Survival Outcomes Affected by Tumor Location in KRAS Wild-Type mCRC

July 12, 2016

The debate between the phase III CALGB/SWOG 80405 and FIRE-3 studies has been settled, as findings of a retrospective analysis of 80405 show that tumor location is significant in determining survival outcomes for patients with KRAS wild-type metastatic colorectal cancer.

Alan P. Venook, MD

In the analysis, which was presented at the 2016 ASCO Annual Meeting, the primary tumor location of all patients enrolled in the 80405 trial was identified either right (293), left (732), transverse (66), or uncertain (46). Across both treatment arms, the median overall survival (OS) was 19.4 months (95% CI, 16.7-23.6) among patients with right-sided tumors compared with 33.3 months (95% CI, 31.4-35.7) for patients with left-sided tumors (HR, 1.60; 95% CI, 1.37-1.86; P <.001). Progression-free survival (PFS) was 8.9 versus 11.5 months, respectively (HR, 1.26; 95% CI, 1.096-1.453; P = .002).¹

Results showed that, among patients who received cetuximab (Erbitux), the median OS was 16.7 months (95% CI, 13.1-19.4) in patients with right-sided tumors versus 36 months (95% CI, 32.6-40.3) in the left-sided tumors (HR, 1.987; 95% CI, 1.60-2.46; P <.001). PFS was 7.7 months versus 11.9 months, respectively (HR, 1.539; 95% CI, 1.259-1.882; P <.001).

The cetuximab outcomes were similar to a previously reported subgroup analysis from the phase III FIRE-3 trial, which compared FOLFIRI plus either bevacizumab (Avastin) or cetuximab in the frontline setting for KRAS wild-type mCRC. In a subgroup of 167 patients from FIRE-3, the median OS was 38.7 months in patients with left-sided tumors (n = 137) versus 16.1 months in patients with right-sided (n = 30) tumors (P <.0001).

Previously, there were contrasting results with the CALGB/SWOG 80405 and FIRE-3 studies—both of which looked at the same patient population with cetuximab and bevacizumab, says Venook, the CALGB/SWOG 80405 lead study author.

In the primary studies, frontline therapy with bevacizumab or cetuximab combined with either FOLFOX or FOLFIRI yielded a comparable survival benefit in CALGB/SWOG 80405 while a survival benefit was seen with cetuximab over bevacizumab in KRAS wild-type patients in the FIRE-3 study.

“Two years ago I was literally in the middle of a mud-slinging fight because people were trying to figure out who was right and wrong; FIRE-3 or 80405,” says Venook, the Madden Family Distinguished professor of Medical Oncology and Translational Research at the University of California in San Francisco. “Now two of these datasets together make a pretty compelling story that patients with right-sided colon cancer shouldn’t be treated with cetuximab, and for patients with left-sided, you could go either way. So we’ve gone from having disparate results where there was a big debate, to realizing that in a number of instances our studies are exactly the same, and I think that’s helped clarify some of these questions.”

OncLive: What started the debate regarding the results of the CALGB/SWOG 80405 and FIRE-3 studies?

In an interview with OncLive, Venook provides a deeper understanding of the impact that tumor location has on patients with mCRC, and the significance of them change the results of the CALGB/SWOG 80405 and FIRE-3 studies.Venook: FIRE-3 and 80405 are very interesting studies. They were started around the same time, in 2003 or 2004. FIRE-3 is a European study that built on FOLFIRI as a backbone. The 80405 study was built on the choice of either FOLFIRI or FOLFOX, because in the United States FOLFOX is much more popular. FIRE-3 had the same randomizations to cetuximab and bevacizumab as we did.

However, the results of the 2 studies were different. Both studies did not select for KRAS originally and in the 80405 study we realized early on that we should be selecting for KRAS wild-type and made that change. FIRE-3’s results showed that cetuximab was superior to bevacizumab in patients in that study by about 3.5 months.

Then, when they analyzed the ‘All RAS’ population, which essentially meant getting rid of the patients from the mix we thought would not respond because of their RAS, the difference was 7.5 to 8 months between cetuximab and bevacizumab. Our study, 80405, was a cooperative group study where our original goal was to look to see a difference. Again, we amended it. We thought patients would live 22 months when we started. They all lived 30 months. In our study we saw no difference at all between the 2 groups.

How has the understanding of the results changed since the initial studies?

Obviously there are other issues. Like it or not, there’s an industry sponsor in Europe that might have had some interest in the results. I don’t know and I don’t think they fabricated things, but they may have put a different spin on it or looked a little differently at things.I said it then and I’ll say it again now; just because you get different data doesn’t mean one’s right and one’s wrong. We needed to understand it. The FIRE-3 investigators have been incredibly collaborative, and we’ve worked together to try to figure this out. We put some of our data together and we have had a couple of publications and a couple of papers.

In fact, what we have figured out over the course of the last few months is that in colon cancer the right and left sides are different, and that may very well be the explanation for why our results are different. Because it turns out, in the FIRE-3 study, 77% of the patients had left-sided cancers, and in 80405, 68% had left-sided cancers. And because of the way the cancers play out—their prognostic foresightedness—we think that may explain much of the difference. There’s also, of course, just different patterns of care and different things that make results different.

In terms of how we interpret it, I think we would interpret things the same way FIRE-3 investigators do now. Now when FIRE-3 came out it was all cetuximab beating bevacizumab, and in the US we didn’t see that. There was a fair bit of conflict. And it also is a function of different societal preferences. In the US, patients mostly don’t like bevacizumab or cetuximab as it has a rash side effect. Doctors probably don’t like to see patients with the rash.

Are there other datasets that you’re planning on looking at to better understand this?

The recent data that I presented, one of the things that made the presentation more powerful, I believe, is that I was provided the internal data from FIRE-3 on the left versus right side question. And that data looked almost identical to our data, even though the patients had a very different outcome overall.There are a couple of important datasets. The CRYSTAL study, which is FOLFIRI with or without cetuximab in KRAS wild-type patients, is very important. I believe we’ll see that the benefit to cetuximab in that study accrues only the people with left-sided disease. That would be confirmatory of our observation.

PEAK is another study that compared bevacizumab to panitumumab. That will be important.

Of the older studies—COIN and OPUS—a lot of them may not help us very much. Again, none of these studies captured the side of the primary in the beginning. It was pretty inept I guess, when you think about it, so part of this is how much effort can we put in to try to figure that out?

But those are the imperative datasets. What our next studies are going to look like, I don’t know. We really don’t know. I don’t believe we should be planning studies until we have a better handle on this because I hate to start a study and then have to stop it midway when maybe we could have said, ‘duh, we should have done something else’.

Venook AP, Niedzwiecki D, Innocenti F, et al. Impact of primary (1º) tumor location on overall survival (OS) and progression-free survival (PFS) in patients (pts) with metastatic colorectal cancer (mCRC): analysis of CALGB/SWOG 80405 (Alliance). J Clin Oncol. 2016 (suppl; abstr 3504).