Sustained viremia during highly active antiretroviral therapy with accelerated proviral DNA decay in the setting of infection with syphilis

Abstract

Highly active antiretroviral therapy (HAART) suppresses viral replication to undetectable levels in most adherent HIV-1-infected individuals. Although intermittent isolated episodes of detectable viremia are common and not necessarily associated with treatment failure [1], persistently detectable viremia is generally regarded as such and often results in a change in regimen [2]. Although sustained viremia during therapy is probably a result of the emergence of a drug-resistant variant [3], non-adherence [4,5] or superinfection with a drug-resistant virus [6] could also lead to viremia during therapy. Measurable HIV-1-RNA levels in the plasma may also reflect the release of particles from recently activated, previously infected cells harboring provirus, including resting lymphocytes and macrophages. We detail the course of a 28-year-old Caucasian gay man, who was found to be viremic on three consecutive monthly visits after 18 months of effective HAART. The viremia resolved without a change in treatment and no viremia has been detectable for the past 17 months. The patient was self-referred to the Aaron Diamond AIDS Research Center on 4 February 2000. The subject had participated in a recombinant envelope vaccine trial, was negative serologically on 13 December 1999 and was newly HIV-1 positive on 26 January 2000. On 8 February 2000 a regimen of abacavir 600 mg, lamivudine 300 mg, indinavir 2400 mg and amprenavir 900 mg a day was initiated. After 16 weeks, indinavir was discontinued and substituted with amprenavir 1200 mg/ritonavir 200 mg a day. On 30 October 2001 the patient presented with complaints of lymphadenopathy in the inguinal region. He described two brief episodes of low-grade fevers in August and September of 2001, and concomitantly reported multiple anal-receptive sexual encounters without the use of condoms. Examination revealed bilateral non-tender lymphadenopathy. A serological test for syphilis was newly positive at a 1 : 64 dilution. The patient was treated with three injections of 2.4 million units of benzathine penicillin. Low-level viremia was detected at 4-week intervals beginning at week 68 (3 July 2001), and continued until week 84 (14 September 2001) of HAART (Fig. 1a). To investigate whether the viremia resulted from superinfection, we compared the V3 region of the envelope gene from plasma virus at baseline and week 84. The V3 region was amplified using a modified nested reverse transcriptase–polymerase chain reaction as previously described [7]. The consensus sequences of the two timepoints were 100% homologous, providing conclusive evidence that the virus that initially infected this patient was the same as that being detected in the patient's plasma during rebound.Fig. 1.: (a) Longitudinal determination of plasma viremia (HIV-1-RNA copies/ml in log 10 ) and CD4 T-cell count (cells/mm 3 ) while on antiretroviral therapy. Viral load (□) is plotted on a semi-log scale, whereas CD4 cell count (●) is on a linear scale. The viral load values at the times of ‘blip’ are denoted as VL. The dotted line indicates the lower limit of detection of HIV-1 RNA (< 50 copies/ml). (b) Proviral DNA (▴) and HIV-1 RNA (□) (copies/ml) of patient from weeks 55 to 110 of highly active antiretroviral therapy. A fivefold drop in proviral DNA is noted immediately before increased viremia. The dotted line indicates the lower limit of detection of HIV-1 RNA (< 50 copies/ml). HAART, Highly active antiretroviral therapy.Having ruled out superinfection, we then asked whether resistance to the antiviral agent was emerging. Clonal sequencing of the HIV-1 protease and reverse transcriptase regions was performed. Six clones at baseline and 10 clones at week 84 were studied. Reverse transcriptase–polymerase chain reaction, cloning and sequencing of pol were performed as described elsewhere [8,9]. No resistance-conferring mutations in protease or reverse transcriptase were detected. We then asked whether the detectable viremia was a result of low-level replication from a sequestered site. We performed longitudinal assessments of the levels of proviral DNA in patient peripheral blood mononuclear cells. We hypothesized that the proviral DNA levels would remain stable or increase slightly if there was ongoing viral replication during HAART. Conversely, if viremia resulted from the expression of HIV-1 proviral DNA without replication, the proviral DNA levels would decrease. Proviral DNA was measured using the Amplicor HIV-DNA Monitor Test (Roche, Alameda, CA, USA) with modifications as published [10]. Proviral DNA levels decreased sharply between weeks 68 and 76 (Fig. 1b). These data suggest that the measurable HIV-1-RNA levels may have originated from activated reservoirs. We hypothesize that this activation is a result of immune stimulation caused by infection with syphilis. This observation suggests that efforts should be made to study apparent virological failure carefully before a particular regimen is abandoned.