IGCS 2012 Vancouver, BC, Canada
Gestational trophoblastic disease tumours in high risk patients
Dr Michael Seckl – Charing Cross Hospital, Imperial College London, UK
Michael, welcome to ecancer television. I’m delighted you’re from my old college, Imperial College, and you’re talking about very, very high risk patients but in an interesting disease, gestational trophoblastic tumours. I’d like to ask you, before I ask you about the high risk patients and how you manage them and how you improve their prospects, what has happened in gestational trophoblastic tumours recently? Has the outlook improved?
Yes, gestational trophoblastic disease is a spectrum of disorders that span from the pre-malignant conditions of complete and partial hydatidiform moles through to the malignant conditions of invasive mole, choriocarcinoma, and the very rare placental site trophoblastic tumours. These disorders used to be almost uniformly fatal but fortunately, over the last fifty years we have developed effective chemotherapy regimens that have completely revolutionised the outcome for women with this.
So hydatidiform moles are benign and yet they can be disastrous?
Yes, I prefer to think of hydatidiform moles as pre-malignant conditions because if you think of them as benign, you might be tempted just to ignore them and all of these women require careful monitoring of their pregnancy hormone.
But the ones that are eventually malignant, many of those have originated from the benign moles, or the so-called benign?
Or the pre-malignant condition.
Pre-malignant, yes.
And overall if you put all moles together, about 10% of them will become malignant, behave in a malignant fashion. But it is also possible to develop a gestational trophoblastic neoplasm, a cancer if you like, after any type of pregnancy: it could have been a term delivery, an ectopic, a miscarriage. So women, for example, will have developed a perfectly healthy child and some months or years later present with widely metastatic disease and unless the doctors concerned are smart enough to measure the HCG they may miss the diagnosis. And that’s important to recognise because it’s very curable.
So the mother, the woman, thinks she’s having another baby?
Yes, the woman may have a completely normal child and be totally unaware that she has, or is harbouring, a problem. This is a cancer of the placenta and it can lie dormant, silent if you like, inside the patient for many years.
Now the standard treatments are to remove it but in some situations you do need to use chemotherapy as well.
For molar pregnancies, the pre-malignant form of the disease, this tends to present in the first twelve weeks of pregnancy and these women present with bleeding in early pregnancy, they see their doctor, have an early ultrasound and the ultrasound shows that the pregnancy looks abnormal. You can’t make the diagnosis on the ultrasound but it just doesn’t look right. So the pregnancy is then terminated and you get a pathological diagnosis saying this is a molar pregnancy and that’s when the monitoring begins. But it’s also possible for a woman to have an entirely normal pregnancy, everything looks fantastic, delivers a normal child at the end of the day and then some months or years later presents with choriocarcinoma or a placental site trophoblastic tumour.
To confuse matters even more, you can have a foetus co-existing with the tumour, can you not?
Yes, that’s true. So we do occasionally see twin pregnancies where we have a completely normal child growing alongside a molar pregnancy. In former times folks used to think that the sensible thing to do would be to terminate the pregnancy because of the risk of malignant change in the mole and the risks that the mole could cause pre-eclampsia or perforate the womb or some other life-threatening event for the mother. And it was considered that it would never go to term, the pregnancy could never deliver the healthy child. What we were able to show in a paper back in 2002 is in fact that about 40% of those pregnancies can actually go all the way through to term and deliver a healthy child and that there was no increased risk of developing malignant change in the mole and there were no maternal deaths. So it seems reasonable to allow such pregnancies to go on. The other situation where you do sometimes see a baby, if you like, or it appears to be a baby alongside this abnormal material is in what we call a partial mole. With a partial mole you can get a variable amount of normal foetus, apparent normal foetus, together with the abnormal placental tissue. I say apparent because in fact it’s triploidy across the genome so in fact the foetus is never viable even if it does get to term.
So in practical terms, what are the clinical challenges facing the doctor?
For molar pregnancies the practical challenges are is there some way in which we could identify which mole is going to be malignant and which one isn’t. At the moment we’re dependent on many months of monitoring with HCG which causes huge amounts of anxiety for the affected women. And although we now cure women with molar pregnancies with virtually a 100% success rate, the problem is that many of these women, about a third of these women, require switches in their treatment; the single drug therapy that they start on isn’t enough and so we need to refine our ability to stratify patients for the most appropriate therapies. So that’s, if you like, the low risk group of women who have a low chance of becoming resistant to single agent therapy. At the other end of the spectrum, of course, we have the high risk women, these are the women that present usually after pregnancies with metastatic disease, with choriocarcinoma. These women used to have a survival rate with multi-agent therapy of around 85-86%; with recent developments that we’ll be talking about at the meeting today we’ve bumped up the survival rate to near enough 95% at five years which is a pretty impressive absolute improvement in survival.
So what are those techniques that have achieved this improvement?
If you look at what causes the deaths in women with high risk disease, it is a combination of early deaths, within the first four weeks of getting to hospital because of overwhelming disease burden, haemorrhagic complications as you start the chemotherapy or multi-organ failure. And later on it’s because of multi drug resistance. A small proportion of cases in fact turn out to be what we call non-gestational choriocarcinomas, so to the pathologist they look just like a gestational choriocarcinoma but actually when you do the genetics you discover that it has no paternal genes in it. This is a disorder of pregnancy, it has to have dad’s genes in it, so if there are no paternal genes present this tells you that this is not a gestational tumour. So, for example, it’s well known that lung cancers, a small proportion of them, or gastric cancers or any other epithelial cancer can occasionally differentiate to look just like a choriocarcinoma.
What are the techniques, then, that you’ve used to actually overcome these problems, like early death?
With the early deaths we realised that patients who were presenting with what I call ultra-high risk disease, disease that’s in the liver, in the brain or overwhelming disease burden in the chest so that they are almost on ventilators, but if you get started with standard doses of chemotherapy it’s just too much for the body to cope with. These are highly vascular tumours and as a consequence of the tumour melting away you then get huge haemorrhagic complications in the body and the body just can’t cope.
So how do you titrate an individualised therapy?
In these circumstances we introduced low dose induction chemotherapy with etoposide and cisplatin, just 100mg/m2 of etoposide and 20mg/m2 of cisplatin on days one and two. That seems to be enough to really almost completely eliminate the early deaths.
So has that been established empirically or do you have hard data on this?
When we established it, if you like, empirically back in about 1994, 1995, we introduced this protocol and we just audited the results in the last year. It’s very clearly shown that in this time period between 1995 and where we are today we’ve had only one early death whereas in the past time period we’d found that we’d had about fifteen early deaths. So we’ve really cut out the early deaths. Now you could say is this because of stage migration, was it that we were simply seeing far more advanced cases in the preceding period? But in fact that’s not the case, we’re seeing just the same degree of advanced cases before as we are now. So that’s not the explanation. I think the explanation is staring us in the face – you can’t give big doses of chemotherapy in people with very advanced disease.
A recommendation, then, might well be to send your patient to a centre of excellence but are there any guidelines that you would give doctors?
The way the service is organised in the UK is exactly what you’ve just intimated which is that the service is centralised. I think that has been crucial for our learning ability. You could imagine if you’re due to have a hip replacement, do you go and see the surgeon who does one operation every five years or do you see the surgeon who does 2,000 a year? It’s not rocket science to work out that if you’re seeing the surgeon who is seeing lots of cases you’re more likely to learn from your mistakes. An example of this was that we had two women who arrived within a month of each other with very advanced lung disease and both of them, I should say, arrived on ventilators so they were being ventilated. The first lady, I couldn’t persuade our anaesthetist to take the ventilator off, they just basically said, “There is no way we can switch this off, she has to keep being ventilated,” and the positive pressures actually burst the blood vessels feeding into the tumours in the lungs, she bled to death into her lungs. The second lady that arrived a month later, the anaesthetists were a bit more convinced that it would be a good idea not to intubate, or keep her intubated. They removed the tube and this woman has survived even though she was terribly hypoxic for about two or three days.
Fascinating. What would be the bottom line, the few messages that you would leave everybody to think about?
I think from our point of view, if you ever see a woman with metastatic disease who is of child-bearing age, always measure the HCG because you might save a life. You don’t need a biopsy to prove the diagnosis, it’s a clinical diagnosis. Don’t delay, start the chemotherapy as soon as possible. This is a rampant disease when it gets going and if you’ve got very extensive disease, start with low dose chemotherapy. The etoposide and cisplatin combination that we’ve developed seems to work very nicely in this situation. Once the disease is under control by all means get started with a more intensive chemotherapy.
Dr Michael Seckl, it’s great to hear from you; fascinating stuff. Thank you for being on ecancer.tv.
