VANCOUVER, Canada-In a
meta-analysis of 13 randomized trials
in patients with non-small-cell lung
cancer (NSCLC), platinum-based regimens
(Gemzar) offered a statistically significant
improvement in overall survival
and progression-free survival, compared
with other platinum-based regimens.
Joan Schiller, MD, professor of
medicine, University of Wisconsin,
Madison, presented the results at the
10th World Conference on Lung Cancer
(abstract O-239) on behalf of the
international team of investigators.
"The gemcitabine-containing regimens
were clearly more effective than
older regimens containing first- or
agents, and at least as effective as those
with other third-generation agents,"
Dr. Schiller said.
Guidelines issued in 1997 by the
American Society of Clinical Oncology
(ASCO) recommended a platinumbased
regimen as optimal therapy in
advanced NSCLC. But there has been
no statistically significant survival benefit
shown for any third-generation
platinum-based regimen, although
showed a small advantage in terms of
time to progression in ECOG 1594,
which compared four regimens, she
of 13 Trials
The study reported at the Vancouver
meeting was a meta-analysis of 13 trials
in which gemcitabine plus cisplatin or
carboplatin (Paraplatin) was compared
with other platinum-based regimens in
the first-line treatment of stage IIIB/IV
NSCLC. It was compiled through a comprehensive
search of published and
unpublished sources of all studies reported
by December 2002.
A pooled hazard ratio was produced
using a fixed-effects meta-analysis. Statistical
heterogeneity was addressed
with a random-effects model when appropriate.
An estimate of absolute
treatment benefit at 1 year was also
Dr. Schiller noted that heterogeneity
of multiple studies constitutes a
potential weakness of any meta-analysis,
but she added that meta-analysis
also offers the strength of statistical
power and the ability to generate a
precise estimate of treatment effect.
This particular meta-analysis was able
to show advantages for gemcitabine
plus a platinum agent that smaller studies
could not detect, she said.
Thirteen eligible studies generated
a pool of 4,556 patients in an analysis
of 17 comparators: 12 against platinum-
based doublets, including vinorelbine
paclitaxel/cisplatin (2), paclitaxel/carboplatin
(2), docetaxel (Taxotere)/cisplatin
(1), and etoposide/cisplatin (1).
Five studies included single-agent or
triplet-agent regimens, including mitomycin
(Ifex)/cisplatin (4) and cisplatin (1).
For overall survival, a slight but
statistically significant reduction in
mortality in favor of the gemcitabinebased
arms was observed, with a hazard
ratio of 0.90. One-year survival
increased from approximately 35% to
39%, for an absolute survival difference
of 3.9% favoring gemcitabine.
Two-year survival rose from 11.6% to
14.2% with gemcitabine plus a platinum,
Dr. Schiller reported.
For progression-free survival as
well, there was a significant improvement
reported with the gemcitabine -
containing regimens, with a hazard
ratio of 0.87 and an absolute benefit at
1 year of 4.2%.
The same trend emerged in separate
analyses of older regimens and
regimens containing third-generation
agents. The hazard ratios were 0.89 for
overall survival and 0.85 for progression-
free survival favoring gemcitabine
arms vs older agents, and 0.93
and 0.84, respectively, vs other thirdgeneration
agents. "Altogether, six of
eight trials of newer agents favored
gemcitabine," Dr. Schiller said.
The findings, she added, were upheld
in a sensitivity analysis in which
all single or triplet comparator arms
were excluded, with hazard ratios of
0.92 for overall survival and 0.88 for