Treasure, Tom MD, MS, FRCS, FRCP*; Fallowfield, Lesley BSc, DPhil, FMedSci†; Lees, Belinda PhD‡
In this supplement of the Journal of Thoracic Oncology that reviews the evidence for the practice of pulmonary metastasectomy, none of the authors have identified any randomized trials. Claims for a survival benefit for patients undergoing this surgery rely on case series. Furthermore, there is little documentation of any symptoms attributable to pulmonary metastases that are alleviated or obviated by metastasectomy. In the particular instance of colorectal cancer, the authors of the review explain the reasons for the perception of benefit,1 but the evidence for cure, life extension, disease modification, or palliation achieved by pulmonary metastasectomy is not of a standard required in evidence-based medicine.2,3
To address these shortcomings in the evidence base, during 2010, a randomized trial called pulmonary metastasectomy in colorectal cancer (PulMiCC) opened to recruitment on March 17, 2010. The study was first proposed in the BMJ in 2007,4 and further soundings were taken by publishing the proposal in the European Journal of Surgical Oncology.3 The study questions whether surgical resection of pulmonary metastases from colorectal cancer lengthens survival and whether surgery makes any difference (positive or negative) to symptoms and quality of life.
In this article, key elements of the PulMiCC trial are described with the rationale for performing the study in the context of the available evidence. It might not be applicable to all pulmonary metastasectomy practices but merits consideration. Trials of surgery within multimodality cancer treatments are not easy but can be achieved.5
PATIENTS ELIGIBLE FOR THE PulMiCC TRIAL
The criteria for pulmonary metastasectomy are in essence those set out very clearly by Kondo et al.6—an update of the well-known Thomford criteria.7
1. The patient must be a good risk for surgical intervention.
2. The primary site is controlled.
3. No other metastatic disease; or, if present, it can be controlled by surgery or another treatment modality.
4. Pulmonary metastases are thought to be completely resectable.
In the context of the PulMiCC trial that is only recruiting patients with colorectal cancer, the third criterion specifically refers to liver metastasis.
Importantly, it should not be assumed that patients who present with a solitary lung nodule against a background of colorectal cancer necessarily have a colorectal pulmonary metastasis. Every such nodule should be considered as a matter for diagnosis and management in its own right, in case it is a primary lung cancer or another pulmonary pathology.8 This should be done by the multidisciplinary team or cancer board designated to deal with thoracic cancer in collaboration with the colorectal team.
It is recognized that all teams performing pulmonary metastasectomy for spread of colorectal cancer select cases on the basis of features that can be identified before surgery.9 For example:
* There are patients in whom the question of pulmonary metastasectomy is ruled out early on. Patients with a large number of metastases appearing very soon after the resection of an aggressive cancer with locoregional nodal involvement and a high carcinoma embryonic antigen (CEA) assay (for example) would be excluded.
* There are patients in whom pulmonary metastasectomy is undertaken by nearly all groups. Most teams would remove the metastases if a patient had one or two nodules, appearing after a long interval of approximately 3 to 4 years, which are growing but at a slower pace.
The important point to note is that if there is concordance that at one end of the spectrum the team will vote “yes” and at the other “no,” there must be a transition zone where there is uncertainty sufficient to merit critical evaluation or even to justify being called equipoise.
The most recently reported multivariable analysis is in 378 patients having pulmonary metastasectomy for colorectal cancer at Memorial Sloane Kettering in a 10-year period from 1998.10 The authors found more than three metastases, a longer interval between primary and metastasectomy surgery, the stage of primary surgery, and the presence of extrathoracic metastases, to be adverse prognostic factors. Of course, these are prognostic factors related to cancer, but they are not predictors of whether the operations work or not. The Memorial Sloane Kettering group also found female sex and age younger than 65 years to be adverse prognostic features. Probably, more informative than the p values are the hazard ratios. Hazard ratios, with their 95% confidence intervals, are 0.68 (0.49-0.94) for male sex and 0.74 (0.54-1.01) for age older than 65 years. These relative differences in the likelihood of living 5 years depending on age and sex are hardly evidence of the effectiveness of the surgery in curing cancer.
PulMiCC is pragmatic and deliberately sets no inclusion or exclusion criteria for cancer-related factors, such as number, size, CEA, the stage or differentiation of the primary cancer, and the timing of the appearance of the metastases in relation to the primary resection. Instead, we recognize that between the group of patients for whom there is a confident decision in favor of pulmonary metastasectomy and those patients for whom there is a confident decision against, there must be some for whom the team debates what to do for the best. This is an expression of uncertainty. We contend that if there is real clinical uncertainty among experts, randomized allocation is as rational a management policy as following a hunch. The situation is summarized in Table 1. Note that the five criteria are presented without boundaries. There is no cutoff for any of them. There are others that can be considered. The clinical decisions weigh these factors separately and in combination.
Cancer teams interested in joining PulMiCC are requested to reflect on their existing practice and to define their “zone of uncertainty.” Table 2 shows how this exercise was used to categorize a series of 45 patients who had undergone pulmonary metastasectomy in the Northern General Hospital, Sheffield, UK (with permission), whose practice is very much in line with that in the many published case series.11
THE TRIAL DESIGN
We have learned a great deal about how this trial might succeed from the Mesothelioma and Radical Surgery trial into which, contrary to all predictions, randomization has proved feasible.5,12 Following that model, we have designed two stages to the consent process. However, to succeed in running a trial, it is vital that the patients are told the truth from the outset and introduced to the concept of uncertainty about the best course of action for them in a neutral way. At present, all too often, a well-meaning nurse or oncologist breaks the news to them and proposes their own solution: “The scans show recurrence of your cancer in the form of little nodules in your lungs, but don't worry—we'll send you to the thoracic surgeon to take them out.” Thoracic surgeons are very familiar with this scenario, turning the patient down for surgery after such an introduction conveys the message of “giving up,” or worse, administering a death sentence by inaction.
We hope to pre-empt expectations of a surgical intervention in the minds of patients by providing them with not only the standard patient information sheet but also a purpose-made explanatory digital video disc. This digital video disc (DVD) supplements the patient information sheet and includes cartoons of the logic and purpose of randomization, schematic diagrams and explanations about different types of surgery, and clips of real doctors answering typical patients' queries. At trial launch, the centers participating in PulMiCC also have an interactive communication skills training session that aims to assist all involved in providing even-handed descriptions of the patients' options including trial entry.
At this point, the question of ethics will come to many readers' minds, and quite rightly. Two ethicists are included in the trial development group: one from legal background and the other a professor of philosophy. The position of the PulMiCC investigators is that if one could trade relative ethics, it is more ethical to come clean and tell the patient that there is uncertainty about whether metastasectomy will cure them rather than to encourage the unproven belief that it will. It is more ethical to find evidence than to perform surgery that is not evidence based. For a more detailed consideration of research ethic and randomization, read the essay by Martyn Evans,13 the philosopher in our trial development group.
At first consent, patients agree to enter the PulMiCC study. They may have probable or confirmed pulmonary metastases and during this phase of the study will have any further investigations required to evaluate the lung nodules that have been identified. The work-up is no more than what is required to meet the updated Thomford criteria.7 In most cases, this will include positron emission tomography (mainly to reduce the chance of missing unsuspected disease at other sites) as well as computed tomography scans. Control at the primary site will be confirmed by the colorectal cancer team. Tissue diagnosis will be obtained if there is clinical doubt about the nature of the pulmonary nodule(s). If chemotherapy is judged appropriate, it can be given during this phase of the study.
The question of liver metastases frequently arises. The incidence of metastatic disease increased from 15% in 1970s to 22% since 2000, based on 34 studies including 2356 patients' records.11 There were about a third in the Sheffield series and 38% in the MSK series.10 PulMiCC asks that if these are considered suitable for resection, this surgery should be performed and they will then be reconsidered.
After this process, some patients will be deemed unsuitable for pulmonary metastasectomy; in some cases, the team will want to offer surgery because they see the patient as matching their existing criteria. These patients will be managed, as agreed between them and their clinicians, but their data will be retained. Patients for whom pulmonary metastasectomy is an option, but for whom there is uncertainty, will be offered randomization within PulMiCC.
Patients consent to be allocated to continued active monitoring or to pulmonary metastasectomy and continued active monitoring. The allocation will be performed offsite and will use minimization in addition to an element of random allocation. Minimization is a means of ensuring balance between the groups for confounding factors, such as the number of metastases, the interval since the primary resection, the CEA, the T and N stage of the colorectal cancer, history of liver metastases, age, and sex. This approach is well established and is particularly valuable where there is heterogeneity and the trial is relatively small.14,15
Reviewers of the literature in this supplement and elsewhere16–19 have been unable to make evidence-based recommendations for whether surgery should be by minimally invasive video surgery or open, unilateral or bilateral, or what surgical technique should be used. We know that practice varies widely.20 There would be no rational basis to do other than to allow these matters to be determined by local preferred practice and what is most appropriate for the individual patient.
Radiofrequency ablation seems to be growing in popularity21 and may be more used in future.22 However, it fails to meet the criteria for the management of metastasectomy in that there is no way of confirming an R0 resection. It also precludes proving that the lesion was a metastasis, and not every metastasis needs to be histologically confirmed before metastasectomy. It may well be that there will be slippage in this direction, but patients are free to seek treatment outside the trial, and PulMiCC trialists recognize that some clinicians will want to do something active for their patients.
Survival is the primary outcome measure. It is essential that the quality of survival is better known than at present, whether there are gains or losses associated with pulmonary metastasectomy. On the positive side, some believe that removal of metastases pre-empts the development of symptoms, so if the operation is not curative, it is still palliative. On the downside there is the inevitable pain of thoracotomy and, with larger resections, loss of lung function, so respiratory function will be measured regularly. Patients may also find the inconvenience of hospitalization impacts on quality of life while those who have only active monitoring may experience more anxiety. Standardized patient-reported outcome measures will be used to assess these issues 3, 6, 12, and 24 months after surgery.
* Functional Assessment of Cancer Therapy—General (FACT-G) and FACT anemia/fatigue subscale (FACT-An) plus selected items from the FACT Lung Cancer Brief Symptom Index.
* Short form of the Spielberger State/Trait Anxiety Inventory.
The health economic implications of pursuing different management policies will also be evaluated via the EQ-5D questionnaire administered at similar time points to the questionnaires listed previously.
Do We Need a Randomized Trial Anyway?
Nearly all surgeries have developed from practical solutions to evident problems. We do not need randomized trials to know that the lame walk better after hip replacement and the blind recover their sight after cataract surgery. The results would have been seen as miraculous in biblical terms.23 There are many instances in surgery from the Thomas's splint for femoral fracture to aortic valve replacement for stenosis, where the benefit attributable to surgery is immediate and irrefutable. This form of evidence is accepted and well defined by evidence-based medicine experts.24,25 However, this is not applicable in all circumstances. The retrospective analysis of surgical series, which is the form of evidence offered in favor of pulmonary metastasectomy, has many pitfalls.26
The paucity of randomized trials in surgery prompted a now classic editorial in The Lancet, comparing surgeons' efforts in research with “comic opera,”27 but still they are commonly rejected as being too difficult, a view that has been countered elsewhere.28 Our view is that the highly selected nature of patients undergoing pulmonary metastasectomy, the great variation in the clinical course of different patients with cancer, and the use of combinations of therapies make it impossible to discern the effect of surgery on survival; we cannot see the signal from the noise in this instance,24 and so we developed PulMiCC.
1. Pfannschmidt, et al. Factors influencing outcome of pulmonary metastasectomy for colorectal cancer. J Thorac Oncol. In press.
2. Hill J, Hunt I. Applying evidence based medicine to cardiothoracic surgery. In T Treasure, B Keogh, I Hunt, et al. (Eds.), The Evidence for Cardiothoracic Surgery. Shrewsbury, UK: tfm Publishing Ltd, 2005. Pp. 1–4.
3. Treasure T, Fallowfield L, Farewell V, et al. Pulmonary metastasectomy in colorectal cancer: time for a trial. Eur J Surg Oncol 2009;35:686–689.
4. Treasure T, Utley M, Hunt I. When professional opinion is not enough. BMJ. 2007 21;334:831–832.
5. Treasure T, Waller D, Tan C, et al. The mesothelioma and radical surgery randomized controlled trial: the MARS Feasibility Study. J Thorac Oncol 2009;4:1254–1258.
6. Kondo H, Okumura T, Ohde Y, et al. Surgical treatment for metastatic malignancies. Pulmonary metastasis: indications and outcomes. Int J Clin Oncol 2005;10:81–85.
7. Thomford NR, Woolner LB, Clagett OT. The surgical treatment of metastatic tumours in the lungs. J Thorac Cardiovasc Surg 1965;49:357–363.
8. Quint LE, Park CH, Iannettoni MD. Solitary pulmonary nodules in patients with extrapulmonary neoplasms. Radiology 2000;217:257–261.
9. Pfannschmidt J, Dienemann H, Hoffmann H. Surgical resection of pulmonary metastases from colorectal cancer: a systematic review of published series. Ann Thorac Surg 2007;84:324–338.
10. Onaitis MW, Petersen RP, Haney JC, et al. Prognostic factors for recurrence after pulmonary resection of colorectal cancer metastases. Ann Thorac Surg 2009;87:1684–1688.
11. Fiorentino F, Hunt I, Teoh K, et al. Pulmonary metastasectomy in colorectal cancer: a systematic review and quantitative synthesis. J R Soc Med 2010;103:60–66.
12. Treasure T, Tan C, Peckitt C, et al. Mesothelioma and radical surgery trial (MARS): the feasibility study process. Lung Cancer 2009;63(Suppl 1):S26.
13. Evans HM. Should patients be allowed to veto their participation in clinical research? J Med Ethics 2004;30:198–203.
14. Treasure T, MacRae KD. Minimisation: the platinum standard for trials? BMJ 1998;317:362–363.
15. Altman DG, Bland JM. Treatment allocation by minimisation. BMJ 2005;330:843.
16. Venuta F, Rolle A, Anile M, et al. Techniques employed in pulmonary metastasectomy. J Thorac Oncol. In press.
17. Molnar TF, Gebitekin C, Tuma A. What are the considerations in the surgical approach in pulmonary metastasectomy? J Thorac Oncol. In press.
18. Naunheim K. Thoracoscopy versus the open approach for resection of solitary pulmonary metastases. In M Ferguson (Ed.), Difficult Decisions in Thoracic Surgery: An Evidence Based Approach. London: Springer-Verlag, 2007. Pp. 151–157.
19. Patel A, de Camp M. Unilateral or bilateral approach for unilateral pulmonary metastatic disease. In M Ferguson (Ed.), Difficult Decisions in Thoracic Surgery: An Evidence Based Approach. London: Springer-Verlag, 2007. Pp. 158–164.
20. Internullo E, Cassivi SD, Van Raemdonck D, et al. Pulmonary metastasectomy: a survey of current practice amongst members of the European Society of Thoracic Surgeons. J Thorac Oncol 2008;3:1257–1266.
21. Gillams AR, Lees WR. Radiofrequency ablation of lung metastases: factors influencing success. Eur Radiol 2008;18:672–677.
22. Pass H. The mets, the scalpels, and possibly the beams. J Thorac Oncol 2008;3:1211–1212.
23. New Testament: The Gospel According to Matthew. Chapter 11, Verse 5.
24. Glasziou P, Chalmers I, Rawlins M, et al. When are randomised trials unnecessary? Picking signal from noise. BMJ 2007;334:349–351.
25. Vandenbroucke JP. Observational research, randomised trials, and two views of medical science. PLoS Med 2008;5:e67.
26. Treasure T, Utley M. Ten traps for the unwary in surgical series: a case study in mesothelioma reports. J Thorac Cardiovasc Surg 2007;133:1414–1418.
27. Horton R. Surgical research or comic opera: questions, but few answers. Lancet 1996;347:984–985.
28. Treasure T. Are randomised trials needed in the era of rapidly evolving technologies? Eur J Cardiothorac Surg 2009;35:474–478.