Toward rational pharmacotherapy for posttraumatic stress disorder: reprise.

Twenty-five years ago, I wrote a special article for the Journal with the same title as this one (1). Posttraumatic stress disorder (PTSD) was only 8 years old as a DSMIII diagnosis, and research on PTSD-related psychobiological alterations was in its infancy. As for randomized controlled trials in PTSD patients, none had been published, although a few were in progress at that time. Instead, the sparse literature on pharmacotherapy for PTSD consisted entirely of open trials and case reports on tricyclic antidepressants, monoamine oxidase inhibitors, antiadrenergics (clonidine and propranolol), lithium, and benzodiazepines. I concluded that article optimistically, stating that there was every reason to hope that emerging psychobiological research would lead to rational pharmacotherapy for PTSD. Since that time, we have learned that individuals with PTSD exhibit dysregulation in a number of key psychobiological systems; most research has focused on the adrenergic and hypothalamic-pituitary-adrenocortical systems, although there have also been studies of serotonergic, dopaminergic, opioid, glutamatergic, GABAergic, cannabinoid, and other mechanisms (2). Unfortunately, it is only recently that such discoveries have stimulated a search for novel pharmacological treatments. Instead of a rational approach, focusing on agents designed to address the unique pathophysiology of PTSD, most randomized controlled trials have followed an industry-dominated empirical approach in which medications with proven effectiveness in other disorders have been tested for their efficacy in PTSD. Randomized controlled trials with antidepressants, most notably selective serotonin reuptake inhibitors (SSRIs) or serotonin-norepinephrine reuptake inhibitors (SNRIs), dominate the list of published trials. Indeed, two SSRIs have been approved by the U.S. Food and Drug Administration as indicated treatments for PTSD (sertraline and paroxetine), although the remission rate after 12 weeks of treatment is only about 30% (3). This compares quite unfavorably with the remission rates achieved with cognitive-behavioral approaches, such as prolonged exposure and cognitive processing therapies (4). There are probably good reasons for the inferior performance of SSRIs and SNRIs. First of all, they are too nonspecific. Given the number and various (sometimes antagonistic) actions of the many different serotonin (5-HT) receptors, it is probably remarkable that any efficacy at all has been demonstrated with these drug classes. Second, these agents appear to have an overall permissive utility that addresses a transdiagnostic irregularity in 5-HT function—hence their efficacy in a variety of mood, anxiety, obsessivecompulsive spectrum, and traumaand stress-related disorders. Belated recognition of this fact has recently led to clinical trials with muchmore selective actions at specific 5-HT receptors. Likewise, clinical trials with antiepileptic drugs acting at glutamatergic or GABA-ergic receptors have been evenmore disappointing, with the possible exception of topiramate (3). To summarize, the hope expressed in my 1988 article has not disappeared but has been lacedwith a strong dose of disappointment, caution, and a healthy appreciation for the psychobiological complexity of PTSD. One thing seems clear: dusting off medications developed for other disorders