HONEST LIMITS

Where chips still fall short

🎲

Cell lines, protocols, and labs differ, so results do not always match

📉

TEER in many models is still below that in the body

🧩

Many chips lack neurons, immune cells, or the brain's own fluid dynamics

🧽

PDMS can absorb some drugs, which skews the numbers

Knowing the limits is part of using the tool well.

A SCORECARD

Four models side by side

Chips combine human cells and flow at moderate throughput. They do not capture whole-body effects, which only an animal or human can.

Each model answers different questions, so they work best as a team.

VALIDATION

Benchmarks Test drugs with known behavior in people and check the ranking

Comparison Match results with brain-to-plasma ratios of unbound drug, written K_p,uu, from animal or clinical data

Reproducibility Repeat across batches, operators, and labs

Standards Agree on how to build and report so results can be compared

A model earns trust by being right about things we already know.

WHAT'S NEXT

🔗

Linking a barrier chip to gut or liver chips to follow a drug through the body

🤖

Higher throughput and standard formats that fit lab robots

🏛️

In 2025 the U.S. FDA announced a plan to reduce animal testing, starting with some antibody drugs, and to encourage new methods such as organ-chips

The tools are maturing, and the standards are catching up.

BEFORE YOU BELIEVE IT

Which cells? Primary, immortalized, or stem-cell derived, and which support cells

What numbers? TEER and tracer leakage, and how they were measured

Which benchmarks? Known drugs that cross and drugs that do not

How often? How many batches and labs got the same result

A chip is only as good as its validation, and you now know the questions to ask.

UNIT 17 STUDY COMPLETE

🧭

You've finished the course. You can now weigh a barrier-chip claim by asking how it was built and checked.