Dr Jayant Vaidya speaks about his work investigating localised intraoperative radiotherapy for breast cancer patients with invasive duct carcinoma. This technique makes use of x-rays rather than radioactive seeds, a feature that reduces radiotoxicity and the results of this study after four years are not inferior to those of traditional radiotherapy.
15th Congress of the European Society of Surgical Oncology (ESSO), 15–17 September 2010, Bordeaux, France
Interview with Dr Jayant Vaidya (University College London, UK)
Targeted intraoperative radiotherapy in low risk breast cancer patients
We have with us Jayant Vaidya from University College, London here in Bordeaux, you’ve stepped out of the meeting at the European Society of Surgical Oncology. Thanks for coming over here, but you’ve been doing an extremely interesting study of breast cancer, can you tell me all about this way of delivering radiotherapy?
Yes, it was back in ’95 that we started the clinical pathological rationale using whole organ analysis of mastectomy specimens that suggested to us that giving radiotherapy only to the area around a tumour is probably going to be enough. So we collaborated with industry and developed a machine to give radiotherapy as a single dose in the operation theatre.
Now this is intraoperative therapy for breast cancer. Which sorts of patients were you looking at?
In the TARGET A trial, which has just been published in the Lancet, we looked at patients who were over the age of 45 with generally small tumours which were not lobular cancers, so invasive duct carcinoma patients more than 45 years of age.
Now you’ve got a good size of sample, over 2,000 patients. What did you do with them and how did you divide them up into the categories of what sort of cancer did they have?
So they had invasive duct carcinoma and they were randomly allocated to receive targeted intraoperative radiotherapy as a single dose or they were receiving the standard control arm, the external beam radiotherapy which was three to six weeks of daily radiotherapy fractions.
And they were randomised?
They were completely randomised and stratified according to the centre.
But there was a situation where some patients could get both therapies?
Yes, this was within the protocol. Within the protocol, if patients were randomised to have targeted intraoperative radiotherapy, that meant that the policy of the randomised experimental arm was this, that they would first get external beam intraoperative radiotherapy and then if post-operatively we found adverse prognostic factors we would add external beam radiotherapy as well and that was this arm, they were not moving to the other arm, that was the arm. We expected this to happen in about 15% of patients and the study was powered to include this; it in fact happened in 14% of patients.
Could you tell me a little bit about the source of radiation because this is a short distance and avoids cardiotoxicity.
Yes. This is not radioactive seeds, this is X-Rays at 50kV and they are generated in a small box, electrons are generated, they hit a target and X-Rays at 50kV are generated. A spherical applicator goes inside the tumour bed to give radiation from within the breast. The dose at the surface of the applicator is 20 Gray and the dose about 1cm from it is about 5 Gray. So your heart, lungs, oesophagus are all spared the radiation that is not necessary.
It sounds very elegant in principle, what are the data coming out of this study?
It was elegant and we needed to prove that in fact it gives the same results and that’s why we did the randomised trial. We had 2,232 patients from 28 centres around the world who participated in the trial; so 1,113 received target as their policy, 1,119 received external beam. We found that the local recurrence rate in the target group was 1.2%, in the EBRT group it was 0.9% at four years. The difference between the two groups at four years was 0.25% with 95% confidence limits of -1 to 1.5. That means the difference, at most, could be -1, that means the target would be 1% better, or +1.5, it could be 1.5% worse. Our non-inferiority margin originally was 2.5% so what we actually found was ten times less than what we accepted as non-inferior. So that’s why we were very pleased with the result, that giving this risk-adjusted radiotherapy compared with one size fits all radiotherapy, we found a risk-adjusted radiotherapy that is delivered with target is as good as whole breast radiotherapy with obvious other advantages.
But in the risk adjusted group they could additionally have external beam radiation?
That is the whole point of risk adjusted. That means that if they were low risk they would receive target as the only treatment, which happened in 80% of patients. If they were high risk they received target plus EBRT, which happened in 20%, or only 14%. So risk adjusted radiotherapy was as whole breast radiotherapy, that is the comparison, not intraoperative alone which is dogmatic in a way. So what we are trying to say is that if this comes into clinical practice and if we want to deliver it to the patients in clinical practice then we must remember that if a patient is found to have a high risk factor, such as major lymph node involvement or if lobular cancer is found, then they should receive external beam radiotherapy as well. Only then we would get the results which are similar to the trial.
Now was the finding at four years sufficient or do you need to wait longer?
It is sufficient enough for us to be quite sure of giving this in the low risk patients and that is what many investigators are starting to think of doing. Many of us have started losing ??? in this regard because you see an 80 year old patient whose local recurrence risk is anyway very low and it would be so much easier for them to give this as the only treatment. What are we risking? At four years the upper limit is 1.5%, even if the difference, which was 0.25%, quadruples at ten years we would still get a difference of 1% local recurrence, which is not really very high. We are indeed looking for long-term data, we are monitoring all the patients, not only for local recurrence in the quadrant but in other quadrants as well and we will keep a watch to see how it goes. So these results are very cautious. One of the reasons why we feel these are important at four years is this: five years and ten years are only because we have five and ten fingers. In terms of the biology of breast cancer most local recurrence occurs in the first four years and that’s what we’ve found in the target trial, the peak occurs at three years. In the fourth year we didn’t have a single local recurrence.
The idea of risk adjusting, this approach, is very appealing and is an example of practical pragmatic medicine. But in fact you will have some critics saying that it would be better to have two completely different groups and stick to that and get very hard data. What do you say to that?
We have such a subgroup as well within the trial. About one third of the patients in the trial were randomised after they had the operation so they were then chosen because they were low risk. In that subgroup they were randomised so half of them received only target, half of them received only external beam. The numbers are enough, they have enough power even within that subgroup, to show non-inferiority. So the follow-up of that will continue and we have equal results in those. We have not done any subgroup analysis at this time because the recurrences are so few, but we have such a subgroup within the trial so we have clean data as well as pragmatic data in which we found together they are equal.
Do you think your evidence so far is enough to prove practice changing?
Some people think it is and it is up to individual clinicians to decide whether they want to wait for many years. Various practices which we have already changed in breast cancer practice, chemoadjuvants, adjuvant therapy, all of those results were published many times at two years, at five years, and they changed practice. So it is up to individual clinicians and to take it really in a very reasonable manner. As long as the risk adjusted strategies apply, I do not think patients will come to harm.
So what would you like to say to busy doctors who are trying to do the best for their patients, many of whom may fall into this category of low risk and in need of radiotherapy? What would you say?
I would say, having examined the evidence carefully, make sure that patients fall into the criteria of the target trial; inform the patient that we have these results at four years where they are equal and that the risk of having trouble is small. As long as they follow this practice it is adaptable.
You published quite recently, have you had any reactions yet?
Yes, we have had four letters to The Lancet, our response to the letters has been quite satisfactory and it will soon be published.
So what sorts of ideas are people coming up with?
One of the ideas is that the dose is too small. Well, the proof of the pudding is in eating. Radiotherapy may work in ways other than killing cancer cells. Today with the surgery we do, we hardly ever have positive margins. What I mean is if we had positive margins we go back in and make them negative. So at the end of the surgery margins are negative, radiotherapy still works so what is it working on? Here is the time that I feel that a small dose may change the tumour micro-environment making it less conducive to tumour growth. That’s why the smaller dose may be enough. So we have to re-think our biological models of how the radiotherapy works; it may be not just killing cancer cells, it may be changing the micro-environment. Therefore, a lower dose may be enough, especially in patients who have got such small cancers in good prognosis tumours.
Well that is fascinating work conducted in nine countries, I believe, and from your centre at University College in London. Jayant, thank you for joining us today.