Journal of Immunological Methods 332 (2008) 53–60www.elsevier.com/locate/jim

Research paper

Development and studies of the anti-R7V neutralizing antibodyELISA test: A new serological test for HIV seropositive patients

Albert Sanchez, François Gemrot, Jean-Manuel Da Costa Castro⁎

IVAGEN, 62, Route Nationale 113, 30620 Bernis, France

Received 1 October 2007; received in revised form 13 November 2007; accepted 14 December 2007Available online 14 January 2008

Abstract

A new peptide-based ELISA test developed for the detection of anti-R7V specific antibodies in the sera of HIV seropositivepatients is described. HIV virus acquires a cellular antigen at the moment the virus is released by budding, the R7Vepitope. Certainpatients produce anti-R7VAb which are described as having the capacity to neutralize in vitro cell infection by HIV. Anti-R7VAbare also detected in asymptomatic patients who have a lower likelihood of progressing to AIDS. Tested with 449 serum samples,the prevalence of anti-R7VAb was 53.2% in HIV positive patients and 5.5% in HIV negative subjects. According to the duration ofinfection, the seroprevalence reaches almost 80% of non-treated long-term infected patients. Other retrospective studies wereconducted on 308 HIV positive samples; the presence of anti-R7V Ab was significantly higher for 177 asymptomatic patients(64.4%) compared to the 131 symptomatic patients (35.1%). Regarding their neutralizing ability, anti-R7Vantibodies were detectedat the highest percentage in asymptomatic HIV-infected patients naive of treatment. Besides conventional biological parameters(CD4 and viral load), the detection of anti-R7Vantibodies could be proposed to clinicians as an additional tool to manage treatmentinitiation and to improve the psychological state of their asymptomatic patients.© 2007 Elsevier B.V. All rights reserved.

Keywords: Anti-R7V antibodies; ELISA; AIDS; HIV; Neutralizing antibodies

Abbreviations:Ab, antibodies; AIDS, acquired immune deficiencysyndrome; AUC, area under curve; β2m, β-2-microglobulin; CDC,Centers for Disease Control and Prevention; CHU, Centre HospitalierUniversitaire (University Hospital Center); CV, coefficient of variation;EBV, Epstein–Barr virus; ELISA, enzyme-linked immunosorbentassay; HAART, highly active antiretroviral therapy; HAV, hepatitis Avirus; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, humanimmunodeficiency virus; OD, optical density; ROC curve, receiveroperating characteristic curve; STI, structured treatment interruption;TMB, 3,3′,5,5′-tetramethylbenzidine.⁎ Corresponding author. Tel.: +33 4 66 73 17 40; fax: +33 4 66 87 08 31.E-mail addresses: a.sanchez@ivagen.com (A. Sanchez),

f.gemrot@ivagen.com (F. Gemrot), jm.dacosta@ivagen.com(J.-M. Da Costa Castro).

0022-1759/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.jim.2007.12.010

1. Introduction

In 2005 more than 7600 people died daily from AIDSrelated causes, and about 38.6 million people worldwideare infected with HIV (UNAIDS, 2006). Followinginfection with HIV, the rate of clinical disease progres-sion varies between individuals. A certain number ofindividuals infected with HIV do not evolve towardsAIDS, termed non-progressors or late progressors (Caoet al., 1995; Pantaleo et al., 1995; Klein and Miedema,1995). Some of these asymptomatic patients have showna lack of progression for more than 10 years, thus raising

54 A. Sanchez et al. / Journal of Immunological Methods 332 (2008) 53–60

the question of possible natural AIDS resistancemechanisms (Buchbinder and Vittinghoff, 1999).

Studies have indicated that host as well as viralfactors are involved in non-progression (Haynes et al.,1996; Hogan and Hammer 2001a,b; Lopalco, 2004).While the mechanism of late progression to AIDS maybe complex and multifactorial, the identification of apredictive marker of this phenomenon would have avery important impact on the treatment and managementof HIV-infected patients. Research has indicated anassociation between non-progression status and thepresence of anti-R7V antibodies (Ab) (Le Contel et al.,1996; Galea et al., 1996).

HIV belongs to a family of retroviruses known aslentiviruses which acquires its envelope from the hostcellular membrane by budding. Studies have shown thatduring this process HIV non-randomly acquires certainhuman cellular proteins such asβ-2-microglobulin (β2m)and incorporates these as integral components of the viralcoat (Hoxie et al., 1987; Arthur et al., 1992; Cantin et al.,1997). A seven amino acid sequence (Arg-Thr-Pro-Lys-Ileu-Gln-Val) derived from β2m, a highly conservedcellular protein, has been identified as an epitope that ispresent at the surface of divergent HIV isolates (Le Contelet al., 1996). This R7V peptide, designated as RTPKIQV,is cryptic on human β2m but is revealed by HIV onincorporation into the envelop and could be targeted bymonoclonal Ab (Galea et al., 1996). Like glycoproteins ofHIV exposed at the viral surface and targeted by neu-tralizing antibodies (Binley et al., 2004), anti-R7Vmonoclonal Ab were shown to have the capacity toneutralize cell infection in vitro by HIV from differentgenotypes (primary isolates of strains A–F from an inter-national panel) as well as from different tropisms (lym-photropic or macrophage tropic laboratory strains) (LeContel et al., 1996). Isolated human anti-R7VAb, purifiedfrom serum obtained from non-progressors, have alsobeen shown to neutralize various strains of HIV-1 in vitro(Galea et al., 1999a,b;Haslin andChermann, 2002). Thus,it is possible that the presence of anti-R7VAb may serveas a potential marker of late progression in HIV sero-positive individuals.

An ELISA test kit has been developed to detect anti-R7VAb in the serum and plasma ofHIV-infected patients.

2. Materials and methods

2.1. Human serum samples

Independent sources of human serum or plasma sam-ples were used for the development of the anti-R7V testand to conduct studies on the ELISA method developed.

2.1.1. Commercial source of serum or plasma samplesThe following samples were provided by SeraCare

Diagnostics (West Bridgewater, MA, USA) or Zepto-Metrix Corp. (Buffalo NY, USA):

– normal human serum samples controlled as negativefor HIV, HBV and HCV (201);

– human HIV-infected serum samples (248) from aclinical collection provided byZeptoMetrix Corp.; theywere documented with demographic data of the HIV-infected patients, date of HIV diagnosis, date of theblood sample, biological data, and antiviralmedication;

– samples obtained from HIV seronegative patients thatmay present potentially interfering concurrent condi-tions (132): anticoagulants (6), hemolyzed samples (9),hyperlipidemia (12), multibirth (9), Lyme's disease (9),lupus erythematosus (11), rheumatoid polyarthritis (7),EBV (13), herpes (15), HAV (9), HBV (10), HCV (11)and rubella (11).

2.1.2. Samples from external clinical sourcesCoded archived serum samples (19) from HIV-

uninfected subjects (6) and HIV-infected patients (13),previously tested for anti-R7VAb by two assays: an ‘inhouse’ anti-R7V research ELISA test (Galea et al., 1996)and the assay in its development stage (6 negative, oneuncertain close to negative and 12 positive), wereobtained fromURRMAR&D,Aubagne, France, courtesyof J.C. Chermann. They were used to assess the precisionof the test.

Archived samples from HIV-infected patients (308)were classified into two groups: an asymptomatic group(class A, CDC classification, 1993, n=177), naive oftreatment; and a symptomatic group (class B and C,n=131). They were kindly selected and tested in thevirology services of six hospitals (five sites in France:CHU de Besançon; CHU de Grenoble, hospital A.Michallon, La Tronche; CHU de Poitiers; CHU deReims, hospital R. Debré, Reims and hospital Delafon-taine, Saint-Denis; and one in Italy: Ospedale Maggioredella Carità, Novarra). Only one frozen serum samplewas used for each of the 308 patients.

Samples were obtained from a group of HIV-1 sero-positive patients with a therapeutic break (21) who parti-cipated in the TRIESTAN study (TReatment Interruptionin Early STArters in The Netherlands). This study was setup to evaluate the safety and efficacy of discontinuingHAART in HIV-1 positive patients who initiated HAARTat a CD4 T cell count N350 cell/mm3. Samples wereselected and tested by M. Cornelissen et al. (Departmentof Medical Microbiology, CINIMA, AMC, Amsterdam,The Netherlands).

55A. Sanchez et al. / Journal of Immunological Methods 332 (2008) 53–60

2.2. Determination of viral load

Plasma viral load was measured at the Italian hospitalsite on the individual patient samples using a commer-cial branched-DNA assay (VERSANT® HIV-1 RNA3.0 Assay (bDNA), Siemens) and a Quantiplex 340system, as specified by the manufacturer.

2.3. Determination of the T-CD4+ lymphocyte count

CD4 were measured at the Italian hospital by flowcytometry assay (Becton Dickinson) using anti-humanCD4 monoclonal antibodies and a BD FACScalibur™flow cytometer.

2.4. Anti-R7V ELISA reagents

– R7V-microplate: peptide R7V (NeoMPS, France) iscovalently bound in every well of the chemicallyactivated microtiter plate (Corning, USA);

– anti-R7V positive control (diluted HIV positive andanti-R7V positive serum);

– anti-R7V negative control (diluted HIV negative andanti-R7V negative serum);

– internal calibrator (diluted HIV positive and anti-R7V positive serum);

– concentrated conjugate: goat anti-human IgG (H+L)linked to horseradish peroxidase (Jackson ImmunoResearch, UK); the conjugate reacts with the heavychains on human IgG and with the light chainscommon to most human immunoglobulins, enablingIgM recognition in the test;

– washing buffer concentrated 5×(phosphate buffersaline, Tween 20);

– diluent for controls, samples and conjugate dilution(phosphate buffer containing 0.1% human albumin);

– chromogenic substrate TMB (Kementec, Denmark);– STOP solution (diluted hydrochloric acid b15%).

2.5. Test procedure

2.5.1. Preparation and incubation of samplesThe samples, internal calibrator and controls were

diluted (1/51) with the diluent supplied. A volume of100 µl was distributed in each well of the R7V-micro-plate. The plate was incubated for 30±5 min at 37±2 °C.

2.5.2. Conjugate incubationTwenty minutes before use, the conjugate was diluted

(1/101) with the diluent supplied. After incubation, themicrowells were washed four times with 300 µl of1×washing buffer. A volume of 100 µl of diluted con-

jugate was distributed in the R7V-microplate. The platewas incubated for 30±5 min at 37±2 °C.

2.5.3. Substrate incubationAfter incubation of the conjugate, the microwells

were washed four times with 300 µl of 1×washingbuffer. A volume of 100 µl of TMB was distributed inthe R7V-microplate and incubated for 10±1min in thedark at room temperature.

2.5.4. ReadingThe optical density (OD) values were read bichro-

matically at 450 nm/630 nm after adding 100 µl of theSTOP solution.

2.5.5. Calculation and interpretation of the resultsAfter subtraction of the blank sample, the mean OD

value calculated for each patient was normalized with themean OD value of the internal calibrator. The interpreta-tion for the presence or absence of anti-R7V Ab tookplace according to the value of the antibody ratio ob-tained. If the ratio was under 0.8, the sample was con-sidered negative. If it was greater than or equal to 1.0, thesample was defined as positive. The range 0.8–1.0 was agrey zone and the sample was considered doubtful.

3. Results

3.1. Determination of the cut-off value

The cut-off value was determined with samples (source:Zeptometrix Corp.) from 201 control individuals (HIVseronegative sera) compared with a reference group of 37HIV-infected patients, potentially with the factors respon-sible for the observed non-progression. Those patientswereconsidered as long-term survivors: asymptomatic (class A,CDC, 1993), formore than 10 years and naive of treatment.

Sensitivity and specificity were estimated by establish-ing a receiver operating characteristic (ROC) curve whichis a plot of the true positive rate (sensitivity) against thefalse positive rate (100−specificity) for the differentpossible cut-off values of the anti-R7V test (Fig. 1). Goodaccuracy of the test can be observed on this graph with anarea under the curve (AUC) of 0.944. The cut-off valuewas determined at a ratio of 1.0 which provides anestimated sensitivity of 78.4% and a specificity of 95.0%;it was also confirmed with a Youden index test (Fig. 2).

3.2. Evaluation of precision

Precision represents the ability to yield the sameresult when the sample is tested in intra- or inter-assays

Fig. 2. Determination of the anti-R7V cut-off, Youden representation.

Fig. 1. ROC curve of the anti-R7V ELISA test (AUC=0.944).

56 A. Sanchez et al. / Journal of Immunological Methods 332 (2008) 53–60

with the same lot of manufacturing tests. Mean anti-R7V Ab ratios and coefficients of variation (CV) werecalculated for each sample tested to assess the precisionof the assay.

Intra-assay precision, the repeatability of the anti-R7VELISA, was evaluated by testing four samples (URRMAR&D, France), one negative sample (no. 1), one doubtfulsample close to a negative (no. 2) and two positivesamples (nos. 3 and 4). Repeatability was evaluated bytesting these samples 20 times in an intra-assay run. Themean ratios were 0.2, 0.8, 1.4 and 1.6; the intra-assayvariability showed a CV b7% for the four samples.

Inter-assay precision, the reproducibility of the anti-R7V ELISA, was evaluated by testing five negativesamples and ten known positive samples (URRMAR&D,France). Inter-assay reproducibility was evaluated withthese 15 samples tested blindly in duplicate, during5 days, three independent runs per day by three differenttechnicians. For the ten positive samples with mean ratiosranging from 1.3 to 2.2, CVs ranged from 9.5 to 16.3%;the overall variability was 12.4%. For the five negativesamples with mean ratios ranging from 0.2 to 0.8, CVsranged from 9.1 to 20.0%; the overall variability was14.5%. The results of the inter-assay evaluation confirmthe reliability of the test; 100% of positive and 80% ofnegative samples were recognized; a negative sample wasfound in the grey zone of the test with a mean ratio valueof 0.8.

3.3. Analysis of potentially interfering conditions

Table 1 shows the results obtained from the anti-R7VELISA test with samples containing potentially interfer-ing substances. Among the 132 samples, 16 presented apositive ratio (12%), whereas 116 presented a negative

ratio or remained doubtful; the overall specificity wasestimated to be 88%. However, in the absence of recentantecedents of rubella and mononucleosis (108 sam-ples), the global specificity reached 93.5%.

3.4. Studies of clinical samples

The clinical specimens were run for anti-R7V Abdetection using the newly developed ELISA method.

An internal study was done on 449 serum and plasmasamples from the US (Zeptometrix samples). All thesamples were blindly tested in duplicate and in twoindependent runs. The results of the anti-R7V tests wereinterpreted from the mean Ab ratio value obtained withthe two assays.

External studies were also performed in six differenthospital sites. A total of 308 additional samples fromHIV patients were tested. Each hospital site selected anaverage sample size of 50 patients who were divided intotwo groups: (1) asymptomatic patients (class A, CDC)naive of treatment; (2) symptomatic patients (class B orC, CDC).

One serum or plasma sample collected at a uniquetime point from each patient was blindly tested, at eachsite, for the presence or absence of anti-R7VAb.

3.4.1. Seroprevalence of anti-R7V Ab in HIV-infectedpatients

Among the 248 HIV positive infected patients and 201HIV negative individuals from the US samples, 132 HIV-

Table 1Results of interference

Samples types Testednumber

Anti-R7V ELISA

Negative Uncertain Positive

Anticoagulant 6 3 1 2Hemolyzed samples 9 7 1 1Hyperlipidemia 12 10 1 1Multibirth 9 8 1 0Lyme's disease 9 7 1 1Lupus erythematosus 11 9 1 1Rheumatoid polyarthritis 7 7 0 0EBV 13 8 2 3Herpes 15 14 1 0HAV 9 8 1 0HBV 10 8 1 1HCV 11 9 2 0Rubella 11 5 0 6

Thirteen concurrent conditions were tested with 132 samples. The resultsof the anti-R7V tests were classified as positive, negative or uncertain.

57A. Sanchez et al. / Journal of Immunological Methods 332 (2008) 53–60

infected patients produced anti-R7V Ab (53.2%) versus11 subjects for the HIV negative population (5.5%).

We observed a significant difference (chi2 test: 18.48,pb0.001) in the anti-R7VAb frequency in the group of88 non-treated patients (71.6%) versus the group of 160patients under HAART (43.1%).

This study allowed us to determine the presence ofanti-R7VAb and the duration of HIV infection using thedate of HIV diagnosis and the date of the sample testedfor the detection of anti-R7V Ab. Fig. 3 shows thedifferences in anti-R7VAb prevalence according to theduration of HIV infection and the presence or absence ofHAART. Among the patients infected with HIV formore than 10 years, the prevalence of anti-R7V wassignificantly higher (chi2 test: 15.86, pb0.001) than inpatients infected for less than 5 years. In this group of

Fig. 3. Percentage of anti-R7VAb according to the duration of HIV infection. Tof years between the date of their HIV diagnosis and the date of tested sampduration for the 248 patients (total HIV+) and in the two groups of 88 patient(treated HIV+).

long-term infected patients, the prevalence was around80.0% when patients were not treated with HAART.

The combined results of the six external studiesshowed a similar seroprevalence of the anti-R7V Ab(51.9%) within the group of 308 HIV-infected patients.Furthermore, the anti-R7V group represented almost64.8% of the patients naive of HAART treatment(n=179). The percentage of anti-R7V Ab was signifi-cantly higher for the 177 asymptomatic patients (64.4%)compared to the 131 patients (35.1%) classified assymptomatic (chi2: 25.88, pb0.001).

3.4.2. Viral load correlationThe anti-R7VAb ratio was compared to the viral load

for the 45 Italian patients for whom data were available.Fig. 4 shows a direct correlation between the viral loadand the anti-R7VAb ratio for the non-treated and non-symptomatic patients (Pearson's coefficient r=0.383,pb0.01). The results confirmed the study of the USZeptometrix samples (data not shown). The anti-R7VAb ratios and the CD4+ T cell counts do not show acorrelation.

The external study using samples from 21 patientsunder therapeutic breaks at the AMC suggests that Abproduction is closely related to the presence of HIV RNA.Fig. 5 and Table 2 show the results of anti-R7V Abreactivity and the viral load variation. The first sample(sample A) was taken before starting HAART. Thirteenpatients were anti-R7VAb positive with a mean viral loadequal to 184,000 copies/ml. The second (sample B) wastaken during HAART; only two of the 21 patients werenot able to decrease their viral load to undetectable levelswhile under therapy (400 copies/ml). Anti-R7V Abdecreased after the start of HAART, which resulted inno reactive samples except for three patients who

he 248 HIV-infected US patients were classed according to the numberles. Percentage of anti-R7VAb was determined in the three ranges ofs naive of therapy (non-treated HIV+) and 160 patients under HAART

Fig. 4. Correlation between the anti-R7V response and the viral load inasymptomatic patients naive of treatment.

Fig. 5. Follow-up of the anti-R7VAb (ratio) and the viral load (log10)for the 21 patients under treatment break. Sample A: samples beforestart of treatment; Sample B, samples drawn during treatment; SampleC, samples drawn during STI; Sample D, samples after reinitiation oftreatment (except for 3 of them). The average duration of infectionbetween the first and the second sample was 30.4 months, between thesecond and the third 19.0 months, and between the third and the fourth32.8 months. Each bar represents the mean value of the parameter. Thegrey zone is in the anti-R7VAb ratio range 0.8–1.0.

58 A. Sanchez et al. / Journal of Immunological Methods 332 (2008) 53–60

remained positive. The third sample (C) taken afterstoppingHAARTresulted in an increase in viral load in allpatients (mean viral load of 154,175 copies/ml). Meanplasma RNA levels during the stop phase of the studystabilized at approximately pre-HAART values. Sevenpatients became anti-R7V Ab positive after stoppingtherapy. Five of those seven were positive in the firstsample (A). The fourth sample (D) was taken afterreinitiating HAART. Eighteen patients were advised torestart HAARTaccording to the current guidelines for theinitiation of HAART. Six patients remained anti-R7VAbpositive; three of these patients did not reinitiate HAART.

4. Discussion

The recommendations (WHO, 2006; Hammer et al.,2006) on antiretroviral therapy among asymptomaticpatients require analysis of the real and potential risksand benefits. According to all experts, the global treatmentof the patients must take into account biological diagnosiscriteria (CD4, viral load) as well as more general elementssuch as their possible access to information and educationin order to optimize their long-term acceptance of thera-peutic strategies. However, even relying on the accumula-tion of data and their analysis, deciding on the mostappropriate time to start a treatment remains one of themost controversial subjects. That is why providing a newcomplementary tool or approach for clinicians andhealthcare staff would be helpful and could be includedin the algorithm of biological detection tests used at clini-cal sites.

The isolated human anti-R7VAb have been shown toneutralize various strains of HIV-1 in vitro includingstrains resistant to antiretroviral drugs (Galea et al., 1996,1999a,b). The authors hypothesized that anti R7V Abspecific humoral response in HIV-1-infected individuals,

including long-term survivors, was related to theirasymptomatic clinical status.

To detect the presence of those Ab in the serum orplasma of HIV-infected patients we developed a prac-tical ELISA immunoassay based on a microtiter plateformat.

The anti R7V serological test developed has a globalspecificity of 93.5%. A small proportion of subjects notinfected with HIVor with recent antecedents of rubella ormononucleosis may give positive results with the anti-R7V ELISA. According to these limits, the test is ex-clusively reserved for HIV positive individuals.

The study based on the serological follow-up of 21HIV seropositive patients under structured treatmentinterruptions (STI) confirmed the specificity of the anti-R7V antibody humoral response. For 77% of patientsfound to be anti-R7V Ab positive before initiation oftreatment, the anti-R7VAb disappeared during success-ful treatment. The STI resulted in the recovery of severalanti-R7V positive Ab patients (54%), in relation to theviral replication.

The disappearance of R7V Ab during successfultreatment can be explained by the fact that the R7V

Table 2Results of the anti-R7VAb and viral load for the 21 patients under treatment break

PatientID

Sample A Sample B Sample C Sample D

Anti-R7V VL (c/ml) Anti-R7V VL (c/ml) Anti-R7V VL (c/ml) Anti-R7V VL (c/ml)

5 + 29,000 − b40 D 333,390 + 72,44218 + 120,000 + 400 + 160,000 + 22,45779 + 423,291 + b50 − 158,357 + 197,75025 + 1000 − b50 D 54,796 − b5047 + 8600 − b50 D 435,590 − b5054 + 120,000 − b50 − 40,120 − b5055 + 810,000 − b50 D 91,815 − b5059 + 66,000 − b5 + 280,000 − b50101 + 489,613 − b50 − 4461 − b50108 + 170,876 − b50 D 5705 − b5040 + 61,000 + b50 + 769,110 + b5078 + 49,000 D b5 + 6500 + b50111 + 43,000 D b40 + 50,451 + b5024 D 110,000 − b50 + 318,874 − b5075 D 6300 − b40 + 6469 − b5010 − 190,000 − b40 − 190,877 − b5029 − 6200 − b50 − 4418 − b5063 − na − 400 − 250,000 − b5081 − N750,000 − b50 − 31,810 − b5077 D 370,000 − b5 − 34,380 − b5019 − N5,000,000 + 50 − 10,561 − b50

Sample A, samples before start of treatment; 13 patients were anti-R7V (+); Sample B, samples drawn during treatment; for 19/21 patients the viralload decreased to undetectable levels and 4 patients were anti-R7V (+); Sample C, samples drawn during STI; for all patients the viral load increased;7/21 patients become anti-R7V (+); Sample D, samples after reinitiating treatment (except for nos. 5, 18 and 79), 6 patients remained anti-R7V (+),including the three who did not reinitiate HAART.

59A. Sanchez et al. / Journal of Immunological Methods 332 (2008) 53–60

epitope is no longer exposed on virus particles and thusnot seen by the host immune system.

Therefore this specific test is better adapted to thedetection of anti-R7VAb in asymptomatic patients, stillnaive of treatment.

In retrospective studies, we assessed the anti-R7Vparameter on serum specimens drawn at only one singlevisit in the course of HIV infection.

With asymptomatic patients, predominantly non-treated, anti-R7VAb was much more frequent than forthe group of patients classed as symptomatic. Amongthose patients with established chronic HIV infection for10 years or more who remained asymptomatic (class A),this specific Ab reached up to 80%. These results corro-borate those obtained using a previously described “inhouse” ELISA research test (Galea et al., 1996). More-over, the optimized reagents of the anti-R7V ELISA testlimit the false positive results (5.5% instead of 22%previously) obtained with a control group of HIVnegative individuals (Galea et al., 1999a,b).

Concerted efforts are underway to continue withexternal trials of this parameter in order to determinewhether patients with anti-R7V antibodies are at lowerrisk of progression to AIDS.

Providing an optimized tool as a biological andcomplementary element, in addition to CD4 count andviral load, would allow a larger number of patients tocontemplate the non-evolution of their disease; as wellas important socio-economic benefits, this would im-prove their psychological state and help make themmore compliant.

Acknowledgements

We thank G. Agius, L. Andreoletti, M. Bakker, C.Chaplain, J.C. Chermann, M. Cornelissen, C. Haslin, GHerbein, S. Jurriaans, P. Morand, and P. Ravanini fortheir collaboration and assistance with the externalstudies.We are very grateful to G. Tixier for his statisticalcontribution.

References

Arthur, L.O., Bess, J.W., Sowder, R.C., Benveniste, R.E., Mann, D.L.,Chermann, J.C., Henderson, L.E., 1992. Cellular proteins bound toimmunodeficiency virus: implication for pathogenesis and vac-cines. Science 258, 1935.

Binley, J.M., Wrin, T., Korber, B., Zwick, M.B., Wang, M., Chappey,C., Stiegler, G., Kunert, R., Zolla-Pazner, S., Katinger, H.,

60 A. Sanchez et al. / Journal of Immunological Methods 332 (2008) 53–60

Petropoulos, C.J., Burton, D.R., 2004. Comprehensive cross-cladeneutralization analysis of a panel of anti-human immunodeficiencyvirus type 1 monoclonal antibodies. J. Virol. 78, 13232.

Buchbinder, S., Vittinghoff, E., 1999. HIV-infected long-term non-progressors: epidemiology, mechanisms of delayed progression,and clinical and research implications. Microb. Infect. 1, 1113.

Cantin, R., Fortin, J.F., Lamontagne, G., Tremblay, M., 1997. Thepresence of host-derived HLA-DR1 on human immunodeficiencyvirus type 1 increases viral infectivity. J. Virol. 71, 1922.

Cao, Y., Qin, L., Zhang, L., Safrit, J., Ho, D.D., 1995. Virologic andimmunologic characterization of long-term survivors of humanimmunodeficiency virus type 1 infection. N. Engl. J. Med. 332, 201.

CDC, 1992. 1993 Revised Classification System for HIV Infection andExpanded Surveillance Case Definition for AIDSAmongAdolescentand Adults. MMWR Recommendations and Reports 41 (RR-17).

Galea, P., Le Contel, C., Chermann, J.C., 1996. Identification of abiological marker of resistance to AIDS progression. Cell. Pharmacol.AIDS Sci. 3, 311.

Galea, P., Le Contel, C., Chermann, J.C., 1999a. A novel epitope R7Vcommon to all HIV-1 isolates is recognized by neutralizing IgGfound in HIV-infected patients and immunized rabbits. Vaccine 17,1454.

Galea, P., le Contel, C., Coutton, C., Chermann, J.C., 1999b. Rationale fora vaccine using cellular-derived epitope presented by HIV isolates.Vaccine 17, 1700.

Hammer, S.M., Saag,M.S., Schechter,M.,Montaner, J.S., Schooley, R.T.,Jacobsen, D.M., Thompson, M.A., Carpenter, C.C., Fischl, M.A.,Gazzard, B.G., Gatell, J.M., Hirsch, M.S., Katzenstein, D.A.,Richman, D.D., Vella, S., Yeni, P.G., Volberding, P.A., InternationalAIDS Society—USA Panel, 2006. Treatment for adult HIV infection:2006 recommendations of the International AIDS Society—USApanel. JAMA 296, 827.

Haslin, C., Chermann, J.C., 2002. Anti-R7Vantibodies as therapeuticsfor HIV-infected patients in failure of HAART. Curr. Opin.Biotechnol. 13, 621.

Haynes, B.F., Pantaleo, G., Fauci, A.S., 1996. Toward an under-standing of the correlates of protective immunity to HIV infection.Science 271, 324.

Hogan, C.M., Hammer, S.M., 2001a. Host determinants in HIVinfection and disease. Part 1: immune responses. Ann. Intern. Med.134, 761.

Hogan, C.M., Hammer, S.M., 2001b. Host determinants in HIV in-fection and disease. Part 2: implications for antiretroviral thera-peutics. Ann. Intern. Med. 134, 978.

Hoxie, J.A., Fitzharris, T.P., Youngbar, P.R., Matthews, D.M.,Rackowski, J.L., Radka, S.F., 1987. Nonrandom association ofcellular antigens with HTLV-III virions. Hum. Immunol. 18, 39.

Klein, M.R., Miedema, F., 1995. Long-term survivors of HIV-1infection. Trends Microbiol. 3, 386.

Le Contel, C., Galea, P., Silvy, F., Hirsh, I., Chermann, J.C., 1996.Identification of the B2m derived epitope responsible forneutralization of HIV isolates. Cell. Pharmacol. AIDS Sci. 3, 68.

Lopalco, L., 2004. Humoral immunity in HIV-1 exposure: cause oreffect of HIV resistance? Curr. HIV Res. 2, 79.

Pantaleo, G., Menzo, S., Vaccarezza, M., Graziosi, C., Cohen, O.J.,Demarest, J.F.,Montefiori,D.,Orenstein, J.M., Fox,C., Schrager, L.K.,et al., 1995. Studies in subjects with long-term nonprogressive humanimmunodeficiency virus infection. N. Engl. J. Med. 332, 209.

UNAIDS, 2006. Report on the global AIDS epidemic. UNAIDS 10thAnniversary Special Edition. UNAIDS, Geneva.

WHO, 2006. HIV/AIDS Programme. Antiretroviral Therapy for HIVInfection in Adults and Adolescents: Recommendations for aPublic Health Approach. WHO, Geneva.