PAT Advances Freeze Dryer Control - Smart Lyophilization

PAT Advances Freeze Dryer Control
2004
Agnes Shanley, Editor in Chief
PharmaManufacturing.com
Process Models and Gas Analysis Help Optimize Lyophilizer Operation
Demand for aseptic packaging and preservation, as well as the rise in production of protein-based drugs, continues to boost global demand for lyophilization capacity. As a consequence, pharmaceutical manufacturers and their equipment suppliers are turning to an array of process analytical technologies (PAT) to help optimize their freeze-drying operations.
"Genomic and designer drug development will continue to drive the market for vial/parenteral systems," says Chuck Dern, lead engineer with SP Industries (Warminster, Pa.). Meanwhile, in tray lyophilization, the need to preserve the form and functionality of therapeutics, including nanotech and bioprocessing products, is pushing growth, says Walt Pebley, director of business development for Oregon Freeze Dry (Albany, Ore.).
Today, freeze-drying equipment is handling larger batches, and portions of loading and unloading have been automated, minimizing operator exposure and the potential for product contamination (Box). Until recently, though, monitoring the process generally meant inserting temperature probes into product vials, then using statistical methods to estimate average process parameters "an imperfect process that isn't well suited to freeze dryers with automated loading or unloading features.
"From the PAT viewpoint, one issue with using resistance probes in vial products is that you are monitoring a very local environment," says Eric Kratschmer, technology manager for SP Industries. Usually the stopper must be left off the vial; otherwise, the leads become a problem during stoppering. The probe itself provides a nucleation point that's not present in other vials, which can affect ice formation. "With probes," Kratschmer explains, "the environment that you are monitoring may not be perfectly representative. At best, you might be able to monitor 16 or 24 points in a lot of 10,000 to 100,000 vials." To a lesser degree, this also holds for bulk tray-dried products, he says.
Today, however, a number of PAT options are being explored that promise to improve users' ability to monitor and optimize the lyophilizaton process. "PAT promises to provide better control and potential economies," says Edward Trappler, principal of Lyophilization Technology, Inc. (Ivyland, Pa.), "and could be used most effectively to monitor the progression of the process."
The Smart Freeze Dryer
Equipment is close to commercialization that will feature built-in computer optimization programs. These control algorithms form the base of the "Smart Freeze Dryer," a program that grew out of research sponsored by the Center for Pharmaceutical Processing Research, and work carried out, initially at Purdue University, then at the University of Connecticut (Storrs). The algorithms can be adapted for use with lyophilization equipment.
FTS Systems (Stone Ridge, N.Y.) has licensed the technology from Purdue and the University of Connecticut and is now incorporating it into freeze drying equipment. The Smart Freeze Dryer algorithms use "manometric pressure," first described in 1997 by Purdue professor Steve Nail, who now works for Eli Lilly, to determine temperature, mass transfer and other key lyophilization process parameters, without sensors. Sensor-free measurement is an important advance, given the trend to automated freeze-drying systems in which probe insertion is cumbersome at best.
"The system runs on feedback, using expert system software that allows the freeze dryer to self-optimize," says University of Connecticut professor Mike Pikal. His research team, led by Charlie Tang, who now works for Amgen, developed the algorithms based on manometric pressure principles. "The operator feeds in internal pressure value, total solids content, then the program runs and gives a printout of the optimized process."
The platform's expert system software determines details of freezing, optimizes chamber pressure and target product temperature and secondary drying conditions, says Tang. Manometric temperature measurement (MTM), meanwhile, allows product temperature and the resistance of the dry product layer to be determined by analyzing "pressure rise" data, allowing calculation of the shelf temperature needed to achieve a given product temperature.
MTM measurements are used to select optimum shelf temperature, determine drying end points and evaluate residual moisture content in real time, Tang explains. Using heat and mass transfer theory, MTM results are then used to evaluate mass and heat transfer rates and to estimate shelf temperatures required to maintain product temperature. Freeze dryer overload conditions can be estimated by calculating heat mass flow at the target product temperature. Results can also be used to indicate the end point of primary drying, Tang says. The concept has been tested on multiple pharmaceutical products, including protein formulations and antibiotics.
FTS believes that its adaptation of the algorithms, which it is calling the "Smart Freeze Dryer Technology," will allow users to better understand their freeze-drying processes, study the impact of freezing on heat transfer, and support cycle changes for freeze-dried drugs. If, for example, a user wants to determine the effect of an annealing step during the initial freezing step for a particular product, the technology will produce process data, showing why a particular freezing protocol would be best for a given product. The technology could also be used in production, to optimize operating conditions and provide feedback for troubleshooting product quality and scaleup issues--for example, evaluating the effect of a higher degree of supercooling, which is normally seen in production environments, rather than the laboratory, says Joseph Brendle, FTS's director of technology and business development.
FTS is finishing software development for its Smart Freeze Drying Technology, and is incorporating it into its LyoStar II research-scale freeze dryer, Brendle says. "We're looking to do beta site tests over the summer and fall," he says. The company plans to install the technology on its production freeze dryers in the future.
Gas Composition Provides Process Clues
PAT techniques based on real-time analysis of lyophilizer gas composition also are being developed, Trappler says. Currently, the two most useful appear to be residual gas analysis (RGA) based on mass spectrometry and infrared (IR) spectrometry. There are some logistics issue involved with mass-spec RGA, he explains, the first being assurance of sterile conditions. "At this point, the instruments are not steam-sterilizeable," Trappler says.
Then, there is the question of interpreting the data. "Both RGA and IR monitor the composition of the atmosphere within the lyophilizer during processing to follow the evolution of water vapor," Trappler says. "But there are questions--for example, where is the best location to sample? Also, it's assumed that the composition of the atmosphere is uniform, but this has not been established yet."
IR is a promising technique but "reflectivity and depth are issues," says Oregon Freeze Dry's Pebley. "It's more a surface, than a penetrating technique."
"RGA is usually done with mass spec analysis, to see vapor in the chamber," adds Joe Brower, director of Operations at Cardinal Health in Raleigh, N.C. However, noise can be a problem. His company is, instead, using pressure-rise measurement to optimize freeze drying, isolating each chamber and taking readings. "It's a straightforward technique," he says. Oregon Freeze Dry uses load-cell measurements, looking at variations by examining product weight loss during the process.