An assay in pharma is any test designed to measure a specific property of a drug, biological sample, or chemical compound, whether that is its potency, purity, concentration in blood, or effect on living cells. Assays show up at every stage of a drug’s life, from the earliest screening of thousands of chemical candidates to the final quality checks on a manufactured pill or vial. They are the measurement backbone of the entire industry, and when an assay fails or misleads, the consequences range from wasted years of research to unsafe products reaching patients.
What Assays Actually Measure
The word “assay” is broad on purpose. In a pharmaceutical context, it can refer to a chemical analysis that confirms how much active ingredient is in a tablet, a biological test that checks whether a protein drug still works the way it should, or a screening experiment that determines whether a new molecule binds to a disease target. What ties them all together is the core idea of quantification: an assay does not just ask “is this substance present?” but “how much, and how active is it?” Assay methods play a central role in characterizing the quality of bulk pharmaceutical substances, serving as one of the primary tools analysts rely on to confirm that a drug meets its specifications.1PubMed Central. The role of assay methods in characterizing the quality of bulk pharmaceuticals
In practice, the type of assay used depends on what question needs answering. A chemistry lab testing a finished tablet might use high-performance liquid chromatography to separate and quantify the active ingredient. A biology lab in early drug discovery might run a cell-based assay to see whether a compound kills cancer cells or blocks a viral enzyme. A clinical pharmacology group might measure how much of a biologic drug is circulating in a patient’s bloodstream after injection. These are all “assays,” but they look nothing alike in the lab.
How Assays Drive Drug Discovery
Drug discovery begins with a vast number of chemical candidates and a need to narrow them down fast. High-throughput screening, or HTS, uses robotic systems and automated detectors to test thousands of compounds against a biological target in a short window of time, measuring which ones show promising activity and flagging those with potential toxicity.2PubMed Central. Adaptation of high-throughput screening in drug discovery-toxicological screening tests Without this kind of rapid screening, researchers would be stuck testing compounds one at a time, a pace incompatible with modern drug development timelines.
Cell-based assays add another layer of realism. Instead of testing a compound against a single purified protein in a test tube, researchers expose whole living cells to the compound and observe what happens. These platforms come in several forms: traditional cell cultures in plates, miniaturized microfluidic setups, biosensor-equipped systems, and chromatography-based methods that use cells as the separation medium. The advantage is biological relevance, since a drug ultimately has to work inside a cell, not just in a vial, and the cost is far lower than jumping straight to animal experiments.3PubMed Central. A review for cell-based screening methods in drug discovery
One persistent headache in screening is false positives. Some chemical structures are notorious for lighting up in almost any assay, binding nonspecifically to targets and giving the misleading impression that they are biologically active. These “pan assay interference compounds,” or PAINS, waste enormous time and resources if they are not caught early. Researchers have developed substructure filters to identify and remove these frequent hitters from screening libraries before they send teams chasing dead-end leads.4PubMed. New substructure filters for removal of pan assay interference compounds (PAINS) from screening libraries and for their exclusion in bioassays This is a good example of how the quality of the assay itself shapes the quality of the science: a screening campaign that does not account for PAINS can produce a list of “hits” that look exciting on paper but go nowhere in follow-up testing.
From Hit Compound to Drug Candidate
Once a compound clears initial screening, it enters lead optimization, a phase where medicinal chemists tweak the molecule’s structure to improve its druglike properties. The key questions here are practical: will the body absorb it, will it distribute to the right tissues, how quickly will the liver break it down, can the kidneys clear it, and will it cause toxicity? These properties, often grouped under the shorthand ADMET (absorption, distribution, metabolism, excretion, and toxicity), are assessed through a battery of in vitro assays that serve as surrogates for what will happen in a living organism.
Researchers have examined the relationship between molecular properties, these surrogate assays, and actual outcomes in animals. In one large-scale analysis spanning thousands of compounds across hundreds of chemical series, the goal was to understand how well the bench-level assays predicted real-world pharmacokinetic and toxicology results in rodents.5PubMed. Relating molecular properties and in vitro assay results to in vivo drug disposition and toxicity outcomes The takeaway is that while no single in vitro assay perfectly mirrors what happens inside a body, using them in combination gives optimization teams an early warning system. A compound that metabolizes too quickly in a liver microsome assay, for instance, will probably have poor oral bioavailability in patients. Catching that in the lab is far cheaper and faster than discovering it in a clinical trial.
Safety-focused assays have become especially prominent. Drug attrition due to safety findings remains one of the biggest sources of loss in the industry, and many companies now position safety assessment at every stage, including discovery, where an early read on potential toxic liabilities is sought at the chemical scaffold level. The logic is straightforward: if you can identify safety red flags when a project costs relatively little, you avoid pouring years and hundreds of millions of dollars into a compound that will eventually fail for toxicity reasons.
Measuring Drug Levels in Patients
Once a drug enters clinical trials in humans, a different class of assays comes into play. Pharmacokinetic (PK) assays measure how much drug is circulating in a patient’s blood over time. This information is critical for setting the right dose: too little and the drug will not work, too much and side effects may become unacceptable. For traditional small-molecule drugs, these measurements are relatively straightforward using mass spectrometry. For biologic drugs like monoclonal antibodies, the measurement challenge is considerably harder because these are large, complex protein molecules.
The standard method for measuring biologic drug levels in blood has historically been an immunoassay called ELISA, which uses antibodies to capture and detect the drug. But as mass spectrometry technology has advanced, hybrid methods that combine an antibody-based capture step with mass spectrometry detection have become widely used, offering improved specificity and the ability to distinguish the drug from the body’s own proteins.6PubMed Central. Immunocapture LC-MS methods for pharmacokinetics of large molecule drugs
The choice between assay platforms is not academic. Different methods can give different numbers for the same patient sample, and those numbers feed directly into dosing decisions. Researchers have compared ELISA, mass spectrometry, and microfluidic immunoassay techniques head to head for quantifying the cancer drug bevacizumab in plasma from lung cancer patients, exploring how results diverge and converge across platforms.7PubMed. Comparison of Bevacizumab Quantification Results in Plasma of Non-small Cell Lung Cancer Patients Using Bioanalytical Techniques Between LC-MS/MS, ELISA, and Microfluidic-based Immunoassay Similar cross-platform comparisons have been done for other antibody drugs, including immunocapture LC-MS methods developed and validated for quantifying the checkpoint inhibitor nivolumab in human plasma.8Talanta. Quantification of nivolumab in human plasma by LC-MS/HRMS and LC-MS/MS, comparison with ELISA When two validated assays disagree on a drug’s concentration, the clinical team has to determine which number to trust, a decision that directly affects whether patients get the right dose.
Immunogenicity Testing
Biologic drugs introduce a problem that small molecules rarely face: the patient’s immune system may recognize the drug as foreign and produce antibodies against it. These anti-drug antibodies (ADAs) can reduce the drug’s effectiveness, alter its clearance rate, or, in rare cases, cause dangerous allergic reactions. Detecting and characterizing ADAs requires its own tiered assay strategy, starting with a screening assay to find samples that may contain ADAs, followed by a confirmatory assay to rule out false positives, and then additional tests to measure how much ADA is present and whether it actually blocks the drug’s activity.9Frontiers in Immunology. Immunogenicity to Biotherapeutics – The Role of Anti-drug Immune Complexes
This tiered approach exists because the stakes of getting it wrong cut both ways. A false positive (flagging a patient as having ADAs when they do not) might lead to unnecessary dose changes or drug withdrawal. A false negative (missing real ADAs) could leave a patient on a drug that is no longer working. The assay design has to balance sensitivity and specificity, and the results feed into regulatory filings that determine the drug’s label and dosing instructions.
Quality Control in Manufacturing
After a drug is approved, assays shift from exploration to enforcement. Every batch of drug product must be tested before it reaches patients, and the tests must show that the batch meets predetermined specifications for identity, purity, potency, and stability. For a small-molecule tablet, this might mean running a chromatographic assay to confirm the active ingredient is present at the right concentration and that degradation products are below acceptable limits.
For biologics, quality control is considerably more involved. A monoclonal antibody is a massive protein molecule whose three-dimensional structure can be altered by subtle changes in manufacturing conditions like temperature, pH, or the composition of the growth medium. Proving that two batches are equivalent requires a battery of assays examining the molecule from multiple angles. For relatively small proteins like insulin, equivalence can be established using a set of well-characterized analytical methods, but for larger proteins, it is almost impossible to guarantee full equivalence through analytical testing alone.10International Journal of Pharmaceutics. Shifting paradigms: biopharmaceuticals versus low molecular weight drugs This is a major reason why bringing biosimilar versions of complex biologic drugs to market is so much harder than making a generic version of a simple chemical pill.
Stability testing adds a time dimension. Drug products must remain safe and effective throughout their labeled shelf life, which means running assays on samples stored under various conditions over months and years. As the industry increasingly works with biologics, biosimilars, and advanced delivery systems like nanoparticle formulations, stability testing is evolving beyond conventional assays toward higher-resolution methods that can capture subtle changes in a product’s physical and functional characteristics.
What Regulators Expect
Pharmaceutical assays do not operate in a regulatory vacuum. The International Council for Harmonisation (ICH) sets guidelines that define how assays should be developed, validated, and documented. The ICH Q2 guideline, recently revised, addresses the validation of analytical procedures and now includes a requirement that confidence interval limits for key performance characteristics like accuracy and precision should be compatible with acceptance criteria.11PubMed. Confidence Intervals for Validation of Analytical Procedures Under ICH Q2(R2) In plain language, this means a company cannot just report that an assay gives the “right answer” on average; it must also show that the range of uncertainty around that answer falls within acceptable bounds.
Validation involves demonstrating that an assay is specific (measures only what it claims to measure), accurate (gives results close to the true value), precise (gives consistent results when repeated), and robust (tolerates small, deliberate changes in conditions without breaking). Every assay used to release a drug product or support a regulatory filing must go through this process, and the documentation can run to hundreds of pages. Regulators at agencies like the FDA and EMA review these validation packages closely, because an unreliable assay undermines every decision made based on its results.
Reagent Quality and Assay Lifecycle
An assay is only as good as its components, and for ligand-binding assays that rely on biological reagents like antibodies and proteins, reagent quality is a persistent vulnerability. The structural integrity and functional quality of these critical reagents is directly linked to assay performance, and physicochemical characterization coupled with functional testing helps ensure the highest degree of reagent quality. Poor reagent management, whether from instability or batch-to-batch variability, can cause costly delays in drug development.12PubMed. Characterization of critical reagents in ligand-binding assays: enabling robust bioanalytical methods and lifecycle management
This is more practical than it sounds. Imagine a company running a clinical trial that depends on an immunoassay to measure drug levels. If the antibody reagent used in that assay degrades or a new lot performs differently from the old one, every sample measured with the bad reagent is suspect. Reanalyzing samples costs time and money; if the problem is not caught, the faulty data could lead to wrong dosing conclusions or a failed regulatory submission. Companies now invest significant effort in reagent characterization, lifecycle management, and bridging studies to confirm that a new lot of reagent performs equivalently to the old one.
Companion Diagnostics
Some assays do not measure the drug at all; they measure the patient. Companion diagnostic assays are tests, often based on immunohistochemistry or molecular profiling, that identify which patients are likely to benefit from a particular therapy. These assays have become essential in oncology, where targeted therapies work only in patients whose tumors express a specific biomarker.
A concrete example is the VENTANA FOLR1 assay, developed as a companion diagnostic for the antibody-drug conjugate mirvetuximab soravtansine (MIRV) in patients with platinum-resistant ovarian cancer. In the SORAYA trial, which used this assay to select patients for treatment, clinically meaningful efficacy was demonstrated, and the assay’s staining acceptability rate was about 98%, confirming it could reliably identify the right patients.13PubMed. Development of an FRα Companion Diagnostic Immunohistochemical Assay for Mirvetuximab Soravtansine If the assay had been unreliable, patients without the target biomarker would have been enrolled, diluting the trial’s results and potentially leading to the wrong conclusion about whether the drug works.
The precision medicine movement has made companion diagnostics standard practice for many new cancer drugs. Regulatory agencies often approve the drug and its companion diagnostic simultaneously, making the assay a gatekeeper for treatment: no test result, no prescription. This means the assay’s accuracy and reproducibility directly affect which patients gain access to a therapy.
Potency Testing for Cell and Gene Therapies
The newest class of medicines, cell and gene therapies, has introduced potency-testing challenges that did not exist a generation ago. Unlike a small-molecule pill with a single active ingredient, a gene therapy product might be a viral vector carrying a corrective gene, or a patient’s own immune cells that have been genetically modified and expanded in a lab. Proving that such products are potent enough to work requires assays that address multiple functional mechanisms, and a single test is rarely sufficient.14PubMed Central. Potency testing of cell and gene therapy products
For cells that have been transduced with a viral vector, potency likely depends on how well the transgene is expressed, but also on how efficiently the vector delivered the gene and how many copies ended up in each cell. Genome editing adds further complexity: the editing process can change multiple characteristics of the cells simultaneously, and the potency assay has to capture those changes in a meaningful way. The field is still working out how to standardize these measurements, and regulatory expectations are evolving in parallel. For companies developing these therapies, designing and validating the right potency assay can be as scientifically challenging as designing the therapy itself.
Emerging Platforms That Could Reshape Assay Work
Traditional cell culture in a flat dish is a poor mimic of what happens inside a human body. Cells grown in a plate do not retain the architecture, mechanical forces, or cell-to-cell interactions they experience in an actual organ. This limitation has driven the development of organ-on-a-chip technology, which uses microfluidic devices to recreate miniaturized versions of human organs on a small chip, complete with fluid flow, mechanical stretching, and multiple cell types interacting in a tissue-like arrangement.15PubMed Central. Organ-on-a-Chip: A New Paradigm for Drug Development These systems show potential for studying how drugs work, prioritizing lead candidates, testing for toxicity, and identifying biomarkers.16PubMed Central. Microfluidic organs-on-chips
Despite the promise, these technologies still face hurdles. Accurately predicting drug efficacy, toxicity, and organ interactions remains difficult because the challenge of maintaining authentic organ function and morphology outside the body is not fully solved.17Drug Metabolism and Pharmacokinetics. Organ/body-on-a-chip based on microfluidic technology for drug discovery Adoption has been gradual, though regulatory agencies have expressed interest in accepting organ-on-chip data as part of drug development packages, which could accelerate their use.
At an even finer resolution, single-cell assays are revealing that individual cells within a tumor can respond very differently to the same drug. Research on breast cancer cells has shown that some cells require more than ten times the drug concentration that the population average would suggest is sufficient for inhibition. In one study, about 11% of cancer cells exposed to the drug omipalisib had sensitivity thresholds far above the typical population-level measurement, suggesting that dosing based on population averages might leave a significant fraction of tumor cells untouched.18npj Systems Biology and Applications. Characterizing heterogeneous single-cell dose responses computationally and experimentally using threshold inhibition surfaces and dose-titration assays This kind of single-cell resolution is exactly the type of insight that traditional bulk assays cannot provide, and it has implications for understanding why some tumors develop resistance to therapy.
Spatial transcriptomics technologies now allow researchers to map drug responses across a tumor’s physical geography, identifying pockets of resistant cells and the local cellular ecosystems that support their survival.19Briefings in Bioinformatics. SpaRx: elucidate single-cell spatial heterogeneity of drug responses for personalized treatment Label-free imaging platforms for drug screening of tumor organoids are also emerging, offering high-throughput readouts that quantify responses at the level of individual organoids, providing insights into heterogeneity that are relevant for predicting how well a treatment will work and whether resistance will emerge.20PubMed. Label-free interferometry platform for drug response profiling of bioprinted tumor organoids at single-organoid resolution These technologies are still largely in the research phase, but they point toward a future where assays do not just tell you whether a drug “works” on average, but whether it works for every subpopulation of cells in a given patient’s disease.
When Assays Replaced Animals
One underappreciated function of assay development is its role in reducing animal testing. A historical example comes from pharmaceutical safety testing for bacterial contamination. For decades, the standard method for detecting pyrogens (fever-causing contaminants) in injectable drugs involved injecting the product into rabbits and monitoring their body temperature. The Limulus Amebocyte Lysate (LAL) test, an in vitro assay using a reagent derived from horseshoe crab blood, was developed as an alternative. Over time, pharmaceutical laboratories in some countries were able to replace the vast majority of rabbit pyrogen tests with the LAL assay without increasing risk to patients. This shift was not just an ethical improvement; it was faster, cheaper, and more standardized. The trajectory from animal model to bench assay is one that the industry continues to pursue across multiple testing areas, with organ-on-chip and other in vitro platforms following the same basic logic: build a better assay, and the need for whole-animal experiments shrinks.