Particle size - meeting the PAT initiative
George Crawley and Andrew Malcolmson, from Malvern Process Systems, examine the role of on-line particle size analysis in enabling companies to implement PAT in accordance with the FDA initiative
George Crawley and Andrew Malcolmson, from Malvern Process Systems, examine the role of on-line particle size analysis in enabling companies to implement PAT in accordance with the FDA initiative
The FDA's PAT initiative1,2 is intended to facilitate the introduction of modern process analytical technology tools and concepts to the manufacturing sector of the pharmaceutical industry.
It aims to encourage innovation while ensuring high quality, and facilitate the adoption of on, in, and at-line analytic technologies which can lead to an improvement in product quality through better process understanding.
With the increasing pressure on regulatory resources, inspection allocations are moving to a risk-based approach focusing on high-risk problem areas, rather than tightly controlled, lower risk processes. Companies using PAT extensively 'build in' quality throughout the manufacturing process. Their processes are therefore more tightly controlled than those companies still using only more traditional laboratory-based testing and are recognised to be of lower risk.
encouraging innovation
Due to the stringent quality control imposed on final grade product, materials released from the pharmaceutical industry are of an extremely high standard. However, relative to other manufacturing sectors, such as the automotive and semiconductor, the pharmaceutical internal production efficiency is not high, with significant batch rejection rates due in part to a lack of true process understanding. According to the Centre for Drug Evaluation and Research (CDER), a current good manufacturing process allows about 10% waste - an error rate that would be unacceptable in many other industries.
The FDA plans to change this with the introduction of its PAT initiative. PAT is linked to the FDA's current good manufacturing practices (cGMP) initiative - quality depends on knowledge and PAT brings more knowledge and understanding of unit operations.
PAT is defined by the FDA as 'systems for analysis and control of manufacturing processes based on timely measurements during processing of critical quality parameters and performance attributes of raw and in-process materials and processes'. In other words - real-time, continuous quality assurance intended to ensure acceptable end-product quality.
Some 70% of drugs are consumed in solid oral dosage form as tablets and capsules. The majority of constituents, or active pharmaceutical ingredients (APIs), and excipients, are particulate in nature. Particle size properties are thus very important characteristics of dosage formulations.
flow optimisation
For example, large particle size helps optimise flow during the granulation process of solid dose forms, while smaller particle sizes of APIs contained in tablets or capsules are known to enhance dissolution and absorption. Particulate processes include spray-
driers, mills, classifiers, and granulators. The continuous and on-line measurement of particle size is one way in which companies can ensure they meet the new FDA PAT criteria.
On-line particle size analysers, such as the Malvern Insitec system, give an important insight into, and understanding of, many particulate processes in the pharmaceutical industry. The Insitec system is routinely used in all areas of pharmaceutical manufacturing to reduce waste and maximise revenue - for the assessment of incoming material quality (where variability leads to many downstream processing problems); optimisation of powder unit operations; improvement of batch-to-batch consistency; and optimisation of process selectivity in real-time to maximise the proportion of in-spec material.
Endpoint approach
Currently processing time is used as the most common endpoint for a batch process. With the introduction of its PAT initiative, the FDA is encouraging the pharmaceutical industry to move towards a process endpoint approach that requires the identification of process-critical control points. For example, granulation is stopped when the optimum granulation size and moisture content is reached, whereas blending is finished when the blend becomes uniform.
One example of the way that real-time on-line particle sizing can be used for endpoint process control is in the creation of intelligent milling systems that measure particle size distribution and automatically adjust the milling parameters.
For example, Intellimill (a joint development between Hosokawa Micron and Malvern Process Systems) allows the closer integration of Insitec dry powder on-line particle size analysers in air classifier mills (ACM) and air jet mills (AFG). This brings the ability to continuously track and control particle size, so that tighter particle size specifications can be achieved. Eliminating out-of-specification materials brings immediate cost savings, both through reduced energy use and less wastage.
energy reduction
Intellimill offers simple operation with dial-in of the desired product grade, allowing the plant to make the transition automatically to the new setpoint. Automatic operation requires minimal operator intervention. Mill performance is optimised, reducing energy costs and wastage, while at the same time maximising throughput. Batch-to-batch consistency is improved, as is batch-to-batch comparability. Product grade changes run smoothly as the transition is fully automated and optimised.
Milling is just one aspect of pharmaceutical manufacturing addressed by the PAT guidance. The PAT conceptual framework (Figure 1) addresses every aspect of pharmaceutical manufacture, from incoming raw materials through optimisation to continuous improvement. It starts with the processability of the incoming raw material, and the FDA hopes that in the future sufficient information about incoming material will be available to allow incoming material attributes to predict or adjust optimal processing parameters. Within the unit operation, on-line measurement is used in real-time quality assurance and process optimisation. Information about end-product quality should be fed back into the process to achieve continuous improvement and process optimisation.
The implementation of PAT also offers the possibility of moving away from batch operation. Historically, the use of batch processing has allowed the subdivision of production runs. This has been necessary to isolate or eliminate off-spec batches, and also to allow the introduction of time lags that enable product quality data to 'catch up' with the product itself. It can be envisioned that smaller, continuous (and less capital intensive) processes, which lend themselves to real-time control, will become more widely adopted.
It is clear that on-line, or at least at-line, assessment of attributes that relate to performance and quality is key to realising all these goals.
On-line particle sizing facilitates real time optimisation of process parameters and configurations. Laser diffraction-based systems, exemplified by the Insitec, offer benefits as the technique requires no calibration and does not need long data acquisition intervals, which can mask process behaviour.
Insitec systems use industry-standard laser diffraction technology engineered into a rugged on-line measurement platform for process. They measure particles from 0.5 to 1000 µm in size and the range includes systems for: 21 CFR Part 11 compliance; GMP compliant processes; for hazardous environments where intrinsic safety is a requirement; and for the measurement of dilute liquid suspensions.
On-line procedures also minimise, or even eliminate, operator exposure to the materials being manufactured, potentially solving some of the practical issues surrounding the production of potent pharmaceuticals.
real-time data flow
The PAT guidance emphasises the importance of a streamlined data flow of information across the manufacturing process. Malvern Link software is the basis of a flexible approach to automation. It delivers the real-time particle size measurements generated by the Malvern Insitec systems to the plant control systems. It also drives automatic sampling and presents the operator with a simplified user interface. In total it enables the use of real-time data to achieve full automatic control, allowing self-regulation of the plant, and maintaining product within specification.
Malvern Link provides connectivity to a wide range of PLCs with full integration into various distributed control systems, for detailed data/instrument status reporting. It also provides improved flexibility in customising the measuring system to meet the needs of individual processes and users.
Two case studies are reported. The first is a mill start-up using off-line size analysis. Figure 2 shows how starting up a new process involves a series of iterative changes to process settings until the particle size is within specification. These iterations are necessary using off-line particle size analysis techniques because time is lost in waiting for the process to reach steady state following any process change (green dots), and the point at which a sample is then extracted (red dots).
on-line measurements
Additionally, there is time consumed while waiting for the off-line results to be completed and communicated to the operator from the laboratory. In this example, the time required to reach the acceptable particle size (defined as a series of two successive in-specification measurements) using off-line techniques was 50 minutes.
Figure 3 shows a comparison of starting up a mill using manual sampling/measurement against on-line analysis. By using in-process real-time particle sizing, there is no need to wait for steady state or for results to arrive from elsewhere. Process changes can be implemented immediately and dynamically to ensure that the desired particle size is achieved. In this example, the process was brought into specification in less than eight minutes.
The resulting reduction in processing time of more than 40 minutes eliminates the need to store, reprocess, or downgrade the out-of-spec material, and reduces the amount of energy consumed in 'tuning' the unit operation.
The second case study looks at mill optimisation. A pharmaceutical manufacturer encountered problems in optimising milling systems for pilot trials of different drugs. Each drug required a different size specification.
The typical cost of each product was US$0.5-1.0m/kg. The micronising system was fitted with a pharmaceutical grade filter to prevent API from being released into the atmosphere. Batches were processed in 500g lots.
There was a poor understanding of the influence of different process parameters on the particle size distribution of the final product, including:
• air flowrate
• air pressure
• feed flowrate
• mill speed
These parameters would be stepped through, each requiring a sample to be taken and measured on an off-line particle size analyser. Process validation was therefore arduous, with long phases of stopping and starting the mill. Typically, using off-line analyses, the milling optimisation study would take about a week. Yields were generally poor, with wastage represented by high amounts of super-fines in the filter, due to overgrinding. This meant the filter would block quickly and would require frequent cleaning.
An Insitec on-line particle sizer was installed in its in-line configuration at the mill exit and used to monitor the entire product as it left the mill (Figure 4). Instantly, instead of measuring small samples (3-5g) infrequently with off-line techniques, the entire batch was being characterised. The mill could then be adjusted by the operator in real-time to the correct particle size specification, without the need to stop the milling operation. This more 'active' role of the operator was seen as a more valued task than simply collecting samples for measurement. Using the Insitec, a more complete study that included more steps in parameters was achieved in less than four hours (a saving of 4.5 work days).
valuable tool
With its PAT guidance, the FDA aims to develop a regulatory framework to facilitate introduction of new manufacturing technologies that enhance process efficiencies and process understanding.
In-process analysers enhance process understanding by monitoring production either in-line or on-line to provide continuous real-time quality information.
Using real-time measurements, closed-loop control is made possible, which can improve batch-to-batch comparability and minimise wastage of valuable product with minimal operator interaction. This simplifies process operation and optimises efficiency.
Laser diffraction measurement technology
Laser diffraction is an established and widely used technology for particle size measurement and is the standard in most particulate processing industries. The technique has significant benefits in that it needs no calibration and does not require long data acquisition intervals that can mask true process behaviour. Its implementation for on-line measurement of dry materials is exemplified in the Insitec system (Malvern Instruments). With laser diffraction techniques, laser light is scattered by particles in the sample and the angular intensity measured by a series of photodetectors. Smaller particles scatter light at greater angles than larger ones and the resulting diffraction pattern is used to measure particle size. By applying the rigorous Mie light scattering theory to the data, the complete particle size distribution can be ascertained. Laser diffraction instruments are based on the principle of single scattering - radiation scattered by a particle goes directly to the detector without encountering another particle. This phenomenon holds true only at dilute particulate loadings. In order to simplify the instrument/process interface, one of the challenges for on-line analysis using light scattering is to accommodate the multiple scattering that takes place at the high particle concentrations encountered in a process line and which are especially significant during plant start-up, shutdown and process changes. Malvern Instruments" development and implementation of patented multiple scattering algorithms ensures the accuracy of particle size measurement regardless of instantaneous process loading.
Hallmarks of a good PAT device
To meet the various criteria for acceptability, an analytical system must possess a number of key attributes that enable its use in process understanding and control. It is required to: - Be an industrially accepted technique based on an established standard - Need no calibration (laser diffraction is a good example of this type of absolute technique) - Demonstrate stable, robust operation - Show negligible drift - Provide real-time information - Meet cGAMP and other regulatory requirements