconverted,
because these were the only untreated control patients from
whom autologous virus isolates were obtained during acute
infection (figure 1B). However, our general experience has been
that anti-gp120 titers in untreated people typically develop to
1: 3 3 10 4–105 [4, 8, 9]. Thus, the titers in the HAART-adherent
patients were less than those in untreated control patients and
markedly so for patients 904, 1304, and 1308.
In general, patients who adhered to treatment did not sustain
an autologous NA response. At most timepoints over the first
2 years, NA titers were IC 50 ! 1 :103 and IC 90 ! 1 :102. After ∼2
years, only patients 313#5 and 313#7 exhibited IC90 titers 1102.
Patients 904 and 1308 never developed appreciable neutralization titers. The situation was similar in the 3 untreated control
patients, although moderate titers developed in patient AD-13
toward the end of the study (figure 1B). Autologous NA responses are generally weak to moderate during untreated HIV1 infection [9]. The occasional strong response (e.g., patient
AD-6 in [9]) is exceptional.
A different pattern was seen in the intermittently adherent
HAART recipients (figure 1C). In all cases, anti-gp120 titers
were detectable when HAART was initiated. Among 3 patients
(patients 313#6, 313#8, and 313#12), increases in anti-gp120
titers were temporally associated with viremic outbursts. Concomitantly, in these 3 cases, autologous neutralization titers
developed or increased to levels that, in our experience, are
unusually high (IC 50 1 1 :104; IC 90 1 1 :103 in patient 313#12).
This was also seen, to a lesser extent, with patient 313#11 (figure
1C).
The presence of antiviral drugs in plasma can sometimes
affect neutralization assays by diminishing virus replication after entry [7]. However, at the plasma dilutions we used (>1:
50), we could measure no NA responses in several chronically
infected HAART recipients either before or after the onset of
therapy (data not shown). This, as well as the lack of neutralization titers in some drug-adherent patients (figure 1), argues against a drug-interference effect on our neutralization
assays.
Effect of HAART on anti-gp120 titers during chronic infection. In chronically infected patients with high and stable
antibody titers, HAART caused a substantial reduction in
plasma viremia, usually to !500 RNA copies/mL, with sub-
JID 2000;182 (September)
Effect of HAART on Anti–HIV-1 Antibodies
947
Figure 1. Effect of highly active antiretroviral therapy (HAART) on the development of antibody responses. Virus load (3), anti-gp120 binding
antibody titer (v), neutralizing antibody (NA) ID50 titer (.), and NA ID90 titer (M) in 15 acutely infected patients are depicted in longitudinal
profiles. A, Patients commenced HAART within 90 days of infection (day 0 is the first day of treatment) and adhered to their drug regimen.
Virus load was, almost without exception, suppressed to levels below detection during the course of the study. B, Patients did not receive antiviral
therapy. Day 0 corresponds to the first clinical visit following infection. C, Patients commenced HAART within 90 days of infection (day 0 is
the first day of treatment) but were intermittently adherent to therapy. Their virus load was therefore only partially suppressed. bDNA, branched
DNA; HIV-1, human immunodeficiency virus type 1.
sequent occasional viremic episodes in some cases (figure 2).
In general, suppression of virus production was not associated
with a rapid decline in anti-gp120 or anti-p24 titers. In 3 patients (patients 509#34, 509#37, and 509#41), there was a gradual decline in antibody titers. Of note is that 2 of these patients
had been infected for only ∼1 year before receiving HAART.
In the other 7 cases, there was little or no change in anti-gp120
titers. Among 2 patients (patients 509#32 and 509#31), a modest increase in anti-p24 titer occurred, perhaps an indirect reflection of a partial restoration of Th cell function and/or
antigen production due to a viremic episode. The lack of relationship between plasma viremia and anti-gp120 and antip24 titers is exemplified by patients 509#33 and 509#40, for
whom pretherapy plasma samples were available. Both patients
maintained a very high anti-gp120 titer after HAART. However,
autologous NA titers in patient 509#33 were both very low
and unaffected by HAART (data not shown).
Discussion
We previously studied the effect of HAART on binding antibody titers to HIV-1 antigens [4–6]. Here we monitor the
generation and maintenance of NAs against autologous virus.
Commencing HAART during acute infection did not significantly improve the development of NAs in adherent patients.
Although HAART may allow for recovery of immune function
[2], virus antigen reduction may become limiting under these
conditions [4–6]. This could be especially important if NAs are
raised against the native envelope glycoprotein complexes on
virions, which may not be abundant during HAART. The inhibitory effect of antigen limitation on humoral immunity is
documented by the gradual decline in anti-gp120 titers that
often occurs when HAART is given within the first 1–2 years
of infection (figures 1 and 2) [4, 5]. In contrast, when HAART
is initiated several years after infection, it usually has very little
948
Binley et al.
JID 2000;182 (September)
Figure 2. Effect of highly active antiretroviral therapy (HAART) on antibody responses during established infection. Virus load (3), antigp120 titer (v), and anti-p24 titer (m) in chronically infected patients are depicted longitudinally. Each patient had been infected for 1–5 years,
as indicated, before initiation of HAART on day 0. Pretherapy data are included for patients 509#33 and 509#40. bDNA, branched DNA.
effect on antibody titers to HIV-1 structural proteins (figure 2).
By this time, there may be sufficient antigen deposits in lymphoid tissues to maintain antibody production, so that a decline
in the amount of circulating antigen has little impact.
We have suggested that anti-envelope antibody responses
may not be homogeneous, because NA responses to virionassociated gp120-gp41 complexes might be more dependent on
T cell help than are binding antibody responses to monomeric
gp120 [8]. The increased anti-p24 titers sometimes observed in
chronically infected HAART recipients whose viremia is well
suppressed demonstrate that the maintenance of immune responses can require very little circulating antigen (figure 2).
Anti-Gag responses are considered to be dependent on intact
Th cell functions [8].
Patients who commenced HAART during acute infection but
then became intermittently nonadherent developed unusually
strong autologous NA responses [5, 6, 10]. We studied too few
such patients for our observations to be other than anecdotal,
but a possible explanation is that effective priming of a humoral
response occurred before the use of HAART, followed by some
recovery of immune competency when viremia was suppressed
[2]. The restoration of antigenemia during a period of nonadherence might then boost both the anti-gp120 and NA titers.
Although viremic episodes can also stimulate Gag-specific T
proliferative responses [10], we could find no direct and consistent correlation between these responses and the generation
of NAs in either the adherent or intermittent HAART recipients. For example, although patient 313#8 (previously reported
as patient A40T [10]) consistently had a relatively high Gagspecific proliferative response, his NA titers were the weakest
of the 4 intermittently adherent cases. Furthermore, the rapid
NA titer increase in patient 313#12 at around day 300 occurred
when proliferative responses to Gag were undetectable [10].
However, the absence of Gag-specific proliferative responses
does not rule out the possibility that Th cell responses to other
viral antigens could be assisting in generating NAs in a similar
JID 2000;182 (September)
Effect of HAART on Anti–HIV-1 Antibodies
way that a period of immune recovery during HAART facilitates responses to neo-antigens, as demonstrated by improved
antibody responses to non-HIV immunogens [15].
Cytotoxic T cell responses are also boosted by antigen production during short periods of nonadherence [6]. Although
HAART controls virus load, HIV-1 antigen-specific responses
can diminish or fail to be generated without a supply of viral
antigen. Since HIV-1 cannot be eradicated from infected patients by HAART, it is now imperative that we learn how to
harness the power of the immune system to control viral replication [1]. The clinical utility of structured treatment interruptions is presently under investigation [6]. Exogenous vaccination may be useful to boost immune responses in HAART
recipients, although this will probably require the development
of better immunogens than those presently available, given the
failure of vaccination to control viremia in patients studied to
date. How structured therapy interruptions or vaccination during HAART can elicit effective immune responses depends on
how reliably these responses can be elicited, their longevity, and
their efficacy at suppressing virus load for prolonged periods,
factors which all remain to be determined.
Acknowledgment
4.
5.
6.
7.
8.
9.
10.
11.
12.
We thank Lynn Morris for useful discussions.
13.
References
1. Gea-Banacloche JC, Lane HC. Immune reconstitution in HIV-1 infection.
AIDS 1999; 13:S25–38.
2. Rosenberg ES, Billingsley JM, Caliendo AM, et al. Vigorous HIV-1–specific
CD41 T cell responses associated with control of viremia. Science 1997;
278:1447–50.
3. Lafeuillade A, Poggi C, Tamalet C, Profizi N, Tourres C, Costes O. Effects
14.
15.
949
of combination of zidovudine, didanosine, and lamivudine on primary
human immunodeficiency virus type 1 infection. J Infect Dis 1997; 175:
1051–5.
Morris L, Binley JM, Clas BA, et al. HIV-1 antigen–specific and nonspecific
B cell responses are sensitive to combination antiretroviral therapy. J Exp
Med 1998; 188:233–45.
Markowitz M, Vesanen M, Tenner-Racz K, et al. The effect of commencing
combination antiretroviral therapy soon after human immunodeficiency
virus type 1 infection on viral replication and antiviral immune responses.
J Infect Dis 1999; 179:527–37.
Ortiz GM, Nixon DF, Trkola A, et al. HIV-1–specific immune responses in
subjects who temporarily contain virus replication after discontinuation
of HAART. J Clin Invest 1999; 104:R13–8.
Dreyer K, Kallas EG, Planelles V, et al. Primary isolate neutralization by
HIV type 1–infected patient sera in the era of highly active antiretroviral
therapy. AIDS Res Hum Retroviruses 1999; 15:1563–71.
Binley JM, Klasse PJ, Cao Y, et al. Differential regulation of the antibody
responses to gag and env proteins of human immunodeficiency virus type
1. J Virol 1997; 71:2799–809.
Moore JP, Cao Y, Ho DD, Koup RA. Development of the anti-gp120 antibody response during seroconversion to human immunodeficiency virus
type 1. J Virol 1994; 68:5142–55.
Binley JM, Schiller DS, Ortiz GM, et al. The relationship between T proliferative responses and plasma viremia during treatment of human immunodeficiency virus type 1 infection with combination antiviral therapy.
J Infect Dis 2000; 181:1249–63.
Wainberg MA, Drodopoulos WC, Salomon H, et al. Enhanced fidelity of 3TCselected mutant HIV-1 reverse transcriptase. Science 1996; 271:1282–5.
Schmidtmayerova H, Lackovicova M, Stankova M, et al. Virus neutralizing
antibodies at different stages of the HIV disease: increased levels after
azidothymidine treatment. Acta Virol 1992; 36:157–65.
Sarmati L, Nicastri E, el-Sawaf G, et al. Increase in neutralizing antibody
titer against sequential autologous HIV-1 isolates after 16 weeks saquinavir (Invirase) treatment. J Med Virol 1997; 53:313–8.
Trkola A, Matthews J, Gordon C, Ketas T, Moore JP. A cell line–based assay
for primary human immunodeficiency virus type 1 isolates that use either
the CCR5 or the CXCR4 coreceptor. J Virol 1999; 73:8966–74.
Kroon FP, Rimmelzwaan GF, Roos MT, et al. Restored humoral immune
response to influenza vaccination in HIV-infected adults treated with
highly active antiretroviral therapy. AIDS 1998; 12:F217–23.