Elsevier

The Lancet

Volume 362, Issue 9388, 20 September 2003, Pages 945-950
The Lancet

Articles
Anatomical loci of HIV-associated immune activation and association with viraemia

https://doi.org/10.1016/S0140-6736(03)14363-2Get rights and content

Summary

Background

Lymphocyte activation, associated with vaccination or infection, can be measured by positron emission tomography (PET). We investigated the ability of PET to detect and measure magnitude of lymph-node activation among asymptomatic HIV-1-infected individuals.

Methods

Initially we assessed PET response in eight HIV-1-uninfected individuals who had received licensed killed influenza vaccine. In an urban teaching hospital, we recruited 12 patients recently infected with HIV-1 (<18 months since seroconversion) and 11 chronic long-term HIV-1 patients who had stable viraemia by RT-PCR (non-progressors). After injection with fluorine-18-labelled fluorodeoxyglucose, patients underwent PET. We correlated summed PET signal from nodes with viral load by linear regression on log-transformed values.

Findings

Node activation was more localised after vaccination than after HIV-1 infection. In early and chronic HIV-1 disease, node activation was greater in cervical and axillary than in inguinal and iliac chains (p<0·0001), and summed PET signal correlated with viraemia across a 4 log range (r2=0·98, p<0·0001). Non-progressors had small numbers of persistently active nodes, most of which were surgically accessible.

Interpretation

The anatomical restriction we noted may reflect microenvironmental niche selection, and tight correlation of PET signal with viraemia suggests target-cell activation determines steady-state viral replication.

Introduction

Increased turnover of uninfected lymphocyte populations is thought to be a major component in the pathogenesis of retroviral infections such as HIV and simian immunodeficiency virus. Thus, despite viral loads as high as 107 copies/mL, sooty mangabeys infected with simian immunodeficiency virus have lymphocyte turnover rates indistinguishable from uninfected animals, and suffer no decline in CD4 T cells or other pathogenic sequelae. This finding contrasts with the roughly four-fold increased lymphocyte turnover in macaques infected with simian immunodeficiency virus.1, 2 Human beings infected with HIV-1 exhibit similarly increased rates of CD4 and CD8 lymphocyte consumption and renewal.3 In quantitative studies, at any given moment, the majority of activated lymphocytes of HIV-1-positive individuals are not infected,3, 4, 5, 6, 7, 8 which suggests a pathogenic role for bystander cell turnover

Although HIV-1-specific proliferating lymphocytes are detected at sites of infection,8, 9 HIV-specific CD8 lymphocytes may be partly excluded from nodes in which virus is replicating,10 and lymphocytes not specific for HIV-1 are also activated by HIV-1 gene products and induced cytokines. Indeed, HIV-1 is capable of attracting and activating lymphocytes non-specifically to infected nodes.11, 12 The chronic availability of susceptible activated targets might be one factor that leads to steady-state virus production in a given host.

In various reports, even under conditions of viral suppression (<50 copies viral RNA/mL) induced by potent antiretroviral therapy for 1 year, CD4 and CD8 lymphocyte turnover have not always completely returned to baseline rates, remaining raised at up to twice normal loads.13 Rare individuals show discordance for viral compared with CD4 lymphocyte responses after treatment interruption.14, 15 Hazenberg and colleagues14, 16 have argued that low immune activation in response to viral rebound preserves normal tissue distribution and stable T-cell concentrations. During discordant responses to antiretroviral treatment, the degree of immune activation may be more predictive than viral load for disease progression.

In positron emission tomography (PET) studies, Scharko and colleagues17 and Wallace and colleagues18 have shown a sequential pattern of lymph-node activation in rhesus macaques infected with simian immunodeficiency virus, starting in the upper torso (mainly axillary nodes) and progressing caudally to involve the lower torso and gut. We used a similar PET method to anatomically locate and measure lymphocyte activation in HIV-1-infected human beings, allowing us to assess this component of pathogenesis in vivo.

Reactive lymph nodes can be quantified by PET because resting lymphocytes switch to glycolysis on activation, and increase their glucose uptake by around 20-fold over 24 h, corresponding to increased metabolism

and progression from G0 to G1 and G2 through mitosis.19, 20, 21 This change can be visualised with use of fluorine-18-labelled fluorodeoxyglucose (FDG), a glucose analogue that accumulates intracellularly in regions of high glucose use22, 23, 24, 25 Lymph-node activation in healthy individuals without intercurrent infections is generally insufficient to generate an FDG signal above background, except in the hilar region, where up to 28% of normal controls have some evidence of mildly increased FDG uptake.26 Regions of the intestine show some increase over background in 11% of healthy adults, positively associated with female sex, older age, and constipation.27 Although PET has been used extensively to identify sites of tumour growth in lymph nodes, only one study has been done on an otherwise healthy patient during acute non-HIV-1 viral infection.28 A boy aged 13 years, who had mononucleosis associated with Epstein-Barr virus, had a strong FDG signal from the reactive mediastinal and cervical lymph nodes, as well as spleen.

We undertook this study to investigate the sensitivity of FDG-PET in detecting activated lymph nodes during early and asymptomatic HIV-1 infection. Because activation and proliferation of lymphocytes are inextricably linked to subsequent programmed cell death in most progeny cells,29 PET signal is, indirectly, a correlate of cell death as well as activation.

Section snippets

Participants

Before studying HIV-1 disease, we studied several healthy, HIV-1-uninfected individuals aged 18–50 years for responses to licensed killed influenza vaccine (FluShield, Wyeth, Marietta, PA, USA), to assess the PET technique in antigenically and temporarily controlled immune stimulation.

We asked individuals recently infected with HIV-1 aged 18–50 years to participate as they enrolled in the Johns Hopkins Acute Infection and Early Disease Research Program, and asymptomatic HIV-1-positive patients

Results

Eight healthy controls, 12 recently infected HIV-1 patients, and 11 chronically infected HIV-1 patients were enrolled. Of the 12 recently infected HIV-1 patients, seven were men and five were women. Risk factors for infection had been sex with other men (four), heterosexual sex with infected partners (three), and injecting drug use (five). These individuals were studied within 2–15 months of seroconversion and none had received previous antiretroviral treatment. Of the 11 long-term asymptomatic

Discussion

Lymph-node metabolism in the absence of intercurrent disease is not sufficient to produce FDG signal. The PET signals detected in normal recipients of influenza vaccine 3–5 days after intramuscular injection (including non-specifically activated bystander cells) were temporally and anatomically related to vaccination, since signal was not detected from contralateral nodes. Similarly, HIV-1-induced immune activation in asymptomatic hosts is attributable to direct and indirect viral effects and

References (32)

  • KovacsJA et al.

    Identification of dynamically distinct subpopulations of T lymphocytes that are differentially affected by HIV

    J Exp Med

    (2001)
  • SchackerT et al.

    Productive infection of T cells in lymphoid tissues during primary and early human immunodeficiency virus infection

    J Infect Dis

    (2001)
  • HosmalinnA et al.

    Dumaurier MJ, et al. HIV-specific effector cytotoxic T lymphocytes and HIV-producing cells colocalize in white pulps and germinal centers from infected patients

    Blood

    (2001)
  • BrodieSJ et al.

    HIV-specific cytotoxic T lymphocytes traffic to lymph nodes and localize at sites of HIV replication and cell death

    J Clin Invest

    (2000)
  • IyengarS et al.

    T cell tropic HIV gp120 mediates CD4 and CD8 cell chemotaxis through CXCR4 independent of CD4: implications for HIV pathogenesis

    J Immunol

    (1999)
  • ChenJJ et al.

    CD4 lymphocytes in the blood of HIV(+) individuals migrate rapidly to lymph nodes and bone marrow: support for homing theory of CD4 cell depletion

    J Leukoc Biol

    (2002)
  • Cited by (0)

    View full text