For decades, the pharmaceutical industry has accepted a fundamental compromise in biologics quality control: medicines developed for humans are often functionally characterized using biological systems that do not fully reproduce human physiology.

Erythropoietin, or EPO, provides an excellent example.

Traditional EPO potency testing has relied on polycythaemic or normocythaemic mouse bioassays. Such assays measure a biologically meaningful erythropoietic response within an intact organism, but they can also be lengthy, costly and biologically variable. The DART EPO case study was developed specifically in response to these limitations and to the broader 3Rs objective of replacing, reducing and refining animal use in batch testing requirements of the old product.

The first transition: from mouse to cell line

The industry has already demonstrated that animal use is not the only way to measure EPO activity.

Human erythroleukemia-derived cell lines such as TF-1 and UT-7 respond to EPO and have been developed into proliferation-based potency assays. Published studies have shown these assays can provide reproducible and sensitive measurements of recombinant EPO activity, and a 2026 study from India’s National Institute of Biologicals reported successful validation of a UT-7-based method against the existing in vivo assay.

That represents important progress.

But replacing a mouse with a transformed cell line does not necessarily complete the journey toward human relevance.

A transformed cell line is optimized for robustness, reproducibility and assay simplicity. Those are valuable QC characteristics. Yet it represents a comparatively narrow biological response, typically proliferation or reporter activation, in an immortalized cellular background.

The next transition: from a cellular signal to human physiology

DART approaches the problem differently.

Rather than asking only whether EPO activates a receptor-dependent response in a transformed cell, DART uses a characterized human hematopoietic in vitro system derived from pooled human umbilical cord blood.

EPO-induced biological activity is evaluated through the generation of human reticulocytes, with culture micrographs analyzed using an AI/ML-trained prediction model.

In the reported case study, DART-generated potency values were bridged against 15 mouse potency values, creating an initial correlation between the established animal assay and the human-relevant platform.

This creates an important conceptual distinction:
Mouse assay: systemic biology, but non-human.
Transformed cell-line assay: human-derived and standardized, but reductionist.
DART: human-derived multicellular hematopoietic biology coupled with quantitative AI analytics.

Why the timing matters

Regulatory science is now moving rapidly toward New Approach Methodologies.

FDA describes NAMs as encompassing human-based in vitro systems, microphysiological platforms and computational approaches capable of supporting safety, efficacy and quality assessment. Its March 2026 draft guidance introduced a validation framework for using NAM data in drug-development submissions.

In September 2026, FDA further clarified through a regulatory update that appropriate non-animal methods may be used in testing drugs and biological products.

This does not mean that every validated animal bioassay can simply be discarded. Regulatory replacement requires fit-for-purpose validation, reproducibility, comparability, predefined acceptance criteria and evidence that the new method measures the biological activity relevant to product quality.

But the direction is becoming clear.

The future of biologics potency testing may not be defined by finding a more convenient substitute for the mouse.

It may be defined by something much more consequential:
measuring the medicine directly against relevant human biology.
That is the opportunity DART is being designed to address.

From animal bioassays, to transformed cell lines, to predictive human biology — potency testing itself is ready to evolve.