Expert Opinion on Pharmacotherapy

ISSN: 1465-6566 (Print) 1744-7666 (Online) Journal homepage: http://www.tandfonline.com/loi/ieop20

Treatment–resistant panic disorder: a systematic

review

Rafael C. Freire, Morena M. Zugliani, Rafael F. Garcia & Antonio E. Nardi

To cite this article: Rafael C. Freire, Morena M. Zugliani, Rafael F. Garcia & Antonio E.
Nardi (2016) Treatment–resistant panic disorder: a systematic review, Expert Opinion on
Pharmacotherapy, 17:2, 159-168, DOI: 10.1517/14656566.2016.1109628

To link to this article: https://doi.org/10.1517/14656566.2016.1109628

Published online: 04 Dec 2015.

Submit your article to this journal

Article views: 470

View related articles

View Crossmark data

Citing articles: 5 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ieop20
EXPERT OPINION ON PHARMACOTHERAPY, 2016
VOL. 17, NO. 2, 159–168
http://dx.doi.org/10.1517/14656566.2016.1109628

REVIEW

Treatment–resistant panic disorder: a systematic review

Rafael C. Freire , Morena M. Zugliani, Rafael F. Garcia and Antonio E. Nardi
Laboratory of Panic and Respiration, Institute of Psychiatry, Federal University of Rio de Janeiro and National Institute for Translational
Medicine (INCT-TM), Rio de Janeiro, Brazil

ABSTRACT ARTICLE HISTORY

Introduction: The prevalence of panic disorder (PD) in the population is high and these patients Received 3 March 2015
have work impairment, high unemployment rates, seek medical treatment more frequently and Accepted 8 October 2015
have more hospitalizations than people without panic symptoms. Despite the availability of KEYWORDS
pharmacological, psychological and combined treatments, approximately one-third of all PD Agoraphobia; anxiety;
patients have persistent panic attacks and other PD symptoms after treatment. clinical trial;
Areas covered: MEDLINE/Pubmed, CENTRAL, PsycINFO and Web of Science databases were cognitive-behavioral
searched for clinical trials in treatment–resistant PD. Only studies published between 1980 and therapy; panic attack
2015, in English, with human subjects, considered “journal articles” and clinical trial were
included. We included trials recruiting only adult subjects with treatment–resistant PD, consistent
with criteria from DSM-III to DSM5. We included all prospective experimental studies. Case, case
series, retrospective studies or studies with <10 PD subjects were not included.
Expert opinion: Only 11 articles were included in this review. There were few quality studies, only
two were randomized, controlled and double blind. Augmentation of the pharmacological
treatment with cognitive-behavioral therapy demonstrated some short-term efficacy in treat-
ment–resistant PD. There were also preliminary evidences of efficacy for monotherapy with
reboxetine and olanzapine, and augmentation with pindolol, divalproex sodium, aripiprazole
and olanzapine in short-term treatment.

1. Introduction responses. If there is a deficiency in the coordination of

stimuli from the cortex and brainstem, this could lead to
Anxiety disorders are very prevalent, with a 12-month
an abnormal activation of the amygdala, with behavioral,
prevalence ranging from 14.0 to 19.9% in the general
autonomic and neuroendocrine activation. Compounds
population.[1,2] Along with chronic pain and mood
that increase the transmission in the serotonin or γ-ami-
disorders, anxiety disorders are one of the most impor-
nobutyric-acid (GABA) systems have an inhibitory effect in
tant causes of disability and work impairment.[3–5] In
the amygdala and on the fear-related structures. Some
the general population the 12-month prevalence of
antidepressants increase noradrenergic activity and mod-
panic disorder (PD) ranges from 1.1 to 2.8% and for
ulate the noradrenaline releases related to stressful situa-
agoraphobia from 0.8 to 2.0%.[1,2,5,6] The lifetime pre-
tions. These drugs probably act in the central nucleus of
valence is even higher, 7.2 – 28.3% for panic attacks, 3.4
the amygdala and its projections by decreasing the sensi-
– 4.7% for panic disorder and 1.9% for agoraphobia.
tivity of the fear network and thus reducing the severity
[5,6] Both patients with clinical and patients with sub-
and frequency of panic attacks.[8,9]
threshold PD show significant work impairment, high
Guidelines, meta-analysis and other review studies
unemployment rates, seek medical treatment more fre-
indicate that several medications are effective in the
quently and have more hospitalizations than people
treatment of PD, including antidepressants, benzodia-
without panic attacks and other PD symptoms.[7]
zepines and other classes of drugs.[2,10–12] However,
It was hypothesized that panic attacks are acute anxi-
many PD patients do not respond to adequate treat-
ety and fear episodes with somatic symptoms that occur
ment with these drugs. In clinical trials with fluvoxa-
in subjects with abnormally sensitive fear networks, such
mine, sertraline, citalopram, escitalopram, paroxetine
as PD patients. In this network, the central nucleus of the
and clonazepam, with duration of 8 – 12 weeks, 17 –
amygdala gathers information from different brain
61% of the PD patients did not respond to treatment.
regions and coordinates autonomic and behavioral
[13,14] In clinical trials with venlafaxine, the response

CONTACT Rafael C. Freire rafaelcrfreire@gmail.com Laboratory of Panic and Respiration, Federal University of Rio de Janeiro, Rua Visconde de Piraja
330/906, Rio de Janeiro, Brazil
© 2015 Taylor & Francis
160 R. C. FREIRE ET AL.
Article highlights.
● There was only one double-blind, randomized, controlled trial
with positive results in treatment–resistant PD, the study of
augmentation with pindolol.
● CBT combined with antidepressants or anxiolytics demonstrated
some efficacy in treatment–resistant PD.
● There were 12 sessions of group CBT in the clinical trials with
this treatment.
● There was preliminary evidence of efficacy for the monotherapy
with reboxetine and olanzapine.
● There was also evidence of efficacy for the combined treatment
of pindolol, divalproex sodium, aripiprazole or olanzapine with
well-established antipanic drugs.
This box summarizes key points contained in the article.
rates were somewhat higher, from 68 to 89%, but from
49 to 64% of the patients did not achieve remission
after 10 – 12 weeks of treatment.[15] In the general
population, almost one-half of PD patients do not
achieve remission in the first year of treatment and
after 2 years of treatment more than one-third of the
patients still have PD symptoms.[16]
The objective of this systematic review is to summar-
ize and discuss the evidences regarding the treatment
of patients with treatment–resistant PD.
2. Methods
Articles were identified by a search of electronic
records, including the databases from MEDLINE/
Pubmed, the Cochrane Collaboration’s Clinical Trials
Register (CENTRAL), PsycINFO and Thomson Reuters’s
Web of Science. The search terms used were: “Panic
disorder” AND (“Treatment–resistant” OR “Treatment
resistance” OR “Pharmacotherapy-resistant” OR
“Pharmacotherapy resistance” OR “Medication-resistant”
OR “Medication resistance” OR “Drug-resistant” OR
“Drug resistance” OR “Refractory” OR “Augmentation”).
Only studies published in the years between 1980 and
2015, in English, with human subjects, considered “jour-
nal articles” and clinical trial were included. We
included trials recruiting only adult subjects with treat-
ment–resistant PD, consistent with criteria from DSM-III
to DSM5.[17–20] All definitions of treatment–resistance
were accepted. We included all prospective experimen-
tal studies including randomized-controlled trials,
quasi-random trials, crossover designs, and single arm
studies, blinded or open label. Case, case series, retro-
spective studies or studies with <10 PD subjects were
not included. Filters with this criteria were applied in
the databases were it was possible.
In the first step of the process, the first author (RCF)
screened the titles and abstracts of the articles and
manually excluded those that did not fit the criteria
mentioned above.
The second step was the study selection. Two inde-
pendent reviewers (RCF and MMZ) assessed the full-text
articles for eligibility, this assessment was made in a
standardized manner. The same reviewers made the
data extraction in an independent manner.
Disagreements between the reviewers regarding the
study selection or data extraction were resolved by
consensus.
As we included trials with different designs, the pre-
sence of a controlling condition, randomization and
subject and observer blindness to treatment were
used as measures of strength of evidence.
The ideal confirmatory format of this review would
consist of meta-analyzes of interaction statistics of pre-
dictors and moderators effects of large, high quality,
randomized controlled trials. Since most trials did not
meet quality requirements, and a broad spectrum of
experimental designs was included, we performed a
qualitative systematic review of the evidence.
3. Results
The searches of databases were conducted in January
of 2015 and yielded 18 articles in CENTRAL, 127 articles
in PsycINFO, 23 in MEDLINE/Pubmed and 128 in Web of
Science. The sum of articles after removing the dupli-
cates was 231, after screening 38 articles remained.
Reviewers examined the full-texts and only 11 articles
met the inclusion criteria. The process of study identifi-
cation and selection is shown in a Preferred Reporting
Items for Systematic Reviews and Meta-Analyses
(PRISMA) flow diagram (Figure 1).[21]
Only two articles were randomized controlled and
blinded clinical trials,[22,23] the other nine were open
studies. In four studies, the treatment was cognitive-
behavioral therapy (CBT), in six studies, the treatment
was pharmacological, and in one study, pharmacologi-
cal treatment and CBT were included. Two articles used
DSM-III-R [18] criteria for PD, all the other articles used
DSM-IV [19] criteria for this disorder. The number of
subjects in each study ranged from 10 to 71.
The criteria for treatment–resistant PD were very
heterogeneous. Some studies defined treatment–resis-
tance as not responding to at least one [24,25] or two
[23,26,27] adequate 8-week treatment trials with drugs
recognized as effective for PD or standard course of
CBT. Other studies used the treatment–resistance cri-
teria of not responding to at least one [22] or two [28]
6-week adequate treatment trials. Incomplete response
after 3 – 4 months of treatment defined treatment–
resistance in four studies.[29–32]
 EXPERT OPINION ON PHARMACOTHERAPY 161

Identification
Records identified through database Additional records identified through
searching other sources
(n = 306) (n = 0)

Records after duplicates removed

(n = 231)
Screening

Records screened Records excluded

(n = 231) (n = 193)

Full-text articles assessed for Full-text articles excluded

Eligibility

eligibility
(n = 38) Not clinical trial 09
Less than 10 PD 07
patients
Not treatment- 07
resistant patients
Retrospective study 02
Full-text not found 01
Not PD patients 01
Included

Studies included in
Total
qualitative synthesis
(n = 27)
(n = 11)

Figure 1. PRISMA diagram of study identification and selection process.

The article from Simon et al. [22] included three
clinical trials, phases I, II and III. Phase I was not with
treatment–resistant PD and was not included in the
review, phases II and III were treated as independent
trials.
The studied drugs were divalproex sodium, reboxe-
tine, pindolol, aripiprazole, olanzapine, sertraline, esci-
talopram and clonazepam. Among the 12 clinical trials,
the studied treatment was an augmentation of another
treatment in nine of them. The trial duration ranged
from 4 to 16 weeks. There were no significant differ-
ences between low and high dose of sertraline or esci-
talopram,[22] otherwise all studied treatments were
considered effective.
There were several limitations regarding the ade-
quacy of treatment including doses and duration of
trials. There were trials with no augmentation, augmen-
tation in part of the patients or in all patients.
Frequently the augmentation strategy varied across
subjects, possibly including bias. In few studies the
sample size was too small. The quality evaluation and
list of limitations were summarized in Table 1.
3.1. Randomized controlled double-blind trials
In the study from Hirschmann et al. [23] pindolol
7.5 mg/day or placebo were added to fluoxetine
20 mg/day in a 4-week clinical trial. Patients on pindolol
and fluoxetine had significant improvements in the
Panic Self-questionnaire (PSQ) (Cohen’s d = 2.72) and
Clinical Global Impression – Improvement scale (CGI-I)
(Cohen’s d = 4.00), compared to patients taking fluox-
etine alone. There were significant improvements in
other panic and anxiety scales, but no difference
regarding depression symptoms. There were already
significant improvements in anxiety symptoms after
2 weeks of treatment.
 162
R. C. FREIRE ET AL.

Table 1. Evaluation of study quality and limitations.

Study TRPD criteria Type Treatment design Analysis design Limitations
Baetz and Bowen 1998 [24] ++ OS ++ + Non-responders to one previous trial were considered treatment–resistant; concurrent use of
benzodiazepines and antidepressants is some of the patients; small sample size
Dannon et al. 2002 [26] +++ OS ++ ++ Short washout before study; short duration of trial and small sample size
Heldt et al. 2003 [32] ++ OS ++ ++ The treatment–resistance criterion was incomplete remission after 4 months of treatment, just one
previous trial; concurrent use of antianxiety or antidepressant medications; small sample size
Heldt et al. 2006(1) [31] ++ OS ++ +++ The treatment–resistance criterion was incomplete remission after 4 months of treatment, just one
previous trial; concurrent use of antianxiety or antidepressant medications
Heldt et al. 2006(2) [30] ++ OS ++ +++ The treatment–resistance criterion was incomplete remission after 4 months of treatment, just one
previous trial; concurrent use of antianxiety, antidepressant medications or lithium
Hirschmann et al. 2000 [23] +++ RCDBCT + ++ Low dose of medication (fluoxetine) and short duration of the trial; small sample size
Hoge et al. 2008 [25] ++ OS ++ + Non-responders to one previous trial were considered treatment–resistant; some patients had
comorbidy with GAD; concurrent use of antidepressant, benzodiazepine or both; very small sample size
Hollifield et al. 2005 [27] +++ OS +++ + Very small sample size
Pollack et al. 1994 [29] ++ OS ++ ++ Treatment resistance was defined by incomplete response after adequate 12-week trial with
medication; concurrent use of clonazepam, alprazolam, fluoxetine, TCA or lithium; very small sample
size
Sepede et al. 2006 [28] +++ OS ++ ++ Concurrent use of antidepressants; small sample size
Simon et al. 2009 – phase II [22] + RCDBCT ++ ++ Treatment–resistance was defined by no response after one 6-week trial with SSRI; short trial duration;
small sample size
Simon et al. 2009 – phase III [22] ++ HHRCT ++ + Treatment–resistance was defined by no response after one 12-week trial with SSRI; concurrent use of
escitalopram or sertraline; very small sample size
GAD: Generalized anxiety disorder; HHRCT: Head-to-head randomized clinical trial; OS: Open study; RCDBCT: Randomized controlled double-blind clinical trial; SSRI: Serotonin selective reuptake inhibitor; TCA: tricyclic
antidepressant; TRPD: Treatment–resistant panic disorder.
Treatment design – adequacy of dose, drug, trial duration and augmentation strategy.
Analysis design – adequacy of scales, sample size, statistical analysis and power.
 EXPERT OPINION ON PHARMACOTHERAPY 163

Table 2. Randomized, controlled and double-blind clinical trials.

Number of Trial
patients duration
Trial TRPD criteria (completed) Drug Dose (in weeks) Outcome Other information
Hirschmann et al. 2000 [23] Previous treatment 26 (25) Fluoxetine + Fluoxetine 4 Superior to PBO; The clinical changes
with at least two pindolol or 20 mg/day; significant differences noted with pindolol
antidepressants fluoxetine + pindolol in PSQ (ES: 2.72), were evident by the
and no response PBO 7.5 mg/day CGI-I (ES: 4.00), second week of the
after a 8-week trial HAM-A, CAS+PA and study
with fluoxetine NIMH Anxiety Scale *
20 mg/day
Simon et al. 2009 – phase II No response after 24 (19) Sertraline or High dose – 6 Not superior to PBO,
[22] 6-week trial with escitalopram sertraline 150 no significant
sertraline (up to high dose or – 200 mg/ differences in CGI-S
100 mg/day) or PBO (no dose day; (ES: 0.16) and PDSS
escitalopram (up increase) escitalopram (ES: 0.01); significant
to 15 mg/day) 20 – 30 mg/ improvement in PDSS
day in both groups **
CAS + PA: Clinical Anxiety Scale with panic attacks; CGI-I: Clinical Global Impression – Improvement; ES: effect size compared to placebo, Cohen’s d; HAM-A:
Hamilton Rating Scale for Anxiety; NIMH: National Institute of Mental Health; PBO: Placebo; PDSS: Panic Disorder Severity Scale; PSQ: Panic Self-
Questionnaire; TRPD: Treatment–resistant panic disorder.
* Response and remission were not defined in this study.
** Remission status was defined as zero panic attacks for at least 1 week and a Clinical Global Impression – Severity score of 1 or 2.

In the study from Simon et al. [22] patients who
tolerated sertraline 100 mg/day or escitalopram
15 mg/day but did not achieve remission after
6 weeks of treatment with these drugs were included
in next phase of this study. In the second phase of the
same study, patients were randomized in double-blind
fashion either to receive a higher dose of selective
serotonin reuptake inhibitor (SSRI) or to remain in the
same dose. Both groups had a decrease in the Panic
Disorder Severity Scale (PDSS) scores, there were no
differences regarding other anxiety and depression
scales. The high SSRI dose group failed to show differ-
ences compared with the low dose group.
Findings from these studies are summarized in Table 2.
3.2. Open pharmacological trials
Among the five open pharmacological trials, three of
them were augmentation studies and there was con-
current use of antidepressants or benzodiazepines.
[24,25,28] Patients treated with aripiprazole, olanzapine
and divalproex sodium had significant improvements
regarding the PD symptoms. In the study from Baetz
et al. [24] only patients with comorbid mood disorder
and self-reported “mood instability” were included.
Besides the improvement of anxiety and panic symp-
toms, there was also improvement of depressive and
mood instability symptoms.
Washout of previous medications was performed in
two open studies.[26,27] The noradrenaline reuptake inhi-
bitor (NRI) Reboxetine 2 – 8 mg/day and the antipsychotic
olanzapine 2.5 – 20 mg/day were effective in treatment–
resistant PD with no concurrent medications. There was
significant improvement in the panic attacks, anticipatory
anxiety agoraphobia, general anxiety and impairment.
Findings from these studies are summarized in
Table 3.
3.3. Cognitive-behavioral therapy studies
Four CBT trials [29–32] did not include comparisons to
placebo, wait list or other treatments. In one rando-
mized trial,[22] CBT and clonazepam were compared
as augmentation to SSRI and both treatments were
equally effective. All studies had limitations regarding
the treatment–resistance criteria and concurrent use of
multiple pharmacological agents such as antidepres-
sants, benzodiazepines and lithium. (Table 1)
All protocols [22,29–32] included 12 sessions of
group therapy. CBT was effective as an augmentation
in all five trials, with improvements in panic attacks,
anticipatory anxiety, agoraphobia, other panic symp-
toms, general anxiety and quality of life. A decrease in
the use of medications was also observed.[31,32] The
improvement of panic symptoms persisted for at least a
couple of months after the end of the CBT.[29] Findings
from these studies are summarized in Table 4.
4. Conclusion
Studies regarding the treatment of treatment–resistant
PD were scant, quality studies regarding this subject
were exiguous. Regarding the pharmacological treat-
ment there was preliminary evidence of efficacy in
treatment–resistant PD for monotherapy with reboxe-
tine and olanzapine. The augmentation of antidepres-
sants, anxiolytics and other drugs with CBT, pindolol,
 164

Table 3. Open pharmacological studies.

Number of
patients Augmentation Trial duration
Trial TRPD criteria (completed) study? Drug Dose (in weeks) Outcome Other information
Baetz and Bowen 1998* [24] Not respond to one trial with 13(10) Yes. 40% of Divalproex sodium Minimum 500 mg/ 8 Significant improvement in PA
CBT and medication subjects using day; serum level (65% decrease in frequency), HAM-
R. C. FREIRE ET AL.

(benzodiazepines or concurrent 45 – 90 μg/ml A, BAI, PGI (70% much or very

antidepressants) benzodiazepines, much improved) and CGI-I (90%
antidepressants much or very much improved)
or both scores **
Dannon et al. 2002 [26] Not respond to two adequate 29(24) No Reboxetine 2 – 8 mg/day 6 Significant improvement in PSQ Short washout
trials with SSRI (paroxetine (90% decrease), HAM-A and GAF (paroxetine – 4
40 mg/day and fluoxetine scores in patients with and without days; fluoxetine –
40 mg/day, duration 8.4 ± 4.3 agoraphobia ** 21 days) before
weeks) study
Hoge et al. 2008 [25] Not respond to one 8-week 10(7) Yes. Concurrent Aripiprazole 5 – 30 mg/day 8 Significant improvement in CGI-S Significant
adequate trial (SSRI, SNRI, use of scores (17% decrease); no improvement in
benzodiazepines, trazodone or antidepressant, difference in PDSS scores (14% CGI-S and HAM-A
bupropion) benzodiazepine decrease); 10% achieved remission in the combined
or both *** group (PD and
GAD)
Hollifield et al. 2005 [27] Not respond to two 8-week 15(10) No Olanzapine 2.5 – 20 mg/day 8 Significant improvement in PAI –
adequate trials (medication or (average 12.3 mg/ PA frequency (ES: 1.83), intensity
CBT) day) (ES: 1.36) and anticipatory anxiety
(ES: 1.18); significant improvement
in PPPDS – anticipatory anxiety (ES:
3.65), phobic avoidance (ES: 2.59)
and impairment (ES: 2.18);
significant improvement in CGI-S
scores (ES: 2.61) **
Sepede et al. 2006 [28] Not responded previously to 31(26) Yes. Concurrent Olanzapine 5.0 mg/day 12 Significant improvement in PAAAS Improvements in
adequate therapeutic trials of use of paroxetine – PA frequency (ES: 2.36), PA HAM-D and GAF
either medication or CBT and 40 mg/day or duration (ES: 3.35) and anticipatory scales; significant
not responded to current sertraline anxiety (ES: 3.74); significant and similar
treatment (at least 6 weeks with 150 mg/day improvement in ACQ and HAM-A improvement in
paroxetine 40 mg/day or scores; response 81%; remission patients with or
sertraline 150 mg/day) 58% **** without
agoraphobia
ACQ: Agoraphobic Cognitions Questionnaire; BAI: Beck Anxiety Inventory; CBT: Cognitive-behavioral therapy; CGI-I: Clinical Global Impression – Improvement; CGI-S: Clinical Global Impression – Severity; ES: Effect size
compared to baseline scores, Cohen’s d; GAD: Generalized anxiety disorder; GAF: Global Assessment of Functioning; HAM-A: Hamilton Rating Scale for Anxiety; HAM-D: Hamilton Rating Scale for Depression; PA: Panic
attack; PAAAS: Panic Attack and Anticipatory Anxiety Scale; PAI: Panic Attack Inventory; PDSS: Panic Disorder Severity Scale; PGI: Patient Global Impression; PPPDS: Physicians Panic and Phobic Disorders Scale; PSQ:
Panic Self-Questionnaire; SSRI: Selective serotonin reuptake inhibitor; TRPD: Treatment–resistant panic disorder.
* Subsample of patients with comorbid mood disorder and self-reported “mood instability”.
** Response and remission were not defined in this study.
*** Remission criteria CGI-S score of 1 or 2.
**** Response criteria: 50% decrease in number of panic attacks and CGI-I score of 1 or 2; remission criteria: no panic attacks and HAM-A score ≤7.
Table 4. Cognitive-behavioral therapy studies.
Number of
patients Augmentation Number/type of Trial duration Other
Trial TRPD criteria (completed) study? sessions Comparison (in weeks) Outcome information
Heldt et al. 2003 [32] Residual symptoms of 32(29) Yes. Concurrent 12/group sessions No 16 Significant improvement in CGI-S (ES: 1.5), Decrease in use
panic attacks, anticipatory use of antianxiety PA frequency (ES: 1.0), PA intensity (ES: of
anxiety and phobic or antidepressant 1.0), agoraphobia (PI) (ES: 1.5), benzodiazepines
avoidance after adequate medications anticipatory anxiety (PI) (ES: 1.0) and HAM-
trial with SSRI; on SSRI for A scores (ES: 1.3); No PA 81%
4 months
Heldt et al. 2006(1) [32] Residual symptoms of 71(64) Yes. Concurrent 12/group sessions No 16 Significant improvement in CGI-S (ES: 2.2), Decrease in use
panic attacks, anticipatory use of antianxiety PA (ES: 0.6), agoraphobia (ES: 1.3) (PI), of
anxiety and phobic or antidepressant anticipatory anxiety (ES: 1.1) (PI) and HAM- benzodiazepines
avoidance after adequate medications A scores (ES: 1.3); no PA 81%; remission and
trial with SSRI; on SSRI for 64% * antidepressants
4 months
Heldt et al. 2006(2) [30] Residual symptoms of 36(32) Yes. Concurrent 12/group sessions No 16 Significant improvement in CGI-S (ES: 2.3),
panic attacks, anticipatory use of PA (ES: 0.5), agoraphobia (ES: 1.0) (PI),
anxiety and phobic antianxiety, anticipatory anxiety (ES: 0.9)(PI), HAM-A
avoidance after adequate antidepressant (ES: 1.1) and quality of life scores (ES: 0.9)
trial with SSRI; on SSRI for medications or (WHOQOL); no PA 75%; remission 44% **
4 months lithium
Pollack et al. 1994 [29] Incomplete response after 18(15) Yes. Concurrent 12/group sessions No 12 Significant improvement in PA (ES: 0.5) Persistent
adequate 12-week trial use of and CGI-S (ES: 1.6); remission 40% *** improvement
with medication clonazepam, after 1 – 8
(imipramine, fluoxetine, alprazolam, months of
phenelzine, clonazepam or fluoxetine, TCA follow-up
alprazolam) or lithium
Simon et al. 2009 – phase III [22] No response after 12-week 19(17) Yes. Concurrent 12/group sessions CLZ 0.5 – 12 No significant difference between the two
trial with sertraline (up to use of 4.0 mg/day treatments; significant improvement in
200 mg/day) or escitalopram CGI-S (ES: 1.0) and PDSS (ES: 0.6);
escitalopram (up to 30 mg/ (15–30 mg/day) remission CLZ 11%, CBT 10% **
day) or sertraline (100
– 200 mg/day)
CBT: Cognitive-behavioral therapy; CGI-S: Clinical Global Impression – Severity; CLZ: Clonazepam; ES: Effect size compared to baseline scores, Cohen’s d; HAM-A: Hamilton Rating Scale for Anxiety; PA: Panic attack; PDSS:
Panic Disorder Severity Scale; PI: Panic Inventory; SSRI: Selective serotonin reuptake inhibitor; TCA: Tricyclic antidepressant; TRPD: Treatment–resistant panic disorder; WHOQOL: World Health Organization Quality of Life
Instrument – Short Version.
* Remission was defined by no panic attacks and CGI-S < 3 for at least 2 months (evaluation made 12 months after the end of the treatment).
** Remission was defined as the absence of PA and CGI-S < 3.
*** Remission was defined as CGI-S < 3.
EXPERT OPINION ON PHARMACOTHERAPY
165
166 R. C. FREIRE ET AL.
divalproex sodium, aripiprazole and olanzapine demon-
strated some efficacy. Dose escalation after initial lack
of response produced little improvement or no
improvement at all. More randomized, controlled and
blind clinical trials are needed to ascertain adequate
strategies for treatment–resistant PD.
5. Expert opinion
The lifetime prevalence of PD in the population is high,
[5,6] among these patients approximately 30% of them
have persistent panic symptoms despite pharmacologi-
cal and psychological treatment.[16,33] Improving the
treatments for PD and finding alternatives for patients
who do not respond to conventional treatments is a
matter of public health with crucial importance.
There is no consensus about the definition of treat-
ment–resistant PD and currently there are no opera-
tional criteria for classifying these patients. We opted
to accept all criteria for treatment–resistance, otherwise
very few articles would be included in this review. The
authors believe that the best criteria for treatment–
resistant PD were those used by Dannon et al. [26]
and Hollifield et al. [27]: not responding to at least
two adequate 8-week treatment trials with drugs recog-
nized as effective for PD in adequate doses or standard
course of CBT. Patients who have poor response to the
first medication trial could respond very well in a sec-
ond trial. In addition, patients who take small doses of
medication and have poor response after 6 weeks of
treatment could improve if the doses were increased
and the trial was prolonged. These patients cannot be
considered treatment–resistant.
Treatment–resistance in PD may be the consequence
of untreated comorbid disorders and patients could
benefit from treatments directed to the causes.
Patients with comorbid bipolar spectrum disorders
would probably benefit from mood stabilizers and anti-
psychotics. Switching the antidepressant – including
tricyclic antidepressants (TCA) and monoamine oxidase
inhibitors (MAOI) in the treatment options – or combin-
ing two antidepressants should be a good strategy for
patients with comorbid major depressive disorder.
Psychotherapy would probably be the best augmenta-
tion strategy for patients with treatment–resistant PD
and comorbid avoidant or dependent personality dis-
orders, agoraphobia or other phobias.
Antidepressants, particularly SSRI and serotonin, and
noradrenaline reuptake inhibitors (SNRI), are highly
recommended by current guidelines [2,11,12] among
the pharmacological agents used in the treatment of PD.
Due to their high efficacy and low risk of adverse reactions
the SSRI escitalopram, citalopram, sertraline, fluoxetine,
paroxetine, fluvoxamine and the SNRI venlafaxine have
the highest recommendation grades. The NRI reboxetine,
the MAOI phenelzine, and the TCA clomipramine, desi-
pramine, imipramine and lofepramine are efficacious, but
they have less favorable side-effects profile. Reboxetine,
moclobemide, duloxetine, milnacipran, nefazodone and
mirtazapine also have some evidence of efficacy in the
treatment of PD.[2,10] There is only one open study with
antidepressants for treatment–resistant PD patients and it
showed that these patients may benefit from augmenta-
tion with reboxetine.[26] Other strategies for treatment–
resistant PD with antidepressants such as switching the
antidepressant to TCA or MAOI, or combining two anti-
depressants are used in clinical practice but have not
been systematically studied. There is preliminary evidence
indicating that escitalopram and venlafaxine are more
effective in the treatment of PD than citalopram and
paroxetine, respectively.[2,13] These evidences make
these antidepressants good candidates as drugs to treat
treatment–resistant PD, but there are still no clinical trials
with these drugs for treatment–resistant PD.
The benzodiazepines alprazolam, clonazepam, diaze-
pam and lorazepam are efficacious in acute treatment of
PD, but have lower recommendation grades due to side
effects and risks.[2,11,12] Despite the fact that benzodiaze-
pines are frequently combined to antidepressants in clin-
ical practice, there are no trials with benzodiazepines as
monotherapy or augmentation for treatment–resistant PD.
Atypical antipsychotics are not recommended by the
current guidelines [2,11,12] as first-line agents for the
treatment of PD, however, aripiprazole, olanzapine, ris-
peridone and sulpiride have demonstrated some efficacy
in this disorder.[10] Quetiapine showed anxiolytic prop-
erties in several studies and is used in mood disorders
with good results,[34,35] for this reason quetiapine is
also a promising agent for the treatment of PD and
treatment–resistant PD. The risks and benefits of atypical
antipsychotics should be weighted in treatment–resis-
tant PD patients because these agents are associated
to increased appetite, weight gain, metabolic abnormal-
ities, amenorrhea/galactorrhea, somnolence, sialorrhea,
sexual dysfunction, somnolence, blurred vision, head-
ache, dizziness, akathisia, insomnia, tremor and other
side effects.[10,36] Increased appetite, weight gain,
metabolic abnormalities and sexual dysfunction are asso-
ciated with both antidepressants and atypical antipsy-
chotics, combining these two compounds could increase
the risk of the mentioned side effects.
According to the guidelines the best treatments for
PD are pharmacological, psychological or a combina-
tion of both.[2] Among psychological treatments for PD
and treatment–resistant PD, CBT has the highest level
of evidence. In most protocols, there are 8 – 12 sessions

once in a week, but shorter CBT programs with only five
sessions may also be effective if combined to the CBT
adjuvant D-cycloserine.[10] The CBT adjuvants are also
promising compounds for treatment–resistant PD.
The study of repetitive transcranial magnetic stimu-
lation and transcranial direct current stimulation had
significant developments in the last decade,[37,38]
and in the future, they may become effective treat-
ments for PD, including treatment–resistant PD.
Currently there are very few low quality studies
addressing the treatment–resistant PD, most of them
are open studies with very small samples of PD
patients. Well-designed, double-blind, randomized and
controlled studies with sufficient number of subjects
are needed to shed light on this subject.
Declaration of interest
Funding for this study was provided by the Brazilian Council
for Scientific and Technological Development (CNPq). RC
Freire has received support from the Brazilian Council for
Scientific and Technological Development (CNPq). A Nardi
has received support from the Brazilian Council for Scientific
and Technological Development (CNPq). The authors have no
other relevant affiliations or financial involvement with any
other organization or entity with a financial interest in or
financial conflict with the subject matter or materials dis-
cussed in the manuscript apart from those disclosed.
ORCID
Rafael C. Freire http://orcid.org/0000-0003-3875-4601
Bibliography
Papers of special note have been highlighted as either of
interest (*) or of considerable interest (**) to readers.
1. Andrade LH, Wang YP, Andreoni S, et al. Mental disorders
in megacities: findings from the Sao Paulo megacity
mental health survey, Brazil. PloS One. 2012;7(2):e31879.
2. Baldwin DS, Anderson IM, Nutt DJ, et al. Evidence-based
pharmacological treatment of anxiety disorders, post-trau-
matic stress disorder and obsessive-compulsive disorder: a
revision of the 2005 guidelines from the British
Association for Psychopharmacology. J Psychopharmacol.
2014;28(5):403–439.
3. Andrade LH, Baptista MC, Alonso J, et al. Days out-of-role
due to common physical and mental health problems:
results from the Sao Paulo Megacity Mental Health
Survey, Brazil. Clinics. 2013;68(11):1392–1399.
4. Lepine JP. Epidemiology, burden, and disability in depres-
sion and anxiety. J Clin Psychiatry. 2001;62(Suppl 13):4–10.
5. Roy-Byrne PP, Stang P, Wittchen HU, et al. Lifetime panic-
depression comorbidity in the National Comorbidity
Survey. Association with symptoms, impairment, course
and help-seeking. Br J Psychiatry. 2000;176:229–235.
6. Kessler RC, Chiu WT, Jin R, et al. The epidemiology of
panic attacks, panic disorder, and agoraphobia in the
EXPERT OPINION ON PHARMACOTHERAPY 167
National Comorbidity Survey Replication. Arch Gen
Psychiatry. 2006;63(4):415–424.
7. Bystritsky A, Kerwin L, Niv N, et al. Clinical and subthres-
hold panic disorder. Depress Anxiety. 2010;27(4):381–389.
8. Gorman JM, Kent JM, Sullivan GM, et al. Neuroanatomical
hypothesis of panic disorder, revised. Am J Psychiatry.
2000;157(4):493–505.
9. Nardi AE, Freire RC, Zin WA. Panic disorder and control of
breathing. Respir Physiol Neurobiol. 2009;167(1):133–143.
10. Freire RC, Machado S, Arias-Carrion O, et al. Current
pharmacological interventions in panic disorder. CNS
Neurol Disord Drug Targets. 2014;13(6):1057–1065.
11. Bandelow B, Zohar J, Hollander E, et al. World Federation
of Societies of Biological Psychiatry (WFSBP) guidelines
for the pharmacological treatment of anxiety, obsessive-
compulsive and post-traumatic stress disorders - first
revision. World J Biol Psychiatry. 2008;9(4):248–312.
12. Katzman MA, Bleau P, Blier P, et al. Canadian clinical
practice guidelines for the management of anxiety, post-
traumatic stress and obsessive-compulsive disorders.
BMC Psychiatry. 2014;14(Suppl 1):S1.
13. Freire RC, Hallak JE, Crippa JA, et al. New treatment
options for panic disorder: clinical trials from 2000 to
2010. Expert Opin Pharmacother. 2011;12(9):1419–1428.
14. Bandelow B, Behnke K, Lenoir S, et al. Sertraline versus
paroxetine in the treatment of panic disorder: an acute,
double-blind noninferiority comparison. J Clin Psychiatry.
2004;65(3):405–413.
15. Kjernisted K, McIntosh D. Venlafaxine extended release
(XR) in the treatment of panic disorder. Ther Clin Risk
Manag. 2007;3(1):59–69.
16. Batelaan NM, De Graaf R, Penninx BW, et al. The 2-year
prognosis of panic episodes in the general population.
Psychol Med. 2010;40(1):147–157.
17. American Psychiatric Association. Diagnostic and statisti-
cal manual of mental disorders: DSM-III. 3d ed.
Washington (DC): American Psychiatric Association; 1980.
18. American Psychiatric Association. Diagnostic and statisti-
cal manual of mental disorders: DSM-III-R. 3rd ed.
Washington (DC): American Psychiatric Association; 1987.
19. American Psychiatric Association. Diagnostic and statisti-
cal manual of mental disorders: DSM-IV. 4 ed.
Washington (DC): American Psychiatric Press; 1994.
20. American Psychiatric Association. Diagnostic and statisti-
cal manual of mental disorders: DSM-5. 5th ed.
Washington (DC): American Psychiatric Association; 2013.
21. Moher D, Liberati A, Tetzlaff J, et al. Preferred report-
ing items for systematic reviews and meta-analyses:
the PRISMA statement. PLoS Med. 2009;6(7):e1000097.
22. Simon NM, Otto MW, Worthington JJ, et al. Next-step
strategies for panic disorder refractory to initial pharma-
cotherapy: a 3-phase randomized clinical trial. J Clin
Psychiatry. 2009;70(11):1563–1570.
** One of the two randomized controlled double-blind
clinical trials for treatment–resistant panic disorder
(phase II). Also includes the only head-to-head rando-
mized clinical trial for treatment resistant panic dis-
order (phase III).
23. Hirschmann S, Dannon PN, Iancu I, et al. Pindolol augmen-
tation in patients with treatment-resistant panic disorder:
a double-blind, placebo-controlled trial. J Clin
Psychopharmacol. 2000;20(5):556–559.
168 R. C. FREIRE ET AL.
* One of the two randomized controlled double-blind
clinical trials for treatment–resistant panic disorder.
24. Baetz M, Bowen RC. Efficacy of divalproex sodium in
patients with panic disorder and mood instability who
have not responded to conventional therapy. Can J
Psychiatry. 1998;43(1):73–77.
25. Hoge EA, Worthington III JJ, Kaufman RE, et al.
Aripiprazole as augmentation treatment of refractory
generalized anxiety disorder and panic disorder. CNS
Spectr. 2008;13(6):522–527.
26. Dannon PN, Iancu I, Grunhaus L. The efficacy of rebox-
etine in the treatment-refractory patients with panic dis-
order: an open-label study. Hum Psychopharmacol.
2002;17(7):329–333.
27. Hollifield M, Thompson PM, Ruiz JE, et al. Potential effec-
tiveness and safety of olanzapine in refractory panic
disorder. Depress Anxiety. 2005;21:33–40.
28. Sepede G, De Berardis D, Gambi F, et al. Olanzapine
augmentation in treatment-resistant panic disorder: a
12-week, fixed-dose, open-label trial. J Clin
Psychopharmacol. 2006;26(1):45–49.
29. Pollack MH, Otto MW, Kaspi SP, et al. Cognitive behavior
therapy for treatment-refractory panic disorder. J Clin
Psychiatry. 1994;55(5):200–205.
30. Heldt E, Blaya C, Isolan L, et al. Quality of life and treat-
ment outcome in panic disorder: cognitive behavior
group therapy effects in patients refractory to medication
treatment. Psychother Psychosom. 2006;75(3):183–186.
31. Heldt E, Manfro GG, Kipper L, et al. One-year follow-up of
pharmacotherapy-resistant patients with panic disorder
treated with cognitive-behavior therapy: outcome and pre-
dictors of remission. Behav Res Ther. 2006;44(5):657–665.
32. Heldt E, Manfro GG, Kipper L, et al. Treating medication-
resistant panic disorder: predictors and outcome of cog-
nitive-behavior therapy in a Brazilian public hospital.
Psychother Psychosom. 2003;72(1):43–48.
33. Bandelow B, Rüther E. Treatment-resistant panic disor-
der. CNS Spectr. 2004;9(10):725–739.
34. Gao K, Wu R, Kemp DE, et al. Efficacy and safety of
quetiapine-XR as monotherapy or adjunctive therapy to
a mood stabilizer in acute bipolar depression with gen-
eralized anxiety disorder and other comorbidities: a ran-
domized, placebo-controlled trial. J Clin Psychiatry.
2014;75(10):1062–1068.
35. Chen YC, Chen CK, Wang LJ. Quetiapine fumarate aug-
mentation for patients with a primary anxiety disorder or
a mood disorder: a pilot study. BMC Psychiatry.
2012;12:162.
36. Hershenberg R, Gros DF, Brawman-Mintzer O. Role of
atypical antipsychotics in the treatment of generalized
anxiety disorder. CNS Drugs. 2014;28(6):519–533.
37. George MS, Padberg F, Schlaepfer TE, et al. Controversy:
repetitive transcranial magnetic stimulation or transcra-
nial direct current stimulation shows efficacy in treating
psychiatric diseases (depression, mania, schizophrenia,
obsessive-complusive disorder, panic, posttraumatic
stress disorder). Brain Stimul. 2009;2(1):14–21.
38. Cristancho MA, Cristancho P, O’Reardon JP. Other ther-
apeutic psychiatric uses of superficial brain stimulation.
Handb Clin Neurol. 2013;116:415–422.