Background: A transrectal (TR) approach for natural orifice translumenal endoscopic surgery (NOTES) makes sense for colorectal surgery because the colotomy can be incorporated into subsequent anastomosis. Because cancer is a primary indication for left-sided colon resection, oncologic standards will have to be met by a NOTES procedure. This study aimed to assess whether pure TR rectosigmoidectomy can be performed with strict adherence to oncologic principles compared with a conventional laparoscopically assisted approach (LAP).
Methods: Human male cadavers were allocated to either TR (n = 4) or LAP (n = 2). A simulated sigmoid lesion was created at 25 cm. Transrectal retrograde mobilization of the rectosigmoid was performed using conventional transanal endoscopic microsurgery (TEM) instrumentation. After ligation of the superior hemorrhoidal artery and further mobilization, the specimen was delivered transanally and divided extracorporeally. Using a circular stapler, NOTES colorectal anastomosis was performed. Lymph node yield, adequate resection margins, and operative time were compared with LAP.
Results: Transrectal retrograde rectosigmoid dissection was achieved in all attempts (4/4) and showed numbers of lymph nodes (median, 5; range, 3-6) similar to the LAP group (median, 4.5; range, 2-7). One pure TR approach failed to resect the lesion. Three TR procedures required additional mobilization via an abdominal approach to provide adequate margins. The mean length of TR specimens was 16 ± 4 cm compared with 31 ± 9 cm achieved by LAP (p < 0.01). The TR operative time was significantly longer (247 ± 15 vs 110 ± 14 min).
Conclusion: Lymph node yield during TR rectosigmoidectomy was similar to that achieved by the LAP approach. However, conventional TEM instrumentation alone did not permit adequate colon mobilization. This indicates a need for flexible instrumentation or other technical solutions to perform true NOTES colectomies.