Esophageal submucosal dissection under steady pressure automatically controlled endoscopy (SPACE): a randomized preclinical trial

• Introduction
• Methods
• The SPACE system
• Experiment 1
• Experiment 2
• Results
• Experiment 1
• Experiment 2
• Discussion
• Conclusions
• References

Background and study aims: A new overtube system has been developed for steady pressure automatically controlled endoscopy (SPACE) in the gastrointestinal tract. The objectives of this study were to validate the feasibility and safety of SPACE in the esophagus, and to evaluate its potential advantages over conventional (manually insufflating) endoscopy in endoscopic submucosal dissection (ESD).

Methods: This was a multicenter preclinical trial using acute porcine models (n = 20). In Experiment 1 (feasibility/safety study), SPACE was attempted in the esophagus with continuous monitoring of cardiopulmonary parameters and intraluminal pressures in the downstream bowel. Different insufflation pressures were tested to optimize the insufflation condition. Each session was video-recorded and scored by blinded reviewers. In Experiment 2 (randomized trial), esophageal ESD was attempted using either SPACE or conventional endoscopy, and results were compared.

Results: In Experiment 1, SPACE was performed safely without intraluminal pressure elevation in the downstream bowel. According to video review, SPACE provided more stable, reproducible, and rapid visualization than conventional endoscopy. The insufflation pressure was optimized at 14 mmHg for esophageal SPACE. In Experiment 2, ESD was completed in all animals. The ESD time was significantly shorter with SPACE compared with conventional endoscopy (1326 vs. 1616 seconds; P = 0.009). Responses to questionnaires showed that 94 % – 100 % of participants considered SPACE to provide improved exposure and more uniform tissue tension than conventional endoscopy. Other data were comparable.

Conclusions: SPACE is feasible, safe, and potentially effective for complicated endoscopic procedures, such as ESD. SPACE improves and standardizes endoscopic exposure and tissue tension. A clinical study is required to further confirm its safety and clinical effectiveness.

Introduction

Laparoscopic surgery requires carbon dioxide (CO₂) insufflation into the peritoneal cavity [1] [2]. This pneumoperitoneum is created using an automatic CO₂ insufflator with meticulous pressure monitoring because the peritoneal cavity is a closed system in which excessive insufflation may lead to serious consequences such as abdominal compartment syndrome [1] [2] [3]. The current computer-mediated, high-flow CO₂ insufflator automatically maintains a steady pressure environment so that an assisting surgeon can improve visualization/exposure with water irrigation, blood suction, and smoke evacuation in an automatically re-distended working space, while an operating surgeon performs bimanual manipulation using surgical energy devices.

In contrast, flexible gastrointestinal endoscopy is performed under on-demand insufflation by an endoscopist without pressure monitoring [1] [2] [3]. Although CO₂ has been increasingly used instead of atmospheric air, the gas is still supplied through the endoscope itself in a manual and blinded manner. This practice has been justified because the gastrointestinal tract allows migration of excessive gas into the upstream/downstream bowel [1]. Consequently, even a high-level team procedure, such as endoscopic submucosal dissection (ESD), is still a “one-person” process, with the operating endoscopist providing visualization, exposure, and instrument manipulation [4]. The role of an assisting endoscopist has therefore been passive and limited.

The introduction of natural orifice transluminal endoscopic surgery (NOTES) has resulted in the rethinking of current endoscopic techniques and instrumentations [4]. This “NOTES revolution” has accelerated the integration process of laparoscopy and flexible gastrointestinal endoscopy [5]. Although integration is required for every aspect, we have focused on insufflation, one of the most fundamental techniques for both modalities but totally different in each [1]. In the era of next-generation endoscopy, insufflation should theoretically be automatic so that the team can perform complex procedures with an evenly divided workload. Blind insufflation should be avoided because the intraluminal procedure will be extended beyond the gut wall [1] [2]. Automatic insufflation with pressure regulation, however, has not been in practical use in current flexible gastrointestinal endoscopy, mainly due to a lack of adequate instrumentation [1].

We have developed a new and simple flexible overtube system that provides a constant pressure environment in the upper gastrointestinal tract using currently available standard flexible gastrointestinal endoscopes and surgical insufflators. This new modality has been named “steady pressure automatically controlled endoscopy (SPACE).” The objectives of this study were to validate the feasibility and safety of SPACE in the esophagus, and to evaluate its potential effectiveness in performing esophageal ESD.

Methods

The SPACE system

The SPACE system consists of a standard commercially available endoscopic overtube (#16630; Top, Co., Ltd., Tokyo, Japan; diameter 19.5 mm, length 210 mm) and a newly developed detachable leak-proof device with an anti-reflux valve and a Luer lock connection (Leak Cutter, #16551; Top; [Fig. 1]). The overtube is introduced orally into the esophagus with the aid of an inner tube under endoscopic guidance. Once the system reaches the mid-portion of the esophagus, the inner tube is removed and Leak Cutter is attached to the proximal end of the overtube. A commercially available automatic surgical insufflator is then connected to the system, and esophageal SPACE is performed with automatic intraluminal CO₂ insufflation utilizing the space between the overtube and the endoscope ([Fig. 2]).

In total, 17 board-certified gastrointestinal surgeons and gastroenterologists from three centers participated in the study and all received formal training prior to the experiments. The following experiments were performed after protocol approval from the Institutional Animal Care and Use Committee.

Experiment 1

Experiments were first conducted to evaluate the feasibility and safety of SPACE in the porcine esophagus and to determine the insufflation pressure to be used in Experiment 2. A total of 10 35-kg female crossbred pigs were divided into two groups: a SPACE group (n = 5) and a manual insufflation (control) group (n = 5). The animals were fasted for 24 hours and premedicated with intramuscular 0.5 mg atropine sulfate. To sedate the animals during endotracheal intubation, a small dose of ketamine (10 mg/kg) with xylazine hydrochloride (2 mg/kg) was administered intramuscularly. The animals were anesthetized using 100 % oxygen and 5 % isoflurane with a flow rate of 1.5 – 2.0 L per minute. Ventilation was maintained by a mechanical ventilator (KV-1a; Kimura Medical, Co., Ltd., Tokyo, Japan) with 100 % oxygen and 2 % isoflurane inhalation under a fixed minute volume. An intravenous line was placed in the right auricular vein to hydrate the animal with a balanced electrolyte solution. CO₂ capnometry (TG-221T; Nihon Kohden, Tokyo, Japan) and portable pulse oximetry (TL-201T; Nihon Kohden) were used to document end-tidal CO₂ (EtCO₂) and percutaneous oxygen saturation (SpO₂) during each session. The left femoral artery was catheterized for blood gas analysis using a handheld blood analyzer (i-STAT; Abbott Point of Care Inc., Princeton, New Jersey, USA).

Midline minilaparotomy was performed as follows: three 2.4-mm diameter polypropylene spray catheters (#16512; Top) were placed in the stomach, duodenum, and jejunum (400 cm distal to the duodenojejunal junction), respectively. Each catheter tip was secured in the bowel wall using sutures, and the proximal ends were exteriorized by minilaparotomy. A digital manometer (MT110; Yokogawa, Tokyo, Japan) with a dedicated data logger (W32-MT210-R; System House Sunrise, Aichi, Japan) was connected to each catheter for continuous pressure monitoring of the downstream bowel during esophageal insufflation ([Fig. 3]). The abdomen was then temporarily closed with a sealing device (LapDisc; Hakko, Nagano, Japan), and a standard flexible gastrointestinal endoscope (EG4 – 50RD5; Fujifilm, Tokyo, Japan) was advanced with the overtube into the mid-portion of the esophagus. Another spray catheter was passed through the endoscope biopsy channel to obtain intraesophageal pressure measurements during esophageal insufflation. LapDisc was opened on completion of each measurement to verify the position and patency of each catheter and to allow manual evacuation of any residual and/or migrated CO₂ in the gastrointestinal tract in order to avoid confounding interaction with the next measurement.

In the SPACE group (n = 5), esophagoscopy was attempted for 20 minutes using the overtube, Leak Cutter, and a surgical insufflator (UHI-3; Olympus Medical Systems, Tokyo, Japan). Three different insufflation pressures were evaluated: 6, 10, and 14 mmHg. All experiments were conducted in a high-flow mode, which can feed 35 L of CO₂ per minute, according to manufacturer’s specifications. In the control group (n = 5), esophagoscopy was performed with manual insufflation through the endoscope using the overtube and a conventional endoscopic CO₂ feeder (GW-1; Fujifilm), which was controlled by an endoscopist who was blinded to the intraluminal pressure data. In place of Leak Cutter, a commercially available membrane-type adapter (#16630; Top) was used to prevent massive gas leakage from the proximal side of the overtube.

Four arms of the experiment were conducted (three SPACE arms and one control arm). Data were obtained and evaluated. A video review of each session was performed using a four-scale scoring system for steadiness of visualization, reproducibility of visualization after suction, degree of distension of the esophageal lumen, and the time to regain exposure after suction. In total, 10 board-certified gastrointestinal endoscopists and surgeons who were blinded to the insufflation settings performed this review. Intraluminal pressure changes in the esophagus, stomach, duodenum, and jejunum were measured every second during esophagoscopy. The mean blood pressure, heart rate, temperature, and EtCO₂ were measured before and every 5 minutes during esophagoscopy. Partial pressure of arterial O₂ (PaO₂) and partial pressure of arterial CO₂ (PaCO₂) were measured before and every 10 minutes during esophagoscopy. Finally, any adverse events were recorded.

Experiment 2

In Experiment 2, 10 additional animals were used to validate the feasibility and safety of esophageal ESD using SPACE, and to evaluate its potential superiority over conventional ESD. Under general anesthesia, two virtual lesions were created at 4 cm and 10 cm above the esophagogastric junction by placing eight burn spots in a 2-cm circle with an electrosurgical knife. Each lesion was then randomly assigned to either a SPACE group or a control group in a crossover fashion. Sufficient distance was maintained between the two lesions to prevent the initial treatment from affecting the subsequent one. An insufflation pressure of 14 mmHg was determined as optimal for SPACE ESD in Experiment 1; this pressure level was therefore used in Experiment 2.

Esophageal ESD was performed using a standard single-channel technique with a soft straight distal attachment (D-201 – 11804; Olympus) on the endoscope tip. An injection catheter with a 25-G needle (#01857; Top) was used for submucosal injection of normal saline with 0.5 % indigotindisulfonate sodium. For mucosal cutting and submucosal dissection, an endoscopic knife (DK2618JB-20; Fujifilm) powered by a high frequency electrosurgical unit (VIO 300D; Erbe Elektromedizin, Tübingen, Germany) was used. The fractionated cutting mode ENDO CUT I (effect 4, duration 4, interval 3) was used for cutting in combination with FORCED COAG (effect 2, 30 W) for dissection and SOFT COAG (effect 6, 80 W) for hemostasis. A hemostatic forceps (FD-410LR; Olympus) was used when indicated.

The following factors were prospectively recorded and compared between the SPACE and control groups: the total ESD time (from mucosal cutting to completion of dissection for both oral and anal lesions), energy device activation time (automatically recorded using dedicated software [VIO Chrono ver. 1.00; Erbe Elektromedizin]), number of forceps exchanges, specimen size, en bloc resection rate (counting numbers of marked spots included in the specimen), vital signs and arterial blood gas analysis, and any intraoperative adverse events. At the end of each experiment, a four-scale questionnaire was distributed to the study participants for qualitative comparison of SPACE and conventional endoscopy. The questions asked about reproducibility of endoscopic visualization, uniformity of mucosal tension during dissection, insufflation power, evacuation power, and endoscope handling.

All data were analyzed using a statistical software package (GraphPad Prism version 5; GraphPad Software Inc., San Diego, California, USA) on a universal personal computer and presented as median (range). Repeated measures analysis of variance (ANOVA) was used to compare parametric data between the SPACE and control arms in Experiment 1, except for the video-reviewing scores, to which the Wilcoxon rank-sum test was applied. Two-way ANOVA was also used to compare ESD data in Experiment 2. A P value of < 0.05 was considered statistically significant.

Results

Experiment 1

Esophagoscopy was successfully completed in all animals in both the SPACE and control arms. [Fig. 4] summarizes the results of the video review by the 10 blinded reviewers. Endoscopic visualization in the SPACE arms, irrespective of the insufflation pressure, was significantly steadier and more reproducible than that in the control arm. Distension of the esophageal lumen in the 14-mmHg SPACE arm was more prominent compared with that in the 10-mmHg SPACE arm (P = 0.005), 6-mmHg SPACE arm (P < 0.001), and control arm (P < 0.0001). Similarly, exposure after suction was regained significantly sooner in the 14-mmHg SPACE arm than in the 6-mmHg SPACE arm (P < 0.0001) and control arm (P < 0.0001). This rapidity of regaining exposure was even significant between the 10-mmHg SPACE and 6-mmHg SPACE arms (P = 0.005), and between the 6-mmHg SPACE and control arms (P < 0.0001).

[Fig. 5] shows the manometric profiles of the esophagus and the downstream bowel during SPACE and conventional esophagoscopy. Every pressure measurement was reproducible. As shown in [Fig. e6] (available online), intraesophageal pressures recorded by the digital manometer were well correlated with the pressures displayed on the UHI-3 front panel, suggesting the reliability of the endoluminal pressure measurement used in this experiment. The intraesophageal pressure fluctuated in the control arm, but remained stable around the preset pressure limit in the SPACE arms ([Fig. 5a]). In both groups, the intragastric pressure increased as the esophagus was insufflated, and both esophageal and gastric pressures gradually became equivalent in the latter half of the session ([Fig. 5b]). The intraduodenal pressure gradually increased in the SPACE arms; however, it remained lower than esophageal and gastric pressures. In contrast, the intraduodenal pressure in the control arm fluctuated and increased over 10 mmHg in 6 – 8 minutes ([Fig. 5c]). No increase in the intrajejunal pressure was observed in the SPACE arms; however, it increased over 10 mmHg in the control arm within 8 – 10 minutes ([Fig. 5 d]).

[Fig. 7] shows the cardiopulmonary changes during SPACE and conventional esophagoscopy. No significant difference was evident in the measured parameters among the four study arms. All the animals tolerated the sessions, and no adverse events occurred.

Given these results, an insufflation pressure of 14 mmHg was determined as optimal for SPACE in Experiment 2.

Experiment 2

Esophageal ESD was successfully completed in both the SPACE and control groups ([Table 1]). The ESD time in the SPACE group was significantly shorter compared with that in the control group (1326 vs. 1616 seconds; P  = 0.009). No difference in the energy device activation time was observed between the two groups. Other data such as the number of forceps exchanges, specimen size, and en bloc resection rate were comparable between the two groups. No adverse cardiopulmonary events or serious intraoperative complications were noted, except for a minor esophageal mucosal injury during advancement of the overtube into the esophagus ([Table 2]).

Questionnaire results (n = 17) are shown in [Fig. 8]. The response rate was 100 %. SPACE was rated as superior to conventional endoscopy in terms of endoscopic visualization and uniformity of esophageal wall distension by 94 % – 100 % of the participants ([Fig. 9]; [Video 1]). The majority (88 %) also felt that the insufflation power was adequate in SPACE, whereas 24 % claimed that the evacuation power was inadequate compared with that in manual endoscopy. Seven participants (41 %) claimed that endoscope handling was impaired because of Leak Cutter.

Discussion

The concept of gastrointestinal insufflation is changing due to the emergence of the next generation of endoscopy, including NOTES [1]. In NOTES, the gut wall is intentionally opened, resulting in rapid gas migration from the gut lumen into the peritoneal cavity. To prevent pressure fluctuation and excess pneumoperitoneum, pressure-regulated automatic endoscopic insufflation has been advocated [1] [2] [3]. However, automatic endoscopic insufflation using current flexible gastrointestinal endoscopes is not straightforward. No gas is supplied through the endoscope when connecting the current surgical insufflator to the insufflating channel of the endoscope [1]. Connecting the insufflator to one of the working channels of a double-channel endoscope does feed CO₂ [1]. This technique is applicable in NOTES cases, where no gas reflux is seen from the abdomen into the gastrointestinal lumen. However, this technique fails to establish a steady pressure environment in the gastrointestinal tract because gas from the surgical insufflator continuously leaks from the mouth. This leakage becomes more significant when the overtube is used for endoscopic intervention. An increasing need for pressure-regulated automatic endoscopic insufflation led to the development of the dedicated overtube system described here. The prototype was first evaluated in the peritoneal cavity to verify its compatibility with NOTES, following which it was tested in the gastrointestinal tract to validate the SPACE concept.

The clinical feasibility of steady-pressure pneumoviscera has already been confirmed in the stomach in a previous study [6]. In laparoscopic intragastric surgery (LIGS), the stomach is insufflated with a UHI-3 insufflation unit connected to a transgastric port at an intragastric pressure of 6 – 8 mmHg. In a study of 15 clinical cases using LIGS, a steady-pressure pneumogastrium was successfully created and maintained for 100 minutes on average without clamping the downstream bowel [6]. No adverse event was noted during LIGS, and no postoperative abdominal distension was observed. The present study demonstrated that a similar steady pressure environment is possible in the esophagus using a flexible endoscopic platform.

In SPACE, endoscopic visualization is automatically obtained once the insufflation pressure and flow rate are established. Exposure is identical to that in conventional ESD and is highly reproducible. Exposure after suction is also regained more quickly than in conventional endoscopy. The flow capacity of current surgical insufflators is higher than that of manual endoscopic insufflators and is considered responsible for the rapid regaining of exposure. For instance, UHI-3 can feed 23.8 L of CO₂ per minute in the high-flow mode and 9.6 L in the medium-flow mode [1]. These flow rates are significantly higher than those of actual endoscopic flow with manual CO₂ insufflation (1.4 L per minute) [1]. In the SPACE experiments reported here, the high-flow mode was used exclusively, for simplicity. In addition, the medium-flow mode might be effective in SPACE because no difference in performance was noted between the flow modes in our pilot study (data not shown). The performance of UHI-3 in any flow mode is considered sufficient to perform SPACE even in a narrow and hollow organ such as the esophagus.

The entire insufflation process is automatic in SPACE. Air/water button manipulation is therefore unnecessary, and the endoscopist is free to concentrate on the intervention itself. In our experimental setting, minimal endoscopic suction was required, resulting in only minor pressure fluctuation even in the control group. In clinical settings where endoscopic suction is applied more frequently, such as in cases of major bleeding, the advantages of SPACE would be even more apparent. In addition, from a safety point of view, SPACE is theoretically beneficial because excess pneumomediastinum and/or pneumothorax can be avoided in the event of iatrogenic esophageal perforation during the intervention.

The intrajejunal pressure was elevated in the control group but not in the SPACE arms. Lack of migration of CO₂ over the proximal jejunum remains unexplained. As shown in [Fig. e10] (available online), the insufflated gas volume was sufficiently low in each SPACE arm, suggesting no major gas migration into the downstream bowel during steady pressure insufflation. The rapid absorptive nature of CO₂ [7] was not responsible for this situation because no significant elevation of CO₂ was noted either in the systemic circulation or in expiration. One possible explanation is the “pinch-cock” phenomenon, which may also explain the clinical success with LIGS. In this phenomenon, the distended upstream bowel (stomach and duodenum) acts as a cock that compresses the downstream bowel, resulting in prevention of massive gas migration. This process may occur earlier and more persistently in SPACE than in conventional endoscopy with slow “on-and-off” insufflation. Another possible explanation is the surface tension in the collapsed gut lumen, which may work as a pressure barrier. Higher insufflation pressure than anticipated may be required to forcefully open the collapsed segment of the downstream bowel.

SPACE significantly reduced the total ESD time in the esophagus compared with conventional endoscopy. Because no difference was observed in the device activation time between the two groups, this time reduction was considered to be mainly because of the reduction in time to establish, re-establish, and maintain endoscopic visualization and working space.

The responses to the questionnaires indicated that the esophageal lumen was more uniformly distended in SPACE than in conventional endoscopy. Consequently, the tension of the esophageal mucosa was also more uniform, thereby facilitating precise and safe submucosal dissection. The rapid and automatic return of endoscopic exposure after suction also contributed to safer and more comfortable ESD because the endoscopist was free to focus on the intervention.

Despite its various advantages, the current system has several drawbacks. First, SPACE requires an overtube. Although no major adverse events were encountered in the present study, the use of an inner tube under adequate endoscopic guidance has always been considered mandatory. Additional care will be required when SPACE is performed in the gastrointestinal tract beyond the esophagus, where a longer overtube will be necessary for effective steady pressure insufflation. Second, as shown in the responses to the questionnaires, the use of Leak Cutter impaired flexible gastrointestinal endoscope handling. The size and rigidity of Leak Cutter should be adjusted to increase practicality. Third, the current surgical insufflators are designed for intraperitoneal use and not for gastrointestinal use. Consequently, as shown in [Fig. 8a] power imbalance was observed between insufflation and evacuation. The development of an automatic CO₂ insufflator optimized for gastrointestinal use is essential to make SPACE more practical and universal.

Conclusions

SPACE is feasible in the esophagus using a currently available flexible gastrointestinal endoscope, a surgical insufflator, and the newly developed overtube system. It is a safe and potentially effective technique for complicated, lengthy procedures such as ESD. Although its true effectiveness must be verified in future clinical trials, SPACE might become a key technology in the next generation of endoscopic interventions, including NOTES.

Competing interests: The authors have received research funding from Top, Co., Ltd. (Tokyo, Japan) and Fujifilm Corp. (Tokyo, Japan).

Acknowledgment

Part of this article was presented at Digestive Disease Week (7 – 10 May 2011; Chicago, Illinois, USA). The authors are grateful to Makoto Miki, Osamu Deguchi, Yuya Ito (Top Co., Ltd., Tokyo, Japan), Nobuyuki Torisawa (Fujifilm, Tokyo, Japan), and Nobuaki Tagi (Amco, Tokyo, Japan) for their technical assistance.

^* K. Nakajima and J. H. Moon contributed equally to this work.

• References

• 1 Nakajima K, Nishida T, Milsom JW et al. Current limitations in endoscopic CO₂ insufflation for NOTES: flow and pressure study. Gastrointest Endosc 2010; 72: 1036-42
  CrossrefPubMedSearch in Google Scholar
• 2 McGee MF, Rosen MJ, Marks J et al. A reliable method for monitoring intraabdominal pressure during natural orifice translumenal endoscopic surgery. Surg Endosc 2007; 21: 672-6
  CrossrefPubMedSearch in Google Scholar
• 3 von Delius S, Huber W, Feussner H et al. Effect of pneumoperitoneum on hemodynamics and inspiratory pressures during natural orifice transluminal endoscopic surgery (NOTES): an experimental, controlled study in an acute porcine model. Endoscopy 2007; 39: 854-61
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Swanstrom LL. NOTES: platform development for a paradigm shift in flexible endoscopy. Gastroenterology 2011; 140: 1150-4
  CrossrefPubMedSearch in Google Scholar
• 5 Rattner DW, Hawes R, Schwaitzberg S et al. The Second SAGES/ASGE White Paper on natural orifice transluminal endoscopic surgery: 5 years of progress. Surg Endosc 2011; 25: 2441-8
  CrossrefPubMedSearch in Google Scholar
• 6 Omori T, Nakajima K, Ohashi S et al. Laparoscopic intragastric surgery under carbon dioxide pneumostomach. J Laparoendosc Adv Surg Tech A 2008; 18: 47-51
  CrossrefPubMedSearch in Google Scholar
• 7 Yasumasa K, Nakajima K, Endo S et al. Carbon dioxide insufflation attenuates parietal blood flow obstruction in distended colon: potential advantages of carbon dioxide insufflated colonoscopy. Surg Endosc 2006; 20: 587-94
  CrossrefPubMedSearch in Google Scholar

Corresponding author

• References

• 1 Nakajima K, Nishida T, Milsom JW et al. Current limitations in endoscopic CO₂ insufflation for NOTES: flow and pressure study. Gastrointest Endosc 2010; 72: 1036-42
  CrossrefPubMedSearch in Google Scholar
• 2 McGee MF, Rosen MJ, Marks J et al. A reliable method for monitoring intraabdominal pressure during natural orifice translumenal endoscopic surgery. Surg Endosc 2007; 21: 672-6
  CrossrefPubMedSearch in Google Scholar
• 3 von Delius S, Huber W, Feussner H et al. Effect of pneumoperitoneum on hemodynamics and inspiratory pressures during natural orifice transluminal endoscopic surgery (NOTES): an experimental, controlled study in an acute porcine model. Endoscopy 2007; 39: 854-61
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Swanstrom LL. NOTES: platform development for a paradigm shift in flexible endoscopy. Gastroenterology 2011; 140: 1150-4
  CrossrefPubMedSearch in Google Scholar
• 5 Rattner DW, Hawes R, Schwaitzberg S et al. The Second SAGES/ASGE White Paper on natural orifice transluminal endoscopic surgery: 5 years of progress. Surg Endosc 2011; 25: 2441-8
  CrossrefPubMedSearch in Google Scholar
• 6 Omori T, Nakajima K, Ohashi S et al. Laparoscopic intragastric surgery under carbon dioxide pneumostomach. J Laparoendosc Adv Surg Tech A 2008; 18: 47-51
  CrossrefPubMedSearch in Google Scholar
• 7 Yasumasa K, Nakajima K, Endo S et al. Carbon dioxide insufflation attenuates parietal blood flow obstruction in distended colon: potential advantages of carbon dioxide insufflated colonoscopy. Surg Endosc 2006; 20: 587-94
  CrossrefPubMedSearch in Google Scholar

• Supplementary Material