Is Microbial Translocation a Cause or Consequence of HIV Disease Progression?

To the Editor—Two recent articles in the Journal of Infectious Diseases described associations between increased levels of microbial translocation as measured by plasma levels of lipopolysaccharide (LPS) and late-stage human immunodeficiency virus (HIV) disease in South Africa [1] and Guinea-Bissau [2]. These cross-sectional studies are seemingly in disagreement with a longitudinal analysis using data from Rakai, Uganda, which found no association between microbial translocation and subsequent HIV-1 disease progression [3, 4]. This apparent discrepancy in the role of microbial translocation in HIV disease has also been observed in other cross-sectional [5–8] and longitudinal human studies [9, 10]. Even in animal models, it was recently shown that LPS levels are not predictive of simian immunodeficiency virus (SIV) disease progression in rhesus macaques [11], although elevated LPS levels prior to SIV infection appear to contribute to the faster disease progression in pigtailed macaques [12]. These discrepancies are primarily due to the fact that, as the 2 groups state, one cannot infer causality from cross-sectional studies; therefore, it is impossible to determine whether the elevation of LPS levels in AIDS is a cause or a consequence of disease progression.

In the Nowroozalizadeh et al [2] report, investigators found differences in microbial translocation only when comparing HIV-negative individuals with patients with AIDS, but not in chronically infected individuals, whether infected with HIV-1 or HIV-2. Additionally, the investigators did not observe a difference in LPS levels between HIV-1 and HIV-2 at either stage of disease. This is somewhat surprising since HIV-1 and HIV-2 differ dramatically in their rates of disease progression.

It was proposed that microbial translocation affects HIV disease progression in part by LPS activation of monocytes, leading to increased production of soluble CD14 (sCD14) [6, 7]. Cassol et al [1] found a significant positive correlation between LPS and sCD14 levels in patients with AIDS who did not have opportunistic infections, but they found a negative correlation between LPS and sCD14 levels in individuals with an opportunistic infection, even though the levels of sCD14 were similar between the 2 groups. It was speculated by the investigators that “[t]his observation suggests that, in untreated patients, the increase in sCD14 was driven predominantly by microbial translocation” [1]. However, a recent study by Rempel et al [13] demonstrated that monocytes from HIV-1-infected individuals with high viral loads are not activated by LPS. Moreover, in our longitudinal study we found no association between disease progression rates and sCD14 levels [3]. This suggests that the increased levels of sCD14 seen in late-stage HIV disease are not directly caused by microbial translocation.

The cross-sectional data from Africa [1, 2, 8], as well as North America and Europe [5–7], demonstrate that there is most likely an association between low levels of microbial translocation and advanced HIV disease, and this has led to the inference that “microbial translocation contributes to immune activation and disease progression during chronic HIV-1 infection” [1]. However, this proposed causal relationship between microbial translocation and HIV-1 disease progression has never been directly demonstrated. In addition, in longitudinal studies, which can address causality, LPS levels did not significantly contribute to HIV-1 disease progression even in the presence of immune activation [3, 4, 10]. Therefore, a direct causal relationship between microbial translocation and HIV disease progression is unlikely.

Alternatively, we propose that increased microbial translocation and LPS levels are a consequence of advanced HIV-1 disease and AIDS. This model is supported by both the cross-sectional and longitudinal data, and may explain why there were no differences in LPS levels between HIV-1 and HIV-2, despite well-established differences in rates of disease progression [2]. Moreover, a consequential relationship is also supported by the finding that LPS levels in HIV-infected individuals successfully treated with antiretrovirals for >5 years are similar to those found in HIV-uninfected individuals [14].

Determining the role of microbial translocation in HIV disease progression is important as it may have a significant impact on future clinical care in a variety of settings. To resolve this issue, we propose these additional studies that should help clarify this matter: (1) examine more longitudinal cohorts for the relationship between disease progression, microbial translocation, antibacterial immune responses, and immune activation; (2) design and test better assays to quantify bacterial products in the bloodstream, as highlighted by Ferri et al [15]; (3) determine the biological relevance of low-level microbial translocation in the absence of HIV; (4) explore the possible clinical benefit of treating low-level microbial translocation directly [7]; and (5) examine if earlier initiation of antiretroviral treatment can minimize this effect and by doing so improve clinical outcomes.

Funding

This work was supported by the Division of Intramural Research, National Institute of Allergies and Infectious Diseases, National Institute of Health (A.D.R. and T.C.Q.).