The Expected Toxicity Rate at the Maximum Tolerated Dose in the Standard Phase I Cancer Clinical Trial Design
- PDF / 633,533 Bytes
- 11 Pages / 504 x 720 pts Page_size
- 99 Downloads / 182 Views
0092-8615i2001
Copyright 0 2001 Drug Information Association Inc.
THE EXPECTED TOXICITY RATE AT THE MAXIMUM TOLERATED DOSE IN THE STANDARD PHASE I CANCER CLINICAL TRIAL DESIGN SEUNG-HOKANG Department of Statistics, Ewha Womens University, Seoul, Korea
CHULAHN Clinical Epidemiology, University of Texas Medical School, Houston, Texas
A main purpose of Phase I cancer clinical trials is to identify the maximum tolerated dose (MTD)of a new agent for experimentation in Phase I1 and I l l studies. The continual
reassessment method has been shown to be superior to the standard design. However; in practice, the standard design has still been widely used. Therefore, it is important to investigate the performance of the standard design accurately. In this paper; we develop an algorithm to compute the exact distribution of the recommended dose level in the standard design. The algorithm is a better tool than simulation in the investigation of the operating characteristics of the standard design, because it does not involve any sampling error and computing time is much shorter than simulation. With the algorithm, the expected toxicity rate at the MTD in the standard design is investigated extensively for some dose-toxicity curves in a certain range. Key Words: Dose finding studies; Cancer; Toxicity; Continual reassessment method
INTRODUCTION A PHASE I CLINICAL, TRIAL is the initial assessment of a new drug in humans at a fixed route and schedule. A main purpose of a Phase I cancer clinical trial is to determine the MTD of a new agent, which will be used as a recommended dose level for experimentation in Phase I1 and I11 studies. Although much of the recent literature reports that the standard design has poor operating characteristics compared with the continual reassessment method (1-6), the standard design has
Reprint address: Chul Ahn, PhD, Clinical Epidemiology, UT-Houston Medical School, 6431 Fannin Street, MSB 1.122, Houston, TX 77030. E-mail: Chul.W.Ahn @uIh.tmc.edu.
been widely used in most practical cases. The reason might be that the standard design does not require elaborate statistical considerations, while the continual reassessment method requires prior specification and dose-response modeling. The standard design is familiar to many clinicians, allowing them to focus on clinical endpoints that are important when investigating new drugs. Therefore, it is important to understand the operating characteristics of the standard design accurately. Simulation which involves sampling error has been the only way to investigate the operating characteristics of the standard design, and just a few scenarios are examined with simulation (4,6,7).Recently, Reiner et al. (8) studied the operating characteristics of the standard design using the exact distribution
1189
Downloaded from dij.sagepub.com at Karolinska Institutets Universitetsbibliotek on May 28, 2015
Seung-Ho Kang and Chul Ahn
1190
of the MTD. However, the diagram in Figure 1 (8) does not delineate the standard design well, although the description and Fi
Data Loading...
