Digital CMC Regulatory Lifecycle (DCRL) Framework

The Digital CMC Regulatory Lifecycle (DCRL) Framework provides a structured, science-based approach to the development, validation, deployment, and lifecycle management of digital predictive tools in pharmaceutical CMC.

It enables industry, regulators, and academia to apply digital tools consistently and confidently while ensuring product quality, patient safety, and regulatory readiness.

What the Framework Supports

The DCRL framework provides guidance across:

  • Context of use definition

  • Risk-proportionate credibility assessment

  • Validation and uncertainty evaluation

  • Lifecycle management

  • Regulatory documentation readiness

  • Inspection and submission preparedness

It promotes predictable, proportionate adoption of digital and AI-enabled tools in regulated environments.

Key Principles

The framework is underpinned by core regulatory and scientific principles:

  • Risk-based implementation

  • Human accountability and oversight

  • Transparency and explainability

  • Robust data governance

  • Lifecycle management and control

  • Regulatory readiness and traceability

These principles align with international guidance including ICH, EMA, FDA, MHRA, and relevant standards and best practice.

Framework Overview

The Digital CMC Regulatory Lifecycle (DCRL) Framework is a practical, science-based approach developed by the Digital CMC CERSI to support the responsible use of digital predictive tools, including AI and machine learning, in pharmaceutical Chemistry, Manufacturing and Controls (CMC). Its purpose is to help innovators, manufacturers and regulators build shared confidence that digital tools are reliable, transparent, proportionate to risk, and suitable for their intended regulatory or manufacturing use.

  • Description text goes here
  • Description text goes here
  • Description text goes here

Figure 1: Informing Digital Predictive Tool Development in CMC

‍ ‍

The framework supports the CERSI’s wider aim: to accelerate safe, sustainable and efficient medicines development while protecting product quality, patient safety and regulatory trust. It provides a common language and workflow for industry, academia and global regulators, reducing uncertainty around evidence expectations, model credibility, documentation and lifecycle oversight.

‍ ‍

‍ ‍

Figure 2: Digital CMC Regulatory Lifecycle (DCRL) Framework

‍ ‍

The DCRL framework has three connected layers:

‍ ‍

Top layer:  Supporting Enablers[IH1]
The DCRL framework is supported by four enablers that facilitate practical adoption and shared understanding across stakeholders. These include 4 regulatory use cases to demonstrate real-world application, Virtual Expert for Regulatory Assistance (VERA), a collaborative Digital CMC Sandbox  for static credibility assessment with an interactive element and evidence generation, and targeted, modular and interactive training platforms (e.g. SkillsFactory) to build capability.

‍ ‍

Together, these enablers provide the consistent language, approach, templates, knowledge for regulators, industry and academia to translate principles into consistent, regulator-ready practice.

‍ ‍

‍ ‍

Middle Layer: Practical workflow[IH2]
The pathway guides digital tools through 4 levels of maturity readiness from early concept through development, validation, regulatory use and lifecycle management: concept and risk analysis; development and verification; validation and credibility assessment; and regulatory use and lifecycle management.

‍ ‍

Ten detailed workflow stages help users define the question of interest, context of use, model type, model risk, data requirements, credibility evidence, regulatory documentation and post-deployment control.

‍ ‍

‍ ‍

Bottom layer:  Foundational regulatory principles and evidence base
The bottom layer anchors the framework in five core principles adapted for Digital CMC which are applicable to knowledge driven, data-driven (incl. AI, and empirical) and hybrid models :

‍ ‍

·         Safety, security and robustness

‍ ‍

·         Transparency and explainability

‍ ‍

·         Fairness and avoidance of bias

‍ ‍

·         Accountability and governance

‍ ‍

·         Contestability, human oversight and redress.

‍ ‍

Not all the principles are applicable to all model types, but by collating them and adapting them for a CMC context, they provide a valuable aide-memoir resource to ensure users consider all aspects to ensure their credibility assessments are regulatory-aligned.  These principles are aligned to collated global regulatory guidance, standards and scientific literature, including ICH guidance, FDA, EMA and MHRA expectations, ASME standards, and peer-reviewed best practice (Figure 3).  

‍ ‍

‍ ‍

Figure 3: Regulatory, Standards and Scientific Best Practice Basis for DCRL Framework

‍ ‍

An interactive DCRL Framework has been developed  to help the user consider the relevant principles that apply totheir computational models at various stages of the workflow and provides key regulatory and standards links

‍ ‍

Embed interactive element into website here – Steve knows how to do it

‍ ‍

https://www.cmac.ac.uk/cersi-home

‍ ‍

‍ ‍

‍ ‍

Together, these layers make the DCRL framework a regulator-ready bridge between digital innovation and pharmaceutical quality assurance. It enables digital tools to be developed and assessed consistently, supports global harmonisation, and helps ensure that digital transformation in CMC delivers faster access to medicines without compromising quality, safety or public trust.

‍ ‍

Digital CMC Regulatory Lifecycle (DCRL) Workflow

The Digital CMC CERSI has developed the Digital CMC Regulatory Lifecycle (DCRL) Workflow to support the regulator-ready development, deployment, and lifecycle management of digital predictive tools used in pharmaceutical CMC.

The workflow provides a structured, risk-proportionate approach that helps organisations move from early-stage digital tool development through to regulatory use and ongoing lifecycle oversight.

The DCRL workflow supports:

  • Definition of the question of interest and context of use

  • Risk assessment and credibility planning

  • Data collection and model development

  • Verification, validation, and uncertainty assessment

  • Documentation for regulatory submission and inspection

  • Lifecycle monitoring and governance

The workflow is designed to improve consistency, transparency, and predictability in the use of digital and AI-enabled tools across regulated pharmaceutical environments. It also helps organisations align with emerging regulatory expectations relating to model credibility, explainability, data governance, and human oversight.

Enablers Supporting Adoption

The Digital CMC CERSI is also developing enabling infrastructure and capability-building activities to support practical adoption.

  • Virtual Expert for Regulatory Assistance (VERA)

An AI-enabled support platform designed to assist with interpretation of regulatory expectations and framework implementation.

  • Case Studies and Demonstrators

Practical examples demonstrating regulator-ready implementation pathways for digital predictive tools and workflows.

  • Digital CMC Sandbox

A collaborative environment for testing, evaluation, and demonstration of digital regulatory science approaches.

  • SkillsFactory and Training

Training programmes and digital capability development resources designed to support workforce readiness and digital literacy across pharmaceutical CMC.

Why This Matters

Digital transformation has the potential to fundamentally improve how medicines are developed, manufactured, controlled, and supplied.

The responsible adoption of digital predictive tools could enable:

  • Faster patient access to medicines

  • More robust and resilient manufacturing

  • Improved sustainability

  • Better process understanding

  • Reduced development costs and timelines

  • Stronger supply chain agility

However, achieving these benefits requires trusted, transparent, and regulator-ready approaches that maintain patient safety and product quality.

The Digital CMC CERSI exists to help bridge the gap between innovation and implementation.

  • Supporting the Future of Pharmaceutical Manufacturing

The Digital CMC CERSI works collaboratively with:

  • Industry

  • Regulators

  • Academia

  • Technology providers

  • Standards organisations

  • International partners

Together, we are helping shape the future of digitally enabled pharmaceutical development and manufacturing.

The Digital CMC Regulatory Lifecycle (DCRL) workflow provides a structured, end-to-end pathway for developing, assessing and deploying digital predictive tools in pharmaceutical CMC. Developed by the Digital CMC CERSI, it directly supports the CERSI mission to enable innovation in regulatory science while maintaining product quality, patient safety and global regulatory confidence.

The workflow translates high-level regulatory principles into a practical, stepwise approach that can be applied consistently by industry, regulators and researchers across the product lifecycle.

Figure 4: Digital CMC Regulatory Lifecycle (DCRL) Workflow

The 10 Stages of the DCRL Workflow

1. Define the Question of Interest (QoI)

Clearly identify the scientific or operational question the digital tool is intended to address.

2. Define the Context of Use (CoU)

Specify how and where the tool will be used (e.g. development, manufacturing, regulatory submission) and its impact on decision-making.

3. Select Model Type

Determine the most appropriate modelling approach (e.g. mechanistic, data-driven, hybrid) based on the question and available data.

4. Conduct Risk Assessment (Model Risk)

Assess the potential impact of model errors on product quality, patient safety and regulatory decisions.

5. Plan Data Collection

Define the data strategy, including sources, quality requirements, and experimental or operational data generation.[IH7.1]

6. Model Development and Initial Credibility Assessment

Develop the model and begin assessing its scientific validity and performance against the intended use.

7. Review Credibility and Assess Fitness for Use

Evaluate whether the model is sufficiently reliable and robust for its defined context of use.

8. Assess Risk of Uncertainty in Decision-Making

Understand how model uncertainty could propagate into manufacturing control or regulatory decisions.

9. Document for Regulatory Use

Prepare clear, transparent documentation to support regulatory submission, inspection, or internal governance.

10. Lifecycle Management

Maintain and monitor the model over time, including updates, revalidation, and ongoing performance oversight.

________________________________________

5 Evidence Base and Sources

The DCRL workflow is grounded in a broad and evolving evidence base. It integrates:

• Global regulatory guidance (e.g. ICH quality guidelines, FDA, EMA, MHRA)

• International standards (e.g. ASME, ISO and related model credibility frameworks)

• Scientific literature and peer-reviewed best practice

• Industry white papers and cross-sector experience

By consolidating these sources into a single, coherent workflow, the DCRL provides a harmonised and practical approach that reduces ambiguity, aligns expectations across regions, and supports consistent regulatory decision-making.

________________________________________

For CERSI, the DCRL workflow is a key mechanism to bridge innovation and regulation—enabling digital and AI tools to be adopted in a way that is predictable, proportionate, and globally aligned. It helps ensure that advances in digital CMC translate into real-world benefits: faster development, more resilient supply chains, and continued assurance of safe, high-quality medicines.