"Data Hormesis": When Data Becomes Toxic

Welcome back to The Nonclinical, a newsletter where we unpack how nonclinical toxicology actually works inside drug development — beyond the protocol and into real decisions.

Believe it or not, having data doesn’t equal having the win.

There is a point in nonclinical development where data stop helping a program and begin to work against it. It’s counterintuitive, especially in a field built on rigor, documentation, and evidence. But it happens more often than teams like to admit.

In toxicology, we have a term for this kind of inflection point: hormesis. At low doses, an exposure can produce a beneficial effect. Push past a certain threshold, though, and the curve pivots. What was once helpful becomes harmful. The dose-response relationship hasn’t disappeared — it has simply crossed a line.

Hormesis: when low levels of drug are beneficial, but high levels are harmful

The same dynamic shows up with data.

Early in development, data do exactly what we want them to do. They reduce uncertainty. They help define boundaries, characterize risk, and support forward movement. They allow teams to have grounded conversations about exposure margins, target effects, and feasibility.

But beyond a certain point, additional data stop clarifying decisions. Instead, they complicate interpretation. Signals become harder to separate from noise. Findings compete for attention rather than converging toward resolution. Teams feel busy and productive, yet less confident than before.

This is usually the moment when the questions subtly change. Instead of asking what decision the data are meant to support, teams begin asking what else they can run. Another endpoint. Another study. Another attempt to make the uncertainty go away.

We can call this "data hormesis":

"Data Hormesis": when the right amount of data is informative, but too much data can be confusing

That pivot isn’t driven by bad science. It’s driven by the absence of a clear strategic spine.

Nonclinical development is often treated as a process of gathering evidence. In practice, it’s a process of making choices under uncertainty. Data are meant to inform those choices — not replace them. When decisions aren’t defined early, data accumulation quietly becomes a substitute for judgment rather than a tool to support it.

When that happens, familiar patterns emerge. Equivocal findings trigger rework instead of interpretation. Borderline results lead to additional studies without clarity on what would actually change. CRO precedent begins to fill the vacuum left when sponsor ownership of decisions is unclear.

The program ends up with extensive datasets and very little clarity about what truly matters. This is where “data hormesis” crosses its toxic threshold - where more information stops helping and starts obscuring the path forward.

Eventually, a point of no return is reached. Conflicting signals pile up. Interpretations fragment. The regulatory narrative becomes harder to defend, not because the science is flawed, but because it was never anchored. The result is pushback from regulators - and sometimes the creeping belief that the drug itself isn’t viable.

In many cases, the drug is viable.

The failure wasn’t the molecule. It was the absence of nonclinical leadership when the decisions mattered most.

---

The programs that move efficiently toward IND aren’t the ones with the most data. They’re the ones that decide, explicitly and early, which risks are acceptable, which questions are worth answering now, and which outcomes would meaningfully change the path forward.

In those programs, data serve their intended role.

→ They don’t create strategy. They make strategy visible.

And when strategy is absent, more data won’t save the program. It will slow it down — often at exactly the moment when clarity matters most.

Short takeaway: This pattern - where data accumulate towards a toxic level - is what led me to write Data Is Not Strategy. The book isn’t about doing less science or lowering rigor. It’s about leveraging the data in it's most beneficial state - and having the restraint to not let it become overgrown. For teams approaching IND, that clarity often matters more than another dataset.

📍If you're building your tox career and want to understand how nonclinical development actually works, The Nonclinical is for you. Free. Every other Thursday. Subscribe below:

Dessi McEntee, MS, DABT is a board-certified toxicologist and the founder of Toxistrategy LLC. She works as a Fractional Head of Toxicology for early-stage biotechs navigating IND-enabling development — bringing senior nonclinical leadership to teams that need it without the cost of a full-time hire.

The Nonclinical is written for scientists, early-career toxicologists, and development teams who want to understand how nonclinical decisions actually get made — and why they matter.

Learn more at www.toxistrategy.com