Objective An increased incidence of cerebrovascular adverse events (CVAEs) in placebo-controlled registration trials of oral risperidone (RIS) has been reported from some, but not all, trials in patients with dementia. This analysis explores data concerning risk of serious cerebrovascular events in patients with dementia.
Methods: Data were derived from studies of RIS in elderly patients with dementia. This included placebo-controlled short-term (<12 weeks) studies conducted in the U.S., Australia, Belgium, and international sites, followed in some cases with longer-term safety extensions. The incidence of CVAEs was assessed from spontaneous reports of adverse events. Risk of CVAEs was compared in the RIS group versus the PBO group. Analyses were performed on various risk factors (when available) and demographic characteristics, and brief clinical narratives were obtained when available. “Serious” (e.g., reported outcomes of death or permanent disability) events were analyzed separately from other events for dementia trials.
Results: Data on 1230 individual patients from four double-blind studies in dementia were available. Mean age in the dementia trials was 82 years. The number of serious cerebrovascular events reported in the dementia trials for RIS and PBO combined was low (15 of 1230; 1.22%). In double-blind phases of two of four studies (one very large U.S., one very small Belgian) in patients with dementia, the rate of serious CVAEs was not higher with RIS compared to PBO. In the remaining two of four studies in dementia, an greater number of serious CVAE was observed in the RIS group. All-cause and stroke-specific mortality were assessed in the 1396 unique patients studied in the aforementioned dementia trials and additionally two safety extension trials and one small (n=18) study that was terminated early. The all-cause mortality for patients taking RIS was 174 per 1000 patient-years compared to 184 per 1000 patient-years for PBO. The stroke mortality for patients taking RIS was 12.4 per 1000 patient-years with RIS compared with 11.5 per 1000 patient years with PBO. Most patients experiencing serious CVAEs in these studies had one or more significant pre-existing cerebrovascular risk factors (e.g., prior stroke, hypertension, and atrial fibrillation). Unlike serious CVAE, the number of “other” CVAE was generally higher among RIS-treated patients. No consistent relationship between CVAE and dose, diagnosis, or co-medications was identified. Time to onset was variable. Case reports of serious and “other” events will be presented.
Conclusions: Risk of serious cerebrovascular events was varied. In dementia, two studies showed no difference in the risk of serious CVAEs with RIS compared to PBO, and two showed an increased risk. When all dementia trial patient exposure is considered, all-cause mortality was comparable between RIS and PBO. A consistent difference between RIS and PBO was observed primarily in “other” CVAE events. When treating this fragile patient population, non-pharmacological interventions should be considered first. When pharmacological intervention is indicated, the risks and benefits of treatment should be carefully assessed for each patient.
Back to PB Tuesday Poster Sessions
Back to The Eleventh International Congress