Cellular HIV-1 DNA levels in patients receiving antiretroviral therapy strongly correlate with therapy initiation timing but not with therapy duration

BACKGROUND: Viral reservoir size refers to cellular human immunodeficiency virus-1 (HIV-1) DNA levels in CD4(+ )T lymphocytes of peripheral blood obtained from patients with plasma HIV-1-RNA levels (viral load, VL) maintained below the detection limit by antiretroviral therapy (ART). We measured HIV-1 DNA levels in CD4(+ )lymphocytes in such patients to investigate their clinical significance. METHODS: CD4(+ )T lymphocytes were isolated from the peripheral blood of 61 patients with a VL maintained at less than 50 copies/ml for at least 4 months by ART and total DNA was purified. HIV-1 DNA was quantified by nested PCR to calculate the copy number per 1 million CD4(+ )lymphocytes (relative amount) and the copy number in 1 ml of blood (absolute amount). For statistical analysis, the Spearman rank or Wilcoxon signed-rank test was used, with a significance level of 5%. RESULTS: CD4 cell counts at the time of sampling negatively correlated with the relative amount of HIV-1 DNA (median = 33 copies/million CD4(+ )lymphocytes; interquartile range [IQR] = 7-123 copies/million CD4(+ )lymphocytes), but were not correlated with the absolute amounts (median = 17 copies/ml; IQR = 5-67 copies/ml). Both absolute and relative amounts of HIV-1 DNA were significantly lower in six patients in whom ART was initiated before positive seroconversion than in 55 patients in whom ART was initiated in the chronic phase, as shown by Western blotting. CD4 cell counts before ART introduction were also negatively correlated with both the relative and absolute amounts of HIV-1 DNA. Only the relative amounts of HIV-1 DNA negatively correlated with the duration of VL maintenance below the detection limit, while the absolute amounts were not significantly correlated with this period. CONCLUSIONS: The amounts of cellular HIV-1 DNA in patients with VLs maintained below the detection limit by the introduction of ART correlated with the timing of ART initiation but not with the duration of ART. In addition, CD4(+ )T lymphocytes, which were newly generated by ART, diluted latently infected cells, indicating that measurements of the relative amounts of cellular HIV-1 DNA might be underestimated.