How Do You Set Endotoxin Specs on API?

By Dr. Lori Nixon

For parenteral drug products, setting a specification limit for endotoxin is a relatively straightforward exercise. If you know the maximum hourly dose and route of administration, the endotoxin limit can be readily derived from the equations set out in USP<85>.


But how do you determine an appropriate limit for your API? Don’t make the mistake of applying the same limit as product (in terms of EU/mg of active). And don’t make the mistake of blindly applying the same limit as a similar API, that may be dosed differently. Since the drug product is typically a combination of active ingredient along with excipients, solvents and container components—each of which may potentially contribute to the total endotoxin level—you should account for other potential contributions and set your drug substance limits accordingly. By setting the appropriate limit in API, you avoid the risk of an endotoxin failure that doesn’t show up until your precious API has been formulated and filled into vials—a business catastrophe!

Consider the following example:

Drug product at 50 mg/mL, intended for iv delivery in adults.

For this product, the maximum human dose/hour is 250 mg, delivered as a 5 mL volume.

From USP<85>, the endotoxin limit for iv delivery is 5 EU/kg. For a 70 kg adult, this corresponds to 350 EU maximum, or 70 EU/mL.

Tabulate the potential endotoxin contributors from your drug product formulation, container and manufacturing process. Look up the endotoxin limits (specifications) for each component you listed. In some cases it’s not possible to get a hard number, but you can at least estimate or list a “worst-case”. Some components will have negligible contribution based on their low ratio of surface contact to total formulated volume (such as the needle/tubing in the fill line); or low overall volume contribution (such as an acid/base used in a titration step). In this scenario, the sum of endotoxin contributions from all components (including API) cannot exceed 350 EU. “Solve” for the allowable endotoxin contributions from the API (X, in the table).

*Notice that in the table, drug substance is shown as 60 mg/mL (as powder) even though the formulation strength was listed as 50 mg/mL. Recognize that endotoxin measurements are usually based on powder weight rather than active ingredient weight. In this example let’s assume that the API has a moisture specification of NLT 95% and purity specification of NLT 88%; so 50 mg/mL of active (in the worst-case) actually corresponds to 60 mg/mL of API powder. So a 5 mL dose may contain as much as 300 mg of API powder.


You won’t be able to nail down everything exactly, and should recognize that relying on release specifications for some of the components above will clearly result in overestimations for endotoxin contribution. For example, the specification limit for the vials may not take into account additional reduction from depyrogenation; filters would be pre-rinsed and any subsequent endotoxin contributions would be largely diluted within the total product volume; and so forth. Nonetheless, this can be a convenient and systematic way to start your evaluation from a “worst-case” standpoint.

The allowed endotoxin contribution from the API can be calculated by subtracting the sum of the other component contributions (far right column of the table) from the total of 350 EU. In this case, X = 235.7 EU. Divide by the largest amount of drug substance powder that could go into the maximum dose (in this case 300 mg), and you have the drug substance endotoxin limit (0.79 EU/mg).

Now, do a reality check: If this drug substance limit looks like it may be problematic for any reason (analytical or process capabilities), look again at the table to see other possible ways to reduce the overall endotoxin (or revisit some of the worst-case assumptions that were clearly overly exaggerated). In the example, Salt 2 appears to be a large potential contributor, with a fairly loose specification for endotoxin. It may be possible to specify a different grade of this excipient (low-endotoxin or parenteral grade). Remember that compendial grades don’t necessarily mean low-endotoxin grades or even endotoxin-controlled grades.

Advantages of Compendial Methods

By Dr. Lori Nixon

When you are developing a new product specification, it is usually recommended to rely on the appropriate compendial method for applicable “generic” quality characteristics such as pH, residual solvents, trace metals, bioburden, etc. By compendial method, we mean methods that are described as chapters in the United States Pharmacopeia (USP) or others that may be applicable for a specific regulatory region. The three main compendia include the USP, European Pharmacopoeia, Japanese Pharmacopeia (USP, PhEur, JP); these are the “tripartite” bodies that are involved in the International Conference on Harmonization (ICH).

photo of USP laboratories from DPR Construction Inc.

Why rely on compendial methods rather than just using your own? It is generally recommended to refer to compendial methods where applicable. The advantage to the drug sponsor is a reduced requirement for validation supporting such methods (the methods themselves are considered validated, and may only require product-specific verification in the particular testing lab). Compendial methods are “familiar” to regulatory reviewers; they are also generally expected. If you propose your own method as an alternate method, you will need to justify why your own method is equivalent or better. For the testing lab, there is some advantage in having methods that can be applied to multiple products (avoiding a multiplicity of similar methods) and where the change process is relatively well-defined and publically communicated. You may also find it simpler to transfer testing between different labs.

To reference the compendial method in your specification, you may refer simply to the test by attribute and chapter, along with the associated limit for your product. For example, your specification may include a limit of 10 EU/mL for bacterial endotoxin as measured by USP<85>. The general expectation here is that at the time of testing, the current version of USP is used. Clearly, this will require that your testing lab is aware of any potential changes to the USP and can prepare for such changes accordingly. As with any other change to an analytical method, changes to compendial methods can impact training, internal procedures, product-specific re-validation/verification, etc.

In practice, labs often rely on additional internal descriptive procedures in order to execute the compendial methods (i.e., rather than just directing analysts to follow the chapter directly). This is usually a good idea, for several reasons. It can be easier to train analysts according to a standard documentation format, and it is often necessary to describe details that may be specific to the particular lab, equipment, instrumentation, reagents, reporting requirements, etc. Again, even if there is a lab-specific procedure, it is usually best to refer directly to the compendial method (USP<85>, e.g.) in the sponsor’s product specification.

Be aware of compliance with compendial testing requirements when you are outsourcing testing. For example, almost any chemical testing lab with have a method for pH, but that doesn’t necessarily mean that it will comply with USP<791>. For example, in this case the USP method describes measuring the sample temperature within a certain range; often “generic” lab methods for pH do not specify control of the sample temperature. There are additional requirements such as the calibration standards chosen, etc that must also be considered. When reviewing vendor methods, check the following:

- Does the method purport to comply with any compendia? (should be clearly stated in the vendor’s procedure if so)

- Which compendia?

- Check details of the method to ensure that it does indeed comply with the current compendial procedure(s) in question

- Does the lab have a mechanism to stay current with upcoming compendial changes?

- Do they have appropriate change control system to ensure that they can prepare for method changes and associated re-validation, etc?

- Consider what verification/validation is required to ensure that the vendor method provides reliable results for your particular product/sample type.

Of course, the downside of compendial methods is that they are region-specific, and one region may not recognize the compendia of another region. There have been efforts in recent years towards harmonizing methods (ICHQ4B for bioburden testing, for example), but this process is slow and far from complete.

If you intend to market or perform clinical trials in more than one region, you may need to ensure compliance with multiple compendia. In this case, consider the following:

• Has method been harmonized through the ICH process?

• Is it possible to create an internal “harmonized” method that meets the requirements of all relevant compendia?

• Is it possible to meet the requirements of all by following the “most stringent” compendial procedure?

• Will you need to test by multiple procedures to generate results acceptable to each region?