The PANAMA trial is a phase II trial, a randomised phase II trial, that addresses the important question of maintenance therapy. We had a certain lack of evidence concerning maintenance therapies with EGFR antibodies, the PANAMA trial was designed as follows. Everyone in the trial that was included received six weeks of FOLFOX plus panitumumab and those patients that were not progressing, that were not eligible for secondary surgery, not dying or not having serious adverse events prior to randomisation were randomised into 5FU/Leucovorin plus panitumumab or 5FU alone which is a bit of a unique design because, at least from my personal point of view, I’ve never seen a 5FU/Leucovorin maintenance alone arm. So this is the first trial providing such an arm.
The primary endpoint was progression free survival of the maintenance treatment. It was possible to do a reintroduction of the induction regimen, so giving FOLFOX panitumumab again which did not really happen in those patients receiving panitumumab maintenance and happened more often, although not perfectly, in patients that received 5FU maintenance. That was the whole structure of the design so it’s not a very randomisation and treatment until progression design, it’s a bit more complex because we have this induction and this randomised part. But if you really want to invest in maintenance that is at least a clear question that we are really analysing maintenance therapy and not induction and maintenance therapy together.
What was the methodology used in the trial?
The hypothesis was superiority of the 5FU panitumumab arm. The hypothesis that we had was a relative improvement of 25% corresponding to a hazard ratio of 0.75. It was 280 events needed for that analysis and, of course, due to the structure of the trial that was really a challenge to estimate the numbers that need to be recruited, need to be randomised and then have an event in that context. So we nearly included 400 patients, 387 to be precise. We had 377 patients treated, of those 265 were randomised and the full analysis set which actually also received maintenance – was not dying or had a resection for obstruction or anything else – was 248 patients.
What were the key results?
The key result, of course, is the primary endpoint. As always with trials it was the progression free survival which was improved relatively by 28% with a hazard ratio of 0.72 and the p-value was 0.014. So the message for the progression free survival was absolutely clear.
They key question with these trials designed for progression free survival is always what does the overall survival report? On that question the overall survival, we have to make that point, is currently immature to have a definite conclusion as only 54% of the events have been recorded. Numerically the overall survival also favours the panitumumab containing maintenance arm but we have to admit that it’s definitely too early to take that as granted. I think we will have to wait another year, maybe 1½ years to record the necessary amount of events to make a definite conclusion to which extent the maintenance therapy really affected and to which extent.
How can these results have an impact on the future treatment of colorectal cancer?
They fill an important gap. For those patients that want to do active therapy the whole question of maintenance therapy is a cornerstone, is regularly asked. We have to be fair and say there are many other situations in metastatic colorectal cancer like secondary resection, stop and go therapies, other induction therapies. So the generalisability is always limited, as always, with cancer trials but if investigators and doctors and patients decide for active therapy as the maintenance therapy of 5FU panitumumab, together with the Italian VALENTINO trial that was published two years ago, it appears as the preferred option. That’s quite clear and I don’t think that will have many more maintenance trials in the future so basically the superior maintenance arm of that trial is rather that that we will practice in the future.
Important to mention of that trial is that there was a comprehensive biobanking which might allow in the future for identification of benefitters, non-benefitters. We have blood sample collections for extensive translational research, some quality of life assessments also focussing on skin toxicity and all these things that are important aspects that will help us with a definite interpretation of the trial in the future. It’s presently too early to have all these data ready because it’s a really young trial so that was the first round of data. Of course the most important set of data but I think there are a few aspects that I’m really curious about that we will present at future meetings.