The first robotic-assisted colorectal surgery with the Hugo™ RAS was a sigmoidectomy performed by our team on February 13th, 2023. It was at the same time the first-in-human colorectal procedure with the Hugo™ RAS was performed in Germany. Patients’ characteristics and outcomes are presented in chronological order of performed procedures in Tables 1 and 2.

Table 1 Patients’ characteristics of 25 colorectal procedures in chronological orderTable 2 Duration of surgery, estimated blood loss, complications and hospital stay

A total of 14 female and 11 male patients underwent ten sigmoid resections, seven low anterior rectal resections, four right and one left hemicolectomies, two Hartmann’s procedure reversals, and one abdominoperineal resection. The median age was 66 years and the median body mass index 26 kg/m2. The most common indication for surgery was colorectal cancer (n = 13), followed by chronic recurring diverticulitis (n = 10) and two patients needing a reversal of a Hartmann’s procedure due to perforated diverticulitis in the past. Seventeen patients had previous history of abdominal surgery causing relevant extent of adhesions in ten of them. A significant proportion of patients (17/25, 68%) suffered relevant comorbidities, seven patients (28%) were on anticoagulation, and four (16%) were on life-long steroid therapy.

Technical performance of the device

We experienced one technical problem with the device in the beginning of this series. In patient Nr. 3, one of the arms repeatedly did not recognize the instruments. The system had to be restarted twice until function was regained. The arm was successfully repaired before the next procedure. Also, a software update was necessary after the fifth procedure to eliminate some bugs. These technical problems led to a time delay, but no adverse clinical events. Subtle arm collisions were repeatedly detected, but tolerated by the device and did not interrupt proper function of the system. Neither instruments nor other hardware parts broke or showed defects. There were no problems with image or video transfer. There were no device-related patient injuries.

Sigmoidectomy and left hemicolectomy

Ten cases of sigmoidectomy for diverticulitis and one left hemicolectomy for cancer were performed based on the setup first proposed by Bianchi (Fig. 1) [8]. In seven patients, mobilization of the left colonic flexure was performed: in two of them fully robotic and in the rest combined robotic/laparoscopic. In four patients, a simultaneous robotic appendectomy was performed, in three cases a time-consuming robotic adhesiolysis was necessary. In two elderly female patients, the coexistent rectal prolapse was treated by a Frykman–Goldberg rectopexy. A 6 cm Pfannenstiel incision was used for specimen extraction. A descendorectostomy was performed at a median of 12 cm from the anal verge. A CEEA stapler was used in five patients, the rest received a two-layer continuous PDS 4-0 hand-sewn anastomosis: both patients with rectopexy, three others because of a narrow intestinal lumen not suitable for a stapler and in a patient with a kinking of the rectum, which made correct insertion of the stapler impossible. Docking time improved from 18 to 7 min over time. Median console time was 120 min. Median total operative time was 201 min. Median postoperative stay was 7 days. No conversions and no severe postoperative complications occurred: a female patient with a concomitant rectopexy was readmitted 3 weeks after surgery with pain in the lower abdomen and slightly elevated CRP—a 5 cm presacral haematoma was drained under CT guidance and antibiotics applied. An overweight male patient suffered in the early postoperative period a beginning of self-limited mild compartment syndrome of the lateral compartment of the right calf, which rapidly improved with mobilization.

Fig. 1

Left hemicolectomy and sigmoidectomy: cart setup (A) and trocar position (B)

Low rectal resections (LAR and APR)

The initial setup was the same as for sigmoid resection. This allowed vascular control of inferior mesenteric vessels as well as medial-to-lateral mobilization of the left colon. However, the port in the upper abdomen was usually far away from the pelvis. For adequate TME dissection down to the pelvic floor, two ports in the left lower abdomen were necessary and redocking performed. The fourth arm remained at its position, while the first three arms were rotated about 45° clockwise, so that the camera was attached to arm 3 between the legs, and arms 1 and 2 were used for the bipolar and the double fenestrated grasper (Fig. 2). The upper abdomen port could then be used to apply upward traction of the rectum, the assistant port for suction, irrigation, additional traction if needed and for stapled transection of the rectum. The two robotic hands allowed traction und countertraction and the dissection itself was done with the shears. In LAR, specimen extraction and insertion of the CEEA-anvil in the colon were performed via a 6 cm Pfannenstiel incision. In patients who underwent LAR and had already had a diverting sigmoidostomy, this step was done through the stomal opening. In APR and in LAR with a coloanal hand anastomosis, the specimen was extracted through the perineal wound or through the anus.

Fig. 2

Rectal resection, APR and Hartmann’s reversal: cart setup (A) and trocar position (B)

Eight patients underwent oncologic resections of the rectum for cancer, seven anterior resections and one abdominoperineal rectal extirpation. Tumors were located at a median height of 5 cm from the anal verge. In five patients, a total neoadjuvant chemoradiotherapy had been performed before surgery. One patient denied chemotherapy and had only short-term radiotherapy. Two patients with tumors in the upper third of the rectum did not receive neoadjuvant therapy. In two cases of ultralow anterior resection, a coloanal anastomosis was necessary, and the rest had stapled anastomosis. A loop ileostomy was diverted as standard in all but one of the patients. Four patients developed postoperative complications which resulted in a median postoperative stay of 11 days. Docking time improved from 15 to 8 min. Median console time was 241 min, while median total operative time was 345 min.

Hartmann’s reversal

In candidates for a Hartmann’s reversal, colostomy was closed, and colon repositioned into the abdomen. The defect was temporarily closed with a small Alexis wound protector and covered with its cap, with one of the left-sided lower abdominal ports inserted through its opening. The port positioning was the same as for LAR. Resected parts of the colon and the rectal stump were extracted through the stomal defect, which was used also for insertion of the CEEA anvil in the colon. We performed two Hartmann reversals with a median console time of 145 min and median total operative time of 295 min. Both patients recovered uneventfully.

Right hemicolectomy

There are two possible, basically different setup configurations for a right hemicolectomy. Our preferred setup was close to that one first reported by Bianchi, with three arms to the right of the patient and one arm at the left hip, connected to three ports in the left hemiabdomen and a port in the right iliac fossa (Fig. 3A and B) [8]. We also used once a modification of the vendor’s setup with a Z-formed port distribution directly to the left of the midline and a butterfly cart positioning; however, we suffered more collisions with it and there was less place and freedom for the bedside assistant (Fig. 3C).

Fig. 3

Right hemicolectomy: cart setup (A), standard trocar position (B) and alternative trocar position (C). C camera trocar, A assistant trocar, LH left hand of surgeon, RH right hand of surgeon, R reserve hand

None of the described setups was satisfactory and we are still in the process of developing our own setup for right hemicolectomy.

Docking time improved from 15 to 10 min. Median console time was 211 min, and median total operative time was 312 min. The long duration of surgery was partially caused by locally advanced tumours with infiltration of adjacent structures, lengthy adhesiolysis and additional pathology such as cirrhosis, liver metastasis or acute cholecystitis. Postoperative recovery was uneventful except for a clostridial infection in one of the patients, treated with oral antibiotics.

Concomitant procedures performed with Hugo RAS during colorectal surgery

A partial or complete takedown of adhesions was robotically performed in ten patients. In two cases, serosal defects occurred which were immediately robotically oversewn. In ten patients a robotic appendectomy was performed, and in the other two a robotic cholecystectomy. Diversion of loop ileostomy was robotically assisted in seven cases.

Systematic review of the literature

A systematic search in PubMed found 80 publications on the clinical use of Hugo™ RAS till April 20th, 2024. Only five of them reported data on a total of 23 colorectal resections [7,8,9,10,11]. Three Italian and two Spanish teams presented case series of one to nine patients. There were nine right and one left hemicolectomies, five ileocaecal, four rectal and four sigmoid resections. The summary of data is presented in Table 3. Port positioning, angles and tilts of arm carts described were not standardized and varied according to authors’ preferences and type of procedure; however, overall setups largely corresponded to our own shown in Figs. 1, 2, 3. Gender distribution (14 male and 9 female), median age of 66 and BMI of 26 kg/m2 did not significantly differ from those in our group. Median docking time was 10 min, median console time was 194 min and median total duration of surgery was 247 min. Median hospital stay was 5 days. No major technical problems with the device and no postoperative complications were reported.

Table 3 Summary of published case series on colorectal resections with Hugo™ RAS