Cell Culture Passage Record

A Cell Culture Passage Record is the structured execution record of a subculture event: detaching or otherwise harvesting cells from a source vessel/bioreactor, optionally fractionating, and reseeding into a destination vessel or bioreactor at a defined seeding density and split ratio. In regulated environments, it operates as a unit-level electronic batch record (eBR) that guarantees full parent–child genealogy, captures critical parameters and materials, and enforces procedural controls, limits, calculations, and signoffs.

The record links to equipment status (cleaning/sterilization and qualification), environmental monitoring, media and supplement lots, and QC release/hold states. It covers adherent and suspension culture passages, cell bank expansions, and pre- and post-passage checks (e.g., viability, microbial controls). Because passage changes cell phenotype and risk profile over time, rigorous documentation is necessary for identity, purity, potency, and safety claims.

• Scope: any subculture/split/expansion step that creates a new lot/child culture
• Applies to: adherent flask passage, microcarrier transfer, spinner/bioreactor inoculation, T/N plate reseeding, cell therapy dose expansion
• Record type: controlled eBMR step under MES with audit trail and e-signatures

In drug manufacturing, 21 CFR 211.188 mandates complete batch production and control records, which extend to unit operations like passages that materially transform intermediates and create new lots. For human cells, tissues, and cellular and tissue-based products (HCT/Ps), 21 CFR 1271.270 requires accurate, indelible, and retrievable records linking all steps to the specific HCT/P, including dates, personnel, equipment, and materials. Electronic systems must comply with 21 CFR Part 11 and EU GMP Annex 11 for trustworthy electronic records and e-signatures, including validated systems, audit trails, and secure user management.

Data integrity expectations (ALCOA+) from MHRA and PIC/S emphasize contemporaneous, attributable, complete, and enduring data. Passage records must support review and decision-making—release, further expansion, or discard—under the site PQS (e.g., ICH Q10 practices), with risk-based controls set according to the process lifecycle.

ISA‑95 frames how the passage record fits across enterprise-to-control integration. At Level 3 (MES), the passage is an operation with material transformations, equipment allocation, and quality data capture that produce a new lot. ISA‑88 complements this by defining recipe structures—unit procedures, operations, and phases—where a passage may consist of phases (e.g., detach, count, dilute, seed) with parameters, setpoints, and procedural logic. This layered model supports enforceable instructions, exception handling, and unambiguous genealogy.

Recipe parameters (target split ratio, target confluency window, seeding density) and procedural interlocks (media prepared and released, incubator qualified and within limits) are expressed as phase parameters and permissives. The MES stores the context: who executed, when, where, and with what materials and equipment—ensuring consistent interpretation across shifts and sites.

A high-fidelity passage record captures data necessary to reproduce outcomes and justify release decisions. Minimum elements include identity and genealogy, environmental conditions, equipment state, materials/reagents lots, in-process checks, and calculations. Calculations must be validated and traceable: e.g., cell yield, viability-adjusted seeding volume, and projected confluency time.

• Identity/genealogy: source lot, daughter lots, pooling/splitting logic, passage number/PDL
• Environment: incubator ID, temperature, CO2 %, O2 % (if controlled), relative humidity, exposure time outside incubator
• Process: detachment method (enzyme/mechanical), contact time, neutralization volume, centrifuge settings
• Counts/viability: method/instrument ID, raw counts, viability %, calculation audit trail
• Seeding: target density, split ratio, destination vessel(s), volume per vessel, occupancy
• Materials: media, supplements, enzymes, buffers—lot/expiry/COA reference and status
• Quality controls: sterility mycoplasma holds, endotoxin (where applicable), in-process morphology notes/images
• People/time: operator(s), verifier/witness, start/stop timestamps, interruptions and reasons

Numeric limits and contextual acceptance criteria should be enforced in real time: warning vs. action limits (e.g., viability < 80% triggers additional wash; < 60% requires supervisor approval or discard). Where images or instrument exports are used (e.g., automated cell counter CSV/PNG), the record should embed immutable attachments with hash or checksum metadata.

Passage events inherently create one-to-many relationships: a parent culture split into multiple vessels/bioreactors. Pooling may also occur (many-to-one). The passage record is the authoritative node that links input lot(s) to output lot(s), each with quantities, locations, and status. This enables backward and forward traceability for deviations, OOS results, complaint investigations, or recalls.

Key patterns include parent–child batch structures, split batch with proportional allocation of in-process attributes, and documented pools with full provenance. Each child lot inherits critical attributes (e.g., cumulative passage number or population doublings) and carries its own subsequent record chain. MES should maintain immutable genealogy graphs supporting queries such as “all lots derived from Master Cell Bank lot X via passages within date range Y–Z,” and “all end products batched with any daughter having viability < threshold at inoculation.”

• Parent–child linkage at each passage node
• Quantitative reconciliation before and after split/pool
• Propagation of constraints (e.g., maximum allowable passage number)
• One-up/one-down trace alignment with warehouse location and WIP status

Part 11/Annex 11 require secure, attributable e-signatures tied to a unique individual with appropriate authorization. Critical steps—enzyme exposure time confirmation, viability acceptance, and seeding density confirmation—should be configured as verifiable checkpoints with enforced signoffs. For high-risk operations, two-person verification is common: one executes, another independently verifies calculations and identity checks before proceeding.

Audit trails must capture who changed what and when, the original value, new value, and reason for change for GMP-relevant fields. Review-by-exception can surface out-of-limit entries, late data entries (post-execution), and missing attachments to quality reviewers, streamlining batch record review and lot disposition.

Under GAMP 5 (2nd ed.), an MES/eBMR passage workflow is typically Category 4/5 (configured/ bespoke) with significant risk to product quality and data integrity. A risk-based approach (aligned to ICH Q9(R1) principles) prioritizes testing of calculation logic (e.g., seeding volume), limit checks, electronic signatures, audit trail behavior, and integration points (LIMS, equipment data).

Validation deliverables should include URS explicitly covering genealogy, pooling/splitting, and passage-number handling; configuration specs for recipe phases and parameters; test protocols demonstrating traceability of all outputs to inputs; and Part 11/Annex 11 controls (access, security, audit trails, e-signatures). Change control must handle procedural changes (e.g., new cell counter model), parameter limit updates, and integration endpoint changes with impact assessment on data integrity and release decisions.

• Risk scenarios: incorrect passage number propagation; time zone misalignment for incubator data; failed LIMS result posting during hold
• Controls: hard stops for exceeding max passage; server-time stamping; message queuing and reconciliation for interfaces
• Evidence: test records showing audit trail content, dual-signature enforcement, and exception reports

Passage records often gate subsequent operations pending QC results: sterility/mycoplasma holds, endotoxin for certain products, identity markers, in-process morphology checks. The MES should create QC samples with chain-of-custody, push testing requests to LIMS, and subscribe to result status updates that automatically change lot state (hold/release/reject) with proper authorization and audit trails.

Environmental monitoring (EM) context—room grade, personnel monitoring, excursion logs—should be referenced to the passage timestamp and location. Where bioreactor or incubator telemetry exists (DO, pH, temperature, CO2), attachments or normalized data links should be embedded. If a QC result later invalidates a passage (e.g., contamination), the genealogy graph supports impact assessment and containment actions across all derived lots.

• Automated sample generation and label printing at passage completion
• Result mapping to defined acceptance criteria and lot-state transitions
• Exception handling for partial result sets and retests

Passage timing is biologically driven (confluency, growth kinetics) and resource constrained (biosafety cabinets, incubators, operators). ISA‑95 Level 3 scheduling should support forecasting windows for each culture based on last known growth rate and passage history, then allocate equipment and personnel with conflict detection. The record must capture start/stop timestamps and any dwell time outside controlled conditions.

Exception rules should flag late or early passages relative to defined windows, prompting risk assessment and additional checks (e.g., viability threshold raised). Integration with maintenance status prevents starting a passage on equipment under calibration/qualification hold. Workload leveling can reduce batch effects across parallel passages, improving consistency.

• Dynamic windows: predicted confluency vs. operational availability
• Interlocks: equipment status qualified; cleaning/sterilization verified
• Escalations: late passage triggers supervisor approval and CAPA evaluation

Frequent issues include mis-numbered passages, manual transcription errors from counters and balances, untracked pooling/splitting, missing attachments (images, counter exports), and incomplete reconciliation of volumes and vessels. Time synchronization problems can lead to mismatched incubator/EM timestamps, undermining root cause analysis.

• Enforce computed passage number from genealogy rather than free text
• Interface cell counters and balances to auto-populate raw values and calculations
• Require explicit split/pool declarations with quantitative reconciliation
• Use server-synchronized timestamps for all devices and MES entries
• Hard-stop on missing critical attachments and unresolved deviations before approval

V5 models passages as ISA‑88 phases within MES operations and maintains a native ISA‑95 genealogy graph. It enforces interlocks to equipment state (Maintenance/Calibration), material status (WMS), and QC holds (LIMS). Data integrity is built-in with Part 11/Annex 11 controls, audit trails on critical fields, and two-person e-signatures where required. Automated sampling, label generation, and LIMS orders are triggered at passage completion, and result feeds drive lot-state transitions under role-based approvals.

The record aggregates evidence on one screen—procedural data, attachments (cell counter exports, images), environmental context, deviations/CAPAs, and training status of operators—supporting review-by-exception and rapid disposition. Parent–child splits and pools are configured templates with validated calculations and reconciliation, minimizing manual effort while preserving flexibility for different cell lines or modalities.