Nov 21, 2026
Biosimilar Analytical Similarity by HRMS: What the Package Should Contain
What a monoclonal antibody biosimilar's comparative analytical assessment should include, and where HRMS fits: structure, PTMs, glycans and lots.
\
For a monoclonal antibody biosimilar, the comparative analytical assessment is the foundation of the biosimilarity demonstration. Under FDA's October 2025 draft guidance, a comparative efficacy study may not be needed when the products are well characterized, the link between quality attributes and clinical performance is understood, and a human PK similarity study is feasible. That raises the bar for the analytical package itself.
This article lists what the structural part of the package typically contains and where high-resolution mass spectrometry (HRMS) adds the most.
Plan the comparison before running it
- Reference product lots: multiple lots, sourced over time and across expiry dates, so the natural variability of the reference product is captured. US-licensed and other-region lots need a bridging comparison if both are used.
- Risk ranking of attributes: each quality attribute is ranked by its potential impact on activity, PK, immunogenicity and safety. Higher-risk attributes get more sensitive methods and stricter comparison criteria.
- Comparison approach: quantitative attributes are commonly compared against quality ranges derived from reference lots; qualitative attributes are compared side by side. The statistical approach should be defined before data are generated.
Structural attributes and the methods that cover them
| Attribute | HRMS approach | Orthogonal methods |
|---|---|---|
| Amino acid sequence | Intact and reduced subunit mass; peptide mapping with multiple enzymes for full sequence coverage | MS/MS confirmation of N- and C-terminal peptides |
| Post-translational modifications (oxidation, deamidation, isomerization, glycation, N-terminal pyroglutamate, C-terminal lysine) | Site-specific relative quantification by peptide mapping | Charge variant analysis; forced degradation to show method sensitivity |
| Disulfide bonds and free thiols | Non-reduced peptide mapping | Free thiol assay; non-reduced CE-SDS |
| N-glycosylation (afucosylation, galactosylation, high mannose, sialylation, non-human glycans) | Glycopeptide and intact or subunit glycoform profiling | Released glycan analysis (HILIC with fluorescence and MS) |
| Charge variants | MS identification of acidic and basic species | cIEF or ion-exchange chromatography |
| Size variants | Native or SEC-MS identification of fragments and aggregates | SEC, CE-SDS |
Higher-order structure and function are compared too: biophysical methods for higher-order structure, and binding and cell-based potency assays aligned with the mechanism of action (for example, Fc receptor binding, ADCC or CDC where relevant). Glycan differences are interpreted together with these functional data, because afucosylation and galactosylation can change effector function.
Where HRMS makes the biggest difference
- Site-level detail. A charge variant profile shows that acidic species differ; peptide mapping shows which residue is modified and by how much.
- Low-abundance species. Sequence variants, glycation and minor glycoforms can be detected at levels that profile-based methods miss.
- Explaining differences. When a difference is found, HRMS identifies the species, which is the starting point for a scientific justification or a process adjustment.
- One dataset, many attributes. A multi-attribute approach can follow many attributes across reference and proposed lots consistently.
Common gaps to avoid
- Too few reference lots, or lots bought at one time, so ranges understate the reference product's variability
- Methods not shown to be sensitive to the differences they are meant to detect
- Glycan and charge data reported without linking differences to function
- Late discovery of a sequence variant or a non-human glycan that could have been caught early
More on this in the six most common analytical gaps in biosimilar programs and what FDA requires to approve a monoclonal antibody biosimilar in 2026.
How Crystal Bio Solutions supports biosimilar programs
At our Natick, Massachusetts laboratory, near Boston, the team develops and qualifies Orbitrap HRMS and orthogonal methods for side-by-side comparison of proposed and reference products: intact and subunit mass, multi-enzyme peptide mapping, PTMs, disulfides, glycans and charge variants, together with cell-based potency and effector function assays. See extended characterization by HRMS. PK similarity and immunogenicity assays for the clinical part of the program are validated and run at our Pleasanton, California bioanalysis laboratory.
Planning a similarity package? Discuss it with Dr. Billy Wu, our Chief Scientific Officer & Head of CMC Analytical
One hour at no cost with the leader who supervises our CMC analytical and HRMS projects, with a CDA in place first.
Frequently Asked Questions
What is a comparative analytical assessment?
A side-by-side comparison of the proposed biosimilar and the reference product across structural, physicochemical and functional attributes, using multiple lots of each.
Why use HRMS for biosimilar similarity?
HRMS confirms the amino acid sequence and measures modifications and glycoforms at the level of individual sites, which helps detect small differences and explain them.
How many reference product lots are needed?
There is no single fixed number for every product; enough lots, sourced over time and expiry, are needed to describe the reference product's variability with confidence.
Can strong analytics replace a comparative efficacy study?
Under FDA's October 2025 draft guidance, a comparative efficacy study may not be needed when specific conditions are met, including a robust comparative analytical assessment and a feasible human PK similarity study. The guidance is a draft, so the strategy should be confirmed with FDA.
Is higher-order structure part of the package?
Yes. Higher-order structure is compared with biophysical methods and supported by functional assays that reflect the mechanism of action.