Only gather the data you need, not the data you can, says a bioprocessing expert. Data is the foundation of biopharmaceutical manufacturing. The only way to understand if a production process is working as intended is to monitor key parameters at regular intervals during production.
Modern biopharmaceutical firms have an array of analytical technologies they can use to track processes in real time and feed information back to central control systems.
And regulators like the FDA and the European Medicines Agency encourage industry to gather manufacturing data to help ensure the quality of the medicines they produce the so-called quality-by-design approach*.
The challenge is knowing when to use analytical technologies and when such systems are unnecessary, according to Joanna Pezzini, president of PAK BioSolutions. She says manufacturers need to find a balance to stop “Big Data” from becoming too big.
“The question is where does new PAT really provide value? Often, we look for fancy ways to implement a PAT instrument into a large-scale manufacturing process where it may not be needed,” explains Pezzini.
“If a process is well defined, minimal instrumentation is needed to operate it. The value in PAT comes when we can eliminate samples that are currently sent to analytical labs—either through online or immediate at-line measurement—because we can then immediately respond to the result and save time on the sample submission process.”
All biopharmaceutical companies also need robust information management systems, irrespective of the extent to which they use PAT.
Manufacturing data from each unit operation is only of value if it can be passed to a central control point and used to model processes. Information siloed in poorly integrated systems is of little use for controlling processes or product quality.
Fortunately, the biopharmaceutical industry is showing signs of moving in the right direction according to Pezzini, who cites willingness to embrace “bioprocess 4.0” ideas as a positive sign.
“Databases that store online instrument data have been in place at most facilities for some time and the use of electronic batch records is growing significantly to record discrete data points about the process,” she continues.
“I expect, piece by piece, we will get to a point where all discrete measurements currently taken are connected to the EBR databases and can automatically record the information required.”
Growing interest in continuous manufacturing will also impact how industry approaches PAT and data management, says Pezzini who cites the demand PAK BioSolutions is seeing for its automated processing systems as evidence.
“We see a demand for implementation of continuous processes, which by their nature require more automation and AI to operate than seen in a typical batch purification process,” she tells GEN.