Effects of High-Dose Intravenous Immunoglobulin on Virus Load in HIV-Infected Children

To the Editor:

Inappropriate, uncontrolled activation of the immune system has been suggested as a major component of HIV pathology ⁽¹⁾, and immunologic downregulation with systemic glucocorticoids has been shown to reduce plasma HIV p24 antigen levels in HIV-infected children acutely ⁽²⁾. High-dose (2 g/kg body weight) of intravenous immunoglobulin (IVIG) has been used to treat diseases characterized by pathologic immune activation ⁽³⁾, and IVIG has been shown to suppress cytokine-dependent human T-cell proliferation in vitro ⁽⁴⁾ and increased tumor necrosis factor-α activity in HIV in vivo ⁽⁵⁾. We hypothesized that IVIG would decrease plasma HIV and that this reduction would correlate with reduced T-cell activation. In this pilot study we measured peripheral blood HIV virus load and activation antigen expression on peripheral blood T cells in stable, HIV-infected children before and after a single, high-dose infusion of IVIG.

Ten HIV-infected children between 2 and 10 years of age (mean, 5 years) were studied. All were on stable antiretroviral regimens for at least 30 days prior to IVIG infusion and had HIV virus loads >10,000 copies/ml (log₁₀ >4.0). The study protocol was approved by the Institutional Review Board and informed consent was obtained from parents or legal guardians and from children 7 years of age or older.

Laboratory studies were obtained before IVIG infusion at two visits 2 to 4 weeks apart (the mean of which was considered baseline); then 1, 3, 7, 14, and 28 days after infusion of 2 g/kg of IVIG (Gammagard-S/D, Baxter Health Care Corp., Glendale, CA, U.S.A.). T-cell subsets were analyzed using standard single and two-color immunofluorescent cytometric techniques for CD3 (T cells), CD4 (helper T cells), CD8 (suppressor T cells), and CD25 and human leukocyte antigen (HLA)-DR (activation antigens). Phenotypic analyses were run within 24 hours of each blood draw; laboratory controls consisted of healthy adult volunteer blood donors. Quantitative plasma HIV RNA determinations were performed using a second-generation Roche kit (Roche Diagnostic Systems, Branchburg, NJ, U.S.A.), and HIV p24 antigen was measured with an immune complex-dissociated enzyme immunoassay using the Coulter ICD-Prep kit (Coulter Corp., Miami, FL, U.S.A.). For each study subject, all plasma specimens were stored frozen until tested concurrently.

Laboratory values at each postinfusion evaluation were expressed as the changes from baseline. In addition, t- tests were performed for the differences between the mean changes preinfusion and postinfusion. Findings that included p values < .05 were considered to be significant.

All infusions were well tolerated and no acute adverse events occurred. CD3⁺, CD4⁺, and CD8⁺ T cells expressing HLA-DR were increased in HIV-infected study subjects relative to those found in normal individuals (2%4 ± 16% versus 4% ± 4%; 12% ± 8% versus 4% ± 3%; and 28% ± 17% versus 7% ± 6%, respectively). In contrast, enhanced expression of CD25 was restricted to CD4⁺ T cells (11% ± 5% versus 2% ± 2% for controls). IVIG did not alter the percentages of circulating CD3⁺, CD4⁺, or CD8⁺ cells that expressed CD25 or HLA-DR over the course of the study (data not shown).

IVIG infusion was not associated with a reduction in HIV in any patient. Mean HIV p24 increased from a baseline level of 669 pg/ml to 1102, 1194, 1254, 1304, and 1225 at 1, 3, 7, 14, and 28 days postinfusion, respectively. These differences reached statistical significance (p < .05 for days 1, 3, 14, and 28). Similarly, mean baseline HIV RNA log₁₀ rose significantly (p < .05) from 4.60 to 4.96, 4.85, and 4.90 at 1, 3, and 7 days postinfusion, respectively (Fig. 1). Mean HIV RNA was not statistically different from baseline at 14 and 28 days postinfusion, 4.79 and 4.77, respectively.

Fig. 1

T cells from patients with HIV express activation-associated antigens to a greater extent than normal individuals. Previously, we have shown that systemic glucocorticoid treatment reduced plasma HIV p24 antigen in a dose-dependent fashion, and that this reduction was associated with decreased expression of cell surface activation markers ⁽²⁾. Therefore, we chose to evaluate the effects of a single, 2 g/kg dose of IVIG upon lymphocyte activation markers (CD25 and HLA-DR) and virus levels (ICD HIV p24 and quantitative HIV RNA) in HIV-infected children. In contrast to our findings with glucocorticoid treatment, no beneficial effect was found following IVIG infusion. Further, in most patients studied, high-dose IVIG was associated with a transient, small increase in circulating virus as measured by ICD p24 and/or quantitative HIV RNA. The mechanism by which IVIG might transiently increase circulating HIV is unknown but may be related to displacement of adherent virions from lymphoid tissue dendritic cells.

* Joseph A. Church

* Sarah Fox

† Edward Gomperts