Systemic toxicity and pyrogenicity answer different questions
Many companies writing a biological evaluation plan put "systemic toxicity" and "pyrogen" on one line of the test request form, on the grounds that both sit under GB/T 16886.11 (corresponding to ISO 10993-11). Grouping them is not wrong, but the two answer completely different questions, and stating the requirement as a single item is a common reason for a plan to come back at review.
Systemic toxicity asks whether substances that can be extracted from the device, once absorbed and distributed in the body, produce toxic effects in organs or systems remote from the point of contact. What is observed is the whole animal: body weight trend, clinical signs, food consumption, and where necessary haematology, clinical chemistry and histopathology. It does not ask whether the patch of skin under the device turned red, which is what irritation testing is for.
Pyrogenicity asks whether a rise in body temperature follows once the device or its extract enters the body. One qualifier here matters a great deal: material-mediated pyrogenicity. It is not the same thing as the bacterial endotoxin everyone is more familiar with, and the two do not even draw on the same standards system. More on that below.
Get this distinction clear and the decisions downstream all fall into line. Whether to test, which tier to run, which extraction media to use, how to prepare samples: all of it follows one thread, namely what can get into the body from this device and for how long.
When it gets required: reason back from the nature of contact
Under the evaluation framework established by ISO 10993-1, biological test items are not ticked off one by one. They are determined jointly by the nature of contact between device and body (surface contact, externally communicating, implant) and by the duration of contact (limited, prolonged, long-term). Systemic toxicity and pyrogenicity belong to the group that has to be considered seriously as soon as absorption into the systemic circulation is on the table.
Typical triggers in practice:
- Devices in direct contact with circulating blood. Blood is a very direct distribution route, and once extractables are released there is essentially no local retention to buffer them.
- Implants, especially those containing degradable or absorbable components, or residues of processing aids.
- Externally communicating devices that contact tissue fluid, cerebrospinal fluid or body cavities.
- Products whose contact area is large relative to body surface area. Even with a low release per unit area, the total may not be negligible.
A short piece of engineering reasoning here (this is analysis of mass transfer and exposure routes, not a statistical conclusion): systemic toxicity risk is essentially the product of three factors, namely the quantity of migratable substances in the material, the rate at which they migrate into the contacting medium, and the efficiency with which that medium transports them into the systemic circulation. If any one of the three approaches zero, overall risk is pushed very low. That is why the same polymer, made into a product with short-term contact with intact skin, usually does not require systemic toxicity, while the same polymer as a long-term implant must be evaluated: the material did not change, the second and third factors did.
Conversely, not every device involving absorption has to go to animal testing. The ISO 10993-1 framework allows conclusions to be supported by existing data, historical data on similar materials, and chemical characterisation combined with toxicological risk assessment. Plenty of companies overlook this route and start with animal work, which is both slow and expensive; others lean on it too heavily while failing to supply complete material information, and end up being asked to run the tests anyway. The test is plain enough: can you describe the extractables clearly and give a toxicological basis for each substance? If not, run the test.
Tiers of systemic toxicity: settle the selection logic, look the duration thresholds up in the standard
Systemic toxicity under GB/T 16886.11 is not one test but a group of tests tiered by exposure duration. The names are familiar enough: acute, sub-acute, sub-chronic, chronic. What actually goes wrong is tier selection. Pick too shallow and you will be asked to repeat the work; pick too deep and you waste money and animals.
Before the selection logic, one common misreading needs clearing away. The four tiers sound as though they are cut along fixed calendar durations, so plans often hard-code a number and match against it. That is unreliable. The duration boundaries of the tiers are expressed relative to the life span of the test animal, and different standard texts do not word the same tier identically. Memorising a set of durations away from the standard text makes it easy to pick the wrong tier, and once the tier is wrong the whole evidence chain in the plan has to be rebuilt.
So this section deals only with judgement logic. How long the exposure period is for a given tier, how many administrations are made, and what acceptance limits apply are governed in every case by the current valid version of the standard text. Confirm them with the laboratory during the evaluation planning phase, and with the reviewer where necessary. Do not reuse an old plan or a comparison table circulating online.
The table below sets out the positioning of each tier and the corresponding submission points. Durations and administration counts are deliberately omitted, leaving only relative characteristics and decision cues.
| Tier | Relative character of exposure | Main question answered | Typical trigger | Information usually missing at submission |
|---|---|---|---|---|
| Acute systemic toxicity | Single administration, or concentrated within a very short window | Are there immediate, dose-related toxic effects | Limited contact devices; routine screening-level evaluation | Rationale for the extraction medium; sterilisation state of the samples |
| Sub-acute systemic toxicity | Repeated administration over a relatively short period | Are there early signs of cumulative effects after repeated exposure | Reusable devices; products used over a course of treatment | Actual clinical frequency of use and duration per use |
| Sub-chronic systemic toxicity | Repeated administration over a markedly longer period | Where are the target organs, and are functional or structural changes detectable | Long-term contact devices; some implants | Material composition list; processing aids and residues |
| Chronic systemic toxicity | Sustained exposure over a long period | Cumulative consequences of long-term low-level exposure | Permanent contact implants | Degradation behaviour data; prior evaluation data on the same family |
In practice, ask yourself what order of magnitude the cumulative exposure of one patient to this device is. Used once and discarded, used repeatedly through a course of treatment, or left in the body permanently? Cumulative exposure sets the tier, not the duration of a single use. This is regularly miscalculated for reusable devices, where each contact is short but a full course of treatment adds up to something quite different.
Two further dimensions get overlooked. First, whether the route of administration and the contact medium match clinical use: within one tier, different routes can give different exposure levels and different target organs, so the plan has to justify the route chosen. Second, whether the product releases or degrades in a time-dependent way: for materials that release quickly and then slowly, a short tier and a long tier can show different things, and such products deserve to be considered one tier deeper.
One more note: tiers do not nest automatically, and running a deeper one does not cover the shallower ones by default. What reviewers look at is whether the plan matches the clinical use scenario, not whose period is longest.
Extraction largely decides whether the result is credible
In the laboratory, most rework on systemic toxicity and pyrogenicity happens not in the animal phase but in extraction. The sample preparation and extraction conditions specified in ISO 10993-12 are the input to the whole chain, and if the input is wrong nothing downstream matters, however well it is executed.
Traps that really do get stepped on:
One, only one extraction medium was used. The standards system requires both polar and non-polar media to be considered, so that extractables with different solubility characteristics are covered. Using only a polar medium amounts to assuming there is nothing lipophilic in the product, an assumption that does not hold for the great majority of polymer parts containing plasticisers, antioxidants or mould release agents. This is a frequent review question.
Two, the extraction ratio is not worked out properly. The standard gives extraction ratios based on device surface area or mass. The rule itself is not complicated; the difficulty is how the area is calculated. For porous materials, foams, textiles and coated complex constructions, geometric area and effective contact area differ widely. Our practice is to ask the client to state in writing the basis and the calculation for area or mass, which the laboratory then checks and records in the raw data. Effort spent here saves a repeat round later.
Three, unsterilised samples are submitted. Systemic toxicity and pyrogenicity evaluate the product in its marketed state. The sterilisation process itself may introduce residues and may change the material surface, so data generated from unsterilised parts is essentially not accepted during review. If the product has several sterilisation methods or several sterilisation batches, the basis for representativeness has to be stated. For the interface with that work, see sterilisation validation and residue testing.
Four, extraction vessels and utensils introduce interference. Non-polar media can leach from or adsorb onto certain labware, and pyrogen-related work additionally requires the labware itself to be non-pyrogenic. This is internal laboratory control, but a client who pre-treats samples before shipping easily breaks the chain. Send the product as it is and leave extraction to the laboratory.
Five, there is not enough sample. Systemic toxicity, pyrogenicity, possible repeat tests and retained samples together often exceed what companies estimate. The number of units is governed by the current valid version of the standard text, but a safe approach is to send the item list to the laboratory before placing the order and let the laboratory work the quantity back from it, instead of estimating it yourself. For sample quantity and state requirements, start with the sample submission requirements.
Pyrogenicity: material-mediated pyrogens and bacterial endotoxin are two different things
This deserves its own section, because the confusion it causes is common.
Material-mediated pyrogenicity measures any substance in the device extract capable of causing a rise in body temperature. It may come from the material composition itself, from monomers, additives or degradation products, or from process residues. It is a whole-response test that asks about effect, not origin.
Bacterial endotoxin measures one specific class of substance, the lipopolysaccharide of the Gram-negative bacterial cell wall, usually by in vitro methods such as the gel-clot method or photometric methods.
The boundary between the standards has to be stated plainly, because this is a common assumption on test request forms. Test methods for material-mediated pyrogens fall within the scope of GB/T 16886.11 (ISO 10993-11). Methods and limits for bacterial endotoxin are neither in that part nor in any other standard listed in this article. They belong to the pharmacopoeial system, which is a separate source of methodology and acceptance criteria.
That has two consequences. First, writing GB/T 16886.11 on the request form does not mean endotoxin is covered along with it; if the biological evaluation report needs endotoxin as evidence, it has to be raised as a separate item with its own cited reference document. Second, the reference document for endotoxin has to be confirmed against the actual situation of the product: which pharmacopoeia applies, which jurisdiction's requirements are being followed, and which edition is in force, all determined by the current valid version and by the submission route applicable to the product. Do not carry the citation over from an old plan or from a peer's plan. This matters especially for projects submitted in several jurisdictions at once, where the endotoxin basis for the same product may not point to the same document; the current requirements published officially by each jurisdiction govern.
The relationship between the two can be understood as follows (the following is analysis based on detection principles, not a statistical conclusion): endotoxin is a subset of pyrogens. A passing endotoxin result shows only that this class of substance is under control, and it does not imply that the material itself is non-pyrogenic. Conversely, a positive material-mediated pyrogen result cannot be attributed to endotoxin directly; you still have to go back to material and process to find the cause.
Three practical suggestions:
- Do not use endotoxin results in place of the material-mediated pyrogenicity evaluation. The two carry different argumentative functions in a biological evaluation report and rest on different standards systems, so list them on separate lines in the plan, each with its own cited source.
- Endotoxin control is largely a manufacturing and release matter, covering environment, water system, raw materials, cleaning process and packaging. It varies from batch to batch and cannot be held in check by a single type test.
- If material-mediated pyrogenicity comes back positive, investigate process contamination first and suspect the material second. Experience says process-side causes, such as incomplete cleaning, the water system, or storage and transport, are more common than the material itself. Following that order shortens the investigation noticeably.
A practical pre-submission list: state these things clearly
For biological evaluation orders, information gaps are more common than sample gaps. The table below puts common practice, likely consequence and recommended practice side by side.
| Aspect | Common practice | Likely consequence | Recommended practice |
|---|---|---|---|
| Description of contact | Only the product name and category | The laboratory cannot confirm the tier, so the plan goes through repeated rounds of confirmation | State the contact site, contact medium, and both single and cumulative contact duration |
| Material information | Only "medical grade polymer" | No basis for choosing extraction media or scoping the risk assessment | Provide grade-level composition, additives and processing aid lists |
| Sterilisation information | Sterilisation method not stated | Sample state differs from the marketed state | State the sterilisation method, whether the sample is finished product, and batch representativeness |
| Existing data | No historical material supplied | Items that already have conclusions are repeated | Provide prior evaluation reports for the same family, literature and chemical characterisation data |
| Area or mass calculation | Left to the laboratory to estimate | Extraction ratio open to question, weak reproducibility | Provide the basis and the calculation, confirmed in writing by both sides |
| Sample quantity | Estimated item by item | Additional samples mid-project and a longer schedule | Work the quantity back from the complete evaluation plan and prepare it in one go |
| Basis for pyrogenicity | Just "do pyrogen" | Material-mediated pyrogens and endotoxin get conflated | Write separate lines stating which one is being tested, each with its source |
| Evaluation plan | Tests first, plan written afterwards | No chain of reasoning behind item selection | Issue the evaluation plan under the ISO 10993-1 framework first, then schedule testing |
On scheduling, one pragmatic reminder: the longer-period tiers of systemic toxicity are among the heaviest items in the whole biological evaluation timeline. If your product needs the sub-chronic or chronic level, start that item first in the registration plan rather than scheduling it after everything else is finished. For scheduling and sample flow, see the testing process.
After the result: an unfavourable finding does not mean the product is unusable
The first reaction to a report with an abnormal finding is usually "that is it, we have to change the material". Not so fast.
The logic of biological evaluation is always risk evaluation, not single-item pass or fail. In the framework established by the GB/T 16886 series (ISO 10993 series), the result of an individual test is evidence, not a conclusion. When something abnormal appears, the sensible order of work is as follows.
Step one, confirm that the test itself holds up. Did the control group behave normally, was extraction performed correctly, were the animals affected by factors outside the test, are the observation records complete? A fair proportion of abnormal findings are traced back to an operational or sample state issue during this review stage alone.
Step two, locate the source. Is it an inherent constituent of the material or a process residue? The more effective tool here is usually not repeating the animal test but chemical characterisation: work out what is actually in the extract first, then judge which constituent may be responsible.
Step three, assess clinical relevance. Test conditions are normally more severe than actual clinical use in extraction ratio, route of administration and mode of exposure. Whether the abnormal finding has real meaning under clinical conditions has to be argued from exposure amount and exposure route, and written into the evaluation report.
Step four, and only then, change the material or the process. And once changed, you have to reconfirm whether the change affects the validity of other items already completed, which is a step that gets missed regularly.
How this connects with the other biological items
Systemic toxicity and pyrogenicity are not isolated items. In the evaluation system they sit downstream: upstream are material information and chemical characterisation; alongside them are cytotoxicity, sensitisation and irritation; downstream is the overall biological evaluation conclusion.
A word on how standards should be cited. When referring to the evaluation system as a whole, writing GB/T 16886 series or ISO 10993 series is appropriate. But as soon as you introduce a specific test or a specific requirement, the citation should go down to the part number: systemic toxicity and pyrogenicity to GB/T 16886.11, sample preparation and extraction conditions to ISO 10993-12, evaluation pathway and item selection to ISO 10993-1. A plan that cites only the series number without the part number will almost certainly attract a request to state which part the requirement comes from; and as for which edition of each part applies, that is governed in every case by the current valid version of the standard text.
The efficient sequence is to get material composition and chemical characterisation solid first, use them to screen out unnecessary animal tests and to define the points of concern for the tests that must be run, then order the items by nature of contact, and finally close everything out in the evaluation report. Doing it the other way round, running a spread of animal tests first and going back for material information afterwards, is not only expensive but also inconsistent with the generally accepted principle of reducing animal use.
For a fuller picture of item selection and the evaluation system, see biocompatibility testing.
What SUNGO Lab can do
SUNGO Lab operates laboratories in Shanghai and Hefei and can take on testing and technical support related to biocompatibility evaluation of medical devices, including working through the evaluation plan, fixing sample preparation and extraction conditions, pre-reviewing the completeness of submission documents, and coordinating the schedule with related work such as sterilisation residues and chemical characterisation. We are accredited by CNAS, CMA and IAS (USA). Please note that an accreditation mark only demonstrates that the laboratory is technically competent within its accredited scope and does not constitute a commitment regarding market access outcomes; registration and submission results still depend on the regulator's review of the complete technical file. Our wider capability is listed under testing services.
If you are unsure whether a product needs systemic toxicity at all, or which tier to run, or you have been asked to add work and cannot see where the gap is, bring the nature of contact, the material information and the existing data and talk to our engineers directly. All test parameters, tier durations and acceptance limits referred to in this article are governed by the current valid version of the standard text; methods and limits for bacterial endotoxin fall outside the standards listed here and have to be confirmed separately against the applicable pharmacopoeial system and the requirements of the jurisdiction of registration.
Call +86 132 4819 8029, or request a quote.