Skip to main content
v2026.11,772 entries · CC-BY 4.0

GMP for ATMPs: How EU Part IV Diverges from Standard Good Manufacturing Practice

How the EU GMP guidelines specific to Advanced Therapy Medicinal Products (EudraLex Volume 4, Part IV) diverge from standard GMP: a risk-based, proportionate approach, donor-to-patient traceability, starting-material control for single-source products, and the single-patient batch definition problem.

Ask CASRAI · included with Regulatory Radar

Ask about GMP for ATMPs: How EU Part IV Diverges from Standard Good Manufacturing Practice

Ask CASRAI answers research-administration questions and cites the passages behind every claim — and says so when the corpus does not cover something, instead of guessing. It comes with a Regulatory Radar subscription at $29 a month, alongside the daily digest of regulatory changes and the dashboard of what changed.

150 questions a day, on this site, over the API, or inside your own tools through the CASRAI MCP server.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays free to read, with no account and no card.

Written and maintained by CASRAI Editorial Board

Last updated

The European Union does not apply its standard Good Manufacturing Practice (GMP) rulebook to advanced therapy medicinal products (ATMPs) unchanged. Since 22 May 2018, ATMPs have been governed by a dedicated document — Part IV of EudraLex Volume 4, the European Commission’s “Guidelines on Good Manufacturing Practice specific to Advanced Therapy Medicinal Products” — rather than the Annex 2 biological-medicinal-products rules that would otherwise apply under Part I. That single substitution reflects a real structural problem: the standard GMP model was built for reproducible, multi-unit batches of a stable chemical or well-characterized biologic, and a meaningful share of ATMPs are neither. This guide covers where Part IV actually diverges from general GMP, and why.

What counts as an ATMP

Regulation (EC) No 1394/2007 defines four ATMP categories: gene therapy medicinal products, somatic cell therapy medicinal products, tissue engineered products, and combined ATMPs (a medical device integrated with cells or tissue, such as a cell-seeded scaffold). What unites them for GMP purposes is not the therapeutic mechanism but the nature of the starting material and manufacturing process — living cells or genetic material, frequently sourced from a single named donor or patient, processed through a chain that cannot simply be re-run if a step fails the way a chemical synthesis batch can.

The risk-based approach: proportionality instead of a single fixed standard

Part IV’s central design choice is that GMP requirements should be applied in a manner proportionate to the specific risk profile of the product and its manufacturing process, rather than imposing one uniform standard across every ATMP regardless of complexity or development stage. Factors the guideline points manufacturers and inspectors toward weighing include the product’s novelty and complexity, the manufacturing scale (frequently small-batch or single-patient production rather than industrial-scale runs), the stage of clinical development, and the origin and variability of the starting material. In practice this means an early-phase autologous cell therapy manufactured for a handful of patients is not held to an identical quality-system burden as a large-scale allogeneic product intended for broad commercial distribution — both must be safe and controlled, but the specific controls, and how much of them, scale to the actual risk.

Traceability: donor-to-patient and back, not just batch-to-shipment

Conventional pharmaceutical GMP traces a batch from raw material through to the pharmacy shelf. ATMP traceability has to run further in both directions: from the individual donor (who may be the patient themselves, in an autologous product) through every processing and storage step, to the specific patient who receives the finished product, and back again if a look-back investigation is ever needed — for example, if a donor is later found to have been unsuitable. Regulation (EC) No 1394/2007 requires this donor-to-recipient and recipient-to-donor traceability data to be retained for a period measured in decades, well beyond the shorter retention windows typical of standard drug-batch recordkeeping (the exact retention figure is set out in the regulation’s traceability provisions and is worth confirming against the current consolidated text before citing a specific number in a compliance filing). Where the starting material is human cells or tissue, this traceability obligation sits alongside — and has to be reconciled with — the donor selection, testing, and procurement requirements of the EU Tissues and Cells Directive (2004/23/EC).

Starting-material control: a supply chain of one

For an allogeneic (donor-derived, one-to-many) ATMP, starting-material control looks recognisably like standard biologics sourcing: qualified donor pools, defined testing panels, documented acceptance criteria. For an autologous (patient-derived, one-to-one) product, the starting material is a specific patient’s own cells or tissue, collected on a specific date under specific clinical conditions that cannot be repeated on demand if the collection is compromised. That single-source, non-substitutable starting material is why Part IV treats supply-chain segregation and positive identification — confirming that what goes into the process, and what comes back out of it, unambiguously belongs to the same person — as a first-order manufacturing control, not a peripheral logistics concern.

Batch definition for personalised products

Standard GMP defines a batch as a quantity of product manufactured in one process run, intended to have uniform character and quality. That definition assumes the batch is larger than one dose and that multiple doses from it can be sampled, tested, and released together. Autologous ATMPs break that assumption: when a single patient’s cells are the entire starting material, the “batch” is frequently a single dose for a single patient, manufactured once, that cannot be resampled or remade if testing consumes the only material available. Part IV’s quality-control and batch-release provisions have to accommodate this — through approaches such as parallel/small-scale process validation runs using surrogate or leftover material, real-time release testing where full release testing would otherwise delay a time-critical infusion, and quality agreements that define exactly what evidence stands in for testing the actual patient dose can’t afford to consume. This single-unit batch problem is one of the most frequently raised practical difficulties in ATMP manufacturing, and it is the main reason a generic validation master plan template written for conventional drug manufacturing needs real rework before it fits a cell-therapy production line.

How this differs from the US framework

The United States has no single unified “GMP for ATMPs” document analogous to EU Part IV. Cell and gene therapy products regulated as biologics are instead subject to the standard pharmaceutical current Good Manufacturing Practice regulations at 21 CFR Parts 210 and 211, layered together with the donor-eligibility and current good tissue practice (CGTP) requirements at 21 CFR Part 1271 Subparts C and D for products that meet the human cells, tissues, and cellular and tissue-based products (HCT/P) definition. FDA addresses risk-based proportionality and the same practical difficulties Part IV names — small-scale manufacturing, patient-specific starting material, batch-of-one release — through guidance documents and phase-appropriate GMP expectations rather than a single dedicated regulation. The underlying manufacturing challenges are the same; the EU has chosen to codify the accommodation in one guideline, while the US layers it across multiple regulations and guidance.

Where this sits in a broader quality system

Part IV does not replace the rest of a GMP quality system — it modifies how that system is applied to ATMP-specific risk. The underlying pharmaceutical quality system expectations (documented procedures, deviation handling, corrective and preventive action, the broader GxP framework GLP and GMP both sit inside, a working GxP training programme, quality agreements with sponsors and testing labs, and facility controls such as those covered for standard cGMP facilities or, for the sterile-manufacturing steps most ATMP processes include, EU GMP Annex 1) still apply. What changes is how proportionately those expectations are scaled, and how the traceability, starting-material, and batch-definition provisions specific to living, patient-linked starting material are layered on top. Teams moving from conventional drug or device manufacturing into ATMP production should expect to rework their quality system’s documentation and validation strategy around these three points specifically, not adopt a wholesale different framework.

Frequently asked questions

Does Part IV apply to all ATMPs, or only some?

It applies to ATMPs manufactured under a marketing authorisation or in the context of a clinical trial within the EU, replacing Annex 2 of Part I for these products specifically. Hospital-exemption ATMPs (prepared on a non-routine basis for an individual patient under a member state’s own hospital-exemption scheme, rather than under a centralised EU marketing authorisation) are instead subject to whatever national GMP-equivalent standard the relevant member state has set, which Part IV does not itself define.

Is a risk-based approach the same as a lower standard?

No. Proportionality changes which controls are emphasised and how they are demonstrated, not whether the product has to be safe, consistent, and traceable. A risk-based justification for a smaller or different validation exercise still has to be documented and defensible to an inspector; it is a scaling mechanism, not an exemption.

Why does autologous manufacturing get treated differently from allogeneic manufacturing?

Because the starting material itself is different in kind, not just in scale. An allogeneic product’s starting material comes from a qualified donor pool that can, in principle, be resampled or substituted; an autologous product’s starting material is one patient’s own cells, collected once, with no substitute available if something goes wrong — which is why traceability, positive identification, and batch-of-one release strategies get the specific attention they do.

For the wider quality-system vocabulary this guide assumes — SOPs, deviation handling, CAPA, and how a research-setting quality management system is typically structured — see the lab-compliance pillar, which indexes the rest of CASRAI’s GxP and quality-systems coverage.

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →

Regulatory Radar

Stop finding out after the fact

$29/month, cancel anytime. Daily digest updates from our analysis, a dashboard holding the same items, and a cited assistant for everything they raise.

  • Federal Register, Federal Register+, Grants.gov, Regulations.gov, NSF News, UKRI, plus CASRAI’s own published content.
  • 72,264 indexed passages, and every answer cites the ones it drew on.