Herpes zoster in HIV-1 infection: the role of CSF pleocytosis in secondary CSF escape and discordance

Abstract

HIV cerebrospinal fluid (CSF) escape is defined by a concentration of HIV-1 RNA in CSF above the lower limit of quantification of the employed assay and equal to or greater than the plasma HIV-1 RNA level in the presence of treatment-related plasma viral suppression, while CSF discordance is similarly defined by equal or higher CSF than plasma HIV-1 RNA in untreated individuals. During secondary CSF escape or discordance disproportionate CSF HIV-1 RNA develops in relation to another infection in addition to HIV-1. We performed a retrospective review of people living with HIV receiving clinical care at Sahlgrenska Infectious Diseases Clinic in Gothenburg, Sweden who developed uncomplicated herpes zoster (HZ) and underwent a research lumbar puncture (LP) within the ensuing 150 days. Based on treatment status and the relationship between CSF and plasma HIV-1 RNA concentrations, they were divided into 4 groups: i) antiretroviral treated with CSF escape (N=4), ii) treated without CSF escape (N=5), iii) untreated with CSF discordance (N=8), and iv) untreated without CSF discordance (N=8). We augmented these with two additional cases of secondary CSF escape related to neuroborreliosis and HSV-2 encephalitis and analyzed these two non-HZ cases for factors contributing to CSF HIV-1 RNA concentrations. HIV-1 CSF escape and discordance were associated with higher CSF white blood cell (WBC) counts than their non-escape (P=0.0087) and non-discordant (P=0.0017) counterparts, and the CSF WBC counts correlated with the CSF HIV-1 RNA levels in both the treated (P=0.0047) and untreated (P=0.002) group pairs. Moreover, the CSF WBC counts correlated with the CSF:plasma HIV-1 RNA ratios of the entire group of 27 subjects (P=<0.0001) indicating a strong effect of the CSF WBC count on the relation of the CSF to plasma HIV-1 RNA concentrations across the entire sample set. The inflammatory response to HZ and its augmenting effect on CSF HIV-1 RNA was found up to 5 months after the HZ outbreak in the cross-sectional sample and, was present for one year after HZ in one individual followed longitudinally. We find that HZ provides a model of secondary CSF escape and discordance. Likely, the inflammatory response to HZ pathology within the CNS provoked local HIV-1 production by enhanced trafficking or activation of HIV-1-infected CD4+ T lymphocytes. Whereas treatment and other systemic factors determined the plasma HIV-1 RNA concentrations, in this setting the CSF WBC counts established the relation of the CSF HIV-1 RNA levels to this plasma set-point. Author summaryHerpes zoster is a neurotropic infection common in people living with HIV. We studied if herpes zoster caused alterations in the cerebrospinal fluid viral (CSF) load when compared with blood levels in patients with and without antiretroviral treatment by studies of the cerebrospinal fluid. HIV-1 CSF escape and discordance were associated with higher CSF white blood cell counts than their non-escape (P=0.0087) and non-discordant (P=0.0017) counterparts, and the CSF white blood cell counts correlated with the CSF HIV-1 RNA levels in both the treated (P=0.0047) and untreated (P=0.002) group pairs. We found that the inflammatory response to herpes zoster pathology within the CSF provoked local HIV-1 production by enhanced trafficking or activation of HIV-1-infected CD4+ T lymphocytes. We suggest that herpes zoster provides a model of secondary HIV CSF escape and HIV CSF/blood discordance, which increase the understanding of HIV central nervous system infection and might be of importance for eradication trials in sanctuary sites such as the central nervous system.