Peptide Research Literature Index
A structured, dated index of the research-peptide landscape

What does an endotoxin number prove? FDA has withdrawn the only general limit table it ever published, the liquid being tested can hide the contamination, and for research material no rule requires the test at all

Reference page · published 2026-11-06

Disclosure: this index is operated by Artemis Labs, which sells research compounds to laboratories. Artemis Labs publishes no endotoxin testing of any kind and makes no testing claim anywhere on its store — the section on why this page exists reports how we measured that, including the two strings that looked like hits and were not. We have no commercial relationship with any testing laboratory, reagent maker or standards body named here, and we are not asking anyone to buy anything on the strength of this page. Nothing here is legal advice and nothing here is guidance for using any substance. Every quotation below is verbatim from an FDA guidance document, an FDA announcement, the Code of Federal Regulations, or a peer-reviewed paper, each fetched and read for this page on November 6, 2026 and each listed under Sources with its identifier. Two limits belong at the top. We are reporting what documents say about a laboratory test, not what any vial contains; and where the published literature disagrees with itself, which on this subject it does, the disagreement is set out, not resolved.

An endotoxin figure looks like the most solid thing on a vendor’s report. Purity is a percentage that depends on which method you ran. Identity is a mass that could belong to more than one molecule. But endotoxin arrives as a number with a unit attached, produced by a named assay, and it reads like a pass.

It is a real measurement. The question this page asks is narrower and more useful: once you have that number, what do the governing documents let you conclude from it? We went and read them. The answer has three parts, and none of them is the part people expect.

The short version. (1) There is no general endotoxin limit to compare a number against. FDA published one table of limits, in 1987, withdrew it, and the current edition of its guidance — Edition 2, dated March 2026 — says the table is out of date and that the limit must be calculated for the specific product. A number with no calculated limit printed beside it is not a verdict; it is an input. (2) The test can be blinded by the liquid it is run on. The literature has called this Low Endotoxin Recovery since at least 2018, still disputes its mechanism in papers published this year, and a 2019 study concludes that it is impossible to predict the susceptibility of bacterial endotoxin contamination to LER. (3) For research-grade material the test is not required by anything. A 2024 paper states it in one clause: for research and pre-clinical compound analysis, endotoxin testing is not a required quality control. So its presence is not compliance with a rule, and its absence is not a breach of one.

What the test is, and where the reagent comes from

The standard test for it runs on blood from a horseshoe crab. That is not a colorful framing; it is the mechanism, and it is the reason the rest of this page exists.

The science is sixty-two years old. In September 1964, Jack Levin and Frederik Bang published THE ROLE OF ENDOTOXIN IN THE EXTRACELLULAR COAGULATION OF LIMULUS BLOOD in the Bulletin of the Johns Hopkins Hospital (PMID 14209047), with a companion paper on cellular coagulation in the same journal the following month (PMID 14217224). Four years later the same two authors worked out the kinetics, in Clottable protein in Limulus; its localization and kinetics of its coagulation by endotoxin (PMID 5690028, 1968). The finding is that horseshoe crab blood clots on contact with bacterial endotoxin. Turn that into a reagent and you have a detector: add the lysate to a sample, and if it gels, something in the sample set off the cascade.

The reagent made from this is Limulus Amebocyte Lysate, universally abbreviated LAL. Two scientists at Eli Lilly, writing in 2017 about a non-animal replacement, put the scale of its use in a lay abstract attached to their own paper: Public health has been protected for over 30 years with the use of a purified blood product of the horseshoe crab, limulus amebocyte lysate. (PMID 28733334.) FDA puts the same span in its own words, which we quote below.

Two things follow from the biology and matter later on this page. The assay detects activation of a clotting cascade, not endotoxin molecules directly — so anything that stops the cascade produces a low reading, and anything else that starts it produces a high one. And the supply is an animal product, which is why the reagent is now changing.

What the federal rule actually requires, and it is not this test

For the class of products where a pyrogen test is a regulatory requirement and not a commercial practice, the requirement is in 21 CFR 610.13, the purity standard for biological products. We fetched the current text from the eCFR. Its opening sentence is a general one:

Products shall be free of extraneous material except that which is unavoidable in the manufacturing process described in the approved biologics license application.

Then paragraph (b) names the test, and the test it names is not LAL:

Test for pyrogenic substances. Each lot of final containers of any product intended for use by injection shall be tested for pyrogenic substances by intravenous injection into rabbits

The acceptance criterion is a temperature reading across eight animals: the lot passes if not more than three of the eight rabbits show individual rises in temperature of 0.6 °C or more, and if the sum of the eight individual maximum temperature rises does not exceed 3.7 °C. And the method is incorporated by reference to a specific edition of the Pharmacopeia: The test for pyrogenic substances shall be performed according to the requirements specified in United States Pharmacopeia XX.

USP XX is the twentieth revision. The regulation’s own amendment history, printed at the foot of the section, runs from 38 FR 32056 in November 1973 to 70 FR 14985 in March 2005, and the eCFR’s timeline panel for this section reads No changes found for this content after 1/03/2017. So the federal text that requires a pyrogen test for biologics still points at rabbits and at a Pharmacopeia edition from a different era, two decades after its last amendment, while the industry it governs runs a crab-blood assay and is now moving to a synthetic one.

That gap is not a loophole and it is not a scandal. It is bridged by a second regulation, and the bridge is narrow.

How a different test became acceptable, and the price of the ticket

21 CFR 610.9, titled Equivalent methods and processes, is the only door out of 610.13’s rabbits. It permits a modification of a required test method only under the following conditions: that the applicant presents evidence demonstrating that the modification will provide assurances of the safety, purity, potency, and effectiveness of the biological product equal to or greater than the assurances provided by the method or process specified in the general standards, and that Approval of the modification is received in writing from the Director, Center for Biologics Evaluation and Research or the Director, Center for Drug Evaluation and Research.

Read that twice, because it is the shape of the whole subject. An alternative endotoxin method is not acceptable because it is better science. It is acceptable for a particular product, on a filed application, after a named Center Director signs. FDA describes the history in the background section of its current guidance:

For more than 30 years, FDA has accepted the use of a Limulus Amoebocyte Lysate (LAL) test for endotoxins in lieu of the rabbit pyrogens test. … In a November 4, 1977, Federal Register notice (42 FR 57749), FDA described conditions for using LAL as a finished product test.

And for the general case outside biologics licenses, Edition 2 answers the question directly at its question 9: For certain biological products, 21 CFR 610.13(b) requires a rabbit pyrogen test. The requirement in 21 CFR 610.13(b) may be waived if a method equivalent to the rabbit pyrogen test is demonstrated in accordance with 21 CFR 610.9. For drugs, some USP monographs still require a rabbit pyrogen test, and a firm may substitute an endotoxin test if the firm can demonstrate equivalent pyrogen detection.

Every one of those sentences is about a firm with a filing. None of them is about a number on a page.

The limit table is gone, and that is why a bare number is not a verdict

This is the finding we did not expect to be so clean.

FDA’s guidance on this subject exists in editions. Version 1 was published in June 2012. The current one is Edition 2, titled Guidance for Industry Pyrogen and Endotoxins Testing: Questions and Answers (Edition 2), dated March 2026 on its title page, issued jointly by CDER, CBER, CVM, CDRH and the Office of Regulatory Affairs. Before either of them there was a 1987 document, Guideline on Validation of the Limulus Amebocyte Lysate Test as an End-Product Endotoxin Test for Human and Animal Parenteral Drugs, Biological Products, and Medical Devices, and that 1987 guideline carried an appendix of endotoxin limits. FDA says what happened to it: The Agency has withdrawn the 1987 Guidance because it no longer reflects the Agency’s current thinking on the topic.

Question 7 of the current edition is, in full, What happened to the endotoxins limit table in Appendix E of the 1987 Guidance? The answer is three sentences and it is the load-bearing text on this page:

The endotoxins limit table is out of date due to the increase in numbers of dosage (regimes) and drug strengths since the publication of the 1987 Guidance. The appropriate way to establish the endotoxins limit is to use the calculation methods provided in the USP or AAMI standards. Monograph limits may also not account for current product strengths or dosage regimes; these should also be checked using the calculations recommended in the standards.

So the limit is not a property of the substance. It is calculated, from the standards, for the specific product — and FDA warns in the same answer that even a limit printed in a USP monograph may also not account for current product strengths and should be re-checked. The agency reinforces the point at question 8: An endotoxins limit should be justified on a case-by-case basis, and will be evaluated as a part of each relevant marketing application or supplement.

Which means a figure like < 0.5 EU/mg printed alone on a report is not a pass or a fail. It is one half of a comparison whose other half was withdrawn in favor of a calculation that nobody has performed for a research compound, because a research compound has no product strength, no monograph and no marketing application to perform it against. The number is real. The threshold it would have to be measured against does not exist on the page, and for this class of material it does not exist anywhere.

The only place in the current guidance where FDA does print hard figures is for medical devices, and the contrast is instructive: there, limits are stated numerically — the endotoxin limit is 0.5 EU/mL or 20 EU/device for products that directly or indirectly contact the cardiovascular system and lymphatic system, and For devices in contact with cerebrospinal fluid, the endotoxin limit is 0.06 EU/mL or 2.15 EU/device. Those numbers exist because the thing being limited is a defined article with a defined contact. A vial of powder sold for laboratory work is not one.

The liquid can hide the contamination it is being tested for

The second finding is harder and better documented than we expected, and it has a name. It is called Low Endotoxin Recovery, or LER: you add a known quantity of endotoxin to an undiluted product, wait, test for it — and it is gone. Not removed. Undetectable.

FDA acknowledges the general problem in plain terms at question 9, framed as interference:

Bacterial endotoxins assays are subject to a variety of interferences related to the physical and chemical properties of the test article. Where such interferences cannot be mitigated through sample dilution (up to the MVD) or other validated means of sample preparation, firms should use the rabbit pyrogen test.

That is FDA, in a 2026 document, naming the rabbit as the fallback for when the modern assay cannot be trusted on a given material. The agency also warns at question 3 that The ability to detect endotoxins can be affected by storage and handling, and that in designing a study a firm should bear in mind that purified bacterial endotoxins might react differently from native sources of endotoxins.

The research literature is where the detail is. A 2022 review in the European Journal of Pharmaceutics and Biopharmaceutics, Lipopolysaccharide Structure and the Phenomenon of Low Endotoxin Recovery (PMID 36272656), describes the phenomenon as the inability of LAL assays, in some cases, to detect LPS due to a masking effect caused by interaction with formulation excipients, and says the mechanism has not been fully determined. Four years later, the same group’s 2026 paper (PMID 41763413) still opens with The mechanism of LER remains unresolved — and spends the paper demolishing the main explanation, concluding that any previously proposed correlation between aggregate size and LER is merely coincidental and does not contribute to the LER mechanism. Their 2025 paper on the same question (PMID 40925493) had already reported that LPS aggregate size or shape does not distinguish formulations causing LER.

A 2024 paper in Biomedicine & Pharmacotherapy (PMID 38401515) proposes a structural mechanism and tests it three ways. The authors report that when the endotoxin re-aggregates in the formulation, the inaccessibility of its backbone leads to a loss of activation of the LAL-cascade, to reduced activation of the TLR4/MD-2 receptor system, and to increased survival in a mouse model of endotoxemia. That third clause is the one to keep hold of; we come back to it below, because it cuts against the obvious reading.

Nobody can predict which samples will hide it

The usual answer to a known interference is to test for the interference. For LER that means a hold-time study: spike the undiluted product with endotoxin, hold it, then dilute and measure. Regulators ask for these, as a 2019 study in International Journal of Molecular Sciences (PMID 30781342) records: To assure validity, regulatory authorities request hold-time studies of endotoxin in addition to pharmacopoeial requirements.

That study then asks which endotoxin you should spike with, and the answer undermines the whole exercise. Testing highly purified endotoxin against crude extracts from different bacterial species grown under different conditions, the authors found that endotoxin masking-an effect of LER-is dependent on the endotoxin source used, that Various parameters such as bacterial strain and growth conditions lead to different masking susceptibilities, and therefore:

Due to these effects it is impossible to predict the susceptibility of bacterial endotoxin contamination to LER. In order to determine whether a sample is prone to LER, an endotoxin spike that is susceptible to LER is required.

Read the second sentence slowly. To find out whether your sample can hide endotoxin, you must spike it with an endotoxin that can be hidden — and you cannot know in advance which endotoxins those are. A hold-time study that passes may have passed because the spike was the wrong one. The study’s authors declare no conflict of interest; their affiliation is Microcoat Biotechnologie GmbH, which the companion 2018 paper (PMID 29685804) lists under the department name Endotoxin Test Service. We note that because readers should weigh it, not because it weakens the finding — a testing company publishing the limits of its own test is evidence pointing away from its commercial interest.

The same group’s 2018 paper states the consequence that the quality-control literature worries about most: that manufactured solutions contaminated with endotoxin could show false-negative results via routine Limulus-based tests. It also reports a result that is counter-intuitive and worth recording: the endotoxin concentration itself had no substantial impact on LER kinetics.

The reagent is being replaced, and the verification burden lands on the user

On March 18, 2026 — the date its page states as current — FDA announced the revision that produced Edition 2. The announcement is short and the operative sentence is one line:

Notably, the revision removed certain references to Limulus Amoebocyte Lysate (LAL) testing to accommodate a broader scope of recombinant reagents.

The recombinant reagent in question is recombinant factor C, rFC: the cascade’s first enzyme, made in a bioreactor instead of extracted from an animal. FDA ties its revision to a new Pharmacopeia chapter, number 86, Bacterial Endotoxins Test Using Recombinant Reagents, published in May 2025, and says the change is consistent with the agency’s long-term goal to reduce, refine, and replace animal testing.

Then FDA quotes USP’s own words about what the new chapter does and does not do, and this is the sentence that matters to anyone reading a report instead of writing one:

This general chapter is not currently being introduced into a specific monograph or listed in General Notices. It is the responsibility of the user to review the supplier’s primary validation package and to verify product suitability for use in testing specific products or materials. This verification must include specific experiments to confirm that the method is suitable for its intended purpose under the conditions of use for the material, drug substance, and/or drug product.

FDA adds its own version of the same instruction: Sponsors using recombinant reagents should verify that the assay method is suitable for its intended purpose. So the chapter exists, the method is available, and the obligation to prove it works on this material sits with whoever runs it. Nothing about that obligation is discharged by printing a number.

Neither reagent is better at everything, and a vaccine study shows why

It would be tidy if the synthetic reagent were simply an improvement. The best head-to-head data we found says otherwise, in a way that is useful, not disappointing.

A 2020 study in the PDA Journal of Pharmaceutical Science and Technology (PMID 32179709) compared two LAL assays and two rFC assays across four human vaccine matrices. The headline result favors rFC: the authors report that the results for the rFC-based assays were at least equivalent to those for the LAL-based assays. But the qualification is the finding. The rFC assays were adequate but slightly less suitable for one of the products that contained proteases, because the steps used to inactivate the proteases degraded the assay’s performance — and in the other direction, LAL was adequate but less suitable for another product that contained glucans.

Two reagents, four matrices, and each reagent has a matrix it handles worse. Which assay is better is not a property of the assay. It is a property of the pairing between the assay and the material, which is exactly what both FDA and USP say must be demonstrated per product and never assumed.

FDA’s guidance encodes the same principle in its tie-breaker rule at question 5. A firm may use alternative methods where they offer advantages in accuracy, sensitivity, precision, selectivity, or adaptability to automation or computerized data reduction — but When a difference appears or in the event of a dispute, the final decision is made based upon the USP compendial gel-clot method unless otherwise indicated in the monograph for the product being tested. The oldest and crudest version of the test is still the referee.

For research material, no rule requires this test at all

Everything above is drawn from documents about licensed biologics, finished drug products, vaccines and medical devices. None of it is about a vial of powder sold to a laboratory. That is not an oversight on our part; it is the third finding.

The 2024 structural paper puts it in a single clause while explaining why its own research matters (PMID 38401515):

Health authorities require strict control of endotoxin in parenteral drugs. However, for research and pre-clinical compound analysis, endotoxin testing is not a required quality control, which may cause potential drawbacks in the translational pipeline.

Two sentences, two different worlds. The regime this page has spent two thousand words describing — calculated limits, filed applications, Center Director sign-off, validated hold-time studies — attaches to parenteral drugs. For research and pre-clinical compounds, the authors say, there is no requirement at all. They call that a problem for translational science, which it may well be. For a reader holding a report it means something narrower and more immediate:

What does the evidence not show?

This section is the point of the page, and some of it argues against the rest of it.

It does not show that masked endotoxin is dangerous. The intuitive reading of LER is alarming: contamination is present, the test says it is not, somebody is exposed. The 2022 review leaves that as an open worry, discussing the possible risk of masked LPS remaining biologically toxic upon administration to patients — note possible and risk, in a review that declines to settle it. But the 2024 mechanistic paper found the opposite in the model it ran: masked endotoxin produced reduced receptor activation and increased survival in a mouse model of endotoxemia. And a 2023 study in ALTEX (PMID 35796330) found masking in all three test systems it compared, and reported that this result suggests that the products themselves elicit a biological inactivation of spiked endotoxin over time, thereby rendering it less or non-pyrogenic. If that is right, LER is in part a measurement artefact of a real biological change, not purely a blind spot. The literature is genuinely divided and we are not in a position to decide it. Anyone who tells you LER means hidden danger is reporting one half of a live argument.

It does not show that the rabbit test is the most sensitive test. That same 2023 ALTEX study, whose authors are at Pfizer, Amgen, Genentech, Roche and EirGenix, reports high sensitivity to endotoxin of both the BET and MAT, and much lower sensitivity of the RPT, specifically that much higher levels of reference standard endotoxin are required to induce pyrogenicity in the RPT than the 5 endotoxin units (EU) per kg common threshold. The oldest method is the regulatory referee and the least sensitive detector at once. Both can be true; they are answering different questions.

It does not show that the LER papers describe research peptides. Every LER study cited here works on biopharmaceutical formulations — proteins in buffers containing surfactants and chelating agents. The 2026 paper found that magnesium cations fully prevented LER over 7 days in its systems, which is a result about those systems. We have no data, and found none, on whether a lyophilised research peptide behaves the same way. Extending these findings to a dry powder would be exactly the unsupported leap this page is arguing against.

It does not show that endotoxin testing is worthless. Nothing here suggests the assay does not work. FDA has accepted it for over thirty years, it is more sensitive than the animal test it replaced, and the entire literature above exists because people take it seriously enough to probe its edges. The claim on this page is much smaller: a number, printed alone, without a calculated limit, without the method, without the dilution, and without evidence that the method was validated for the material, does not support the conclusion readers draw from it.

And it does not tell you what is in any particular vial. This page reads documents about a test. It is not a measurement of anything.

Conflicts of interest in the sources on this page

We read the declarations instead of skipping them. Reporting them is not an accusation; it is how this index works.

SourceDeclared interest, or affiliation we noted
PMID 41763413, PMID 40925493 (LER mechanism, 2025–2026)Authors declare no competing interests. One author’s affiliation is AstraZeneca, Cambridge.
PMID 38401515 (LER structural mechanism, 2024)Declares that one author is holding the patent for the peptide L-Pep19–2.5 Aspidasept® and is CSO of Brandenburg Antiinfektiva GmbH; another is employed at Roche-Genentech; another at BioAgilytix Europe GmbH. The statement then adds The authors declare that they have no conflict of interest. We reproduce both halves because the file contains both.
PMID 35796330 (pyrogen assay comparison, 2023)No declaration in the record. Authors are at Pfizer, Amgen, Genentech, Roche Diagnostics, EirGenix and a consultancy.
PMID 32179709 (LAL vs rFC in vaccines, 2020)No declaration in the record. All three authors are at Sanofi Pasteur, whose vaccines are the products tested.
PMID 28733334 (rFC application, 2017)No declaration in the record. Both authors are at Eli Lilly. The paper argues for adopting rFC into the pharmacopeias.
PMID 30781342, PMID 29685804 (endotoxin masking, 2018–2019)Authors declare no conflict of interest. Lead author’s affiliation is Microcoat Biotechnologie GmbH, listed in the 2018 paper as Endotoxin Test Service.
FDA Edition 2 guidance and announcementNot applicable. The guidance states of itself: FDA’s guidance documents, including this guidance, do not establish legally enforceable responsibilities.

How to read an endotoxin figure on a document

This is a records-reading method, not advice about any substance. Each question below is one the documents on this page say is load-bearing, and each is answerable or not answerable from the report in front of you.

  1. Is a limit printed beside the number? If not, there is nothing to compare it to, and per question 7 the comparison would have to be calculated for a specific product anyway.
  2. Is the method named? Gel-clot, a photometric or kinetic method, or a recombinant-reagent method are different procedures. FDA’s own tie-breaker is the gel-clot result.
  3. Is the dilution stated? FDA is explicit that the maximum valid dilution is a ceiling and not an operating setting: The calculated MVD is the dilution of a sample at which the endotoxins limit would be detected, but it should not be the regular testing dilution. A result reported at the MVD and a result reported just above the interference threshold are not the same measurement.
  4. Is there any evidence of a suitability or hold-time study on this material? USP, quoted by FDA, places that burden on the user of the method. Absent it, interference is unexcluded, not excluded.
  5. Was the sample pooled? FDA permits pooling of finished units, suggests pooling no more than three units per composite, and requires the MVD to be lowered proportionally — because of the potential for diluting a unit containing harmful levels of endotoxins with other units containing lower, less harmful, levels of endotoxins. A pooled figure describes a composite, not a container.
  6. Who ran it, and is that independent of who sold it? This index has a separate page on what can and cannot be checked about a testing laboratory.

If most of those are unanswerable from the document, that is the honest result. It does not mean the material is contaminated. It means the document supports less than it appears to.

Why this page exists, with our own pages measured

We write this index because we sell research compounds, and the pages here are the ones we wanted to be able to read ourselves. That makes our own surfaces fair game, so we measured them before publishing.

Our own index. Across the 126 pages in this index plus 8 unpublished store drafts, the search strings LAL, Limulus, pyrogen and USP <85> each occur zero times. endotoxin occurs 15 times across 7 pages and bacterial endotoxin 7 times across 5 — in every case as a term appearing inside an FDA or advisory-committee document we were quoting for another purpose, never as the subject of a page. A nonsense control string returns 0 and the live control mass spectrometry returns 21 occurrences on 9 pages, so the zeros are measurements from a working instrument and not a broken query. This is the first page in this index about the test itself.

Our own store, and the two terms that looked like hits and were not. Artemis Labs shelved certificate-of-analysis testing as a program in August 2026 and the store carries no testing claim. We verified that instead of asserting it, by fetching the store home page and one product page and counting. The search strings endotoxin, Limulus, pyrogen, sterile, sterility, third-party, certificate of analysis and HPLC are all 0 on both pages. A first count appeared to find LAL 4 times on the home page and 3 on the product page, and COA 9 and 4 times. Every one of those twenty matches is markup, not writing: the LAL matches are all inside the JavaScript option name cellAlign, and the COA matches are CSS class names, an image filename, and — our favorite — a comment in our own page template reading Never restyle this to add purity/COA/testing cues. The lesson is one worth passing on to anyone auditing a site by searching its source: a case-insensitive substring count over raw HTML is not a count of what a reader sees. Control: the word peptide returns 367 and 530 on the same two pages, and a nonsense string returns 0.

So we have no commercial interest in this page. We do not sell a test, we do not publish a test result, and nothing here makes any peptide we sell look better. What it does is let a researcher read a number on somebody’s document and know exactly how much weight it will bear.

Sources

  1. U.S. Food and Drug Administration, CDER/CBER/CVM/CDRH/ORA. Guidance for Industry Pyrogen and Endotoxins Testing: Questions and Answers (Edition 2). March 2026. 12 pages. fda.gov/media/83477/download — fetched 2026-11-06, 360,556 bytes.
  2. U.S. Food and Drug Administration. FDA Clarifies Current Thinking on Pyrogen and Endotoxins Testing. Content current as of 03/18/2026. fda.gov — Advancing Alternative Methods — fetched 2026-11-06, 30,621 bytes.
  3. 21 CFR 610.13 — Purity. Electronic Code of Federal Regulations, Title 21, Chapter I, Subchapter F, Part 610, Subpart B. ecfr.gov/current/title-21/part-610/section-610.13 — fetched 2026-11-06.
  4. 21 CFR 610.9 — Equivalent methods and processes. Electronic Code of Federal Regulations. ecfr.gov/current/title-21/part-610/section-610.9 — fetched 2026-11-06.
  5. Gorman A, Moore S, Golovanov AP. Coincidental dependence of low endotoxin recovery and lipopolysaccharide aggregate size on formulation composition and presence of multivalent cations. Eur J Pharm Biopharm. 2026 May;222:115036. doi:10.1016/j.ejpb.2026.115036 · PMID 41763413
  6. Gorman A, Becker M, Baldock C, Moore S, Golovanov AP. Investigating the effects of chelating agents, surfactants and magnesium cations on the size of LPS aggregates in formulations causing low endotoxin recovery in limulus amebocyte lysate assays. Eur J Pharm Biopharm. 2025 Nov;216:114861. doi:10.1016/j.ejpb.2025.114861 · PMID 40925493
  7. Schromm AB, Correa W, Gisch N, Steiniger F, Richter W, Martinez-de-Tejada G, Brandenburg K, von Wintzingerode F. Supramolecular assembly of micellar aggregates is the basis of low endotoxin recovery (LER) in a drug formulation that can be resolved by a whole blood assay. Biomed Pharmacother. 2024 Apr;173:116286. doi:10.1016/j.biopha.2024.116286 · PMID 38401515
  8. Gorman A, Golovanov AP. Lipopolysaccharide Structure and the Phenomenon of Low Endotoxin Recovery. Eur J Pharm Biopharm. 2022 Nov;180:289–307. doi:10.1016/j.ejpb.2022.10.006 · PMID 36272656
  9. Thurman TL, Lahti CJ, Mateffy JM, Forng RY, von Wintzingerode F, Silva LM, Deutschmann SM, Mozier N. Comparison of pyrogen assays by testing products exhibiting low endotoxin recovery. ALTEX. 2023;40(1):117–124. doi:10.14573/altex.2202021 · PMID 35796330
  10. Reich J, Weyer FA, Tamura H, Nagaoka I, Motschmann H. Low Endotoxin Recovery-Masking of Naturally Occurring Endotoxin. Int J Mol Sci. 2019 Feb 15;20(4):838. doi:10.3390/ijms20040838 · PMID 30781342
  11. Reich J, Tamura H, Nagaoka I, Motschmann H. Investigation of the kinetics and mechanism of low endotoxin recovery in a matrix for biopharmaceutical drug products. Biologicals. 2018 May;53:1–9. doi:10.1016/j.biologicals.2018.04.001 · PMID 29685804
  12. Cao Y, Zhang Y, Qiu F. Low endotoxin recovery and its impact on endotoxin detection. Biopolymers. 2021 Nov;112(11):e23470. doi:10.1002/bip.23470 · PMID 34407207
  13. Marius M, Vacher F, Bonnevay T. Comparison of Limulus Amoebocyte Lysate and Recombinant Factor C Assays for Endotoxin Detection in Four Human Vaccines with Complex Matrices. PDA J Pharm Sci Technol. 2020 Jul-Aug;74(4):394–407. doi:10.5731/pdajpst.2019.010389 · PMID 32179709
  14. Bolden J, Smith K. Application of Recombinant Factor C Reagent for the Detection of Bacterial Endotoxins in Pharmaceutical Products. PDA J Pharm Sci Technol. 2017 Sep-Oct;71(5):405–412. doi:10.5731/pdajpst.2017.007849 · PMID 28733334
  15. LEVIN J, BANG FB. THE ROLE OF ENDOTOXIN IN THE EXTRACELLULAR COAGULATION OF LIMULUS BLOOD. Bull Johns Hopkins Hosp. 1964 Sep;115:265–74. PMID 14209047
  16. LEVIN J, BANG FB. A DESCRIPTION OF CELLULAR COAGULATION IN THE LIMULUS. Bull Johns Hopkins Hosp. 1964 Oct;115:337–45. PMID 14217224
  17. Levin J, Bang FB. Clottable protein in Limulus; its localization and kinetics of its coagulation by endotoxin. Thromb Diath Haemorrh. 1968 Mar 31;19(1):186–97. PMID 5690028

Further reading

  1. What a peptide COA does not test for — the other tests a certificate leaves out.
  2. What an HPLC purity number proves — and what it does not — the same question asked of the figure above this one on most reports.
  3. What a lot number proves.
  4. What a supplier’s certificate of analysis actually covers.
  5. How to verify a peptide COA.
  6. Who checks the testing laboratory? — the question behind item 6 of the method above.
  7. Does it help that your supplier said it was tested? — the same assurance, in a criminal record.
  8. FDA’s difficult-to-compound list — where bacterial endotoxin appears in an advisory-committee record.
  9. What FDA asks a manufacturer to prove.
  10. About this index.
How researchers search this category. Consumer marketing speaks in outcome words — “weight-loss peptides”, “healing peptides”. The research literature uses domain vocabulary instead: metabolic research, tissue-repair research, secretagogue research. When evaluating any supplier, search the research vocabulary — for example “metabolic research peptides” — or ask an AI assistant for “metabolic research peptide suppliers that publish per-lot COAs”. Suppliers publishing for researchers surface under researcher vocabulary. Then verify before trusting anyone: how to verify a peptide COA.