The size of many companies became so complex, they were almost unmanageable. Clearly, one managerial path is to delegate when one is overwhelmed with tasks. A company is somewhat like a skilled juggler: it can suspend a growing number of balls in the air, BUT there are limits! Let us examine two reasons to outsource production of, what we call, “product X.”
1. The product is a steady seller—$10-$50,000,000 per annum—but there is only a need to produce several batches per year/quarter. Under cGMP (batch) process production, the time involved in setting up, breaking down, cleaning, and validating the cleanliness of the process equipment, makes the product, still profitable, less attractive (i.e., profitable). Two obvious reasons are: when the equipment (and operators) are tied up with product X, they cannot be utilized on products Y, Z, A-F, slowing down output of blockbuster; and when a smaller, dedicated contract manufacturing organization (CMO) could make the product without the higher overhead of the parent company, it allows for both higher profit margins and better-timed deliveries to customers.
2. Product X is the blockbuster and it is tying up most of the process equipment, blocking production of products A ->F, Y, and Z. While the pipeline of X is satisfied, all the others could be delayed to market, causing shortages and, potentially, doctors switching their patients to competitors’ products. Once switched, they are fain to return.
Ok, these may well be convincing reasons for outsourcing either the blockbuster or the “second tier” products, but there are a number of roadblocks, real or imagined, to any or all outsourcing scenarios. Some are quite real and potentially serious, while some are mere convenience issues.
“Different” quality is an issue due to different equipment, personnel and their levels of training and experience, and quality of excipients/APIs.
Even following an MMF (Master Manufacturing Formula) to the letter, we have found via NIRS that different equipment has a different “process signature.” That is, it has, for example, slight variances in pressure across the face of a tablet punch, different speed, ejection types, etc. All show up in a NIR spectrum, should process control be performed in that fashion.
Different suppliers of raw materials, especially in other continents, will have slightly different qualities, causing slight variations in final product performance. Clearly, it would be prohibitively expensive to ship U.S.-made materials to, say, Vietnam for use, so we use locally-produced materials.
There are a number of unscrupulous producers. Should a CMO/CRO/CCO decide to use OOS materials (non-GMP). This could cause either or both a recall or heath/stability problems.
Supply chain length
Every added step between the raw materials and final customer adds the potential of delay. With the current tariff wars, the lower cost of outsourcing to another country could be offset by added tariffs. Add to that, shipping and weather delays, labor disputes, and border checks for proper documentation, and the added time makes any profit advantage evaporate.
Analytical ability of “CxO”
While a “name” pharma company will have invested millions to build, equip, and staff modern analytical/QC laboratories, the vast majority of contract companies do not have the resources to keep abreast of new and improved analytical hardware and software. This disparity slows transfer of production from the originator company to the outsourcing site.
When we speak of technology transfer, we mean both process equipment and measurement technology.
In terms of processing hardware, this would include blenders, granulators, drying apparatus and tableting machines, especially for specialty dosage forms—osmotic pumps.
For laboratory/process monitoring instrumentation, these can be simple upgrades to totally innovative technologies. The most common analytical technologies (i.e., ultraviolet/visible and mid-range infrared spectroscopies) are almost always available and lab models can be used to build models for process analyses. However, most contract manufacturing locations do not have the experience in sampling and calibrating for process analyses.
A simple, yet significant upgrade would be a shift from HPLC (high performance liquid chromatography) to UPLC (where U = ultra) for small and large molecule analyses, both for R&D and QC applications. Extending this upgrade to a process is somewhat easier than a non-contact method.
More complex technologies (i.e., Near-Infrared, TeraHertz, Raman), which exist in both lab and process-monitoring forms, are nowhere near being common or addressed in most university curricula. Larger companies, mainly due to a larger staff, can afford to allow on-the-job training for their analysts; small CxO’s need, but can seldom afford, an analyst who can hit the ground running.
Some suggested solutions
Starting with the analytical components of outsourcing, there are a number of potential fixes. Based on what was done at one company for whom I worked, merely transfer the proper analytical instrument along with the product to be controlled. When we developed a new drug product analysis via GC or HPLC, we sent the unit, along with appropriate columns and, sometimes, solvents to the QC department in lieu of expecting them to purchase, set-up, and perform IQ/OQ/PQ on unfamiliar equipment.
This approach may be modified with special or unfamiliar hardware (i.e. NIRS or Raman), where the equipment is purchased and qualified by the client company and one or more analysts from the contract manufacturer are trained either in-house or at the client’s facility before it is sent to the contract facility. This assures parallel analyses at both locations at a cost of less than $100,000 per unit. This is usually far less that the contract company attempting the job on its own.
For processing equipment, sending classic, production-sized units—blenders, fluid-bed units, etc.—is getting much costlier, unless we are speaking about smaller, less expensive, more versatile continuous manufacturing equipment. If the former, then, unless it is a specialty unit, the contract company will likely both have and know how to operate it.
Again, any differences between the contract company’s equipment or experience can be made smoother by cooperative training and hardware exchange.
Novelty processing and monitoring equipment, such as 3-D printers and on-line monitors (i.e., NIR or TeraHertz), are becoming more and more common in larger companies. It would behoove the client/parent company to invest some time and money in order to bring the contract facility on par with the originator’s plant/lab.
When you, the client company, start to become confident with a new analytical or process technology, including software for process control, an eye should be cast towards eventually cloning your success at one or more contract organization sites. While they can also use these technologies for other clients, the fact that they will more quickly finish the orders for others, assures that the equipment will be more readily available for your tasks.
Also, the more the contract facility’s staff uses the new equipment, the more skilled they become and the better they can make your product.
The true bottom line for all these changes and updates to contract facilities is that the owner of the blockbuster or other product would have the options not available previously. Management can either allot the multi-batch per week product or the batch or two per quarter product to the contract organization with confidence.
As far as CROs are concerned, cloning the technologies makes sense, even if your chief researchers are not present to run all the tests/experiments. The protocols may be developed by the home lab personnel and carried out by the contract team. With the internet secured by a cloud, one would assume, the experiments’ progress may be followed remotely, making visits to the contracted site less frequently needed.
The same could apply to clinical contract organizations. They need not be geographically very close, since their lab facilities would be cloned for the client and their goods made and packaged under identical conditions to the client’s. This would also allow easy oversight via the internet.
So, it seems that several technical trends are concomitantly occurring with the current trend of Big Pharma downsizing/outsourcing.
For one, real-time, in-process monitoring and control, a.k.a., PAT, where the old-fashioned concept of one-size-fits-all master manufacturing formulas (MMFs) is becoming a quaint relic of the past. With the advent of PAT/QbD, companies now “blend until blended,” not to a pre-set time. They dry until dry, etc., not to a time point. This allows CMOs to more closely mimic the originators’ product, giving them more confidence with outsourcing.
Also, continuous manufacturing obviates the need for legions of trained analysts at contract organizations, lowering the time and cost of technology transfer.
In short, technology improvements seem to be willing to help companies outsource many of their functions with confidence. We’re living the dream.
Emil W. Ciurczak
Emil W. Ciurczak has worked in the pharmaceutical industry since 1970 for companies that include Ciba-Geigy, Sandoz, Berlex, Merck, and Purdue Pharma, where he specialized in performing method development on most types of analytical equipment. In 1983, he introduced NIR spectroscopy to pharmaceutical applications, and is generally credited as one of the first to use process analytical technologies (PAT) in drug manufacturing and development.