Projects: HITKIP

Suchtverhalten und riskanter Substanzkonsum sind nicht durch individuelle Faktoren wie Impulsivität oder Stresssensitivität gekennzeichnet, sondern es spielen dabei auch sozioaffektive Faktoren wie soziale Normen und Möglichkeiten sowie sozial Kontextfaktoren wie Stress in der Familie oder der Bezug zu Freunden und Gleichaltrigen. Die COVID-19 Pandemie hat nun durch den verhängten Lockdown, den limitierten Aktionsradius und die Einschränkung sozialer Begegnungen erhebliche soziale Veränderungen mit sich gebracht, die sich je nach individueller Lage und über alle Altersspannen hinweg erstreckt. Zuvor identifizierte normative und nicht-normative Risikoverhaltensweisen und -bedingungen für einen Substanzkonsum könnten sich somit ebenso verändert haben. Durch den Rückgriff auf bereits existierende längsschnittliche Kohorten (IMAGEN, ROLS, MARS) und querschnittlicher Datensätze, bei denen auch Informationen direkt in alltäglichen Situationen erhoben wurden (IMAC-Mind) sowie COVID-19 bezogene Erhebungen zu Gesundheit, sozialen Faktoren und Verhaltensweisen während des Lockdowns, werden wir multivariate Analysen durchführen, um die Stabilität von in früheren Studien identifizierten Maßen unter COVID-19 zu schätzen und zu sehne, wie sich diese über die Zeit hinweg nochmals verändern. Hierfür verwenden wir Daten zu Gehirnstruktur und -funktion, Sensor-basierte Verhaltensdaten in alltäglichen Situationen sowie Maße der Affektregulation („mindfulness“), die wir auch über verschiedenen Alterspannen hinweg, kreuzvalidieren. Faktoren, die zuvor als Schutzmaßnahmen klassifiziert wurden könnten sich abschwächen oder verstärken und andere Mechanismen und Prozesse könnten sich als zentral herausstellen. Dies könnte auch Unterschiede zwischen den Geschlechtern betreffen. Durch das aktuelle Projekt könnten wertvolle Einblick in die gesundheitsbezogenen Konsequenzen einer solchen Pandemie übe verschiedene Lebensphasen hinweg gewonnen werden.“

Schizophrenia (SZ) and major depressive disorder (MDD) are common and severe mental disorders that constitute an extraordinarily high public health burden. To facilitate the development of effective, individualized therapeutic interventions, robust biomarkers are needed that index specific functional deficits relevant for a given patient. Among these, psychomotor abnormalities represent a core clinical feature with transdiagnostic importance in SZ and MDD, but their relevance for individualizing therapeutic options remains unexplored. Psychomotor functioning is defined through the interaction of primary sensorimotor function (e.g., the dopaminergic-based subcortical-cortical motor circuit) and non-motor function, including cognition and emotion, and changes in the underlying neural processes are known to cross diagnostic boundaries of mental illness. This project will provide the basis for characterizing psychomotor mechanisms in SZ and MDD and their downstream translation into clinically useful predictors of therapeutic response. For this, we will integrate a harmonized battery of behavioural assessments targeted at psychomotor functioning with a multimodal neuroimaging approach, in order to provide a detailed characterization of psychomotor alterations in SZ and MDD, and to generate a data backbone tailored towards application of machine learning. Using such multimodal machine learning, we will identify neurobehavioral signatures predictive of treatment response in psychomotor domains 12 weeks after an acute illness episode within and across SZ and MDD. Through such neural and behavioral characterization of psychomotor mechanisms, this study will contribute to the dimensional dissection of severe mental illness and provide preliminary markers for individualization of therapy in the psychomotor domain.

To date, reward discounting but not aversion discounting was examined in SUD. Our working hypothesis of increased temporal aversion discounting in AUD patients will be tested by novel tasks for reliable and quantitative assessment of aversion discounting in humans and animal models. We will study the underlying neurobiology of aversion discounting by fMRI in humans and calcium imaging microendoscopy in rats. Computational analyses will be used to model the decision-making processes and deliver a detailed and formal parametrization of aversion discounting on multiple levels of analysis. In the future, such information can be used for the development of therapeutic approaches that strengthen self-regulation and cognitive control

Recurrent neural networks (RNN) are a class of artificial neural networks that can be trained with empirical behavioral data to emulate in silico how the brain performs complex cognitive tasks. This training is achieved by adapting a complex topological structure that is optimized to perform algorithmic computations through a nonlinear input-output mapping. The resulting properties of RNN can be analyzed with tools from network neuroscience to gain insights into the neural mechanisms of high-order cognition, and have shown remarkable similarities to the topological properties of the living brain. The proposed project aims to establish a framework to enable the use of RNN as preclinical models of cognition in health and disease. Capitalizing on schizophrenia as an example of cognitive dysfunction, the project will examine the organizational properties of RNN trained on aggregated cognitive performance data derived from a battery of cognitive tests in controls and patients with schizophrenia. We expect that RNN trained on human healthy data will show network properties that support a balance between segregated and integrated information processing, while RNN trained on schizophrenia data will show a shifted balance towards integrated processing. To further validate the utility of RNN as cognitive models and demonstrate their sensitivity to inter-individual differences, we will collect highly sampled longitudinal cognitive test data from 5 healthy subjects over the course of one year, and train individual RNN on those test data. We expect individual models to show topological variation that explains individual variation in independent cognitive measures.

Psychotic disorders, including schizophrenia and bipolar disorders, comprise some of the most severe mental illnesses that cause an enormous clinical and healthcare burden, costing close to €100 billion annually in Europe alone. The diagnostic delineation of these conditions is clinically defined and does not index appropriately the underlying biology. Patients are treated largely by a “one-fits-all” approach, despite substantial clinical heterogeneity in course, treatment response and presence of somatic comorbidities that include type 2 diabetes, cardiovascular diseases and neurodegenerative processes. There is a strong need to identify biological means to stratify patients with psychotic disorders and identify the biological basis of somatic comorbidity. This will allow improved clinical delineation of psychotic illnesses and facilitate novel intervention strategies targeted at the minimization of comorbidity risk, reducing mortality and morbidity. To address this, COMMITMENT will develop an innovative computational Framework for stratification of psychotic disorders and identification of biological domains shared with comorbidities. COMMITMENT builds on distributed machine learning that integrates mechanistic information mined from the literature. It extends approaches successfully used for oncological systems-medicine investigations, to optimally extract disease signatures from partially overlapping multi-OMICs data. We will use massive-scale genetic and neuroimaging data to explore the lifespan trajectories of stratification and comorbidity profiles, to identify age periods with pronounced comorbidity risk and to disentangle state- from trait-effects. We will explore the predictivity of comorbidity profiles for illness course, treatment response and occurrence of comorbidity during early illness phases. With this, COMMITMENT will provide the basis for biologically-informed clinical tools for improved personalized care of patients with psychotic disorders.

Das SYNSCHIZ Projekt stellt eine Zusammenarbeit von Experten aus Norwegen, Deutschland, der Schweiz, Finnland, Rumänien, und den Niederlanden dar, mit dem Ziel synaptische Dysfunktion als Risikomechanismus der Schizophrenie (SCZ) zu untersuchen. WP5 hat das Ziel die kommerzielle Verwertung der in SYNSCHIZ gewonnen Erkenntnisse sicherzustellen und damit einen entscheidenden Beitrag zum translationalen Aspekt von SYNSCHIZ zu leisten. Wir gehen davon aus, dass mehrere Forschungsergebnisse von hohem Wert für die Forschungsgemeinschaft erzielt werden. Insbesondere die Entwicklung eines in vitro Modells für SCZ kann im Bereich Wirkstoffscreening und dem Testen neuer Pharmazeutika von großem Wert sein.Die kommerzielle Auswertung kann sowohl durch die Ausgründungen von Firmen oder durch Lizenzvereinbarungen mit Pharmafirmen erfolgen.

Psychotische Störungen gehören zu den führenden Ursachen für den Verlust an gesunder Lebenszeit, Arbeitsunfähigkeit und Frühberentung. Die derzeitige Diagnostik basiert auf Symptom-Verlaufskriterien geringer biologischer Valididät, was die evidenzbasierte Auswahl geeigneter Therapieoptionen erheblich erschwert. Eine Vielzahl von Studien zeigt entsprechend, dass die Gruppe der Psychosen ätiologisch heterogen ist, und zwischen Schizophrenie und der bipolaren Störung erhebliche Überschneidungen genetischer, molekularer oder bildgebender Faktoren aufweist. Da es sich um hochgradig heritable Gehirnerkrankungen handelt, sollte sich aus genom-weiten genetischen Daten in Kombination mit einer Charakterisierung der Hirnfunktion und Struktur eine verbesserte, biologisch basierte und therapierelevante Taxonomie dieser Erkrankungsgruppe entwickeln lassen. Hierzu werden in diesem Vorhaben neue multivariate statistische Verfahren entwickelt und eingesetzt, um diagnose-übergreifende Untergruppen psychotischer Patienten mit Hilfe polygener genetischer Signaturen zu identifizieren und deren neurobiologische Grundlagen auf der Systemebene durch multimodale Bildgebung zu untersuchen.

Despite remarkable advances in molecular and imaging technologies and nearly 15,000 articles on schizophrenia and depression (S&D) every year, there have been few truly innovative new chemical entities (NCEs) which have made it to the clinic. While there has been a tremendous explosion of new knowledge: dozens of single-nucleotide polymorphisms linked to disease, hundreds of new molecules and pathways identified, numerous imaging findings differentiating patients from controls, yet, it has been hard to take these findings from the bench to the clinic. We think there are three major bottlenecks that are holding the field back: i) a lack of pathophysiologically-accurate animal models guiding the drug discovery of NCEs; ii) a lack of tools and tests in healthy volunteers that can provide early indication of efficacy; and iii) the reliance of clinical trials on symptom-based DSM-categories which inevitably lead to biologically heterogeneous groups of patients. To overcome these limitations, we have brought together a consortium of six leading European and an Israeli academic institution (which bring expertise in animal models, genetics, functional MRI and PET imaging, clinical settings and analysis methods) and two SMEs (which bring expertise in high-throughput genetics, transcriptomics and proteomics) who will partner with the dozen EFPIA partners in the NEWMEDS consortium. To specifically target the challenges identified in Call 10, the NEWMEDS consortium will: a) develop animal models that focus on reliable cross-species endophenotypes (e.g., cognitive function, electrophysiology) and use crossspecies methods (small-animal MRI, EEG and micro-PET) to bring animal models closer to clinical endpoints; b) validate the use of fMRI-based paradigms as early and surrogate markers for efficacy; and to combine this with PET approaches for measuring changes in endogenous transmitters – thus providing new methods that can be implemented in small Phase 1B studies in healthy volunteers to provide guidance for drug development; and c) identify pharmacogenetic biomarkers that can be used to stratify patients within an umbrella DSM-diagnosis, thus allowing for targeted clinical trials, individualized treatment and back-translation of subgroup-specific biomarkers into preclinical drug discovery. To increase the chance of a breakthrough we will implement new analytical approaches – the use of support vector machine learning algorithms for image analyses; the use of Bayesian and growth mixture models for more meaningful analyses of clinical trial data. The project will be delivered through a series of integrated workpackages organized in three clusters – preclinical models, imaging methods, and biomarker development as exemplified in Figure 1. Our consortium has achieved its 1:1 in-kind match, indicative of the involvement and commitment of all EFPIA partners. One of Europe’s leading scientific management SMEs (GABO:mi) will facilitate the management of NEWMEDS and a distinguished international Scientific Advisory Board will provide input and guidance. To ensure that we maximally integrate with other ongoing international initiatives, we have commitments of collaborations from several international consortia and experts (e.g. MATRICS, NIH Biomarkers Consortium). By the end of the 5 year project we expect to provide ready to use new cross-validated animal models, new fMRI methods with dedicated analysis techniques, new PET radioligands, as well as new genetic and proteomic biomarkers for patient-segmentation or individual response prediction. These tools should provide our EFPIA partners with an added competitive advantage in developing new drugs for S&D.