665
Longitudinal Analysis of CD4 T Cell Counts, T Cell Reactivity, and Human
Immunodeficiency Virus Type 1 RNA Levels in Persons Remaining AIDS-Free
despite CD4 Cell Counts õ200 for ú5 Years
I. P. M. Keet, M. Janssen, P. J. Veugelers, F. Miedema,
M. R. Klein, J. Goudsmit, R. A. Coutinho, and F. de Wolf
Division of Public Health, Municipal Health Service, Central
Laboratory of the Netherlands Red Cross Blood Transfusion Service,
and Department of Virology, Human Retrovirus Laboratory, University
of Amsterdam, Amsterdam, Netherlands
CD4 cell count, T cell reactivity, and human immunodeficiency virus type 1 (HIV-1) RNA load
were examined in 10 persons remaining free of AIDS while CD4 cell counts were õ200/mm3 for
ú5 years, so-called low CD4 cell count long-term survivors (low CD4 LTS), and 10 matched rapid
and 10 matched intermediate progressors (AIDS within 1 and 1 – 3 years). Longitudinal analysis of
CD4 cell count and T cell reactivity revealed slower declines in low CD4 LTS. HIV-1 RNA load
was 1 log/mL lower in the low CD4 LTS compared with the rapid progressors, with a similar linear
increase over time in low CD4 LTS and progressors of 0.1 log/year. There was considerable overlap
between groups. The constant difference between low CD4 LTS and progressors probably reflects
HIV RNA load early after infection, but low CD4 LTS and progressors show similar slopes of HIV1 RNA load rise.
Several recent studies suggest that human immunodeficiency
virus (HIV) load is a very early predictor of disease progression
and that virus load in blood remains remarkably stable over
time during asymptomatic HIV-1 infection, the so-called set
point hypothesis [1 – 7]. In addition, the use of virus load has
been recommended as a surrogate marker to assess the need
for antiretroviral treatment or for changes therein [8 – 10].
Whether the hypothesis of stability of virus load, described as
a steady state [1], holds in prolonged asymptomatic HIV-1
infection remains to be determined. We performed a longitudinal analysis of virus load in low CD4 cell count long-term
survivors (low CD4 LTS), persons who remain AIDS-free in
spite of low CD4 cell counts. These low CD4 LTS were compared with matched progressor controls. To relate longitudinal
patterns of virus load with markers of immune deficiency, we
first analyzed patterns of CD4 cell decline and loss of T cell
reactivity over time in the same groups.
Many different and sometimes mutually exclusive case
definitions have been used for LTS [11]. These definitions have
in common that they identify persons with an exceptionally
Received 12 November 1996; revised 3 April 1997.
Presented in part: XI International Conference on AIDS, Vancouver, Canada,
July 1996.
Written informed consent was obtained from all participants of the Amsterdam cohort study.
Financial support: Advisory Council on Health Research, Programme Committee for AIDS Research, Ministry of Public Health (118775), and Netherlands
Foundation of Preventive Medicine (28-2370).
Reprints or correspondence: Dr. Ireneus P. M. Keet, Municipal Health Service, Dept. of Public Health and Environment, P.O. Box 20244, 1000 HE
Amsterdam, Netherlands.
The Journal of Infectious Diseases 1997;176:665–71
q 1997 by The University of Chicago. All rights reserved.
0022–1899/97/7603–0016$02.00
/ 9d31$$se09
07-16-97 22:27:01
jinfa
benign course of HIV-1 infection in various stages of HIV-1
infection. Three categories of definitions can be distinguished:
lack of clinical and immunologic progression [12, 13], longterm survival after clinical progression [14], and lack of clinical
progression in spite of a low CD4 cell count. The present study
analyzes the third category and uses the case definition for low
CD4 LTS proposed by Schrager et al. [11]: persons who have
remained free of clinical AIDS for §5 years after a decrease
in CD4 cell count to õ200 1 106/L.
The analysis of this subgroup may help to identify factors
that sustain immune function in spite of low CD4 cell counts
and help to identify correlates of arrested disease progression.
In a previous study, we reported that independent predictors
of AIDS-free survival following a CD4 cell decline below
200 1 106/L are absence of syncytium-inducing (SI) variants,
absence of HIV p24 antigen, and relatively preserved T cell
reactivity after stimulation with phytohemagglutinin [15].
Subjects and Methods
Study population. The study subjects were all homosexual men
who participated in the Amsterdam cohort study on HIV infection
among homosexual men, which was started in October 1984 [16].
Participants were seen every 3 months. For the present study, we
identified a subset of HIV-infected participants who had had at
least one CD4 cell count ú200/mm3 after entry or seroconversion
and who subsequently had a CD4 cell count õ200/mm3 at least
two times within 12 months. CD4 cell counts ú200/mm3 between
these two visits were allowed. The time of the first CD4 cell count
õ200/mm3 has been designated as time zero in the analysis. In a
nested case-control analysis, we compared men with an AIDS-free
survival according to the 1987 Centers for Disease Control AIDS
case definition of ú5 years (low CD4 LTS) with men who developed AIDS within 1–3 years (intermediate progressors, IP) and
within 1 year after time zero (rapid progressors, RP). Low CD4
UC: J Infect
666
Keet et al.
LTS, RP, and IP were matched for early treatment, defined as
antiretroviral treatment that was started before a clinical AIDS
diagnosis. Early treatment was started in the Amsterdam cohort
study in 1991 in persons with CD4 cell counts õ200/mm3. Because
of the differences in duration of follow-up, matching for early
treatment was performed as follows: low CD4 LTS who never
used early antiretroviral treatment could be matched only with IP
and RP who likewise never received early antiretroviral treatment.
Low CD4 LTS who started early treatment were preferably
matched with IP and RP who started treatment at the same moment
before or after time zero within a maximum interval of 6 months.
If such controls were not available, low CD4 LTS were matched
with IP and RP who developed AIDS before the low CD4 LTS
started early treatment. The same procedure was used to match
for primary prophylaxis against Pneumocystis carinii pneumonia
with either cotrimoxazole or pentamidine, which was started in
1990. In addition, low CD4 LTS and controls were matched for
age at time zero and calendar year of first CD4 cell count õ200
1 106/L.
Laboratory methods. Stored serum samples frozen at 0707C
were used for measurements of HIV-1 RNA load with a quantitative RNA assay (NASBA; Organon, Oss, Netherlands) [17–19].
Results are expressed as number of HIV RNA copies per milliliter
and were log-transformed for analysis. The quantitation limit of
this assay is 1000 RNA copies/mL, and sera with undetectable
levels were assumed to have 1000 copies/mL in the analysis. Sera
were tested for the presence of antibodies to HIV-1 with a commercially available EIA (Abbott, Abbott Park, IL) and confirmed with
Western blotting.
T cell subsets were tested with fresh cells. CD4 and CD8 T cells
were enumerated by direct immunofluorescence using monoclonal
antibodies and a flow cytometry system. Cocultivation of cryopreserved peripheral blood mononuclear cells with MT-2 cells was
used to detect the presence of SI HIV-1 variants [20, 21]. To
measure T cell reactivity, fresh cells were stimulated in a whole
blood culture system with CD3 monoclonal antibodies [22–24].
Reactivity is expressed in counts per minute per well. HIV p24
antigen was detected in fresh sera with a solid-phase, sandwich
type immunoassay (Abbott).
Statistics. We studied CD4 cell count, T cell function, and
HIV load over time by use of the proc Mixed module of SAS
(SAS Institute, Cary, NC) in three separate analyses. To meet the
criteria of multivariate normality, these data were log-transformed.
Independent variables were group (low CD4 LTS, IP, or RP), years
since time zero, and matching indicator.
Two potential covariance structures between the observations
were explored: compound symmetry, which is characterized by a
constant correlation between observations from the same person
over time, and an autoregressive structure assuming a decrease in
the correlation with longer time periods. The covariance structure
resulting in the highest likelihood was selected. Second, we analyzed the presence of an interaction between time and group. In
the absence of an interaction, we examined whether there was an
overall time effect. In case a time effect was detected, we analyzed
with the likelihood ratio test whether this was best described by
a linear, second-order, or third-order model. All data were used to
define the model. Subsequently, outlying observations were identified by visual inspection of the data, by use of two criteria. First,
observations that were outliers from the model were selected. Sec-
/ 9d31$$se09
07-16-97 22:27:01
jinfa
JID 1997;176 (September)
ond, we explored whether these observations were also outliers
within the individual. If the latter was the case, the observations
were removed; if not, all observations of the individual remained
in the model.
Individual virus load trajectories of the low CD4 LTS were
analyzed by use of linear regression. Finally, the Pearson correlation coefficient (r) was calculated. The square of the correlation
coefficient (r2) represents the proportion of variation of a dependent
variable that is explained by an independent variable, the so-called
explained proportion. Here, we use it to analyze which proportion
of virus load variation is explained by variation in CD4 cell count
and T cell function.
Results
Characteristics at Baseline and Follow-Up
In 186 men (mean age, 38 years), CD4 cell counts dropped
below 200/mm3 during follow-up. The probability of remaining
AIDS-free ú5 years was 16% (95% confidence interval, 10% –
23%), and 10 men were identified as low CD4 LTS. Subsequently, each low CD4 LTS was successfully matched with 1
IP and 1 RP. The mean time between the first HIV-positive
visit and the first CD4 cell count õ200/mm3 was 3.1 years in
the low CD4 LTS, 3.0 years in the IP, and 3.0 years in the
RP. SI variants were present at time zero in 0 of 10 low CD4
LTS, 4 of 10 IP, and 5 of 9 RP. Two low CD4 LTS, 6 IP, and
4 RP were positive for HIV p24 antigen. Five years after time
zero, SI variants were present in 6 low CD4 LTS. Early antiretroviral treatment was given to 8 low CD4 LTS: zidovudine
monotherapy, which was later replaced in 2 men by zidovudine
plus didanosine and zidovudine plus zalcitabine. Two of the
IP and none of the RP started early treatment. P. carinii pneumonia prophylaxis was given to 8 low CD4 LTS (cotrimoxazole, 6; pentamidine, 2), to 1 IP (pentamidine), and to none of
the RP. These differences were expected because of the longer
duration of follow-up in the low CD4 LTS and were accounted
for in the matching procedure.
Immunologic Parameters
CD4 cell count. For modeling of the CD4 cell count in
low CD4 LTS, IP, and RP, 728 values could be used. Table 1
shows the group medians and interquartile ranges of the CD4
cell counts around time zero in low CD4 LTS and progressors
and at year /5 in the low CD4 LTS. Note that the individual
CD4 cell count at a given time point is the mean CD4 cell
count in the 6-month interval around that time point (3 months
before until 3 months after the time point). This explains why
the median CD4 cell count at time zero can be ú200.
Figure 1 is a graphic presentation of the course of the crude
log-transformed CD4 cell counts over the entire follow-up period in the low CD4 LTS, IP, and RP. These data were used for
modeling the CD4 cell counts in each group. Fifteen outlying
observations (2%) were removed before the final analysis was
UC: J Infect
JID 1997;176 (September)
Low CD4 Cell Count Long-Term Survivors
667
Table 1. Median values and interquartile ranges of CD4 cell count, T cell reactivity, and HIV-1 RNA load at different time points in longterm survivors with CD4 cell counts õ200/mm3 (low CD4 LTS), intermediate progressors (IP), and rapid progressors (RP).
CD4 cell count
Year
01
0
/0.5
/1
/5
T cell reactivity
HIV RNA load
LTS
IP
RP
LTS
IP
RP
LTS
IP
RP
315
(280 – 404)
220
(171 – 240)
247
(196 – 273)
210
(158 – 260)
140
(45 – 210)
372
(323 – 528)
188
(173 – 206)
162
(106 – 203)
116
(85 – 158)
393
(299 – 528)
177
(148 – 215)
80
(50 – 138)
*
1316
(350 – 3250)
490
(210 – 2380)
450
(210 – 1050)
308
(64 – 448)
1267
(420 – 2650)
378
(219 – 605)
110
(80 – 180)
8535
(2500 – 32,500)
16,862
(7500 – 64,000)
17,673
(9000 – 70,000)
23,367
(11,325 – 75,000)
49,295
(11,750 – 305,000)
58,023
(27,250 – 155,500)
85,016
(43,000 – 93,500)
56,924
(34,000 – 101,500)
103,514
(64,500 – 175,000)
75,701
(24,500 – 395,000)
82,567
(16,000 – 405,000)
340,017
(124,250 – 725,000)
3003
(630 – 3325)
1820
(1404 – 2757)
1560
(400 – 1960)
1407
(305 – 2250)
NOTE. Year: no. of years before or after initial CD4 cell decline below 200/mm3. T cell reactivity was measured after stimulation with anti-CD3 antibodies.
Individual CD4 cell count at given time point is mean CD4 cell count in 6-month interval around that time point (3 months before until 3 months after); thus,
median CD4 cell count at time zero can be ú200.
* Only 2 observations were available for T cell reactivity in LTS at year 01.
done. The autoregressive covariance structure was found to be
most appropriate for these data. CD4 cell decline in the low
CD4 LTS can best be described by a linear function with a
mean decline of 0.08 log/year (figure 2). The CD4 cell decline
in progressors is accelerated and can best be described by a
second-order function for IP and a third-order function for RP.
The correlation coefficient (r) between CD4 cell counts and
HIV-1 RNA load was .1760 (P Å .003), resulting in an explained proportion (r2) of .03.
T cell reactivity. The raw data on T cell reactivity, of which
450 counts were available, show a decline in T cell reactivity
in the groups (table 1). Six extreme observations were removed
(1.3%). The autoregressive covariance structure was most appropriate for modeling the T cell reactivity data. The low CD4
LTS had a decline of 0.16, IP of 0.46, and RP of 0.64 log cpm/
year (figure 3).
The correlation between T cell reactivity and HIV-1 RNA
load was 00.5387 (P õ .001), resulting in an explained proportion of .29.
HIV-1 RNA Load
HIV-1 RNA load was measured at 6-monthly intervals in
the time interval between 02 and /2 years and at 12-monthly
Figure 1. Group profiles of mean
log-transformed CD4 cell count as
observed in years preceding and following CD4 cell count decline below
200 1 106/L.
/ 9d31$$se09
07-16-97 22:27:01
jinfa
UC: J Infect
668
Keet et al.
JID 1997;176 (September)
Figure 2. Predicted CD4 cell
count: modeling of CD4 cell count
as dependent variable of time in
years since CD4 cell count decline
below 200 1 106/L. Years 02 to 06
are years preceding; years 2 – 8 are
years following CD4 cell count decline below 200 1 106/L.
intervals in the interval 07 to 02 and /2 to /5 years, provided
sera were available. In total, 277 samples were tested for this
study, of which the group medians around time zero and at
year 5 are presented in table 1 and figure 4. The correlation
between the individual observations over time was best described by the compound symmetry structure. Six outlying
observations (2%) were removed before the final analysis was
done. There was a significant effect of time (P õ .001) and
group (P õ .001) but no interaction effect between group and
time (P Å .77), implying that the group effect is constant over
time. Hierarchic testing of the three time functions showed that
the fit of a linear model was better than that of a second- or
third-order function. Therefore, we used linear modeling (figure
5). The constant difference between RP and low CD4 LTS in
log load was 0.96. The intercept (the estimated load at time
zero) was 4.12 in the low CD4 LTS, 4.79 in the IP, and 5.08
Figure 3. Modeling of T cell reactivity after stimulation with monoclonal antibodies against CD3 as dependent variable of time in years
since CD4 cell count decline below
200 1 106/L. Years 01 through 03
are years preceding; years 2 – 8 are
years following CD4 cell count decline below 200 1 106/L.
/ 9d31$$se09
07-16-97 22:27:01
jinfa
UC: J Infect
JID 1997;176 (September)
Low CD4 Cell Count Long-Term Survivors
669
Figure 4. Group profiles of observed virus load: mean log-transformed HIV-1 RNA levels as observed in years preceding and
following CD4 cell count decline below 200 1 106/L.
in the RP. The differences in load between RP and low CD4
LTS (P õ .001) and between IP and low CD4 LTS (P õ .001)
were clearly larger than between the 2 groups of progressors
(P Å .12). The slope of load increase over time was the same in
the 3 groups: 0.10 log HIV-1 RNA copies/mL/year. Individual
trajectories of virus load in the low CD4 LTS were analyzed
with linear regression. The median value was /0.15 log HIV1 RNA copies/mL/year, with a range of 00.13 to /0.33. Nine
of the 10 low CD4 LTS had a positive virus load slope over
time.
Discussion
We explored longitudinal differences between HIV-infected
low CD4 LTS and matched progressors regarding HIV-1 RNA
load, CD4 cell count, and T cell reactivity. We analyzed quantitative and qualitative differences, the first referring to different
rates of increase or decline and the latter to different trends:
linear or higher-order functions. Only quantitative differences
were found for HIV-1 RNA load, which was low but not stable
in the low CD4 LTS. Moreover, no indication for a differential
Figure 5. Predicted virus load:
modeling of HIV-1 RNA levels as
dependent variable of time in years
since CD4 cell count decline below
200 1 106/L. Years 02 to 06 are
years preceding; years 2 – 6 are years
following CD4 cell count decline below 200 1 106/L.
/ 9d31$$se09
07-16-97 22:27:01
jinfa
UC: J Infect
670
Keet et al.
increase between low CD4 LTS and progressors was found.
The group medians and interquartile ranges showed considerable overlap between low CD4 LTS and progressors regarding
the level of HIV-1 RNA load. However, the real difference
between low CD4 LTS and progressors may have been underestimated in this study because of the detection limit of the
assay. Differences in HIV-1 RNA load between low CD4 LTS
and progressors already appear early after infection, but these
data clearly show that load does not remain stable throughout
the asymptomatic stage of HIV-1 infection, as has been suggested in the set point hypothesis [1 – 7, 25]. The increase in
load over time was similar in the 3 groups, resulting in a delay
of 6 – 10 years before the low CD4 LTS reached the baseline
levels of load of the progressors. The increase in load over
time reported in this study may have been preceded by a longer
interval of a stable increasing or even decreasing load following
seroconversion, which needs further exploration.
With respect to the CD4 cell profiles, both quantitative and
qualitative differences were found between groups. In the low
CD4 LTS, the CD4 cell decline following and preceding the
CD4 cell drop below 200 is best described by a linear function.
However, in the RP and IP, the CD4 cell decrease shows a
marked acceleration as the CD4 cell count comes close to 200.
Regarding T cell reactivity, there was a clear difference in
the rate of decline in the 3 groups, with the RP and IP showing
a much faster drop in T cell reactivity when the CD4 cell
counts fell below 200/mm3. A better discrimination can be
expected in a longitudinal analysis of T cell reactivity after
stimulation with phytohemagglutinin [15]. However, because
of a change in the method to measure T cell reactivity after
phytohemagglutinin stimulation in 1992, this analysis was not
possible. The removal of outlying observations had no impact
on the estimator of the virus load rise or CD4 cell count and
T cell function decrease (data not shown).
Comparison of the correlation coefficients shows that the
association between HIV-1 RNA load and T cell reactivity is
stronger than between load and CD4 cell count in this population with low CD4 cell counts. The lower correlation with CD4
cell count may be explained by the fact that fluctuation in HIV1 RNA load is not associated with short-term changes in CD4
cell count [26]. Apparently, HIV-1 RNA load is a better predictor of T cell reactivity than of CD4 count at a given time
point.
There were several arguments to use repeated-measurements modeling as opposed to more conventional methods in
the present study. First, a crude analysis of the data at different
time points would have provided much less information. Statistical inference would have been hampered by multiple testing in a small data set. In contrast, the modeling in the present
study markedly raised the power by examining longitudinal
patterns for each marker in a single analysis. Also, these
analyses provide insight into qualitative differences of temporal changes. Use of the conventional least-square regression
analysis in this respect would have been inappropriate, since
/ 9d31$$se09
07-16-97 22:27:01
jinfa
JID 1997;176 (September)
persons differ with respect to the number of observations.
Here, the benefits of the application of repeated measurements
were shown. Other markers and determinants can be studied
as well with this analysis, such as the role of cytotoxic lymphocyte response [27].
Nevertheless, the results have to be interpreted with caution
because of the small numbers, and confirmation in other and
larger studies is necessary.
The results of the analysis show that the 3 groups are strongly
discriminated by different trends of CD4 cell decline. In contrast, analysis of HIV-1 RNA load reveals a similar trend of
linear increase of load over time in the 3 groups. Presumably,
HIV-1 RNA load is discriminatory throughout HIV-1 infection,
with a constant difference between low CD4 LTS and progressors.
In summary, we showed that HIV load in low CD4 LTS is
low in comparison with that in progressors. However, virus
load goes up over time in this group, and there is an ongoing
decrease in CD4 cell count and T cell reactivity. In addition,
SI variants appeared in 6 of 10 low CD4 LTS. Progression in
this group appears to be delayed but not arrested. Therefore,
as in all persons with low CD4 cell counts, appropriate antiretroviral and prophylactic treatment is necessary for these persons.
Acknowledgments
We thank Margreet Bakker for HIV RNA load testing, Ronald
Geskus, Erik van Ameijden, and Maria Prins for critical reading
of the manuscript and statistical advice, and Ethel Boucher for
data management.
References
1. Ho DD. Viral counts count in HIV infection. Science 1996; 272:1124 – 5.
2. Mellors JW, Rinaldo CR Jr, Gupta P, White RM, Todd JA, Kingsley LA.
Prognosis in HIV-1 infection predicted by the quantity of virus in
plasma. Science 1996; 272:1167 – 70.
3. Mellors JW, Kingsley LA, Rinaldo CR Jr, et al. Quantitation of HIV-1
RNA in plasma predicts outcome after seroconversion. Ann Intern Med
1995; 122:573 – 9.
4. O’Brien TR, Blattner WA, Waters D, et al. Serum HIV-1 RNA levels and
time to development of AIDS in the Multicenter Hemophilia Cohort
Study. JAMA 1996; 276:105 – 10.
5. Jurriaans S, Van Gemen B, Weverling GJ, et al. The natural history of HIV1 infection: virus load and virus phenotype independent determinants of
clinical course? Virology 1994; 204:223 – 33.
6. Goudsmit J. The role of viral diversity in HIV pathogenesis. J Acquir
Immune Defic Syndr Hum Retrovirol 1995; 10(suppl 1):S15 – 9.
7. Daar ES, Chernyavskiy T, Zhao JQ, Krogstad P, Chen ISY, Zack JA.
Sequential determination of viral load and phenotype in human immunodeficiency virus type 1 infection. AIDS Res Hum Retroviruses 1995;
11:3 – 9.
8. Saag MS, Holodniy M, Kuritzkes DR, et al. HIV viral load markers in
clinical practice. Nature Med 1996; 2:625 – 9.
9. Harrigan R. Measuring viral load in the clinical setting. J Acquir Immune
Defic Syndr Hum Retrovirol 1995; 10(suppl 1):S34 – 40.
10. Steele FR. Viral load: to treat or not to treat? Nature Med 1995; 1:980.
UC: J Infect
JID 1997;176 (September)
Low CD4 Cell Count Long-Term Survivors
11. Schrager LK, Young JM, Fowler MG, Mathieson BJ, Vermund SH. Longterm survivors of HIV-1 infection: definitions and research challenges.
AIDS 1994; 8(suppl 1):S95 – 108.
12. Buchbinder SP, Katz MH, Hessol NA, O’Malley PM, Holmberg SD. Longterm HIV-1 infection without immunologic progression. AIDS 1994; 8:
1123 – 8.
13. Keet IPM, Krol A, Klein MR, et al. Characteristics of long-term asymptomatic infection with human immunodeficiency virus type 1 in men with
normal and low CD4 cell counts. J Infect Dis 1994; 169:1236 – 43.
14. Hardy AM. Characterization of long-term survivors of acquired immunodeficiency syndrome. J Acquir Immune Defic Syndr 1991; 4:386 – 91.
15. Keet IPM, Krol A, Koot M, et al. Predictors of disease progression in
HIV-infected homosexual men with CD4/ cell õ200 1 106/L but free
of AIDS defining clinical disease. AIDS 1994; 8:1577 – 83.
16. de Wolf F, Lange JMA, Houweling JTM, et al. Numbers of CD4/ cells
and the levels of core antigens of and antibodies to the human immunodeficiency virus as predictors of AIDS among seropositive homosexual
men. J Infect Dis 1988; 158:615 – 22.
17. Van Gemen B, van Beuningen R, Nabbe A, et al. A one-tube quantitative
HIV-1 RNA NASBA nucleic acid amplification assay using electrochemiluminescent (ECL) labelled probes. J Virol Methods 1994;49:157–67.
18. Van Gemen B, Van de Wiel P, van Beuningen R, et al. The one-tube
quantitative HIV-1 RNA NASBA: precision, accuracy, and application.
PCR Methods Applications 1995; 4:S177 – 84.
19. Van Gemen B, Kievits T, Nara P, et al. Qualitative and quantitative detection of HIV-1 RNA by nucleic acid sequence-based amplification. AIDS
1993; (suppl 2):S107 – 10.
/ 9d31$$se09
07-16-97 22:27:01
jinfa
671
20. Koot M, Vos AH, Keet IPM, et al. HIV-1 biological phenotype in longterm infected individuals evaluated with an MT-2 cocultivation assay.
AIDS 1992; 6:49 – 54.
21. Koot M, Keet IPM, Vos AH, et al. Prognostic value of HIV-1 syncytiuminducing phenotype for rate of CD4/ cell depletion and progression to
AIDS. Ann Intern Med 1993; 118:681 – 8.
22. Van Lier RA, Boot JH, Verhoeven AJ, De Groot ER, Brouwer M, Aarden
LA. Functional studies with anti-CD3 heavy chain isotype switch-variant monoclonal antibodies. J Immunol 1987; 139:2873 – 9.
23. Schellekens PTA, Roos MTL, de Wolf F, Lange JMA, Miedema F. Low
T-cell responsiveness to activation via CD3/TCR is a prognostic marker
for acquired immunodeficiency syndrome (AIDS) in human immunodeficiency virus-1 infected men. J Clin Immunol 1990; 10:121 – 7.
24. Roos MTL, Miedema F, Koot M, et al. T cell function in vitro is an
independent progression marker for AIDS in human immunodeficiency
virus – infected asymptomatic subjects. J Infect Dis 1995; 171:531 – 6.
25. Piatak M Jr, Saag MS, Yang LC, et al. Determination of plasma viral
load in HIV-1 infection by quantitative competitive polymerase chain
reaction. AIDS 1993; 7(suppl 2):S65 – 71.
26. Paul DB, Kuhns MC, McNamara AL, Pottage JC Jr, Spear GT. Shortterm stability of HIV provirus levels in the peripheral blood of HIVinfected individuals. J Med Virol 1995; 47:292 – 7.
27. Klein MR, van Baalen CA, Holwerda AM, et al. Kinetics of Gag-specific
cytotoxic T lymphocyte responses during the clinical course of HIV-1
infection: a longitudinal analysis of rapid progressors and long-term
asymptomatics. J Exp Med 1995; 181:1365 – 72.
UC: J Infect