Data Sharing in Personalised Medicine Clinical Research

he ERA PerMed project Artipro (Artificial intelligence for personalised medicine in depression – analysis and harmonization of clinical research data for robust multimodal patient profiling for the prediction of therapy outcome)

Julia Stingl, Institute of Clinical Pharmacology, University Hospital Aachen and Maria Giulia Baccalini, IRCCS, Istituto delle Scienze Neurologiche di Bologna, in behalf of all Artipro partners

The European collaborative project Artipro is coordinated by the Institute of Clinical Pharmacology under the leadership of Prof. Stingl. Partners  are Maria Giulia Bacalini, PhD from the IRCCS Istituto delle Scienze Neurologiche di Bologna from Italy, Assoz. Prof. Roberto Viviani, PhD from the University of Innsbruck from Austria, Prof. Noam Shomron, PhD from Tel Aviv University from Israel, Espen Molden, PhD from Diakonhjemmet Hospital at the University of Oslo from Norway, Dr. rer. nat. Catharina Scholl from the Federal Institute for Drugs and Medical Devices (BfArM) in Bonn, Germany, and Nada Bozina, MD, PhD from the Faculty of Medicine at the University of Zagreb in Croatia. These researchers have decided to explore the use artificial intelligence and higher-level biomarker analyses to find out whether it will be possible in the future to predict which patients will respond best to which antidepressant treatment for depression. In the long term, the experts hope to develop personalised treatment methods in which each patient can be given the therapy that best suits them.

Depression is one of the most common mental illnesses in the world. Treatment is often long and complex. Less than half of patients respond adequately to their first antidepressant, posing a major challenge for healthcare professionals. In the future, personalised medicine approaches will be used to help patients. These represent a paradigm shift from a generalised approach to treating a disease to one that focuses on a person’s unique characteristics and needs. Strategies for preventing, diagnosing and treating disease are thus oriented towards the patient, who is at the centre of healthcare. Medical models based on the characterisation of an individual’s phenotype and genotype (e.g. lifestyle data, pharmacogenetic or molecular information, or medical imaging) will help to provide treatment strategies and prevention approaches that are tailored to each individual at the right time. Researchers are aiming to increase the effectiveness of treatment and avoid over-medication and unnecessary interventions, which should improve overall care and quality of life.

To this end, the Artipro collaborative project – “Artificial intelligence for personalised medicine in depression – analysis and harmonisation of clinical research data for robust multimodal patient profiling for prediction of therapy outcome” – is establishing a data and analysis platform that will bring together data from existing clinical research projects across Europe on individual response to treatment in depression. The aim is to identify robust multimodal biomarkers and predictive profiles for depression medication response using artificial intelligence methods.

The development of the ArtiPro platform poses important ethical and legal challenges, as the project will re-use data previously collected or generated in different and unrelated projects, which will be shared and transferred across the consortium. The re-use of existing data for research purposes is strongly encouraged by the World Health Organisation and the European Union in line with the FAIR (Findability, Accessibility, Interoperability, and Reusability) principles. At the same time, however, in Europe, the processing of personal data (including its transfer and re-use) should comply with the General Data Protection Regulation (GDPR), which became binding law in the EU in 2018. However, individual member state can override the provisions of the GDPR with their own legislation. This aspect poses a challenge for the re-use of collected data in future studies. In this scenario, researchers can feel confused, as if they are being pulled by two opposing forces.

On the one hand, the need and willingness to share data with project partners in order to advance research in the field of interest; on the other hand, the need to comply with ethical, legal and social regulations and the possible consequences of violating the national laws. 

The preparation of the Data Management Plan in the first months of the project was an important moment of reflection on the legal basis and the modalities of data sharing, transfer and re-use within ArtiPro. This phase benefited from the presence of partners with different expertise within the consortium. Indeed, ArtiPro includes researchers with a legal and digital background and researchers with a more medical/biological background, with expertise in clinical trials, patient recruitment and interaction with the local ethics committee.

In order to ensure that applications to the Ethics Committee/DPO are consistent across the ArtiPro partners, a template has been prepared containing a short summary of the project and information on how data will be shared and stored. In particular, it is planned to store clinical data using the RedCap platform, a secure web application for building and managing databases; a common data dictionary will be created based on the original data to ensure data harmonisation between the different cohorts. Molecular data will be stored in a secure server hosted at the IRCCS Istituto delle Scienze Neurologiche di Bologna.

To ensure compliance with GDPR, each partner has been provided with the following guidance:

  • verify that existing consents cover the transfer and re-use of data for research purposes compatible with those envisaged by ArtiPro, possibly by consulting the local ethics committee or data protection officer (DPO).
  • In case the original consent does not cover data transfer and re-use, consider possible alternative legal bases, such as another provision of the GDPR, in which case it is strongly recommended to consult the Ethics Committee and/or DPO.

To this end, Artipro is combining data on the course of treatment and response from the existing clinical research projects of the partners in the collaborative project and integrating them into the complex data platform for analyses. In particular, AI will be used to expand the data on biomarkers and close existing data gaps. By systematically pooling the data from the individual studies, a comprehensive overall data set will be created that is homogeneously suitable for multimodal and transdiagnostic evaluations. This will significantly improve the power of the analyses compared to the original individual data. Big data approaches and artificial intelligence will also help to identify novel biomarker index profiles that are predictive of therapy and provide a basis for the development of decision support systems for personalised therapy.

Download project presentation