EndotoxinBacterial Endotoxins Test (USP <85>)

Endotoxin is the lipopolysaccharide (LPS) component of the outer membrane of Gram-negative bacteria. When the bacterium is killed by terminal sterilisation, filtration or harsh manufacturing conditions, the LPS is released and remains pyrogenic — capable of inducing fever, septic shock, organ failure, and (in IV products) potentially fatal reactions in nanogram-per-kilogram doses. A sterile product can still be unsafe if it contains endotoxin above the per-product limit, because sterilisation kills the organism but does not destroy the LPS.

The maximum allowed endotoxin per dose is K/M, where K is the threshold pyrogenic dose per kilogram body weight (5 EU/kg for IV products, 0.2 EU/kg for intrathecal, 0.2 EU/kg for ophthalmic) and M is the maximum dose per kilogram body weight per hour. For a typical adult (70 kg) IV product dosed at 10 mL/kg/hour: K/M = 5 EU/kg ÷ 10 mL/kg = 0.5 EU/mL. The endotoxin limit on the spec is then 0.5 EU/mL for that product. Different routes have different K values; intrathecal products are tightly limited because the blood-brain barrier offers no clearance pathway.

• Gel-clot — qualitative or semi-quantitative; the simplest LAL test. Mix sample with LAL reagent, incubate 60 min, invert tube — a firm clot = positive at the labelled sensitivity. Cheap, robust, slow to interpret quantitatively. Still common for water-system release and routine batch testing of well-characterised products.
• Turbidimetric — kinetic (turbidity development rate vs time) or end-point quantitative LAL test. Higher throughput than gel-clot, full quantitation.
• Chromogenic — kinetic or end-point; LAL reagent contains a chromogenic substrate that releases a yellow chromophore on clotting cascade activation. Most popular routine quantitative method.
• Recombinant Factor C (rFC) — uses recombinant horseshoe-crab Factor C produced in cell culture instead of LAL extracted from live crabs. Equivalent or superior performance, no live-crab supply chain. USP <86> (2020+) provides a harmonised compendial method; EMA, FDA and PMDA all accept rFC.

Method suitability for BET requires demonstration that the product, at the chosen test dilution, doesn't inhibit or enhance the LAL reaction. The Maximum Valid Dilution (MVD) is the most-diluted point at which the test can be performed and still detect endotoxin below the product limit: MVD = endotoxin limit × concentration of sample / λ (the labelled sensitivity of the reagent). Inhibition/enhancement testing spikes the product (at the chosen dilution) with a known low endotoxin amount and demonstrates recovery within ±25 % to ±100 % depending on the method. A Positive Product Control (PPC) is run alongside every test to confirm the day's sample preparation didn't introduce inhibition.

1. Endotoxin limit calculated against the wrong M (e.g. used label dose rather than worst-case dosing per the prescribing information).
2. MVD pushed too high to mask inhibition, with the result that the test loses sensitivity to the actual product limit.
3. Inhibition/enhancement not re-validated after a formulation, container or filter change.
4. Reliance on the test alone for endotoxin control — no upstream Gram-negative bioburden trend, no WFI endotoxin trending, no investigation of slowly rising endotoxin in raw materials.
5. OOS on endotoxin handled as a re-test loop without quarantining the batch or investigating the upstream water system.
6. Switching from LAL to rFC without updating method suitability and the validated MVD.
7. Stability protocol omits periodic endotoxin time-points for parenteral products with known LPS-binding excipients.

• Product spec module holds the calculated endotoxin limit per product/dose/route with the K/M derivation as a controlled record; the limit recalculates automatically if the dose schedule changes (with change-control routing).
• LIMS captures every BET result as a critical release attribute, the method, the MVD, the PPC recovery and the reagent lot — a regulator sees the full chain without assembling spreadsheets.
• Water-system endotoxin trending (PW, WFI loop points) feeds the same trending engine as product-release BET so an excursion at a loop point is visible against batches that pulled from that loop on that day.
• OOS workflow: a positive or out-of-limit result triggers batch quarantine, an investigation with documented retest decisions, and recall-readiness if released stock could be impacted.
• rFC adoption: the platform supports the LAL → rFC transition as a method change-control event, with the parallel-testing record visible during the transition window.