Outsourcing-Pharma (OSP) recently discussed continuous manufacturing (CM) trends with two DFE Pharma leaders: Brian Carlin, director of QbD/regulatory; and Mara van Haandel, innovation manager.
OSP: For readers that might not be well versed, could you please explain what the key advantages are of opting for a continuous-manufacturing process over batch manufacturing?
DFE: There are several key advantages of using CM as opposed to traditional batch manufacturing:
- Output can be tailored to demand by changing the run time, whereas the specific quantity associated with traditional batch manufacturing is fixed.
- There is no scale-up and the same continuous line can be used to produce development and production batches.
- Output can be further increased by increasing run rate, installing multiple identical lines (scale-out), or true scale scale-up by installing a larger continuous processor.
- CM is material sparing compared to full scale batch manufacturing which is important in early stage development when API quantities may be limited.
- CM under automated process control can explore a greater number of formula and process variants, allowing a more comprehensive DOE where the data is not subject to the confounding effect of scale-up.
- CM and the associated Process Analytical Technologies (PAT) facilitate Continuous Process Validation (CPV), making it easier to demonstrate a state of control and reducing the need for regulatory oversight.
- Smaller facilities footprint, possibly modular and portable
- Reduced segregation potential for powders, extending the scope of direct compression at low drug loadings.
- Continuous API synthesis allows use of chemistries too hazardous to operate in batches, affording higher yields and purities.
OSP: How has the adoption of CM grown/evolved in the pharma industry over the years?
DFE: Pharma CM has evolved slowly over the last 25 years culminating in the approval of Orkambi (Vertex) in 2015. There are now seven approved CM products on the market. The Pharmaceutical industry is notoriously risk averse and conservative, and CM adoption has been slow for several reasons:
- Lack of consensus on benefits of CM.
- Perception that CM is only for high volume production.
- Differing requirements across regulatory jurisdictions globally.
- Low efficiency operating CM intermittently for short run times. Cleaning and reassembly can take longer than the run time.
- Uncertainty of financial and regulatory benefits.
The arrival of CM in 5-7 years has been heralded for over 20 years but a more realistic time scale was given by Janet Woodcock (FDA) at the 2011 AAPS Annual meeting: “It is predicted that manufacturing will change in the next 25 years as current manufacturing practices are abandoned in favor of cleaner, flexible, more efficient continuous manufacturing.”
The rate of adoption is expected to increase with greater technical and regulatory familiarity with CM and the greater availability of integrated equipment lines designed for CM.
OSP: What are some of the challenges and key considerations involved in CM?
DFE: They include:
- Avoiding batch mentality. Definition of a CM batch size is still an issue after 20 years and some regulatory jurisdictions have batch size incorporated in their legal requirements
- Lower number of ingredients. In batch manufacturing there is no limit but in CM the number of feeders is often only five or six.
- Understanding residence time distributions is essential for traceability of ingredients and how transient disturbances propagate through the process. If a batch process takes longer than the residence time of a continuous processor CM may not be feasible.
OSP: Similarly, how are the aspects of CM different for pharma environments—what factors do professionals in this industry need to consider that people in other industries might not have to?
DFE: The pharma industry is compliance driven, which is not always congruent with product quality. Coupled with a lack of fundamental understanding of product performance the environment is inhibitory to innovation and product/process improvement, including switching to CM. In other industries scaling up in batch mode and then switching to CM maximizes efficiency.
In contrast, pharma companies are forced to choose between starting continuous (inefficient if intermittent/short run time but reaping the benefit of avoiding scale-up) versus scaling up in batch mode then switching to CM, running the risk of failing to demonstrate bioequivalence.
OSP: How can pharma professionals properly weigh the impact of excipients and related variables on CM, and vice versa?
DFE: Pharma professionals cannot unilaterally weigh the full impact of excipient variability. It is advisable to discuss your CM application with your excipient suppliers as a potential critical material attribute (CMA) may not be included in the pharmacopoeial or supplier specification. General purpose excipients may be designed for general applications but not for a specific product.
Many organizations will assess the impact of excipient variability by testing multiple batches of excipients. However, this is a time-consuming, labor-intensive process that is not always executed diligently.
Instead, DFE Pharma advocates the use of multivariate data analysis (MVA) statistical techniques, which allow the simultaneous analyses of multiple variables to investigate patterns or clusters in the excipient dataset.
Our researchers work with our pharma partners to identify the major relevant sources of variation in the large, complex datasets, as well as noise variables that have no effect.
Batches near the Hotelling’s T2 95% confidence limit are then selected as a representative of a wide range of possible variation. Termed ‘stretch batches’, these application-specific batches of excipients, based on DFE Pharma historical data, are truly representative of the possible variation, while remaining well within the specification limits of the CoAs.
The stretch batch approach uses large data sets to enable MVA to identify major sources of variation, and facilitate a greater understanding of the relationships between all design inputs, from the desired quality attributes to process parameters and raw material attributes.
Insights gleaned from these stretch batches contribute to quality by design (QbD) by informing risk assessments and focusing the design of experiments on the most important parameters.
By effectively managing the variations that can impact on product performance, developers can slash the level of experimentation needed during product development, especially if CM/PAT facilitates automated DOE.
OSP: Do you have any resources put out by your company that might be useful for people looking to gain a greater understanding of CM?
DFE: Yes:
- DFE Pharma has an R&D team dedicated to CM, developing applications data on understanding how excipients from a mechanistic point of view interact with CM components such as feeders, blenders, twin-screw extruders and roller compactors.
- Understanding excipient variability is integral to the CM process but it is extremely complex. The parameters that dictate control vary from product to product, from grade to grade and even from batch to batch.
- DFE Pharma can supply stretch batches of excipients, which are those batches at or beyond the 95% confidence limits on a principal component analysis. Using such batches in their DOE allows the CM formulator to rapidly assess the impact of excipient variability on their CM process and product to ensure product robustness.
- DFE Pharma can provide expert advice on specification of excipients, including CMAs.
OSP: What else would you like to tell us that we didn’t touch upon above?
DFE: The complexities of incorporating the impact of excipient variabilities into pharmaceutical product design are illustrated by the recent publication from the IPEC Federation of their guide on “Incorporation of Pharmaceutical Excipients into Product Development using Quality-by-Design (QbD).”
Complexity is associated with too many degrees of freedom so another reason for using CM is the fewer degrees of freedom. The degrees of freedom are reduced for a small working volume (relative to traditional batch sizes) and the elimination of scale-up. Fewer degrees of freedom reduces the risk of product failure.