Robotic Full Lobe Hepatectomy With Natural Orifice Extraction: Case Series Describing the Novel Technique of Robotic Major Hepatectomy and Transvaginal Specimen Extraction.

A 35-year-old female (body mass index [BMI] 23.1 kg/m 2 ) with a previous hysterectomy for uterine fibroids presented with a symptomatic 14.4 × 10.6 cm cavernous hemangioma diagnosed by magnetic resonance imaging (Fig. 2 A). She underwent a right hepatectomy with early ligation and division of the inflow vasculature. The total operating time (incision to extubation) was 8.17 hours. Estimated blood lost (EBL) was 1475 mL. Postoperative day 2 (POD2) average pain score (on visual analog scale between 0 and 10) was 3.5. Her postoperative course was uncomplicated, and she was discharged home 4 days after surgery. Preoperative imaging. A, Patient 1 magnetic resonance imaging demonstrating large right lobe hemangioma. B, Patient 2 arterial enhancing exophytic left lobe lesion on CT. C, Patient 3 CT of 10-cm intrahepatic angiomyolipoma in the left lobe. D, Patient 4 right lobe adenoma demonstrated on CT. CT, computed tomography scan. A 52-year-old female (BMI 19.3 kg/m 2 ) with a previous hysterectomy for uterine fibroids presented with a 5.1 × 4.6 × 3.5 cm exophytic arterial enhancing lesion in the left lobe consistent with an adenoma (Fig. 2 B). This mass had increased in size on serial imaging studies. The left hepatic artery and portal vein were isolated for inflow occlusion during the parenchymal transection. The total operating time was 3.37 hours. Estimated blood loss was 50 mL. POD2 average pain score was 2.8. Her postoperative course was uncomplicated, and she was discharged home 3 days after surgery. A 51-year-old female (BMI 53.0 kg/m 2 ) presented with a previous hysterectomy for abnormal uterine bleeding. A 10-cm fat-containing mass was incidentally identified in the left lobe of the liver on cross-sectional imaging. Biopsy confirmed an intrahepatic angiomyolipoma (Fig. 2 C). She was also found to have a 4.8-cm ovarian mass. She underwent a combined left hepatectomy and left salpingo-oopherectomy. The total operating time was 6.4 hours. Estimated blood loss was 700 mL. POD2 average pain score was 6.4. Her postoperative course was uncomplicated, and she was discharged home 3 days after surgery. A 46-year-old female (BMI = 25.3 kg/m 2 ) with a previous hysterectomy who presented with an 8.5 × 5.5 cm adenoma in the right lobe (Fig. 2 D). She underwent preoperative right portal vein embolization to allow for an adequate liver remnant. A robotic right hepatectomy was performed along with a wedge resection of a small left lobe lesion. The total operating time was 6.95 hours. Estimated blood loss was 500 mL. Final pathology demonstrated a telangiectatic type adenoma. POD2 average pain score was 4.25. Her postoperative course was uncomplicated, and she was discharged home 5 days after surgery. Table 1 compares our series of 4 patients with transvaginal extraction to a group of 30 consecutive females who underwent robotic right or left hepatectomies with specimen extraction from a pfannensteil incision between 2016 and 2020. In an analysis using the Wilcoxon rank test, there were no significant differences between the transvaginal or pfannensteil extraction groups with regards to total operative time (mean 6.22 vs. 5.88 hours, P = 0.49), EBL (mean 681 vs 605 mL, P = 0.65), POD2 average pain score (mean 4.50 vs 3.98, P = 0.38), time to return of bowel function (mean 3.25 vs 3.91 days, P = 0.18), or LOS (mean 3.75 vs 4.97, P = 0.08). Operative and Postoperative Outcomes Following Transvaginal Extraction Compared to Pfannensteil Extraction

Robotic and laparoscopic liver resection has risen in popularity due to an overall reduction in perioperative morbidity combined with evidence supporting equivalent oncologic outcomes for resection of malignant tumors when compared to open liver resection. 1 However, the traditional approach either uses an existing hand-assist incision for specimen extraction or requires extending a trocar incision. Both diminish the full potential benefits of a minimally invasive approach. The specimen extraction site has been shown to account for the majority of laparoscopic wound infections, incisional hernia, and postoperative pain. 2 – 4 To avoid this incision, an alternative is to utilize a natural orifice for specimen removal. Natural orifice specimen extraction has been described by a variety of surgical specialties including colorectal surgery, urological surgery, gynecologic surgery, hepatobiliary surgery, and abdominal transplantation. 5 – 9 The first transvaginal extraction for a cholecystectomy was performed in 1993 and nephrectomy with transvaginal extraction in 2002. 10 , 11 Since that time, there have been numerous case reports and series demonstrating the safety of transvaginal specimen extraction. Here, we present a series of 4 women with a history of hysterectomy who underwent robotic total right or left hepatectomy with transvaginal specimen extraction.

For robotic left hepatectomies, the patient is placed in the supine lithotomy position. For right hepatectomies, the patients are placed in a modified flank position to allow for better visualization behind the right lobe (Fig. 1 A). The Pfannenstiel incision is marked in case transvaginal extraction is not possible. When transitioning to the extraction portion of the procedure, the robot is undocked, bed tilted to a deep Trendelenburg position, and then redocked targeting the pelvis. A Veress needle is used for insufflation and a 5-mm trocar is introduced under direct visualization with a 0-degree laparoscope at the level of the umbilicus. Four 8-mm robotic trocars are placed across the mid abdomen (Fig. 1 B) with variable angulation depending on the patient anatomy and location of the liver lesion. Patient positioning and port placement. A, For right lobe liver lesions, the patient is placed in a modified decubitus position to provided excellent exposure of the structures posterior to the right lobe. B, The Pfannenstiel incision is marked in case transvaginal extraction is not possible. Four 8-mm robotic trocars are placed across the mid abdomen with variable angulation depending on the patient anatomy and location of the liver lesion. Cholecystectomy was performed in all cases. The hepatic artery and portal vein branches supplying the lobe being resected are then circumferentially dissected for early ligation. The hepatic artery is double tied on the proximal end, whereas the portal vein is tied early for inflow control and later divided with an endovascular stapler during parenchymal transection. Ultrasound examination of the whole liver is also performed to rule out additional lesions not identified on cross sectional imaging. The lobe is then fully mobilized before parenchymal transection to visualize the vena cava. The direct hepatic branches are tied with 2-0 silk on the vena cava, and divided using the Harmonic Scalpel (Ethicon Endo-Surgery Inc, Cincinnati, OH). Larger branches are oversewn with 4-0 prolene on the vena cava and hemolock clips applied on the hepatic end before division. Once the right hepatic vein is visualized, we proceed with parenchymal transection. Intraoperative ultrasound is used to map out the resection plane and to ensure adequate margins are obtained. The hepatic parenchyma is divided using the Harmonic Scalpel. This allows us to identify portal and hepatic vein branches that are either clipped, oversewn with 4-0 prolene, or divided directly with the Harmonic Scalpel. An Echelon Flex GST White load linear stapler is used to divide the right portal vein, bile duct, the right hepatic vein, and large branches. Once the liver is transected, the specimen is placed in a 12/15 mm endobag (Applied Medical #CD004). Hemostasis is ensured, and the cut edge of the liver is surveyed for bile leak before turning to the transvaginal extraction portion of the operation. Once the robot is docked with the patient in Trendelenburg position, the bowel is swept out of the way and adhesiolysis is performed as needed. The ureters and bladder are clearly identified. The apex of the vagina is placed on tension with a colpoprobe. The bladder is freed from the anterior vaginal wall. Posteriorly, the rectovaginal space may be entered to free the vaginal wall and reflect the rectum posteriorly. The course of ureters is kept in view during this process. The shears with monopolor energy is used to make a transverse colpotomy against the colpoprobe. The bagged liver specimen is then grasped through the colpoprobe and delivered transvaginally. The vagina is closed robotically using a self-locking running suture. Pneumoperitoneum may be maintained with a bulb from an asepto syringe inserted into the vagina. The pelvis is copiously irrigated. The patient is then repositioned for the final evaluation of the liver. A drain is placed along the cut surface of the liver, and the ports are removed under direct visualization.

Liver resection, including major hepatectomy for both benign and malignant indications, can be safely approached with a minimally invasive technique. 12 – 14 A meta-analysis comparing laparoscopic versus open hepatic resection showed comparable long-term survival and no difference in disease recurrence, when resections were performed for malignant indication. 1 In addition, the laparoscopic group had a shorter hospital length of stay as well as fewer days of intravenous or oral narcotic use. Despite these benefits with the laparoscopic approach, the specimens are often quite large, necessitating a mini-laparotomy for removal. Although transvaginal specimen extraction in itself is not a new concept, using this technique for removal of specimen following major hepatectomy has not been previously described. 15 – 18 In our case series of transvaginal hepatectomy extraction, we did not find significant differences in operative time, EBL, or time to return of bowel function when compared to a group of 30 females who underwent robotic hepatectomies with specimen extraction through a pfannensteil incision. Average pain score and LOS were shorter in the transvaginal group, but these differences were not significant likely due to small sample size. Other surgical fields have evaluated transvaginal specimen extraction with varying results. Awad and Griffin 8 evaluated outcomes following natural orifice specimen extraction during laparoscopic right hemicolectomy. Though the natural orifice specimen extraction group had better cosmetic scores, they found no difference in length of stay, oncologic outcome, and rates of hernia formation or surgical site infection. One patient developed dyspareunia that took 1 year to resolve. Of note, their study was not appropriately powered to demonstrate a statistically significant difference between the 2 groups for the majority of the outcomes evaluated. A randomized study looking at natural orifice specimen extraction versus conventional extraction for left-sided colon resection found significant improvement in postoperative pain, with no difference in complication rates or hospital length of stay. 19 Other studies have also demonstrated a reduction in postoperative pain scores, improved cosmetic outcome, and overall satisfaction, without additional blood loss or operative time, when natural orifice specimen extraction was compared to transabdominal incisional extraction. 9 , 20 , 21 There are theoretical benefits to natural orifice specimen removal including less postoperative pain, reduced surgical site infections, and lower rate of incisional hernia formation. However, transvaginal extraction although rare, does carry its own risk profile including dyspareunia, pelvic infection, dehiscence, and injury to the bladder, ureter, or rectum. 19 , 20 , 22 – 25 The risk of infection is greatly reduced with preoperative antibiotics and a thorough vaginal preparation. A study examining the microbiological colonization and contamination after transvaginal extraction cholecystectomy showed that vaginal disinfection with hexetidine tablets and octenidine significantly decreased the vaginal bacterial load and bacterial growth in routine cultures. In 27 patients who underwent transvaginal extraction, no postoperative surgical site infections occurred indicating a low risk of peritoneal contamination caused by the transvaginal access. 21 To reduce injury to the surrounding organs, the bladder is freed of the anterior vaginal wall and rectum may be deflected posteriorly by entering the rectovaginal space. The course of the distal ureters is noted during the dissection and during the colpotomy incision. In addition, this approach may be more technically challenging in the setting of an obliterated cul de sac from adhesions or endometriosis. Colpotomy and transvaginal tissue extraction is generally safe providing the following surgical practices are met: Prepping the vagina and using preoperative antibiotics to reduce the risk of infection. Freeing the bladder and the rectum from the vaginal wall. Identifying and knowing the course of the ureters. Providing good elevation and traction to the apex of the vagina during colpotomy incision. Although we elected to perform transvaginal extractions only in women who had hysterectomies, transvaginal extraction has been successfully performed in women with uterus. 26 Specimen sizes as large as 9 cm were successfully removed. 9 Recently an ovarian tumor 13 × 10 × 7 was removed intact following a 4- to 5-cm colpotomy incision (personal observation) illustrating the compliance of the vaginal wall in many patients. Although transvaginal extraction of smaller specimens such as partial hepatectomies might be possible without history of prior hysterectomy, this approach may not be feasible following large hepatic lobectomies and may depend on specimen characteristics and patient body habitus. This case series represents a proof of concept for natural orifice specimen extraction following major liver resection and was not intended to establish superiority in outcomes compared to a standard extraction approach. This operative approach warrants further research on a larger scale to clearly define the risks and benefits of this operative approach.

The authors would like to acknowledge Marcelle Philosophe for the medical illustrations provided.