Do or doughnut: A systematic review and pooled analysis on the utility of pathological evaluation of the anastomotic doughnut in oncological colorectal operations.

Abstract

End-to-end anastomosis staplers are frequently used in colorectal surgery, generating two anastomotic doughnuts. Whether pathological evaluation of the doughnut changes clinical practice remains unclear. We aim to identify any effects of pathological evaluation of anastomotic doughnuts after oncological colorectal surgery. We performed a systematic literature search utilizing PubMed, Clinicaltrials.gov, Cochrane, Embase and Web of Science databases and selected studies on evaluation of the anastomotic doughnut after oncological colorectal surgery with stapled end-to-end anastomosis. Outcome measures included: involved distal margin on the oncological sample, histological involvement of the doughnut, clinical change in management from a positive doughnut and study recommendations. Of the 5761 studies identified, eight studies encompassing 1754 patients were evaluated. Most operations were for primary colon (37.5%) or rectal adenocarcinoma (37.5%). Incidence of distal margin involvement of the oncological sample was reported in three papers, with six positive cases (1.1%). Of the 1754 doughnut pairs evaluated, five were positive for neoplasia (0.29%), three for adenomas (0.18%) and one for metaplastic polyp (0.06%), none of which changed postoperative treatment. Four studies recommended abandoning routine histopathological evaluation of anastomotic doughnuts, while the remaining four recommended evaluation only under certain criteria, including gross distal margin <2cm (one study), gross distal margin <3cm (one study), tumours undetected on gross examination (one study), 'histologically aggressive cancers' or grossly involved distal margin (one study). Routine evaluation of anastomotic doughnuts should be reconsidered, as <1% are positive for neoplasia. Exceptions may include specific scenarios where histopathology is likely to be clinically useful.