AssayIndex
Microbiological

Bacterial endotoxin testing

Kinetic chromogenic, turbidimetric and gel-clot LAL; why the index specification is a screening threshold and not a limit; and the dose arithmetic that gives a real limit.

method reference3 sections4 pharmacopoeial references

Bacterial endotoxin is lipopolysaccharide from the outer membrane of gram-negative bacteria. It is pyrogenic, it is not removed by sterilising filtration, it is stable to autoclaving, and no chromatographic result bears on it whatsoever. A 99.4 %-pure peptide can carry a substantial endotoxin burden and a purity certificate will never show it. It is the measurement whose absence from a certificate is least defensible, and it is absent from roughly three quarters of the material in this index.

The three methods

MethodPrincipleSensitivityOutput
Kinetic chromogenicLAL cascade cleaves a chromogenic substrate; onset time is read against a standard curve0.005 EU/mLQuantitative, wide range
Kinetic turbidimetricCoagulation increases turbidity; onset time read photometrically0.01 EU/mLQuantitative
Gel-clotPresence or absence of a firm gel at the labelled sensitivity λ0.03 EU/mL typicalLimit test — pass/fail only
Recombinant factor CRecombinant cascade, fluorogenic read-out; no horseshoe-crab lysate0.005 EU/mLQuantitative

A gel-clot result is a pass or a fail against a stated limit and cannot be converted into a number. Where a certificate reports "< 10 EU/mg by gel-clot", that is the whole information content; it does not mean the material is at 9.9 EU/mg and it does not mean it is at 0.1 EU/mg.

Interference is the hard part

Peptides interfere with the LAL cascade in both directions — cationic peptides in particular inhibit it, and some samples enhance it. The test is therefore invalid without an inhibition/enhancement control: a known spike of standard endotoxin added to the sample must recover at 50–200 %. Where it does not, the sample must be diluted until it does, subject to the maximum valid dilution. Reports that omit the recovery figure are recorded by the index as endotoxin-not-validated.

The specification versus a real limit

The index applies a screening threshold of 5.0 EU/mg across all classes. That number is the index's own and it is chosen so that the flag is informative across this dataset. It is not a pharmacopoeial limit and it is not a safety statement.

A real limit is dose-dependent. The pharmacopoeial threshold for parenteral administration is K = 5.0 EU per kilogram of body weight per hour (0.2 EU/kg for intrathecal routes). Converting that to a per-milligram specification requires knowing the dose:

Endotoxin limit  =  K × W / M

  K = 5.0 EU/kg/h        threshold pyrogenic dose, parenteral
  W = body weight (kg)
  M = maximum mass of substance administered per kg per hour (mg)

Worked: a 70 kg subject, 2.5 mg of compound per dose:

  maximum permitted endotoxin per dose = 5.0 × 70   = 350 EU
  limit for the material               = 350 / 2.5  = 140 EU/mg

Worked: the same subject, 500 mg of compound per dose:

  maximum permitted endotoxin per dose = 350 EU
  limit for the material               = 350 / 500  = 0.70 EU/mg

Same threshold. Two limits differing by 200-fold.
The dose, not the compound, sets the specification.

Normative references

Cited because they are the documents the acceptance criteria above are taken from. The index applies them as written and states every deviation.

  • USP <85> — Bacterial endotoxins test
  • Ph. Eur. 2.6.14 — Bacterial endotoxins
  • Ph. Eur. 2.6.30 — Monocyte activation test
  • USP <1085> — Guidelines on endotoxin testing

Related methods

Sterility testing

Membrane filtration and direct inoculation to USP <71>: the media, the fourteen days, and why a passed sterility test on one vial says very little.

Microbiological