Improving management of bowel cancer

By informing clinical decision-making at several stages along the patient pathway, MRI is providing an opportunity to potentially downstage even poor prognosis rectal tumours and improve patient outcomes.

Consultant radiologist Dr Gina Brown, the NHS lead in imaging research at The Royal Marsden, has pioneered the application of MRI in planning rectal cancer surgery. In the mid-nineties, Dr Brown was working as a Research Fellow at the University Hospital of Wales in Cardiff: “In 1995, the landscape of colorectal surgery was changing. We were moving from a situation where general surgeons performed bowel surgery, to there being specialised colorectal surgeons. This change coincided with an increasing awareness of Professor Bill Heald’s total mesorectal excision (TME) technique.” Now the ‘gold standard’ for rectal cancer surgery, TME involves precisely removing the optimal block of tissue to safely encompass the cancer – including its local spread. “TME gave researchers the opportunity to study an excellent pathology specimen – the full ‘anatomical package’. You learn what you are seeing on an MRI from the pathology specimen, by precisely correlating MRI features with the detailed pathology. For the first time we could rigorously compare whole pathology specimens with pre-surgery MRI scans, to understand more about tumour staging.” At the same time, MRI scanning itself was improving. “Innovations such as a pelvic phased array coil placed locally to the pelvis, meant that the machine sensors were getting a far better signal. The images improved, and the process was more pleasant for patients.” The alternative was a cumbersome endo-rectal coil. Today, the technique for staging rectal cancers involves taking a series of five key sequences of MR images using a pelvic phased array coil, each showing different aspects of the pelvis. This usually takes up to 45 minutes, with the patient in a supine position. Practically speaking, as no bowel preparation, contrast agents, or air insufflation is required it is a patientfriendly procedure. “The fact is that in the early nineties nobody really understood pelvic anatomy in its true complexity, so these technical advances were opening up a new world to us,” said Dr Brown, who was, at the time, working in close collaboration with Professor Lennart Blomqvist at the Karolinska University Hospital in Stockholm, Sweden, who brought his knowledge and enthusiasm to bear. A pivotal moment came when Dr Brown was a year into her colorectal research, with preliminary data. Prof. Heald invited her to present at a TME workshop at the Royal College of Surgeons on the role of MRI in rectal cancer staging. Little had been published on the subject at the time. Attending the presentations, Dr Brown saw the TME operation performed via live links for the first time. “Seeing the pelvic anatomy being demonstrated through live surgery was extraordinary – and hugely educational. To put it in perspective – the mesorectum was not even mentioned in many text books of the time. It brought it all to life.” At the meeting, Dr Brown met Dr Phil Quirke, a pathologist from Leeds, who was proselytising the practice of thoroughly auditing the TME specimen, checking the margins of removed tissue for microscopic traces of cancer which might indicate whether malignant cells had been left behind. If cancer is detected at the outer margin of the removed specimen, then it is far more likely to recur, leading to a poor chance of survival in many cases. “The possibility of identifying, preoperatively, those rectal cancers which could threaten the surgical margins using MRI would enable a team to make appropriate pre-operative treatment decisions for individual patients. I wrote a revised plan of research on the train home, and redrafted my thesis as a result,” said Dr Brown. Two years later, Dr Brown’s research showed that it was indeed possible to correlate MR images with the pathology, making accurate disease staging presurgery possible: MRI could inform pre-operative treatment decisions and surgical planning by predicting whether a margin was likely to be threatened. This could facilitate accurate treatment – audited by pathology – improving patient survival. However, this was a single centre study: a larger scale study was needed to demonstrate the accuracy, feasibility and reproducibility of MRI as a method of assessing rectal cancer.

Evidence for the use of MRI