Deep Phenotyping for Clinical Inferring Response in Treatment Resistant Depression

Overview

DECIDE- Deep phenotyping for clinical inferring response in treatment resistant depression -Study Building upon the "Biobanking" initiative at the Max Planck Institute of Psychiatry, the present project aims to identify clinically relevant subtypes of treatment-resistant depression (TRD) through Clinical Deep Phenotyping (CDP). According to clinical trials, 30-40% of the patients suffering from TRD benefit from lithium treatment. By collecting multimodal biological and clinical-diagnostic markers, such as structural and functional brain imaging via magnetic resonance imaging (MRI), brain signals from electroencephalography, comprehensive blood tests, assessment of perception and cognition through neuropsychological testing, as well as the evaluation of specific depression symptoms and psychological and other comorbidities using standardized questionnaires, a bio-clinical signature will be identified using multivariate machine learning algorithms as an integration method. This signature aims to predict the response to lithium therapy in TRD. Prospectively, such an algorithm could later personalize the treatment decision of 'lithium administration in TRD'. This concept is in line with the Research Domain Criteria (RDoC) of the National Institute of Mental Health (NIH) and aims to offer lithium therapy as a personalized treatment strategy for TRD. Specifically, this means that the likelihood of treatment response can be estimated before administration based on the results of the present study, thus enabling lithium to be offered specifically to those patients who are likely to benefit from it. The study design is non-interventional, meaning the decision for lithium treatment is made for patients according to clinical routine in accordance with the recommendation of the German National Treatment Guideline (NVL) independent of study enrollment. Study participation does not influence treatment decisions for the patients.

DECIDE- Deep phenotyping for clinical inferring response in treatment resistant depression -Study Background: TRD is a form of major depressive disorder (MDD) typically characterized by an inadequate response (partial or no response) to first-line antidepressants (AD) therapy of adequate dose and duration. This variant of depression, common in psychiatry, affects approximately 30-50% of all patients with depression and is the focus of the present DECIDE-study. In treating TRD, augmenting existing antidepressant pharmacotherapy with lithium proves to be an effective option, recommended in the German national care guideline. The precise definition of TRD within this study encompasses the following criteria and is based on Bauer et al. 2013: 1. Confirmation of MDD diagnosis according to DSM-V, verified through the Mini Neuropsychiatric Interview (MINI)-Interview. 2. Inadequate response to treatment after ≥4 weeks of antidepressant (AD) therapy. Inadequate response is defined as not achieving remission from depressive symptoms following ≥28 days of AD usage prior to the study. This determination includes the exclusion of pseudo-resistance through measurement of medication plasma levels. 3. Stage I TRD is characterized by the failure of ≥1 adequate trial using one major class of AD. Major classes are selective serotonin reuptake inhibitors (SSRIs), serotonin and norepinephrine reuptake inhibitors (SNRIs), tricyclic and tetracyclic antidepressants (TCAs), Monoamine oxidase inhibitors (MAOIs) and atypical antidepressants. 4. Stage II TRD is characterized by the failure of two distinct classes of major AD. A Montgomery-Åsberg Depression Rating Scale (MADRS) total score of ≥25 serves as an additional qualifying criterion. Importantly, while only around 30% patients experience benefits after 4 to 6 weeks of lithium treatment, those who do respond often experience significant relief. In the NVL, it is recommended to assess the effectiveness of lithium therapy after 2-4 weeks. If the patient does not benefit, lithium should be discontinued. However, despite substantial benefits by the recommended lithium application in clinical use, the augmentation strategy of lithium in TRD faces limitations in clinical practice due to concerns by physicians about lithium's potential side effects. Therefore, there is a risk that the use of lithium treatment in TRD remains under-exploited despite evidence for its significant potential to aid patients with TRD. Having a predictive tool indicating lithium efficacy in specific TRD patients before therapy initiation could aid psychiatrists in deciding to pursue lithium treatment. DECIDE - the here described prospective diagnostic study aims to explore whether in-depth clinical and biological phenotyping of TRD before lithium treatment could provide the basis to classify TRD into responders and non-responders to lithium. This concept aligns with the RDoC, the research criteria established by the NIH. Various studies have investigated predictors of lithium response in bipolar disorder (BP). Given the clinically observed heritability of lithium responsiveness, initial approaches involved analyzing genetic markers. As the field of lithium prediction evolved, subsequent investigations delved into epigenetic markers and genome-wide methylation data associated with lithium response. Genome-wide association studies (GWAS) identified genetic loci that were associated with lithium (non-)response but yielded inconclusive predictions in pooled datasets. Still, there was supporting evidence linking polygenic risk scores for schizophrenia and major depression to poor lithium response. Behind genetics, a recent study, that focused on serotonergic receptors and transporters, explored differences between responders and non-responders using nuclear medical imaging techniques. Only a limited number of studies have focused on analyzing noncoding RNA levels in cells from patients and their impact on the response to lithium. Valuable insights might arise from investigations focusing on the microRNA (miRNA) levels found in peripherally derived exosomes. Exosomes, secreted by diverse neuronal cell types, facilitate intercellular communication and signaling. Their capacity to traverse the blood-brain barrier allows for analysis through blood sampling. Initial results regarding the application of miRNAs from extracellular vesicles as BD biomarkers are promising, however the investigation of exosomes as a source of biomarkers in lithium response remains largely unexplored. While there have been studies on biomarker signatures of lithium responses in MDD, these studies have not established an algorithm for personalized treatment decisions so far. However, the precise molecular mechanisms behind lithium augmentation in TRD are still undergoing intensive investigation. Most previous studies have focused on individual clinical or biological markers, likely explaining only a fraction of mental illness diversity. Integrating clinical and multimodal response markers could enable the development of a composite 'prediction algorithm' and the testing of it in subsequent projects, applicable to new populations. The differentiation between nonspecific and clinically relevant molecular mechanisms of lithium response remains too intertwined to be approached solely through clinical studies. Therefore, in addition, this study aims to stratify a subgroup of the deeply clinical classified TRD patients, generate patient-derived induced pluripotent stem cells (iPSCs), and forward these to iPSC-based technology. In vitro studies utilizing neurobiological test systems derived from iPSC obtained from patients could make a substantial contribution to advance our understanding of psychiatry diseases in general, and the mechanism of lithium response in particular. Interestingly, the cellular response to lithium mirrored the clinical response, indicating significant neurobiological mechanisms in bipolar disorder. The goal here is to examine whether lithium response is also reflected on a cellular level in TRD, and to analyze the specific pathways active in different types of TRD. Moreover, the iPSC-based part of DECIDE aims to translate clinical and iPSC-based phenotypes in order to deliver neuroscience-based methods to improve the prediction of treatment response and non-response in TRD. Statistical analysis concept: The statistical analysis is conducted at both the individual and multi-marker levels using linear and logistic regression models. Machine learning techniques such as clustering methods are utilized to create the biosignature. The analyses are not intended solely at the end of data collection; interim analyses are also planned. Study protocol: After positive evaluation of study inclusion at baseline, the following examinations are to be conducted subsequently. After the initiation of lithium therapy, routine serum level monitoring occurs, with Follow-up 1 defined as the timepoint when an effective lithium level is achieved (target range of 0.6-0.8 mmol/L). Two weeks later (at Follow-up 2), MADRS assessments are conducted to capture therapy response. If, during weeks 3-6 (Follow-up 3, Follow-up 4, Follow-up 5, Follow-up 6), the endpoint MADRS response (≥50% reduction in baseline MADRS total score) is achieved, optional additional examinations will be conducted. If not, the optional additional examinations will take place in week 6 (Follow-up 6). If the treating psychiatrist decides to discontinue lithium earlier (due to non-response or other reasons), the optional additional examinations will be conducted at that timepoint. If the clinical treatment team decides to change the antidepressant medication, the patient must be withdrawn from the study. Follow up 7 will take place after 52 weeks for all included patients, independent of initial classification of response/non-response. Clinical characterization: The clinical characterization includes the MINI Interview according to DSM-V to confirm the diagnosis of MDD. The MINI Interview stands as the globally most recognized and extensively employed psychiatric structured diagnostic interview. Information regarding patients' psychiatric history and medication history will be evaluated using a designed questionnaire, sociodemographic information and family history will be assessed additionally. Biosamples: Biosamples will be collected from all participants and used for the generation of a biosinature, namely for RNA isolation, Proteomics, Metabolomics, Exosomes, Genotyping, Methylation, drug levels and PBMC isolation. Multimodal MRI (mMRI): The mMRI includes anatomical techniques, functional MRI measurements and single voxel proton MR spectroscopy at different locations. Electroencephalography (EEG): All subjects recruited into the study will undergo digitized EEG recordings of about 20 minutes using a standardized set-up with scalp electrodes exactly as it is recommended in clinical routine during Lithium therapy. Neurocognitive assessment (Cognition): The standardized Brief Assessment of cognitive function in schizophrenia (BACS) battery will be administered to all patients. The BACS includes six tasks evaluating four cognitive domains deemed crucial for clinical trials by the MATRICS Neurocognition Committee. For the sake of comparability with other studies that focuses on Treatment resistance involving schizophrenic patients, this DECIDE Study will also use the BACS. These domains are processing speed (verbal fluency in category instances such as supermarket items, and letter fluency using F and S words, Token Motor Task, Symbol Coding), reasoning and problem-solving (Tower of London test), verbal memory (list learning), and working memory (digit sequencing). Additionally, we enhance the assessment with the widely used Montreal Cognitive Assessment (MoCA) for a quick and comprehensive neurocognitive evaluation.