Conference Schedule

8:15 – 9:00
Registration and Continental Breakfast

9:00 – 9:10
Introductory Remarks
Boris Iglewicz, Temple University

9:10 – 9:50
Fitting Ill- conditioned Data Sets in Censored Data Regression: Shrinking, Transforming, and Repairing Stepwise Fitting
David Harrington, Harvard University

9:50 – 10:30
Innovations in Clinical Trial Design and Analysis
Donald Berry, M.D.Anderson Cancer Center

10:40 – 10:40

10:40 – 11:00

11:00 – 11:40
The Analysis of Case-Control Data to Detect Candidate Genes
Robert Elston, Case Western Reserve University

11:40 – 12:20
Data Mining, Clustering, and Robust Partial Mixture Estimation
David W. Scott, Rice University

12:20 – 12:30

12:30 – 1:40

1:40 – 1:50
What is SPAIG?
Robert Starbuck, Wyeth Research

1:50 – 2:20
Can Statisticians Have A Higher Impact on Drug Discovery and Development Process?
Frank Shen, Bristol-Myers Squibb

2:20 – 2:30
Bayes and Biopharmaceutics?
Lawrence Gould, Merck Research Laboratories

2:30 – 2:40
Molecular Profiling in the Pharmaceutical Industry: A statistician’s view
Daniel Holder, Merck Research Laboratories

2:40 – 2:50
Statistical Challenges for Vaccine Clinical Trials
Devan Mehrotra, Merck Research Laboratories

2:50 – 3:10
Frank Rockhold, GSK; Jerry Schindler, Wyeth; Joseph Heyse, Merck

3:10 – 3:25
Floor Discussion


Fitting lll-conditioned Data Sets in Censored Data Regression:
Shrinking, Transforming, and Repairing Stepwise Fitting

David Harrington, Harvard University
(Joint with Jie Huang and Chang-Heok Soh)

In linear regression, variance inflation and increased estimation or prediction error have been well studied when predictors are either highly correlated or the number of candidate predictors is large compared to the number of cases. This talk discusses analysis methods that can be used when the same situation arises with right-censored data. These extensions include the penalized partial likelihood in proportional hazards model, extensions of partial least squares to linear regression with censored data, and resampling methods to adjust coefficients estimated in model selection methods, such as stepwise regression. The methods all use resampling methods for censored data either to choose shrinkage parameters, select linear transformations, or correct stepwise estimates. The methods will be illustrated with applications to multiple myeloma and HIV.

Innovations in Clinical Trial Design and Analysis

Donald Berry, M.D. Anderson Cancer Center

I will describe recent Bayesian innovations in the design and analysis of clinical trials. The goals are (i) more efficient clinical trials and clinical development programs, and (ii) treating patients more effectively, both those in and those outside of clinical trials. These innovative trial designs have been effected at my home institution, in national oncology studies and in pharmaceutical and medical device industry-sponsored trials. The revolutionary aspect is hardly earth shattering in non-medical science: we pay heed to the accumulating data and let it guide the course of the trial! I will provide some background on the Bayesian approach and give case studies showing its use in actual designs and in analyses presented to the FDA. These examples include the possibility of early stopping, assigning patients to better performing therapies, and variations on themes such as seamless phases II and III trials with sequential sampling and using early endpoints. The savings of such an approach in terms of effectively using patient resources are substantial. In addition, within the context of some of the examples I will address the role of statistical decision analysis in the pharmaceutical industry.

The Analysis of Case-Control Data to Detect Candidate Genes

Robert C. Elston, Case Western Reserve University

Case-control data have historically been used to detect environmental effects causing disease, but are now being used to detect genetic effects as well, typing the cases and controls for candidate gene loci. For each candidate gene the resulting measure on each subject is a genotype, comprising two alleles. Because the alleles within a person are typically not independent, the analysis poses both difficulties and opportunities. Armitage’s trend test can be used to determine if a particular allele is more frequent among cases than controls, and a test for departure from Hardy-Weinberg equilibrium proportions can be used to provide a further indication that the typed locus marks a genomic region where a disease susceptibility gene lies. These two pieces of information are pooled to obtain a powerful test that better maintains validity in the presence of population structure/heterogeneity.

Data Mining, Clustering, and Robust Partial Mixture Estimation

David W. Scott, Rice University

Mining large datasets successfully requires careful application of statistical modeling tools. Such data are seldom clean and robust statistical methods are especially appropriate to automatically cope with large numbers of outliers. We discuss in particular robust normal mixture estimation. Such models are useful for clustering by associating a cluster with each component of the mixture model. Finally, we present a number of examples including simple regression, robust covariance estimation, incomplete model specification, lightning detection, particle physics detection, and finding the largest eigenvector in a mixture dataset.

Can Statisticians Have A Higher Impact on Drug Discovery and Development Process?

Frank Shen, Bristol-Myers Squibb

Drug Discovery to Development is a ten-year bet industry-wide. Discovery, which includes target identification through clinical proof of principle, accounts for half that time. In the old world, companies were profitable as long as they had occasional winners. But the new world has become much more competitive and tougher. While the development cost for each NCE climbed to at least 800 million, the U.S. Food and Drug Administration approved only 21 new chemical entities (NCEs) last year, marking a steady decline since a peak of 53 in 1996. The process of generating a steady pipeline requires much more predictability and productivity. Companies have rushed to find their solutions by reorganization, consolidation, licensing, and outsourcing. This talk will shed some light on driving forces that change our working environment today and in the future, and share some personal thoughts on how statisticians can “step out of their box” and take more ownership to have higher impact.


David Harrington has been the Chair of the Department of Biostatistical Science at the Dana-Farber Cancer Institute since 1998 and is currently Professor of Biostatistics in the Department of Biostatistics at the Harvard School of Public Health. He received his doctoral degree in statistics from the University of Maryland in 1976, and has held academic appointments at the University of Virginia and the Harvard School of Public Health. Dr. Harrington conducts statistical research in clinical trials, in survival analysis and in longitudinal data arising in clinical and observational studies. He has worked collaboratively in interdisciplinary research in lymphoma, leukemia, lung cancer, and myeloma. He is also the coauthor of the Wiley book, “Counting Processes & Survival Analysis.”

Professor Harrington served as the Principal Statistician for the Eastern Cooperative Oncology Group from 1990 to 2000, an organization of approximately 300 treatment sites conducting clinical and basic research in all adult malignancies. He is currently the Director of the Biostatistics Core Facility for the Dana-Farber/Harvard Cancer Center, a consortium of Harvard Medical School teaching affiliates, academic departments and laboratories with more than 750 investigators directly involved in cancer research. He is also the principal investigator for the Statistical Coordinating Center for the Cancer Care Outcomes Research and Surveillance Consortium (CanCORS), a recently funded network of cancer registries and cancer centers organized to study barriers to effective cancer care in the US populations.

Dr. Harrington is a fellow of the American Statistical Association, the Institute of Mathematical Statistics, and an elected member of the International Statistical Institute.

Donald Berry holds the Frank T. McGraw Memorial Chair for Cancer Research at The University of Texas M. D. Anderson Cancer Center, where he is chairman of the Department of Biostatistics and Applied Mathematics. In addition, he serves as the faculty statistician on the Breast Cancer Committee of the Cancer and Leukemia Group B (CALGB), a national oncology group. A native of Massachusetts, Dr. Berry received his Ph.D. in statistics from Yale University, and previously served on the faculty at the University of Minnesota and at Duke University, where he held the Edger Thompson Professorship in the College of Arts and Sciences. The author of more than 200 published articles as well as several books on biostatistics in medical research, Dr. Berry has been the principal investigator for numerous medical research programs funded by the National Institutes of Health and the National Science Foundation. Dr. Berry is a statistics editor for the Journal of the National Cancer Institute and associate editor of Breast Cancer Research and Treatment and also of Clinical Cancer Research. He is a Fellow of the American Statistical Association and of the Institute of Mathematical Statistics. Through Berry Consultants, LLC, he and his son, Scott Berry, build innovative designs and analyses for drug and device trials.

Robert C. Elston serves as Professor, Department of Epidemiology and Biostatistics, Case Western Reserve University. He is the Director, Division of Genetic and Molecular Epidemiology at Case Western Reserve University. Professor Elston received his Ph.D. from Cornell University in 1959, was on the faculty of the University of North Carolina, Chapel Hill from 1960-1962 and 1964-1979, Head of the Department of Biometry and Genetics at the Louisiana State University Medical Center from 1979-1995, and joined Case Western Reserve University in 1995. Professor Elston is a Fellow of the American Statistical Association and the Institute of Mathematical Statistics. He served as President of the International Genetic Epidemiology Society and was the lead invited speaker at the 2000 grand opening of the new Sir Henry Wellcome Building of Genomic Medicine, Oxford University.

David W. Scott is a Noah Harding Professor of Statistics at Rice University. He is the past-editor of the Journal of Computational and Graphical Statistics and author of the Wiley text “Multivariate Density Estimation.” He is a member of the National Academy’s Committee on Applied and Theoretical Statistics. He is an elected member of the IMS Council and Fellow of the ASA, IMS, and AAAS. His research interests include nonparametric density estimation, data visualization, statistical computing, and applications.

Dr. Frank Shen serves as Executive Director of Exploratory Development, Global Biostatistics and Programming at Bristol-Myers Squibb Co. Frank received his B.S. and M.S. degrees in Chemical Engineering from Chung-Yuan University (Taiwan) and his Ph.D. in statistics from Temple University. He is a Fellow of the American Statistical association, President of the International Chinese Statistical association, and has served as ViceChair of Biostatistics and Data Management Technical Group of US Pharmaceutical Research and Manufacturing Association (PhRMA).

A. Lawrence Gould is Senior Director, Scientific Staff, Biostatistics and Research Data Systems, Merck Research Laboratories. He received his Ph.D. in Biometry from Case Western Reserve University. Dr. Gould is a Fellow of the American Statistical Association, has served on a number of grant review panels, and has served in a variety of positions with the Biopharmaceutical Section of the ASA and the Biometric Society, ENAR. He served as Editor of the Journal of Biopharmaceutical Statistics, received the 1991 Donald Francke Award for most outstanding article published in DIA Journal that year, and the 1994 best presentation award from the Biopharmaceutical Section of ASA. His research interests tend to be driven by problems arising in drug development, and include blinded sample size re-estimation, Bayesian methods, meta-analysis, bioequivalence, analysis of safety data, data mining, clinical trial simulation, and management science.

Daniel Holder is a Director in the Biometrics Research department of Merck Research Laboratories. He earned an MS in Statistics from the University of Wisconsin-Madison in 1988, and a PhD in Statisics from Temple University in 1993. He worked in Phase I/II clinical biostatistics at Merck before moving to the preclinical area. In the clinical area his main focus was on PK models including bioequivalence. Since moving to the preclinical area, he has concentrated on molecular profiling, including the genetic basis for HIV resistance, gene expression assays, and proteomics. He has authored or co-authored over 50 publications including articles in Biometrika, the Journal of the American Statistical Association, the Journal of the American Medical Association, the New England Journal of Medicine, and Neuroscience.

Devan V. Mehrotra has worked as a biostatistician in the pharmaceutical industry for over 13 years, and is currently Director, Clinical Biostatistics at Merck Research Laboratories. Devan received is Ph.D. in statistics from University of Delaware. He has published articles on a diverse array of topics, and is frequently invited to present at conferences and in academia. He holds adjunct faculty appointments at the University of Pennsylvania and Villanova University. He is an Associate Editor for the Journal of Biopharmaceutical Statistics and the Biometrical Journal. His other professional activities have included serving as President and Vice President of the Philadelphia Chapter of the American Statistical Association, serving on the steering committee of the annual FDA/Industry Statistics Workshop, and serving as a consultant to the Center for Scientific Review at the National Institutes of Health.


This event has passed.


640 W. Germantown Pike, Plymouth Meeting PA 19462
(610) 834-8300

From Airport: Take 95 South to 476 North to the last exit #20(Germantown Pike-West). Merge with Germantown Pike and follow for 3 lights. Make a right onto Hickory Rd. at the 3rd light. The hotel is the 3rd building on the left.

New York/ New Jersey Turnpike: Take the New Jersey Turnpike to exit #6, which is PA turnpike. Go west to exit #333- Norristown. Follow signs to Plymouth Rd. Go to the 1st light and make a left. Go to the next light and make a right onto Germantown Pike. Go to the second light and make a right on Hickory Rd. The hotel is the second driveway on the left.

Washington D.C., Wilmington, and Delaware: Take I-95 North to Route 476 North. Take Route 476 to the Germantown Pike West exit #20. Go to the third light, Hickory Rd., and make a right. The hotel is the 2nd driveway on the left.

Route 476: Take 476 to the Germantown Pike West exit #20. Go to the third light, Hickory Rd., and make a right. The hotel is the 2nd driveway on the left.

From downtown Philadelphia: I-76 west Plymouth Meeting exit #331B (Route 476). Take Route 476 north to Germantown Pike exit. Go to the third light, Hickory Rd., and make a right. The hotel is the 2nd driveway on the left., Temple University