See discussions, stats, and author profiles for this publication at: [Link]

net/publication/247530172

Recent Issues in Stability Study

Article  in  Journal of Biopharmaceutical Statistics · January 2003

DOI: 10.1081/BIP-120022758

CITATIONS READS
5 387

1 author:

Yi Tsong
U.S. Food and Drug Administration
148 PUBLICATIONS   3,224 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Development of Statistical Methods for Analytical Similarity Assessment View project

Family Health Project View project

All content following this page was uploaded by Yi Tsong on 08 March 2016.

The user has requested enhancement of the downloaded file.

 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

JOURNAL OF BIOPHARMACEUTICAL STATISTICS

Vol. 13, No. 3, pp. vii–ix, 2003

INTRODUCTION

Recent Issues in Stability Study

Yi Tsong*

Office of Biostatistics, Office of Pharmacoepidemiology and Statistics Sciences,

Center for Drug Evaluation and Research, Food and Drug Administration,
Rockville, Maryland, USA

The stability of a drug substance or drug product is the capacity of the drug substance or
drug product to remain within the established acceptance criteria to ensure its identity,
strength, quality, and purity during a specified period of time. The Food and Drug
Administration (FDA) requires that adequate testing be performed by the applicant to
demonstrate the stability of the drug substance or product in support of the approval of a
new drug application (NDA), abbreviated new drug application (ANDA), biologics
license application (BLA), product license application (PLA), and certain postapproval
supplemental changes. 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
environmental factors, such as temperature, humidity, and light, and to establish the
recommended storage conditions and a shelf life for the drug product. A shelf life (also
referred to as expiration dating period) is the period of time during which a drug product is
expected to remain within the approved shelf life specification, provided that the drug
product has been stored under the conditions defined on the container label. After this
period, a drug product batch must be discarded.
Application of statistical expertise is needed in designing the experiment; modeling
the changes of the values of the assay, purity, and chemical content; estimating the
parameters of interest; and determining the shelf life. The foundation of the modeling can
be traced back to Bancroft’s article on preliminary test as early as 1944 (Bancroft, 1944).
For decades, a single-factor design for a given drug product with multiple batches was

*Correspondence: Yi Tsong, Quantitative Methods and Research Staff, Office of Biostatistics,

Office of Pharmacoepidemiology and Statistics Sciences, Center for Drug Evaluation and Research,
Food and Drug Administration, Rockville, Maryland, USA; E-mail: tsong@[Link].

vii

DOI: 10.1081/BIP-120022758 1054-3406 (Print); 1520-5711 (Online)

Copyright q 2003 by Marcel Dekker, Inc. [Link]
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

viii Tsong

used as the standard in stability study. It was the only study design covered in the 1987
FDA Guidelines for Submitting Documentation for the Stability of Human Drugs and
Biologics (FDA, 1987). Complicated design such as multiple-factors, bracket, and matrix
designs were introduced to the pharmaceutical community in United States in the early
1990s. It is interesting to note that articles on statistical modeling of these multiple-factor
designs were limited in the literature. One of the explanations is that, in practice, such
designs were only seldom applied to New Drug Application (NDA) data in the United
States until the late 1990s. In contrast to the 1987 FDA Guidance, the International
Conference on Harmonization (ICH) Guidance (ICH, 1994; 2003a,b) on Stability Study
developed recently, a variety of multiple factor designs became one of the major focal
points. Correspondingly, a statistical working group on stability issue was formed in the
FDA in 2002 to identify and address the design and analysis issues associated with the
multiple factors designs. Those issues also motivated us to propose the Journal of
Biopharmaceutical Statistics for this special issue on stability study.
This special issue consists of nine important papers covering designs, analysis of
stability data, and postmarketing stability monitoring. The members of the FDA CDER
Statistical Working Group on Stability Study contributed four of the papers. This working
group was formed to identify and address the issues raised in developing the ICH
Guidance. Drs. Daphne Lin and Chi-wen Chen’s paper “Overview of Stability Designs”
covers the various designs of stability study that were the focus of many stability papers in
the last 10 years. In Drs. Wen Jen Chen and Yi Tsong’s paper “Significance Levels of
Stability Pooling Test: A Simulation study,” they addressed the issue of choice of
significance levels in stability study. They reiterate the fact pointed out by Bancroft in
1944 that the slope or intercept pooling tests in stability study are the preliminary tests.
Hence, the significance level of the pooling test should be set in order to control the type I
error rate of the primary test.
In comparison to the number of articles published on multiple factor stability designs,
the number on the analysis of multiple factor design is small. Dr. Yi Tsong et al.’s paper
“ANCOVA Approach for Shelf Life Analysis of Stability Study of Multiple Factor
Designs” pointed out the difficulties of extending the single stage analysis of covariance
(ANCOVA) modeling approach using the type I sum of squares to multiple factor design.
Hence, they proposed the approach of stepwise ANCOVA modeling using type I sum of
squares. Alternative approaches based on equivalence assessment were discussed in Dr. Yi
Tsong et al.’s paper “Shelf Life Determination Based on Equivalence Assessment.” For
nonnormal stability measurements, Drs. Shein Chung Chow and Jun Shao contributed a
paper “Stability Analysis with Discrete Responses” and Dr. Ying Qing Chen et al.
contributed a paper “Rank Regression in Stability Analysis” in the special issue.
Although postmarketing stability monitoring is as important as premarketing study,
there were few articles on postmarketing stability surveillance published in the literature.
Fortunately, we included two articles on this subject in the special issue. In Dr. William
Fairweather’s paper “A Quantitative Assessment of Factors Influencing the Probability of
Post-Marketing Out-of-Specification Observations,” the author examined the relationship
between the probability of out-of-specification sample and the influencing factors. He also
provided recommendations on actions that can be taken to reduce the probability.
Dr. William Fairweather et al.’s paper “Monitoring the Stability of Human Vaccines” is
unique in its focus on postmarketing monitoring of the stability of vaccine product.
 MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016
©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.

Recent Issues in Stability Study ix

Also included in this special issue is Drs. Stan Altan and Damaraju Raghavarao’s
paper “A Note on Kinetic Modeling of Stability Data and Implications on Pooling.” They
discussed the relationship between the underlying kinetic model and the statistical model
used to study degradation.
We wish to thank the authors for their contributions to make this special issue a
successful one.
Finally, we wish to thank the referees, Dr. Peter Lachbruch of Center for Biologics
Evaluation and Research (CBER), FDA, Drs. Musfiqur Rashid and Wen Jen Chen of
Center for Drug Evaluation and Research (CDER), FDA, Dr. William Fairweather of
Flower Valley Consultants, Prof. Jen Pei Liu of Cheng-Kong University of Taiwan,
Republic of China, for their careful reading and comments that lead to the improvement of
the presentation of these papers; to Dr. Shein Chung Chow, the Chief Editor of the Journal
of Biopharmaceutical Statistics for giving me the opportunity to put together this
important special issue and his encouragement in the process.

REFERENCES

Bancroft, T. A. (1944). On biases in estimation due to the use of preliminary tests of

significance. Annal. Math. Stat. 15:190 – 204.
FDA. (1987). Guidelines for Submitting Documentation for the Stability of Human Drugs
and Biologics. Rockville, MD: USA Food and Drug Administration, Center for Drugs
and Biologics.
ICH. (1994). Q1A(R) Stability Testing of New Drug Substances and Products. Geneva,
Switzerland: Conference on Harmonization. ([Link]
[Link]).
ICH. (2003a). Q1D Bracketing and Matrixing Designs for Stability Testing of New Drug
Substances and Products. Geneva, Switzerland: Conference on Harmonization.
([Link]
ICH. (2003b). Q1E Stability Data Evaluation. Geneva, Switzerland: Conference on
Harmonization. ([Link]

View publication stats