Objective:

To evaluate immune reconstitution within HIV-infected lymphoid tissue during highly active antiretroviral therapy (HAART).

Design and methods:

In situ cellular responses were studied in sequential tonsillar biopsies in three asymptomatic HIV-infected (CD4 cells greater than 400× 10 6/l) antiretroviral treatment-naive volunteers enrolled in a clinical trial to determine the early effect of HAART. Computerized image analysis was used to study immunohistochemically stained sequential tonsil sections for the patterns of local cytokine production, chemokine receptor expression and cellular distribution. Replicate quantitative assessments of samples before and after 4 weeks of therapy were used for the evaluation of drug effects and compared with four uninfected controls. Tonsillar HIV proviral-DNA was determined by fluorescent in situ 5′-nuclease assay.

Results:

HIV-infected tonsil tissue was characterized by extensive pro-inflammatory and type 1 cytokine expression. A five-to 15-fold elevation of interleukin (IL)-1α, IL-12, IL-2 and interferon (IFN)-γ protein expression was found compared with controls, and each encompassed a mean of at least 4.5% of the tissue compartment. This was reduced by 20–90% in all individuals after 4 weeks of HAART. In contrast, type 2 cytokine expression (IL-4, IL-10), plus tumour necrosis factor (TNF)-α, remained low throughout the study. HAART reduced, by 40%, the expression of HIV co-receptors, CCR5 and CXCR4, which initially were elevated four to six times over the control values. In addition, the myelomonocytic inflammatory proteins, CD68 and calprotectin, diminished by 26–83% after therapy. The HIV RNA was reduced to undetectable levels in plasma by HAART. However, a large pool of tonsil cells (2–7%), remained HIV DNA positive after 4 weeks of therapy.

Conclusions:

Although immune activation may be the direct consequence of HIV replication, HAART-associated reconstitution begins with a reduction in inflammatory cytokine production which precedes the elimination of local proviral reservoirs.