INTACT Trial Analyses Fail to Find Prognostic Factors for Gefitinib in Non–Small-CellLung Cancer

CHICAGO-Multivariateand subset analyses of the two INTACT(Iressa NSCLC Trial AssessingCombination Treatment) trials haveso far failed to identify any new prognosticfactors for improved survival ofadvanced non-small-cell lung cancer(NSCLC) patients treated with gefitinib(ZD1839, Iressa) and platinumbasedchemotherapy (ASCO abstracts2522 and 2523).Investigators are intrigued, however,by a trend toward improved survivalin a subset of adenocarcinoma patientswho were on chemotherapy for90 days or more. Patients given 250mg of gefitinib daily in INTACT 2 hada median survival of 17.1 months comparedto 16.1 months for patients on a500-mg dose and 13.6 months for patientson placebo.The trend suggests that giving gefitinibas a maintenance drug mightbenefit some patients who do well onchemotherapy, according to investigatorRoy S. Herbst, MD, chief of thoraciconcology at the University ofTexas M. D. Anderson Cancer Center."There's a hint that there might be aneffective maintenance therapy followingchemotherapy. And that shouldhopefully lead to future trials," he toldONI."This is only hypothesis generating,"Dr. Herbst cautioned. "It's onlya hypothesis-a subset analysis to atrial. New trials are concurrently beingdesigned to follow-up on this finding,"he said.No Consistent Effects
Although gefitinib produced significantantitumor activity in singleagentphase I and II trials in patientswith advanced or metastatic NSCLC,it failed to improve survival when combinedwith standard chemotherapy inthe phase III INTACT trials. Patientsreceived a gemcitabine (Gemzar)/cisplatindoublet in INTACT 1 or acarboplatin/paclitaxel doublet inINTACT 2. They were also randomizedto receive either 250 mg or 500mg of gefitinib daily or a placebo (Figure1).Median survival in INTACT 1 was9.9 months for both gefitinib groupsvs 10.9 months for patients on placebo.In INTACT 2, median survivalwas 9.8 months with 250 mg of gefi-tinib, 8.7 months with 500 mg, and 9.9months with placebo.Despite findings linking survival togender and histology in the smallersingle-therapy trials with gefitinib, thenew analyses did not turn up a relationshipin INTACT 1 or 2. Multivariateanalysis with eight prognostic factorsalso failed to show any consistentable effects.Tissue samples are still being analyzedfor expression of the HER1/epidermalgrowth factor receptor (HER1/EGFR) tyrosine kinase targeted by gefitinib,as investigators are interestedin whether positive status could be aprognostic factor. They speculated thatpatients might not have benefited becausethey were chemotherapy-naive,the dose and scheduling were not op-timal,or patients were insensitive togefitinib.

"In both trials we didn't see a benefitfrom giving the drug up front.Possibly the sequencing was not ideal.These were good trials," Dr. Herbstsaid, adding that benefit might begreater if the investigators knewhow to select patients who would respondto the agent. He plans to collecttissue samples in future trials to lookat HER1/EGFR status and other potentialbiomarkers.In a discussion of NSCLC trials,Alex Adjei, MD, PhD, of the MayoClinic in Rochester, Minnesota, sup-opportedfurther investigation, possiblyusing proteomics, to identify subsetsof patients who might benefit fromgefitinib and other novel agents. Hewarned the investigators to "beware ofspurious associations," however. Dr.Adjei explained that "correlation doesnot equate to causation" and "the factthat tumors express EGFR does notmean EGFR has a causal and criticalrole in cancer cell growth."Nonetheless, he agreed that in theabsence of prognostic factors, giving anovel agent to patients who did wellon chemotherapy might be an approachworth trying.