Endoscopic submucosal dissection specimens in early colorectal cancer: lateral margins, macroscopic techniques, and...

Endoscopic submucosal dissection (ESD) allows en-bloc resection of superficial gastrointestinal tumors, providing specimens on which lateral margin analysis can be performed reliably. Positive lateral margins have been linked to higher rates of recurrence/residual tumor. There are no guidelines for macroscopic processing of lateral margins. Currently, most institutions use parallel lateral sections, which are difficult to interpret. We use perpendicular lateral sections, hypothesizing that it decreases potential artifactually positive lateral margins. We analyzed positive lateral margin rates in colorectal ESD specimens according to sectioning method. We also looked at morphological factors associated with margin positivity as a function of technique used. We studied 166 ESD specimens, on which parallel sectioning practiced from 2006 to 2011 (n = 75). Perpendicular sectioning was used from 2010 to 2015 (n = 91). We recorded the number of positive margins, along with grade of dysplasia/carcinoma. Other information such as histopathological type, specimen size, lesion location, and patient follow-up was also recorded for evaluation. Forty of seventy-five (63%) margins were positive for parallel sections. In contrast, perpendicularly cut margins were significantly less frequently positive: 22/91 (24%) (p = 0.0001). Positive margins were found significantly more frequently in tubulo-villous lesions compared to tubular lesions in both the parallel and perpendicular groups (p = 0.03 and p = 0.02, respectively). Specimen size was not significantly associated with positive margins. Using perpendicular sectioning of colorectal ESD specimens, the proportion of cases with a positive lateral margin was significantly lower than when parallel sectioning was used. We suggest perpendicular sectioning to improve accuracy in histopathological analysis. This method is particularly important to use in future studies, as it may prevent authors from making conjectures based on overestimation of positive lateral margins.