Anxiety disorders are a common, heterogeneous group of illnesses marked by varied clinical presentations, ranging from relatively moderate symptom levels to severe functional impairment and profound disability. Pharmacotherapy, often coupled with behavioral treatments, is a standard approach to treating anxiety disorders. First-line pharmacological treatments for the majority of anxiety disorders are the selective serotonin reuptake inhibitors and serotonin norepinephrine reuptake inhibitors (1). Sedative-hypnotics are commonly employed for anxiety disorders but are frequently reserved for short durations and targeted symptoms (e.g., insomnia) because of the potential for abuse, dependency, withdrawal reactions, and cognitive impairment (1). The role for antipsychotic drugs in the treatment of anxiety disorders, however, is less clear because of the relative paucity of well-designed trials to assess their safety, efficacy, and effectiveness.

The article by Comer et al. (2) in this issue of the Journal provides an invaluable contribution to our knowledge of the pharmacotherapy of anxiety disorders by underscoring the high and growing utilization of second-generation antipsychotic prescriptions for anxiety disorders in the office-based psychiatry setting. A strength of the study is that it uses data from psychiatrists as opposed to primary care physicians and other nonpsychiatric prescribers because of the greater likelihood of accurate diagnoses of anxiety disorders. Using an extensive database of 4,166 office visits from 1996 to 2007, Comer et al. found that the proportion of visits for anxiety disorders in which an antipsychotic was prescribed increased from 10.6% to 21.3% over the 12-year observation period. This trend was paralleled by the introduction of several second-generation antipsychotics and the widespread adoption of this class of agents for psychotic and mood disorders. The second-generation antipsychotics have fewer or less severe motor side effects than first-generation agents, which may account for their increased use for anxiety disorders despite the high rate of weight gain and metabolic side effects associated with some of the second-generation agents and the risk of cerebrovascular events in the elderly.

The Comer et al. article raises many critically important questions that have implications for clinicians, patients, and health care systems. The authors used the National Ambulatory Medical Care Survey database, which is an excellent resource for the aims of the study, but the database has a number of limitations that prevent a full interpretation of the results presented. Perhaps the most important limitation is that it does not provide extensive information about the clinical status and treatment outcomes of the patients in the database. Thus, we do not know if the patients referred to the outpatient psychiatry practices included in the study were, for example, particularly complicated cases or had inadequate responses to first-line treatments.

There is a lack of evidence in the scientific literature for the use of antipsychotics as a first-line treatment for anxiety disorders, but there is support for their role as an adjunctive treatment for treatment-refractory obsessive-compulsive disorder and posttraumatic stress disorder and for cases complicated by comorbid conditions (3, 4). The largest increase in antipsychotic treatment reported in the study was for panic disorder, which is another anxiety disorder that is associated with inadequate efficacy for many patients treated with first-line therapies. The Comer et al. study provides hints that at least a portion of the patients were complicated cases and perhaps candidates for alternative treatments. The majority of patients treated with antipsychotics had other diagnoses in addition to their anxiety disorders. Only 8.5% of the anxiety patients treated with antipsychotics in the final year of the study had only an anxiety disorder, with no additional nonanxiety disorders noted. There were also significant trends of increases in frequency of antipsychotic prescriptions for patients with comorbid mood, nonmood, and nonpsychotic disorders, suggesting a complicated clinical presentation. No increases were observed in frequency of antipsychotic prescriptions for patients with comorbid psychotic disorders, however, suggesting that antipsychotic treatment in this subgroup remained relatively stable over the observation period. The database did not provide information related to whether patients had failed to benefit from first-line pharmacotherapy before antipsychotics were prescribed.

Another factor that may have contributed to the rise in antipsychotic prescriptions during the time frame of the study's data collection is widespread and inappropriate marketing practices. While psychiatrists are entitled to use medications for nonapproved conditions such as anxiety disorders, U.S. Food and Drug Administration (FDA) regulations forbid pharmaceutical companies from promoting the use of medications for nonapproved conditions—yet there is documentation that the marketing of second-generation antipsychotics for nonapproved conditions took place during the period of data collection for this study (5).

The use of antipsychotics for anxiety disorders is not approved by the FDA. This point raises a dilemma for clinicians, patients, and payers as to what steps to take when FDA-approved medications fail to deliver adequate clinical outcomes. This scenario is commonly faced in the field of psychopharmacology as prescribers must decide to go off-label and, for example, use higher than approved doses, add combination therapy, or switch to other medications that may not be FDA approved. Indeed, there are numerous DSM-IV disorders for which there is no FDA-approved pharmacotherapy. Moreover, there are very few examples in mental health of FDA-approved medications for patients with treatment-refractory illness—those who have not achieved an adequate response with FDA-approved treatments and are still in need of symptom relief and functional improvement. One exception is clozapine for treatment-refractory schizophrenia; another is aripiprazole or quetiapine in combination with antidepressants for major depressive disorder when efficacy is inadequate with antidepressant monotherapy. For clinical conditions for which there are no FDA-approved treatments or for which FDA-approved treatments have failed and there is still a clinical need for treatment, it is incumbent on the academic and clinical fields to provide guidance in the form of expert consensus guidelines and treatment algorithms that are supported by comprehensive reviews of the clinical literature. Two examples are the Schizophrenia Patient Outcomes Research Team (PORT) (6) and the Davidson et al. treatment guidelines for generalized anxiety disorder (7).

Perhaps the greatest value of the Comer et al. study is that it calls attention to the pressing need for more research to evaluate antipsychotics for use in patients with anxiety disorders given the high and growing use of these agents in this patient group. We need both well-controlled randomized clinical trials that evaluate the safety and efficacy of antipsychotics for anxiety disorders and large-scale effectiveness studies that determine outcomes in real-world practice settings. Key questions include the following: Do antipsychotics, either as monotherapy or in combination with first-line agents, provide a favorable safety/efficacy profile for anxiety disorders? Is there a unique role for low-dosage antipsychotics in the treatment of anxiety disorders? Do the second-generation antipsychotics that have the most favorable safety profile (the lowest rates of weight gain, metabolic complications, and motor side effects) offer clinical advantages compared with other antipsychotics for patients suffering from anxiety disorders? What treatment approaches are most effective for treatment-refractory anxiety disorders? Given the heterogeneity of anxiety disorders, which individual disorders and symptom complexes, if any, are best suited for antipsychotic therapy?

Once the clinical trial database of antipsychotics for anxiety disorders is enriched, clinicians will be able to make rational, evidenced-based decisions about their clinical use for this important group of disorders.