HIV-1 and HIV-2 are genetically related lentiviruses that are capable of causing CD4 T-cell depletion and AIDS in infected persons [1,2]. Compared with HIV-1 infection, infection with HIV-2 is associated with a slower course of clinical disease, and lower rates of vertical and horizontal transmission [3,4]. HIV-2 infection is also associated with lower plasma and cellular viral loads [5–7] which may explain the slower course of disease and lower rates of transmission. In West Africa, both HIV-1 and HIV-2 are present, thus some persons are HIV-1 and HIV-2 dually seropositive. By PCR testing, about 70% of persons who are dually seropositive are truly dually infected with the two viruses [8–10]. Few data exist regarding immunologic and virologic consequences of dual infection in vivo. We have shown previously that dual infection has no impact on HIV-1 plasma viral load  but one limited study found lower HIV-1 viral loads in dually infected patients . However, limited published data exist on the effect of dual infection on HIV-2 viral replication and immunologic parameters in vivo. We hypothesized that in dually infected persons HIV-1 may interact and affect HIV-2 in vivo in two possible ways: the faster disease producing HIV-1 virus may out-replicate HIV-2, or the presence of HIV-1 may lead to viral synergism that may favor up-regulation of the slowly replicating HIV-2 virus leading to hyper-immune activation, such as increased levels of expression of HLA-DR and CD38 on CD8 cells resulting in poorer disease prognosis among dually infected persons. Indeed, increased level of expression of CD38 on CD8 T cells is an independent prognostic marker for the development of AIDS . Addressing these hypotheses could provide insights into the effect of a more pathogenic HIV-1 on a lesser pathogenic HIV-2, which may be valuable for vaccine design and therapeutic approaches. Here we compared HIV-2 RNA plasma viral load and immune activation markers in persons infected with HIV-2 alone and persons dually infected with HIV-1 and HIV-2 in Abidjan, Côte d'Ivoire.
Between September 1996 and February 2000, we retrospectively analyzed plasma and blood samples obtained from consecutively enrolled persons from different on-going studies at Projet RETRO-CI in Abidjan, who were infected with HIV-1, HIV-2, or were HIV dually seropositive. Only persons with available plasma samples were included in this study: female sex workers attending a confidential clinic, patients seen at an HIV outpatient clinic, and pregnant women attending an antenatal clinic. None of the participants was receiving HIV antiretroviral therapy at the time of specimen collection. The study was approved by the ethical committee of the Côte d'Ivoire Ministry of Health and the Institutional Review Board of the Centers for Disease Control and Prevention, Atlanta, Georgia, USA.
Whole blood was drawn from participants into EDTA tubes (Becton Dickinson, San Jose, California, USA). Within 4 h, plasma was separated from cells by centrifugation at 200 × g, and then aliquoted and stored at −70°C. We determined HIV antibody status using an enzyme-linked immunosorbent assay (ELISA)-based testing parallel algorithm . For HIV type-specific serodiagnosis, we used a combination of monospecific ELISA . DNA PCR testing was performed on uncultured peripheral blood mononuclear cells. In brief, total cellular DNA was prepared by lysing 4 × 106 peripheral blood mononuclear cells/ml in lysis buffer (10 mM Tris–HCl pH 8.3, 0.05% Triton X-100), adding 10 mg proteinase K/ml, and then incubating at 56°C for 1 h, followed by 95°C for 10 min to inactivate the proteinase K. For HIV-1 and HIV-2 PCR testing, we used specific PCR primers from the protease gene as reported elsewhere [15,16].
RNA viral load assay
We quantified HIV-1 RNA viral load in plasma by the reverse transcription (RT)–PCR Amplicor HIV-1 Monitor Assay, version 1.5 (Roche Diagnostic Systems, Branchberg, New Jersey, USA). This assay accurately quantifies HIV-1 subtype A/G recombinant viruses, which is the predominant subtype in Côte d'Ivoire .
HIV-2 RNA plasma viral load was quantified by an RT–PCR prototype assay (Roche Diagnostic Systems). In brief, RNA from 300 μl plasma along with a spiked quantification standard (50 copies) was extracted. A single-tube RT–PCR was performed with rTth DNA polymerase (Roche Molecular Systems, Alameda, California, USA) using 50 μl specimen equivalents per reaction as previously reported . The limit of detection of this assay is 100 copies/ml.
For CD4 T-cell count determination, three-color flow cytometric measurements were carried out with a FACScan flow cytometer (Becton Dickinson) on fresh peripheral whole blood within 4 h of collection into an EDTA tube, using Tritest kit and Multiset software (Becton Dickinson) for analysis. For analysis of immune activation markers, aliquots of cells were stained with commercially available monoclonal antibodies (Becton Dickinson). The markers analyzed were CD3, CD4, CD8, CD38, and HLA-DR. Levels of CD38 and HLA-DR on CD8 T cells are markers of immune activation in HIV infected patients and are prognostic markers for disease progression [12,18]. CD38 is re-expressed on primed cells upon activation in HIV-infected persons; its expression on CD8 T cells increases significantly with disease progression. HLA-DR is a class II antigen that is expressed on activated T cells and in HIV-positive persons, its expression is significantly increased on both CD4 and CD8 T cells.
Analysis of data
A person was considered dually infected if both HIV-1 and HIV-2 DNA PCR testing were positive. The detection limit of the HIV-1 Amplicor Monitor assay was 2.3 log10 copies/ml, and that for the HIV-2 plasma load assay was 2.0 log10 copies/ml. Immunologic data were analyzed as the percentage of the major lymphocyte subsets (CD4 and CD8 T cells) stained with each or a combination of monoclonal antibodies (activation markers).
Log–normal regression models, with left censoring of viral load measurements at detectable limits, were used to estimate differences in HIV-2 viral load for HIV-2 and dually infected persons. These CD4-stratified models (0–199, 200–499, 500+) included covariates for sex, age, and CD4 T-lymphocyte count. Model estimates were log10 transformed for interpretation.
Log–normal regression, with left censoring of undetectable viral load measurements, was also used to estimate the log10 change in HIV-2 viral load associated with a 10 unit increase of CD38 and HLA-DR expression in CD8 T cells. To determine whether the association between HIV-2 viral load and immune activation was comparable, this analysis was stratified by infection status (HIV-2, Dual). In addition, we tested for differences in the association between immune active markers and HIV-2 viral load by infection status, by inclusion of two-way interaction terms in the log–normal regression model.
Mann–Whitney non-parametric methods were used to test for differences in the distribution in levels of CD38 and HLA-DR expression in CD8 T cells. All statistical tests were two-sided with a significance level set at 0.05.
Characteristics of study populations
Between September 1996 and February 2000, 1947 outpatients (1783 HIV-1, 53 HIV-2, 111 HIV dually seropositive), 425 pregnant women (387 HIV-1, 26 HIV-2, 12 HIV dually seropositive), and 322 female sex workers (240 HIV-1, 6 HIV-2, 76 HIV dually seropositive) were tested under different study protocols in Abidjan. Of these persons, sufficient plasma samples were available for 200 individuals. Of these individuals, 52 (26%) were from persons infected with HIV-2 only; 104 (52.0%) were dually seropositive patients, of which 75 (72.1%) were confirmed by DNA PCR testing as dually infected with HIV-1 and HIV-2; and 44 (22.0%) were from HIV-1 infected patients (Table 1). The proportion of males and median age was comparable among persons dually infected and those infected with HIV-2 only. However, persons infected with HIV-2 only had significantly higher median CD4 T-cell counts (582 × 106/l) than HIV dually infected persons (165 × 106/l; P < 0.001; Table 1). Compared with HIV dually infected persons, median CD4 T-cell counts were lower in the 44 patients infected with HIV-1 only (P = 0.046), whereas HIV-1 RNA plasma viral load was similar for HIV-1 infected and HIV dually infected persons (P = 0.26; Table 1).
HIV-2 plasma viral load in dually infected versus HIV-2 only infected patients
In a univariate analysis, among persons with < 200 × 106 CD4 T cells/l, HIV-2 RNA viral load was significantly more likely to be undetected in HIV dually infected persons compared with those infected with HIV-2 only (P = 0.0026) and the distribution of viral load was lower (Table 2). Among persons with 200 × 106 – 500 × 106 CD4 T cells/l, although there was a higher percentage of dually infected persons with undetectable HIV-2 RNA viral load, this finding was not statistically significant (P = 0.57). Interestingly, among persons with > 500 × 106 CD4 T cells/l, HIV-2 RNA viral load was 11.5 times more likely to be undetectable in HIV-2 infected persons (P = 0.0007).
In a CD4-stratified multivariate analysis (Table 2), adjusting for CD4, age and sex, at CD4 T-cell counts < 200 × 106/l HIV-2 RNA viral load was 2.0 log10 copies/ml lower in persons who were dually infected than in those who were infected with HIV-2 only (P < 0.0001). However, at CD4 T-cell counts > 500 × 106/l, HIV-2 viral load was 0.9 log10 copies/ml higher among dually infected persons compared with those infected with HIV-2 only (P < 0.0001). There were no differences in viral load at CD4 T-cell counts 200 × 106 − 499 × 106 cells/l (Table 2). Overall, at lower levels of CD4 T-cell counts, HIV-1 seems to out-replicate HIV-2.
Association between HIV RNA viral load and markers of immune activation
We hypothesized that if HIV-1 out-replicates HIV-2 in dually infected persons as disease progresses, the strength of association between immune activation markers and HIV-2 viral load will be weaker among dually infected patients compared to those infected with HIV-2 only. We found that among persons infected with HIV-2 only, a 10 unit increase in HLA-DR on CD8 T cells resulted in a 0.7 log10 increased in HIV-2 viral load (P = 0.0008). However for dually infected persons, a 10 unit increase in HLA-DR on CD8 T cells resulted in only 0.3 log10 increase in HIV-2 viral load (P = 0.09). A test for interaction between HLA-DR on CD8 T-cell and infection status was not statistically significant (P = 0.100). Concerning CD38 expression on CD8 T cells, in patients infected with HIV-2 a 10 unit increase in levels of expression of CD38 on CD8 T cells resulted in 1.3 log10 increase in HIV-2 viral load (P < 0.0001). However, in HIV dually infected patients, a 10 unit increase in levels of expression of CD38 on CD8 T cells resulted in only 0.4 log10 increase in viral load (P = 0.09). A test for interaction between CD38 on CD8 T cells was significant (P = 0.03). Thus, overall, the strength of the association between immune activation markers and level of HIV-2 viral load was greater for persons who were infected with HIV-2 infected only compared to those who were dually infected, particularly the correlation between CD38 on CD8 T cells and HIV-2 viral load.
Differences in immune activation in persons with quantifiable HIV-2
We reasoned that because persons in whom HIV-2 could be quantified may have actively replicating viruses, no differences may exist in immune activation compared to patients infected with HIV-1 and HIV dually infected patients. We found that among persons with quantifiable HIV-2 viral load, levels of CD38 on CD8 T cells were similar among HIV-1, HIV-2, and HIV dually infected persons (Fig. 1a). However, among persons in whom HIV-2 viral load was not quantifiable, levels of CD8 T cells expressing CD38 were significantly higher among HIV dually infected patients [median, 96%; interquartile range (IQR), 93–98%] compared with those who were infected with HIV-2 only (median, 82; IQR, 72–88%; P = 0.01; Fig. 1b). The proportion of CD8 T cells expressing HLA-DR was higher, albeit insignificantly, among HIV dually infected persons (median, 66%; IQR, 58–73%) compared to those who were infected with HIV-2 only (median, 57%; IQR, 43.5–71%; P = 0.07; Fig. 1c), but significantly higher than in HIV-1 only infected persons (median, 52%; IQR, 40–59%; P < 0.001). Among persons in whom HIV-2 viral load was not quantifiable, the proportion of CD8 T cells expressing HLA-DR was significantly higher among HIV dually infected persons (median, 62%; IQR, 47–68%) than for HIV-1 (median, 52%; IQR, 40–59%; P = 0.05), and HIV-2 infected patients (median, 37%; IQR, 28–48%; P = 0.01; Fig. 1d).
Our results show that after adjusting for CD4 T-cell counts, sex, and age, levels of HIV-2 plasma viral load in dually infected patients differ from those infected with HIV-2 only depending on the level of CD4 T-cell counts. Significantly lower levels of HIV-2 viral load were observed among dually infected patients when CD4 T-cell counts were less than 200 × 106/l. However, at higher levels of CD4 T-cell counts (> 500 × 106 cells/l), HIV-2 viral load was higher in dually infected persons. Our findings also demonstrated that the strength of association between immune activation markers and level of HIV-2 viral load was greater for patients infected with HIV-2 only compared with those who were dually infected, particularly the relationship between CD38 on CD8 T cells and HIV-2 viral load. In addition, the difference in levels of expression of immune activation depended on whether HIV-2 plasma viral load was quantifiable or not, with significantly higher levels observed only among dually infected persons when HIV-2 viral load was non-quantifiable.
There are at least two ways in which dual infection with HIV-1 and HIV-2 may affect HIV-2 replication in vivo: the faster replicating HIV-1 virus may out-replicate the slowly replicating HIV-2 virus; or viral synergism may occur, favored by the strong HIV-1 induced immune activation, thereby up-regulating the less pathogenic HIV-2 virus. Indeed, one study has reported viral synergism of HIV-1 strains isolated from a person dually infected with two genetically and phylogenetically distinct HIV-1 subtype B viruses . Our results that at lower CD4 T-cell counts HIV-2 viral load was significantly decreased in dually infected persons supports the possibility that, in vivo, as disease progresses and CD4 T-cell counts decrease, HIV-1 seems to out-compete HIV-2 in HIV dually infected persons. This assertion is further strengthened by our observation that at lower CD4 T-cell counts, the probability to detect HIV-2 in dually infected persons was significantly decreased (Table 2). Our findings confirm and extend a previous study by Sarr and colleagues who reported lower HIV-2 proviral loads in HIV dually infected persons in Senegal . How can these findings be explained? Unlike HIV-2, HIV-1 has a high turnover rate of about 1 × 109 virions per day [21,22]. This high turnover rate may enable HIV-1 to out-compete HIV-2 for available target CD4 T cells resulting in lower HIV-2 plasma viral load in dually infected persons.
Interestingly, at higher CD4 T-cell counts (> 500 × 106/l), we observed that HIV-2 viral load was significantly higher in dually infected patients than in those infected with HIV-2 only. It is possible that the elevated levels of immune activation observed in dually infected persons may provide a favorable environment for the HIV-2 present in these persons to replicate to significantly higher levels. Alternatively, the dually infected patients we analyzed may have been infected for a longer period of time than the HIV-2 patients as the overall CD4 cell counts levels for the dually infected patients was 3.5 times lower than levels in HIV-2 only patients (Table 1). However, our study may be limited by lack of knowledge of the length since seroconversion for each virus, and the order of infection of the two viruses.
A noteworthy aspect of our study is that when HIV-2 RNA viral load was quantifiable, levels of expression of markers of immune activation were remarkably similar among HIV-2 only and HIV dually infected persons (Fig. 1a and c). These results suggest that active HIV-2 replication leads to similar stimulation of the immune system as HIV-1 or HIV dual infection and may lead to similar rates of disease progression. Indeed, a study conducted in the Gambia showed that HIV-2 infected patients with high RNA plasma viral load lost CD4 T cells rapidly and died faster than those with lower RNA HIV-2 levels . The difference observed when HIV-2 viral load is not quantified may thus be mainly due to the presence of HIV-1 in dually infected patients, as levels of CD38 on CD8 T cells were similar between HIV-1 and dually infected patients. Others have also reported lower levels of immune activation among HIV-2 infected persons than among those infected with HIV-1 . We have also shown previously that levels of expression of HLA-DR on CD8 cells were significantly higher in dually infected persons than in those who were infected with HIV-1 only . Activated T cells are an important determinant of the magnitude of on-going HIV replication because of the close association between viral load and the level of activated CD8 T cells . Moreover, such cells increase the size of the pool of target cells that are susceptible to HIV infection, leading to increased levels of plasma HIV RNA .
Our finding of low levels of HIV-2 plasma viral loads (median, 4.0 log10 copies/ml) among patients with detectable virus is consistent with what we and others have reported [6,11,22,23] and has important implications for clinical care of HIV-2 infected patients. First, although there is increased impetus to introduce antiretroviral drug therapy in Africa, no guidelines exist for the its use in treating HIV-2 and HIV dually infected patients. Our finding, that HIV-2 plasma viral load was more likely to be detectable in the plasma of persons with CD4 T-cell counts < 200 × 106/l (62%) than in persons with higher CD4 T-cell counts is consistent with other studies [5,6,26]. A recent study in France reported median HIV-2 viral loads of 2.1 log10 copies/ml among asymptomatic patients and 3.1 log10 copies/ml among AIDS patients . These findings suggests that use of antiretroviral drug therapy in HIV-2 infected patients could be delayed until CD4 T-cell counts are around 200 × 106/l. This argument is further strengthened by studies of Ariyoshi and colleagues who reported that HIV-2 infected patients with low or undetectable levels of plasma viral load had a low or no disease progression . Secondly, the significantly lower levels of HIV-2 plasma viral load among dually infected persons suggest that in vivo HIV disease in dually infected patients is driven by HIV-1 and may partly explain the similarity in the clinical course of HIV-1 and dually infected persons . Thus, it could be appropriate to treat dually infected patients according to guidelines used for treating persons infected with HIV-1 .
In summary, we have shown that in HIV dually infected persons, levels of HIV-2 plasma viral load depend on the immune status of the patient, with lower viral load levels observed at decreased CD4 T-cell counts. These findings suggest that in dually infected patients, HIV-1 out-replicates HIV-2 in vivo as disease progresses. Thus, at lower CD4 cell counts, dually infected patients should be managed as those infected with HIV-1 only.
We thank E. Boateng, B. Vuylsteke, and D. Adams for technical assistance. We thank also C. Phares and E. Pascal for statistical analysis support. Thanks to S. Kwok and C. Christopherson for providing kits and reagents for RNA quantification from Roche Molecular Systems.
Sponsorship: Supported by the Division of HIV/AIDS Prevention, Surveillance and Epidemiology, National Center for HIV, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia, USA. Part of this work was also supported by the Belgian ‘Fonds voor Wetenschappelijk Onderzoek (F.W.O.) Vlaanderen’ grants no. G.0396.99.
1. Clavel F, Guetard D, Brun-Vezinet F, Chamaret S, Rey MA, Santos-Ferreira MO, et al
. Isolation of a new human retrovirus from West African patients with AIDS.Science
2. Kanki PJ, Barin F, M'boup S, Alan JS, Romet-Lemonne JS, Marlink R, et al
. New human T – lymphotropic retrovirus related to simian T-lymphotropic virus type III (STLV-IIIAGM).Science
3. Adjorlolo-Johnson G, De Cock KM, Ekpini E, Vetter KM, Sibailly TS, Ou CY, et al
. Prospective comparison of mother-to-child transmission of HIV-1 and HIV-2 in Abidjan, Ivory Coast.JAMA
4. Marlink RP, Kank PJ, Thior I, Travers K, Eisen G, Siby T, et al
. Reduced rate of disease development after HIV-2 infection as compared to HIV-1.Science
, 1994 265
5. Popper SJ, Sarr AD, Travers KU, Gueye-Ndiaye A, Mboup S, Essex ME, et al
. Lower human immunodeficiency virus (HIV) type 2 viral load reflects the difference in pathogenicity of HIV-1 and HIV-2.J Infect Dis
6. Shanmugam V, Switzer MW, Nkengasong JN, Garcia-Lerma G, Green TA, Ekpini E, et al
. Lower HIV-2 plasma viral loads may explain differences between the natural histories of HIV-1 and HIV-2 infections.J Acquir Immune Defic Syndr
7. Simon F, Matheron S, Tamalet C, Loussert-Ajaka I, Batczak S, Pepin JM, et al
. Cellular and plasma viral load in patients infected with HIV-2.AIDS
8. Ishikawa K, Fransen K, Ariyoshi K, Nkengasong JN, Janssens W, Heyndrickx L, et al
. Improved detection of HIV-2 proviral DNA in dually seroreactive individuals by PCR.AIDS
9. Sarr AD, Hamel DJ, Thior I, Kokkotou E, Sankale JL, Siby T, et al
. HIV-1 and HIV-2 dual infection: Lack of HIV-2 provirus correlates with low CD4+ lymphocyte counts.AIDS
10. Nkengasong JN, Kestens L, Ghys PD, Koblavi-Deme S, Otten RA, Bile C, et al
. Dual infection with human immunodeficiency virus type 1 and type 2: impact on HIV type 1 viral load and immune activation markers in HIV-seropositive female sex workers in Abidjan, Ivory Coast.AIDS Res Human Retroviruses
11. Andersson S, Norrgren H, da Silva Z, Biague A, Bamba S, Kwok S, et al
. Plasma viral load in HIV-1 and HIV-2 singly and dually infected individuals in Guinea-Bissau, West Africa: significantly lower plasma virus set point in HIV-2 infection than in HIV-1 infection.Arch Intern Med
12. Carbone J, Gil J, Benito JM, Navarro J, Munoz-Fernandez A, Bartolome J, et al
. Increased levels of activated subsets of CD4 T cells add to the prognostic value of low T cell counts in a cohort of HIV-infected drug users.AIDS
13. Nkengasong JN, Maurice C, Koblavi S, Kalou M, Yavo D, Maran M, et al
. Evaluation of HIV serial and parallel serologic testing algorithms in Abidjan, Côte d'Ivoire.AIDS
14. Nkengasong JN, Maurice C, Koblavi S, Kalou M, Bile C, Yavo D, et al
. Field evaluation of a combination of monospecific enzyme linked immunosorbent assays for type-specific serodiagnosis of human immunodeficiency virus type 1 (HIV-1) and HIV-2 infections in HIV seropositive persons in Abidjan, Côte d'Ivoire.J Clin Microbiol
15. Ellenberger DL, Pieniazek D, Nkengasong JN, Luo CC, Devare S, Maurice C, et al
. Genetic analysis of human immunodeficiency virus in Abidjan, Ivory Coast reveals predominance of HIV type 1 subtype A and introduction of subtypes G.AIDS Res Hum Retroviruses
16. Pieniazek D, Ellenberger D, Janini LM, Ramos AC, Nkengasong JN, Sassan-Morokro M, et al
. Predominance of human immunodeficiency virus type 2 subtype B in Abidjan, Ivory Coast.AIDS Res Hum
17. Nkengasong JN, Chi-Cheng L, Abouya L, Pieniazek D, Maurice C, Sassan-Morokro M, et al
. Distribution of HIV-1 subtypes among HIV-seropositive patients in the interior of Côte d'Ivoire.J Acquir Immune Defic Syndr
18. Kestens L, Vanham G, Vereecken C, Vandenbruaene M, Vercauteren G, Colebunders RL, et al
: Selective increase of activation antigens HLA-DR and CD38 on CD4+ CD45RO+ T lymphocytes during HIV-1 infection.Clin Exp Immunol
19. Wang B, Lal RB, Dwyer DE, Miranda Saksena M, Boadle R, Cunningham AL, et al
. Molecular and biological interactions between two HIV-1 strains from a co-infected patient reveal the first evidence in favor of viral synergism.Virology
20. Sarr AD, Popper S, Thior I, Hamel DJ, Sankale JL, Siby T, et al
. Relation between HIV-2 proviral load and CD4+ lymphocyte count differs in monotypic and dual HIV infection.J Hum Virol
21. Ho DD, Neuman AU, Perelson AS, Chen W, Leonard JM, Markowitz M. Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection.Nature
22. Popper SJ, Sarr AD, Gueye-Ndiaye A, Mboup S, Essex ME, Kanki PJ. Low plasma human immunodeficiency virus type 2 viral load is independent of proviral load: low virus production in vivo.J Virol
23. Ariyoshi K, Jaffar S, Alabi AS, Blanchard T, Schim van der Loeff M, Awasana AA, et al
. Plasma RNA viral load predicts the rate of CD4 T cell decline and death in HIV-2-infected patients in West Africa.AIDS
24. Michel P, Toure Balde A, Roussilhon C, Aribot G, Sarthou JL, Gougeon ML. Reduced Immune activation and T cell apoptosis in human immunodeficiency virus type 2 compared with type 1: Correlation of T cell apoptosis with β2 microglobulin concentration and disease evolution.J Infect Dis
25. Finzi D, and Siliciano RF. Viral dynamics in HIV-1 infection.Cell
26. Damond F, Gueudin M, Pueyo S, Farfara I, Robertson DL, Descamps D, et al
. Plasma RNA viral load in human immunodeficiency virus type 2 subtype A and subtype B infections.J Clin Microbiol
27. Kassim S, Sassan-Morokro M, Ackah A, Abouya LY, Digbeu H, Yesso G, et al
. Two-year follow-up of persons with HIV-1 and HIV-2 associated pulmonary tuberculosis treated with short-course chemotherapy in West Africa.AIDS
28. Yeni PG, Hammer SM, Carpenter CC, Cooper DA, Fischl MA, Gatell JM, et al
. Antiretroviral treatment for adult HIV infection in 2002: updated recommendations of the International AIDS Society- USA Panel.JAMA