Emil W. Ciurczak, DoraMaxx Consulting04.03.17
Having just attended IFPAC (International Forum Process Analytical Technology) in Maryland and then PittCon in Chicago, I had the opportunity to speak with the “movers and shakers” of the Pharma industry. There are several things upon which they agree: 1) continuous manufacturing (CM), including 3-D printing, is the future of pharmaceuticals; and 2) even the biggest name manufacturers are moving towards becoming “virtual” companies. Becoming virtual means that Big Pharma is shrinking—not in profits, just headcount. That is, they are, more and more, moving to outsourcing in clinical trials, manufacturing, and packaging while enhancing their core. Even the pure research is being augmented by university research and smaller, independent Pharma companies, so the future will look far different from the past.
Now that Big Pharma is seeking partners for the grunt work (i.e., CROs, S+CMOs, and generic partners), the contract world needs to embrace a new paradigm and see it as a boon to the bottom line. To attract new monies from clients, we need to embrace the quote from Field of Dreams: “If you build it, he will come.” The tried and true approach of transferring production from the parent company to the contract manufacturing organization (CMO) will need to be updated. No longer will having mirror hardware of mixers, granulators, and tablet presses be sufficient. Since most of the larger companies (e.g., Pfizer, GSK, Merck, and Novartis) are embracing CM, having 1980s equipment is not what will attract new business. But, as I mentioned the hardware needs previously, I would like to address the other elephant in the room: building a QRM—needed for PAT and QbD—in a small company.
What, you might ask, is QRM? Quality Risk Management is, in its simplest form, a mindset and atmosphere where we learn what could possibly go wrong and proceed to address it. Notice I did not say “fix,” just “address.” A properly designed and well-oiled QRM program follows the adage, “Let me fix what I can, accept what I can’t, and the wisdom to know which is which.” One of the questions that must be asked and seldom is by newcomers to the QRM field is, “Will fixing the problem cause more harm than good?”
The FDA certainly believes QRM is key and the agency represented half the speakers in the dedicated sessions. Clearly, the large Pharma and BioPharma companies take it seriously and understand it is key to successful PAT/QbD programs. Now, we need to familiarize the CMOs and CROs with the need and relative simplicity of managing risk. Simple? Well, simple is relative. When the concept of QRM was advanced by the FDA and ICH, there were few software tools on the market to help. Over the last few years, any number of software packages have been released to the market. These range from bullet types—individual pieces, such as Ishakawa fishbone charts, numerous models of ranking risks such as FMEA, and design of experiment (DoE) software—to combinations of the pieces.
Several companies who make larger statistical packages (e.g., CAMO and Umetrics, now MKS) also make DoE programs. Most are quite useful, if not specialized. Other, smaller packages are also available, where just Ishikawa (fishbone) diagrams, where all steps and sub-steps are graphed to more easily ascertain what could possibly contribute to a poorly behaving product (see Figure 1). An alternate way of displaying these points of potential risk is a simple block diagram (Figure 2), where each step of a batch process—easily converted to CM process—is a shopping list of contributions/parts in that step.
While many larger companies have full-time staff to perform QRM programs, smaller companies may have to enlist part-timers or consultants to perform theirs. However, after seeing and participating in several exercises, the staff of these companies should be able to perform this on their own from there on. The most difficult part, in my opinion, is building risk values. Each small step is evaluated for three attributes: 1) How serious is any occurrence?; 2) What are its chances of occurring?; and 3) How easily can we detect it? Usually this is displayed as a couple of charts where each is paired off with one of the others (see Figure 3).
These values for occurrence/seriousness/detectability can also be displayed on a single graph as seen in Figure 4. These values (ranging from 1-10) are placed into the ranking chart, often a FMEA program. The software can then represent the risks from highest to lowest and suggest which parameters to use in a formal DoE.
Normally, these different steps may be purchased as smaller, individual software programs. And, just about any vendor’s works as advertised. However, the simpler the program, the more skill/experience is needed by the operator of that program. In a large, established organization, this is usually no problem. But, in a CMO/generic firm, the “warm bodies” are seldom just sitting in the break room, waiting to be consulted. But, do not fret. Newer, more operator friendly programs are being written and sold, even as we speak. One such package was shown at IFPAC this year.
Named iRISK it was designed to ensure the organization follows proper, pre-defined sequences of risk management steps that include:
Quality Target Product Profile (QTPP); Process Mapping; Cause-Effect Analysis; Failure Modes and Effects Analysis (FMEA); and on to DoE.
It takes advantage of the linkage between the QRM tools available in the package to minimize re-work and boost risk-based decisions. These “workflows” are configured by the organization for the several Lifecycle stages of products/processes, as well as equipment and suppliers, to ensure the best fit with all QRM activities.
When setting up the program, the block (step-wise) diagram is used. Clicking on each step provides a pull-down list of contributions. This is where your consultants or in-house staff insert all the steps affecting the particular stage of the process (e.g., for raw materials: density, particle size, moisture, flowability, crystallinity, or polymorphic form).
When these boxes are populated, a click will convert the block diagram to an Ishakawa chart. When this “fishbone” presentation is satisfactory, a FMEA-type chart is automatically populated. The “experts” then use the values for each step based on severity, occurrence, and detectability to give them numerical values. These values are then converted to a Pareto chart for determination of which parameters should be evaluated in a DoE.
What I liked about this package is that a small, less experienced company doesn’t need to purchase several programs or lean quite so much on consultants to succeed in developing a QRM program, which places them on par to succeed in gathering more contract manufacturing work from the “big guys.”
Live long and prosper.
Now that Big Pharma is seeking partners for the grunt work (i.e., CROs, S+CMOs, and generic partners), the contract world needs to embrace a new paradigm and see it as a boon to the bottom line. To attract new monies from clients, we need to embrace the quote from Field of Dreams: “If you build it, he will come.” The tried and true approach of transferring production from the parent company to the contract manufacturing organization (CMO) will need to be updated. No longer will having mirror hardware of mixers, granulators, and tablet presses be sufficient. Since most of the larger companies (e.g., Pfizer, GSK, Merck, and Novartis) are embracing CM, having 1980s equipment is not what will attract new business. But, as I mentioned the hardware needs previously, I would like to address the other elephant in the room: building a QRM—needed for PAT and QbD—in a small company.
What, you might ask, is QRM? Quality Risk Management is, in its simplest form, a mindset and atmosphere where we learn what could possibly go wrong and proceed to address it. Notice I did not say “fix,” just “address.” A properly designed and well-oiled QRM program follows the adage, “Let me fix what I can, accept what I can’t, and the wisdom to know which is which.” One of the questions that must be asked and seldom is by newcomers to the QRM field is, “Will fixing the problem cause more harm than good?”
The FDA certainly believes QRM is key and the agency represented half the speakers in the dedicated sessions. Clearly, the large Pharma and BioPharma companies take it seriously and understand it is key to successful PAT/QbD programs. Now, we need to familiarize the CMOs and CROs with the need and relative simplicity of managing risk. Simple? Well, simple is relative. When the concept of QRM was advanced by the FDA and ICH, there were few software tools on the market to help. Over the last few years, any number of software packages have been released to the market. These range from bullet types—individual pieces, such as Ishakawa fishbone charts, numerous models of ranking risks such as FMEA, and design of experiment (DoE) software—to combinations of the pieces.
Several companies who make larger statistical packages (e.g., CAMO and Umetrics, now MKS) also make DoE programs. Most are quite useful, if not specialized. Other, smaller packages are also available, where just Ishikawa (fishbone) diagrams, where all steps and sub-steps are graphed to more easily ascertain what could possibly contribute to a poorly behaving product (see Figure 1). An alternate way of displaying these points of potential risk is a simple block diagram (Figure 2), where each step of a batch process—easily converted to CM process—is a shopping list of contributions/parts in that step.
While many larger companies have full-time staff to perform QRM programs, smaller companies may have to enlist part-timers or consultants to perform theirs. However, after seeing and participating in several exercises, the staff of these companies should be able to perform this on their own from there on. The most difficult part, in my opinion, is building risk values. Each small step is evaluated for three attributes: 1) How serious is any occurrence?; 2) What are its chances of occurring?; and 3) How easily can we detect it? Usually this is displayed as a couple of charts where each is paired off with one of the others (see Figure 3).
These values for occurrence/seriousness/detectability can also be displayed on a single graph as seen in Figure 4. These values (ranging from 1-10) are placed into the ranking chart, often a FMEA program. The software can then represent the risks from highest to lowest and suggest which parameters to use in a formal DoE.
Normally, these different steps may be purchased as smaller, individual software programs. And, just about any vendor’s works as advertised. However, the simpler the program, the more skill/experience is needed by the operator of that program. In a large, established organization, this is usually no problem. But, in a CMO/generic firm, the “warm bodies” are seldom just sitting in the break room, waiting to be consulted. But, do not fret. Newer, more operator friendly programs are being written and sold, even as we speak. One such package was shown at IFPAC this year.
Named iRISK it was designed to ensure the organization follows proper, pre-defined sequences of risk management steps that include:
Quality Target Product Profile (QTPP); Process Mapping; Cause-Effect Analysis; Failure Modes and Effects Analysis (FMEA); and on to DoE.
It takes advantage of the linkage between the QRM tools available in the package to minimize re-work and boost risk-based decisions. These “workflows” are configured by the organization for the several Lifecycle stages of products/processes, as well as equipment and suppliers, to ensure the best fit with all QRM activities.
When setting up the program, the block (step-wise) diagram is used. Clicking on each step provides a pull-down list of contributions. This is where your consultants or in-house staff insert all the steps affecting the particular stage of the process (e.g., for raw materials: density, particle size, moisture, flowability, crystallinity, or polymorphic form).
When these boxes are populated, a click will convert the block diagram to an Ishakawa chart. When this “fishbone” presentation is satisfactory, a FMEA-type chart is automatically populated. The “experts” then use the values for each step based on severity, occurrence, and detectability to give them numerical values. These values are then converted to a Pareto chart for determination of which parameters should be evaluated in a DoE.
What I liked about this package is that a small, less experienced company doesn’t need to purchase several programs or lean quite so much on consultants to succeed in developing a QRM program, which places them on par to succeed in gathering more contract manufacturing work from the “big guys.”
Live long and prosper.
