Lung Metastasectomy: a Working Group of the European Society of Thoracic Surgeons
The European Society of Thoracic Surgeons is running a Lung Metastasectomy Work Group immediately before its Annual Meeting in May 2006 at Cluj-Napoca in Romania. I have been asked to chair the session on Trials and Registries and I am getting some ground work done. In this blog I will set out how I see the issues and would welcome information and comments.
What is meant by “lung metastasectomy”?
A feature of cancer is that it spreads. In some cases the primary cancer invades blood vessels, fragments break off and are carried in the blood. All the blood passes through the lungs before being recirculated and cancer cells are filtered out. From these seeds new islands of cancer grow. These are called “secondary deposits” or “metastases” - I will use the words interchangeably. The suffix “ectomy” refers to any surgical excision and so we get lung metastas-ectomy.
To narrow the subject to something manageable, I will limit this discussion to colorectal cancer. The two commonest primary sites for cancers are lung and breast but we do not knowingly perform metastasectomy for those; the most typical context for metastasectomy is in colorectal cancer.
The current practice of lung metastasectomy in colorectal cancer
When patients come to us with lung metastases we discuss them in our multidisciplinary team meetings. In advising whether to operate we apply widely accepted selection criteria:
1. fewer metastases
2. longer intervals since the primary cancer resection and
3. less aggressive Duke stage.
I am uncertain about how much good we do by removing secondaries and I believe that a proper randomised trial is needed. My colleagues argue:
· that the evidence is there
· that the patients and their referring doctors want the cancer out and
· a trial would be impossibly difficult.
We have evidence? What is it?
There are no randomised trials. The evidence for a survival benefit depends on observational case series or registries. Survival graphs are presented for different groups, separated by the presence or absence of predictive factors. Patients with favourable factors (listed above: fewer secondaries, longer intervals and less aggressive cancer) live longer.
The flaw in this evidence
Suppose I have a cohort of 100 students and I ask them to form a line, shortest at the front, tallest at the back. I then send the first 25 into class 1 and so on, with the last 25 into class 4. An unsuspecting statistician makes measurements and finds that there are significant differences between the heights of the students in the four classes. Well – no one would be that stupid. Let me make it a little more interesting. I have reason to believe that above average weight and shoe size, and male sex are predictive of greater height. For each of the students I note their weight and shoe size and calculate the mean weight and the median shoe size for the 100. The predictive features I choose are above average weight and shoe size and being male. I send those with three of these features into class 1, with any two features into class 2 and so on until the light and small footed females end up in class 4. I find a suitably blinded statistician and what is she likely to find? The tallest students tend to be in class I. That is absolutely fine but it is not a chance finding – I set it up that way.
Now suppose that I had previously engaged in some legerdemain involving snake oil to make children grow and I claim that my findings are the living proof that it works, and that it works better in heavier males with big feet which is why class 1 is taller than class 4. Replace shoe size, weight, and sex with the number of metastases, time interval and Duke stage, replace snake oil with surgery, and the height with survival and you have the evidence you were looking for. I have not proven that there is no survival benefit but I am unsatisfied with the quality of the evidence offered in its favour. I am not alone. See Torkel Åberg’s writings in the Annals of Thoracic Surgery in 1980 and again in 1997.
But patients want their secondary cancer out!
It sounds better than leaving cancer in there – it stands to reason – or does it? What is patients’ perception of the gain? They must have picked up some ideas from their doctors, so what do doctors think is the gain? The data are presented in terms of survival. The patients’ hopes and expectations are presumably for cure. No indication is given of whether surgery increases or reduces well being.
What do the doctors think?
Surgeons should have a clear idea of what exactly we hope to achieve with every operation we perform.
· In some circumstances it is for a survival advantage; examples are asymptomatic aortic aneurysms, left main stem coronary narrowing, and aortic stenosis. All cause death and in these instances death can be averted by surgery,
· In other instances it is entirely for symptoms; hip replacement and cataract surgery allow the lame to walk and the blind to see.
In cancer we might influence survival or symptoms - either, neither, or both - depending on circumstances. In lung cancer we very rarely operate unless we believe we are able to cure the cancer. On the other hand, in colorectal cancer there are reasons for removing the primary even when cure is impossible. If you have only months or a few years to live it would still be preferable to be spared bowel obstruction, tenesmus, rectal bleeding, sacral pain and fistulae into the vagina and bladder. There is an imperative for colorectal surgeons to operate to control the primary bowel cancer. It does not follow that thoracic surgeons have good reason to take out asymptomatic lung secondaries.
Experts advise it – so it must be right.
My mathematician colleagues, I discovered, believed that because metastasectomy operations are done, there must be a good reason. Let us just burst that belief bubble. In other times doctors as intelligent, as rational and as compassionate as ourselves have done things that we do not do now, including blood letting, cupping, organopexy, colectomy for constipation, industrial scale tonsillectomy in children, ever more radical surgery for breast cancer, and bed rest for everything. These are not now practised. It follows that doctors of the future will not do all the things we do now, but we do not know which interventions they will look back on as unavailing. But you can be sure that there will be some.
Do the expectations of patients and of their doctors match?
Let me challenge another implicit belief: doctors and their patients have a very clear, explicit and shared understanding of what will be achieved by an intervention. Wrong. Modern Western medicine is great at dealing with parts of the problem while missing the patient’s bigger picture. Anne Fadiman’s extraordinary book “The spirit catches you and you fall down” (1997) should be compulsory reading for residents – if only they had time.
I place great emphasis on the patients’ wishes but they must be well informed by doctors, who in turn must be well informed. The surgeon should be able to explain what that individual patient stands to gain or to lose by having the proposed operation. For surgeons to justify removing asymptomatic secondaries for so called “psychological” reasons seems to me to cross a boundary that it would be wiser to recognise and respect.
So what about a trial?
The first response when established practice is challenged is to say that a trial would be impossibly difficult. Difficult always, impossible never. Here is how I would approach it.
· Suppose a patient is sent to a surgeon five years after removal of a very favourable colorectal cancer and has been found to have a single lung metastasis which can easily be removed by minimal access (VATS) surgery. What would we do? Most would take it out even in the absence of sound evidence of a survival benefit.
· Suppose a patient is sent to a surgeon five weeks after an operation for advanced stage colorectal cancer and the chest x-ray shows a dozen metastases. What would we do? A clam shell incision to resect them? Given the absence of sound evidence for a survival benefit, most would not.
So here is the concept. Between these two extreme ends of the lung metastasectomy spectrum there must be patients for whom an oncologist, a surgeon, or a team feels uncertain about whether they are doing more good than harm. If you say “yes” in some cases and “no” to others, there must surely be a zone of uncertainty around the cross over. The trial need only include, indeed can only include, patients where there is uncertainty, that is to say equipoise. How wide the band of uncertainty is and whether it is at the same point for different teams does not matter in the trial design, provided there is unbiased allocation to surgical resection or not. It would be a pragmatic trial – all other treatments would be allowed. The research question is whether the surgery adds any advantage. And let us not forget, there must be some measure of well being apart from just survival. I do not want a futile thoracotomy in the last months of my life just so that some doctor can go home content with the day’s work saying “At least we tried.”