The latest clinical trial looking at combining vascular endothelial growth factor (VEGF) and epidermal growth factor receptor (EGFR) inhibition in metastatic colorectal cancer adds little to our overall understanding of the mechanisms for optimizing selection of patients to receive such therapies. While we continue to see small gains in outcomes when targeted agents are combined, we have failed to demonstrate a clear, clinically useful benefit for our patients.
Search for Biomarkers
Our current model for developing anticancer agents involves conducting very large randomized clinical trials of treatment A vs treatment B. These trials are large because we know that we will get very small differences in the primary and secondary endpoints. To date, we have valued these small changes, and as a result, medicines have been approved and their high price is being covered, but for only small gains. However, we are increasingly intolerant of this approach, mostly due to toxicity and cost for patients and society, and we have come to conclude that while findings can be statistically significant, often they are clinically irrelevant.
To try and compensate for this, we are increasingly looking for biomarkers. So now, we still do large clinical trials and find small differences but look backwards through the rearview mirror in the hopes of finding the drivers. In a few cancers, this has been successful; where luck has prevailed, driver mutations have been discovered due to clever observations by those evaluating clinical trial results.
In gastrointestinal cancers, we have not been nearly so successful or lucky. Frankly, in gastrointestinal cancers, we have yet to identify a significant driving mutation. Even KRAS, for all the hype, has failed to deliver the same types of dramatic improvements that we have observed with HER2 targeting in breast cancer, for example.
I believe part of our problem relates to the fact that we continue to look at old banked tumors and not fresh tissue biopsies in order to measure our biomarkers. It is possible that gastrointestinal tract cancers are much more complicated, with no clear patterns of driver mutations. Of course, if this is true, we have a very steep uphill battle ahead.
Novel Trial Designs
The trial reported by Siu and colleagues as well as other trials clearly point to the need for novel clinical trial designs. Small prospectively enriched studies with fresh tumor samples, using biomarkers as the entry criteria, will be critical for us moving forward.
The Otto J. Ruesch Center at Georgetown, which I am proud to direct, is focused on linking centers around the world interested in performing such clinical trials. Our model will be to obtain tissue and do biomarker analysis on all patients prospectively and assign them to treatments, both standard and experimental, based on their biomarker and oncogene expression.
Only by working together and on a global scale will we define the intricacies of metastatic gastrointestinal cancers. We will no longer tolerate the statistically significant but clinically irrelevant results of large randomized clinical trials that are unenriched for the proper patients.
There is hope going forward, but it will require a global collaboration and partnership among our patients, our clinical research partners, and our academic institutions. ■
Dr. Marshall is Chief, Division of Hematology/Oncology, and Director, Otto J Ruesch Center for the Cure of GI Cancers, Georgetown University, Washington, DC.
Disclosure: Dr. Marshall reported no potential conflicts of interest.
