How to Choose the Right Implant Test Site: A Practical Guide to ISO 10993-6:2026

Choosing the wrong implant test site is one of the more avoidable mistakes in biocompatibility study design and one of the more consequential ones. If your test doesn't reflect how the device actually contacts tissue in clinical use, your data may not support the conclusions you need, and FDA reviewers will notice.

Under ISO 10993-6:2026, the test sample must be implanted into or onto the tissues most relevant to the intended clinical use of the material. The choice of tissue, implantation site, and study duration must all be documented and justified in your study documentation. 

Here's what that means in practice and where manufacturers and CROs most commonly get it wrong.

Why the Implant Site Has to Match Clinical Use

The principle behind implantation testing is comparison. ISO 10993-6:2026 evaluates local biological effects by comparing the tissue response to your test sample against the response to a control — a material with established clinical acceptability. That comparison only means something if the tissue environment is appropriate for both.

Different tissues respond differently to the same material. Subcutaneous tissue, muscle, bone, and neural tissue each have distinct cellular environments, vascularization, immune cell populations, and healing dynamics. A material that produces an acceptable response in subcutaneous tissue may behave differently in bone or neural tissue. Running the wrong study doesn't just waste resources, it leaves you with data that can't actually support your biocompatibility conclusions.

The justification for your implantation site choice must be recorded in the study documentation. If you choose a site that doesn't align with how the device is used clinically, the standard requires a specific scientific justification. "It's easier to implant in subcutaneous tissue" is not that justification.

The Four Normative Implantation Test Sites in ISO 10993-6:2026

The standard includes four normative annexes covering specific implantation methods. Each annex applies to a different tissue type and carries its own requirements for sample dimensions, animal model selection, number of implants, and surgical procedure.

Subcutaneous Tissue (Annex A)

Subcutaneous implantation is appropriate for devices implanted into subcutaneous tissue or those with long-term contact to mucosal membranes. It's one of the most commonly used sites because it's relatively accessible and well-characterized.

When using discs, samples should be 10 mm to 12 mm in diameter and 0.3 mm to 1.0 mm thick. Rod and cylinder samples should be 1.5 mm to 2 mm in diameter and 5 mm to 10 mm in length with rounded ends. Non-solid materials including powders, gels, liquids, and particulates can be implanted directly into subcutaneous tissue, with a location marker recommended for absorbable materials so the site can be identified at retrieval.

A minimum of three animals per material should be used, with sufficient implant sites to yield 10 test samples and 10 control samples for each material and each implantation time point.

Subcutaneous tissue is also useful for comparing different surface textures or conditions of the same material, or assessing the effect of treatments or surface modifications.

Muscle (Annex B)

Muscle implantation is appropriate when the device has direct contact with muscle tissue in its intended clinical application. The paravertebral muscles of rabbits are the primary recommended site, though gluteal muscles in rats or the quadriceps femoris or gluteal muscles in rabbits may also be used for smaller samples.

For rabbit paravertebral muscle, implants of 1 mm to 3 mm in width and approximately 10 mm in length are typically used. Larger samples up to 10 mm in diameter and 3 mm thickness can also be surgically implanted. All samples should have rounded edges.

Implantation in muscle is performed by hypodermic needle or trocar, with larger implants requiring appropriate surgical techniques. When the trocar method is used, ISO 10993-6:2026 recommends verifying complete ejection and appropriate placement of test and control articles, or adding additional implant sites to account for potential misplacement — a detail that's easy to overlook in study planning.

Like subcutaneous testing, muscle implantation requires at least three animals per material and 10 test and 10 control samples per time point.

Bone (Annex C)

Bone implantation applies to devices intended to contact bone — orthopedic implants, dental implants, bone screws, and related materials. The standard allows implantation in cancellous (trabecular) or dense compact bone, with the site selection dependent on the clinical use of the device.

Sample dimensions in ISO 10993-6:2026 are specific to both the animal model and the bone location. For cancellous bone in rabbit lateral condyles, cylindrical implants of 2 mm to 6 mm in diameter and 6 mm to 10 mm in length are typical. Larger animals — dogs, sheep, goats — use larger dimensions. Screw-shaped implants are preferred when feasible because they provide initial stability; cylindrical implants may be used when screw preparation is not practical.

The standard specifies that the femur and tibia are typically used, with implant sites contralateral between test and control samples. A maximum of six implant sites per rabbit and twelve per dog, sheep, goat, or pig is specified.

One important practical point: ISO 10993-6:2026 specifies that bone drilling should be performed at low speed with intermittent drilling and profuse irrigation with physiological saline. Overheating the bone causes local tissue necrosis, which confounds the tissue response evaluation.

Bone implantation studies also require skeletally mature animals in most cases, with the note that bone quality can vary between non-purpose-bred animals of the same species. Bone densitometry can be needed to identify suitable test animals.

Neural Tissue (Annex D)

Neural implantation applies to devices intended to contact brain tissue, spinal cord, dura, or cerebrospinal fluid — electrodes, shunts for hydrocephalus, and similar devices. This is the most demanding of the four normative test sites.

One important scope clarification in ISO 10993-6:2026: neuro-interventional devices that contact the vessel wall but not neural tissue directly are evaluated under ISO 10993-4, not this standard.

Sample sizes are small, typically rod or wedge-shaped implants of 1 mm x 1 mm or less in cross-section and 2 mm to 6 mm in length, or disc-shaped implants of 8 mm diameter depending on the configuration and number of sites.

A critical requirement that differs from the other annexes: test and control samples shall not be implanted in the same animal for neural tissue studies. This reflects the sensitivity of the tissue and the potential for cross-site confounding.

Beyond the Four Sites: When Clinical Use Requires Something Different

ISO 10993-6:2026 acknowledges that not every device falls neatly into the four normative annexes. If an implantation site other than subcutaneous tissue, muscle, bone, or neural tissue is clinically appropriate, the standard permits alternative sites but requires that the general scientific principles from Annexes A through D still be followed, and a written justification must be provided explaining why the chosen anatomical location doesn't align with the normative sites.

For some devices, vertical standards prescribe specific implant study designs. Intraocular lens implants are covered by ISO 11979-5. Dental devices fall under ISO 7405. Studies designed to satisfy those device-specific standards can also satisfy ISO 10993-6:2026 requirements.

Additionally, Annex E provides guidance for devices contacting peripheral nerve tissue, which doesn't have a normative annex but presents unique considerations around surgical technique, neurophysiology, and interpretation of results.

Test Periods: Site-Dependent and Clinical Use-Driven

Implantation time points aren't arbitrary. The standard requires that the test period be determined by the likely clinical exposure time or continued until a steady-state biological response has been reached — whichever requires longer study duration.

One nuance that affects bone implantation specifically: ISO 10993-6:2026 notes that bone tissue can need longer observation periods before steady-state is reached than muscle or connective tissue. In muscle and connective tissue, depending on the species, device design, and severity of surgical trauma, a steady-state can take 9 to 12 weeks. For bone, observation periods are typically longer.

For non-absorbable materials, short-term responses are normally assessed from 1 to 4 weeks, with long-term responses evaluated in studies exceeding 12 weeks. The standard includes a table of example test periods for common species including mice, rats, guinea pigs, rabbits, dogs, sheep, goats, and pigs, with time points ranging from 13 weeks out to 104 weeks depending on species and study objectives.

For absorbable materials, the test period framework is more complex and must reflect the degradation profile of the specific material including early, mid, and late-stage degradation time points.

Common Study Design Mistakes in Implant Site Selection

A few issues come up regularly when manufacturers and CROs are planning implantation studies:

Using subcutaneous tissue as a default when clinical use involves other tissue types.  Subcutaneous implantation is well-characterized and relatively convenient, which makes it tempting to use even when the device contacts muscle, bone, or neural tissue. ISO 10993-6:2026 is clear that clinical use drives site selection, not convenience.

Insufficient justification for site deviations. If you're using a non-normative site, the justification needs to be substantive and scientifically grounded. Regulatory reviewers will look for it.

Not accounting for tissue-specific steady-state timing in study duration. Setting a study duration based on general industry practice rather than the specific tissue type and device design can result in data collected before a steady-state is reached, leaving you with ambiguous results and the potential need to repeat the study.

Mismatching control geometry to test article geometry. The 2026 edition reinforces that comparative control samples should be matched as closely as reasonably possible for physical properties and geometrical characteristics. A significant mismatch between your test article and control article shapes means you're not comparing equivalent tissue environments.

What This Means for Medical Device Consulting

Study design decisions made early in device development are difficult and expensive to correct later. Choosing the right implant site, the right control, the right time points, and the right animal model before the study begins — and documenting the rationale clearly — determines whether the data you generate will actually support your biocompatibility conclusions in a regulatory submission.

If the study is poorly designed, the data may not answer the right questions regardless of how well it's executed.

Expert Implantation Study Strategy and Biocompatibility Consulting for Medical Devices

Getting implantation study design right from the start requires deep knowledge of ISO 10993-6:2026 requirements, regulatory expectations, and how study data will be interpreted during FDA review.

At JL Tox Consulting, we help medical device manufacturers and CROs develop implantation study designs that reflect clinical use, satisfy ISO 10993-6:2026 requirements, and support defensible biocompatibility conclusions.

Our implantation testing and biocompatibility consulting services include:

• Biological evaluation planning that determines whether implantation testing is required under ISO 10993-1:2025's risk-based framework

• Implantation study design including test site selection, control selection, sample preparation strategy, and time point justification

• Study protocol review assessing scientific rigor and alignment with ISO 10993-6:2026 before studies begin

• Toxicological risk assessment to complement implantation data and support complete biocompatibility conclusions

• FDA submission support integrating implantation data into biocompatibility evaluation reports for 510(k), PMA, and De Novo applications

• Deficiency response preparation addressing FDA questions about implantation study design and interpretation

With over a decade of specialized experience in medical device biocompatibility and FDA regulatory submissions, Dr. James Lyons and the JL Tox team provide the toxicological and regulatory expertise needed to build implantation testing strategies that hold up under FDA review.

Contact JL Tox Consulting to develop an implantation study strategy that supports your regulatory submission:

Email: info@JLTox.com

Phone: (877) 899-6568