AbstractBackgroundTeprenone (geranylgeranylacetone), a gastric mucosal protective agent, has been reported to inhibit the accumulation of coagulated amyloid‐β by increasing heat shock protein 70 (HSP70) in mice.ObjectiveWe conducted a randomized, double‐blind, placebo‐controlled study to ascertain teprenone’s therapeutic ability for AD.MethodsPatients with mild to moderate AD, with a Mini‐Mental State Examination (MMSE) score of 13 to 26, were randomly allocated into two groups depending on the administered drug: donepezil + placebo (placebo group) and donepezil + teprenone (teprenone group). The primary and secondary endpoints included changes in scores of the Japanese version of the AD Assessment Scale‐cognitive subscale (ADAS‐J cog) and other evaluations, respectively, including MMSE scores, during a 12‐month period after the first administration.ResultsForty‐two and thirty‐seven patients were allocated to the teprenone and placebo groups, respectively. ADAS‐J cog score changes were not different between groups (placebo, 0.6±0.8; teprenone, 0.4±0.8; p = 0.861). However, MMSE scores significantly improved in the teprenone group (placebo, ‐ 1.2±0.5; teprenone, 0.2±0.5; p = 0.044). Subgroup analysis considering the severity of medial temporal area atrophy revealed that this improvement by teprenone was significant in patients with mild (p = 0.013) but not with severe atrophy (p = 0.611). Adverse events were observed in 17.8 and 10.4% of patients in the placebo and teprenone group, respectively.ConclusionTeprenone may be effective for AD when administered before atrophy progression in the medial temporal areas. (J Alzheimers Dis. 2019;71:1187‐1199)