Plantwide Dynamic Simulation and Model-Based Control of a Continuous Pharmaceutical Process

ABSTRACT

Continuous Pharmaceutical Manufacturing (CPM) constitutes a groundbreaking paradigm in modern pharmacopeia: multinational pharmaceutical industries face unprecedented profitability gaps, due to decreasing drug prices, increasing Research & Development (R&D) and operating expenditures. These business challenges emerge in the era of globalization  and can be attributed predominantly (but not exclusively) to the transformation of patent legislation, the vast reduction of patent life, the harmonization of regulatory specifications and the advent of generics; new plants must rely on optimization, standardization and modularization ideas (Behr et al., 2004).

Batch pharmaceutical processes are widespread, forming the overwhelming majority of installed manufacturing plants; nevertheless, most are based on suboptimal recipes focused on conforming with the strict licensing requirements for drug product purity, stability and efficacy (ICH, 2009). Continuous pharmaceutical processes, though, need rely on systematic (not empirical) process understanding and R&D, therefore necessitating the employment of process modeling towards the articulation of Design Space (Yu, 2008) and the use of Process Analytical Technology (PAT) (Munson, 2006); this may turn out a time-consuming endeavor for novel manufacturing lines. The U.S. Food & Drug Administration (FDA) thus spearheads the pursuit of CPM initiatives, recognizing their potential to reduce operating cost at high quality (FDA, 2008) as confirmed in comparative technical (Betz et al., 2003) and modeling (Gerogiorgis & Barton, 2009) studies. 

Modern batch manufacturing processes are frequently operated at maximum unit efficiency via dynamic optimization models, relying on mathematical rigor to capture critical phenomena,  empirically or mechanistically (Chekhova et al., 2000). However, process scale-up hinders model adaptation. Optimal external input profiles can be computed and applied during each batch processed (low-level regulatory control), and statistical methods for offline quality assessment (QA) can reveal deviation trends (high-level supervisory control). Product quality and uniformity are evaluated a posteriori: often, drug substance (DS) or drug product (DS) lots must be discarded due to their poor Critical Quality Attributes (CQA). The characteristic batch plant time for a given lot size is the sum of processing and idle times required for its production.

Continuous manufacturing lines, in contrast to batch tank farms, eliminate idle times between processing units by definition; however, deeper understanding is required for their operation. Plantwide dynamic model building requires significant effort  and parameter estimation implies arduous, costly experiments which are not a likely luxury if rapid launching is of the essence. The characteristic timescale for a novel CPM plant of given capacity is harder to define, since it depends on structural complexity, unit interdependence, recycles, rate data and timescale spectra. Benign (dilute) or challenging (dense) flow chemistries and multicomponent mixtures affect reaction and separation times, as pointed out in a CPM miniplant study (Roberge et al., 2008). Online sensing and manipulation is possible via modern PAT sensors and actuators, but empirical models can neither explain nor predict how local disturbances may affect plantwide quality performance (Ward et al., 2010), as fundamental phenomena are often poorly understood.  CPM plantwide dynamics, therefore, remain a hitherto elusive (hence challenging) objective.

The present study focuses on plantwide dynamics and control of a novel upstream (DS) process which is aimed at the continuous production of a novel Active Pharmaceutical Ingredient (API). The CPM flowsheet comprises 3 organic synthesis reaction and 4 separation steps; individual dynamic models for all respective unit operations are proposed on the basis of experimental data. Open-loop plantwide process simulation in MATLAB (Simulink) allows the determination of  timescales, illustrating how common feed disturbances propagate to CQA (e.g. product quality). Control loop pairings can be identified by implementing a RGA method (Skogestad, 2005) and closed-loop plantwide process simulations provide insight into their efficiency and robustness.

REFERENCES

Behr, A., Brehme, V.A., Ewers, C.L., Grön, H., Kimmel, T., Küppers, S., Symietz, I., New developments in chemical engineering for the production of drug substances. Eng. Life Sci. 4(1): 15-24 (2004).

Betz, G., Junker-Purgin, P., Leuenberger, H., Batch and continuous processing in the production of pharmaceutical granules. Pharm. Dev. Technol. 8(3): 289-297 (2003).

Chekhova, E., Barton, P.I., Gorak, A., Optimal operation processes of discrete-continuous biochemical processes. Comput. Chem. Eng. 24(2-7): 1167-1173 (2000).

European Medicines Agency (EMEA), Impurities: Residual solvents, VICH Topic GL18 (plus annexes), http://www.emea.europa.eu/pdfs/vet/vich/050299en.pdf (2000).

Gerogiorgis, D.I., Barton, P.I., Steady state optimization of a continuous pharmaceutical process. Computer-Aided Chemical Engineering (Proceedings of PSE 2009) 27(1): 927-932, Elsevier, Amsterdam (2009).

International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals (ICH), Guidance Documents Q8 (Pharmaceutical Development), Q9 (Quality Risk Management), Q10 (Pharmaceutical Quality System), http://www.ich.org (2010).

Munson, J., Stanfield, C.F., Gujral, B., A review of process analytical technology (PAT) in the US pharmaceutical industry. Curr. Pharm. Anal. 2(4): 405-414 (2006).

Roberge, D.M., Zimmermann, B., Rainone, F., Gottsponer, M., Eyholzer, M., Kockmann, N., Microreactor technology and continuous processes in the fine chemical and pharmaceutical industry: Is the revolution underway?. Org. Proc. Res. Dev. 12(5): 905-910 (2008).

Skogestad, S., Postlethwaite I., Multivariable feedback control: Analysis & design. Wiley, Chichester (2005).

U.S. Food and Drug Administration (FDA-CDER), Innovation & continuous improvement in pharmaceutical manufacturing, www.fda.gov/cder/gmp/gmp2004/manufSciWP.pdf (2008).

Ward, J.D., Yu,C.-C., Doherty, M.F., Plantwide dynamics and control of processes with crystallization, Comput. Chem. Eng. 34(1): 112-121 (2010).

Yu, L.X., Pharmaceutical Quality by Design: Product and process development, understanding, and control. Pharm. Res. 25(4): 781-791 (2008).