ASCO 2012, Chicago, USA

Effects of panitumumab against oesophageal-gastro cancer and carcinomas

Dr Tom Waddell – Royal Marsden Hospital, Sutton, UK

Dr Waddell, you’ve been talking about oesophago-gastric cancer, oesophago-gastric adenocarcinoma and, in particular, patients do sometimes have, between a quarter and a half, I gather, the EGFR receptor. Now in view of the outlook of this disease that could be a ray of hope and you’ve been looking at that, haven’t you? Tell me about the study you’ve done.

So we knew going into the study that EGFR over-expression did occur, we knew from some series that it correlated with poor prognosis so there was a potential that that may represent a viable therapeutic target. Advanced gastro-oesophageal cancer has a very poor prognosis currently with, on average, less than twelve months survival from diagnosis and that’s with the best chemotherapy regimens that we currently have. So it was felt that may be a way to try and target that pathway and improve outcomes and the RIO 3 trial was designed therefore to evaluate panitumumab, which is a monoclonal antibody directed against EGFR, in combination with what would be regarded as the UK standard chemotherapy regimen, epirubicin, oxaliplatin and capecitabine or EOC.

So tell me what you did in the study.

Patients with previously untreated advanced tumours of the oesophagus, gastro-oesophageal junction or stomach were randomly allocated in a one to one fashion to either standard chemotherapy or chemotherapy in combination with panitumumab. And, of note, the combination therapy arm, the chemotherapy doses did have to be reduced in order to combine it with panitumumab due to some overlapping toxicity, primarily diarrhoea.

So you selected your 500 or so patients on the basis that they were over-expressing EGFR?

No, the recruitment was actually an unselected population because we didn’t have data strong enough, really, going in to it of any really validated predictive markers. So although we knew over-expression occurred, amplification occurs, to low levels of mutation, similarly there wasn’t really sufficient data going into the trial to say that only one of those populations should be treated. So all patients were treated with a planned accompanying biomarker analysis to look once we have the results at which subsets or biomarkers might have predicted benefit.

Can you give me the data on what happened with or without panitumumab?

Yes, so unfortunately the trial was halted early due to review of the unblinded data. We had an independent data monitoring committee who annually reviewed the unblinded data to ensure safety, really, throughout and at one of the reviews in October of last year the data suggested an inferior survival outcome in the experimental group. So the addition of panitumumab was certainly not being associated with any benefit and actually that group seemed to be doing slightly worse than the control arm. So we obviously at that time had a discussion about what to do and it was agreed that we should stop the trial to further recruitment, withdraw the panitumumab and cross patients over who were on the experimental arm to the standard dose EOC. And then, as I say, in the fullness of time we will yet look at the various biomarkers which we can study and see if there were subpopulations who were actually deriving benefit despite the fact that in the unselected population there certainly wasn’t benefit seen.

What do you think could be the effect if you had been able to select out the EGFR expressing patients, over-expressing patients?

The hope would be that you can find a population where the survival is increased with the use of the antibody treatment and certainly the TOGA trial provides a very good biological rationale for that based on they confirmed that if you do a HER2 selected population that you then target with an antibody that you can significantly improve survival outcomes. So that’s what we’re looking for but, I guess based on this result, if there are subpopulations who do benefit they’re likely to be relatively small populations of the order of maybe 5% or so because if it was something more frequent than that we would have expected it to have a more positive effect overall.

Now, could it be that you were on the right track but that the disease was so advanced that the specific therapy, the anti-EGFR therapy, didn’t have much of a hope of working?

It would certainly be standard to start investigating this sort of therapy in an advanced disease setting so you couldn’t really investigate it…

Any other way.

… in the first instance, in an adjuvant setting or an operable disease setting because of the risks of harm. So I don’t think there’s anything really that would firmly say this was the wrong setting. As I say, I’m not sure that we can, at the moment, fully explain the result that we have until we’ve got all of the full accompanying tissue and translational work.

Do you think that this particular molecular feature could be exploited in the future?

Yes, I think perhaps not EGFR over-expression as such, and certainly in other cancers such as lung cancer it’s known that EGFR over-expression in itself is not a very good biomarker for therapy and lung cancer mutation certainly seems to be important. And in other cancers amplification, which is a much less frequent finding, is seen to be important. So possibly these less frequent events which will obviously, therefore, be in smaller subpopulations may be the biomarkers that we’re looking for.

Now in terms of practical messages for managing oesophago-gastric cancer, what do you think are the messages that doctors should be learning from your experience here, because this is a big study?

Yes, I think the first thing to say is that the control arm performed very consistently, so EOC as a chemotherapy standard regimen does seem to be quite efficacious. In the RIO 2 trial that had an overall survival of 11.2 months and now in this trial it’s 11.3 months in the control group so that does seem to confirm that that does represent a good standard first line chemotherapy regimen. As regards to panitumumab, I think this trial confirms that panitumumab, at the moment, should not be combined with chemotherapy in an unselected population. There is an on-going study, the EXPAND trial, which is examining in a relatively similar population the addition of cetuximab to a cisplatin based chemotherapy backbone so it’s actually asking a very similar question but with slightly different drug combinations. That will obviously add more information and evidence as to the usefulness overall of anti-EGFR antibodies in combination with chemotherapy in the setting of oesophago-gastric cancer.

You’ve established, then, that this quite powerful therapy of three very strong drugs, epirubicin, oxaliplatin, capecitabine, is actually doing a lot of good in this disease.

Yes, and we knew for some time that without any chemotherapy the average survival is somewhere round about 4-6 months at best. Previous chemotherapy standards such as ECF are somewhere round about 9-10 months and this certainly seems to be at least as good as those previous standards and performing perhaps slightly better in these two trials. So I think it certainly supports EOC as a chemotherapy standard but perhaps not the addition of panitumumab.

And your brief take home message for doctors then, to sum it up?

I think that we’re awaiting the further biomarker analyses, there may yet be some positive results or some subgroups who do benefit but at the moment the combination of anti-EGFR therapy with chemotherapy does not seem to be beneficial in this setting and, as I say, further results from the on-going EXPAND trial will also add further information on that.

Tom Waddell, great to see you. Thank you very much.

Thank you.