The rectosigmoid problem: defining the rectum and optimising staging of sigmoid cancers with imaging

Abstract

Introduction The aim of this study was to research whether imaging could improve the outcomes of sigmoid colon cancer treatment, by better pre-operative staging and consistent differentiation of the rectum from the sigmoid. Methods I investigated sigmoid cancer outcomes by systematic review and meta-analysis. I performed a literature search and evidence synthesis on definitions of the rectum, and compared their accuracy in a radiological study. I led a Delphi consensus to agree a definition of the rectum international with colorectal experts. I designed a prospective, ethics-approved joint radiological-tissue study to both investigate the concept of a sigmoid mesenteric waist and validate an imaging-based definition of the rectum. I created and analysed a multi-institutional database to identify prognostic variables for colon cancer and compare Computed Tomography (CT) staging systems. Results The pooled local recurrence rate for sigmoid cancer on meta-analysis is high at 10.5%. A wide range of definitions of the rectum are employed by national associations and guidelines. In-vivo radiological comparison of all definitions confirmed they did not accurately localise tumours. Through a Delphi consensus of 132 international experts, the imaging landmark of the sigmoid take-off was agreed to define the rectum. A mesenteric waist was identified at the rectosigmoid junction that could serve as a pathological definition of the rectum. Radiological and specimen analysis validated the sigmoid take-off on imaging as a definition of the rectum. 5 In 414 patients the local recurrence rate of sigmoid cancer remained high at 12.6%, and was associated with CT-disease features. Identification of tumour deposits and extramural venous invasion on CT were associated with higher rates of recurrence than TNM or T3 substage based CT-systems. Conclusion There are a wide range of non-anatomical definitions of the rectum currently in use, which cannot accurately localize tumours. An international Delphi consensus definition of the rectum was the sigmoid take-off, which was validated on specimen analysis. Local recurrence of sigmoid cancer is unexpectedly high. Pre-operative imaging can identify patients that are at higher risk of recurrence.