Understanding The New U.K. MHRA ‘Draft Guideline on Individualized mRNA Cancer Immunotherapies’

By Tiffany Lucas, Ph.D., and Ifty Saiyed, Eliquent Life Sciences

The U.K. Medicines and Healthcare products Regulatory Agency (MHRA) has released the Draft Guideline on Individualised mRNA Cancer Immunotherapies,¹ open for public comment from February 3 to March 31, 2025. The draft guideline outlines best practices for drug product design and manufacturing of mRNA cancer immunotherapies using lipid nanoparticle (LNP) delivery systems. It also provides insights into chemistry, manufacturing, and controls (CMC) related aspects of nonclinical, clinical, post-authorization considerations as well as information for patients and healthcare providers.

This review highlights key CMC aspects that applicants may overlook in mRNA therapeutic applications. Rather than reiterating the full draft guideline, we focus on critical areas where compliance with MHRA expectations is essential. When the MHRA specifies that applicants “should” provide certain data, it is generally required for evaluation. If compliance is not feasible, a robust justification will likely be necessary.

The MHRA references additional regulatory documents that provide crucial guidance. Applicants should review these alongside the draft guideline and seek expert regulatory advice to ensure your approach is accurate, appropriate, and timely for your specific product.

mRNA Cancer Immunotherapies: Truly Personalized Medicine

Individualized mRNA cancer immunotherapies, also referred to as bespoke or individualized medicines, are often developed through four critical steps:

1. isolation of genetic material from a patient’s tumorigenic and healthy tissues or cells,
2. a computational analysis to determine novel tumor-specific proteins (i.e., neoantigen, foreign, or non-self-proteins),
3. computational design of one or more mRNAs that encode a therapeutic protein that recognizes the neoantigen(s), and
4. manufacture of a drug product that contains mRNAs encapsulated in a lipid nanoparticle (LNP) delivery system for administration.

The draft guideline does not address alternative technologies, such as peptides, polymer-based delivery systems, or non-integrating DNA, although it commits to providing future updates. It also excludes discussions on continuous learning algorithms, which are not aligned with U.K. medical device regulations. Additionally, while the draft guideline does not define platform technology terminology or applicability, it confirms that developers may submit prior knowledge data for MHRA evaluation to determine its relevance in supporting proposed approaches.

Novelty And Significance Of Draft Guideline

This draft guideline is significant in clarifying two areas of industry interest. First, it clarifies the opportunity to submit a single marketing application (MA) for a product with a variable active component. For example, seasonal influenza vaccines allow for annual product changes to align with circulating influenza variants. In this respect, the mRNA cancer immunotherapies represent a truly progressive approach to product variability. Second, it clarifies current thinking on the use of artificial intelligence (AI), bioinformatics (BI), or other complex computer algorithms in generating the variable active component of a specific drug product. This draft guideline is timely and relevant to several clinical trials in multiple global markets, offering assurance of a viable path toward market approval.

Terminology And Communication

Currently, this draft guideline’s scope is limited to an mRNA cancer immunotherapy defined as an mRNA (active substance) encapsulated in an LNP. The mRNA may contain a variable component that is tailored to a patient’s neoantigen, and the final product may contain more than one mRNA design. The MHRA carefully notes that mRNA cancer immunotherapies do not meet Human Medicines Regulations’ (HMR) definition of a vaccine and therefore the term mRNA vaccine should not be used. The draft guideline specifically excludes guidance on non-lipid nanoparticle delivery or other delivery modalities or the use of circulating tumor DNA profiling for tumor-specific mutations.

To improve regulatory communications with the MHRA, we recommend that applicants use the provided definitions and organize their applications accordingly. Currently, the CMC regulatory view of mRNA cancer therapeutics is that they share strong similarities with advanced therapy medicinal products (ATMPs), specifically gene therapy applications. Therefore, in many cases, similar thinking can be applied to regulatory applications for both product classes.

However, with consideration for the mechanism of action (MOA) and the risk profile, the MHRA is evaluating whether mRNA cancer immunotherapies may be more like mRNA vaccine products. Therefore, the MHRA is considering the addition of a new ATMP sub-classification of nucleic acids because a chemically synthetized mRNA would not be considered a biologic.

The drug product is considered the final formulated mRNA encapsulated in an LNP with any excipients intended for administration. The mRNA is considered an active substance and should be fully described in the application. Starting materials for mRNAs may be either the nucleotide substrate or the ribonucleotides, depending on how the mRNA is manufactured.

The MHRA notes there is no clear definition of what is considered AI and/or ML (machine learning), but the applicant should determine which regulations apply to their specific AI/ML approach (e.g., SaMD) and should provide justification for their assessed use of AI/ML. However, the application’s bioinformatics analysis should typically include the sequencing and preliminary raw data analysis, variant calling, and a software-based model approach for filtering and selecting neoantigens (i.e., AI, ML, or AI/ML).

Aligning Applications With MHRA Expectations

The applicant should showcase a “process as the product approach” by demonstrating that a known manufacturing process can predictably yield drug products with consistent quality attributes. In this respect, the preclinical work may be more robust than a defined gene therapy. For example, an anti-CD19 targeting moiety is easier to characterize than potentially hundreds of unknown, yet-to-be-biologically understood neoantigen binding moieties.

If a program is feeling overwhelmed by unknowns, it may be beneficial to seek outside expertise to provide clarity. We are providing the following highlights for consideration to encourage applicants to design successful clinical trials applications.

Segregation, Traceability, and Identity Across Manufacturing

Ensuring traceability, chain of identity, and segregation throughout the manufacturing process is critical for both therapeutic benefit and patient safety. Hypothetically, a mix-up of products could potentially generate an inappropriate immune response in an unintended recipient (e.g., autoimmune response). Therefore, the applicant must establish a holistic control strategy spanning patient sample collection to final product administration.

Chain-of-identity (COI) must align with CMC chain-of-custody (COC), beginning at sample collection and extending through administration. A single, unique patient identifier is recommended by the MHRA, although other regulatory agencies may want multiple identifiers. The assignment, management, and documentation of these identifiers should be clearly described throughout manufacturing. Since unique identifiers may change at different stages, such as from tumor samples to raw sequencing data to the final product lot, this transition must be well documented. During manufacturing, identifiers must ensure that operators use the correct patient-specific components, including data files, plasmids, and mRNAs.

Traceability must be maintained across all reagents, starting materials, and samples, including software versions, ensuring tracking throughout the batch record. In the event of a safety signal or reagent recall, traceability allows for the rapid identification of affected lots, facilitating prompt corrective and preventive actions (CAPAs) and regulatory communication.

A robust segregation strategy is essential to prevent cross-contamination. Applicants should identify risk points and provide a clear segregation plan, particularly when patient-specific lots are manufactured concurrently or in multi-product facilities. At all times, segregation and unique identifiers must ensure that patient-specific materials are not mixed-up during manufacturing, holding, or storage.

Some examples include, but are not limited to:

• Patient samples (e.g., biopsy tissues, blood)
• Extracted nucleic acids
• Sequencing data
• Bioinformatics data
• Patient-specific manufacturing materials (e.g., plasmids, mRNAs)
• Any drug substances or manufacturing intermediates
• Final drug product

Applicants should also specify whether product contact materials are single-use or shared-use. For shared-use equipment, justification and data demonstrating adequate cleaning between lots must be provided. These expectations align with those for autologous ATMPs, where product mix-ups pose substantial patient risk. Applicants should engage early with contract manufacturing and testing organizations to prevent delays in regulatory compliance and product development.

In-Process Controls, Operating Ranges, and Parameters

For mRNA cancer therapies targeting multipole neoantigens, in-process controls are essential for providing assurance of the process. Multiple checkpoints should be present across the pipeline to determine whether manufacturing should proceed. The acceptance criteria and justification should be provided.

The following are some examples of in-process controls where acceptance criteria can be implemented:

• A tissue sample should be inspected for signs of improper storage or contamination prior to sample preparation.
• While not required, a morphological control may be appropriate, and applicants may find this additional information beneficial as they refine their approach.
• A nucleic acid sample is evaluated for concentration prior to use in sequencing.
• The software analysis should evaluate data quality metrics to determine suitability for downstream analysis, such as read quality, coverage, and read depth.
• Normal operating ranges should be established for equipment based on manufacturing experience to date and may be further narrowed with additional experience.

In early phase studies, it may not be possible to understand all parameters that ultimately ensure the commercial product’s critical quality attributes. Therefore, characterization data should be collected and the MHRA specifically notes that the applicant is required to perform characterization of the samples. Additionally, characterization data can be leveraged to demonstrate comparability across manufacturing changes, which inevitably will occur.  

Justify Selected Approaches

In multiple areas, the MHRA is clear that they will not take a prescriptive approach for CMC. This includes genomics and bioinformatics equipment, kits, reagents, libraries, software, or methods used in generation of mRNA cancer therapeutics. Unfortunately, some applicants mistake this to mean the regulatory agency does not consider the supporting information important or that a regulatory agency lacks the expertise to evaluate the material. The result is often an incomplete application.

Regulatory agencies expect the applicant to know the product best. Therefore, the applicant should craft an application narrative that demonstrates expertise through experience and data. Regulatory agencies have in-house genomics and bioinformatics expertise and are fully capable of reviewing complex omics-based and bioinformatics-informed approaches for mRNA cancer therapeutics.

Here are examples where justification should be provided:

• How variables (parameters) were selected for inclusion in algorithm development, such as analysis of data sets or biological knowledge
• All sequencing and software-based approaches have biases that should be clearly presented, and a discussion of risk mitigation should be provided
• If research-use-only (RUO) instruments and software will be used, justify why it is acceptable by providing robust validation data and explaining how GMP requirements will be met
• A library quantification method should be established and used to quantify each library. Known concentrations for nucleic acids should be established. Biases and limitation of the library and methods should be known, and mitigation approaches should be taken and discussed in the submission
• The design of patient lots and the definition of the final product lot should include account for planned variability. For example, product design and characterization could include a defined numerical range of target neoantigens based on BI analysis results, specified ratios of multiple mRNA-LNPs for a single patient product lot, or parameters related to drug product dosing, such as the total mRNA-LNPs delivered or the minimum dose delivery of a single mRNA-LNP. The approach should ensure collection of safety and efficacy data

Conclusion

The MHRA draft guideline on mRNA cancer therapeutics offers meaningful guidance for regulatory submissions for these unique drugs. While this review is unable to capture the entire draft guideline, it underscores key CMC aspects that applicants may overlook in mRNA therapeutic applications and discusses ways to effectively communicate and demonstrate CMC readiness.

Reference

1. Draft Guidance on Individualised mRNA Cancer Immunotherapies; Medicines and Healthcare Products Regulatory Agency. Published 3 February 2025: https://www.gov.uk/government/consultations/draft-guidance-on-individualised-mrna-cancer-immunotherapies

About The Authors

Tiffany Lucas, Ph.D., is a principal consultant at Eliquent Life Sciences in Washington, D.C. She leverages her six years of experience at the U.S. FDA as a reviewer for cell and gene therapy products, along with her experience as an investment analyst, to solve complex issues related to regulatory affairs. She offers strategic chemistry, manufacturing, and control (CMC) guidance to improve regulatory submissions across all stages of product development, from pre-IND through post-commercialization.

Ifty Saiyed is the head of the European regulatory affairs practice at Eliquent Life Sciences in London, U.K. He is a global regulatory affairs leader with over 25 years of experience across multinational pharmaceutical companies, contract research organizations (CROs), and international regulatory agencies. His expertise spans a wide array of regions, including ICH markets, LATAM, and MENA, where he has consistently navigated complex regulatory landscapes and provided strategic guidance on diverse, high-impact global projects.