DR GOLDBERG: It’s an interesting study, a very large study, 908 patients randomized. And the randomization was to no further therapy versus 2 years of imatinib at the usual dose. Patients could be escalated or they actually started out with 400 patients to accrue, and then they escalated the accrual to 900. And I guess the presumption was to increase the power of the study.

DR LOVE: I think they were getting lower-risk patients than they thought they were going to get.

DR GOLDBERG: Right. And the interesting thing about it was the 5-year overall survival in both arms, which was 100% for the imatinib arm and 99% for the nontreatment arm. And they coined a new endpoint, which was imatinib failure-free survival. And the interesting thing about that is, of course, patients who relapsed got imatinib at relapse. And patients who were treated got imatinib immediately. And my take on this was that there wasn’t a real big difference between treatment now and treatment later, except that some of the patients never relapsed and wouldn’t have needed treatment. So I would say that this maybe is an argument for showing that there’s not a big difference in 5-year overall survival for a cytostatic agent that’s delayed until time of relapse. And so it actually, to some degree, contradicts the conclusions that we’ve drawn from the other randomized trials, which are that we ought to use imatinib in everybody with high risk. And I think the difference is that every time there’s been a randomized trial of imatinib, people who got it for longer have done better. And there’s been an excess in relapses in the year after the imatinib’s been stopped. And I think that that’s because it is not a cytotoxic agent, but rather is a cytostatic agent.

DR LOVE: Yes. It kind of reminds me — you may not know, but it’s like really a long time ago, way more than 20 years ago, one of the first adjuvant breast cancer trials of tamoxifen, was the so-called Scottish trial. I think it was the only adjuvant study I know of in solid tumors where the randomization was to tamoxifen up front, adjuvantly or on relapse. So it was kind of addressing the same thing in the long run. Of course, there they saw survival benefit.

How do you put this together, Randy, in terms of the impact of adjuvant therapy and, particularly, the question of whether it affects survival and whether, by going to 3 years, which is one of the other studies looked at, you would see a survival advantage, particularly in higher-risk patients?

DR HECHT: So I mean, I actually agree with a lot of what Rich said, because that’s always been in the back of our minds, because when it was approved it was really for relapse-free survival. But if you looked at the original randomized trial of 1 year versus nothing, the curves all come together. And those who are going to relapse are going to relapse. And it was always kind of “pay me now or pay me later.”

I was actually surprised at the 3- versus 1-year, because there was a survival difference with that. But the survival difference was relatively small. And if you parse the data, some of the survival difference was not due to cancer. So it wasn’t all cancer deaths, because the number of people who had died was still relatively small. So I think this trial actually is very interesting.

Now, I understand that this is a better group of patients and that maybe there’s a subgroup of patients — the ones who we wouldn’t treat didn’t benefit, but we wouldn’t treat those people necessarily anyway. And it’d be hard — it’s just a presentation — to peel out the ones that we would treat. But I think it does give you pause, because imatinib is not without tox — I mean, it’s well tolerated compared to FOLFIRINOX, but on the other hand, patients who are on for 3 years, they often know it. I mean, they have edema. They have fatigue. They have GI side effects. They have skin — they often have skin rash. And it’s not uncommon for people to come off. And I think this trial gives you at least a bit of a leg to stand on, particularly the patient you’re not quite as worried about, that maybe you don’t absolutely have to give imatinib in the adjuvant setting.

I’d like to see more data when they peel it out.

DR LOVE: Johanna, how do you approach this in your patients? Do you feel that it maybe is not necessary to use it up front, wait until relapse? What are the up sides or down sides of not even using adjuvant therapy?

DR BENDELL: Yes. I mean, that’s been the “jillion-dollar” question. I mean, we saw the imatinib now or imatinib later. That’s been over and over again. And then you look at your patient in front of you and you see the 3-year data and there seems to be some survival data. They’ve done studies where they’ve looked at intermittent imatinib dosing in patients in the metastatic setting and actually found that to be inferior to continuously giving imatinib, which makes one think that, should you just go ahead and give it? I was thinking about a patient of mine who has high-risk GIST. And I have prescribed for her 3 years of adjuvant imatinib therapy, to which she’s thrilled, because she’s having a bunch of facial swelling right now and more fatigue. But I think for those young patients with the high risk, particularly now the intermediate risk, I might be more inclined to say we can either wait or I wouldn’t feel as bad if we had to stop. But for the patients with the high risk, I may just continue them on for the moment.

DR LOVE: When you talk about high and low risk, how do you convert that into risk of relapse?

DR BENDELL: So what you’re looking at is you’re looking at the patients who have the higher number of mitotic figures on histology. And you’re looking for the patients with the larger tumors. The other thing that I’m looking at in terms of risk — and more, this is risk of imatinib dosing, which we haven’t talked about yet and actually was not presented in the EORTC data, is the whole issue of exon 11 versus exon 9 mutations. And that’s something that I’m actually checking in my patients, because even though there hasn’t been a prospectively defined trial, there’s been enough data coming out of subset analyses from much larger randomized Phase IIIs that really suggest, if you have an exon 9 mutation, 400 mg of imatinib is not enough. And you have to give them 800.

And so a lot of these studies — and because there is such a relatively high prevalence of exon 9 mutations, without seeing that data, to put that into some context too. Was there an imbalance between the arms? It makes it harder for me to say I shouldn’t give any adjuvant therapy.

DR LOVE: But again, talking about high risk, people are usually thinking about risk of relapse. So, yes, you have the different factors. You plug it into a nomogram. But what risk of relapse, in your mind, is what you’re calling high risk? And what risk of relapse is what you’re calling low risk?

DR BENDELL: So to me — and I’m going to have to ask for some help here, also, from my colleagues, is, to me, a high risk of relapse is, in GIST, it’s like 50%.

DR LOVE: And you’re going to tell me that the low risk is 25%.

DR BENDELL: Yes.

DR LOVE: Because those are always the numbers I hear. Is that right?

DR HECHT: The studies vary as to — the adjuvant studies don’t always — and that’s a problem. So all these different studies have different criteria for high risk or low risk, perhaps. The thing that we use the most, there’s a very useful table that comes from a very large data set. And the other thing that also kicks in — so it’s mitoses per high-powered field, size and also location.

And so, if you have a small tumor in the stomach that has very low mitotic rate, that’s very different than, for example, than someone who has a high-grade tumor in the rectum, for example, that’s large. So what we do is that we have this printed out, because we use it so much. And sometimes I’ll use it to show the patient, “This is — you fit in here. These are the risks of what you have,” and then we have a discussion about adjuvant.

DR LOVE: The thing that I’ve always found a little bit strange — again, comparing to breast cancer, where people get treated routinely, 10% risk of relapse — is that I’ve often heard — and I’m curious from your thoughts — that if you have a patient with GIST whose risk of relapse calculates to be maybe 25%, which sounds pretty high, you all consider it low risk.

DR HECHT: So the other question is, what is the benefit? And so in the breast cancer data, I mean, you’re right. And sometimes you get in trouble for talking about breast cancer adjuvant. I don’t, but other people do an awful lot to women to get a 3% improvement. I think Rich had mentioned number needed to treat, which is always a useful thing, so that your number needed to treat it would be 33 to get that 3%. We’re still, as this paper shows, we can tell you what the risk of relapse is. What’s a lot harder to say is what is the benefit of treatment.

Colorectal cancer

Gastric cancer

Pancreatic cancer

Hepatocellular carcinoma

Gastrointestinal neuroendocrine tumors

Gastrointestinal stromal tumors