Biologics Specifications and Potency — Building a BLA-Ready Release Package

3.2.S.4 Biologics Specifications: Setting Potency, Purity, and Safety Limits That FDA Will Accept

“The potency assay acceptance criterion of 50–200% (relative to reference standard) is too wide to provide meaningful assurance of biological activity — please provide a scientific justification for this range or tighten the specification.” This deficiency is among the most common in BLA specification review — and the underlying problem is almost never the assay. It is the absence of a documented rationale connecting the clinical exposure-response relationship to the acceptance criterion.

That distinction matters because it defines the entire remediation path. A sponsor who receives this deficiency and interprets it as a request to improve the assay will spend months redesigning a bioassay that may already be performing adequately. A sponsor who understands that the deficiency is a documentation failure — specifically, the failure to connect clinical data to the acceptance criterion — can respond with a targeted justification document that addresses the reviewer’s concern directly. The assay precision, the clinical batch range, and the clinical exposure-response relationship were all available in the submission; they were simply never assembled into a coherent argument for why the criterion is clinically meaningful. The resolution is not a new assay. It is a specification justification document that the submission should have contained from the outset.

ICH Q6B, finalized in 1999 and still the governing guidance for biologics specifications, establishes the foundational principle that acceptance criteria for biological products must be justified on the basis of safety and efficacy data from clinical and nonclinical studies, analytical data generated during development, and relevant process knowledge. The guidance explicitly rejects the approach of setting limits by analytical convention or by a fixed percentage around the nominal value without clinical support. For potency in particular, Q6B requires that the test reflect the biological activity relevant to clinical use and that the acceptance criteria be capable of detecting clinically meaningful differences in product activity. A 50–200% relative potency range set without reference to the clinical exposure-response relationship fails this requirement on its face — not because the range is inherently too wide or too narrow, but because no documented link exists between the range and any clinical outcome.

The technically correct approach to potency acceptance criterion derivation begins with the clinical manufacturing batch history. The relative potency values of all clinical batches — measured against the primary reference standard using the mechanism-of-action-based functional assay that will serve as the release test — define the observed clinical batch range. This range is not the acceptance criterion; it is the starting point for derivation. To this range, the sponsor adds the contribution of assay precision: the combined effect of intra-assay variability and inter-assay variability, typically quantified as the geometric coefficient of variation from a properly designed assay precision study conforming to ICH Q2(R2) (final FDA guidance, March 2024) criteria. The precision contribution is conventionally expressed as plus or minus two standard deviations of the log-transformed relative potency distribution, which captures approximately ninety-five percent of the expected assay variation around a true potency value. The resulting expanded range — clinical batch range widened by the assay precision component — represents the outer boundary of what the process has historically produced and the assay can reliably measure.

The final step is the clinical relevance statement, and this is where most submissions fall short. The sponsor must connect the derived acceptance criterion boundary to the clinical exposure-response relationship for the product. If the product has a demonstrated therapeutic window — the concentration or activity range within which efficacy is maintained and safety is acceptable — then the specification limits must sit within that window with adequate margin. FDA’s 2011 guidance on potency tests for cellular and gene therapy products, though directed at a specific product class, articulates this principle broadly: potency acceptance criteria should reflect the activity levels present in clinical material that demonstrated efficacy and safety. For a monoclonal antibody, the relevant clinical data include the dose-response characterization from Phase 1 and Phase 2 studies, the PK/PD modeling that relates exposure to pharmacodynamic endpoints, and any clinical evidence that variation in product activity within the observed range did not produce variation in clinical outcomes. If all clinical batches fell within 80–120% relative potency and no exposure-response relationship suggesting activity-dependence at those levels was identified, that is an affirmative clinical relevance statement — the criterion is scientifically justified because the clinical program was conducted entirely within it.

Purity specifications present a parallel challenge but with a different analytical framework. The major purity attributes for a biological drug substance — high molecular weight species measured by SEC-HPLC, protein fragmentation measured by non-reducing capillary electrophoresis SDS, and charge variant distribution measured by imaged capillary isoelectric focusing — each require an acceptance criterion that is derived from the clinical batch profile and justified by the available safety or clinical data. The conventional error is to set SEC-HPLC aggregate limits at the maximum observed value in any clinical batch, with no margin and no clinical justification for why that level of aggregation is acceptable. This approach anchors the limit to a single batch outlier rather than to a statistically derived clinical batch distribution, and it provides no documentation of why the aggregate level in that outlier batch did not raise immunogenicity concerns. FDA’s 2014 guidance on immunogenicity assessment for therapeutic proteins establishes the basis for the safety justification: the relationship between aggregate content and immunogenicity risk is product-specific, route-of-administration-dependent, and must be assessed on the basis of the immunogenicity data from the clinical program. A specification limit for aggregates derived from the clinical batch range combined with a statement that no correlation between aggregate content within that range and anti-drug antibody incidence was observed in clinical studies constitutes a defensible justification. An aggregate limit set at the maximum observed clinical batch value without this clinical immunogenicity data linkage does not.

Charge variant specifications by icIEF — percent acidic species, percent main peak, percent basic species — require the same clinical anchoring approach applied to a more complex multi-component profile. Because charge variants include deamidation and oxidation products that affect potency, receptor binding, and potentially immunogenicity, the specification for each component of the charge distribution should be justified both analytically, by reference to the observed clinical batch profile with statistical confidence intervals, and functionally, by reference to the characterization data demonstrating the biological impact of charge variants at the levels observed. The EMA Guideline on Development, Production, Characterization and Specifications for Monoclonal Antibodies takes an integrated view of the charge variant profile, treating it as a quality attribute panel that must collectively reflect the activity and safety envelope of the clinical product rather than as a set of independent analytical checkpoints.

Safety attribute specifications — endotoxin, host cell protein, and host cell DNA — are governed by risk-based derivation frameworks that differ from the clinical batch anchoring approach used for potency and purity. Endotoxin limits by limulus amebocyte lysate assay are calculated from the dose and route of administration using the K/M formula specified in USP <85>, a chapter harmonized across the three ICH pharmacopoeias — USP <85>, Ph. Eur. 2.6.14, and JP 4.01 — under ICH Q4B Annex 14: the endotoxin limit equals K, the threshold pyrogenic dose per kilogram of body weight, divided by M, the maximum clinical dose in milligrams per kilogram per hour, with the parenteral route applying a K value of five endotoxin units per kilogram and the intrathecal route applying the more conservative K value of 0.2 endotoxin units per kilogram, both values specified within USP <85> itself. This is a deterministic calculation, not a clinical batch distribution problem — the limit is set by the clinical dose and route, not by process capability, and a specification limit that is more permissive than the calculated pharmacopoeial threshold is not compliant regardless of what the clinical batches showed.

Host cell protein limits are not governed by a universal numerical standard. FDA’s default position, derived from safety assessments performed under the ICH Q6B framework, treats HCP content as a product-specific safety attribute whose specification limit must be justified by the immunogenicity risk assessment. The HCP profile — which proteins are present at what levels — is determined by a combination of the purification process design and the specificity of the HCP ELISA used for detection. Because most HCP ELISAs developed against a host cell lysate do not achieve one hundred percent coverage of all HCP species, the specification limit for total HCP should be accompanied by orthogonal characterization data from mass spectrometry-based HCP identification studies demonstrating the identity and relative abundance of residual HCP species. HCP species with known adjuvant activity or known immunogenicity should be specifically identified and their concentrations justified by comparison to safety thresholds from the literature or from nonclinical immunogenicity studies. Host cell DNA limits follow FDA’s default of ten nanograms per dose with fragment size below two hundred base pairs — a risk-based threshold that traces to FDA’s February 2010 guidance, Characterization and Qualification of Cell Substrates and Other Biological Materials Used in the Production of Viral Vaccines for Infectious Disease Indications, which formalized the World Health Organization Expert Committee on Biological Standardization’s 1996 upward revision of the acceptable residual DNA mass per dose and added the two-hundred-base-pair fragment-size ceiling as a second, independent control on oncogenic risk. This ten-nanogram, two-hundred-base-pair default is more permissive than the specific one-hundred-picogram-per-dose figure in FDA’s earlier 1997 Points to Consider on monoclonal antibody products, and sponsors should cite the 2010 guidance — not the 1997 document — as the operative source when justifying the ten-nanogram default, unless the sponsor has safety data supporting a higher limit for the specific product and patient population.

USP chapters 1032, 1033, and 1034 provide the statistical framework for design and interpretation of bioassays, including the relative potency model, the parallel line assay, the slope ratio assay, and the methods for calculating confidence intervals on relative potency estimates. These chapters are not optional references for biological assay development — they define the analytical validation requirements that must be met before a relative potency result can be considered reliable, and the ICH Q2(R2) (final FDA guidance, March 2024) guidance on analytical procedure validation reinforces their applicability within the fit-for-purpose validation paradigm.

The consistent pattern in biologics specification deficiencies is not technical failure. It is documentation failure: the clinical data exist, the analytical methods are sound, the process is controlled, but the submission contains no document that assembles those data sources into an explicit derivation of each acceptance criterion. That document — the specification justification — is not a standard CTD section. It is an interpretive document that must be deliberately authored, structured around the three justification pillars of clinical relevance, analytical performance, and risk assessment, and submitted as part of the specification package for 3.2.S.4. Specifications submitted without it will generate deficiencies. Specifications submitted with it will not.

The XGene Biologics Specification Justification Framework

Clinical-Anchored Justification — Three Attribute Classes, Three Derivation Pathways

POTENCY — Acceptance Criterion Derivation Step 1: Compile the relative potency values (% vs. reference standard) for all clinical drug substance batches from the primary mechanism-of-action functional assay (mechanism-of-action-based bioassay preferred; binding assay acceptable as orthogonal confirmation per ICH Q6B). Construct the distribution and identify the observed clinical batch range with the mean and 95% confidence interval. Step 2: Quantify assay precision from the validated relative potency assay using the ICH Q2(R2) precision hierarchy (repeatability, intermediate precision, reproducibility). Express precision as the geometric coefficient of variation. Calculate the ±2SD precision contribution on the log-transformed potency scale. Step 3: Expand the clinical batch range by the ±2SD precision contribution to define the candidate acceptance criterion boundaries. Step 4: State the clinical relevance: cite the exposure-response characterization from clinical studies, confirm that no clinically significant variation in efficacy or safety was observed across the clinical batch potency range, and document that the derived criterion boundaries are consistent with the therapeutic window established in clinical development. Every potency acceptance criterion in the specification is traceable to a specific set of clinical batch potency measurements, a quantified precision estimate, and a clinical exposure-response statement.

PURITY — Shelf-Life Limit Derivation Limits for SEC-HPLC aggregates, nrCE-SDS fragments, and icIEF charge variants are derived from the degradation profile of clinical drug substance lots stored under the proposed commercial storage condition. The derivation uses the time-zero clinical batch distribution plus the expected degradation rate extrapolated to the proposed shelf life. The limit is set to accommodate the maximum projected end-of-shelf-life value across the clinical batch distribution, with the clinical immunogenicity data (for aggregates) and the functional characterization data (for charge variants and fragments) cited as the safety and activity justification for that level.

SAFETY — Risk-Based Limit Setting Endotoxin: Calculated limit per USP <85> K/M formula (K = 5 EU/kg ÷ M = maximum dose in mg/kg/hr for parenteral; K = 0.2 EU/kg for intrathecal), harmonized with Ph. Eur. 2.6.14 and JP 4.01 under ICH Q4B Annex 14. Not a process capability problem — the clinical dose and route determine the limit. HCP: Limit justified by the immunogenicity risk assessment: HCP ELISA method coverage assessment by mass spectrometry, identification of residual HCP species, immunogenicity risk classification of identified species, and confirmation that total HCP content in clinical lots did not correlate with elevated ADA incidence in clinical immunogenicity monitoring. HCD: FDA default ≤10 ng/dose with fragment size <200 bp per FDA’s February 2010 guidance on Characterization and Qualification of Cell Substrates and Other Biological Materials Used in the Production of Viral Vaccines (formalizing WHO ECBS’s 1996 revision); or product-specific limit supported by nonclinical safety data if the default cannot be met.

The XGene Biologics Specification Justification Framework produces a three-page specification justification document — one page per attribute class — that provides the BLA reviewer with a traceable, quantitative derivation of every acceptance criterion in 3.2.S.4. This document does not replace the specification itself; it is the scientific narrative that makes the specification defensible.