Holistic Design of Experiments Using an Integrated Process Model | JoVE Visualize

Area of Science:

Biopharmaceutical manufacturing
Process engineering
Statistical experimental design

Background:

Current biopharmaceutical process characterization relies on statistical experimental designs (e.g., factorial, optimal, definitive screening) that analyze unit operations in isolation.
These methods do not account for the interconnected nature of multi-step manufacturing processes, limiting holistic understanding and optimization.

Purpose of the Study:

To develop a novel experimental design methodology, termed holistic Design of Experiments (hDoE), for biopharmaceutical process characterization.
To enhance information gain regarding final product quality by strategically placing experimental runs within integrated process models.
To minimize the simulated out-of-specification rate of a multi-step process model.

Main Methods:

Developed an integrated process model comprising a chain of regression models linking process parameters to critical quality attributes for each unit operation.
Connected unit operation models by simulating the transfer of response as a load parameter between sequential steps, mimicking real-world manufacturing.
Benchmarked the proposed hDoE method against standard process characterization techniques using in silico simulation studies with diverse ground truth processes.

Main Results:

The hDoE approach demonstrated a reduction of over 50% in the number of required experiments compared to standard methods, even in simpler scenarios.
The method effectively minimized the simulated out-of-specification rate, indicating improved control over final product quality.
Validation across various simulated process scenarios confirmed the robustness and broad applicability of the hDoE strategy.

Conclusions:

The holistic Design of Experiments (hDoE) offers a more efficient and effective approach to biopharmaceutical process characterization.
By integrating unit operations, hDoE enables better prediction and control of final product quality.
This method supports the primary objectives of process development, validation, and manufacturing for consistent product quality delivery.