Cell Therapy Potency — The Most Consequential and Difficult CMC Problem in Your Program

Potency for a cell therapy product is not a release test. It is the primary regulatory evidence that manufacturing produced cells that can do what the label claims. CBER treats it accordingly — and most sponsors do not.

The consequence of that misalignment is not a form 483 observation and a corrective action plan. It is a clinical hold at the end of Phase 2, when your BLA-enabling trial is enrolling, because you cannot demonstrate that the product released to patients three years ago was potent in any mechanistically meaningful sense. Potency is the CMC domain where early-stage scientific compromises become late-stage regulatory catastrophes — and where the recovery cost is measured in years, not months.

Why Potency Is the Most Consequential — and Most Difficult — CMC Problem in Cell Therapy

Cell therapy products are not defined by a chemical structure or a binding affinity constant. They are defined by what they do — and under 21 CFR 610.10, potency is the quantitative measure of that function. Unlike small molecules, where physical characterization and purity testing provide strong surrogates for activity, a viable, CD56-positive NK cell that fails to kill its target cell at the relevant effector-to-target ratio has no value as a therapeutic regardless of how pure, sterile, or morphologically intact it is. The biological activity of the cell is the product. That is why ICH Q6B, which governs biological product specifications, places potency at the center of the product characterization framework — not as one attribute among many, but as the anchor of clinical relevance.

The analytical challenge that follows from this regulatory logic is formidable. A cytotoxicity assay — for example, a 4-hour chromium-51 release assay or the non-radioactive DELFIA EuTDA assay using K562 target cells — must be robust enough to generate reproducible, lot-differentiating results against the background variability of living effector cells that have been cryopreserved, thawed, and rested before testing. Specification-setting requires that the sponsor define a clinically anchored lower limit: for NK cell products, an accepted industry benchmark requires ≥20% specific lysis at an effector-to-target ratio of 5:1 using K562 cells, with testing typically performed across a panel of E:T ratios including 1:1, 5:1, 10:1, and 20:1 to characterize the full cytotoxicity curve. Achieving assay precision sufficient to enforce that threshold across manufacturing lots, across donor variability in autologous programs, and across product generations in a comparability exercise is an undertaking of a different order than anything encountered in monoclonal antibody CMC.

The Regulatory Definition of Potency for Cell Therapy Products: What CBER Requires

The FDA’s 2011 guidance, Potency Tests for Cellular and Gene Therapy Products, is the controlling document for this domain. Its requirement is unambiguous: sponsors must develop a potency test that measures the relevant biological activity linked to the mechanism of action of the product. The guidance does not accept surrogate markers of manufacturing success — transduction efficiency, vector copy number, transgene expression by flow cytometry — as standalone potency assays. It accepts them as interim measures during early Phase 1 only when a functional assay is being developed in parallel, with a defined timeline for functional assay implementation. Sponsors who read this guidance in 2011 and filed INDs in 2015 with “CAR expression ≥20% by flow cytometry” as the sole potency assay were making a deferred regulatory bet that CBER has consistently declined to honor at BLA review.

For NK cell products — whether derived from peripheral blood mononuclear cells, umbilical cord blood units, or iPSC-derived platforms — the identity standard is CD56+CD3−, with the CD3-negativity criterion critical to distinguish true NK cells from NKT cells, which are CD56+CD3+. That identity criterion, however, is not a potency criterion. CBER’s expectation for non-CAR NK cell products is potency demonstrated by cytotoxicity assay, with IFN-γ secretion as a complementary functional endpoint. For CAR-NK products, cytotoxicity by a target-cell-specific assay is mandatory — the CAR transgene endows the cell with a redirected killing function, and that function must be measured, not assumed. FDA’s Chemistry, Manufacturing, and Controls (CMC) Information for Human Gene Therapy INDs guidance (2020) reinforces this point specifically in the context of genetically modified cellular products, making explicit that the potency assay must reflect the function conferred by the genetic modification, not merely the presence of the modification.

The further regulatory complication — one that the 2011 CBER potency guidance addresses directly — is the reference standard requirement. A potency assay without a reference standard is a test with no anchor. The numerical result generated on a 96-well plate on any given testing day is meaningful only in relation to a characterized biological reference material against which the lot result can be normalized. For early-phase programs, this is typically a well-characterized internal cell bank preparation whose potency is assigned and used to normalize lot results over time. Failing to establish a reference standard at IND is not merely a data quality problem — it renders the historical release testing dataset non-comparable at the point in the program when comparability data become the primary CMC argument for Phase 3 manufacturing scale-up.

Potency Assay Development for Cell Therapy: The MOA-Based Assay Requirement and Its Technical Challenges

The mechanism-of-action-linked potency requirement creates a specific development sequence that many cell therapy sponsors abbreviate to their eventual detriment. ICH Q2(R2)/Q14, which governs analytical procedure development and validation, requires that method suitability be demonstrated before the method is used to support regulatory decisions. For a cytotoxicity assay — whether the 51Cr release format or the DELFIA EuTDA non-radioactive alternative — method qualification must address specificity (the assay measures killing of target cells by effector cells, not background cell death), linearity across the E:T ratio range, intermediate precision across operators and days, and robustness to effector cell preparation variables including cryopreservation and post-thaw rest time. The K562 cell line, used as a standard target because of its well-characterized NK sensitivity and lack of HLA class I expression, must itself be qualified as part of the assay system — passage number, growth characteristics, and baseline viability at time of use all contribute to assay variability that propagates directly into lot release result variability.

The operational failure pattern in this domain is consistent and documented: a sponsor develops a cytotoxicity assay for research use during process development, transfers it to the GMP analytical laboratory without a formal technology transfer study, uses it to generate IND lot release data on twenty lots across two clinical sites, and then presents that data at a pre-BLA meeting only to discover that CBER reviewers have identified excessive lot-to-lot variability that the sponsor cannot explain because assay precision was never characterized under the ICH Q2(R2)/Q14 framework. The root cause is not that the assay was wrong — it is that a living-cell assay operated without reference standard normalization and without intermediate precision data can produce a coefficient of variation that makes the underlying manufacturing process appear more variable than it is. Distinguishing process variability from assay variability is impossible without method qualification data, and without that distinction, comparability arguments for any manufacturing change made during clinical development are analytically unsupportable.

FDA’s Considerations for the Design of Early-Phase Clinical Trials of Cellular and Gene Therapy Products guidance (2015) addresses this challenge by acknowledging that potency assay development may be iterative during Phase 1, but insists that the development roadmap — what assay will be used, how it will be qualified, when it will replace the interim surrogate, and what acceptance criterion will apply — must be described in the IND. USP <1046>, Cell-Based Advanced Therapies and Tissue-Based Products, provides the compendial framework that aligns with CBER’s expectations, offering a practical structure for describing assay system components including target cells, effector cell preparation procedures, assay incubation conditions, and result calculation methodology. Sponsors who use USP <1046> as an organizational scaffold for their potency assay description in Module 3.2.P.5 often produce packages that hold up at review precisely because the compendial structure forces completeness in a way that sponsor-developed formats rarely achieve.

Building a Cell Therapy Potency Testing Program That Will Hold Up Through BLA Submission

The XGene Cell Therapy Potency Assay Development Architecture is a stage-gated potency assay development roadmap for cell therapy products — from IND surrogate marker strategy through Phase 2/3 cytotoxicity assay qualification to BLA full validation — with CBER interaction milestones and reference standard establishment requirements built into each gate.

Step 1 — IND Surrogate Mapping with MOA Anchor: At IND filing, define the interim surrogate assay (e.g., CAR expression by flow cytometry or CD56+CD3− enumeration) and simultaneously document the mechanism-of-action-linked potency assay that will replace it — including the target cell line (K562 or antigen-positive target for CAR-NK), the E:T ratio panel (1:1, 5:1, 10:1, 20:1), the assay format (4-hour 51Cr release or DELFIA EuTDA), and the intended acceptance criterion (e.g., ≥20% specific lysis at E:T 5:1). This mapping document becomes the regulatory commitment that CBER will track through the IND lifecycle.

Step 2 — Reference Standard Establishment Before Phase 1 Lot Release: Designate and characterize an internal reference standard cell bank preparation before the first GMP lot is released to clinical use. Assign a potency value to the reference standard using the qualified cytotoxicity assay, and establish the normalization procedure by which every subsequent lot result will be expressed relative to the reference standard. This step is non-negotiable — without it, your Phase 1 and Phase 2 lot release datasets are non-comparable and cannot support a manufacturing comparability argument at BLA.

Step 3 — Cytotoxicity Assay Qualification Under ICH Q2(R2)/Q14 Before Phase 2 Enrollment: Execute a formal analytical procedure qualification study that characterizes specificity, intermediate precision (across operators, instruments, and days), and robustness to effector cell preparation variables. Document target cell qualification parameters including passage number limits and pre-assay viability acceptance criteria. Generate a qualification report that can be submitted as a Module 3 amendment before Phase 2 IND lot release data are generated. This report is the evidentiary foundation for every comparability exercise the program will conduct from this point forward.

Step 4 — BLA-Ready Validation Package with CBER Interaction Milestone: Before initiating BLA-enabling Phase 3 manufacturing, conduct a pre-BLA meeting with CBER specifically to align on potency assay validation scope — including acceptance criteria, reference standard expiry and bridging study requirements, and the statistical model for specification-setting based on Phase 2 lot history. This interaction prevents the most common late-phase failure mode: a BLA potency section that CBER considers analytically insufficient because the validation package was designed without agency input on the validation acceptance criteria.

The output of the XGene Cell Therapy Potency Assay Development Architecture is a BLA-ready potency dossier that maps every regulatory requirement from the 2011 CBER potency guidance, ICH Q6B, and ICH Q2(R2)/Q14 to a specific assay document, qualification record, or specification in Module 3 — not a gap analysis, but a close-out package organized around the arguments CBER reviewers will apply at BLA review.

The programs that arrive at pre-BLA meetings without a qualified, MOA-linked potency assay and without a reference standard are not programs that lacked technical capability — they are programs that made deliberate early-stage decisions to defer analytical investment in favor of clinical speed. That deferral does not save time. It transfers the cost of the work to the most expensive point in the development timeline, where every month of analytical remediation is a month of BLA delay, and where the absence of reference standard-normalized historical data may require repeating manufacturing comparability studies that were already conducted. Potency is the CMC domain where the cost of early-stage compromise is not paid early — it is paid at the end, in full, with interest.

For your cell therapy product, can you identify today whether your potency assay measures biological activity linked to the mechanism of action (e.g., antigen-specific cytotoxicity), name the cell line and E:T ratio used, identify the acceptance criterion in your current lot release specification, and state whether a reference standard has been established for potency normalization?