TK Profiles: The Good, The Bad, and The Ugly

Every animal survived. Body weights stable. No major clinical signs. You're practically popping champagne — and you're about to get an FDA hold. Because buried in the appendices is a TK table that proves your drug never actually made it into the bloodstream. In this episode, we break down toxicokinetics — what it is, why it underpins every dose justification you'll ever make, and what good, bad, and ugly TK profiles actually look like in practice.

Key takeaways:

• TK is just pharmacokinetics in the context of a toxicology study — Cmax, Tmax, AUC, and T½ are the bridge between the dose you give and the effects you see
• Good TK tells a clean, cohesive story: consistent exposure, dose-proportional increases, well-timed sampling — it validates your NOAEL and supports your IND
• Bad TK doesn't fall apart completely, but introduces enough uncertainty to make interpretation tricky — inconsistent exposure, non-linear AUC increases, and exposure overlap between dose groups
• Ugly TK undermines the entire study — no systemic exposure at high dose, formulation failure, unexpected accumulation — and can force you to repeat the study entirely
• TK is not a supporting actor. It's part of the main story. Review it alongside findings, not as an afterthought

Links:

• Early access — The Complete Guide to Nonclinical Development: https://www.nonclinical.academy/

The Nonclinical is hosted by Dessi McEntee, MS, DABT — board-certified toxicologist and Fractional Head of Toxicology. Subscribe to the newsletter on LinkedIn, take the course at nonclinical.academy, or work with Dessi at toxistrategy.com.

Transcript

SPEAKER_00 0:00

So picture this. You are uh staring at a draft toxicology report for your lead asset. Every single animal survived.

SPEAKER_01 0:08

Which is always the first thing you look for.

SPEAKER_00 0:10

Oh, absolutely. I mean no major clinical signs, the body weights are totally stable. You are practically popping the champagne. You don't get because you think you're heading straight to the clinic with this massive, perfectly safe dosing window.

SPEAKER_01 0:22

Yeah. But that is exactly where the trap is.

SPEAKER_00 0:25

Right. Because this exact report is about to get your clinical trial slapped with a devastating FDA hold. And uh why? Because in your rush to celebrate all that survival data, you skipped over a table buried deep in the appendices. Happens all the time. It does. A table that proves your so-called safe drug wasn't actually safe. It uh it just never made it into the animal's bloodstream.

SPEAKER_01 0:48

It is the ultimate false sense of security. I mean, I see biotech leaders fall into this trap constantly. Oh, yeah. They look at the superficial clinical observations, they see healthy animals, and they just assume the drug is benign. Meanwhile, the underlying pharmacokinetic data is quietly screaming that the formulation is fundamentally broken, or, you know, the biology is just wildly unpredictable.

SPEAKER_00 1:10

Okay, let's untack this because we are looking at an absolute masterclass on this exact blind spot today. We're pulling from the July 3, 2025 edition of the Toxin Trials newsletter.

SPEAKER_01 1:20

By Desi Mackinty.

SPEAKER_00 1:21

Exactly, Desi Maginty. She operates as a head of talks, a biotech board member, and uh she's the founder of Non-Clinical Academy. She has been in the room when these reports are built.

SPEAKER_01 1:31

And she knows exactly what causes them to completely implode under regulatory scrutiny.

SPEAKER_00 1:36

Totally. So our mission for this deep dive is to completely shift how you, as a biotech leader managing early stage programs, view this data. We want to take toxicokinetics or TK out of that boring regulatory checkbox category and uh really position it as the foundational pillar of your entire clinical strategy.

SPEAKER_01 1:54

Aaron Powell Which is absolutely critical for anyone managing an early stage pipeline. Misunderstanding the story your TK data is telling isn't just like an academic error. Right. It is quite literally the difference between a cleared IND and a multimillion dollar clinical delay. I mean, you cannot defend a clinical dose if you cannot defend the toxicology data it's based on.

SPEAKER_00 2:16

And to defend it, you really have to understand the fundamental mechanics of what TK is actually doing. Right. Look, if you're leading drug development, you already know the basic alphabet soup.

SPEAKER_01 2:25

Aaron Powell The classics, CMAX, T Max.

SPEAKER_00 2:27

Right. You know, CMAX is your peak plasma concentration. T Max is the time you hit that peak. AUC is your total exposure over time. And, you know, your half-life dictates how long the drug hangs around. But there is a massive psychological trap here. Trevor Burrus, Jr.

SPEAKER_01 2:41

Because those are the exact same metrics we use in standard pharmacokinetics, PK.

SPEAKER_00 2:45

Exactly. It is incredibly easy to just look at TK as uh PK done on a rat, but the source makes it clear that the strategic goal of TK is entirely inverted.

SPEAKER_01 2:54

Aaron Powell Yeah. If we connect this to the bigger picture, the difference lies entirely in the burden of proof. Aaron Powell Well, when you run standard PK in your efficacy models, you are trying to prove a positive. You want to show the drug gets to the target receptor and you know actually creates a therapeutic benefit.

SPEAKER_00 3:11

You want to see it work.

SPEAKER_01 3:12

Right. But in a toxicology study, you are doing the exact opposite. You are introducing the concept of the N-O-A-E-O, the no observed adverse effect level.

SPEAKER_00 3:22

The magical acronym.

SPEAKER_01 3:23

The one that dictates everything. Your objective there is to desperately prove that the toxicity you see at high doses and crucially the lack of toxicity you see at lower doses is mathematically linked to the drug's actual presence in the systemic circulation.

SPEAKER_00 3:39

Okay. I like to think of this using like a forensic accounting analogy.

SPEAKER_01 3:43

Oh, I like that.

SPEAKER_00 3:44

Yeah. So the clinical observations, the fact that the animal survived and low healthy, that's your top line revenue. It looks great on a billboard. But the TK data that is the forensic audit.

SPEAKER_01 3:53

It's digging into the receipts.

SPEAKER_00 3:55

Yes. It is the hardcore, unglamorous math that proves whether your revenue is real or if it's just, you know, a hollow accounting trick.

SPEAKER_01 4:02

Aaron Powell That is the perfect way to frame it. Because regulators are essentially auditors. When you submit your investigational new drug application, your IND, you don't just walk in and say, hey, we gave the animal 100 milligrams and they didn't die, so 100 milligrams is safe.

SPEAKER_00 4:17

They would laugh you out of the building.

SPEAKER_01 4:18

Exactly. The FDA does not care about the raw dose administered. They care about the exposure. You have to use your TK data to prove that a specific concentration of drug in the blood is safe. Right. Every single clinical margin you calculate for your human trials is based on that internal exposure, not the external dose.

SPEAKER_00 4:37

Aaron Powell, which brings us to the ideal scenario. When the biology and the chemistry align perfectly, you know, what does that actually look like? The new letter refers to this as the good. And it really boils down to telling a cohesive, mathematically predictable story. Trevor Burrus, Jr.

SPEAKER_01 4:54

And predictability is the holy grail of drug development. Good TK means your systemic exposure is beautifully consistent across all the animals in your intended dose levels.

SPEAKER_00 5:04

So no wild variations.

SPEAKER_01 5:05

Right. If you have five animals in a group, their blood work looks relatively similar. You don't have one animal absorbing 10 times more drug than its cagemate.

SPEAKER_00 5:15

And you see dose proportional increases. This is the part that makes the clinical modeling so much easier, right? If you double the dose from 10 milligrams to 20 milligrams, you want to see roughly double the AUC and double the CMAX.

SPEAKER_01 5:28

Exactly. It means the absorption pathways aren't getting saturated, the metabolic clearance is functioning normally, the drug is behaving exactly the way the math says it should.

SPEAKER_00 5:38

Yes, it's just clean.

SPEAKER_01 5:38

It's beautiful. And beyond the physiological behavior, good TK also proves that your study design was fundamentally sound. Like your time points are perfectly placed to actually catch the TMAX.

SPEAKER_00 5:49

Oh, this is huge. Yeah.

SPEAKER_01 5:50

Because if your drug hits its peak concentration at hour two, but your toxicologists only drew blood at hour one and hour six.

SPEAKER_00 5:57

You completely missed the peak exposure. You're flying blind.

SPEAKER_01 6:00

You really are. Good TK proves your sampling represented the full dosing interval accurately. Furthermore, it shows that steady-state exposure matches your initial models with no hidden massive accumulation over, say, weeks of daily dosing.

SPEAKER_00 6:17

Okay, here's where it gets really interesting. When you have that predictable linear story, your TK profile stops being just a data table. It transforms into an ironclad insurance policy for your clinical margin.

SPEAKER_01 6:30

It's your best defense.

SPEAKER_00 6:31

Because if a regulator questions a minor finding, say a slight elevation in liver enzymes at your high dose, you can point directly to the TK and say, yes, that is a real finding, but look at the massive drug exposure required to trigger it. Our clinical dose will be nowhere near that.

SPEAKER_01 6:45

It removes all the ambiguity, it tells the FDA that your study was meaningful, your mathematical models are sound, and your safe level, your NOAAL is an absolute rock.

SPEAKER_00 6:54

Regulated love of a rock.

SPEAKER_01 6:55

They do, they hate surprises. When you give them dose proportional, clean data, you remove the regulatory friction that slows programs down.

SPEAKER_00 7:02

But uh let's step out of the utopian ideal for a second. Because anyone who has spent more than a week in biotech knows biology is messy.

SPEAKER_01 7:11

It's always messy.

SPEAKER_00 7:12

Right. What happens when that perfect linearity starts to fray? This is what the source calls the bad. And it seems like this is where leaders really start losing sleep because the data doesn't completely fail. It just gets incredibly murky.

SPEAKER_01 7:26

The defining characteristic of bad TK is the sudden introduction of uncertainty. You start seeing wild inconsistencies. Like maybe animal A and animal B are in the same exact dose group, but their internal exposures are orders of magnitude apart. Wow. Or worse, you start seeing nonlinear absorption. You double the raw dose, but the total exposure in the blood only goes up by like 10%.

SPEAKER_00 7:49

Usually because something in the biological pathway is maxed out, right? Like the gut transporters are saturated, or the drug is precipitating out of solution in the stomach before it can even be absorbed.

SPEAKER_01 8:00

Precisely. Or perhaps it's a clearance issue. The drug induces metabolic enzymes, so by week four, the animal is clearing the drug three times faster than it was on day one, which just tanks your overall exposure. Or you see mild, unanticipated accumulation where the trough levels just keep creeping up day after day.

SPEAKER_00 8:19

Okay, let me push back on this a bit though. Because if I am sitting in the boardroom looking at a slightly messy graph, my first instinct is to say, look, we are dealing with living, breathing, genetically diverse animals. Aren't we demanding mathematical perfection from a notoriously chaotic system?

SPEAKER_01 8:38

It's a tempting argument.

SPEAKER_00 8:39

Right. If there is a general upward trend in the exposure, even if it's a little noisy, why isn't that enough for the agency to just rubber stamp the safety margin? We know the drug is generally getting in.

SPEAKER_01 8:50

That is the exact rationalization that leads to brutal clinical holds.

SPEAKER_00 8:55

Really?

SPEAKER_01 8:55

Yes, you cannot rely on a general trend when human safety is on the line. The newsletter highlights a specific phenomenon in bad TK that destroys the biological variability argument. Exposure overlap. Specifically, overlap between your NOAL group, your safe dose, and your LOAL group.

SPEAKER_00 9:13

The lowest observed adverse effect level.

SPEAKER_01 9:15

Right, where the animals actually start getting sick.

SPEAKER_00 9:17

Walk me through the mechanics of that, because that sounds like a paradox. How can the safe group and the sick group overlap?

SPEAKER_01 9:23

It happens because of that exact variability and nonlinear absorption you were just defending. Imagine you dose group two at 50 milligrams and they are perfectly healthy. You dose group three at 100 milligrams and they develop severe organ toxicity. Okay. On paper, 50 milligrams is your safe NOAL. But when you run the forensic audit, when you look at the TK, you realize the highest exposed animal in group two actually had more drugs circulating in its blood than the lowest exposed animal in group three.

SPEAKER_00 9:53

Oh man. So the internal environment of a healthy animal was actually subjected to more drug than a sick animal.

SPEAKER_01 9:59

Yes. And the moment that happens, your entire safety argument collapses. You can no longer draw a definitive line in the sand until the FDA, hey, toxicity only happens when the blood concentration hits this specific threshold.

SPEAKER_00 10:11

Because you just proved it doesn't.

SPEAKER_01 10:13

Exactly. The findings look dose-related on the clinical chart, but the chemistry tells a conflicting, unresolvable story.

SPEAKER_00 10:21

Which means the toxicologists and the regulators are forced into a corner. When the data is blurry, they are legally and ethically obligated to assume the absolute worst-case scenario.

SPEAKER_01 10:32

And for a biotech leader, that translates directly into shrinking therapeutic windows. You thought you were going into the clinic with a 50x safety margin, but because of that exposure overlap, the regulators force you to calculate your starting dose using the most conservative possible interpretation of the data.

SPEAKER_00 10:49

So your margin gets crushed.

SPEAKER_01 10:50

Suddenly you have a 5x margin. You might not even be able to reach a therapeutically active dose in humans without hitting a safety ceiling imposed by your noisy rat data.

SPEAKER_00 11:00

That margin killer is painful to manage. But as much of a headache as bad TK is, at least you are still, you know, moving forward.

SPEAKER_01 11:08

Barely, but yes.

SPEAKER_00 11:10

The newsletter then shifts into the true boardroom nightmares. The scenarios that don't just shrink your margins, they actively destroy your entire development timeline. Let's talk about the ugly.

SPEAKER_01 11:21

Ugly TK is an existential threat to an asset. This isn't about messy interpretation or slight overlap. This is when the foundational premise of the study entirely falls apart.

SPEAKER_00 11:32

It's the point of no return.

SPEAKER_01 11:34

Usually rendering the entire multimillion dollar toxicology package completely useless.

SPEAKER_00 11:40

The source outlines several terrifying features here. The most glaring one is zero systemic exposure in your mid or high dose groups. You are feeding the animals massive amounts of the active pharmaceutical ingredient, but the blood work comes back flat. Nothing is absorbing.

SPEAKER_01 11:54

And that almost always points to a catastrophic formulation failure. The vehicle you use to deliver the drug in the TOC study might have caused it to crash out of solution the second it hit the low pH of the stomach.

SPEAKER_00 12:06

It just becomes a break in their stomach.

SPEAKER_01 12:07

Basically, or perhaps the first pass metabolism in the liver is so aggressive that it obliterates 100% of the drug before it ever reaches systemic circulation.

SPEAKER_00 12:16

Another feature is massive unexplained accumulation. You might have a 13-week study running smoothly, and suddenly at week 12, animals start dropping from severe, bizarre toxicities that weren't there at week four.

SPEAKER_01 12:30

It's terrifying when that happens.

SPEAKER_00 12:31

I bet. And the ugly TK reveals that the drug has a massive half-life you didn't account for, and it has just been silently building up in the deep tissues for three months.

SPEAKER_01 12:42

You also see wild, unexplainable outliers that completely skew the mean exposure. If you have five animals and four of them have an AUC of a hundred, but one animal has an AUC of 10,000.

SPEAKER_00 12:54

Wow.

SPEAKER_01 12:55

Yeah. Your statistical average is now completely divorced from biological reality.

SPEAKER_00 12:59

But the one that really stands out from the source is the flatline exposure profile across escalating doses. You dose 10 milligrams, 50 milligrams, and 100 milligrams. But the blood levels for all three groups are identical.

SPEAKER_01 13:10

It's maddening.

SPEAKER_00 13:11

The absorption pathway was maxed out at 10 milligrams, and the rest just passed straight through the animal.

SPEAKER_01 13:15

Which creates a massive logical black hole if you see toxicity at the high dose. If the blood exposure is exactly the same across all three groups, but only the high dose animals are getting sick, what is actually causing the toxicity? Right. It can't be systemic exposure to the drug. Is it localized irritation in the gut? Is it toxicity from the excipients in the formulation?

SPEAKER_00 13:39

So what does this all mean? Let's go back to our opening scenario. You are a leader holding that pristine survival report. The high dose animals didn't get sick, they look perfectly healthy, their coats are shiny, they're gaining weight.

SPEAKER_01 13:51

The dream.

SPEAKER_00 13:51

But if your ugly TK reveals that their blood exposure was barely above background noise, you are living in a dangerous delusion. You absolutely cannot go to an investor or a regulator and claim we observed no adverse effects at the higher dose.

SPEAKER_01 14:05

Because you didn't actually test the drug. You just fed them highly expensive, completely unabsorbed powder.

SPEAKER_00 14:10

Exactly.

SPEAKER_01 14:10

The drug never engaged with their biology. Of course they're healthy.

SPEAKER_00 14:13

What's fascinating here is how ugly TK doesn't just raise red flags, it actively erases answers. If you have a messy study, you can sometimes negotiate or model your way out of it. If you have an ugly study, you don't even have a foundation to stand on. The data doesn't exist.

SPEAKER_01 14:29

And the consequence of that erase data is study repetition. Repeating a pivotal toxicology study is a catastrophic event for an early stage biotech.

SPEAKER_00 14:37

Gotta be devastating.

SPEAKER_01 14:38

It means a massive unplanned cash burn. It guarantees deep regulatory skepticism moving forward, and it guarantees serious clinical delays that can allow your competitors to beat you to the market. A study that looked pristine on the surface completely unravels because the chemistry proves the drug was never there.

SPEAKER_00 14:57

That is a very sobering reality check. We have moved from the margin-shrinking headaches of bad data to the absolute existential dread of unabsorbed powder. It's a dark place. So, how does a proactive biotech leader ensure their development teams steer clear of these landmines? Let's transition into the playbook. What are the actionable strategies the source provides to prevent this?

SPEAKER_01 15:19

The overarching theme Desi Mackinty pushes is that TK interpretation requires heavy context and intense cross-functional collaboration. You cannot view the TK table in a vacuum, and you certainly cannot wait until the final draft report is sitting in your inbox to look at it.

SPEAKER_00 15:35

It's too late by then.

SPEAKER_01 15:36

Way too late.

SPEAKER_00 15:37

The newsletter gives us five highly actionable steps for success. Step one, run early PK to design appropriate TK time points. We touched on this earlier. Do not guess when your T Max is going to hit.

SPEAKER_01 15:59

But if this new asset is absorbed incredibly quickly and peaks at 15 minutes, your generic time points will completely miss the CMAX. You have to run small non-GLP pharmacokinetic pilot studies to understand the drug's basic behavior before you lock in the design for the massive, expensive pivotal TACAS study.

SPEAKER_00 16:17

That makes total sense. Okay, step two. Use a sufficient number of animals to avoid unhelpful pooling or massive variability. This feels like it directly addresses those wild outliers we just talked about.

SPEAKER_01 16:29

It does. There is always pressure from the finance side to use the absolute minimum number of animals to save money and resources.

SPEAKER_00 16:36

Of course.

SPEAKER_01 16:37

But if you have high biological variability, a small sample size is a massive risk. If you only have three animals per group and one has a weird metabolic quirk that spikes its exposure, your entire safety margin is ruined. You need a robust enough sample size to see the true statistical reality of how the drug behaves across a population.

SPEAKER_00 16:56

Gotta spend money to save money. Step three, review the TK data alongside the clinical findings, never as an afterthought.

SPEAKER_01 17:04

They are inextricably linked. The clinical finding, the pathology report, tells you what happened to the animal's organs. The TK data is the only thing that tells you why it happened and whether the drug was actually the culprit. You cannot finalize one without understanding the other.

SPEAKER_00 17:17

Which leads right into step four. Tie exposure to effect before officially calling anything adverse.

SPEAKER_01 17:23

This is how you protect your asset's reputation. If a pathologist notes a minor liver enzyme elevation, but the TK audit shows that specific animal had incredibly low systemic exposure, you have to hit pause.

SPEAKER_00 17:35

Wait, really? Even if the enzyme is elevated?

SPEAKER_01 17:38

Yes, because you have to investigate if something else caused that elevation, like a totally unrelated infection, before you permanently brand your new drug with an adverse liver finding in your FDA investigator's brochure.

SPEAKER_00 17:51

Oh wow. That makes so much sense. And finally, step five: never ignore accumulation, especially in longer-term studies. If your drug's half-life is longer than your daily dosing interval, the math dictates that it is going to build up.

SPEAKER_01 18:04

It has nowhere else to go.

SPEAKER_00 18:05

Right. You have to model that trajectory out mathematically before you start a three-month study so you aren't blindsided by sudden toxicity at week 12.

SPEAKER_01 18:13

The ultimate takeaway for a biotech leader is that you cannot be a passive recipient of this data. You need to be actively questioning your toxicology and pharmacology teams about these specific five points months before the pivotal studies begin.

SPEAKER_00 18:26

Lead from the front.

SPEAKER_01 18:27

Exactly. You have to set the expectation early that TK is not a supporting actor. It is the main narrative. It validates, it questions, or it entirely unravels your toxicological package based on how rigorously this playbook is applied.

SPEAKER_00 18:40

So let's briefly recap this journey. We started by reframing TK not as the basic pharmacokinetic exercise, but as the forensic audit that proves your clinical safety margins. We saw how good TK builds those beautiful wide therapeutic windows by telling a mathematically predictable story.

SPEAKER_01 18:58

Like a holy grail.

SPEAKER_00 18:59

Then we explored how bad TK introduces uncertainty, creating that dreaded overlap between safe and toxic doses that ultimately shrinks your dosing limits. And we confronted ugly TK, the unabsorbed powder and formulation failures that force you to burn investor cash and start entirely over.

SPEAKER_01 19:14

It really requires a fundamental paradigm shift in how executives review their non-clinical data.

SPEAKER_00 19:19

It really does. Thank you for joining us on this deep dive. Our main takeaway for you today is simple but vital. The next time that draft talks report pings your inbox, do not scroll past the executive summary straight to the survival data. Your job is to flip straight to that toxicokinetics table first, not last.

SPEAKER_01 19:39

And as you review those tables, I want to leave you with something to mull over. We just discussed how ugly TK, things like entirely erratic exposure or zero absorption at the highest dose levels, frequently stems from formulation failures. Right. It raises a profound strategic question. As a biotech leader managing early stage development, how much of your drug's ultimate valuation is actually based on the brilliant chemistry of the active ingredient and how much is secretly being held hostage by the seemingly mundane science of how that drug was formulated for a rat?