Objectives. To evaluate different study designs and the general utilit
y of phase II trials on prostate cancer. Methods. Extensive literature
studies and correspondance within the working group during 1 year wer
e summarized in a preliminary manuscript. The manuscript was finalized
at a I day meeting and is presented here as a consensus document. Res
ults. The main objectives of phase II studies are to assess whether a
treatment is sufficiently active to justify comparative phase III stud
ies, and to obtain further information on adverse reactions. Bidimensi
onally measurable lesions are traditionally studied, allowing objectiv
e criteria for response and progression to be applied. However, as ske
letal metastases do not fulfill the criteria for such lesions, the maj
ority of patients with metastatic prostate cancer are not eligible for
traditionally-designed phase II trials. Therefore, ancillary response
parameters, especially serum prostate-specific antigen (PSA), have be
en proposed for use. For the evaluation of adverse reactions, the crit
eria of the World Health Organization were proposed for use. A review
of various statistical designs was presented, with a focus on their ad
vantages and disadvantages in phase II trials. Conclusions. The role o
f PSA in phase II trials has not yet been firmly established. Further
study of its correlation with other endpoints is needed. In future pha
se II trials, a shift to softer endpoints than traditionally used may
enhance the process of evaluation of new antitumor drugs. Phase II stu
dies may even be replaced by early phase III studies, especially in si
tuations where new drugs do not have very heavy adverse effects. (C) 1
997, Elsevier Science Inc.