Shein-Chung Chow

Adaptive design methods in clinical trials – a review

In recent years, the use of adaptive design methods in clinical research and development based on accrued data has become very popular due to its flexibility and efficiency. Based on adaptations applied, adaptive designs can be classified into three categories: prospective, concurrent (ad hoc), and retrospective adaptive designs. An adaptive design allows modifications made to trial and/or statistical procedures of ongoing clinical trials. However, it is a concern that the actual patient population after the adaptations could deviate from the originally target patient population and consequently the overall type I error (to erroneously claim efficacy for an infective drug) rate may not be controlled. In addition, major adaptations of trial and/or statistical procedures of on-going trials may result in a totally different trial that is unable to address the scientific/medical questions the trial intends to answer. In this article, several commonly considered adaptive designs in clinical trials are reviewed. Impacts of ad hoc adaptations (protocol amendments), challenges in by design (prospective) adaptations, and obstacles of retrospective adaptations are described. Strategies for the use of adaptive design in clinical development of rare diseases are discussed. Some examples concerning the development of Velcade intended for multiple myeloma and non-Hodgkins lymphoma are given. Practical issues that are commonly encountered when implementing adaptive design methods in clinical trials are also discussed.

Extended Valganciclovir Prophylaxis to Prevent Cytomegalovirus After Lung Transplantation: A Randomized, Controlled Trial

BACKGROUND Cytomegalovirus (CMV) is the most prevalent opportunistic infection after lung transplantation. Current strategies do not prevent CMV in most at-risk patients. OBJECTIVE To determine whether extending prophylaxis with oral valganciclovir from the standard 3 months to 12 months after lung transplantation is efficacious. DESIGN Randomized, clinical trial. Patients were randomly assigned by a central automated system to treatment or placebo. Patients and investigators were blinded to treatment status. (ClinicalTrials.gov registration number: NCT00227370) SETTING Multicenter trial involving 11 U.S. lung transplant centers. PATIENTS 136 lung transplant recipients who completed 3 months of valganciclovir prophylaxis. INTERVENTION 9 additional months of oral valganciclovir (n = 70) or placebo (n = 66). MEASUREMENTS The primary end point was freedom from CMV disease (syndrome or tissue-invasive) on an intention-to-treat basis 300 days after randomization. Secondary end points were CMV disease severity, CMV infection, acute rejection, opportunistic infections, ganciclovir resistance, and safety. RESULTS CMV disease occurred in 32% of the short-course group versus 4% of the extended-course group (P < 0.001). Significant reductions were observed with CMV infection (64% vs. 10%; P < 0.001) and disease severity (110 000 vs. 3200 copies/mL, P = 0.009) with extended treatment. Rates of acute rejection, opportunistic infections, adverse events, CMV UL97 ganciclovir-resistance mutations, and laboratory abnormalities were similar between groups. During the 6 months after study completion, a low incidence of CMV disease was observed in both groups. LIMITATION Longer-term effects of extended prophylaxis were not assessed. CONCLUSION In adult lung transplant recipients who have received 3 months of valganciclovir, extending prophylaxis by an additional 9 months significantly reduces CMV infection, disease, and disease severity without increased ganciclovir resistance or toxicity. A beneficial effect with regard to prevention of CMV disease seems to extend at least through 18 months after transplantation.

Adaptive Design Methods in Clinical Trials

Introduction What Is Adaptive Design Regulatory Perspectives Target Patient Population Statistical Inference Practical Issues Aims and Scope of the Book Protocol Amendment Introduction Moving Target Patient Population Analysis with Covariate Adjustment Assessment of Sensitivity Index Sample Size Adjustment Concluding Remarks Adaptive Randomization Conventional Randomization Treatment-Adaptive Randomization Covariate-Adaptive Randomization Response-Adaptive Randomization Issues with Adaptive Randomization Summary Adaptive Hypotheses Modifications of Hypotheses Switch from Superiority to Noninferiority Concluding Remarks Adaptive Dose-Escalation Trials Introduction CRM in Phase I Oncology Study Hybrid Frequentist-Bayesian Adaptive Design Design Selection and Sample Size Concluding Remarks Adaptive Group Sequential Design Sequential Methods General Approach for Group Sequential Design Early Stopping Boundaries Alpha Spending Function Group Sequential Design Based on Independent P-Values Calculation of Stopping Boundaries Group Sequential Trial Monitoring Conditional Power Practical Issues Statistical Tests for Seamless Adaptive Designs Why a Seamless Design Is Efficient Step-Wise Test and Adaptive Procedures Contrast Test and Naive P-Value Comparisons of Seamless Design Drop-the-Loser Adaptive Design Summary Adaptive Sample Size Adjustment Sample Size Re-Estimation without Unblinding Data Cui-Hung-Wangs Method Proschan-Hunsbergers Method Muller-Schafer Method Bauer-Koehne Method Generalization of Independent P-Value Approaches Inverse-Normal Method Concluding Remarks Two-Stage Adaptive Design Introduction Practical Issues Types of Two-Stage Adaptive Designs Analysis for Seamless Design with Same Study Objectives/Endpoints Analysis for Seamless Design with Different Endpoints Analysis for Seamless Design with Different Objectives/Endpoints Concluding Remarks Adaptive Treatment Switching Latent Event Times Proportional Hazard Model with Latent Hazard Rate Mixed Exponential Model Concluding Remarks Bayesian Approach Basic Concepts of Bayesian Approach Multiple-Stage Design for Single-Arm Trial Bayesian Optimal Adaptive Designs Concluding Remarks Biomarker Adaptive Trials Introduction Types of Biomarkers and Validation Design with Classifier Biomarker Adaptive Design with Prognostic Biomarker Adaptive Design with Predictive Marker Concluding Remarks Appendix Target Clinical Trials Introduction Potential Impact and Significance Evaluation of Treatment Effect Other Study Designs and Models Concluding Remarks Sample Size and Power Estimation Framework and Model/Parameter Assumptions Method Based on the Sum of P-Values Method Based on Product of P-Values Method with Inverse-Normal P-Values Sample Size Re-Estimation Summary Clinical Trial Simulation Introduction Software Application of ExpDesign Studio Early Phases Development Late Phases Development Concluding Remarks Regulatory Perspectives - A Review of FDA Draft Guidance Introduction The FDA Draft Guidance Well-Understood Designs Less Well-Understood Designs Adaptive Design Implementation Concluding Remarks Case Studies Basic Considerations Adaptive Group Sequential Design Adaptive Dose-Escalation Design Two-Stage Phase II/III Adaptive Design Bibliography Subject Index

Invasive Bacterial and Fungal Infections Among Hospitalized HIV-Infected and HIV-Uninfected Adults and Adolescents in Northern Tanzania

BACKGROUND few studies describe patterns of human immunodeficiency virus (HIV) co-infections in African hospitals in the antiretroviral therapy (ART) era. METHODS we enrolled consecutive admitted patients aged ≥ 13 years with oral temperature of ≥ 38.0°C during 1 year in Moshi, Tanzania. A standardized clinical history and physical examination was done and hospital outcome recorded. HIV antibody testing, aerobic and mycobacterial blood cultures, and malaria film were performed. HIV-infected patients also received serum cryptococcal antigen testing and CD4(+) T lymphocyte count (CD4 cell count). RESULTS of 403 patients enrolled, the median age was 38 years (range, 14-96 years), 217 (53.8%) were female, and 157 (39.0%) were HIV-infected. Of HIV-infected patients, the median CD4 cell count was 98 cells/μL (range, 1-1,105 cells/ μL), 20 (12.7%) were receiving ART, and 29 (18.5%) were receiving trimethoprim-sulfamethoxazole prophylaxis. There were 112 (27.7%) patients who had evidence of invasive disease, including 26 (23.2%) with Salmonella serotype Typhi infection, 24 (21.4%) with Streptococcus pneumoniae infection, 17 (15.2%) with Cryptococcus neoformans infection, 12 (10.7%) with Mycobacterium tuberculosis complex infection, 8 (7.1%) with Plasmodium falciparum infection, and 7 (6.3%) with Escherichia coli infection. HIV infection was associated with M. tuberculosis and C. neoformans bloodstream infection but not with E. coli, S. pneumoniae, or P. falciparum infection. HIV infection appeared to be protective against Salmonella. Typhi bloodstream infection (odds ratio, .12; P = .001). CONCLUSIONS while Salmonella Typhi and S. pneumoniae were the most common causes of invasive infection overall, M. tuberculosis and C. neoformans were the leading causes of bloodstream infection among HIV-infected inpatients in Tanzania in the ART era. We demonstrate a protective effect of HIV against Salmonella. Typhi bloodstream infection in this setting. HIV co-infections continue to account for a large proportion of febrile admissions in Tanzania.

On Sample Size Calculation in Bioequivalence Trials

Sample size calculation plays an important role in bioequivalence trials. In practice, a bioequivalence study is usually conducted under a crossover design or a parallel design with raw data or log-transformed data. In this paper, we discuss the differences in sample size calculation between a crossover design and a parallel design with raw data or log-transformed data. Formulas for sample size calculation under a crossover design and a parallel design with raw data or log-transformed data are derived. A brief discussion for the relationship among these formulas is given.

Sample size determination for the two one-sided tests procedure in bioequivalence.

Approximate formulae of sample sizes for Schuirmanns two one-sided tests procedure are derived for bioequivalence studies with the 2×2 crossover design. These formulae are simple enough to be carried out with a pocket calculator.

On the Regulatory Approval Pathway of Biosimilar Products

Biosimilars (or follow-on biologics) are a new class of medicine which enters the market subsequent to a previously approved version. They have demonstrated similarity to innovator biologic products in terms of quality, safety, and efficacy. The EMA has taken the lead in the regulatory approval framework for biosimilar products, and WHO has published guidelines on the evaluation of biosimilars in order to facilitate the global harmonization. Based on EMA and WHO guidelines, many other countries such as Canada, Japan and Korea have also issued their own guidance for evaluating follow-on biologics. The US FDA was authorized to approve follow-on biologics by the BPCI Act passed by the US Congress on March 23, 2010, and has just issued a draft guidance in early 2012. The basic concepts and main principles of approving biosimilars are similar among various nations, notwithstanding some differences in regard to the scope, the choice of reference product, and the data requirement. This article reviews the regulatory approval pathway of biosimilar products in different regions.

Statistical Consideration of Adaptive Methods in Clinical Development

ABSTRACT In recent years, the use of adaptive methods in clinical development based on accrued data has become very popular due to its flexibility in modifying trial procedures and/or statistical procedures of on-going clinical trials. However, it is a concern that the actual patient population after the modifications could deviate from the originally targeted patient population. Major modifications of trial procedures and/or statistical procedures of on-going trials may result in a totally different trial, which is unable to address the scientific/medical questions that the trial intends to answer. In this article, the impact on the target patient population, statistical inference, and power analysis for sample size adjustment after changes or modifications made to an on-going clinical trial is studied.

ASSESSING SENSITIVITY AND SIMILARITY IN BRIDGING STUDIES

In pharmaceutical industry, the sponsors are interested in bringing their drug products from one region (e.g., the United States of America) to another region (e.g., Asian Pacific) to increase the exclusivity of the drug products in the marketplace. However, it is a concern whether the clinical results can be extrapolated from the target patient population in one region to a similar but different patient population in a new region due to a possible difference in ethnic factors. The International Conference on Harmonization (ICH) recommends that a bridging study may be necessarily conducted to extrapolate the clinical results between regions. However, little or no information regarding the criterion for determining whether a bridging study is necessary based on the evaluation of the complete clinical data package is provided by the ICH. Furthermore, no criterion on the assessment of similarity of clinical results between regions is given. In this paper, we propose the use of a sensitivity index as a possible criterion for regulatory authorities in the new region to evaluate whether a bridging clinical study should be conducted and the sample size of such a bridging clinical study. A criterion and a statistical method for assessment of similarity of clinical results between regions are also proposed, using the concept of population bioequivalence [FDA. Guidance for Industry—Statistical Approaches to Establishing Bioequivalence, Center for Drug Evaluation and Research, Food and Drug Administration: Rockville, MD, 2001] assuming that study site is random.

Invasive bacterial and fungal infections among hospitalized HIV-infected and HIV-uninfected children and infants in northern Tanzania.

Objective  To describe the contribution of paediatric HIV and of HIV co‐infections to admissions to a hospital in Moshi, Tanzania, using contemporary laboratory methods.