More Endpoints Not Necessarily Merrier in NSCLC Drug Trials

May 13, 2015
Dennis Bittner, PhD

Special Reports, NSCLC (Issue 7), Volume 7, Issue 2

The US Food and Drug Administration (FDA) released new guidance on the use of alternative endpoints, in addition to overall survival (OS), for clinical trials of non–small cell lung cancer (NSCLC) therapies.

James P. Stevenson, MD

The US Food and Drug Administration (FDA) released new guidance on the use of alternative endpoints, in addition to overall survival (OS), for clinical trials of non—small cell lung cancer (NSCLC) therapies.1Pharmaceutical companies will be allowed to use several types of clinical trial endpoints to demonstrate efficacy of drugs intended to treat NSCLC.

In 2003, the FDA announced that surrogate endpoints such as progression-free survival (PFS), time-to-progression (TTP), and objective response rate (ORR) could be used to gain special approval. In its April 2015 clarification of that guidance, the FDA noted, “Based on these recommendations from the 2003 advisory committee meeting, we have continued to recommend OS as the primary endpoint for NSCLC clinical trials. However, a clinical trial demonstrating a large improvement in PFS, with acceptable toxicity, could potentially lead to regular approval, particularly in an unmet medical need population.”

The FDA announcement also included guidance on patient-reported outcome (PRO) measures, stating that this measure could provide direct evidence of treatment. The guidance encouraged the development of “well-defined and reliable PRO instruments that capture the essential treatment benefit concepts in the targeted population,” noted the FDA announcement. It added that effective interpretation of PRO data requires a complete record of all potentially confounding concomitant medications that might impact anticancer effects.

“This is a welcome announcement. PFS has been used successfully as a primary endpoint in breast cancer for drug approval because of the number of available therapies there. The pace of development for lung cancer drugs has increased to a point where this makes sense for NSCLC, too,” said James P. Stevenson, MD, of the Taussig Cancer Institute at the Cleveland Clinic in Ohio, in an interview with Targeted Oncology. The guidelines for approval based on extending PFS an additional 2 to 3 months at statistical significance are reasonable, according to Stevenson. The inclusion of endpoints for PRO measures in the FDA’s announcement will be advantageous in helping select treatments for his patients based on potential toxicity, Stevenson stated. “The more options, the better,” he said, adding that validated quality of life (QOL) measurement tools are already available to apply to such data.

Commenting on the announcement, Chicago-area radiation oncologist Kamal Patel, MD, explained, “OS will always be considered a gold standard measure as an endpoint. However, there are some problems with looking at OS; it does not take into account the patient’s QOL and it requires a large number of patients and longer follow-up than other endpoints. With a big demand to get drugs to market ASAP, studies are shifting to other endpoints such as PFS, TTP, and ORR. Although PFS data can be evaluated much earlier, thus allowing drugs to come to market sooner, PFS and similar metrics may be subject to measurement error, subjectivity of measurement, or a predisposition to different types of biases. Knowing the advantages, disadvantages, and limitations of each endpoint is key in trial design.”

Helping to clarify trends in NSCLC studies, Adrian G. Sacher, MD, and colleagues at the University of Toronto examined more than 200 phase III NSCLC clinical trials performed in the 3 decades between 1980 and 2010 in an article published last year.2Although OS remained the most common primary endpoint for these trials, the portion of trials using PFS in place of OS increased significantly, to 13% of the total in the last decade of the study. An increase in the percentage of trials reporting a positive outcome without meeting their primary endpoint also increased significantly in the last decade, to more than 50%. Other trends identified included a decreasing magnitude of survival gain in positive trials, leading the researchers to suggest that their findings raised questions about the design and interpretation of phase III trials in NSCLC.

In an accompanying editorial, David Johnson, MD, chair of Internal Medicine at University of Texas Southwestern Medical Center, Dallas, wrote that increased use of PFS in the most recent NSCLC trials was likely due to the availability of multiple agents and the 2003 guidance permitting PFS as a surrogate endpoint.3However, Johnson said that more worrisome trends are the increase in trials declared positive despite failure to meet a primary statistical endpoint and a tendency to attach clinically meaningful value to progressively lower survival advantages. “Unabated, such a trend could lead to an ever increasing number of trials achieving so-called statistically significant results that turn out to be clinically irrelevant and costly,” said Johnson.

Howard L.(Jack) West, MD, Swedish General Hospital, Seattle, Washington, countered Johnson’s editorial, saying that it implied that this relaxation of standards was “regrettable.”4West cited trials that could not report OS because patients were surviving so long without progression, and that the improvement in survival was “self-obvious.” West said that the results of a single-arm, phase I/II trial of crizotinib in ALK-positive patients revealed a median PFS of 8 months to 10 months, about twice that expected from chemotherapy, yet the OS evidence for crizotinib’s superiority was not available until years later. West asked rhetorically how many patients with ALK-positive mutations would have been denied a clearly beneficial therapy during the additional 3 years needed to reach a randomized trial data that he said would be “…the equivalent of running a trial to confirm the value of parachutes for people falling out of an airplane.” However, West emphasized the importance of the difference in magnitude of PFS improvements, between those offering only 1 to 2 months’ difference in PFS and those that “profoundly change the trajectory of the disease and are much more likely to track with a clinically meaningful OS difference.” West believes in the benefit of balancing treatment costs. He thinks that while patients might argue the value of a few extra weeks before being told that their scan looks worse, there needs to be compelling value for cancer drugs, “when everyone ends up sharing the cost burden by paying taxes or rising insurance premiums.”

Sacher, the lead author of the 2014 study, currently at Dana-Farber Cancer Institute, Boston, offered his view on the FDA’s announcement. “Any new cancer treatment is only useful if it helps patients live longer or live better—ideally both,” he said. “This concept has been distilled into the clinically meaningful endpoints of OS and QOL.” While acknowledging that the convenience of surrogate endpoints such as PFS and ORR allow clinical trials to be completed more quickly, expediting the development of effective therapies for aggressive malignancies, Sacher also warned that such endpoints are potentially subjective and need to be interpreted in the context of each clinical trial. He added, “A strong PFS benefit from a new treatment increases the likelihood that this treatment may also improve OS and/or QOL, but PFS benefit alone is not itself a clinically meaningful endpoint.”

Both Johnson and Sacher endorsed recommendations5restricting regulatory approval based on PFS to treatments considered “paradigm changing” for a difficult cancer like NSCLC, defined as those that would be expected to double PFS and result in at least a 50% increment in median or 2-year survival rates, at a minimum.


1. Accessed May 1, 2015.

2. Sacher AG, Le LW, Leighl NB. Shifting patterns in the interpretation of phase III clinical trial outcomes in advanced non—small-cell lung cancer: The bar is dropping.J Clin Oncol. 2014;32:1407-1411.

3. Johnson DH. Setting the bar for therapeutic trials in non-small-cell lung cancer: how low can we go?J Clin Oncol.2014 32:1389-1391.

4. Accessed May 6, 2015.

5. Sobrero A, Bruzzi P. Incremental advance or seismic shift? The need to raise the bar of efficacy for drug approval.J Clin Oncol.2009;27:5868-5873.