Process Drift: When Do We Detect it? Richard L. Friedman Director, DMPQ CDER/Office of Compliance PQRI Process Drift Workshop December 1, 2010

Transcription

1 Process Drift: When Do We Detect it? Richard L. Friedman Director, DMPQ CDER/Office of Compliance PQRI Process Drift Workshop December 1, 2010

2 Overview Goal of Manufacturing Central Question: Why is process output not always predictable? Problem detection on stability program...too late! Examples: Why did the failures occur? Poorly understood processes = uncontrolled variability Testing Reliability Recommendations

3 Goal of a Manufacturing Organization Provide a consistent, defect-free product to the marketplace via consistent manufacturing operations For a Drug Manufacturer, this means Assure safety & efficacy every day, every dose. But not every company has established adequate quality practices, or achieved this predictable output. Some companies, products, or processes make more mistakes and defective units than others. Why?

4 More questions Why due routine stability studies find defects and result in recalls? This includes: Potency Content Uniformity Dissolution Impurities Physical characteristics (e.g., viscosity of ointment)

5 Impact What is the risk to the patient if a lot must be recalled? Incorrect dose, poorly dissolved product, high impurity ineffective or unsafe Shortage product not available What are the possible root causes? What is the conventional wisdom? Degradation! What are other possible reasons?

6 Two Overall Root Causes 1. The classical stability issue is chemical and physical degradation during storage. Some products are by nature unstable and have a shorter shelf life, but are expected to be able to meet their labeled expiration dating as supported by previous estimates/predictions. 2. The second issue to be considered is the inherent manufacturing/process variability due to inadequate process design and manufacturing.

7 Examples

8 Importance of Excipients: Some Examples of Recalls Involving Excipients ( ) Excipient Product Reason Root Cause Malt Syrup Oral Powder Product (Granules) Mold Long excipient storage period Sorbitol Syrup OOS Stability Different excipient supplier Magnesium Stearate 5, 10, 15, and 30 mg Tablet Product Dissolution Failures Rendered drug hydrophobic due to nonuniform mix issues (excipient itself was not defective). HPMC Extended Release Tablet Dissolution Failure (Drug release rate too rapid) Excipient functionality led to this Recall 8

9 Case Study #1 Dissolution Failure (multivariate cause?) 9 It was concluded that the cause of the dissolution failure was a combination of factors, e.g., a formulation change, specifically a 1% increase in lubricant a subtle change in the effective density of the tablets which was apparently affected by the bulk density of microcrystalline cellulose (another ingredient variable) mixer change (different mixing principle) Dissolution testing on stability ultimately detected the problem. only 1 batch/year placed on stability, but extra testing done after the failure 3 batches failed

10 Case Study #2 Subpotency (multivariate cause?) Tablet product Assay failure on Stability (9 months) Firm s investigation concludes that product stability needs to be improved by: 1.changing to a different API source 2.modification of formula 10 3.improved container/closure system

11 Ointment drug product Case study #3 Assay, Appearance NDA sold, and product now being made by contract manufacturer Firm receives numerous complaints of product being thinner consistency, watery, or liquidy. Intermittent failures to meet specification for assay, appearance and/or propylparaben content Process involves complex set of process steps: melting and mixing, dispersion, dissolving, heating, cooling for specified times/temps. Transfer of materials to and from 5 different vessels. Product and Process Design flaws. Firm issued WL.

12 Variability

13 Variability? Have sources of variability been identified and minimized? API, Excipients Formulation Process Container-Closure System Sampling and Measurement methods

14 Development Do companies collect enough data, within and between batches, to really understand the stability profile, and transition to typical one lot/year? Do companies truly understand first principles the physicochemical reasons that contribute to stability failure? If so, why are batches vulnerable to stability failure during shelflife not routinely identified (prediction capability) prior to distribution?

15 Formulation Development Do companies routinely do sufficient preformulation studies of excipients and API s to understand their behavior? 1975 journal article used DOE for preformulation excipient compatibility. Studied 5 factors and 10 two-factor interactions to get the best combination. Do companies follow the QbD concepts in ICH Q8 for formulation design? Is DOE used routinely?

16 Raw Material Variability Have firms routinely identify all of the raw material attributes that are important? How has the variability been identified, measured and minimized? Many excipients (e.g., cellulose-based) are natural materials, with unmeasured variability... Are incoming ingredient batches sufficiently analyzed using adequate tests? e.g., HPMC

17 Raw Material Variability Changes in particle size of some excipients, for example, may affect content uniformity. In other cases, a change in the supplier of an excipient or lubricant may affect dissolution or bioavailability. The failure to specify the amount of granulating solution, resulting in over wetting and dissolution failures of aged batches. FDA Guide to Pre/Post-Approval Inspections of Solid Oral Dosage Forms (1994)

18 Excess Manufacturing Variability? The root cause of root causes is often the failure of management to focus on minimizing unwanted variability, differences, and discrepancies throughout the product life cycle. There is a need to fully implement (a). to validate the performance of those manufacturing processes that may be responsible for causing variability

19 Management s Role It is good management to continually reduce the variation of any quality characteristic - Deming This builds robustness and ruggedness into the process and product, increases product quality, and can reduce costs.

20 Representative Samples The sample must represent the batch physically. For example, beginning, middle, end of the batch, and... The sample must represent the variability in the batch This applies to all samples, whether raw material, in-process, QC, or stability...

21 Representative Samples This is a key concept and assumption in the CGMP s (b)(21) (b) (a) (b) (e)(1) Definitions Testing Materials Inspection Review

22 Testing Programs

23 QC Release: Quality System Detection of Variation & Defects before Distribution Test of a firm s Quality System is if it will promptly catch a problem in a batch vs. discovering only after it is marketed. 1. Mistakes are, in many cases, not caught by the individual making the error, but instead through final inspection or QC test!! 2. QC testing is of limited sample size intended to assess a chemical, microbiological, or physical attribute. 3. To avoid detecting mistakes or defects only after a drug product has been distributed: Use Redundancy of Controls or PAT

24 Testing After Distribution: How Much is Enough? Normally, 3,6,9,12,18,24... But based on experiences with some products, some firms have had to test more often than the usual intervals. Major Migraine drug: was tested every month. Female Hormone Product: put every fifth lot on stability due major dissolution problems, then ultimately every lot pending reformulation. (Not the only such example)

25 Stability Studies and Variability As number of tested samples increases... More tests lead to higher probability of failure for unstable processes or products. Excessive variability in batches results in higher probability of failure.

26 E One statistical approach for developing inhouse specifications for USP Standards now gaining attention and acceptance is ASTM E11 E2709 adopted in 2009 It is a statistical procedure that evaluates the variability in the data and calculates probability of passing the specification (assay, content uniformity, dissolution) with a specified confidence.

27 E : Probability of Passing So instead of a pass/fail response from using the USP Standard as a specification, this statistical method gives a probability. For example, we would be able to say: we are 99% confident that the probability of the batch passing the USP Standard in the future is 99% or better. If a batch just barely passes the dissolution or content uniformity criteria once (e.g., the USP Standard), what is the probability of passing 1 to 9 more times? Depending on the overall inherent batch variability, it can range from circa 1% to 100%!

28 For Example (E2709 Analysis) A batch with a 90% probability of passing (i.e. 10% fail) the test the first time, has almost a 60% chance of failure if tested eight times on a stability program. A batch with a 99% probability of passing the test the first time, still has almost a 9% chance of failure if tested nine times.

29 Summary / Recommendations Unwanted variability is the root cause! raw materials, product stability, and process stability Passing the specifications once often gives little assurance that it will pass again The batch has to be robust to be assured of passing repeatedly on a stability program

30 Summary/Recommendations 1. Preformulation DOE and history Learn and routinely use DOE in formulation 2. Process/product control Learn and adjust under Quality System 3. Raw material change control 4. Follow GMPs to assure representative sampling 5. Minimize variability 6. Batch history 7. Process capability

31 Building Knowledge Process Validation Lifecycle Replication at full scale provides initial assurance of commercial process reliability. Validation includes lifecycle monitoring. Post-market information gathering, promotes maintenance of a stable process and identifies areas for continual improvement and adaptation. Our Compliance Policy Guide on Process Validation, and the draft Process Validation Guidance, recognize the value of advanced engineering principles and control technologies. Design Assess Confirm Monitor

32 CGMP: Every batch, Every day We rely upon the manufacturing controls and standards to ensure that time and time again, lot after lot, year after year the same clinical profile will be delivered because the product will be the same in its quality We have to think of the primary customers as people consuming that medicine and we have to think of the statute and what we are guaranteeing in there, that the drug will continue to be safe and effective and perform as described in the label. - Janet Woodcock, M.D.

33 Reference Six Sigma in the Pharmaceutical Industry: Understanding, Reducing, and Controlling Variation in Pharmaceuticals and Biologics By Brian Nunnally and John McConnell CRC Press, 2007

34 Acknowledgments: Lynn Torbeck For More Information: