Portal de Periódicos da CAPES

Rebound of plasma HIV viral load following prolonged suppression with combination therapy

1998; Lippincott Williams & Wilkins; Volume: 12; Issue: 11 Linguagem: Inglês

10.1097/00002030-199811000-00028

1473-5571

Julio Montaner, Marianne Harris, T Mo, P. Richard Harrigan,

HIV/AIDS Research and Interventions

Recent reports [1–3] have provided evidence of latent HIV reservoirs not eliminated by current antiretroviral therapies. This report provides clinical support for these laboratory observations, suggesting that HIV eradication [4] may not be readily achievable. A 42-year-old HIV-positive male injecting drug user (mean pretherapy plasma viral load, 2142 copies/ml, Amplicor; Roche Diagnostics, Mississauga, Canada; CD4 cell count, 425 × 106/l) commenced therapy on 21 March 1995 with zidovudine, didanosine, and nevirapine within the INCAS trial (Italy, The Netherlands, Canada and Australia) [5,6]. By week 2 plasma viral load had decreased below the limit of quantification of the Ultra Direct assay (Roche Molecular Systems, Alameda, California, USA) [7] and remained below detection for 28 months. In June 1995, he discontinued didanosine due to gastrointestinal intolerance. In June 1996, lamivudine was added. On 2 October 1997 he was found unconscious following a suspected cocaine overdose. On 22 October 1997 (26 h before death from anoxic brain injury), plasma viral load was 123 036 copies/ml. Antiviral therapy had been stopped for a minimum of 3 weeks (since 2 October 1997) and a maximum of 9 weeks (since 20 August 1997). Sequence analysis of plasma viral reverse transcriptase region (codons 23–236) indicated no mutations characteristic of resistance to nevirapine, didanosine, zidovudine or lamivudine upon viral rebound. Only two amino-acid changes (N162D and K173G, neither associated with known drug resistance) were noted after 2 years of therapy. This lack of selection of mutations conferring resistance to lamivudine or nevirapine (either of which require only a single nucleic acid sub-stitution for high level resistance) upon viral rebound has been taken as evidence that significant viral replication had not occurred during therapy. There was no evidence of changes in the predominant viral sequences in the V3 loop of the viral envelope over this time period, providing strong evidence of viral rebound rather than reinfection. The possibility of fully suppressing HIV replication with current antiretroviral therapy has led to the development of the eradication hypothesis [4]. Briefly, if HIV replication is halted, the time to eradication (estimated to be 28–37 months) is dictated by the half-life of the longest-lived virus-producing cells [4]. This patient experienced a rapid viral rebound to substantially above pretherapy levels, despite 28 months of viral suppression below the limits of the most sensitive assay. The magnitude of the increase could have been in part influenced by the patient's condition. We were only able to test plasma viral load once after therapy discontinuation and did not have access to other tissue samples. Of note, a single virus-producing cell could produce approximately 1013 virions in 3 weeks, sufficient to explain the observed plasma viral load rebound. This is based on the assumption that multiple rounds of infection occur with a rate constant k = ln(n)/t, where n is the number of infectious virions produced per cell (approximately 100), and t is the time required for one infection cycle (approximately 3 days) [8,9]. In summary, our case, together with recent reports of HIV recovery after prolonged non-quantifiable plasma viral load during therapy, suggests that viral eradication of HIV may not be a realistic goal at present.