A randomised controlled trial of fMRI-neurofeedback in depression

Neurofeedback (NF) enables patients to develop personal strategies that are
most effective in self-regulating brain areas and networks associated with
mental imagery through the feed-back of signals which reflect their own
specific neural activation patterns. Previous studies suggest that participants
can learn to control activity of brain areas involved in emotion processing
within one or a small number of sessions of NF using real-time functional
magnetic resonance imaging (rt-fMRI) signals. Because approximately a third of
patients with depression do not respond satisfactorily to currently available
(pharmacological or psychological) treatments it will be important to explore
the use of this technique as a potential new add-on treatment for depression,
which is one of the most common mental disorders. The PI*s group has recently
completed a randomised controlled trial comparing an fMRI-neurofeedback
(fMRI-NF) intervention targeting emotion networks with a control intervention
targeting an area involved in the processing of scenes and places (both on top
of standard medication). Although clinical benefits were considerable (over 40%
improvement on a clinical rating scale), they did not differ between the
groups, leaving open the possibility that patients would also have improved to
the same degree with just their standard care (Mehler et al., 2018). The next
step in the evaluation of this new technique will thus be to evaluate the
fMRI-NF protocol targeting emotion networks (NFE) (plus standard care) against
standard care alone. We hypothesise that training to gain control over the
affective system and reinforcing neural correlates of positive emotions will
elevate mood and diminish depressive symptoms that would otherwise remain with
the current standard care.

Primary Objective: To determine whether NFE + standard care is superior to
standard care alone in the treatment of depression.
Secondary Objective(s): To assess correlations between NFE success and brain
activation / clinical changes; to assess changes in brain structure and
activation over the course of the NFE training; to assess NFE effects on other
clinical domains; to assess long-term (6-month follow-up) effects of NFE; to
find predictors of NFE success.

This is a single-blind interventional study, a randomised controlled clinical
trial of NFE + standard care vs. standard care.

NFE will be delivered in 5 sessions. After contact has been made through one of
the collaborating clinical teams, a suitably qualified member of the study team
will provide information about the study. One week after this patients will
give consent and the screening assessment will be conducted to determine
inclusion and exclusion criteria. Patients will then be invited to attend the
baseline assessment after appr. one week and randomized. This also provides
some delay for them to be able to re-consider their agreement and withdraw
their consent before the actual start of the intervention if desired.
Participants will additionally perform valence ratings of positive images used
for localization of neurofeedback regions at the end of this session. After the
baseline assessment, participants will undergo an experience sampling procedure
for one week. For the NFE group this will be followed by four further visits
for fMRI-NF sessions (for the NFE group) in appr. weekly intervals, and a fifth
fMRI-NF session after a further month. Participants will train the
self-regulation strategies used during the neurofeedback also at home. After
appr. two months participants will undergo the experience sampling procedure
again for one week before the post-study visit. The intervention will thus be
concluded after appr. 10 weeks. A further follow-up visit (6 months after
baseline assessment) will conclude the trial. The control group will only
attend the screening, baseline, post-study and follow-up assessments. Before
the start of the full trial we will pilot the assessment and scanning procedure
in up to ten patients. This will only entail the five neurofeedback sessions
and the baseline and follow-up assessment. This phase will allow us to adjust
localizer and feedback parameters for optimal feasibility.

Patients of the neurofeedback group take part in five neurofeedback sessions
(four weekly sessions and one session four weeks after the fourth session). The
fMRI-NF sessions will include anatomical scans, an fMRI localizer (visual
presentation of standardized positive and neutral images to map emotion areas)
and neurofeedback runs during which participants see a thermometer on a screen
and are instructed to upregulate activation in brain areas that were identified
based on their response to the positive images in the localizer procedure.
Success of this upregulation is signaled by the level of the thermometer.
Patients will receive some guidance as to potential mental strategies (e.g.
remembering the positive pictures just seen) but no firm instructions. The
principle of neurofeedback training is that patients optimize the strategies
for self-regulation themselves. Patients will be asked to practice the
successful up-regulation strategies at least twice per week for 10 minutes each
and keep a diary about this.

Both groups will undergo an initial screening of appr. 1.5 hours duration,
three clinical/psychometric assessments of appr. 1.5 hours duration (baseline,
end of intervention (2 months from baseline), follow-up (6 months from
baseline) and two weeks of experience sampling (one week following the baseline
measurement and one week preceding the end of intervention measurement)
consisting of ten brief daily questionnaires (daily load about 10m for six days
per week) with a total duration of appr. 1hour per week. Participants will
additionally rate positive images during the baseline measurement (0.5h).
Participants of the neurofeedback group will undergo 5x2hour scanning sessions
including psychological assessments (debriefing and mood scales). Patients of
the neurofeedback group will also be asked to practice the successful
up-regulation strategies at least twice per week for 10 minutes each and keep a
diary about this (for the eight intervention weeks about 1.5h in total). The
overall time commitment (excluding travel) will thus be appr. 20 hours for the
intervention group and 8.5 hours for the standard care group. Patients will be
compensated for their time (160¤ neurofeedback group, 110¤ control group) and
travel costs. Expected benefits are clinical improvement over and above
standard care in the intervention group. There are no known safety issues
arising from fMRI-neurofeedback over and above general MRI safety requirements
(for which strict guidelines implemented at Scannexus will be followed). The
risk of adverse effects of the fMRI-neurofeedback procedure on patients*
wellbeing is minimal and will be monitored through debriefing after each
session.