Richard M. Goldberg, MD, of The Ohio State University Medical Center, chaired the steering committee of the 2012 Gastrointestinal Cancers Symposium in San Francisco, which attracted approximately 4,000 registrants who viewed data from some 700 scientific abstracts. The ASCO Post asked Dr. Goldberg to describe his “take-aways” from the meeting, and how the data fit with the emerging landscape of GI cancer:
It is becoming clear from recent studies, such as those presented at the 2012 Gastrointestinal Cancers Symposium, that gastrointestinal tumors we call one entity (based on the organ in which they originate) are actually very heterogeneous. Our understanding of GI cancers will probably become as complicated as our taxonomy in lymphoma. Tumors will be described by their behavior, which will be predicted by molecular markers and not just by the classical anatomic markers like grade and stage. Markers such as KRAS probably will be joined by dozens of other factors that will help us choose drugs.
I believe in the next 10 to 15 years we will be referring to tumor subtypes differentiated by their genetic profiles. At that point, our strategies may become more effective, with one strategy good for one subtype and a different strategy for another. We are already seeing this, for instance, in neuroendocrine tumors. Everolimus (Afinitor) and sunitinib (Sutent) are effective in pancreatic neuroendocrine tumors, but there is no evidence of an effect in carcinoid tumors, and all are called neuroendocrine tumors. We heard at this meeting, in a study by Yao et al,1 that certain carcinoid subsets might benefit from these agents, and this has relevance to our therapeutic approach, especially when we use a drug that has significant toxicity.
In keeping with these genetic advances, we are seeing new genetic tests to assess colorectal cancer recurrence risk. The ColoPrint test, with data presented by Tabernero et al,2 is the next step in what I think is a refinement of these decision-making tools. They are “on the verge” of becoming useful, and this will grow over time.
We also heard from Dr. David Ahlquist about a stool test that evaluates fecal DNA for a marker of a precancerous or cancerous condition.3 The idea behind this—that polyps and tumors are constantly exfoliating cells—is fascinating. This technique does not depend on bleeding and does not involve bowel prep, appears to be much more sensitive than the current fecal occult blood test, and detects low-profile serrated polyps, which can be problematic to identify with colonoscopy. If we could use serial stool DNA testing like serial Pap smears instead of colonoscopy, the cost savings would be tremendous. From all these perspectives, this technology would be welcomed. The data presented by Dr. Ahlquist was very impressive, and this is a very exciting development.
New Drugs: Some Wins, Some Disappointments
We also heard encouraging data for the new multitargeted tyrosine kinase inhibitor regorafenib.4 This is a very interesting compound that has the potential to affect a number of pathways in colon cancer. The data presented by Grothey et al showed an overall survival difference with regorafenib as a single agent in very heavily pretreated patients. The data look very similar to the initial data in the third-line setting for cetuximab (Erbitux) and panitumumab (Vectibix), before we identified the subset of patients (with KRAS wild-type) who actually benefit.
Regorafenib is a drug worthy of additional testing, though we need to find biomarkers that will help us discern what subset of patients are most likely to gain from this agent. The 1.4 months of overall survival difference was modest, but it’s a signal. By refining how we use this agent we will likely be able to exploit it to greater benefit.
Two other studies, however, were disappointing. In advanced gastric cancer, the GRANITE-1 study found no overall survival benefit for everolimus as a single agent.5 While the shape of the progression-free survival curve suggested there may be a responsive subset, my thought is we should focus on other targeted agents rather than developing iterative studies of agents for whom the trial endpoints are negative.
The other novel tyrosine kinase inhibitor, brivanib, also failed to improve overall survival in combination with cetuximab, vs cetuximab alone, in advanced colorectal cancer.6 But should we conclude that brivanib is a useless drug? I hope we don’t. Perhaps, we should instead try to narrow down and enrich the population to increase the likelihood of a response in future trials, in other words, restrict the drug to patients most likely to benefit. Biomarker studies must also inform us on this.
So my take-home lessons from this meeting are that we must return to our laboratories to sequence DNA and RNA and must exploit other analytic techniques to get a firmer grasp of tumor biology. This, in turn, will permit us to optimize drug development with the ultimate and attainable goal of improving patient outcomes. Studies from this meeting point to looming discoveries toward incremental progress. ■
Disclosure: Dr. Goldberg has received research funding from Abbott, Genentech, Bayer, Myriad, and sanofi-aventis. He has been an unpaid consultant to Bayer, Myriad, and sanofi-aventis, and has been paid as a consultant by Genomic Health, Genentech, and Lilly.
1. Yao J, Hainsworth JD, Wolin EM, et al: Multivariate analysis including biomarkers in the phase III RADIANT-2 study of octreotide LAR plus everolimus or placebo among patients with advanced neuroendocrine tumors. 2012 Gastrointestinal Cancers Symposium. Abstract 157. Presented January 20, 2012.
2. Tabernero J, Moreno V, Rosenberg R, et al: Clinical and technical validation of a genomic classifier (ColoPrint) for predicting outcome of stage II colon cancer patients. 2012 Gastrointestinal Cancers Symposium. Abstract 384. Presented January 21, 2012.
3. Ahlquist DA: Stool DNA testing: A step closer to eradicating colorectal cancer. Invited presentation at the 2012 Gastrointestinal Cancers Symposium. Presented January 21, 2012.
4. Grothey A, Sobrero A, Siena S, et al: Results of a phase III randomized, double-blind, placebo-controlled, multicenter trial (CORRECT) of regorafenib plus best supportive care versus placebo plus BSC in patients with metastatic colorectal cancer who have progressed after standard therapies. 2012 Gastrointestinal Cancers Symposium. Abstract LBA385. Presented January 21, 2012.
5. Van Cutsem E, Yah KH, Bang YJ, et al: Phase 3 trial of everolimus in previously treated patients with advanced gastric cancer: GRANITE-1. 2012 Gastrointestinal Cancers Symposium. Abstract LBA3. Presented January 19, 2012.
6. Siu LL, Shapiro JD, Jonker DJ, et al: Phase III randomized trial of cetuximab plus either brivanib alaninate or placebo in patients with metastatic chemotherapy refractory K-RAS wild-type colorectal carcinoma: The NCIC Clinical Trials Group and AGITG CO.20 trial. 2012 Gastrointestinal Cancers Symposium. Abstract 386. Presented January 21, 2012.
Findings presented at the 2012 Gastrointestinal Cancers Symposium will impact the future care of patients with these malignancies. The ASCO Post has summarized some of the most newsworthy data in the following briefs.
In the phase III GRANITE-1 trial (n = 656), single-agent treatment with...