Journal of Psychiatric Research 47 (2013) 168e174

Contents lists available at SciVerse ScienceDirect

Journal of Psychiatric Research

journal homepage: [Link]/locate/psychires

Augmentation of exposure therapy with post-session administration

of D-cycloserine
Candyce D. Tart a, *, Pamela R. Handelsman a, Lindsey B. DeBoer a, David Rosenfield a, Mark H. Pollack c,
Stefan G. Hofmann b, Mark B. Powers a, Michael W. Otto b, Jasper A.J. Smits a
a
Department of Psychology, Southern Methodist University, USA
b
Department of Psychology, Boston University, USA
c
Rush University Medical Center, USA

a r t i c l e i n f o a b s t r a c t

Article history: Background: Pre-session administration of D-cycloserine (DCS) has been found to augment exposure
Received 16 July 2012 therapy outcomes in a variety of anxiety disorders. To be able to enhance learning only for successful
Received in revised form exposure sessions, it would be beneficial to have the option of administering DCS after rather than before
26 September 2012
the session, a strategy encouraged by pre-clinical work. We believe the present study is the first pub-
Accepted 27 September 2012
lished report on the efficacy of post-session administration of DCS in humans.
Method: Adults (N ¼ 29) with a DSM-IV diagnosis of acrophobia were randomized to receive two sessions
Keywords:
of virtual reality exposure therapy (VRE) in combination with placebo or 50 mg of DCS. Instead of
Augmentation
CBT
administering the pill prior to each of the sessions, as has been done in extant work, we administered the
Cognitive behavioral therapy pill immediately following each session. Measures of acrophobia severity were collected at baseline, at
D-cycloserine each treatment session, 1-week post-treatment, and at 1-month follow-up.
Exposure treatment Results: Mixed-effects repeated-measures ANOVAs and GLMMs revealed significant improvement in all
Randomized controlled trial outcome measures over time, but no between-group differences were observed. At post-treatment,
Acrophobia 63.5% of patients in the placebo condition vs. 60.0% of those in the DCS condition were in remission.
At 1-month follow up, 63.4% of those in the placebo condition vs. 66.6% of those in the DCS condition
were in remission.
Conclusions: These findings do not support the application of post-session DCS administration for aug-
menting the efficacy of exposure-based treatments. Possible reasons for these findings are discussed.
Trial Registry: The Trial is registered at [Link] (NCT01102803).
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction augmentation of exposure treatment for anxiety disorders

(Hofmann et al., 2011).
An exciting success of translational research is the use of D- In the initial clinical trial, Ressler et al. (2004) enrolled 27 adults
cycloserine (DCS) as an augmentation strategy for exposure-based with DSM-III-R acrophobia into a 2-session virtual reality exposure
treatment (Anderson and Insel, 2006). Following pre-clinical (VRE) therapy protocol. The investigators selected this VRE protocol
research implicating the N-methyl D-aspartate (NMDA) receptor in because VRE allows for carefully controlling exposure procedures
fear conditioning and extinction learning (Baker and Azorlosa, and two 45-min sessions is considered a suboptimal dose, thus
1996; Davis and Myers, 2002; Falls et al., 1992; Lee and Kim, leaving sufficient room to detect augmentation effects. The efficacy
1998). Walker et al. (2002) demonstrated that rats administered of DCS augmentation was tested by randomly assigning patients to
DCS, a partial agonist at the NMDA receptor, prior to extinction take either (a) pill placebo (PBO); (b) 50 mg of DCS; or (c) 500 mg of
training displayed enhanced dose-dependent extinction relative to DCS 2e4 h prior to each of the 2 sessions. Relative to patients
rats administered saline. Replicated since (Myers and Davis, 2007), receiving PBO, those receiving DCS showed significantly greater
these findings encouraged the application of DCS for the reductions on in acrophobia severity during the 2-session inter-
vention and the 3-month follow-up period with no differences
* Corresponding author. Tel.: þ1 214 768 4125; fax: þ1 214 768 3910.
between the two DCS doses.
E-mail addresses: candycetart@[Link] (C.D. Tart), jsmits@[Link] The findings of the Ressler et al. (2004) study have since been
(J.A.J. Smits). replicated outside a VRE paradigm and extended across the anxiety

0022-3956/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
[Link]
 C.D. Tart et al. / Journal of Psychiatric Research 47 (2013) 168e174 169

disorders (Hofmann et al., 2006; Kushner et al., 2007; Otto et al., month follow-up. We hypothesized that, relative to participants
2010; Wilhelm et al., 2008). In a meta-analysis of 10 trials receiving placebo-augmented VRE, those receiving DCS-
involving clinical samples, Norberg et al. (2008) reported a signifi- augmented VRE would show decreased acrophobia symptoms at
cant advantage of DCS over PBO augmentation for exposure-based post-treatment and follow-up.
treatment efficacy (d ¼ .60).
In all trials of DCS augmentation to date, the administration of 2. Materials and method
DCS has been before instead of after the exposure session. However,
post-session administration of DCS would offer the clinician and 2.1. Participants
patient the advantage of select use of this augmentation strategy
(i.e., only after sessions deemed successful). Indeed, the judicious Participants (N ¼ 29; Mage ¼ 33.38) with acrophobia were
use of DCS e administration following only the most successful recruited from Southern Methodist University and the greater
exposure sessions e is supported by observations that repeated Dallas area. Inclusion criteria were: (a) age 18e65; and (b) meeting
dosing appears to rapidly cause tolerance to DCS (Lopes et al., 1997; DSM-IV-TR criteria for acrophobia. Exclusion criteria were: (a)
Parnas et al., 2005; Quartermain et al., 1994). Importantly, DCS a subjective distress score (SUDS) of <50 on the BAT; (b) a lifetime
dosing only after successful exposure sessions addresses the clin- history of bipolar disorder, schizophrenia, psychosis, delusional
ical concern that DCS may be applied to exposure sessions that disorders, or obsessive-compulsive disorder; current or recent
accidentally sensitized the patient to fear cues. Recent research in diagnosis of substance or alcohol abuse or dependence; (c) current
fear reconsolidation validates this clinical concern. Specifically, or recent suicidality or suicidal behavior; (d) current or recent
when memories are retrieved, they become temporarily destabi- diagnosis of PTSD, panic disorder, or eating disorder; (e) any
lized and vulnerable to intervention that may attenuate, modify, or physical or psychiatric condition that could interfere with the
stabilize the original memory (Lee, 2008; Monfils et al., 2009; capacity to engage in therapy; (f) history of head trauma causing
Schiller et al., 2010). As such, during this time-limited state, loss of consciousness, seizure, or ongoing cognitive impairment; (g)
memories are amenable to updating with new information, and pregnancy; (h) concurrent psychotropic medication or psycho-
hence sensitizing events or inadequate learning during exposure therapy or prior non-response to exposure therapy for acrophobia.
procedures may lead to reconsolidation. Research further suggests
that, when the period of the initial conditioning sessions is brief, 2.2. Experimental design
reconsolidation processes are dominant, whereas extinction
processes dominate in longer sessions (Eisenberg et al., 2003; Lee Eligible participants were assigned to one of two blinded arms e
et al., 2006). Therefore, DCS has the potential to augment extinc- VRE therapy plus pill placebo (N ¼ 14) or 50 mg of DCS (N ¼ 15).
tion learning when the conditioned stimulus (CS) is re-exposed Randomization was done by research staff not involved in the trial
sufficient times for extinction to occur. In contrast, if stimulus re- using minimization procedures (Scott et al., 2002) and stratifying
exposure during memory reactivation is relatively brief compared on gender, therapist, and time of day of treatment sessions (before
to the strength of conditioning, little extinction is induced and DCS or after 4 pm). Assessments occurred at baseline, prior to each
may augment reconsolidation (i.e., if the memory is in the labile treatment session, at one-week post-treatment, and at 1-month
state) (Lee et al., 2006). Although the bounds around reconsolida- follow-up. Consistent with the original study (Ressler et al.,
tion effects in humans have yet to be investigated, research 2004), the primary outcome measures were the BAT, AAVQ, AAQ
suggests that this period of lability, in which reconsolidation can (Cohen, 1977), ATHQ (Abelson and Curtis, 1989), and the CGI (Guy,
occur, persists for approximately 6 h in animals (Nader et al., 2000). 1970).
Hence, the application of post-session DCS administration only
when exposure is judged to be adequate has the potential for 2.3. Procedures
avoiding unwittingly strengthening fear associations (Litz et al.,
2012). 2.3.1. Screening
Pre-clinical data offers evidence in support of the potential Respondents to study advertisements were briefly screened by
efficacy of post-session DCS augmentation. For example, a research assistant before being invited to the clinic for formal
Ledgerwood et al. (2003) found that post-extinction subcutaneous eligibility screening. After signing informed consent, participants
administration of DCS in rats facilitated extinction and these effects were administered the Structured Clinical Interview for DSM-IV
were comparable to that observed with pre-extinction adminis- (SCID) (First et al., 2001) and clinician-rated instruments. The
tration. Timing, similar to dosage, followed a linear trend such that study physician reviewed the medical history information and
peak extinction facilitation was observed when DCS was adminis- prescribed all study medications. The senior author (JAJS) verified
tered either immediately afterward or 30 min post-training, and assessment of entry criteria and diagnoses.
diminished results were found at 120 min post-training adminis-
tration. No benefits of DCS were observed at 240 min post-training 2.3.2. Treatment
administration (Ledgerwood et al., 2003). Parnas et al. (2005) have The two-session VRE protocol was identical to the protocol used
since replicated these findings. by Ressler et al. (2004). Treatment started one week following
The present study investigates whether post-session oral baseline assessment. Assessment of clinical status and safety
administration of DCS facilitates exposure-based CBT outcomes for occurred prior to each session. In the first session (60 min), ther-
anxiety disorders. Replicating the procedures of the first study apists reviewed the rationale for exposure and the cognitive-
examining the efficacy of pre-session DCS in humans (Ressler et al., behavioral model of acrophobia. Therapists then led participants
2004), we randomized 29 adults with height phobia to receive two through their first graded VRE exposures. The virtual reality envi-
sessions of VRE in combination with either PBO or 50 mg of DCS. ronment simulated a hotel with a glass elevator, and a series of
Instead of administering the pill two to 4 h before each of the two catwalks, balconies, and rooftop in which participants could stand
VRE sessions, as was done in the Ressler et al. (2004) study, we and look around while wearing a virtual reality helmet and goggles.
administered the pill immediately after each of the two VRE Participants reported their subjective fear levels (i.e., subjective
sessions. Measures of acrophobia severity were collected at base- units of distress [SUDS]) (Wolpe, 1958) on a 0e100 scale (0 indi-
line, at each treatment session, 1-week post-treatment, and at 1- cates no fear, 100 indicates extreme fear or panic) every 5 min
170 C.D. Tart et al. / Journal of Psychiatric Research 47 (2013) 168e174

throughout the exposures. Exposure sessions began at the floor 3.4. Attitudes toward heights questionnaire (ATHQ)
where the participants first reported SUDS of 50 or greater on the
BAT. Therapists then led participants through each exposure (i.e., This questionnaire (Abelson and Curtis, 1989) includes six
teaching them to approach, not avoid, and challenging beliefs), heights situations and assesses attitudes toward these situations
until meaningful SUDS decrease occurred (>50% reduction), at using a 0e10 scale. Reported internal consistency is sound (a ¼ .81)
which time the therapist would lead the participant up to the next and validity is acceptable (Coelho et al., 2008; Davis et al., 2006;
floor and repeat these procedures for a total of 30 min of VRE Rothbaum, 1995).
exposures. At the second session, completed one week after the
first session, participant completed another 30 min of VRE expo- 3.5. Clinical global impressions severity and improvement scales
sures. At the conclusion of each of the two sessions, participants (CGI-S and CGI-I)
were administered the pill (DCS or PBO). Blind to the condition,
therapists were advanced doctoral-student level therapists trained The CGI-S and CGI-I are widely used measures of global
and supervised by the senior author (JAJS). psychopathology severity and improvement initially developed for
the study of psychotropic drugs (Guy, 1970). In order to obtain CGI
2.3.3. Medication ratings, the therapists (blind to study condition) interviewed the
The 50 mg DCS and PBO capsules (size #0) were compounded by participant and used the SCID (including the specific phobia
the Abrams Royal Pharmacy in Dallas, TX. The DCS capsules con- module) as well as the additional measures of acrophobia symp-
tained D-cycloserine 50 mg (derived from Seromycin 250 mg toms (BAT, AAQ, AAVQ, and ATHQ). Response was defined as either
capsules) and polyethylene glycol 3350 powder. The matching “very much improved” or “much improved” on CGI-I (score  2).
placebo capsules contained only the polyethylene glycol 3350 Remission was defined as either “normal” or “minimally ill” on CGI-
powder. We selected a 50 mg dose of DCS because this dosage was S (score  2).
as effective as a large dose (500 mg) in the original clinical trial we
replicated (Ressler et al., 2004) and because this dosage has been
applied successfully in other trials by our group examining the 4. Data analysis
effects of pre-session administration of DCS (Hofmann et al., 2006;
Otto et al., 2009, 2010). All capsules were identical in appearance to All data were evaluated for differences between the groups at
maintain the blind. baseline (DCS vs. PBO; see Table 1). Study hypothesis was tested
using a series of mixed-effects repeated-measures analyses of
3. Measures variance (ANOVAs), which allows all participants to be included in
the analyses regardless of missing data and allows specification of
3.1. SCID a more complex covariance matrix for the correlation of the errors
of the repeated measures (Singer and Willett, 2003). A growth
The SCID was administered for inclusion and exclusion DSM-IV curve model (e.g., linear or quadratic) was not employed since
criteria at screening (First et al., 2001). The specific phobia module recent research has shown that constraining changes over time to
of the SCID was completed again at the post-treatment and follow- fit a specific growth curve can lead to misleading results (Liu et al.,
up assessments in order to inform clinical global impression 2012). Time (baseline, week 1, week 2, post-treatment, and 1 month
ratings. All SCID interviews were conducted by certified doctoral follow-up) was entered as the within-person factor, and Treatment
students. The certification procedure required (a) observing 3e4 Condition (0 ¼ PBO, 1 ¼ DCS) as the between-subjects factor. Hence,
videotaped and live SCID administrations by trained and certified a significant Time by Treatment Condition interaction on outcome
interviewers with the comparison of the trainees’ ratings to that of measures would provide support for the study hypothesis. For
the trained and certified interviewer, and (b) administer 3e5 SCID outcomes that were dichotomous (response and remission),
interviews in the presence of the trained and certified interviewer generalized linear mixed models (GLMMs) were performed, using
with the requirement of the trainees’ diagnosis match that of the a logistic linking function for the dependent variable. Accordingly,
trained and certified interviewer on at least 3 consecutive all analyses were based on an intent-to-treat design, including all
administrations. randomized individuals (N ¼ 29).
An a-priori power analysis revealed that a sample size of 16 (for
3.2. Behavioral avoidance test (BAT) 5 time points; CGI-S)e22 (for 3 time points; BAT) yields adequate
power (i.e., .80) to detect a medium effect size (f2 ¼ .25) at p < .05
During the initial screen, at post-treatment, and at follow-up, for withinebetween interaction in a two-group design using
participants underwent a behavioral avoidance test in the virtual
reality height environment. Participants reported on a 0e100 scale Table 1
(100 being the most intense fear) their SUDS for floors 1, 2, 3, 4, 9, 19 Baseline data.
of the virtual glass elevator and balconies. This test has been used
Variable DCS PBO P value
successfully as a measure of treatment gains in previous studies of M (SD) M (SD)
acrophobia research (Ressler et al., 2004). For the outcome anal- (n ¼ 15) (n ¼ 14)
yses, we included the level of fear reported at the highest floor of Age 29.33 (14.67) 37.71 (16.81) .16
the virtual elevator environment (19th floor). Behavioral avoidance 68.07 (14.78) 71.29 (19.94) .62
test (BAT)
Acrophobia avoidance 20.73 (7.74) 23.06 (10.60) .50
3.3. Acrophobia questionnaire with avoidance (AAVQ) and anxiety
questionnaire (AAVQ)
(AAQ) scales Acrophobia anxiety 48.27 (20.98) 58.36 (25.71) .26
questionnaire (AAQ)
This questionnaire (Cohen, 1977) describes 20 situations and Attitude toward heights 45.67 (8.53) 47.71 (12.98) .62
assesses levels of avoidance (0e3) and anxiety (0e6). These scales questionnaire (ATHQ)
Clinical global impression 4.13 (.52) 4.14 (.36) .96
widely used measure of acrophobia with adequate retest reliability of severity (CGI)
(r ¼ .82e.86) and validity (Baker et al., 1973).
 C.D. Tart et al. / Journal of Psychiatric Research 47 (2013) 168e174 171

repeated-measures ANOVAs (G*Power 3) (Faul et al., 2007). We and thus included in the analyses. On average, the sample pre-
selected a sample size (N ¼ 29) that exceeded the minimum needed sented as “moderately ill” at baseline (see Table 1). There were no
to detect clinically meaningful differences in order to approximate differences between DCS and PBO on any of the measures at
the methodological approach employed in the work being repli- baseline (see Table 1). By post-treatment, a total of 3 patients in the
cated (Ressler et al., 2004). PBO condition (21.4%) and none in the DCS condition had dropped
out. At follow-up, 4 patients in PBO (28.6%) and 3 in DCS (20.0%)
had dropped out (see Fig. 1). None of these differences was signif-
5. Results
icant (ps > .10).
5.1. Study flow and baseline characteristics
5.2. Treatment effects
Five hundred and sixty-two people completed the Internet
screen and 76 came in for a clinical interview (see Fig. 1). Of these, Means and standard deviations for all outcome measures at the
29 participants were eligible and randomized (PBO ¼ 14; DCS ¼ 15) different time points are reported in Table 2.

Fig. 1. Patient flow diagram.

172 C.D. Tart et al. / Journal of Psychiatric Research 47 (2013) 168e174

Table 2 5.2.4. Changes in cognitions related to heights

Means and standard deviations of dependent variables at each time point. There was a significant main effect of Time [F(3,72) ¼ 28.53,
Measure Mean (SD) p < .001] such that negative cognitions about height situations
Baseline Session 1 Session 2 Post-treatment Follow-up
(ATHQ) decreased over time. There was no significant Time by
Treatment Condition interaction [p ¼ .83], suggesting that there
BAT
DCS 68.07 e e 29.73 25.59 were no differences between DCS and PBO in ATHQ scores over
(14.78) (25.67) (23.98) time.
PBO 71.29 e e 35.55 23.80
(19.94) (25.18) (19.86)
5.2.5. Changes in clinician-rated improvement of acrophobia
AAVQ
DCS 20.73 e 16.87 9.00 6.83
There was a significant main effect of Time on CGI-I
(7.74) (9.90) (7.41) (7.76) [F(3,44) ¼ 33.92, p < .001] such that clinician-rated global
PBO 23.06 e 14.85 12.38 7.73 improvement scores of acrophobia decreased over time, but there
(10.60) (10.75) (9.72) (6.28) were no differences between treatment conditions on CGI-I scores
AAQ
changes [p ¼ .90]. GLMM analyses showed an 81.8% response rate
DCS 48.27 e 37.67 25.47 17.75
(20.98) (18.66) (14.83) (12.74) for PBO and 66.7% response rate for DCS at post-treatment. At 1-
PBO 58.36 e 41.15 33.08 26.27 month follow up, response rates were 80.1% and 75.0% for PBO
(25.71) (20.44) (22.54) (16.80) and DCS, respectively. Between-group differences were not signif-
ATHQ icant [ps  .40].
DCS 45.67 e 38.13 45.67 26.83
(8.53) (13.28) (8.53) (13.11)
PBO 47.71 e 37.08 47.71 27.55 5.2.6. Changes in clinician-rated severity of acrophobia
(12.98) (11.06) (12.98) (10.30) There was a significant main effect of Time [F(4,86) ¼ 32.47,
CGI-S p < .001] such that participants’ clinician-rated global severity
DCS 4.13 4.15 3.60 2.53 2.17
scores (CGI-S) of acrophobia decreased over time. Again, there was
(.52) (.38) (.91) (1.13) (.94)
PBO 4.14 4.09 3.62 2.36 2.50 no significant interaction between Time and Treatment condition
(.36) (.54) (.65) (.81) (.97) [p ¼ .49]. GLMM analyses indicated remission rates of 63.5% for PBO
CGI-I and 60.0% for DCS condition at possttreatment. At 1-month follow
DCS e 4.00 3.00 2.27 1.92 up, remission rates were 63.4% and 66.6% for PBO and DCS,
(.00) (.93) (.96) (1.00)
PBO e 3.91 2.92 2.00 1.90
respectively. These between-group differences were not significant
(.30) (.86) (.89) (.99) [ps > .75].
Note. BAT ¼ Behavioral avoidance test, AAVQ ¼ Acrophobia avoidance question-
naire; AAQ ¼ Acrophobia anxiety questionnaire; ATHQ ¼ Attitudes toward heights 6. Discussion
questionnaire; CGI-S ¼ Clinician’s global impression of severity; CGI-I ¼ Clinician’s
global impression of improvement. To our knowledge, this is the first published examination of
whether post-exposure session administration of DCS augments
fear extinction in humans. Specifically, we randomized 29 indi-
5.2.1. Changes in behavioral measure of fear of heights
viduals with height phobia to receive 2 sessions of VRE plus either
There was a significant main effect of Time on BAT SUDs
placebo or 50 mg of DCS administered post-session, in contrast to
[F(2,49) ¼ 59.26, p < .001] such that participants’ self-reported fear
the Ressler et al. (2004) in which DCS was administered pre-
on the top floor of the virtual reality environment decreased over
session. We hypothesized that those in the DCS condition would
time. There was no significant interaction between Time and
demonstrate greater reductions in acrophobia symptoms and
Treatment Condition, suggesting no differences between treatment
severity compared to those in the PBO condition. Our results indi-
conditions in improvement on the BAT over time [p ¼ .76].
cated that participants in both groups improved significantly on all
measures of acrophobia severity, but participants receiving DCS did
5.2.2. Changes in self-reported avoidance of height situations
not outperform participants receiving PBO.
There was a significant main effect of Time [F(3,73) ¼ 22.22,
Our findings reflect a failure to translate results from animal
p < .001] such that self-reported avoidance of height situations
studies showing that post-extinction training administration of
(AAVQ) decreased over time. The Time by Treatment Condition
DCS facilitates extinction learning (Ledgerwood et al., 2003; Parnas
interaction was not significant [p ¼ .07], indicating that there were
et al., 2005). Pre- rather than post-session administration of DCS in
no differences between DCS and PBO in improvements on AAVQ
humans was first adopted because of the time taken between oral
scores.1
administration and bioavailability, with peak plasma levels at
approximately 90e120 min following oral administration (D’Souza
5.2.3. Changes in self-reported fear of heights situations
et al., 2000; van Berckel et al., 1997). In animal studies investigating
There was a significant main effect of Time [F(3,72) ¼ 25.90,
post-extinction administration of DCS, DCS has been administered
p < .001] such self-reported fear of height situations (AAQ)
subcutaneously (Ledgerwood et al., 2003; Parnas et al., 2005),
decreased over time. There was no significant Time by Treatment
which results in peak plasma levels within as little as 15 min (Wlaz
Condition interaction [p ¼ .83], suggesting that there were no
et al., 1994). As such, post-session oral administration of DCS may
differences between treatment conditions in AAQ scores over time.
simply fail to arrive at the salient site of action in a time course
necessary for therapeutic effect.
Nonetheless, there are a number of other important considerations
for interpreting the null findings in our study. First, the failure of DCS to
augment exposure treatment with post-session DCS administration
1
Given the trend for an interaction effect, we conducted additional analyses. The may be specific to phobic samples that are not severely disabled (i.e.,
findings revealed that DCS outperformed PBO at some time points (post-treatment)
but underperformed it at others (session 2). Also, the interaction effect remained
adults with a diagnosis of acrophobia). As suggested by Grillon (2009),
non-significant when we removed the 1-month follow-up assessment from the individuals with “lower” levels of fear may improve in exposure-based
analyses. protocols via more high-level cognitive processing (i.e., cognitive
 C.D. Tart et al. / Journal of Psychiatric Research 47 (2013) 168e174 173

restructuring) instead of amygdala-mediated circuits (i.e., NMDA- the manuscript. P.R.H., L.B.D. M.H.P., S.G.H., M.B.P. and M.W.O.
mediated), while those with greater levels of fear may improve in assisted with the design and implementation of the study, inter-
exposure therapy protocols via more automatic learning processes pretation of the data and the writing of the manuscript. D.R.
(i.e., habituation) that are driven by amygdala-mediated circuits, thus participated in the data analyses strategies, interpretation of data
supporting the indication of DCS-augmented treatment for more and writing of the manuscript. All authors contributed to and
severely disabled patients. Evidence consistent with this hypothesis approved the final manuscript.
comes from the failure to observe augmentation effects with pre-
session DCS dosing in samples with “low-level” fears actively
Conflict of interest
recruited from the community (Guastella et al., 2007a, 2007b) and
a recent report showing stronger augmentation effects of pre-session
Dr. Tart, Ms. Handelsman, Ms. DeBoer, Dr. Rosenfield and Dr.
DCS administration among severe relative to less severe patients with
Powers report no financial relationships with commercial interests.
post-traumatic stress-disorder (de Kleine et al., 2012). In terms of our
Dr. Pollack’s disclosures over the last three years include: Advisory
purpose of replicating Ressler et al. (2004), albeit with post-session
Boards and Consultation: Eli Lilly, Medavante, Otsuka, Targia
dosing, the mean severity scores for the DCS and PBO groups on the
Pharmaceuticals Research Grants: Bristol Myers Squibb, Euthymics,
AAQ measure were 48.27 (SD ¼ 20.98) and 58.36 (SD ¼ 25.71),
Forest Laboratories, GlaxoSmithKline, NCCAM, NIDA, NIMH Equity:
respectively, in the present study vs. 65.8 (SD ¼ 6.2) and 73.4
Medavante, Mensante Corporation, Mindsite, Targia Pharmaceu-
(SD ¼ 5.6), respectively, in the Ressler et al. (2004) study. Accordingly, it
tical Royalty/patent: SIGH-A, SAFER interviews. Dr. Hofmann has
is possible that the absent DCS augmentation effect in the present
served as a consultant for Merck and Schering-Plough and receives
study was linked to recruitment of participants with lower baseline
book royalties from various book publishers for work unrelated to
height phobia severity than those in the Ressler et al. (2004) study.
this study. Dr. Otto serves as consultant for MicroTransponder, Inc.
A third possible explanation for the null findings is that, despite
and receives royalties from various book publishers for work
providing suboptimal dosing of VRE, many patients responded very
unrelated to this study. Dr. Smits receives royalties from various
well to treatment with VR and placebo, thus perhaps creating
book publishers for work unrelated to this study.
a ceiling effect that left little room for the demonstration of addi-
tional improvement (with DCS). As indicated by data from the CGI,
almost 3 out of 4 participants in the present trial were rated by Acknowledgments
a clinician as “much improved” or better at post-treatment and 1-
month follow-up. Interestingly, using the same manualized VRE This report was supported by Diversity Supplement to
protocol and dosing, Ressler et al. (2004) appeared to have R01MH075889 from the National Institute of Mental Health
observed weaker overall treatment effects, with only approxi- (NIMH). We appreciate the support of the NIMH. The content is
mately 1 of 4 participants rated, by self-report in this study, as solely the responsibility of the authors and does not necessarily
“much improved” at post-treatment and 3-month follow-up, represent the official views of the NIMH or the National Institutes of
thereby perhaps leaving the room necessary for augmentation to Health.
exert additional positive effects.
The present study has a number of limitations that deserve
mention. First, we did not include a pre-session pill administration References
manipulation. Had we included this, we would have been able to
Abelson JL, Curtis GC. Cardiac and neuroendocrine responses to exposure therapy in
test some of the hypotheses for our null findings and demonstrate height phobics: desynchrony within the ‘physiological response system’.
whether the failure to replicate the Ressler et al. (2004) study was Behaviour Research and Therapy 1989;27:561e7.
specific to a post-session administration of DCS. Second, our CGI Anderson KC, Insel TR. The promise of extinction research for the prevention and
treatment of anxiety disorders. Biological Psychiatry 2006;60:319e21.
ratings were completed by a clinician who was, although blind to
Baker BL, Cohen DC, Saunders JT. Self-directed desensitization for acrophobia.
group assignment, not an independent evaluator. While this Behaviour Research and Therapy 1973;11:79e89.
methodological feature cannot explain the absence of between- Baker JD, Azorlosa JL. The NMDA antagonist MK-801 blocks the extinction of
Pavlovian fear conditioning. Behavioral Neuroscience 1996;110:618e20.
group differences, it may have introduced a bias resulting in
Coelho C, Silva C, Santos J, Tichon J, Wallis G. Contrasting the effectiveness and
larger overall clinician-rated improvement rates. Third, we only efficiency of virtual reality and real environments in the treatment of acro-
examined one dose of DCS (50 mg). We selected this dose because it phobia. PsychNology Journal 2008;6:203e16.
had shown efficacy in the Ressler et al. (2004) trial and subsequent Cohen D. Comparison of self-report and overt-behavioral procedures for assessing
acrophobia. Behavior Therapy 1977;8:17e23.
trials by our group (Hofmann et al., 2006; Otto et al., 2010), but it is D’Souza DC, Gil R, Cassello K, Morrissey K, Abi-Saab D, White J, et al. IV glycine and
possible that positive effects of post-session DCS dosing would have oral D-cycloserine effects on plasma and CSF amino acids in healthy humans.
been evident had we used a different dose. Finally, unlike Ressler Biological Psychiatry 2000;47:450e62.
Davis M, Myers KM. The role of glutamate and gamma-aminobutyric acid in fear
et al. (2004), we did not include a physiological measure of fear, extinction: clinical implications for exposure therapy. Biological Psychiatry
and can therefore make no inferences with respect to the effects of 2002;52:998e1007.
post-session DCS administration on objective measures of fear. Davis M, Ressler K, Rothbaum BO, Richardson R. Effects of D-cycloserine on
extinction: translation from preclinical to clinical work. Biological Psychiatry
Despite these shortcomings, this study provides a meaningful 2006;60:369e75.
contribution to the growing literature on DCS augmentation of de Kleine RA, Hendriks GJ, Kusters WJ, Broekman TG, van Minnen A. A randomized
exposure treatment. Our findings indicate that post-session oral placebo-controlled trial of D-cycloserine to enhance exposure therapy for
posttraumatic stress disorder. Biological Psychiatry 2012;71:962e8.
DCS administration of 50 mg may not be effective for augmenting
Eisenberg M, Kobilo T, Berman DE, Dudai Y. Stability of retrieved memory: inverse
the efficacy of exposure-based treatments; however, these initial correlation with trace dominance. Science 2003;301:1102e4.
negative findings should be re-examined with multiple treatment- Falls WA, Miserendino MJ, Davis M. Extinction of fear-potentiated startle: blockade
by infusion of an NMDA antagonist into the amygdala. Journal of Neuroscience
seeking samples that span a greater range of anxiety severity.
1992;12:854e63.
Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power
Contributors analysis program for the social, behavioral, and biomedical sciences. Behavior
Research Methods 2007;39:175e91.
First MB, Spitzer RL, Gibbon M, Williams JBW. Structured clinical interview for
C.D.T and J.A.J.S. designed the study, conducted the literature DSM-IV axis-I disorders. Clinician version. Arlington, VA: American Psychiatric
searches, participated in the data analyses, and wrote a first draft of Publishing, Inc.; 2001.
174 C.D. Tart et al. / Journal of Psychiatric Research 47 (2013) 168e174

Grillon C. D-cycloserine facilitation of fear extinction and exposure-based therapy Norberg MM, Krystal JH, Tolin DF. A meta-analysis of D-cycloserine and the facili-
might rely on lower-level, automatic mechanisms. Biological Psychiatry 2009; tation of fear extinction and exposure therapy. Biological Psychiatry 2008;63:
66:636e41. 1118e26.
Guastella AJ, Dadds MR, Lovibond PF, Mitchell P, Richardson R. A randomized Otto MW, Basden SL, McHugh RK, Kantak KM, Deckersbach T, Cather C, et al. Effects
controlled trial of the effect of D-cycloserine on exposure therapy for spider of D-cycloserine administration on weekly nonemotional memory tasks in
fear. Journal of Psychiatric Research 2007a;41:466e71. healthy participants. Psychotherapy and Psychosomatics 2009;78:49e54.
Guastella AJ, Lovibond PF, Dadds MR, Mitchell P, Richardson R. A randomized Otto MW, Tolin DF, Simon NM, Pearlson GD, Basden S, Meunier SA, et al. Efficacy of
controlled trial of the effect of D-cycloserine on extinction and fear condi- d-cycloserine for enhancing response to cognitive-behavior therapy for panic
tioning in humans. Behaviour Research and Therapy 2007b;45:663e72. disorder. Biological Psychiatry 2010;67:365e70.
Guy W. Clinical global impression. Manual for the ECDEU assessment battery; 1970. Parnas AS, Weber M, Richardson R. Effects of multiple exposures to D-cycloserine
Hofmann SG, Meuret AE, Smits JA, Simon NM, Pollack MH, Eisenmenger K, et al. on extinction of conditioned fear in rats. Neurobiology of Learning and Memory
Augmentation of exposure therapy with D-cycloserine for social anxiety 2005;83:224e31.
disorder. Archives of General Psychiatry 2006;63:298e304. Quartermain D, Mower J, Rafferty MF, Herting RL, Lanthorn TH. Acute but not chronic
Hofmann SG, Smits JA, Asnaani A, Gutner CA, Otto MW. Cognitive enhancers for activation of the NMDA-coupled glycine receptor with D-cycloserine facilitates
anxiety disorders. Pharmacology Biochemistry and Behavior 2011;99:275e84. learning and retention. European Journal of Pharmacology 1994;257:7e12.
Kushner MG, Kim SW, Donahue C, Thuras P, Adson D, Kotlyar M, et al. D-cycloserine Ressler KJ, Rothbaum BO, Tannenbaum L, Anderson P, Graap K, Zimand E, et al.
augmented exposure therapy for obsessive-compulsive disorder. Biological Cognitive enhancers as adjuncts to psychotherapy: use of D-cycloserine in
Psychiatry 2007;62:835e8. phobic individuals to facilitate extinction of fear. Archives of General Psychiatry
Ledgerwood L, Richardson R, Cranney J. Effects of D-cycloserine on extinction of 2004;61:1136e44.
conditioned freezing. Behavioral Neuroscience 2003;117:341e9. Rothbaum. Effectiveness of computer-generated (virtual reality) graded exposure in
Lee H, Kim JJ. Amygdalar NMDA receptors are critical for new fear learning in the treatment of acrophobia. The American Journal of Psychiatry 1995;152:
previously fear-conditioned rats. Journal of Neuroscience 1998;18:8444e54. 626e8.
Lee JL. Memory reconsolidation mediates the strengthening of memories by addi- Schiller D, Monfils MH, Raio CM, Johnson DC, Ledoux JE, Phelps EA. Preventing the
tional learning. Nature Neuroscience 2008;11:1264e6. return of fear in humans using reconsolidation update mechanisms. Nature
Lee JL, Milton AL, Everitt BJ. Reconsolidation and extinction of conditioned fear: 2010;463:49e53.
inhibition and potentiation. Journal of Neuroscience 2006;26:10051e6. Scott NW, McPherson GC, Ramsay CR, Campbell MK. The method of minimization for
Litz BT, Salters-Pedneault K, Steenkamp MM, Hermos JA, Bryant RA, Otto MW, et al. allocation to clinical trials. a review. Controlled Clinical Trials 2002;23:662e74.
A randomized placebo-controlled trial of d-cycloserine and exposure therapy Singer JD, Willett JB. Applied longitudinal data analysis: methods for studying
for posttraumatic stress disorder. Journal of Psychiatric Research 2012;46: change and event occurrence. New York: Oxford University Press; 2003.
1184e90. van Berckel BN, Lipsch C, Timp S, Gispen-de Wied C, Wynne H, et al. Behavioral and
Liu S, Rovine MJ, Molenaar PC. Selecting a linear mixed model for longitudinal data: neuroendocrine effects of the partial NMDA agonist D-cycloserine in healthy
repeated measures analysis of variance, covariance pattern model, and growth subjects. Neuropsychopharmacology 1997;16:317e24.
curve approaches. Psychological Methods 2012;17:15e30. Walker DL, Ressler KJ, Lu KT, Davis M. Facilitation of conditioned fear extinction by
Lopes T, Neubauer P, Boje KM. Chronic administration of NMDA glycine partial systemic administration or intra-amygdala infusions of D-cycloserine as
agonists induces tolerance in the Porsolt swim test. Pharmacology Biochemistry assessed with fear-potentiated startle in rats. Journal of Neuroscience 2002;22:
and Behavior 1997;58:1059e64. 2343e51.
Monfils MH, Cowansage KK, Klann E, LeDoux JE. Extinction-reconsolidation Wilhelm S, Buhlmann U, Tolin DF, Meunier SA, Pearlson GD, Reese HE, et al.
boundaries: key to persistent attenuation of fear memories. Science 2009;324: Augmentation of behavior therapy with D-cycloserine for obsessive-compulsive
951e5. disorder. American Journal of Psychiatry 2008;165:335e41.
Myers KM, Davis M. Mechanisms of fear extinction. Molecular Psychiatry 2007;12: Wlaz P, Baran H, Loscher W. Effect of the glycine/NMDA receptor partial agonist, D-
120e50. cycloserine, on seizure threshold and some pharmacodynamic effects of MK-
Nader K, Schafe GE, LeDoux JE. The labile nature of consolidation theory. Nature 801 in mice. European Journal of Pharmacology 1994;257:217e25.
Reviews Neuroscience 2000;1:216e9. Wolpe J. Psychotherapy by reciprocal inhibition. Palo Alto, CA: Stanford; 1958.