A randomized, double-blind, placebo-controlled trial of resveratrol for Alzheimer disease

Classification of evidence.

This study provides Class II evidence that for patients with AD resveratrol is safe, well-tolerated, and alters some AD biomarker trajectories. The study is rated Class II because more than 2 primary outcomes were designated.

Study design.

A multicenter, double-blind, placebo-controlled trial was conducted June 2012–March 2014 with participants recruited from 26 US academic clinics affiliated with the Alzheimer's Disease Cooperative Study (ADCS). The enrollment target was 120 (60 per group) randomized to drug or placebo. Actual enrollment was 119. A subgroup of 15 participants enrolled in a randomized 4:1 (n = 15, 12 treated plus 3 placebo) study for 24-hour pharmacokinetics at selected sites. For these individuals, blood samples were collected at times 0, 0.17, 0.33, 0.5, 0.67, 1, 1.5, 2, 2.5, 3, 4, 6, 8, 12, and 24 hours. Measurements included resveratrol, 3-O-glucuronidated-resveratrol (3G-RES), 4-O-glucuronidated-resveratrol (4G-RES), and 3-sulfated-resveratrol (S-RES). These participants completed 24-hour pharmacokinetics at each dosage: after the first dose following baseline, after the first dose with each increment (weeks 13, 26, and 39), and after the final dose (week 52). The afternoon dose of resveratrol was withheld during the 24-hour blood sampling.

Standard protocol approvals, registrations, and patient consents.

This study was conducted in accordance with Good Clinical Practice guidelines. Informed consent was obtained from participants and study partners. The study was conducted under local institutional review board supervision, under Food and Drug Administration IND 104205, and registered at ClinicalTrials.gov (NCT01504854).

Study visits.

Visits occurred at screening, baseline, and weeks 6, 13, 19, 26, 32, 39, 45, and 52. Visits included concomitant medications and adverse events (AEs) review, physical and neurologic examination, urinalysis, pill count, and venipuncture for laboratory tests, pharmacokinetics, and biomarker analyses. Brain MRIs were obtained at baseline, week 13, and week 52. ECGs and CSF collections were performed at baseline and week 52. Oral glucose tolerance tests, with peripheral blood mononuclear cell collections at 0 and 120 minutes, were performed at screening and week 52 (except in participants enrolled in the 24-hour pharmacokinetics substudy).

Participants and randomization.

Inclusion criteria for enrollment included age >49 years, fluent in English or Spanish, diagnosis of probable AD by National Institute of Neurological and Communicative Disorders and Stroke–Alzheimer's Disease and Related Disorders Association criteria,¹⁶ Mini-Mental State Examination (MMSE) score¹⁷ 14–26 at screening, modified Hachinski Score¹⁸ <5, normal laboratory values, stable medications for 4 months, and stable use of cholinesterase inhibitors or memantine. Exclusion criteria included non-AD dementia, Down syndrome, sensory impairments precluding participation, pregnancy, contraindication to lumbar puncture or MRI, >4 microhemorrhages on a recent MRI, treated diabetes mellitus, use of resveratrol-containing supplements, and unsuitable disorder or laboratory finding. Participants were assigned to resveratrol or placebo using a stratified permuted block method with an allocation ratio of 1:1. Assignment to groups was stratified by site. After participants signed informed consent and eligibility was confirmed, study sites received randomization numbers from the Informatics Core. A subgroup (n = 15) was randomized 4:1 (12 treated, 3 placebo) for 24-hour pharmacokinetics. Sample sizes (60 per group) were determined from power analyses utilizing published data on CSF biomarkers in AD trials, and a predicted 20% dropout rate.

Study medication.

Aptuit Laurus, Inc. (Kansas City, MO, now Catalent, Inc., Somerset, NJ) synthesized and encapsulated resveratrol (trans-3,5,4′-trihydroxystilbene) and provided identical placebo, according to current Good Manufacturing Practices. The dose escalation was in 500-mg increments every 13 weeks as follows: 500 mg QAM, 500 mg BID, 1,000 mg QAM and 500 mg QPM, and 1,000 mg BID.

Outcomes.

Primary outcomes were levels of plasma Aβ40 and Aβ42, CSF Aβ40, Aβ42, tau, and phospho-tau 181, and volumetric MRI (rate of whole brain volume change, rate of ventricular volume change, rate of hippocampal volume change, and rate of entorhinal cortex volume change). Additional outcomes included safety and tolerability (AEs, physical examinations, neurologic examinations, clinical laboratory results) and pharmacokinetics. Secondary outcomes included scores on the MMSE,¹⁷ Alzheimer's Disease Assessment Scale–cognitive (ADAS-cog),¹⁹ ADCS Activities of Daily Living Scale (ADCS-ADL),²⁰ Clinical Dementia Rating–sum of boxes (CDR-SOB),²¹ and Neuropsychiatric Inventory (NPI),²² APOE genotype, and insulin and glucose metabolism (including oral glucose tolerance tests).

Safety assessments.

Participants received physical and neurologic examinations and vital signs at each visit. Site investigators classified AEs by severity and causality. If a participant withdrew, an early termination visit similar to a baseline visit was scheduled. An independent Data and Safety Monitoring Board reviewed data quarterly.

Liquid chromatography–mass spectrometry conditions for resveratrol plasma and CSF pharmacokinetics.

Stock solutions of resveratrol, hexestrol (Sigma Aldrich, St. Louis, MO), trans-resveratrol-3-O-β-D-glucuronide, trans-resveratrol-4-O-β-D-glucuronide, and trans-resveratrol-3-sulfate sodium salt (Toronto Research Chemicals Inc., Canada) were suspended to 2 mg/mL in dimethyl sulfoxide (DMSO). Working stocks of 0.1 mg/mL were made in DMSO. Calibrator standards and quality controls were made in human pooled plasma (Gemini Bioproducts, Sacramento, CA). Plasma samples were mixed with acetonitrile with 0.1% formic acid and centrifuged for 2 minutes to remove precipitate. Pharmacokinetic analysis was performed using an AB Sciex (Framingham, MA) QTRAP 5500 and Shimadzu UFLC XR rack changer liquid chromatography–mass spectrometry system. Data acquisition and analysis was made using Analyst 1.6 (AB Sciex). Quantification was done in Analyst (AB Sciex) by using linear regression with a 1/x² weighing factor based on goodness-of-fit criteria and coefficient of determination (r²). The CSF conditions were the same as plasma except the injection volume was 20 μL. Calibrator standards were made in 50:50 (v/v) acetonitrile and water with 0.1% formic acid.

Bioassays.

The Biomarker Core assayed Aβ40 and Aβ42 in plasma and CSF and tau and phospho-tau 181 in CSF.²³ Validated assay platforms from Meso Scale Discovery (Rockville, MD) were used to detect Aβ isoforms and total tau. Innotest pTau181 was used for phospho-tau 181. Internal standards were used to adjust for plate-to-plate variation and assess freezer storage effects.²³ The Biomarker Core performed APOE genotyping using real-time PCR restriction fragment length polymorphism analysis. Genomic DNA from blood was extracted using QIAamp DNA blood maxi kit (Qiagen, Venlo, Netherlands) and APOE genotyping performed using Applied Biosystems (Foster City, CA) TaqMan SNP Genotyping Assay. The assay was run on a Bio-Rad (Hercules, CA) CFX96.

MRI methods.

MRIs at baseline, 3 months, and 12 months were acquired using GE (Cleveland, OH), Philips (Best, the Netherlands), or Siemens (Munich, Germany) 1.5 and 3.0 T scanners. Site personnel scanned participants longitudinally on the same scanner using a consistent protocol and quality checks confirmed that parameters held constant. Volumetric MRI analysis was performed on 3D T1-weighted volumes acquired sagittally with imaging parameters modeled on the nonaccelerated T1-weighted sequence from the Alzheimer's Disease Neuroimaging Initiative. NeuroQuant (CorTechs Labs, San Diego, CA) image preprocessing and automated segmentation was used to measure whole-brain, hippocampus, and entorhinal volumes^24,–,26 and other methods were as described.^27,–,32 Regional deformation was quantified and averaged within all segmented areas; however, to reduce multiple comparisons for primary analysis, only ventricular volume change was assessed using longitudinal registration.

Statistical analyses.

Mixed-model repeated-measures (MMRM) analyses were used to assess between-group differences in change scores from baseline to week 52. The dependent variable in each MMRM analysis was change from baseline. Fixed effects included baseline scores on outcome measures, age at baseline, group assignment, study visit, and treatment-by-visit interaction. Additionally, covariates that were significantly associated with the response measure (p < 0.15) and were out of balance at baseline (p < 0.2) were included as fixed effects. Study visit was modeled as a categorical variable; an autoregressive (order 1) covariance structure was specified. Variables considered as potential covariates in each model included biomarkers CSF total tau, CSF phospho-tau 181, CSF Aβ40, CSF Aβ42, plasma Aβ40, plasma Aβ42, brain volume, ventricular volume, hippocampal volume, and entorhinal thickness; insulin and glucose; APOE; clinical measures: baseline ADAS-cog, baseline MMSE, and baseline CDR-SOB.

The key efficacy analysis was based on a modified intention-to-treat (ITT) population, which included all randomly assigned participants with at least one postbaseline observation. Secondary analyses were performed for completers, for APOE4 carriers and noncarriers, and for mild AD (MMSE > 19). The baseline characteristics of the study groups were compared with the use of Fisher exact test for categorical variables and a Wilcoxon signed-rank test for continuous variables. Safety analyses were based on summary listings of AEs, with Fisher exact test used for pairwise comparisons. Safety analyses were based on the full ITT population. R version 3.1 (www.r-project.org) was used for all analyses. A p value <0.05 indicated significance. All testing was 2-sided. Blinding of investigators was maintained until outcomes were determined. Results for testing several endpoints were not adjusted.