Cancer vaccines have been extensively tested in animal models, and in human
s. Initial studies focused on first generation vaccines based on whole cell
preparations or tumor lysates derived from autologous or allogeneic tumors
. Clinical studies conducted with such candidate vaccines contributed to es
tablish the feasibility of immunizing cancer patients against their own tum
ors. Significant clinical benefits were observed, both in terms of long ter
m survival and recurrence rate, in some of these trials. More recently, how
ever, cancer vaccines targeting well-characterized tumor-associated antigen
s, i.e. molecules selectively or preferentially expressed by cancer cells b
ut not by normal cells, have been designed and tested in humans. Results ob
tained as of today with these second-generation vaccines suggest that they
are safe and that they can elicit humoral and cellular responses against tu
mor-specific antigens, without inducing unacceptable clinical signs of auto
immunity. Advances in tumor biology and tumor immunity have helped to bette
r understand the mechanisms displayed by a number of tumors to escape host
immunity. This bulk of new knowledge will be used to design future cancer v
accines, which will likely target multiple TAAs, presented by different ant
igen presentation platforms. in association with synthetic adjuvants and/or
immunostimulatory cytokines. Lastly, specific tools allowing to assess in
a qualitative and quantitative manner immune responses are critically neede
d in order to establish correlates between clinical and immune responses in
patients receiving experimental vaccines. (C) 2001 Elsevier Science Ltd. A
ll rights reserved.