RESUMO
RATIONALE: Schizophrenia is associated with impairments in cognitive functioning yet there are no approved drugs to treat these deficits. OBJECTIVES: Based on animal models, we investigated the potential for roflumilast, a selective inhibitor of phosphodiesterase type 4 (PDE4), to improve cognition, which may act by increasing intracellular cyclic adenosine monophosphate in brain regions underlying cognitive deficits in schizophrenia. METHODS: This study consisted of a randomised, double-blind, placebo-controlled, crossover design involving 15 schizophrenia patients. In 3 treatment periods, patients were given 8 days of placebo or one of the two doses of roflumilast (100 and 250 µg daily) with 14 days of washout between treatments. The primary endpoints were dorsolateral prefrontal cortex (DLPFC) activation during a visuospatial working memory task measured with fMRI on dosing day 8 and verbal memory and working memory performance change from baseline to day 8. Least square mean change scores were calculated for behavioural outcomes; fMRI data were analysed in SPM12 with bilateral DLPFC as regions of interest. RESULTS: Verbal memory was significantly improved under 250 µg roflumilast (effect size (ES) = 0.77) compared to placebo. fMRI analyses revealed that increasing dose of roflumilast was associated with reduction of bilateral DLPFC activation during working memory compared to placebo, although this was not statistically significant (ES = 0.31 for the higher dose). Working memory was not improved (ES = 0.03). CONCLUSIONS: Results support the mechanistic validation of potential novel strategies for improving cognitive dysfunction in schizophrenia and suggest that PDE4 inhibition may be beneficial for cognitive dysfunction in schizophrenia. TRIAL REGISTRATION: NCT02079844 .