The American Society of Clinical Oncology is pressing cancer researchers to rethink the design of future clinical trials to achieve larger gains in four common cancers.
The final recommendations, which come after months of deliberations and public comment, try to hit the sweet spot between proposing guidelines that are not obtainable, and thus ignored, and having ambitious yet realistic goals.
For pancreatic cancer, for example, the experts recommended that clinical trials seek to improve median overall survival by 50%, or 4-5 months, for patients eligible for FOLFIRINOX (leucovorin, fluorouracil, irinotecan, and oxaliplatin) and by 3-4 months for those eligible for gemcitabine (Gemzar) with or without nab-paclitaxel (Abraxane).
Overall survival (OS) was selected over progression-free survival as the primary endpoint, although it was acknowledged that OS poses challenges such as the need for longer follow-up, the potential confounding effect of post-study therapies, and use of second-line therapies for secondary mutations identified after progression during first-line targeted therapy.
Ultimately, an improvement in median OS of 2.5-6 months, depending on the setting, was identified as the minimum incremental improvement over standard therapy that would define a clinically meaningful outcome.
The recommendations, published March 17 in the Journal of Clinical Oncology (J. Clin. Oncol. 2014 Mar. 17 [doi:10.1200/JCO.2013.53.8009]), also note that incremental improvements should be accompanied by little to no added toxicity over current treatments, and that a highly toxic regimen should produce the greatest OS gains to be considered clinically meaningful.
"We expect that sponsors will appreciate the need for raising the bar with regard to clinical trial goals, but that they will be conservative in their adoption of the recommendations," Dr. Lee M. Ellis, committee chair and professor of surgery at the University of Texas M.D. Anderson Cancer Center, Houston, said in an interview. "Trials designed with less ambitious goals may still be of benefit to individual patients if trial endpoints are met and if we can develop methods to identify patients most likely to benefit from the intervention."
Achieving the "smaller and smarter" trials envisioned by the committee rests on the ability to select patients for targeted therapy based on the molecular drivers of their tumors, rather than enrolling all comers. Unfortunately, in many cases, targeted agents continue to be developed without complete understanding of the drug target and, therefore, companion diagnostics to aid in patient selection, the experts observed.
"It is difficult to hit a target when it is not certain where it is or if it is valid," agreed Dr. David M. Dilts, codirector of the Center for Management Research in Healthcare, Oregon Health & Science University, Portland, in an accompanying editorial (J. Clin. Oncol. 2014 Mar. 17 [doi:10.1200/JCO.2013.54.5277]). "This, not insubstantial risk, should be ameliorated in the near future as major clinical research organizations are banking specimens, some of which are highly annotated, and as technology to analyze such specimens becomes faster, better, and cheaper."
To further this goal, the expert committee calls on trial sponsors to develop comprehensive biospecimen banks for each trial.
"Obstacles to developing these banks include cost and the willingness and ability of trial sponsors to foot the bill," Dr. Ellis said. "However, we believe the investment will pay off in increasing our ability to understand the molecular drivers of cancer and, as a result, more appropriate targeted therapies for people with cancer."
Though quality of life was a common theme that arose in all working group discussions, the recommendations lack hard targets in this area. Instead, the working groups cited the 2011 approval of the Janus kinase 1 and 2 inhibitor ruxolitinib (Jakavi) for myelofibrosis as an example of how serial assessment of specific cancer-related symptoms can define a clinically meaningful outcome for patients.
"It is not enough to just mention how important quality of life is. A clinical trial must be designed with a suite of thoughtful, feasible, validated patient-reported outcome measures that capture clinical benefit," Ms. Musa Mayer, a long-time advocate for patients with metastatic breast cancer, said in an interview. "Observed adverse events can never fully account for the lived experience of a given treatment."
For breast cancer, the committee selected metastatic triple-negative breast cancer that was previously untreated for metastatic disease. They recommend clinical trials aim for an increase in OS of 4.5-6 months, although it was noted that consensus was not achieved by the breast cancer group on the magnitude of the benefit that would be considered clinically meaningful. The current median overall survival in this poor-prognosis population is 18 months.
The committee addressed two lung cancer populations: nonsquamous cell carcinoma and squamous cell carcinoma. They recommend clinical trials seek to improve OS by 3.25-4 months and by 2.5-3 months, respectively. Current baseline median OS in these groups is 13 and 10 months.