correlations of soluble interleukin-2 and tumor necrosis factor type ii receptors with immunologic...

Correlations of Soluble Interleukin-2 and Tumor Necrosis FactorType II Receptors with Immunologic and Virologic Responsesunder HAART

F. BONNET,1,4 M. SAVES,2 PH. MORLAT,1 C. DROZ,2 G. CHENE,2 E. PEUCHANT,3 N. BERNARD,1

D. LACOSTE,1 R. SALAMON,2 and J. BEYLOT1

Accepted: November 7, 2001

We assessed the correlations between some plasma markers ofimmune activation (soluble receptors of interleukin 2 (sIL2-R)and TNF�p75 (sTNFII-R) and usual markers of HIV infectionin patients treated with protease-inhibitors (PI). Forty-six PI-naive HIV-1-infected adults were included in a 1-year prospec-tive cohort from the initiation of a PI-containing regimen (M0).Measurements of CD4�cell count, plasma HIV-RNA, sIL2-Rand sTNFII-R were performed at M0, M6, and M12. Theevolution of sIL2-R from baseline to M12 was significantlydifferent between immunological responders (IR) (CD4�countabove 200/mm3 for subject having less than 200 CD4�/mm3

at inclusion, or increase of at least 50 CD4�/mm3 for others)(58 UI/ml) and non-IR (�28 UI/ml) (P �0.01). Theevolutionof sTNFII-R between M0 and M12 was significantly differentbetween virological responders (VR) (plasma HIV-1 RNA lessthan 500 copies/ml at M12) (�2.5 ng/ml) and non-VR (�0.2ng/ml) (P �0.02). Ourstudy shows significative correlationsbetween the evolutions of soluble interleukin-2 and TNFR-IIreceptors and those of CD4�T-lymphocytes or HIV-RNAresponses in patients under HAART.

KEY WORDS: HIV; marker of immune activation; protease inhibitor;soluble tumor necrosis factor� receptor; soluble interleukin-2 receptor.

INTRODUCTION

Human immunodeficiency virus (HIV) RNA andCD4�cell count are highly predictive of progression

to acquired immune deficiency syndrome (AIDS) ordeath in HIV-infected patients (1, 2). However, vari-ations of these markers do not explain all variations ofdisease progression, and the value of additive markersof immune activation has to be considered. Previousstudies have reported elevated levels of plasma im-mune activation markers, soluble interleukin-2 recep-tor (sILR-2) and soluble tumor necrosis factor� typeII receptors (sTNFR-II), during the course of infectionwith HIV (3, 4). sTNFII-R is already considered as aprognostic marker of AIDS and rapid CD4�cell countdecline (5–7). sIL-2R also has been identified as aprognostic marker of AIDS, correlated with the evo-lution of plasma HIV-RNA and inversely correlatedwith CD4�cell count (4, 8, 9). Most of these studieswere conducted before the era of highly active anti-retroviral therapy (HAART). Recently, protease inhib-itors (PI)-containing regimen has dramatically im-proved prognostic of HIV-infected patients bysuppressing HIV replication and then increasing CD4�cell count (10). The CD4�cell count is the only immu-nologic marker routinely measured, but the kinetics of itsmodification is slow and varies much among PI-treatedpeople. Moreover, CD4�cell count does not reflectactivation status of immune system, which could be ofinterest, particularly in cases of immunovirologic-discordant responses. Then, plasma immune markers assTNFR II and sILR-2 might be valuable candidates toevaluate more accurately than CD4�cell count alone theimpact of PI-containing regimen on the activation statusof immune system. So, after having demonstrated theprognostic value of these biomarkers before the era ofHAART, we aimed to study the evolution of sILR-2 andsTNFR-II after the initiation of a PI-containing regimenand its correlation with variations of usual immunologicand virologic markers of HIV infection.

1Service de Me´decine Interne et Maladies Infectieuses, Hoˆpital Saint-Andre, 1 rue Jean Burguet, 33075 Bordeaux cedex, France.

2INSERM U330, Universite´ Victor Segalen Bordeaux 2, 146 rue Le´oSaignat, 33076 Bordeaux cedex, France.

3Laboratoire de Biochimie, Hoˆpital Saint-Andre´, 1 rue Jean Burguet,33075 Bordeaux cedex, France.

4To whom correspondence should be addressed at Service de Me´decineInterne et Maladies Infectieuses, Hoˆpital Saint-Andre´, 1 rue JeanBurguet, 33075 Bordeaux cedex, France; tel: 00 33 556 79 58 23; fax:00 33 556 79 58 22; e-mail: [email protected]

Journal of Clinical Immunology, Vol. 22, No. 2, March 2002 (©2002)

750271-9142/02/0300-0075/0 © 2002 Plenum Publishing Corporation

METHODS

Study Population

A prospective study of 46 HIV-1-infected adult pa-tients was set up from initiation of PI-containing regimen(between March 1996 to September 1997) in the outpa-tient HIV clinic of Saint-Andre Hospital (Bordeaux,France). The patients had CD4�cell count between 50and 250 � 106/Liter, plasma HIV RNA above 500copies/ml, and no sign or symptom of an opportunisticinfection or acute complication. They were all naive ofPI and received 1 PI (indinavir, 800 mg 3 times/day, n �27; saquinavir hard gel, 600 mg 3 times/day, n � 14;ritonavir, 600 mg 2 times/day, n � 3; nelfinavir, 750 mg3 times/day, n � 2) in combination with one nucleosideanalogue (NA) (3 patients) (zidovudine, 250 mg 2times/day, n � 1; stavudine, 40 mg 2 times/day, n � 1,lamivudine, 150 mg 2 times/day, n � 1) or two (43patients) (zidovudine 250 mg and lamivudine 150 mg 2times/day, n � 20; zidovudine 250 mg 2 times/day anddidanosine 400 mg 1 time/day, n � 1; zidovudine 250mg 2 times/day and zalcitabine 0.75 mg 3 times/day, n �3; stavudine 40 mg and lamivudine 150 mg 2 times/day,n � 18; stavudine 40 mg 2 times/day and didanosine400 mg 1 time/day, n � 1). Data were recorded at 6(M6) and 12 months (M12) of follow-up.

Laboratory Measurement

Measurement of plasma cytokines was performedfrom frozen samples obtained at the initiation of PI-containing regimen, at M6 and M12, with enzymeamplified sensitivity immunoassay (MEDGENIX EA-SIA kit, Biosource Europe S.A., Nivelles, Belgium)according to manufacturer’s instructions. The assays takeinto account free and bound receptors. Minimal detect-able concentrations were 0.1 ng/ml for sTNFR-II and 24IU/ml for sILR-2. The intra- and interassay coefficientsof variation were less than 10% for all EIA.

CD4�cell count was measured by flow cytometrysystem. Measurement of plasma HIV RNA was obtainedwith bDNA Chiron test 2.0 (lower limit of detection: 500copies/ml, Quantiplex, Chiron Inc.) according to themanufacturer’s instructions.

Statistical Analysis

When plasma HIV RNA or cytokines levels werebelow the limits of detection, limit of detection wasapplied. Data are described on each point as medians and25th–75th percentiles. At baseline, we studied the asso-

ciation between plasma concentrations of each immunemarker and clinical status using the Kruskall–Wallis testand biological paramaters using the Mann–Withney test.For each parameter, levels after therapy initiation (M6and M12) were compared with baseline levels by thenonparametric Wilcoxon matched pairs test. Correlationsbetween evolutions of plasma markers of immune acti-vation and other parameters were calculated using Spear-man’s rank correlation test.

We used the nonparametric Wilcoxon test to comparethe evolutions of plasma immune markers betweensubgroups of good (defined by an HIV RNA less than500 copies/ml at 12 months) and poor virologic respond-ers, and between subgroups of good (defined by anincrease above 200 CD4�T lymphocytes at 12 monthsfor patients having less than 200 CD4�T lymphocytes atinclusion or increase of more than 50 CD4�T-lymphocytes for patients having more than 200 CD4�Tlymphocytes) and poor immunologic responders. Anal-ysis were performed on Statistical Analysis SystemSoftware (6.12; SAS Institute, Cary, North Carolina).

RESULTS

Baseline Characteristics

Forty-six patients (34 men and 12 women) wereincluded in the study. Median age was 38 years (inter-quartile range, 34–46). Median plasma level of CD4�cell counts and HIV RNA were 135 � 106/1 (77–194)and 4.3 log10/ml (3.8–4.9), respectively. Twelve patientswere classified as asymptomatic HIV-infected (CDCgroup A), 28 as symptomatic non-AIDS HIV-infected(CDC group B) and 6 as having AIDS (CDC group C).Among these patients, 44 (95.6%) had received priorantiretroviral therapy: mono (4.3%) or dual NA therapy(91.3%). Median plasma levels of sTNFR-II was 8.7ng/ml (6.2–12.0), and median plasma level of sILR-2was 795 IU/ml (515–1063). sTNFR-II was significantlyhigher in patients with a low baseline CD4�cell count[11.2 (7.5–12.9) versus 7.1 (4.8–10.9); P � 0.05], or ahigh plasma HIV-RNA level [10.1 (7.9–13.6) versus 6.6(4.8–11.2); P � 0.02]. No association was observedbetween level of sILR-2 and other biological markers.For both cytokines, no significant association was ob-served between cytokines plasma levels and CDC clini-cal stages.

Evolution of Plasma Levels of the Different Parameters

After initiation of a PI-containing regimen, there wasa significant decrease of plasma HIV RNA level between

76 BONNET ET AL.


M0 and M6 (�0.99 log10; P � 10�4) and no significantevolution between M6 and M12 (�0.25 log10; P �0.11). Concomitantly, there was an increase of CD4�cell count between M0 and M6 (�62/ml; P � 10�4),sustained between M6 and M12 (�41/ml; P � 0.02).Plasma levels of sTNFR-II and sILR-2 did not decreasesignificantly throughout the whole follow-up in the 46patients (respectively, �1.18 ng/ml; P � 0.18 and�0.32 UI/ml; P � 0.44 between M0 and M12).

Correlations between Evolution of Plasma LevelsMarkers during Follow-up

From M0 to M6, we identified a negative strongcorrelation between evolution of plasma levels ofsTNFR-II, sILR-2, and CD4�cell count (r � 0.39 andr � 0.38, respectively, both P � 0.05) (Figs. 1 and 2).Evolution of the two soluble receptors were positivelycorrelated (r � �0.77; P � 10�3). There was nocorrelation between HIV RNA and sTNFR-II or sILR-2between M0 and M6 (r � �0.16 and 0.19, respectively,both P � 0.05).

From M0 to M12, there was a significant positivecorrelation between evolution of sTNFR-II or sILR-2and those of HIV RNA (r � �0.43 and r � �0.36,respectively, P � 0.05) (Figs. 3 and 4). No significantcorrelation was found between evolutions of both solublereceptors and those of CD4� (r � 0.02 for sTNFR-IIand CD4�, r � 0.21 for sILR-2 and CD4�, both P �

0.05). We noticed again a significant positive correlationbetween evolutions of the two soluble receptors (r ��0.48; P � 0.05)

Comparison of Good and Poor Responders

Among 30 men and 7 women, 12 (32%) were goodvirologic responders. At M12, the levels of sTNFR-IIand sILR-2 were not statistically different between goodand poor virologic responders, although there was a trendfor lower plasma levels of immune markers in thesubgroup of good virologic responders (6.4 vs. 8.8ng/ml, respectively, P � 0.06 for sTNFR-II; and 674 vs.801 UI/ml, P � 0.22 for sILR-2).

Between M0 and M12, there was a decrease ofsTNFR-II plasma levels and of sILR-2 in good virologicresponders, while sTNFR-II remained stable and sILR-2tended to increase in poor virologic responders (Table I).The evolution of plasma levels of sTNFR-II between M0and M12 according to the profile response was signifi-cantly different between good virologic responders(�2.5 ng/ml) and poor virological responders (�0.2ng/ml) (P � 0.02). The same analysis about sILR-2 didnot show any significant association.

The immunologic response was studied in 30 men and10 women, among which 23 (58%) were consideredgood immunologic responders and 17 poor immunologicresponders at M12. At M12, the levels of sTNFR-II andsILR-2 were not statistically different between good and

Fig. 1. Correlation between plasma level evolution of sTNFR-II and lympho-cytes CD4�count [difference between the value at 6 months (M6) and the valueat the initiation of protease inhibitor therapy (M0)].

PLASMA IMMUNE MARKERS UNDER HAART 77


poor responders, although there was a trend for lowerplasma levels of immune markers in good compared topoor immunologic responders (6.6 vs. 8.9 ng/ml, respec-tively, P � 0.09 for sTNFR-II; and 714 vs. 838 IU/ml,P � 0.22 for sILR-2). Plasma sILR-2 significantlydecreased in the group of good immunologic responders[�58 UI/ml (�323; �8); P � 0.03] and tended toincrease in the group of poor immunologic responders.The evolution of sILR-2 plasma level between M0 and

M12 was significantly different between good immuno-logic responders (�58 IU/ml) and poor immunologicresponders (�28 IU/ml) (P � 0.01) (Table II). Therewas no difference between the two subgroups for evolu-tion of plasma level of sTNFR-II at 12 months.

DISCUSSION

In this study, evolutions of sTNFR-II and sILR-2 werenegatively correlated with the evolution of CD4�T-lymphocytes between M0 and M6 and positively corre-lated with the evolution of HIV RNA between M0 andM12. Nevertheless, at M12 changes in HIV-RNA weremore closely correlated with changes in plasma sTNFR-II, while changes in CD4�cell count were closelycorrelated with changes in plasma sILR-2.

The prognostic value of sTNFR-II in the progressionof HIV infection is already known (5, 6). Our results atbaseline are in accordance with previous studies: thehigher the plasma levels of sTNFR-II, the lower theCD4�T lymphocytes and the higher the HIV RNA.Nevertheless, in the entire group of our patients we didnot find a significant decrease of plasma level of immunemarkers under therapy or a correlation between theevolution of these markers and those of plasma HIVRNA between M0 and M6 as described by other authors(11, 12). These discordant results could be explained bythe disappointing efficacy of our antiretroviral treatmentor by the kinetics of sampling. Indeed, our patients were

Fig. 2. Correlation between plasma level evolution of sILR-2 and lymphocytesCD4�count [difference between the value at 6 months (M6) and the value at theinitiation of protease inhibitor therapy (M0)].

Fig. 3. Correlation between plasma level evolution of sTNFR-II p75and viral load (VL) [difference between the value at 12 months (M12)and the value at the initiation of protease inhibitor therapy (M0)].

78 BONNET ET AL.


markedly advanced in the HIV infection with a lowCD4�cell count and a high plasma HIV RNA and mostof them received prior mono- or dual-nucleosidic ther-apy. Moreover, Aukrust et al. (13) has shown thatdecrease of sTNFR-II after therapy is rapid (3 months),followed by an increase as early as the 4th month. Thedifferences in the tenor and interpretation of our data andthose reported by Aukrust could be explained by a timingdifferences in sample collections and by the fact that ourstudy has focused mainly on patients with advanced HIVdisease.

We have shown in this study that the decrease ofplasma level of sTNFR-II was significantly higher invirologic responders and that its evolution between M0and M12 was correlated with changes of plasma HIVRNA. Conversely, the difference of sTNFR-II plasmalevel between good and poor immunologic responderswas not significant. Moreover, evolution of sTNFR-IIplasma levels was more closely related to virologic statusthan to immunologic status. These findings can beexplained by the relations between the TNF system andHIV replication. HIV enhances TNF-� production invarious cells (14, 15). This activation of TNF systeminduces activation of NF-�B, enhancing HIV replicationthrough an amplification circle. It is hypothesized thatdecreased TNF activity may lead to low-grade viralreplication in patients under HAART (13). Nevertheless,increased TNF system activation may be either the causeor the result of an increase of HIV-RNA. If HIVreplication is increased by TNF in vitro, the high level ofTNF also could mean a defense mechanism by inducingapoptosis of HIV-infected lymphocytes (16, 17). More-

over, It has been suggested that persistant activation ofthe TNF system may be involved in treatment failure andenhance HIV replication (18). Therefore, evaluation ofplasma levels of sTNFR-II could reflect the TNF systemactivation and reflect the viral status of extravascularcompartments; in this context, the impact of anti-inflammatory and anti-TNF drugs might be of interest asthalidomide showed, in vitro, a decrease of HIV type 1replication in human macrophages (19, 20). Unfortu-nately, the effectiveness of thalidomide in counteringoral esophageal ulcers in 29 HIV-infected patients wasnot associated with a reduction of plasma levels of HIVRNA, TNF� or its soluble receptors (21). Specificexperiments on HIV reservoir markers should be consid-ered in patients having a high plasma level of sTNFR-IIand a low HIV-RNA. Conversely, patients with low HIVRNA and low plasma levels of sTNFR-II could beselected for therapeutic reductions. Moreover, it could bean earlier marker for virologic treatment failure orrelapse as Salazar–Gonzalez et al. (22) have shown thatincrease of plasma soluble markers preceded for severalmonths the increase of HIV RNA and the decrease ofCD4�lymphocytes.

High concentration of sILR-2 is associated with alonger delay to AIDS and is negatively correlated withchanges of CD4�T lymphocytes (4, 8, 9). Therefore, webelieve that it could be a good marker of lymphocyteactivation, although it has never been studied in patientsunder HAART like in our study. We found higherplasma levels of sILR-2 in the subgroup of poor immu-nologic responders, while the evolution of this markerwas not correlated with virologic response status. The

Fig. 4. Correlation between plasma level evolution of sILR-2 et and viral load(VL) [difference between the value at 12 months (M12) and the value at theinitiation of protease inhibitor therapy (M0)].



Tab

leI.

Com

pari

son

ofE

volu

tion

ofPl

asm

aL

evel

sof

sTN

FR-I

Ian

dsI

LR

-2,

inG

ood

and

Poor

Vir

olog

ical

Res

pond

ers

afte

r12

Mon

ths

ofPI

The

rapy

Vir

olog

ical

resp

onse

onM

12a

Goo

d(n

�12

)Po

or(n

�25

)

Pb

M0

M12

M12

-M0

PM

0M

12M

12-M

0P

STN

FR-I

I(n

g/m

l)M

edia

n(1

st–3

rdqu

artil

es)

10.7

(6.2

–12.

3)6.

4(5

.1–

8.5)

�2.

5(�

4.3;

�1.

7)0.

028.

8(6

.6–1

2.0)

8.8

(6.6

–14.

7)�

0.2

(�2.

5;1.

9)0.

540.

02SI

LR

-2(U

I/m

l)M

edia

n(1

st–3

rdqu

artil

es)

767

(576

–110

0)67

4(5

03–1

073)

�23

(�81

;151

)0.

0976

7(51

5–98

3)80

1(6

75–1

039)

�65

(�28

7;�

82)

0.72

0.17

aG

ood

viro

logi

cre

spon

se:

unde

tect

able

vira

llo

adat

12m

onth

s.bC

ompa

rais

onbe

twee

npo

oran

dgo

odre

spon

ders

.

Tab

leII

.C

ompa

riso

nof

Evo

lutio

nof

Plas

ma

Lev

els

ofsT

NFR

-II

and

sIL

R-2

,in

Goo

dan

dPo

orIm

mun

olog

ical

Res

pond

ers

afte

r12

Mon

ths

ofPI

The

rapy

Imm

unol

ogic

alre

spon

seat

M12

a

Goo

d(n

�23

)Po

or(n

�17

)

Pb

M0

M12

M12

-M0

pM

0M

12M

12-M

0p

STN

FR-I

I(n

g/m

l)M

edia

n(1

st–3

rdqu

artil

es)

9.6

(5.2

;�13

.6)

6.6

(5.0

;�14

.0)

�0.

6(�

3.7;

1.2)

0.20

11.2

(7.5

;12.

7)8.

9(7

.3;1

4.3)

�1.

4(�

3.6;

0.7)

0.65

0.83

SIL

R-2

(IU

/ml)

)M

edia

n(1

st–3

rdqu

artil

es)

767

(587

;116

1)71

4(5

15;1

034)

�58

(�32

3;�

8)0.

0382

2(5

13;9

82)

838

(522

;121

6)28

(�68

;234

)0.

110.

01

aG

ood

imm

unol

ogic

resp

onse

:in

crea

seab

ove

200

CD

4�T

lym

phoc

ytes

at12

mon

ths

for

patie

nts

havi

ngle

ssth

an20

0C

D4�

Tly

mph

ocyt

esat

incl

usio

nor

incr

ease

ofm

ore

than

50C

D4�

Tly

mph

ocyt

esfo

rpa

tient

sha

ving

mor

eth

an20

0C

D4�

Tly

mph

ocyt

es.

bC

ompa

rais

onbe

twee

npo

oran

dgo

odre

spon

ders

.

80 BONNET ET AL.


increase of plasma level of sILR-2 in nonimmunologicresponders could mean a relative deficiency in IL-2production. This finding could allow one to evaluate theendogenous production of IL-2 by a simple and cheapermethod than culture and measurement in vitro of IL-2production by T cells. Nonspecifically, sILR-2 plasmalevel might be of interest in immunologic nonrepondersand discordant responders to provide arguments in theindication of a substitutive treatment with IL-2. Salazar-Gonzalez et al. (22) showed that sILR-2 levels increasedsignificantly 2.5 years before AIDS diagnosis, suggest-ing that evaluation of this plasma immune marker couldhelp the clinician earlier than CD4�T lymphocytescount alone.

Then, we believe that measurement of these immuneactivation markers seems to be of interest. They might beearlier and more accurate marker of immunologic impactof therapy than CD4�cell count alone and might in-crease the knowledge about temporal relationships be-tween immunologic and virologic features of HIV infec-tion. Moreover, it is not clear if persistent immunologicactivation may play a pathogenic role or not in HIVinfection. Some authors suggest that the maintenance ofanti-HIV immune stimulation could have some interestfor controlling viral replication in the long term (23).This persistent immune stimulation could be assessed bythe measurement of some of these markers. Some of thiscomments could be applied to immunovirologic-discordant responders where antiretroviral strategies areoften difficult to manage (24).

In conclusion, we have shown that changes of solubleplasma immune markers were correlated with immuno-logic or virologic responses under HAART. We alsoshowed, in these patients with advanced HIV disease,that the decrease of sTNFR-II was higher in goodvirologic responders and that the decrease of sILR-II washigher in good immunologic responders. Therefore, themeasurements of these immunologic markers may be ofinterest as an early, economical, and additionnal deter-mination of the impact of PI-containing regimen in thecase management of patients and in the development ofnew antiretrovirals.

ACKNOWLEDGMENTS

The authors would like to thank the following indi-viduals for participating in the monitoring or data col-lection: Jean Agbemenou, Christine Bigaillon,Veronique Fournis, Agnes Galimard, Patrick Mercie,Serge Niamke, Emmanuelle Salord, Ingrid Steinbruck-ner, Wanarou Traore, Remi Vatan, and all the secretary

and nurse staff of the out patient HIV clinic of Saint-Andre Hospital.

REFERENCES

1. Hughes MD, Johnson VA, Hirsh MS, Bremer JW, Elbeik T, EriceA, Kuritzkes DR, Scott WA, Spector SA, Basgoz N, Fischl MA,D’Aquila RT: Monitoring plasma HIV-1 RNA levels in addition toCD4�lymphocyte count improves assessment of antiretroviraltherapeutic response: ACTG Protocol Virology Substudy Team.Ann Intern Med 126:929–938, 1997

2. Mellors JW, Munoz A, Giorgi JV, Margolick JB, Tassoni CJ,Gupta P, Kingsley LA, Todd JA, Saah AJ, Detels R, Phair JP,Rinaldo CR Jr: Plasma viral load and CD4�lymphocytes asprognostic markers of HIV-1 infection. Ann Intern Med 126:946–954, 1997

3. Fahey JL, Taylor JM, Detels R, Hofmann B, Melmed R, NishanianP, Giorgi JV: The prognostic value of cellular and serologicmarkers in infection with human immunodeficiency virus type 1. NEngl J Med 322:166–172, 1990

4. Godfried MH, van der Poll T, Weverling GJ, Mulder JW, JansenJ, van Deventer SJ, Sauerwein HP: Soluble receptors for tumornecrosis factor as predictors of progression to AIDS in asymptom-atic human immunodeficiency virus type 1 infection. J Infect Dis169:739–745, 1994

5. Saves M, Morlat P, Chene G, Peuchant E, Pellegrin I, Bonnet F,Bernard N, Lacoste D, Salamon R, Beylot J: Prognostic value ofimmune activation in patients with advanced HIV disease treatedby combination antiretroviral therapy. Clin Immunol 99:347–352,2001

6. Fahey JL, Taylor JM, Manna B, Nishanian P, Aziz N, Giorgi JV,Detels R: Prognostic significance of plasma markers of immuneactivation, HIV viral load and CD4 T-cell measurements. AIDS12:1581–1590, 1998

7. Matsuyama T, Kobayashi N, Yamamoto N: Cytokines and HIVinfection: is AIDS a tumor necrosis factor disease? AIDS 5:1405–1417, 1991

8. Noronha IL, Daniel V, Schimpf K, Opelz G: Soluble IL-2 receptorand tumor necrosis factor-alpha in plasma of haemophilia patientsinfected with HIV. Clin Exp Immunol 87:287–292, 1992

9. Sulkowski MS, Chaisson RE, Karp CL, Moore RD, Margolick JB,Quinn TC: The effect of acute infectious illnesses on plasmahuman immunodeficiency virus (HIV) type 1 load and the expres-sion of serologic markers of immune activation among HIV-infected adults. J Infect Dis 178:1642–1648, 1998

10. Sepkowitz KA: Effect of HAART on natural history of AIDS-related opportunistic disorders. Lancet 351:228–230, 1998

11. Nokta M, Rossero R, Loesch K, Pollard RB: Kinetics of tumornecrosis factor alpha and soluble TNFRII in HIV-infected patientstreated with a triple combination of stavudine, didanosine, andhydroxyurea. AIDS Res Hum Retroviruses 13:1633–1638, 1997

12. Chougnet C, Fowke KR, Mueller BU, Smith S, Zuckerman J,Jankelevitch S, Steinberg SM, Luban N, Pizzo PA, Shearer GM:Protease inhibitor and triple-drug therapy: Cellular immune param-eters are not restored in pediatric AIDS patients after 6 months oftreatment. AIDS 12:2397–2406, 1998

13. Aukrust P, Muller F, Lien E, Nordoy I, Liabakk NB, Kvale D,Espevik T, Froland SS: Tumor necrosis factor (TNF) system levelsin human immunodeficiency virus-infected patients during highlyactive antiretroviral therapy: Persistent TNF activation is associ-



ated with virologic and immunologic treatment failure. J Infect Dis179:74–82, 1999

14. Merrill JE, Koyanagi Y, Chen IS: Interleukin-1 and tumor necrosisfactor-A can be induced from mononuclear phagocytes by humanimmunodeficiency virus type 1 by binding to the CD4 receptor.J Virol 63:4404–4408, 1989

15. Westendorp MO, Shatrov VA, Schulze-Osthoff K, Frank R, KraftM, Los M, Krammer PH, Droge W, Lehmann V: HIV-1 Tatpotentiates TNF-induced NF-kappa B activation and cytotoxicityby altering the cellular redox state. EMBO J 14:546–554, 1995

16. Duh EJ, Maury WJ, Folks TM, Fauci AS, Rabson AB: Tumornecrosis factor alpha activates human immunodeficiency virus type1 through induction of nuclear factor binding to the NF-kappa Bsites in the long terminal repeat. Proc Natl Acad Sci USA86:5974–5978, 1989

17. Matsuyama T, Hamamoto Y, Soma G, Mizuno D, Yamamoto N,Kobayashi N: Cytocidal effect of tumor necrosis factor on cellschronically infected with human immunodeficiency virus (HIV):Enhancement of HIV replication. J Virol 63:2504–2509, 1989

18. Stein DS, Lyles RH, Graham NM, Tassoni CJ, Margolick JB, PhairJP, Rinaldo C, Detels R, Saah A, Bilello J: Predicting clinicalprogression or death in subjects with early-stage human immuno-deficiency virus (HIV) infection: A comparative analysis of quan-tification of HIV RNA, soluble tumor necrosis factor type IIreceptors, neopterin, and beta2-microglobulin. Multicenter AIDSCohort Study. J Infect Dis 176:1161–1167, 1997

19. La Maestra L, Zaninoni A, Marriott JB, Lazzarin A, Dalgleish AG,Barcellini W: The thalidomide analogue CC-3052 inhibits HIV-1and tumour necrosis factor-alpha (TNF-alpha) expression in

acutely and chronically infected cells in vitro. Clin Exp Immunol119:123–129, 2000

20. Moreira AL, Corral LG, Ye W, Johnson B, Stirling D, Muller GW,Freedman VH, Kaplan G: Thalidomide and thalidomide analogsreduce HIV type 1 replication in human macrophages in vitro.AIDS Res Hum Retroviruses 13:857–863, 1997

21. Jacobson JM, Greenspan JS, Spritzler J, Ketter N, Fahey JL,Jackson JB, Fox L, Chernoff M, Wu AW, MacPhail LA, VasquezGJ, Wohl DA: Thalidomide for the treatment of oral aphthousulcers in patients with human immunodeficiency virus infection.National Institute of Allergy and Infectious Diseases AIDS Clini-cal Trials Group. N Engl J Med 336:1487–1493, 1997

22. Salazar-Gonzalez JF, Martinez-Maza O, Nishanian P, Aziz N,Shen LP, Grosser S, Taylor J, Detels R, Fahey JL: Increasedimmune activation precedes the inflection point of CD4 T cells andthe increased serum virus load in human immunodeficiency virusinfection. J Infect Dis 178:423–430, 1998

23. Ortiz GM, Nixon DF, Trkola A, Binley J, Jin X, Bonhoeffer S,Kuebler PJ, Donahoe SM, Demoitie MA, Kakimoto WM, Ketas T,Clas B, Heymann JJ, Zhang L, Cao Y, Hurley A, Moore JP, HoDD, Markowitz M: HIV-1-specific immune responses in subjectswho temporaly contain virus replication after discontinuation ofhighly active antiretroviral therapy. J Clin Invest 104:R13–18,1999

24. Kaufmann D, Pantaleo G, Sudre P, Telenti A: CD4-cell count inHIV-1-infected individuals remaining viraemic with highly activeantiretroviral therapy (HAART). Swiss HIV Cohort Study. Lancet351:723–724, 1998

82 BONNET ET AL.


correlations of soluble interleukin-2 and tumor necrosis factor type ii receptors with immunologic...

Documents