Overview of germ cell tumour research in North America

Share :
Published: 29 May 2013
Views: 6206
Dr Stewart Goldman - Children's Memorial Hospital, Chicago, Illinois, USA

Dr Stewart Goldman presents data on developing research in North America at the 3rd International CNS Germ Cell Tumour Symposium in Cambridge, UK.

Our assignment was to give the different perspectives from both Europe, Asia and the US and so I thought I would give actually one man’s North America perspective, emphasizing some opportunities that we’ve lost, some lessons we’ve learned and some new emphasis for the future.

I did not put a conflict slide in there, as Dr Matsutani did. I have many psychologic conflicts but none that involved drug companies or off label use. And I am willing to discuss those over the DNA beer.

From the US perspective, or the North American perspective, we’ve actually had very few national trials. We are now in a new era of biologic emphasis that we’ll hear about later today. That we’ve had some consistency in treating relapse with new protocols has not been our past experience but I think with some pioneering work done, especially through LA and other sites, we are now going to have that in the near future.

Looking back at some of the national trials we’ve done, Dr [?? 1:09] from the Paediatric Oncology Group, which is now combined with the Children’s Cancer Group for the COG, she did a study, a somewhat small study, but looking at patients with germinoma with pathologic confirmation and HCG of less than 50 versus high risk patients who were those with the higher HCG or Alpha-fetoprotein or NGGCT. Actually the key message here that we’re going to focus on is it was using a cisplatinum based regimen and whether you’re low or high risk an increase in the dose. Post-induction response at week 12 to the chemotherapy was noted and then patients had a slight change in their radiation. The results of this trial showed that 83% of the low risk patients were responders with 33% as a CR and 50% as a PR with good survival statistics. High risk patients had fourteen patients included whereas one toxic death, four were non-evaluable, two CRs and three PRs. From this first experience we learned that national consortium trials in the North American experience are possible, that risk stratification is also possible and beware DI, beware DI, beware DI and cisplatinum. I think that that has actually affected many of the protocols we’ve done since. One of the questions we had to face was is carboplatinum equally effective as cisplatinum and cisplatinum is certainly a little harder to manage with patients with diabetes insipidus.

When Dr Matsutani and Dr Calaminus and I talked no-one was going to grab the international studies because they didn’t fit in just one continent. So I will briefly mention that this was another historically important, there were concerns for long-term sequelae of both the cognitive and growth and endocrine function as well, hearing loss and secondary malignancies. So there was an effort to eliminate radiation and many of the people in the room were directly involved with writing and running these studies so I’ll just briefly mention their overall importance. They were based on an excellent response rate to platinum compounds for systemic GCT, especially cisplatinum. They looked at responses to four cycles of chemotherapy and that dictated the further treatment. This was one of the first studies where second-look surgery and intensified chemotherapy really came into rage.

Now, to give a long three wonderful studies a very short conclusion, the conclusion is that chemotherapy alone was inferior to chemotherapy plus radiation regimens. Now, I know Dr Finlay will raise his hand once it’s over because there certainly are a significant proportion of patients that we can treat without radiation but I think that is the accepted conclusion of these studies. And it is shocking here, I see Dr Finlay’s head shaking in agreement with me.

One of the great other moments in North America came from some wonderful work done by Dr Allen and the group at Beth Israel where they looked at a cohort of 38 patients with germinomas. Actually it’s interesting that 55% of them had either small amounts of disease, M disease, and again looked at how they could get patients into complete remission after two cycles of a cisplatinum based chemotherapy. Those who didn’t got two more cycles and ultimately most of the patients could get a response, either a CR or a PR or down to minimal residual disease, with this chemotherapeutic regimen. In this regimen if those who got to CR or minimal residual disease got 30Gy involved field RT and a median follow-up of 46 months they had an 80% relapse-free survival rate.

I think that this important study done by Dr Allen and colleagues led to the management of germinomas, our first COG trial, ACNS0232. I bring this study up because I think this is, unfortunately for us in the North American perspective, an opportunity lost. This was a really great novel study that I think deserves a lot of credit. Because in there the group including Dr Allen looked at looking at modifications of occult multifocal disease, looking at the patients who had slight disease or who have DI with a pineal lesion. Very interesting, really important work. They looked at changing the radiation regimen, it was a randomisation between radiation therapy alone and chemo plus RT including patients with occult multifocal disease or disseminated disease and choosing the radiation dose and field. I don’t put this slide up there to give the exact… doses aren’t important, just to show the concept of how this was changed. Using chemotherapy with carboplatinum based and if you didn’t respond completely to get cisplatinum and Cytoxan as dose intensification. Those patients who then had regimen B, again their location, their dose of radiation was affected by their response.

Now this was a great study that had important questions to ask; it was our first COG germinoma trial. I think it showed an intriguing precedence for staging. All patients had central review and more than one time for both pathology and radiographic findings. Unfortunately accrual was slow and really my question is were we ready as a national consortium for this trial because this was a really beautifully written trial.

But we are now moving towards the future and a little bit later this morning Dr Bartels will talk about the new COG trial for germinoma. I’m not going to speak on her trial as I know she will do it herself. In the COG we were able to perform a phase II trial for nongerminomatous germ cell tumours, secreting tumours. This is the trial I ran; the basic outline is that patients got induction chemotherapy using carboplatinum BP 60 and Ifos BP 60 based on the French regimen. Those who went to a CR went to get radiation therapy, those on less than a CR second look surgery was not mandated but strongly encouraged. Those who then, after second look surgery, were brought to our definition of a good responder, meaning a 65% decrease in tumour volume, normalisation of markers and no progression or dissemination, went on to craniospinal radiation. Those who were not could have a test of further intensification using consolidation chemotherapy with a high dose regimen of thiotepa and BP 16. Out of 102 eligible patients we only had two patients that went on on study to go for high dose chemotherapy. Both of those are long term survivors. Unfortunately you learn how to write your study after your study is open and patients who progressed on trial, many of them did go on for high dose chemotherapy and some of those were salvaged and are doing well but by study criteria they’re called progressive disease and the rest of the data is not really captured. That was a mistake too late to recuperate from.

I don’t want to go over the eligibility but just to show that we used 50 as our HCG cut-off and this is one of the issues, I think, as we compare trials between the different continents, that we’re using different terminologies for what we consider high risk germinoma, something  I hope that maybe a consensus can be made for future trials that we use the same definitions.

We had 104 patients enrolled, two were ineligible. Just as Dr Matsutani talked about the basal ganglia thalamic tumours, the hypothalamic tumours were present and I think a subset analysis will show some difficulties. One thing about the pie chart I’m going to quickly mention, one of the other foils on our study is if you came in as a marker only you were graded as a mixed germ cell tumour. So I think we lose some of the histologic significance and when we do our analysis by histology it’s lost in our trial because patients were coded if they came on as marker only as a mixed germ cell tumour.

One of the other lessons I learned on this trial is when you look at it in the North American style about 30% of our patients underwent a biopsy or resection at the time of diagnosis. 29 of those 32 actually had resections, three were biopsy only, one gross total, one extensive, eight subtotal. So basically about 68-70% of our patients were enrolled basically on marker elevation alone which leads to another lesson learned and one of the difficulties in all national trials is some of the surprises we see. About 70% of our patients had complete serum and CSF marker evaluation performed at the time of enrolment, 30 were missing at least one marker. Now certainly I don’t have to say to anyone in the audience there are patients where it’s unsafe to do a lumbar puncture but what’s interesting is 13 of these patients had no CSF markers, two which also had failure to obtain serum AFP but had a significant HCG so the patient could go on trial. Four were missing beta HCG but had everything else; four others were enrolled without CSF, AFP and everything else. Three patients had serum markers, without serum markers but CSF markers, and four were missing only serum AFP and two others were missing serum beta HCG. Is this critical to ruin our trial? No, but I do think it makes some of our evaluations more difficult and this is only at baseline.

Our response to the chemotherapy, our event free survival and overall survival on this trial is quite good. We had institutional reports of their CR and PR rate and then a central review by and we had a discordance where the institutional reports actually downgraded response, I was shocked to find out, and our independent panel said that about 87% of patients had either a CR or PR to our initial chemotherapy.

If we look at our overall survival and event free survival rates we see that five year event free survival is 84 ± 4% and overall survival of 93%. Now I do have some concerns about these survival curves that I’ll mention in a second. When we looked at the HCG using a number of 100 we could not find a statistically significant difference when we looked at 1,000. Again the p-value is not significant but we move forward, and AFP is our most significant finding of a high AFP at diagnosis.

Now this led to some really great work by our investigators, Dr Bartels, Dr Dhall and Dr Khattab who have now put together the next trial to discuss. But they went and looked at a subset analysis and they looked at event free survival and overall survival for localised patients considered by the treating site, remember they had even a less objective response rate, as having a CR and PR and those who were found to have fibrosis or teratoma or no tumour on second look surgery. If you look at that subset the overall survival was 97.9% with event free survival of 91.6%. So I do think one of the lessons we’ve learned from 0122 is we can pick out a population of patients who we can then further try to reduce therapy.

At second look surgery much has been found in other sites. We have a high propensity of teratomas or fibrotic tissue and only two of our fifteen patients still had either a mixed germ cell tumour or embryonal carcinoma. Those who had second look surgery at the time of progression still had, four of them had, a growing teratoma syndrome. Once again they were taken off study and coded as PD as we did not make a spot for that on the trial.

So what lessons have we learned from 0122? It has a very good five year event free and overall survival. We do feel that we’ve shown that carboplatinum can replace cisplatinum. We do think that second look surgery is extremely important as we move forward. We think growing teratoma syndrome is important and we have had one late recurrence at 5.75 years who was coded pathologically by their institution as an adenocarcinoma consistent with the original tumour. It didn’t quite make sense to me but I believe since the original tumour had teratoma in there this is really a malignant transformation.

This leads to the new study that Dr Fangusaro will be discussing this afternoon, ACNS1123, which will move forward our field. So, from the North American experience I believe we have a new emphasis on late effects and quality of life and that’s one thing I think the new study that will be discussed later today really is emphasising.

I think that we show that we need to have multi-institutional trials for the relapsed patients and again Dr Finlay and Dhall with gempox we’ll hear about dasatinib with Dr Osorio and Finlay and the importance of biology and I know that Dr Lau will be speaking of this later today.

We have many questions for the future, neurocognitive function, what are the baseline issues. As we know, many of these patients have issues even prior to therapy, and what do we save with our different regimens? What stratifications make for the best and least toxic outcomes? Can we stratify better and what are the next therapies? Before Dr Calaminus asks us the same question I want to mention that as we were just finishing the manuscript and hoping it’s gone for ever, as it keeps coming back from COG, the next thing we’re going to do is compile the endocrine data to look at the endocrine follow-up on our patients. Thank you.