Download University of Groningen The prediction of nonresponse to

University of Groningen
The prediction of nonresponse to pharmacotherapy in panic disorder
Slaap, Bernhard Reinier
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to
cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2001
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Slaap, B. R. (2001). The prediction of nonresponse to pharmacotherapy in panic disorder s.n.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 18-06-2017
Chapter 1
The prediction of nonresponse to
pharmacotherapy
in panic disorder:
A review
Bernhard R. Slaap M.A.
Johan A. den Boer M.D. Ph.D.
Depression and Anxiety: in press
Abstract
Background: Several effective pharmacotherapeutic treatments exist for panic disorder,
however, not all patients respond to treatment: between 20% to 40% are nonresponders.
Recent studies have reported several predictors of nonresponse to pharmacotherapy. In this
review two questions are addressed: is there consensus with respect to predictors of nonresponse, and are there any differences between short-term and long-term predictors?
Methods: In this review both short-term and long-term outcome studies are discussed.
Studies were included if at least DSM-III criteria were used and baseline variables were
investigated as possible predictor of response, or nonresponse, to pharmacotherapy. Of
each clinical predictor, tallies were made of the particular predictors employed and of those
predictors that predicted nonresponse.
Results: It appears that a long duration of illness and severe agoraphobic avoidance are
robust predictors of nonresponse, particularly in long-term studies. Personality disorders, or
even personality traits, are possibly the most robust predictors of nonresponse.
Conclusions: several factors appear to be robust predictors of nonresponse: factors that are
present before treatment and exert their influence on short-term and long-term treatment
outcome. Prospective studies are needed to further investigate these factors and to test
whether it is viable to intervene in an attempt to increase treatment response.
14
Introduction
In Panic Disorder (PD), with or without agoraphobia, it has been well established that
pharmacotherapy is highly effective in reducing anxiety and the frequency of panic attacks
(Feighner, 1999). Treatment with different compounds like tricyclic antidepressants (TCAs)
(Modigh, 1987; Liebowitz, 1989), selective serotonin reuptake inhibitors (SSRIs)(den Boer
and Westenberg, 1990; Westenberg and den Boer, 1994), high-potency benzodiazepines
(Jonas and Cohon, 1993) and monoamine oxidase inhibitors (MAOIs)(Tyrer and Shawcross,
1988; van Vliet et al., 1993; van Vliet et al., 1996) have been proven effective. However, not
all patients respond to treatment: between 20% to 40% are nonresponders.
Predicting nonresponse to pharmacotherapy would be very helpful for both patients and
clinicians, since it takes several weeks before a clinical effect can be expected for most of
these drugs. Only the high-potency benzodiazepine alprazolam (Andersch et al., 1991;
Rickels and Schweizer, 1998) is characterized by a rapid onset of therapeutic action. In view
of the delayed onset and side effects associated with pharmacotherapy (e.g. anticholinergic
effects, sedation or nausea, depending on the compound), it is an important clinical task to
identify factors which may hamper effective pharmacotherapy.
In this review we will survey the available data on predictors of nonresponse to pharmacotherapy in PD, both in short-term and in long-term studies. Two questions will be addressed: is there consensus in respect to predictors of nonresponse, and are there any differences between short-term and long-term predictors.
Method
In this review short-term studies, with a duration between 6 and 16 weeks (Table 1), and
long-term studies, with a duration of 1 to 7 years (Table 2), will be discussed.
A Medline and EMBASE search was conducted on the following key words: “panic”,
“agoraphobia”, “predict”, “depression”, “personality” and “follow-up”. The timeframe of
this literature search was 1986 through 2000. To further insure completeness, the search
results were cross-checked with cited references in these studies.
For inclusion, studies had to investigate baseline variables as possible predictors of
response, or nonresponse, to pharmacotherapy. Studies were included in this review if
patients were treated with an adequate dose of adequate medication for an adequate duration and if at least DSM-III criteria were used. Cowley and co-workers (1997) have demonstrated that it is essential to ensure adequacy of treatment, duration and dose before
attempting any prediction of response. In their sample of 106 PD patients, previously
treated by their community practitioners, only 59 (23%) of 252 documented medication
trials were adequate in medication, dosage and treatment duration. Other researchers have
reported similar findings (Taylor et al., 1989; Bandelow et al., 1995b). In determining the
adequacy of medication, dose and treatment duration the “Practice Guideline for the
Treatment of Patients with Panic Disorder”, issued by the American Psychiatric Association
15
(Work Group on Panic Disorder.American Psychiatric Association, 1998), were followed.
For inclusion in this review, the duration of treatment had to be at least 6 weeks.
In order to predict nonresponse to treatment, it is essential to define criterion variables,
that will determine the expected response or nonresponse (Steketee and Chambless, 1992).
Unfortunately, there is no consensus on how to measure response (Shear and Maser, 1994;
Shear et al., 1998). The DSM-IV doesn’t offer criteria for response or remission and the
literature on pharmacotherapy in PD abounds with many different scales which are used as
criterion variables. Consequently, the studies in this review do not report on nonresponse in
a uniform fashion. This problem hampers a methodologically flawless comparison of studies
which investigated nonresponse to pharmacotherapy in PD. In this review nonresponse is
defined in a general way as a lack of response to adequate treatment. Nonresponse can also
mean that patients were unable to achieve remission (Pollack et al., 1990; Rickels et al., 1993;
Warshaw et al., 1997; Toni et al., 2000). As the DSM-IV does not include a definition on
remission in PD, remission is defined as an asymptomatic state with a duration of at least
two (Pollack et al., 1990; Warshaw et al., 1997; Toni et al., 2000) or three months (Rickels et
al., 1993).
Studies were also included if they investigated the effect on treatment outcome of a
comorbid disorder, such as major depression or a personality disorder.
In order to investigate the strength of each predictor, we compared the number of studies
which investigated a particular predictor with the number of studies which actually found
that variable to predict nonresponse (Table 3).
In most short-term studies, treatments are pooled when predictors of nonresponse are
investigated. An exception to this rule are studies which investigated a large number of
patients; these studies reported on predictors of nonresponse for each treatment separately
(Lesser et al., 1988; Maier et al., 1991; Keller et al., 1993; Pollack et al., 1994; Woodman et al.,
1994; Hofmann et al., 1998). The larger sample sizes of these studies enabled the researchers
to investigate possible differential effects of predictors of nonresponse. When pertinent
these differential effects are discussed.
In long-term studies, treatment is often naturalistic, which means that for the prediction
of nonresponse treatments are also pooled. In this review results are discussed for all
pharmacological treatments, irrespective of the type.
Predictors of nonresponse
In Table 3 the most prominent baseline predictors are depicted. These are subdivided in
three categories: demographic variables, illness variables and comorbidity. As can be seen in
the table, it is clear that not all predictors are equally robust, or as widely investigated as
others.
It has to be taken into account that these results may be biased. Demographic variables,
for instance, were probably analysed in more studies than those who actually reported on
these predictors. This means that the importance of some predictor variables could actually
16
be even less than some authors have suggested (Mavissakalian and Michelson, 1986;
Faravelli et al., 1995).
Demographic variables
Demographic characteristics, such as gender, age or marital status do not appear to be predictors of nonresponse. In most studies these variables were investigated as potential predictors of nonresponse, but the results are inconsistent at best. Two studies have reported
that gender is associated with nonresponse. In a short-term study, Mavissakalian and
Michelson (1986) have found that male gender predicted a poor outcome. In contrast, Maier
and Buller (1988) reported in a long-term study that female gender was a predictor of
nonresponse.
In only one short-term study it was found that a younger age predicted nonresponse
(Woodman et al., 1994). In a recent long-term study by Faravelli and co-workers (1995) the
opposite was reported. Additionally, this study is also the only study that found marital
status to be a predictor of nonresponse. A larger proportion of patients who were not married at baseline were doing well at follow-up, as compared to married patients.
Social-economic class may be the best demographic predictor, but is has not been widely
investigated. In two long-term studies, Noyes, Jr. and co-workers (1990; 1993) investigated
this predictor and in both studies they found that a lower social economic class predicted
nonresponse. Recently, Warshaw et al. (1997) reported in a long-term study that they also
found a lower social-economic class to predict nonresponse.
Illness variables
Duration of illness and age at onset
Duration of illness and age at onset have both been investigated thoroughly as likely predictors of nonresponse. A longer duration of illness has been found to predict nonresponse
in more than half of the long-term studies (Noyes, Jr. et al., 1989; Noyes, Jr. et al., 1990;
Noyes, Jr. et al., 1993; Pollack et al., 1993; Albus et al., 1995; Faravelli et al., 1995; Katschnig et
al., 1995; Scheibe and Albus, 1997; Shinoda et al., 1999; Toni et al., 2000) (see Table 3).
Noyes, Jr. et al. (1993) reported a mean ( ± SD) duration of illness for patients with a
marked improvement of 7.7 (± 8.7) years and 12.5 (± 12.0) years for patients without
marked improvement. Faravelli et al. (1995) reported an even more dramatic difference;
patients who were fully recovered or improved had a mean duration of illness of 3.1 (± 3.8)
years versus 10.7 (± 10.6) years for patients who had a recurrent course of illness or poor
outcome.
This predictor does not seem to be as robust in short-term studies, since in only a minority of the studies a longer duration of illness was found to predict nonresponse
(Mavissakalian and Michelson, 1986; Basoglu et al., 1994). It appears that the influence of a
long duration of illness is less evident on the short-term treatment effect, indicating that
17
chronic patients also benefit from short-term treatment. The effect of a longer duration of
illness becomes more evident when the long-term treatment gains are investigated.
Age at onset appears to be a less important factor, since it was only found to predict
nonresponse in a minority of the long-term studies (Noyes, Jr. et al., 1990; Scheibe and
Albus, 1997; Warshaw et al., 1997; Toni et al., 2000). In these studies it was found that a
younger age at onset was associated with nonresponse. Age at onset did not predict nonresponse in any of the short-term studies.
Global illness severity
Global illness severity, as measured by a global scale such as the Clinical Global Impression
(CGI) (Guy, 1976), has been investigated primarily in short-term studies as a possible
predictor of nonresponse. In two short-term studies it was found that a higher score on a
global illness scale predicted nonresponse (Pollack et al., 1994; Pollack et al., 1993). Three
other short-term studies did not show this relation (Liebowitz et al., 1986; Mavissakalian and
Michelson, 1986; Ito et al., 1995). In long-term studies, this relation was only found in the
study of Noyes, Jr. and co-workers (1989). Two other studies were unable to replicate this
finding (Faravelli and Albanesi, 1987; Pollack et al., 1993).
Anxiety
Free-floating, chronic anxiety has been found to predict nonresponse in about half of the
short-term (Liebowitz et al., 1986; Rosenberg et al., 1991a; Scheibe et al., 1992; Woodman et
al., 1994; Sharp and Power, 1999) and long-term studies (Faravelli and Albanesi, 1987;
Lelliott et al., 1987; Noyes, Jr. et al., 1990; Rickels et al., 1993; Albus et al., 1995). Nonresponders to pharmacotherapy tended to have higher baseline scores on scales such as the
Hamilton Anxiety Scale (HAS) (Hamilton, 1959) or the Clinical Anxiety Scale (CAS) (Snaith
et al., 1982). Whether these higher scores reflect a greater illness severity or a (sub-threshold)
comorbid generalized anxiety disorder is difficult to determine since these scales measure
both generalized anxiety and panic anxiety. Due to the high overlap of symptoms, these
scales are not sufficiently specific to differentiate these two disorders (Bech et al., 1993).
Not only can differences in HAS scores be quite dramatic between responders and nonresponders, but also between studies. In a short-term study Liebowitz and co-workers
(1986) reported that nonresponders had a baseline HAS score (mean ± SD) of 22.5 ± 5.3,
whereas responders had 14.4 ± 5.6. In a long-term study Noyes, Jr. et al. (1990) reported a
mean baseline HAS score of 14.6 for patients with high self-rated anxiety at follow-up,
versus 8.3 for patients with low self-rated anxiety.
Phobic avoidance
As reported in epidemiological studies, about a third to one half of the patients with PD also
have agoraphobia (Klerman et al., 1991; Eaton et al., 1994; Weissman et al., 1997). This
percentage is substantially higher in clinical samples, and the studies reviewed here, where
some 80% of the patients have PD and agoraphobia. Patients with PD and agoraphobia are
18
reported to be more ill than patients with PD only: they have more panic symptoms
(Starcevic et al., 1993; Goisman et al., 1994) and a higher score on the HAS and the HDS
(Lesser et al., 1988).
Phobic avoidance can be measured with scales such as the Fear Questionnaire (FQ)
(Marks and Mathews, 1979) or derivatives like the Phobia Scale (Ballenger et al., 1988).
Severity of phobic avoidance appears to be one of the most robust predictors of nonresponse, particularly in long-term studies. Almost all long-term studies have found that
nonresponders have higher scores at baseline. Lelliott et al. (1987) and Faravelli et al. (1995)
are the only researchers who did not find agoraphobic avoidance to predict nonresponse.
Sample characteristics may explain why Lelliott and co-workers (1987) reported different
findings. They investigated a sample of 40 patients with severe agoraphobia. Both
responders and nonresponders alike, all had a very high score on the FQ, which makes it
difficult, if not impossible, to find significant differences between these groups. The sample
in the study Faravelli et al. (1995) was a mix of PD patients with and without agoraphobia:
38% of the patients had a diagnosis of PD without agoraphobia. Since the authors
themselves made no attempt, an explanation to these deviant findings is cumbersome.
In a majority of short-term studies, it was found that severity of agoraphobic avoidance
predicted nonresponse (Mavissakalian and Michelson, 1986; Rosenberg et al., 1991a; Pollack
et al., 1994; Basoglu et al., 1994; Woodman et al., 1994; Sharp and Power, 1999). Slaap et al.
(1995) could not replicate this finding, but they did find that nonresponders had a higher
score on the blood-injury phobia subscale of the FQ, indicative of phobic avoidance of
another type.
Two studies reported on differential effects of phobic avoidance on pharmacotherapy.
Maier et al. (1991) reported that alprazolam and imipramine were more effective than placebo, particularly in patients with severe phobic avoidance. In terms of patients with less
severe avoidance, active medication was about as effective as placebo. In contrast, Pollack et
al. (1994) reported opposite differential effects: in patients with severe agoraphobic
avoidance they found no evidence of superiority of alprazolam or imipramine over placebo.
The samples of both studies are not directly comparable, as in the study of Pollack et al.
(1994) only PD patients with agoraphobic avoidance were included. Furthermore, this study
had a much smaller sample size (n=126 vs. n=1134) which influences the power of such
comparisons. In conclusion: the relation between severity of agoraphobic avoidance and
outcome of pharmacotherapy is very robust, although possibly not as strong in short-term
studies.
Frequency & intensity of panic attacks
The accurate measurement of the frequency and intensity of panic attacks is one of the
challenges PD researchers encounter. It has been noted that not only large variations in
panic frequency exist between patients, but also within patients (Shear and Maser, 1994;
Bouchard et al., 1997). This phenomenon reduces the power to discriminate between
19
treatments and makes the frequency and intensity of panic attacks a less attractive efficacy
measure (Bandelow et al., 1995a).
This variable has been studied as a possible predictor of nonresponse to pharmacotherapy. It appears that nonresponders are characterized by a greater panic frequency or intensity at baseline in about half of the short-term (Mavissakalian and Michelson, 1986;
Rosenberg et al., 1991a; Woodman et al., 1994; Slaap et al., 1995; Sharp and Power, 1999) and
long-term studies (Noyes, Jr. et al., 1989; Noyes, Jr. et al., 1990; Noyes, Jr. et al., 1993; Rickels
et al., 1993). The aforementioned problems with the measurement of panic frequency and
intensity, together with differences in patient samples, may explain why this predictor was
not found in more studies.
Comorbidity
Comorbid Depression
The prevalence of comorbid depression in patients with PD ranges between 23% and 53%
(Klerman, 1990; Gorman and Coplan, 1996; Pelissolo and Lepine, 1998). The average reported lifetime occurrence of major depression in patients with PD is also high: between
40% and 50% (Lydiard, 1991; Gorman and Coplan, 1996; Pelissolo and Lepine, 1998). In
large epidemiological studies, such as the National Comorbidity Study (NCS) (Kessler et al.,
1998; Roy Byrne et al., 2000) and the World Health Organization (WHO) Collaborative
Study on Psychological Problem in General Health Care (Lecrubier and Ustun, 1998), and
in several clinical studies (Grunhaus et al., 1994; Chambless and Gracely, 1988), it has been
reported that comorbid patients tend to have more severe symptoms and more impairment.
An important question is whether this implies less response to treatment.
In short-term studies, a comorbid depression has been found to predict nonresponse in
two out of the six studies which investigated this predictor (Reich, 1988; Pollack et al., 1993).
The results of the study of Reich (1988) need to be interpreted carefully, because in his
study the outcome of treatment of only four PD patients with a secondary depression is
compared to the outcome of 45 patients without depression. Four other short-term studies,
with larger patient samples, did not find a relation between a comorbid depression and
treatment outcome (Lesser et al., 1988; Keller et al., 1993; Basoglu et al., 1994; Rosenberg et
al., 1991b). Rosenberg and co-workers (1991b) reported that some 20% of their sample met
the DSM-III criteria for a diagnosis of current or past depression. These patients were more
ill at baseline, but their improvement was not significantly different after eight weeks of
treatment with either alprazolam or imipramine.
In about half of the long-term studies, a comorbid depression was found to predict nonresponse (Maier and Buller, 1988; Nagy et al., 1989; Noyes, Jr. et al., 1990; Albus and
Scheibe, 1993; Noyes, Jr. et al., 1993; Albus et al., 1995). Maddock and Blacker (1991) reported, that as a group, patients with a comorbid diagnosis of current or lifetime history of
depression improved as much as the patients without depression. This did not apply to all
patients; the patients with a primary depression, who had their first episode before the onset
20
of panic disorder, had a worse outcome. Pollack and co-workers (1993) reported that a
comorbid diagnosis of dysthymia was not a predictor of long-term treatment response,
while it did predict nonresponse to short-term treatment (see above). In another study,
Pollock and co-workers (1990) reported that there was no relation between a lifetime or
current diagnosis of depression and failure to achieve remission (Pollack et al., 1990). They
did find that a diagnosis of comorbid depression was associated with other predictors of
nonresponse, for example a high panic frequency or a comorbid anxiety disorder. This
suggests that there is a relation between a comorbid depression and nonresponse, albeit not
a direct relation.
In long-term studies which found that a comorbid depression was a predictor of nonresponse, patients were more ill at follow-up. These patients had higher HAS and HDS
scores (Nagy et al., 1989; Albus et al., 1995), more disability (Noyes, Jr. et al., 1990; Albus and
Scheibe, 1993) and more panic symptoms (Noyes, Jr. et al., 1993). Albus and co-workers
(1995) remark in their study that patients with a comorbid depression had a poorer
outcome. However, no differences were found on several relevant illness variables, such as
agoraphobic avoidance and the number of panic attacks. The authors speculate that this
finding may be due to the fact that 45% of the patients in the comorbid group and 33.3% of
the pure panic group were still on medication.
Three studies investigated whether depressed patients with a comorbid panic disorder had
a worse prognosis that those with depression only (Grunhaus et al., 1986; Grunhaus et al.,
1988; Coryell et al., 1988). In both short-term studies (Grunhaus et al., 1986; Grunhaus et al.,
1988) and the long-term study (Coryell et al., 1988), depressed patients with a comorbid PD
had a worse treatment outcome.
It appears that the combination of PD and depression, irrespective of primary diagnosis,
is a factor to be reckoned with. Comorbid patients seem to have a reduced chance of getting
well. However, there is evidence that the effect of comorbidity on the short-term treatment
effect is less robust, meaning that patients with a comorbid condition may equally benefit
from pharmacotherapy, as compared to patients with a single diagnosis.
Comorbid Anxiety Disorder
The available literature on comorbidity suggests that a significant percentage of patients with
PD has an additional diagnosis. The percentage of PD patients with a diagnosis of an
additional anxiety disorder ranges between 44% (Di Nardo and Barlow, 1990) and 83.3%
(Starcevic et al., 1992). In the above cited studies, as well as in other papers (Brown and
Barlow, 1992; de Ruiter et al., 1989; Sanderson et al., 1990; Marshall, 1996; Hoffart et al.,
1995) it appears that generalized anxiety disorder, social phobia and simple phobia are the
most frequently occurring additional diagnoses in PD patients. Percentages ranging between
19% (Hoffart et al., 1995) and 51.9% (Starcevic et al., 1992) have been reported for a
comorbid generalized anxiety disorder. For a comorbid social phobia the percentages vary
between 9.6% (de Ruiter et al., 1989) and 40.7% (Starcevic et al., 1992) The percentages for a
comorbid simple phobia vary between 6.3% (Brown and Barlow, 1992) and 46.6% (de
21
Ruiter et al., 1989). Differences in reported percentages reflect differences in diagnostic
criteria and differences between samples.
In two short-term studies the influence of a comorbid anxiety disorder on outcome was
investigated (Pollack et al., 1993; Pollack et al., 1994). Both studies reported that the presence
of a comorbid social phobia or generalized anxiety disorder did not influence treatment
outcome.
In long-term studies the opposite has been found: in three studies the presence of a
comorbid anxiety disorder predicted nonresponse (Pollack et al., 1990; Pollack et al., 1993;
Scheibe and Albus, 1997). Pollack and co-workers (1990) reported that the presence of any
comorbid anxiety disorder predicted failure to achieve remission. The same group, in another study, reported that a comorbid diagnosis of social phobia predicted nonresponse
(Pollack et al., 1993). This diagnosis did not influence short-term treatment success, see
above. Scheibe and Albus (1997) recently reported that a comorbid generalized anxiety
disorder was the most robust predictor of nonresponse in their sample. Warshaw et al.
(1997) were surprised by their own contradictory findings: in their study, patients with a
comorbid social phobia were more likely to have experienced periods of remission.
The study of Toni and co-workers (2000) has to be interpreted with care. They did not
investigate whether the presence of a comorbid anxiety disorder influenced the likelihood of
achieving remission. They investigated whether comorbid anxiety disorders influenced the
total time spent in remission. In their study, the presence of these comorbid conditions,
diagnosed at baseline, had no effect on the duration of remission. Predicting the total time
spent in remission is different from predicting the likelihood of achieving remission. It is
imaginable that patients with a comorbid condition have a much reduced chance of
achieving remission, while the duration of the period of remission is not shorter than in
patients without this comorbid condition. This makes it hard to compare the results of this
study with other studies.
Only a few studies have investigated the influence of a comorbid anxiety disorder on the
effect of pharmacotherapy in PD. Nevertheless, it appears that a comorbid anxiety disorder
may have a great influence on treatment outcome in the long term. More studies are needed
to verify this tentative conclusion.
Comorbid Personality Disorder
Personality disorders are not uncommon in PD patients: using categorical measures, incidence rates have been reported ranging from 27% to over 50% (Mavissakalian and Hamann,
1986; Reich et al., 1987; Friedman et al., 1987; Green and Curtis, 1988; Reich, 1988;
Mavissakalian and Hamann, 1988). These estimates greatly exceed those of the general
population, which range from 5.9% to 18% (Zimmerman and Coryell, 1989; Reich et al.,
1989; Samuels et al., 1994; Lenzenweger et al., 1997). Differences in instruments used and
investigated populations probably caused the variation in estimates (Noyes, Jr. et al., 1991).
In PD it has been repeatedly found that a certain amount of covariation exists between
the severity of personality disorder characteristics and severity of illness (Mavissakalian and
22
Hamann, 1986; Mavissakalian and Hamann, 1987; Wittchen et al., 1991; Hofmann et al.,
1998). The impairment in personality functioning may worsen on an exacerbation of PD
symptoms and may improve on effective treatment of PD. This suggests that personality
disorder characteristics are in part state dependent, because they can change significantly
with effective treatment (for a overview see Hofmann et al., (1998)). The measurement of
personality disorder characteristics in acute PD patients is therefore not independent of
illness severity, as they are likely to show more Axis II traits or diagnoses, compared to
when they are in remission.
The influence of a comorbid personality disorder on the outcome of pharmacotherapy
has been investigated in ten studies. Both personality traits and personality disorder diagnoses have been investigated. In these studies a diagnosis is established by means of a
structured clinical interview, such as the Structured Interview for DSM-III Personality
Disorders (SIDP) (Pfohl et al., 1982), or by means of a self-report questionnaire such as the
Personality Diagnostic Questionnaire (PDQ) (Hyler et al., 1983), or the Wisconsin Personality Disorders Inventory (WISPI) (Klein et al., 1990). Traits can be established by using
scales such as the Minnesota Multiphasic Personality Inventory (MMPI) (Hathaway and
McKinley, 1945).
The presence of a comorbid personality disorder, or even traits, is one of the most robust
predictors of nonresponse in PD. Almost all studies that have investigated the role of this
predictor have found it to predict nonresponse, in both short-term (Mavissakalian and
Hamann, 1987; Green and Curtis, 1988; Reich, 1988; Ito et al., 1995) and long-term studies
(Faravelli and Albanesi, 1987; Noyes, Jr. et al., 1990; Lydiard et al., 1992; O'Rourke et al.,
1997).
There are two notable exceptions: a short-term study of Hofmann and co-workers (1998)
and a long-term study by Toni et al.(2000). Hofmann and co-workers (1998) did not find
that personality disorder characteristics, as measured by the WISPI, predicted nonresponse
to either 12 weeks of imipramine or 11 sessions of individual cognitive-behavior therapy.
The authors acknowledge that their study is limited by the fact that it is unclear whether the
WISPI is a valid instrument to assess and diagnose personality disorder. Toni and coworkers (2000) did not find any influence of personality disorder characteristics on the total
time spent in remission. As discussed above, the design of their study hampers direct
comparison with other studies.
Ito et al. (1995), in a short-term study, and Faravelli and Albanesi (1987), in a long-term
study, investigated MMPI traits as possible predictors of nonresponse. Both groups reported
that nonresponders had higher scores on the Psychopathic Deviate and the Hysteria scales.
The authors merely state that these personality traits are associated with nonresponse, but an
interpretation of these results is not given.
Most studies that have investigated the influence of a comorbid personality disorder on
outcome reported that any personality disorder would interfere with treatment, irrespective
of cluster or exact diagnosis. This applies to studies that used structured clinical interviews,
e.g. the SIDP (Noyes, Jr. et al., 1990; Green and Curtis, 1988; Reich, 1988), or self-report
23
questionnaires, e.g. the PDQ (Pollack et al., 1990; O'Rourke et al., 1997; Mavissakalian and
Hamann, 1987; Reich, 1988). Both Green and Curtis (1988) and O’Rourke et al. (1997) have
reported that comorbid disorders of the anxious-fearful cluster were most prominent in
nonresponders. Reich (1988) reported that nonresponders were most notably characterized
by personality disorders of the dramatic cluster.
Discussion
In this review evidence has been presented for several factors which have a negative influence on the outcome of pharmacotherapy in PD patients. These factors are present
before treatment and exert their influence on short-term and long-term treatment outcome.
There does not appear to be much evidence for the existence of factors which specifically
predict nonresponse to a particular treatment. Six short-term studies investigated possible
treatment specific predictors of nonresponse. Two studies reported a differential predictive
effect of agoraphobia, albeit in opposite directions. All other predictors, investigated in these
studies, failed to show any differential effects. Based on the scantily available data, it appears
more probable that factors exist which hamper pharmacotherapy, irrespective of type.
Two questions were addressed in this review: is there consensus in respect to predictors
of nonresponse, and are there any differences between short-term and long-term predictors?
It appears that patients who are more ill at baseline, and who have been ill for a long time,
have a worse prognosis. The severity of illness is evident from higher scores on scales that
measure phobic avoidance, anxiety, panic frequency and from the presence of comorbid
disorders. The CGI, a global measure of illness severity, was able to detect differences
between responders and nonresponders in only two short-term studies which investigated
this predictor. The CGI is probably not sensitive enough to detect differences in illness
severity between responders and nonresponders at baseline.
Illness variables, such as panic frequency, anxiety and phobic avoidance appear to exert
their detrimental influence on both short-term and long-term treatment outcome. Anxiety
and panic frequency (or intensity) predicted nonresponse in about half of the studies which
investigated these predictors. Agoraphobic avoidance predicted nonresponse in a majority
of the short-term studies and in almost all long-term studies. It is not clear whether the
influence of severe agoraphobia is less substantial on short-term treatment success; further
research is needed to make a definite statement about this. It does seem to be the case that
severe agoraphobia is not controlled by the treatments offered in the studies in this review.
Several variables do not appear to influence short-term treatment success, but show their
harmful influence when the long-term treatment effect is evaluated. A longer duration of
illness, a comorbid anxiety disorder, and probably a comorbid depression, do not appear to
disrupt the short-term treatment effect. These variables did predict nonresponse to treatment in a majority of the long-term studies. This suggests that chronic patients, and/or
patients with a comorbid affective disorder, equally benefit from pharmacotherapy in the
short-term, as compared to patients without these conditions. However, in the long-term,
24
the treatment gains of these patients are not sustained. The mechanism, by which a chronic
condition and/or comorbidity disrupts long-term treatment outcome is unclear and warrants further research. Also of note is the relative lack of studies which investigated the
influence of comorbid anxiety disorders. This condition often co-occurs in PD patients, yet
only a few studies have investigated its influence on treatment outcome.
A comorbid diagnosis of a personality disorder, or even personality traits, is possibly the
most robust predictor of nonresponse in this review. Almost all studies which investigated
this predictor found it to predict nonresponse. This is a worrisome conclusion, and similar
findings have been reported on the influence of personality disorders on the treatment of
Axis-I disorders in general (Reich and Green, 1991; Reich and Vasile, 1993). Standard
treatment for PD is clearly less effective against PD with a comorbid personality disorders,
or even traits, which makes it evident that these patients should receive alternative treatments, or a combination of treatments, to reduce the influence of this comorbid condition.
Roy-Byrne and Cowley (1994) have reviewed follow-up studies in PD and have meticulously described some of the methodological problems involved. These problems also apply
to the prediction of nonresponse and all deal with differences between studies. These
differences make comparison cumbersome and definitive conclusions of predictors of
nonresponse to remain tentative.
To name the most prominent problems involved in the comparison of studies on the
prediction of nonresponse:
The lack of consensus on how to measure response to treatment in panic disorder,
meaning that a variety of instruments is used.
The patient samples differs between studies; not only in size, but also in different inclusion and exclusion criteria.
To date, all studies used a retrospective prediction. In other words, at the end of the
study, after the data had been collected, investigators examined the association between
potential predictors and outcome.
Future research would benefit greatly if consensus on definitions of response and remission could be reached. Consensus would make a vast difference in this field, as it would
make methodologically flawless generalization of findings possible. In spite of all the
problems mentioned above, it is our firm belief that the tentative conclusions from this
review are of value, in the absence of more definitive prospective studies. The clinical issue
here is predicting poor treatment response and the next step, based on the tentative
conclusions of this paper and others, would be a prospective trial, in which these hypotheses
can be tested. Ultimately, studies are essential which test whether it is possible to intervene,
based on baseline predictor variables, in an attempt to increase response to pharmacotherapy.
25
References
Albus M, Scheibe G. 1993. Outcome of panic disorder with or without concomitant depression: a 2-year prospective follow-up study. Am J Psychiatry 150:1878-1880.
Albus M, Scheibe G, Scherer J. 1995. Panic disorder with or without concomitant depression 5 years after treatment: A prospective follow-up. J Affect Disord 34:109-115.
Andersch S, Rosenberg NK, Kullingsjo H, Ottosson JO, Bech P, Bruun Hansen J, Hanson
L, Lorentzen K, Mellergard M, Rasmussen S. 1991. Efficacy and safety of alprazolam,
imipramine and placebo in treating panic disorder. A Scandinavian multicenter study. Acta
Psychiatr Scand Suppl 365:18-27.
Ballenger JC, Burrows GD, Du pont RL, Noyes R, Pecknold J, Rifkin A, Swinson R. 1988.
Alprazolam in panic disorder and agoraphobia; Results of a multi-center trial. I. Efficacy in
short term treatment. Arch Gen Psychiatry 45:413-422.
Bandelow B, Hajak G, Holzrichter S, Kunert HJ, Ruther E. 1995a. Assessing the efficacy of
treatments for panic disorder and agoraphobia.1. Methodological problems. Int Clin
Psychopharmacol 10:83-93.
Bandelow B, Sievert K, Rothemeyer M, Hajak G, Ruther E. 1995b. What treatments do
patients with panic disorder and agoraphobia get? Eur Arch Psychiatry Clin Neurosci
245:165-171.
Basoglu M, Marks IM, Swinson RP, Noshirvani H, Osullivan G, Kuch K. 1994. PreTreatment Predictors of Treatment Outcome in Panic Disorder and Agoraphobia Treated
with Alprazolam and Exposure. J Affect Disord 30:123-132.
Bech P, Malt UF, Dencker SJ, Ahlfors UG, Elgen K, Lewander T, Lundell A, Simpson GM,
Lingjaere O. 1993. Scales for assessment of diagnosis and severity of mental disorders. Acta
Psychiatr Scand Suppl 372.
Bouchard S, Pelletier MH, Gauthier JG, Cote G, Laberge B. 1997. The assessment of panic
using self-report: A comprehensive survey of validated instruments. J Anxiety Disord 11:89111.
Brown TA, Barlow DH. 1992. Comorbidity among anxiety disorders: implications for
treatment and DSM-IV. J Consult Clin Psychol 60:835-844.
Buller R, Maier W, Goldenberg IM, Lavori PW, Benkert O. 1991. Chronology of panic and
avoidance, age of onset in panic disorder, and prediction of treatment response. A report
from the Cross-National Collaborative Panic Study. Eur Arch Psychiatry Clin Neurosci
240:163-168.
26
Chambless DL, Gracely EJ. 1988. Prediction of outcome following in vivo exposure
treatment of agoraphobia. In: Hand I, Wittchen HU, editors. Panic & Phobias 2. New York:
Springer Verlag. p 209-220.
Coryell W, Endicott J, Andreasen NC, Keller MB, Clayton PJ, Hirschfeld RM, Scheftner
WA, Winokur G. 1988. Depression and panic attacks: the significance of overlap as reflected in follow-up and family study data. Am J Psychiatry 145:293-300.
Cowley DS, Ha EH, Roy Byrne PP. 1997. Determinants of pharmacologic treatment failure
in panic disorder. J Clin Psychiatry 58:555-561.
de Ruiter C, Rijken H, Garssen B, van Schaik A, Kraaimaat F. 1989. Comorbidity among
the anxiety disorders. J Anxiety Disord 3:57-68.
den Boer JA, Westenberg HGM. 1990. Serotonin function in panic disorder: a double blind
placebo controlled study with fluvoxamine and ritanserin. Psychopharmacology 102:85-94.
Di Nardo PA, Barlow DH. 1990. Syndrome and symptom comorbidity in the anxiety disorders. In: Maser JD, Cloninger CR, editors. Comorbidity in anxiety and depression.
Washington DC: American Psychiartic Press. p 205-230.
Eaton WW, Kessler RC, Wittchen HU, Magee WJ. 1994. Panic and Panic Disorder in the
United States. Am J Psychiatry 151:413-420.
Faravelli C, Albanesi G. 1987. Agoraphobia with panic attacks: 1-year prospective follow-up.
Compr Psychiatry 28:481-487.
Faravelli C, Paterniti S, Scarpato A. 1995. 5-year prospective, naturalistic follow-up study of
panic disorder. Compr Psychiatry 36:271-277.
Feighner JP. 1999. Overview of antidepressants currently used to treat anxiety disorders. J
Clin Psychiatry 60 Suppl. 22:18-22.
Friedman CJ, Shear MK, Frances A. 1987. DSM-III personality disorders in panic patients. J
Personal Disord 1:132-135.
Goisman RM, Warshaw MG, Peterson LG, Rogers MP, Cuneo P, Hunt MF, TomlinAlbanese JM, Kazim A, Gollan JK, Epstein-Kaye T, Reich JH, Keller MB. 1994. Panic,
Agoraphobia, and Panic Disorder with Agoraphobia - Data from a Multicenter Anxiety
Disorders Study. J Nerv Ment Dis 182:72-79.
Gorman JM, Coplan JD. 1996. Comorbidity of depression and panic disorder. J Clin
Psychiatry 57:34-43.
Green MA, Curtis GC. 1988. Personality disorders in panic patients: response to termination of antipanic medication. J Personal Disord 2:303-314.
27
Grunhaus L, Harel Y, Krugler T, Pande AC, Haskett RF. 1988. Major depressive disorder
and panic disorder. Effects of comorbidity on treatment outcome with antidepressant
medications. Clin Neuropharmacol 11:454-461.
Grunhaus L, Pande AC, Brown MB, Greden JF. 1994. Clinical Characteristics of Patients
with Concurrent Major Depressive Disorder and Panic Disorder. Am J Psychiatry 151:541546.
Grunhaus L, Rabin D, Greden JF. 1986. Simultaneous panic and depressive disorder:
response to antidepressant treatments. J Clin Psychiatry 47:4-7.
Guy W. 1976. Early Clinical Drug Evaluation Unit (ECDEU). Rockville, MD: National
Institute of Mental Health.
Hamilton M. 1959. The assessment of anxiety states by rating. Br J Med Psychol 32:50-55.
Hathaway SR, McKinley JC. 1945. Minnesota Multiphasic Personality Inventory. New York:
The Psychological Corporation.
Hoffart A, Thornes K, Hedley LM. 1995. DSM-III-R Axis I and II disorders in agoraphobic
inpatients with and without panic disorder before and after psychosocial treatment.
Psychiatry Res 56:1-9.
Hofmann SG, Shear MK, Barlow DH, Gorman JM, Hershberger D, Patterson M, Woods
SW. 1998. Effects of panic disorder treatments on personality disorder characteristics.
Depress Anxiety 8:14-20.
Hyler SE, Reider RO, Spitzer RL. 1983. Personality Diagnostic Questionnaire (PDQ). New
York: New York State Psychiatric Institute.
Ito LM, Gorenstein C, Gentil V, Miyakawa E. 1995. Minnesota Multiphasic Personality
Inventory correlates of panic disorder with agoraphobia: Changes with treatment. Braz J
Med Biol Res 28:961-965.
Jonas JM, Cohon MS. 1993. A comparison of the safety and efficacy of alprazolam versus
other agents in the treatment of anxiety, panic and depression: a review of the literature. J
Clin Psychiatry 54 (suppl.):25-45.
Katschnig H, Amering M, Stolk JM, Klerman GL, Ballenger JC, Briggs A, Buller R, Cassano
G, Garvey M, Roth M, Solyom C. 1995. Long-term follow-up after a drug trial for panic
disorder. Br J Psychiatry 167:487-494.
Keller MB, Lavori PW, Goldenberg IM, Baker LA, Pollack MH, Sachs GS, Rosenbaum JF,
Deltito JA, Leon A, Shear MK, Klerman GL. 1993. Influence of depression on the treatment of panic disorder with imipramine, alprazolam and placebo. J Affect Disord 28:27-38.
28
Kessler RC, Stang PE, Wittchen HU, Ustun TB, RoyBurne PP, Walters EE. 1998. Lifetime
panic-depression comorbidity in the National Comorbidity Survey. Arch Gen Psychiatry
55:801-808.
Klein MH, Benjamin LS, Treece C, Rosenfeld R, Greist J. 1990. The Wisconsin Personality
Disorder Inventory (Available from: M.H. Klein, Department of Psychiatry, University of
Wisconsin, School of Medicine, Madison, WI 53706).
Klerman GL. 1990. Depression and panic anxiety: the effect of depressive co- morbidity on
response to drug treatment of patients with panic disorder and agoraphobia. J Psychiatr Res
24 Suppl 2:27-41.
Klerman GL, Weissman MM, Ouellette R, Johnson J, Greenwald S. 1991. Panic attacks in
the community. Social morbidity and health care utilization. JAMA 265:742-746.
Lecrubier Y, Ustun TB. 1998. Panic and depression: a worldwide primary care perspective.
Int Clin Psychopharmacol 13 Suppl 4:S7-11.
Lelliott PT, Marks IM, Monteiro WO, Tsakiris F, Noshirvani H. 1987. Agoraphobics 5 years
after imipramine and exposure. Outcome and predictors. J Nerv Ment Dis 175:599-605.
Lenzenweger MF, Loranger AW, Korfine L, Neff C. 1997. Detecting personality disorders
in a nonclinical population. Application of a 2-stage procedure for case identification. Arch
Gen Psychiatry 54:345-351.
Lesser IM, Rubin RT, Pecknold J, Rifkin A, Swinson RP, Lydiard RB, Burrows GD, Noyes
R, Jr., DuPont R. 1988. Secondary depression in panic disorder and agoraphobia. I. Frequency, severity, and response to treatment. Arch Gen Psychiatry 45:437-443.
Liebowitz MR. 1989. Antidepressants in panic disorders. Br J Psychiatry Suppl 155:46-52.
Liebowitz MR, Fyer AJ, Gorman JM, Campeas R, Levin A, Davies SR, Goetz D, Klein DF.
1986. Alprazolam in the treatment of Panic Disorders. J Clin Psychopharmacol 6:13-20.
Lydiard RB. 1991. Coexisting depression and anxiety: special diagnostic and treatment
issues. J Clin Psychiatry 52 Suppl:48-54.
Lydiard RB, Lesser IM, Ballenger JC, Rubin RT, Laraia M, DuPont R. 1992. A fixed-dose
study of alprazolam 2 mg, alprazolam 6 mg, and placebo in panic disorder. J Clin Psychopharmacol 12:96-103.
Maddock RJ, Blacker KH. 1991. Response to treatment in panic disorder with associated
depression. Psychopathology 24:1-6.
Maier W, Buller R. 1988. One-year follow-up of panic disorder. Outcome and prognostic
factors. Eur Arch Psychiatry Clin Neurosci 238:105-109.
29
Maier W, Roth SM, Argyle N, Buller R, Lavori P, Brandon S, Benkert O. 1991. Avoidance
behaviour: a predictor of the efficacy of pharmacotherapy in panic disorder? Eur Arch
Psychiatry Clin Neurosci 241:151-158.
Marks IM, Mathews AM. 1979. Brief standard self-rating for phobic patients. Behav Res
Ther 17:263-267.
Marshall JR. 1996. Comorbidity and its effects on panic disorder. Bull Menninger Clin
60:A39-53.
Mavissakalian M, Hamann MS. 1986. DSM-III personality disorder in agoraphobia. Compr
Psychiatry 27:471-479.
Mavissakalian M, Hamann MS. 1987. DSM-III personality disorder in agoraphobia. II.
changes with treatment. Compr Psychiatry 28:356-361.
Mavissakalian M, Hamann MS. 1988. Correlates of DSM-III personality disorder in panic
disorder and agoraphobia. Compr Psychiatry 29:535-544.
Mavissakalian M, Michelson L. 1986. Agoraphobia: relative and combined effectiveness of
therapist-assisted in vivo exposure and imipramine. J Clin Psychiatry 47:117-122.
Modigh K. 1987. Antidepressant drugs in anxiety disorders. Acta Psychiatr Scand Suppl
335:57-74.
Nagy LM, Krystal JH, Woods SW, Charney DS. 1989. Clinical and medication outcome
after short-term alprazolam and behavioral group treatment in panic disorder. 2.5 year
naturalistic follow-up study. Arch Gen Psychiatry 46:993-999.
Noyes R, Jr., Clancy J, Woodman C, Holt CS, Suelzer M, Christiansen J, Anderson DJ.
1993. Environmental factors related to the outcome of panic disorder. J Nerv Ment Dis
181:529-538.
Noyes R, Jr., Garvey MJ, Cook BL. 1989. Follow-up study of patients with panic disorder
and agoraphobia with panic attacks treated with tricyclic antidepressants. J Affect Disord
16:249-257.
Noyes R, Jr., Reich J, Christiansen J, Suelzer M, Pfohl B, Coryell WA. 1990. Outcome of
panic disorder. Relationship to diagnostic subtypes and comorbidity. Arch Gen Psychiatry
47:809-818.
Noyes R, Jr., Reich JH, Suelzer M, Christiansen J. 1991. Personality traits associated with
panic disorder: change associated with treatment. Compr Psychiatry 32:283-294.
O'Rourke D, Fahy TJ, Brophy J, Prescott P. 1997. The Galway study of panic disorder III:
Outcome at 5 to 6 years. Br J Psychiatry 168:462-469.
30
Pelissolo A, Lepine JP. 1998. Epidemiology of depression and anxiety disorders. In:
Montgomery SA, den Boer JA, editors. SSRI's in depression and anxiety. New York: Wiley
& Sons Inc.
Pfohl B, Stangl D, Zimmerman M. 1982. The structured interview for DSM-III personality
disorders (SIDP). Iowa City, Iowa: University of Iowa Hospitals and Clinics.
Pollack MH, Otto MW, Rosenbaum JF, Sachs GS, O'Neil C, Asher R, Meltzer Brody S.
1990. Longitudinal course of panic disorder: findings from the Massachusetts General
Hospital Naturalistic Study. J Clin Psychiatry 51 Suppl A:12-16.
Pollack MH, Otto MW, Sachs GS, Leon A, Shear MK, Deltito JA, Keller MB, Rosenbaum
JF. 1994. Anxiety psychopathology predictive of outcome in patients with panic disorder
and depression treated with imipramine, alprazolam and placebo. J Affect Disord 30:273281.
Pollack MH, Otto MW, Tesar GE, Cohen LS, Meltzer Brody S, Rosenbaum JF. 1993. Longterm outcome after acute treatment with alprazolam or clonazepam for panic disorder. J
Clin Psychopharmacol 13:257-263.
Reich J, Noyes R, Jr., Troughton E. 1987. Dependent personality disorder associated with
phobic avoidance in patients with panic disorder. Am J Psychiatry 144:323-326.
Reich J, Yates W, Nduaguba M. 1989. Prevalence of DSM-III personality disorders in the
community. Soc Psychiatry Psychiatr Epidemiol 24:12-16.
Reich JH. 1988. DSM-III personality disorders and the outcome of treated panic disorder.
Am J Psychiatry 145:1149-1152.
Reich JH, Green AI. 1991. Effect of personality disorders on outcome of treatment. J Nerv
Ment Dis 179:74-82.
Reich JH, Vasile RG. 1993. Effect of personality disorders on the treatment outcome of axis
I conditions: an update. J Nerv Ment Dis 181:475-484.
Rickels K, Schweizer E. 1998. Panic disorder: Long-term pharmacotherapy and discontinuation. J Clin Psychopharmacol 18:12S-18S.
Rickels K, Schweizer E, Weiss S, Zavodnick S. 1993. Maintenance drug treatment for panic
disorder. II. Short- and long-term outcome after drug taper. Arch Gen Psychiatry 50:61-68.
Rosenberg R, Bech P, Mellergard M, Ottosson JO. 1991a. Alprazolam, imipramine and
placebo treatment of panic disorder: predicting therapeutic response. Acta Psychiatr Scand
Suppl 365:46-52.
Rosenberg R, Bech P, Mellergard M, Ottosson JO. 1991b. Secondary depression in panic
disorder: an indicator of severity with a weak effect on outcome in alprazolam and imipramine treatment. Acta Psychiatr Scand Suppl 365:39-45.
31
Roy Byrne PP, Cowley DS. 1994. Course and outcome in panic disorder: a review of recent
follow-up studies. Anxiety 1:151-160.
Roy Byrne PP, Stang P, Wittchen HU, Ustun B, Walters EE, Kessler RC. 2000. Lifetime
panic-depression comorbidity in the National Comorbidity Survey - Association with
symptoms, impairment, course and help-seeking. Br J Psychiatry 176:229-235.
Samuels JF, Nestadt G, Romanoski AJ, Folstein MF, McHugh PR. 1994. DSM-III personality disorders in the community. Am J Psychiatry 151:1055-1062.
Sanderson WC, DiNardo PA, Rapee RM, Barlow DH. 1990. Syndrome comorbidity in
patients diagnosed with a DSM-III-R anxiety disorder. J Abnorm Psychol 99:308-312.
Scheibe G, Albus M. 1996. Predictors of outcome in panic disorder: A 5-year prospective
follow- up study. J Affect Disord 41:111-116.
Scheibe G, Albus M. 1997. Predictors and outcome in panic disorder: A 2-year prospective
follow- up study. Psychopathology 30:177-184.
Scheibe G, Nutzinger D, Buller R, Walther AU. 1992. Pretreatment anxiety level as differential predictor in outpatients with panic disorder. Arzneimittelforschung 42:1090-1094.
Sharp DM, Power KG. 1999. Predicting treatment outcome for panic disorder and agoraphobia in primary care. Clinical Psychology and Psychotherapy 6:336-348.
Shear MK, Clark D, Feske U. 1998. The road to recovery in panic disorder: Response,
remission, and relapse. J Clin Psychiatry 59:4-10.
Shear MK, Maser JD. 1994. Standardized assessment for panic disorder research - A conference report. Arch Gen Psychiatry 51:346-354.
Shinoda N, Kodama K, Sakamoto T, Yamanouchi N, Takahashi T, Okada S, Noda S,
Komatsu N, Sato T. 1999. Predictors of 1-year outcome for patients with panic disorder.
Compr Psychiatry 40:39-43.
Slaap BR, van Vliet IM, Westenberg HGM, den Boer JA. 1995. Phobic symptoms as predictors of nonresponse to drug therapy in panic disorder patients (a preliminary report). J
Affect Disord 33:31-38.
Snaith RP, Baugh SJ, Clayden AD, Husain A, Sipple MA. 1982. The Clinical Anxiety Scale:
an instrument derived from the Hamilton Anxiety Scale. Br J Psychiatry 141:518-523.
Starcevic V, Kellner R, Uhlenhuth EH, Pathak D. 1993. The phenomenology of panic
attacks in panic disorder with and without agoraphobia. Compr Psychiatry 34:36-41.
Starcevic V, Uhlenhuth EH, Kellner R, Pathak D. 1992. Patterns of comorbidity in panic
disorder and agoraphobia. Psychiatry Res 42:171-183.
Steketee GS, Chambless DL. 1992. Methodological issues in prediction of treatment outcome. Clin Psychol Rev 12:387-400.
32
Taylor CB, King R, Margraf J, Ehlers A, Telch M, Roth WT, Agras WS. 1989. Use of
medication and in vivo exposure in volunteers for panic disorder research. Am J Psychiatry
146:1423-1426.
Toni C, Perugi G, Frare F, Mata B, Vitale B, Mengali F, Recchia M, Serra G, Akiskal HS.
2000. A prospective naturalistic study of 326 panic-agoraphobic patients treated with antidepressants. Pharmacopsychiatry 33:121-131.
Tyrer P, Shawcross C. 1988. Monoamine oxidase inhibitors in anxiety disorders. J Psychiatr
Res 22, suppl. 1:87-98.
van Vliet IM, den Boer JA, Westenberg HGM, Slaap BR. 1996. A double blind comparative
study of brofaromine and fluvoxamine in outpatients with panic disorder. J Clin
Psychopharmacol 16:299-306.
van Vliet IM, Westenberg HGM, den Boer JA. 1993. MAO-inhibitors in panic disorder:
clinical effects of treatment with brofaromine, a double blind placebo controlled study.
Psychopharmacology 112:483-489.
Warshaw MG, Massion AO, Shea MT, Allsworth J, Keller MB. 1997. Predictors of remission in patients with panic with and without agoraphobia: Prospective 5-year follow-up data.
J Nerv Ment Dis 185:517-519.
Weissman MM, Bland RC, Canino GJ, Faravelli C, Greenwald S, Hwu HG, Joyce PR,
Karam EG, Lee CK, Lellouch J, Lepine JP, Newman SC, Oakleybrowne MA, RubioStipec
M, Wells JE, Wickramaratne PJ, Wittchen HU, Yeh EK. 1997. The cross-national epidemiology of panic disorder. Arch Gen Psychiatry 54:305-309.
Westenberg HGM, den Boer JA. 1994. The neuropharmacology of anxiety; a review on the
role of serotonin. In: den Boer JA, Sitsen JMA, editors. Handbook of depression and
anxiety. New York: Marcel Dekker Inc. p 404-446.
Wittchen HU, Essau CA, Krieg JC. 1991. Anxiety disorders: similarities and differences of
comorbidity in treated and untreated groups. Br J Psychiatry Suppl23-33.
Woodman CL, Noyes R, Ballenger JC, Lydiard RB, Sievers G, Mihalko D. 1994. Predictors
of Response to Alprazolam and Placebo in Patients with Panic Disorder. J Affect Disord
30:5-13.
Work Group on Panic Disorder.American Psychiatric Association. 1998. Practice guideline
for the treatment of patients with panic disorder. Am J Psychiatry 155:1-34.
Zimmerman M, Coryell W. 1989. DSM-III personality disorder diagnoses in a nonpatient
sample. Demographic correlates and comorbidity. Arch Gen Psychiatry 46:682-689.
33
Table 1: An overview of short-term studies which investigated nonresponse to
pharmacotherapy in PD.
Study
Duration
n
Treatment
(weeks)
Liebowitz et al. (1986)
12
30
alprazolam
Mavissakalian and
12
62
imipramine or flooding, combined
Michelson (1986)
with programmed practice or
placebo
Mavissakalian and Hamann
16
33
antidepressants and exposure
Green and Curtis (1988)
8
30
alprazolam, imipramine or placebo
Lesser et al. (1988)
8
481
alprazolam or placebo
Reich (1988)
8
52
alprazolam or diazepam
Buller et al. (1991)
8
1168
alprazolam, imipramine or placebo
Maier et al. (1991)
8
1134
alprazolam, imipramine or placebo
Rosenberg et al. (1991a;
8
123
alprazolam, imipramine or placebo
Scheibe et al. (1992)
8
103
alprazolam, imipramine or placebo
Keller et al. (1993)
16
126
alprazolam, imipramine or placebo
Pollack et al. (1993)
6
59
alprazolam, clonazepam or placebo
Basoglu et al. (1994)
8
144
alprazolam or placebo, combined
(1987)
1991b)
with exposure or relaxation
Pollack et al. (1994)
16
126
alprazolam, imipramine or placebo
Woodman et al. (1994)
8
506
alprazolam or placebo
Ito et al. (1995)
8
33
imipramine, clomipramine or placebo
Slaap et al. (1995)
12
44
brofaromine or fluvoxamine
Hofmann et al. (1998)
12
93
imipramine or cognitive behavioral
therapy
Sharp and Power (1999)
12
149
fluvoxamine or placebo, alone or
combined with cognitive behavioral
therapy (CBT), CBT alone
34
Table 2: An overview of long-term studies which investigated nonresponse to
pharmacotherapy in PD.
Study
Duration
a
n
Treatment
53
antidepressants, alprazolam or
(years)
Faravelli and Albanesi
1
(1987)
benzodiazepines; psychotherapy if
needed
Lelliott et al. (1987)
5
40
imipramine or placebo, combined
with exposure or relaxation for 28
weeks
Maier and Buller (1988)
Nagy et al. (1989)
1
77
2.5
60
benzodiazepines or antidepressants
behavioral group therapy and
alprazolam for 16 weeks, alprazolam
if necessary during follow-up
Noyes, Jr. et al. (1989)
1-4
107
tricyclic antidepressants, sometimes
combined with a benzodiazepine
Noyes, Jr. et al. (1990)
3
89
alprazolam, diazepam, placebo for 8
weeks (up to 6 months)
Pollack et al. (1990)
2
100
benzodiazepines or antidepressants
Maddock and Blacker (1991)
1
30
naturalistic treatment
Albus and Scheibe (1993)
2
52
imipramine or doxepin for 8 weeks,
supportive psychotherapy for 6 to 8
months, as needed
Noyes, Jr. et al. (1993)
Pollack et al. (1993)
7
69
naturalistic treatment
1.5
59
alprazolam, clonazepam or placebo
for 6 weeks
Rickels et al. (1993)
1
48
alprazolam, imipramine or placebo
for 8 months
35
Table 2: continued
Study
Duration
a
n
Medication
50
imipramine or doxepin for 8 weeks,
(years)
Albus et al. (1995)
5
supportive psychotherapy for 6 to 8
months, as needed
Faravelli et al. (1995)
5
99
naturalistic treatment
Katschnig et al. (1995)
4
423
alprazolam, imipramine or placebo
for 8 weeks
O'Rourke et al. (1997)
5.3
68
lofepramine, clomipramine or
placebo
Scheibe and Albus (1996)
5
50
imipramine or doxepin for 8 weeks,
supportive psychotherapy for 6 to 8
months, as needed
Scheibe and Albus (1997)
2
53
antidepressants for 8 weeks,
supportive psychotherapy for 6 to 8
months, as needed
Warshaw et al. (1997)
5
412
naturalistic treatment
Shinoda et al. (1999)
1
65
naturalistic treatment
Toni et al. (2000)
3
326
naturalistic treatment
Legend: a In most studies a protocolized treatment is given initially. After this period
treatment is naturalistic, unless otherwise stated.
36
Table 3: Baseline clinical predictors of nonresponse to pharmacotherapy in panic
disorder a.
Predictor of nonresponse
Short-term treatment
Long-term treatment
(6 –16 weeks)
(1 - 7 years)
Demographic variables
Female gender
0/7
b
1/12
Higher age
0/7
1/11
Married
0/3
1/6
Longer duration of illness
2/9
10/15
Lower age at onset
0/3
4/10
Higher CGI score
2/5
1/3
Higher Anxiety score
5/9
5/9
Higher Phobic Avoidance score
7/10
13/15
Higher Panic score
5/9
4/11
Comorbid depression
2/6
6/11
Comorbid anxiety disorder
0/2
3/5
Comorbid personality disorder
4/5
4/5
Illness variables
Comorbidity
Legend: a: variables reported on in three or more studies; b: 0/7 means that this variable
was investigated in seven studies, but in none of the studies it was found to be a predictor of
nonresponse.
37