Prognostic/Predictive Value of Tumor-Sidedness in CRC
Panelists: John L. Marshall, MD, Georgetown University Hospital; Dirk Arnold, MD, PhD, Instituto CUF de Oncologia; University of Hamburg; Fortunato Ciardiello, MD, PhD, Seconda Università di Napoli; Paul R. Helft, MD, Indiana School of Medicine of the Indiana University Melvin and Bren Simon Cancer Center; Tanios Bekaii-Saab, MD, Mayo Clinic Cancer Center in Arizona
John L. Marshall, MD: In patients who are unfortunately diagnosed with cancer, I think the biggest evolution that’s happening is our understanding that sidedness of where the tumor arose from actually matters. Fortunato, why don’t you start us off with this, and just remind everybody about the trials and the results from those trials that show the distinction?
Fortunato Ciardiello, MD, PhD: We have to remember that the colon is a very long organ. We traditionally call the cancer left- or right-sided. What is right? What is left? And what is rectum? It is complex to define, but there is biology and epidemiology behind it.
Basically, true of the colon, the function of the mucosa is physiologically different. And the way carcinogenesis works and develops is considerably different in the right side versus the left side. We have knowledge of different gene evolutions, and the roles, and carcinogenesis according to the sidedness, but we still lack the global picture. In other words—and we can discuss that later—even if we take out the known genes like RAS, BRAF, and whatever else we understand is important for the biology of this disease, there is still something lacking in our knowledge in saying the right side is different from the left. For a longer period of time, we knew that, prognostically, tumors are rising on the right side.
John L. Marshall, MD: Can I interrupt? We people around this table designed clinical trials. What you’ve just said is an incredibly important thing. We’ve known, for what, 10, 15 years…
Tanios Bekaii-Saab, MD: Even longer.
John L. Marshall, MD: …that right-sided colon cancers are worse than left-sided. Why weren’t we stratifying, prospectively, in our clinical trials? That’s my biggest self-check on us—those of us who designed these studies. Why did we choose to ignore it?
Dirk Arnold, MD, PhD: We did this experiment a couple of days ago where we had a list of 9 factors that we would stratify according to these factors—including molecular factors, clinical factors, and also sidedness. And of these 9 factors, which would you pick to stratify coding? The finding was that this differs from which drug is used. And for sidedness, I don’t see that risk. With this binary marker in nearly two-thirds of patients, or one-third of the distribution that are statistically.…
John L. Marshall, MD: Will come out.
Dirk Arnold, MD, PhD: It’s kind of a balancing act. I don’t have to lose my stratification factor from the statistical power to these factors.
John L. Marshall, MD: Part of me wonders, if we had been doing this and looking at it through our development of our 10 drugs in colorectal cancer, would we know different things today? But anyway, I like your answer—that it will come out in the wash. You don’t spin statistical power.
Dirk Arnold, MD, PhD: It may be relevant for different drugs, but for the general history of trials, I’m not so certain.
Tanios Bekaii-Saab, MD: The initial data were mostly in the early stages, not in the later stages. The thought was very simplistic—that the right takes a longer time to be diagnosed than the left. You present with the symptoms later, and thus, it was blamed on our timing.
John L. Marshall, MD: Stage migration.
Tanios Bekaii-Saab, MD: Yes. The thought was mostly that it was stage migration, so it didn’t get integrated as much as a prospective thought.
Fortunato Ciardiello, MD, PhD: More with TNM classification rather than biology.
Tanios Bekaii-Saab, MD: Yes.
John L. Marshall, MD: I interrupted you. Why don’t you continue.
Fortunato Ciardiello, MD, PhD: Basically, what happened last year—but this started before, also, with some preliminary studies—was that people started to ask if this difference could be related not only in a different prognosis but also in a different behavior. People asked about the molecular target agents we are usually using, mostly in first-line metastatic colorectal cancer—namely, anti–EGFR therapies and bevacizumab as antiangiogenic monoclonal agents. What came out, in a very cautious consideration, was a pooled analysis of 6 old trials that were looked at, last year, during our ESMO congress in Copenhagen, in which data on anti-EGFR drugs were compared with chemotherapy alone or chemotherapy plus anti-EGFR drugs. 3 trials contained the comparator with chemotherapy plus bevacizumab. The overall picture was confirming the prognostic difference between left-sided and right-sided colon cancer, regardless of treatment. John L. Marshall, MD: And it’s not a small difference, right? It’s about a year to a year and a half? It’s what, 18 months versus 25, 30 months?
Fortunato Ciardiello, MD, PhD: It’s clinically evident and clinically meaningful evidence. Then there was very strong evidence, at least within these 6 trials—5 of them were first-line trials and 1 was a second-line trial—that if we don’t see the experimental arm, chemotherapy plus either cetuximab or panitumumab on the left side was really superior in terms of overall survival, PFS, and response rate to the comparator—either chemotherapy or chemotherapy plus bevacizumab. The other side, the right side, besides being prognostic, was predictive of lack of efficacy of anti-EGFR drugs.
From our analysis, we were not able to understand which regimen was the best. For chemotherapy alone, there was a trend in favor of adding bevacizumab to chemotherapy. But still, another strange phenomenon came out that could be understood by biology—the response rate, even on the right side, was slightly in favor of the cetuximab or panitumumab-containing chemotherapy combination. This could say that, in the RAS wild-type patients, maybe there is still a small proportion of cancer cells that are responsive to EGFR inhibition—we have a better response, but it doesn’t last. Is that relevant, clinically, or not? This is the question, and in most cases, it’s not.
John L. Marshall, MD: For those of us who don’t spend all day thinking about colon cancer, let’s figure out what this means. Let’s say I have a right-sided colon cancer patient. I’m going to make this patient have BRAF, RAS, and everything as wild-type. Is there ever a role for using EGFR therapy in that patient? You’re teasing us by saying that maybe there’s a response delta there. The National Comprehensive Cancer Network is basically saying, “No. Let’s steer away from front line.” One of the studies, in that analysis, is the second-line study, which shows the same thing with panitumumab. If you think you have a right-sided RAS wild-type, BRAF wild-type, tumor, are you ever giving that patient EGFR therapy?
Tanios Bekaii-Saab, MD: Yes, as third-line therapy.
John L. Marshall, MD: Third line. Are you optimistic? Or are you going to try it because you don’t know of anything better?
Tanios Bekaii-Saab, MD: Both. I think the data are pretty suggestive that these VEGF inhibitors will work better on the right side. There’s no doubt, and it’s not because VEGF inhibitors work better on the right side but because EGFR inhibitors really do not do as well on the right side.
To the point that was just made, EGFR inhibitors, in a small population of patients, still benefit those patients on the right side. We don’t know which ones exactly. We don’t know if there’s biology that’s changing. We don’t know, essentially, if we’re putting these tumors under pressure from first-line, or second-line, therapy. And then, we get some emerging clones of EGFR amplification, where something else is happening that may make those tumors a little bit more susceptible. But it doesn’t matter. We don’t have other chemotherapy options before we get to the other agents we have. I think, absent of more data, we still have an obligation to put those patients on, or attempt to try, EGFR inhibitors in that setting because we don’t have data that say we’re taking away benefit and because we also think that some patients may still benefit.
John L. Marshall, MD: In later-lines therapy.
Tanios Bekaii-Saab, MD: Beyond second line.
John L. Marshall, MD: I sort of agree with Tony. This is what I’m doing. Does anybody have a difference of opinion? Paul?
Paul R. Helft, MD: No. I actually wanted to agree with Tony. We get fooled, a little bit, by thinking we have an embarrassment of riches of therapies for these patients—even for those who are completely wild-type. But in fact, we run out of therapies sooner than we would like to. I agree that the preponderance of evidence is that the right-sided patients, in general, don’t respond to EGFR inhibition therapy. In the absence of lots of other wonderful therapies beyond second line, which we really don’t have at this point, I’m still willing to try it on them. But I’m not as hopeful.