Introduction

Through opposite trends of invasiveness, gastrointestinal (GI) endoscopy and laparoscopic surgery have converged to produce an intriguing but still largely experimental field termed natural orifice transluminal endoscopic surgery (NOTES). Since the publication in 2004 by Kalloo and colleagues¹ on transgastric peritoneoscopy in the porcine model, NOTES continues to develop as a minimally invasive, “scar-less” technique for intraperitoneal and more recently intrathoracic surgeries. The theoretical advantages of NOTES include cosmesis; decreased anesthesia time; shortened postoperative recovery; and elimination of postsurgical complications associated with transabdominal and transthoracic incisions, such as skin infections, wound dehiscence, and hernias. Early questionnaire studies indicated that patients would prefer a NOTES approach over a traditional laparoscopic approach primarily because of issues surrounding pain and visible scars.²

The definition of the NOTES concept is in flux and currently spans a continuum. Initially, the concept referred to surgeries performed purely via a natural orifice; and these procedures are referred to in this chapter as pure NOTES. More recently, the concept of hybrid NOTES has evolved to allow for concomitant assistance provided by laparoscopy as a bridge to pure NOTES. In addition, the concept of NOTES now also encompasses single-incision laparoscopic surgery (SILS) (i.e., single-port laparoscopy or transumbilical laparoscopy), which technically is not surgery via an orifice but rather the coalescing of several laparoscopic ports into a single multilumen transabdominal trocar accommodating rigid laparoscopic instruments.

Moving forward, how might gastroenterologists fit within the broad and shifting NOTES landscape? Which NOTES procedures might become realistic procedures for GI endoscopists? An analysis of the published literature provides clues to factors that may influence the answer.

Current State of Natural Orifice Transluminal Endoscopic Surgery (NOTES) Research

Since the first NOTES publication by Kalloo and colleagues in 2004,¹ there has been progress toward human NOTES in numerous international centers but only a handful of U.S. centers. As of July 2009, the Natural Orifice Surgery Consortium for Assessment and Research NOTES registry had 166 human cases from eight U.S. sites and two international sites, 162 of which were approved by an institutional review board.³ SILS cases were not included in this registry. Most of these natural orifice cases were performed via the transvaginal route because of the historical precedent of culdoscopy performed by gynecologists, a reduced concern regarding sterility, and a decreased need for “secure” closure of the posterior fornix access point compared with enteral closure. Most of these cases also required laparoscopic assistance and would be more accurately categorized as hybrid NOTES procedures. Of the few pure NOTES procedures, Swanstrom and coworkers⁴ succeeded in performing transgastric cholecystectomy. Other pure NOTES procedures on closer examination have been hybrid transgastric procedures. There have been two small published case series of transgastric cholecystectomies assisted by the use of one to three transabdominal trocars.^5,6 There has also been a series of 10 diagnostic transgastric peritoneoscopies performed after laparoscopic exploration before Whipple’s surgery,⁷ but the challenge of access closure was not directly addressed. Generally, most pure NOTES publications to date have been animal and human cadaveric studies.

Of the eight fundamental obstacles to NOTES identified in the original 2005 White Paper (Box 54.1),⁸ the most challenging obstacles to pure NOTES human studies have been the ones heavily dependent on device development.⁹ Two specific areas of technical need are highlighted here. First, current closure devices remain in various stages of development with none emerging as the clear favorite. Viscerotomy closure must be 100% secure and ideally be easy to perform and reproducible for any chance of clinical adoption of NOTES. No device has yet fulfilled all of these requirements. The second and probably the most rate-limiting step in the clinical adoption of NOTES is the commercial development of a robust, flexible, versatile multitasking platform. Complex surgical maneuvers, even in surgeries as commonplace as cholecystectomies, have technical requirements that cannot be met by a double-channel endoscope, such as triangulation of instruments, aggressive traction and countertraction of tissue, position fixation, and reliable force transmission over instruments. The NOTES platform should be versatile enough to manage iatrogenic intraperitoneal complications, such as hemorrhage, perforation, or organ injury. It is not surprising that there has been such a preponderance of hybrid NOTES studies and a general tendency to retain the laparoscopic paradigm insofar as SILS is concerned.

Future Indications for the Gastroenterologist

Despite the many limitations on the translation of NOTES from animal and cadaveric studies to actual human trials, endoluminal GI endoscopy has steadily grown bolder in procedures involving and sometimes traversing the gut wall in ways not heretofore considered as NOTES per se. Endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) have naturally evolved to attempts at endoscopic full-thickness resection (EFTR) for early GI cancers. Endoscopic treatment of pancreatic disease has expanded from endoluminal treatment of the pancreatic duct (endoscopic retrograde pancreaticocholangiography [ERPC]) to transgastric or transduodenal needle puncture for sampling of pancreatic lesions (endoscopic ultrasound [EUS]–guided fine-needle aspiration) and for endoscopic pseudocyst drainage. Transgastric retroperitoneal necrosectomies have also been increasingly performed.

This chapter elaborates on five areas we believe are potential NOTES procedures that will be performed by gastroenterologists in the future (Box 54.2). These procedures share one or more of the following characteristics: (1) surgical sites more easily reached using a flexible endoscope or flexible NOTES platform compared with a laparoscope or via open access, (2) surgical procedures of low to medium complexity, and (3) probable improved outcomes from adjunct use of EUS for access or directed biopsies or resections.

Endoscopic Full-Thickness Resection

EFTR represents the next step in the natural evolution of endoscopic procedures aimed at removing GI cancers and cancerous precursor lesions. EFTR follows in the line of polypectomy, EMR, and ESD. For early stage gastric cancers, ESD cures certain T1, N0 adenocarcinomas. EFTR may find a role in the treatment of T2, N0 adenocarcinoma and small GI stromal tumors.¹⁰ In 2006, Kaehler and associates¹¹ reported EFTR in two patients (one with early gastric cancer and the other with carcinoid tumor) using a flexible stapler. There have been other case reports of gastric and duodenal EFTR but with laparoscopic assistance.^12,13 In these cases, resection was performed endoscopically, but closure was performed laparoscopically. A full-thickness suturing device has been used to close comparably smaller gastric defects during or after EMR of gastric tumors¹⁴ and a perforation resulting from GI stromal tumor resection using ESD.¹⁵

Although experience of EFTR in other organs has been reported mostly in animal studies, human studies are likely to be performed in the near future. Successful EFTR has been reported in an animal survival study.¹⁶ Colonic full-thickness resection has also been successfully shown in the animal model using various closure strategies, including but not limited to T-tags,¹⁷ clips,¹⁸ nitinol compression clamps,¹⁹ and staplers.²⁰

Diagnostic Peritoneoscopy and Lymph Node Harvesting

NOTES peritoneoscopy with lymph node harvesting has potential as a minimally invasive and accurate means of intraabdominal cancer staging. In patients in whom operability is uncertain, NOTES peritoneoscopy and lymph node mapping may be preferable to staging laparoscopy. In pancreatic, biliary, or gynecologic cancers, transgastric peritoneoscopy could be used to assess quickly for peritoneal and liver implants.²¹ Diagnostic transgastric endoscopic peritoneoscopy has already been performed in humans for staging of pancreatic head masses. In this particular study, transgastric peritoneoscopy corroborated laparoscopic findings for surgical decision making in 19 of 20 patients.²²

NOTES locoregional sentinel node sampling could potentially be combined with EFTR in the same procedure.⁹ In early stage gastric cancer, gastric EFTR could be performed with perigastric lymph node harvesting. In early stage colon cancer, direct NOTES sentinel node sampling could potentially save a patient from radical mesenteric lymphadenectomy.²³ Cahill and colleagues^24,25 reported the technical feasibility of lymphatic mapping and node biopsy by NOTES in a porcine model. They performed transvaginal endoscopic peritoneoscopy for gastric lymphatic mapping²⁴ and transgastric endoscopic peritoneoscopy for lymphatic mapping in the sigmoid colonic mesentery.²⁵ Methylene blue was injected into the submucosal layer for identification of lymphatic channels and lymph nodes. Excisional biopsy was performed using a coagulating electrode and an endoscopic grasper.

Finally, lymph node harvesting is of particular interest for staging of neoplasias of the chest and mediastinum as an alternative to thoracoscopy and mediastinoscopy, which can be quite painful.²⁶ The adjunctive use of EUS is likely to be important for safe mediastinal entry. In addition to lymph node harvesting, more recent human cadaveric studies have shown that other procedures may be possible, such as pleural biopsy, vagotomy, thymectomy, thoracic duct ligation, thymectomy, and pericardial window formation.²⁷

Liver Biopsies

When directed core liver biopsies are required, the transgastric approach may play a role. Steele and coworkers²⁸ showed transgastric liver biopsy in three patients undergoing laparoscopic gastric bypass surgery, adding only 4 to 6 minutes to the primary procedure. Segments II, III, IVb, V, and VI (partial) were able to be visualized. Self-limited minor bleeding was encountered after obtaining liver biopsy specimens using standard endoscopic biopsy forceps. Because this study protocol was added to gastric bypass surgeries, gastrostomy closure was not specifically addressed. In another study evaluating liver biopsies in a porcine model, animals were randomly assigned to liver biopsy by laparotomy, laparoscopy, and transgastric NOTES. Both the incidence and the severity of adhesions were lowest with transgastric NOTES.²⁹

Bariatric Surgery

Natural orifice bariatric procedures may not be transluminal per se but are likely to represent a significant opportunity for therapeutic gastroenterologists in the future given both the obesity pandemic and the need for a safe and durable alternative to bariatric surgery. Ultimately, avoidance of the transabdominal route makes intuitive sense in morbidly obese patients. Numerous strategies are available for endoluminal treatment of obesity, including suturing, stapling, implantable polymers and prostheses (e.g., sleeves), electrical stimulation, and mucosal ablation.³⁰ These strategies have been developed to replicate the understood mechanisms of bariatric surgery, classically described as restrictive, malabsorptive, or both.³¹ Preclinical studies have shown promising early data regarding safety and efficacy (percent excess weight loss 24% to 58% depending on the device).^32,33 Understanding of the pathophysiology of obesity and the mechanisms of bariatric surgery is in its infancy. The complex interplay between anatomic alterations and neurohormonal modulation remains to be worked out, but the elucidation of critical mechanisms promises to yield bariatric devices and procedures for the gastroenterologist’s armamentarium.³²

Anastomosis

Creation of an endoscopic anastomosis has been a goal of NOTES since its inception. Endoscopic gastrojejunostomy for palliation of malignant gastric outlet obstruction would theoretically be a superior alternative to surgical gastrojejunostomy at the end of life. Similarly, endoscopic creation of a cholecystogastrostomy could enable direct endoscopic treatment of gallbladder disease or allow for prophylactic gallstone removal and mucosal ablation. Early feasibility work in endoscopic anastomosis creation has featured the use of T-tags or endoscopic suturing devices.^34,35 Additional work is under way in our laboratory featuring injectable, self-assembling magnets, which could significantly simplify endoscopic anastomotic creation (unpublished data).

Equipment

The evolution of NOTES depends on technical development. The most pressing need is a safe and reproducible means of closing perforations. Several strategies are available for perforation closure—suturing, tissue anchors, staples, clips, occluding devices.³⁶ Requirements would differ depending on the tissue involved (e.g., gastric vs. colonic vs. esophageal). Similarly, there is still a need for robust, flexible, multitasking platforms. Promising prototypes of direct drive systems and multitasking flexible endoscopes are now being tested, and it is hoped that these new tools would enable more complex surgical maneuvers.³⁷

Anatomic Considerations

The transgastric approach provides excellent access to the lower abdomen and pelvis (Table 54.1). To view the upper abdomen from a transgastric access point, the endoscope must be retroflexed, which can limit maneuverability and reach. Peritoneal access is targeted along the anterior gastric body guided by transabdominal palpation. When creating the gastrostomy, the endoscopist should be mindful of the gastroepiploic arteries running along the greater curvature, the left lobe of the liver, and the anterior abdominal wall. The most frequently used technique for access involves using a needle-knife attached to cautery. A wire is passed through the needle-knife into the peritoneal cavity. An endoscopic dilating balloon is advanced over the wire to dilate the gastrostomy tract. The endoscope is advanced into the peritoneal cavity. Alternatively, the gastrostomy can be created using the percutaneous endoscopic gastrostomy tube technique whereby a transabdominal needle delivers the wire over which a balloon is advanced to dilate the gastrostomy. Another option is to use the tunneling method whereby a 5- to 15-cm submucosal tunnel is created before serosal exit to create a gastric flap that self-approximates and seals more easily (Figs. 54.1, 54.2, and 54.3).

Fig. 54.1 Natural orifice transluminal endoscopic surgery (NOTES) gastrostomy creation. Laparoscopic view with endoscopic view (inset) in human cadaver.

Fig. 54.2 Natural orifice transluminal endoscopic surgery (NOTES) gastrostomy access. Laparoscopic view in human cadaver.

Fig. 54.3 Transgastric appendectomy. Endoscopic view with laparoscopic view (inset) in human cadaver.

In contrast to the transgastric approach, the transcolonic approach provides excellent access to the upper abdominal organs, but it requires retroflexion to visualize and engage the lower abdominal organs (see Table 54.1). Usually an entry point is chosen in the proximal anterior rectum, just proximal to the superior valve of Houston, because the proximal third of the rectum is covered by visceral peritoneum.³⁸ The peritoneal reflection is generally 7 to 9 cm from the anal verge in men and 5 to 7.5 cm from the anal verge in women. This anterior peritonealized space is the rectovesicle pouch in men and the rectouterine pouch in women. As with transgastric access, the needle-knife is used to create the initial transmural cut. However, there is usually no need to balloon dilate as in transgastric access because of the comparatively thinner colonic wall. Other sites of inferior access include transvaginal and transvesicular, which are briefly mentioned here. As described previously, most recent NOTES literature has used the transvaginal route given a historical precedent in the form of culdoscopy, decreased concerns of sterility, and the adequacy of hand-sewn sutures for closure of the posterior fornix. However, this technique is available only to female patients, and gastroenterologists would likely not be involved in this approach. The transvesicular approach has not seen much clinical use.

The transesophageal approach (see Table 54.1) is usually guided by the use of EUS to identify an area of the mediastinum beyond the esophageal wall that appears suitable for entry, with no vessels. For viscerotomy creation, most published studies (all in animals or cadavers) use either the direct needle-knife technique or a modified tunneling method for “flap” access. As with the transcolonic approach, the transesophageal approach is associated with concerns of infection (i.e., mediastinitis). Closure methods for transesophageal access in animal studies have included clips and a T-tag suturing system, but a more fail-safe closure method is likely required before embarking on human trials.

Peritoneal and mediastinal anatomy would initially be foreign to the endoscopist accustomed to endoluminal navigation. To “recalibrate” the endoscopist to new anatomic perspectives, part of the answer may lie in a formal NOTES training fellowship.³⁹ Although such a fellowship would incorporate fundamentals of surgical training (e.g., surgical exposure, understanding tissue planes during dissection), NOTES would cover new ground even for the surgeon. NOTES allows for full prone positioning of the patient, which in the porcine model has shown full exposure of the retroperitoneum, a view not seen even by laparoscopic surgeons.

Additionally, part of the answer may lie with complementary, real-time image-guided navigational systems to compensate for difficulties with spatial positioning and orientation, identification of anatomy, and localization of pathology.⁴⁰ A real-time three-dimensional CT guidance system has been employed for navigational assistance during colonoscopy, ERCP and EUS, and NOTES procedures in human cadaveric studies.⁴¹ This system can track endoscope position with a three-dimensional volumetric reconstruction of a preoperative computed tomography scan (Fig. 54.4).

Fig. 54.4 Image registered gastroscopic ultrasound (IRGUS) system.

Case Preparation

It is unlikely that patient preparation for NOTES procedures would be greatly different from preparation protocols for current endoscopic procedures. For upper GI procedures, patients would likely be placed on nothing by mouth (NPO) status for 6 to 12 hours and may be required to undergo an additional purge preparation depending on the intended procedure (e.g., NOTES bypass). Intravenous antibiotics would likely be administered. Early NOTES procedures would also probably require patients to undergo general anesthesia to satisfy institutional review board requirements. Sedation protocols may transition to intravenous conscious sedation as the body of safety data accumulates.

During the procedure, carbon dioxide insufflation would likely be required to replicate laparoscopic insufflation conditions. Intraperitoneal pressures would require monitoring, either through an endoscope-based system or via a transabdominal Veress needle connected to a laparoscopic insufflator.

Precautions

The concern regarding infection is paramount in NOTES. More recent studies in patients have shown that intravenous antibiotics are sufficient in preventing peritonitis after transgastric peritoneoscopy.^42,43 However, the mean intraperitoneal time in these studies was less than 21 minutes. Transmural breach involving longer durations, organs with greater microbial load (e.g., transcolonic or even transesophageal), and prosthetic implantation may require additional mucosal antiseptics, such as chlorhexidine or povidone-iodine, which have been shown to reduce mucosal microbial burden to near zero when used in conjunction with intravenous antibiotics.⁴⁴

Conclusion

The arrival of NOTES allowed therapeutic endoscopy to move beyond the GI lumen. Breaching the gut wall—previously a forbidden act to the endoscopist and a reflexive, urgent call to the surgeon—now opens doors to intraperitoneal endoscopic procedures. It became conceivable and reasonable that gastroenterologists could be outside the GI tract. Looking ahead, the following NOTES surgeries have potential as procedures performed by interventional gastroenterologists: (1) EFTR, (2) diagnostic peritoneoscopy with or without lymph node resection, (3) directed liver biopsies, (4) bariatric procedures, and (5) anastomosis creation. These procedures share the following characteristics: (1) surgical sites more easily reached using a flexible endoscope or flexible NOTES platform compared with a laparoscope or via open access, (2) surgical procedures of low to medium complexity, and (3) probable improved outcomes from adjunct use of EUS for access or directed biopsies.

From a technical perspective, the advancement of NOTES depends on technical developments in the areas of viscerotomy closure and multitasking flexible platforms. From a training perspective, gastroenterologists are likely to need to undergo formalized training incorporating the fundamentals of extraluminal anatomy, dissection techniques, and management of complications. Infection control is of paramount importance. Still in its infancy, human NOTES research needs to be performed under oversight of institutional review boards to allow for data collection and safety evaluation in a rigorous manner.