March 1, 2014 (Vol. 34, No. 5)
Kathryn Martin Remington, Ph.D. Principal Scientist BioReliance
Chas Hernandez Senior Scientist & Virologist SAFC
Facilitating Performance Assessment of Complex Media Components or Formulations
High-temperature short-time treatment (also known as HTST or flash pasteurization technology), a mature technique in the food industry, is quickly becoming an integral part of many biopharmaceutical companies’ viral risk-mitigation strategies.
HTST involves rapidly heating a material to a predetermined temperature and holding it at that temperature for a specified time (typically 10–15 seconds) to inactivate adventitious viral agents that may be present in the product.
Before any risk-mitigation strategy is implemented in a process, it must be properly evaluated. In the biopharmaceutical sector, it is important to know that not only are adventitious viruses effectively inactivated by the treatment, but the thermal processing does not impart significant physical or performance changes in the product.
The evaluation itself can be problematic, as it is rarely feasible to carry out viral clearance studies at pilot or commercial scale, primarily because of the hazards of handling relatively large quantities of virus stocks. Furthermore, introduction of virus to manufacturing equipment is an unacceptable risk, so the viral reduction must be evaluated on a small scale, with results extrapolated to larger-scale processes.
Optimization of HTST processing parameters is required to ensure treatment will provide sufficient inactivation against a broad spectrum of viruses, which may include the more resistant nonenveloped viruses such as parvoviruses or porcine circoviruses (PCV). Recent PCV contamination events in the vaccine industry highlight the importance of pretreating media and cell culture components, as the contamination was believed to have entered the process via porcine-derived trypsin.
Processing conditions need to be sufficiently robust to ensure sufficient inactivation of adventitious viruses without imparting physical changes to the medium, which might significantly alter performance of medium in the cell culture process. Media makeup is variable in different processes, so each medium must be evaluated independently.
When carrying out a small-scale viral-clearance study, the goal is to replicate key operating parameters used at the larger scale. In this situation, a small-scale model system is developed to closely mimic the HTST procedure at pilot or commercial scales. Relevant analytical and biological parameters demonstrate the comparability.
Consequently, viral-clearance data from the small-scale experiments are shown to be valid representations of the clearance that could be expected at full scale.
After evaluating commercially available benchtop systems for performing viral-clearance studies, BioReliance decided to develop an in-house benchtop system using static immersion. In this model, virus-spiked test material is contained in a small stainless steel tube, which is then immersed in a circulating hot oil bath. The temperature of the bath and the internal temperature of the tubes are monitored.
Control tubes are also prepared, but not thermally processed. Once the tubes reach the desired temperature, they are maintained for a specified period of time—typically 102°C for 10 seconds. The tubes are then removed from the oil bath, and immediately cooled in an ice bath. Infectious virus in the contents of the cooled, treated tubes are measured, and compared with the untreated control.
To test the benchtop system, a 50% glucose solution was used as a worst-case solution. Glucose solutions are more difficult to process by HTST because their viscosity is higher than typical medium, causing changes to heat conductivity in a product.
The thermal treatment is more problematic than for media solutions whose viscosities are more similar to water. The glucose solution was spiked with feline calicivirus (FCV), a surrogate for vesiviruses, and an aliquot of this mixture was transferred into four sterile stainless steel tubes. Two were thermally processed to the desired temperature, and held for 10 seconds before cooling in an ice bath; the other two were held at 25°C as controls. Three different temperatures were assessed: 97°C, 102°C, and 107°C. Virus titrations were carried out on all samples using Crandell Rees feline kidney in 96-well plates, and plates were scored for cytopathic effect at days two and three post-infection.
A considerable decrease in infectious virus was achieved at all three temperatures (Table 1). Furthermore, complete inactivation of FCV was achieved at both of the higher temperatures.
The next step was to assess the effect of processing on the quality of medium.
This was done by HTST, treating larger, unspiked batches of 50% glucose solution at the pilot scale and comparing it with untreated batches. The HTST treatment was carried out in an Armfield FT74X UHT pilot-scale system, which uses a water-heated shell-in-tube heat exchanger, with the glucose solution heated to and held at 102°C for 10 seconds. Both the treated and untreated batches were sterile filtered, and then packaged into 10 L EVA bags.
The heating and processing profiles for both the pilot and benchtop scale were similar. The cooling time post-processing in the pilot-scale unit was significantly longer than in the benchtop system. As it remains at a higher temperature for longer, the potential for higher levels of viral inactivation may be achieved in this larger-scale system; the benchtop system represents a worst-case scenario.
Performance of both treated and untreated glucose solution in cell cultures was then evaluated. It was added at a concentration of 5.5 g/L to glucose-free, chemically defined CHO medium, and inoculated with SAFC’s CHOZN® GS–/– ZFN modified CHO cells at a density of 3.0 x 105 cells/mL. Cultures were counted daily to determine growth and viability.
The cultures were maintained in a Multitron incubator at 37°C and 5% CO2 until viabilities dropped below 70%. Spent media samples were collected on days 7, 10, 12, and 14 for analysis.
The growth curves for medium that contained glucose that was HTST-treated at pilot scale showed similar growth kinetic profiles to medium containing the nonthermally processed solution, with culture viabilities comparable across treated and nontreated media. HTST processing of the 50% glucose solution had no effect on the growth characteristics of the medium.
A diverse set of media and concentrates were also performance tested following HTST treatment at 102ºC for 10 seconds. Serum-free and chemically defined media, supplements such as vitamins, hydrolysate concentrates and amino acid concentrates were all included. In each case, no significant differences were observed in analytical data or cell culture performance using CHO cells following HTST treatment.
When introducing a new media or feed to a process involving HTST, or integrating HTST into an existing process, it is critical to test the compatibility of the media and supplements. Some components such as phosphates and high concentrations of divalent cations can precipitate at elevated temperatures, so verification at pilot scale is essential.
These results demonstrate that the small-scale, static, benchtop model, when used in concert with the pilot-scale HTST system, facilitates viral clearance evaluation and a performance assessment of complex media formulations or media components.