Robotic-assisted total mesorectal excision with the single-docking technique for patients with rectal cancer

We included 95 patients with stage I-III rectal cancer (adenocarcinoma) who underwent totally robotic-assisted TME with the single-docking technique with the da Vinci® surgical system at a single-institution between September 2013 and December 2016. This study was approved by the institutional review board of the Kaohsiung Medical University Hospital (KMUHIRB-E-20150003). Written informed consent to participate was obtained from each patient before performing the robotic surgery.

All patients routinely underwent preoperative colonoscopy and abdominal and pelvic computed tomography (CT) or magnetic resonance imaging (MRI) for preoperative staging. On the basis of the distance from the anal verge, rectal cancer was categorized into upper (11–15 cm), middle (6–10 cm), and lower (≤ 5 cm). Patients with T3, T4, or N+ rectal cancer received preoperative CCRT. Furthermore, the 5-fluorouracil, leucovorin, and oxaliplatin (FOLFOX) regimen was prescribed for patients with T4 or N+ rectal cancer and a fluoropyrimidine-based regimen was prescribed for patients with T3 N0 rectal cancer. Long-course radiotherapy (LCRT, total 5000 cGy in 25 fractions) was concurrently administered. Totally robotic-assisted TME with the single-docking technique was scheduled after more than 6 weeks after radiotherapy completion.

Clinicopathological features and perioperative parameters or outcomes were collected and evaluated, including age; sex; histological type; tumor, node, and metastasis (TNM classification); perineural invasion; vascular invasion; time interval between completion of preoperative radiotherapy and robotic surgery; tumor location (distance from anal verge); pre-CCRT, preoperative, and postoperative serum carcinoembryonic antigen (CEA) levels; and body mass index (BMI). The TNM classification was defined according to the criteria of the American Joint Commission on Cancer (AJCC)/International Union Against Cancer (UICC) [16]. The tumor regression grade (TRG) was evaluated according to the AJCC system [17]. Perioperative outcomes were collected and evaluated, including surgical procedures, docking time, console time, operation time, estimated blood loss, time of the first flatus passage, time of resuming soft diet, duration of postoperative hospital stay, and postoperative first day visual analog scale (VAS) pain score.

Patients were regularly followed up, including the collection of their clinical outcomes and survival statuses. History-taking and physical examinations were performed postoperatively every 3 months during the first 2 years and then every 6 months during the following 3 years. Measure of serum CEA levels were performed every 2–3 months postoperatively. A colonoscopy was performed approximately 1 year after surgery. Repeat colonoscopy was typically recommended at 3 years, unless follow-up colonoscopy indicated advanced adenoma (villous polyp, polyp > 1 cm, or high-grade dysplasia). Abdominal and pelvic CT scans were annually performed during postoperative 3 years in patients with stage II–III disease.

For all patients, we performed laparoscopic examinations to initially examine the intra-abdominal cavity. If an adhesion was observed, we performed laparoscopic lysis. Subsequently, we performed robotic surgery. The single-docking technique with five or six ports (Fig. 1) was used as the docking method, as described in our previous studies [18, 19]. One 12-mm camera port was placed 2 cm superior to the umbilicus. One 8-mm port (Arm 1 port) was inserted approximately 2 cm inferior to the line between the location of the camera port down to the right anterior superior iliac spine and slightly medial to the right mid-clavicular line (MCL). One 8-mm port (Arm 3 port) was inserted right laterally 8 cm from the Arm 1 port. One 12-mm port (assistant port) was inserted at the right MCL, approximately 4 cm inferior to the right costal margin. One 8-mm port (Arm 2 port) was inserted left laterally 8 cm from the camera port. One 12-mm port (assistant port) was inserted at the left MCL, approximately 2–4 cm inferior to the left costal margin. A monopolar permanent cautery spatula (Intuitive Surgical) was used in Arm 1, a Maryland bipolar forceps (Intuitive Surgical) was used in Arm 2, and a double fenestrated grasper (Intuitive Surgical) was used in Arm 3. The da Vinci® Si Surgical System was docked over the left flank of a patient. We performed medial to lateral dissection. Peritoneal incision at the level of the sacral promontory was performed first. The dissection was extended downward and then upward to the root of the IMA. We performed so-called high dissection and low ligation [19] in the form of D3 lymph node dissection and low-tie ligation of the IMA by using endo clips (Hem-O-Lok, Weck Closure Systems, NC) with preservation of the left colic artery in all patients. The inferior mesenteric vein (IMV) was also recognized, but was not ligated and divided instantly. If there was tension during the colonic anastomosis, the IMV would be ligated by using endo clips (Hem-O-Lok, Weck Closure Systems, NC) and divided. The splenic flexure of the colon was not routinely mobilized, if its mobilization was dependent on the tension of the anastomosis. Totally robotic-assisted TME with single-docking technique was performed in all patients.

After the sigmoid or descending colon, mesocolon, entire rectum and mesorectum were mobilized completely, low anterior resection (LAR) with the double-stapled technique, intersphincteric resection (ISR) with coloanal anastomosis and loop colostomy, or abdominoperineal resection (APR) was accordingly performed [18, 19]., LAR with the double-stapled technique was used for a tumor located in the upper and mid rectum. The rectum was divided by the assistant using an Endo GIA stapler (Endo GIA™ Reinforced Reload with Tri-Staple™ Technology, Medtronic) or ECHELON FLEX™ Powered ENDOPATH® stapler (Ethicon US, LLC) with one to three 60-mm reloads before the da Vinci® Si Surgical System was undocked. We extracted the specimen through the extended camera port wound with the Alexis® wound proctor and resected it. We then re-established the pneumoperitoneum and performed laparoscopic anastomosis by using a circular EEA stapler. Intraoperative dye test [20] was routinely performed to examine potential anastomotic leakage after LAR using the double-stapled technique. For a tumor located in the low rectum, ISR with coloanal anastomosis and loop colostomy was used. We used the Lone Star Retractor System® (Lone Star Medical Products Inc., Houston, TX) for ISR and subsequently we extracted the specimen and resected it transanally (natural orifice specimen extraction). Coloanal anastomosis was performed using the hand-sewn method. A protective loop colostomy of transverse colon was performed. Finally, we checked for bleeding in the abdominal cavity by using the traditional laparoscope and placed a drain tube in the pelvic cavity.

The baseline characteristics and perioperative outcomes of 95 patients with rectal cancer who underwent totally robotic-assisted TME with the single-docking technique were summarized in Table 1. The median age and BMI of the patients was 62 (range, 28–88) years and 23.54 (range, 17.20–34.02) kg/m², respectively. Of the 95 patients, 48 (50.5%), 30 (31.6%), 17 (17.9%) had lower, middle, and upper rectal cancers, respectively. The median distance of the tumor from the anal verge was 5.5 (range, 1.0–15.0) cm.

The most frequent surgical procedure was LAR (59/95, 62.1%). ISR with coloanal anastomosis was performed in 32 (33.7%) patients, and APR was performed in 4 (4.2%) patients. Moreover, of the 32 patients undergoing ISR, 3 underwent transabdominal ISR and their tumor distances from the anal verge were 2–4 cm. Positive dye leakage after the completion of anastomosis was identified in six patients who had undergone LAR. Protective colostomies were performed accordingly. Finally, protective diverting loop transverse colostomy was performed in 38 (43.9%) patients, including 32 patients and 6 patients who underwent ISR and LAR, respectively. Sphincter preservation rate was 95.8%. The median estimated blood loss including tissue fluid after CCRT was 80 mL. The median time of the first flatus passage and resuming soft diet postoperatively was 2 and 4 days, respectively. The median duration of postoperative hospital stay was 6 days (range, 5–32).

The postoperative complications are summarized in Table 2. Postoperative complications were observed in 14 patients with 17 episodes (17.9%). Three patients who developed intraabdominal abscess, CT-guided pigtail drainage were subsequently performed in 2 patients. Anastomosis leakage was observed in 5 (5.4%) patients who underwent LAR with the double-stapled technique, and loop colostomy of transverse colon was subsequently performed. Four (4.2%) patients developed stenosis of coloanal anastomosis and underwent dilation using a colonoscope. Urethral injury during ISR was noted in one (1.0%) patients. According to the Clavien-Dindo Classification, all post-operative ileus, urinary tract, and pulmonary complications were of grades I, and the patients recovered after conservative treatment. Moreover, no 30-day hospital mortality occurred.

The pathological characteristics and oncological outcomes of all 95 patients are listed in Table 3. Preoperative clinical staging demonstrated that the majority of the patients had locally advanced rectal cancers including T3 in 61 (64.2%) patients, T4 in 13 (13.7%) patients, or N+ in 57 (60.0%) patients. Therefore, preoperative CCRT was performed in 75 (78.9%) patients, including FOLFOX regimen in 58 (77.3%) patients with cT4 or cN+ disease, fluoropyrimidine-based regimen in 17 (22.7%) patients. The median number of harvested lymph nodes and apical lymph nodes was 9 (range, 0–36) and 2 (range, 0–15), respectively. However, positive apical lymph node metastasis was observed in only 3 (2.9%) patients. The median distance of the distal resection margin (DRM) and circumferential resection margin (CRM) was 2.30 and 1.0 cm, respectively. CRM and DRM were positive in 2 (2.1%) and 1 (1.1%) patients, respectively. R0 resection for primary rectal cancer was performed in 92 (96.8%) patients. Of the 75 patients who received preoperative CCRT, a pathologic complete response (pCR) of the primary tumor was observed in 27 (28.4%) patients. 28 (37.3%), 30 (40.0%), 11 (14.7%), and 6 (8.0%) patients exhibited complete response (TRG 0), moderate response (TRG 1), minimal response (TRG 2), and poor response (TRG 3), respectively. The median time interval between radiotherapy completion and robotic surgery was 82 (range, 41–203) days.

The median follow-up duration of 95 patients from the primary treatment was 25.6 (range, 6.6–52.2) months. One patient undergoing local excision of primary tumor and radiotherapy at other hospital underwent chemotherapy with FOLFOX regiment and robotic ISR and coloanal anastomosis at our hospital after local recurrent tumor developed. We excluded this patient to analyze the oncological outcomes of the patients with undergoing R0 resection. Of 91 patients undergoing R0 resection, local recurrence and distant metastases were noted in 5 (5.5%) and 10 (11.0%) patients, respectively. At a median follow-up duration of 25.6 months, the 2-year OS was 94% and 2-year DFS was 83% (Fig. 2). Furthermore, 2-year local control rate and 2-year distant metastasis control rate were 95% and 90%, respectively.

The learning curves in terms of console and operation time are presented in Fig. 3. The median docking, console and operation time was 5 (range, 3–22), 200 (range, 130–435), and 325 (range, 210–795) minutes, respectively. A linear regression analysis indicated a decreasing trend for console time. The first plateau of console time was observed after 32 patients. The mean console time for the first 32 patient was significantly longer than that of the remaining patients (270.09 ± 64.830 vs 200.27 ± 42.080 min, P < 0.001). The linear regression analysis of operation time also indicated a decreasing trend for operation time. The mean operation time for the first 32 patient was significantly longer than that of the remaining patients (516.09 ± 11.460 vs 306.03 ± 6.804 min, P < 0.001).