Endotoxin Test Routing

Endotoxin test routing is the MES-governed definition and execution flow for bacterial endotoxin testing across the manufacturing lifecycle. It specifies when sampling is triggered (e.g., pre-sterile filtration, post-fill, point-of-use WFI), how samples are identified and handled, which LAL method/version is applied, what limits govern accept/reject, and how results gate status changes such as line clearance, lot release, or device DHR completion.

Routing includes integration patterns to LIMS, assignment of instruments and analysts, result verification, handling of inhibition/enhancement, and decision branches for OOS or invalid results. It binds quality test definitions to ISA‑88/ISA‑95 master data so that every test is contemporaneously attributable to a batch, unit, or device lot with full genealogy and data integrity controls.

For parenterals, 21 CFR 211.167 requires appropriate laboratory testing to determine conformance to specifications for sterility, pyrogens, and particulate matter. FDA’s Pyrogen and Endotoxins Testing Q&A clarifies acceptable LAL methods, controls for inhibition/enhancement, system suitability, and validation expectations, and recognizes harmonized pharmacopeial approaches. EU GMP Annex 1 expects robust bioburden and endotoxin controls in sterile manufacturing and associated utilities (e.g., WFI). For medical devices with blood or cerebrospinal fluid contact, FDA expectations include justified endotoxin limits and validated extraction and test methods as part of premarket and process controls.

"Appropriate laboratory testing shall be conducted to determine conformance to final specifications for sterility and pyrogens for parenteral drug products."

• Scope: Parenteral drug products, certain implantable/critical medical devices, and sterile manufacturing utilities.
• Methods: Gel-clot, turbidimetric, chromogenic LAL—validated per method and product matrix.
• Controls: System suitability, positive product controls (spikes), and inhibition/enhancement assessments.

Endotoxin test routing is a Level 3 (Manufacturing Operations Management) quality operation in ISA‑95, intersecting Production Operations via holds, interlocks, and release-by-test. It is defined in Operations Definitions (quality test specifications, sampling plans) and executed within Operations Requests triggered by the batch/lot eBMR or device eDHR. ISA‑88 batch models link the routing to unit procedures (e.g., Fill, Filter, Aseptic Hold) where permissives enforce no-start/no-release until results are accepted.

A robust routing decomposes the sampling and testing lifecycle into deterministic steps governed by specifications and risk. Define the sampling triggers (time- or event-based), sample containers and handling, label schema, and chain-of-custody. Reference the product- and matrix-specific LAL method and its validated parameters (λ sensitivity, MVD, spike recovery). Establish alert/action limits and equivalence rules (e.g., EU/mL, EU/mg, EU/device) aligned with intended use and compendial guidance.

• Sampling points: WFI storage/POU, pre-filter bulk, post-filter bulk, fill tank, finished units, post-cleaning rinse for devices.
• Sample equivalence: Define lot/test unit boundaries and pooling rules; enforce maximum pool sizes per validation.
• Gating: Electronic holds on batch/lot until accepted results; permissives to start/continue critical steps (e.g., aseptic filling).
• Retest policy: Pre-defined criteria for invalid assays (e.g., failed controls) vs. OOS, linked to QMS deviations.

At execution, MES raises LIMS test orders with all critical metadata (batch, lot, matrix, method version, limits, sampling point). LIMS registers samples, prints labels, and orchestrates the LAL run with instrument connectivity. Results, system suitability, and inhibition/enhancement data flow back to MES for automated evaluation against routing rules; any failures auto-create deviations and enforce continued hold. ISA‑95/B2MML-style interfaces or modern APIs enable reliable, timestamped, and attributable exchanges.

• Uniqueness: Sample IDs are immutable and link to batch genealogy and equipment/utility states.
• Completeness: Result packets include raw signals, calculations, analyst ID, instrument ID, and audit trail references.
• Timeliness: Store-and-forward with queuing ensures no data loss during network disruptions (aligned to NIST ICS resilience).

Routing encodes the decision tree from system suitability through final batch disposition. Acceptance must consider method-specific criteria (λ confirmation, spike recovery within limits, blank controls), applicable specification units (EU/mL, EU/mg, EU/device), and matrix-specific MVD. The MES should evaluate the entire result set—including invalid/aborted runs—to avoid cherry-picking and to drive the correct quality workflow.

1. Verify system suitability and controls; if failed, mark assay invalid and trigger retest policy.
2. Check inhibition/enhancement; if out-of-range, apply validated dilution series per method; if unresolved, escalate deviation.
3. Evaluate reportable result vs. specification; if OOS, place/maintain batch hold and open deviation/OOS per QMS.
4. On acceptance, auto-release gate(s) and update batch record; propagate to ERP/ERP-QP release where applicable.

Electronic routing and results must satisfy Part 11 and Annex 11 expectations: unique user credentials, role-based access, secure time-synchronized audit trails for all GMP-significant actions, and binding e-signatures for review/approval. PIC/S data integrity guidance and MHRA principles reinforce ALCOA+ across sampling, instrument operation, result calculation, and transcriptions. Maintain configuration baselines for methods/limits; changes are governed by change control and captured in the audit trail with rationale and impact assessment.

• Audit trail review: Periodic, risk-based review of critical events (result edits, method/limit changes, reprocessing).
• Attributable records: Sample labels link to operator, time, place; raw data retained in original format and human-readable form.
• Security: Segregation of duties—analyst, independent reviewer, QA approver; enforced via RBAC and dual e-signatures when required.

Apply a GAMP 5, risk-based lifecycle to endotoxin routing spanning MES, LIMS, and instrument interfaces. Define configuration specifications for sampling plans, limits, and decision logic; qualify interfaces and instrument dataflows; and verify end-to-end in PQ with realistic scenarios (invalid runs, inhibition, OOS, partial results, network outage). Leverage supplier documentation where appropriate, but retain user requirements, traceability, and test evidence in a controlled repository.

• IQ/OQ/PQ: Include interface failure modes (duplicate samples, delayed results) and authorization boundary tests.
• Periodic review: Confirm method versions, λ values, MVD, and limits remain current and effective.
• Change control: Impact assess to batches-in-flight; migrate routing updates with effective dates and versioned records.

Well-designed routing shortens release lead time without compromising control. Track KPIs such as sample-to-result cycle time, test first-pass yield, retest/invalid rate, inhibition occurrence rate, and contribution to batch release critical path. Trend utility (WFI) endotoxin results for state-of-control and tie alert limits to preventive maintenance and sanitization triggers.

• Unit mismatches (EU/mL vs EU/mg vs EU/device) embedded incorrectly in routing logic.
• Outdated λ sensitivity or MVD not aligned to current method validation.
• Pooling beyond validated limits or changing pool composition without change control.
• Transcription of results without independent verification or audit trail review.
• Release before completion of spike recovery or system suitability.
• Ambiguous sample equivalence causing under-testing of split or merged lots.

V5 models endotoxin routing as versioned, executable specifications at ISA‑95 Level 3, linking sampling events in MES to LIMS test orders and back to eBMR/eDHR release gates. Results, raw data references, and review/approval e-signatures live on one record; OOS, invalid assays, and re-sampling spawn QMS workflows automatically, and ERP disposition updates occur only after QA approval. Instrument connectors capture LAL outputs with metadata to sustain Part 11/Annex 11 expectations and simplify audit trail review.