Champions Oncology gives cancer drugs a rehearsal

Before Champions Oncology kept a bank of human tumors, its corporate predecessor had a rather different collection: sports-bar rights. The restaurant concept went to Marriott in 1997. Biotechnology arrived in 2007. Somewhere between those two businesses sits an unexpectedly useful question: what, exactly, is worth keeping?

For a cancer researcher, a tumor sample is valuable. A sample accompanied by its molecular profile, treatment history and experimental responses is more useful still. Champions has built its business around keeping those pieces together. Its archive can become another experiment.

The tumor comes with a history

The company’s foundational technique is the patient-derived xenograft, or PDX. Researchers implant human tumor tissue into immunodeficient mice, expand it and compare how treated tumors grow against controls. Champions calls its platform TumorGraft. The practical aim is to learn which experimental therapies deserve further investigation.

The interesting detail is the annotation. Two tumors can carry the same cancer label while behaving differently after treatment. Knowing which drugs a patient previously received, and which molecular features accompany resistance, changes the experiment a scientist might order. Champions’ current PDX page advertises more than 1,500 models, including tumors from previously treated patients.

Its customers are pharmaceutical and biotechnology teams trying to choose compounds, combinations and patient populations. Those teams need evidence before committing to the next development step. A model that helps identify an unpromising direction can be useful even when the drug produces a disappointing result.

A screen, a mouse, a second question

Imagine a developer comparing several drug combinations. Starting every possibility in an animal study would make for a cumbersome first pass. Champions’ TumorGraft3D platform grows PDX-derived cells as three-dimensional tumor cultures, allowing researchers to screen candidates outside a living animal. Matched TumorGraft models offer a route into subsequent animal testing.

The sequence matters. A screen narrows the field; the next experiment asks whether the result survives a different biological setting. The company also offers blood-cancer screens, humanized models and syngeneic models, alongside assays that measure molecular changes. Each supplies a different view of the same development problem.

There is a boundary here that deserves plain language. A conventional PDX lacks a complete human immune system. Tumors do not always engraft, and transplantation can change their surroundings. Champions-associated research published in Annals of Oncology describes these limitations. A response in a mouse supports a hypothesis; human clinical testing must establish whether a treatment helps patients.

Nor does Champions own this approach exclusively. Crown Bioscience advertises more than 3,000 PDX models. The useful comparison is whether a supplier has the right tumor, treatment history, measurements and study design for the decision at hand. A larger catalog alone cannot settle that.

Consider the distinction between a tumor that has never encountered a particular therapy and one that survived it. Both may sit under the same diagnosis in a catalog. For a developer investigating resistance, that difference is the starting point. The experiment needs to reflect the problem the eventual patient population actually presents.

Lumin supplies the computational side. Researchers can explore genomic, transcriptomic and proteomic information, compare response patterns and select models. Its pitch includes AI-assisted analytics, but the tangible attraction is access to curated information linked to experimental biology. An observation on screen can suggest something to test in the laboratory.

The expensive part is making it repeat

Building the platform required outside capital. A March 2015 private placement raised $14 million, led by New Enterprise Associates, with existing investors participating. That figure included conversion of a $2 million management bridge. The announced uses included expanding the tumor bank and investing in technology.

Today the main commercial engine is contracted research. Champions describes typical studies around $125,000, with an increasing number between $250,000 and $500,000. It also licenses data and sells annual Lumin subscriptions. These are different buying decisions: commissioning an experiment, accessing information and subscribing to analytical software.

The commercial friction shows up in the numbers. Data-license revenue fell from about $4.7 million in fiscal 2025 to $764,000 in fiscal 2026. The earlier year contained a large single-customer transaction; the next contained smaller contracts. Overall revenue nevertheless reached $59.4 million. Champions also reduced its focus on flow cytometry. The evidence points to a business adjusting its mix, rather than a smooth expansion of every service.

Follow the drug. Then ask what happened.

Radiopharmaceuticals provide a concrete example of the newer direction. These agents carry radioactive material attached to a targeting molecule. Researchers need to understand where the compound travels as well as how the tumor responds. Champions partnered with Medicines Discovery Catapult to connect imaging expertise with its tumor models.

In September 2026, it announced in-house SPECT/CT imaging, bringing repeated measurements of uptake closer to biodistribution and efficacy studies. Integration gives a sponsor fewer handoffs between those measurements.

“Our vision is to make radiopharmaceutical development more predictive.”

Rob Brainin · CEO · 2026 imaging announcement

The next financial snapshot offered some encouragement. For the quarter ended July 31, 2026, Champions reported revenue of $15.2 million, up 8.8% from a year earlier. Oncology services margin reached 51%, while adjusted EBITDA was $671,000. Those figures measure operating progress; they do not establish that the newer data and discovery initiatives will become dependable businesses.

The lesson other businesses can copy is practical: retain the context that makes an asset useful, then make the next action easier. For Champions, that means connecting a tumor’s history to a model choice, a model choice to an experiment, and an experiment to a development decision. The value of the library appears when someone knows what to ask it next.