Understanding the FDA’s thinking is central to defending an improvement of corrective action to a regulated process.  The pharmaceutical industry’s relationship with the FDA and its sister regulatory organizations around the world have been tenuous at best.  For years perceived to be in conflict with industry, FDA has struggled to establish a framework with industry that was not strictly punitive but fostered some facility for dialogue. One example of this is the plethora of guidance documents issued by the agency in an effort to provide direction and articulate what the FDA would like to see in every facet of the drug development business.

While this makes sense in theory, the reality is that drug development is a highly complex undertaking, complicated by a myriad of specific details which can dramatically impact compliance exposure and public risk. As a result, FDA compliance inspectors were required to be experts in a vast array of guidances in order for this approach to be effective in moving the industry to some common level of understanding.  Practically, this was not easy to implement. As a result, enforcement was uneven across industry, promoting confusion and lack of consensus among inspectors and industry.

In an effort to promote common understanding, the FDA chose to focus on specific areas within the industry through a combination of focused surveillance inspections and presentations at industry professional organizations. In the 1980s, the focus was on water system design and testing and computerized system validation. Both industry and the agency were beginning to integrate risk as a central part of the compliance argument.

FDA’s diminishing ability to depend upon inspection and documentation as its backbone for risk management drove it to become more and more risk averse, making drug development costs skyrocket.  The FDA argued that this was not the case, as the average time for NDA approvals was holding steady, partly due to additional resources the agency could apply to the new process, as a result of Commissioner Kessler’s prescription drug user fee act (PDUFA) fee paid by manufacturers upon regulatory filing. 

However, what was not being tracked were the additional hurdles being thrown at the industry at the IND levels. API’s being applied in combination therapies or utilizing improved drug delivery platforms had been in the marketplace for decades, and suddenly companies were being asked to run additional carcinogenicity and mutagenicity studies on them, in a desperate effort by the agency to drive its risk exposure as low as possible. The result was fewer and fewer new drugs coming on to the market with less time for manufacturers to recoup their investments.

The other reality is that this approach was not really driving down risk, as it did not address the underlying sources of variation in these therapies which could represent a threat to the patient.

Issues came to a head in 2004, when David Graham, an Associate Director of FDA’s Office of Drug Safety, used the protection of the Whistleblower Act to address the public risk from Vioxx.

Dr. Graham had previously been successful in removing the unsafe drugs Omniflox, an antibiotic, Rezulin, a diabetes treatment, Fen-Phen and Redux, weight-loss drugs, and phenylpropanolamine, an over-the-counter decongestant, from the U.S. market and in restricting the use of Trovan, an antibiotic, to hospital settings. He had also played a part in the removal of Lotronex, Baycol, Seldane, and Propulsid.

Dr. Graham cited the fact that the FDA succumbed to the industry pressures of a potential blockbuster drug and disregarded the warnings of its own scientist sin terms of the safety risks. The agency knew this could not continue and began a wholesale transformation for defining acceptable drug development, one that was based on scientific understanding and risk management as a foundation for demonstrating safety and efficacy.

The FDA launched its Critical Path Initiative in an effort to “lean out” the regulatory approval process. Specifically the objective was to modernize the techniques and methods used to evaluate the safety, efficacy, and quality of medical products as they move from candidate selection and design to mass manufacture through better predictive and evaluative tools, including:

• innovative trial design, new statistical tools and analytic methods,
• use of modeling and simulation;
• establishing and qualifying predictive biomarkers for specific conditions;
• less stringent Current Good X Practice (cGxP)1 regulations for IND exploratory studies and clinical trials.

In order to achieve this, the pharmaceutical industry had to change the way it developed medical products and established Quality Assurance. FDA looked for guidance from best practices across industries and around the world. It had participated heavily in the International Committee on Harmonization (ICH) of Technical Requirements for Registration of Pharmaceuticals for Human Use, which had attempted to consolidate best practices across the industry.

Landmark Guidance Documents
Out of this effort came several landmark guidance documents, which became the foundation for most global regulatory philosophy. Three critical guidance documents were ICH Q8, ICH Q9, and ICH Q10.

ICH Q8 provided guidance in terms of the Pharmaceutical Development section of an ICH regulatory submission. It indicated areas where the effective application of risk management and pharmaceutical and manufacturing sciences can create a basis for flexible regulatory.

ICH Q9 provided guidance in terms of establishing a quality risk management framework. It provided guidance on the principles and some of the tools of quality risk management that can enable more effective and consistent risk-based decisions, both by regulators and industry, regarding the quality of drug substances and drug (medicinal) products across the productlifecycle. ICH Q10 attempted to integrate these two new concepts as part of pharmaceutical quality system.

With these best practices in place, the FDA issued a series of new guidance documents, which were intended to integrate this new thinking. The first landmark guidance was issued in September 2003 by the agency and was entitled “Pharmaceutical cGMPs for the 21st Century—A Risk-Based Approach.”

This guidance laid out the vision for the agency and attempted to describe the benefit to industry in terms of reduced regulatory burden and potentially faster time to market. This was quickly followed between September 2003 and March 2005 by a series of additional guidance documents:

• Guidance for Industry—Quality Systems Approach to Pharmaceutical Current Good Manufacturing Practice Regulations
• Guidance for Industry Sterile Drug Products Produced by Aseptic Processing–cGMP
• Guidance for Industry—Process Analytical Technology (PAT) a Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance

This guidance was designed to encourage the industry to adopt online testing, characterization and control

• Guidance for Industry—Premarketing Risk Assessment
• This guidance lays the groundwork for risk assessment during product development, with emphasis on Phase III clinical trials.
• Guidance for Industry—Development and Use of Risk Minimization Action Plans

This guidance documents additional steps that companies can take for patient risk that may not mitigated by traditional product labeling.

Together, these guidance documents represented a wholesale shift in quality thinking, one based on process understanding and predictability (as opposed to product performance) as the primary metrics for product safety and efficacy.  ICH Q9 fundamentally embraces the concept that every step of the drug development process involves some element of risk.

A typical risk management framework is shown in Figure 1. In order to apply a risk management framework effectively, the technical and quality organization must consistently apply risk management tools to support its decision-making process.   Some of the risk management tools that can be applied are discussed in the following section.

Fault Tree Analysis (FTA)
FTA is a team-based method used to identify the causal chain that creates a hazard or a failure mode (effects are typically ignored). It represents the sequence and combination of possible events that may lead to a failure mode. Once the causes are identified, preventive action can be taken.

It is composed of a series of Events and Gates. An “Event” is a cause or an effect. A “Gate” defines the conditional relationship between causes and effects, between x and y (what must happen for the effect to occur). FTAs are often used:

• when conducting a risk analysis of a new facility or equipment;
• when multiple causes of a failure mode are suspected;
• when an interaction of causes is suspected;
• as an input to a FMEA;
• as an input to an experimental design to characterize a process or to determine variables and levels that will create a failure mode.

FTAs do have some limitations, they:

• are time and resource intensive;
• require expert knowledge of system under study;
• can lead to paralysis by analysis
• require Microsoft Visio or other specialized software to document;
• are more useful for problem solving than prevention.

An example of a FTA is shown in Figure 2.

Hazard Analysis and Critical Control Points (HACCP)
A HACCP is a method of identifying and controlling sources of variation in critical process steps that could lead to a hazardous condition. It is similar to a control plan. It cannot be used effectively without manual or automated process control methods, including statistical process control (SPC).

FDA has recognized HACCP as a useful and effective risk management tool for some time.  It became a mandatory component of the FDA’s regulation for all seafood processing plants in 1997 and has been mandatory for canned food manufacturers since 1973.

The basic steps for conducting a HACCP are to:

• develop a flow chart for the process;
• identify any hazards that must be prevented. eliminated, or reduced to acceptable levels;
• identify variables that can cause the hazard (Design of Experiments or other tools);
• identify the critical control points in the process where those variables are impacted/affected;
• establish critical limits for those variables beyond which the hazard is created;
• determine and implement effective monitoring procedures at critical control points;
• determine corrective actions when monitoring that indicate that a critical control point is not under control (adjustment, maintenance, and so on);
• establish records of control and correction.

HACCP analyses are often used with new manufacturing processes or equipment.

However, HACCP analysis requires:

• excellent process knowledge to be effective;
• an FMEA to identify critical hazards/ failure modes (a HAACP could be an action to reduce the risk in an FMEA);
• use of more complex statistical tools to be effective (including characterization designs of experiment (DOEs,) SPC, Capability Matrices).

An example of a typical HACCP form for a compression operation is shown in Table 1.

Hazard and Operability Method (HAZOP)
A HAZOP is a team-based risk management tool designed to identify hazards and deviations from normal operations, determine the hazard level, and brainstorm actions and recommendations to prevent the hazard from occurring, or to minimize its impact.
In a HAZOP, the key parameters of a particular process design are identified. In a pharmaceutical process this would be the Critical Process Parameters (CPP) that affect process predictability. For a chemical reaction vessel, for example, these parameters could be temperature, mixing time, pressure, and time.

After parameters are identified, the HAZOP leader applies a series of guide words to the critical parameters and the team determines if any apply to the critical parameters in terms of a possible hazard.  Table 2 is a sample Guide Word table.

The hazard severity and frequency are also assessed in order to establish the hazard levels. The risk matrix for this assessment may already exist for the process and product or may have to be developed by the team as part of the HAZOP exercise. An example of a risk matrix is shown in Figure 3.

An example of a HAZOP for a chemical reaction vessel is shown in Table 3. HAZOPs are often used for new or current manufacturing processes or equipment and can be an excellent planning tool for maintenance groups.

However, they have the following limitations when compared to other risk management tools:

• risk level is more subjective than FMEA;
• typically there is no reassessment of risk after controls applied.

Failure Modes and Effects Analysis (FMEA)
FMEA is widely used in the pharmaceutical industry today, to:

• minimize the impact or severity of the risk;
• prevent the causes of risk from occurring; or to
• detect the risk early in its lifecycle to minimize its effect.

FMEAs are cross-functional and require representation from all departments that may be impacted by the risk. The FMEA should be updated throughout the life of the product and the process development. It is best to conduct an FMEA as early as possible, preferably during product design and development.

Before beginning the FMEA the team must establish agreed-upon metrics, called ranking tables, for assessing risk severity, frequency of occurrence, and likelihood of detection. Examples of severity and occurrence ranking are shown in Tables 3 and 4.
Based upon these agreed-upon ranking tables, the team will then step through each process and determine if there is a possible failure mode and assign a severity, occurrence, and detection ranking. These values when multiplied together comprise the Risk Priority Number (RPN). The RPN is used to prioritize mitigation actions for each identified risk. An example of an FMEA Ranking Table is shown in Table 5. 

---

These excerpts from Chapter 2 (entitled Trouble in Paradise and Catalyzing Change) of Mr. Chatterjee’s new book,”Applying Lean Six Sigma in the Pharmaceutical Industry” are reproduced by permission of the author and publisher. The book is published by Gower Publishing, Ltd., ISBN 978-0-566-09204-6