Among HIV-infected individuals with detectable plasma VL, we found higher CSF VL in those with serologic evidence of neurosyphilis (NS+) than in subjects with either systemic syphilis alone or no syphilis coinfection. The pattern of results was the same after adjusting for important covariates such as plasma VL and CSF pleocytosis. Previous studies have demonstrated that vaccination and acute systemic infections including syphilis are associated with increases in plasma VL. Acute CNS opportunistic infections also increase CSF VL, but this study is the first to show that chronic neurosyphilis is associated with elevated CSF VL.
We cannot attribute the increase in CSF VL to active neurosyphilis for several reasons. First, most NS+ subjects reported a prior diagnosis of and appropriate treatment for syphilis. Second, these were ambulatory volunteers undergoing routine lumbar punctures for research purposes who reported no neurological symptoms suggestive of active neurosyphilis. Finally, indicators of disease activity such as pleocytosis and elevated protein levels were no more common in the NS+ group than in those without syphilis. Thus, the increases in CSF VL in subjects with positive CSF VDRL occurred in the absence of clinical or inflammatory indicators of active neurosyphilis.
As CSF viral loads did not differ by regimen type, regimen type could not account for the group differences in CSF viral load. We also considered the possibility that differential CNS penetration of the antiretroviral regimens prescribed to subjects in the three groups might explain higher CSF viral load in the neurosyphilis group. For example, if subjects in the NS+ group had more poorly penetrating regimens than those in the other groups, this could be reflected in higher CSF viral loads. However, CNS penetration was not significantly associated with CSF VL and subjects in the NS+ group did not have significantly lower penetration scores than those in the Syph+ or Syph− group.
Higher CSF HIV VL levels may be a consequence of intrathecal immune activation and subsequent amplification of HIV replication in CSF. Increased cellular activation by coinfection may enhance the surface expression of HIV receptors and coreceptors (CD4, CCR5, CXCR4), facilitating HIV cell entry. Furthermore, transcription of viral genes—even in the absence of replication—may be up-regulated by cytokines expressed in response to coinfection.
These immune effects may persist substantially beyond the period of active coinfection. The adverse effect of CNS neurological coinfections such as neurosyphilis on clinical outcomes of HIV CNS infection may be explained in part by persisting intrathecal immune disturbances that enhance CNS HIV replication. Such increased CNS replication might promote genotypic diversity (Buchacz et al. 2004
), increasing the likelihood that transient CNS infection will evolve into autonomous infection (Staprans et al. 1999
), with subsequent increased risk of developing an HIV-related neurocognitive disorder including dementia.
coinfection may be particularly likely to interact with HIV because the two pathogens share the same antigen-presenting cells. Thus, CNS tissue macrophages and microglia may present treponemal antigens to lymphocytes. These cells constitutively express major histocompatibility complex (MHC) class I, and after stimulation by coinfection or cytokine activation (e.g., interferon [IFN]-γ) may also express MHC class II (Kreutzberg, 1996
; Sedgwick, 1995
; Sedgwick et al. 1998
). Additionally, endothelial cells of the blood-brain barrier (BBB) constitutively express MHC classes I and II, thus being able to act also as antigen-presenting cells (Sedgwick, 1995
Coinfections such as syphilis and other chronic bacterial, parasitic, and helminthic infections tend to be more prevalent in developing countries. The majority of HIV-1–infected individuals live in these developing countries, where viral replication and disease may be accelerated by chronic immune activation caused by coinfecting pathogens (Lawn, 2004
; Pennycook et al. 2000
). Efforts should be geared to target treatments for both HIV itself and coinfections in subpopulations of HIV-1–infected individuals with coinfections.
Similar potentiation of HIV replication in the CNS by neurosyphilis could predispose to a higher frequency of neurocognitive impairment (Brew et al. 1997
; Ellis et al. 1997
; McArthur et al. 1997
). Although neuro-cognitive evaluation was not the objective of this study, a previous study by Wallace et al
found that HIV-infected subjects with a history of either syphilis or gonorrhea had poorer neurocognitive performance testing than those with no such coinfection (Wallace et al. 1997
). This difference was not explained by education, age, race, or CD4 count and was not observed in the HIV-uninfected control subjects. The study by Wallace et al
did not address CSF VL. Our finding that CSF VL is higher in systemic syphilis than in controls is consistent with the findings of Wallace et al
and suggests that systemic inflammation acting on the CNS could be the mechanism. Future studies should evaluate whether syphilis coinfection is associated with higher rates of neurocognitive impairment.