Overcoming Obstacles in Process Analytical Technology

Process analytical technology (PAT) is becoming more widely used in solid-dosage drug manufacturing. PAT’s use is primarily in batch manufacturing, which still makes up the majority of processes, but also in continuous manufacturing. On-line near infrared (NIR) spectroscopy, for example, is being used to measure critical quality attributes (CQAs) such as blend uniformity and moisture content of tablets.

Solving PAT limitations

With increased use, however, some of the limitations of current technology are becoming evident. “The industry is starting to take a serious look at where current PAT is reaching its limit and what the alternatives are,” notes Doug Hausner, Associate Director for Industrial Relations and Business Development at the Engineering Research Center for Structured Organic Particulate Systems (C-SOPS), an academic–industry consortium based at Rutgers University. One of the challenges is that using PAT commercially, over a long time period, requires a significant amount of personnel hours for maintenance-checking equipment and keeping calibrations up to date, notes Hausner.

One reason why existing calibration curves are a burden to maintain is that they lack robustness, says Hausner. “A solution is to build better calibration curves using more datapoints and using design of experiments to identify the edges of failure,” he explains. Hausner notes that continuous manufacturing allows process engineers to run experiments more quickly, with less material compared to batch, so that collecting more data is not as costly. Another benefit of continuous manufacturing is that, unlike batch processes where laboratory equipment is smaller than commercial equipment, the continuous manufacturing process is already at scale, so the data collected in development experiments are relevant to commercial manufacturing conditions.

Another solution seeks to use “soft” (i.e., software or virtual) sensors, which use existing, direct process measurements (e.g., temperature, pressure, feed rate) to calculate the desired CQA or critical process parameter, instead of using on-line spectroscopic PAT. Measurements from these sensors do not need complex multivariate models that require developing, validating, and maintaining, explains Jim Holman, process development manager, APC Pharma Solids, at GEA UK. “The methods are robust to potential variation in incoming raw materials and scalable over throughput ranges.” 

Another problem is that some PAT techniques may be too slow for some processes. For potent drugs, for example, the amount of API in the blend is typically low, so there can be a lot of “noise” in spectroscopy and a longer sample time is required. Light-induced fluorescence (LIF) spectroscopy is one of the solutions being considered to evaluate API concentrations that are too low to be measured with NIR or Raman spectroscopy, although it also has limitations.

PAT enables real-time release and closed-loop control

Real-time release (RTR) is often discussed as a goal for more efficient manufacturing that applies to both batch and CM processes. Regulators have not yet signalled acceptance of soft sensors for product release, and spectroscopic PAT would likely still be needed, notes Hausner.  RTR for drug tablets, for example, would need PAT to confirm the chemical identity of the API at the correct concentration in finished tablets exiting the tablet press, as well as to predict hardness and weight or thickness.

Communication between PAT sensors and systems enables visibility of information about a reaction in real time and can eventually allow movement to automated, closed-loop control of a small or large-molecule reaction using PAT, says Ernie Hillier, principal systems manager at Waters Corporation. Hillier will discuss the lessons learned from ongoing industry collaborations in sampling and sensing in a presentation at the IFPAC conference (Feb. 28, 2017).  Waters’ PATROL ultraperformance liquid chromatography (UPLC) process analysis system allows nearly real-time chemical analysis so that scientists can know in greater detail what is happening in a reactor at a given point of time. Data from UPLC is used in R&D, process development, and manufacturing, notes Hillier.

FOYA category winner highlights PAT use

Eli Lilly and Company’s continuous direct compression process for oral solid dosage drug manufacturing won the International Society for Pharmaceutical Engineering (ISPE) Facility of the Year Awards category award for Process Innovation, in part, for a control scheme anchored by advanced automation and PAT (1). The process uses soft sensors to estimate CQAs, and NIR integrated with a real-time residence distribution time model to allow real-time release. Spectroscopic PAT combined with automation logic can be used to reject portions of the batch suspected as being out of limits, explained the company (1).

PAT advances at IFPAC

The upcoming IFPAC conference (Feb. 27–Mar. 2, 2017, Washington DC) is dedicated to advances in PAT, quality by design, and process control for the pharmaceutical, biotechnology, and other industries. Topics will include RTR and soft sensors in continuous manufacturing and advances in PAT technologies such as LIF.

Reference

• ISPE, “Category Winner for Process Innovation,” www.facilityoftheyear.org/winners/2017-process-innovation, accessed Feb. 6, 2017.