Transcranial magnetic stimulation (TMS) for patients with exposure therapy-resistant obsessive-compulsive disorder (OCD): TETRO - a multicenter randomized controlled trial

Obsessive-compulsive disorder (OCD) is a serious and disabling mental disorder
with a lifetime prevalence of 2% (Ruscio et al. 2010). It is characterized by
obsessions and compulsions and is associated with substantial comorbidity and
morbidity (Stein et al. 2019). Obsessions are repetitive and persistent
thoughts, images, impulses or urges that are intrusive and unwanted, and are
commonly associated with anxiety. Compulsions are repetitive behaviors or
mental acts, that the individual feels driven to perform in response to an
obsession. 50% of patients experience initial symptoms already during
childhood. Approximately 50% of patients treated with standard treatments
(exposure therapy with/without medication) fail to respond fully, resulting in
chronicity and poor participation in social and educational/occupational
domains. Disability adjusted life years (DALY) for OCD amount to 329,684 in
Europe (rate 7.9 per 10,000 DALYs, Wittchen et al. 2011). In the US, the
largest contributor to the total cost incurred by OCD was lost productivity
costs (74%), and not health-care costs (DuPont at al. 1995); Dutch data show
that lost productivity costs from several mental disorders account for
approximately 85% of total costs (Smit et al. 2006).
In this study, we aim to fill the gap between the standard treatments
(exposure therapy with/without medication) on the one side and invasive
end-stage strategies (brain surgery) on the other side, using a non-invasive
alternative: repetitive transcranial magnetic stimulation (rTMS). Despite
proven efficacy (Zhou et al. 2017; Rehn et al. 2018), rTMS for OCD is not yet
covered by the Dutch insurance system while rTMS for treatment resistant
depression is. In case of proven cost-effectiveness it will lead to the
addition of rTMS as insured health care for patients with OCD as well.

Usual care & guidelines
The 1st choice treatment for OCD is exposure therapy with response prevention
(ERP), with an effect size of 1.33 (Ost et al. 2015). Although less effective
and relatively unfavorable at the long-term, the use of selective serotonergic
antidepressants (SSRIs) is the pharmacological alternative. Higher doses are
used for OCD than for other anxiety disorders or depression and associated with
more dropout due to adverse effects. The switch to clomipramine (tricylic
antidepressant) or augmentation with antipsychotics are commonly used
pharmacological strategies for ERP- and SSRI-resistant patients, at the cost of
additional side effects. The added effect of clomipramine or antipsychotic to
ERP is limited (Foa et al. 2005; Simpson et al. 2013), while conversely ERP has
a significant added value in cases that already use serotonergic or
antipsychotic medication (Simpson et al. 2013). Brain surgery in the form of
deep brain stimulation (DBS) or lesion surgery is a last-resort invasive
treatment option that is reserved for the extreme treatment-resistant cases
(<1%). This shows the enormous gap between the 50% of patients helped by ERP
(with/without medication) and the 1% receiving DBS.

Existing evidence of effect, usual care
SSRIs have a small to medium effect (weighted mean difference 3.21; overall
effect size d=0.43) in OCD (Soomro et al. 2008), relapse risk following
discontinuation is high (Batelaan et al. 2017) and non-response following
re-instating the antidepressant may occur (Bosman et al. 2018). Antipsychotic
augmentation has a similar-sized effect (weighted mean difference 2.34; overall
effect size d=0.4), with only 1/3 of patients with SSRI-resistant OCD showing a
clinically meaningful effect (Veale et al. 2014; Bloch et al. 2006).
Antipsychotics give serious side effects (e.g, weight gain and metabolic
dysregulation). Within the small group of selected patients with severe
refractory OCD (<1%), approximately 30-50% respond to brain surgery (Hamani et
al. 2014). Possible side effects include intracerebral bleeding/infection,
epilepsy, fatigue, memory problems, irritability, and disinhibition (Luyten et
al. 2016).

The intervention to be investigated, rTMS
Non-invasive neuromodulation using rTMS has been used in OCD on various targets
(Zhou et al. 2017; Rehn et al. 2018). Variation across studies also concerns
stimulation parameters (high vs. low frequency, % resting motor threshold,
number / frequency of sessions). Until recently, in most studies rTMS has been
used as monotherapy. We expect better effects when rTMS is used as adjuvant
therapy to ERP (as is currently also the practice in ongoing experimental rTMS
studies). We here therefore propose to use low-frequency rTMS over the pre-SMA
in combination with ERP.

Existing evidence of effect, rTMS
rTMS as mono-therapy has shown to be effective for OCD, with recent
meta-analyses showing an effect size of 0.71 (Zhou et al. 2017) and 0.79 (Rehn
et al. 2018) compared to sham stimulation. It is unknown what the additive
effect is of rTMS above the effect of ERP alone.

Anticipated cost-effectiveness
No (inter)national studies exist that assess the cost-effectiveness of rTMS in
the treatment of OCD. However, OCD has a similar health-care burden as
depression (Skapinakis et al. 2016) and the treatments for MDD and OCD are
similar (e.g. SSRI*s, CBT and rTMS). For depression, there are three cost
effectiveness studies that concluded that rTMS in the treatment of MDD is
cost-effective after a single failed treatment with an antidepressant (Voigt et
al. 2017; Simpson et al. 2009) and outperforms antidepressants after two failed
antidepressants (Nguyen et al. 2015). TMS is thus a cost-effective treatment
for depressed patients who have not received sufficient benefit from
antidepressant pharmacotherapy and significant cost savings may be expected
relative to the current standard of care.
Although rTMS is expected to be more expensive than usual care on the short
term, we expect that the new intervention will result in higher recovery rates
and better quality of life than usual care. This is expected to result in
reduced OCD treatment duration and increased societal/occupational
participation that will lead to reduced healthcare costs and lost productivity
costs in the long term. In some cases the TMS treatment will replace the
expensive and invasive alternative, i.e. brain surgery/DBS. Taken together, we
expect that the new intervention is cost-effective in comparison with usual
care.

Added value
The proposed treatment intends to fill the gap between standard therapies (ERP
with / without medication) and end-stage invasive strategies. The greatest
added value will be the reduction in disease chronicity, resulting in decreased
morbidity, improved quality of life, increased social and
education/occupational participation and decreased healthcare and productivity
costs for the society. Moreover, I will reduce demoralization in patients and
therapist, who deal with these difficult to treat symptoms.

We aim to fill the gap between the 50% of patients helped by standard therapies
(ERP with/without SSRIs) and the 1% receiving surgery by using a non-invasive
alternative: repetitive transcranial magnetic stimulation (rTMS) to potentiate
the effects of ERP. We will establish the added value of 1Hz rTMS applied over
the preSMA (versus sham rTMS) when combined with ERP in OCD patients, who show
no/insufficient response to ERP (alone or combined with medication).

We propose to conduct a multi-center, double-blind, randomized, sham-controlled
rTMS trial, to determine the efficacy of 1Hz rTMS over the pre-supplementary
motor area (pre-SMA) as adjuvant treatment to ERP in patients with OCD who did
not respond (sufficiently) to ERP alone (with or without medication). Alongside
this randomized controlled trial we will also conduct an economic evaluation.
Assessments will be conducted at baseline, weekly during the 5 (to 7) weeks
treatment, after last treatment session and at 3, 6 and 12 months follow-up.

PICOT:
Population: 250 adult (18 years and older) patients with treatment-resistant OCD
Intervention: 20 (or 24 or 28) sessions of 1-Hz rTMS over the pre-SMA followed
by intensive ERP
Comparator: sham-rTMS with ERP
Outcomes: pre-to-post standardized mean difference in OCD severity (as measured
with the Yale-Brown Obsessive-Compulsive Scale - Y-BOCS), compared to the sham
condition.
Time horizon: we will measure effect of treatment before and immediately after
treatment (i.e. 5, 6 or 7 weeks), and after 3, 6 and 12 months follow-up.

This is a 2-arm RCT with an active adjuvant rTMS treatment to daily ERP; the
comparison condition will use sham rTMS combined with daily ERP. Treatment
duration is adaptive and depending on clinical improvement (see under
Intervention) and will contain 20, 24 or 28 combined rTMS-ERP sessions.
Patients will undergo a baseline structural and functional MRI scan, for
localizing the TMS target region using neuronavigation and for prediction
analyses of response and relapse. rTMS will be performed in a single-blind
fashion, as the experimenter cannot be blind to the stimulation coil; patients,
ERP therapists, clinical assessors will remain blind to treatment conditions
during treatment and follow-up. Outcome measures will be assessed pre-treatment
and post-treatment and at 3, 6 and 12 months follow-up. A selection of the
outcome measures will be assessed weekly during treatment (see outcome
measures).

Treatment will consist of almost daily (4 days/week) rTMS sessions immediately
followed by a 90-minute ERP session. This offers the logistic benefit of
catching up on a missed treatment during a treatment week. The treatment
follows an adaptive design, offering a minimum of 20 sessions to all patients
with a possible extension of 4 or 8 sessions for patients who continue to show
improvement and/or are motivated to continue therapy for optimal treatment
effect. The weekly improvement is measured during the treatment by a weekly
administration of the YBOCS to obtain a severity score after every four
individual sessions.
After the 5 weeks of combined treatment, patients who have achieved
remission (YBOCS<=12) on two consecutive weeks will discontinue the treatment;
patients who have at the end of week 5 reached (partial) response, defined as
>=25% improvement on YBOCS score (Mataix-Cols et al. 2016), will continue, if
motivated, for another week (week 6); patients who have not reached (partial)
response but feel that they can still profit from continued treatment will also
continue. Patients who worsened during treatment will be discontinued. At the
end of week 6 we will again monitor progress using the same criteria, allowing
patients who continue to show steady (partial) response the opportunity to
undergo a seventh week of treatment, amounting to a maximum of 28 sessions per
participant. This design allows patients who are experiencing protracted and
slowly accruing benefit from the repeated combined sessions to maximize their
treatment effect; and it accounts for the slow and linear benefit of repeated
rTMS sessions that has been observed in patients with OCD, who still improve at
or after 20 sessions (Donse et al. 2018, Mantovani et al. 2010). The decision
process on (dis)continuation will follow shared decision making; in case of
doubt, the case will be discussed in the weekly indication & treatment meeting.
Previous meta-analyses have shown conflicting and inconclusive results
about the efficacy of different stimulation targets in OCD (Berlim et al. 2013,
Rehn et al. 2018, Zhou et al. 2017, ). We base our choice of stimulation
location, intensity, duration and frequency on recommendations from
international guidelines and following the updated meta-analysis that we
carried out (Fitzsimmons et al., in preparation). Treatment will thus consist
of low-frequency (1-Hz) rTMS over the pre-supplementary motor area (pre-SMA) in
time-locked combination with ERP, while either unmedicated or on a stably
established dose of medication.

rTMS parameters
The intensity of stimulation will be set at 110% of resting motor threshold,
ascertained using EMG recording of the first dorsal interosseus muscle under
handheld single-pulse stimulation by trained experimenters. Resting motor
threshold is defined as 2 motor evoked potentials of at least 50µV peak-to-peak
amplitude in a series of 4 consecutive single-pulse stimulations. In order to
monitor changes in motor threshold during treatment, we will repeat the
assessment of the motor threshold after 2 and 4 weeks of treatment.
In the verum rTMS condition, we will deliver 1500 continuous 1-Hz
pulses to the pre-SMA, i.e. 25 minutes of stimulation under neuronavigation
(see below). These protocols allow us to obtain a maximum stimulation effect
while conforming to internationally published safety guidelines (Rossi et al.
2009).
The comparison group will thus receive 4 days/week sham-rTMS (using a
sham coil) followed by ERP (unmedicated or on stable dose of medication for at
least 8 weeks) similar to the other treatment arm. Sham-rTMS delivers TMS
pulses of ineffective intensity to the brain while maintaining the scalp
sensation, and therefore is considered a null-treatment, leaving possible
clinical response in this group to the ERP. The location of the sham treatment
will be performed over the same location, i.e. pre-SMA, to further ensure
blinding of the patient to the treatment condition.
TMS coil placement using neuronavigation
Precision and specificity of the rTMS intervention will be ascertained and
maintained throughout the treatment using neuronavigation (Sack et al. 2009),
available in all collaborating centers. The procedure of neuronavigation
requires the use of a high-resolution T1 MRI scan, that will be registered to
the patient*s head upon co-registering visible landmarks on the head of the
patient with landmarks on the MRI scan. This will allow us to perform
stereotaxic placement of the center of the stimulation coil on the
pre-identified target. The use of neuronavigation software will additionally
allow us to record the location of every single pulse of stimulation throughout
the treatment, enabling us to reposition the coil during treatment if needed,
and for post-hoc verification of the precise stimulation location in the
subject*s individual MRI in native space, and at group level in standard space.
Precision of stimulation location will therefore also be available for use in
post-hoc analyses of treatment efficacy. Finally, the neuronavigation
recordings are used to reposition the coil at the sequential visits with
minimal time investment. The use of neuronavigation allows us to place the coil
at the stimulation target with millimeter precision and resolution.
The positioning of the TMS coil on the scalp will be based on the EEG
10-20 coordinate system. In parallel we define three alternative individual
MRI-based pre-SMA coordinates, using:
1) pre-identified coordinates in standard space derived from group activation
studies that activate the pre-SMA. This standard target will then be converted
to each patient*s individual native space, using the inversion of the
conversion matrix from native to standard space. Based on the literature, the
pre-defined target coordinates in standard space for the left pre-SMA are:
x=-4, y=14, z=58 (Norman et al. 2019; Picard and Strick. 2001).
2) coordinate in pre-SMA with strongest anti-correlation with the amygdala
during resting state
3) coordinate in pre-SMA with strongest anti-correlation with the amygdala
during emotional processing (symptom provocation task).
This will allow us to post-hoc map the optimal individualized MRI-based
coordinates relative to the actually stimulated coordinate in order to relate
treatment success to stimulation location (distance between actual and optimal
coordinate). This approach will serve to inform us on the added value of
MRI-based individualized targeting using neuronavigation versus
non-neuronavigated rTMS treatments. This is relevant for use in future clinical
practice.

Requirements for the TMS and neuronavigation devices
In this study we will use CE-certified rTMS equipment with similar pulse shapes
and pulse widths to ascertain an acceptable degree of standardization. Systems
using biphasic TMS pulses and pulse widths between 280-330 µs will be deemed
acceptable (this includes Magstim (330 µs), MagVenture (280 µs) and Deymed
DuoMag (280 µs) systems). All the TMS systems should be outfitted with a MEP
unit to quantify the motor threshold.
For the placebo condition *passive* placebo coils will be used, i.e. coils
identical to the stimulation coils in appearance, but with a built-in metal
plate that effectively blocks the active stimulation while maintaining
mechanical scalp sensation. These placebo coils are available for the systems
mentioned.
For neuronavigation several systems and approaches exist. In this study
we will include only CE-certified systems that achieve coil placement using
frameless stereotaxy, guided by an infrared-emitting and receiving stereoscopic
camera. Available setups in the participating centers are Brainsight (Brainbox
Ltd, UK) and TMSnavigator (Localite, Germany).

Patients will receive a minimum of 20 and maximum of 28 sessions of rTMS (30
minutes per session) adjuvant to ERP (90 minutes per session), 4 days/week for
5 (to 7) weeks. Prior to treatment they will get an MRI (max 1 hour). Screening
is 1 hour, baseline assessment takes half a day (3.5 - 4 hours). Pre-treatment,
weekly during treatment, and post-treatment and at 3, 6 and 12 months
follow-up, they will be assessed for about 1 hour.

Benefits experimental group: severity of OCD (and comorbid affective) symptoms
will probably decrease.
Benefits sham group: severity of OCD (and comorbid affective) symptoms will
probably decrease, even in this group, thanks to the intensive ERP program (4
days/wk).
Risks experimental group: rTMS-related side effects, see paragraph of side
effects (see below).
Risks sham group: risk of suboptimal treatment effect, due to fact that they
only receive intensive ERP.

Side effects of rTMS
TMS is considered safe and generally tolerable. When following the
international safety guidelines (Rossi et al. 2021), the risk to induce an
epileptic convulsion is extremely low. Hearing protection is achieved by
wearing ear plugs during stimulation. Possible adverse events of TMS have been
frequently mentioned but a validated checklist to monitor and quantify adverse
effects does not exist. We recently performed a pilot study with the aim to
systematically objectify possible adverse events of rTMS during treatment for
OCD, using an in-house-made questionnaire (currently used in ongoing
proof-of-concept RCT in OCD as part of the VIDI project of OA van den Heuvel).
This questionnaire consists of three parts, assessing possible adverse events
during, directly after and during the days after an rTMS-treatment session. We
see that adverse events, or co-occurrences, during treatment in OCD patients
are rare, and consist mostly of headache, local scalp pain and sleepiness.
Frequency and severity of these adverse events differs widely between subjects,
and has not led to subjects withdrawing from our earlier trials.