August 2008

System Integration

Extreme flexibility

Targeted medicines mean plants must integrate new processes to meet demand

Fast Forward

  • Regulatory changes mean higher costs.
  • Globalization forcing change in drug production.
  • Manufacturing plants face hiked complexity.
  • Manufacturers will need to become more nimble.
 
By Frede Vinther

Pharmaceutical and biotech companies over the last decade have been changing their ways of running the business, especially with regulatory compliance issues coming into play.

Companies have felt the impact from regulatory changes from the Food and Drug Administration in the U.S. and The European Medicines Agency. These regulations have not only had an impact on the cost of building and running production facilities, but also on research and development from the beginning all the way through to approved product. All of this occurred without any real increase in approvals of new molecular entities (NMEs).

The result is drug prices for consumers continue to skyrocket.

Governments have already started questioning the price for medicines, and they will push much harder for lower prices in the years to come. But pushing Pharma prices does not do it alone; governments have played a major role in pushing the medicine prices to the current level. Governments will also have to update legislation to adhere to and enable a more global approach to thinking with respect to medicine and fighting diseases to enable faster, more reliable, cheaper drugs, well functioning and very dynamic drug development and production.

One example of legislation obstacles in bringing new medicine to market is the Japanese legislation that today requires clinical trials for NMEs on native Japanese people and therefore adds another five to seven years of clinical trials on top of any U.S. or European approval before the drug can hit the Japanese market.

Basically, the whole way the public and private healthcare is organized and the way treatment of diseases, understanding diseases and development and production of medicine will have to change, for consumer prices of medicine to decline. 

The human factor

When turning the focus toward the manufacturing of drugs and the whole supply chain and its logistics, the globalizing world and its demands for medicine will need to foster a dramatic change. The era of Big Pharma searching for blockbusters and building facilities solely to manufacture these will vanish.

Instead, medicines will undergo development for disease subtypes and therefore will target different patient subpopulations. These medicines will only undergo testing for patients that suffer specific conditions, thereby reducing the number of trials and size of the trials required to prove efficacy. With proper legislation, the development process of the future will be much more refined.

Targeted treatment will drastically reduce the time to market for a specific drug, but at the same time drastically increase the number of closely related but still distinguished treatments and medicines.

The targeted treatment will have an impact on all we currently know, do, and use as best practices in relation to manufacturing and supply of drugs.

Manufacturing facilities will have to be extremely flexible; they will most probably have to be geographical distributed to secure a safe supply chain to the individual patients and at the same time ensure high security for the medicine. The entire manufacturing and supply chain will be much more complex, and to secure patient safety and fast delivery, it has to be automated, fully integrated, and interoperable. The pace of establishing manufacturing facilities and to adjust established facilities to enable production of related and non-related medicines requires the manufacturer to use commercial off-the-shelf (COTS) technologies for their processes.

The use of COTS really hits the head of the nail with respect to major obstacles for the Pharma industry and its sub-suppliers. To achieve this goal, the companies and the people working there have to overcome the "Not Invented Here (NIH) religion."

Flexibility, flexibility, flexibility

To enable manufacturing facilities flexibility, they will have to undergo some changes. From the traditional build to order, manufacturing have to convert into a modular process. Where each process stage is a complete module, the vendor's package has to include all mechanical, instrumentation, and specific controls needed to perform the process functionality.

Each (and every) of the vendor's packaged equipment (VPEs) or process modules shall be self-contained with respect to processing. The interface between these process modules should be very simple.

Besides process modules performing the actual chemical or microbial processes, there must be product storage modules. We can call these storage modules farm modules. They exist in two types: one dedicated to a specific process module or set of identical process module types located near the same storage station and that primarily have product buffering or holding functionality (the farm module); and one type more flexible that can support multiple types of process modules and can move between storage stations (the smart farm module).

Both types of farm modules should have the capability of controlling and maintaining product as well as environmental conditions.

system 23

Operations management

From operational perspective, the manufacturing scheme clearly adds higher demands for flexibility.

A study for a Pharma user aimed to incorporate flexibility from a schedule and cost avoidance perspective. The study looked at the change of the production-setup of a manufacturing facility from one product to another similar but distinct product, where time to market for a product was the focus. The research studied from the start of changeover from the previous product to when the next product was produced and released.

It is clear a reduction in downtime really lowers the cost of manufacturing as the utilization of a manufacturing facility will annually increase from 65% to 95% if you can propagate the downtime and operate times. The reduction in lead time does not really affect the utilization of the facility, but solely the time to market.

The reduction in downtime for product changeover down to two days is possible, partly due to a sliding changeover of equipment and if the manufacturing operation management (MOM) and its associated controls and quality related activities required to perform the product changeover come with pre-validated process modules including their controls (i.e. the VPEs), and if the manufacturing facility MOM has plug-n-play for interfacing and integrating with the VPEs control.

The study was able to identify a number of architecture scenarios, ranging from existing low automation level up to what is barely possible with today's MES/PCS technology and definitely not possible with today's VPE control and interface capabilities and people's NIH mindset.

When going to procedural coordination only, all the coordination of the process modules, including any inter-process module coordination, shall be done by the MOM. This puts very high demands on the MOM.

It is questionable if the MOM as sketched above at all can be described as a "system." The MOM will most likely be a combination of COTS applications running on a common infrastructure and use this infrastructure for information exposure and gathering. The concepts and technology for the MOM must have its roots in the IT world and not in the traditional automation or batch world.

system 22

Standards come in play

Over the last decade or so, ISA as a community has developed ISA88 for Batch Control and ISA95 for Enterprise/Control System Integration, and the Pharma industry has adopted these standards.

The good part: Numerous applications have been put into operation covering functionality for integrating the business systems with the MES and PCS systems with respect to scheduling of production orders, discrete and batch execution, and monitoring the production performance for business use. And the functionalities have been divided according to the ISA95 functional hierarchy model, thus using standard definitions and terminology.

The bad part: Most of these installations and applications have been built to order. Another problem with these standards is they are not aligned to complement each other; instead there is some overlap as described by the ISA Joint Working Group 88-95: "ISA TR 88-95.01: Alignment of ISA88 and ISA95." This gives a number of the architecture scenarios described above as really not possible to define or to apply with consistency to either of the two standards.

On the Enterprise/Control System Integration side, initiatives started to extend the standard to also include methodology for the actual computer based integration of business to manufacturing transactions (ANSI/ISA-95.00.05-2007). Still, this standard is more a skeleton/framework for integration rather than defining an actual plug-n-play functionality. As such, it will still foster applications to build to order and not COTS.
In the same range of standards, the ongoing ISA88.00.05 (Make2Pack) is perhaps more interesting as it strives to address and standardize the interface for Machines to Control Systems (i.e. a sub-set of the VPEs).

Path forward

With the challenges that will arise within a short time, integrating systems vertically (shop floor to top floor) and horizontally (globalized supply chain) where everything must function in near real time and have a full-fledged plug-n-play nature and interface and communication capabilities will have to change. Changes will also have to occur to provide security and robustness for execution as well as retrieving and documenting the specific medicine supplied to the individual patient.

From the automation viewpoint, a number of initiatives will have to start to create a path forward that will enable a more flexible approach. Some of these initiatives could be:

Shift into a much higher gear on standardization work.

Create a common standard for ISA88 and ISA95 areas or redefine the ISA95 standard to defining the complete manufacturing operations management as a generic business and software model not specifically targeted toward batch, but that will also accomplish the task of batch control.

Investigate and explore technologies and standards used in other IT areas to create a basis for understanding how to enable integration across systems and software platforms and technologies.

Extend the current work done by Make2Pack and create a real compliance standard for integration of VPEs to the MOM (the plug-n-play standard).

Secure "buy-in" from automation and IT system and application vendors because without them we can do nothing.

The technological initiatives can only be successful if all participating stake holders will truly support the overall vision and make initiatives to break the current culture that avoids us all from using already existing products, research, or knowledge because of its different origins, also known as the NIH syndrome. This may quite well be the most difficult task of all to overcome.

ABOUT THE AUTHOR

Frede Vinther (frha@nnepharmaplan.com) is a senior automation specialist with Copenhagen, Denmark-based NNE Pharmaplan A/S, a services provider. This report is from a paper presented at the 2008 WBF North American Conference in Philadelphia.

Resources