Preoperative staging for rectal cancer

Colorectal cancer is the second most common cause of death from malignant disease in the United Kingdom, with about 20 000 deaths each year. Around one million new cases (9% of all cancers) are diagnosed each year worldwide (CANCERMondial; www-dep.iarc.fr). As the UK population ages the incidence is predicted to rise.¹ At present the only realistic prospect of cure is complete surgical resection of the primary tumour. The restricted anatomical space in the pelvis makes this technically easier to achieve for cancers of the colon than the rectum. Consequently, local recurrence rates after surgery for rectal cancer have been as high as 50%. Local recurrence is a devastating complication as it is invariably fatal even without disseminated disease. Local recurrence can be reduced by two methods—surgical technique and radiotherapy. Currently, radiotherapy is given to most patients even though only a subgroup will benefit. There is increasing interest in the use of preoperative staging to target high risk patients who will benefit most from radiotherapy. A study in this week's BMJ shows the value of magnetic resonance imaging in the preoperative staging of rectal cancer.²

Local recurrence rates can be reduced to less than 5% if the surgeon removes the rectum en bloc with the mesorectum using precise anatomical dissection (total mesorectal excision).³ Crucially, the surgeon must ensure that the circumferential resection margin is clear of tumour because a positive margin predicts a high risk of local recurrence.⁴ The quality of this surgery, which is technically demanding, can be determined by pathological examination of the integrity of the mesorectum in resected specimens. In Sweden, the adoption of total mesorectal excision by surgeons who had attended a training programme reduced the local recurrence rate after surgery for rectal cancer from 20% to 8% and increased survival.⁵

Randomised trials have shown that preoperative radiotherapy reduces local recurrence and is superior to postoperative treatment.⁶,⁷ The Swedish rectal cancer study showed that “short course” preoperative radiotherapy reduced local recurrence and improved survival, but this study was subsequently criticised because total mesorectal excision was not used.⁸ In a later Dutch trial all patients had total mesorectal excision and were randomised to preoperative radiotherapy or surgery alone.⁹ As expected with high quality surgery the benefit from radiotherapy was less; local recurrence was reduced from 8% to 2%. Furthermore, the size of the benefit depended on the stage of the primary tumour—0.2% for T1 tumours, 5% for T2, and 12% for T3 primary lesions. These data can also be expressed as the proportion of patients who received unnecessary radiotherapy—99.8% of patients with a T1 tumour and 95% of those with T2 lesions. Even for patients with relatively high risk T3 tumours the figure was 88%. This is important because the Swedish and Dutch trials reported higher morbidity, including bowel dysfunction and incontinence, in patients treated with radiotherapy.¹⁰,¹¹ Ideally, patients at high risk of a positive resection margin because of advanced primary disease should be identified before surgery so they can be targeted with preoperative chemotherapy or radiotherapy.

The study in this week's issue from the MERCURY Study Group shows that pelvic magnetic resonance imaging performed before surgery for rectal cancer has a sensitivity of 94% and a specificity of 92% for predicting negative circumferential margins after surgery.² Crucially, the study showed that when radiologists undergo specific training and the technique is standardised, results are reproducible between centres. The technique was less reliable at predicting positive resection margins mainly due to localised tumour perforation or the presence of adjacent lymph nodes, but this does not limit the clinical value of pelvic imaging.

These data have important implications for the management of patients with rectal cancer. Firstly, although a previous study suggested that preoperative staging may be used to target radiotherapy to high risk patients, the number of patients included was small.¹² The MERCURY Study Group however has now confirmed in a multicentre observational study that magnetic resonance imaging may be used for this purpose. This means that patients with rectal cancer at low risk of local recurrence can avoid unnecessary and harmful chemotherapy or radiotherapy. Also resources can be focused towards those patients at high risk.

Secondly, future studies of chemotherapy and radiotherapy can now be targeted towards a homogeneous group of patients known to be at high risk of local recurrence after surgery. This should make it easier to design trials with sufficient power to determine the most effective treatments in high risk groups, the benefits of short course radiotherapy compared with longer courses, and the role of salvage surgery in both responders and non-responders.

Finally, preoperative staging provides an objective yardstick against which the quality of the surgical technique may be audited. When used with histological assessment of the integrity of the mesorectum in resected specimens, surgeons will no longer be able erroneously to attribute positive resection margins after rectal cancer surgery to advanced disease rather than poor surgical technique.

Preoperative staging has been advocated in guidelines for the management of colorectal cancer and many clinicians are already using it.¹³,¹⁴ The task ahead is to make magnetic resonance imaging mandatory for all patients with rectal cancer before treatment decisions are made at multidisciplinary meetings.