The best magazine
Methods
This Phase 2, multicenter, randomized, double-blind, placebo-controlled, parallel, three-arm, multiple dose study was conducted at 40 study sites in the United States between January 2007 and December 2010 (ClinicalTrials.gov identifier NCT00566397.) The study was approved by each study site's Local Institutional Review Board (see Additional file 1). Each patient provided written informed consent. If patients had impaired decisional capacity, caregivers provided consent and patients provided assent.
Patients
Eligible patients were aged ≥ 50; met the criteria for a diagnosis of probable AD; had a MMSE score between 14 and 26, had a modified Hachinski (Rosen) score ≤4, were receiving treatment with a stable dose of an acetylcholinesterase inhibitor and/or memantine for ≥ 4 months prior to randomization. Patients were excluded for clinically significant neurologic, psychiatric or other diseases contributing to his/her dementia, MRI and/or CT evidence of stroke or significant cerebrovascular disease, uncontrolled hypertension, unstable cardiac or pulmonary disease, diabetes (or hemoglobin A1c at screening > 6%).
Study Design and Treatment
Enrollment targeted 399 patients (133 per group), randomized (1:1:1) to placebo, or TTP488 20 mg daily (after a loading dose of 60 mg daily for 6 days), or 5 mg daily (after a loading dose of 15 mg daily for 6 days), for 18 months. An independent Data and Safety Monitoring Board (DSMB) monitored the safety of subjects in the trial.
Study visits occurred at screening, baseline, then at four weeks, 3, 6, 9, 12, 15, 18 months, with a safety follow-up visit at 21 months. Visits included clinical and safety evaluations, blood draw for plasma biomarker and pharmacokinetic analysis, and pill counts to assess compliance. Brain MRIs were obtained at baseline, 12 and 18 months. Lumbar punctures for CSF biomarkers were performed at baseline and 12 months on a subgroup of subjects.
Outcome Measure
The primary efficacy measure was the ADAS-cog. The ADAS-cog/12-item (scored 0–80) scale was administered before the first dose, and at 3, 6, 9, 12, 15 and 18 months with the pre-specified analyses being on the ADAS-cog/11-item scale (Scored 0–70). The key secondary clinical measure was the CDR-sb. The ADCS-ADL was included as a secondary measure. Both CDR-sb and ADCS-ADL were administered prior to dosing and at months 6, 12 and 18.
Pharmacokinetic Assessments
Blood samples for TTP488 PK analysis were collected prior to dosing at Week 1, at Months 1, 3, 6, 9, 12, 15, 18, and 21 and at Early Termination.
Statistical Analysis
Populations. The full analysis set (FAS) consisted of all subjects who received at least 1 dose of study medication, and had a baseline (if applicable for the endpoint being analyzed) and post-baseline observation for the measurement of interest. The results of this analysis have been presented previously. The on-treatment analysis set was defined as all available on treatment data, where "on treatment" was defined as date of last dose plus 45 days (this definition was used because the drug has a long ½ life of 10–20 days).
Primary Analysis
The primary analysis planned in the study protocol compared differences in mean treatment effect using 5 statistical methodologies that cope with missing data in different ways, with multiple-imputation methods demarked as primary and others as supportive (supportive methods included endpoint analysis, observed cases, generalized estimating equations (GEE), and mixed-models repeated measures on the longitudinal data). Point estimates, standard errors, confidence intervals, and p-values were computed using the statistical models as planned. For all analyses alpha = 0.05, as the supportive analyses were planned to ensure robustness against missing data.
Baseline measures of the variable of analysis are recommended covariables for statistical modeling. Subgroup analysis for covariables of baseline severity of AD can be based on MMSE or ADAS-cog, the latter of which is the variable of analysis. Use of the baseline ADAS-cog can reduce heterogeneity, thereby increasing the sensitivity of detecting delineation between treatments.
Pharmacokinetic/Pharmacodynamic Analyses
Blood samples (5 mL) for TTP488 plasma concentrations were collected in dipotassium (K2) EDTA tubes prior to dosing at Week 1, months 1, 3, 6, 9, 12, 15, 18, 21 and at Early Termination. Samples were centrifuged at approximately 1700 g for about 10 minutes at 4°C with plasma stored in polypropylene tubes at approximately -20°C within 1 hour of collection. Plasma samples were analyzed for TTP488 concentrations using a validated HPLC-MS/MS method.
Exploratory analyses relating TTP488 plasma concentration (including the 20 mg/day and 5 mg/day dose groups) to ADAS-cog values, and changes over time, utilized deriving a subject-level value by two methods: (1) deriving the subject level value by the maximum of the trough concentration values for that subject over the 18-month period, and (2) deriving the subject level value by taking the median concentration value for that individual. Analyses were done at the subject level.
Subjects were classified into concentration groups according to quintile cut-points in the distribution of concentration values ignoring administered dose.
Analysis of Mild vs. Moderate Subgroups
Protocol-planned analyses included subgroup analysis based on baseline severity of AD. ADAS-cog and CDR-sb changes from baseline were analyzed (ANCOVA adjusted for baseline main effects) by baseline disease severity using an MMSE based definition of mild AD (MMSE ≥ 20). ADAS-cog, CDR-sb and ADCS-ADL were additionally evaluated (2-sample t-test) using an ADAS-cog based definition of mild AD (ADAS-cog ≤ 19). An ADAS-cog value of 19 was selected based on conversion of a traditional cut-off of an MMSE value of 20, used in the analyses reported above, to a corresponding ADAS-cog value using the previously described linear relationship between ADAS-cog and MMSE (ADAS-cog = 56.4–1.86*MMSE).
Sample Size
With approximately 133 subjects per group, the primary study had 80% power to detect a 3 point difference in change from baseline to 18 months in ADAS-cog scores between a TTP488 dose group and placebo, allowing for 25% missing data and two interim analyses. ADAS-cog 18- month changes from baseline were assumed to have a standard deviation of 6.5 points for all treatment groups. A total experiment-wise Type 1 error rate of α= 0.05 was targeted.
Source: ...