Association between herpes simplex virus infection and Alzheimer's disease biomarkers: analysis within the MAPT trial

Abstract

In vitro and animal studies have suggested that inoculation with herpes simplex virus 1 (HSV-1) can lead to amyloid deposits, hyperphosphorylation of tau, and/or neuronal loss. Here, we studied the association between HSV-1 and Alzheimer's disease biomarkers in humans. Our sample included 182 participants at risk of cognitive decline from the Multidomain Alzheimer Preventive Trial who had HSV-1 plasma serology and an amyloid PET scan. Plasma Aβ42/40 ratio, neurofilament light chain and p-tau181 were also available for a sub-sample of participants. Multivariate linear regressions were performed and stratified by APOE4 genotype. The median age was 74.0 years, 85.2% were infected with HSV-1. Infected participants tended to have a lower cortical amyloid load than uninfected participants (β = -0.08, p = 0.06), especially those suspected of reactivating HSV-1 most frequently (i.e. with a high anti-HSV-1 IgG level; n = 58, β = -0.09 p = 0.04). After stratification, the association was only significant in APOE4 carriers (n = 43, β = -0.21 p = 0.01). No association was found with the plasma biomarkers. The trend toward lower cortical amyloid load in HSV-1-infected participants was unexpected given the pre-existing literature and may be explained either by a modified immune response in HSV-1 infected subjects which could favour the clearance of amyloid deposits or by a selection bias.

Fig. 1

Study design and flow chart. The main sample of our study includes 182 subjects with both HSV-1 serology at baseline and amyloid PET scans that may have been performed throughout follow-up. Two subsamples were used to assess associations between HSV-1 infection and plasma biomarkers: i) a subsample of 164 subjects with dosages of Aβ42/40 ratio and NfL at 12 months and ii) a subsample of 138 subjects with two dosages of p-tau 181 at baseline and 36 months. HSV, herpes simplex virus; NfL, neurofilament light chain; PET, positron emission tomography; p-tau, phosphorylated-tau.

Fig. 2

Association between HSV-1 serostatus and Alzheimer’s disease biomarkers: Multivariate regression models. MAPT trial. *To examine associations between HSV-1 serostatus and AD biomarkers, multivariate linear regression models (or logistic regression models for the binary PET amyloid positivity variable) were performed on either i) the presence of anti-HSV-1 IgG (IgG +) or ii) the level of anti-HSV-1 IgG considered in terciles (T1, T2 and T3 corresponding to 1^st, 2^nd and 3^rd terciles, respectively). In both cases, the reference category was the absence of anti-HSV-1 IgG (IgG-). Analyses were adjusted for baseline age, sex, presence of at least one APOE4 allele and level of education. The results are presented in the form of β (and their 95% confidence intervals in the figure), standard deviations and p values for the linear regressions or adjusted odds ratios and their 95% confidence intervals, as well as p values for logistic regressions. ^§in pg/ml, log transformed. Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; HSV-1, herpes simplex virus 1; IgG, immunoglobulin G; NfL, neurofilament light chain; p-tau181, phosphorylated tau 181; std, standard deviation; SUVR, standard uptake value ratio; T1, T2 and T3, 1^st, 2^nd and 3^rd terciles, respectively.