BLUE rubber bleb nevus syndrome, a rare vascular malformation involving the gastrointestinal tract, may result in clinically significant acute and chronic bleeding. The bleeding is generally chronic and unremitting, with continual melena necessitating transfusion or surgical management. Definitive surgical management is frustrating because the lesions are usually multiple (often numbering in the hundreds) and may be found anywhere in gut derivatives from the mouth to the anus, most often in the small intestine. The length of the instruments available and the risks of incomplete removal and perforation limit endoscopic treatment such as sclerotherapy or laser photocoagulation as an alternative. Combined approaches using transoral and transrectal endoscopic examination and endoscopic examination via laparotomy/enterotomy permit visualization of the entire gastrointestinal tract.1–6
An 11-yr-old, 29.6-kg girl with blue rubber bleb nevus syndrome underwent simultaneous upper and lower endoscopy and exploratory laparotomy for resection of multiple vascular lesions. She had experienced gastrointestinal bleeding 2 weeks before admission.
After intravenous premedication with midazolam, anesthesia was induced with propofol, fentanyl, and pancuronium. The trachea was intubated with a 6.0-mm cuffed orotracheal tube, and anesthesia was maintained with an air–oxygen mixture and 0.5–1.2% end-tidal isoflurane. Ventilation at a tidal volume of 320 ml and a rate of 9 breaths/min was controlled by a rising-bellows pneumatic anesthesia ventilator. Standard noninvasive monitors were used, and in addition, a radial artery line was placed. An infrared spectrometer was used (Datex Capnomac Ultima; Datex Medical Instrumentation, Tewksbury, MA), and end-tidal carbon dioxide and isoflurane concentrations were recorded.
Approximately 3 h into the procedure, after laparotomy, esophagogastroduodenoscopy, and transoral proximal enteroscopy with manual assistance to advance the telescope and while undergoing colonoscopy (Olympus PCF-160-AL video colonoscope; Olympus America Inc., Melville, NY), the patient had a decrease in oxygen saturation to 88–94%, a decrease in end-tidal carbon dioxide to 22 mmHg, and hypotension (51/31 mmHg; baseline, 93/43 mmHg). Simultaneously, there was an increase in the peak inspiratory pressure (33 cm H2O; baseline, 22 cm H2O) and mean airway pressure (17 cm H2O; baseline, 5 cm H2O). The small bowel was distended approximately 6 cm and attenuated. The insufflation was stopped, the bowel was allowed to decompress, the isoflurane and air in the inspired gas were discontinued, and the oxygen saturation improved to 99–100%. A bolus of lactated Ringer’s solution (19 ml/kg) was given. After the patient became hemodynamically stable, the surgeon examined the bowel and found small air bubbles with a “Rice Krispies®” appearance in the mesentery (fig. 1).
Gas delivery system dysfunction accounting for a sudden decrease in end-tidal carbon dioxide was quickly checked for and ruled out: There was no leak in the endotracheal tube cuff, the patient did not show evidence of spontaneous breathing either by physical examination or notching of the capnographic waveform, and there were no disconnects or ventilation system obstructions. No further treatment was required, and the case proceeded uneventfully to the conclusion of surgery, emergence, and the perioperative period.
Venous air embolism (VAE), now commonly reported with many surgical procedures, has rarely been reported with gastrointestinal endoscopy. This case offered a unique opportunity to examine the external surface of the bowel and mesentery during upper and lower gastrointestinal endoscopy. The most likely cause of the VAE was direct transmucosal to intravascular introduction of enteric air under high pressure, with additional entrapment in the mesentery.
The pathophysiology of VAE has been well described7; typical intraoperative signs include hypotension, a decrease in end-tidal carbon dioxide, and alterations in airway pressure. The biochemistry of these multisystem findings may result from the release of vasoactive mediators, complement, and cytokines.8–10However, early detection remains the basis for monitoring in higher risk cases. Although not usually employed during lower risk cases, transesophageal echocardiography, precordial Doppler ultrasonography, or a pulmonary artery catheter, in descending level of sensitivity, reveal evidence of VAE.11The pneumomesentery with crunchy “Rice Krispies®”-like pockets of air and the rapid resolution of the signs of VAE after decompression of the bowel and removal of the colonoscope serve to further support the intraoperative diagnosis of VAE.
The choice of insufflation gas has been a matter of considerable investigation. Almost all surgical endoscopy at this point is conducted with carbon dioxide as the insufflation gas, whereas the gastrointestinal endoscopy insufflation gas most commonly used continues to be air. Other gases less frequently used include nitrous oxide, helium, argon, nitrogen, and oxygen. Although nitrous oxide is somewhat less soluble than carbon dioxide (130 vs. 170 ml/100 ml water), it supports combustion. Even the use of nitrous oxide as part of the anesthetic technique is less frequently chosen because of its possible contribution to expanding any entrained or insufflated air.12The choice of carbon dioxide as an insufflation gas is based on its high aqueous solubility and lack of support of combustion. The magnitude of physiologic derangement attributable to carbon dioxide is 6.5 times less than that of air.13Neither helium nor argon support combustion, but argon has been associated with myocardial depression, and helium has been associated with subcutaneous emphysema.
This intraoperative endoscopic approach to the multifocal vascular malformations of blue rubber bleb nevus syndrome has rarely been described,3,14,15although in 10 previous cases at our institution by the same surgeon and gastroenterologist, VAE did not occur. However, this case raises an intriguing question. Because transoral and transrectal flexible endoscopy is so commonly performed and intraoperative enteroscopy is rarely used, it is likely that technical aspects of the latter technique contributed to the occurrence of VAE. Continuous insufflation through the endoscope while maintaining manual compression of the distal bowel optimizes visualization and the likelihood of identifying small lesions. However, this technique may increase the intraluminal pressure beyond that caused by standard closed-abdomen endoscopy. As collaborations in the operating room continue to develop between surgeons and procedurally trained medical colleagues, consideration should be given to how techniques developed outside the operating room should be modified, such as conversion to carbon dioxide rather than air for gastrointestinal insufflation. VAE is rare in current clinical practice,16,17but as practice and new collaborations evolve, previous practices should be a cause for reflection if not active, prospective study.