EGFR-TK Mutations Underlie Majority of Responses to Gefitinib

NEW ORLEANS-Mutationsin the epidermal growth factor receptor(EGFR) may predict for significantresponse to drugs targeted at this molecule in non-small-cell lung cancer(NSCLC), according to presentationsat a special educational sessionof the American Society of ClinicalOncology 40th Annual Meeting("EGFR Mutations Predict Responseto Gefitinib-Now What?")The session was called after reportsfrom several groups, dealing with smallnumbers of patients, that strong response to EFGR inhibition correlateswith the presence of mutations in thereceptor's tyrosine kinase (TK) domain.The results were discussed byThomas Lynch, MD, of MassachusettsGeneral Hospital (MGH) and PasiJnne, MD, PhD, of the Dana-FarberCancer Institute. Carlos L. Arteaga,MD, of the Vanderbilt-Ingram CancerCenter at Nashville, Tennessee, provideda concluding discussion.Strong Responses to GefitinibDr. Lynch gave an update of publishedwork (N Engl J Med 350:2129-2139, 2004) in which 25 (9%) of 275patients showed a strong responseto single-agent gefitinib mesylate(Iressa), a small molecule inhibitor ofthe TK domain of the EGFR.Tumor samples from 10 of thesepatients have been assayed. All hadbeen diagnosed with adenocarcinoma or bronchoalveolar carcinoma(BAC), and were never or formersmokers. Eight of these 10 samplesshowed somatic mutations in thissame TK domain, located in the ATPbinding site. Such mutations were absent from seven nonresponders togefitinib treatment.In addition, a survey of tumor materialfrom 25 NSCLC patients nottreated with gefitinib revealed twomore cases of mutated TK domainsin the EGFR.Dr. Lynch also cited in vitro workby Jeff Settleman, PhD, of the MGHCancer Center. In cell lines, transfectionof the mutated EGFRs intoCos-7 led to enhanced activation inresponse to EGF, as well as increasedsensitivity to gefitinib; both phenomenaare of direct relevance to the developmentof the cancer and theresponse to gefitinib. [Dr. Settleman'swork was published onlinein Science July 29, 2004.]Dr. Lynch suggested that these relationshipsmight justify sequencedependentdose modification protocols,in which mutation-negativepatients would receive higher doses,while mutation-positive patientscould possibly benefit from reductionsif toxicity were proving problematic.He also urged first-line trials of EGFRinhibitors in mutation-positiveNSCLC patients.Several clinical reports have shownthat the frequency of strong responsesto gefitinib is higher in women thanin men, in Japanese vs Caucasian patients,in adenocarcinoma vs otherhistologies, and in nonsmokers vsformer or current smokers.In a sequencing program ofunselectedNSCLC specimens, donein collaboration with Japanese researchersat Nagoya City University,Dr. Jnne and his colleagues foundthat the frequency of EGFR mutations"very much mirrored the populationof patients that derive the greatestbenefit, in terms of response, fromgefitinib." In one striking example,54% of women nonsmokers with adenocarcinomashad mutated EGFRs,compared with only 11% of womensmokers with adenocarcinoma.In vitro analysis of NSCLC cell lineswith the EGFR L858R mutationrevealed increased sensitivity togefitinib, compared with wild-typecell lines. In addition, Dr. Jnne andhis colleagues showed that this enhancedsensitivity is accompanied bya dramatic increase in gefitinib-mediatedapoptosis, suggesting that themutant receptor is important for cellsurvival in NSCLC.Dr. Lynch urgedfirst-line trials ofEGFR inhibitorsin mutation+NSCLC patientsDetailed analysis of individual patientswho responded to gefitinibyielded results similar to those reportedby Dr. Lynch; all seven patientswith a strong response to gefitinib hadEGFR mutations, while none wasfound in six patients progressing ongefitinib."These comparisons are highly significantand, combined with the invitro findings, suggest that the mutationsin these patients are determinantsof sensitivity to gefitinib," Dr.Jnne said.Finally, EGFR sequence analysis of227 tumor blocks from the TRIBUTEtrial, in which patients received chemotherapywith or without erlotinib(Tarceva), another small moleculeEGFR-TK inhibitor, revealed that inclusionof erlotinib led to significantlygreater response rates among the29 patients with mutated receptors vsthe 199 patients with wild-typeEGFRs. These responses in patientswith mutations have not yet translatedinto improved survival.He noted that survival was significantlylonger in patients with mutatedvs nonmutated cancers regardlessof treatment, suggesting that mutationsmay not only predict responseto EGFR inhibitors may but also beprognostic for survival.Underestimates?Dr. Arteaga pointed out that theestimates of mutation frequencies discussedabove (about 10%) may beunderestimates. In most cases, theentire gene was not sequenced; moreover,the use of material from paraffin-embedded sections may entaillower detectability, since the qualityand amount of DNA from such sourcesare not optimal. In addition, theuse of diagnostic tumor material priorto the appearance of metastatic diseasemay mean that subsequent transformationevents, including mutationof the EGFR, may not have been detected.Ongoing investigations shouldhelp to assess the relevance of theseconcerns. There is certainly no shortageof potential mechanisms throughwhich these mutations can be transforming.The EGFR is only the first ofa family of four receptors, erbB1through erbB4, which form homodimersand heterodimers in respondingto a variety of ligands.In principle, mutations could leadto enhanced downstream signalingthrough a variety of potential mechanisms,including: (1) constitutive receptordimerization, especially inmore active heterodimer combinations;(2) enhanced response toligands; (3) more effective engagementwith downstream pathways; and(4) less sensitivity to endocytosis anddegradation. ATP mimetics such as gefitinib or erlotinib could be effectivein countering several of these possibilities,especially if the EGFR mutationalso confers enhanced affinityto the inhibitor.Possibility of ScreeningWhile the evidence for the relevanceof EGFR mutations is convincing, Dr.Arteaga argued that wide screening forEGFR mutations is of limited utility atpresent. In addition to the question ofthe number of likely mutations, it isdifficult to conceive of a test that couldinclude them all with current technology,he said. In fact, mutational screeningis not done in clinical laboratories at all, and remains very much a research-based enterprise. Practical matterssuch as tissue availability alsoremain problematic."However," Dr. Arteaga said, "ifsurgical tissue was available and therewas access for EGFR sequencing, thereare good reasons why this should bedone. These studies will inform furtheron the natural history of mutantEGFR lung cancers as well as potentiallyidentify candidates for randomizedadjuvant studies with EGFR inhibitorsthat are in early phases ofdevelopment." This would be particularlytrue, he said, for certain classesof patients, such as nonsmoking womenwith adenocarcinoma, for whomthe probability of EGFR mutations isapproximately 50%."For these patients," Dr. Arteagatold Oncology News International,"mutational screening, as well asentrance into clinical trials of firstlinetherapy with EGFR inhibitors,might well be a reasonable proposition,even at our current state ofknowledge."