Adaptive designs for early phase clinical trials in oncology with an application to myeloma

Dr Andrew Hall


We are funding a part time PhD for Dr Andrew Hall at the University of Leeds, supervised by Dr Sarah Brown.

Lay Summary

The clinical development of treatments for cancer has traditionally followed sequential stages.  Phase I trials look at the safety of a regimen usually at multiple doses.

Phase II studies are the first formal look at activity to determine if there is sufficient evidence to justify a later Phase III trial.

Over the period of 1993 – 2004 only 13.4% of cancer drugs that made it to clinical testing received a license predominantly due to lack of effectiveness or toxicity.

We want to pay more attention to early phase trials, in particular the dosing often determined by just 15-20 patients in Phase I to reduce the drop-out rate (the proportion of people not adhering to the regimen, e.g. by choosing to stop taking the drug) in Phase III.

Traditionally, Phase I trials would determine the maximum tolerated (MTD) dose of a drug and start to document the safety profile. Due to the poor prognosis of most cancers, very severe toxicity is considered tolerable in the short term in order for longer-term survival benefit.

The objective is to give as much drug as possible to be effective but without causing life-threatening consequences.

Questions to be answered in Phase I and Phase II have changed as advances in treatment have come from targeted agents which may be expressed differently in each patient.

Early phase trials now are less general and try to answer more complex questions.

Phase I objectives are changing from simply the maximum amount of a drug that can be administered, to seeking to find the most successful dose.  This approach looks for a balance between increasing effectiveness and toxicity to selects the optimal dose to take forward.

This project over 5 years will address the following:

  • Better characterise the shift in optimal dosing objectives between cytotoxic agents and targeted therapies. Specifically looking in the context of myeloma. In addition, look at the increase in the use of dose expansion cohorts and what the objectives are.
  • Explore the statistical designs of longer-term tolerability, joint modelling of efficacy and toxicity and how larger randomised Phase III dose ranging studies apply in a Phase I/Phase II setting when the ethical considerations of underdoing are different.
  • Describe how the current designs achieve the clinical objectives of non-cytotoxic agents and their shortcomings.
  • Develop designs that will achieve the objectives
  • Feasibility of utilising survival methodology in the dose optimising setting.
  • Joint modelling of efficacy and toxicity in this setting is very limited and under developed in the literature particularly missing data and differing time periods for toxicity and efficacy.
  • Explore a statistical justification for expansion cohorts.