The Perennial Perils of Aseptic Manufacturing

By Mark I. Schwartz –

Unlike solid oral dosage form drugs that don’t require sterilization, generally, parenteral drug products must be sterilized and, due to their nature, many parenteral drug products cannot be terminally sterilized. This leaves aseptic manufacturing. In the aseptic process, the drug product and the container/closure are first subjected to sterilization methods separately, and then are brought together. Given that, in these circumstances, there is no process to sterilize the drug product in its final container, the final product must be filled and sealed in an exceedingly high quality environment. See FDA, Guidance for Industry: Sterile Drug Products Produced by Aseptic Processing – Current Good Manufacturing Practice (Sept. 2004), at 2.

As an example of what can go wrong in an aseptic environment, at a recent FDA Team Biologics inspection at a facility in Europe, investigators noted in the 483 a wide variety of what they concluded were microbiological-based manufacturing issues, including: a viral contamination event; mold-like material in the fermentation and purification areas; inadequate CAPAs related to contamination events involving a spore-forming microorganism and one from the species Methylobacterium; as well as a CAPA from a contamination event related to a soil bacteria (Bacillus thurengiensis); cleaning and cleaning validation issues; and inadequate environmental and microbial controls. (The 483 also contained three purportedly significant labeling observations, which we won’t dwell on here, given the subject matter of the posting.)

Theoretically, any one of these events, assuming they actually occurred as suggested by the FDA investigators, could have resulted in product contamination and recall, though it is not even clear that the products manufactured during this timeframe were ever released from the facility.

Given that much can go wrong in an aseptic manufacturing process, and microorganisms are ubiquitous in our environment, many firms find that significant microbiological contamination issues occur at facilities engaged in such manufacturing. Indeed, FDA has stated that the vast majority of drugs recalled due to nonsterility or a lack of sterility assurance in the U.S. have been produced via aseptic processing. Id. at 3.

And certainly much is alleged to have gone wrong at the European site inspected by FDA’s Team Biologics late last year. However, our ability to understand precisely what transpired is hampered by the liberal redactions that FDA made to the 483 before it was cleared for release by the agency. Nevertheless, the following has been alleged: first, the purported inadequate microbial control supposedly consisted of a failure to perform bioburden and endotoxin testing at certain steps in the manufacturing process.

Regarding the corrective and preventative actions that “…were found to be inadequate…”, 21 CFR Part 211 does not even speak of corrective and preventative actions (i.e., CAPAs). So, the investigators’ conclusion that they were inadequate is somewhat of a red herring. Rather, 21 CFR 211.192 requires that “…[a]ny unexplained discrepancy…or failure of a batch or any of its components to meet any of its specifications shall be thoroughly investigated. . . . ” So a more appropriate observation by FDA would have been that these investigations had allegedly not been thoroughly investigated.

One of the purportedly inadequate investigations involved a water breach from a cooling system that leaked into the adjacent holding tank due to an improper weld. The cooling liquid was analyzed as part of the investigation and the soil bug, Bacillus thurengiensis, was recovered. Yet, FDA investigators claimed that no CAPAs had been initiated “…to clean, disinfect, sanitize, or sterilize the cooling system.”

Another purportedly inadequate investigation was opened as a result of a contact sample which allegedly grew 96 colony forming units (presumably well exceeding the action level), including a species of Methylobacterium and certain spore forming rods.

The viral contamination event is not well described by FDA in the 483 but, to the extent that it actually occurred, it likely contaminated the cell line, and apparently the firm was allegedly unable to detect where the virus had allegedly entered the manufacturing process. In addition, the FDA investigators criticized the firm because, allegedly, “…[a]dditional samples are not tested from the previous harvests in an effort to verify that virus is not present in low concentrations or non-uniformly throughout the fermenter.”

Next, cleaning and cleaning validation are significant programs at aseptic processing facilities. When either is inadequate the result is likely to provide evidence of microbiological contamination, because dirt and residue can be vehicles for, and oftentimes nutrition for, microbes. In this case, regarding cleaning validation, there was allegedly no procedure or requirement to repeat it or to re-evaluate it, even when circumstances may have warranted. Allegedly, a non-conformance was opened after two valves on a tank holding sterile product were observed to have a residue (believed to be an aggregated protein). Yet, allegedly, no samples were taken of the residue for definitive identification or for microbiological testing.

Regarding the cleaning operations, FDA concluded that equipment which had been previously cleaned, and verified as cleaned by an operator, was observed to allegedly contain a dried white residue. There was also allegedly inadequate cleaning of the fermentation and purification areas of the facility, where investigators noted “…[t]he trench next to the steam condenser in fermentation contained growth-like blobbed material as well as black mold-like material. One leg for the tank was observed to have black mold-like material on it.”

Finally, there were several alleged instances where equipment appeared to be wet and where the residual moisture was not evaluated for microbial contamination. Specifically, a tank was “…observed to contain a significant amount of residual liquid which covered more than 80% of the bottom of the vessel…” and in another situation, “…residual water was observed inside piping for tanks…and rouge was observed…[yet]…[t]here were no swabs taken inside the tanks or piping for microbiological testing. There was no bioburden or bacterial endotoxin testing of the remaining liquid. . . .”

Industry practice dictates, and FDA strongly recommends, that sterile drugs be manufactured using aseptic processing only when terminal sterilization is not feasible. Id. In other words, aseptic manufacturing is fraught with problems that are absent when manufacturing products that can be terminally sterilized. These inspectional observations would purport to buttress this proposition. Of course, biological products, like the drugs manufactured at this facility in Europe, cannot be terminally sterilized.

We’ll have to wait and see whether CBER determines that the observations in this 483 meet the threshold of regulatory significance to warrant a Warning Letter. Rest assured, you will be kept posted on all developments.