National Cancer Institute - NCI Division of Cancer Treatment and ...

target, is sensitive to the agent. To achieve
efficiency, the design parameters must be
carefully structured to provide adequate
enrichment of the randomly assigned
patients.
Freidlin B, Simon R. An evaluation of randomized
discontinuation design. J Clin Oncol
2005:23;5094–8.
Surrogate Endpoints
In many clinical trials, it would be useful
to have a surrogate endpoint that could
be measured earlier or less invasively than
the definitive endpoint. Three BRB statisticians—Drs.
Korn, Albert, and McShane—
described statistical methods for using the
surrogate and definitive endpoint results
from a series of previously completed trials
to assess whether the surrogate endpoint
could be used for a future trial.
Korn EL, Albert PS, McShane LM. Assessing surrogates
as trial endpoints using mixed models
(with discussion). Stat Med 2005:24;163–90.
Early Release of Interim Data
in Randomized Clinical Trials
Standard data monitoring procedures
for clinical trials only allow release of
interim efficacy results at the end of the
trial or earlier if the results have crossed
a data monitoring boundary. Drs. Korn,
Hunsberger, and Freidlin, BRB, in collaboration
with Drs. Malcolm Smith and Jeffrey
Abrams, CTEP, suggest specific clinical
situations in which it might be preferable
to release interim efficacy results even
though no boundary has been crossed.
The situations are such that the interim
release of data will not interfere with the
final analysis of the trial but will potentially
offer a significant benefit to the public.
22 ■ P R O G R A M A C C O M P L I S H M E N T S 2 0 0 6
■ ■ ■
In specific clinical situations… the interim release of data will
not interfere with the final analysis of the trial but will
potentially offer a significant benefit to the public.
Korn EL, Hunsberger S, Freidlin B, Smith MA,
Abrams JS. Preliminary data release for randomized
clinical trials of noninferiority: a new
proposal. J Clin Oncol 2005:23;5831–6.
Multiple Comparisons
and Clinical Trials
Multiple comparison issues arise in
clinical trials with subgroup analysis,
multiple variables, interim monitoring,
and data-driven choice of hypotheses.
It has been suggested that a nonstandard
type of analysis of clinical trial data
(“likelihood-based methods”) can
eliminate the problems with multiple
comparisons. Drs. Korn and Freidlin
examine this proposition in detail and
find it to be lacking.
Korn EL, Freidlin B. The likelihood as statistical
evidence in multiple comparisons in clinical
trials: no free lunch. Biom J (In press).
Sample Size Calculations for
Trials with Historical Controls
In the 1980s, Dr. Simon and his colleagues
showed that it was incorrect to ignore the
variability of the historical control data
when performing sample size calculations
for trials using historical controls. More
recently, BRB staff members have shown
how these widely used methods from the
1980s can be improved upon.
Korn EL, Freidlin B. Conditional power calculations
for clinical trials with historical controls.
Stat Med (In press).
Rubinstein LV, Korn EL, Freidlin B, Hunsberger S,
Ivy SP, Smith MA. Design issues of randomized
phase II trials and a proposal for phase II screening
trials. J Clin Oncol 2005:23;7199–206.