"The in vitro data look excellent."
That statement comes up often in drug discovery. The candidate shows strong biochemical potency, favorable cellular activity, good selectivity, high metabolic stability, and limited CYP inhibition. On paper, the molecule looks ready to progress into animal pharmacology and pharmacokinetic studies.
Why, then, can a candidate with strong in vitro data still perform poorly in vivo?
A small molecule may show excellent microsomal stability but unexpectedly high clearance in animals. An antibody may show strong target binding and cellular potency but fail to maintain sufficient target coverage in vivo. An ADC may retain favorable antibody exposure while releasing its payload prematurely. A peptide may look stable in vitro yet be eliminated rapidly in vivo. For nucleic acid therapeutics, plasma concentrations can decline quickly even though pharmacological activity persists for weeks or months.
These outcomes are not contradictions. They reflect a basic principle of DMPK: in vitro assays isolate individual mechanisms, while in vivo pharmacokinetics reflects the behavior of the entire biological system. The relevant question is not whether a drug is "stable" or has a "long half-life," but whether its disposition produces sufficient exposure at the site, and in the molecular form, required for pharmacological activity.
From In Vitro Properties to In Vivo Drug Behavior
After administration, a drug passes through a sequence of connected processes: absorption, distribution, metabolism, elimination, tissue penetration, target interaction, and downstream pharmacology.
An in vitro assay examines only one component of this system. Microsomal stability, for example, describes intrinsic metabolic turnover under defined experimental conditions. By itself it does not determine renal clearance, transporter-mediated elimination, tissue distribution, plasma protein binding, or the contribution of extrahepatic metabolism.
The same point applies to more complex modalities. For antibodies, systemic exposure depends not only on FcRn-mediated recycling but also on target expression and target-mediated disposition. For ADCs, total antibody exposure does not necessarily represent exposure to intact, pharmacologically competent ADC. For peptides and nucleic acids, tissue distribution and intracellular persistence can matter more than plasma stability.
DMPK interpretation therefore has to move from isolated parameters toward a mechanism-based exposure framework:
Molecular properties → absorption and disposition → tissue exposure → target engagement → pharmacodynamic response
Failure at any stage breaks the relationship between in vitro potency and in vivo efficacy.
Small Molecules: Microsomal Stability Is Only One Determinant of Clearance
For conventional small molecules, early DMPK programs commonly include liver microsomal stability, hepatocyte stability, CYP inhibition or induction, plasma stability, and permeability assessments.
Microsomal stability is useful because it characterizes a compound's susceptibility to hepatic oxidative metabolism. High microsomal stability, however, should not be read as low systemic clearance.
In vivo clearance can also involve renal filtration and secretion, biliary excretion, transporter activity, extrahepatic metabolism, and other elimination mechanisms. Distribution can further change the observed plasma concentration-time profile.
Half-life also needs careful reading. Because:
t₁/₂ = 0.693 × Vd / CL
a long terminal half-life can result from low clearance, a large apparent volume of distribution, or both. A short half-life, in turn, does not mean rapid metabolic degradation alone.
The distinction matters when extrapolating across species. Drug-metabolizing enzymes and transporters differ among rodents, dogs, non-human primates, and humans. Animal PK cannot always be predicted from a single in vitro metabolic assay, or from another species, without integrating several lines of evidence.
Small-molecule DMPK should therefore answer two separate questions:
How stable is the compound?
And what mechanisms dominate its in vivo clearance and exposure?
The questions are related but not interchangeable.
Antibodies: Exposure Is Not the Same as Target Coverage
Large-molecule therapeutics operate under a different DMPK framework.
For conventional IgG antibodies, FcRn-mediated recycling contributes substantially to prolonged systemic persistence. Systemic half-life alone, however, does not determine whether an antibody produces adequate pharmacological activity.
Target expression, tissue distribution, receptor-mediated internalization, and target-mediated drug disposition (TMDD) can all influence antibody exposure. This matters most for antibodies directed against highly expressed or rapidly internalizing targets, where target binding becomes part of the elimination pathway and produces nonlinear or dose-dependent pharmacokinetics.
Antibody development should therefore move past the question of how long the antibody stays in plasma, and ask whether the exposure maintains sufficient target coverage for the required duration. The distinction is central to PK/PD interpretation: an antibody can have a favorable systemic half-life yet still fall short pharmacologically if tissue penetration or target engagement is insufficient.
ADCs: Measuring the Antibody Does Not Tell the Whole Story
Antibody-drug conjugates add another layer of complexity, because the therapeutic entity is not the antibody alone.
An ADC contains at least three functionally important components: the antibody, the linker, and the payload. After administration their behavior can diverge.
The circulating species may include intact ADC, partially deconjugated antibody, free payload, and active or inactive catabolites. Total antibody concentration can stay relatively high even as the concentration of pharmacologically competent ADC declines.
This creates an important analytical distinction:
Total antibody exposure ≠ intact ADC exposure ≠ active payload exposure
ADC studies therefore need to measure parameters such as intact or conjugated antibody, drug-to-antibody ratio (DAR), free payload, payload-related metabolites, and, where relevant, tissue exposure.
DAR also shows why optimizing a single parameter can mislead. Increasing payload loading raises drug content per antibody, but excessive DAR changes physicochemical properties, including hydrophobicity, aggregation propensity, tissue distribution, and clearance.
The central DMPK question for an ADC is whether intact ADC can reach the relevant tissue, stay sufficiently stable in circulation, and release an active payload at the right site and time.
Peptides: Half-Life Extension Requires More Than Proteolytic Stability
Peptide therapeutics present another distinct DMPK challenge.
Many endogenous peptides are degraded rapidly by proteases, and because of their small molecular size they also undergo substantial renal clearance. Extending peptide exposure often requires changing the overall disposition profile, rather than only increasing resistance to enzymatic degradation.
Semaglutide is a useful example. Its prolonged systemic exposure is associated with structural modification that promotes albumin binding, which reduces renal elimination and extends circulation time.
The broader lesson is that peptide DMPK should consider several connected properties: proteolytic stability, protein binding, renal clearance, tissue distribution, and receptor engagement. A peptide that is highly stable in an isolated protease assay can still have an unfavorable in vivo profile if its dominant clearance mechanism lies elsewhere.
Nucleic Acids: Plasma PK May Not Reflect Pharmacological Duration
Small interfering RNA (siRNA) and antisense oligonucleotides (ASOs) show the clearest limitation of conventional plasma PK interpretation.
For many traditional small molecules, plasma concentration is reasonably connected to exposure at the pharmacological site. For nucleic acid therapeutics, that connection can be much weaker.
After administration, an oligonucleotide may distribute into tissues, undergo cellular uptake and intracellular trafficking, and accumulate in pharmacologically relevant compartments. Biologically meaningful exposure can persist after plasma concentrations have declined.
Inclisiran illustrates this. As a GalNAc-conjugated siRNA, its activity depends on delivery to hepatocytes, engagement of the intracellular RNA interference machinery, and sustained suppression of PCSK9 expression. The duration of pharmacodynamic activity cannot be read from its plasma concentration-time profile.
The same principle is even clearer for CNS-directed ASOs such as nusinersen, where cerebrospinal fluid and neural tissue exposure matter more to pharmacological activity than plasma concentration.
For nucleic acid therapeutics, the relevant chain is:
Dose → tissue distribution → cellular uptake → intracellular persistence → target RNA modulation → PD duration
This is fundamentally different from evaluating a conventional small molecule mainly through plasma PK.
One DMPK Framework Does Not Fit Every Drug Modality
The differences among modalities can be summarized by asking what the most relevant pharmacological exposure actually is.
| Drug modality | Common early DMPK focus | Common interpretation trap | More informative question |
|---|---|---|---|
| Small molecules | CL, F, t₁/₂, CYP, stability | High stability = low clearance | What mechanisms determine free drug and tissue exposure? |
| Antibodies | CL, t₁/₂, FcRn | Long half-life = sufficient efficacy | Is target coverage maintained? |
| ADCs | Antibody PK | Antibody exposure = ADC exposure | How much intact ADC and active payload reach the target? |
| Peptides | Stability, t₁/₂ | Proteolytic stability = long exposure | How do binding and clearance determine systemic persistence? |
| siRNA / ASO | Plasma PK, tissue distribution | Plasma disappearance = loss of activity | Is productive intracellular exposure sustained? |
The table makes a broader point: DMPK endpoints should be selected according to the biology of the therapeutic modality, rather than applied uniformly across drug classes.
What Should DMPK Ask Before a Candidate Moves Forward?
When reviewing a DMPK package, teams often focus first on AUC, Cmax, clearance, and half-life. These parameters matter, but they do not explain by themselves whether the drug will work.
A more useful assessment starts with four questions.
Where does the drug need to act? The relevant compartment may be plasma, tumor tissue, liver, CNS, or an intracellular compartment.
What molecular species is pharmacologically active? For an ADC, it may be intact conjugate or released payload. For a nucleic acid therapeutic, total concentration may not represent productive intracellular exposure.
Can PK explain PD? If systemic exposure rises without a corresponding pharmacodynamic response, the problem may lie in tissue distribution, target engagement, intracellular delivery, or downstream biology, rather than in systemic exposure itself.
Does the selected animal model reproduce the relevant disposition mechanism? Species differences in enzymes, transporters, target expression, FcRn biology, tissue distribution, and disease state can all affect translational interpretation.
These questions shift DMPK from a collection of numerical parameters toward a mechanistic understanding of drug behavior.
From Exposure to Translation
The goal in DMPK is not to maximize a single parameter. It is to establish a coherent relationship between dose, exposure, distribution, target engagement, pharmacodynamics, and ultimately efficacy and safety.
This matters more as development expands beyond conventional small molecules toward antibodies, ADCs, peptides, oligonucleotides, cell and gene therapies, and other complex modalities.
For some programs, plasma PK remains the dominant translational endpoint. For others, tissue exposure, target coverage, intact drug concentration, intracellular persistence, or imaging-based distribution provide more informative evidence. Nonclinical imaging can complement conventional PK by giving longitudinal information on biodistribution, target engagement, and pharmacodynamic changes in living subjects.
In non-human primate studies, integrating PK/PD with tissue-level measurements, biomarkers, imaging, and pharmacological endpoints helps establish whether the candidate behaves according to its intended mechanism. Prisys' translational research platform combines NHP pharmacology with PK/PD, biomarker analysis, pathology, and clinical-equivalent imaging where scientifically appropriate, supporting a more integrated interpretation of drug disposition and pharmacology.
The lesson is straightforward: good in vitro data are necessary, but they do not guarantee good in vivo pharmacology. The most informative DMPK strategy follows the drug from molecular design to its actual site of action:
Molecular properties → In vivo disposition → Relevant tissue exposure → Target engagement → PD → Efficacy and safety
When this chain is understood, unexpected in vivo results stop being "failures of prediction" and become mechanistic evidence that can guide candidate optimization, study design, and translational decision-making.
FAQ
Q: Why can a compound with high microsomal stability still have high in vivo clearance?
A: Microsomal stability primarily reflects metabolic turnover under controlled experimental conditions. In vivo clearance can also involve renal elimination, biliary excretion, transporter-mediated processes, extrahepatic metabolism, and distribution-related effects. Therefore, microsomal stability should be interpreted as one component of overall clearance rather than a direct surrogate for systemic PK.
Q: Why is plasma half-life not sufficient for evaluating biologics?
A: For biologics, pharmacological activity depends not only on systemic persistence but also on tissue distribution, target expression, target-mediated disposition, and target engagement. A long half-life does not necessarily ensure adequate exposure at the pharmacological site.
Q: What makes ADC PK different from conventional antibody PK?
A: An ADC contains an antibody, linker, and payload whose disposition can change independently after administration. Total antibody concentration therefore does not necessarily reflect intact ADC or active payload exposure. ADC studies should consider the integrity of the conjugate and payload-related species in addition to conventional antibody PK.
Q: Why can nucleic acid drugs remain active after plasma concentrations decline?
A: siRNA and ASO therapeutics can accumulate in tissues and cells and may produce prolonged intracellular pharmacological effects. Consequently, plasma concentration can decline substantially before the relevant intracellular mechanism has returned to baseline.
Q: What is the most important DMPK endpoint?
A: There is no universal endpoint. The appropriate endpoint depends on the therapeutic modality and mechanism of action. For some small molecules, free drug exposure and tissue concentration may be critical; for antibodies, target coverage may be more informative; for ADCs, intact ADC and payload exposure may matter; and for nucleic acids, productive intracellular exposure may be the key determinant of pharmacological duration.
