Organ‑on‑a‑Chip and 3D Tissue Models: The Future of Biocompatibility Testing

Organ‑on‑a‑chip and 3D tissue models add human‑relevant evidence to a risk‑based ISO 10993 plan. Today, in‑vitro irritation is recognized under ISO 10993‑23; most other endpoints are supportive NAMs that complement (not replace) core testing. FDA’s current ISO 10993‑1 guidance and its NAM roadmap set the tone: use science‑based alternatives where validated, and keep the plan risk‑justified. 

What these models are and why they matter

They’re microphysiological systems that mimic human tissues under flow and mechanical cues. They help you see human biology early, reduce animal use, and focus resources on real risks. ASTM even standardized the shared vocabulary for these systems, which helps reviewers align on terms.

A common question pops up: “Are chips ‘regulatory‑grade’ yet for devices?” The short answer is that they can be, where a standard or clear context of use exists, and they’re persuasive as supporting evidence elsewhere under a risk‑based plan. FDA’s new roadmap to reduce animal testing is explicit about advancing NAMs like organ‑on‑chips.

If you’re mapping endpoints, start with a clear matrix using biocompatibility testing for medical devices, so contact and duration drive the plan.

Where they are accepted today

The fastest path is in vitro irritation using reconstructed epidermis per ISO 10993‑23. FDA recognizes 10993‑23, so when your device qualifies, this can replace animal irritation. 

For other endpoints, chips and 3D tissues are strong supporting tools. Use them to: 

• clarify mechanisms
• rank risks
• set smarter follow‑ups

FDA’s position is to welcome validated alternatives inside a risk‑managed ISO 10993‑1 framework.

When irritation is in scope for your device, align your approach with in‑vitro irritation testing under ISO 10993‑23.

How chips fit into an ISO 10993 plan

Keep ISO 10993 as the backbone. 

Run ISO 10993‑18 chemical characterization first, so chemistry defines what biology must answer. Then build a Biological Evaluation Plan (BEP) that shows where you’ll use recognized in‑vitro irritation, where you’ll keep classical testing, and where chips add human‑relevant context.

Close the loop with a toxicological evaluation under ISO 10993‑17 and a final Biological Evaluation Report (BER) that connects chemistry → tox → biology → performance. 

FDA’s 2023 biocompatibility guidance reinforces this risk‑based, evidence‑linked story.

Hemocompatibility under flow

Engineers often ask: “Will a microfluidic flow loop replace ISO 10993‑4?” Not today. Vessel‑on‑a‑chip models under physiologic shear are great for insight into platelet activation and thrombosis, but they supplement rather than replace the hemolysis/coagulation/platelet/complement panel defined in 10993‑4. Use them to explain edge cases and refine decisions.

When blood contact is part of use, anchor your file with hemocompatibility testing per ISO 10993‑4 and bring chip data as mechanism‑level support.

Systemic and chronic effects

Can liver‑on‑a‑chip data shrink systemic tox? 

Liver chips can reveal metabolism or low‑level liabilities and help justify scope, but they don’t replace repeat‑dose systemic toxicity for devices at this time. They’re best used to focus risk, refine materials, and support your justification narrative.

FDA’s ISTAND and NAM initiatives confirm the direction of travel, especially for DILI in drugs. Use that momentum as a supportive context in device files. 

If local tissue response matters for your implant, plan subcutaneous implantation studies under ISO 10993‑6 and cite chip data to explain the mechanisms you observe histologically.

Study designs that work

Pick studies that answer a real question, fast. Pair standardized tests with focused chip studies that match your risk.

• irritation with RhE using polar and non‑polar extracts
• vessel‑on‑chip shear screen, then classic hemocompatibility endpoints
• liver chip metabolism screen to guide systemic tox scope
• chip co‑cultures to flag inflammatory signaling before implantation

For standards context across the test menu, this primer on ASTM tests in medical device biocompatibility is a helpful refresher.

Standardization is moving fast

Europe released an Organ‑on‑Chip standardization roadmap through CEN‑CENELEC and the European Commission’s JRC. It highlights shared terminology, validation needs, and links to a new ISO subcommittee for microphysiological systems. That progress will make data more comparable and reviews smoother. ASTM F3570 is already locked in common terms.

Practical tips you can act on

• write the BEP early and list where chips reduce uncertainty
• pre‑align extraction and media with your irritation model
• set shear ranges that mirror real use before vessel‑on‑chip runs
• show how chip data changed your test scope, not just that you ran it

Use chips to answer a risk, not to add noise. Start small. Tie every result back to your clinical scenario. FDA reviewers care most about that link.

FAQs

Can ISO 10993‑23 let us skip rabbit irritation in a 510(k)?

Yes, if your extraction and method fit the standard and recognition notes. Check FDA’s SIS first; as of July 28, 2025, the in‑vitro irritation clause in ISO 10993‑23 is not recognized, so most files still rely on recognized in‑vivo irritation or a risk‑based alternative.

Do organ‑on‑a‑chip or 3D tissues replace standard ISO 10993 endpoints today?

Not broadly. FDA’s roadmap encourages NAMs, but use is case‑by‑case under a risk‑based ISO 10993‑1 plan.

Can a vessel‑on‑a‑chip study replace ISO 10993‑4 hemocompatibility?

No. It can support your story, but 10993‑4 remains the anchor for blood‑contact devices.

Can liver‑on‑a‑chip shrink my systemic toxicity program?

It can inform scope and help explain risk, but it does not replace repeat‑dose systemic tox for devices at this time.

How should I document chip data in my submission?

Tie each result to a specific risk and show how it changed your plan, then connect chemistry → tox → biology in a risk‑based narrative per ISO 10993‑1.

What extraction setup should I use for irritation models?

Follow ISO 10993‑12 for sample prep and extraction, and align details with ISO 10993‑23 and the FDA’s recognition notes before you test.

Work with NABI

If you want a clean path, we can blueprint the plan, run the right in‑vitro irritation, build your chem‑char and tox evaluation, execute the hemocompatibility and implantation work where needed, and wrap it with a BER that reads like a story, not a stack of reports.

Phone: +1 407‑278‑6815 

Email: contact@nabi.bio