QC of biologics: what is tested and why

A small-molecule drug such as paracetamol is one defined chemical structure. It can be fully described by a few tests, and an impurity is simply a different molecule. A biologic — a monoclonal antibody, a hormone, an enzyme or a vaccine antigen — is a large protein made by living cells. Every batch is a mixture of closely related forms: some molecules carry slightly different sugars, some have lost a C-terminal lysine, a few have deamidated or oxidised, and a small fraction stick together as aggregates.

This natural variation is called microheterogeneity. QC for biologics does not try to prove the product is a single molecule. It shows that the mixture is consistent from batch to batch, matches the material used in clinical trials, and stays within limits that are known to be safe and effective. The manufacturing process itself is part of the product's definition, which is why process control and process-related impurities get so much attention.

A common phrase in the industry is 'the process is the product'. Changing the cell line, the culture conditions or the purification steps can change the product, even if the amino acid sequence stays the same.

ICH Q6B sets out how specifications for biotechnological and biological products are built. Release tests are grouped by what they tell you about the product. Each group usually needs more than one method, because no single technique sees everything.

Identity tests must be specific enough to tell the product apart from other products made in the same facility — often other antibodies with very similar structure. Peptide mapping is the classic identity test: the protein is cut with an enzyme and the peptide pattern is compared with the reference standard. With mass spectrometry it also confirms the amino acid sequence.

Many labs add a second, orthogonal identity test, such as the charge profile by cIEF or a binding test, because a mix-up between two similar antibodies must never pass release.

ICH Q6B separates the forms of the product into three groups. The desired product is the intended structure. Product-related substances are variants that form during manufacture or storage but have activity, efficacy and safety comparable to the desired product — many charge variants fall here. Product-related impurities are variants that do not have comparable properties, such as aggregates, fragments or inactive oxidised forms.

Because a variant can be a substance or an impurity depending on its effect, laboratories characterise variants carefully during development: isolate them, test their activity, and then decide how tightly each needs to be controlled.

Process-related impurities come from the way the product is made rather than from the product itself. They are mainly removed by the purification process, and the level remaining must be shown to be low and consistent.

• Host cell proteins (HCP): proteins from the production cells. Some may cause immune reactions; some enzymes can degrade the product or the polysorbate in the formulation. Usually measured by a sandwich ELISA and reported in ng per mg of product (ppm).
• Residual host cell DNA: DNA from the production cells. Measured by qPCR and reported per dose; a commonly applied limit is 10 ng per dose (typical, from WHO guidance).
• Leached process reagents: for example Protein A ligand leaching from the affinity column, measured by a specific ELISA.
• Media and process additives: antifoams, antibiotics, inducers, detergents or solvents used in culture or purification, often controlled by process validation rather than routine testing.
• Many process-related impurities are tested at drug substance and may move to periodic testing once the process has been shown to clear them reliably — this needs regulatory agreement.

Process validation studies often spike an impurity into an early process step and measure how much each purification step removes. This 'clearance' data supports reduced routine testing later.

Potency is the measure of biological activity linked to the product's intended action. Because two batches can look the same chemically but differ in activity, ICH Q6B expects a potency test for release. Potency is usually expressed relative to a reference standard (for example as % relative potency), not as an absolute number.

Potency assays may be cell-based (measuring a response of cultured cells to the product), binding-based (measuring binding to the target, for example by ELISA or surface plasmon resonance) or, for enzymes, a measure of enzymatic activity. A dose–response curve for the sample is compared with that of the reference standard, and the curves must be parallel for the comparison to be valid. Details of each assay are product-specific and are set out in the approved procedure.

Quantity is usually protein concentration measured by UV absorbance at 280 nm with the product's extinction coefficient, because it is fast, accurate and needs no standard curve. Colorimetric total protein assays (BCA, Bradford, Lowry) are used where A280 is unsuitable, for example in early process samples. Protein content is also needed to calculate other results, such as HCP in ng per mg.

Most biologics tests are comparative, so the in-house reference standard is central. It is a well-characterised batch, ideally linked to material used in clinical studies, stored under controlled conditions and replaced only through a documented qualification against the old standard.

Specifications under ICH Q6B are set from development data, batch history, clinical experience, stability data and method capability. Limits are therefore product-specific; the values quoted in method pages on this site are typical examples, not universal limits.

Manufacturers change processes over a product's life — scale-up, a new site, a new raw material supplier, a modified purification step. ICH Q5E describes how to show the product before and after the change is comparable: highly similar, with no adverse impact on safety or efficacy.

1. 1Assess the change and the quality attributes it might affect.
2. 2Test pre-change and post-change batches side by side with release methods and extended characterisation (e.g. peptide mapping with MS, glycan profiling, higher-order structure).
3. 3Compare against pre-set criteria, often derived from the historical range of pre-change batches.
4. 4Include stability or forced-degradation comparisons, because differences sometimes appear only under stress.
5. 5Decide whether quality data alone are enough or whether non-clinical or clinical data are also needed.

Proteins degrade by several routes at once — aggregation, fragmentation, deamidation, oxidation — so stability programmes for biologics (ICH Q5C) rely on a set of stability-indicating methods rather than one assay. Most products are stored at 2–8 °C, or frozen for drug substance, with accelerated and stress conditions used to understand degradation pathways.

• Typical stability-indicating tests: SEC-HPLC, charge variants, CE-SDS, potency, sub-visible particles, appearance and pH.
• Real-time data at the recommended storage condition support the shelf life; extrapolation is used cautiously for proteins.
• Forced degradation (heat, light, oxidation, agitation, pH) shows which methods detect which changes.