Posttraumatic stress disorder (PTSD) is a difficult-to-treat condition that over a lifetime affects approximately 10% of the general population 1. The condition develops after traumatic events such as combat, terror activities, disaster, or rape, and has 3 main features: (1) reexperiencing the trauma through recollection, dreams, and reliving, (2) avoidance of thoughts, activities, and emotions associated with the trauma, and (3) hyperarousal 2. PTSD is usually a chronic disorder, with one third of patients displaying symptoms for ≥ 10 years after experiencing the traumatic event 34. Generally the response to pharmacotherapy has been poor, with many patients completely unresponsive and others only marginally responsive 5. Some tricyclic antidepressants, monoamine oxidase inhibitors, and selective serotonin reuptake inhibitors have demonstrated efficacy in double-blind trials 3. The complex neurobiology of PTSD involves a number of systems, including dopaminergic, serotonergic, sympathetic, hypothalamic-pituitary-adrenal, and various anatomic regions of the amygdala and other parts of the limbic system 3. It has been suggested that after traumatic events limbic nuclei may become kindled or abnormally sensitized 6, leading to increased susceptibility for psychic and physical arousal and psychiatric disturbance 2. Because of the suggested involvement of the kindling phenomenon, several anticonvulsants have been assessed in the treatment of PTSD, including carbamazepine, valproate, lamotrigine, gabapentin, tiagabine, and topiramate 3.

Topiramate has a broad spectrum of pharmacologic properties, including Na⁺ channel blockade 7891011, inhibition of some high voltage-activated Ca²⁺ channels 12, enhanced γ-aminobutyric acid (GABA) neuroinhibition at novel GABA_A receptors 1314, glutamate inhibition at kainate and α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors 1516, and promotion of protein phosphorylation of neuronal conductance channels 15. These properties, together with inhibitory activity in animal kindling models 1718, suggest that topiramate may have therapeutic potential in PTSD.

Treatment with topiramate has been reported to improve reexperiencing symptoms associated with civilian PTSD 6. Results from that study, which included a number of patients who were classified as treatment resistant, suggested that topiramate fully or partially suppressed both intrusions (distressing recollections or nonhallucinatory flashbacks) and nightmares, if present, in 89% of patients with nonhallucinatory PTSD.

Although the DSM-IV definition of PTSD lists hallucinations among reexperiencing symptoms 19, because a few patients displayed varying degrees of impaired reality testing of hallucinations, it was difficult to determine if hallucinations were due to PTSD, to an independent psychotic disorder, or, in some individuals, to both. Although psychotic variants of PTSD have been reported in as many as 40% of veterans with chronic combat-associated PTSD without evidence of a primary psychotic disorder 2021, it is also possible for patients with psychotic disorders to have comorbid PTSD, leaving questions about whether specific hallucinations should be attributable to the PTSD reexperiencing cluster or to a psychotic disorder. Because the prior study found a more robust effect in the group of PTSD patients without hallucinations 6, it was decided to focus further on the core group of PTSD patients without ambiguous symptoms confounded by the possible presence of independent psychotic disorders.

Based on these earlier results of topiramate in PTSD, a fully prospective open-label study was begun to see whether the previously observed signal of a therapeutic benefit of topiramate for chronic civilian PTSD could be confirmed. None of the patients reported in the earlier study were included in the present study. In order to address methodological limitations of the prior study, the current study is modified in 4 major ways: (1) it excludes patients with hallucinations, (2) it limits the study to 12 weeks, (3) it prospectively employs the PTSD Checklist-Civilian Version (PCL-C) as a validated clinical assessment instrument in all patients, and (4) it determines a clinical response rate as of week 4, using a conventional, predetermined definition for positive response as a ≥ 30% improvement in a standardized clinical rating scale.

This sample included all consecutive adults (n = 33) meeting the DSM-IV criteria for chronic civilian PTSD in an outpatient private practice who started open-label topiramate between January 2000 and November 2002 in the course of clinical care. The study excluded patients with concurrent auditory or visual hallucinations associated with either PTSD or a possible psychotic disorder. Topiramate, administered either as monotherapy (n = 5) or added to the patients' existing therapeutic regimen (n = 28) (Table 1), was initiated at a starting dosage of 12.5 to 25 mg/day and increased in 25- to 50-mg/day increments every 3 to 4 days as tolerated until a clinical response was achieved.

The study was performed according to the Declaration of Helsinki. Each participant provided verbal informed consent based on disclosure of off-label usage, the availability of alternative medications for PTSD such as sertraline and paroxetine, and expected adverse events and risks. Verbal consent was deemed appropriate and sufficient because of the nonexperimental nature of the treatment. Because this was a descriptive study using aggregate data gathered in the course of treating patients in clinical practice for their own benefit and not an experimental design, there was no protocol that required Institutional Review Board approval. Patients were asked to measure their symptoms using standardized clinical instruments already routinely used as a part of individual care. The study included no features of clinical studies, such as randomization, concealment of treatment, or surrender of patient decision-making rights, as an integral part of conducting the study.

The primary outcome was a change in PTSD symptoms calculated from change in PCL-C score from baseline through 4 weeks of treatment 22. The PCL-C, a validated self-report instrument, strongly correlates (k~0.90) with the Clinician-Administered PTSD Scale (CAPS), a structured clinical interview widely used in PTSD studies 23. The PCL-C is scored on a Likert scale of 1 to 5 (1 = not at all; 5 = extremely) for each of 17 items that cover all 3 symptom clusters (B, reexperiencing; C, avoidance/numbing; D, hyperarousal), yielding a score range of 17 to 85. A total score of ≥ 50 meets the threshold for active PTSD. In accordance with conventional usage, a 30% reduction in clinical symptoms of PTSD was defined as a positive clinical response. The secondary measures included were Cluster B symptoms of reexperiencing of trauma, including trauma-related intrusions and nightmares. Secondary measures were limited to Cluster B symptoms because of the practical difficulty patients in clinical practice can have linking a symptom (such as avoidance, numbing, hyperarousal, insomnia, or startling) specifically to a trauma as opposed to some other source, as required by the DSM-IV definitions. Also, Cluster B symptoms appeared to be among the most distressing and salient for many patients, thereby making tracking these distinctive symptoms comparatively easy.

Data on patient demographics, concomitant medication, comorbidities, primary trauma, and duration of PTSD were collected at baseline. Assessments were made through clinical interviews over a 12-week period. Cessation of reexperiencing symptoms was considered partial if subjects reported a definite reduction in intensity and frequency of nightmares or intrusive recollections or flashbacks. Full cessation was defined as complete elimination of nightmares and intrusions for a sustained period. Adverse effects were recorded if they resulted in treatment discontinuation.

Statistical analyses were performed using Jandel SigmaStat^® Version 2.0 (Jandel Corporation, San Rafael, CA, USA). A single paired t-test of before and after total PCL-C scores was used to analyze the data. Subscale scores for PCL-C Clusters B, C, and D were also evaluated.

Results of this prospective open-label study suggest that the use of topiramate for the treatment of chronic PTSD in civilians is indeed promising and warrants controlled clinical trials. Topiramate was able to improve symptoms of reexperiencing, avoidance, and hyperarousal by 49% overall, compared with baseline, and decreased intrusions in 94% of patients and nightmares in 79% of those with this symptom.

Although there are numerous other reports describing the use of antiepileptic agents for the treatment of PTSD, most are case reports and case series, and some lack the use of standardized PTSD scales. The first published trials using anticonvulsants date back almost 20 years, when a 5-week trial of carbamazepine in 10 combat veterans produced a 36% reduction in a nonvalidated interviewer-scored instrument 24. Studies on adjunctive therapy with valproate reported improved global scores in avoidance and hyperarousal, but not reexperiencing symptoms in veterans over a 14-month period 25, whereas a subsequent 8-week open-label trial reported a small, 17% reduction in CAPS score in 13 subjects with significant reductions in all 3 cluster criteria 26. Two studies in which valproate was used as monotherapy reported varying results, with one showing a 30% reduction in total CAPS scores in 28 veterans 27, whereas the other in 10 civilian patients with non-combat-related PTSD showed no improvement 28.The only double-blind placebo-controlled trial evaluating an anticonvulsant in PTSD was conducted with lamotrigine 29. Overall, after 10 weeks of treatment, 5 of 10 subjects were characterized as responders, compared with 1 of 4 subjects receiving placebo. However, a last observation carried forward analysis revealed only a 23% decrease in mean pretreatment scores with lamotrigine, compared with 20% in the placebo group. A retrospective clinical series of adjunctive gabapentin therapy in 30 consecutive patients with PTSD assessed improvement of target symptoms such as nightmares, insomnia, and irritability. It was reported that the majority of patients (77%) showed moderate or greater improvement in duration of sleep, and most notably a decrease in the frequency of nightmares 30. These studies highlight the need for double-blind placebo-controlled trials using standardized and validated questionnaires.

It is encouraging that the results of this current study with topiramate were consistent with the earlier, preliminary data obtained in patients with PTSD. Percentage responders (77% versus 75%), median time to full response (9 d versus 8 d), suppression of intrusions and nightmares (88% versus 89%), and discontinuation rate (39% versus 32%) were all similar. Overall, the open-label experience with topiramate, based on patient statements and self-rating reports, suggests that the drug may have a rapid rate of response with limited risk of clinically significant adverse events and no evidence of tolerance developing over time 631. Moreover, from topiramate studies in epilepsy [Topamax prescribing information] it is known that the drug is not associated with the cardiac, pancreatic, and hematologic toxicity found with valproate or carbamazepine 3. In addition, topiramate is not associated with weight gain, a side effect reported with valproate that can predispose patients to diabetes mellitus 32. Topiramate also offers potential clinical advantages over serotonin reuptake inhibitors (SSRIs) because it does not appear to destabilize mood in patients with comorbid bipolar disorder and is relatively devoid of common adverse effects of SSRIs, such as sexual dysfunction, weight gain with chronic use, or sedation when used as monotherapy.

Limitations of the present study include: (1) the absence of structured assessment data following week 4 to test the maintenance of response over time, (2) use of self-report instruments that, although validated and correlated with results of structured clinical interviews, may be less accurate than structured interviews, and (3) the absence of clinical assessment scores at the time patients discontinued the trial. From a broader perspective, the greatest limitation of this study is characteristic of all open trials: the lack of standard features of clinical trials such as placebo controls, randomization, and blinding of raters.

Fundamentally, the findings of current and prior topiramate studies are convergent and consistently signal a potentially beneficial therapeutic effect for all 3 criteria clusters of chronic PTSD in civilian adults. In both studies, there is a positive response in a large proportion of patients, at dosages considerably below the usual anticonvulsant dosage levels of 200 to 600 mg/day, and a rapid onset. The effect appears independent of comorbidity with bipolar disorder, age, sex, duration of symptoms, baseline severity of illness, or administration alone or with other psychotropic medications. The purpose of this study was to prospectively test the results of an earlier open-label trial in a new sample; however, the limitations of those open studies will have to be addressed in double-blind, placebo-controlled studies, which are currently in progress.

The author is a consultant to and a participant in the Speaker's Bureau for Ortho-McNeil Pharmaceutical, Inc. He also holds a licensing agreement with Ortho-McNeil based on issuance of a United States Government patent for topiramate therapy of PTSD.

JLB conceived of the study and its design, coordination, statistical analysis, and drafted the manuscript.