Sertraline in the Treatment of Panic Disorder: A Double-Blind Multicenter Trial

The study was conducted at 10 sites. Patient recruitment at each site was performed in accordance with the procedures of each site’s institutional review board. After a thorough description of the study to potential subjects, written informed consent was obtained. The study group consisted of male and female outpatients who were 18 years of age or older and met the DSM-III-R criteria for panic disorder, as assessed with the Structured Clinical Interview for DSM-III-R—Patient Version, part I (16).

Subjects were required to have a minimum of four panic attacks, at least one of which was unanticipated, during the 4 weeks before day 1 of the placebo lead-in and at least three, but no more than 100, DSM-III-R-defined panic attacks during the 2-week single-blind lead-in phase. In addition, subjects had to have total scores of 18 or higher on the Hamilton Anxiety Rating Scale (17) and 17 or lower on the 21-item Hamilton Depression Rating Scale (18) at the baseline randomization visit. Patients with current DSM-III-R major depression, organic mental disorders, bipolar disorder, obsessive-compulsive disorder, schizophrenia, or other psychotic disorder were excluded from the trial, as were those with a principal diagnosis of DSM-III-R dysthymic disorder, any anxiety disorder other than panic disorder, or a personality disorder. Psychoactive substance use disorder within the past 6 months and a lifetime history of schizophrenia or other psychotic disorder were other reasons for exclusion.

Subjects were required to stop taking all psychotropic medication, with the exception of chloral hydrate for sleep, before study entry. They were excluded if they had received regular daily therapy with any benzodiazepine within 1 month before the first medication dose in the double-blind phase or had a positive urine screen for benzodiazepines or a positive serum screen for alprazolam on either day 1 or 8 of the placebo lead-in period. Patients with any medical conditions deemed to be clinically significant were excluded; women of childbearing potential were required to use an effective means of birth control.

All patients received a single placebo tablet daily, identical in appearance to the 25-mg sertraline tablet, during the 2-week single-blind placebo lead-in. During the double-blind phase, the patients randomly assigned to sertraline took 25 mg of sertraline daily for the first week and 50 mg for the second week. After this, the daily dose was increased by 50 mg each week for patients who had not yet achieved a satisfactory clinical response. Any dose increase depended on the absence of dose-limiting side effects. The maximal daily dose was 200 mg; patients who responded well to less than the maximal dose continued with this dose for the remainder of the study. The patients took their medication once daily with the evening meal. The dose could be decreased at any time because of adverse events. The patients randomly assigned to placebo during the double-blind phase took a corresponding number of matching placebo tablets.

During the double-blind phase, the patients were evaluated weekly for 4 weeks, then every 2 weeks for three more visits. Each patient maintained a daily diary of symptoms; this information was converted into weekly ratings on the Panic and Anticipatory Anxiety Scale (19) of D.V. Sheehan. In the patient diary, “panic attacks” were defined by four or more and “limited-symptom attacks” by one to three of the criterion C panic disorder symptoms in DSM-III-R. Daily each patient recorded 1) the number of panic and limited-symptom attacks and 2) the duration of the anticipatory anxiety regarding having an attack. Panic attacks and limited-symptom attacks were categorized as unexpected (no cause) or situational (with cause). At each study visit during the double-blind phase the patients also completed a self-evaluation of their improvement, using the anchors of the Clinical Global Impression (CGI) (20) improvement scale. At the beginning and end of the double-blind phase they also completed the Quality of Life Enjoyment and Satisfaction Questionnaire (21), a validated quality of life scale. At each visit, clinicians assessed the patients with the following instruments: the Multicenter Panic Anxiety Scale (22), the Hamilton Anxiety Rating Scale, and the CGI severity of illness and global improvement scales (the global improvement scale was not completed at baseline). The Multicenter Panic Anxiety Scale assesses five aspects of panic disorder (frequency of attacks, degree of distress during attacks, anticipatory anxiety, phobic avoidance of situations, and impairment/interference with work functioning); each aspect is rated by using a 5-point scale (0 represents “none” or “not present”; 4 indicates “extreme frequency,” “distress,” “worry,” “avoidance,” or “impairment”). The CGI improvement scales were used to rate global improvement and also improvement in five discrete panic disorder symptom domains: panic attacks, anticipatory anxiety, phobic avoidance, social/family functioning, and occupational functioning.

Physical examinations were done at the beginning and end of the study; blood pressure, heart rate, and body weight were assessed at each visit. The following tests were performed at the beginning of the placebo lead-in and at the end of weeks 2 and 10 of the double-blind period: ECG, complete blood cell count with differential and platelet counts, chemistry screen, urinalysis, and testing of serum beta human chorionic gonadotropin for pregnancy in women of childbearing potential.

In addition, T3 uptake ratio and T4 were measured at day 1 of the placebo washout. A serum alprazolam screen and a urine drug screen (for benzodiazepines or drugs of abuse) were done on day 1 and day 8 of the placebo lead-in and at the end of weeks 2 and 10 of the double-blind period. Adverse events during treatment were documented by observation and open interview.

The safety analyses included all patients who took at least one dose of medication during the double-blind phase and provided any follow-up data; patients included in the safety analysis who had baseline and postrandomization efficacy data were included in the efficacy analyses. All statistical tests were two-sided.

During the double-blind phase, 168 patients were administered medication; 88 received placebo, and 80 received sertraline. For one patient taking placebo and one patient taking sertraline, follow-up efficacy data were not available. Thus, the efficacy analyses included 166 patients, 87 of whom were treated with placebo and 79 with sertraline.

The patients randomly assigned to sertraline and placebo did not differ on baseline demographic features (sex, racial background, and age), clinical characteristics (duration of illness, weekly numbers of panic and limited-symptom attacks, extent of anticipatory anxiety, and scores on the Multicenter Panic Anxiety Scale, Hamilton anxiety scale, CGI severity rating, and Hamilton depression scale), or social functioning as measured by the Quality of Life Enjoyment and Satisfaction Questionnaire. Overall, 57% of the patients were female (96 of 168), 88% were white (147 of 168), and they had a mean age of 37.5 years (SD=11.5). The mean lengths of illness were 9.5 years (SD=10.1) and 8.6 years (SD=8.2), respectively, for those receiving sertraline and placebo. At baseline, their overall symptoms were of moderate to marked severity; the mean CGI severity score was 4.3 (SD=0.7) for the patients taking sertraline and 4.3 (SD=0.6) for those taking placebo. They were experiencing frequent panic attacks: a mean of 6.4 (SD=7.1) and 5.2 (SD=5.7) panic attacks per week for the sertraline and placebo groups, respectively; the corresponding numbers of limited-symptom attacks were 9.2 (SD=9.7) and 7.2 (SD=7.3), respectively. In addition to their symptoms during attacks, at baseline the patients spent a mean of 31% of their time (SD=26%) worrying about having another attack. At baseline, 48 of the sertraline-treated and 51 of the placebo-treated patients met the DSM-III-R criteria for agoraphobia in addition to panic disorder. In contrast to their substantial anxiety symptoms, the patients had little depression: their mean scores on the Hamilton Depression Rating Scale were 10.9 (SD=4.2) for the sertraline group and 11.4 (SD=3.8) for the placebo group.

Thirty-six patients withdrew from treatment during the study; 74% of the sertraline-treated patients (N=59) and 83% of the placebo-treated patients (N=73) completed the full 12 weeks of double-blind treatment. The most frequent reason for discontinuation of sertraline treatment was adverse experiences; 9% of the sertraline-treated patients (N=7) stopped taking the medication because of adverse experiences, in comparison to 1% of the placebo-treated patients (N=1). In the placebo group, insufficient clinical response was the most common reason for discontinuation; 7% of the placebo-treated patients (N=6) and 1% of the sertraline-treated patients (N=1) stopped treatment because of lack of efficacy.

The mean duration of therapy was 58.2 days (SD=23.2) for the sertraline group and 64.2 days (SD=18.3) for the placebo-treated group. At endpoint, the mean dose received by the sertraline-treated patients was 126 mg/day (SD=62). By week 4, all of the sertraline-treated patients were taking at least 50 mg/day.

Benzodiazepines demonstrate significantly greater efficacy on most measures than placebo after 1–2 weeks (24-26); tricyclics do so after 3–4 weeks (4). In the current study sertraline demonstrated significant efficacy somewhat sooner than has been reported for the tricyclics—by week 3 for the primary outcome measure of major panic attacks. By week 2, 61% of the patients had achieved a 50% reduction in the weekly number of panic attacks. In the sertraline treatment group there was no attrition in the first 3 weeks because of lack of efficacy, suggesting that, despite the initial severity of the disorder in this patient group and the high level of medical help seeking among panic patients in general, the improvement was sufficiently rapid.

High rates of response to placebo are frequently found in panic treatment studies, around 50% according to a recent review (27). This makes it especially difficult to demonstrate efficacy in short-term treatment studies of panic disorder. The 46% placebo response observed at endpoint in the current study is consistent with this finding. At least one naturalistic follow-up of placebo responders in a panic disorder treatment trial (28) demonstrated that most of these patients eventually begin medication treatment. This is consistent with the recurrent course of panic disorder (29).

Overall, the patients tolerated sertraline well; only 9% dropped out of treatment because of side effects. The side effects of sertraline in this study were similar to those reported in studies of sertraline for depression and obsessive-compulsive disorder. A clinically relevant complication frequently reported during treatment of panic disorder with antidepressants is a jitteriness syndrome, consisting of adrenergic symptoms such as nervousness, sweating, and tremor (2). The jitteriness syndrome has been observed to occur in the first 2–3 weeks of treatment with both tricyclics (2, 3) and fluoxetine (30), and it may lead to medication discontinuation.

In the current study, the patients taking sertraline did not report the presence of a jitteriness syndrome during the initial 3 weeks of treatment, and only two sertraline-treated patients discontinued treatment because of any adverse experiences during the first week of treatment, neither of which was compatible with jitteriness. The absence of a jitteriness syndrome in our patients may have been due to the use of a low starting dose (25 mg/day) in our study during the initial week of treatment.

In conclusion, sertraline appears to be a safe, well-tolerated, and highly effective treatment for panic disorder. Therapeutic response occurs relatively rapidly and without notable risk of jitteriness with a 25-mg starting dose. The effective dose appears to be in the range of 50–200 mg/day. Sertraline treatment resulted in significant reductions in symptom severity in many clinically relevant dimensions of panic, including quality of life.