A 22-year old gay man presented on July 25, 2002, with gastrointestinal symptoms, sweats, rash, and lymphadenopathy. He had tested HIV-negative 3 months previously but reported recent unprotected anal intercourse with multiple partners. His regular partner was HIV-positive. An EIA screening test for HIV-1 and HIV-2 antibody (Vironostika) was weakly reactive (Table 1). An additional test for HIV 1/2 antibody (Abbott IMx) was negative. P24 antigen was not detected. HIV antibody tests performed by a UK reference center, the Health Protection Agency Centre for Infections (HPA), were all negative (Innotest HIV 1/2, Murex HIV antigen/antibody, in-house GACPAT HIV-1 and GACPAT HIV-2 assays1).
The patient was advised that the HIV test was indeterminate. On August 15, 2002, a further blood sample produced the same weakly reactive result (Vironostika) and was negative by the IMx assay.
Blood drawn on August 20, 2002, was reactive on both HIV antibody tests (Vironostika and IMx). A confirmatory immunoassay (Inno-LIA) showed a pattern consistent with antibody to HIV-1 (gp120, gp41, p24 and p17 positive; gp31, gp105, and gp36 negative). His CD4 count was 718 × 109 cells/mL, and HIV viral load 147 copies/mL (Bayer bDNA assay).
On August 23, 2002, a further blood sample confirmed reactivity in 2 tests for anti-HIV-1/-2 antibodies (Vironostika and IMx). His CD4 count was 1158 × 109 cells/mL and viral load below the limit of detection of 50 copies/mL (Bayer bDNA). This sample was later (February 2004) sent to HPA and reactive results obtained by Anti-HIV 1/2 (DetectHIV IEA, Adaltis), Anti-HIV 1/2 and p24 antigen (Abbott Murex Combined EIA), in-house GACPAT HIV-1 and in-house GACPAT HIV-2 assays (Table 2). Abbott Murex Wellcozyme Anti-HIV-1 Recombinant and Anti-HIV-2 ICE EIAs were negative. Western Blot (Genelabs) was positive for antibodies to p24 and gp160, showed a “trace” of antibodies to p66 and gp120, and was negative for antibodies to p17, p55, p31, p51, gp41, and gp36 (HIV-2). See Table 3 for details of antibody tests. The viral load was <250 copies/mL (Roche Cobas Monitor), and p24 antigen (Biorad EIA) was not detected.
On the basis of the results from the specimen collected on August 23, 2002, apparently showing an evolving anti-HIV antibody response, the patient was informed that his confirmatory HIV test was positive. His symptoms resolved in 2 weeks. Over the following 8 months, he was reviewed regularly. His CD4 count stayed high, at above 890 × 109 cells/mL. HIV RNA remained undetectable (<50 copies/mL) by the Bayer assay, the Roche ultrasensitive RT-PCR, and the Nuclicens assay (NASBA). In view of these persistently undetectable viral loads, the HIV antibody test was repeated in June 2003 and was negative.
HIV diagnostic testing was repeated (using HIV antibody tests A and B, see Table 1 footnote) in December 2003, March 2004, July 2004, and November 2005, during which period it has continued to be negative. His CD4 count has remained high and HIV RNA undetectable by Bayer and Roche assays. In August 2004 and November 2005, proviral DNA was not detected. In October 2004, testing for HIV-1 RNA by qualitative PCR was also negative.
Acute phase serum was negative for EBV and CMV IgM. Sampling error and contamination were excluded by forensic DNA profile testing of multiple samples. Unfortunately, the patient has declined further investigations at our center.
This man was at high risk of HIV acquisition and had symptoms consistent with HIV seroconversion. His HIV antibody response evolved from negative to equivocal to a combination of reactions considered to be positive, and the interpretation was that he had developed HIV infection. However, persistently undetectable viral loads prompted repeat antibody tests 10 months later, which were negative. P24 antigen and proviral DNA could not be detected.
Seroreversion refers to the loss of previously documented HIV antibody. Individuals with advanced HIV disease may lose detectable antibody, permanently or transiently, as a result of impaired immunity.2,3 Seroreversion has also occurred in acute HIV-1 infection treated early with HAART,4 when it seemed that durable virological suppression resulted in incomplete evolution of the HIV antibody response, or even loss of specific antibody.
In our case, repeat antibody testing was prompted by persistently undetectable viral loads. However, it is well established that some HIV-positive individuals maintain undetectable viral loads without treatment. A French study followed 2800 HIV-positive patients for 10 years and found 5 untreated patients who never had a viral load above 50 copies/mL.5 They were described as “spontaneous controllers.” HIV infection was confirmed by finding very low and stable proviral DNA in peripheral blood monocytes. HIV-specific CD8+ T cells were detected, and HIV antibodies remained positive. Several factors could result in such spontaneous viral control, including attenuation of the HIV virus, reduced susceptibility of the host CD4 cells to HIV entry, and strong HIV-specific immune responses.
Our patient did not have immunosuppression or antiretroviral drugs to account for his seroreversion. Nor did he have detectable proviral DNA to indicate he was infected with HIV. His results are therefore intriguing. The patient has proposed that he was infected with HIV but cleared it with his immune system. This possibility should be given serious consideration, but there are several arguments against. First, the antibody tests became reactive in a sequence that is not typical for HIV seroconversion: a specimen collected 4 weeks after presentation with a suspected seroconversion illness showed a combination of strong reactivity in both HIV-1 and HIV-2 GACPAT assays and weak Western Blot reactivity, together with a negative result in the Wellcozyme HIV-1 Recombinant EIA (a competitive assay). The GACPAT assays are primarily intended to discriminate between HIV-1 and HIV-2 infection by detecting the proportion of the antibody response that is type-specific. They show lower sensitivity early in seroconversion than commercial screening tests (such as Wellcozyme HIV Recombinant EIA), and moreover would not be expected to be so strongly reactive in the light of the very restricted and weak Western blot reactivity. Reactions with the p24 band in the absence of HIV infection occur frequently, and false gp160 reactions also occur. The Murex HIV-1 and HIV-2 type-specific EIAs failed to confirm the GACPAT reactivities. Second, p24 antigen is usually present before anti-HIV antibodies, which it was not in this case. Third, during seroconversion, HIV viral load is typically high, in the order of 104 to 105 copies/mL.6,7 Low level false-positive HIV viral loads are occasionally found in uninfected individuals, generally at levels of up to 2000 copies/mL.8,9 In this context, the initial viral load of 147 copies/mL is likely to represent a false-positive result.
It is possible that this man developed antibodies when challenged with an HIV-1 antigenic stimulus, but that infection did not occur. His immune system may have responded in a highly unusual manner, producing low levels of antibody in an atypical sequence, but rapidly controlling and then clearing the virus. We are not aware that such a “near-miss” HIV infection, with seroconversion and seroreversion, has been documented before. This theory would have been more convincing, although still highly controversial, if the patient had consented to further investigations and we had found strong HIV-specific cytotoxic T lymphocytes (CTLs). In other multiply exposed but uninfected individuals, HIV-specific CTL responses have been associated with recent exposure to HIV-1, although they have not been found to be protective against infection.10 It would also have been informative to have performed ultrasensitive PCR tests for HIV DNA, which can demonstrate extremely low levels of HIV-1 when proviral DNA is negative by routine techniques.11 It has also been possible in rare cases to use limiting-dilution PCR to demonstrate extraordinarily low level of persistent virus in the resting CD4+ T cells of some exposed but seronegative individuals.12
Another possibility is that of passive antibody acquisition, but it is difficult to imagine that sufficient transfer occurred for antibody to be detectable without resulting in infection.
Alternatively, the reactive HIV tests may have been due to nonspecific reactivity. If this is the case, it is a most unusual pattern of reactivity, in that it affected a broad range of HIV antibody assays, including reactions on immunoblots sufficient to fulfil criteria for anti-HIV-1 positivity.
This case highlights the importance of questioning a diagnosis of HIV infection in the presence of low or undetectable viral load. Although “HIV controllers” do exist, HIV antibodies would be expected to remain positive. In this exceptional case, HIV antibody reactivities disappeared, and careful analysis of the clinical and immunologic picture led us to conclude that it is unlikely that this man was ever infected with HIV.
This case received international publicity and caused public anxiety about the reliability of HIV tests. While “false positivity” in one HIV antibody test does occur infrequently, laboratories perform multiple different antibody tests on screen-reactive samples, all of which should be clearly positive before concluding established HIV infection is present. This minimizes, but does not entirely exclude, the possibility of a false-positive diagnosis. In this case, most of the antibody tests that were applied during the month after presentation were reactive, but further investigation points heavily towards an absence of HIV infection. These findings appear to be exceptionally rare.
1. Parry JV, Connell JA, Garcia AB, et al. GACPAT HIV 1 + 2: A simple, inexpensive assay to screen for, and discriminate between, anti-HIV 1 and anti-HIV 2. J Med Virol 1995; 45:10–16.
2. Montagnier L, Brenner C, Chamaret S, et al. Human immunodeficiency virus infection and AIDS in a person with negative serology. J Infect Dis 1997; 175:955–959.
3. Zaaijer HL, Bloemer MH, Lelie PN. Temporary seronegativity in a HIV type 1-infected man. J Med Virol 1997; 51:80–82.
4. Kassutto S, Johnston MN, Rosenberg ES. Incomplete HIV type 1 antibody evolution and seroreversion in acutely infected individuals treated with early antiretroviral therapy. Clin Infect Dis 2005; 40:868–873.
5. Lambotte O, Boufassa F, Madec Y, et al. HIV controllers: A homogenous group of HIV-1-infected patients with spontaneous control of viral replication. Clin Infect Dis 2005; 41:1053–1056.
6. Hubert JB, Burgard M, Dussaix E, et al. Natural history of serum HIV-1 RNA levels in 330 patients with a known date of infection. AIDS 2000; 14:123–131.
7. Lyles RH, Munoz A, Yamashita TE, et al. Natural history of human immunodeficiency virus type 1 viraemia after seroconversion and proximal to aids in a large cohort of homosexual men. J Infect Dis 2000; 181:872–880.
8. Daar ES, Little S, Pitt J, et al. Diagnosis of Primary HIV-1 Infection. Ann Intern Med 2001; 134:25–29.
9. Rich JD, Merriman NA, Mylonakis E, et al. Misdiagnosis of HIV infection by HIV-1 plasma viral load testing: A case series. Ann Intern Med 1999; 130:37–39.
10. Kaul R, Rutherford J, Rowland-Jones SL, et al. HIV-1 Env-specific cytotoxic T-lymphocyte responses in exposed, uninfected Kenyan sex workers: A prospective analysis. AIDS 2004; 18:2087–2089.
11. Kloosterboer N, Groeneveld PHP, Jansen CA, et al. Natural controlled HIV infection: Preserved HIV-specific immunity despite undetectable replication competent virus. J Virol 2005; 339:70–80.
12. Zhu T, Corey L, Hwangbo Y, et al. Persistence of extraordinarily low levels of genetically homogeneous human immunodeficiency virus type 1 in exposed seronegative individuals. J Virol 2003; 77:6108–6116.