ROCKVILLE, Maryland- Members of the Oncology Drugs Advisory Committee (ODAC) have recommended that the Food and Drug Administration (FDA) expand the clinical study endpoints it accepts in approving drugs for treatment of non-small-cell lung cancer (NSCLC). ODAC overwhelmingly urged the FDA to use improvement in diseasefree survival (DFS) resulting from adjuvant chemotherapy to support regular approval for NSCLC drugs and to consider progression-free survival (PFS) in granting accelerated approval to such agents. By a narrower vote, ODAC members also recommended 11 to 8 the use of PFS as an endpoint for granting regular approval of a drug for first-line treatment of metastatic NSCLC. The ODAC meeting on new NSCLC endpoints, was its first in a series of sessions to evaluate possible new endpoints for approving drugs for specific cancers. It followed a workshop to address the issue as it relates to NSCLC. FDA convened the day-long public session in consultation with the American Society of Clinical Oncology (ASCO) and the National Cancer Institute (NCI). FDA has focused primarily on survival and secondarily on improved quality of life as approval endpoints: In the classic assertion by FDA official Robert Temple, MD, "Survival trumps everything." To shorten the drug testing period and in response to the need for new ways to determine the efficacy of targeted cancer therapies, however, the FDA is investigating which additional approval endpoints might be scientifically valid. The ODAC members readily approved DFS improvement from adjuvant chemotherapy as a supportive endpoint to give regular approval for NSCLC drugs. David Johnson, MD, director, Division of Hematology and Oncology, Vanderbilt University, and a voting consultant to ODAC, cited evidence from two studies to support his vote-an international study presented at the 39th annual meeting of the American Society of Clinical Oncology and a Japanese trial. In both studies, DFS closely tracked the ultimate survival rate. PFS as an Endpoint During its discussion and voting, ODAC dropped the term "time to progression" and adopted "progression- free survival" instead. According to an FDA document, although PFS has been proposed generally as a surrogate endpoint for regular approval of cancer drugs, researchers have not rigorously validated it as such. In favor of PFS, the agency said:

• It measures tumor effect in all patients rather than just in a subset.
• Oncologists and patients widely view progression as an indicator of disease worsening that necessitates a change in therapy.
• Tumor progression is in the direct causal path of morbidity and death.

On the Con Side

• PFS is an indirect measure of patient benefit.
• The clinical meaning of a small difference in PFS remains unclear.
• Researchers have questioned the reliability of PFS in an unblinded setting.
• PFS findings are difficult for independent reviewers and the FDA to verify.

Although its members strongly favored PFS as a supporting endpoint for granting accelerated approval to an NSCLC drug, the group was split on its adequacy as an endpoint for regular approval. At issue was the quality of evidence presented regarding PFS as a valid surrogate for survival. "I would be interested in knowing whether there is more evidence than I have heard so far," said ODAC voting consultant Thomas Fleming, PhD, professor and chair, Department of Biostatistics, University of Washington. He was joined by Richard Gralla, MD, president of the Multinational Association of Supportive Care in Cancer, a nonvoting guest speaker, who noted that most of the data came from phase II studies. "That is not the kind of data you need to validate a surrogate because that is just getting at a correlation of response and an outcome," Dr. Gralla said. "In my heart of hearts, response really does agree with survival. The question is, are the data robust enough at this time." Committee members, by a threevote margin, recommended positive PFS results as an endpoint for considering regular approval of a new NSCLC drug for patients with metastatic disease. However, they overwhelmingly rejected the use of PFS as an approval endpoint in patients with inoperable locally advanced NSCLC. ODAC also discussed whether a predetermined duration of PFS should be established for randomized clinical studies in which it is used as an endpoint. Periods of 2 or 3 months' superiority were proposed, but in the end, members decided against voting to recommend a specific time span. HRQOL Data Considered In their report of the workshop on NSCLC endpoints, panelists noted that more than 90% of stage III and IV lung cancer patients report two or more disease-related symptoms, including pulmonary effects, fatigue, pain, and anorexia, as well as high degrees of psychological distress. "Consequentially," they concluded, "in addition to survival outcomes, information about treatment effects on patient-reported outcomes of health-related quality of life (HRQOL) and symptom benefit is important." In the last 8 years, most studies presented to ODAC to support the approval of cancer drugs have included HRQOL data. However, committee members have repeatedly questioned the quality and reliability of the findings because of poor study design, missing data, and the number of patients lost to follow-up. The panel spent some time discussing whether researchers should routinely include HRQOL questionnaires in clinical studies of NSCLC drugs. In the end, without a formal vote, its members voiced a consensus that such instruments would be helpful, but FDA should not mandate them.