ISO 10993 Biological Endpoint Selection: Quick Reference

What This Quick Reference Does—and Does Not Do

This page is an original, non-normative decision guide. It deliberately does not reproduce the copyrighted tables, wording, or requirements of ISO 10993-1, and it does not reproduce FDA's endpoint tables. Use it to organise the analysis before consulting the purchased standard, current regulator sources, applicable device-specific guidance, and qualified specialists.

Endpoint selection is only one part of biological evaluation. It does not by itself determine a laboratory method, test article, extraction design, acceptance criterion, toxicological strategy, or final biological-safety conclusion. Those decisions depend on the device, the evidence and the regulatory pathway.

Current Status: ISO, FDA, and EU MDR

ISO 10993-1:2025 is the current sixth edition. ISO records it as published in November 2025 and lists the 2018 edition as withdrawn. The 2025 scope places biological evaluation within risk management, addresses direct and indirect body contact, and considers device constituents, tissue-device interaction, foreseeable use conditions, and lifecycle change.

Part 3 publication update—checked August 23, 2026: ISO 10993-3:2026 is the current fourth edition for evaluating genotoxicity, carcinogenicity, reproductive toxicity and developmental toxicity after the need for those evaluations has been established under ISO 10993-1. ISO's public preview identifies chemical characterization and toxicological risk assessment as an approach for these effect areas, but it does not provide the operative detail needed for clause-level implementation. Publication by ISO does not itself establish FDA recognition, EU harmonisation or national adoption; verify the current market position and use the licensed final text.

The Endpoint-Selection Workflow

1. Define the evaluation unit. Identify the marketed configurations, accessories, variants, reusable life, target populations, indications, and markets covered.
2. Map every body-contact route. Record direct and indirect contact for each relevant component, including fluid or gas pathways and practitioner contact where applicable.
3. Describe actual exposure. Capture anatomical site, tissue or fluid, frequency, single-use duration, repeated or intermittent use, cumulative exposure assumptions, and reasonably foreseeable misuse.
4. Characterise the final finished device. Include material grades, additives, colourants, coatings, adhesives, process aids, cleaning, sterilisation, packaging interaction, aging, wear, corrosion, and degradation where relevant.
5. Identify biological hazards and endpoint families. Ask what harm could arise from constituents or tissue-device interactions instead of copying a test list.
6. Review evidence before requesting new work. Assess final-device data, justified comparisons, chemistry, toxicology, literature, clinical experience, complaints, and post-market information for applicability and quality.
7. Record an endpoint disposition. For each relevant effect, explain the evidence route, limitations, uncertainty, conclusion, and any additional characterisation, analysis, or testing.
8. Apply the market overlay and lifecycle triggers. Check current FDA or EU conditions, device-specific requirements, and changes that could invalidate the conclusion.

Endpoint-Family Decision Table

The table below contains decision questions and evidence routes—not “required” or “not required” answers. The triggers are examples that can increase relevance; they are neither exhaustive nor automatic instructions to test.

Material-mediated pyrogenicity should not be confused with bacterial-endotoxin control. ISO's published scope information treats infectious-agent risks outside ISO 10993-1 and points bacterial endotoxin questions to other standards. The two controls can coexist but answer different questions.

Contact and Duration Are Inputs, Not the Whole Decision

Contact classification remains a powerful screening tool, but the product name is not the classification. A catheter can contain a skin-contacting hub, an indirectly contacting fluid path, an adhesive joint, and a blood-contacting distal section. Each component can contribute a different route and biological question. The analysis should show the map, not compress the device into one convenient label.

FDA's public endpoint resource uses three duration bands: limited contact of 24 hours or less, prolonged contact of more than 24 hours through 30 days, and long-term or permanent contact of more than 30 days. Those are FDA framework labels. Do not present them as a complete statement of every exposure-duration provision in ISO 10993-1:2025.

Repeated and intermittent exposure needs device-specific reasoning. Record the actual clinical schedule, expected number of uses, intervals, constituent persistence, and any potential accumulation. FDA's 2026 recognition entry notes that further technical work is developing on intermittent contact, bioaccumulation, foreseeable misuse, and lifecycle application. Where classification materially affects a US submission, discuss uncertainty with the appropriate review office or through the Q-Submission process.

Addressing an Endpoint Does Not Always Mean Testing It

FDA's own endpoint pages state that their tables are a framework rather than a testing checklist. They also state that recommended endpoints should be addressed through existing data, endpoint-specific testing, or a rationale explaining why additional assessment is unnecessary. The strength of a no-new-testing conclusion therefore depends on the evidence—not on the elegance of the sentence.

• Final-device data: usually persuasive when the article, processing, sterilisation, contact and use conditions represent the current device.
• Comparison or bridging: useful only when formulation, supplier, processing, surface, geometry, exposure and other relevant differences are disclosed and scientifically assessed.
• Chemistry and toxicology: can address constituent-exposure questions when the analytical coverage, exposure assumptions, identification, uncertainty, and toxicological reasoning are fit for purpose.
• Literature and clinical history: can add valuable evidence when the device or material match, methods, exposure, adverse findings, and access to underlying details are made clear.
• New testing: may be proportionate when a consequential hazard remains unresolved, the existing evidence does not represent the marketed device, or novelty and uncertainty are material.

What to Record for Every Endpoint

FDA-Specific Cautions

• Partial recognition: FDA recognition 2-313 excludes the phrase “consumer products” in clause 6.5.11.3 and clause 6.9 on biological risk estimation.
• Genotoxicity alignment: FDA states that additional genotoxicity provisions in the 2025 edition might not align with FDA Attachment A for every prolonged-contact device and encourages early review-office discussion.
• Transition wording: July 1, 2029 is the end of FDA's stated period for accepting declarations of conformity to recognition 2-258. Do not call it an expiry date for every 2018-era report.
• Final finished form: FDA evaluates the device as supplied, including sterilisation where applicable, and considers materials, manufacturing, interactions, and residuals—not isolated raw-material labels.
• Device-specific sources: applicable FDA guidance and special controls can add or refine biocompatibility expectations.

FDA's September 2024 guidance on chemical analysis for biocompatibility remains labelled Draft—Not for Implementation on FDA's official page as of this review. It may inform awareness, but it should not be presented as final FDA policy.

EU MDR-Specific Cautions

MDR Annex I section 10 addresses chemical, physical, and biological properties, including material choice, toxicity, compatibility with tissues, cells and body fluids, intended purpose, process effects, and relevant physical properties. Annex II requires technical documentation to include biocompatibility information and identification of all materials in direct or indirect contact with the patient or user. Where no new testing is performed, the technical documentation still needs a rationale.

EN ISO 10993-1:2025 is harmonised under the MDR, but harmonised standards remain voluntary. Conformity provides a presumption only for the requirements the standard covers. A biological evaluation therefore still needs explicit links to applicable General Safety and Performance Requirements, the device's risk management, preclinical evidence, clinical evaluation, and lifecycle information.

A Worked Reasoning Example—Without Prescribing a Test Battery

Consider a sterile, single-use fluid-delivery catheter family. The first task is not to copy the catheter row from a table. It is to map the components: a hub contacting the practitioner and possibly skin; a fluid pathway that can transfer constituents indirectly; bonded joints with adhesive; and a distal section with tissue or blood contact. Next, define the maximum procedure time, repeated exposures across clinical use, patient population, delivered fluids, sterilisation, shelf life, and the worst-case family configuration.

The team then asks which endpoint families are triggered by those routes, materials, processes, flow conditions, and uncertainties. Existing data are checked against the current adhesive, tubing grade, colourant, sterilisation dose, surface area, extraction article, and marketed geometry. Chemistry may resolve constituent questions; hemocompatibility evidence must reflect the actual blood-contact situation; a prior study may be bridgeable only after differences are analysed. The result is an endpoint-by-endpoint evidence map with gaps—not a generic list of tests for “a catheter.”

Common Endpoint-Selection Failures

• Treating every marked cell as a mandatory new test.
• Treating an unmarked cell as permission to ignore a device-specific hazard.
• Using one contact classification for a multi-component device without a contact map.
• Applying FDA duration bands without labelling them as FDA's framework.
• Calling a base polymer “ISO 10993 certified” or “FDA-approved.”
• Bridging from a predicate or legacy device without supplier, formulation, process, sterilisation, geometry, surface, and exposure comparison.
• Ignoring repeated use, degradation, wear, corrosion, aging, packaging, or residues.
• Writing “not applicable” or “passed previously” without evidence and applicability analysis.
• Claiming FDA fully recognizes ISO 10993-1:2025 or that EU harmonisation makes it legally mandatory.
• Using one endpoint strategy globally without a documented jurisdiction check.

Frequently Asked Questions

Limits of This Guide

This information is educational and does not provide a device-specific test plan, regulatory determination, legal opinion, or assurance of regulator or notified-body acceptance. Standards, recognition conditions, guidance, harmonisation decisions, and device-specific expectations change. Verify the controlled primary sources at the time of each evaluation and use qualified biological-safety, toxicology, clinical, and regulatory expertise where appropriate.