A Step-by-Step Approach to Effective Endotoxin Testing

Regardless of where the sample comes from, it is sampled in a defined manner under defined conditions and, if it has been to the lab before, has a validation/verification package and an official routine test protocol derived from it describing how it is to be tested based on the method developed.  This is true for not only drug samples but also for in-process and waters, components or raw materials etc.  

The idea of sample receipt views the sample as having a specific chain of custody where its history from origin to receipt is referenced.  It also comes with the knowledge of how it should be properly stored (refrigerated, frozen, room temperature, etc.).  A laboratory log is used to capture the time and initials of the receiver who stores the sample appropriately.

Storage is an important aspect of sample receipt as an improperly stored sample may invalidate any subsequent test. The storage will have been demonstrated to preserve the potential for contaminant detection. For example, if freezing diminishes the recovery of sample contaminant content, then another manner of storage or limitations on the storage time must be made.

This will vary from test type to test type and depends on the software but will include a series of concentrations with a lower limit (lambda) and a higher limit. Results below the lower limit will give a “less than” value (e.g. <0.005 EU/mL) and results above the higher limit will need to be further diluted and tested again to quantify the value.

A series of standard curve solutions must be created with two exceptions. The first is in using the ENDOZYME® II GO plate which has standards and sample spikes already dried in the plate and the second is in the use of an archive standard curve. The latter can be either a factory standard curve or a user-performed standard curve that is performed without samples on the plate and the curve saved to be applied to future sample tests. The standard application to the plate is critical because any error here will cause the necessity of repeat testing of the entirety of the samples on the plate.

This has always been the cornerstone of lab work for endotoxin testing in that a good method developed for a test makes for a good test over time. Interfering factors must be overcome by either dilution or treatment or both. Examples of interfering factors include chelation of divalent cations needed for endotoxin and rFC/LAL reaction, pH range adjustment, endotoxin aggregation (which needs disaggregation) and a host of issues that revolve around solubility and suspension. The official protocol must be followed and the protocol should allow for no ambiguity in that each analyst should be performing it in the same manner. If analysts are getting disparate results, then the problem may lie in the wording of the protocol. For example, if it says “transfer” a certain amount of sample but it is a viscous sample, then one analyst may rinse the pipette tip while another may not (see the next blog on technical writing). 

The actual test seems one of the simplest aspects of testing where the standards and samples applied to the plate are pre-incubated and the test is initiated with automatic, temperature-controlled, reading over time and automatic posting of test results in the software (extrapolation of sample values from associated standard curve values).

Retest is a critical part of any out-of-range acceptance criteria or, more importantly, for any out-of-specification (OSS) result.  This is where testing becomes most highly scrutinized so that analysts are not viewed as “testing into compliance” which, in simple terms would open one up to the charge of testing until you get the answer you want. During the initial test the sample must be withdrawn in an aseptic manner so as not to contaminate the sample for current or potential future testing. After an initial test a sample must be stored as per the defined storage conditions to preserve its integrity in case of retest. After all, the retest of the original solution is the soundest way to demonstrate that an initial test was inadvertently contaminated in the lab.

The dilution factor and appropriate application of the units constitute the final result as entered into the LIMS reporting mechanism that will accrue into the batch record with other quality attributes (sterility, etc.).  

Test release is the electronic approval of a numerical value with appropriate units. At this point all test steps will have already been second person verified, the test data archived, and all ancillary data also verified (sample weight entries, pH, etc.). The approval of the reported entry corresponds to the release of the product lot for a specific item in terms of this specific quality attribute (endotoxin).