Brain (18)F-FDG PET of SIV-infected macaques after treatment interruption or initiation

BACKGROUND: Although rates of severe HIV-associated neurocognitive disorders have declined in the post-antiretroviral treatment (ART) era, subtle deficits persist, possibly exacerbated by treatment non-adherence. The actual effects of ART interruption/initiation on brain glucose metabolism as a reflection of viral replication and neuroinflammation remain unclear. Our study investigates how treatment initiation and interruption alter brain glucose metabolism in SIV-infected macaques, using (18)F-FDG PET in correlation with plasma and CSF viral loads (VL) and cytokine levels. METHODS: SIV-infected macaques (n = 7) underwent ART initiation only, ART interruption only, or both. Five uninfected animals served as controls. (18)F-FDG PET imaging was performed at baseline and 1, 3, and 6 months after treatment modification. Mean and maximum standardized uptake values (SUV) for the whole-brain and subregions were calculated. Plasma and CSF VL and cytokine levels were measured. Paired t tests evaluated acute changes in whole-brain SUV from baseline to 1 month, while mixed-effect linear regression models evaluated changes over multiple timepoints and correlated SUV values with disease markers. RESULTS: ART interruption was associated with increased SUVmean and SUVmax acutely, after 1 month (SUVmean 95% CI [0.044–0.786 g/ml], p = 0.037; SUVmax 95% CI [0.122–3.167 g/ml], p = 0.041). The correlation between SUV and time, however, was not significant when evaluated across all timepoints. Increased SUVmean and SUVmax correlated with decreased CD4+ and CD8+ T-cell counts and increased plasma VL. SUVmax was positively associated with increases in CSF VL, and there were borderline positive associations between SUVmax and IL-2, and between SUVmean and IL-15. The treatment initiation group showed no associations between imaging and disease biomarkers despite viral suppression, reduced cytokine levels, and increased CD4+ and CD8+ T-cell counts. CONCLUSIONS: ART interruption is associated with increased brain glucose metabolism within 1 month of treatment cessation, which, in concert with increased levels of pro-inflammatory cytokines in the CSF, may reflect neuroinflammation in the setting of viral rebound. Although we cannot assert neurologic damage in association with cerebral hypermetabolism, it is a concerning outcome of ART non-adherence. Treatment initiation, meanwhile, did not result in significant changes in brain metabolism. HIV-induced neuroinflammation may require a longer period to abate than our follow-up period allowed. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12974-018-1244-z) contains supplementary material, which is available to authorized users.