Novel approaches to identify optimal outcome measures in phase II cancer trials

There are four stages of the drug development process, known as phases I to IV. Phase III of the process considers whether the treatment under investigation is better than, or at least as good as, the current standard treatment for the disease, and this is assessed in a large scale clinical trial. Phase II clinical trials are designed to find out if a new treatment works well enough to test in a larger phase III trial, or to find out about the types of cancer the treatment works for, and any side effects of the treatment and how to manage them. If the results of a phase II trial show that a new treatment may be as good as, or better than, an existing treatment, a phase III trial may be undertaken. When we carry out a phase II trial, we need to make sure that the measures we use to find out if the new treatment works well, are the most appropriate measures to make an accurate decision about whether to test the treatment further in a large phase III trial. Currently, many phase II cancer trials use a measure called response to assess whether a new treatment may be as good as or better than an existing treatment. This is a measure of how much the cancer has shrunk. However for some new treatments, this may not be the best way to measure how good the treatment is at controlling the cancer. By using more appropriate measures in phase II trials this will allow faster, more accurate, development of new treatments. This research aims to look at other measures to use in phase II trials of colorectal cancer, for example looking regularly at the actual length of the cancer, to assess whether treatments may be as good as or better than existing treatments, so that we don?t carry out large phase III trials unnecessarily, and so that we don?t miss treatments in a phase II trial that may be worth further investigation. A team of clinical trial researchers at the University of Leeds will carry out this research, using data from colorectal cancer trials that have already been carried out.

The aim of this research is to optimise the design of phase II studies by identifying alternative outcome measures to use in their design and analysis. Phase II studies assess the short-term efficacy of treatments, and the potential for further evaluation of longer-term efficacy. Outcome measures used in phase II studies should be sufficiently associated with longer-term phase III outcome measures such as overall (OS) and progression-free survival (PFS), and should act as a sufficient screen of potential new therapies. The strength of association required between phase II and phase III outcomes is unknown. Response to treatment in many cancers is typically measured using the Response Evaluation Criteria in Solid Tumors (RECIST). There are however, issues with the use of response rate as an outcome measure in phase II trials. Clinical benefit may occur without tumour regression, for newly developing treatments, and response rate may no longer be the most appropriate outcome measure. Using data from a number of recent advanced colorectal cancer (CRC) trials, the following alternative phase II outcome measures will be explored: tumour measurement as a continuous variable, measured at various time-points; dichotomisation of tumour measurements; PFS assessed at various time-points. We will assess the likely ranges of correlations between RECIST and OS and PFS, and consider the relationship between each of the alternative outcome measures and OS and PFS. Surrogacy methodology that considers the association between outcomes at both the individual patient level and the trial level will be used and SAS programs produced by Hasselt University surrogate endpoint group to carry out this surrogacy methodology will be further developed for this research. Using as a benchmark the relationship between RECIST and OS and PFS, we will compare the strength of association of the alternative outcome measures with this, to identify the outcome measure most strongly associated with OS and PFS. To determine that these new outcome measures should be adopted for use in clinical trials, these new outcome measures will be validated using data from phase III studies to conduct simulations of phase II studies. Results of this research will be published in peer-reviewed journals, and SAS programs produced will be made freely available. By focusing on colorectal cancer this will act as an exemplar for consideration of such issues in other tumour types.