The FDA’s August 2026 draft guidance Potency Assessment of Active Immunotherapy Products emphasizes that potency assessment for active immunotherapy products should operate within a broader potency-assurance strategy rather than depend solely on a lot-release assay.
MOA and potency-related CQA should guide analytical procedure development so that methods interrogate biologically meaningful product characteristics.
Manufacturing-process understanding, material controls, in-process testing, and assay-reagent strategy can all contribute to potency assurance alongside finished-product testing.
Personalized neoantigen therapies demonstrate why potency evidence may extend across bioinformatics, manufacturing, physicochemical characterization, and release testing rather than reside in a unique functional assay for every patient-specific lot.
A well-designed potency strategy should remain informative across the product life cycle, particularly when manufacturing changes require stability or comparability assessments.
From Potency Test to Potency Assurance
The U.S. Food and Drug Administration’s (FDA’s) new draft guidance for active immunotherapies (Potency Assessment of Active Immunotherapy Products) is nominally about potency assessment, but its more consequential message extends beyond the assay itself.1 For products designed to treat an existing disease or condition by inducing, stimulating, or modulating immune effector cells through disease-associated antigens, intended biological activity ultimately depends on a response generated within the patient. Demonstrating potency therefore requires more than identifying a reproducible laboratory readout.
The August 2026 draft guidance covers vectored, peptide- and protein-based, and cell-based active immunotherapy products (ACTIMPs), including personalized products. It also extends a regulatory trajectory that predates the new document. Earlier guidance for cell and gene therapy products recognized that potency may require extensive product characterization and complementary methods, while subsequent guidance described potency assurance as encompassing manufacturing-process design and control, material controls, in-process testing, and lot-release assays.2,3
The new guidance applies that philosophy more specifically to active immunotherapies and makes its practical implications clearer. A potency assay remains a critical part of the control strategy where appropriate, but the larger task is to establish why the selected measurements matter and what other evidence supports confidence in the product’s biological capacity.
A release result can show that a lot met a defined criterion. Potency assurance requires an explanation of why that criterion, together with the surrounding controls and accumulated product knowledge, provides sufficient confidence in biological activity.
Start With the Biology, Not the Assay
The starting point is the active ingredient and its mechanism of action (MOA). The FDA recommends using that understanding to identify potency-related critical quality attributes (CQAs), which then helps determine which product characteristics must be measured, controlled, or otherwise understood.1
Analytical technology should therefore follow the biological question rather than define it. Selecting a convenient platform and then determining what it can measure risks allowing the available technology to shape the potency strategy. A more defensible approach first asks which attributes are meaningfully connected to biological activity and then determines how best to interrogate them.
Those relationships may be informed by product characterization, nonclinical and proof-of-concept studies, clinical experience, prior knowledge, and monitoring of immune responses.1 As development proceeds, additional product, process, and clinical knowledge can clarify or revise the relationship between particular attributes and biological activity.
This approach aligns with the life cycle framework for analytical procedure development. Analytical procedures should be fit for their intended purpose, and product and process understanding should inform the quality attributes that need to be tested. Development of the procedure and validation of its performance are related but distinct activities: one establishes what information the method needs to generate, while the other establishes whether it can generate that information reliably.4,5
For active immunotherapies, that distinction can be especially important because understanding of the relationship between product characteristics and immune activity may mature throughout development. The potency strategy should be able to evolve with that knowledge rather than becoming fixed around an early analytical choice that later proves too narrow.
One Number May Not Capture the Relevant Biology
Complex active immunotherapies may contain multiple components or depend on several biological steps, making it difficult for one analytical result to capture every feature relevant to potency.
The FDA expects potency assays to be quantitative and sufficiently precise to distinguish sufficiently active from sub-potent product, but the appropriate analytical architecture can differ substantially among modalities. Depending on the product, developers may rely on a quantitative bioassay, complementary functional assays, combinations of biological and physicochemical methods, or, when adequately justified, physicochemical measurements that address the relevant potency-related attributes.1
For defined multicomponent peptide- or protein-based products, for example, an assessment of immunological potential may need to be complemented by quantitative measurements of the individual components. The objective is not to maximize the number of assays; it is to obtain enough analytical resolution to detect changes that matter without generating large amounts of characterization data that have no clear connection to biological activity.
The same principle appeared in earlier potency guidance, which recognized that multiple complementary methods may be needed when a single assay cannot adequately reflect relevant biological activity.2
A multivalent genetic cancer vaccine provides a practical example. Its developers established a quantitative reverse-transcription polymerase chain reaction (RT-PCR) assay capable of measuring expression of multiple encoded transgenes and detecting loss of activity affecting an individual component.6 The example is product-specific, but it illustrates the underlying analytical problem: an aggregate readout can obscure differences among biologically important elements of a complex product.
Potency assessment therefore depends not only on whether an assay produces a reliable number but also on whether the analytical approach can discriminate changes in the attributes that matter.
Manufacturing Control Can Carry Part of the Potency Burden
One of the most consequential elements of the new guidance is its recognition that a potency-related CQA does not necessarily require its own lot-release assay. If process design or the control strategy can adequately ensure that an attribute remains within acceptable limits, the FDA indicates that a separate release measurement may not be necessary.1
This places process understanding inside the potency strategy. When manufacturing controls are expected to assure an attribute that is not independently measured at release, developers need sufficient knowledge of the relationship among the process, the resulting product attribute, and biological activity to support that approach.
The concept is consistent with the potency-assurance framework for cell and gene therapy products, which treats process design, process controls, material controls, in-process testing, and release assays as complementary elements rather than independent activities.3 It also aligns with the FDA’s life cycle approach to process validation, in which evidence accumulated from process design through routine manufacturing establishes that a process can consistently deliver product meeting its quality requirements.7
Process development data can therefore become part of the rationale supporting potency assurance. If a process parameter or set of parameters is demonstrably linked to a potency-related attribute, effective process control may provide evidence that would otherwise require additional end-product testing.
Assay reagents create another connection between analytical strategy and operational control. The FDA recommends identifying sources for necessary potency-assay reagents and qualifying them for their intended purposes early in clinical development.1 Reagent availability, lot-to-lot variability, qualification, and replacement can all affect whether a method remains reliable through later stages of development.
A release assay can still provide essential evidence about the finished product, but confidence in that result may depend on controls established much earlier. Potency assurance is therefore partly analytical and partly a function of how well the manufacturing system is understood and controlled.
Personalized Immunotherapies Reveal the Limits of the Conventional Model
Personalized neoantigen therapies make this concept unusually visible because product composition can change from patient to patient. The control strategy must govern a manufacturing platform that repeatedly produces different individualized products rather than identical copies of a fixed formulation.
The FDA explicitly acknowledges that developing a separate product-specific bioassay for every patient lot of a personalized peptide- or protein-based ACTIMP may not be practical. The guidance instead describes a framework in which analytical, computational, and manufacturing evidence can contribute to potency assurance as development advances.
During phase I development of personalized neoantigen products, potency may in some circumstances be inferred from confirmation of the intended sequence when the approach is supported by appropriate scientific justification of the bioinformatics pipeline. In this setting, confidence in potency depends partly on whether the computational system correctly selected the intended sequence and whether manufacturing produced that sequence as designed.
Later in development, the evidence can shift. Once the manufacturing process, including the relevant bioinformatics pipeline, has been appropriately qualified, physicochemical testing of potency-related CQAs may in some circumstances provide sufficient support for lot release without creation of a unique bioassay for every individualized product.
This places the bioinformatics pipeline inside the quality logic of potency assurance. Computational decisions determine which antigenic sequences enter manufacturing, so changes to the pipeline can potentially change the product that reaches the patient. The FDA accordingly notes that changes to a bioinformatics pipeline or manufacturing process that could alter predicted epitopes or potency-related CQAs may require additional studies.
Published development programs show why this framework is necessary. A personalized peptide vaccine platform integrated tumor sequencing and bioinformatics with patient-specific peptide selection, high-throughput synthesis, good manufacturing practice production, and analytical control. Because peptide composition differed among patients, analytical method development had to accommodate product variability from batch to batch, leading to a risk-based manufacturing and analytical approach rather than a conventional model built around a fixed composition.8
A separate personalized neoantigen messenger RNA program addressed the same basic manufacturing challenge using a different modality. The developers established a small-batch, good manufacturing practice-compliant process for individualized mRNA, subjected manufactured material to extensive quality assessment, and demonstrated reproducible production in validation runs.9
Both examples underscore the importance of controlling the systems that generate individualized products. Sequence selection, computational processing, manufacturing consistency, product characterization, and release testing can each contribute evidence at different stages.
That does not lower the standard for potency. Instead, it changes where the evidence supporting potency resides.
A Potency Strategy Has to Survive Manufacturing Change
Potency-related CQAs matter beyond release. The FDA identifies them as relevant to stability and comparability, meaning that attributes selected early in development may later become central to determining whether a manufacturing change has altered the product in a meaningful way.1
For biological products generally, comparability assessments evaluate whether manufacturing changes adversely affect product quality, safety, or efficacy.10 For products with complex biological activity, potency becomes especially important because a manufacturing change can affect attributes that routine structural or physicochemical measurements may not fully reflect.
The regulatory framework is particularly explicit for cell and gene therapy products. Comparability evaluations should incorporate understanding of CQAs, accumulated manufacturing experience, and MOA, and quantitative potency assessment should form part of analytical comparability. When a single method is imprecise or cannot assess all relevant aspects of MOA potentially affected by a manufacturing change, several analytical methods may be useful. The FDA also warns that excluding potency analysis can compromise the conclusions drawn from a comparability exercise.11
Those recommendations apply specifically to cell and gene therapy products and should not be generalized automatically to peptide- or protein-based ACTIMPs. They nevertheless illustrate why early potency decisions can have consequences much later in the product life cycle. A method selected only because it meets an immediate release need may prove less useful later if it cannot detect biologically meaningful changes introduced through process modification.
Personalized products extend the same principle into computational change. A modification to a bioinformatics pipeline can alter predicted epitopes, while a manufacturing-process change may affect potency-related CQAs, potentially creating a need for additional evaluation.1
A potency strategy should therefore retain enough biological and analytical resolution to remain informative as manufacturing evolves. Developers cannot anticipate every future modification, but they can avoid defining potency so narrowly that later comparability depends on methods incapable of detecting the attributes most likely to matter.
Analytical Validation Cannot Rescue an Irrelevant Measurement
Method validation remains essential, but it answers a different question from biological relevance. Q14 distinguishes analytical procedure development from validation, while Q2(R2) provides the framework for demonstrating that a method performs appropriately for its intended purpose.4,5
A method can therefore be analytically strong yet still provide incomplete potency assurance if the characteristic it measures does not adequately reflect the biologically important attributes of the product. Precision, accuracy, and robustness cannot substitute for a weak connection between the analytical readout and the activity that needs to be controlled.
The order of operations matters. Product and process knowledge should inform the attributes of interest, those attributes should define the intended purpose of the analytical procedure, and method development and validation should follow from that purpose. As knowledge develops, the analytical approach may also need to change.4
Validation confirms that the chosen method performs as intended. It does not establish that the chosen measurement is biologically informative.
Potency Assurance Depends on Information Continuity
The evidence supporting potency often spans biological, analytical, manufacturing, and quality functions, and personalized products can add bioinformatics to that chain.
The FDA’s process-validation framework emphasizes multidisciplinary participation spanning areas such as process engineering, analytical chemistry, manufacturing, and quality assurance.7 Potency assurance provides another reason for those disciplines to remain connected because decisions made by one group can determine what another group needs to measure or control.
Outsourcing can complicate those interfaces. A drug developer may use one organization for process development and manufacturing, another for specialized analytical testing, and additional partners for characterization or computational work. The regulatory framework does not prescribe a particular outsourcing model, but fragmentation raises a practical question: does each party have enough product and process context to understand how its work contributes to potency assurance?
An external analytical laboratory that receives only a method and specification may be able to execute the test correctly without understanding why the measurement matters. That may be adequate for routine execution, but it provides less context when troubleshooting unexpected results, assessing method changes, or supporting comparability. Likewise, a manufacturing partner cannot fully evaluate the potency implications of process variability if the potency-related CQAs and their biological rationale have not been communicated.
Information continuity becomes especially important when conditions change. An unexpected analytical trend, process modification, reagent replacement, or revised understanding of the product can require analytical, manufacturing, and quality teams to determine whether a measurable difference is also biologically consequential.
For drug developers working across internal and external organizations, potency assurance therefore depends not only on technical capability but also on whether the biological rationale behind the control strategy survives each handoff.
From a Passing Assay to a Defensible Potency Strategy
The new guidance does not diminish the importance of potency assays. Where potency is assessed through a release assay, that method must provide an appropriately quantitative and discriminating measure of the relevant activity. What the broader approach changes is how much that single result is expected to prove.
For many active immunotherapies, confidence in potency will emerge from connected forms of evidence that include MOA, potency-related CQAs, product characterization, analytical measurements, process controls, material controls, and accumulated manufacturing knowledge. Personalized products make this distributed model especially visible because computational selection and manufacturing-platform control can themselves become part of the evidence supporting a patient-specific product.
The release question remains essential: does the lot meet its established potency requirements? The larger development question is whether those requirements, and the systems supporting them, adequately reflect the biological properties that need to be preserved.
The release assay is therefore not the potency strategy itself. It is one of the places where that strategy becomes measurable.
