OR WAIT null SECS
© 2023 MJH Life Sciences™ and Pharmaceutical Technology. All rights reserved.
Analytical methods need a boost to ensure quality control for cell and gene therapies.
To date, the industry struggles to detect and classify contaminants in cell and gene therapies (CGTs). Key contaminants found within CGT therapies include—but are not limited to—bacteria, mycoplasma, viral, peptides/proteins, cytokines, growth factors, antibiotics, beads, serum, dead cells, unwanted mammalian cells, non-viable particulates, endotoxins, pyrogens, and bioburden. The lack of standards surrounding many of the potential contaminants is largely due to the novel nature in which these components are being utilized, according to a PLOS One article (1). In addition, the lack of standards for CGT comes from the naivety of the industry, with prior drugs often using endpoint sterilization as a highly successful method to reduce the contamination risk, especially in the pharmaceutical industry. This leaves the burden, or potential opportunity, for the CGT developer to develop their own analytical method based on knowledge and risk within their individual therapies. As a result, detecting any contamination in the final product after the therapy has been produced leads to high manufacturing failure rates and increases in manufacturing costs.
Unfortunately, sterilization methods cannot be adopted for the final product in CGT due to the living nature of the product. Sterilization by filtration, heat, radiation, or chemical would render the product ineffectual. Therefore, manufacturing requires highly controlled environments to limit the contamination risk. However, even with all these controls, there remains a high chance of contamination, no more than from entry to the cleanroom of personnel and materials (2,3). As a result of this risk, regulators are adamant about the use of good manufacturing practices (GMPs) as stated in 21 Code of Federal Regulations Parts 210 and 211. Even with GMPs in place, there are many causes of contamination and, as such, there remains an expectation that final product testing for contamination is conducted on every batch due to the potentially fatal consequence of administering a contaminated product.
Perhaps the greatest impediment to analytics used for contamination in the CGT industry is directly tied to the time and limited throughput of release assays, closely followed by their highly manual and laborious nature. The standard reference techniques (e.g., United States Pharmacopeia [USP] <71> 14-day test for sterility, USP <63> 28-day test for mycoplasma) are designed for the pharmaceutical industry where large batch sizes and freedom of time allow for them (4,5). Some rapid microbial methods (RMMs) for sterility exist (BACTEC and BacT/ALERT), but detection can take five to seven days. Furthermore, RMMs require extensive validation (6). In addition, standard reference techniques were developed for products where any living cell is often considered a contaminant. The final product is a living product, and yet there is a need to characterize and differentiate between the desired living cells (such as T-cells and Haematopoietic cells) from the unwanted living component (such as bacteria and fungi). Especially for autologous therapies, such as chimeric antigen receptor T cells (CAR T), it is necessary to test every individual batch that is produced, each of which is of limited volume, and critically time is of the utmost essence for delivery of the therapy back to the patient.
A paper published by the BioPhorum Operations Group highlighted the impediment in time-to-result as being one of the main areas of dissatisfaction of current GMP release tests across a survey of biopharmaceutical end-users (3). Further, to deliver required treatments to the number of patients awaiting therapies, scaling out in parallel to have many different and unique batch products is essential. For the CGT industry, the most likely target for overcoming this current roadblock is through technological innovations.
Beyond microbial contamination, non-viable particulates can be shed from any number of raw materials or be present in raw materials as manufacturing by-products. Therefore, particulate identification must be built into development work to inform testing methods and timing. Standard nonviable particulate testing for injectable biologics is achieved through light microscopy, USP <788>, although the technique is extremely manual, tedious, and subjective (7). In addition, all therapy products themselves are formed from particulates (e.g., cells and protein clusters). As such, determining undesirable particulates from the therapeutic product remains a challenge. Once particulates are detected, identification of material can be accomplished through scanning electron microscopy (SEM) and/or Fourier transform infrared spectroscopy—both destructive methods that require highly skilled analysts. These destructive tests lead to, in the case of autologous therapies, a patient not receiving their therapeutic.
The lack of standards for many possible contaminates also offers an opportunity for new advances in technology to play a significant role in ensuring the safety of therapeutics. As analytical power and capabilities have advanced, more attention is being focused to apply new techniques to address this quality control bottleneck. One logical approach is through molecular analytical methodologies for microbial contamination detection. The majority of these take a polymerase chain reaction (PCR) approach, where detection can be achieved at high levels of sensitivity and in relatively short assay periods (3). PCR analysis is not without limitation, as high levels of intra- and inter-laboratory variability in these assays pose a problem for comparative analyses. Digital droplet PCR (ddPCR) is also an extension technology of growing interest in CGT analytics. ddPCR can show higher sensitivity and accuracy in detecting pathogen load, and thus could provide an advantage over more traditional RT‑qPCR assays (8).
One area of notable mention is the focus on next-generation sequencing (NGS) for contamination tracing. In its role of bridging burgeoning technologies to solve industry challenges, the Cell and Gene Therapy Catapult partnered with a major cell therapy developer and a technology provider to collaborate on a multi-year project with a focus to develop novel rapid sterility tests for CGTs. This was focused on sequencing of target DNA and the use of highly sensitive digital PCR capable of detecting contamination to a single copy number, ideally also discriminating between live and dead microbes (6). In an article published in BioRxiv, experts have reported on the development of a sterility assay based around NGS and its application to testing in mammalian culture samples (9).
Beyond the traditional approaches described above, there remains a great desire to shorten overall test time, limit necessary sampling burden, and also apply quality-by-design (QbD) principles to contamination testing such that monitoring can be done more regularly throughout CGT manufacture—rather than waiting on the testing result for the final product. Advancement in technologies is enabling new methods for assay development. One such is microfluidics, which has shown incredible versatility in control and application to diagnostic platforms in recent decades. An article in SLAS Technology described a microfluidic separation and detection system for rapid microbiological screening which can be applied directly on mammalian cell samples (10).
In terms of non-viable particulates, if development work can be done to understand typical particulate sources during the manufacturing process, non-viable particulate testing could theoretically be moved to in-process to either replace or inform final product testing. Flow imaging may be a viable method, although product-specific algorithms may be needed depending on the cell product type (11). Alternatively, current cartridge-based cell counters or flow cytometers could be configured to distinguish particulates from cells.
Although each of the discussed approaches to contamination identification holds great promise to simplify and reduce quality control (QC) burden, realizing these analytical capabilities in the CGT space remains a challenge. Certainly, at a commercial manufacturing level, there is still a heavy reliance on methods that will be ineffective to meet the growing demands of the industry. These therapies have very unique testing challenges, and the time to result is especially critical (12). The ability to move testing upstream is an approach that alleviates time delays with inline sensing to provide real-time detection—another advancement that, to date, is not technologically possible. Around these approaches, the use of automation can considerably alleviate timelines of CFR/USP compendial test methods, along with enhancing any upcoming technology (12,13). The application of developing techniques and novel technologies has also facilitated massive strides forward, and as new technology comes to market, it is likely more specific solutions tailored for the CGT industry will be developed. However, these are no small factors in evaluating new analytics for QC release.
The ability to successfully measure contaminants in the final product is crucial to ensure safe products reach patients, but it feels more like a bandage than a solution. The primary aim of QbD is to reduce the likelihood of contamination throughout the process, rather than evaluate the final product, although in CGT it is likely both situations will occur, reducing and final product testing.
Measuring for contamination ensures the safety of drug products, but any positive results (whether true or not) lead to manufacturing failure as well as failure to treat patients. Flipping the model suggests that we need to design an industry where contamination is minimized even before manufacturing begins. The growing shift from open manufacturing systems to closed platforms revolutionizes the approach. By moving to a (successful) fully closed system, by definition, would ensure no environmental exposure to the therapy—thereby removing one of the largest sources of contamination. However, this is just one driver for moving toward closed systems, as a reduction in environmental costs due to lower grade facilities, but also implementation of automation is easier for a closed system. For this reason, more often than not, closed and automated systems are used in combination. But this does not solve other contamination issues such as those from raw materials or consumables.
For this, CGT developers need to work closely with vendors of such supplies to satisfy quality requirements. Sterility has and needs to be of utmost necessity. But other CMAs (non-viable particulates, pyrogens, bioburden specifications) expected from vendors should be defined. The current burden of cost, time, and accuracy for therapies are easier on raw materials, often manufactured in large lots, spreading the cost of testing across many products, along with longer stability times as to not create supply challenges. Only through demanding this from raw material suppliers can the end-to-end contamination risks be further minimized.
Even with full automation, process closure, and contamination specification on raw materials, there will always be the need to further reduce any risk of contamination during the entire manufacturing process. There will remain the need for contamination testing, but moving the testing upstream alleviates some of the current challenges on CGT manufacturing. These analytical methods continue to take massive strides forward. By increasing control and implementing QbD principles, efforts can be made to reduce the overall burden of analytical release testing on products. For required QC results, faster, more efficient, and cheaper techniques can be sought out, which can be automated to a large extent and remove much of the highly skilled labor costs. Additionally, through the adoption of new assay methodologies, solutions may emerge which provide much higher levels of sensitivity and accuracy for contamination detection, which ultimately increase assurances around product safety—and ultimately patient safety as well.
David Smith, PhD, Vice President of Technical Operations, Ori Biotech. Courtney LeBlon, PhD, Director of Technical Operations, Ori Biotech. Stephen Shapka, PhD, Lead Automation Engineer, Ori Biotech