Statistical Methods for Biotechnology Products Regulatory Requirements of Drug Stability by Jen-pei Liu, PhD, Professor Division of Biometry, Department of Agronomy National Taiwan University and Division of Biostatistics and Bioinformatics National Health Research Institutes Copyright by Jen-pei Liu, PhD 1 Background Outlines - What？ - Why？ - Definitions FDA Guideline - Basic requirements - Batch similarity - Container, drug product, and sampling time consideration International Harmonization - International Conference on Harmonization - ICH stability guideline Current Issues - Batch similarity - Design issues - International estimate Copyright by Jen-pei Liu, PhD 2 - Frozen drug products What is stability Historical milestone Pre-1975 No requirements( some drug products are known to be unstable, e.g., Penicillin ) 1975 First USP with expiration dating period clause 1984 FDA issued stability guideline 1987 Current FDA guideline 1993 WHO draft guideline 1993 ICH issued stability guideline 1994 FDA guideline being revised 2003-2005 ICH stability guidances Q1A, Q1D, Q1E, Q1F Copyright by Jen-pei Liu, PhD 3 What is stability？ FDA Guideline (1987) The capacity of a drug product to remain within specifications established to ensure its identity, strength, quality, and purity. ICH Guideline (2003) The purpose of stability testing is to provide evidence on how the quality of a drug substance or drug product varies with time under the influence of a variety of environmental factors such as temperature, humidity, and light (storage conditions) and to establish a shelf life and recommended storage conditions. Prior to expiration date, all characteristics of drug product must remain within approves specification. Copyright by Jen-pei Liu, PhD 4 Why do we need stability？ FDA requirement - For every drug product on the market, an expiration date must be indicated on the immediate container label Purposes - To characterize the degradation of a drug product - To establish an expiration dating period applicable to all future batches of the drug product - To ensure that the drug product will retain its identity, strength, quality, and purity through out the expiration period Copyright by Jen-pei Liu, PhD 5 Definitions Formal stability studies Long-term and accelerated (and immediate) studies undertaken on primary and/or commitment batches according to a prescribed stability protocol to establish or confirm the shelf life of a drug product. Long-term testing Stability studies under the recommended storage conditions for the shelf life proposed (or approved) for labeling establish or confirm the shelf life of a drug product. Accelerating testing Studies designed to increase the rate of chemical degradation or physical change of a drug product by using exaggerated storage conditions Immediate testing Studies conducted at 30oC/65%RH and designed to moderately increase the rate of chemical degradation or physical change of a drug product intended to be stored at 25oC. Copyright by Jen-pei Liu, PhD 6 Definitions Shelf-life (expiration dating period) Time interval during which a drug product is expected to remain within the approved shelf life, provided that it is stored under the conditions defined on the container label. Expiration date The date placed on the container label of a drug product designating the time prior to which a batch of the product is expected to remain within the approved shelf life, if stored under defined conditions, and after which it must not be used. Copyright by Jen-pei Liu, PhD 7 Definitions Drug product The dosage form (tablet, capsule, solution, cream, etc) in the final immediate packaging intended for marketing Primary batch A batch of a drug product used in a formal stability study, from which stability data are submitted in a registration application for the purpose of establishing a shelf life Container closure system The sum of packaging components that together contain and protect the dosage form Copyright by Jen-pei Liu, PhD 8 Definitions Quantitative chemical attributes Chemical attributes that can be quantitatively measured such as assay, degradation products, preservative content, etc. The values of quantitative attributes at all times points should be reported as measured such as assay as percent of label claim. Drug product are formulated with intent to provide 100% of the label amount of the drug substance at the time of batch release. Copyright by Jen-pei Liu, PhD 9 How to determine shelf-life FDA guideline (1987) P.31 “ An acceptable approach for drug characteristics that are expected to decrease with time is to determine the time at which the 95% one-sided lower confidence limit … for mean degradation curve intersects the acceptable lower specification limit.” P. 32 “ … , we may be 95% confident that the average drug product characteristic (e.g., strength) of the dosage units in the batch is within specifications up to the end of the expiration dating period.” PhD Copyright by Jen-pei Liu, 10 How to determine shelf-life ICH Q1A(R2) guidance (2003) P.16 “ An approach for analyzing data of quantitative attribute that is expected to change with time is to determine the time at which the 95% one- sided confidence limit for the mean curve intersects the acceptance criterion” ICH Q1E guidance (2004) p.11 A two-sided 95% confidence interval or 95% one-sided upper or lower confidence interval can be also used. One-sided lower limit: known degradation One-sided upper limit: known impurities Two-sided interval: unknown situation about increase or decrease of the assay with the time Copyright by Jen-pei Liu, PhD 11 How to determine shelf-life？ degradation curve 95% lower % of label confidence bound claim lower specification limit 0 3 6 9 12 storage time (month) Copyright by Jen-pei Liu, PhD 12 FDA Guideline FDA stability guideline (1987) Guideline for Submitting Documentation for the Stability of Human Drugs and Biologics Purposes - To provide recommendations for the design and analysis of stability studies to establish an expiration dating period and product requirements - To provide recommendations for the submission of stability information and data to FDA for investigational and new drug applications and product license applications Copyright by Jen-pei Liu, PhD 13 Basic requirements Protocol Stability design - Factorial designs - Matrixing or bracketing design Representative sample - Individual dosage units - Containers within a batch - Batches Statistical analysis - Validity - Accuracy Copyright by Jen-pei Liu, PhD 14 - Reliability Batch sampling consideration At least three primary batches and preferably more should be tested Justification - To allow for some estimates of batch-to- batch variability - To test the hypothesis that a single expiration dating period for all future batches is justifiable Copyright by Jen-pei Liu, PhD 15 Batch-to-batch variation Common intercept Different intercepts Common slope Common Slope Common intercepts Different intercepts Different slopes Different slopes Copyright by Jen-pei Liu, PhD 16 Container, drug product, and sampling-time consideration Container and drug product sampling - At least as many containers should be sampled as the number of sampling times - Sampling of at least two containers for each sampling time is encouraged Sampling-time consideration - 3-month intervals during the first year, 6-month intervals during the second year, and annually thereafter, e.g., for a 4-year duration, 0, 3, 6, 9, 12, 18, 24, 36, and 48 Copyright by Jen-pei Liu, PhD 17 International Harmonization Background - Different regulatory requirement in different areas - May have to repeat similar studies for different marketplaces - Strong industrial/regulatory interest for Harmonization of requirements Copyright by Jen-pei Liu, PhD 18 International Harmonization Least stable batch - EC least stable batch - US minimum of individual shelf-lives Least protective packaging - US sampling of at least two containers each packaging material for each sampling time in all cases - Japan prefer using the least stable packaging material instead of testing the product in all packages Replicates - US testing an increased number of replicates at later sampling times, particularly at latest sampling time - Japan each test must be repeated three times without Copyright by provision for scientific andJen-pei Liu, PhD justification statistical 19 International Harmonization Minimum duration of stability testing - EC at least six months - US a minimum of 12 months - Japan also 12 months Minimum number of batches - EC only two batches - US at least three batches - Japan also thee batches Definition of room temperature - EC between 15o to 25oC - USP/NF between 15o to 30oC - Japan between 1o to 30oC Extension of shelf-life - EC does not accept an extension of shelf-life beyond real-time data submitted - US accepts an extension of shelf-life based on limited data. However, Copyright by Jen-pei Liu, PhD 20 stability result up to the expiration date must be submitted International Harmonization International Conference on Harmonization (ICH) - European Community - Japan (Ministry of Health and Welfare) - U.S. (Food and Drug Administration) Workshop on stability, November 5-7, 1991 Brussels, Belgium Expert Working Group (EWG) developed a tripartite guideline on stability ICH Stability Guideline issued in 1993 - The Tripartite Guideline for the Stability Testing of New Drug Substances and Products Copyright by Jen-pei Liu, PhD 21 International Harmonization The information on stability generated in any of the three areas of the European Community, Japan and the U.S. would be mutually acceptable in both of the other two areas. Test conditions are based on an analysis of the effects of climatic conditions in the three areas - Mean kinetic temperature can be derived from climatic data EWG meeting held in Washington DC recommended a three-year phase-in period - ICH requirements must be met in January 1997 Copyright by Jen-pei Liu, PhD 22 International Harmonization Testing frequency (Sampling times) - Long-term storage conditions (LSC) For a product with a proposed shelf life of at least 12 months at the time of submission - 0, 3, 6, 9, 12, 18, 24, 36, 48, … - Accelerated storage conditions (ASC) a minimum of 3 time points, including the initial and final time points (0, 3, 6 months) - Immediate storage conditions (ISC) – a significant change from ASC a minimum of 4 time points, including the initial and final time points (0, 3, 6, 12 months) Number of batches - At least three primary batches Copyright by Jen-pei Liu, PhD 23 International Harmonization Storage Conditions (LSC) Study Storage conditions Minimum time period covered by data at submission Long-term 25oC2oC/60%RH5%RH 12 months or 30oC2oC/65%RH5%RH Immediate 30oC2oC/65%RH5%RH 6 months Accelerated 40oC2oC/75%RH5%RH 6 months Copyright by Jen-pei Liu, PhD 24 International Harmonization Storage Conditions (LSC) Study Storage conditions Minimum time period covered by data at submission Refrigerator Long-term 5oC3o 12 months Immediate 25oC2oC/60%RH5%RH 6 months Freezer Long-term -20oC5oC 12 months Copyright by Jen-pei Liu, PhD 25 International Harmonization A Significant change A 5% change in assay Degradation exceeding its acceptance criterion Failure to meet acceptance criteria or appearance, physical attributes, functionality test Failure to meet acceptance criteria for pH or dissolution for 12 dosage units Statistical analysis - Determination of drug shelf-life - Goodness of fit of the data on all batches and combined batches Chow, S.C. and Liu, J.P. (1995). Statistical Design and Analysis in Pharmaceutical Science. pp 584, Marcel Dekker, Inc., New York, New York. Chow, S.C. and Pong, A. (1995). Current issues in regulatory requirements of drug Copyright by Jen-pei Liu, PhD 26 stability. Submitted to Journal of Food and Drug Analysis for publication. Current Issues Batch similarity - Data pooling - Random batches？ Design issues - Design factors - Response variables - Matrixing/breaking designs Interval estimate - Underestimate or overestimate？ - The bias of an estimates shelf-life Frozen drug products - Several storage temperatures - Multiple-phase stability study Other issues Copyright by Jen-pei Liu, PhD 27 Batch similarity FDA guideline P.35 “Combining the data should be supported by preliminary testing of batch similarity. The similarity of the degradation curves … by apply statistical test if the equality of slopes and of zero time intercepts.” “Bancroft recommended … a level of significance of 0.25 … the data from batches would be pooled.” ICH Q1E guidance p. 12 “ANCOVA can be employed … a significance level of 0.25….” Copyright by Jen-pei Liu, PhD 28 Batch similarity Comments - Wrong hypothesis for similarity “Similarity” = “Equality” “Similarity” = “Equivalence” ？ - Failure of rejection of the null hypothesis of equality does not prove equality of slopes - Why 0.25？ Penalize good studies？ References Chow, S.C. and Shao, J. (1989). Test for batch-to batch variation in stability analysis. Statistics in Medicine, 8, 883-890. Copyright by Jen-pei Liu, PhD 29 Batch similarity Copyright by Jen-pei Liu, PhD 30 Batch similarity Table 11.4.1 ANCOVA Table for Data Set 1 Source of variation df Sum of Squares F Statistic P value Intercepts 5 0.60 2.85 0.038 Time 1 21.52 503.24 <0.001 Different in Slopes 5 0.35 1.64 0.186 Error 25 1.07 Source: Ruberg and Stegeman (1991). Ruberg, S. and Stegeman, J.W. (1991). Pooling data for stability studies: testing the quality of batch degradation slopes. Biometrics, 47, 1059-1069 Copyright by Jen-pei Liu, PhD 31 Batch similarity Copyright by Jen-pei Liu, PhD 32 Batch similarity Table 11.4.2 ANCOVA Table for Data Set 2 Source of variation df Sum of Squares F Statistic P value Intercepts 5 15.93 4.01 0.009 Time 1 4.14 5.21 0.032 Different in Slopes 5 4.60 1.16 0.359 Error 25 18.28 Source: Ruberg and Stegeman (1991). Copyright by Jen-pei Liu, PhD 33 Penalize good studies？ Table 11.4.1 ANCOVA Table for Data Set 1 Source of variation df Sum of Squares F Statistic P value Intercepts 5 0.60 2.85 0.038 Time 1 21.52 503.24 <0.001 Different in Slopes 5 0.35 1.64 0.186 Error 25 1.07 Source: Ruberg and Stegeman (1991). Do not pool the data Table 11.4.2 ANCOVA Table for Data Set 2 Source of variation df Sum of Squares F Statistic P value Intercepts 5 15.93 4.01 0.009 Time 1 4.14 5.21 0.032 Different in Slopes 5 4.60 1.16 0.359 Error 25 18.28 Source: Ruberg and Stegeman (1991). Pool the data Copyright by Jen-pei Liu, PhD 34 Batch similarity FDA guideline P.36 “… the overall expiration dating period may depend on the minimum time a batch may be expected to remain within acceptable limits.” P.29 “The expiration period should be applied to all future batches …” Comments - The minimum approach lacks statistical justification - Batch should be treated as a random variable References Chow, S.C. and Shao, J. (1991). Estimating drug shelf-life with random batches. Biometrics, 47,1071-1079. Shao, J. and Chow, S.C. (1994). Statistical inference in stability. Biometrics, 50, 753-763 by Jen-pei Liu, PhD Copyright 35 Design Factors Strength Package type Batch or lot Sampling (or storage) times Storage conditions Replicates Analyst Location : etc. Copyright by Jen-pei Liu, PhD 36 Response Variables Potency Dissolution Appearance Hardness Color Moisture : etc. Copyright by Jen-pei Liu, PhD 37 Design objective Minimize total sample size Reduce number of assays - limited laboratory capacity - limited resources Have valid statistical results Comments Take fraction of combinations of package by batch by strength and/or sample chosen combination at a portion of time points Reference Nordbrock, E. (1992). Statistical comparison of stability Copyright by Statistics, study designs. J. Biopharm.Jen-pei Liu, PhD 2, 91-113. 38 Matrixing design Definition Any subset of a full factorial design is considered a matrix design Example Table 10.3.2 Two-Thirds Matrixing Design Dosage strength/lot of granulation Package 50 mg 75 mg 100 mg type A B C A B C A B C Blister + + (+) (+) + + + (+) + HDPE1 (+) + + + (+) + + + (+) HDPE2 + (+) + + + (+) (+) + + Source: Helboe (1992). Copyright by Jen-pei Liu, PhD 39 (+), not tested at this time point. Bracketing design Definition The design of a stability schedule such that at any time point only the samples on the extremes of container size and/or dosage strengths are tested Example TABLE 10.3.3 Example of Bracketing Design Storage Testing interval (month) Temp Relative 3 6 9 12 18 24 36 48 (oC) humidity (%) 21 45 (+) (+) (+) (+) (+) (+) (+) (+) 25 60 + + + + + + + + 30 35 (+) (+) (+) (+) (+) (+) (+) (+) 30 70 + + + + + + + + Copyright by corresponding The parentheses indicates that theJen-pei Liu, PhD stability tests are omitted.40 Matrixing and bracking designs Remarks - A matrixing or bracketing design may be applicable to strength if there is no change in proportion of active ingredients, container size, and intermediate sampling time points (Liu, 1994) - A matrixing design is not recommended if there is a significant change in proportions of active ingredients. - A matrixing design should not be applied to sampling times at two endpoints (i.e., the initial and the last sampling time points) and at any time points beyond the desired expiration date - A matrixing design should be avoided if the drug product is sensitive to temperature, humidity, and light Copyright by Jen-pei Liu, PhD 41 Interval estimate Is it safe to take the drug product when it is just expired？ Does the labeled shelf-life overestimate or underestimate the true shelf-life FDA might prefer a conservative approach which is to underestimate rather than overestimate the true shelf-life It is suggested the bias of an estimated drug shelf-life obtained according to the method suggested by the FDA be evaluated. Is an interval estimate for drug expiration dating period more appropriate than the point estimate？ Reference Sun, Y., Chow, S.C., Li., G. and Chen, K.W. (1995). Assessing distribution of estimated drug shelf-life in Copyright by Jen-pei publication stability analysis. Submitted for Liu, PhD 42 bx a η b 2G2 (bx a η) 2G1 k ~ x t b b2 1 a ˆ k (a a) ˆ 2 k (b b) o(1) b b Hence ~ a d b 2G2 (bx a ) 2 G1 N 2 kt , b b2 where ~ a e2 e2 2 x 2 b2 j 1 ( x j x ) n b 2 nb 2 t e 2 2 G1 j 1 n ( x j x )2 t e 2 2 G2 n ˆ t e 2 ˆ2 G1 j 1 ( x j x ) 2 n ˆ t e ˆ2 2 Copyright by Jen-pei Liu, PhD 43 G2 n Frozen drug products Frozen products usually stored at several temperature such as –20oC, 5oC, and 25oC A typical shelf-life statement for frozen drug products may be - 24 months at –20oC followed by either one month at 5oC or 2 days at 25oC - 24 months at –20oC followed by two weeks at 5oC or 1 days at 25oC The drug shelf-life is determined based on two-phase stability study - Frozen study – to estimate shelf-life under frozen storage condition - Thawed study – to estimate shelf-life under refrigerated or ambient conditions Copyright by Jen-pei Liu, PhD 44 Copyright by Jen-pei Liu, PhD 45 Other issues Shelf-life determination for discrete response variables - particle size - color Alternative methods - Inverse method Stability for multiple active ingredients - Premarin (estrone, equilin and 17 -dihydroquilin) References Chow, S.C. and Ju, H. (1995). The application of biopharmaceutical statistics in drug development. J. Chinese Statistical Association, 33(2), 1-23. Chow, S.C. and Liu, J.P. (1995). Statistical Design and Analysis in Pharmaceutical Science. Marcel Dekker, Inc., New York, Copyright by Jen-pei Liu, PhD 46 New York.