If you’re developing a medical device, sooner or later you’ll confront the alphabet soup of ISO 10993-18 requirements. Regulators in the U.S., Europe, Japan, and China all lean on this standard, and each one has its own nuances. Getting chemical characterization right the first time can mean the difference between a smooth submission and months of deficiency letters.
This guide explains what ISO 10993-18 actually requires, what regulators are specifically looking for, and how to build a defensible testing strategy from day one.
What is ISO 10993-18?
ISO 10993-18:2020 (formally titled “Biological Evaluation of Medical Devices – Part 18: Chemical Characterization of Medical Device Materials Within a Risk Management Process”) defines how manufacturers should identify, measure, and evaluate the chemical substances present in or migrating from their devices.
It was substantially revised in 2020, moving from a general document to a risk-based framework. This shift mirrors the evolution of ISO 10993-1, that, starting in 2009, repositioned material characterization as the starting point — not an afterthought — of the biological evaluation process.
“Chemical characterization involves the identification of a material and the identification and quantification of its chemical constituents as part of an assessment of the overall biological safety of a medical device.” — ISO 10993-18:2020
Where ISO 10993-18 Fits in Your Biocompatibility Strategy
The biological evaluation of medical devices conducted within a risk management process as described by ISO 10993-1:2025, indicates that gathering physical and chemical information is a crucial first step in the biological evaluation process. As indicated in ISO 10993-1:2025 and in ISO 10993-18:2020, the extent of physical and chemical information required depends on the nature and duration of body contact. By gathering information that is currently available, one can determine if there are gaps which require additional data to be generated. Chemical characterization testing may be one source of this additional information. If chemical characterization testing is performed, it can help confirm which biological tests are necessary, set the scope for the toxicological risk assessment (TRA), and become an integral part of the Biological Evaluation Report (BER).
Think of ISO 10993-18 as the intelligence-gathering phase. At a minimum it can be used as a reference for the process of gathering information which supports the use of certain materials, processes or chemicals in a device. Depending on the nature and duration of patient contact and other factors, it may then extend to the generation of data to support the information gathered or to fill gaps with regards to the information available. This is the point at which chemical characterization testing, often extractables testing, becomes part of a Biological Evaluation Plan (BEP). The BEP defines how the chemical characterization process fits into the overall assessment for a device, so that defensible decisions about biological safety can be made.
A few things this framework makes clear:
Chemical characterization may reduce the extent of biological testing. The outputs of ISO 10993-18 (the list of identified extractables and leachables) supported by a TRA based on ISO 10993-17 may reduce the required in vitro and in vivo testing. Skip or shortcut this step, and your biological test plan may lose its scientific rationale.
ISO 10993-18 connects directly to ISO 10993-17. Every chemical identified above the AET (Analytical Evaluation Threshold) needs a toxicological risk assessment under ISO 10993-17.
The BER is only as strong as its chemical foundation. Regulators reviewing a submission will scrutinize whether the chemical characterization study was adequately designed, whether the AET was properly justified, and whether every material in the device was accounted for. Gaps here cascade into gaps in the BER (and deficiency letters).
The Three Core Approaches Under ISO 10993-18
The standard provides manufacturers with three distinct pathways for chemical characterization. You don’t always need all three; the right approach depends on your device type, materials, and clinical exposure.
- Compositional Evaluation: Review material composition data, often from supplier documentation, existing literature, or databases. May be sufficient for well-characterized materials with limited contact or have identical materials, manufacturing, patient exposure, and packaging as previously approved devices.
- Extractables Evaluation: Lab-based extraction under exaggerated or exhaustive conditions (aggressive solvents, elevated temperature) to identify all potential chemical constistuents above the AET in a worst-case scenario.
- Leachables Evaluation: Testing under simulated clinical use conditions. Measures which chemicals migrate into patients during normal device operation.
In practice, extractables and leachables (E&L) testing is the most common path for devices with direct or indirect patient contact, especially when the exposure is prolonged (>24 hours, to 30 days) or long-term (>30 days). The extractables study establishes an upper bound on chemical risk while the leachables study confirms what reaches the patient.
The Analytical Evaluation Threshold (AET): A Critical Concept
One of the most important additions in the 2020 revision is the formal definition and use of the Analytical Evaluation Threshold (AET). This concept bridges chemistry and toxicology — and misunderstanding it is one of the most common reasons manufacturers receive regulatory deficiencies.
What is the AET?
The AET is the concentration threshold (expressed as µg/mL or µg/device) below which a detected chemical need not be formally identified or reported. It is calculated from a toxicologically valid exposure limit value (Threshold of Toxicological Concern (TTC), Tolerable Intake (TI), etc.), and it must be established before the chemistry study begins, not retroactively. If your analytical method isn’t sensitive enough to detect compounds at or below the AET, your data may not be accepted by regulators.
NAMSA’s team of chemists and toxicologists works collaboratively to set the AET before any lab work begins — ensuring your study is designed with regulatory defensibility built in.
What Regulators Are Looking For
The ISO standard is globally referenced, but regional regulatory bodies each add their own layer of expectations. Here’s a comparison of the most prominent global markets:
| Regulator | Key Expectations |
|---|---|
| FDA (U.S.) | – Use of polar, mid-polar and non-polar extraction vehicles. Any deviation must be justified. – Exhaustive extractions are expected unless the duration of extraction exceeds or meets the duration of exposure on a cumulative basis – Correct selection of dose-based threshold to calculate the AET – Use the FDA 2024 draft guidance as a reference alongside ISO 10993-18. |
| EU (MDR / Notified Bodies) | – Physical and chemical information which is part of chemical characterization must be part of a BEP – The presence of CMR substances (carcinogens, mutagens, reproductive toxins) and endocrine disrupters must be justified if ≥0.1% concentration. |
| Japan (MHLW/PMDA) | – Recommends that if chemical characterization testing is part of your BEP, a consultation meeting with PMDA should be considered to discuss the approach. |
| China (NMPA) | – Acceptance of chemical characterization with a TRA to offset some biological endpoints can be questioned – Reviewer discretion varies – May require reference to equivalent Chinese standards. |
Year of the last major ISO 10993-18 revision
Characterization approaches defined by the standard
Major global regulators referencing ISO 10993-18
Most common root cause of biocompatibility deficiencies
Designing a Defensible Chemical Characterization Study
The most common mistake manufacturers make is treating chemical characterization as a checkbox exercise rather than a scientific investigation. Regulators have become increasingly sophisticated and so have their questions. Here’s what a well-designed study includes:
Material Inventory
Before any testing, compile a complete list of all device materials, including polymers, adhesives, colorants, lubricants, coatings, additives, and potential processing residues. Supplier documentation should be gathered early; gaps here often delay the entire program.
Appropriate Extraction Conditions
Extraction solvent polarity, temperature, and duration must reflect the device’s clinical exposure and consider the purpose of the study being performed. A polar aqueous solvent mimics body fluid contact; a nonpolar organic solvent captures fat-soluble compounds. Most studies use a combination of polar, mid-polar and non-polar. The FDA’s 2024 draft guidance emphasizes that extraction conditions should be tailored to device type and duration of use.
Multi-Detector Analytical Methods
Comprehensive chemical characterization requires multiple analytical techniques — typically GC-MS, LC-MS/MS, ICP-MS, and NVR (non-volatile residue) gravimetry — to detect the full spectrum of volatile, semi-volatile, non-volatile, and metallic compounds. No single technique covers everything.
Toxicological Risk Assessment (TRA)
Every identified compound above the AET must be evaluated against established toxicological limits via a formal TRA, conducted in accordance with ISO 10993-17:2023.
Related Reading
ISO 10993-18 Primer: The Complete Guide to Medical Device Chemical Characterization
Go deeper on regulatory expectations, EU MDR nuances, and how to prepare your chemical characterization strategy for multi-region submissions.
Common Regulatory Deficiencies — and How to Avoid Them
Based on submission experience across FDA, EU Notified Bodies, and PMDA, these are the recurring deficiencies that delay approvals:
- Inadequate AET justification: The AET must be derived from a toxicologically valid value, documented, and agreed upon before the study. Post-hoc AET adjustments raise red flags with reviewers.
- Incomplete material information: Missing data on colorants, processing aids, or secondary materials is a frequent FDA query topic. Every component that contacts the patient — directly or indirectly — must be accounted for.
- Use of outdated standards: ISO 10993-18:2020 superseded the 2005 version significantly. EU Notified Bodies will require a gap analysis if data referencing the older version is submitted.
- Missing NVR analysis: The FDA’s 2024 draft guidance specifically calls out non-volatile residue (NVR) analysis as a required step for devices with prolonged or permanent contact, especially in determining an exhaustive endpoint and correlating NVR to the quantities of identified constituents in the analytical methods used. It’s not optional.
- Siloed chemistry and toxicology: Chemistry and toxicology teams must co-design the study — the AET is a prime example of where disconnected teams create regulatory problems.
NAMSA’s ISO 10993-18 testing services are designed around integrated chemistry and toxicology, so your study is submission-ready from day one.
When Do You Need Chemical Characterization?
Chemical characterization under ISO 10993-18 is required or strongly advisable in several scenarios. Each of these needs to be considered in the context of the device and its exposure to the patient and note that not all will necessarily involve chemical characterization testing:
- New device submissions — required as part of the biocompatibility evaluation package for regulatory submissions
- Material changes — even a supplier change for the same nominal material may introduce different extractable profiles requiring reassessment.
- Manufacturing process changes — new sterilization methods, coatings, or processing aids can alter a device’s leachable profile significantly.
Frequently Asked Questions (FAQs)
Do regulators always require extractables testing under ISO 10993-18?
Not always. ISO 10993-18 supports a stepwise approach that may include compositional evaluation, extractables, and/or leachables depending on what information is needed to identify hazards and estimate risk. The justification must match device-specific context and risk.
What’s the fastest way to trigger regulator questions on chemical characterization?
The most common triggers include unclear worst-case device selection, weak extraction justification (especially for prolonged/long-term devices), and missing/unclear AET logic showing that method sensitivity is adequate.
Where does chemical characterization “end” and the TRA begin?
Chemical characterization generates and organizes chemical identity/quantity information (ISO 10993-18). That dataset feeds the toxicological risk assessment framework (ISO 10993-17) and ultimately supports biological evaluation conclusions (ISO 10993-1).
How does FDA’s chemical analysis draft guidance relate to ISO 10993-18?
FDA’s draft guidance provides detailed recommendations on information gathering, extraction procedures, chemical analysis, and reporting. It is explicitly aligned with ISO 10993-18 concepts and aims to promote more consistent and reliable analytical chemistry studies.
How does ISO 10993-18 relate to EU MDR compliance?
The EU Medical Device Regulation (MDR) does not prescribe specific test standards but requires manufacturers to demonstrate biological safety as part of their clinical evaluation and technical documentation. ISO 10993-18 is the de facto standard accepted by EU Notified Bodies for chemical characterization. There must be evidence of its use in the BEP. The MDR adds additional requirements: if substances classified as CMR (carcinogenic, mutagenic, or toxic to reproduction), and/or endocrine-disrupting substances are present at ≥0.1% concentration in any device component, their presence must be justified. Any testing performed to older versions of ISO 10993-18 (pre-2020) will require a formal gap analysis for MDR submissions.
