Critical Issues in the Clinical Testing of Cancer Vaccines
Summary of a Panel Discussion at a Symposium Entitled
“Cancer Vaccines: Are They Here Yet?”
Conducted in Toronto, Canada, December 2002
How should we design and develop trials to
examine the potential beneﬁts of cancer vaccines
in the treatment of solid tumors such as breast,
prostate, colon, and lung cancer?
It is important to deﬁne the best and most appro-
priate population in which to test the vaccine. We
need to take into account established therapies for
this patient population, and develop a strategy to
combine this therapy with the vaccine therapy.
In early phase I clinical trials, we need to estab-
lish the best surrogate endpoints to optimize vac-
cine activity; immunology endpoints are mecha-
nistically relevant. These early trials should have
small patient groups and should be designed to
optimize combination therapies using these surro-
In these trials, it will be important to randomize
early to address questions about the best treatment
schedules and the best antigens. Statistical meth-
ods can be used to assess these immunologic end-
Finally, we should use the best arm to carry to
the next phase of clinical testing.
For cancer vaccines to be considered therapeu-
tic, we need to demonstrate clear clinical end-
points. We need to show increased survival rates,
tumor eradication or tumor control. To achieve
these endpoints, a strong commitment to carrying
out randomized clinical trials needs to be made.
The difﬁculty with making decisions to proceed
to advanced clinical testing of cancer vaccines is
the fact that we can always improve the vaccine by
adding a new antigen adjuvant or improving the
dosing schedule. However, there is no reason why
the vaccine reﬁnement work cannot continue
while we move forward to test clinical activity of
promising vaccines in these large randomized clin-
There has been a tremendous amount of work
completed to date in terms of evaluating im-
mune responses as surrogate endpoints. But
these are only surrogate endpoints. We need to
move the work from the laboratory into the
clinical setting to evaluate the clinical beneﬁts
of these vaccines.
Are we ready for large-scale, randomized, phase
III clinical trials to evaluate the efﬁcacy of cancer
If we are going to pattern cancer vaccine ther-
apies on viral immunology, we need to overcome
the tolerance of the immune responses to cancer
self-antigen targets. We need to see responses that
are protective, last for long periods of time, and
trigger high numbers of circulating tumor-reactive
T cells or antigen-speciﬁc T cells that have func-
Until we can accomplish this, we should not
embark on larger scale trials.
Do we need to see some level of immune re-
sponse maintained for some period of time?
We know that in the development phase of
other cancer drugs, we do not see the response
rates that we eventually see in clinical trials. We
do know that observing some response to the
therapy is a positive predictor of survival improve-
ment outcomes in recurrent disease.
If we were evaluating the number of vaccine
From Aventis Pasteur, Toronto, Ontario.
Dr Berinstein is an employee of Aventis Pasteur.
Address correspondence to Neil Berinstein, MD, Aventis Pas-
teur, 1755 Steeles Ave, Toronto, Ontario M2R 3T4.
© 2003 Elsevier Inc. All rights reserved.
37Seminars in Oncology, Vol 30, No 3, Suppl 8 (June), 2003: pp 37-38