|
|
|
|
|
|
|
How many patients should be entered in Phase II? In the Early Clinical Trials Group (ECTG) of the EORTC, such trials usually have a two-stage design. In the first stage, 14 evaluable patients are entered for a particular tumor type. If no response is observed the trial is closed. This approach ensures that if the aim is to identify a drug with a least 20% response rate, the chance of rejecting it wrongly is 0.044 [23]. In the second stage (if one or more responses are seen in the first stage) up to 11 additional patients are entered to estimate the response rate with a standard error of 10%. Whilst this approach is commendably rapid, and allows ECTG to run multiple Phase II trials, it may miss activity less than the 20% level, which may still be better than existing drugs (e.g., in colorectal cancer). Also low-level activity may prompt a re-examination of the preclinical and Phase I data to seek better dosing/schedules and may encourage programs of analogue development to attempt to improve on anticancer activity. It is therefore not uncommon to observe Phase II trials involving fifty or more subjects, and the limits are again arbitrary or dictated by economic constraints. |
|
|
|
|
|
|
|
|
Since the demonstration of anticancer activity is so essential at this stage, great care should be taken to ensure that only those patients with measurable disease be entered and that methods of assessment be applied meticulously, preferably with external independent review of radiological images, etc. |
|
|
|
|
|
|
|
|
Most Phase II trials have been open label, nonrandomised studies. In those drugs that have ready made comparators (i.e., prodrugs or analogues of existing drugs) it may be appropriate to consider a randomised design of the new drug versus the standard agent, thus allowing an early comparison of therapeutic index [24], which is usually a registration prerequisite for such drugs. |
|
|
|
|
|
|
|
|
Increasing amounts of resource and effort have been put into the use of pharmacokinetics and pharmacodynamic modeling in early clinical trials of anticancer drugs. Such detailed study has definite benefits if the compound progresses to product license application [25], but is expensive, and particularly so if the compound is abandoned. Much recent effort has gone into developing limited sampling strategies as early as possible to reduce the burden on clinical and analytical resource. This has been particularly true of one of the taxanes, Taxotere, which is currently in global Phase II-III programmes [13,26]. I think it is fair to say that this scale of pharmacokinetic/dynamic input is unique and is watched with great interest by others to determine the cost benefit of this exercise, before it could be accepted as the norm. |
|
|
|
|
|
|
|
|
Few drugs reach this stage in comparison to the huge numbers of compounds screened for potential anticancer effects! |
|
|
|
|
|