The newer anticonvulsants in the treatment of generalised anxiety disorder: a systematic review and meta-analysis

Generalised anxiety disorder (GAD) is a common, chronic and debilitating mental disorder impairing quality of life and functioning. The 1st line treatments for GAD include the selective serotonergic reuptake inhibitors (SSRIs) and the selective serotonergic noradrenergic reuptake inhibitors (SNRIs). However, they have rates of non-response ranging from 25 to 40%. There is justification to search for new and more efficacious GAD medication. It has hypothesised anticonvulsants possess anxiolytic properties based on animal studies and epilepsy trials. There is inconsistent evidence that anticonvulsants are efficacious in GAD. It was considered useful and timely to investigate this further. The newer anticonvulsants (felbamate, gabapentin, lamotrigine, levetiracetam, oxcarbazepine, pregabalin, tiagabine, topiramate and zonisamide) were investigated as they were considered to have a more benign side effect profile and fewer drug interactions than older anticonvulsants. This study is a systematic review and meta-analysis of the newer anticonvulsants in the treatment of GAD. The main objective was to use randomised controlled trial (RCT) data to estimate efficacy of the newer anticonvulsants in GAD. using A search strategy was designed and three separate searches conducted by the Cochrane Depression Anxiety and Neurosis Group of the Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register (CCDANCTR) and clinicaltrials.gov (the last search in May 2013). An updated, independent, search was conducted in May 2016 with no additional citations retrieved. 287 citations were retrieved and screened in total. Two independent raters assessed citations using the abstracts and selected trials that satisfied the inclusion criteria. 12 RCTs were included with eight using pregabalin and four using tiagabine. A single rater collated data from RCTs assisted by Covidence Systematic Review Software. All statistical analyses were performed using Review Manager. A random effects meta-analysis was performed expressing summary statistics as effect estimates with 95% confidence intervals (CI). There were 4001 participants in total with 2516 in the anticonvulsant group and 1485 in the placebo. Primary outcomes were reduction in symptom severity using the Hamilton Anxiety Rating Scale (HAM-A) and treatment response using the Clinical Global Impressions Scale-Improvement item (CGI-I). Secondary outcome was medication acceptability. Reduction of symptom severity on the HAM-A for: the anticonvulsant group (pregabalin and tiagabine combined) was significantly favourable with a mean difference (MD) of -2.10 ([-2.83, -1.36] 95% CI); pregabalin was significantly favourable (MD -2.86 [-3.52, -2.21] 95% CI) tiagabine was statistically insignificant (MD - 0.58 [-1.41, 0.25] 95% CI). The risk ratio (RR) of treatment response using the CGI-I (RR >1 favours the anticonvulsant) for: the anticonvulsant group was significantly favourable (RR 1.23 [1.12, 1.35] 95% CI); pregabalin was significantly favourable (RR 1.35 [1.21, 1.50] 95% CI) tiagabine was statistically insignificant (RR 1.09 [0.98, 1.22] 95% CI). The RR of treatment acceptability (RR >1 favoured placebo) for: the anticonvulsant group was significantly unfavourable (RR 1.49 [1.18, 1.88] 95% CI); pregabalin was statistically insignificant (RR 1.23 [0.92, 1.65] 95% CI) tiagabine was significantly unfavourable (RR 1.95 [1.29, 2.93] 95% CI). In conclusion, this systematic review of the newer anticonvulsants included only RCTs of pregabalin and tiagabine. The main finding was that pregabalin showed significant efficacy in reducing symptom severity and improving treatment response in GAD. Tiagabine failed to show significant efficacy in primary outcomes. Further work is needed to better clarify the place of the newer anticonvulsants in the treatment armamentarium of GAD