ICH's Q5 series covers the quality of biotechnological products end to end — but Q5B is the one guideline in that series most CMC teams treat as a formality: a one-time sequencing report filed before the real cell-bank work begins. That gets the sequence backwards. Q5B exists because every claim Q5D's cell-substrate characterization makes about identity, purity, and genetic stability rests on one unexamined assumption — that the expression construct inside the production cell is exactly the one you characterized in the first place.
What Q5B actually characterizes
Q5B reached ICH Step 4 on November 30, 1995, and describes the expression construct as the expression vector containing the coding sequence of the recombinant protein. Its scope is narrow and deliberate: nucleic acid techniques — restriction endonuclease mapping and sequencing of the coding region, regulatory elements, and flanking sequences — applied to the construct itself, used in conjunction with testing on the purified recombinant protein. It covers both eukaryotic and prokaryotic expression systems, and it says nothing about cell-bank derivation, adventitious agents, or tumorigenicity — that work belongs to Q5D.
Why teams conflate Q5B and Q5D
Q5D gets the attention: it governs tumorigenicity and adventitious-agent testing, it shows up constantly in cell and gene therapy programs, and it carries the regulatory weight that keeps CMC teams focused on the cell bank. Construct-level work gets folded into the same deliverable as an afterthought — a sequencing certificate attached to the cell-banking report rather than a characterization package in its own right. The practical risk is that every downstream claim inherits the gap. Q5D's genetic-stability testing and the Q5A(R2) viral-safety evaluation both assume the construct is fixed, known, and fully described — and neither can verify that assumption on its own.
- Q5B asks what is in the cell. Structure, sequence, copy number, and physical state — integrated or episomal — of the expression construct itself.
- Q5D asks what the cell substrate is and does. Identity, purity, and genetic stability of the bank, plus adventitious-agent and tumorigenicity testing where relevant.
- Q5B's record is the reference point. Q5D's genetic-stability testing and Q5E's comparability assessments both measure against the construct Q5B characterized, not a general description of the cell line.
- Neither substitutes for the other. A complete cell-bank characterization package with a thin construct-analysis section is still an incomplete package.
A cell bank is only as well-characterized as the construct inside it. Skip the construct-level work and every later genetic-stability claim is resting on an assumption, not a record. Why construct analysis comes first
Carrying the construct record forward
A promoter swap, a codon-optimized recoding, a vector-backbone change — any of these is a change to the expression construct, and each one is exactly the kind of event ICH Q5E's comparability framework exists to assess. That assessment is only tractable if there is a rigorous Q5B baseline to compare against; sponsors who treated the original construct analysis as a formality end up re-characterizing from scratch instead of running a targeted comparability study. The same logic increasingly shows up outside Q5B's original scope: cell and gene therapy sponsors routinely apply its sequence-confirmation and copy-number techniques to plasmid and viral-vector constructs, even though Q5B itself was written for recombinant protein products.
- Confirm the construct structure. Restriction mapping and sequencing of the coding region and regulatory elements, before the construct goes into the production cell.
- Establish copy number and physical state. Integrated versus episomal, and how many copies — both feed Q5D's later genetic-stability testing.
- Carry the record into cell banking. Treat the Q5B characterization as the baseline Q5D's bank testing verifies against, not a standalone filing.
- Route construct changes through Q5E. Any later change to the construct is a comparability event measured against the original Q5B record.
None of this requires new testing infrastructure. It requires treating the construct-analysis package as a first-class deliverable — sequence-confirmed, copy-number-characterized, and filed where the cell-bank and comparability teams can actually use it — rather than a sequencing report that gets attached and forgotten. Programs that get this right spend less time re-deriving construct history when a change comes, because the Q5B baseline was built to be compared against from the start. Our biotechnology regulatory strategy work starts exactly here, before the cell-bank characterization is built on top of it.
Frequently asked questions
Is ICH Q5B the same as ICH Q5D?
No. Q5B governs analysis of the expression construct itself — the vector and coding sequence introduced into the production cell, verified through restriction mapping and nucleic acid sequencing. Q5D governs derivation and characterization of the resulting cell substrate: the master and working cell banks, their identity, purity, genetic stability, and — where relevant — adventitious agent and tumorigenicity testing. Q5B is the construct-level record Q5D's cell-bank claims depend on.
Does ICH Q5B apply to cell and gene therapy products?
Q5B was written in 1995 for recombinant DNA-derived protein products made in eukaryotic and prokaryotic expression systems. Sponsors commonly extend its construct-characterization logic — sequence confirmation, copy number, physical state — to plasmid and viral-vector constructs in cell and gene therapy, but Q5B itself does not govern those modalities directly. Confirm current FDA and ICH expectations for your specific construct type rather than assuming Q5B's scope covers it by default.
What happens if the expression construct changes after initial characterization?
A change to the expression construct is a manufacturing change, and it is exactly the kind of change ICH Q5E's comparability framework exists to assess. The original Q5B characterization is the reference point that comparability testing measures the post-change construct against — which only works if that original record is complete enough to compare to.
Sources & further reading
- FDA. Guidance for Industry — Q5B Quality of Biotechnological Products: Analysis of the Expression Construct in Cells Used for Production of r-DNA Derived Protein Products (ICH, Step 4, Nov. 30, 1995). fda.gov
- EMA. ICH Q5B — Analysis of the expression construct in cells used for production of r-DNA derived protein products — scientific guideline. ema.europa.eu
This article is provided for general informational purposes and reflects the regulatory landscape as of October 2026. It is not legal or regulatory advice. Confirm current ICH Q5B, Q5D, and Q5E expectations with FDA, EMA, or qualified counsel before acting.
