Process Analyzers for the Pharmaceutical Industry Since 1984

Historically, pharmaceutical GMP’s assumed that if you validate a process and never change anything you will achieve the same quality product every time. Over the past 30 years we have learned that this isn’t an accurate assumption, which is why the FDA introduced in August 2002 the initiative Pharmaceutical cGMP’s for the 21st Century: A Risk Based Approach. Process Analytical Technology or PAT is a key component of that initiative with process analyzers, or process analytics, providing timely chemical information.

In general, a PAT program is designed to “enhance understanding and control the manufacturing process”. The major value of process analytics is different depending on what stage in the process development cycle it is being applied. The major value to the process development lab and the pilot plant is process understanding while, to manufacturing, the value is in attaining high precision, efficiency and safety. Of the tools amenable to process analytics, high performance liquid chromatography (HPLC) offers selectivity, sensitivity and robustness not achievable with other measurement tools.

Dionex has been working with pharmaceutical companies to develop and manufacture on-line LC-based process analyzers since 1984.

Understanding Critical Parameters in Process Optimization

In the process development lab, the major goal is to develop the optimum process (e.g., minimize impurities, maximize yield, use inexpensive reagents, minimize unique equipment, etc.) by identifying what parameters (e.g., reaction times, reagent addition rates, reflux rates, temperature, etc.) are important. Some labs may start to investigate process control at this scale, but this is not common.

The greatest benefit that process analytics provides process development labs is the in-depth understanding of the factors that impact product quality, and it does so more efficiently than can be achieved using off-line assays. The company can then use that process understanding to identify critical process parameters (CPP) that need control in order to maintain a consistent level of quality. As raw materials or conditions change, they can make changes on the fly and still maintain product quality.

Scale-Up Process Analytics from the Bench to Pilot Plant

Once the optimum conditions or series of conditions for a given process are identified, the process moves to the pilot plant. Here, they try to run the process at a scale that is typically about 10% of full-scale manufacturing to determine if there are any scalability issues. The process is further refined to maximize the process. It is typically at the pilot scale where the engineers begin to develop control algorithms; i.e., feedback control of reagent feed rates, temperature ramp rates, etc. Once the process is believed to be optimized, they begin to see how reproducible and robust the process is and address any issues that contribute to poor reproducibility. They will also refine the process controls so that information can be transferred to the process engineers and control engineers designing the full scale manufacturing process.

Cycle time reduction is one of the potential benefits that can be derived from properly implemented process analytics. As an example, Dionex has been able to help one pharmaceutical company with the development and implementation of an on-line HPLC system, increasing the throughput for one purification step by a factor of 10.

Dionex Process Analytics Help Reduce Process Variability

Once a pharmaceutical process is transferred to manufacturing, the process will typically be validated and locked down. No more changes. An example that gets used a lot is a pharmaceutical blending process. The typical way to run blending is to put the ingredients in a blender, run the process while taking samples that are run off-line, and then determine when the blend is uniform. The problem is, there can and will be slight variations that impact achieving a uniform mixture even though the blend time was exactly the same.

What is needed is a timely direct measurement of the critical quality attribute (blend uniformity) that can be used to determine the process endpoint. If you can measure the critical quality attribute in a timely manner and understand what parameters impact that critical quality attribute, you can tweak those parameters on the fly to achieve the critical quality attribute and use this approach to reduce variability in the product. Direct measurement of the critical quality attribute so it can be used to control the process (i.e., reduce variability) is the major benefit of process analytics at manufacturing scale. The lower the variability, the lower the risk of exceeding a product specification. This is where  a 2-3 sigma process can be moved to a 6 sigma process.

With Dionex IC and LC systems support and service, implementing process analytics can be achieved with a minimum of pain and maximum of productivity.