(4ic) Systems Engineering for Manufacturing of Advanced Biotherapeutics

My objective is to become a leader in the development and application of mathematical tools to improve the design, optimization, and control of manufacturing systems for advanced biotherapeutics, such as cell therapies and genetic medicines.

Rapid advancements in the fields of cell therapies and genetic medicines have led to the recent approval of treatments for diseases that were previously incurable, such as certain types of cancer, blindness, and muscular dystrophy­. However, most advanced biotherapeutics are produced through manufacturing processes that present low efficiency and high costs. With over 1000 ongoing clinical trials for cell and gene therapies (clinicaltrials.gov), the current manufacturing yields will soon be insufficient to provide approved treatments to all patients (Destro et al., 2024a; Elverum and Whitman, 2020). At the same time, the low manufacturing efficiency contributes to the high price of advanced therapeutics (exceeding $3 million per dose for certain gene therapies). To deliver cutting-edge biotherapies to patients, it is necessary to dramatically increase the efficiency of current biomanufacturing systems.

Process systems engineering approaches have successfully been used to optimize the design and the operation of several manufacturing sectors. During my PhD, I worked in research groups leading the application of process systems engineering to the enhancement of pharmaceutical manufacturing (see, for instance, Destro et al., 2022; Destro and Barolo, 2022). For my PhD research's contribution to advancing pharmaceutical manufacturing, I received the 2024 EFCE Excellence Award in Recognition of an Outstanding PhD Thesis on Computer Aided Process Engineering. During my postdoctoral research at MIT, I used my experience in traditional (i.e., small-molecule) pharmaceutical manufacturing to pioneer the enhancement of gene therapy manufacturing through process systems engineering. Building on my experience, my independent group will create and apply a computational framework to enhance the manufacturing of innovative biotherapeutics, with a focus on lentiviral vectors and chimeric antigen receptor (CAR) T-cells.

Future research directions will encompass applying the same paradigm and the developed computational tools to optimize the manufacturing processes of next-generation cell therapies, genetic medicines, and vaccines.

Compared to other manufacturing systems, ­biomanufacturing processes pose unique challenges, such as high intrinsic stochasticity, limited data availability, typically high measurement error, and complex system dynamics. Consequently, the design, optimization, and control of biomanufacturing systems require innovative approaches. Drawing from my expertise in mechanistic, data-driven and hybrid modeling (Destro et al., 2020; 2021), I plan to develop theory and algorithms for addressing these challenges.

My goal is to deliver computational tools for automating model development and deployment in biomanufacturing. Within this context, a core objective will involve developing novel techniques tailored for biomanufacturing systems, focusing on key areas such as model-based design of experiments, model selection, and design space determination. The algorithms will be rooted in advanced Bayesian approaches, to determine the optimal tradeoff between exploration and exploitation when developing and using a model for the design and the optimization of a biomanufacturing process. Algorithms will be developed to: (i) obtain the best model of a process for a given purpose, in terms of accuracy, efficacy, and parsimony, and (ii) design experiments enabling a synergistic advancement in both model development and process optimization, at the same time. A specific objective will be the development of modeling pipelines that leverage recent advancements in sensor technology for biomanufacturing processes. This research area will build upon my current work in developing machine learning algorithms that optimize biomanufacturing systems using advanced tensorial data, such as single-cell biophysical signatures and measurements from imaging flow cytometry (Destro et al., 2024b).

Novel cell therapies and genetic medicines have recently received approval by regulators for, among other applications, immunotherapies and treatment of genetic disorders. Cell therapies involve the use of living cells as therapeutic agents, while genetic medicines harness the potential of nucleic acids to prevent or treat diseases. The most prominent categories of genetic medicines are gene therapy, gene editing, and mRNA therapeutics. Several scale-up and manufacturing challenges have to be solved for making these advanced therapeutics available to all patients in need.

Certain cell therapies and genetic medicines have reached an advanced stage of development, enabling process engineering studies that can enhance manufacturability and optimize scale-up and cost-effectiveness. In my postdoctoral work (Destro et al., 2023; Destro and Braatz, 2024; Destro et al., 2024a, Destro et al., 2024b), I have used model-based techniques for improving the efficiency of a manufacturing process for recombinant adeno-associated virus (rAAV), a popular vector for in vivo gene therapy. Building on this expertise, my lab’s research will optimize the biomanufacturing process of (i) CAR T-cells and (ii) lentiviral vectors. CAR T-cells are the therapeutic agents in the first FDA-approved gene-modified cell therapy, an immunotherapy for relapsing B-cell acute lymphoblastic leukemia. Lentiviral vectors are the commonly used vectors in gene therapy [1], and play a significant role in CAR T-cell manufacturing, contributing towards their high cost.

My lab will develop mathematical models to drive the design, optimization, and control of the key unit operations for CAR T-cells and lentivirus manufacturing. Multiscale models, based on mechanistic, machine learning, and hybrid approaches, will connect the performance of unit operations with the dynamics of small-scale phenomena, such as the intracellular pathways. The models will play a crucial role in optimizing the process by suggesting improved designs, as well as molecular biology enhancements, for the biomanufacturing process. The novel computational tools developed within the research group, together with well-established techniques from the process systems engineering expertise, will provide a powerful framework for addressing the current and future challenges in advanced biotherapeutics manufacturing.

In spring of 2024, I was an instructor for MIT's 10.551 Systems Engineering graduate class, where I gave lectures on systems modeling, optimization, process control, and data analytics to chemical engineering graduate students.

Building on this experience, along with my previous roles as a teaching assistant during my PhD, I am interested in teaching and developing instructional materials in two key areas for my future positions: (i) foundational chemical engineering courses, such as transport phenomena, and (ii) courses that apply models to solve engineering tasks, such as process control. In the foundational courses, students will develop a solid understanding of the underlying principles that govern physical and chemical processes, and how to translate such principles into mathematical models. In the application-oriented courses, such as process control, I aim to emphasize the application of modeling for solving engineering problems. Students often lack the skill of developing models at the rapid pace dictated by an industry environment. However, simple models can be highly effective for a wide range of tasks in process design and control. To bridge this gap, I will create teaching resources to aid students in acquiring proficiency in model-based computing, enabling them to exploit the full potential of digital tools in their future careers. In addition, I intend to design a process systems engineering course. This application-oriented course will provide several examples highlighting how models can be swiftly developed and utilized across various applications. The course will also cover data-driven and hybrid modeling, which are often given much less importance than mechanistic modeling in chemical engineering programs.

Selected publications

Destro, F. and Braatz, R. D. (2024). Efficient simulation of viral transduction and propagation for biomanufacturing. Submitted.

Destro, F., Wu, W., Srinivasan, P., Joseph, J., Bal, V., Neufeld, C., Wolfrum, J.M., Manalis, S.R., Sinskey, A.J. Springs, S.L., Barone, P.W., and Braatz, R.D (2024a). The state of technological advancement to address challenges in the manufacture of rAAV gene therapies. Submitted.

Destro, F., Wu, W., Joseph, J., Wolfrum, J.M., Sinskey, A.J. Springs, S.L., Barone, P.W., Manalis, S.R., and Braatz, R.D (2024b). Real-time prediction of recombinant adeno-associated virus production in suspended cultures using single-cell measurements and machine learning. In preparation.

Destro, F., Joseph, J., Srinivasan, P., Kanter, J.M., Neufeld, C., Wolfrum, J.M., Barone, P.W., Springs, S.L., Sinskey, A.J., Cecchini, S., Kotin, R.M. and Braatz, R.D (2023). Mechanistic modeling explains the production dynamics of recombinant adeno-associated virus with the baculovirus expression vector system. Mol. Ther. Methods Clin. Dev. 30, 122-146.

Destro, F. and Barolo, M. (2022). A review on the modernization of pharmaceutical development and manufacturing – Trends, perspectives, and the role of mathematical modeling. Int. J. Pharm. 620, 121715.

Destro, F., Nagy, Z.K. and Barolo, M. (2022). A benchmark simulator for quality-by-design and quality-by-control studies in continuous pharmaceutical manufacturing ‒ Intensified filtration-drying of crystallization slurries. Comput. Chem. Eng. 163, 107809.

Destro, F., Hur, I., Wang, V., Abdi, M., Feng, X., Wood, E., Coleman, S., Firth, P., Barton, A., Barolo, M. and Nagy, Z.K. (2021). Mathematical modeling and digital design of an intensified filtration-washing-drying unit for pharmaceutical continuous manufacturing. Chem. Eng. Sci. 244, 116803.

Destro, F., Facco, P., Munoz, S.G., Bezzo, F. and Barolo, M. (2020). A hybrid framework for process monitoring: Enhancing data-driven methodologies with state and parameter estimation. J. Process Control 92, 333-351.

Elverum, K. and Whitman, M. (2020). Delivering cellular and gene therapies to patients: solutions for realizing the potential of the next generation of medicine. Gene Ther. 27, 537–544.

[1] alongside rAAVs and adenoviruses