Statins Differentially Affect Amyloid Precursor Protein Metabolism in Presymptomatic PS1 and Non-PS1 Subjects FREE

The putative potential of statins to retard the onset and progression of Alzheimer disease (AD) remains controversial.^1- 2 Statin therapy may have the potential to increase nonamyloidogenic soluble amyloid precursor protein α (sAPPα), thereby reducing β-amyloid 42 (Aβ₄₂) and the downstream markers of neurodegeneration phospho-tau (p-tau) and total tau in the cerebrospinal fluid, and thus potentially slow the onset and progression of AD.^3- 5 Thus, in a small but unique cohort of cognitively normal subjects with presenilin 1 (PS1) mutations, we have taken the opportunity to conduct an hypothesis-generating pilot study to assess the effects of intensive statin therapy, using both a lipophilic (simvastatin) and a hydrophilic (atorvastatin) statin. We also studied a second group of subjects without such mutations but with hyperlipidemia (low-density lipoproteins [LDL]>100 mg/dL) and/or who are heterozygous for apolipoprotein ε4 (APOEε4).

Presymptomatic subjects with known PS1 mutations from 2 different kinships (Table) participated in these studies carried out under approval from an institutional review board. After they gave informed consent, subjects meeting the inclusion criteria (PS1 mutation, APOEε4, or APOEε3 with hyperlipidemia [LDL > 100 mg/dL]) were included. We performed sequential lumbar punctures as well as lipid profiles at baseline and at 6 and 12 months. Cerebrospinal fluid was immediately frozen at −80°C and so maintained prior to testing. Lumbar punctures at each visit were performed for each subject at a standard time of day to eliminate diurnal variability in any analyte. Athena Diagnostics (Worcester, Massachusetts) carried out the Innogenetics methods (Innogenetics, Ghent, Belgium) for determining levels of phospho-tau (hyperphorylated tau, eg, phosphorserine 181), total tau, and Aβ₄₂. Concentrations of sAPPα, and sAPPβ were measured by enzyme-linked immunosorbent assay.⁶ All samples were analyzed together to avoid interrun variability. The study design allowed all subjects to act as their own controls. We used an intent-to-treat analysis assessing changes in analytes vs changes in serum lipid levels. Treatment outcomes were evaluated using general linear mixed models to estimate whether there had been a significant change in slope for values of analytes from prestatin levels to those at 6 and 12 months.

All subjects showed profound decreases in both total cholesterol (P < .001) and LDL (P < .001) levels at 6 and 12 months after initiating statin therapy (Table). In PS1 subjects, statin therapy induced significant decreases of both sAPPα and sAPPβ. Combining results for both statins, sAPPα dropped by 16.5% (P = .001) and sAPPβ dropped by 21.2% (P < .001). When the effects were categorized according to drug type, simvastatin decreased sAPPα by −26.5% (P < .001) and sAPPβ by −31.5% (P < .001). There were no significant changes in the PS1 group treated with atorvastatin for either sAPPα or sAPPβ. Ratios of sAPPα to sAPPβ remained unchanged after treatment. Statin therapy reduced p-tau values in 5 of 6 PS1 subjects by −8.3% and approached significance (P = .08). There were no significant changes in total tau or Aβ₄₂.

In the non-PS1 subjects and when controlling for reductions in total cholesterol, sAPPα was increased by 9.6% (P = .008) by atorvastatin and 23.7% (P < .001) by simvastatin. Increases in sAPPβ were not significant, but as with the PS1 group, sAPPα to sAPPβ ratios did not change after statin therapy. There were no significant changes in Aβ₄₂, p-tau, and total tau.

These results are not easily interpretable based on the assumption that an increase in sAPPα occurs with a corresponding decrease in the available APP for cleavage by β-secretase. Instead, our results are more consistent with the interpretation that treatment increased either the production or the availability of the APP substrate in non-PS1 subjects and decreased either factor in subjects with PS1 mutations. It remains to be determined whether these differential effects are related to the lipid-lowering or pleiotropic effects of statins.

Correspondence: Dr Majaz Moonis, Department of Neurology, University of Massachusetts Medical School, 55 Lake Ave N, Worcester, MA 01655 (monism@ummhc.org).

Author Contributions:Study concept and design: Hinerfeld, Moonis, St. George-Hyslop, and Pollen. Acquisition of data: Hinerfeld, Moonis, Swearer, Caselli, Rogaeva, and Pollen. Analysis and interpretation of data: Hinerfeld, Moonis, Baker, St. George-Hyslop, and Pollen. Drafting of the manuscript: Hinerfeld, Moonis, Swearer, Baker, St. George-Hyslop, and Pollen. Critical revision of the manuscript for important intellectual content: Hinerfeld, Moonis, Baker, Caselli, Rogaeva, St. George-Hyslop, and Pollen. Statistical analysis: Baker. Obtained funding: Moonis and Pollen. Administrative, technical, and material support: Hinerfeld, Moonis, Swearer, Caselli, and Pollen. Study supervision: Hinerfeld, Moonis, and Pollen.

Financial Disclosure: None reported.

Funding/Support: This study was supported by a gift from the Edith Marrone Memorial Foundation and from an anonymous donor.

Additional Information: Drs Pollen and Swearer are in the Department of Neurology, University of Massachusetts Medical School, Worcester. Mr Baker is in the Bioinformatics Unit, Information Service and Department of Cell Biology, University of Massachusetts Medical School. Dr Hinerfeld worked on the project while at the University of Massachusetts but has since moved to the Jackson Laboratory, Bar Harbor, Maine. Drs Hyslop and Rogaeva are in the Division of Neurology, Department of Medicine, University of Toronto and Toronto Western Hospital, Toronto, Ontario, Canada. Dr Caselli is in the Department of Neurology, Mayo Clinic, Scottsdale, Arizona.