Open-Label Topiramate as Primary or Adjunctive Therapy in Chronic Civilian Posttraumatic Stress Disorder: A Preliminary Report

Background: The hypothesis that exposure to traumatic events may sensitize or kindle limbic nuclei has led to efforts to treat posttraumatic stress disorder (PTSD) with anticonvulsants. Based on the kindling hypothesis of PTSD, this open-label study assesses clinical response to topiramate as a potential treatment for DSM-IV PTSD.

Method: A naturalistic data review was conducted of medical records of all adult outpatients (9 men, 26 women symptomatic for a mean±SD of 18±15 years with DSM-IV chronic civilian PTSD) treated with topiramate, 12.5 to 500 mg/day, as add-on (N=28) or monotherapy (N=7). The last 17 patients completed the PTSD Checklist-Civilian Version (PCL-C) before treatment and at week 4. Dosage titration started at 12.5 to 25 mg/day and increased in 25- to 50-mg increments every 3 to 4 days until a therapeutic response was achieved or the drug was no longer tolerated. The mean duration of treatment was 33 weeks (range, 1-119 weeks).

Results: Topiramate decreased nightmares in 79% (19/24) and flashbacks in 86% (30/35) of patients, with full suppression of nightmares in 50% and of intrusions in 54% of patients with these symptoms. Nightmares or intrusions partially improved in a median of 4 days (mean=11±13 days) and were fully absent in a median of 8 days (mean=35±49 days). Response was seen in 95% of partial responders at a dosage of 75 mg/day or less, and in 91% of full responders at a dosage of 100 mg/day or less. Mean reductions in PCL-C score from baseline to week 4 were highly significant (baseline score=60 vs. week 4 score=39, p2-year study period. Except for a single instance of acute secondary narrow-angle glaucoma, there were no serious side effects.

Conclusion: Topiramate appeared effective as add-on or monotherapy for chronic PTSD. It demonstrated a rapid onset of action and minimally serious, dose-related side effects without the development of tolerance. Double-blind studies are indicated.