ISO 22442: Animal Tissue Medical Devices Under MDR and FDA

Introduction

Full conformity with ISO 22442 is the part of an animal-tissue file that nobody argues about, and it is not the part that decides whether the device gets certified. The series describes how to manage the risk. It does not describe the procedure a European notified body must follow before it may issue a certificate, it does not classify the device, and it does not tell an FDA reviewer what to expect in a premarket submission. Manufacturers who read the three parts, implement them properly and assume the regulatory question is answered are the ones who discover, late, that a statutory consultation with a competent authority sits between them and the CE mark.

This is a documentation and sequencing problem far more often than a scientific one. The hazards are well characterised, the control measures are conventional, and the laboratories that run clearance studies are competent. What fails is the join between the standard and the law: a risk management file built to ISO 14971 that never states the justification the regulation demands, or a general safety and performance requirements checklist that treats Section 13.2 as a cross-reference rather than as a set of obligations with their own evidence.

This article covers what the series does and does not do, the current status of each of its three parts, how the word "utilising" decides classification, the sourcing controls that determine the outcome before any validation is run, the evidence routes open to an inactivation claim, the Regulation (EU) No 722/2012 consultation and its statutory clock, what the FDA asks instead, the obligations that survive the certificate, and the six gaps that recur in files that are otherwise sound.

Table of Contents

• Why ISO 22442 conformity is not a conformity route
• The three parts, and the one that has aged badly
• Part 1 makes ISO 14971 do work it does not do alone
• Part 2 turns sourcing into a process rather than a purchase
• Part 3 has not been revised since 2007
• The word "utilising" is what pulls a device into Class III
• Sourcing decides the outcome before a validation is run
• Where the inactivation evidence usually stops short
• The consultation is a scheduling problem first
• What the FDA asks instead, and where the systems diverge
• The obligations that survive the certificate
• Six gaps that recur in animal-tissue files
• Frequently asked questions
• Conclusions

Why ISO 22442 conformity is not, on its own, a conformity route

ISO 22442 is a risk management instrument. Part 1 states this plainly: it applies in conjunction with ISO 14971, and it specifies a procedure to identify hazards and hazardous situations, estimate and evaluate the resulting risks, control them, and monitor the effectiveness of that control. It then sets out how to decide whether the residual risk is acceptable, balancing that residual risk against the expected medical benefit compared with available alternatives. What it deliberately does not do is stipulate levels of acceptability, except for a small set of derivatives dealt with in an annex — tallow derivatives, animal charcoal, milk and milk derivatives, wool derivatives and amino acids.

That reticence is correct standards practice and it is also the source of the misunderstanding. A standard that declines to set acceptance thresholds cannot, by construction, tell a manufacturer whether a given device will be certified. In the European Union that decision is made against the general safety and performance requirements, with the classification rules deciding the conformity assessment route and Regulation (EU) No 722/2012 imposing a procedure on top. In the United States it is made inside a premarket submission against the agency's own expectations. The series feeds both. It replaces neither.

There is a second, more practical reason not to lean on the series alone. A presumption of conformity exists only where the reference to a harmonised standard has been published in the Official Journal in support of the applicable regulation. The lists published under the old directives do not carry over, and the Regulation's own list has been built up through a sequence of Commission implementing decisions rather than issued once. Any claim that a given edition of a part of this series confers a presumption of conformity should be checked against the current consolidated list the day it is written.

The three parts, and the one that has aged badly

The series is published under the general title Medical devices utilizing animal tissues and their derivatives. Three parts are International Standards and a fourth is a technical report. Every part excludes in vitro diagnostic medical devices from its scope, and every part excludes the use of human tissues. Both exclusions matter: an IVD manufacturer using animal-derived antibodies or serum in a reagent is outside this series entirely and has to reason from first principles and from the IVD Regulation.

Part 1 makes ISO 14971 do work it does not do alone

A generic risk management file identifies hazards, estimates risk and documents controls. Part 1 requires the same process but forces it through a set of hazard categories that a generic file will not surface: contamination by bacteria, viruses, transmissible spongiform encephalopathy agents, moulds, yeasts and parasites, and the pyrogenic, immunological and toxicological reactions that animal material can provoke independently of any infectious agent. It also requires the decision to use animal material at all to be justified, which is not something ISO 14971 asks for.

The practical consequence is that the hazard analysis has to be re-run, not annotated. A hazard inventory built for a synthetic device and then extended with a row for "TSE risk" is the signature of a file that has treated Part 1 as a formality. The benefit-risk analysis has the same problem in reverse: it has to conclude on the animal material specifically, against named alternatives, and not simply on the device as a whole.

Part 2 turns sourcing into a process rather than a purchase

Part 2 specifies requirements for controls on the sourcing, collection and handling of animals and tissues, where handling includes storage and transport. It applies where the risk management process of Part 1 requires it, which means the scope of Part 2 is itself an output of the Part 1 analysis rather than a fixed list. Selective sourcing is singled out as particularly important where the tissue or derivative originates from bovine, ovine and caprine species, deer, elk, mink or cats.

Part 2 does not require a full quality management system, but it does specify quality system elements, and the standard notes that those elements can form part of an ISO 13485 system. In a company that already holds an ISO 13485 certificate this is the cheapest part of the whole exercise, because it is mostly a matter of extending existing supplier controls to cover species, source country and abattoir rather than building anything new.

Part 3 has not been revised since 2007

Parts 1 and 2 were both reissued as third editions in 2020. Part 3 was published in 2007, reviewed and confirmed in 2022, and remains current while a revision works its way through the ISO process. It governs the validation of the elimination or inactivation of viruses and TSE agents during manufacture, and it explicitly does not cover other transmissible and non-transmissible agents.

This asymmetry is worth stating to a project team early. The part of the series that carries the heaviest technical load, and that supports the claims most likely to be challenged, is the part that has stood still while reviewer expectations have moved. Conformity with the 2007 text is necessary and it is not, by itself, a persuasive answer to a well-briefed assessor asking why a particular model agent was chosen. Where the standard is silent or dated, the technical report in Part 4, the ICH viral safety guideline and the published scientific opinions carry the argument.

The word "utilising" is what pulls a device into Class III

Under Rule 18 of Annex VIII to the EU Medical Device Regulation, all devices manufactured utilising tissues or cells of human or animal origin, or their derivatives, which are non-viable or rendered non-viable, are classified as Class III. The single exception is narrow: devices manufactured utilising animal-origin material that are intended to come into contact with intact skin only.

The operative word is "utilising", not "containing". A device manufactured using animal-derived material falls under the rule whether or not that material remains in the finished product. This is the provision that surprises people. Fetal calf serum in a cell culture step, animal-derived enzymes in a purification process, tallow derivatives used as lubricants or release agents, gelatine in a coating that is later removed — each of these is a route into Class III for a device the manufacturer has always described as synthetic. The bill of materials is not where this is found; the process description is.

The exclusion that most often applies is not in the classification rule at all. Regulation (EU) No 722/2012 does not apply to tallow derivatives processed under conditions at least as vigorous as those in Section 3 of its Annex I: transesterification or hydrolysis at not less than 200 °C for not less than 20 minutes under pressure, saponification with 12 M sodium hydroxide, or distillation at 200 °C. A supplier declaration that names the process and its parameters can therefore take a device out of the consultation procedure entirely. A declaration that says only "meets ISO 22442" cannot.

Two scope limits are worth holding in mind together, because they are not the same limit. The series applies to animal material generally and is not written around a species list. Regulation (EU) No 722/2012 applies only to tissues and derivatives originating from bovine, ovine and caprine species, deer, elk, mink and cats. A porcine-derived device is squarely inside the series and squarely inside Rule 18, and it is outside 722/2012. That distinction changes the project timeline by months, and it is routinely got wrong in both directions.

Sourcing decides the outcome before a validation is run

Section 13.2 of Annex I to the Regulation sets three obligations for devices manufactured utilising animal tissues or cells, or their derivatives, that are non-viable or rendered non-viable. Where feasible for the species concerned, the material must originate from animals subjected to veterinary controls adapted to the intended use of the tissues, and information on the geographical origin of the animals must be retained by the manufacturer. Sourcing, processing, preservation, testing and handling must be carried out so as to provide safety for patients, users and, where applicable, other persons. And where the material falls within Regulation (EU) No 722/2012, the particular requirements of that Regulation apply.

Those particular requirements are where the detail lives. Annex I to Regulation (EU) No 722/2012 requires the manufacturer to justify the decision to use animal tissue at all, specifying species, tissue and sourcing, and weighing clinical benefit and residual risk against suitable alternatives such as lower-risk tissues or synthetic materials. It then sets out the factors that determine safety, and each one has to be analysed, evaluated and managed:

• Animals as a source of material, with risk animals — fallen stock, emergency slaughtered animals and TSE-suspected animals — excluded outright, and sourcing from young healthy animals treated as risk-reducing
• Geographical sourcing, assessed against the BSE status established for countries and regions by Commission Decision 2007/453/EC
• The nature of the starting tissue, against the WHO guidelines on tissue infectivity distribution, with only Category 3 material under Regulation (EC) No 1069/2009 permitted, and the specified risk material list of Regulation (EC) No 999/2001 treated as potentially of high infectivity
• Slaughtering and processing controls to prevent cross-contamination through collection, processing, handling, storage and transport
• The quantity of tissue required per unit of device, and whether the process can concentrate any infectivity present
• The quantity, area, type, condition and duration of contact with the patient or user, and the number of devices that might be used in one procedure or across a patient's lifetime
• The route of administration as stated in the product information

Read as a list this is a checklist. Read as a sequence it describes the only lever a manufacturer has once the material is chosen. High-infectivity tissue must not be sourced unless exceptional circumstances apply, justified by important patient benefit and the absence of an alternative. Everything downstream can reduce risk but cannot undo that choice.

Where the inactivation evidence usually stops short

Regulation (EU) No 722/2012 splits devices into two populations before any study is designed. For devices that cannot withstand an inactivation or elimination process without undergoing unacceptable degradation, the manufacturer must rely principally on the control of sourcing. For everything else, if a claim is made that the manufacturing process removes or inactivates TSE infectious agents, that claim must be substantiated by appropriate documentation.

The first population is the one that generates avoidable findings. A collagen or pericardium device whose structure would be destroyed by a rigorous inactivation step is not deficient for having no clearance study. It is deficient for not saying so. Where the file simply omits the validation without explaining why, an assessor reads an omission rather than a design constraint, and the question comes back as a deficiency instead of being closed at the first review.

For the second population, two routes are open. Relevant information from an analysis of appropriate scientific literature can support inactivation and elimination factors, provided the processes described in that literature are genuinely comparable to those used for the device. That analysis must also cover the scientific opinions adopted by European and international bodies, which the Regulation names: the former Scientific Steering Committee, EFSA, the EMA, the World Organisation for Animal Health and the WHO. Where those opinions conflict, they serve as the reference point. A literature review built only from journal articles is not what the Regulation describes.

Where the literature does not substantiate the claim, a specific inactivation or elimination study is required, and the Regulation lists what has to be considered in it: the hazard identified for the tissue, the identification of relevant model agents, the rationale for the particular combination of model agents chosen, the step or stage selected to eliminate or inactivate the agents, documentation of the validation parameters, and the calculation of the reduction factors. The final report must identify the manufacturing parameters and limits that are critical to the effectiveness of the process, and documented procedures must ensure those validated parameters are applied in routine manufacture.

The item most often challenged is the third one: the rationale for the combination of model agents. A study that names its model agents without linking each to a hazard identified for that specific tissue has produced numbers that a reviewer cannot connect to anything. This is also the item that laboratories cannot supply, because the link runs back to the manufacturer's own hazard analysis.

The consultation is a scheduling problem first

This is the step that has no equivalent anywhere else in the world, and it is the one most likely to be discovered too late. Before issuing the certificate, the notified body must inform the competent authorities of the other Member States and the Commission of its assessment, through its own competent authority acting as coordinating competent authority, by means of a summary evaluation report. The Regulation prescribes the content of that report in its Annex II: nineteen numbered fields covering the notified body's designation, the product description and composition, the intended use, whether an EDQM TSE certificate of suitability is available, the nature of the starting tissue, the species and geographical sources, the key elements adopted to minimise the risk of infection, an estimate of the TSE risk, the justification for using animal material at all, and the approach taken to auditing source establishments and suppliers.

The competent authorities may then submit comments, and the deadlines are the numbers that belong in a project plan. Four weeks where a TSE certificate of suitability has been submitted for the starting material. Twelve weeks where it has not. The authorities and the Commission may agree to shorten those periods, but that is a possibility, not a plan. The notified body must give due consideration to any comments received and must convey an explanation of that consideration, including a justification where it decides not to take account of a comment.

Under the MDR the requirement is unchanged and is anchored in Annex IX, with the notified body required to apply the relevant requirements of the Regulation and to carry out the consultation. Annex VII requires notified bodies to hold documented procedures that fulfil those requirements, including for the preparation of the summary evaluation report — which is a designation question, and a reason to confirm early that the chosen notified body is actually designated for this.

MDCG 2020-12 closes off the assumption that most affects legacy devices. For the first certification under the MDR, the notified body must submit the full summary evaluation report to the competent authorities, even where a consultation already took place under the Directives. Where nothing in the manufacturer's 722/2012 documentation has changed, the report may be accompanied by a declaration from the notified body stating which elements have remained identical, and where the notified body's own assessment is also unchanged that can be included too. That is a real mitigation and it is worth preparing deliberately, because it is the only thing standing between a well-run transition and a second twelve-week wait.

What the FDA asks instead, and where the systems diverge

The FDA's position is set out in a final guidance issued in March 2019, Medical Devices Containing Materials Derived from Animal Sources (Except for In Vitro Diagnostic Devices), which replaced a text dating from November 1998 and followed a draft published in January 2014. Its stated purpose is to provide clarification and updated information on the use of ISO 22442-1, ISO 22442-2 and ISO 22442-3 to support applications to the agency, and it covers four things: what should be documented at the manufacturing facility, what should be included in a premarket submission, how specific aspects of the quality system regulation should be applied to control and document manufacture, and approaches for determining whether manufacturing methods eliminate viral contamination from the final product.

Two differences matter more than the rest. The first is that the FDA guidance weights viral pathogens as heavily as TSE agents, and covers a broader range of both than the 1998 text did. European practice, shaped by Regulation (EU) No 722/2012, is heavily TSE-centric, and a file assembled for a European submission can arrive in the United States with its viral safety argument comparatively thin. The second is structural: there is no consultation step. The assessment sits inside the premarket submission and the review clocks are the ordinary ones for the submission type. A US programme built on the assumption that the EU timeline transfers will be too long; an EU programme built on the assumption that the US timeline transfers will be too short.

The mechanism for relying on the series also differs. In the United States the parts of ISO 22442 are recognised consensus standards, and a declaration of conformity may refer only to a standard for which the agency has assigned a recognition number. Recognition is granted to a specific edition and to a stated extent — full, partial, or with agency-identified deviations — and the supplemental information sheet sets out the conditions and any transition dates. Extent of recognition is not uniform across the three parts, and it changes. Checking the current entry in the recognised consensus standards database is a task for the week the submission is compiled, not a fact to be carried over from the last project.

Canada is the useful third comparison because it does something neither of the other two does: it tiers the data requirements by material. Health Canada's guidance distinguishes limited, reduced and full biological safety data, so highly processed lower-risk materials such as collagen from hides and skins do not attract the package demanded for higher-risk starting tissue. It also brings viable animal tissue into scope, which the ISO series does not. That tiering is why the Canadian dossier cannot simply be a translation of the European one.

The obligations that survive the certificate

Regulation (EU) No 722/2012 requires the manufacturer to establish and maintain a systematic procedure to review information gained about the device or similar devices in the post-production phase, and it names three triggers for feeding that information back into risk management: a previously unrecognised hazard is identified, the estimated risk arising from a hazard has changed or is no longer acceptable, or the original assessment is otherwise invalidated. Where any of these occurs, the risk control measures must be reconsidered and, if residual risk or its acceptability may have changed, the impact on measures already implemented must be re-evaluated and justified.

Alongside that sits a change-control obligation with teeth. Any change to sourcing, collection, handling, processing, inactivation or elimination, and any new information on TSE risk relevant to the device, must be transmitted to the notified body and, where applicable, approved by it before implementation. Changing an abattoir is not an internal supplier change in this context. If the accumulated information leads to an increase in the overall TSE risk, the whole Article 5 procedure applies again, consultation included.

Certificate renewal has its own review. For an extension of up to a further five years the notified body must review at minimum an updated justification for using animal tissue against lower-risk and synthetic alternatives, an updated risk analysis and clinical evaluation, updated test data or rationales against current harmonised standards, any changes since the last renewal that could affect TSE risk, and evidence that the design dossier remains state of the art. That last item is the one that ages.

None of this warrants a parallel system. The animal-tissue triggers belong inside the existing post-market surveillance plan, with a named owner and a stated review frequency for the inputs a general PMS process will not otherwise capture: changes to the BSE status of source countries, zoonotic reports affecting the source species, and scientific opinions from the bodies the Regulation names.

Six gaps that recur in animal-tissue files

The findings below are documentation and sequencing failures rather than scientific ones, which is exactly why they survive into files that are technically competent. A laboratory can produce an excellent clearance study; nobody outside the manufacturer can write the justification that connects it to the device.

Of the six, the one with the largest commercial consequence is the third. A consultation discovered after design freeze does not add four or twelve weeks to the programme; it adds those weeks plus the time needed to assemble a summary evaluation report to a standard the notified body is willing to submit under its own name, plus the time to resolve any technical documentation nonconformities first, because a notified body will not normally start the consultation while findings remain open against the file. Twelve weeks of statutory clock routinely becomes six months of elapsed time.

Frequently asked questions

Does ISO 22442 apply to in vitro diagnostic devices?

No. Every part of the series excludes in vitro diagnostic medical devices from its scope, and so does the FDA guidance that supports it. An IVD manufacturer using animal-derived antibodies, serum or enzymes has to build the equivalent argument from the IVD Regulation and from general biological safety principles instead. The absence of a purpose-written standard is a gap in the international framework, not a sign that the risk is lower.

Does a device made only from porcine tissue require the Regulation (EU) No 722/2012 consultation?

No. That Regulation applies to tissues and derivatives originating from bovine, ovine and caprine species, deer, elk, mink and cats. Porcine material is outside it, so there is no summary evaluation report and no competent authority consultation. The device is still classified under Rule 18, Section 13.2 of Annex I still applies, and ISO 22442 still applies in full. Only the consultation procedure and its statutory clock fall away.

How long does the 722/2012 competent authority consultation take?

Competent authorities may submit comments within four weeks where a TSE certificate of suitability issued by the EDQM has been submitted for the starting material, and within twelve weeks where it has not. The authorities and the Commission may agree to shorten those periods, and MDCG 2020-12 recommends expediting the review where most elements of the report are unchanged, but neither is something a project plan can assume.

Does ISO 22442 apply when animal material is used only during manufacturing?

Yes, and so does Rule 18. The classification rule covers devices manufactured utilising animal tissues or their derivatives, not only devices containing them. Fetal calf serum in a culture step, animal-derived enzymes, and tallow derivatives used as processing aids all bring the device within scope even when the material is not present in the finished product. The check has to be run against the process description, not the bill of materials.

Which edition of each part of ISO 22442 is current?

Part 1 and Part 2 are both 2020 third editions. Part 3 remains the 2007 edition, reviewed and confirmed in 2022, with a revision in drafting at ISO. Part 4 is a 2010 technical report and is informative rather than normative. Citing a superseded edition of Part 1 or Part 2 is a common and easily avoided finding.

Are tallow derivatives excluded from Regulation (EU) No 722/2012?

They are excluded where they have been processed under conditions at least as vigorous as those set out in Section 3 of Annex I to that Regulation, which specifies transesterification or hydrolysis at not less than 200 degrees Celsius for not less than 20 minutes under pressure, saponification with 12 M sodium hydroxide under defined batch or continuous conditions, or distillation at 200 degrees Celsius. The exclusion depends on evidence of the actual process, which means a supplier declaration naming the parameters.

Conclusions

Treat the series as the technical spine and the regulation as the schedule. Three decisions determine whether an animal-tissue programme runs to plan, and all three are made before design freeze: whether the material is in scope of Regulation (EU) No 722/2012 at all, whether the device can withstand an inactivation process or must rely on sourcing, and whether an EDQM certificate of suitability will be available. The first decides whether there is a consultation, the second decides what evidence has to be generated, and the third decides whether the wait is four weeks or twelve.

Write the three arguments the regulation asks for and write them explicitly: why animal material at all, against which named alternatives; why this sourcing strategy, against the BSE status of the source country and the infectivity of the starting tissue; and why this inactivation evidence, against the hazards identified for that specific tissue. The Risk Management Documentation Kit on MD Regulatory provides the plan, report, hazard analysis, FMEAs and GSPR checklist those arguments have to live in — aligned with EN ISO 14971:2019/A11:2021 and MDR Annex I, and ready to drop into an existing ISO 13485 quality management system.

Related articles

• Biocompatibility Testing for Medical Devices
• ISO 10993-1:2025 Biological Evaluation
• EU MDR Clinical Evaluation

Primary sources: Commission Regulation (EU) No 722/2012 on EUR-Lex, MDCG 2020-12 on transitional provisions for consultations, and the FDA guidance on medical devices containing materials derived from animal sources.