This is a ready-to-use risk assessment for scoping a container closure integrity testing (CCIT) sampling strategy against the real batch population size of the product it protects, rather than applying a conventional AQL-based sampling table unmodified regardless of how many units a batch actually contains. It states, for each batch-size tier a program manufactures across, the sampling approach, the non-destructiveness and compatibility basis that makes that approach defensible, and the disposition path when a batch is too small to support a retest. Replace every <<FILL: ...>> placeholder with your own specifics. A filled specimen follows.
Document control header
| Field | Entry |
|---|---|
| Document title | CCIT Sampling Strategy and Batch-Size Tier Justification for <<FILL: product / process>> |
| Document number | <<FILL: e.g. CCIT-SAMP-2026-01>> |
| Version | <<FILL>> |
| Effective date | <<FILL>> |
| Linked CCIT study protocol | <<FILL: document number>> |
| Linked method selection and MALL risk assessment | <<FILL: document number>> |
| Document owner | <<FILL: e.g. Validation Lead / Quality Assurance>> |
1. Purpose
This assessment states the CCIT sampling approach for each batch-size tier <<FILL: product name>> is manufactured at, documents why each approach is statistically and physically defensible at that tier’s population size, and states the disposition path for a batch too small to support a conventional retest.
2. Batch-size tiers manufactured for this product
| Tier | Typical batch population | Manufacturing context |
|---|---|---|
<<FILL: e.g. Commercial multi-unit lot>> | <<FILL: e.g. 200 to 2,000 units>> | <<FILL: e.g. allogeneic drug product fill-finish campaign>> |
<<FILL: e.g. Small allogeneic lot>> | <<FILL: e.g. 5 to 40 units>> | <<FILL: e.g. single donor-derived intermediate>> |
<<FILL: e.g. Autologous, patient-specific>> | <<FILL: e.g. 1 to 3 units>> | <<FILL: e.g. one patient's own final cell therapy fill>> |
Mark not applicable for any tier this product does not manufacture at; do not leave a real tier blank.
3. Sampling approach and basis by tier
| Tier | Sampling approach | Statistical or physical basis |
|---|---|---|
| Commercial multi-unit lot | <<FILL: e.g. AQL sample per [cite the specific sampling table and inspection level used]>> | <<FILL: state why the sample size remains statistically representative of the released batch at this population>> |
| Small allogeneic lot | <<FILL: e.g. reduced sample size per risk assessment, or 100 percent non-destructive testing>> | <<FILL: state the risk basis for any reduced sample, or the non-destructiveness and compatibility data basis if 100 percent testing is used>> |
| Autologous, patient-specific | <<FILL: 100 percent non-destructive testing of every unit in the batch>> | <<FILL: cite the method validation report and product compatibility study proving the test itself does not consume or alter the dose>> |
4. Non-destructiveness and compatibility basis for 100 percent testing tiers
Complete this section for any tier using 100 percent testing of product-bound units rather than a sample drawn from a separate population.
| Field | Entry |
|---|---|
| CCIT method used | <<FILL: e.g. vacuum decay, mass extraction>> |
| Confirmed non-destructive to the container | <<FILL: cite method validation report>> |
| Confirmed not to alter or degrade the product itself (not just the container) | <<FILL: cite product compatibility study; state the specific exposure, e.g. vacuum level and duration, challenged>> |
| Units used to build the statistical validation (POD curve, MALL margin) | <<FILL: state that these were dedicated qualification/surrogate units, matrix-matched to the real product and package, never intended for patient administration>> |
5. Disposition path with no retest capacity
Complete this section for any tier where a batch may consist of a single unit or too few units to permit a second test.
| Question | Answer |
|---|---|
| Does any tier in Section 2 have batches where a failing or borderline result leaves no second unit to retest | <<FILL: yes/no>> |
| If yes, what is the pre-agreed disposition for a failing result on the only unit tested | <<FILL: state explicitly; do not leave this to be decided in the moment>> |
| If yes, what is the pre-agreed disposition for a genuinely borderline, unresolved result (see the investigation decision plan referenced below) | <<FILL>> |
| Who is notified, and on what timeline, when a zero-retest disposition is invoked | <<FILL: role and timeline>> |
6. Approvals
| Role | Name | Signature | Date |
|---|---|---|---|
| Author (Validation Lead / Statistician) | <<FILL>> | ||
| Reviewer (Method Owner / MSAT) | <<FILL>> | ||
| Approver (Quality Assurance) | <<FILL>> |
Filled specimen
Extract for an example autologous CAR-T product manufactured alongside a small-scale allogeneic comparator program. Details are illustrative.
| Tier | Typical batch population | Manufacturing context |
|---|---|---|
| Small allogeneic lot | 12 to 20 units | Single donor-derived intermediate, process development comparator runs |
| Autologous, patient-specific | 1 unit (occasionally 2, split dose) | One patient’s own final CAR-T cell suspension, cryopreserved in infusion bags |
Section 3, autologous tier: sampling approach is 100 percent non-destructive vacuum decay testing of every bag in the batch (one bag, or two for a split dose), with no sampling fraction, because the batch has no population larger than the dose itself to sample from. Basis: vacuum decay validated non-destructive to the cryobag per CCIT-VAL-2026-04, and product compatibility study CCIT-COMPAT-2026-02 confirmed the validated vacuum exposure (18 inHg, 12 seconds) produces no measurable change in post-thaw viability or potency assay result versus unexposed control bags from the same manufacturing run.
Section 4: the probability-of-detection curve and MALL margin supporting this method were built from 40 dedicated surrogate bags filled with placebo cryopreservation medium during method validation and PPQ, laser-drilled to certified hole sizes bracketing the MALL; none of those surrogate bags were patient material.
Section 5: a failing vacuum decay result on the single bag constituting a patient’s batch is an automatic reject with no retest, because no second bag exists; the attending treating site and the sponsor’s medical and quality leadership are notified within 2 hours of a confirmed failing result, per the escalation table in the linked CCIT investigation decision plan. A genuinely borderline, unresolved result (for example, a signal inside instrument noise with no confirmed assignable artifact) is also treated as a reject for this tier, since the clinical consequence of releasing an unconfirmed integrity result on a single-unit, non-retestable batch outweighs the consequence of a conservative reject.
Common inspection findings this justification prevents
- A conventional AQL sampling table applied to a patient-specific batch with no documented basis for why it fits a population of one to a few units.
- 100 percent testing of product-bound units with no stated proof the method is non-destructive to the product itself, not just the container, leaving an implicit, unverified assumption behind a practice the program is actually relying on.
- A disposition plan for a failing or borderline CCIT result that assumes a retest is always possible, with no stated path for the single-unit batch where one is not.
- Statistical validation (POD curve, MALL margin) built by consuming units from an actual production or patient batch rather than from dedicated qualification or surrogate material.
How to adapt this justification
- List every batch-size tier the product is actually manufactured at in Section 2; do not describe only the commercial tier if a clinical-stage or patient-specific tier also exists.
- For any 100 percent testing tier, confirm Section 4’s compatibility basis is a real, cited study, not an assumption carried over from a different method or product.
- Complete Section 5 for every tier where a batch could be too small to retest, even if the current pipeline has not yet hit that scenario; a gene or cell therapy program that is allogeneic today can add an autologous program later.
- Route this document through the same approval chain as the CCIT study protocol it supports, and cite it from that protocol’s sampling rationale rather than duplicating the tier table in two places.