AI Archives - European Industrial Pharmacists Group (EIPG)

Some perspectives on green pharmaceuticals


by Giuliana Miglierini The central role the green agenda plays within the EU Commission’s transformative policies impacts also on the development and availability of pharmaceutical products characterised by a improved sustainability. The concept of “Pharmaceuticals in the environment” (PiE) is Read more

European Council’s conclusions on the European Innovation Agenda and research infrastructures


by Giuliana Miglierini The European socio-economic framework is undergoing a profound transformative moment, as a result of the new vision impressed by the von der Leyen Commission, with its goals in the field of the Digital and Green transitions. The Read more

EMA’s new Quality Innovation Expert Group (QIG)


by Giuliana Miglierini Innovative approaches to the development manufacturing and quality control of medicines are becoming the new paradigm to be faced both from an industrial and regulatory perspective. Not only innovative technologies for delivery, such as mRNA vaccines, many Read more

A concept paper on the revision of Annex 11

, , , , , , , , , , , , , , , , , , , , , , ,

This concept paper addresses the need to update Annex 11, Computerised Systems, of the Good Manufacturing Practice (GMP) guideline. Annex 11 is common to the member states of the European Union (EU)/European Economic Area (EEA) as well as to the participating authorities of the Pharmaceutical Inspection Co-operation Scheme (PIC/S). The current version was issued in 2011 and does not give sufficient guidance within a number of areas. Since then, there has been extensive progress in the use of new technologies.

Reasons for the revision of Annex 11 include but are not limited to the following (in non-prioritised order):

  • The document should be updated to replace relevant parts of the Q&A on Annex 11 and the Q&A on Data Integrity on the EMA GMP website
  • An update of the document with regulatory expectations to ‘digital transformation’ and similar newer concepts will be considered
  • References should be made to ICH Q9
  • The meaning of the term ‘validation’ (and ‘qualification’), needs to be clarified
  • Guidelines should be included for classification of critical data and critical systems
  • Important expectations to backup processes are missing e.g. to what is covered by a backup, what types of backups are made, how often backups are made, how long backups are, retained, which media is used for backups, or where backups are kept
  • The concept and purpose of audit trail review is inadequately described
  • Guidelines for acceptable frequency of audit trail review should be provided
  • There is an urgent need for regulatory guidance and expectations to the use of artificial intelligence (AI) and machine learning (ML) models in critical GMP applications as industry is already implementing this technology
  • FDA has released a draft guidance on Computer Software Assurance for Production and Quality System Software (CSA). This guidance and any implication will be considered with regards to aspects of potential regulatory relevance for GMP Annex 11

The current Annex 11 does not give sufficient guidance within a number of areas already covered, and other areas, which are becoming increasingly important to GMP, are not covered at all. The revised text will expand the guidance given in the document and embrace the application of new technologies which have gained momentum since the release of the existing version.

If possible, the revised document will include guidelines for acceptance of AI/ML algorithms used in critical GMP applications. This is an area where regulatory guidance is highly needed as this is not covered by any existing regulatory guidance in the pharmaceutical industry and as pharma companies are already implementing such algorithms.

The draft concept paper approved by EMA GMP/GDP IWG (October 2022) and by PIC/S (November 2022) and released for a two-months consultation until 16 January 2023.


Draft topics for the first IHI calls for proposals

, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

by Giuliana Miglierini

The Innovative Health Initiative (IHI) published on its website the first draft topics which may be part of the first two calls for proposals, scheduled in June 2022. Interested parties may start the activities needed to build and formalise the research consortia, taking into consideration that the announced topics are draft, pending their approval by the IHI Governing Board and their final version may differ from the drafts.

IHI call 1 shall focus on innovative technologies for the development of decision-support system for improved care pathways, next generation imaging, personalised oncology and access and integration of heterogeneous health data in areas of high unmet public health need.

IHI call 2 shall address cardiovascular diseases and the development of a harmonised methodology to promote early feasibility studies.

The draft topics of IHI call 1

Innovative decision-support systems are important to make available improved care pathways for patients with neurodegenerative diseases and comorbidities. The actions to be undertaken include the enhanced cross-sectoral and sustainable collaboration between healthcare industries, academia and other stakeholders in order to exchange data (through the new European Health Data Space), analytical tools and material for training and professional development of personnel.

Earlier diagnosis should lead to more clinically effective interventions and reduced hospitalisation and facilitate the adherence to therapy. Clinical outcomes are expected to support a better patient stratification, which is needed to develop more patient-adapted interventions, therapeutics and cost-effective pathways for the management of neurodegenerative diseases. Among the expected outcomes is the development of a re-usable, interoperable, easily adaptable, and scalable digital platform, initially targeted to support patients in this therapeutic area, but further expandable in the future to other areas of interest. The action should also involve the development of agreed standards and guidelines to support data collection and operational features of the digital platform. New algorithms may provide (near) real time feedback on health interventions and support the constant monitoring of the patient’s status.

The second topic is focused on the development of high-quality tools, high-quality data, advanced patient imaging and image-guided technologies and processes for improved early diagnosis, prognosis, staging, intervention planning, therapy and management of cancer. Imaging can be part of new combined cancer therapies (e.g., theranostics, chemotherapy, targeted therapy including immunotherapy, radiotherapy and/or surgery). The call should also include the development of improved validation and evaluation methodologies specific to artificial intelligence (AI)and machine learning (ML), with a particular attention to the creation of new solutions that automatically link images to clinical data. This could be applied, for example, to develop minimally invasive interventions guided by medical imaging, or image-driven planning and predictive tools.

The third topic is also aimed to tackle cancer through the development of personalised interventions. This action should contribute to break down silos that are often still characterizing medicine and technological areas. The availability of harmonised approaches should lead to safe and effective innovative health technologies, to the integration of future products, services and tools and the development of more patient-centred tools. Here again, an expected outcome is represented by a dynamic platform for R&I collaboration across different sectors and stakeholders, focusing on the early stages of applied clinical research on cancer and on the testing and validation of multi-modal therapeutic approaches, including novel or emerging technical and clinical concepts and the possible contribution arising from in vitro diagnostics.

Topic 4 of IHI 1 addresses the integration of future products, services and tools along the healthcare pathway to better respond to specific patients’ needs. The availability of interoperable, quality data which reflect the FAIR principles (Findability, Accessibility, Interoperability, Reusability) is central to this action, as well as the development of advanced analytics/artificial intelligence supporting health R&I. Among the main expected outcomes is the long-term access to diverse types of data enabled by the linkage and integration of novel and cross-sectoral sources. Access to interoperable tools should also become possible for citizens and patients to support the self-management of health and the joint decision making process between healthcare professionals and patients.

The draft topics of IHI call 2

Cardiovascular diseases (CDV) remain one of the main causes of death; the development of new tools for the primary and secondary prevention of CDV is the main focus of Topic 1, to be pursued by the identification of existing comprehensive CVD and heart failure (HF) patient datasets, in order to facilitate the diagnosis of atherosclerosis and HF. These data shall be also integrated with those captured by diagnostic tools (e.g., wearables, imaging devices, bio samples/biopsies).

Classical diagnostic screening, in-vitro- diagnostics, ‘multi-omic’ platforms (e.g. genomic, transcriptomic, proteomic and multimodality imaging data), continuous glucose monitoring (CGM) data, continuous electrocardiogram (ECG) from wearable, HF and activity data, wearable devices and digital health applications are all possible sources for the data. Projects may also leverage data in currently available IMI federated databases in compliance with the GDPR regulation governing protection of personal data.

The utility of already existing or new biomarker combinations shall be assessed to detect patients at risk, also making use of AI models to analyse data. Validated data referred to patient reported outcome and experience measure (PROMs and PREMs) may also be considered for use in the clinical setting.

The second draft topic is targeted to establish a harmonised methodology to promote the diffusion of Early Feasibility Studies (EFS) among healthcare professionals. Once again, the availability of digital technologies easily accessible by patients shall be key to this action. Among expected outcomes are the improvement of the quality of clinical evidence on health technology innovation generated through earlier clinical experience, together with the increase of the attractiveness of clinical research for healthcare technologies in the EU.

These activities are essential to enable the fast translation of innovation into the clinical practice, improving access to patients especially where there are only limited or no alternative therapeutic options. This approach to the development of innovative technologies may also benefit regulators and health technologies assessment (HTA) bodies, as well as notified bodies. All stakeholders involved in clinical practice and research may contribute to the early generation of quality data, so to achieve a better understanding of diseases management and treatment options and to support the future development of new medical guidelines.

The creation of hubs of clinical excellence to attract investment may also be considered under this topic, with involvement of developers of medical devices, drug-device combination products, imaging equipment, in-vitro diagnostics, and SMEs.


Artificial intelligence in medicine regulation

, , , , , , , , , , , ,

The International Coalition of Medicines Regulatory Authorities (ICMRA) sets out recommendations to help regulators to address the challenges that the use of artificial intelligence (AI) poses for global medicines regulation, in a report published on 16 August 2021.

AI includes various technologies (such as statistical models, diverse algorithms and self-modifying systems) that are increasingly being applied across all stages of a medicine’s lifecycle: from preclinical development to clinical trial data recording and analysis, to pharmacovigilance and clinical use optimisation. This range of applications brings with it regulatory challenges, including the transparency of algorithms and their meaning, as well as the risks of AI failures and the wider impact these would have on AI uptake in medicine development and patients’ health.

The report identifies key issues linked to the regulation of future therapies using AI and makes specific recommendations for regulators and stakeholders involved in medicine development to foster the uptake of AI. Some of the main findings and recommendations include:

  • Regulators may need to apply a risk-based approach to assessing and regulating AI, which could be informed through exchange and collaboration in ICMRA;
  • Sponsors, developers and pharmaceutical companies should establish strengthened governance structures to oversee algorithms and AI deployments that are closely linked to the benefit/risk of a medicinal product;
  • Regulatory guidelines for AI development, validation and use with medicinal products should be developed in areas such as data provenance, reliability, transparency and understandability, pharmacovigilance, and real-world monitoring of patient functioning.

The report is based on a horizon-scanning exercise in AI, conducted by the ICMRA Informal Network for Innovation working group and led by EMA. The goal of this network is to identify challenging topics for medicine regulators, to explore the suitability of existing regulatory frameworks and to develop recommendations to adapt regulatory systems in order to facilitate safe and timely access to innovative medicines.

The implementation of the recommendations will be discussed by ICMRA members in the coming months.

Source: EMA