Abstract
For release of parenteral drug products, bacterial endotoxin testing is one of a panel of necessary tests. In order to ensure the validity of such tests, various controls are performed, including demonstration of compendial method suitability or method qualification. In addition to compendial suitability testing, quality control (QC) sample hold-time studies are requested by authorities like the Food and Drug Administration (FDA) as described in "Guidance for Industry: Pyrogen and Endotoxins Testing." It is requested to be determine whether the ability to detect endotoxins can be affected by storage and handling of the sample to be tested. To accomplish these studies, endotoxin is introduced or spiked into the undiluted product and held for a certain period of time in process-representative containers. This time period reflects procedural maximum QC sample hold time from sampling until analysis. Inadequate detection of endotoxin can be caused by adsorption of endotoxin to container surfaces or molecular masking effects, in which the binding sites on the endotoxin molecules are prevented from triggering the enzymatic cascade necessary in the assay, are obscured. The endotoxin may form macromolecular structures, such as sheets or blebs, or the binding sites may otherwise be rendered unavailable due to the sample matrix composition. In either case, the endotoxin assay may yield falsely low results if and when masking occurs. In this work, the QC sample hold times of different in-process controls within the production process of a biopharmaceutical product were analyzed. One out of eight different samples showed a strong masking of endotoxin. Analysis of the sample composition revealed that either kifunensine, mycophenolic acid (MPA), or lauryl-N, N-dimethylamine oxide (LDAO) was responsible for masking. Further analysis clearly identified LDAO as the root cause for masking. A novel one-step mechanism for LDAO-induced endotoxin masking is proposed. The principle is similar to an already-proposed two-step mechanism for endotoxin masking, but the LDAO case combines these two steps: the disturbance of the salt bridges and hydrophobic interactions with LPS in one molecule. These molecular interactions occur quickly when both endotoxin and LDAO are present in the same matrix. Thus, depending on the masking agents, low endotoxin recovery (LER) can occur regardless of the QC sample hold duration.