Major Depressive Disorder (MDD) is one of the most severe and frequently occurring brain disorders worldwide. It has been linked to serotonergic dysfunction, sexual dysfunction, vulnerability to stress and neuro-inflammation. However, at the same time the etiological understanding is limited. Most antidepressants act on the serotonin (5- HT) system, yet between 30-50 % of patients with MDD does not respond successfully to 5-HT acting drugs. Recent experimental models from our group suggest that cerebral 5-HT levels in vivo can be indexed through molecular brain imaging of the 5-HT 4 receptor (5-HT4R) with a novel Positron Emission Tomography (PET) ligand (11C-SB207145). Also, our human studies have confirmed that cerebral synaptic 5-HT is inversely related to 5-HT4R binding and this technique thus can be used to investigate the role of 5-HT tone in the brain in MDD with differential responses to standard antidepressant treatment. By using multimodal neuroimaging technology, we aim to determine the status of the 5-HT system prior to and after either successful or failed neuropharmacological intervention in a non-randomized longitudinal open clinical trial. 100 untreated patients with moderate to severe MDD will be included. Data collection from various neurobiological domains (i.e, 5-HT4R PET imaging, Magnetic Resonance Imaging (MRI), functional MRI (fMRI), electroencephalogram (EEG), psychometrics, neuropsychological tests, and peripheral biomarkers) will be conducted before, during and after 12 weeks of antidepressant treatment. The objective is to identify predictors of pharmacological antidepressant treatment response in depressed individuals before and after 8 weeks of antidepressant treatment.

Study population and study program:

Patients will be recruited through a unique new central referral site for "depression packages" in The Mental Health Services in the Capital Region of Denmark. 100 patients with MDD, 18-65 years of age, with moderate to severe single or recurrent episode of MDD (Hamilton 17 item (HAMD-17) score > 17) will be recruited through this portal. All diagnoses will be confirmed by a specialist in psychiatry.

Before initiation of pharmacological antidepressant treatment with escitalopram, patients will receive baseline examinations as follows: 1) 5-HT4R imaging with 11C-SB207145 PET-scan, 2) EEG examinations, 3) structural MRI, 4) functional MRI , 5) neuropsychological testing, 6) peripheral markers of immune-active cell responses, oxidative stress, cortisol-levels, RNA, genotypes and epigenetic factors will be measured in urine, saliva and peripheral blood at baseline and across the study period. Repeated measures across the study period include: 7) psychometrics by using self-reported questionnaires covering trait and state, including mental distress related to depression and other psychopathology, 8) neuropsychological examinations as well as 9) clinical follow-up with interview based ratings of mental status.

Patients will be treated with escitalopram at flexible doses of 10-20 mg/day adjusted depending on effects and side effects, and participate in clinical follow-up sessions at week 1, 2, 4 and 8. Patients with no response to escitalopram after 4 weeks will be shifted to a secondary pharmacological treatment (duloxetine). A final visit to determine longer-term clinical outcome will be performed at week 12. Compliance, side-effects to antidepressant treatment, and depressive symptoms will be monitored at each follow-up session.

The Hamilton 6 items (HAMD-6) subscale has recently shown to be more sensitive to antidepressant response (Østergaard et al), and will be used to identify treatment response in patients. Patients with > 50 % reduction in HAMD-6 after 4 weeks will be defined as early responders, and those with additional < 5 points on the HAMD-6 scale after 8 weeks will be considered in remission. Patients with >25% response at week 4 and < 50% reduction in HAMD-6 after full intervention will be considered non-responders. The assessment program including brain imaging with PET 11C-SB207145 will be conducted before drug intervention is initiated and, depending on treatment outcome, again after 8 weeks of antidepressant treatment in 20 patients in remission (remitters) and 20 non-responding patients (non-responders).

UPDATE: As per April 3, 2017 the rescanned group has been expanded to include various response patterns in order to capture rescan data from patients on a spectrum from poor to excellent treatment response, since out of the first 17 included patients only 1 patient fulfilled the non-responder criterion defined above. Accordingly, the clinical outcome parameter in the context will be changes in HAM-D6 from baseline at week 8.

Healthy controls:

From previous studies conducted at the Neurobiological Research Unit, there is access to brain imaging data and baseline PET-SB207145 brain images among healthy controls as well as an associated biobank with stored blood specimens. During the trial, additional healthy controls will be included and will receive baseline examinations and repeated neuropsychological testing after 12 weeks.

Hypotheses:

1. 5-HT4R binding (as imaged by 11C-SB207145 PET) before treatment will predict antidepressant treatment outcome (remission vs non-response) in depressed patients; higher binding at baseline is hypothesized to be associated with a better treatment outcome. Likewise, secondary, 5-HT4R binding (as imaged by 11C-SB207145 PET) before treatment is expected to be associated with the magnitude of change from baseline in Hamilton score (HAMD-6) at 8 weeks.
2. Remitters will display a greater reduction in 5-HT4R binding (as imaged by 11C-SB207145 PET) after antidepressant treatment relative to non-responders.
3. Patients with MDD will have a higher 5-HT4R binding (as imaged by 11C-SB207145 PET) relative to healthy controls.
4. High amygdala reactivity in an "emotional faces" fMRI paradigm before treatment predicts remission in response to antidepressant treatment.
5. Amygdala response to exposure to negative emotional faces (fear and anger) and cerebral 5-HT4R binding is expected to be associated, both in cross-sectional comparison before treatment and with regard to longitudinal changes (changes from baseline).
6. In resting state fMRI (rsfMRI), we expect that changes from baseline in a functional network centered around dorsomedial prefrontal cortex and other networks centered around anterior cingulate cortex and posterior cingulate cortex (default mode network) respectively will predict treatment outcome.
7. Low activity in a brain network engaged in reward processing (as assessed by fMRI) will predict antidepressant treatment response (i.e. positive association between low activity and remitter- relative to non-responder group) both in cross-sectional comparison before treatment and with regard to longitudinal changes (changes from baseline).
8. We expect reduced evoked gamma-activity measured with EEG and "Event-related-potentials" (ERP) in MDD patients compared to controls.
9. Using EEG/ERP, we expect reduced alpha and theta cordance and higher theta band activity among remitters measured at baseline.
10. Higher levels of systemic inflammation will predict poor cognitive function at baseline, and a poor antidepressant treatment outcome.
11. A blunted cortisol awakening response (CAR) at baseline will predict a positive antidepressant treatment outcome.
12. Restored HPA-axis dynamics (increased CAR compared to baseline) at week 8 will be associated with a positive treatment outcome.
13. There will be a significant association between functional networks found by EEG and fMRI, both at baseline and as contrast between remitters and non-responders after 8 weeks.
14. Hippocampal volume will predict treatment outcome.
15. Patients with a low 5-HT4R binding (as imaged by 11C-SB207145 PET) at baseline before treatment will have a lower libido relative to patients with a high binding.
16. Patients who respond to antidepressant treatment with the most pronounced changes in 5-HT4R binding (as imaged by 11C-SB207145) will experience a larger degree of sexual side-effects; i.e., the magnitude of 5-HT4R binding changes from baseline will be positively associated with the degree of sexual side-effects.
17. Childhood abuse (defined by a composite measure of early life stress questionnaires) and anxious depression (defined as anxiety/somatization factor score in HAMD-17 > 7) predicts poor antidepressant treatment outcome.
18. We expect urinary markers of oxidatively generated DNA/RNA damage to increase more in responders than in non-responders.
19. With an exploratory approach it will be possible to identify a set of predictors of antidepressant treatment outcome. This explorative analysis includes genetics, epigenetics, peripheral and central neurobiological characteristics, neuropsychology testing outcomes, psychometry including self-reported mental distress, both cross-sectional at baseline and as longitudinal measures.

Ethical Aspects:

The study protocol complies with the Declaration of Helsinki II and approval by all relevant authorities will be obtained before initiation. All human volunteers will receive oral and written information about the given study and provide written informed consent before enrolment. The trial is monitored by a Good Clinical Practise unit for the relevant domains.

Inclusion Criteria:

• Moderate to severe depression
• Age 18-65 years
• No previous antidepressant treatment the last 2 months
• Informed and signed consent

Exclusion Criteria:

• More than one previous attempt with antidepressant drugs
• Duration of current depression more than 2 years
• Current or previous psychiatric severe co-morbidity
• Acute suicidal ideation
• Psychotic
• Previous non-response to Selective Serotonin Reuptake Inhibitor (SSRI)
• Contraindication for SSRI treatment
• More suitable with treatment of alternative anti-depressive drug.
• Severe somatic co-morbidity
• Somatic medicine that can influence the trial
• Contraindications for MR-scanning
• Previous exposure to radioactivity > 10 milli sievert (mSv) within the last year
• Alcohol or drug abuse
• Previous severe head trauma
• Pregnancy
• Breast-feeding
• Insufficient Danish skills

Via database of Center for Integrated Molecular Brain Imaging (Knudsen et al 2016, NeuroImage) data will be available for neuroscience research community contingent on approval by scientific board.