FDA Urged to Add Endpoints for Approving NSCLC Drugs

ROCKVILLE, Maryland-Members of the Oncology Drugs AdvisoryCommittee (ODAC) have recommendedthat the Food and DrugAdministration (FDA) expand theclinical study endpoints it accepts inapproving drugs for treatment ofnon-small-cell lung cancer (NSCLC).ODAC overwhelmingly urged theFDA to use improvement in diseasefreesurvival (DFS) resulting from adjuvantchemotherapy to support regularapproval for NSCLC drugs andto consider progression-free survival(PFS) in granting accelerated approvalto such agents. By a narrower vote,ODAC members also recommended11 to 8 the use of PFS as an endpointfor granting regular approval of a drugfor first-line treatment of metastaticNSCLC.The ODAC meeting on newNSCLC endpoints, was its first in a series of sessions to evaluate possiblenew endpoints for approving drugsfor specific cancers. It followed aworkshop to address the issue as itrelates to NSCLC. FDA convened theday-long public session in consultationwith the American Society ofClinical Oncology (ASCO) and theNational Cancer Institute (NCI).FDA has focused primarily on survivaland secondarily on improvedquality of life as approval endpoints:In the classic assertion by FDA officialRobert Temple, MD, "Survival trumpseverything." To shorten the drug testingperiod and in response to the needfor new ways to determine the efficacyof targeted cancer therapies, however,the FDA is investigating whichadditional approval endpoints mightbe scientifically valid.The ODAC members readily approvedDFS improvement from adjuvant chemotherapy as a supportiveendpoint to give regular approval forNSCLC drugs. David Johnson, MD,director, Division of Hematology andOncology, Vanderbilt University, anda voting consultant to ODAC, citedevidence from two studies to supporthis vote-an international study presentedat the 39th annual meeting ofthe American Society of Clinical Oncologyand a Japanese trial. In bothstudies, DFS closely tracked the ultimatesurvival rate.PFS as an EndpointDuring its discussion and voting,ODAC dropped the term "time toprogression" and adopted "progression-free survival" instead. Accordingto an FDA document, althoughPFS has been proposed generally as asurrogate endpoint for regular approvalof cancer drugs, researchershave not rigorously validated it assuch. In favor of PFS, the agency said:

• It measures tumor effect in allpatients rather than just in a subset.
• Oncologists and patients widelyview progression as an indicator ofdisease worsening that necessitates achange in therapy.
• Tumor progression is in the directcausal path of morbidity anddeath.

On the Con Side

• PFS is an indirect measure ofpatient benefit.
• The clinical meaning of a smalldifference in PFS remains unclear.
• Researchers have questioned thereliability of PFS in an unblindedsetting.
• PFS findings are difficult forindependent reviewers and the FDAto verify.

Although its members strongly favoredPFS as a supporting endpointfor granting accelerated approval toan NSCLC drug, the group was spliton its adequacy as an endpoint forregular approval. At issue was thequality of evidence presented regardingPFS as a valid surrogate forsurvival."I would be interested in knowingwhether there is more evidence than Ihave heard so far," said ODAC votingconsultant Thomas Fleming, PhD,professor and chair, Department ofBiostatistics, University of Washington.He was joined by Richard Gralla,MD, president of the MultinationalAssociation of Supportive Care in Cancer,a nonvoting guest speaker, whonoted that most of the data came fromphase II studies. "That is not the kindof data you need to validate a surrogatebecause that is just getting at acorrelation of response and an outcome,"Dr. Gralla said. "In my heartof hearts, response really does agreewith survival. The question is, are thedata robust enough at this time."Committee members, by a threevotemargin, recommended positivePFS results as an endpoint for consideringregular approval of a newNSCLC drug for patients with metastaticdisease. However, they overwhelminglyrejected the use of PFS asan approval endpoint in patients withinoperable locally advanced NSCLC.ODAC also discussed whether apredetermined duration of PFS shouldbe established for randomized clinicalstudies in which it is used as anendpoint. Periods of 2 or 3 months'superiority were proposed, but in theend, members decided against votingto recommend a specific time span.

HRQOL Data Considered

In their report of the workshop onNSCLC endpoints, panelists notedthat more than 90% of stage III andIV lung cancer patients report two ormore disease-related symptoms, includingpulmonary effects, fatigue, pain, and anorexia, as well as highdegrees of psychological distress."Consequentially," they concluded,"in addition to survival outcomes,information about treatment effectson patient-reported outcomes ofhealth-related quality of life(HRQOL) and symptom benefit isimportant."In the last 8 years, most studiespresented to ODAC to support theapproval of cancer drugs have includedHRQOL data. However, committeemembers have repeatedly questionedthe quality and reliability ofthe findings because of poor studydesign, missing data, and the numberof patients lost to follow-up.The panel spent some time discussingwhether researchers shouldroutinely include HRQOL questionnairesin clinical studies of NSCLCdrugs. In the end, without a formalvote, its members voiced a consensusthat such instruments would be helpful,but FDA should not mandatethem.