The conclusion: extraction conditions are derived, not copied
Choose the wrong extraction conditions and the report can be entirely compliant in itself yet still be unable to support your product. One line in a review — extraction conditions not applicable to this device — and the whole report is wasted. Worse, this kind of write-off is usually discovered late, and by the time the work is repeated, samples, laboratory slots and the registration milestone all have to be rearranged.
Only three inputs decide extraction conditions: the contact nature and duration of the final product, how much the material itself can tolerate, and what question this particular test is meant to answer. Any approach that bypasses those three and lifts somebody else's plan wholesale carries high risk. The GB/T 16886 and ISO 10993 series provide a selection framework and a set of principles, not a universal recipe card.
| Decision axis | Your situation | How to handle it |
|---|---|---|
| Purpose of the test | A biological endpoint test, cytotoxicity for example | The medium has to obey the requirements of the test system itself, so that the cells or biological system remain viable |
| Purpose of the test | Chemical characterization screening | The media have to cover a range of polarities, and conditions lean towards the exaggerated end |
| Nature of contact | Direct contact with blood or body fluid | Medium and ratio have to stay close to the actual mode of contact, with simulated fluids where necessary |
| Material tolerance | Heat-sensitive, prone to hydrolysis or degradation | Conditions must not exceed the stability range of the material, or you manufacture degradation products that would never occur clinically |
| Product form | Porous, absorbent, irregular geometry | Take the ratio by mass rather than by surface area, and write the reasoning into the test plan |
1. Media: how much polarity you cover decides what you can see
The logic for choosing media differs between endpoint tests and characterization work, and this is where people mix things up from the outset.
On the cytotoxicity side, the ISO 10993-5 system places hard constraints on the medium: serum-containing culture medium extracts hydrophobic substances more effectively but also brings more background interference, while polar media such as physiological saline mainly pull out water-soluble substances. Which one you pick depends on the mode of contact and on the class of leachables you want covered — subject to one condition, that you do not wreck the cell system in the pursuit of more extraction. Once the system breaks down, the results cannot be interpreted.
On the chemical characterization side the goal is the opposite: bring out as many potential leachables as possible, so the media have to span polar, mid-polarity and non-polar. Running a single medium and drawing a conclusion from it is a routine reason for being asked to repeat the work.
The rule of thumb comes in three lines:
- For endpoint tests, the medium serves the test system.
- For characterization, the medium serves coverage, and it is better to add one more than to run one too few.
- Where the clinical contact is with an unusual fluid — alcohol-containing preparations, lipid vehicles and the like — consider a simulated fluid separately; a conventional medium is not a substitute.
2. Ratio: surface area method or mass method
| Product form | Method | Watch out for |
|---|---|---|
| Regular, smooth-surfaced solid parts with measurable area | Surface area method | Films and sheets contacted on both faces have to be counted on both sides |
| Porous, foamed, woven or absorbent materials | Mass method | Effective surface area simply cannot be measured accurately, and forcing a conversion clearly underestimates it |
| Assemblies made from several materials | Component-level or whole-device evaluation, depending on how the evaluation subject is defined | Extract everything together and a problem in one material gets diluted away |
| Very thin or very fine products | The basis for the chosen method has to be stated in the test plan | When the area-to-mass relationship is unusual, the two methods give markedly different results |
There is a practical trap here: absorbent materials soak up part of the medium, so the volume actually taking part in extraction does not match the design value, the recovered extract falls short, and the tests you lined up afterwards cannot all be completed. With materials like this, raise the sample quantity in advance. This is not a question of being conservative or not; it is a question of having enough liquid. Additional samples requested because the first shipment was too small often mean waiting for the next production batch, which costs far more than shipping a few extra units would have.
3. Conditions: exaggerated versus simulated use, and why the wrong direction fails in opposite ways
The two condition families serve different purposes. Exaggerated conditions pair a higher temperature with a shorter duration, or a lower temperature with a longer duration, in order to pull out as many potential leachables as possible for screening and risk identification. Simulated use conditions stay close to the clinical mode and duration of contact, and answer the question of roughly how much really comes out.
The consequences of choosing the wrong direction are symmetrical:
- Use simulated use conditions where exaggerated conditions were needed, and you under-detect: risk identification is incomplete and you get asked to repeat the work.
- Use exaggerated conditions where simulation was needed, and you detect a pile of substances that would never appear clinically, each of which then needs an exposure justification. The workload goes up, not down.
There is also a hard line: extraction conditions must not push past what the material can tolerate. If the sample melts, softens or deforms noticeably, degrades appreciably or changes colour abnormally during extraction, the conditions are no longer applicable, and the results represent neither the clinical situation nor a usable basis for evaluation. The correct handling is to declare the conditions inapplicable on the spot and go back for reconfirmation, rather than pressing on and sorting it out later. Ask about this during test plan discussion: whether your material is stable under the proposed conditions is something you know better than the laboratory does.
4. Sample state: if it is not the final product, none of it counts
Extraction has to be performed on the final product state, and three conditions all apply:
- Sterilized by the actual production process. Change the sterilization method and you change what can be extracted; related preparation is covered under sterilization validation services.
- Presented in the actual packaging. Migrants from packaging materials are part of the contact path; related verification is covered under packaging validation services.
- If the evaluation has to cover the end of shelf life, aged samples are also required, otherwise the data represent the as-manufactured state only.
Of those three, sending unsterilized semi-finished parts is by far the most frequent. The reasoning is usually that the sterilization line has no slot yet, so testing goes first to save time. The time saved comes back doubled.
5. Five situations that void a report outright
| Situation | Direct consequence | What to do instead |
|---|---|---|
| Extraction performed on unsterilized samples | Results do not represent the marketed form | Submit final product state samples; where that is genuinely impractical, supply evidence that sterilization introduces no change |
| Extraction ratio chosen by habit, with no basis in the test plan | No way to show the extraction volume relates to real device contact | Choose the method from the product form and state the reasoning in the plan |
| Chemical characterization with a single medium | Polarity coverage is incomplete and whole classes of leachables are missed | Run several media in parallel to cover different polarities |
| Sample deforms or degrades during extraction | Conditions inapplicable, results unusable | Reduce the severity or move to simulated use conditions, and record how the call was made |
| Extract stored too long before analysis, or held in unsuitable containers | Unstable substances are lost, containers adsorb or leach and interfere | Set time limits for storage and analysis, and run a labware blank to establish background |
These share a common trait: nothing looks wrong when the report comes out, and the data may look clean. The problem is whether that data represents your product at all. Which is why the test plan review matters considerably more than the laboratory work.
Pre-submission self-check
- Is the evaluation subject clearly defined: whole device, individual component, or grouped by material?
- Has the contact site and contact duration been written down for every component under test?
- Are the samples in final product state: sterilized, packaged, and aged where necessary?
- Is the basis for the extraction ratio method — surface area or mass — recorded in the test plan?
- Has material stability under the proposed conditions been confirmed?
- Has the sample quantity been increased to allow for absorption, losses and retest reserve?
- Does the plan include blanks and a labware background design?
Every line on this list costs essentially nothing to confirm at planning stage. Once samples have shipped and work has started, any one of them failing means new samples and a new slot in the queue.
Extraction is execution; the question above it is the evaluation route
Worth flagging: extraction conditions are an execution-level decision. The question above it is which evaluation route this device should follow, which items can be closed out on paper, and which have to be measured. That is a separate body of logic, covered in biocompatibility testing services and in the material under testing knowledge. The relationship between the two is straightforward: the route decides which items get run, and the extraction conditions decide whether the results of those items can be used.
One note on report acceptance: an accreditation mark only demonstrates that the laboratory holds the relevant technical competence within its accredited scope; it is not a commitment regarding market access in the target market, and the applicable acceptance position remains subject to review by the receiving authority. Where a standard is cited, the currently effective version of the standard text governs. Writing the reasoning behind your extraction conditions clearly into the test plan does more to cut down review cycles than any mark does.
Spending a little time on the plan beats repeating the work
If your product form is on the awkward side — porous, absorbent, a combination of materials, or clinically in contact with a non-standard fluid — walk through the test plan before shipping samples. Send us the product construction, the material list, the sterilization method and the actual mode of contact, and we can settle the reasoning for the extraction method, media and conditions first, then schedule sample submission, so you do not finish the work only to find the conditions were inapplicable. Call +86 132 4819 8029, or simply request a quote.