Some notable properties of the standard oncology phase I design

Some notable properties of the standard oncology phase I design Gregory J. Hather Department of Statistics, University of California, Berkeley, California, USA Howard Mackey Genentech Inc., South San Francisco, California, USA October 22, 2018 Abstract We identify three properties of the standard oncology phase I trial design or 3 + 3 design. We show that the standard design implicitly uses isotonic regression to estimate a maximum tolerated dose. We next illustrate the relationship between the standard design and a Bayesian design proposed by Ji et al. (2007). A slight modification to this Bayesian design, under a particular model specification, would assign treatments in a manner identical to the standard design. We finally present calculations revealing the behavior of the standard design in a worst case scenario and compare its behavior with other 3 + 3-like designs. 1 Introduction The main goal of an oncology phase Ia clinical trial is to assess the safety of a drug which has not yet been tested in humans (Arbuck, 1996). A well designed oncology phase I trial should yield enough information to determine a safe dose, or range of safe doses, to use in further trials, while maintaining a reasonably low level of risk to the patients in the study. The dose or doses to be used for further study should be low enough to be safe for most patients, but high enough to be efficacious, since higher doses are often more effective. While the ‘more is better’ paradigm is well accepted in the case of chemotherapeutic agents, it is not clear that this paradigm should 1 arXiv:0808.0359v2 [stat.AP] 7 Jun 2009 hold for newer, targeted, non-chemotherapeutic cancer agents. That is not to say that many or most of the newer molecularly targeted agents are not more efficacious in higher doses, but that it is not a given that the ‘more is better’ paradigm always holds. The primary scientific question in an oncology phase Ia study dictates to some extent the type of patient who would enroll in such a study. Patients who have standard of care treatment available to them are less likely to participate in a phase I oncology study, and as a result, the patient population is often a heterogeneous group of patients with different types of late stage cancer. Different types of cancer often suggest different risk-benefit tolerances by both the patient and the treating clinician. This may result in a difficulty in selecting a trial design due to the existence and validity of multiple risk-benefit ratios. For example, a patient with metastatic pancreatic cancer may be willing to risk more toxicity than a patient with newly diagnosed late stage prostate cancer. This may result in the uncomfortable situation where the same adverse reaction has different implications for future development depending on the type of disease the patients has. In general, Phase I protocols report so called, dose-limiting toxicities (DLTs), irrespective of the type of disease the patient has. Additionally, the developers of a new oncology therapy may not know, before the first human data is generated in phase Ia, all the types of cancer to target in future phase II and III studies. This may be due to unexpected, dramatic results in phase Ia and/or changes in the financial resources of the entity developing the therapy. Another complicating reality is that assigning attribution of patient outcomes to the therapy under investigation, a patient’s cancer, or a patient’s concomitant medication is not an exact science. This is especially true when therapeutic agents are being tested for the first time in humans. Early on in testing, it is not uncommon for a patient’s outcome to be deemed a DLT only after a number of other patients have experienced the same outcome and/or degree of severity. An outcome determined to be a DLT after patients have started treatment in higher dose levels may result in a protocol specified action which is different than the action already taken had the DLT been identifier earlier. These realities are rarely discussed in the literature for consideration of a phase I design, which further complicates the mission of designing the ‘best’ phase Ia trial. Unlike the phase II or III paradigm, where the objective is to assess efficacy while obtain- ing valuable safety information, phase I studies often necessitate the administration of unsafe amounts of drug to some patients in order to determine a maximum tolerable dose, or range of safe doses. And as in all clinical studies, the trial should not involve too many patients or take 2 too long to complete. One can find many methods for dose finding in the literature, the “stan- dard design” (Storer, 1989) being the oldest of the commonly employed phase I designs. Other designs contained in the literature include continual reassessment (O’Quigley et al., 1990), ran- dom walk (Durham and Flournoy, 1994), escalation with overdose control (Babb et al., 1998), and cumulative cohort designs (Ivanova et al., 2007). For a more extensi

…(Full text truncated)…

This content is AI-processed based on ArXiv data.