GMP Cell-Therapy QC Counting: What Makes a Count Audit-Ready

Audit-ready cell counting is a count whose result and full data history you can prove — to Quality, to IT and data-integrity, and to an FDA reviewer.

That bar exists because in GMP cell-therapy manufacturing the count and viability go on the certificate of analysis. In an autologous CAR-T run the lot size is often one, so every number stands on its own, with no batch to average against and no second draw of the material.

That reframes what you are buying. A counting step that used to be housekeeping can now carry release-grade weight, and the question shifts from how accurate the counter is to whether its count will survive the audit. Two things settle that. The count has to be provable, its full data history holding up under 21 CFR Part 11, and it has to be trusted, concording with the reference method your lab already runs.

Up front: the instrument is research-use-only. It gives you a count, a viability read, and an identity check you can prove. Your validated method and quality system own the release decision.

What does audit-ready cell counting for GMP cell-therapy QC require?

An audit-ready count clears two bars at once: it has to be provable, and it has to be trusted. Provable means 21 CFR Part 11 data integrity. Trusted means concordance with the reference method you already run. Accuracy on its own is table stakes for any serious instrument.

That resolves into five requirements you can score an instrument against:

1. Native data integrity. An audit trail, electronic signatures, and role-based access native to the instrument, rather than assembled from third-party middleware.
2. Concordance with your reference method. The count and viability agree with the sizer or counter your lab has already validated.
3. Precision on hard samples. A tight, repeatable count on small cells, debris, and low concentrations, where a wide coefficient of variation turns two reads of one lot into a question you have to answer for an auditor.
4. Viability plus a light identity check in one read. Live/dead and a single CD4 or CAR marker at-line, without booking a core run for a routine confirmation.
5. A real validation path. An IQ/OQ package and a documented calibration cadence, so you can bring the instrument under GMP control.

Each requirement has a way to test it. The full scorecard, with the test for each, lives in the audit-ready cell counting guide. To score a single candidate against the data-integrity items on one page, use the 21 CFR Part 11 cell-counter evaluation checklist. The two that most often decide the purchase — provability and precision — get their own answers below.

Why is the cell count now a release-grade measurement in CAR-T manufacturing?

Because the count and viability are release criteria on the certificate of analysis, and in autologous CAR-T the lot is often a single patient's cells. When the batch size is one, there is no lot-to-lot average to settle a soft number. The count you report is the count of record.

That puts three readers behind every result. Quality has to accept the method; IT and data-integrity, the record. An FDA reviewer has to accept both, sometimes years later, from the audit trail alone. A count that reads correctly but cannot show who produced it, when, and what changed along the way is the gap an inspection is built to find.

So "accurate enough" stops being the whole question. A manual hemocytometer count, a borrowed core instrument no one validated for release, or a bench counter with no audit trail can each produce a plausible number. None of them produces a number you can prove after the fact.

Is accuracy the deciding factor when choosing a GMP cell counter?

No. Between serious instruments, accuracy is table stakes, and the decision turns on precision and provability instead.

Start with what is settled. In a peer-reviewed Penn Medicine CAR-T QC study, a benchtop volumetric analyzer showed no significant difference from a reference volumetric sizer on both count and size (Pajarillo et al., Cytotherapy 2024). When a candidate instrument already concords with the method your lab trusts, a further accuracy contest between the two is not where the risk lives, and a superiority claim on accuracy is not worth chasing.

Precision is where methods separate. A volumetric count holds a validated coefficient of variation under 2%, where image-based counters run 20 to 30%. That gap matters most where cell-therapy samples are hardest — on small cells, debris, and low concentrations, where an image-based count drifts. At the release gate, a wide CV is the reason two reads of the same lot disagree, and the disagreement you then have to explain. Which method to trust, and why the mechanism behind that CV gap favors direct volume measurement, is the subject of volumetric vs image-based cell counting for CAR-T QC.

What does "provable" mean for a cell count under 21 CFR Part 11?

A provable count is one whose every data-modifying action is recorded. Under 21 CFR Part 11, that means an audit trail that logs who did what and when, an electronic signature on every change, and role-based access that separates the operator from the reviewer and the administrator. Deletion is the fourth requirement, restricted to an administrator and logged when it happens.

Deletion is the detail that decides real audits. An audit trail that records a result but not its removal is exactly the gap a reviewer flags, and the concern cell-therapy QC teams raise most often when they vet a counter. A record has to show the number you kept and anything changed or discarded on the way to it.

Where compliance comes from matters as much as whether it is present. Data integrity built into the instrument is one validated system. Compliance assembled from third-party middleware adds a second system to validate, maintain, and audit. Of the Moxi family, only the Moxi GO II offers a 21 CFR Part 11 "Secure Mode" — an activated, paid audit-trail module (not enabled by default) that puts those Part 11 controls in the instrument's firmware rather than bolting them on (the Moxi V and Z do not). Treat that on-device record as one input to your data-integrity program, not a replacement for it. What Secure Mode does and does not cover, and how to score it against your own SOPs, is laid out in what 21 CFR Part 11 requires of a cell counter.

Can one instrument return count, viability, and identity in a single run?

Yes, with one limit. One run on a benchtop cell analyzer returns a volumetric count, a viability read, and one identity marker (a CD4 or a CAR marker) in about 10 to 15 seconds at the bench.

The mechanism is why the count holds up. The Moxi GO II measures every particle's true volume by impedance, up to 40,000 particles per test, rather than imaging a few hundred and inferring the rest. The same run adds 488 nm fluorescence for viability and a single identity marker. In the Penn Medicine study, that fluorescence channel tracked a reference fluorescence analyzer closely: viability at an r² of 0.97, CD4 at 0.997, and CAR19 at 0.90 (Pajarillo et al., Cytotherapy 2024).

The "one marker" is literal. The instrument has two fluorescence channels, and a viability stain consumes one, so a single run gives you viability plus one identity marker, not a multi-color panel. That is enough for a routine at-line confirmation that the cells are alive and are the right cells. The full identity or potency panel is a different job.

Where does a benchtop cell analyzer stop and a flow cytometry core begin?

A benchtop cell analyzer covers routine at-line counts, viability, and a one- or two-color identity check. A flow cytometry core covers the high-parameter identity and potency panels. The two are complements, not substitutes.

The dividing line is the number of parameters and the question you are asking. When you need to confirm at the bench that this lot is the right count, alive, and expressing its marker, a benchtop read gives you that in seconds and keeps the sample in the suite. When you need a full multi-color identity or potency characterization for release or a regulatory filing, that is core-facility work, and the benchtop does not replace it.

Instrument	What it covers
Benchtop cell analyzer	Routine at-line counts, viability, and a one- or two-color identity check — in seconds, keeping the sample in the suite.
Flow cytometry core	The high-parameter identity and potency panels for release or a regulatory filing.

Used together, the benchtop analyzer triages the routine in-process and at-line checks. That frees the core instrument and its schedule for the high-parameter work only it can do. If your primary release assay has to be a full high-parameter panel, a core-facility analyzer is the right path, and you will know that quickly.

How to pressure-test a counter before the audit does

Start with the count you already run. Score it against the five requirements, and put the two that decide most purchases — provability and precision — under the most pressure. The 21 CFR Part 11 cell-counter evaluation checklist turns the data-integrity items into a one-page score sheet, and the audit-ready cell counting guide carries the full five-requirement scorecard with the test for each.

When you are ready to bring your IT and Quality colleagues into the conversation, talk to a specialist to walk them through Secure Mode — the activated Part 11 audit-trail module — and the IQ/OQ package, so the count is proven long before an inspection asks the question.