Endotoxin and Sterility Tests: Why Passing One Is Not Enough

Two results, two different claims

A passing bacterial endotoxins test does not establish sterility. A passing sterility test does not establish endotoxin conformity or the absence of all pyrogens. For a QC reviewer reading a sterile injectable batch record, the results supply different evidence about different attributes. Neither result alone is sufficient for a release decision. Each must be interpreted against the applicable specification, a justified sampling plan, a suitable method and the manufacturing record. 21 CFR 211.165; 21 CFR 211.167.

The bacterial endotoxins test, or BET, detects or quantifies endotoxin from Gram-negative bacteria against a product-specific limit. The sterility test looks for viable microorganisms able to grow under its test conditions. Endotoxin can remain after microorganisms have been killed or removed, so a result about growth cannot answer the endotoxin question. Equally, an acceptable endotoxin result supplies no demonstration that viable microorganisms are absent. This distinction remains true when a manufacturer changes its endotoxin reagent or uses an authorized release arrangement. USP endotoxin chapter overview; FDA aseptic-processing guidance.

The matrix is an interpretation aid for reading the evidence, rather than a sampling plan or a laboratory procedure. Its distinction between the tested sample and the batch matters: a sample result can support a batch conclusion when the surrounding method, sampling and process evidence justify that inference. It cannot expand the attribute that the assay measures. FDA endotoxin Q&A, Edition 2; FDA aseptic-processing guidance.

Why sterilization does not establish endotoxin control

Sterilization and depyrogenation address different objectives. FDA’s 2004 aseptic-processing guidance explains that sterilizing-grade filters and moist-heat sterilization have not been shown effective at removing endotoxin. A record showing that a sterilization process met its parameters therefore cannot, by itself, demonstrate acceptable endotoxin levels. The practical review question is which evidence establishes each objective, rather than whether one processing step sounds sufficiently severe. FDA aseptic-processing guidance.

The distinction does not mean that every control must be physically separate. FDA notes that dry-heat treatment of glass containers generally accomplishes both sterilization and depyrogenation. That is a shared process serving two objectives, supported by appropriate validation. A reviewer should identify the evidence for the endotoxin objective instead of assuming that the word sterile covers it. Conversely, a satisfactory endotoxin-removal record does not establish control of subsequent microbial contamination. FDA aseptic-processing guidance.

FDA’s endotoxin-control discussion connects prevention to drug components, containers, closures, equipment and storage time. This makes a final BET result one element of a control strategy rather than a substitute for upstream controls. A clean result at the end of the batch record does not explain away an unresolved manufacturing concern; the concern still needs an evidence-based assessment. The same logic applies to a no-growth sterility result when the manufacturing process has departed from its validated conditions. FDA aseptic-processing guidance.

For the related utility-control question, see pharmaceutical water results and microbial evidence. Water-system evidence belongs in the manufacturing context; it does not replace either finished-product result.

Sampling limits: what the result can miss

FDA’s 2004 guidance gives two published illustrations of the limited detection power of sterility sampling. They should be read as illustrations from that guidance, not as a current universal sample-size rule or a prediction for a particular manufacturing lot. In particular, the 20-unit example does not establish that every 10,000-unit batch should be sampled in exactly that way. FDA aseptic-processing guidance.

The direction of the probability matters. FDA describes the chance of obtaining a passing test under a specified contamination scenario. It does not estimate the chance that an actual batch is contaminated after a pass. Reversing those statements would require additional information about the manufacturing process and contamination likelihood. The useful conclusion for a reviewer is narrower: no growth in sampled containers cannot certify the condition of every container. FDA aseptic-processing guidance.

Representative sampling is therefore part of the interpretation. FDA recommends sterility samples from the beginning, middle and end of the aseptic operation and in conjunction with processing interventions or excursions. A reviewer needs to connect the sample record to the actual filling history. A total sample count alone does not show that the sampling covered the relevant operating conditions. FDA aseptic-processing guidance.

BET has different sampling limitations. In Question 1 of its March 2026 Q&A, FDA describes sampling as a plan informed by process understanding. Relevant factors include potential contamination in materials and finished product, process consistency, in-process hold times, endotoxin-removal steps and finished-product specifications. FDA recommends beginning with maximum coverage and adjusting as confidence in prevention grows, with appropriate regulatory updates. That is not permission to reduce sampling solely because recent results passed. FDA endotoxin Q&A, Edition 2.

Question 3 addresses sample storage and handling, including mixing. Laboratory data should support the stability of assayable endotoxin during the handling interval. Question 4 addresses pooling: where pooling is appropriate, the adjusted maximum valid dilution, or MVD, must account for the risk of diluting a contaminated unit with cleaner units. Some products are unsuitable for pooling. These checks determine whether the test result can support the intended interpretation; a pass from an unsuitable sample arrangement is not an answer to a sampling concern. FDA endotoxin Q&A, Edition 2.

What the rules require, and where

In the United States, 21 CFR 211.167(a) requires appropriate laboratory testing for each batch purporting to be sterile and/or pyrogen-free, using written procedures that are followed. It establishes the conformance obligation for those attributes. It does not declare the sterility and endotoxin methods interchangeable, nor does it prescribe a single universal assay for every product and every pyrogen risk. 21 CFR 211.167.

The broader release framework appears in 21 CFR 211.165. Paragraph (a) requires appropriate laboratory determination of conformance to final specifications before release, subject to its stated radiopharmaceutical exception. Paragraph (c) requires written sampling and testing procedures identifying the sampling method and number of units per batch. Paragraph (d) addresses adequate acceptance criteria and appropriate statistical quality-control criteria. Paragraph (e) requires established and documented accuracy, sensitivity, specificity and reproducibility of methods. Paragraph (f) requires rejection of products failing established standards, specifications or other relevant quality-control criteria; it also addresses reprocessing and subsequent conformance. 21 CFR 211.165.

Those are binding US CGMP regulations. FDA’s 2004 aseptic-processing guidance and March 2026 endotoxin Q&A provide nonbinding recommendations explaining the agency’s thinking. USP chapters are compendial standards, with applicability determined in the relevant product and regulatory context. USP’s public harmonization notice for Chapter <85> identifies an adopted standard; that notice is not a substitute for the applicable chapter and monograph when establishing laboratory procedures. FDA endotoxin Q&A, Edition 2; USP harmonization notice.

EU GMP Annex 1 is a separate EU framework. Section 10.5 places the finished-product sterility test within a series of controls and says it cannot compensate for unmet design, procedural or validation parameters. Section 10.2 addresses microbial, particulate and endotoxin/pyrogen limits where monitoring or the contamination control strategy indicates the need. Section 10.6 calls for representative sterility samples, including higher-risk portions of the batch. EU GMP Annex 1.

Section 10.3 generally calls for bioburden testing on each batch of aseptically filled and terminally sterilized products, with results considered at final review. It also specifies scheduled monitoring where overkill terminal-sterilization parameters apply. That qualification matters when reading the paragraph as a whole. These EU provisions should be applied within their own framework, rather than presented as additional text inside the US regulation. EU GMP Annex 1.

The 2026 method update stays within the endotoxin question

FDA’s guidance record identifies the current Q&A as the March 2026 final Level 2 revision, Edition 2. The agency’s accompanying explanation says that it removed certain references to Limulus Amoebocyte Lysate, or LAL, testing to accommodate a broader scope of recombinant reagents. FDA also connects the revision to USP Chapter <86> and emphasizes suitability of the assay for its intended purpose. This updates the method discussion; it does not broaden an endotoxin result into sterility evidence. FDA guidance record; FDA revision explanation.

USP’s injectables overview lists both Chapter <85>, Bacterial Endotoxins Test, and Chapter <86>, Bacterial Endotoxins Test Using Recombinant Reagents. The existence of both chapters does not mean that every manufacturer can exchange methods without addressing product suitability and regulatory commitments. For the reader interpreting a batch packet, the question is which method was applicable and supported for that product, rather than which reagent category sounds newer. USP endotoxin chapter overview; FDA revision explanation.

Questions 5 and 6 of the FDA Q&A discuss alternative assays and transitions between BET methods. Validation and equivalence are central, including comparison of detection capability and inhibition or enhancement. The guidance also discusses the regulatory submission process. A method-change record therefore needs its own supporting evidence; an acceptable result under a new method does not itself establish that the transition was properly justified. FDA endotoxin Q&A, Edition 2.

The broader pyrogen question also deserves care. In Question 9, FDA explains that rabbit pyrogen testing remains relevant for certain biological products and some USP monographs, while discussing waivers or alternatives where equivalence is demonstrated. It also addresses non-endotoxin pyrogen risks and circumstances where BET interference cannot be resolved by permitted dilution or other validated preparation. These are choices within the pyrogen-testing assessment. Neither a BET method change nor a different pyrogen method supplies the missing sterility evidence. FDA endotoxin Q&A, Edition 2.

Exceptions change the evidence route or timing

A statement that both completed laboratory results must always precede release would be too broad. Under 21 CFR 211.165(a), specified batches of short-lived radiopharmaceuticals may be released before sterility and/or pyrogen testing is complete, provided the testing is completed as soon as possible. This is an explicit sequencing provision, not a conclusion that one passing test makes the other unnecessary. It should not be generalized to products outside that provision. 21 CFR 211.165.

In the EU framework, Annex 1 section 10.7 addresses short shelf life that prevents a sterility result before release and calls for documented risk mitigation. Section 10.4 addresses products already authorized for parametric release, including a supporting per-batch pre-sterilization bioburden programme and endotoxin/pyrogen monitoring where appropriate. It is not a stand-alone authorization to omit testing or to infer endotoxin conformity from sterilization parameters. EU GMP Annex 1.

For a reviewer, the distinction is between evidence supplied through a justified, applicable release arrangement and evidence that is simply missing. Neither speed of manufacture nor the presence of another passing test establishes an exception. The batch packet should identify the basis for its release route and the evidence required under that route. A general comparison article cannot determine whether a particular product qualifies.

Reading a batch record with both results

The following review questions translate the source documents into a way of reading a batch packet. They do not set product acceptance criteria or decide disposition. Their purpose is to identify which claim each record supports and where the quality unit needs the applicable procedure, validation record or investigation before drawing a conclusion.

Connect each result to its intended claim

Identify the product specification and the result being used to support it. Keep sterility evidence, endotoxin evidence and any broader pyrogen assessment distinct. Confirm that the record identifies the applicable method rather than relying on a generic statement that microbiology passed. Where an alternative method is involved, locate the suitability and transition evidence supporting its use for that product. 21 CFR 211.167; FDA endotoxin Q&A, Edition 2.

Connect the samples to the manufacturing history

Compare the sample identities and collection record with the written sampling plan and the filling history. For sterility testing, ask whether the record reflects relevant interventions and operating periods. For BET, examine the documented sampling rationale, handling interval and any pooling arrangement. This is a traceability question: can the reviewer follow the chain from the units selected to the result reported and the batch inference being proposed? 21 CFR 211.165; FDA aseptic-processing guidance; FDA endotoxin Q&A, Edition 2.

Keep process concerns and failures visible

Read the results alongside the manufacturing and contamination-control evidence. FDA treats a positive sterility test as a serious CGMP issue requiring thorough investigation. Its guidance limits invalidation to conclusive, documented evidence that the contamination arose during testing; a later negative result is not, by itself, that evidence. A passing BET cannot resolve a sterility positive, just as no-growth sterility samples cannot resolve an endotoxin failure. FDA aseptic-processing guidance.

If the packet proposes omitting a result, identify the applicable authorization or timing provision rather than accepting the other test as justification. If it proposes dismissing a process deviation because both results passed, ask what evidence addresses the deviation itself. This preserves the value of each analytical result without assigning it a claim that the method cannot support. 21 CFR 211.165; FDA aseptic-processing guidance.

For supplier and technical-agreement questions beyond this result-interpretation task, see the separate guide to sterile fill-finish CDMO evidence and contamination-control responsibilities. That is a different decision from interpreting the two test packets for a batch.

Sources

• Pyrogen and Endotoxins Testing: Questions and Answers (Edition 2), Guidance for Industry, March 2026

• Pyrogen and Endotoxins Testing: Questions and Answers (guidance landing page)

• FDA Clarifies Current Thinking on Pyrogen and Endotoxins Testing

• Sterile Drug Products Produced by Aseptic Processing — Current Good Manufacturing Practice (September 2004)

• 21 CFR 211.165 — Testing and release for distribution

• 21 CFR 211.167 — Special testing requirements

• EudraLex Volume 4, Annex 1: Manufacture of Sterile Medicinal Products (2022)

• <85> Bacterial Endotoxins — Notice of Adoption of Harmonized Standard

• USP Reference Standards — Injectables (Endotoxins Testing Solutions)