A Randomised Trial of Pulmonary Metastasectomy in Colorectal Cancer

Colorectal cancer is one of the three commonest solid tumours. Approximately 100 cases are diagnosed each day in the UK. In 2006, 37,514 cases of colorectal cancer were diagnosed. The increasing use of CT scans in routine follow up of patients has led to earlier diagnosis of isolated and/or subclinical lung nodules and there has been widespread adoption of pulmonary metastasectomy in selected cases. A survey of practice in Europe showed that pulmonary metastasectomy is very commonly performed but criteria vary widely. About 300 such operations were being carried out annually in the UK when PulMiCC opened as a 'feasibility' study. Although well-established in clinical practice there have been no previous randomised controlled trials of the effect of lung metastasectomy on survival and there is doubt as to its clinical effectiveness. The most recent systematic review and meta-analysis concludes that pulmonary metastasectomy is associated with a high likelihood of recurrence, with doubling of the recurrence rate with each of three adverse prognostic factors: (i) more than one metastasis, (ii) an interoperative interval (primary resection to lung metastasectomy) less than three years, (iii) elevated CEA. In a prospective registry capturing more than 60% of Spanish practice for a two year period these limits were commonly exceeded. In an RCT it has been shown that recurrence can be detected earlier with CT and/or CEA surveillance and as a result more liver and lung metastasectomy operations were performed. However, these operations did not provide any survival benefit, within the RCT. The uncertainty about the practice has been set out in the British Medical Journal in 2014. Pulmonary metastasectomy is thus part of a bigger question about effectiveness of surgery for metastases from colorectal cancer. There are known unfavourable prognostic factors which include the number of metastases and the length of time they take to become evident radiologically: (a) a solitary nodule appearing after a long interval is will be removed in most instances, (b) surgery is rarely advocated when there are multiple metastases present at the time of surgery on the primary colorectal cancer or appearing soon after. These scenarios represent opposite ends of the continuum of favourable to adverse factors for survival after resection of pulmonary metastases. Most patients operated on fall between these extremes and it is evident that if there is a "yes" towards one end of the continuum and "no" towards the other that there must be cross over zone where there is clinical uncertainty. PulMiCC aims to investigate the outcome following pulmonary metastasectomy in patients where there is accepted clinical uncertainty, to provide evidence to guide practice in the future. It has also been shown previously that trials requiring patients to consider randomisation between treatments that appear very different, as in this instance between surgery or no surgery, can experience recruitment difficulties. This is because potential participants may have acquired a strong preference for one of the treatments over the other and they are not willing to accept the possibility of being allocated to the less non-preferred option alternative. There can also be problems of bias if one treatment is mentioned to the patient before the option of a trial is introduced; the first mentioned treatment may be perceived to be preferred between doctor and patient. This trial has been designed to overcome these potential difficulties and maximise recruitment as follows. Colorectal cancer patients presenting with pulmonary metastases will first be consented for registration into the study. This consent is to undergo evaluation according to the trial protocol as part of the work up for consideration by the MDT. Following evaluation, the MDT will consider how they would normally treat each patient according to their standard local practice. Patients eligible for randomisation will be those for whom clinical uncertainty exists as to whether surgery would be of benefit. This trial design was successful in a study of the feasibility of randomising patients in a trial of mesothelioma surgery. The time taken for patients to undergo the full range of tests for evaluation gives them time to think carefully about the possible treatment options and discuss them with their doctor. In addition to the patient information leaflet, patients will be invited to take home a DVD explaining the trial in detail, which can be watched as often as the patient likes. It is hoped that those who are ultimately identified as eligible for randomisation will have fewer anxieties because of their extended opportunity to consider the trial. Local site research staff involved in recruiting patients will attend a training session to learn the best method of informing patients about the trial before recruitment commences.