PROSPER - Prediction and Outcome Study in PTSD and Personality disorders

Evidence-based treatments for posttraumatic stress disorder (PTSD), such as Eye
Movement Desensitization and Reprocessing (EMDR) and Imagination and
Rescripting Therapy (ImRs), are highly effective treatments in the majority of
the PTSD patients. PTSD is highly comorbid with personality disorders (PD),
especially borderline personality disorder (BPD), and cluster C - avoidant,
dependent, or obsessive-compulsive - personality disorders (CPD). It is not
clear yet what treatment is most effective for those who suffer from both PTSD
and PD.
There is growing motivation in clinicians to offer PTSD treatments to PTSD with
comorbid PD, because these treatments are highly effective, relatively short
(weekly sessions, 3-6 months) and there is some evidence that with PTSD
treatment, comorbid PD symptoms might resolve as well. PTSD are less
time-consuming than PD treatments and - at least in the short term -
financially attractive. However, at least 30-44% PTSD patients do not
sufficiently respond to these treatments. Moreover, a high number of PTSD
patients are excluded from these therapies because of suicidality,
self-destructive behaviour or other personality problems. Therefore, it might
be more efficient to add a PD treatment at the same time. Evidence-based
treatments for personality disorders (PD), such as dialectical behaviour
treatment (DBT) for BPD, and schema-focused treatment (SFT) for CPD are well
established. These treatments are more intensive (twice a week for at least one
year) than PTSD treatments. There is some evidence that integrated PTSD-PD
treatment is twice as effective on reducing PTSD symptoms than PD treatment
alone, but integrated PTSD-PD treatment is not yet directly compared to PTSD
treatment alone. This study will address this knowledge gap, including
secondary outcome measures on functioning, quality of life and
cost-effectiveness.
The result of this study might be that one or the other treatment works better,
depending on the personal profile of the patient. So far, some psychological
factors have been found to be associated with worse outcome of PTSD treatment.
These are cognitive (educational level, working memory emotion regulation),
affective (anger, sleep problems, dissociation), and relational factors
(therapeutic alliance, attachment, social support). In addition,
neurobiological factors are found to predict PTSD treatment outcome, such as
increased activity connectivity of the limbic network and decreased activity
and connectivity of the cognitive control networks, and disturbed hormonal
levels and epigenetic factors (5-HTTLPR, BDNF, cortisol/FKBP5-methylation,
oxytocin/OXTR) and possibly mediate treatment outcome (BDNF, cortisol and
FKBP5). Because these candidate predictors and mediators are found on a group
level (in non-responders vs. responders), they cannot directly be used on an
individual level. By using machine-learning techniques we might use these
candidate predictors and mediators on an individual level to guide treatment
choices and thereby personalise psychiatry.

In the current project, we will firstly study effectiveness of PTSD-treatment
compared to integrated PTSD-PD-treatment in treatment-seeking, adult patients
with comorbid PTSD and PD in a wide range of severity (minimally 4 criteria of
a personality disorder). Secondly, we will investigate psychological
(cognitive, affective, and relational) and neurobiological candidate predictors
and mediators of treatment outcome, and use them in a machine-learning paradigm
to predict and explain which treatment works best for which specific
individual, thereby presonalizing psychiatry.

Two parallel RCTs will be conducted with predictor analyses.

In patients with PTSD and BPD: EMDR (6 months plus 6 months treatment pause)
compared to integrated DBT-EMDR-treatment (12 months);
in patients with PTSD and CPD: ImRs (6 months plus 6 months treatment pause)
compared to integrated ImRs-SFT treatment (12 months).

The burden and risks associated with participation in this study is reasonable.
All patients will receive psychotherapy, which is considered to be the most
effective treatment for PTSD and/or PD. There is evidence that both
interventions are therapeutic in patients with PTSD and PD. There is no
evidence yet what treatment (PTSD or integrated PTSD-PD) is more effective.
Suicidality or self-injurious behaviour is very common in the study population
included for this study. It is also known that starting a new treatment, such
as EMDR or DBT could increase symptoms in the beginning, which can develop into
suicidal behavior. To monitor the possible increased symptoms, patients have to
fill in a questionnaire on self-injury at the assessments every three months.
In between the assessments the therapists will monitor SAE*s and take
appropriate action.
Total time of the clinical interviews/questionnaires is approximately 12.5 hour
in 1.5 year per patient with three visits to the Sinai Center and online
filling in of questionnaires. If the respondent also participates in the
blood/hair and/or MRI study, resp. 30 min (visit to the local hospital) and 180
min (2 visits to VUmc for MRI) is added. Questionnaires, which are partly part
of the routine outcome measurements (ROM), can be filled in online at home. The
burden and risks - fatigue - are acceptable while the benefits are expected to
be considerable. On the one hand, extended clinical interviews and self-rating
scales can be felt as disturbing because taken time and emotional burden. On
the other hand, patients may feel well recognized by the time taken by
specialized clinicians for an extensive assessment. Assessors will be
well-trained with close connection with the clinicians and treatment teams.
For the identification of predictors and mediators of the treatment response,
biological and genetic measures are integrated in the study. These measurements
include physical examination, blood samples and hair samples. The burden and
risk associated with the baseline blood sample and hair sample is reasonable.
For the subgroup of MRI research, participants will have a 60-minutes MRI
session during which they will recall traumatic events, and perform some
cognitive-affective tasks during scanning. Functional MRI is a commonly used
technique that is considered to be safe if you follow the safety instructions
(e.g. no metal objects in the MRI room) and contraindications (e.g. no metal
implants, no pregnant, no seriously claustrophobia). Lying in the scanner
and/or performing affective-laden tasks in the scanner can occasionally give
patients uncomfortable feelings of anxiety and distress by reliving of their
traumatic experiences. During and after the scan procedure a debriefing will be
held to cover this by the executor of the scan protocol. The principal
investigators of this study have long experience with symptom provocation in
the scanner (Thomaes: early traumatized PTSD patients with comorbid personality
disorders: only 1 in 33 patients had a panic attack; OA van den Heuvel in
patients with panic disorder, PTSD, OCD, Tourette, Parkinson, hypochondriasis:
panic attacks were rare and not more frequently than healthy controls). In all,
we consider the risk and burden associated with participation to be low.
Benefits for PTSD patients as a whole are that this study will provide
important information about best treatment choice in patients with both PTSD
and personality problems. It will help profiling patients and predict response
or non-response on an individual basis, to know what works for whom and be able
to provide personalize mental health care that can be as short as possible and
at the main time most effective. It will help to understand why (working
mechanisms) what treatment works best for whom.