Send to

Choose Destination
Ann Surg. 2016 Apr;263(4):751-60. doi: 10.1097/SLA.0000000000001193.

Prospective Validation of a Low Rectal Cancer Magnetic Resonance Imaging Staging System and Development of a Local Recurrence Risk Stratification Model: The MERCURY II Study.

Author information

*Pelican Cancer Foundation, Basingstoke, United Kingdom †North Hampshire Hospital, Basingstoke, United Kingdom ‡Dresden-Friedrichstadt General Hospital, Dresden, Germany §Pathology & Tumor Biology, Leeds Institute of Cancer and Pathology, University of Leeds, Leeds, United Kingdom ¶Salisbury District Hospital, Salisbury, United Kingdom ||Vivantes im Friedrichshain, Berlin, Germany **The Royal Marsden NHS Foundation Trust, London, United Kingdom ††Aarhus University Hospital, Aarhus, Denmark ‡‡Frimley Park Hospital, Frimley, United Kingdom §§The Academic Department for Radiology, The Royal Marsden Hospital, Sutton, United Kingdom.



This study aimed to validate a magnetic resonance imaging (MRI) staging classification that preoperatively assessed the relationship between tumor and the low rectal cancer surgical resection plane (mrLRP).


Low rectal cancer oncological outcomes remain a global challenge, evidenced by high pathological circumferential resection margin (pCRM) rates and unacceptable variations in permanent colostomies.


Between 2008 and 2012, a prospective, observational, multicenter study (MERCURY II) recruited 279 patients with adenocarcinoma 6 cm or less from the anal verge. MRI assessed the following: mrLRP "safe or unsafe," venous invasion (mrEMVI), depth of spread, node status, tumor height, and tumor quadrant. MRI-based treatment recommendations were compared against final management and pCRM outcomes.


Overall pCRM involvement was 9.0% [95% confidence interval (CI), 5.9-12.3], significantly lower than previously reported rates of 30%. Patients with no adverse MRI features and a "safe" mrLRP underwent sphincter-preserving surgery without preoperative radiotherapy, resulting in a 1.6% pCRM rate. The pCRM rate increased 5-fold for an "unsafe" compared with "safe" preoperative mrLRP [odds ratio (OR) = 5.5; 95% CI, 2.3-13.3)]. Posttreatment MRI reassessment indicated a "safe" ymrLRP in 33 of 113 (29.2%), none of whom had ypCRM involvement. In contrast, persistent "unsafe" ymrLRP posttherapy resulted in 17.5% ypCRM involvement. Further independent MRI assessed risk factors were EMVI (OR = 3.8; 95% CI, 1.5-9.6), tumors less than 4.0 cm from the anal verge (OR = 3.4; 95% CI, 1.3-8.8), and anterior tumors (OR = 2.8; 95% CI, 1.1-6.8).


The study validated MRI low rectal plane assessment, reducing pCRM involvement and avoiding overtreatment through selective preoperative therapy and rationalized use of permanent colostomy. It also highlights the importance of posttreatment restaging.


[Indexed for MEDLINE]
Free full text

Supplemental Content

Full text links

Icon for Wolters Kluwer Icon for White Rose Research Online
Loading ...
Support Center