Post Hoc Subgroup Analysis by Disease Severity in a Non-Inferiority Trial of Donepezil 27.5 mg Transdermal Formulation in Japanese Patients With Mild-to-Moderate Alzheimer Disease.

BACKGROUND

The efficacy of cholinesterase inhibitors varies depending on the severity of Alzheimer disease. We examined the effects of the disease severity on the efficacy of donepezil 27.5 mg patches in Japanese patients with Alzheimer disease.

METHODS

A post hoc analysis of covariance using per-protocol set in a noninferiority study of donepezil 27.5 mg patches with donepezil hydrochloride 5 mg tablets (JapicCTI-194582) was conducted after imputation of missing data using multiple imputations.

RESULTS

No apparent imbalances in the extracted confounding factors were observed between the 2 treatment groups. The baseline value of the Alzheimer Disease Assessment Scale (Japanese version) cognitive subscale was adjusted using analysis of covariance. Least-squares mean of changes from baseline in the Alzheimer Disease Assessment Scale (Japanese version) cognitive subscale at week 24 for each group were -1.661 (-2.681 to -0.640) (donepezil patch) and 0.065 (-0.987 to 1.117) (donepezil hydrochloride tablet). The difference in the least-squares mean (95% CI) between 2 groups was -1.726 (-3.1913 to -0.2602, P =0.021).

CONCLUSIONS

A post hoc analysis of covariance using the per-protocol set suggested that donepezil patches may be more effective than donepezil tablets in slowing cognitive decline in patients with mild Alzheimer disease.