Mark Nachtigall, PhD, is senior scientist, Manufacturing Science and Technology, Pfizer CentreOne.

Mark Nachtigall, PhD

	Lyophilization is an important method for producing stable drug products that otherwise would not be stable under common storage conditions, and thus expands opportunities for the development of life-saving medicines. The process of developing a lyophilization cycle is complicated, however, and requires in-depth knowledge of the drug substance and other excipients in each formulation, the container/closure system used, and the lyophilization equipment and conditions. When a product is developed in the laboratory setting or using small-scale equipment, such expertise is particularly valuable during technology transfer and scale-up of lyophilization processes. While these operations generally run smoothly, there are times when conditions in the laboratory-scale setting cannot be duplicated at commercial scale; production, formulation, container, or equipment-related problems arise. Effective troubleshooting strategies, the ability to collaborate closely with vendors of lyophilization equipment and consumables, and comprehensive development programs are essential to successful implementation of commercial lyophilization processes.

Avoiding pitfalls and overcoming challenges


	The pitfalls and challenges associated with technology transfer and scale-up in lyophilization are often due to the fact that laboratory-scale processes cannot exactly mimic processes at larger scales. Meanwhile, site-to-site transfers at the same scale can suffer from even slight variations in procedures that at first glance seem inconsequential. These issues can arise even though great care is taken to evaluate the cycle, process, container, and closure to minimize risks during scale-up and commercialization.

	Fortunately, with in-depth understanding of the chemical and physical properties of the product and its formulation, a thorough understanding of the impact of scale on process parameters, a robust development process, and the crafting of detailed and explicit procedures, unexpected and unavoidable issues that arise during the transfer and scale-up of lyophilization processes can be rapidly and successfully addressed. Indeed, expert upfront analysis and effective troubleshooting capabilities can serve as effective risk management tools, reducing the likelihood of problems during tech transfer and scale-up of lyophilization processes and ensuring consistent and timely supply of life-enhancing medicines.

	The following case studies highlight specific problems that may occur during the transfer and scale-up of lyophilization processes, even when all practical steps have been taken to avoid them. In each example, the authors present a description of the problem, their response, and the lessons learned.

Case study one: The value of effective troubleshooting


	For some processes, although no issues are observed when performing the proposed lyophilization cycle at development scale, problems occur when a full-scale engineering run is conducted in a production-scale lyophilizer. Without any indication of potential problems, these issues cannot be anticipated. Therefore, having an established troubleshooting strategy in place is crucial for rapidly resolving such problems and maintaining the commercialization schedule.

	In a project at Pfizer CentreOne, 11% vial damage (including breakage) was observed in the full-scale engineering run using a proposed cycle for which no vial damage occurred at the development scale (see 

). This problem is significant because any loss of vials due to damage reduces the yield of the process and has a negative impact on productivity.

	It was noted that in the engineering run, more vials were damaged on shelves located higher in the chamber (

). This behavior was attributed to the residual heat from the sterilization cycle, which was significantly different than that observed at laboratory scale. The residual heat had the effect of providing a more aggressive lyophilization cycle for the vials higher in the chamber.

: Percent damage to vials in an engineering lyophilization run as a function of shelf position.			

	A review of the development history was conducted to identify potential causes for the higher breakage rates in the production lyophilizers. This troubleshooting exercise revealed that issues had been observed for the development of this particular cycle. The cake height was above the middle of the vial, a generally avoided practice; however, the fill volume was not changeable and using a wider vial reduced the number of vials in the chamber, reducing the efficiency. Also, the most aggressive cycles were most likely to damage vials. In addition, ready-to-use vials used straight out of the packaging were less likely to be damaged than production-prepared vials (obtained to mimic production conditions more closely). Efforts were made to reduce the incidence of damaged vials by using production-prepared vials and reducing the aggressiveness of the cycle. At the time the development work was completed, the vial damage was not considered to be an issue for production because it had been resolved. The rate of damage in the production lyophilizer, however, was significantly greater than that observed for the development lyophilizer.

	Three options for resolving this type of issue were considered: use of a less-aggressive cycle; longer cooling of the chambers to remove residual heat from the steam-in-place (SIP) process; or use of a stronger vial that can withstand greater stress. The first two options were not appropriate for this particular project because the purpose of the project was to reduce the cycle length using an aggressive cycle. Lengthening of the post-SIP cool down would have required modification of the programming by the lyophilizer manufacturer, and was therefore not practical because to do so would have added significant time to the project.

	The most attractive option was to identify a stronger vial that could withstand the aggressive cycle at production scale. Developing this solution required working closely with the vial vendor, which added time to the project. Fortunately, however, evaluation of the new vials was already in process, and therefore, the time to implement this solution was less than might have been expected.

	The company worked in close collaboration with the glass supplier to identify several possible vials for a second engineering run. Two alternate vials were selected and used in the same batch: medium-strength (designated as C1) and higher-strength (designated as D1) compared to the original vial. Both provided significantly reduced vial damage during the engineering run (

). The vial that was chosen made it possible to reduce the vial damage rate to 0.2%.

: Results of a second engineering run with alternate vials. Due to automated loading systems and manufacturing constraints, the C1 vials were loaded only on the upper shelves and the D1 vials only on the lower shelves of the chamber.			

	This experience underscored that known problems observed at development scale may occur at a much greater magnitude at production scale due to factors such as residual heat from SIP and glass damage during the washing, depyrogination, filling, and loading steps. Vial damage during laboratory runs should be closely monitored and its potential impact at production scale carefully considered, even if conditions are ultimately identified under which no damage occurs. In general, if damage is observed at the laboratory scale, problems should be anticipated at production scale.

	Furthermore, the rapid resolution of this issue was achieved due to the contract manufacturer’s ability to collaborate closely with the vial vendor. Fostering of such close relationships with key suppliers can have a significant impact on the ability of contract manufacturers to ensure the on-time delivery of on-specification drug products to their biopharma partners.

	It should also be noted that vial damage can occur by two possible mechanisms: thermal expansion of the material in the vial with changes in temperature and excipient crystallization. The latter issue occurs for excipients that adopt amorphous structures when they initially freeze, but then subsequently crystallize, leading to a change in their molecular size that can place significant and unexpected stress on vials. Therefore, careful consideration of the choice of excipients and having a clear understanding of their freezing/crystallization behavior can aid in the development process. For instance, when using an excipient that is known to undergo two-phase freezing, an annealing step may be introduced during which the excipient will crystallize when in a softer state, rather than later when the forces involved will be much greater.

Case study two: Paying attention to lab anomalies


	Many manufacturing complications are caused by specific manufacturing procedures that are essential for maintaining aseptic processing conditions. These types of stringent and highly regulated conditions are not required during the development phase. The second case study involves the sticking of vial stoppers to the underside of the shelf by the end of the lyophilization cycle. It is important to prevent stoppers sticking because sticking can lead to breakage of the stopper seal and loss of aseptic integrity and/or oxygen ingress, which can lead to degradation of the drug product, reducing the yield and impacting productivity. This problem is fairly common for small, lightweight vials that require larger stoppers that have a large surface area that can potentially come in contact with the upper shelf during manufacture. This issue is rarely observed at the development scale and thus not initially considered to be of consequence.

	In the present case, the stoppers were purchased as pre-washed and pre-siliconized, with no coating and minimal silicon applied by the supplier. As a result, the level of siliconization was ineffective and the stoppers were sufficiently tacky to stick to the underside of the shelf during stoppering.

	It was quickly determined that choosing an alternate, less-tacky stopper would lengthen the project due to the need to repeat stopper compatibility and stability studies. The on-site addition of silicone to the tops of the stoppers was unappealing because to do so would defeat the purpose of purchasing a ready-to-use (RTU) stopper. In addition, higher levels of silicone have the potential to generate high particulate counts within the drug product and may cause the stoppers to back out once seated. The chosen remedy involved the application by the supplier of a crosslinked silicone coating to the top surfaces of the stoppers that had already been evaluated. This solution avoided any product contact concerns and allowed the continued use of the RTU stoppers.

	In addition to clarifying that larger stoppers used with smaller, lightweight drug-product vials should be purchased with a pre-applied outer coating to reduce tackiness, this experience underscored the importance of monitoring anomalies observed during laboratory runs. Although the sticking of the stoppers occurred rarely during development runs, the issue was clearly significant on the production scale. Collaboration with the supplier was an important component of the solution to this case study.

	It is worth noting that lyophilization equipment manufacturers are aware of the issue of stopper stickage. Tray manufacturers now offer various solutions to the problem. Trays are now available that are designed to prevent stoppers from sticking through the application of either coatings on their bottoms or textured undersides. The use of these trays eliminates the need to coat the stoppers. It was also observed that for lyophilization chambers requiring manual tray loading, placement of a plastic cover over the top of the tray is an effective solution for preventing stopper stickage. This approach has the added benefit of providing an additional protective barrier for the vials.

Case study three: Guiding engineering runs with development studies


	In the third example, the cycle time of an engineering run on the production scale was longer than expected (as indicated by temperatures measured using thermocouples located inside the lyophilizer) based on the development data (see 

	Robustness studies had been conducted during development at shelf temperatures and chamber pressures both higher and lower than the values considered to be optimum. A comparison of the production and development data revealed that the thermocouple temperatures in the engineering run were more similar to the conservative development cycle than the optimum cycle.

	This difference was thought to be due to freezing of the product. The production environment is much cleaner than the laboratory environment, with few if any particles present to aid ice nucleation of the drug product. Supercooling prior to freezing is one possible consequence of such a clean environment, which can lead to the formation of smaller ice crystals and a cake that is more resistant to drying. 

	Because the conditions during the engineering run were found to be closer to the conservative development cycle, the shelf temperature and chamber pressure were both increased-within the design space defined during development-in order to achieve the desired cycle time. A second engineering run confirmed that the modification to the lyophilization parameters was successful.

	This project clearly demonstrates that the design space for a given process at the laboratory scale can be different from the design space at the production scale. It also underscores the fact that it is often not possible to duplicate the production environment in a laboratory. It is therefore best to be prepared for differences between engineering and development runs.

	The most important lesson learned from this experience, however, is that robust well-documented development studies can be invaluable for rapidly resolving unexpected lyophilization behavior at the commercial scale. The most benefit can be gained if the same team is involved in running the development and production batches, because their knowledge can facilitate rapid comparison of the two data sets and identification of the root cause of the unexpected performance results. In addition, performance of engineering runs in commercial equipment guided by thorough development studies can enable accelerated implementation of optimal lyophilization cycles with the most efficient cycle times.

	As a final point, wireless thermocouples are ideal for use in lyophilizers equipped for automated loading. These thermocouples can be placed in multiple locations with a lyophilization chamber and provide invaluable data during engineering runs.


	Lyophilization is a complicated process that can pose challenges during tech transfer and scale-up to commercial manufacturing, but issues that do arise can be resolved if appropriate risk-mitigation measures are taken in advance. A few key steps include:

		Perform engineering batches in commercial chambers, preferably with the actual drug product, but using a placebo if necessary.

		Perform full-load runs in both the laboratory and production lyophilizers for worst-case conditions to obtain robust data for useful comparison and statistical evaluation.

Articles

Troubleshooting and collaboration are essential in implementing commercial lyophilization processes.