ISO 10993Biological evaluation of medical devices

ISO 10993 is a 23-part international standard family (10993-1 through 10993-23) covering the biological evaluation of medical devices. Part 1 — ISO 10993-1:2018 — is the framework standard: it sets out the risk-management approach to biocompatibility, classifies devices by nature and duration of patient contact, and drives the decision about which biological endpoints (Parts 3 through 23) must be evaluated through test, literature, justification or chemical characterisation.

FDA recognises ISO 10993-1:2018 as a consensus standard and issued a 2023 guidance giving the agency's specific expectations. EU MDR Annex I §10 makes biocompatibility a General Safety and Performance Requirement, and the harmonised version of ISO 10993-1 is the route to GSPR conformity. Health Canada, PMDA, TGA, ANVISA and essentially every other regulator follow the same framework.

10993-1 categorises devices on two axes: nature of body contact and duration of contact.

The intersection of these two axes produces a cell in Annex A Table A.1 of 10993-1, which lists the biological-effect endpoints to be evaluated — typically eight to twelve endpoints for a long-term implant and one or two for a transient intact-skin contact.

ISO 10993-18:2020 reset the discipline. Rather than reflexively running every animal test on every endpoint, modern programmes start with chemical characterisation: identify the materials of construction, the manufacturing process, the cleaning and sterilisation regime, then perform extractables and leachables (E&L) studies under exaggerated and simulated-use conditions to identify and quantify every compound that could leach into the patient.

The extractables list goes to a toxicological risk assessment (ISO 10993-17:2023) where each compound is evaluated against published toxicity data and a calculated Acceptable Exposure (AE) is compared to estimated patient exposure. Endpoints where the toxicological risk assessment shows margin can be addressed without biological testing — directly satisfying the 3Rs (replacement, reduction, refinement of animal testing) and the EU MDR Annex I §10.4 hierarchy of evidence.

10993-1 §5 requires a Biological Evaluation Plan (BEP) and Biological Evaluation Report (BER). The BEP is written before testing starts: it captures the device categorisation, the endpoints identified, how each endpoint will be addressed (test, literature, characterisation + risk assessment, justified absence), and the test methodologies and acceptance criteria. The BER is the closeout: it presents the evidence against each endpoint and concludes on biological safety.

The BER is the document the Notified Body or FDA reviewer reads. It must be self-contained — every endpoint addressed with traceable evidence, every justification scientifically defensible, and the conclusion explicitly linking the biological evaluation to the device's intended use and the ISO 14971 risk file.

1. Device categorisation wrong — sites classify a blood-path device as external-communicating-indirect when it is direct, missing the haemocompatibility programme.
2. Biological evaluation copied from a predecessor device without justification that manufacturing process, materials, sterilisation and shelf life are equivalent.
3. Chemical characterisation (Part 18) extractables study run under conditions less aggressive than worst-case clinical use.
4. Toxicological risk assessment (Part 17) AE values derived from non-pharmacopoeial sources without traceability.
5. Sensitisation addressed with LLNA only — FDA increasingly expects guinea pig maximisation for skin-contact polymeric materials.
6. Sterilisation residuals (EtO, ECH) not addressed under ISO 10993-7 when EtO is the chosen sterilisation modality.
7. Particulate matter from polymer-on-polymer wear addressed by literature only, no Part 6 implantation study with histopathology.
8. BER concludes biological safety but the underlying chemical-characterisation report has gaps the BER does not acknowledge.

• BEP template captures categorisation, endpoint matrix and test plan, with cross-links to the ISO 14971 risk file.
• Material declarations are first-class objects — change to a polymer compound triggers a biocompatibility-re-evaluation ticket.
• Sterilisation modality is tracked alongside the device record so EtO devices automatically flag the -7 residuals programme.
• Test reports from contract labs (Toxikon, NAMSA, WuXi AppTec, Eurofins) are filed against their endpoint, with expiry-based re-test reminders.
• BER drafts pull every endpoint conclusion from the same dossier so the closeout document cannot lose synch with the underlying reports.
• Chemical characterisation (Part 18) and toxicological risk assessment (Part 17) outputs are versioned and linked to the analytical-method qualifications.