How many facilities? Centralized vs. decentralized manufacture

Author: Natasha Davie, 03/05/13

When it comes to the cell therapy commercialization crunch there are some fundamental questions that need to be answered. Is your product going to be frozen or fresh? How will the product be administered? What is your dosing strategy? And the one that I’m going to focus on in this blog: Will your manufacturing strategy be centralized or decentralized?

Having a centralized manufacturing strategy means having one principal facility from which your product is transported and distributed. It makes ensuring consistent quality easier as you only need to validate one process, and removes the need for complicated technology transfer to additional sites. However, this system only works in conjunction with a heck of a cold storage chain, which is pricey in itself. Furthermore, products manufactured in this way are often frozen prior to transportation, and as a result tend to require some sort of end point manipulation by the administrator – an additional source of variability.

A decentralized strategy consists of multiple satellite sites covering a large area, enabling manufacturing to occur closer to the final point of administration. Decentralized strategies minimize the length of the distribution chain and, as such, are invaluable for time-critical products such as fresh autologous therapies. However, decentralized facilities generally require a greater up front capital investment to set up, in addition to the substantial challenge of demonstrating product and process consistency across multiple manufacturing sites.

So, which is better suited to regenerative medicine?

As the industry grows and our understanding of products improves, we are beginning to learn our manufacturing limits. For the centralized argument, one of the issues raised by the UK-based VALUE consortium (you can find the final report ‘Regenerative Medicine: Navigating the Uncertaintieshere) was that ‘establishing and maintaining comparability between sites is likely to become impossibly burdensome beyond two or three sites’.

you want to ensure that what is administered in Tokyo is the same as what’s administered in Toronto.”

This is because each product has defined characteristics, whether it’s the level of expression of a cell surface marker or a cytokine profile, and these characteristics are expected to be near identical for every product and every batch. This is particularly true for allogeneic products, where all products are derived from the same donor cells and therefore should be able to demonstrate the same key features.

Due to the inherent sensitivity of cells, when a new facility is brought online the manufacturing environment needs to be virtually identical to the original in order to generate a comparable product. Sophisticated technology transfer initiatives exist to try to overcome this, but if for some reason the same product made in two different facilities is not ‘substantially similar’, as defined by the FDA, then there is little that can be done. You simply can’t sell two different products as the same thing. And as you can imagine, the complexity of this operation increases exponentially with the number of facilities. This notion is supported, perhaps unintentionally, by Dendreon’s recent decision to sell one of its three Provenge manufacturing facilities to Novartis.

With a single manufacturing facility, you can focus on making sure that a product is exactly within your specifications. The time and money saved by not overseeing several facilities can be invested in making minor process tweaks (within reason) to optimize production and improve product quality. This investment is well worth making, especially for allogeneic products, where you want to ensure that what is administered in Tokyo is the same as what’s administered in Toronto.

But don’t rule out decentralized manufacturing just yet. The beauty of manufacturing regenerative products is that so many of them are made using single-use technologies. This means that new facilities can be assembled at a fraction of the time and cost that it takes for conventional pharmaceuticals (see more information here). Decentralized manufacture of sorts has been successfully used by the blood and bone marrow transplant industries for years. The fewer the number of patients processed at one site, the lower the risk of confusion or cross contamination. In addition, the variability between sites is less of an issue for autologous therapeutics, where there is a greater tolerance for product heterogeneity, particularly if there is a favourable trade-off such as removing the need to add a toxic cryopreservative.

What works better for regenerative medicine? The pros and cons of both strategies need to be weighed to determine which is more appropriate for a particular product. The trend is that autologous therapies tend to suit decentralized manufacturing, whereas allogeneic therapies tend to suit a centralized facility, but this is by no means gospel. In the end, the answer, like so many questions in this field, is a teasing ‘it depends’.

What’s your opinion? Have I missed any arguments that would steer one towards centralized over decentralized or vice versa?

 

The following two tabs change content below.

Natasha Davie

Natasha Davie is part of the Centre for Accelerating Medical Innovations at Oxford University, where she is pursuing a doctorate in Clinical Laboratory Sciences. She has been involved in regenerative medicine since 2002, when she worked with the London Regenerative Medicine Network on numerous projects analysing cell therapy translation, and gaining expertise in clinical trials, regulation, manufacture and commercialization. She completed her Masters in Biochemical Engineering at University College London in conjunction with the Harvard Stem Cell Institute and Harvard Medical School. Follow Natasha on Twitter @natashadavie
Tags: , ,

One Response

Leave a Reply

*
*