Double-blind comparison of 30 and 60 mg tranylcypromine daily in patients with panic disorder comorbid with social anxiety disorder

Psychiatry Research 175 (2010) 260-265

Contents lists available at ScienceDirect

Psychiatry Research j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p s yc h r e s

Double-blind comparison of 30 and 60 mg tranylcypromine daily in patients with panic disorder comorbid with social anxiety disorder Antonio E. Nardi ⁎, Fabiana L. Lopes, Alexandre M. Valença, Rafael C. Freire, Isabella Nascimento, Andre B. Veras, Marco A. Mezzasalma, Valfrido L. de-Melo-Neto, Gastão L. Soares-Filho, Anna Lucia King, Leila O. Grivet, Arabella Rassi, Marcio Versiani Laboratory of Panic & Respiration. Institute of Psychiatry. Federal University of Rio de Janeiro. R Visconde de Piraja, 407/702. Rio de Janeiro. Brazil

A R T I C L E

I N F O

Article history: Received 15 May 2007 Received in revised form 2 September 2007 Accepted 12 June 2008 Keywords: Anxiety disorder Phobic disorder Antidepressant Monoamine oxidase inhibitor Tranylcypromine Prognosis

A B S T R A C T Our objective was to explore the dose–response relationship in patients with panic disorder and social anxiety disorder comorbidity (DSM-IV). After 1 week of no-drug washout, 36 such patients were assigned to a double-blind controlled comparison of the effects of 30 mg and 60 mg of tranylcypromine, and were followed up for 12 weeks. The main instrument used to measure the number of panic attacks was the Sheehan Panic and Anticipatory Anxiety Scale. The primary outcome measure for social anxiety disorder symptoms was the mean change from baseline in the Liebowitz Social Anxiety Scale (LSAS). After 12 weeks of treatment, panic attacks were reduced 69.6% from baseline in the 30-mg group (n = 19) compared with a 74.8% reduction in the 60-mg group (n = 17). Twelve patients (70.6%) of the higher dose group and 14 patients (68.4%) of the lower dose were completely free of panic attacks. There was no difference in efficacy between the tranylcypromine groups in the panic disorder symptoms. The 60-mg dose was more efficacious as measured by the LSAS scores, showing a significant difference in relation to the lower group. Mean change from baseline in LSAS total score (mean ± SD) for 30-mg group was − 17.9 ± 14.7 and for the 60-mg group was − 35.0 ± 14.8. The social anxiety symptom scale showed a two-fold greater change with the 60-mg dose, and the 30-mg dose group could be considered the equivalent of a placebo control group. Tranylcypromine – 60 mg daily – was found effective in the treatment of panic disorder and social anxiety disorder comorbidity. © 2008 Elsevier Ireland Ltd. All rights reserved.

1. Introduction The concept of comorbidity has clinical and theoretical implications (Frances et al., 1990). Situational panic attacks are common in social anxiety disorder (Liebowitz et al., 1985), and non-situational panic attacks occur frequently in anxiety disorders as well as in panic disorder (Liebowitz et al., 1985; Frances et al., 1990). The diagnostic dilemma presented by patients who exhibit both panic disorder and agoraphobia and who avoid certain performance or social situations has been addressed by Liebowitz et al. (1985). The term “primary social anxiety disorder” was suggested for those patients who fear scrutiny or evaluation by others and whose anxiety is confined to social situations or anticipation. “Secondary social anxiety disorder” would apply to those patients who have panic attacks in nonsocial situations and fear/avoid situations where an exit is not available. Occasionally, cases of anxiety disorder comorbidity are described as a clinical curiosity (Goldstein, 1987). Goldstein (1987) presented three

⁎ Corresponding author. Laboratory of Panic & Respiration, Federal University of Rio de Janeiro, R. Visconde de Pirajá, 407/702, Rio de Janeiro — RJ — 22410-003, Brazil. Tel.: +55 21 2521 6147; fax: +55 21 2523 6839. E-mail address: [email protected] (A.E. Nardi). 0165-1781/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.psychres.2008.06.025

cases in which the DSM-III classificatory distinctions are blurred, and the course of the illness seems to involve temporary overlapping manifestations of panic attacks, social anxiety disorder and agoraphobia. The association between lifetime anxiety disorders and personality disorders among adults in the community was explored by Goodwin and Hamilton (2003). Their data were drawn from the National Comorbidity Survey (n = 5877), a representative community sample of adults aged 15–54 in the 48 US states. Out of the 3.3% of adults with antisocial personality disorder (ASPD), over half (54.33%) had a comorbid anxiety disorder (lifetime). Similarly, 42.31% of adults with a history of conduct disorder (CD) (9.4%) who did not meet criteria for ASPD had a lifetime anxiety disorder. Social anxiety disorder and post-traumatic stress disorder were associated with significantly increased odds of ASPD. The comorbidity of anxiety disorders and ASPD was associated with significantly higher odds of major depression, substance use disorders, and suicide ideation and attempt compared with odds among those without both disorders. These data provide initial evidence of an association between posttraumatic stress disorder and social anxiety disorder, and an increased likelihood of ASPD among adults in the community, after adjustment for comorbid affective and substance use disorders. Adults with ASPD and comorbid anxiety had significantly higher levels of comorbid

A.E. Nardi et al. / Psychiatry Research 175 (2010) 260-265

major depression, alcohol dependence, and substance dependence and substantially higher rates of lifetime suicidal ideation and suicide attempts compared with adults with ASPD or anxiety disorders alone or with neither disorder. Over the last two decades the combination of selective serotonin reuptake inhibitors (SSRIs) with cognitive-behavioural therapy (CBT) has been considered an efficacious option for the treatment of anxiety disorders (Roy-Byrne et al., 2006). The clear efficacy and increased tolerability of the SSRIs over previous treatments (tricyclics and monoamine oxidase inhibitors — MAOI) have led to the widespread conclusion of their being the first-line choice in all these disorders (Roy-Byrne et al., 2006). In a clinical setting the treatment of choice depends on a number of issues, including adverse effects, efficacy, and the presence of concomitant syndromes. It is crucial to provide a riskbenefit assessment for each patient, and the treatment choice should be based on individualized assessment. From this point of view, there is still a place for older drugs such as the MAOIs in certain cases, and they should be studied in special populations. Arnot (1960) was the first to describe the anxiolytic effect of a MAOI. Sheehan et al. (1980) showed that phenelzine was superior to imipramine in relation to several panic disorder measures, including reduction of disability, and this finding renewed interest in this class of drug for treating panic disorder. Since then, MAOIs have been shown to be efficacious in social anxiety disorder (Deltito and Perugi, 1986; Liebowitz et al., 1986; Versiani et al., 1988; Liebowitz et al., 1988; Versiani et al., 1992) and panic disorder or comparable disorders (Sargant, 1969; Tyrer et al., 1973; Sheehan et al., 1980; Versiani et al., 1987; Tyrer and Shawcross, 1988). The first double-blind study with patients with social anxiety disorder (Liebowitz et al., 1988) compared phenezine and atenolol with placebo in the treatment of 74 patients with social anxiety disorder. The average dose was 76 mg/day for phenelzine. At the end of the 16-week treatment period, about 2/3 of the patients treated with phenelzine showed a significant improvement, whereas less than 1/3 improved with atenolol and placebo. Although MAOIs are an effective antipanic agent (Sargant, 1969; Sheehan et al., 1980; Versiani et al., 1987; Tyrer and Shawcross, 1988), they are generally used as second line agents due to their associated adverse reactions and dietary restrictions (Jann and Kurtz, 1987; Bechelli et al., 1989). A disadvantage of the conventional irreversible MAOIs is the risk of hypertensive crises when combined with dietary tyramine (Tyrer and Shawcross, 1988; Bechelli et al., 1989). There are no important therapeutic differences between the MAOIs, and there is growing evidence that MAOIs are somewhat more effective than tricyclic antidepressants in the treatment of anxiety disorders and when phobic anxiety is an important component of a depressive disorder (Jann and Kurtz, 1987; Tyrer and Shawcross, 1988). Patients that present the characteristic features of both social anxiety disorder and panic disorder (Liebowitz et al., 1985) might benefit most from a treatment suited for both disorders. This study was designed to explore the dose–response relationship for tranylcypromine in patients with panic disorder (with and without agoraphobia) with social anxiety disorder comorbidity — DSM-IV. Our hypothesis is that a higher dose of tranylcypromine may be necessary to the treatment of social anxiety disorder symptoms than for panic disorder symptoms considering not only the phobic avoidance but the panic attack and the anticipatory anxiety. This different response to a lower and a higher dose may be associated with the dopaminergic system in social anxiety disorder (Ressler and Nemeroff, 2000).

261

for panic disorder, with or without agoraphobia, and social anxiety disorder generalized subtype, both determined by the Structured Clinical Interview (First et al., 1997) for DSM-IV. Patients needed to have a minimum of four panic attacks, at least one of which was unanticipated, during the 4 weeks before the initiation of the non-drug washout period, and at least three panic attacks during the 1-week non-drug washout before the open treatment. At the end of the washout, patients needed to score at least 18 on the Hamilton Anxiety Rating Scale (HAMA; Hamilton, 1959) and below 17 on the 21-item Hamilton Depression Rating Scale (HAMD) (Hamilton, 1960). Patients had to have a negative result for a urinary screen for benzodiazepines and barbiturates on baseline. Women of childbearing potential had to be using an effective method of birth control. Pregnant or nursing women were excluded from participation. Patients who met DSM-IV criteria for current major depression, bipolar disorder, obsessivecompulsive disorder, schizophrenia, delusional or psychotic disorders, organic brain syndrome, severe personality disorder, epilepsy, or substance abuse or dependence (during the previous year) were also excluded. Patients with comorbid dysthymia, generalized anxiety disorder or past major depression could be included if the panic disorder and social anxiety disorder were judged to be the principal diagnosis. Other reasons for exclusion included unstable medical conditions; hypersensitivity or other medical contraindications to MAOI therapy; participation in an investigational drug study within 6 months of study entry; previous treatment with tranylcypromine; concomitant treatment with any psychotropic drug or psychotherapy during the study; use of any regular antipsychotic, antidepressant, regular benzodiazepine or nonbenzodiazepine anxiolytic medication within 4 weeks, or fluoxetine within 5 weeks of the first administration of study medication; or the presence of suicidal risk. After complete description of the study to the subjects, written informed consent was obtained. The protocol complying with the principles laid down in the Declaration of Helsinki was approved by our local Ethics Committee. 2.1. Drug administration In a double-blind design, patients were randomly assigned to 12 weeks of treatment with either tranylcypromine 30 mg/day or tranylcypromine 60 mg/day. They were informed about the probable side effects including the symptoms of hypertensive crisis and precautions that they should follow (diet restrictions, drug interactions, nifedipine for hypertensive crisis, etc.). Study capsules contained either 10 mg of tranylcypromine or placebo to preserve the double-blind condition after the 30 mg daily. Patients were instructed to take the capsules twice daily after breakfast and after lunch. Those randomized to tranylcypromine 30 mg/day received 20 mg/day for the first week; in the absence of dose-limiting adverse experiences, this was increased to 30 mg/day for the second week and this dose was continued for the rest of the study. The patients randomized to tranylcypromine 60 mg/day received 20 mg/day for the first week, 40 mg/day for the second week, and 60 mg/day for the third week, and this dose was continued for the rest of the study. All patients received the same number of capsules per day. The doses were not decreased in the case of limiting adverse experiences (any side effect that disturbed the patient and could not be easily managed); if adverse events occurred, the patients were dropped from the study. 2.2. Efficacy assessments This is a double-blind trial as the raters were blinded to the dose group and the patients and the medical assistant were also blinded to the dose level to which patients were assigned. The raters were blind during the analysis of the diaries and throughout the statistical analysis. Patients were seen for evaluation at baseline and at the end of weeks 1, 2, 4, 6, 8, and 12.

2. Methods

2.2.1. Primary instruments All independent raters were trained in the instruments used in this trial in our Laboratory. Before the trial began and twice during it, we had theoretical and practical sessions of training to standardize our recordings. The main efficacy instrument for panic disorder symptoms was the Sheehan Panic and Anticipatory Anxiety Scale (Sheehan, 1983), from which the principal efficacy measure – the percentage of patients with no panic attacks – was obtained. Throughout the study, participants maintained a daily diary, in which the frequency of panic and limited symptom panic, panic attacks, and duration of anticipatory anxiety were recorded by the patient. These data were subsequently reviewed by the investigators and used to complete the Panic and Anticipatory Anxiety Scale scores. Baseline Panic and Anticipatory Anxiety Scale scores were based on the 1-week non-drug washout period. Anticipatory anxiety was recorded as the percentage of time in each 24 h spent worrying about having a panic attack. The outcome measures for the social anxiety symptoms were the mean change from baseline in the Liebowitz Social Anxiety Scale (LSAS) total score (Liebowitz et al., 1988), and the patient-rated visual analogue scale scores reflecting self-confidence in social interactions, anticipatory anxiety, acute anxiety reactions in social situations, and dysphoria following anxiety reactions.

The patients were randomly selected at the Laboratory of Panic & Respiration of the Institute of Psychiatry of the Federal University of Rio de Janeiro. Our clinic receives many anxiety disorder patients, and we had a reasonable number of newly admitted panic disorder and social anxiety disorder comorbid patients to choose for this trial. Participants were men and women, aged between 18 and 60, who met DSM-IV criteria

2.2.2. Secondary instruments Clinicians rated global severity by means of the Clinical Global Impression Severity Scale (Guy, 1976) (CGI-S, ranging from 1 — not at all ill to 7 — extremely ill). Global change from the baseline assessment was rated by means of the Clinical Global Impression Improvement Scale (Guy, 1976) (CGI-I, ranging from 1 — very much

262

A.E. Nardi et al. / Psychiatry Research 175 (2010) 260-265

improved to 7 — very much worse). The CGI referred to the overall evaluation of both panic disorder and social anxiety disorder. The frequency of full panic attacks (more than 4 symptoms) and with the CGI-S score were examined both as continuous variables and as indicators of panic-free status and of efficient end-state functioning. Panic-free status at the end-point was defined as 2 weeks with no full panic attacks, and efficient end-state functioning was defined as a CGI-S score of 1 or 2 occurring at the end-point in conjunction with panic-free status. Clinicians also rated the overall severity of the panic disorder by means of the seven-item Panic Disorder Severity Scale (Shear et al., 1997), assessing frequency of attacks, degree of distress during attacks, anticipatory anxiety, phobic avoidance of situations, phobic avoidance of sensations, impairment or interference with work functioning, and impairment or interference with social functioning. Severity of anxiety was assessed with the 14-item Hamilton Anxiety Rating Scale (HAMA; Hamilton, 1959). All patients provided Patient Global Evaluation ratings of improvement at every visit after baseline (ranging from 1 — very much improved to 7 — very much worse). Patients also completed the Quality of Life Enjoyment and Satisfaction Questionnaire (Endicott et al., 1993), rating 16 aspects of quality of life, including physical health, mood, activities of daily living, and overall life satisfaction. The quality of life scale was completed by the patient at baseline and at the end of week 12 of the treatment period (or at study discontinuation). 2.3. Safety assessments A physical examination was performed on day 1 of washout and at the end of the 12-week period (or at study discontinuation). Blood pressure and heart rate were measured at every visit. A 12-lead electrocardiogram and laboratory test (i.e., hematology, chemistry, urinalysis, thyroid hormones, and pregnancy test) were obtained on day 1 of the washout and at the end of week 12 or at study discontinuation. A urine drug screen for benzodiazepines and drugs of abuse was done at baseline and at every evaluation. Adverse events were recorded either after inquiry to the patient, observation in the clinical exam, or a spontaneous report by the patient (including onset duration, severity, cause in the investigator's judgment, action taken, and outcome). All adverse events were tabulated regardless of their assessed severity or relationship to study drug. Multiple episodes of the same complaint were counted only once, although rated at the greatest level of severity. 2.4. Statistical analyses Patients who took at least one dose of the medication and completed any additional assessment were included in the analysis of safety and efficacy. Age, weight, HAMD, duration of illness, and baseline Sheehan Panic and Anticipatory Anxiety Scale and HAMA were compared among treatment groups at baseline using analysis of variance (ANOVA) with terms for treatment group. Gender differences were compared using χ2 tests. Parallel analyses of efficacy parameters were performed weekly — including only those data actually collected and at end-point with the last observation carried forward. The Sheehan Panic and Anticipatory Anxiety Scale measures (number of panic attacks, number of episodes of anticipatory anxiety, and percentage of time spent having anticipatory anxiety) were not normally distributed, owing to marked between-patient variability in the frequency and intensity of attacks; therefore, values were expressed as a ratio to individual baseline values, and were log-transformed prior to analysis. Although these ratios are adjusted for baseline severity, their means can be distorted by very large values for which reason the geometric mean was also used to estimate central tendencies for the four treatment groups; this parameter is more appropriate for log-transformed data than the arithmetic mean is. Analyses of panic attack were performed using analysis of covariance (ANCOVA) models with log of the baseline as the co-variate. Pair-wise comparisons among the treatment groups were performed at end-point using Fisher's protected least significant difference method. The Kruskal– Wallis test was applied to confirm analyses of the three main efficacy parameters. Changes in HAMA and HAMD scores from baseline were compared between treatments by ANCOVA. Scores for CGI global improvement were directly compared between treatment groups by ANOVA, and pair-wise comparisons at end-point were performed using Fisher's protected least significant difference method. The non-parametric Wilcoxon rank-sum (Mann–Whitney) test was used for all social anxiety efficacy variables. All p values are two-tailed, and statistical significance was set at 5% level (p b 0.05).

3. Results

60-mg group were 10 women (mean age ± SD= 35.9 years ± 10.5) and 7 men (mean age ± SD = 34.8 years ± 9.1). The age difference between the groups was not statistically significant (ANOVA, F = 0.01, df = 1, P = 0.857). 3.1. Efficacy Both doses – 30 and 60 mg – were equally effective, with no statistical difference between them at the end-point in reducing the number of weekly panic attacks (Table 1). The main efficacy measure, percentage of patients with no panic attacks, was observed in 12 patients (70.6%) of the higher dose group and in 14 patients (68.4%) of the lower dose group. At the end of the study – week 12 – more than 68% in each tranylcypromine group were free of panic attacks (Table 1). The percentage of subjects reporting zero panic attacks weekly is shown in Table 1 and there is no difference between the two groups over the course of the trial. Overall, ANCOVA demonstrated a significant treatment effect of tranylcypromine (F = 5.22, df = 1, P b 0.01), but there were no significant differences between the 30mg and the 60-mg groups. Our report represents intent-to-treat samples with the last observation carried forward. Including the eight patients who dropped out due to inefficacy or side effects, at the end-point, 11 (57.9%) tranylcypromine 30-mg patients and 11 (64.7%) tranylcypromine 60-mg patients were considered moderately or completely recovered (CGI severity = 1 or 2). Both of the tranylcypromine-treated groups were significantly different from baseline data (χ2 = 6.47, df = 2, P = 0.023), but there was no difference between them (χ2 = 0.546, df = 2, P = 0.324). An estimate of effect size is established by “relative rate”, the rate of panic attacks at end-point in the 30-mg group and the 60-mg group relative to the rate of panic attacks at baseline (after 1-week washout). The tranylcypromine, 30 mg daily, group had a relative rate of 0.65, confidence interval (CI) 0.38–0.92; P = 0.435; for the tranylcypromine, 60 mg daily, group the value was 0.70, CI 0.35–0.95, P = 0.056. No significant differences were noted between the two dose groups. The number of episodes of anticipatory anxiety at end-point was also reduced significantly in subjects taking the drugs, as shown in Table 2 (F = 3.34, df = 1, P b 0.001). The baseline values (mean± SD) for the 30-mg group in the LSAS were 65.4 ± 18.3 for total score, 34.9 ± 19.5 for the fear score, and 30.5 ± 21.8 for the avoidance score. The baseline values (mean± SD) for the 60mg group in the LSAS were 67.2 ± 23.9 for total score, 35.3 ± 24.2 for the fear score, and 32.3 ± 19.5 for the avoidance scores. All scores for the LSAS and self-rated visual analogue scales for social anxiety symptoms showed differences between the treatments (Table 3). Both scales showed the same pattern of response. The difference in the total LSAS score was significant for the 60-mg daily group. Two subjects in the 30mg daily group and one in the 60-mg daily group reported worsening of social anxiety and avoidance symptoms. All the secondary efficacy variables demonstrated a therapeutic response in both groups at end-point (Table 3). We observed a

Table 1 Percentage of subjects reporting zero panic attacks, by week. Week

We randomly selected 51 patients and only 36 patients fulfilled the criteria for entering the trial. All patients met the diagnostic criteria for either panic disorder with or without agoraphobia and social anxiety disorder — DSM-IV. No patient dropped out during the 1week no-drug washout period. The patients that entered the open phase were 21 women and 15 men with a mean (±SD) age of 36.4 (±9.8) years. In the tranylcypromine 30-mg group were 11 women (mean age ± SD = 37.9 ± 6.5 years) and 8 men (mean age ± SD = 41.2 ± 12.6 years); in the tranylcypromine

1 2 4 6 8 12 End-point

Tranylcypromine

Tranylcypromine

% (n = 19)

% (n = 17)

30 mg/day

60 mg/day

15.8 (3/19) 22.2 (4/18) 50.0 (7/14) 57.1 (8/14) 71.4 (10/14) 92.8 (13/14) 68.4 (13/19)

11.8 (2/17) 26.7 (4/15) 57.1 (8/14) 71.4 (10/14) 78.6 (11/14) 85.7 (12/14) 70.6 (12/17)

P (χ2)

0.769 0.647 0.471 0.318 0.792 0.735 0.916

A.E. Nardi et al. / Psychiatry Research 175 (2010) 260-265

significant difference between the tranylcypromine doses in scales that evaluate both panic symptoms and social anxiety symptoms such as the CGI, PGI, and the quality of life score. Scales that evaluate panic symptoms such as the Panic Disorder Severity Scale and scales that evaluate depressive or global anxiety symptoms such as the HAMD and the HAMA, respectively, showed no significant differences. The respective HAMD scores at baseline were 12.5 ± 3.2 for the lower dose group and 11.4 ± 3.8 for the higher dose group (ANCOVA, F = 0.23, df = 1, P = 0.639). 3.2. Adverse events From the 36 subjects included in the trial, eight (22.2%) withdrew from the study, five (13.9%) due to adverse events and three (8.3%) due to insufficient clinical response. Discontinuations for tranylcypromine for the tranylcypromine 30-mg group and the 60-mg group were 26.3% (n = 5) and 17.6% (n = 3), respectively. The difference in dropouts between the groups was not statistically significant χ2 = 0.733, df = 1, P = 0.645). In the 30-mg group, two patients (10.5%) discontinued due to adverse experience and three (15.8%) due to insufficient clinical response. In the 60-mg group, the reasons for dropping out were all due to adverse experiences (3–17.6%) (p = 0.458 when comparing the overall discontinuation rate between the groups — Fisher's exact test). The adverse events were usually mild with a few of moderate intensity. No severe adverse event occurred during the study. The side effects mostly tapered off after the first 4 weeks of treatment. The dropout rate for adverse events was small in both groups. This difference was not statistically significant (p = 0.533, Fisher's exact test). All the subjects who dropped out due to adverse events did so in the first 4 weeks. A total of 14 (73.7%) subjects in the 30-mg tranylcypromine group experienced at least one adverse event compared with 15 (88.2%) in the 60-mg group, a statistically significant difference statistically significant (χ2 = 25.13, df = 1, Pb 0.001). The most common adverse events recorded during the study (N10%) were in the 30-mg group (%): orthostatic dizziness (47.3%), decreased libido (42.1%), insomnia (31.6%), headache (31.6%), constipation (21.0%), somnolence (21.0%), nausea (15.8%), dizziness (15.84%) and dry mouth (10.5%). In the 60mg group, the most common adverse events were: orthostatic dizziness (58.8%), insomnia (47.1%), decreased libido (41.2%), somnolence (35.3%), headache (23.5%), dizziness (23.5%), nausea (17.6%), dry mouth (17.6%), constipation (11.8%), fatigue (11.8%), and ejaculatory delay [7 men] (28.5%). The severity did not differ between

Table 2 Panic and Anticipatory Anxiety Scale mean, S.D. and maximum values at baseline and end-point (last observation carried forward, LOCF) and week 12 (subjects who completed study only). Tranylcypromine 30 mg/day

Tranylcypromine 60 mg/day

N

Max.

N

Mean

S.D.

Max.

87 23 2

17 17 14

25.2 7.3 2.9

18.3 8.1 2.6

78 25 4

76 18 3

17 17 14

21.0 4.6 2.5

17.3 5.4 1.5

67 15 4

71 26 4

17 17 14

22.7 5.6 5.6

12.5 6.6 4.5

64 38 8

95 23 8

17 17 14

25.2 7.4 3.8

18.2 3.6 3.1

90 28 10

Mean

S.D.

Panic attacks Baseline 19 32.9 22.5 End-point (LOCF) 19 5.2 12.4 Week 12 14 1.6 1.1 Unexpected attacks Baseline 19 20.5 17.9 End-point (LOCF) 19 7.6 5.6 Week 12 14 2.2 2.0 Limited symptom attacks Baseline 19 31.3 25.7 End-point (LOCF) 19 9.2 10.5 Week 12 14 2.2 1.5 Number of episodes of anticipatory anxiety Baseline 19 29.5 20.5 End-point (LOCF) 19 7.4 5.6 Week 12 14 4.9 3.8 The minimum for all variables is zero.

263

Table 3 Summary of efficacy variables at end-point (mean ± SD). Efficacy variable

Tranylcypromine Tranylcypromine F (n = 19) (n = 17) 30 mg/day

CGI-S (change from baseline) −2.34 ± 1.71 CGI-I (rating at end-point) 2.27 ± 1.43 Patient Global Evaluation 2.45 ± 1.11 (rating at end-point) Panic Disorder Severity Scale 1.76 ± 0.86 (change from baseline) HAMD score (change from − 2.48 ± 2.24 baseline) HAMA score (change from − 13.42 ± 6.99 baseline) LSAS (change from baseline) − 17.9 ± 14.7 Total scorea Fear − 8.7 ± 11.4 Avoidance − 11.5 ± 8.2 Self-rated visual analogue scales (change from baseline) Self-esteem − 1.4 ± 1.2 Anticipatory anxiety − 0.3 ± 1.9 Anxiety reactions − 0.5 ± 2.0 Dysphoric reactions − 0.8 ± 3.1 Quality of life score (change from baseline) Total 2.06 ± 3.96 Overall satisfaction 0.38 ± 0.91

P

60 mg/day − 3.19 ± 1.35 1.52 ± 0.74 2.76 ± 1.63

15.851 0.002 13.990 0.003 12.579 0.004

1.45 ± 1.24

2.451

0.126

− 2.92 ± 2.33

2.237

0.273

− 10.65 ± 8.04

1.345

0.341

−35.0 ± 14.8 − 16.1 ± 12.6 −14.7 ± 13.6

− 2.5 ± 2.4 −3.7 ± 3.0 − 3.8 ± 4.2 − 3.1 ± 3.5

5.45 ± 3.38 1.68 ± 0.58

13.245 0.003 8.734 0.002 7.976 0.001

9.674 5.634 6.005 2.341

0.003 0.017 0.005 0.036

18.645 0.001 16.245 0.003

CGI-S indicates Clinical Global Impression Severity Scale; CGI-I, Clinical Global Impression Improvement Scale; HAMD, Hamilton Depression Rating Scale; and HAMA, Hamilton Anxiety Rating Scale. LSAS — Liebowitz Social Anxiety Scale. a Fisher's protected least significant difference (30 mg vs. 60 mg, F = 25.06, df = 2, P = 0.0003).

tranylcypromine groups. A dose–response relationship was noted only for the incidence of orthostatic dizziness and insomnia. No statistically significant differences were found in the incidence of clinically significant laboratory abnormalities between the groups. No hepatic dysfunction was observed at the end of the trial in either of the two groups. No clinically significant ECG abnormalities were noted. 4. Discussion Although very important clinically, the psychopharmacological treatment of anxiety disorders comorbid with other conditions has seldom been very rarely studied with a systematic methodology (Deltito and Perugi, 1986; Jann and Kurtz, 1987; Perugi et al., 1999). This is the first time that a sample of comorbid panic disorder and social anxiety disorder patients participated in a trial. In common with pure panic disorder samples in other studies (Versiani et al., 1987; Tyrer and Shawcross, 1988; Shear et al., 1997), the majority of patients were middle-aged women. Tranylcypromine has not yet been compared with placebo in a published trial in panic disorder. For social anxiety disorder, there exists only one open trial (Versiani et al., 1988). There was one poster by Versiani et al. (1987) that compared tranylcypromine, alprazolam, imipramine, and placebo in panic disorder. In this study (Versiani et al., 1987), the active drugs were clearly more efficacious than placebo. This result justifies our decision to conduct a study without a placebo control. The results of our trial indicate that 60 mg daily of tranylcypromine may be an effective and satisfactorily tolerated treatment for panic disorder with social anxiety disorder comorbidity. At the end-point, the patients exhibited significantly greater reductions in panic attack frequency and global severity of illness than a previous study comparing tranylcypromine with placebo in panic disorder (Versiani et al., 1987). The patients treated with tranylcypromine, 60 mg daily,

264

A.E. Nardi et al. / Psychiatry Research 175 (2010) 260-265

showed a substantial reduction in social and avoidance anxiety, consequently being able to better perform their social and professional activities. We could question the power to detect differences between the two active treatment groups. In order to increase the anticipated power, we used a double-blind design with independent raters, different scales, a treatment period of 12 weeks, and a sample carefully selected by the inclusion and exclusion criteria. We observed similar results at both doses in panic disorder, but a significant difference in social anxiety disorder. This is very rare in clinical psychopharmacological trials with two active-drug groups but shows the power to detect differences in our trial. Although there was a significant reduction in the number of full and limited panic attacks, and in the level of social anxiety symptoms, and although some patients achieved a panic-free status with 60 mg daily, this does not means that tranylcypromine should be used as a first-line drug for panic patients with social anxiety disorder comorbidity. MAOIs can cause distressing and dangerous adverse events (Tyrer et al., 1973; Tyrer and Shawcross, 1988; Bechelli et al., 1989; Quitkin et al., 1990). In our trial, most of the side effects were mild to moderate, but seven patients discontinued tranylcypromine due to adverse effects. Initiation of the treatment with lower doses than the ones used in this trial – 20 mg/day – can be a useful strategy to minimize early treatment discontinuation due to adverse effects. Perhaps other antidepressants with a safer profile should also be studied in the case of anxiety disorder comorbidities. The fixed-dose design used in this study permits assessment of the comparative efficacy of different dose levels of tranylcypromine and a specific pattern has been observed with the higher dose being more efficacious in our comorbid sample for social anxiety disorder symptoms. Although depression is a common comorbidity in anxiety disorder patients (Stein et al., 1990), the present study excluded patients with current major depression or marked depressive symptoms, ensuring that the patients' improvement was not the result of amelioration of depression. It was confirmed by a small change and no statistical difference in the HAMD scores among the groups. The goals of future research will be to broaden and refine our understanding of the psychological, neuroendocrine, and genetic factors that might contribute to higher rates of comorbidity in patients with panic disorder. Many double-blind, placebo-controlled trials in panic disorder have found a high placebo response (Rapaport et al., 2000), and their data are similar data in our 30-mg group. A precise diagnosis excluding comorbidities, such as personality disorders and drug abuse, can strengthen our results in the absence of placebo group. The strongest part of the study is that the social anxiety symptom scale showed a two-fold greater change with the 60-mg dose, and the 30mg dose was ineffective; the 30-mg group could be considered the equivalent of a placebo control group. The positive response in panic disorder symptoms may also be meaningful, and is in agreement with several prior studies (Rapaport et al., 2000). Our results have strength as they are statistically and clinically relevant, and they provide encouragement to carry out new trials. Panic disorder has been the topic of much research dealing with the neurochemistry of anxiety disorders. The γ-aminobutyric acid (GABA) and the monoamine neurotransmitters, norepinephrine (NE) and serotonin [5-hydroxy-tryptamine (5-HT)], have been the focus in the neurochemistry of panic disorder (Gorman et al., 2000). These neurotransmitters are closed linked to the discovery of drugs that can treat or worsen PD symptoms by altering the function of the receptors upon which these neurotransmitters naturally bind (Ressler and Nemeroff, 2000). Whilst noradrenergic, serotonergic, and GABAergic systems have remained the focus of attention, other transmitters and modulators (e.g. dopamine, cholecystokinin) are becoming more prominent, while researchers also try to assess the importance of the balance between these systems (Ressler and

Nemeroff, 2000). The involvement of dopamine (DA) in anxiety disorders has not been as well established (Nutt and Lawson, 1992; Gorman et al., 2000), but it may be related to the biochemical difference between panic disorder and social anxiety disorder. DA is catabolized by MAO and COMT, which may explain in some way the efficacy of the MAOIs. The importance of DA in social anxiety disorder can also be observed in the poor response to tricyclic antidepressants (Versiani et al., 1988; Simpson et al., 1998). The increased prevalence of social anxiety disorder in patients who later developed Parkinson's disease, as well as the exacerbation of social anxiety disorder in response to dopamine blockers and the clinical observation of increased anxiety in patients on antipsychotic drugs, can be interpreted as indirect evidence for an association between dopaminergic dysfunction and social anxiety disorder (Liebowitz et al., 1985; Liebowitz et al., 1986). Taken together with trials showing the efficacy of MAOIs in social anxiety disorder, these data suggest that the dopamine system may well play an important role in mediating the symptoms of social anxiety disorder and be responsible in our trial with an MAOI for the difference in response between the symptoms of panic disorder and the symptoms of social anxiety disorder. Follow-up studies would be useful to examine the ability of tranylcypromine to maintain treatment benefits over time. In an open trial, Versiani et al. (1988) treated social anxiety disorder patients with tranylcypromine for 1 year and noted maintenance of response. There are no long-term data with tranylcypromine in panic disorder. Tranylcypromine, due to its unfavorable side-effect profile, is only an option for some refractory patients. Many past published studies of MAOIs have yielded poor results because the drugs have been prescribed for an insufficient time (less than 4 weeks) or at too low dosage (Liebowitz et al., 1988; Tyrer and Shawcross, 1988). Balon et al. (1999) made a survey of prescribing practices for MAOIs and explored reasons for the widely noted decline in their use. A questionnaire was sent in 1997 to 1129 members of the Michigan Psychiatric Association. A total of 717 responses were received, for a response rate of 64%. Only data from the 573 psychiatrists who were currently practicing were used. Twelve percent of the respondents had never prescribed MAOIs, 27% had not prescribed them for at least 3 years, and 17% had prescribed them from 1 to 3 years ago. The most frequent reasons for not prescribing the drugs were side effects and interactions with other medications (46%), preference for other medications (30%), and dietary restrictions necessary for patients taking MAOIs (19%). Fifty-six percent of responders believed that MAOIs were useful for panic disorder and 44% for social anxiety disorder. However, only 3% would use them as a first-line treatment of social anxiety disorder. The results document the commonly held view that psychiatrists believe MAOIs are efficacious but use them infrequently, primarily due to concerns about side effects and drug interactions. Our results need to be interpreted cautiously; although we included in an intention-to-treat analysis, our completion rate was less than 80%. Future studies might confirm our observation that although panic symptoms disappear at a low dose – 30 mg daily, tranylcypromine is only efficacious in comorbid cases with symptoms of both panic disorder and social anxiety disorder at a higher dose level – 60 mg daily. Perhaps, other trials could also observe the same pharmacological pattern of response with other drugs. Acknowledgments This research was supported by the Brazilian Council for Scientific and Technological Development (CNPq). Grant 554411/2005-9.

References Arnot, R., 1960. Calming effects of the monoamine oxidase inhibitors. Diseases of the Nervous System 11, 448. Balon, R., Mufti, R., Arfken, C.L., 1999. A survey of prescribing practices for monoamine oxidase inhibitors. Psychiatric Services 50, 945–947.

A.E. Nardi et al. / Psychiatry Research 175 (2010) 260-265 Bechelli, L.P., Nardi, A.E., Alves, A.B., 1989. Review of the use of the irreversible monoamine oxidase inhibitors (MAOIs) in psychiatry. Current Medical Research Opinions 11 (S1), 38–52. Deltito, J.A., Perugi, G., 1986. A case of social phobia with avoidant personality disorder treated with MAOI. Comprehensive Psychiatry 27, 255–258. Endicott, J., Nee, J., Harrison, W., Blumenthal, R., 1993. Quality of Life Enjoyment and Satisfation Questionnaire: a new measure. Psychopharmacology Bulletin 29, 321–326. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.M., 1997. Structured Clinical Interview Diagnostic (SCID) for DSM-IV Axis I Disorders — Clinician Version (SCID-CV). American Psychiatric Press, Washington, DC. Frances, A., Widiger, T., Fyer, M.R., 1990. The influence of classification methods on comorbidity. In: Maser, J.D., Cloninger, C.R. (Eds.), Comorbidity of Mood and Anxiety Disorders. American Psychiatric Press, Washington, DC, pp. 41–60. Goldstein, S., 1987. Three cases of overlap between panic disorder, social phobia, and agoraphobia. Journal of Clinical Psychiatry 48, 452–453. Goodwin, R.D., Hamilton, S.P., 2003. Lifetime comorbidity of antisocial personality disorder and anxiety disorders among adults in the community. Psychiatry Research 117, 159–166. Gorman, J.M., Kent, J.M., Sullivan, G.M., Coplan, J.D., 2000. Neuroanatomical hypothesis of panic disorder, revised. American Journal of Psychiatry 157, 493–505. Guy, W. (Ed.), 1976. ECDEU Assessment Manual for Psychopharmacology, Revised. DHEW Publication. Superintendent of Documents, Government Printing Office, Washington, DC. Hamilton, M., 1959. The assessment of anxiety states by rating. British Journal of Medical Psychology 32, 50–55. Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery, and Psychiatry 23, 56–62. Jann, M.W., Kurtz, N.M., 1987. Treatment of panic and phobic disorders. Clinical Pharmacology 6, 947–962. Liebowitz, M.R., Fyer, A.J., Gorman, J.M., Campeas, R., Levin, A., 1986. Phenelzine in social phobia. Journal of Clinical Psychopharmacology 6, 93–98. Liebowitz, M.R., Gorman, J.M., Fyer, A.J., Campeas, R., Levin, A.P., Sandberg, D., Hollander, E., Papp, L., Goetz, D., 1988. Pharmacotherapy of social phobia: an interim report of a placebo-controlled comparison of phenelzine and atenolol. Journal of Clinical Psychiatry 49, 242–257. Liebowitz, M.R., Gorman, J.M., Fyer, A.J., Klein, D.F., 1985. Social phobia. Review of a neglected anxiety disorder. Archives of General Psychiatry 42, 729–736. Nutt, D.J., Lawson, C., 1992. Panic attacks: a neurochemical overview of models and mechanisms. British Journal of Psychiatry 160, 390–396.

265

Perugi, G., Nassini, S., Socci, C., Lenzi, M., Toni, C., Simonini, E., Akiskal, H.S., 1999. Avoidant personality in social phobia and panic-agoraphobic disorder: a comparison. Journal of Affective Disorders 54, 277–282. Quitkin, F.M., McGrath, P.J., Stewart, J.W., Harrison, W., Tricamo, E., Wager, S.G., OcepekWelikson, K., Nunes, E., Rabkin, J.G., Klein, D.F., 1990. Atypical depression, panic attacks, and response to imipramine and phenelzine: a replication. Archives of General Psychiatry 47, 935–941. Rapaport, M.H., Pollack, M., Wolkow, R., Mardekian, J., Clary, C., 2000. Is placebo response the same as drug response in panic disorder. American Journal of Psychiatry 157, 1014–1016. Ressler, K.J., Nemeroff, C.B., 2000. Role of serotonergic and noradrenergic systems in the pathophysiology of depression and anxiety disorders. Depression and Anxiety 12 (suppl. 1), 2–19. Roy-Byrne, P.P., Craske, M.G., Stein, M.B., 2006. Panic disorder. Lancet 368, 023–1032. Sargant, W., 1969. Treatment of the phobic anxiety state. British Medical Journal 3, 49. Shear, M.K., Brown, T.A., Barlow, D.H., Money, R., Sholomskas, D., Woods, S.W., Gorman, J.M., Papp, L.A., 1997. Multicenter Collaborative Panic Disorder Severity Scale. American Journal of Psychiatry 154, 1571–1575. Sheehan, D.V., 1983. The Anxiety Disease. Charles Scribners & Sons, New York. Sheehan, D.V., Ballenger, J., Jacobson, G., 1980. Treatment of endogenous anxiety with phobic, hysterical and hypochondrial symptoms. Archives of General Psychiatry 37, 51–59. Simpson, H.B., Schneier, F.R., Campeas, R.B., Marshall, R.D., Fallon, B.A., Davies, S., Klein, D.F., Liebowitz, M.R., 1998. Imipramine in the treatment of social phobia. Journal of Clinical Psychopharmacology 18, 132–135. Stein, M.B., Tancer, M.E., Gelernter, C.S., Vittone, B.J., Uhde, T.W., 1990. Major depression in patients with social phobia. American Journal of Psychiatry 147, 637–639. Tyrer, P., Shawcross, C., 1988. Monoamine oxidase inhibitors in anxiety disorders. Journal of Psychiatric Research 1 (22 Suppl), 87–98. Tyrer, P.J., Candy, J., Kelly, D., 1973. A study of the clinical effects of phenelzine and placebo in the treatment of phobic anxiety. Psychopharmacologia 32, 237–254. Versiani, M., Costa e Silva, J.A., Klerman, G.L., 1987. Treatment of Panic Disorder with Alprazolam, Imipramine, Tranylcypromine or Placebo. Annual Meeting of the American College of Neuropsychopharmacology, p. 276. San Juan, Puerto Rico. Versiani, M., Mundim, F.D., Nardi, A.E., Liebowitz, M.R., 1988. Tranylcypromine in social phobia. Journal of Clinical Psychopharmacology 8, 279–283. Versiani, M., Nardi, A.E., Mundim, F.D., Liebowitz, M.R., Amrein, R., 1992. social phobia: a controlled study with moclobemide and phenelzine. British Journal of Psychiatry 161, 353–360.