Journal of Virological Methods 103 (2002) 183– 190 www.elsevier.com/locate/jviromet

Reliability of rapid diagnostic tests for HIV variant infection Maria Makuwa a,*, Sandrine Souquie`re a, Marie-There`se Niangui a, Pierre Rouquet a, Cristian Apetrei a, Pierre Roques a, Franc¸ois Simon a,b a b

Centre International de Recherches Me´dicales, B.P. 769 France6ille, Gabon Laboratoire de 6irologie, CHU Charles Nicolle, 76031 Rouen Cedex, France

Received 23 October 2001; received in revised form 29 January 2002; accepted 30 January 2002

Abstract The sensitivity of one ELISA method, six HIV-1/HIV-2 rapid screening tests, and one confirmatory test was evaluated in comparison with a third-generation EIA method (taken as the ‘gold standard’) and Western blot on well-characterized panels of sera. HIV diversity was represented by 50 HIV-1 group M subtype A to H, nine HIV-1 group O, 12 HIV-2, two HIV1 +2 positive and six indeterminate Western blot profiles. Sensitivity during HIV-1 seroconversion was studied on 39 serial samples collected from six patients during early primary infection. Serial samples obtained from two primates during experimental primary SIV infection were used to mimic HIV-2 seroconversion samples. The sensitivity ranged from 100 to 94.6% according to the test. During seroconversion, rapid tests became positive 2–8 days later than the third-generation EIA. This reveals a major limitation of rapid tests, which are being recommended for use in developing countries. The lack of sensitivity seen during early HIV-1 seroconversion and/or limited specificity in some of the evaluated tests present serious limitations to their use in countries with high HIV incidence and variability. It is suggested that, as soon as possible, less sensitive rapid tests for blood bank screening should be abandoned in favor of highly sensitive rapid tests and/or more robust, more sensitive and cheaper ELISAs. These results stress the need for better screening tools and specific local evaluations. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Rapid tests; HIV/SIV diversity; Local serum panels; Seroconversion panels

1. Introduction Several diagnostic strategies based on rapid tests have been developed during the last decade,

* Corresponding author. Tel.: +33-241-67-7092; fax: + 33241-67-7295. E-mail address: [email protected] (M. Makuwa).

and some have been recommended for use in developing countries (WHO, 1997). Initial HIV screening tests were evaluated using sera from HIV-1 subtype B-infected patients. The discovery of HIV-1 group O prompted manufacturers to modify their tests in order to cover a broader strain diversity (Loussert-Ajaka et al., 1994). The recent description of HIV-1 group N and new simian immunodeficiency viruses (SIVs) has raised

0166-0934/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 0 9 3 4 ( 0 2 ) 0 0 0 2 6 - 5

184

M. Makuwa et al. / Journal of Virological Methods 103 (2002) 183–190

concerns that current strategies may fail to detect divergent variants with a potential for widespread dissemination (Simon et al., 1998; Corbet et al., 2000; Souquie`re et al., 2001). The sensitivity of HIV screening tests is still being tested, with comparisons of tests based on synthetic peptides (SP), tests based on recombinant proteins (RP), or combinations of the two. Some third-generation assays and tests based on combinations of HIV p24 antigen and antibodies can detect the low antibody levels that occur during seroconversion. Previous studies have highlighted the importance of p24 antigen detection for early diagnosis, and have demonstrated the improved increased sensitivity of combined antigen– antibody assays over antibody-only tests (Ly et al., 2001; Laperche et al., 2000; Courouce´ et al., 1999). The sensitivity of HIV screening tests varies when samples of different geographical origins are tested, underlining the need for regular testing of local sera. The sensitivity of seven screening assays that are used widely in Central Africa was evaluated by testing a local panel of diverse human and primate samples. The study was carried out at Centre International de Recherches Me´ dicales in Franceville, at the request of the Gabonese Ministry of Public Health, as part of a local quality control of HIV assays marketed in this country. 2. Materials and methods

2.1. Panels of sera Four panels of sera were prepared and tested with rapid tests. Results were compared with those obtained with our ‘gold standard’ — a thirdgeneration EIA method (Genscreen HIV-1/2 2e vs, Biorad).

2.1.1. A panel of human sera representati6e of HIV di6ersity in Africa Genotyping was carried out by heteroduplex mobility assay and/or sequencing (Makuwa et al., 2000). This panel comprised 79 genotyped samples. Fifty HIV-1 subtyped samples (26 A, one B, one C, six D, two F, four H and 10 indeterminate).

Six late HIV-1 seroconversion samples on which genotyping and Western blot were indeterminate but yielded a profile suggestive of late seroconversion (no antibodies against HIV RT or integrase) Twelve HIV-2-positive sera (eight A, four B) and two dually reactive sera (HIV-1 M+ HIV-2 B) Nine HIV-1 group O sera. Given the limited number of HIV-1 group N samples, tests were limited to the semi-rapid assay (Particle Agglutination) and two rapid tests (Recombigen and Determine) which are used widely in Gabon. As a specificity control, 56 samples from HIV-seronegative Gabonese patients attending a sexually transmitted disease clinic were also studied.

2.1.2. Human serocon6ersion panel with samples representati6e of HIV-1 strain di6ersity collected 6ery early after primary infection This panel consisted of 39 serial samples (spanning at least 1 month) from six patients infected by HIV-1 subtype A, B or C. Eighteen of these serial samples were positive for p24 antigen (ELAVIA Ag I, Sanofi Pasteur). 2.1.3. Primate serocon6ersion panel Samples were obtained very early after experimental primary SIV infection of two primates. As we had no human HIV-2 seroconversion samples, samples were obtained very early after experimental SIV infection (macaques infected by the SIVmac251 prototype strain). Experimental primary infection of a red-capped mangabey (Cercocebus torquatus torquatus) with SIVrcm was used to mimic infection by non HIV-1/non HIV-2 immunodeficiency viruses (Georges-Courbot et al., 1998). 2.1.4. A simian panel with samples representati6e of the range of lenti6irus di6ersity This panel consisted of 23 SIV positive samples from six species (African green monkeys, chimpanzees, mandrills, macaques, baboons and sooty mangabeys). All these samples were collected from naturally or experimentally infected animals housed in our primate research center.

M. Makuwa et al. / Journal of Virological Methods 103 (2002) 183–190

185

Table 1 Characteristics of the assays Test

Genscreen 2 Genelavia SFD HIV 1/ 2 PA ImmunoComb Genie II DoubleCheck Recombigen Determine CombFirm

Manufacturer

Coating on

Biorad U plate Biorad U plate Biorad/Fujirebio Gelatine particles PBS Orgenics Plastic card Biorad Membrane PBS Orgenics Membrane Cambridge/ Membrane PBS Abbott Nitrocellulose strip PBS Orgenics Plastic card

Type of test

Time of procedure

Type of antigen

P24

Gp41 Gp160

Gp36

ELISA 3 ELISA 2 PA

1 h 30 min 1 h 30 min 2h

RP – RP

– SP RP

RP RP –

SP SP RP

SRA SRT SRT SRT

40 min 10 min 15 min 5–10 min

– RP – –

SP RP SP RP

SP – SP RP

SP SP SP RP

SRT

15 min

RP/SP

RCA

1h

RP–p24, p31, gp120, SP–gp41, gp36

U plate-microplate with flat-bottomed wells; ELISA 3, 3rd generation assay; ELISA 2, 2nd generation assay; PA, particle agglutination; SRA, semi-rapid assay; SRT, simple rapid test; RCA, rapid confirmatory assay; RP, recombinant protein; SP, synthetic peptide.

2.2. Assays The tests were selected on the basis of their format and antigen composition (SP, RP or a combination of the two) (Table 1). The Genscreen HIV1/2-2nd version third-generation assay is highly sensitive (Ly et al., 2001; Laperche et al., 2000; Courouce´ et al., 1999) and was used as the ‘gold standard’ for this study. The rapid confirmatory assay ImmunoComb II HIV 1&2 was evaluated as an alternative to classical Western blot. The different rapid tests are listed in Table 1

of marked variability. In order to improve their reliability, tests were considered positive if two technicians noted independently even faint reactivity. Equivocal samples were re-tested once, samples with discordant results were considered indeterminate. Sensitivity and specificity were calculated using the following formulas, excluding indeterminate results:

Table 2 Assay performance on a human serum panel

2.3. Testing and quality assurance

Reference test: GENSCREEN 1/ 2 2nd version

The positivity criteria result used here were as follows: human samples: Genscreen positive and WB positive (according to WHO criteria). Seroconversion panels: Genscreen positive and p24Ag positive, or PCR positive. Immunoassays, such as particle agglutination assays, rapid tests and Western or line blots, are scored or read subjectively, introducing a source

Assays

Sensitivity (%)

Specificity (%)

Genelavia Recombigen DoubleCheck Immunocomb Determine Particle Agglutination Genie II CombFirm

100 100 100 100 100 100 97.4 (2 FN) 100

100 100 100 98.2 98.2 94.6 100 94.6

J0 J2 J6 J30 J0 J4 J5 J7 J30 J0 J2 J3 J5 J0 J2 J3 J5 J16 J22 J23 J36 J90 J0 J1 J2 J5 J8 J16 J31 J90 J0 J1 J3 J4 J8 J11 J15 J17 J28

B-1 B-2 B-3 B-4 CL-1 CL-2 CL-3 CL-4 CL-5 Z-1 Z-2 Z-3 Z-4 H-1 H-2 H-3 H-4 H-5 H-6 H-7 H-8 H-9 HS-1 HS-2 HS-3 HS-4 HS-5 HS-6 HS-7 HS-8 O-1 O-2 O-3 O-4 O-5 O-6 O-7 O-8 O-9

− − − − + + + + − + + + + + + + + − − − − − + + + − − − − − + + + − − − − − −

p24

Ag

+ + + + − + o + + + + + + − + + o + + o o + + o o + + o o + − + o + + + o + o

GS

+ + + + − − − − + − − − − − − − − + + + + + o − − − + + + + − − o − + + + + +

GM

Assay

+ + + + − − − − + − − f f − − f +/− + + + + + − − − +/− +/− + + + − − − f + + + + +

RE + + + + o o o o o o o o o o o o o o o o o o − − − − − + + + o o o o o o o o o

DC + + + + − − − − + − − − − f f +/− +/− + + + + + − − − f + + + + − f f f + + + + +

IC + + + + − − o − + − − − − − − − o + + + + + − o o f + ++ ++ + − − o f ++ ++ ++ ++ ++

DT + + + + +/− + + + + +/− +/− +/− +/− +/− + + + + + + + + − − +/− + + + + + − +/− + + + + + + +

PA + + + + − − − − + − − − − − − − IND IND IND IND o o o o o o o o o o − − − − o o o o o

CF p25, fp40, p55, fgp160 p25, fp40, p55, fgp160 fp18, p25, p40, fgp41, p55, fgp120, fgp160 fp18, p25, p40, fgp41, p55, fgp120, fgp160 ne´ gatif p18, fp34, fp55 p18, fp34, fp55 p18, fp34, fp55 p18, p25, fp34, p40, fgp4, p55, fp68, fgp120, gp160 negative fp25, fp34 fp25, fp34 fp25, fp34 negative negative negative negative p18, fp25, fp55, fgp160 p18, fp25, fp55, fgp160 p18, fp25, fp55, fgp160 p18, p25, p34, p40, fgp41, p55, fgp120, fgp160 p18, p25, p34, p40, fgp41, p55, fgp120, fgp160 Negative negative fp25 p18, p25, p34, p40, gp41, p52, p55, p68, gp120, gp160 p18, p25, p34, p40, gp41, p52, p55, p68, gp120, gp160 p18, p25, p34, p40, gp41, p52, p55, p68, gp120, gp160 p18, p25, p34, p40, gp41, p52, p55, p68, gp120, gp160 p18, p25, p34, p40, gp41, p52, p55, p68, gp120, gp160 negative fp25, p34, fp40, p55 fp25, p34, fp40, p55 p18, p25, p55, gp160 fp18, p25, p55, fgp160 fp18, p25, fp40, p55, gp160 fp18, p25, fp40, p55, gp160 fp18, p25, fp34, fp40, fp52, p55, gp160 fp18, p25, fp34, fp40, fp52, p55, gp160

W. Blot VIH1

A

A

B

B

C

HMA

GS, Genscreen; GM, Genelavia Mixt; RE, Recombigen; DC, DoubleCheck; IC, ImmunoComb; DT, Determine; PA, Particle Agglutination; CF, CombFirm; HMA, heteroduplex mobility assay, o, not tested, ++, strong positive; +, positive; +/−, weak positive; f, faint reaction; −, negative.

Day

Number

Table 3 Assay performance on a human seroconversion panel

186 M. Makuwa et al. / Journal of Virological Methods 103 (2002) 183–190

M. Makuwa et al. / Journal of Virological Methods 103 (2002) 183–190



Sensitivity =



Specificity =

n

True Positive (True Positive + False Negative)

n

True Negative True Negative+ False Positive

187

Table 4 Assay performance on a simian panel

× 100

× 100

3. Results

3.1. Human panel All the tests were highly sensitive and recognized correctly HIV-1-positive Gabonese sera; however, Genie II gave false-negative results for two HIV-1 genotype A late seroconversion samples, and was thus excluded from further evaluation. All the other assays, including the confirmatory test CombFirm, recognized correctly all HIV variants (M, O, HIV-2, HIV1+2). Recombigen, Particle Agglutination and Determine all correctly recognized HIV-1 group N samples. This study was not designed to evaluate specificity, but a certain lack of specificity was observed with some assays in the small panel of 56 negative samples (Table 2).

3.2. Human serocon6ersion panel The rapid tests lacked sensitivity compared with the ‘gold standard’. The first date of Genscreen positivity was taken as day 1 of seroconversion. Rapid tests were unreliable with pre-seroconversion samples (p24-positive, Western blot-negative) and some seroconversion samples (Western blot reactive). The Particle Agglutination test performed better, although we have previously noted the relatively poor specificity of this assay (95%). ImmunoComb Bi-Spot performed less well than Particle Agglutination. Samples from subtype C seroconverters were collected later during seroconversion and were recognized correctly by all the screening assays. One of the subtype B seroconversion samples was only detected by the Particle agglutination and Recombigen tests. The remaining assays were only positive between day 3 and 7.

Reference test: Genscreen 1/ 2 2nd version Evaluated assays

Sensitivity

Specificity

Genelavia Recombigen ImmunoComb Particle Agglutination Determine DoubleCheck Combfirm

100 95.8 100 95.8 91.6 75 Nt

100 100 100 100 100 100 Nt

Nt, not tested.

With the human seroconversion panel, CombFirm was less sensitive than Western blot, missing at least three of the five seroconversion samples tested. CombFirm also failed to detect p25 and p34, which were very faintly positive in Western blot assays of p24 Ab-positive sera (Table 3).

3.3. Primate serocon6ersion panels SIVmac251 primary infection of macaques was detected reliably by nearly all the tests, probably because they contain HIV-2 antigens; DoubleCheck only became positive on day 28, and CombFirm was not tested (Table 5). Some rapid tests, lacking specific SIV antigens, lacked sensitivity for mangabey SIVrcm primary infection, which was used to mimic infection by non HIV-1/non HIV-2 immunodeficiency viruses (Table 6). Table 5 Assay performance on a primate seroconversion panel (SIVmac251 in macaques) Assays

Genscreen Genelavia Recombigen Immunocomb Particle Agglutination Determine DoubleCheck

Days post-infection D0

D7

D14

D21

D28

− − − − − − −

− − − − − − −

+ + + + + + −

+ + + + + + −

+ + + + + + +

188

M. Makuwa et al. / Journal of Virological Methods 103 (2002) 183–190

Table 6 Assay performance on a primate seroconversion panel (SIVrcm9 in mangabeys) Assays

Genscreen Genelavia Recombigen Immunocomb Particle Agglutination Determine DoubleCheck

Days post-infection D0

D7

D14

D30

D45

− − − − − − −

− − − − − − −

+ + + + + + −

+ + + + + + −

+ + + + + + +

3.4. Primate panel Only SIVcpz and SIVmac251 samples were recognized correctly by all the assays based on a combination of HIV-1 and HIV-2 antigens (CombFirm was not tested). These simian viruses are closely related to HIV-1 and HIV-2, respectively. In contrast, SIVmnd and SIVagm form independent groups of simian viruses distinct from HIV-1 and HIV-2. Four of the seven assays tested gave false-negative results with SIVmnd and SIVagm sera (Table 4).

4. Discussion A large number of commercial assays designed to detect antibodies to HIV viruses type 1 and 2 have been marketed. ELISA methods are used most widely, but rapid tests are used increasingly for screening and diagnosis in developing countries. HIV-1 is characterized by its high genetic variability, rapid evolution and diversification, making it difficult to design reliable tests for specific antibodies. Such tests may miss HIV-1 subtype O strains (Loussert-Ajaka et al., 1994), HIV-1 M strains (Apetrei et al., 1996), HIV-2 infection (Cristiansen, et al., 1996), and HIV seroconversion. The recent identification of HIV group N and new SIVs complicates further the issue (Simon et al., 1998; Corbet et al., 2000; Souquie`re et al., 2001).

Several factors can influence the performance of serological tests, such as the geographical origin and condition of the blood sample, the environmental conditions at the site of testing, human interpretation of the results, and the inherent qualities of the test. Immunoassays such as particle agglutination assays, rapid tests and western or line blots are scored or read subjectively, with a risk of intra- and inter-reader variability, as well as intra- and inter-laboratory variability (Constantine et al., 1994; Dax and O’Connell, 1999). The sensitivity of one ELISA method, six rapid screening assays, and one rapid confirmatory test was evaluated by testing a panel of Gabonese human sera, seroconversions panels, and a panel of primate sera. The Genscreen HIV1/2 2nd version ELISA (Biorad) was used as the ‘gold standard’ in this evaluation, because of its high sensitivity and also because all pre-seroconversion and seroconversion samples used in the panel had been previously assessed with Elavia Ag I and Genscreen HIV1/2 2nd version assays (Biorad). One of the evaluations of the p24 Ag/anti-HIV combined tests showed that they permit an earlier diagnosis of HIV infection than third generation assays, and that they are less sensitive than p24 Ag-only screening assays. Consequently, the p24 antigen–antibody assays have not been used for the diagnosis of a primary infection as an alternative to p24 antigen screening test and, they must be considered as good tools for the detection of HIV infection (Laperche et al., 2000). Recently, a number of newer fourth-generation HIV screening assays have been developed that offer increased sensitivity over earlier tests by combining detection of anti-HIV antibodies with detection of the p24 viral antigen. However, the results emphasize the detection limitations of different assays and suggest improvement for future HIV screening assays (Ly et al., 2001). All the tests had very good sensitivity for selected genotyped Gabonese HIV-1 variant samples for which Western blot profiles were almost complete. Two HIV-1 genotype A late seroconversion samples in the Gabonese panel were not recognized by the Genie II assay. This lack of sensitivity might be due to the use of SP spanning regions that are not shared by all strains, and/or

M. Makuwa et al. / Journal of Virological Methods 103 (2002) 183–190

to the low antibody levels present during primary infection (Thorstensson et al., 1998). Four assays lacked specificity when used to test a panel of 56 seronegative samples, regardless of their antigen composition. It has been reported previously (Bredberg-Raden et al., 1988; Van Kerckhoven et al., 1991; Constantine et al., 1997) that many of the tests evaluated here have significantly lower specificity when used to test African sera. Non specific reactivity with African specimens is mainly due to high IgG levels, but freeze–thawing of the sera may have been contributory. In this study, specificity (95– 100%) was relatively homogeneous, allowing us to compare all the tests. P24 antigenemia early during seroconversion, just prior to antibody emergence, has been documented in African patients (Simon et al., 1996). African serum panels can be expected to contain a large number of samples from patients recently infected by non B subtypes, and are useful for evaluating sensitivity. The best type of panel for sensitivity testing consists of serial seroconversion samples (Thorstensson et al., 1998; Wagner et al., 1992; Sapan and Stramer, 1992; Gallarda et al., 1992; Zaaijer et al., 1992; Simon et al., 1996). This human seroconversion panel included 39 serial genotyped samples collected from six patients infected by HIV-1 subtypes A, B and C, 18 of which were positive for p24 antigen. The test became positive between 3 and \ 7 days after day 1 of Genscreen positivity (‘gold standard’). These results are consistent with previous data (Healey et al., 1989; Vercauteren et al., 1995; Poljak et al., 1997) suggesting that the Serodia Particle Agglutination assay is one of the most sensitive, followed by ImmunoComb BiSpot. CombFirm performed less well than Western blot on the seroconversion panel, possibly because it is unable to detect anti-HIV-1 antibodies in samples containing p24 antigen. CombFirm also missed at least three of the five western blot p24 Ab-positive seroconversion samples. The panel representative of primate lentiviral diversity consisted of samples collected from naturally or experimentally infected primates. SIVmac primary infection was reliably detected, probably owing to the presence of HIV-2 antigen in the test. Conversely, false-negative results ob-

189

tained with SIVagm and SIVmnd-positive sera, together with the difficulty of identifying SIVrcm primary infection, indicate a need for specific SIV tests for primate epidemiological studies and for tests capable of recognizing cross-species transmission. None of the rapid tests was very sensitive during HIV or SIV seroconversion, and nor in terms of detection of SIV as representative or potential lentiviral diversity. The lack of sensitivity seen during early HIV-1 seroconversions and/or limited specificity in some of the evaluated tests present serious limitations to their use in countries with high HIV incidence and variability. Combinations of rapid tests using different antigen preparations might improve reliability, but this approach would be more costly than using more robust, sensitive and cheaper ELISAs. The rapid spread of HIV in developing countries, and the potential emergence of new variants in Central Africa, seriously restrict the utility of rapid screening tests. Above all, HIV screening assays must be tested on local serum panels. Acknowledgements This study was supported by the Centre International de Recherches Me´ dicales (Franceville, Gabon) founded by the Gabonese government, by ELF Gabon and the French Ministry of Cooperation. The assays were provided free of charge by PBS Orgenics/Cambridge Biotech (Courbevoie, France), Biorad/Fujirebio (Marnesla-Coquette, France) and Abbott Laboratories (Abbott Park, USA). References Apetrei, C., Loussert-Ajaka, I., Descamps, D., Damond, F., Saragosti, S., Brun-Vezinet, F., Simon, F., 1996. Lack of screening test sensitivity during HIV-1 non-subtype B seroconversions. AIDS 10, F57 – F60. Bredberg-Raden, U., Kiango, J., Mhalu, F., Biberfeld, G., 1988. Evaluation of commercial immunoassays for antiHIV-1 using East African sera. AIDS 2, 281 – 285. Constantine, N.T., Van der Groen, G., Belsey, E.M., Tamashiro, H., 1994. Sensitivity of HIV-antibody assays determined by seroconversion panels. AIDS 8, 1715 – 1720.

190

M. Makuwa et al. / Journal of Virological Methods 103 (2002) 183–190

Constantine, N.T., Zekeng, L., Sangare, A.K., Gurtler, L., Saville, R., Anhary, H., Wild, C., 1997. Diagnostic challenges for rapid human immunodeficiency virus assays. Performance using HIV-1 group O, HIV-1 group M, and HIV-2 samples. J. Hum. Virol. 1, 45 –51. Corbet, S., Mu¨ ller-Trutwin, M.C., Versmisse, P., Delarue, S., Ayouba, A., Lewis, J., Brunak, S., Martin, P., BrunVe´ zinet, F., Simon, F., Barre´ -Sinoussi, F., Maucle`re, P., 2000. Env sequences of Simian immunodeficiency virus from Chimpanzees in Cameroon are strongly related to those of Human immunodeficinecy virus group N from the same area. J. Virol. 74, 529 –534. et le groupe de travail Re´ trovirus de la Socie´ te´ franc¸ aise de transfusion sanguine, Courouce´ , A.M., 1999. Sensibilite´ des trousses de de´ pistage des anticorps anti-VIH. Re´ e´ valuation 1999. Transfus. Clin. Biol. 6, 381 –394. Cristiansen, C.B., Jessen, T.E., Nielsen, C., Staun-Olsen, P., 1996. False negative anti-HIV-1/HIV-2 ELISAs in acute HIV-2 infection. Vox Sang 70, 144 –147. Dax, E.M., O’Connell, R., 1999. Standardisation of subjectively scored HIV immunoassays: developing a quality assurance program to assist in reproducible interpretation of results using an anti-HIV particle agglutination as a model. J. Virol. Methods 82, 113 –118. Gallarda, J.L., Henrard, D.R., Liu, D., et al., 1992. Early detection of antibody to human immunodeficiency virus type 1 by using an antigen conjugate immunoassay correlates with the presence of immunoglobulin M antibody. J. Clin. Microbiol. 30, 2379 –2384. Georges-Courbot, M.C., Lu, C.Y., Makuwa, M., Telfer, P., Onanga, R., Dubreuil, G., Chen, Z., Smith, S.M., Georges, A., Gao, F., Hanh, B.H., Marx, P.A., 1998. Natural infection of a household pet Red capped mangabey (Cercocebus torquatus torquatus) with a new simian immunodeficiency virus. J. Virol. 72 (1), 600 –608. Healey, D.S., Maskill, W., Gust, I.D., 1989. Detection of anti-HIV immunoglobulin M by particle agglutination following acute HIV infection. AIDS 3, 301 –304. Laperche, S., Maniez-Montreuil, M., Courouce, A.M., 2000. Screening tests combined with p24 antigen and anti-HIV antibodies in early detection of HIV-1. Transfus. Clin. Biol. 7 (1), 18s – 24s. Loussert-Ajaka, I., Ly, T.D., Chaix, M.L., Ingrand, D., Saragosti, S., Courouce, A.M., Brun-Vezinet, F., Simon, F., 1994. HIV-1/HIV-2 seronegativity in HIV-1 subtype O infected patients. Lancet 343, 1393 –1394. Ly, T.D., Laperche, S., Courouce, A.M., 2001. Early detection of human immunodeficiency virus infection using thirdand fourth-generation screening assays. Eur. J. Clin. Microbiol. Infect. Dis. 20 (2), 104 –110.

Makuwa, M., Souquie`re, S., Apetrei, C., Tevi-Benissan, C., Bedjabaga, I., Simon, F., 2000. HIV prevalence and strain diversity in Gabon: the end of a paradox. AIDS 14, 1275 – 1276. Poljak, M., Zener, N., Seme, K., Kristancie, L., 1997. Particle agglutination test ‘Serodia HIV-1/2’ as a novel anti-HIV-1/ 2 screening test: comparative study on 3311 serum samples. Folia Biol. 43, 171 – 173. Sapan, C.V., LIU, D. Stramer, S.L. 1992. Seroconversion performance characteristics of HIV-1 compared to HIV-1/ HIV-2 assays. VIII International Conference on AIDS/ STD World Congress. Amsterdam, July 1992, Abstr. PuC8198. Simon, F., Descamps, D., Bouchaud, O., Lacassin, F., Bedos, J.P., Dussaix, E., Brun-Vezinet, F., 1996. p24 antigenemia in African patients during primary HIV infection. J. Acquir. Immune. Defic. Syndr. Hum. Retrovirol. 11, 204 – 205. Simon, F., Maucle`re, P., Roques, P., et al., 1998. Identification of a new human immunodeficiency virus type 1 distinct from group M and group O. Nature Medicine 4, 1032 – 1037. Souquie`re, S., Bibollet-Ruche, F., Robertson, D.L., Makuwa, M., Apetrei, C., Onanga, R., Kornfeld, C., Plantier, J.C., Gao, F., Abernethy, K., White, L.J.T., Karesh, W., Telfer, P., Wickings, E.J., Maucle`re, P., Marx, P.A., Barre´ Sinoussi, F., Hahn, B.H., Mu¨ ller-Trutwin, M.C., Simon, F., 2001. Wild Mandrillus sphinx are carriers of two types of Lentivirus. J. Virol. 75, 7086 – 7096. Thorstensson, R., Andersson, S., Lindba¨ ck, S., Dias, F., Mhalu, F., Gaines, H., Biberfeld, G., 1998. Evaluation of 14 commercial HIV-1/HIV-2 antibody assays using serum panels of different geographical origin and clinical stage including a unique seroconversion panel. J. Virol. Methods 70, 139 – 151. Van Kerckhoven, I., Vercauteren, G., Piot, P., van der Groen, G., 1991. Comparative evaluation of 36 commercial assays for detecting antibodies to HIV. WHO Bull. 69, 753 – 760. Vercauteren, G., Beelaert, G., Van der Groen, G., 1995. Evaluation of an agglutination HIV-1 + 2 antibody assay. J. Virol. Methods 51, 1 – 8. Wagner, C., Polesky, H.F., Hanson, M., 1992. gp41 synthetic peptide ELISA shows reactivity in seroconversion before the detection of anti-gp41 by western blotting. VIII International Conference on AIDS/STD World Congress. Amsterdam, July 1992, Abstr. PuB73335. WHO, 1997. Revised recommendations for the selection and use of HIV antibody tests. Week. Epidemiol. Rec. 72, 81 – 88. Zaaijer, H.L., van Exel-Oehlers, P., Kraaijeveld, T., et al., 1992. Early detection of antibodies to HIV-1 by a thirdgeneration assays. Lancet 340, 770 – 772.