Julien Taieb, MD, discusses the prognostic value of primary tumor location in colon cancer.
Julien Taieb, MD
Right-sided tumors in colon cancer have been associated with poorer outcomes than left-sided tumors.
In a study presented at this year’s World Congress on Gastrointestinal Cancer, investigators aimed to evaluate the prognostic value of primary tumor location in patients with fully resected colon cancer, as well as its relationship to microsatellite instability-high (MSI-H), RAS, and BRAF mutational status.
Of 40% of patients who had a right-sided tumor, 10% were MSI-H, 50% were RAS-mutant, and 11% were BRAF-mutant. Right-sided tumors were associated with a shorter overall survival (OS) (HR, 1.25; CI, 1.02 - 1.54, P = .03). Similar results were demonstrated for MSI-H and microsatellite stable (MSS) patients.
OncLive: Could you provide an overview of this study?
In an interview with OncLive during the congress, lead author Julien Taieb, MD, head of the gastroenterology and GI oncology department of the Georges Pompidou European Hospital, Paris Descartes University, discussed the prognostic value of primary tumor location in colon cancer.Taieb: For the past year and a half, there has been a big buzz around sidedness of primary tumor location in colon cancer patients in the metastatic setting. First because it seems like a good prognostic marker—if you have a right-sided tumor you have a poor prognosis as compared to a left-sided tumor. And second, it has been shown by the German study FIRE-3/AIO first and then confirmed in the United States, that left colon cancer will maybe respond well to anti-EGFR therapies and maybe right colon cancers respond better to anti-VEGF therapies. So, it is also a kind of predictive marker to choose the best targeted agent in each patient.
What were the significant findings?
Moreover, we are at the time of molecular classification, but of course this classification is not available everywhere in the world. So, by knowing what to do based on right or left colon cancer is a big advantage in developing countries. Here, for the first time, we have looked at a clinical trial prospectively performed on more than 2000 patients with stage III colon cancer—so, before the occurrence of metastatic disease. We asked, “Is sidedness predictive, and is sidedness prognostic?”
We had a trial testing anti-EGFR plus chemotherapy, versus chemotherapy. This trial was negative, as anti-EGFR did not improve survival. This trial has been used with assessments for all mutations status like BRAF, RAS, with NRAS and KRAS, MSI status—it characterizes each tumor.
Then, we moved to sidedness, and what we found was that disease-free survival (DFS) with no recurrence is not impacted by the sidedness. But, if you relapse, the survival after relapse is clearly impacted by the sidedness, and the right-sided tumors—as previously described—have poor prognosis.
When looking at sidedness and the effect of anti-EGFR in RAS wild-type patients, what we have seen is that you have a positive effect of cetuximab in this purified patient—which is not significant but better than chemotherapy alone—and this is true in right and left, so there is no predictive value either.
But, surprisingly, as the patients were very well characterized, we looked in different molecular groups of what is happening on recurrence by sidedness. The first point is that MSI/MSS is exactly the same, no influence of sidedness on DFS. But, if you look at BRAF and KRAS mutant, it seems like having this mutant population, having a right-sided tumor is a good prognostic factor—which is very surprising and has never been described before.
Does this bring us closer to tailoring treatment for right- and left-sided tumors in colorectal cancer?
On the contrary, if you look at the double-wild type for BRAF and RAS, you have bad prognosis of right-sided tumors which is more usual. So, now we will have to understand in future studies—as this is a huge study of almost 1900—that we will have preclinical studies exploring why this mutation made the sidedness associated with a different prognosis.I think it is currently in some guidelines—that the treating physician should take into account the sidedness of the primary tumor before deciding on treatment. I know it is in the United States, but not in Europe or Asia yet. I think that sidedness reflects something at the molecular level, and we have to understand what this is. It is clearly not sufficient to have BRAF, KRAS, MSI—what we are doing currently for colon cancer. We may be able to find different markers that are differently expressed in right- and left-sided tumors which would allow us to fine-tune our assessment of each patient, and not only the side, which is a bit disappointing in a time of whole-genome sequencing and whatnot in cancer.
Where does that leave us on treatment options?
However, it can be useful today, and in some patients where bevacizumab (Avastin) and cetuximab (Erbitux) for example do the same, I will maybe help myself with the sidedness if I have a choice to make.Clearly, with the results in the metastatic setting, the best one has been published recently, showing that in left-sided tumors, if you compare upfront antiangiogenic and anti-EGFR with chemotherapy—anti-EGFR with chemotherapy is the preferred choice: more tumor shrinkage, better PFS, and better OS. The recent publication of the US trial also goes in that way with FOLFOX/cetuximab doing better than FOLFOX/bevacizumab in left-sided colon cancer KRAS wild-type for the United States trial, RAS wild-type for the European.
On the contrary, for right-sided, there is a real-trend toward the benefit of antiangiogenic, but the problem is that you still have more…response…with the anti-EGFR. Shrinking a tumor and survival benefit is not so clear to us, so I think we have to explore more right-sided tumors and launch more clinical trials comparing and stratifying from that to improve our knowledge on that topic.
J Taieb, RK Hampig, FE Jean, O-015 Prognostic value of primary tumor location in stage III colon cancer is associated with RAS and BRAF mutational status. Ann Oncol 2017; 28 (suppl_3): mdx262.014. doi: 10.1093/annonc/mdx262.014.