Methods used to determine the contamination risk posed by the plastic components of single-use systems are resource-intensive and limited in scope, say researchers behind an alternative, whole process simulation-based approach.
Growing drug industry adoption of single-use systems has made leachability analysis vital, according to lead author Maximilian Bossong, from the Institute of Pharmaceutical and Biomedical Sciences at Johannes Gutenberg University in Mainz, Germany.
“Leachables, or process equipment-related leachables (PERLs) from single-use (SU) components, are trace chemicals that can transfer into process fluids.
“Even at low levels, they can affect product quality, process performance—for example, cell growth or yield—and patient safety. They can also trigger unexpected investigations or manufacturing delays. Because biomanufacturing processes commonly use many SU components and buffers, exposure needs to be risk assessed,” he told GEN.
At present, drug companies rely on a combination of supplier data, risk assessments, and targeted studies under “worst-case” or process-simulating conditions to determine the risk posed by leachables.
But, although such methods are effective, they are rather limited in scope, according to Bossong.
“Testing is resource-intensive and often only isolated unit steps are considered. In addition, it may not capture dynamic behaviors like adsorption to biomass or accumulation/dilution across steps,” he said.
The other issue is that assessments need to be repeated when processes are modified.
Simulation
As an alternative, Bossong and colleagues developed a simulation-based protocol which, they claim, significantly reduces the necessity for practical testing and provides more detailed information.
“We couple a mechanistic film-diffusion source term—release from the polymer into liquid with partitioning—to a dynamic “box” model that tracks PERL fate through a process flowsheet. This lets us include sinks and sources such as adsorption to cells/debris, dilution, equilibration, and hold-times.”
The new approach relies on several key inputs, such as information about liquid partitioning and diffusion parameters, as well as initial load estimates, which are derived from extractables studies carried out by suppliers.
It also takes device geometry and process conditions, like surface area, thickness, volumes, and contact times, into account.
To gather this information, Bossong and colleagues use a range of techniques—everything from targeted HPLC, GC-MS, or LC-MS—in combination with process metadata. They also carry out simple adsorption and binding studies to determine how those leachables that are present are likely to interact with cell lines or drug products.
“There are numerous potential advantages,” Bossong said. “The approach supports scenario prediction without running costly studies. It also enables the identification of low-risk steps and critical conditions and gives users the ability to integrate supplier extractables data with process metadata and provides guidance to focus on analytics where they are feasible and add the most value.”