When donating tissues to the human tissue bank, informed consent has to be obtained from the patients. The consent is usually general, and not specific. Regulations for medical research require consent for each specific research, which means that repeated consent process is needed when the research is conducted using banked human tissue. However, the consent requirement for human tissue banking has been changed recently to permit research with the general consent. The objective of the present study was to clarify patients' attitude toward general consent by questionnaire, on whether they feel repeated consent is necessary. The survey revealed that most of the tissue donors thought repeated consent (specific consent for research) was not always important. On the other hand, they showed interests and concerns regarding details of the research and the results, and how the tissues contributed to the progress of medicine. We conclude that while it is possible to utilize the banked resources without obtaining specific consent for each research, greater efforts have to be made to publicize the research contents and results.

View full abstract

The main goals of a phase I oncology trial are to evaluate safety and to identify the recommended dose for phase II trials. This recommended dose is typically defined by the maximum tolerated dose (MTD). The standard design in a phase I oncology trial is 3+3 design to estimate the MTD. The 3+3 design proceeds in cohorts of three patients each. Dose escalation and reduction are conducted per the cohort, and the MTD level is estimated. The advantage of this method is the simplicity for setting dose escalation rules and stopping rules. However it has the disadvantage that MTD level is not estimated based on statistical evidence. O'Quigley et al. introduced the continual reassessment method (CRM) to overcome these issues. The CRM models the dose-toxicity relationship based on prior information, and sets target toxicity level as the probability of toxicity occurring at MTD. After initiation of the trial, the CRM updates the model parameter of the dose-toxicity relationship as a Bayesian estimate using the accumulated data, and estimates MTD level from the target toxicity level. Compared to the 3＋3 design, the CRM may improve the accuracy of MTD level estimation, and is expected to increase the proportion of patients administered the dose level estimated to be closest to the MTD. To address the criticism and issues of the CRM proposed by O'Quigley et al., a variety of improved methods are now available. Also, extensions of the original CRM have been proposed with the aim to develop more practical methods taking into account the features of the trial design. This paper reviews the development of CRM over the past two decades, including the authors' studies, and discusses future prospects.

View full abstract

An increasing number of clinical trials recently require the submission of examination data as sample to check the feasibility of the study before the trials are started. However, procedures to obtain informed consent have not been standardized. Therefore, we investigated various procedures by reviewing past cases and guidelines on research ethics and personal information protection. Based on this review, we developed the procedures for four types of study classified according to the prospective nature and degree of invasiveness of the examination. In type A, when the sample data is prospective and requires implementation of invasive examination not conducted as part of routine medical care, written informed consent is mandatory and each examination must be approved by the institutional review board (IRB). In type B, when the sample data is prospective and requires implementation of less invasive or non-invasive examination not conducted as part of routine medical care, use of pre-approved template of written informed consent that includes explanations of foreseeable risks and inconveniences accompanying the examination is required. In type C, when the sample data is prospective and uses results of examination conducted as part of routine medical care, informed consent can be obtained either orally or in writing at the discretion of the investigator. In type D, when the sample data uses preexisting medical data, efforts have to be made to obtain oral informed consent as far as possible. However, if informed consent is not possible, individual informed consent can be omitted but general notice should be given to patients regarding use of their medical records for purposes including answering inquires on medical services from other medical institutions. This classification system, which includes flowcharts and templates for informed consent, has been approved by the IRB at the University of Tokyo Hospital and will facilitate proper handling of sample data.

View full abstract