✕
☰

Digitalisation in HPAPI development & manufacturing – hype or valuable toolbox?

by ,

How a growing digital toolbox is enabling pharmaceutical developers to advance high potent API production.

Authors: Niklaus Künzle, Head, Process Technology & Innovation & Conrad Roten, Group Leader, API Development Services Lonza AG Visp, Switzerland.

Highly potent API (HPAPI) are playing a growing role in the small molecule pipeline due to their usefulness in treating cancer and other indications. As these compounds have a physical and clinical effect at very low dose, they pose potential occupational health risks. Personnel in development and production facilities must be protected at all times from the products they manufacture. The sustainable production of highly potent APIs (HPAPIs) therefore requires specific precautions for operator health and safety on the one hand and for product quality on the other (Fig. 1).

However, while protecting workers is a critically important element of HPAPI production, biopharma innovators need access to many other specialised capabilities to bring their HPAPI innovations from concept to commercialisation. As HPAPI molecules grow in prevalence, the pharma & biotech industry is seeking new ways for efficient and safe development and production methods to meet rising demand [1,2]. A rapidly growing toolbox of digital technologies can be applied in many areas of both process development and production, which may help accelerate improvements in HPAPI production.

Digital Technologies in HPAPI Development & Production

In the process of making a drug substance (DS), a chemical process generally leads from different starting materials, through several intermediates, to the API as the end product. The intermediates involved are not necessarily as potent as the API. The starting materials are often not highly potent and only become highly potent once the API core structure is formed. On the other hand, there are examples where highly potent materials are converted to non-potent materials, or where potency varies largely throughout the chemical pathway. The required efforts also depend largely on whether or not a certain intermediate must be isolated in its pure form. As a result, not all steps require the same type of measures, which means it is possible to execute non-HPAPI steps in a non-HPAPI facility. 

A range of technological tools can help approach the wide variety of HPAPI process development needs. Digital technologies that can help safely and effectively advance HPAPI production at various points in the development process include:

Digital modelling and simulation

Modelling and simulation of toxicological, chemical and physical properties of compounds are helpful tools to support development and manufacturing activities, especially for early phase projects with very limited experimental data. When such information is available before extensive testing and measuring can be done, technicians have greater understanding right from the start, which may enable a more efficient process execution within an appropriate setting. Handling of highly potent substances can therefore be minimised, or the required material consumption reduced, since availability of highly potent compounds is often limited at this phase anyway. 

Modeling and simulation reduce the amount of lab experiments but do not make them obsolete. Therefore, it is necessary to get as much information as possible from the remaining lab tests, especially during late phase development. Sensors and online spectroscopy with direct links to data management and analysis systems (e.g., OSI-PI) and online modeling tools (e.g., CAMO ProcessPulse) can help do so. This method can provide data acquisition and analysis with a short sampling rate even for overnight experiments or experiments carried out in containment labs with restricted access (typical for HPAPI applications).

Process Analytical Technology (PAT)

The application of advanced Process Analytical Technology (PAT) is an important development tool, allowing development projects to be assessed rapidly at certain clinical phase milestones. PAT relies on tools like multivariate data acquisition and data analysis tools, which aid in design of experiments, collection of raw data and statistical analysis of this data in order to determine critical process parameters as well as other continuous improvement and knowledge management tools. When it comes to HPAPI processes and related process development activities, PAT tools belonging to the process analytical chemistry category are of additional value, as they enable data acquisition and increased process understanding with the potential of very limited exposure of assets and personnel. This enables cost and time reduction in dealing with offline samples analysed in dedicated HPAPI analytical assets. Interesting examples of application are scale-down models evaluating process parameters to control particle size and crystal morphology, or measurements which enable optimised IPC time and sampling measurement later during manufacturing.

Besides basic univariate technologies, PAT may rely on multivariate spectroscopic measurements such as NIR, MIR, RAMAN, UV/VIS and MS, which allow formulators to seamlessly follow the reaction progress in reaction vessels. In combination with additional process, quality and compliance data, these tools are the basis for real-time batch release. As strong progress in the reliability and compliance of such techniques is observed, many developers believe PAT measurements for process control will have a big impact in highly potent production, especially in late and commercial phase manufacturing. This could also lead to reduced investment costs for containment setup in plant areas where analytical assessment or sampling has to take place. (Fig. 2)

Automated visual assessment and documentation

Visual assessment and documentation of process mixture and unit operation is an important aspect of familiarisation/transfer activities for chemical processes. Such behavior can well be captured via cameras taking pictures of the process at various stages or videos of ongoing processing. This approach can reduce or even remove the need for scientists, process givers or production personnel to observe the scale down model and transfer of a process to a high-containment lab, which requires training and gowning to enter. For example, slow filtration of a resin could be monitored efficiently via camera from an easily accessible area outside the high-containment lab or from the normal lab. Attention must be paid to ensure workers’ privacy rights are protected by concentrating the recording on the actual processing devices.

Virtual and augmented reality

Even though HPAPIs are often produced in modern and automated facilities, some production steps (especially in clinical phase manufacturing) are still performed manually and require highly skilled and trained operators. For this reason, trainings based on virtual and augmented reality (VR, AR) can help users learn realistic handling, cleaning and movement techniques. 

Machine learning and data analysis

One of the big challenges of process digitalisation is the alignment of various new data sources such as data from production equipment, analytical instruments, planning and QA systems and there is still a big amount of dedicated programming and validation effort necessary to connect systems. To get information out of huge amounts of data, there is a broad technology platform where visualisation, multivariate data analysis and machine learning algorithms are used to get real-time information about running processes. This understanding can be used to improve planning, predict and prevent deviations, to optimise processes and to support problem solving. Machine learning algorithms can also be applied to make accurate predictions or increase the accuracy of conventional methods in order to speed up development time. [3]

Conclusion

As HPAPI products become more common and drug developers seek new ways to bring products to patients, digital tools can help. Process digitalisation should be considered wherever possible in the production of highly potent APIs, although this seems less logical for multi-purpose facilities running campaigns with limited number of batches. Manufacturers may explore several levels of automation, such as MES, in-line process analytical technologies (PAT) or real-time release-testing (RTRT). Avoiding paper-based documentation in high-containment facilities seems to be a final step in containing the HPAPI to where it is desired. 

The key for biopharma innovators is to find the right balance between flexibility and automated robustness, when implementing digitalisation in a multi-purpose facility. The digital toolbox is a valuable addition to HPAPI development and manufacturing that is expected to grow in importance in the future.

References 

[1] Walker, N. “HPAPI Market Trends” (2017). Available at: https://www.contractpharma.com/issues/2018-09-01/view_features/hpapi-market-trends/. Accessed July 3, 2020.

[2] Janssen, M. “Containment strategy for highly potent API manufacturing.” Available at https://www.cleanroomtechnology.com/news/article_page/Containment_strategy_for_highly_potent_API_manufacturing/149845. Accessed July 3, 2020.

[3] Raies, Arwa B., and Vladimir B. Bajic. "In silico toxicology: computational methods for the prediction of chemical toxicity." Wiley Interdisciplinary Reviews: Computational Molecular Science 6.2 (2016): 147-172.

Back to topbutton