17th Congress of EHA, Amsterdam, 14-17 June 2012
Lenalidomide in low-risk MDS del(5q)
Professor Aristoteles Giagounidis – St John’s Hospital, Duisburg, Germany
Professor Giagounidis, it’s really great to have you on ecancer.tv because you are the expert on MDS and I gather that at the moment low risk MDS is really in the focus. You can do something about it, you’ve done a study, the MDS-004 study. Can you tell me why, first of all, you wanted to study low risk disease?
Yes, so the interest in low risk myelodysplastic syndrome is that these patients are expected to have a relatively long survival and most of those patients at some point will become transfusion dependent. We know from pre-clinical data and from registry analyses that patients who are transfusion dependent have a bad overall survival and a higher leukaemic transformation rate. So what we are interested in is to study this population and to see whether we can do something about those patients to prevent them becoming transfusion dependent or reverse transfusion dependence if they are transfusion dependent at baseline. So one sub-group of myelodysplastic syndrome patients, a relatively large sub-group, 15% of those patients have the so-called del(5q) chromosomal abnormality which means they are del(5q) MDS low and intermediate-1 risk according to IPSS, and all of those patients in the lenalidomide MDS-004 study were transfusion dependent. So studying lenalidomide’s effect in patients with transfusion dependent low or intermediate-1 risk MDS was important to see: Is it safe? Does it lead to transfusion independence? What are the side effects, and what do these patients die of?
And why did you restrict yourselves to deletion (5q) patients?
There is another study, the MDS-005 study, on-going at the moment studying non-del(5q) patients. This is an on-going study, I can’t comment at the moment about that study. What we know is that in previous studies, in the so-called MDS-002 study, non-del(5q) patients have been entered and they do also show responses but del(5q) has been more in the focus in the past because lenalidomide seems to be especially efficacious in this sub-group of MDS patients.
OK, it’s quite an elegant study. The MDS-004 has around 200 intention to treat patients and split into three groups; could you explain what you did?
So what actually happened is that the study was designed to have three arms: 10mg of lenalidomide 21 days out of 28; 5mg 28 days out of 28 and placebo. Now for ethical reasons in a transfusion dependent patient population it was difficult to say that the placebo patients would stay entirely on placebo throughout the clinical trial so a cross-over was allowed after four months. So after sixteen weeks these patients got a bone marrow and we looked at response rate, progression rate etc etc etc. And then those patients on placebo could cross over to 5mg and were allowed then to take lenalidomide also.
Now what was already known about the potential effect of lenalidomide in those patients?
OK, the very first study of lenalidomide in myelodysplastic syndrome included a relatively high proportion, 25%, of del(5q) patients and it showed an erythroid response rate of 83%. So there was a previous study, the MDS-003 study, which was a phase II study and every single patient went on lenalidomide, there was no placebo arm. And in that trial again 65% of MDS patients with del(5q), transfusion dependent MDS patients with del(5q), received lenalidomide. Now this was enough for the American Food & Drug Administration to approve lenalidomide in myelodysplastic syndrome with del(5q) but the European Agency said, “You know, we would like to see a comparative trial to placebo”, so MDS-004 was set up. Now MDS-004 showed that in the 10mg and the 5mg arm the response rate was significantly higher than the placebo arm; it was around 60% for transfusion independence and relatively long term transfusion independence in the treatment arms compared to about 5% in the placebo arm. After four months obviously the patients from the placebo arm could cross over and then they displayed the same response rate as those who were on lenalidomide at the beginning.
But you have been able to do the statistics and you can show a strong statistical significance of the superiority of the lenalidomide group?
Absolutely, so 10 and 5 were statistically significantly superior with a P of < 0.0001.
Hence you allowed crossover?
No, no, no, in the crossover, during the crossover treatment period of sixteen weeks this is crystal clear, and afterwards those patients who cross over, the patients who cross over, also have a response. So the drug works and there was no question about that. The question that the EMA really raised is does lenalidomide show any signs of leukaemogenicity? So could it be that in fact what happens is you treat the patients with lenalidomide and while a non-treated patient would be transfusion dependent but remain relatively stable, with lenalidomide you get a nice increase in haemoglobin but then at some point the patient actually transforms to AML and this is the impact of the drug treatment.
And what did you find; because you had two different dose levels as well of lenalidomide?
Right, so we had 10mg and 5mg and placebo and one important information is that there were eleven patients who never crossed over, eleven patients who never got lenalidomide because of several reasons, one of which was that, for instance, they were not transfusion dependent, they entered the trial and remained relatively at a baseline of 8-9 g/dl and never got transfusions and in those eleven patients who never crossed over 36%, four patients, actually developed acute myeloid leukaemia. Now, in the other groups there was no difference in the transformation rate; after three years, for instance, we found about a quarter of patients that transformed to acute myeloid leukaemia and interestingly this was exactly the same for the 10mg treated people, for the 5mg treated people and for those who crossed over after placebo treatment showing, or at least pointing to the fact, that if a patient has a response, and this could be shown in the trial too, if a patient has a response he is much more unlikely to transform to acute leukaemia than if he has no response and especially if the patient has a cytogenetic response this is a very predictive factor showing that the patient would not transform.
So what shakes out of all of this in terms of dose and also the scheduling, because one group of patients had four weeks out of four, the other had three weeks out of four?
Right. It seems that the dosage that a patient gets within the first two cycles does predict the response, so if you give the higher dose, this is my personal recommendation, if you give the higher dose 10mg, 21 out of 28 days, if you give the higher dose this is probably the most efficacious dose. There are several hints to this: a) the transfusion independent rate was higher than in the 5, although not statistically significant, but also the cytogenetic remission rate was higher for the 10mg dose than for the 5mg dose. So I do think that it’s safe to say that the higher the dose the better for the patient.
Now there is some historical data on this as well because there is MDS-003 and 4, a registry study, am I right?
And also there has been an evolving understanding of 5q deletion disease. How would you sum up what’s happening and where it’s going; where it’s got indeed?
OK, let me sum this up in a very short and clear way. So a) the first thing that happened is it has been shows that del(5q) disease is much more complex than people previously thought. Previously we thought that if you have a del(5q) patient he is a good risk patient, don’t worry about him and he will be transfused but he will probably survive for a very long time. What has emerged is that del(5q) is very heterogeneous: a) there are patients who have additional chromosomal abnormalities, at higher risk of acute myeloid leukaemia transformation and of death; b) transfusion dependent patients have a much worse overall survival and progression to acute myeloid leukaemia than non-transfusion dependent patients and c) patients who have del(5q) and have an increase in medullary blasts have a very bad overall survival.
So what is the call for action for cancer doctors?
So I think what we can say is that, this is from registry analysis what I just said, it is from registry analysis. Now treating those patients with lenalidomide gives you two-thirds of those patients’ erythroid response rate and about 25-40% complete cytogenetic response rate. Those patients have a really relatively good overall survival and are at lower risk of transformation to acute myeloid leukaemia. The registry analysis you alluded to compared patients who were treated with lenalidomide to about 400 patients who never received lenalidomide and they could show that the question of leukaemogenicity of lenalidomide is probably not true, so that patients who were treated with lenalidomide did not have more leukaemogenesis than patients who were in the registry and who never received lenalidomide.
Now purists would say that, even though you had the difficulties lenalidomide had been licensed in the US, it would still have been better to do a more pure study with a control group.
So do you think other studies need to be done and what would you summarise doctors should now be doing with their patients who have relatively low risk myelodysplastic syndrome with del(5q)?
There is a lot of debate around this topic and we had a lunch debate at EHA where Eva Hellstrom-Lindbergh from Sweden, and Pierre Fenaux from Paris were actually debating exactly around this question: what’s the indication for lenalidomide in the del(5q) patient population? In my opinion it’s safe to say that lenalidomide does lead to good responses, that leukaemogenicity is probably not lenalidomide induced and if you ask me what to do with a patient with a single del(5q) transfusion dependent who comes to a doctor’s practice and presents at that practice I would say, OK, you put this patient on lenalidomide and you try to get this patient into complete cytogenetic remission. Preliminary data show that if you can sustain complete cytogenetic remission for a while you can even safely stop treatment with lenalidomide after six to twelve months in a complete cytogenetic remission. So even if you were concerned about leukaemogenicity of the drug you probably can even come off the drug and keep the patient transfusion independent for a very, very long time.
So there’s room for clinical judgement here and not all of the evidence is in but you’ve got some good ideas. What about the future though: other IMiDs or other agents that might help clarify everything?
Yes, so I don’t think we need an additional study because first of all the del(5q) patients, this is not a very big patient population and if you want to do long term comparative trials, placebo controlled etc. it’s not going to give us any information before 2020 or so. So I think the evidence is hard enough to say that lenalidomide can safely be given to those patients. On the other hand, what happens to patients who fail lenalidomide? What happens to patients who respond to lenalidomide and then lose their response? All this needs to be looked at and preliminary data show that for those patients who lose the response to lenalidomide, if you give them a drug holiday of four to six months you can restart the drug and for those patients who fail lenalidomide we actually don’t have any evidence that any other drug treatment could help but, as you alluded to, pomalidomide or other IMiDs may have a role here and I am sure that studies will come up discussing this.
Very briefly though, finally, toxicity comes into it. These are low risk patients so are you justified in using an IMiD at all?
A very good point. These drugs lead to neutropenia and thrombocytopenia in about two-thirds of the patients. These drugs can also lead to diarrhoea, skin changes, muscle cramps etc. The main concern though is thrombocytopenia and neutropenia. If you go through the first eight weeks without serious thrombocytopenia and neutropenia then the patient is actually relatively safe. During the first eight weeks these patients need to be followed up every week and you need to look at their blood values and you may need to give them growth factors for neutropenia, you may need to drug reduce or even stop the drug for a while if they are very thrombocytopenic. So I agree with you - it’s a low risk population, don’t put them at risk, watch your patients but what we have seen from the 300 odd patients treated in the different trials is that the toxicities are manageable and therefore I think the drug can actually safely be given.
Aristotoles, thank you very much.
It’s a pleasure.