Downstream sample preparation for subsequent quality attribute analysis represents “a significant bottleneck in process development for non-antibody biologics,” wrote Kevin Brower, PhD, global head of purification development—Mammalian Platform, at Sanofi in a recent journal article. Bioprocessors could lessen the impact with process analytics, but that is the topic, perhaps, of a future post.

The multi-step chromatography steps typically required lead to limited throughput and long lead times before product may be released and, generally, more resources than one might suspect. In his article Brower described an immunoaffinity purification technology that achieves single-step purification of two therapeutic enzymes (one, Fabrazyme®, which is commercial) using polyclonal and monoclonal antibodies as ligands. Brower claimed his work was a powerful tool demonstrating the suitability of analytics to predict critical quality attributes of a commercial enzyme biologic. “These models,” Brower wrote, ” empower process development decision-making with drug substance-equivalent product quality information without generation of actual drug substance.”

GEN asked Brower what happened to “dilute and shoot” methods? He responded that yes, this approach often works for Fc-containing proteins and that many quality attributes in this molecular class can be measured without cleanup.

However, the lower enzyme titers relative to antibody processes means a different approach is required. “The enzymes are of approximately equal prevalence to host cell proteins in the bioreactor, and enzymes are structurally similar to many HCPs. Without affinity purification, the level of purity even after non-affinity capture chromatography can be too low to enable confident measurement of product quality. ”

Brower’s method is not magic, as it requires first raising antibodies to the target enzyme, creating an affinity resin, and validation. “This technique requires up-front investment which, depending on the anticipated number of samples could be too much effort for the return.”

The method mentions isolation of “target molecules…without generation of actual drug substance.” Which raise the question of whether “drug substance,” a regulatory term, and “target molecules” are one and the same.

“There is some nuance here, at least in my opinion,” Brower tells GEN. “Drug substance is the target molecule, but it is contextual. Drug substance is the target molecule one obtains as a result of executing a specified process. This difference may require some bridging between the product quality produced by executing a non-affinity process that is used to make drug substance and the product quality obtained with an affinity purification. There may be no differences for some attributes, whereas for others one or both of the purification methods may alter product quality along the way. The work described is the combination of a sample preparation method and an approach to predict drug substance product quality produced by a non-affinity purification process.”

Previous articleBioindustry Ready to Adopt Continuous Processing within Five Years
Next articleMersana Lives in a QbD World