For anything intended to be injected, the single most important safety question is one that a purity percentage cannot answer. A vial can be 99 percent pure, correctly identified by mass spectrometry, and still carry enough bacterial endotoxin to cause fever or, at high dose, far worse. Endotoxin testing is the assay that closes that gap, and this guide explains what endotoxin is, why an identity or purity certificate does not cover it, and how the USP <85> Bacterial Endotoxins Test actually measures it.
What bacterial endotoxin actually is
Bacterial endotoxins are lipopolysaccharides (LPS), the major glycolipid of the outer membrane of Gram-negative bacteria. Structurally, an LPS molecule is built from three regions: the O-antigen (the distal polysaccharide), the core oligosaccharide, and lipid A, the hydrophobic anchor that embeds the molecule in the bacterial membrane. Lipid A is the most bioactive component of LPS and is responsible for its endotoxic activity (Biochemistry, Lipopolysaccharide, StatPearls).
LPS is a potent pyrogen, meaning a fever-causing agent. Through recognition by the TLR4/MD-2 receptor complex, it triggers the production of cytokines such as TNF, IL-1, IL-6, and IL-8, which cause fever and, at high dose, endotoxic (septic) shock. Because endotoxin is a structural molecule shed from the bacterial membrane rather than a living organism, it is a distinct contaminant from viable microbes: it is a separate question from sterility, measured by a separate test.
Why a purity or identity COA does not cover it
This is the crux for anyone reading a certificate. A standard analytical panel confirms what a molecule is and how pure it is, but those are chromatographic and mass-spectrometric measurements of composition, not biological safety assays. A high purity result does not imply a low endotoxin load; they are entirely separate assays measured by entirely different methods.
This blind spot is documented, not subtle. Our guide to reading a peptide Certificate of Analysis notes that research-only material is often not tested for endotoxin at all, a gap worth knowing about. The companion analysis of third-party testing claims makes the same point in stronger terms: a purity or identity COA does not by default cover endotoxin, sterility, or heavy metals, and it is a gap that no amount of marketing language can close. Injection safety in fact decomposes into several independent tests, a framework set out in our overview of what makes a research peptide safe, and endotoxin is the one attribute a purity number is least able to stand in for.
USP <85>, the compendial reference method
The compendial reference method for endotoxin is USP General Chapter <85>, the Bacterial Endotoxins Test. It is a harmonized standard, agreed through the Pharmacopeial Discussion Group (PDG) with the Japanese Pharmacopoeia acting as the coordinating pharmacopoeia (USP, Bacterial Endotoxins harmonization).
The test uses Limulus Amebocyte Lysate (LAL), an aqueous extract of the circulating amebocytes (the blood cells) of the horseshoe crab. When LAL contacts endotoxin, an enzymatic cascade is activated, and USP <85> defines three techniques for reading that reaction (USP General Chapter <85>, full text).
The three LAL techniques
Under USP <85>, the three techniques trace the same underlying cascade but differ in how the result is detected and quantified:
- Gel-clot detects or quantifies endotoxin based on clotting of the lysate reagent. It is the simplest format, read as the presence or absence of a firm gel at a defined sensitivity.
- Turbidimetric is based on the development of turbidity after cleavage of an endogenous substrate in the lysate.
- Chromogenic is based on the development of colour after cleavage of a synthetic peptide-chromogen complex.
Which method decides a dispute
USP <85> is explicit about precedence. In case of dispute, the final decision is based on the gel-clot technique unless otherwise indicated in the monograph. That makes gel-clot the referee method, even though the turbidimetric and chromogenic formats are often preferred in routine work for their quantitative range and throughput.
The endotoxin limit: EL = K/M
An endotoxin result only means something when it is read against a defined limit. USP <85> sets the endotoxin limit for a product using the formula EL = K/M, where K is the threshold pyrogenic dose of endotoxin per kilogram of body weight per hour and M is the maximum recommended human dose of product per kilogram of body weight per hour. The units are endotoxin units, which USP expresses as USP-EU.
The value of K depends on the route of administration. For parenteral routes other than intrathecal, K = 5 USP-EU per kilogram of body weight per hour. For intrathecal administration, which places the product in contact with cerebrospinal fluid, K = 0.2 USP-EU per kilogram, a far stricter ceiling that reflects how sensitive that compartment is to pyrogens. Because the limit scales inversely with dose (M), a product given at a higher dose per kilogram has a lower allowable endotoxin concentration, which is why the limit is always calculated for the specific product and route rather than quoted as a single universal number.
When endotoxin testing is required
The Bacterial Endotoxins Test is not an optional extra in regulated manufacturing. It is performed as lot-release testing for injectable pharmaceutical products, and for medical devices that have direct or indirect contact with the cardiovascular system, the lymphatic system, or cerebrospinal fluid. "Lot release" is the operative phrase: the test is run on each batch before that batch can be released, because endotoxin contamination is a batch-specific property that cannot be inferred from an earlier run.
On the regulatory side, the FDA's guidance for industry, Pyrogen and Endotoxins Testing: Questions and Answers, addresses the testing recommendations and acceptance criteria in USP <85> and in USP <161>, the chapter covering bacterial endotoxin and pyrogen tests for medical devices (FDA guidance).
From the rabbit test to recombinant reagents
Endotoxin testing has a history that explains why the LAL test exists at all. The Bacterial Endotoxins Test was added to the USP in 1980 (USP 20) and has largely replaced the in vivo Rabbit Pyrogen Test of USP <151>. The rationale is that endotoxins from Gram-negative bacteria are the most potent and prevalent pyrogens encountered in parenteral manufacturing, so an in vitro assay aimed specifically at endotoxin covers the dominant risk; a validated equivalent in vitro pyrogen or bacterial endotoxin test may be used in place of the rabbit test where appropriate (phasing out the rabbit pyrogen test, meeting report).
The method continues to move away from animal-derived reagents. USP introduced General Chapter <86>, Bacterial Endotoxins Test Using Recombinant Reagents, covering recombinant Factor C (rFC) and Recombinant Cascade Reagent (rCR) as non-animal alternatives to horseshoe-crab LAL. It was published for early adoption on 22 August 2023 (USP-NF notice for <86>). For a reader, the practical point is that a modern endotoxin result may be generated by LAL under USP <85> or by a recombinant reagent under USP <86>, and the method used should be stated on the report.
Reading an endotoxin result in practice
Given all of the above, a credible endotoxin line on a certificate should let you answer three questions. First, which method produced it: gel-clot, turbidimetric, or chromogenic under USP <85>, or a recombinant reagent under USP <86>. Second, what the numeric result is and in what units, since a bare "pass" with no value and no calculated limit is not verifiable. Third, whether the result is tied to the specific batch in front of you, because endotoxin is a lot-specific property that a report from another run cannot vouch for. Treating a laboratory line as evidence rather than reassurance is the same discipline that underlies reading any analytical report. An endotoxin figure with no method, no unit, and no batch reference tells you very little, however reassuring the surrounding language.
The bottom line
Endotoxin testing measures a specific and serious hazard that identity and purity assays are structurally unable to detect. USP <85> defines the reference method: an LAL reagent read by the gel-clot, turbidimetric, or chromogenic technique, with gel-clot as the deciding method in a dispute, and a product-specific limit set by EL = K/M (K = 5 USP-EU/kg for most parenteral routes, and 0.2 USP-EU/kg for intrathecal). For any material meant to be injected, an endotoxin result is not a nicety layered on top of a purity certificate. It is a separate answer to a separate question, and its absence is a real gap rather than a rounding error.