A Test That Needs to Mean Something

Dissolution testing is one of the most widely performed analyses in pharmaceutical development. It is also one of the most frequently misunderstood.

A dissolution test that cannot distinguish a good batch from a bad one is not a quality control tool. It is a formality. And a test that shows 98% release in 30 minutes in pH 6.8 phosphate buffer, but the drug is barely absorbed in patients, is actively misleading. The goal is a method that reflects what actually happens when the patient swallows the tablet.

Getting that right requires understanding the biology of oral drug absorption, the limitations of the most common test methods, and how to match the method to the molecule and the formulation.

The BCS Framework as Your Starting Point

The Biopharmaceutics Classification System (BCS) divides drugs into four classes based on solubility and permeability. The class tells you which absorption step is rate-limiting, and therefore what your dissolution method needs to capture.

BCS Class	Solubility	Permeability	Rate-Limiting Step	Dissolution Test Priority
Class I	High	High	Gastric emptying	Simple: discriminate fast from slow release; biowaver often available
Class II	Low	High	Dissolution in GI fluid	Critical: method must reflect solubility-limited dissolution; biorelevant media often needed
Class III	High	Low	Permeation across intestinal wall	Less critical for formulation discrimination; API properties dominate
Class IV	Low	Low	Both dissolution and permeation	Most challenging: biorelevant media important; in vitro-in vivo correlation difficult

Standard Buffer vs. Biorelevant Media: When It Matters

Standard Buffer Methods

Standard USP buffer systems at pH 1.2, 4.5, and 6.8 are the default for most regulatory dissolution work. They are simple, reproducible, and well-understood. For BCS Class I drugs and many conventional formulations, they do the job.

For BCS Class II and IV drugs, they often do not. Aqueous buffers lack the surfactant and bile salt content of real gastrointestinal fluid. A poorly soluble API may show 100% release in a standard buffer with a high concentration of solubilising excipients, then fail to achieve adequate exposure in patients because the dissolution medium in the GI tract has very different solubilising power.

Biorelevant Media

Biorelevant media such as FaSSIF (fasted-state simulated intestinal fluid) and FeSSIF (fed-state simulated intestinal fluid) contain bile salts and lecithin at physiologically relevant concentrations. They better mimic the solubilising capacity of the intestinal environment and are far more discriminating for BCS Class II molecules.

They are also more variable to prepare, more expensive, and not required by regulators for routine release testing. The practical approach is to use biorelevant media during development to understand in vivo-predictive performance, then design a simpler validated method for GMP release that is correlated to the biorelevant results.

In Vitro-In Vivo Correlation (IVIVC): The Regulatory Goal

An IVIVC links the in vitro dissolution profile to the in vivo absorption profile, creating a model that allows dissolution data to predict bioavailability. A validated Level A IVIVC, where the in vitro dissolution curve maps point-by-point to the in vivo absorption curve, is the most valuable outcome: it allows dissolution testing to serve as a surrogate for bioavailability studies when the formulation changes. The FDA’s IVIVC guidance for extended release oral dosage forms outlines the requirements for establishing and using an IVIVC in regulatory submissions.

Establishing an IVIVC is not always feasible, particularly for Class II/IV molecules with complex absorption behaviour. But even a partial correlation, or a mechanistic understanding of why in vitro and in vivo profiles differ, is more valuable than a dissolution method selected without considering the biology.

Dissolution Testing for Modified Release Products

Extended release, delayed release, and pulsatile release dosage forms require dissolution methods that capture the full release profile over the intended release duration. A method that tests at a single timepoint, or that uses conditions where the release mechanism does not function correctly, will not discriminate failing from passing formulations.

For enteric-coated products, the dissolution method must include a two-stage test: acid stage exposure at pH 1.2 (to confirm the coating resists gastric conditions) followed by buffer stage at pH 6.8 (to confirm complete release in intestinal conditions). The duration and test conditions of each stage must reflect the gastric residence time expected in the target patient population.

Ardena’s Dissolution Development Expertise

Ardena’s analytical teams at Ghent develop dissolution methods as an integrated part of formulation development programmes, not as a separate analytical exercise. Method design is informed by the molecule’s BCS classification, the intended formulation, and the clinical context, ensuring that the method reflects in vivo performance and supports the regulatory filing.

Why CQAs Matter More for LNPs Than for Conventional Drug Products

For a conventional small molecule tablet, the critical quality attributes that determine in vivo performance, dissolution rate, disintegration time, content uniformity, are relatively straightforward to measure and control. For a lipid nanoparticle (LNP) system, the relationship between physicochemical properties and in vivo behaviour is more intricate, and the analytical tools needed to characterise it are more demanding.

An LNP that looks identical to a passing batch in a visual inspection can behave very differently in vivo if its particle size distribution has shifted, its encapsulation efficiency has dropped, or its surface charge has changed. The four critical quality attributes discussed in this article, particle size, polydispersity index (PDI), encapsulation efficiency, and zeta potential, are the core measurements that define the identity and performance of an LNP product, and they must be reliably measured and controlled throughout development and into GMP manufacturing.

The 4 Critical Quality Attributes

CQA	What It Measures	Typical Target Range	Why It Matters
Particle size (Z-average)	Mean hydrodynamic diameter of the particle population, measured by DLS	50-200 nm for most therapeutic applications	Size determines circulation time, tissue distribution, and cellular uptake. Particles above 200 nm clear rapidly; below 50 nm may pass through kidney filtration.
Polydispersity index (PDI)	Width of the particle size distribution (0 = monodisperse; 1 = polydisperse)	Below 0.2 for clinical LNP products	High PDI indicates a heterogeneous population. Broader distributions correlate with variable in vivo performance and reduced reproducibility between batches.
Encapsulation efficiency (EE%)	Proportion of the nucleic acid or small molecule payload that is encapsulated within the LNP rather than free in solution	Greater than 85% for mRNA LNPs; target depends on payload and indication	Free (unencapsulated) payload is bioavailable but unprotected, degrades rapidly, and contributes to off-target effects. Low EE% reduces effective dose and increases immunostimulation risk for nucleic acid payloads.
Zeta potential	Net surface charge of the particle at the slipping plane, measured by electrophoretic light scattering	Near-neutral for ionisable LNPs in physiological conditions; negative in buffer (typically minus 5 to minus 20 mV)	Surface charge affects colloidal stability, protein adsorption (corona formation), and cellular uptake. Highly charged particles aggregate more readily and may trigger immune activation.

Target ranges above are representative of typical clinical-stage LNP products based on published literature. Programme-specific specifications must be established based on the molecule and intended clinical use.

Measuring Particle Size and PDI: Dynamic Light Scattering

Dynamic light scattering (DLS) is the standard technique for measuring LNP particle size and PDI. DLS measures the Brownian motion of particles in suspension and uses the diffusion coefficient to calculate the hydrodynamic diameter via the Stokes-Einstein equation. It is fast, non-destructive, and highly sensitive to small shifts in particle size, making it well suited for both development screening and GMP release testing.

DLS has important limitations. It is an intensity-weighted measurement and is therefore disproportionately sensitive to larger particles, which scatter more light. A small number of aggregates can significantly shift the reported Z-average upward even when the bulk of the population is within specification. For LNP batches where aggregation is a concern, complementary techniques such as nanoparticle tracking analysis (NTA) or asymmetric flow field-flow fractionation (AF4) provide a number-weighted or separated view of the size distribution.

Measuring Encapsulation Efficiency: The RiboGreen Assay

For nucleic acid payloads including mRNA and siRNA, encapsulation efficiency is most commonly measured using the RiboGreen fluorescence assay. The assay compares the fluorescence signal from the nucleic acid in an intact LNP sample (where encapsulated nucleic acid is inaccessible to the dye) with the signal from a sample disrupted with a detergent such as Triton X-100 (where all nucleic acid is accessible). The ratio of the two signals gives the encapsulation efficiency.

For small molecule payloads, encapsulation efficiency is typically measured by separating the encapsulated and free fractions using size exclusion chromatography or ultracentrifugation and then quantifying each fraction by LC-UV or LC-MS/MS.

Integrating CQA Measurement into the Development and GMP Lifecycle

Development Phase

During formulation development, CQA measurements are used to screen formulation variables including lipid ratios, drug-to-lipid ratios, and processing conditions. A design of experiments (DoE) approach linking formulation parameters to CQA outcomes allows the formulation space to be explored efficiently and a robust formulation to be identified before GMP scale-up.

GMP Manufacturing

At GMP manufacturing scale, CQA measurements are used as in-process controls during manufacture and as release tests that every batch must pass before it can be used in a clinical study. The specification ranges for each CQA are derived from the development data, with limits set to ensure that batches released for clinical use are within the range of material that has demonstrated acceptable in vivo performance.

LNP Characterisation at Ardena Oss

Ardena’s GMP nanomedicine facility in Oss is equipped with DLS instruments, RiboGreen assay capability, and the analytical infrastructure needed to characterise LNP products across all four critical quality attributes during development and GMP manufacturing. The analytical team works alongside the formulation scientists to integrate CQA monitoring into the development workflow from early screening batches through to GMP release testing.

The Question Nobody Asks Early Enough

When you are developing a nanoparticle drug product, the bioanalytical question you need to answer is not simply ‘how much drug is in the blood?’ It is ‘which form of the drug are you measuring, and what does that tell you about what is actually happening in the patient?’

For a conventional small molecule, total plasma drug concentration is a reliable surrogate for the concentration available to interact with the target. For a nanoparticle product, the drug in the plasma exists in at least two distinct populations: drug encapsulated within intact nanoparticles (not directly bioavailable until the particle releases it), and free drug (the fraction that has been released and is available for cellular uptake or systemic distribution). These two fractions have different pharmacokinetic profiles, different safety implications, and different relationships to efficacy. Measuring only the total gives you a misleading picture of both.

The Three Fractions and What They Tell You

Analyte Fraction	What It Represents	Clinical Relevance	Measurement Approach
Total drug (encapsulated + free)	Everything in the sample, regardless of form	Overall systemic exposure; comparison to free drug PK for benefit of encapsulation	Extract with detergent or organic solvent before analysis; LC-MS/MS or validated immunoassay
Free (unencapsulated) drug	Drug released from the particle in circulation	Off-target toxicity risk; available for direct cellular uptake; clearance profile	Separate free fraction by ultrafiltration or SEC before extraction; subtract from total
Encapsulated drug	Drug retained within intact nanoparticles	Depot of drug still in delivery system; circulation half-life of the carrier	Calculated as total minus free; or direct measurement after nanoparticle enrichment

How to Separate the Fractions

Ultrafiltration

Centrifugal ultrafiltration is the most commonly used approach for separating free drug from encapsulated drug. The sample is centrifuged through a membrane with a molecular weight cut-off appropriate to retain the nanoparticles while allowing free drug to pass into the filtrate. The free drug is then quantified in the filtrate, and total drug is measured in a separately extracted aliquot of the unprocessed sample.

The main limitation of ultrafiltration is the potential for non-specific binding of the drug to the membrane, which can lead to underestimation of the free fraction. Membrane compatibility must be assessed during method development using spiked samples at relevant concentrations.

Size Exclusion Chromatography

Size exclusion chromatography (SEC) physically separates nanoparticles from free drug based on size. Nanoparticles elute in the void volume; free drug elutes later. Fractions can be collected and analysed separately. SEC is gentler than ultrafiltration for fragile nanoparticle formulations and avoids membrane binding artefacts, but is lower throughput and requires careful method development to ensure complete separation.

Matrix-Specific Challenges for Nanoparticle PK Assays

Nanoparticle products interact with blood components in ways that complicate sample handling. Plasma proteins adsorb to nanoparticle surfaces to form a protein corona within seconds of contact with blood, changing the particle’s surface properties and, in some assays, interfering with the detection of the encapsulated payload. Lipid nanoparticles can fuse with endogenous lipoproteins in plasma, causing transfer of lipid payload and apparent changes in the free-to-encapsulated ratio that are artefacts of the in vitro processing rather than the in vivo behaviour.

These matrix interactions must be characterised during method development and controlled by standardised sample collection and processing procedures. Samples for nanoparticle PK analysis should be processed promptly after collection, and freeze-thaw stability of the sample before the separation step must be validated to confirm that particle integrity is maintained under study storage conditions.

Regulatory Expectations for Nanoparticle PK Bioanalysis

The FDA and EMA have not yet published dedicated guidance specifically on bioanalytical methods for nanoparticle drug products, but the general principles of ICH M10 apply. Sponsors developing nanoparticle products are expected to define the analyte clearly in the validation protocol, to justify the choice of measurement (total versus free versus encapsulated), and to demonstrate that the sample preparation procedure does not introduce systematic bias through nanoparticle disruption or membrane binding. FDA review comments on approved nanoparticle products provide useful precedent for the level of characterisation expected.

Ardena’s Nanoparticle Payload Bioanalysis at Assen

Ardena’s bioanalytical team in Assen develops and validates methods for nanoparticle payload quantification using ultrafiltration and SEC-based fractionation combined with LC-MS/MS detection. The team has experience with liposomal, LNP, and polymeric nanoparticle products, and can advise on bioanalytical strategy design to ensure the PK data package supports both dose selection decisions and the regulatory filing.

The Role of PK Data in Early Clinical Development

Pharmacokinetic data is the foundation on which dose selection decisions are built in early clinical development. Before the first human dose is administered, the preclinical PK profile provides predictions of human exposure based on allometric scaling or physiologically based pharmacokinetic (PBPK) modelling. After the first dose in humans, that prediction is tested against reality, and the clinical PK data that emerges drives every subsequent dosing decision.

For a single ascending dose (SAD) study, PK data from each cohort determines whether escalation to the next dose level is appropriate. For a multiple ascending dose (MAD) study, trough concentrations at steady state confirm whether the dosing interval produces adequate drug exposure between doses. For a food effect study, the comparison of AUC and Cmax in fed and fasted states informs whether the drug needs to be taken with or without food. In each case, the quality of the bioanalytical data directly limits the quality of the clinical decision.

Building the Analytical Method for a New Molecule

Selectivity and Matrix Selection

The first task in developing a bioanalytical method for a new small molecule is establishing selectivity: can the method measure the drug accurately in the presence of the endogenous components of the biological matrix? For most plasma or serum methods, selectivity is demonstrated by showing that the analyte signal in drug-free matrix samples from at least six individual donors is within 20% of the lower limit of quantification (LLOQ), confirming that matrix components do not produce a false signal at the analyte’s retention time and mass transition.

Sensitivity and LLOQ

The LLOQ defines the lowest concentration that can be measured with acceptable precision and accuracy, typically defined as a coefficient of variation below 20% and a bias within plus or minus 20% of the nominal value. Setting the LLOQ appropriately requires an estimate of the lowest plasma concentration expected in the clinical study, which for a SAD study is typically the Cmax at the lowest dose level adjusted for the expected elimination over the last sampling timepoint. For molecules with very long half-lives or very wide dose ranges, the LLOQ requirement may span several orders of magnitude.

Metabolite Coverage

For many small molecules, metabolites are present in plasma at concentrations that can be clinically significant, either because the metabolite is pharmacologically active or because it is associated with toxicity. Regulatory guidance, including the FDA’s 2020 guidance on safety testing of drug metabolites, requires that metabolites present at greater than 10% of parent drug exposure are characterised and assessed for safety. The bioanalytical method must be capable of measuring the relevant metabolites in addition to the parent compound if their concentrations are expected to be clinically relevant.

LC-MS/MS: The Gold Standard for Small Molecule PK

Liquid chromatography tandem mass spectrometry (LC-MS/MS) is the primary analytical platform for quantitative small molecule bioanalysis in regulatory studies. The combination of chromatographic separation with mass selective detection provides the specificity needed to measure a drug and its metabolites at nanomolar or sub-nanomolar concentrations in complex biological matrices, with precision and accuracy that consistently meets ICH M10 acceptance criteria across a wide range of analytes.

Key method development parameters for an LC-MS/MS assay include the choice of ionisation mode (positive or negative electrospray ionisation based on the molecule’s ionisable groups), the chromatographic conditions (stationary phase, mobile phase composition, gradient profile), the mass transitions monitored (typically the precursor ion to at least two fragment ions for confirmation of identity), and the sample preparation approach (protein precipitation, liquid-liquid extraction, or solid-phase extraction based on the required sensitivity and selectivity).

From Single Dose to Steady State: What the PK Data Tells You

PK Study Phase	Key Measurements	Clinical Decision Supported
Single ascending dose (SAD)	Cmax, AUC0-t, AUC0-inf, t1/2, Tmax at each dose level	Dose proportionality; human half-life; dose selection for MAD study
Multiple ascending dose (MAD)	Cmin (trough) at steady state; AUCss; accumulation ratio	Dosing interval confirmation; time to steady state; accumulation characterisation
Food effect study	AUC and Cmax in fed vs fasted state; 90% CI for the ratio	Fasting or fed administration instruction; effect of food on variability
Special population PK	Clearance and exposure in renally or hepatically impaired subjects	Dose adjustment recommendations for labelling
Drug-drug interaction studies	PK of victim drug with and without perpetrator	Labelling of clinically relevant DDIs; dose adjustment requirements

Sample Management and Chain of Custody

The quality of bioanalytical data depends not only on the analytical method but on the handling of the samples from the moment of collection to the moment of analysis. Blood must be collected into the correct anticoagulant tube, processed to plasma or serum within the validated time window, aliquoted correctly, and stored at the validated temperature. Deviations from the validated sample handling procedure can result in analyte degradation that produces systematically low concentrations, invalidating the data.

Ardena’s clinical team at Assen provides sample management services that coordinate sample receipt from clinical sites, log chain of custody, and confirm that sample handling conditions were within validated parameters before analysis proceeds. Samples that have experienced documented excursions are flagged for scientific review before inclusion in the dataset.

Ardena’s Small Molecule PK Bioanalysis at Assen

Ardena’s bioanalytical laboratory in Assen provides fully validated LC-MS/MS methods for small molecule PK studies from early non-clinical through to Phase III. The laboratory operates under GLP and GCP conditions for regulated studies, with ICH M10 compliant validation packages and ISR programmes as standard. The team has experience with plasma, serum, urine, cerebrospinal fluid, and tissue matrices across a wide range of therapeutic areas.

Why Bioequivalence Studies Depend on Exceptional Analytical Quality

A bioequivalence (BE) study answers one specific question: does a test formulation, whether a generic product, a reformulated branded product, or a product manufactured at a new site, deliver the same drug exposure to the body as the reference formulation? The answer is expressed in terms of the 90% confidence interval for the geometric mean ratio of the primary PK parameters, AUC and Cmax, between the test and reference. For a BE study to be accepted, both confidence intervals must fall within the 80 to 125% acceptance range.

The regulatory consequence of this criterion is that the bioanalytical method used in a BE study must be exceptionally precise and accurate. The confidence interval calculation is sensitive to variability in the PK data, and a significant proportion of that variability is analytical in origin. A method with poor precision inflates the variability of the concentration-time data, widens the confidence interval, and increases the risk of a study failing to demonstrate BE even when the formulations are truly equivalent.

How BE Bioanalytical Requirements Differ from Early-Phase Clinical Work

Requirement	Early-Phase Clinical Bioanalysis	Bioequivalence Bioanalysis
Validation standard	ICH M10; fit-for-purpose acceptable for exploratory endpoints	Full ICH M10 validation required; no fit-for-purpose concession
Precision (CV)	15% across calibration range; 20% at LLOQ	Same thresholds; but BE study success is highly sensitive to precision at Cmax concentration range
Incurred sample reanalysis (ISR)	Required; minimum 10% of study samples	Required; particular scrutiny on ISR pass rate in BE submissions
Selectivity	Tested in representative matrix samples	Must demonstrate selectivity from endogenous compounds at relevant concentrations; critical for low-dose products
Stability	Bench-top, freeze-thaw, long-term frozen	As per early-phase plus specific validation of stability under study-specific conditions
Regulatory scrutiny	Moderate; method validation report reviewed at submission	High; FDA and EMA actively flag BE studies with bioanalytical issues

Incurred Sample Reanalysis in BE Studies

Incurred sample reanalysis (ISR) is a key quality indicator for bioanalytical data in BE studies. ISR involves re-analysing a subset of study samples, typically 10% or more of the total, from re-thawed aliquots. The results of the re-analysis must agree with the original analysis within 20% for at least two-thirds of the re-analysed samples. ISR failures, where the agreement between original and repeat results is poor, indicate instability of the analyte under the study conditions, assay imprecision, or matrix effects that were not adequately controlled during validation.

The FDA’s guidance on bioanalytical method validation and the EMA’s guideline on bioanalytical method validation both address ISR requirements in detail, and a poor ISR outcome in a BE study submission will typically result in a regulatory query or a request to repeat the study.

Special Considerations for Specific BE Scenarios

Highly Variable Drugs

Some drugs show high intrasubject variability in PK parameters, typically defined as an intrasubject coefficient of variation greater than 30% for AUC or Cmax. For these drugs, demonstrating BE within the standard 80 to 125% acceptance window requires an extremely large study population. Regulatory agencies including the FDA and EMA have provisions for scaled average bioequivalence approaches for highly variable drugs, which widen the acceptance window in proportion to the reference drug’s variability. These approaches require specific study designs and statistical analyses and are not applicable in all circumstances.

Narrow Therapeutic Index Drugs

For drugs with a narrow therapeutic index, where small differences in exposure can have significant clinical consequences, the standard 80 to 125% acceptance range is tightened. The FDA’s guidance on narrow therapeutic index drugs requires a 90 to 111.11% acceptance criterion for certain drugs, which significantly increases the analytical and statistical rigour required to demonstrate BE successfully.

Endogenous Compounds

Some drugs are endogenous compounds or closely related to endogenous compounds that are present at measurable concentrations in the biological matrix. For these analytes, the background level of the endogenous compound must be characterised and subtracted, and the assay must be validated to demonstrate that it can accurately measure the drug above this background. Biotin, melatonin, and certain amino acid derivatives all present this challenge.

Ardena’s Bioequivalence Bioanalysis Services

Ardena’s bioanalytical facility in Assen provides fully validated LC-MS/MS and ligand-binding assay methods for BE studies, with full ICH M10 validation packages and ISR programmes as standard. The laboratory operates under GLP conditions for regulated bioanalysis, and study reports are prepared to the format required for FDA, EMA, and other major regulatory agency submissions.

Ardena’s experience spans both reference-listed drug BE studies for generic development and formulation bridging BE studies for branded pharmaceutical programmes. The team works closely with clinical pharmacology partners to ensure that the bioanalytical method design is optimised for the specific concentration ranges and matrices relevant to the study, and that the data quality is sufficient to support the regulatory submission.

Why Stability Data Is Non-Negotiable

Every regulatory filing for a drug substance or drug product, from the earliest IND or IMPD through to a marketing authorisation application, requires stability data. That data answers one fundamental question: for how long can the product be stored under defined conditions and remain within specification? The answer determines the retest period for the drug substance, the shelf life of the drug product, and the conditions under which both must be stored and shipped.

Building a stability programme that generates the right data, at the right time, to support each regulatory filing is a practical challenge that requires both scientific rigour and careful project planning. Getting it wrong means either filing with insufficient data and receiving a regulatory query, or generating redundant data that consumes resources without adding information.

The ICH Q1 Framework

The ICH Q1A(R2) guideline on stability testing of new drug substances and products is the primary regulatory framework for pharmaceutical stability programmes. It defines the storage conditions for long-term, intermediate, and accelerated stability studies based on the intended climatic zone of the market, the required study duration, and the minimum number of batches that must be included in a registration stability package.

Study Type	Condition	Intended Duration	Primary Purpose
Long-term (Zone I/II)	25 degrees C / 60% RH	12 months minimum for registration; typically to proposed shelf life	Primary basis for shelf life assignment
Long-term (Zone IVb)	30 degrees C / 65% RH or 40 degrees C / 75% RH	12 months minimum for registration	Required for markets in climatic Zone IVb (e.g. India, parts of Africa)
Intermediate	30 degrees C / 65% RH	6 months minimum	Required when significant change observed at accelerated condition
Accelerated	40 degrees C / 75% RH	6 months minimum	Supports shelf life prediction; early stability screening
Refrigerated products	5 degrees C +/- 3 degrees C (long-term)	12 months minimum	For products stored at 2-8 degrees C
Frozen products	Minus 20 degrees C (long-term)	Programme defined by product type	For frozen drug substances and biologics

Real-Time vs. Accelerated Stability: What Each Can and Cannot Tell You

Real-Time Stability

Real-time stability studies store samples under the intended long-term storage conditions and test them at defined intervals to assess whether the product remains within its specification over time. Real-time data is the definitive basis for shelf life assignment: a product can only be assigned a shelf life supported by real-time data collected at or beyond the proposed expiry date.

The limitation of real-time data is obvious: it takes time. A product intended to have a two-year shelf life requires two years of real-time data before that claim can be fully supported. For early clinical filings, extrapolation from available real-time data combined with accelerated data is typically accepted, but the real-time data must eventually be generated.

Accelerated Stability Studies

Accelerated stability studies store samples at elevated temperature and humidity conditions, typically 40 degrees Celsius at 75% relative humidity, to accelerate the rate of chemical and physical degradation. The results allow scientists to predict long-term stability behaviour based on the degradation kinetics observed under stress conditions, using the Arrhenius relationship between temperature and reaction rate.

Accelerated data is valuable for comparing formulation options quickly, for generating supportive data for early clinical filings, and for predicting the degradation pathways that will need to be monitored in the long-term programme. However, accelerated data cannot substitute for real-time data in a registration-stage shelf life claim, and for some degradation mechanisms, particularly physical instability in amorphous formulations or moisture-driven hydrolysis, the Arrhenius relationship does not hold.

Stress Testing: Beyond ICH Stability

ICH Q1A(R2) stability studies are designed to generate the data needed for regulatory filings. Stress testing, as described in the related ICH Q1B guideline on photostability and in general scientific practice, goes further: it exposes the product to extreme conditions, including high temperature, oxidation, acid and base hydrolysis, and intense light, to force degradation and identify the degradation products that will appear in real-time stability samples.

Stress testing is typically conducted on the drug substance before formulation work begins, and again on the drug product once the formulation is established. The degradation products identified in stress studies inform the choice of stability-indicating analytical methods, the excipient compatibility assessment, and the packaging selection. A formulation that fails rapidly under oxidative stress conditions is not suitable for storage in a packaging system that allows oxygen ingress.

Stability Testing Across Ardena’s Network

Ardena operates GMP stability storage facilities across its network of sites, including temperature-controlled chambers and freezers covering the ICH long-term, intermediate, and accelerated conditions for ambient, refrigerated, and frozen products. Stability samples are managed under a formal stability protocol with a documented pull schedule, and analytical testing is conducted by the relevant site’s analytical team using validated stability-indicating methods.

For programmes that require stability storage across multiple product types or at multiple sites, Ardena’s project management structure ensures that the stability programme is tracked centrally and that any out-of-trend or out-of-specification stability results are escalated promptly.

Why Concentration Data Alone Is Not Enough

A bioanalytical study produces concentration-time data: measurements of drug (and sometimes metabolite) levels in biological matrices at defined timepoints. That data is necessary, but on its own it does not answer the questions that matter for drug development. How does exposure change with dose? What is the relationship between exposure and effect? What dose is needed to achieve the target concentration in the intended patient population? How often does the drug need to be dosed to maintain effective concentrations?

Pharmacokinetic (PK) and pharmacodynamic (PD) modelling provides the framework for answering those questions. PK modelling describes how the body handles the drug over time: absorption, distribution, metabolism, and excretion. PD modelling describes the relationship between drug exposure and biological effect. Together, PK/PD modelling transforms concentration data into the quantitative understanding needed to make dose selection and clinical trial design decisions with confidence.

The PK Parameters That Drive Drug Development Decisions

PK Parameter	Definition	Why It Matters
AUC (area under the curve)	Total drug exposure over a defined time period	Primary measure of systemic exposure; used to assess dose proportionality and accumulation
Cmax	Peak plasma concentration after a dose	Related to acute tolerability and, for some drugs, to efficacy and toxicity thresholds
Tmax	Time to peak concentration	Reflects rate of absorption; relevant to onset of effect
t1/2 (elimination half-life)	Time for drug concentration to halve during the elimination phase	Determines dosing interval and time to steady state (approximately 5 x t1/2)
CL (clearance)	Volume of plasma cleared of drug per unit time	Determines the dose required to achieve a target steady-state exposure
Vd (volume of distribution)	Apparent volume in which the drug distributes	High Vd indicates extensive tissue distribution; affects loading dose requirements
Bioavailability (F)	Fraction of dose reaching systemic circulation	Critical for oral and other non-intravenous routes; determines dose needed for a given exposure target

Non-Compartmental vs. Compartmental Analysis

Non-Compartmental Analysis (NCA)

Non-compartmental analysis is the standard approach for summarising PK data from individual studies. It makes no assumptions about the underlying PK model and derives parameters such as AUC, Cmax, t1/2, and clearance directly from the observed concentration-time data using trapezoidal integration and regression. NCA is used to characterise PK in each arm of a clinical study and to assess dose proportionality across dose levels.

Population PK Modelling

Population PK modelling uses a mixed-effects approach to simultaneously analyse PK data from all subjects in a study, or across multiple studies, and to identify covariates, such as body weight, renal function, or hepatic function, that explain variability in drug exposure between individuals. Population PK models are increasingly required as part of the regulatory submission package. The FDA’s guidance on population pharmacokinetics describes the expectations for population PK analysis in NDA and BLA submissions.

PK/PD Modelling for Dose Selection

PK/PD modelling links drug exposure to a pharmacodynamic effect, whether a biomarker response, a clinical outcome, or a safety signal, through a mathematical relationship. The most commonly used models for concentration-effect relationships are the Emax model and its variants, which describe the sigmoidal relationship between drug concentration and effect with parameters including the maximum effect (Emax), the concentration producing 50% of maximum effect (EC50), and the Hill coefficient describing the steepness of the relationship.

Building a PK Study Package That Answers Regulatory Questions

A well-structured PK programme in early clinical development generates the data needed to make dose selection decisions for subsequent studies and to address the PK questions that will arise in regulatory review. Key components of an early clinical PK package include a single ascending dose study, a multiple ascending dose study, and often a food effect study for oral products.

The FDA’s M3(R2) guideline on non-clinical safety studies and the EMA’s guidance on first-in-human studies both address the PK characterisation expected before clinical dose escalation, and the FDA’s clinical pharmacology guidance documents set out the expectations for the PK data package supporting NDA submissions. Designing a study programme with the regulatory endpoint in mind from the outset avoids the need for additional studies to fill data gaps at the time of submission.

Ardena’s PK/PD Services at Assen

Ardena’s bioanalytical team in Assen provides the sample analysis services that generate the concentration-time data underpinning PK/PD programmes, using validated LC-MS/MS and ligand-binding assay methods. The team supports NCA reporting and works with clinical pharmacology partners to provide the bioanalytical component of population PK and PK/PD analyses.

Integrated PK and immunogenicity data from the same study is a particular capability: understanding whether elevated or reduced drug exposure in a subset of subjects is linked to ADA development is a question that requires both datasets to be available and interpreted in parallel.

What Flow Cytometry Offers That Other Techniques Cannot

Flow cytometry measures multiple physical and chemical characteristics of individual cells as they pass, one by one, through a laser beam. Each cell scatters light in a pattern determined by its size and internal complexity, and emits fluorescence from labelled antibodies or other probes bound to specific markers on its surface or inside the cell. A modern flow cytometer can measure ten, fifteen, or more parameters simultaneously on each cell, generating a dataset that describes the phenotype and functional state of thousands of individual cells in a single sample.

For clinical trials that involve the immune system, this single-cell resolution is what makes flow cytometry irreplaceable. Bulk methods such as ELISA or gene expression arrays describe the average behaviour of a cell population. Flow cytometry resolves that population into its constituent subsets, revealing changes in the relative proportions of T cell subsets, NK cell activation states, or myeloid cell phenotypes that would be invisible in aggregate data.

Clinical Applications of Flow Cytometry

Immunophenotyping for Immunotherapy Trials

Immune checkpoint inhibitors, CAR-T cell therapies, bispecific antibodies, and other immunotherapy modalities all act by modifying the immune system. Understanding how they change the composition and activation state of immune cell populations in peripheral blood or tumour tissue is essential for interpreting clinical responses and adverse events. Flow cytometry panels for immunotherapy trials typically characterise T cell subsets including CD4, CD8, regulatory T cells, and exhaustion markers, as well as NK cells and myeloid populations.

Pharmacodynamic Monitoring

Flow cytometry is widely used to measure the pharmacodynamic effects of biologic drugs that target immune cell populations. For a drug targeting CD20-positive B cells, flow cytometry provides direct evidence of B cell depletion in peripheral blood. For a drug intended to expand a specific T cell population, multi-parametric immunophenotyping demonstrates the intended pharmacological effect and supports dose selection.

Minimal Residual Disease (MRD) Assessment

In haematological malignancies including leukaemia and multiple myeloma, flow cytometry is used to detect residual tumour cells at levels below the threshold of morphological assessment. Multi-parametric MRD panels using eight or more markers can detect one tumour cell in ten thousand or more normal cells, providing a sensitive endpoint for assessing depth of response to treatment.

Building a Validated Multi-Parametric Panel

Development Step	Purpose	Key Considerations
Panel design	Select fluorochrome-antibody combinations that minimise spectral overlap and maximise signal resolution	Use brightest fluorochromes for low-density targets; apply compensation controls for every fluorochrome in the panel
Titration optimisation	Determine optimal antibody concentration for each reagent	Under-titration loses signal; over-titration increases background; titrate in the intended matrix
Specificity testing	Confirm each antibody detects the intended target	Use positive and negative control cell populations with known phenotype
Sensitivity / LLOQ	Determine the lowest detectable frequency of positive cells	Critical for rare cell populations and MRD applications
Inter-operator and inter-instrument precision	Demonstrate reproducibility across analysts and instruments	Use standardised bead-based calibration; include reference samples across runs
Fit-for-purpose validation	Demonstrate panel performance meets requirements for intended use	Scope determined by data criticality; follow EuroFlow or ISAC guidelines as appropriate

Practical Considerations for Clinical Sample Handling

Flow cytometry results are sensitive to pre-analytical variables including time from blood collection to processing, storage temperature, and the use of anticoagulants. Whole blood samples for immunophenotyping are typically processed within four to six hours of collection, using lyse-no-wash protocols that minimise cell activation and loss. For clinical trials where samples are collected at remote sites and shipped to a central laboratory, the pre-analytical handling conditions must be validated to demonstrate that the analytical results are not affected by the transport time and conditions.

Ardena’s clinical team at Assen works with clinical operations teams to design sample handling procedures that are practical for site staff while ensuring data quality at the central laboratory. Stability data supporting the validated shipping conditions is documented and available for regulatory review.

Ardena’s Flow Cytometry Capabilities

Ardena’s flow cytometry laboratory in Assen operates multi-laser instruments capable of panels up to fifteen or more parameters, with dedicated capacity for clinical trial sample analysis. The team develops and validates multi-parametric immunophenotyping panels, MRD panels, and functional assays including intracellular cytokine staining and proliferation assays.

Flow cytometry services at Ardena are integrated with the wider bioanalytical platform, allowing immune cell data to be interpreted alongside PK, PD, and immunogenicity data from the same study, providing the multi-dimensional dataset that characterises the immune pharmacology of complex therapeutics.

Biomarkers Are No Longer Optional in Drug Development

A decade ago, biomarkers were an add-on to many clinical development programmes: interesting, potentially informative, but rarely central to regulatory decision-making. That has changed. The growth of precision medicine, the expansion of companion diagnostic requirements, and the increasing use of biomarker-defined patient populations in oncology and rare disease trials have made biomarker analysis a core element of the clinical development package for many new drugs.

Regulatory agencies have responded. The FDA’s biomarker qualification programme, the EMA’s qualification of novel methodologies procedures, and the extensive biomarker content of ICH E16 all reflect an expectation that biomarker data generated in clinical trials is produced to a level of analytical rigour appropriate for the decisions it will support.

Qualification vs. Validation: Understanding the Distinction

The terms biomarker qualification and biomarker validation are used inconsistently in the literature and in industry practice. The most widely used framework, from the FDA’s 2018 biomarker qualification guidance, distinguishes between the two as follows:

Concept	Definition	When It Applies
Biomarker qualification	A regulatory conclusion that a biomarker can be relied upon to have a specific interpretation in a specific context of use	When a biomarker is intended to support regulatory decisions across multiple drug development programmes
Biomarker assay validation (fit-for-purpose)	Demonstration that an assay is suitable for its intended use in the specific programme	The standard for most clinical trial biomarker assays; scope of validation determined by context of use
Full analytical validation	Complete validation to the standards applied to PK assays under ICH M10	Required when biomarker data is used for primary regulatory endpoints or safety decisions

For most clinical biomarker assays, a fit-for-purpose validation, with the scope of experiments calibrated to the criticality of the data, is the appropriate standard. A biomarker used to support patient stratification decisions in a pivotal Phase III trial requires more rigorous validation than the same biomarker used for exploratory characterisation in a Phase I study.

Categories of Clinical Biomarkers

Pharmacodynamic Biomarkers

Pharmacodynamic biomarkers measure biological changes that occur as a direct consequence of drug exposure. They provide evidence that the drug is engaging its intended target and producing the expected biological response. In early clinical trials, PD biomarkers are used to establish proof of mechanism, select doses for later studies, and define the dosing schedule that achieves the desired level of target engagement.

Predictive Biomarkers

Predictive biomarkers identify patient populations that are more or less likely to respond to a specific treatment. They are the foundation of precision medicine. The most commercially significant predictive biomarkers, such as HER2 amplification in breast cancer or EGFR mutation status in lung cancer, are used to define the registered indication and are assessed as companion diagnostics subject to their own regulatory requirements.

Safety Biomarkers

Safety biomarkers detect the early onset of drug-induced organ toxicity or other adverse effects. Established safety biomarkers such as alanine aminotransferase for hepatotoxicity and serum creatinine for nephrotoxicity have been used in clinical monitoring for decades. Novel safety biomarkers, such as kidney injury molecule-1 (KIM-1) for renal tubular injury, are being increasingly incorporated into clinical monitoring programmes as their qualification evidence accumulates.

Prognostic Biomarkers

Prognostic biomarkers predict the natural course of disease in the absence of treatment. They help in the design of clinical trials by identifying patient populations with the appropriate rate of clinical events to power efficacy endpoints, and in interpreting trial results by accounting for baseline differences between treatment arms.

Key Assay Performance Parameters for Biomarker Validation

Parameter	Definition	Typical Approach for Fit-for-Purpose Validation
Precision	Reproducibility of repeated measurements	Within-run and between-run precision at low, mid, and high concentrations
Accuracy / Trueness	Agreement of measured value with true concentration	Recovery assessment using spiked samples or reference materials
Sensitivity (LLOQ)	Lowest concentration measurable with defined precision and accuracy	Determined during assay development; at least 5x lower than lowest expected study sample
Selectivity	Ability to distinguish analyte from matrix components	Tested in representative matrices including haemolysed and lipaemic samples
Stability	Analyte stability under relevant conditions	Bench-top, freeze-thaw, long-term frozen; specific to the matrix and storage conditions
Dilutional linearity	Proportional response upon sample dilution	Important for high-concentration samples exceeding the calibration range

Ardena’s Biomarker Services at Assen

Ardena’s bioanalytical centre in Assen provides biomarker assay development, qualification, and fit-for-purpose validation services using MSD electrochemiluminescence, ELISA, flow cytometry, and qPCR platforms. The team has experience with pharmacodynamic, safety, and predictive biomarker assays across oncology, immunology, and rare disease programmes.

Ardena takes a context-of-use approach to biomarker validation, working with clients to define the appropriate scope of assay performance experiments based on the intended use of the data, the phase of development, and the regulatory context. This avoids over-validation of exploratory assays while ensuring that decision-critical biomarker data meets the standards required.

Why Immunogenicity Matters for Biologic Drugs

Biologic drugs, including monoclonal antibodies, fusion proteins, ADCs, and gene therapy vectors, are structurally complex molecules derived from biological systems. When administered to patients, they can trigger an immune response. The body may produce antibodies directed against the therapeutic molecule itself: anti-drug antibodies, or ADAs.

The clinical consequences of ADA development range from inconsequential to serious. At one end of the spectrum, low-titre ADAs may have no effect on drug pharmacokinetics or efficacy. At the other end, high-titre neutralising ADAs can accelerate drug clearance, abolish efficacy, and, in rare cases, trigger severe hypersensitivity reactions or cross-react with an endogenous protein. Regulators require a systematic immunogenicity assessment for all biologic drug programmes, and the data package must be in place before Phase I dosing begins.

The Tiered Immunogenicity Testing Strategy

The regulatory-recommended approach to immunogenicity testing uses a tiered strategy designed to balance sensitivity and specificity efficiently. The FDA’s guidance on immunogenicity testing for therapeutic protein products and the EMA’s guideline on immunogenicity assessment of biotechnology-derived therapeutic proteins both describe this framework, and the ICH M10 bioanalytical method validation guideline provides the validation requirements applicable to immunogenicity assays.

Tier	Assay Purpose	Acceptance Criterion	What Happens Next
Tier 1: Screening	Detect all potentially ADA-positive samples with high sensitivity	Cut-point set at approximately 5% false positive rate	Positive samples proceed to Tier 2
Tier 2: Confirmatory	Confirm true positives by drug competition	Typically greater than 20-25% inhibition signals confirmation	Confirmed positives proceed to Tier 3
Tier 3: Titration	Quantify ADA titre in confirmed positive samples	Serial dilution to endpoint	High-titre samples may proceed to neutralisation assay
Tier 4: Neutralisation	Determine whether ADAs block drug activity	Cell-based or competitive ligand-binding assay	Neutralising ADA data informs clinical risk assessment

Cut-Point Setting: The Statistical Foundation of Immunogenicity Assays

The screening cut-point is the signal threshold above which a sample is classified as potentially ADA-positive. Setting this threshold correctly is critical: too low and you generate an unmanageable number of false positives that consume confirmatory assay capacity; too high and you miss true positive samples.

Cut-point setting uses a statistical approach based on the distribution of signal responses in a panel of drug-naive individuals from the target population. Typically a parametric or non-parametric approach is used to set the cut-point at a level corresponding to approximately a 5% false positive rate. The cut-point must be evaluated separately for each matrix used in the study, and a normalised or floating cut-point approach is commonly used to account for plate-to-plate variation.

Drug Tolerance: The Biggest Technical Challenge in ADA Assays

Drug tolerance refers to the ability of an immunogenicity assay to detect ADAs in the presence of circulating drug. This is a fundamental challenge because clinical samples collected during a dosing study will contain the therapeutic molecule, which can bind to any ADAs in the sample and prevent them from being detected by the assay. An assay with poor drug tolerance will produce false negative results in samples collected near Tmax, leading to underestimation of the true incidence of immunogenicity.

Strategies for improving drug tolerance include acid dissociation pre-treatment of samples to disrupt drug-ADA complexes before analysis, the use of assay formats with inherently higher drug tolerance such as bridging ELISA configurations, and the optimisation of sample dilution and blocking strategies. Drug tolerance is one of the key parameters evaluated during immunogenicity assay validation.

Immunogenicity Programme Design Considerations

Sample Collection Timing

Immunogenicity samples must be collected at pre-dose and at defined intervals throughout and after the dosing period. The timing of samples should reflect the expected time course of ADA induction: too few samples and you may miss the peak of the immune response; too many and you impose unnecessary burden on the clinical site and the patient.

Matrix Selection

Most immunogenicity assays use serum or plasma as the primary matrix. The choice between serum and plasma can affect assay performance because of differences in protein composition and the presence of clotting factors. The matrix used for assay development and validation must match the matrix collected in the clinical study.

Risk-Based Approach to Neutralisation Testing

Not all ADC or biologic programmes require a full four-tier immunogenicity programme including cell-based neutralisation assays. A risk-based approach, considering the target, the patient population, the dosing regimen, and the clinical consequences of neutralising ADAs, is used to determine which tiers are required and when neutralisation data is needed.

Ardena’s Immunogenicity Testing Services

Ardena’s bioanalytical team in Assen provides complete immunogenicity testing programmes for biologic drug development, including screening, confirmatory, titration, and neutralisation assays. The team is experienced in developing and validating assays on MSD ECL and ELISA platforms, applying appropriate cut-point setting methodologies, and producing validation reports that meet ICH M10 and agency-specific requirements.

Ardena offers integrated immunogenicity and PK bioanalysis, enabling the simultaneous characterisation of drug exposure and immune response data that is required for a complete safety and efficacy assessment.