Evidence-based pharmacotherapy of panic disorder

International Journal of Neuropsychopharmacology (2005), 8, 473–482. Copyright f 2005 CINPdoi:10.1017/S1461145705005201 Abraham Bakker1, Anton J. L. M. van Balkom2 and Dan J. Stein3 1 National Centre for Eating Disorders, Robert Fleury Stichting, Leidschendam, The Netherlands2 Vrije Universiteit Amsterdam and Institute for Research in Extramural Medicine, Amsterdam, The Netherlands3 University of Stellenbosch, Cape Town, South Africa This paper reviews the literature on the pharmacotherapy of panic disorder, in order to address thequestions (1) what is the first-line pharmacotherapy of choice for panic disorder?, (2) for how long shouldmaintenance pharmacotherapy be continued, and (3) what is the optimal approach to the treatment-refractory patient with panic disorder. A MEDLINE search (1966–2003) was undertaken to collate random-ized controlled trials of pharmacotherapy in panic disorder. A review of the evidence indicates thatSSRIs are currently the first line agent of choice in panic disorder, and that pharmacotherapy should becontinued for at least 1 year. There has been relatively little research on the pharmacotherapy of treatment-refractory panic disorder, and this area requires future attention.
Received 16 May 2004; Reviewed 8 June 2004; Revised 14 October 2004; Accepted 17 October 2004 Key words : Agoraphobia, panic disorder, pharmacotherapy, treatment.
attack is ‘ a discrete period of intense fear or dis-comfort, in which four or more of the following Panic disorder (PD) with or without agoraphobia is symptoms develop abruptly and reach a peak within one of the most prevalent of the anxiety disorders. The 10 minutes ’. The symptoms listed are: palpitations, disorder is also accomplished by significant morbidity sweating, trembling or shaking, sensations of short- and comorbidity. Fortunately, a number of effective ness of breath or smothering, feeling of choking, chest treatments for PD are available. This chapter focuses pain or discomfort, nausea or abdominal distress, on the pharmacotherapy, assessing the evidence feeling dizzy or faint, derealization or depersonaliz- base in order to address questions about (1) the opti- ation, feeling of losing control or going crazy, fear mal first-line pharmacotherapy of PD, (2) the optimal of dying, paresthesias and chills or hot flushes.
duration of pharmacotherapy, and (3) the optimal Agoraphobia is defined as : ‘anxiety about being in approach to pharmacotherapy in the treatment- places or situations from which escape might be diffi- refractory patient. In order to ensure that all relevant cult (or embarrassing) or in which help may not be randomized controlled trials were considered, a available in the event of an unexpected or situationally MEDLINE search (1966–2003) was undertaken using the predisposed panic attack. Agoraphobic fears typically key words ‘panic’ and ‘treatment ’. In addition, recent involve characteristic clusters of situations that in- meta-analyses of PD, and treatment guidelines on PD clude being outside the home alone, being in a crowd were reviewed. We begin by briefly discussing the or standing in a line, being on a bridge, and travelling diagnosis and target symptoms of PD.
Given that panic attacks and agoraphobia occur in different anxiety disorders, a first issue in the assess-ment of panic symptoms is the accurate differentiation Panic attacks and agoraphobic avoidance are defined of PD from these other entities. DSM-IV-TR empha- in the DSM-IV-TR (APA, 2000) as follows : A panic sizes that PD is diagnosed when the attacks experi-enced are unexpected and when at least one of the Address for correspondence : Dr A. Bakker, National Centre for attacks has been followed by 1 month (or more) of Eating Disorders, Robert-Fleury Stichting, PO Box 422, 2260 AK persistent concern about having additional attacks, worry about the implications of the attacks or their Tel. : (31)704441351 Fax : (31)704441008E-mail : A.Bakker@robertfleury.nl consequences, or a significant change in behaviour related to the attacks (APA, 2000). Patients suffering during a panic attack, has the potency to provoke from social phobia or PTSD may also suffer from panic anxiety. The most important mechanism for this is attacks, but these attacks are not unexpected. They are probably the cognitive misinterpretation of these related to specific situations : a difficult social situation bodily sensations (Clark, 1986). The use of psychoac- or a situation that refers to a traumatic event.
tive substances like caffeine, cannabis or cocaine may When a patient with PD is also extremely afraid of also produce panic attacks. Withdrawal from benzo- going into public places like shopping centres, trains, diazepines or alcohol is another cause of panic symp- cinemas or other situations from which escape would toms. In all these situations, the patient should be be difficult or in which help would not be available in diagnosed not with PD, but rather with ‘anxiety dis- case of a panic attack, this person is said to have PD order due to a general medical condition with panic with agoraphobia. There is an ongoing debate about attacks/phobic symptoms’ or ‘ substance-induced the possibility that patients suffer from agoraphobia anxiety disorder with panic attacks/phobic symp- without panic attacks. In everyday clinical practice it is toms ’. In some cases, substance abuse may be denied, found very often that severe agoraphobic avoidance and drug screening may be required before an accurate without present panic attacks only developed after the occurrence of at least one unexpected panic attack.
Target symptoms in PD include not only panic at- Phobic behaviour is also common among patients tacks and agoraphobia, but also comorbid symptoms suffering from other anxiety disorders, but in these and associated impairment. PD is frequently associ- disorders it is related to the avoidance of specific ated with mood disorders, other anxiety disorders situations that relate to that anxiety disorder, e.g. get- (e.g. social anxiety disorder), and substance-related ting dirty in patients suffering from OCD, or going disorders. The clinical relevance of diagnosing PD to meetings in patients suffering from social phobia.
with or without comorbid disorders lies in the fact A second important diagnostic issue is the differ- that patients with comorbidity are more severely and entiation of PD from general medical disorders. Given chronically ill, more disabled, utilize services more the high prevalence of PD and its frequent presen- frequently and are more difficult to treat (Roy-Byrne tation in medical settings, it is important to have a high et al., 2000). Formulated in other words, ‘comorbidity ’ index of suspicion for this disorder. As many as 3.5 % points to a more severe subgroup of PD. PD is of the general population are estimated to suffer from characterized by significant distress and functional PD (Eaton et al., 1994 ; Katerndahl and Realini, 1993), impairment, and it is important these features are also and 1-month prevalence rates are 0.7–2.2 % for females targeted by treatment interventions.
and 0.4–0.8 % for males (Bijl et al., 1997; Eaton et al.,1991). The prevalence of PD in medical specialities like cardiology and otolaryngology is even much higher :15 % and higher percentages have been reported Although the pathogenesis of PD remains incom- (Chignon et al., 1993; Stein et al., 1994) and it is, pletely understood, a range of effective pharmaco- therefore, important to rule out PD in patients pres- enting with unexplained physical complaints.
treatments have been developed in the past three For agoraphobia, high prevalence rates are also decades. Pharmacological treatments mainly comprise found with a 1-month prevalence for females of 4.4 %, treatment with high-potency benzodiazepines and and for males 1.6 % (Bourdon et al., 1988). PD usually antidepressants. Other medications that have been develops during the third decade of life, with a mean suggested for the treatment of PD include b-adrener- age of onset of 28 yr (Marks, 1987). People with panic gic blocking agents (e.g. propranolol), a2-adrenergic attacks may present to primary-care practitioners or to receptor agonists (e.g. clonidine), mood stabilizers a range of medical specialists. Unnecessary special (e.g. carbamazepine, lithium), and antipsychotic investigations are frequently ordered.
agents. Propranolol has been found ineffective in con- Conversely, however, a comprehensive medical trolled trials, while most of these other agents have not history and examination may be needed to exclude the presence of physical disorders that can cause the Behavioural and cognitive therapies have also been symptoms of a panic attack (Raj and Sheehan, 1987).
found effective for the treatment of PD (Bakker et al., Well-known causes of panic-like symptoms are hy- 1999; van Balkom et al., 1997). Questions remain about perfunction of the thyroid, hypoglycaemia and pheo- how best to sequence and combine pharmacotherapy chromocytoma. In general, any clinical condition that and psychotherapy. In practice, they are frequently is associated with physical signs that can occur also combined, with the rationale that this may lead not Evidence-based pharmacotherapy of panic disorder only to symptom improvement, but also to a more Table 1. Start, mean and maximum dosage of drugs persistent recovery. Although psychodynamic psy- chotherapeutic intervention strategies have beendeveloped (Milrod et al., 1997), the efficacy of non- directive therapies has not been documented incontrolled studies. Brief psychodynamic psycho- therapy has, however, been reported to be helpful inreducing relapse rates following treatment with an antidepressant (Wiborg and Dahl, 1996).
In the treatment of PD, separate options for panic attacks and agoraphobic avoidance behaviour can be distinguished. Pharmacological and cognitive– behavioural therapy (CBT) for panic attacks diminish frequency and severity of panic attacks (Bakker et al., 1999; van Balkom et al., 1997). With pharmacological treatments, accompanying avoidance behaviour also improves, as do comorbid general anxiety symptoms, and, in case of treatment with an antidepressant, there is also reduction in comorbid depressive symptoms (van Balkom et al., 1997). Behavioural treatment strategies without cognitive therapeutic elements, e.g.
‘exposure in vivo ’, may be effective in treating agor- aphobic avoidance behaviour, but are not effective for panic attacks (van Balkom et al., 1997). In clinicalpractice, pharmacological panic management is, SSRIs, Selective serotonin re-uptake inhibitors ; TCAs, therefore, frequently combined with exposure in vivo.
tricyclic antidepressants; MAOIs, monoamine oxidase In the remainder of this paper, we focus, however, on 2000; Gorman, 1997). These agents have a slower onset High-potency benzodiazepines and antidepressants of action than the benzodiazepines, and an initial trial are the best studied pharmacotherapy options for PD.
The high-potency benzodiazepines (e.g. alprazolam MAOIs such as phenelzine have shown efficacy in and clonazepam) have been extensively researched, the treatment of PD, but are not used on a regular basis and appear to be more effective than placebo in the since patients need to be on low-tyramine diets to short-term treatment of this disorder (Beauclair et al., avoid hypertensive crises (Rosenberg, 1999). Within 1994; Jonas and Cohon, 1993). Dosage of the high- the TCAs, both imipramine and clomipramine have potency benzodiazepines in PD is higher than the been studied (Cross National Collaborative Panic usual dosages used in generalized anxiety disorder Study, Second Phase Investigators, 1992; Papp et al., (see Table 1). Low-potency benzodiazepines like dia- 1997). The currently available SSRIs citalopram, es- zepam may also only have an anti-panic effect at citalopram, fluvoxamine, fluoxetine, paroxetine, and higher doses than normally prescribed for other dis- sertraline have all been proven more effective than orders like generalized anxiety disorder. When effec- pill-placebo in reducing symptomatology in PD tive, panic and phobia symptoms improve soon after (Ballenger et al., 1998a; Black et al., 1993; Hoehn-Saric the administration of benzodiazepines (Burrows and et al., 1993; Lecrubier et al., 1997; Michelson et al., 1998; Rapaport et al., 1998; Sharp et al., 1996; Wade The majority of studies investigating medication et al., 1997). The daily dosages of these antidepressants therapy in PD have focused on treatment with anti- are similar to those used in the treatment of major depressants. Tricyclic antidepressants (TCAs), selec- tive serotonin reuptake inhibitors (SSRIs), and ir- Indeed, antidepressants that influence the sero- reversible monoamine oxidase inhibitors (MAOIs) tonergic system have consistently been shown to have been proven to be effective in PD (Bakker et al., have efficacy in the treatment of PD. In contrast, data on noradrenaline reuptake inhibitors for PD have Lecrubier et al., 1997; Otto et al., 2001; Wade et al., been less consistent. For example, in a double-blind 1997). Some data suggests that SSRIs have a more comparison between the SSRI fluvoxamine and the rapid onset of action than TCAs (Lecrubier et al., noradrenaline uptake inhibitor maprotiline, only 1997), and that SSRIs are associated with a lower fluvoxamine demonstrated good anti-panic properties drop-out rate (Bakker et al., 2002).
Side-effects that occur during the first weeks of treatment with antidepressants can easily be mis-interpreted as symptoms of a panic attack, e.g. palpi- In everyday clinical practice combinations of different tations, sweating and nausea. In patients unaware of medications are frequently used, as well as combi- the possibility that such side-effects may occur, an nations of pharmacotherapy and CBT. However, the apparent increase in the frequency and intensity of number of controlled studies that include combination panic attacks may be seen. The best way to prevent such an outcome is to provide patients with sufficient The combination of a SSRI with a benzodiazepine is information about the working mechanisms and po- particularly widely used in clinical practice. Early co- tential side-effects of antidepressants before treatment administration of (high-potency) benzodiazepines like with this agents is initiated. Early drop-out, non- alprazolam and clonazepam may prevent the initial compliance and suboptimal treatment outcome can worsening of anxiety symptoms reported during the also be enhanced with this strategy. Outcome can be first weeks of treatment with a SSRI. The number of evaluated properly only after 6 wk of treatment.
well-designed studies that have investigated thisstrategy is, however, limited. The most recent and important study was carried out by Goddard et al.
(2001). They studied double-blind, placebo-controlled A number of studies have compared different phar- co-administration of clonazepam (1.5 mg/d) with macological therapies for PD, with the majority of open-label sertraline for the first 4 wk of treatment.
these focusing on the comparison between anti- Fifty patients were randomized, and 34 completed the depressants and benzodiazepines. These studies have trial. There was no significant difference in drop- consistently found similar efficacy. In the largest of out rate between the sertraline/clonazepam and the these studies, both imipramine and alprazolam were sertraline/placebo condition (25 % vs. 38 %). The superior to placebo for most outcome measures (Cross intent-to-treat analysis found a higher percentage of National Collaborative Panic Study, Second Phase responders in the sertraline/clonazepam group at the Investigators, 1992). To date, a total of nine studies end of both weeks 1 and 3 of the trial (41 % and 63 %) in comparing imipramine with high-potency benzodiaz- comparison to the sertraline/placebo-treated subjects epines (alprazolam, clonazepam) have been published (4 % and 32 %). The authors concluded that rapid (van Balkom et al., 1995). Both classes of agent appear stabilization of PD can be achieved safely with a effective for panic and phobic symptoms. Differences sertraline/clonazepam combination, supporting the were observed in the time to response (earlier with clinical utility of this type of regimen.
benzodiazepines) and drop-out rate (lower with ben- With respect to the combination of medication and zodiazepines). Despite these differences, intent-to- psychotherapy, the distinction between benzodiaz- treat analyses revealed no significant differences in epines and antidepressants appears relevant. In a large 8-wk study Marks et al. (1993) found no differ-ences in efficacy between alprazolam and exposure, alprazolam and relaxation, placebo and exposure, Relatively few studies have compared different anti- and placebo and relaxation. However, there were depressants in PD (Bakker et al., 1999; Cassano et al., longer-lasting gains with exposure alone than with 1988; den Boer et al., 1987 ; den Boer and Westenberg, alprazolam following withdrawal of the medication.
1988; Lecrubier et al., 1997; Modigh et al., 1992; Nair In contrast, there is evidence that SSRIs+CBT appear et al., 1996; Seedat et al., 2003 ; Tiller et al., 1997; Wade et al., 1997). In most studies, different antidepressants Thus, in early work (de Beurs et al., 1995) using a have shown equal efficacy in reducing the total num- double-blind, placebo-controlled design, fluvoxamine ber of panic attacks. In particular, comparison of TCAs followed by exposure in vivo demonstrated efficacy and SSRIs has not found differences in efficacy be- superior to that of psychological panic management tween these classes of medication (Bakker et al., 1999; followed by exposure, and exposure in vivo alone.
Evidence-based pharmacotherapy of panic disorder Similarly, a study by Sharp et al. (1996) included with or without agoraphobia. Included were 52 treat- conditions with combinations of placebo with CBT ment conditions with medication (28 high-potency and of fluvoxamine with CBT and the largest and most benzodiazepines, 24 antidepressants), with 1653 consistent treatment gains were found in the fluvox- patients at pre-test and 1324 at post-test. Pre/post- amine with CBT group. Oehrberg and co-workers effect size Cohen’s d were calculated within the treat- (1995)investigatedparoxetine+standardizedcognitive ment conditions. Seven large treatment conditions therapy (CT) vs. pill-placebo+CT; paroxetine+CT were used in the main analyses, including high- was significantly more effective than placebo+CT on potency benzodiazepines and antidepressants. Both benzodiazepines and antidepressants were superior toa control condition, consisting of pill-placebo, atten-tion placebo (a non-specific conversation on a regular base with a ‘therapist ’ that does not focus on any Since comparisons within one study between different specific symptom of the psychiatric disorder) and pharmacotherapies are relatively scarce, additional waiting list for both panic attacks and agoraphobic information on the differential efficacy of different avoidance. A comparison between high-potency ben- agents on panic attacks and agoraphobic avoidance zodiazepines and antidepressants found no differ- can arguably be derived from comparisons between ences in efficacy. In this meta-analysis the combination studies. These comparisons can be performed in a of antidepressants with exposure in vivo was found to quantitative manner by means of meta-analytical be the most potent short-term treatment of PD with or without agoraphobia, especially with respect to Several meta-analyses comparing different phar- macological treatments for PD have been published Longer-term follow-up data in the studies included (Bakker et al., 1998; Boyer, 1995; van Balkom et al., in the meta-analysis of van Balkom et al. (1997) were 1997; Wilkinson et al., 1991). More recently a number reported seperately (Bakker et al., 1998). Eight studies of meta-analyses focusing on the comparison of TCAs reported on high-potency benzodiazepines and five on and SSRIs have been published (Bakker et al., 2002; antidepressants. The results were consistent with Otto et al., 2001). The most relevant results of these A more recent meta-analysis exclusively focused on The meta-analysis of Wilkinson et al. (1991) the short-term efficacy of SSRIs vs. TCAs (Bakker et al., included 19 double-blind, placebo-controlled trials of 2002). Included were 43 studies, published prior to antidepressants (n=13) and benzodiazepines (n=6) or during 1999 (34 randomized, nine open), including for patients with PD. It showed that active treatment 53 treatment conditions, 2367 patients at pre-test and had a 25 % greater success rate than placebo over 1804 at post-test. Outcome was measured by the pro- a mean duration of 14 wk. There were no statistic- portion of patients becoming panic-free, and with pre/ ally significant differences observed between anti- post-Cohen’s d effect sizes, calculated for four clinical variables: panic, agoraphobia, depression, and general The meta-analysis of Boyer (1995) reviewed 27 anxiety. The results are summarized in Table 2, published or presented placebo-controlled, double- and indicated no differences between SSRIs and TCAs blind studies of PD. The serotonin reuptake inhibitors on any of the effect sizes, with both groups of anti- included clomipramine, fluvoxamine, paroxetine, and depressants equally effective in reducing panic symp- zimelidine. The comparison treatments were imipra- toms, agoraphobic avoidance, depressive symptoms mine and alprazolam. All three treatments were sig- and general anxiety. Also the percentage of patients nificantly superior to placebo in alleviating panic. The free of panic attacks at post-test did not differ across serotonin reuptake inhibitors were also significantly treatments. As mentioned earlier, the number of drop- superior to both imipramine and alprazolam. The outs was significantly lower in the group of patients superiority of the serotonin reuptake inhibitors treated with SSRIs (18 %) vs. TCAs (31 %). The main remained, but was less pronounced, when they were conclusion was that SSRIs and TCAs have equal effi- compared to the studies which used higher doses of cacy in the treatment of PD, but SSRIs are better The meta-analysis of van Balkom et al. (1997) It can be concluded from these data that treatment evaluated the short-term efficacy of benzodiazepines, with antidepressants may result in complete remission antidepressants, psychological panic management, of both panic attacks and agoraphobic avoidance.
exposure in vivo, and combination treatments in PD Addition of exposure in vivo may help to overcome Table 2. Characteristics of treatment conditions and effect-sizes at post-test d, Cohen’s d effect size; S.D., standard deviation.
* Number of treatment conditions providing data with respect to this item.
# x2=32.8, d.f.=1, p<0.001 agoraphobic avoidance that does not respond to of antidepressants that are used most frequently, monotherapy with medications (van Balkom et al., SSRIs and TCAs (Bakker et al., 2002; Otto et al., 2001).
However, there are indications that the drop-out ratediffers significantly in favour of the SSRIs (Bakker etal., 2002). Drop-outs may occur for different reasons, varying from ineffectiveness to serious side-effects.
There are two categories of medications with sufficient The lower drop-out rate of SSRIs is most likely due to a evidence to support their use as a first-line treatment more tolerable side-effect profile. Side-effects of TCAs for PD: high-potency benzodiazepines and anti- may also prevent therapeutic doses of these drugs depressants. There are several reasons for choosing being given. Taken together, this has led to a pre- antidepressants rather than benzodiazepines. Perhaps ference for SSRIs over TCAs in both clinical practice the most important is the adverse effect profile of and in consensus guidelines (Ballenger et al., 1998b; the benzodiazepines, including problems with with- drawal. Furthermore, in contrast to antidepressants, There is no data that show clear advantages of one benzodiazepines are not effective in reducing the de- of the SSRIs over the others. All six that are currently pressive symptomatology that often accompanies PD available (citalopram, escitalopram, fluvoxamine, flu- (van Balkom et al., 1997). Finally, the risk of relapse or oxetine, paroxetine, and sertraline) have been found to recurrence of PD after discontinuation of the benzo- be effective in the treatment of PD in double blind, diazepines is relatively high. It is not surprising, therefore, that antidepressants have become the first- During the first weeks of treatment with a SSRI it line pharmacotherapy of choice in the treatment of PD.
may be helpful to add a low dose of a high-potency The relative efficacy of different groups of anti- benzodiazepine (alprazolam, clonazepam) as an ad- depressants in the treatment of PD and related symp- ditional medication (Goddard et al., 2001). This may toms is, however, still a matter of debate (Bakker et al., result in more rapid stabilization and higher response 2002; Otto et al., 2001). To date, no differences in effi- rates during the first weeks of treatment. The risk of cacy have been demonstrated between the two groups an initial worsening of anxiety symptoms is probably Evidence-based pharmacotherapy of panic disorder reduced by this regimen as well. The main problem Table 3. Treatment options for refractory patients that may result from prescribing a benzodiazepine as aco-medication is that patients refuse to stop taking this agent. This may lead to dependency and other prob- SSRI (for at least 4 wk in a therapeutic dosage) lems related to long-term use of benzodiazepines. In our clinical practice we mention the possibility of taking a benzodiazepine during the first 2 or 3 wk of Tricyclic antidepressant (clomipramine or imipramine) treatment with an antidepressant. We tell the patients that they can decide themselves whether they want High potency benzodiazepine (in a high dosage) this co-medication, but that the duration of the prescription is strictly limited to 3 wk.
It has been suggested that the treatment of PD requires a lower starting dose of the antidepressant * Early co-administration of clonazepam or alprazolam may compared with depression to prevent an initial worsening of symptoms and reduce the rate of drop-out due to adverse effects. In patients who have pre- should be continued for at least 1 yr. There are also viously been unable to tolerate a standard dose of papers that have reported positively on long-term antidepressant medication, this would certainly seem treatment with the SSRIs fluoxetine and paroxetine a rational approach. However, explaining to the (Lydiard et al., 1998; Michelson et al., 1999). Important patient that such a regimen may be associated with a consensus guidelines have reached the conclusion that delayed response, may lead to the patient to choose to medication should be taken for at least 1 yr (Ballenger start a therapeutic dose sooner. In such cases, patients et al., 1998b; Bandelow et al., 2002).
should be made aware of the possibility of early Nevertheless, a more recent paper by the same transient side-effects, and the possibility of lowering authors (Mavissakalian and Perel, 2002) concluded that neither the duration of treatment with imipraminenor the method of discontinuation were predictorsof relapse. In this study, the rate of relapse after only How long should maintenance pharmacotherapy 6 months of treatment was identical to the rate of relapse after 12–30 months of treatment. The main As PD runs a chronic course, long-term treatment limitation of these findings is the limited sample outcome may be more important than short-term effi- size: only 51 patients were included in the analyses.
cacy. CBT appears to have long-term benefits, and Relapse prevention for patients who want to dis- generally, the short-term results are maintained.
continue their medication can potentially be enhanced Naturalistic follow-up studies of psychotherapy have by the addition of psychotherapy. An interesting been published as long as 9 yr after short-term treat- study by Wiborg and Dahl (1996) reported that The long-term effect of psychopharmacological treatments for PD has received less attention. A major lower relapse rates 18 months after treatment than clinical problem is the fact that treatment gains patients treated with clomipramine alone. Research on may disappear after tapering off the medication.
long-term treatment and discontinuation of therapies This is perhaps especially true for the high-potency requires more attention, for pharmacotherapy as well benzodiazepines. Relapse rates up to 80 % have as psychotherapy. The conflicting results of the limited been reported following complete withdrawal from data that are available with respect to the optimal duration of pharmacotherapy underlines this need.
There is relatively little data that addresses the Mavissakalian and Perel (1992) reported that when responders to a 6-month trial of imipramine weretreated at half-dose for another 6 months, they main- Despite advances in the pharmacotherapy of PD, not tained their improvement. This group of patients all patients respond to the first trial of medication.
showed significantly lower relapse rates than a group Unfortunately, there is very little persuasive data with of patients treated with imipramine for 6 months only.
respect to this topic. We suggest that when an SSRI These data suggest that successful pharmacotherapy fails, irrespective of the reason, a second SSRI is a reasonable next step to take next. Co-medication with Agoraphobic behaviour in particular may respond a benzodiazepine, preferably a high-potency benzo- well to this combined strategy. It is, however, unclear diazepine like alprazolam or clonazepam, can also be how long and at what dosage the medication should considered on a case by case basis, depending on the be continued, and what the optimal duration and frequency of the exposure treatment is.
If a second SSRI is not effective, then a third choice Another important issue that remains to be fully of medication may be a TCA, e.g. clomipramine or resolved is the costs that accompany different treat- imipramine. Such an agent can also be combined with ment strategies. Drop-out due to adverse side-effects a benzodiazepine. Subsequent options for pharmaco- of medication, lack of efficacy, lack of compliance and logical treatment of refractory patients include a high relapse after cessation of medication are important dose of high-potency benzodiazepines (up to 6 mg factors that increase the total cost of treatment with alprazolam or the equivalent), and treatment with pharmacological agents. Psychotherapeutic inter- an irreversible MAOI, e.g. phenelzine.
ventions also have associated costs. Further work on The addition of CBT to non- or partial medication the cost-efficacy of the treatment of PD is required.
responders should also be considered. Table 3 In the interim, we suggest that when patients suffer summarizes the strategies in case of non-response to from panic attacks with no or only limited avoidance then treatment of first choice is an SSRI, but that whenpatients suffer from agoraphobic avoidance due totheir panic attacks, exposure should be added. Our advice is to give the antidepressant first, and start When untreated, PD may run a chronic course, with a formal exposure after the first positive effects of the waxing and waning of symptoms. The high preva- medication have occurred. The antidepressant should lence and significant disability associated with PD, as be continued at least 9 months after maximal efficacy well as the frequent co-occurrence of PD with mood has been attained. Psychotherapy should continue disorders and substance abuse disorders underline the throughout this whole period, and can be stopped importance of having effective treatments available.
after patients have been free of medication for at least Fortunately, recent decades have witnessed the development of a number of effective treatments,including psychopharmacological interventions.
With regard to the short-term treatment of PD, several standard pharmacological and psychological interventions show equal efficacy in reducing panic,agoraphobia and related complaints. Nevertheless,SSRIs are increasingly viewed as a pharmacotherapy of choice, partly because benzodiazepines have rela- tively little effect on the depressive symptoms thatoften complicate PD. Short-term co-medication withhigh-potency benzodiazepines may be a useful strat- APA (2000). Diagnostic and Statistical Manual of Mental For both pharmacotherapy and psychotherapy it is Disorders (4th edn, text revision). Washington, DC: still unclear how long treatment should be continued, in what dose medication should optimally be con- Bakker A, van Balkom AJLM, Spinhoven P (2002). SSRIs tinued after remission of symptoms has been versus TCAs in the treatment of panic disorder: a meta- achieved, and what the optimal approach to refractory analysis. Acta Psychiatrica Scandinavica 106, 163–167.
patients is. After withdrawal of short-term benzodiaz- Bakker A, van Balkom AJLM, Spinhoven P, Blaauw BMJW, epine and antidepressant treatment there are rela- van Dyck R (1998). Follow-up on the treatment of panic tively high percentages of relapse. Data for valid disorder with or without agoraphobia : a quantitativereview. Journal of Nervous and Mental Disease 186, 414–419.
comparisons of the differences in relapse between Bakker A, van Balkom AJLM, van Dyck R (2000). Selective different medication classes is not yet available.
serotonin reuptake inhibitors in the treatment of panic With regard to the combination of different treat- disorder and agoraphobia. International Clinical ment strategies, there is some evidence that the com- Psychopharmacology 15 (Suppl. 2), S25–S30.
bination of an antidepressant and exposure in vivo Bakker A, van Dyck R, Spinhoven P, van Balkom AJLM (1999). Paroxetine, clomipramine and cognitive therapy in Evidence-based pharmacotherapy of panic disorder the treatment of panic disorder. Journal of Clinical agoraphobia: comparison of fluvoxamine, placebo, and psychological panic management combined with exposure Ballenger JC, Davidson JRT, Lecrubier Y, Nutt DJ, Baldwin and of exposure in vivo alone. American Journal of DS, den Boer JA, Kasper S, Shear MK (1998b). Consensus statement on panic disorder from the international den Boer JA, Westenberg HGM (1988). Effect of a serotonin consensus group on depression and anxiety. Journal of and noradrenaline uptake inhibitor in panic disorder ; a Clinical Psychiatry 59 (Suppl. 8), 47–54.
double-blind comparative study with fluvoxamine and Ballenger JC, Wheadon DE, Steiner M, Bushnell W, Gergel maprotiline. International Clinical Psychopharmacology 3, IP (1998a). Double-blind, fixed-dose, placebo-controlled study of paroxetine in the treatment of panic disorder.
den Boer JA, Westenberg HGM, Kamerbeek WDJ (1987).
American Journal of Psychiatry 155, 36–42.
Effect of a serotonin uptake inhibitor in anxiety disorders : Bandelow B, Zohar J, Hollander E, Kasper S, Moller HJ a double-blind comparison of clomipramine and (2002). World Federation of Societies of Biological fluvoxamine. International Clinical Psychopharmacology 2, Psychiatry (WFSBP) guidelines for the pharmacological treatment of anxiety, obsessive-compulsive and Eaton WW, Dryman A, Weissman MM (1991). Panic and posttraumatic stress disorders. World Journal Biological phobia. In : Robins LN, Regier DA (Eds.), Psychiatric Disorders in America (pp. 155–179). New York: Free Press.
Beauclair L, Fontaine R, Annable L, Holobow N, Chouinard Eaton WW, Kessler RC, Wittchen HU, Magee WJ (1994).
G (1994). Clonazepam in the treatment of panic disorder: a Panic and panic disorder in the United States. American double-blind, placebo-controlled trial investigating the Journal of Psychiatry 151, 413–420.
correlation between clonazepam concentrations in plasma Emmelkamp PMG, Bouman T, Scholing HA (1992). Anxiety and clinical response. Journal of Clinical Psychopharmacology disorders: A Practitioner’s Guide. Chichester: Wiley.
Goddard AW, Brouette T, Almai A, Jetty P, Woods SW, Bijl RV, van Zessen G, Ravelli A (1997). Psychiatric Charney D (2001). Early coadministration of clonazepam morbidity among adults in the Netherlands: the NEMESIS with sertraline for panic disorder. Archives of General study. Prevalence of psychiatric disorders. Nederlands Tijdschrift voor Geneeskunde 141, 2453–2460.
Gorman JM (1997). The use of newer antidepressants for Black DW, Wesner R, Bowers W, Gabel J (1993).
panic disorder. Journal of Clinical Psychiatry 58 (Suppl. 14), A comparison of fluvoxamine, cognitive therapy, and placebo in the treatment of panic disorder. Archives of Hoehn-Saric R, McLeod D, Hipsley PA (1993). Effect of fluvoxamine on panic disorder. Journal of Clinical Bourdon KH, Boyd JH, Rae DS, Burns BJ, Thompson JW, Locke BZ (1988). Gender differences in phobias: results of Jonas JM, Cohon MS (1993). A comparison of the safety the ECA community survey. Journal of Anxiety Disorders 2, and efficacy of alprazolam versus other agents in the treatment of anxiety, panic, and depression: a review of Boyer W (1995). Serotonin uptake inhibitors are superior to the literature. Journal of Clinical Psychiatry 54 (Suppl. 10), imipramine and alprazolam in alleviating panic attacks: a meta-analysis. International Clinical Psychopharmacology 10, Katerndahl DA, Realini JP (1993). Lifetime prevalence of panic states. American Journal of Psychiatry 150, 246–249.
Burrows GD, Norman TR (1999). The treatment of panic Lecrubier Y, Bakker A, Dunbar G, Judge R, and the disorder with benzodiazepines. In : Panic Disorder: Clinical Collaborative Paroxetine Panic Study Investigators Diagnosis, Management and Mechanisms (pp. 145–158).
(1997). A comparison of paroxetine, clomipramine and placebo in the treatment of panic disorder. Acta Psychiatrica Cassano GB, Petracca A, Perugi G, Nisita C, Musetti L, Mengali F, McNair DM (1988). Clomipramine for panic Lydiard RB, Steiner M, Burnham D, Gergel I (1998). Efficacy disorder, I: the first 10 weeks of a long-term comparison studies of paroxetine in panic disorder. Psychopharmacology with imipramine. Journal of Affective Disorders 14, Marks IM (1987). Fears, Phobias and Rituals. Oxford, NY: Chignon JM, Lepine JP, Ades J (1993). Panic disorder in cardiac outpatients. American Journal of Psychiatry 150, Marks IM, Swinson RP, Basoglu M, Kuch K, Noshirvani H, OkSullivan G, Lelliott PT, Kirby M, McNamee G, Clark DM (1986). A cognitive approach to panic. Behaviour Sengun S, Wickwire K (1993). Alprazolam and exposure alone and combined in panic disorder with agoraphobia.
Cross National Collaborative Panic Study, Second Phase British Journal of Psychiatry 162, 776–787.
Investigators (1992). Drug treatment of panic disorder.
Mavissakalian M, Perel JM (1992). Clinical experiments in British Journal of Psychiatry 160, 191–202.
maintenance and discontinuation of imipramine therapy de Beurs E, van Balkom AJLM, Lange A, Koele P, in panic disorder with agoraphobia. Archives of General van Dyck R (1995). Treatment of panic disorder with Mavissakalian M, Perel JM (2002). Duration of imipramine multicenter flexible-dose trials. Psychopharmacology Bulletin therapy and relapse in panic disorder with agoraphobia.
Journal of Clinical Psychopharmacology 22, 294–299.
Rosenberg R (1999). Treatment of panic disorder with Michelson D, Lydiard RB, Pollack MH, Tamura RN, Hoog tricyclics and MAOIs. In : Panic Disorder: Clinical Diagnosis, SL, Tepner R, Demitrack MA, Tollefson GD and the Management and Mechanisms (pp. 125–144). London: Martin Fluoxetine Panic Disorder Study Group (1998).
Outcome assessment and clinical improvement in panic Roy-Byrne PP, Stang P, Wittchen HU, Ustun B, Walters EE, disorder: evidence from a randomized controlled trial of Kessler RC (2000). Lifetime panic-depression co-morbidity fluoxetine and placebo. American Journal of Psychiatry 155, in the National Comorbidity Survey. British Journal of Michelson D, Pollack M, Lydiard RB, Tamura R, Tepner R, Seedat S, van Rheede van Oudtshoorn E, Muller JE, Tollefson G and the Fluoxetine Panic Disorder Study Mohr N, Stein DJ (2003). Reboxetine and citalopram in Group (1999). Continuing treatment of panic disorder after panic disorder: a single-blind, cross-over, flexible-dose acute response : randomised, placebo-controlled trial with pilot study. International Clinical Psychopharmacology 18, fluoxetine. British Journal of Psychiatry 174, 213–218.
Milrod BL, Busch FN, Cooper AM, Shapiro T (1997). Manual Sharp DM, Power KG, Simpson RJ, Swanson V, Moodie E, of Panic-Focused Psychodynamic Psychotherapy. Washington, Anstee JA, Ashford JJ (1996). Fluvoxamine, placebo, and DC : American Psychiatric Association.
cognitive behaviour therapy used alone and in Modigh K, Westberg P, Eriksson E (1992). Superiority of combination in the treatment of panic disorder and clomipramine over imipramine in the treatment of panic agoraphobia. Journal of Anxiety Disorders 10, 219–242.
disorder: a placebo-controlled trial. Journal of Clinical Stein MB, Asmundson GJ, Ireland D, Walker JR (1994).
Panic disorder in patients attending a clinic for vestibular Nair NPV, Bakish D, Saxena B, Amin M, Schwartz G, West disorders. American Journal of Psychiatry 151, 1697–1700.
TEG (1996). Comparison of fluvoxamine, imipramine, and Tiller JWG, Bouwer C, Behnke K (1997). Moclobemide for placebo in the treatment of outpatients with panic anxiety disorders: a focus on moclobemide for panic disorder. International Clinical Psychopharmacology 12 Noyes R, Garvey MJ, Cook B, Suelzer M (1991). Controlled discontinuation of benzodiazepine treatment for patients van Balkom AJLM, Bakker A, Spinhoven Ph, Blaauw with panic disorder. American Journal of Psychiatry 148, BMJW, Smeenk S, Ruesink B (1997). A meta-analysis of the treatment of panic disorder with or without Oehrberg S, Christiansen PE, Behnke K, Borup AL, Severin agoraphobia: a comparison of psychopharmacological, B, Soegaard J, Calberg H, Judge R, Ohrstrom JK, cognitive-behavioral, and combination treatments. Journal Manniche PM (1995). Paroxetine in the treatment of panic of Nervous and Mental Disease 185, 510–516.
disorder. A randomised, double-blind, placebo-controlled van Balkom AJLM, Nauta MCE, Bakker A (1995).
study. British Journal of Psychiatry 167, 374–379.
Meta-analysis on the treatment of panic disorder with Otto MW, Tuby KS, Gould RA, McLean RY, Pollack MH agoraphobia: review and re-examination. Clinical (2001). An effect-size analysis of the relative efficacy and Psychology and Psychotherapy 2, 1–14.
tolerability of serotonin selective reuptake inhibitors for Wade AG, Lepola U, Koponen HJ, Pedersen V, Pedersen T panic disorder. American Journal of Psychiatry 158, (1997). The effect of citalopram in panic disorder. British Journal of Psychiatry 170, 549–553.
Papp LA, Schneier FR, Fyer AJ, Liebowitz MR, Gorman JM, Wiborg IM, Dahl AA (1996). Does brief dynamic Coplan JD, Campeas R, Fallon BA, Klein DF (1997).
psychotherapy reduce the relapse rate of panic disorder? Clomipramine treatment of panic disorder: pros and cons.
Archives of General Psychiatry 53, 689–694.
Journal of Clinical Psychiatry 58, 423–425.
Wilkinson G, Balestrieri M, Ruggeri M, Bellantuono C Raj A, Sheehan DV (1987). Medical evaluation of panic (1991). Meta-analysis of double-blind placebo controlled attacks. Journal of Clinical Psychiatry 48, 309–313.
trials of antidepressants and benzodiazepines for Rapaport MH, Wolkow RM, Clary CM (1998).
patients with panic disorders. Psychological Medicine 21, Methodologies and outcomes from the sertraline

Source: http://dare2.ubvu.vu.nl/bitstream/handle/1871/23235/186720.pdf?sequence=1