Eur J Cardiothorac Surg 2008;33:300-302. doi:10.1016/j.ejcts.2007.10.015
Copyright © 2008, European Association for Cardio-thoracic Surgery. Published by Elsevier. All rights reserved.
Pulmonary metastasectomy for colorectal cancer: weak evidence and no randomised trials
Tom Treasure*
Clinical Operational Research Unit, Department of Mathematics UCL, Gower Street, London WC1E 6BT, UK
Received 31 July 2007;
accepted 21 October 2007.
* Corresponding author. Tel.: +44 7957 168754. (Email: tom.treasure{at}gmail.com).
 |
Abstract
|
|---|
The practice of pulmonary metastasectomy is widespread. A web-based survey run as part of the ESTS Working Group on pulmonary metastasectomy has provided some insights into similarities and variations in belief and practice [Internullo E, Cassivi S, Van Raemdonck D, Friedel G, Treasure T. Pulmonary metastasectomy: the state of the practice in Europe. Interact Cardiovasc Thorac Surg 2007;6(Suppl. 2):S182. Ref Type: Abstract.]. What we do know from the work of the group and from a systematic review [Pfannschmidt J, Dienemann H, Hoffmann H. Surgical resection of pulmonary metastases from colorectal cancer: a systematic review of published series. Ann Thorac Surg 2007;84(July(1)):324–38.] is that there are no randomised outcome data to guide us. We rely on presumed benefit based on comparison with poorly characterised survival estimates for other patients with advanced disease. In this review I consider the evidence to date and conclude that a trial is needed.
Key Words: Colorectal cancer • Metastases • Metastasectomy
|
1. Introduction
|
|---|
As far as surgery for cancer is concerned, it can be argued that there have been more retreats than advances over the last 40 years. Surgical attempts to cure cancer ran up against unyielding obstacles some years ago and the obstacles remain. If the cancer is disseminated we cannot encompass it in a surgical field and then extirpation of all the cancer by surgery is not possible. Let us consider some of these retreats.
|
2. Surgical retreats
|
|---|
- 1. Lymphoma. There was a time when lymphoma was operated on and regarded as a surgical disease. Inspection of the results of surgery, understanding its behaviour, and the advent of better chemotherapy removed lymphoma from attempted surgical cure. Our role as surgeons is in biopsy, staging and some palliation. The reason is that we now view lymphoma as a systemic disease from the outset, no more operable than leukaemia. Surgery is limited to what is anatomically accessible but the bloodstream delivers chemotherapy to every cell. So lymphoma treatment is in the hands of medical oncologists who have made remarkable progress in that group of diseases.
- 2. Small cell lung cancer. A similar retreat occurred for small cell lung cancer and for a similar reason. It was the recognition of the failure in most cases of attempts at surgical cure, juxtaposed to RCT proven responsiveness to chemotherapy, albeit short lived in most cases, that led to a new terminology in lung cancer. Rather than being classified under a number of descriptive histological types (squamous cell, oat cell, adenocarcinoma and large cell undifferentiated) the diagnostic frames [3] have become simpler: we now classify the disease into small cell or non-small cell lung cancer. This was a new operational classification determined not by clinical features or histological appearance per se, but groups of patients best treated in different ways.
- 3. Breast cancer. As a young surgeon in the 1970s, I performed radical mastectomy for breast cancer, removing pectoralis major and clearing the axilla leaving bare the axillary vein. Some surgeons even opened the sternum to dissect out the internal mammary chain of lymph nodes. If the cancer was medially sited in the breast this followed the same perceived logic that applied to axillary clearance. This radical locoregional surgery was abandoned and breast cancer came to be viewed as a systemic disease from the outset. Put in contemporary jargon, R0 resection (that is removal of all cancer) came to be deemed impossible once metastatic disease was shown. It is strange that metastasectomy for breast cancer, which was really not considered an option then, is now apparently increasing in practice. It is noteworthy that one of the greatest exponents of this surgery includes breast cancer in his indications and has removed as many as 124 metastases in a bilateral operation in women with breast cancer and includes this operation in the R0 (no residual disease) group [4].
Are we entering a new era of multimodality cancer therapy in which surgery will have a resurgence as one component amongst other therapies? If so, there would be reason to revisit old lessons and ask again if partial removal of cancer has some merit. Or are we, and our patients, behaving like drowning men clutching at straws? If treatments are cheap and harmless why withhold anything which gives comfort and solace to those with incurable cancer, but metastasectomy is not without harm and carries considerable cost. It involves major surgery, sometimes bilateral, sometimes repeated, and is used in combination with other noxious therapies. For reasons of humanity and economy, we should not operate on these cancer patients unless we can show evidence of benefit.
|
3. The objectives of pulmonary metastasectomy
|
|---|
The results sections of surgical reports are dominated by, or are devoted exclusively to, survival data. The data are most commonly presented as Kaplan–Meier survival graphs. This is the language of life saving or life extending treatment; the words cure or curative are often used. The distinction between R0 and R1 resection is usually made and the premise on which we operate is that all cancer can be removed. If this is the case then the measure of success of metastasectomy should be that the disease should be eradicated and a proportion of patients live for an open-ended time with no sign of the cancer. Since thoracotomy and loss of lung parenchyma will shorten some lives, we need to be assured that there is a net gain.
Is there any other objective? Some clinicians state that metastasectomy is palliative not curative. If so what symptom is being palliated or pre-empted? I have yet to find papers reporting any measure of health gain such as quality of life (QoL) or patient reported outcomes (PROs). There is no established practice of resecting incurable lung cancer with the object of improving or pre-empting symptoms. Why should asymptomatic secondary cancer in the lung be considered for resection with such an objective in mind?
In clinical discussions it is not always adequately clear what drives the decision to operate on metastases. If we ask the patients, they hope for cure. Amongst protagonists there is often talk of psychological benefit but there are no measures of psychological well being to support this argument nor are any psychologists or psychiatrists amongst the authors. The evidence of efficacy offered in the reports is longer survival—that is the currency of Kaplan–Meier analysis. In any clinical trial this would have to be accompanied by formal assessment of quality of life or patient reported outcomes to correct for the inevitable detrimental effect of thoracotomy and loss of lung parenchyma. In the metastasectomy literature it is just survival time which is reported, without any adjustment for its quality. If there is palliative benefit it is yet to be demonstrated in colorectal cancer.
What determines survival time after pulmonary resections for colorectal metastases? Many authors (there are over 70 reports of colorectal cancer metastasectomy) have performed multivariate analysis to identify factors such as the disease-free interval, the number of metastases, levels of carcinoma embryonic antigen, Dukes stage and the degree of differentiation of the primary which emerge rather variably as being significantly associated with survival time. The inconsistencies are not very important; whether or not a factor emerges as being significant is highly dependent on how many cases are in the series, and whether the factor under consideration was widely represented in the data set of patients undergoing metastasectomy. Thus if high CEA, short disease-free interval and multiplicity of lesions were relative exclusion criteria in the clinical practice being reported, they are less likely to emerge in an analysis. What they have in common and runs consistently through the reports is that these are all descriptors of the aggressiveness of colorectal cancers so are likely to be associated with survival time regardless of whether metastases are resected or not.
There are no randomised trials to help us so we have only non-randomised studies [2]. There are serious shortcomings in the use of surgical series as evidence on which to base practice [5]. In the case of metastases the populations put forward for comparison are very poorly characterised and often from some years ago. The patients operated on are selected and indeed the recommendations for future practice usually include words such as ‘for a limited subset of patients’ and emphasis is placed on the importance of ‘highly selected patients’. It is concerning that authors have been using these phrases for years but there is no evidence that surgery is being more selective; if anything indications are being broadened [1]. Current practice seems to be if anything less selective, with a greater extent of disease being included and a shorter period of apparent freedom from disease being no contraindication.
My opening remarks may be interpreted as rather negative, or even nihilistic, as though my opinion is based upon some fundamental difference in the way I looked at the world. ‘Some see the glass half-empty, some see it half-full’ is the English saying that sums up the difference between people's view of the world. I do not see this as a matter of subjective judgement but of quantification first and foremost. Better that we consider the facts and take a realistic view. I propose a very positive solution to the question; which is to run a trial to find evidence for benefit and to help define groups where there is no advantage so that they may be spared unavailing surgery.
|
4. A randomised clinical trial
|
|---|
The first task is to define the PICO. This is an acronym popularised by health service researchers to summarise the elements of a trial.- P patients or the population under investigation;
- I intervention under test;
- C comparison or control treatment;
- O outcome.
|
5. Patient population
|
|---|
Although practice varies, there is some common ground and respondents indicate that 90% of these patients are discussed in multidisciplinary team meetings [1].- • There are patients for whom a team does not hesitate to offer surgery and they may well continue to do so. These are likely to be young and fit with one or few isolated accessible metastasis.
- • On the one hand multidisciplinary teams will have agreed criteria for excluding patients. For example if the patient has uncontrolled primary cancer, is known to have cancer at other inaccessible sites, or is unfit for the surgery they will be excluded.
My proposal is that thoracic surgeons and the MDTs with whom they work randomise patients in whom they feel uncertain about the best course of action. The randomisation would be between surgical metastasectomy combined with any other treatment modalities vs any other management apart from surgery. My prediction is that once there is a trial, people will become less confident that they can make a clinical judgement—the acceptance of uncertainty broadens the potential pool of patients. We have noted that effect in the MARS trial for mesothelioma [6].
The patient must be fit to undergo the required surgery but may be in a zone of uncertainty as far as the number of metastases or the disease-free interval are concerned. In practice, we have to make a decision and surgeons by their nature do not shrink from this. If some patients are accepted and others rejected, it follows that there is a position where the answer changes from yes to no. If there is a spectrum of cases and some are accepted for metastasectomy and some are rejected, it must be true that there is some point in the middle when the decision is finally balanced. At this point of cross-over there is uncertainty. Since this will depend on many combinations of factors it will not be easy to define these by explicit criteria to which all trialists can adhere in all cases. However, a reading of the available literature indicates that many patients operated on are not cured. If a group of patients have an equal likelihood of harm as benefit, it seems to be as reasonable to allocate these patients by randomisation. The ethics of this have been very carefully explored [7] and we know that in cancer trials patients allocated to the ‘new intervention’ group are disadvantaged as often as they benefit [8].
The question arises ‘should these just be colorectal cancer patients?’ It is not essential. Provided it is a factor that is entered into the process of unbiased allocation they could be included in the same study. The process of allocation by minimisation [9] can be used to achieve that.
|
6. Intervention
|
|---|
The way in which the surgery is performed is governed by the site and number of metastases and the preference of the team. Carefully performed and authoritative systematic reviews have not found evidence to make a recommendation in favour of thoracotomy or VATS, or for a bilateral as opposed to a unilateral approach [10,11]. However the objective should be an R0 resection or the benefits of metastasectomy, as practised are not being tested.
The surgery could be combined with any adjuvant therapy; it is the surgical resection which is under test. If there is no evident difference the patient should be spared the operation.
|
7. Control group
|
|---|
It is extremely difficult to run a trial leaving half the patients feeling that they are untreated. Chemotherapy may well be given to both groups. Radiotherapy, radiofrequency ablation or any other measure may be used in the control group or at some stage in the course of care of the surgical metastasectomy group. If surgery is superior, that should become evident if the effect is substantial.
|
8. Outcome
|
|---|
Clearly survival will be a central issue. This should be all cause or it will neglect the possibility that patients may die sooner as a consequence of having less lung. As important will be a measure of harm or benefit. Quality of life has never been an easy measure to employ but patient reported outcomes (PROs) for breathlessness, pain and other symptoms must surely be included.
http://www.fda.gov/fdac/features/2006/606_patients.html.
|
Footnotes
|
|---|
\#9734; Presented as part of the postgraduate course at the 15th European Conference on General Thoracic Surgery, Leuven, Belgium, 3–6 June 2007.
|
References
|
|---|
- Internullo E, Cassivi S, Van Raemdonck D, Friedel G, Treasure T. Pulmonary metastasectomy: the state of the practice in Europe. Interact Cardiovasc Thorac Surg 2007;6(Suppl. 2):S182Ref. Type: Abstract.[CrossRef]
- Pfannschmidt J, Dienemann H, Hoffmann H. Surgical resection of pulmonary metastases from colorectal cancer: a systematic review of published series. Ann Thorac Surg 2007;84(July (1)):324-338.[Abstract/Free Full Text]
- Rosenberg C, Golden J. Framing disease. New Brunswick, New Jersey: Rutgers University Press; 1992.
- Rolle A, Pereszlenyi A, Koch R, Richard M, Baier B. Is surgery for multiple lung metastases reasonable? A total of 328 consecutive patients with multiple-laser metastasectomies with a new 1318 nm Nd:YAG laser. J Thorac Cardiovasc Surg 2006;131(6):1236-1242.[Abstract/Free Full Text]
- Treasure T, Utley M. Ten traps for the unwary in surgical series: a case study in mesothelioma reports. J Thorac Cardiovasc Surg 2007;133(June (6)):1414-1418.[Free Full Text]
- Treasure T, Tan C, Lang-Lazdunski L, Waller D. The MARS trial: mesothelioma and radical surgery. Interact Cardiovasc Thorac Surg 2006;5:58-59.[Abstract/Free Full Text]
- Evans HM. Should patients be allowed to veto their participation in clinical research?. J Med Ethics 2004;30(April (2)):198-203.[Abstract/Free Full Text]
- Kumar A, Soares H, Wells R, Clarke M, Hozo I, Bleyer A. Are experimental treatments for cancer in children superior to established treatments? Observational study of randomised controlled trials by the Children's Oncology Group. BMJ 2005;331(December 3 (7528)):1295.[Abstract/Free Full Text]
- Treasure T, MacRae KD. Minimisation: the platinum standard for trials? Randomisation doesn’t guarantee similarity of groups; minimisation does. BMJ 1998;317(August 8 (7155)):362-363.[Free Full Text]
- Naunheim K. Thoracoscopy versus the open approach for resection of solitary pulmonary metastases. In: Ferguson M, editor. Difficult decisions in thoracic surgery: an evidence based approach. Springer; 2007. pp. 151-157.
- Patel A, DeCamp M. Unilateral or bilateral approach for unilateral pulmonary metastatic disease. In: Ferguson M, editor. Difficult decision in thoracic surgery: an evidence based approach. Springer; 2007158–64.
This article has been cited by other articles:
|
|
|
|
 
A. Nakas, M. N. Klimatsidas, J. Entwisle, A. E. Martin-Ucar, and D. A. Waller
Video-assisted versus open pulmonary metastasectomy: the surgeon's finger or the radiologist's eye?
Eur. J. Cardiothorac. Surg.,
September 1, 2009;
36(3):
469 - 474.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
1. Introduction
