Changes in the design and interpretation of phase 3 trials for advanced non–small-cell lung cancer (NSCLC) have begun to emerge regarding statistical power, sample size, and the primary end points used. In a new study, researchers explored how the design and interpretation of these trials has changed over the past 3 decades. The findings point to a disquieting shift in NSCLC trials (Sacher AG, et al. J Clin Oncol. 2014;32:1407-1411).
Using PubMed to conduct a literature search of all phase 3, randomized, placebo-controlled clinical trials evaluating systemic treatment for advanced NSCLC between 1980 and 2010, the researchers identified 245 trials, 203 of which met the inclusion criteria. Review of the included trials revealed several notable trends.
The number of NSCLC phase 3 trials increased dramatically over time, from 32 trials conducted in the 1980s to 53 studies in the 1990s, to 118 trials from 2001 to 2010. The sample size also increased from a median of 152 patients in the 1980s, to 184 patients in the 1990s, to 413 patients between 2001 and 2010.
The distribution of treatments has also shifted from multiagent triplet chemotherapy in the 1980s to predominately doublet therapy in the 1990s, and then to a sizable increase in targeted therapies in the most recent decade.
The researchers also found a significant shift in the primary study end point. In most of the studies conducted before 2000, the primary end point was overall survival (OS), with 97% of trials reporting median OS in the 1980s and 96% in the 1990s. Significantly fewer trials used OS as the primary end point between 2001 and 2010 (81%; P <.002).
Instead, progression-free survival was the primary end point in 13% of trials conducted during this period. Although the percentage of trials with statistically significant improvement in the primary outcome has remained stable over time, the percentage of trials reporting positive outcomes without achieving that end point increased from 31% in 1980 to 1990 to 53% in 2001 to 2010.
In 60 of the trials reporting a statistically significant improvement in survival, there was a decreasing trend for an improved median net survival benefit over time (3.9 months in 1981-1990 compared with 2.4 months in 1991-2000 and 2.5 months from 2001-2010). When all trials deemed positive were assessed, this changed from 3.9 months in 1980 to 1990 to 2 months from 1991 to 2000 and 0.9 months from 2001 to 2010. Despite the declining net survival, the findings showed that average median survival across trials in each decade increased from 6.7 months in 1980 to 1990, to 7.9 months from 1991 to 2000, to 9.5 months from 2001 to 2010.
The researchers concluded that important changes in the design and interpretation of phase 3 clinical trials in patients with advanced NSCLC have occurred over the past 30 years. Although the size of the trials has increased significantly, they have less clinically meaningful end points and are less effective in identifying the clinical benefit of new therapies for the management of patients with cancer.
