Sertraline Treatment of Panic Disorder: Response in Patients at Risk for Poor Outcome

Background: More than one third of panicdisorder patients have a chronic and/or recurrent form of thedisorder, accounting for much of the individual and societal costassociated with the illness. Six clinical variables have beenmost consistently identified as high-risk predictors of pooroutcome: (1) panic severity, (2) presence of agoraphobia, (3)comorbid depression, (4) comorbid personality disorder, (5)duration of illness, and (6) female sex. No published researchhas systematically examined the differential antipanic efficacyof selective serotonin reuptake inhibitors in patients at highrisk for poor outcome.

Method: Data were pooled (N = 664) from 4double-blind, placebo-controlled studies of the efficacy ofsertraline for the treatment of DSM-III-R panic disorder. Two ofthe studies were 12-week fixed-dose studies with starting dailydoses of sertraline, 50 mg, and 2 were 10-week flexible-dosestudies with starting daily doses of sertraline, 25 mg. All otherstudy design features were the same, except for the exclusion ofwomen of childbearing potential in the 2 fixed-dose studies.Exclusion of patients with marked personality disorders anddepression meant that only 4 of the poor-outcome variables couldbe evaluated.

Results: Clinical improvement was similarfor patients treated with sertraline whether or not they carriedan agoraphobia diagnosis, had a duration of illness > 2 years,or were female. Patients with high baseline panic severity hadsignificantly (p = .01) less improvement on the endpoint ClinicalGlobal Impressions-Improvement (CGI-I) scale than patients withmoderate severity, although the Clinical GlobalImpressions-Severity of Illness scale change score was higher inthe patients with high severity (-2.00 vs. -1.31). For patientswith 3 or more high-risk variables, there was a modest, butstatistically significant, tendency for reduced globalimprovement (endpoint CGI-I score of 2.7 for the high-risk vs.2.4 for the non-high-risk group; p = .017), although thehigh-risk group actually had a similar endpoint reduction infrequency of panic attacks (82%) compared with the non-high-riskgroup (78%).

Conclusion: Treatment of panic disorder withsertraline was generally effective, even in the presence ofbaseline clinical variables that have been associated with poortreatment response. The main limitations of the analysis were thereliance on pooled data from 4 studies (even if the designs weresimilar) and our inability to examine the impact of depressionand personality disorders on response to treatment because of theexclusion criteria of the clinical trials.