Effect of phosphodiesterase-5 inhibitors on cerebral blood flow in human beings: a systematic review

Abstract

BackgroundPhosphodiesterase-5 inhibitors (PDE5i) are established in the treatment of erectile dysfunction and pulmonary hypertension. Agents that augment cerebral blood flow could be a potential treatment for vascular cognitive impairment. The aim of this review was to assess the published effects of PDE5i on cerebral blood flow in human beings. MethodsInterventional and observational studies investigating the effects of PDE5i on cerebral blood flow in adults were considered for inclusion. Embase, Medline, and the Cochrane Library Trials databases were searched between Aug 14 and Aug 25, 2015; on April 4, 2016; and again on Nov 14, 2016. Reference lists of retrieved articles were also checked. Risk of bias in individual studies was assessed using the Cochrane bias assessment tool. Findings15 studies with a total of 339 participants were included. Six studies were double-blind, randomised controlled studies (103 participants), one was a non-blinded randomised study (30), and two were non-blinded non-randomised controlled studies (68 in total). Six observational studies (138 in total) were also included. Methodological quality was variable. Techniques used to measure cerebral blood flow were heterogeneous, the commonest being middle cerebral artery velocity by insonation. In studies mainly using transcranial doppler imaging, PDE5i did not alter basal middle cerebral artery blood flow. However, there was evidence that PDE5i improved cerebrovascular reactivity (measured by the response to carbon dioxide inhalation, for example) in conditions causing endothelial dysfunction. No serious adverse effects were reported. InterpretationOur systematic review shows that the effect of PDE5i on cerebral blood flow remains incompletely understood and that evidence is limited by studies in disparate populations using heterogeneous techniques. Cerebrovascular reactivity, but not basal cerebral blood flow, was improved by PDE5i in the presence of endothelial dysfunction, perhaps because deficient nitric oxide-mediated signalling disproportionately affects responsiveness compared with the resting state. That PDE5i did not produce increases in cerebral blood flow detectable with middle cerebral artery insonation at rest does not preclude the possibility that resting cerebral blood flow at the level of the small arterioles, such as those involved in small vessel disease, is altered, because these are not assessed on transcranial doppler imaging. Future studies using better surrogates of deep cerebral blood flow measurements (eg, arterial spin labelling MRI) are warranted to explore effects at the arteriolar level. Future studies should also correlate changes in cerebral blood flow with clinical parameters such as cognitive function. FundingAlzheimer's Drug Discovery Foundation, Alzheimer's Society (UK).