Biological
Psychiatry
Celebrating
50 Years:726
BiologArchival ReportA Randomized Trial Directly Comparing Ventral
Capsule and Anteromedial Subthalamic Nucleus
Stimulation in Obsessive-Compulsive Disorder:
Clinical and Imaging Evidence for Dissociable
Effects
Himanshu Tyagi, Annemieke M. Apergis-Schoute, Harith Akram, Tom Foltynie,
Patricia Limousin, Lynne M. Drummond, Naomi A. Fineberg, Keith Matthews,
Marjan Jahanshahi, Trevor W. Robbins, Barbara J. Sahakian, Ludvic Zrinzo, Marwan Hariz,
and Eileen M. JoyceABSTRACT
BACKGROUND: Deep brain stimulation (DBS) is an emerging treatment for severe obsessive-compulsive disorder
(OCD). We compared the efficacy of ventral capsule/ventral striatal (VC/VS) and anteromedial subthalamic nucleus
(amSTN) DBS in the same patients and tested for mechanistic differences on mood and cognitive flexibility and
associated neural circuitry. The possible synergistic benefit of DBS at both sites and cognitive behavioral therapy
was explored.
METHODS: Six patients with treatment-refractory OCD (5 men; Yale-Brown Obsessive Compulsive Scale score .32)
entered double-blind counterbalanced phases of 12-week amSTN or VC/VS DBS, followed by 12-week open phases
when amSTN and VC/VS were stimulated together, in which optimal stimulation parameters were achieved and
adjunctive inpatient cognitive behavioral therapy was delivered. OCD and mood were assessed with standardized
scales and cognitive flexibility with the Cambridge Neuropsychological Test Automated Battery Intra-Extra
Dimensional Set-Shift task. Diffusion-weighted and intraoperative magnetic resonance imaging scans were
performed for tractography from optimally activated electrode contacts.
RESULTS: DBS at each site significantly and equivalently reduced OCD symptoms with little additional gain following
combined stimulation. amSTN but not VC/VS DBS significantly improved cognitive flexibility, whereas VC/VS DBS
had a greater effect on mood. The VC/VS effective site was within the VC. VC DBS connected primarily to the medial
orbitofrontal cortex, and amSTN DBS to the lateral orbitofrontal cortex, dorsal anterior cingulate cortex, and
dorsolateral prefrontal cortex. No further improvement followed cognitive behavioral therapy, reflecting a floor effect
of DBS on OCD.
CONCLUSIONS: Both the VC/VS and amSTN are effective targets for severe treatment-refractory OCD. Differential
improvements in mood and cognitive flexibility and their associated connectivity suggest that DBS at these sites
modulates distinct brain networks.
Keywords: Anteromedial subthalamic nucleus, DBS, Deep brain stimulation, Obsessive-compulsive disorder, OCD,
Ventral internal capsule
https://doi.org/10.1016/j.biopsych.2019.01.017Obsessive-compulsive disorder (OCD) has a lifetime prevalence
of 1% to 2%. Effective treatments include cognitive behavioral
therapy (CBT) and serotonin reuptake inhibitor medication (1),
but there remains a severely impaired, treatment-refractory
subgroup (2) for whom anterior cingulotomy or capsulotomy is
an option (3).
Deep brain stimulation (DBS) for OCD was introduced as an
alternative to neurosurgical ablation because it is reversible
and adjustable (4). Four studies have included a randomized
controlled trial of DBS at the site of the anterior capsulotomySEE COMMENTARY
ª 2019 Society of Biological Psychiatry. This is an open access ar
CC BY license (http://creativecommons.org/licenses/by/4.0/).
ical Psychiatry May 1, 2019; 85:726–734 www.sobp.org/journal(ventral capsule/ventral striatum [VC/VS]). Earlier, small studies
found that relatively few patients responded to VC/VS DBS
(5,6), but more recent, larger studies found level 1 evidence in
favor of DBS (7,8). Neurosurgical targeting in these studies
changed over time (9), and there were also differences in the
choice of DBS electrodes, all of which had four contacts at
their tip but variable distances between them. It is likely that
such factors led to discrepancies in the location and/or spread
of stimulation and may explain the inconsistent response rates.
For example, within the area of the VC/VS, it is not clearON PAGE 706
ticle under the
ISSN: 0006-3223
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 Yearswhether stimulation of capsular white matter and/or nearby
gray matter provides the optimal response (7,8).
The anteromedial subthlamic nucleus (amSTN) is an alter-
native DBS target for OCD following a randomized controlled
trial by Mallet et al. (10), who found significant symptomatic
reduction. Which of the two sites is better for ameliorating
severe OCD has yet to be established. In addition, preliminary
evidence suggests that the mechanisms of action at the two
sites may differ. VC/VS DBS studies have reported a pro-
nounced improvement in mood (5,7,8) not evident following
the amSTN study (10), leading Denys et al. (7) to suggest that
different neural circuits may mediate OCD, with one stimulated
by VC/VS DBS primarily improving mood and another affected
by amSTN DBS primarily modulating compulsive behavior.
In this study, using a randomized, double-blind, counter-
balanced design, we investigated the effects of VC/VS and
amSTNDBS in the samepatients to 1) establishwhether one site
is more efficacious than the other in improving OCD, 2) deter-
mine the precise neuroanatomical locations for optimal effects
by calculating volumes of tissue activation (VTAs) at each site,
and 3) test the hypothesis of Denys et al. (7) by first employing
ratings of mood and a test of cognitive flexibility previously
shown to measure an endophenotype of OCD (11) and then
investigating the neural circuitry associated with the DBS at
each site with magnetic resonance imaging (MRI) tractography.
The potential of synergistic benefit of DBS at both sites and
adjunctive CBT was additionally explored in open phases
when both DBS sites were stimulated.
METHODS AND MATERIALS
Participants
Six patients were recruited from OCD specialized services.
Eligibility required ICD-10 OCD, age.20 years, illness duration
.10 years, unremitting symptoms for 2 years, Yale-Brown
Obsessive Compulsive Scale (Y-BOCS) score $32, and DSM-
IV General Assessment of Function Scale score #50. Patients
were ineligible if they were pregnant or had ICD-10 substance
misuse, organic brain syndrome, adult personality disorder
(except obsessive-compulsive type), pervasive developmental
disorder, schizophrenia or bipolar disorder, or contraindications
to neurosurgery. Treatment resistance was defined as no sus-
tained benefit from 1) at least two serotonin reuptake inhibitors
for a minimum of 12 weeks at optimal doses; 2) augmentation
of serotonin reuptake inhibitor treatment with antipsychotics or
extension of the serotonin reuptake inhibitor dose beyond rec-
ommended limits; and 3) two trials of CBT, one as inpatient,
lasting 10 hours minimum. All patients provided written
informed consent. The clinical trial was registered with the UK
Clinical Research Network (No. 13158) and the ISRCTN registry
(No. 18430630) and received ethical approval (clinical study:
National Health Service Health Research Authority No. 12/LO/
1087; MRI: National Health Service West London Research
Ethics Committee No. 10/H0706/68).
Study Design
Eligible participants were offered stereotactic ablation or entry
to the DBS trial. Prescribed medications were kept constant
unless clinically indicated. Before surgery, all patientsBiologicaunderwent 3T diffusion-weighted MRI for tractography anal-
ysis (Magnetom Tim Trio; Siemens, Berlin, Germany).
Following surgery, stimulation remained off for at least 4
weeks. At the beginning of each phase, participants were
admitted to a neuropsychiatry ward for 2 weeks for DBS
adjustment. Participants remained on optimized settings for 12
weeks with a contingency for readjustment if clinically indi-
cated (Supplemental Figure S1).
The initial two phases were double blind, randomized, and
counterbalanced; each phase lasted 12 weeks. Participants
received stimulation of either the amSTN (n = 3) or the VC/VS
(n = 3) followed by the alternate condition. A 12-week open
phase followed, during which electrodes at both sites were
active (combined stimulation [COMB] phase). There were two
additional 12-week open phases when optimized stimulation
settings were delivered using data from previous phases (OPT
phase), followed by the participants’ receiving CBT/exposure
and response prevention in an inpatient unit (OPT plus
adjunctive CBT phase) while continuing with the OPT DBS
settings. Clinical and cognitive assessments were performed
before surgery (baseline) and after each phase.
Neurosurgery
Surgery was performed under general anesthesia. Patients
underwent stereotactic 1.5T MRI (Magnetom Avanto; Siemens)
for planning of amSTN and VC/VS coordinates and trajectories
(12). Through 14-mm frontal bilateral burr holes, 1.5-mm-
diameter radiofrequency electrodes were introduced to each
target under dynamic impedance monitoring. These were
replaced with DBS leads through the same trajectory to the
target. Separate corticotomies, within the same burr hole, were
used to implant the twoDBS leads. Quadripolar DBS leadswere
used with 0.5-mm separation between contacts for amSTN
leads (3389; Medtronic, Minneapolis, MN) and 1.5-mm sepa-
ration between contacts for VC/VS leads (3387; Medtronic). The
VC/VS lead was planned to locate one contact within the nu-
cleus accumbens core, one contact within its shell, and the
upper two contacts in the most ventral aspect of the anterior
limb of the internal capsule. An immediate stereotactic MRI
verified targeting accuracy (13). Two neurostimulators (Activa
PC; Medtronic) were placed subcutaneously below the collar-
bone, one on each side, and each was connected to bilateral
leads from one of the electrodes via subcutaneous cables.
Randomization and Blinding
Computer-generated pairwise randomization was used so that
equal numbers were recruited to receive amSTN or VC/VS
stimulation first, in a counterbalanced order. Two unblinded
clinicians (TF, PL) held the randomization list and adjusted DBS
parameters. All other team members, ward staff, and partici-
pants were blinded to allocation.
Stimulus Adjustments
Optimal DBS parameters were derived in an iterative fashion
over 2 weeks. Initially, each contact was screened with volt-
ages up to 4 V (amSTN) or 8 V (VC/VS) using monopolar
stimulation (pulse width 60 ms and frequency 130 Hz). Imme-
diate clinical effects from stimulation delivered through each
contact in turn and the threshold associated with positive andl Psychiatry May 1, 2019; 85:726–734 www.sobp.org/journal 727
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 Yearsnegative effects were noted. Anticipated adverse effects of
stimulation included hypomania and anxiety (9,10) and were
documented with a visual analog scale. Stimulation parame-
ters were refined daily according to patient feedback, visual
analog scale, and clinical assessment, including the use of
stimulation through multiple monopolar electrode contacts per
lead or using a bipolar configuration.
Clinical and Cognitive Assessments
The Y-BOCS was the primary outcome measure to test VC/VS
and amSTN DBS effects on OCD symptoms. To test the
mechanistic hypothesis, secondary measures were the
Montgomery–Åsberg Depression Rating Scale (MADRS) to
evaluate mood and the Cambridge Neuropsychological Test
Automated Battery Intra-Extra Dimensional Set-Shift (IED) task
to evaluate cognitive flexibility. In the IED task, participants
progress through nine stages assessing the ability to learn and
reverse rules governing correct responses using computer
feedback (14). In stages 1 to 7, responses to a specific stim-
ulus dimension are correct. The ability to shift attention away
from the previously correct stimulus dimension to a different
dimension (i.e., cognitive flexibility) is tested in extradimen-
sional set-shifting task stage 8 (EDS).
Volumes of Tissue Activation
SureTune, Version 2 (Medtronic), a DBS therapy planning plat-
form, was used to model activation volumes around individual
contacts. This applies neuron models coupled to finite element
simulations to generate DBS therapy activation volumes (15).
The preprocessed postoperative magnetization prepared rapid
acquisition gradient echo scans were manually aligned with the
preimplantation magnetization prepared rapid acquisition
gradient echo scans (see Supplement). Automatic coregistra-
tion was carried out with a restricted volume of fusion cantered
on the mesencephalon, diencephalon, and VS to minimize
registration error resulting from brain shift incurred during sur-
gery despite minimal brain shift with our surgical technique (16).
Registration accuracy was inspected and the process iterated if
necessary. All volumeswere realignedwith a planeparallel to the
anterior commissure-posterior commissure line.
Thepostimplantationmagnetizationprepared rapidacquisition
gradient-echo scan was used to fit the DBS leadmodel within the
MRI artifact produced by the leads. For each patient, DBS pa-
rameters during the 12-week phase when each set of electrodes
was active and optimized were used to generate activation vol-
umes around active DBS contacts in the amSTN and VC/VS with
corresponding voltages. Binary image files of activation volumes
with corresponding transformation matrices were exported and
processed in MATLAB, version R2016A (The MathWorks, Inc.,
Natick,MA) using an in-housesoftware to generateNeuroimaging
Informatics Technology Initiative volumes for further analysis.
Volumes corresponding to the active electrodes for each target
weremerged using Fslmerge (FSL v5.0) into a 4-dimensional data
file for each hemisphere. Fslmaths (FSL v5.0) was then used to
generate group average volumes of tissue activation.
Connectivity Analysis
Preprocessed data (see Supplement) were fed into BedpostX
(FSL v5.0) to estimate fiber orientations. Up to three crossing728 Biological Psychiatry May 1, 2019; 85:726–734 www.sobp.org/joufibers were estimated in each brain voxel using model 2 and
graphics processing unit parallelization (17,18). Probtrackx
was used on these estimates to obtain global connectivity
(i.e., the probability of the existence of a path through the
diffusion field between any two distant points, a surrogate
measure of anatomical connectivity) (19). Probabilistic trac-
tography was generated in ProbtrackX2 graphics processing
unit version (FSL v5.0) (number of samples = 5000, curvature
threshold = 0.2, step length = 0.5 mm, subsidiary fiber volume
fraction threshold = 0.01) (19). The process repetitively sam-
ples from the distributions of voxelwise principal diffusion
directions generated in BedpostX, each time computing a
streamline through these local samples to generate a prob-
abilistic streamline or a sample from the distribution on the
location of the true streamline, building up a spatial connec-
tivity distribution. Streamlines truly represent paths of minimal
hindrance to diffusion of water in the brain, but they are
reasonable indirect estimates of long-range white matter
connections (20).
Probabilistic tractography streamlines were generated for
each patient in native space. Individual patient DBS VTAs were
used as seed masks resulting in four tracts per patient from the
amSTN and the VC/VS bilaterally. The corpus callosum was
used as an exclusion mask. Cerebrospinal fluid termination
masks were used to exclude false positive streamlines. Using
the obtained transformations to and from standard space,
resulting streamlines were transformed to Montreal Neurolog-
ical Institute space, and group averages were generated. The
Human Central Nervous System: A Synopsis and Atlas (21) was
used to corroborate the relevant anatomical structures.
Statistical Analysis
Friedman’s test was used to test for DBS effects during the
double-blind crossover phases comparing baseline, amSTN,
and VC/VS. Significance levels were adjusted for multiple
comparisons using the false discovery rate (FDR) method of
Benjamini and Hochberg (22). Post hoc pairwise Conover tests
for significant effects were used (23), with FDR corrections also
applied. For each variable, the effect of time, stimulation at
both sites, and adjunctive CBT was assessed with Friedman’s
test across all six time points independent of stimulation type.
To test the effect of DBS of amSTN and VC/VS together,
Wilcoxon’s tests were used to compare phases when one site
was stimulated (time 2 1 time 3) and when both sites were
stimulated (time 4 1 time 5) and the effect of CBT, by
comparing time 5 and time 6. FDR corrections of significance
values were applied throughout. To substantiate our effects,
we repeated these analyses using more powerful parametric
statistics, again correcting for multiple comparisons (see
Supplement).
RESULTS
Six participants were recruited from NHS England Specialised
OCD Services by HT, NF, LMD, and KM. Five were men; the
age range was 38 to 62 years and the duration of illness was
20 to 30 years (Table 1). The study took place between January
10, 2013, and October 25, 2016. During the trial, patient 2
required further DBS adjustment during the amSTN, VC/VS,
and COMB phases and required additional medication forrnal
Table 1. Patient Details
Patient
No./Sex
Age at
Onset,
Years
Age at
Surgery,
Years
Illness
Duration,
Years Disability at Study Entry Medication
1/Female 16 38 22 Living in inpatient unit; failed supportive
accommodation
Escitalopram 20 mg; risperidone 0.5 mg; trazodone
150 mg; clomipramine 200 mg
2/Male 16 38 22 Extreme avoidance; impaired social function Tramadol 200 mg; memantine 35 mg; pregabalin
300 mg; citalopram 120 mg; quetiapine 25 mg;
clonazepam 0.75 mg
3/Male 32 62 30 Largely housebound; requiring help with ADLs Aspirin 75 mg; omeprazole 20 mg; sitagliptin 100 mg;
nitrazepam 10 mg as needed
4/Male 17 37 20 Largely housebound; severely impaired in ADLs Escitalopram 40 mg; aripiprazole 10 mg
5/Male 32 55 23 Largely housebound; unable to live independently Aripiprazole 20 mg; chlorpromazine 400 mg; pregabalin
300 mg; propranolol 40 mg; sertraline 400 mg;
zopiclone 7.5 mg
6/Male 15 43 28 Extreme avoidance; impaired social function Pregabalin 600 mg; aripiprazole 20 mg; sertraline 200 mg
ADLs, activities of daily living.
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 Yearsmood instability. Patient 6 required DBS adjustment during the
OPT phase because of worsening of OCD. Patient 1’s OCD
symptoms improved during VC/VS DBS but worsened when
switched to amSTN. Stimulation adjustment over 2 weeks
made no improvement. Because of patient distress, stimula-
tion reverted to the VC/VS site. The scores entered for the
amSTN stage for this participant were at the point of switching
from amSTN back to VC/VS, akin to last entry carried forward.
This participant continued with subsequent trial phases per
protocol.
Comparison of VC/VS and amSTN DBS
All participants completed the IED task, and the only signifi-
cant DBS effect was an improvement in EDS errors. There
were significant improvements in Y-BOCS score (Table 2),
MADRS score, and EDS errors following DBS (Figure 1). All
Friedman tests were significant, controlling for FDR: Y-BOCS
score (c22 = 9.0, p = .017), MADRS score (c
2
2 = 10.33; p =
.017), EDS errors (c22 = 7.00, p = .03). Y-BOCS scores
significantly improved following both amSTN and VC/VS DBS
(baseline vs. amSTN: p , .001; baseline vs. VC/VS: p , .001;
amSTN vs. VC/VS: p = 1.00). Changes in MADRS scores were
significantly different from baseline for both amSTN and VC/VS
DBS. The magnitude of the amSTN effect was significantly
greater than VC/VS (baseline vs. amSTN: p = .023; baseline vs.Table 2. Y-BOCS Scores at Baseline and During Five Stimulatio
Patient Baseline amSTN VC/V
1 38 32 (16)a 22 (4
2 34 26 (23) 29 (1
3 37 17 (55) 18 (5
4 38 20 (47) 13 (6
5 34 23 (32) 17 (5
6 36 1 (97) 3 (9
Mean 6 SEM 36.17 6 0.75 19.83 6 4.32 17.00 6
Values in parentheses indicate % reduction from baseline. amSTN deep
AdCBT, optimal combined settings plus adjunctive cognitive behavioral
phase; OPT, optimal combined settings; VC/VS, ventral capsule/ventral str
aThis score is last entry carried forward. Y-BOCS scores range fro
0–7 = subclinical; 8–15 = mild; 16–23 = moderate; 24–31 = severe; 32–40 =
BiologicaVC/VS: p, .001; amSTN vs. VC/VS: p = .001). Changes in EDS
errors were significant for amSTN but not VC/VS DBS (baseline
vs. amSTN: p = .003; baseline vs. VC/VS: p = .157; amSTN vs.
VC/VS: p = .018). Parametric analyses substantiated these
effects other than the post hoc comparison of baseline and
amSTN DBS on the MADRS, which was not significant (see
Supplement).
Previous studies have defined a DBS response as $35%
reduction in baseline Y-BOCS (5–8). The proportion of patients
achieving responder status at the end of each phase was the
following: amSTN phase, 3 of 6; VC/VS phase, 5 of 6; COMB
phase, 5 of 6; OPT phase, 6 of 6; OPT plus adjunctive CBT
phase, 6 of 6.
Effects of Time, Combined DBS, and Adjunctive
CBT
Following FDR corrections across all three Friedman’s tests,
there was a significant improvement in Y-BOCS and MADRS
scores over time, and EDS errors did not change: Y-BOCS
score (c25 = 24.11, p , .001), MADRS score (c
2
5 = 16.49,
p = .006), EDS errors (c25 = 6.59, p = .253). Following FDR
corrections of post hoc comparisons, there were no statisti-
cally significant changes when combined DBS of both sites
was compared with single-site stimulation (Y-BOCS: p = .116;
MADRS: p = .146) or when DBS at both sites was comparedn Phases
S COMB OPT AdCBT
2) 17 (55) 18 (53) 13 (66)
5) 23 (32) 20 (41) 21 (38)
1) 12 (68) 10 (73) 2 (95)
6) 16 (58) 10 (74) 7 (82)
0) 17 (50) 15 (56) 13 (62)
2) 0 (100) 13 (64) 0 (100)
3.57 14.17 6 3.18 14.33 6 1.69 9.33 6 3.21
brain stimulation was the initial condition for patients 4, 5, and 6.
therapy; amSTN, anteromedial subthalamic nucleus; COMB, combined
iatum; Y-BOCS, Yale-Brown Obsessive Compulsive Scale.
m 1 to 40 and are categorized according to severity as follows:
extreme.
l Psychiatry May 1, 2019; 85:726–734 www.sobp.org/journal 729
OCD Symptoms
(Y-BOCS)
Depressed Mood
(MADRS)
Cognitive Inflexibility
(EDS Log Errors)
NS
NS
Baseline
amSTN
VC/VS
Both
Optimal
Post CBT
0 2010 30 40 0 20 40 60 0 0.2 0.4 0.6 0.8 1 1.2 1.4
Figure 1. Mean Yale-Brown Obsessive Compulsive Scale (Y-BOCS) score, Montgomery–Åsberg Depression Rating Scale (MADRS) score, and extra-
dimensional set-shifting (EDS) log errors at baseline and following deep brain stimulation phases: anteromedial subthalamic nucleus (amSTN), ventral capsule/
ventral striatum (VC/VS); combined amSTN and VC/VS DBS (Both); optimal combined settings (Optimal); optimal combined settings plus cognitive behavioral
therapy (Post CBT). The amSTN and VC/VS phases followed a randomized counterbalanced design. Both the Optimal and Post-CBT phases were open and
sequential. Y-BOCS scores range from 1 to 40: 0–7 = subclinical; 8–15 = mild; 16–23 = moderate; 24–31 = severe; 32–40 = extreme. MADRS scores range from
1 to 60: 0–6 = normal; 7–19 = mild; 20–34 = moderate; 35–60 = severe. *p , .05; **p , .01; ***p # .001. NS, not significant; OCD, obsessive-compulsive
disorder.
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 Yearswith adjunctive CBT (Y-BOCS: p = .092; MADRS p = .223).
Parametric analyses substantiated these findings (see
Supplement).
Amplitude of Active Contacts VTAs
Postoperative stereotactic MRI verified that amSTN electrodes
had at least two contacts within the amSTN and that VC/VS
electrodes had two or three contacts within the VC and one or
two contacts in the nucleus accumbens. For VC/VS, the most
dorsal contacts produced the optimal clinical response for all
patients, and the mean stimulation amplitude was 5.85 6 1.2 V
(Supplemental Tables S1 and S2). For the amSTN, the deepest
contacts were most effective for all patients, and the mean
stimulation amplitude was 1.56 6 0.82 V (Supplemental Tables
S1 and S2). The average VC/VS VTA was centered on the white
matter in the ventral portion of the anterior limb of the internal
capsule and encroached slightly on adjacent portions of the
nucleus accumbens, head of caudate, globus pallidus, and
putamen (Figure 2). The average amSTN VTA was centered on
the anterior-inferior medial border of the STN spreading into
the ventral tegmental area (Figure 2). Exploratory multiple re-
gressions showed that changes in Y-BOCS score, MADRS
score, or EDS errors did not predict STN and VC volumes of
tissue activation following STN and VC DBS, respectively
(range of B values: 20.750 to 0.625; range of t values: 21.988
to 1.254).
Tractography
The average streamlines generated from individual amSTN
VTAs were connected to the lateral orbitofrontal cortex (OFC),
dorsal anterior cingulate cortex (DACC), dorsolateral prefrontal
cortex (DLPFC), and medial forebrain bundle. The average
streamlines generated from individual VC VTAswere connected
to the medial OFC, the mediodorsal thalamus, the amygdala via
the amygdalofugal pathway, the hypothalamus, and the habe-
nula via the habenulointerpeduncular tract (Figure 3A, B). Indi-
vidual patient streamlines are shown in Supplemental Figure S2.730 Biological Psychiatry May 1, 2019; 85:726–734 www.sobp.org/jouAdverse Events
There were no adverse events associated with surgery. The
most common adverse event during the DBS trial was hypo-
mania (elevated mood or irritability, racing thoughts, disinhi-
bition), occurring within hours after stimulation adjustment and
remedied by further adjustment. This was witnessed twice
each during amSTN and VC/VS DBS and three times during
the COMB phase. One patient reported hypomania-like
symptoms during the OPT phase on a more sustained basis
(racing thoughts and urges to steal) and required frequent
admissions for adjustment of DBS for unstable mood. The
VC/VS battery became depleted and was replaced once in 3
patients and twice in 1 patient during open phases of the
study, without surgical complications (Supplemental Table S3).
DISCUSSION
This is the first study to compare, in the same patients, DBS of
two brain sites for severe OCD. In a within-subjects, random-
ized, double-blind, counterbalanced comparison of amSTN
and VC/VS stimulation in 6 patients, we were able to address
important questions regarding the efficacy and mechanism of
DBS at each site. All patients had been ill for at least 20 years
and failed to respond to high doses of medication plus inten-
sive CBT. DBS significantly reduced OCD symptoms during
stimulation of the VC/VS and amSTN, and the magnitude of the
reduction at each site, measured with the Y-BOCS, did not
differ. In contrast, there were different effects of DBS on mood
and cognition. DBS of the amSTN, but not the VC/VS, signif-
icantly improved cognitive flexibility (EDS errors). DBS of the
VC/VS elicited a striking improvement in mood, measured by
the MADRS, which was to a greater degree than amSTN DBS.
Tractography from optimally activated electrode contacts at
each site suggests that these dissociated effects reflected
DBS modulation of distinct brain networks.
Many studies suggest that orbitofrontal (OFC) cortico-
striato-thalamic circuitry is dysfunctional in OCD. It is now
recognized that the medial and lateral aspects occupyrnal
Figure 2. Average deep brain stimulation volume of tissue activation (VTA) in the ventral capsule (VC-VTA) and anteromedial subthalamic nucleus
(amSTN-VTA).
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 Yearsseparate trajectories within the OFC cortico-striato-thalamic
circuitry and are functionally different (24). Abnormal func-
tional connectivity between lateral OFC (Brodmann areas 10,
11, 47) and caudate nucleus is associated with EDS errors in
OCD (25). Compatible with this, amSTN DBS improves glucose
metabolism in OFC (Brodmann areas 10, 11) in association
with better Y-BOCS scores (26). The EDS tests the ability to
adapt to changing circumstances by shifting the focus of
attention and acquiring new responses (i.e., cognitive flexi-
bility); this ability is a putative endophenotype of OCD, being
also impaired in first-degree unaffected relatives (11). These
findings are compatible with the observations of this study that
the amSTN DBS site was associated with improved EDS
performance and that tractography streamlines from activated
contacts connected to the lateral OFC.
Connectivity analysis also showed streamlines linking the
DLPFC and DACC with the amSTN. In addition to the OFC, the
function of these two cortical areas has been linked to
the severity of OCD symptoms measured by the Y-BOCS. For
example, in a functional MRI study, reduced functional con-
nectivity between the DLPFC and putamen was associated
with both impaired goal-directed planning and OCD symptom
severity measured by the Y-BOCS (25). In addition, ablation of
the DACC, performed during cingulotomy, is an established
neurosurgical procedure shown to alleviate symptoms in pa-
tients with medically refractory OCD (27).
In nonhuman primates, the amSTN receives uninterrupted
projections from the OFC, DLPFC, and DACC, considered toBiologicaform the limbic and cognitive components of the hyperdirect
pathway (28). The STN is integral to basal ganglia circuitry, and
input from the hyperdirect pathway provides a mechanism
whereby cortical influences can rapidly suppress the mani-
festation of behaviors that are already being programmed (29).
Taking all findings together, amSTN DBS may regulate aber-
rant information processing in the hyperdirect pathway from
the lateral OFC, DLPFC, and DACC (28) and enable patients to
interrupt their compulsive cycle of repetitive acts and thoughts
through improving cognitive flexibility and goal-directed
planning.
For all patients, the VC/VS dorsal-most electrode contacts
were the most effective, suggesting that stimulation of the
ventral anterior limb of the internal capsule (VC), rather than the
VS/nucleus accumbens, mediated the clinical effect of
improved OCD and mood. This was confirmed by the volume
of tissue activation, which was centered on the VC with only
minor encroachment on the surrounding gray matter. Other
studies support this conclusion (8,30).
Nonhuman primate tracing studies, shown to substantiate
human tractography (31), suggest that VC DBS captures fibers
from the medial OFC to the thalamus and more posterior areas
(32,33). This is in keeping with the tractography finding here of
a streamline extending anteriorly to the medial OFC and pos-
teriorly to the dorsomedial thalamus, as well as streamlines
involving the amygdalofugal pathway and habenulointerpe-
duncular tract. The medial OFC is hyperactive during the early
processing of threat-related stimuli in OCD (34). Thel Psychiatry May 1, 2019; 85:726–734 www.sobp.org/journal 731
Figure 3. (A) Group average streamlines from
anteromedial subthalamic nucleus volumes of tissue
activation (amSTN-VTA). (B) Group average stream-
lines from ventral capsule volumes of tissue activa-
tion (VC-VTA). DACC, dorsal anterior cingulate
cortex; DLPFC, dorsolateral prefrontal cortex; IC,
internal capsule; LAT, lateral; MED, medial; MFB,
medial forebrain bundle; OFC, orbitofrontal cortex.
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 Yearsamygdalofugal pathway is an output tract from the basolateral
nucleus of the amygdala to the mediodorsal thalamus and
OFC. A recent functional MRI study of OCD found specific,
abnormal connectivity between the basolateral amygdala and
medial OFC in OCD, which was predictive of successful
treatment with CBT (35). The habenulointerpeduncular tract
has been implicated in the development of depression via in-
hibition of brainstem serotonergic raphe nuclei (36). In OCD,
selective serotonin reuptake inhibitors increase serotonin
neurotransmission and are the mainstay of pharmacological
treatment because they improve both mood and core OCD732 Biological Psychiatry May 1, 2019; 85:726–734 www.sobp.org/jousymptoms (37). Thus, modulation by VC DBS of circuitry
involving the medial OFC, the associated amygdalofugal
pathway, and the habenulointerpeduncular tract may mediate
the striking improvement in mood as well as OCD symptoms.
We anticipated that a positive response to DBS would
enable patients to utilize CBT and boost the OCD effect further
(38), but this was not found. However, inspection of individual
Y-BOCS changes before CBT showed that 4 of 6 patients
were already at a mild level of OCD symptom severity, thus
making it difficult for further measurable gains to be made
during CBT.rnal
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 YearsAdverse effects were mainly stimulation-induced hypoma-
nia, also reported in other studies. These were not more
common during stimulation of either site or combined
stimulation.
There are several limitations to the study, the main one
being the small sample size. However, when comparing the
efficacy of the two DBS sites, patients served as their own
controls in an innovative design, and the conclusions were
robust to adjustment for multiple comparisons and parametric
and nonparametric analyses. Nevertheless, it would be
important to test our hypothesis in a larger group of patients
when the mechanistic actions of STN and VC DBS on recovery
from OCD can be evaluated in more detail. Another limitation is
the possible confounding effect of combined stimulation and
CBT with time. Future studies could compare the effect of
additional CBT at an earlier stage.
In summary, DBS of the VC and amSTN significantly alle-
viated OCD symptoms, and the magnitude of effect did not
differ between these sites, suggesting that both targets are
equally efficacious. The finding that amSTN but not VC DBS
improved cognitive flexibility and that the effect of DBS on
mood was significantly greater for VC DBS implicates the
involvement of different neural circuitries associated with
distinct symptoms in OCD. Tractography findings revealed that
VC and amSTN DBS modulate distinct brain networks impli-
cated in OCD and are compatible with these clinical and
cognitive observations.ACKNOWLEDGMENTS AND DISCLOSURES
This work was supported by Medical Research Council Grant No. MR/
J012009/1, National Institute for Health Research University College London
Hospitals Biomedical Research Centre (to EMJ), the Monument Trust and
the Parkinson’s Appeal UK (to MH, LZ, TF, PL), the Brain Research Trust (to
HA), Wellcome Trust Award Grant No. WT 104631/Z/14/Z (to TWR), and the
National Institute for Health Research Cambridge Biomedical Research
Centre mental health theme (to TWR, BJS, AMA-S).
We thank the patients who participated in the study, Lesley Morton
and staff on Hughlings Jackson ward, and the Queen Square DBS
nursing team. The full trial protocol is available at https://www.ucl.ac.uk/
ion/research/departments/motor-neuroscience-and-movement-disorders/
research/professor-eileen-joyce.
TF has received honoraria for speaking at meetings sponsored by Bial,
Profile Pharma, and Britannia Pharmaceuticals and serving on advisory
boards for Bial and Oxford Biomedica. NAF has received honoraria for
speaking at meetings from Abbott, Wiley, and the British Association
for Psychopharmacology; financial support to attend scientific meetings
from the International Society for Affective Disorders, International Forum
of Mood and Anxiety Disorders, European College of Neuro-
psychopharmacology, British Association for Psychopharmacology, and the
Royal College of Psychiatrists; financial royalties for publications fromOxford
University Press; and payment for editorial duties from Taylor and Francis.
KM has received honoraria for chairing advisory boards sponsored by
Medtronic; received research project funding from Merck Serono, Lundbeck,
Reckitt Benckiser, St. JudeMedical (Abbott), and Indivior; and received travel
and accommodation support from Medtronic and St Jude Medical (Abbott)
to attend scientific meetings. TWR has received royalties from Cambridge
Cognition; consulted for Cambridge Cognition, Lundbeck, Mundipharma,
and Unilever; and received research grants from Lundbeck and Shionogi.
BJS has received honoraria as a consultant to Cambridge Cognition, PEAK,
and Mundipharma. MH, PL, and LC have has received honoraria and travel
expenses from Medtronic, St. Jude, and Boston Scientific for speaking at
meetings. EMJ has received funding from Boston Scientific, the National
Institutes of Health, the International College of Neuropharmacology, and the
European College of Neuropsychopharmacology to attend meetings. TheBiologicaother authors report no biomedical financial interests or potential conflicts of
interest.
ISRCTN: Deep Brain Stimulation for Severe Obsessive Compulsive
Disorder; http://www.isrctn.com/ISRCTN18430630; ISRCTN18430630.ARTICLE INFORMATION
From the Department of Clinical and Movement Neurosciences (HT, HA, TF,
PL, MJ, LZ, MH, EMJ), University College London Queen Square Institute of
Neurology, London; The National Hospital for Neurology and Neurosurgery
(HT, HA, TF, PL, LZ, MH, EMJ), London; Highly Specialised Service for OCD
and BDD (England) (LMD), SW London and St George’s NHS Trust, London;
Departments of Psychology (AMA-S, TWR) and Psychiatry (BJS), Behav-
ioural and Clinical Neuroscience Institute, University of Cambridge, Cam-
bridge; Highly Specialised Service for OCD and BDD (England) (NAF),
Hertfordshire Partnership University NHS Foundation Trust, Welwyn Garden
City; Centre for Clinical & Health Research Services (NAF), School of Life
and Medical Sciences, University of Hertfordshire, Hatfield; Division of
Neuroscience (KM), School of Medicine, University of Dundee, Dundee; and
Advanced Interventions Service (KM), NHS Tayside, Ninewells Hospital and
Medical School, Dundee, United Kingdom.
Address correspondence to Eileen M. Joyce, FRCPsych, Ph.D., Box 19,
National Hospital for Neurology and Neurosurgery, Queen Square, London
WC1N 3BG, United Kingdom; E-mail: e.joyce@ucl.ac.uk.
Received Mar 29, 2018; revised Dec 13, 2018; accepted Jan 3, 2019.
Supplementary material cited in this article is available online at https://
doi.org/10.1016/j.biopsych.2019.01.017.REFERENCES
1. Skapinakis P, Caldwell DM, Hollingworth W, Bryden P, Fineberg NA,
Salkovskis P, et al. (2016): Pharmacological and psychotherapeutic
interventions for management of obsessive-compulsive disorder in
adults: A systematic review and network meta-analysis. Lancet Psy-
chiatry 3:730–739.
2. Boschen MJ, Drummond LM (2012): Community treatment of se-
vere, refractory obsessive-compulsive disorder. Behav Res Ther
50:203–209.
3. Matthews K, Eljamel MS (2003): Status of neurosurgery for mental
disorder in Scotland. Selective literature review and overview of cur-
rent clinical activity. Br J Psychiatry 182:404–411.
4. Nuttin B, Cosyns P, Demeulemeester H, Gybels J, Meyerson B (1999):
Electrical stimulation in anterior limbs of internal capsules in patients
with obsessive-compulsive disorder. Lancet 354:1526.
5. Abelson JL, Curtis GC, Sagher O, Albucher RC, Harrigan M, Taylor SF,
et al. (2005): Deep brain stimulation for refractory obsessive-compulsive
disorder. Biol Psychiatry 57:510–516.
6. Goodman WK, Foote KD, Greenberg BD, Ricciuti N, Bauer R, Ward H,
et al. (2010): Deepbrain stimulation for intractable obsessive compulsive
disorder: Pilot study using a blinded, staggered-onset design. Biol
Psychiatry 67:535–542.
7. Denys D, Mantione M, Figee M, van den Munckhof P, Koerselman F,
Westenberg H, et al. (2010): Deep brain stimulation of the nucleus
accumbens for treatment-refractory obsessive-compulsive disorder.
Arch Gen Psychiatry 67:1061–1068.
8. Luyten L, Hendrickx S, Raymaekers S, Gabriels L, Nuttin B (2016):
Electrical stimulation in the bed nucleus of the stria terminalis alleviates
severe obsessive-compulsive disorder. Mol Psychiatry 21:1272–1280.
9. Greenberg BD, Gabriels LA, Malone DA Jr, Rezai AR, Friehs GM,
Okun MS, et al. (2010): Deep brain stimulation of the ventral internal
capsule/ventral striatum for obsessive-compulsive disorder: World-
wide experience. Mol Psychiatry 15:64–79.
10. Mallet L, Polosan M, Jaafari N, Baup N, Welter ML, Fontaine D, et al.
(2008): Subthalamic nucleus stimulation in severe obsessive-
compulsive disorder. N Engl J Med 359:2121–2134.
11. Chamberlain SR, Fineberg NA, Menzies LA, Blackwell AD,
Bullmore ET, Robbins TW, et al. (2007): Impaired cognitive flexibility
and motor inhibition in unaffected first-degree relatives of patients with
obsessive-compulsive disorder. Am J Psychiatry 164:335–338.l Psychiatry May 1, 2019; 85:726–734 www.sobp.org/journal 733
Ventral Capsule and Subthalamic Nucleus DBS for OCD
Biological
Psychiatry:
Celebrating
50 Years12. Holl EM, Petersen EA, Foltynie T, Martinez-Torres I, Limousin P,
Hariz MI, et al. (2010): Improving targeting in image-guided frame-
based deep brain stimulation. Neurosurgery 67:437–447.
13. Zrinzo L, Yoshida F, Hariz MI, Thornton J, Foltynie T, Yousry TA, et al.
(2011): Clinical safety of brain magnetic resonance imaging with
implanteddeepbrain stimulationhardware: Largecase series and review
of the literature. World Neurosurg 76:164–172; discussion 169-173.
14. Sahakian BJ, Owen AM (1992): Computerized assessment in neuro-
psychiatry using CANTAB: Discussion paper. J R SocMed 85:399–402.
15. Astrom M, Zrinzo LU, Tisch S, Tripoliti E, Hariz MI, Wardell K (2009):
Method for patient-specific finite element modeling and simulation of
deep brain stimulation. Med Biol Eng Comput 47:21–28.
16. Petersen EA, Holl EM, Martinez-Torres I, Foltynie T, Limousin P,
Hariz MI, et al. (2010): Minimizing brain shift in stereotactic functional
neurosurgery. Neurosurgery 67:ons213–ons221; discussion ons221.
17. Jbabdi S, Sotiropoulos SN, Savio AM, Grana M, Behrens TE (2012):
Model-based analysis of multishell diffusion MR data for tractography:
How to get over fitting problems. Magn Reson Med 68:1846–1855.
18. Hernandez M, Guerrero GD, Cecilia JM, Garcia JM, Inuggi A, Jbabdi S,
et al. (2013): Accelerating fibre orientation estimation from diffusion
weighted magnetic resonance imaging using GPUs. PLoS One
8:e61892.
19. Behrens TE, Berg HJ, Jbabdi S, Rushworth MF, Woolrich MW (2007):
Probabilistic diffusion tractography with multiple fibre orientations:
What can we gain? Neuroimage 34:144–155.
20. Jbabdi S, Johansen-Berg H (2011): Tractography: Where do we go
from here? Brain Connect 1:169–183.
21. Nieuwenhuys R, Voogd J, van Huijzen C (2008): The Human Central
Nervous System: A Synopsis and Atlas, 4th ed. Berlin, Germany:
Springer, 2008.
22. Benjamini Y, Hochberg Y (1995): Controlling the false discovery rate: A
practical and powerful approach to multiple testing. J R Stat Soc
Series B Stat Methodol 57:289–300.
23. Pohlert T (2014): The Pairwise Multiple Comparison of Mean Ranks
Package (PMCMR). R Package. Available at: https://CRAN.R-project.
org/package=PMCMR. Accessed August 7, 2018.
24. Fettes P, Schulze L, Downar J (2017): Cortico-Striatal-thalamic loop
circuits of the orbitofrontal cortex: Promising therapeutic targets in
psychiatric illness. Front Syst Neurosci 11:25.
25. Vaghi MM, Vertes PE, Kitzbichler MG, Apergis-Schoute AM, van der
Flier FE, Fineberg NA, et al. (2017): Specific frontostriatal circuits for
impaired cognitive flexibility and goal-directed planning in obsessive-
compulsive disorder: Evidence from resting-state functional connec-
tivity. Biol Psychiatry 81:708–717.
26. Le Jeune F, Verin M, N’Diaye K, Drapier D, Leray E, Du Montcel ST,
et al. (2010): Decrease of prefrontal metabolism after subthalamic734 Biological Psychiatry May 1, 2019; 85:726–734 www.sobp.org/joustimulation in obsessive-compulsive disorder: A positron emission
tomography study. Biol Psychiatry 68:1016–1022.
27. Dougherty DD, Baer L, Cosgrove GR, Cassem EH, Price BH,
Nierenberg AA, et al. (2002): Prospective long-term follow-up of 44
patients who received cingulotomy for treatment-refractory obsessive-
compulsive disorder. Am J Psychiatry 159:269–275.
28. Haynes WI, Haber SN (2013): The organization of prefrontal-
subthalamic inputs in primates provides an anatomical substrate
for both functional specificity and integration: Implications for Basal
Ganglia models and deep brain stimulation. J Neurosci 33:4804–
4814.
29. Aron AR, Herz DM, Brown P, Forstmann BU, Zaghloul K (2016):
Frontosubthalamic circuits for control of action and cognition.
J Neurosci 36:11489–11495.
30. van den Munckhof P, Bosch DA, Mantione MH, Figee M, Denys DA,
Schuurman PR (2013): Active stimulation site of nucleus accumbens
deep brain stimulation in obsessive-compulsive disorder is localized in
the ventral internal capsule. Acta Neurochir Suppl 117:53–59.
31. Safadi Z, Grisot G, Jbabdi S, Behrens TE, Heilbronner SR,
McLaughlin NCR, et al. (2018): Functional segmentation of the anterior
limb of the internal capsule: Linking white matter abnormalities to
specific connections. J Neurosci 38:2106–2117.
32. Haber SN, Knutson B (2010): The reward circuit: Linking primate
anatomy and human imaging. Neuropsychopharmacology 35:4–26.
33. Lehman JF, Greenberg BD, McIntyre CC, Rasmussen SA, Haber SN
(2011): Rules ventral prefrontal cortical axons use to reach their
targets: Implications for diffusion tensor imaging tractography and
deep brain stimulation for psychiatric illness. J Neurosci 31:10392–
10402.
34. Apergis-Schoute AM, Gillan CM, Fineberg NA, Fernandez-Egea E,
Sahakian BJ, Robbins TW (2017): Neural basis of impaired safety
signaling in obsessive compulsive disorder. Proc Natl Acad Sci U S A
114:3216–3221.
35. Fullana MA, Zhu X, Alonso P, Cardoner N, Real E, Lopez-Sola C, et al.
(2017): Basolateral amygdala-ventromedial prefrontal cortex connec-
tivity predicts cognitive behavioural therapy outcome in adults with
obsessive-compulsive disorder. J Psychiatry Neurosci 42:378–385.
36. Hikosaka O (2010): The habenula: From stress evasion to value-based
decision-making. Nat Rev Neurosci 11:503–513.
37. Fineberg NA, Reghunandanan S, Simpson HB, Phillips KA,
Richter MA, Matthews K, et al. (2015): Obsessive-compulsive disorder
(OCD): Practical strategies for pharmacological and somatic treatment
in adults. Psychiatry Res 227:114–125.
38. Mantione M, Nieman DH, Figee M, Denys D (2014): Cognitive-
behavioural therapy augments the effects of deep brain stimulation in
obsessive-compulsive disorder. Psychol Med 44:3515–3522.rnal