A Pure Natural Orifice Transrectal Approach for Oncologic Resection of the Recto-Sigmoid: an Experimental Randomized Comparison to Conventional Laparoscopy

Erwin Rieder, MD, Georg O Spaun, MD, Yashodhan S Khajanchee, MD, Danny V Martinec, BS, Brittany N Arnold, BS, Lee L Swanstrom, MD, Mark H Whiteford, MD. MIS Program, Legacy Health, Portland, OR

Objective: Natural orifice translumenal endoscopic surgery (NOTES) has long been criticized for breaching an otherwise intact organ. A transrectal approach though, makes immense sense for incisionless colorectal surgery as the access colotomy is incorporated into subsequent anastomosis. As cancer is one of the primary indications for left sided colon resections, surgical oncological requirements also have to be met by a NOTES procedure. The aim of this experimental cadaver-study was to assess whether a pure transrectal (TR) recto-sigmoidectomy using standard transanal endoscopic microsurgery (TEM) instrumentation can be performed with strict adherence to oncological principles similar to a conventional laparoscopic-assisted approach (LAP).
Methods: Frozen then thawed male cadavers were randomized to either TR (n=4) or LAP (n=2) sigmoidectomy. A simulated polyp was created about 20 cm from the anal verge. For the transrectal procedure standard TEM instrumentation was used. After transluminal access through the mesorectum and into the abdominal cavity, retrograde en bloc mobilization of the recto-sigmoid was attempted. After ligation of the superior hemorrhoidal artery, sigmoid mobilization was continued until instrument length limited further dissection. The specimen was delivered transanally and subsequently resected. For the creation of a stapled colorectal anastomosis a 29-mm EEA stapler anvil was introduced into the proximal colon and secured with a purse-string suture with a long tail left attached. Consecutively, a loose rectal purse-string suture was sewn at the proximal rectum with the reinserted TEM platform. When the anvil was delivered into the rectum by using the attached suture-tail as a handle, the loose rectal purse-string suture was tightened. The stapler was placed, mated with the anvil and fired. Results regarding lymph node yield, adequacy of margins, and operative time were compared with conventional LAP utilizing an additional fenistil incision for specimen retrieval.
Results: Transrectal recto-sigmoid dissection by TEM was achieved in all (4/4) attempts. En bloc sigmoid resection revealed similar numbers of lymph nodes in the TR group (median: 5; range: 3-6) compared to the LAP group (median 4.5; range: 2-7). One pure NOTES approach using standard instrumentation failed to resect the previously created lesion due to limited specimen length (10 cm) after transanal delivery. Three TR procedures required additional laparoscopic assistance to adequately mobilize the sigmoid colon for transanal resection and pure NOTES colorectal anastomosis. Mean length of the TR resected specimen was 16±4 cm compared to 31±9 cm achieved by LAP (p<0.01). Insufficient anastomosis was detected in one pure TR attempt (1/4). Transrectal operative time was significantly longer compared to LAP (247 ± 15 min vs. 110 ± 14 min, p<0.01).
Conclusion: Lymph node yield during TR-sigmoidectomy was found to be similar to that obtained by the conventional laparoscopic assisted approach. Standard TEM instrumentation, however, did not allow adequate colon mobilization, which was the most pressing technical challenge and has to be addressed in additional studies. A TR hybrid approach, using laparoscopic assistance, could initially overcome several drawbacks and currently appears to be the more appealing option for transrectal colorectal resection.