Cerebrospinal fluid neopterin: an informative biomarker of central nervous system immune activation in HIV-1 infection

Abstract

HIV-1 invades the central nervous system (CNS) in the context of acute infection, persists thereafter in the absence of treatment, and leads to chronic intrathecal immunoactivation that can be measured by the macrophage activation marker, neopterin, in cerebrospinal fluid (CSF). In this review we describe our experience with CSF neopterin measurements in 382 untreated HIV-infected patients across the spectrum of immunosuppression and HIV-related neurological diseases, in 73 untreated AIDS patients with opportunistic CNS infections, and in 233 treated patients.In untreated patients, CSF neopterin concentrations are almost always elevated and increase progressively as immunosuppression worsens and blood CD4 cell counts fall. However, patients with HIV dementia exhibit particularly high CSF neopterin concentrations, above those of patients without neurological disease, though patients with CNS opportunistic infections, including CMV encephalitis and cryptococcal meningitis, also exhibit high levels of CSF neopterin. Combination antiretroviral therapy, with its potent effect on CNS HIV infection and CSF HIV RNA, mitigates both intrathecal immunoactivation and lowers CSF neopterin. However, despite suppression of plasma and CSF HIV RNA to below the detection limits of clinical assays (<50 copies HIV RNA/mL), CSF neopterin often remains mildly elevated, indicating persistent low-level intrathecal immune activation and raising the important questions of whether this elevation is driven by continued CNS infection and whether it causes continued indolent CNS injury.Although nonspecific, CSF neopterin can serve as a useful biomarker in the diagnosis of HIV dementia in the setting of confounding conditions, in monitoring the CNS inflammatory effects of antiretroviral treatment, and give valuable information to the cause of ongoing brain injury.

Figure 1

Induction of neopterin formation in brain cells. Pro-inflammatory cytokines like interferon-γ (IFN-γ) induce expression of GTP-cyclohydrolase I in various brain cells. As an intermediate product 7,8-dihydroneopterin-triphosphate is produced which is further converted by pyruvoyl-tetrahydropterin synthase (PTPS) to form 5,6,7,8-tetrahydrobiopterin (BH4), the cofactor of several aromatic amino acid monooxygenases that are involved in the production of tyrosine, L-DOPA, serotonin and nitric oxide. Different from neurons, monocytic cells possess only low constitutive activity of PTPS. Thus, 7,8-dihydroneopterin-triphosphate does not undergo conversion to BH4, rather it is dephosphorylated and oxidized to neopterin in non-enzymatic reactions.

Figure 2

Cross-sectional analysis of CSF neopterin in HIV disease in the context of other CSF and blood measurements. Included are 53 HIV-seronegative volunteers; untreated HIV positive neurologically asymptomatic (NA) subjects; 53 with CD4+ counts <50 cells/μL (mean age 38.9); 69 subjects with 50-199 cells/μL (mean age 38.6); 69 with counts 200-349 cells/μL (mean age 38.8); and 108 with CD4+ counts >350 cells/μL. Untreated patients with ADC were divided into 30 with Stage 1, and 53 with Stage 2-4. Treated subjects included 150 with plasma HIV RNA suppressed below 50 copies/mL (treatment successes) and 83 with >50 copies/mL (treatment failures) after >6 months of treatment. The OI group included 73 patients with CNS opportunistic diseases (see text). The boxes show the 25-50^th quartile with median bar and mean +, while the whiskers show the 10-90^th quartile. A. CSF neopterin. Overall ANOVA P < 0.0001, Dunn's post hoc comparisons showed that HIV- group differed from all HIV+ groups (P < 0.001 except Sucesses P < 0.5); ADC 2-4 differed from all NAs (P < 0.01- 0.001) but not from ADC 1 group; the ADC 1 group differed from the NA CD4 >350 (P < 0.05) and 200 - 349 (P < 0.001) but not from other NA groups. The treated successes differed both from all the untreated HIV-infected groups (P < 0.001) and the HIV negatives (P < 0.05), while the treated failures also differed from the untreated HIV-infected (P < 0.05- 0.001), except those with CD4 >350, and from the HIV- (P < 0.001). The OI group differed from the NAs with CD4>200 and treated groups but not from those with lower counts or from ADC groups. B. Plasma neopterin. Statistical analysis was similar to CSF except that ADC 2-4 differed only from the two higher CD4 NAs (P < 0.01- 0.001) and the ADC 1 only from the CD4 >350, and the treatment successes did not differ from the HIV seronegatives while the failures did (P < 0.001). C. CSF HIV RNA. D. Plasma HIV RNA. E CSF WBC counts. F. Blood CD4+ T cell counts. Abbreviations: HIV-, HIV seronegative control group; NA, neurologically asymptomatic; ADC, AIDS dementia complex; Rx Success, treated with plasma suppression to <50 copies HIV RNA per mL; Rx Failure, treated with continued plasma viremia with ≥ 50 copies HIV RNA per mL.

Figure 3

CSF neopterin concentrations in the 73 subjects with CNS opportunistic diseases (OIs) included 16 with progressive multifocal leukoencephalopathy (PML), 13 with cytomegalovirus encephalitis (CMV-E), 18 with toxoplasmic encephalitis (toxo), 16 with cryptococcal meningitis (crypto) and 10 with primary CNS lymphoma (PCNSL), and for comparison, neuroasymptomatic HIV positive (NA) subjects with <200 blood CD4 counts (collapsed from two groups in Figure 2) and ADC 1-4 (also collapsed from two groups in Figure 2). The box and whiskers and statistical methods are as described for Figure 2. The CSF neopterin in the PML group differed from the CMV-E (P < 0.001) and ADC 1-4 group (P < 0.01) but not from the other OI groups or from the NA group. The CMV encephalitis patients had the highest levels and in addition to differing from the PML group, differed from the toxoplasmosis patients (P < 0.01) and neuroasymptomatics (P < 0.001), but not the ADC group. The toxoplasmosis group, in addition to differing from the CMV group also differed from the ADC group (P < 0.05) but not the other OIs or NAs. The cryptococcal meningitis group differed from the NA with low CD4 T cells/μl (P < 0.05) while the PCNSL group did not differ from the other groups.

Figure 4

CSF neopterin in A. Successes and B. Failures. Relation to the revised 2010 CPE rank scores [40]. There were no significant differences in CSF among these groups, nor was there a correlation when all ranks were considered as a continuous variable. Symbols show the medians and lines the intraquartile range for each group. Additionally, no correlation was found using the older CPE score system (not shown) [41].

Figure 5

Four subjects studied longitudinally. For each of the four subjects (A-D) the top panel shows the HIV RNA concentrations and treatment intervals and the bottom panel the CSF and blood neopterin levels. The symbol definitions in the two A panels apply to all four subjects.

Figure 6

ROC curves for two comparisons. A. ADC 2-4 was compared to the four groups of untreated neuroasymptomatics. B ADC 1-4 (ADC 1 group and 2-4 group combined) was compared to the four groups of NAs. AUC, area under the ROC concentration curve where 1 is high and 0.5 not different from random.