Can mechanical balloon dissection be applied to cleave fibrotic submucosal tissues? A pilot study in a porcine model.

BACKGROUND AND STUDY AIMS: Removal of a lesion containing an ulcer scar is one of the most challenging applications of endoscopic submucosal dissection (ESD). The present study examined whether a novel balloon dissector could cleave fibrotic submucosal tissue beneath ulcer scars. METHODS: Six pigs were studied. Endoscopic mucosal resection (EMR) with ligation was performed at 7 or 8 sites in the stomach for each animal; 4 weeks later, 23 sites with a visible scar were selected for submucosal dissection. The procedure involved first creating a submucosal fluid cushion (SFC) by injecting either saline mixed with mesna or pure saline. A slender, compliant balloon with a diameter of 8, 13, or 18 mm was inserted into the SFC. The balloon was unfolded and thrust forward to cleave the fibrotic submucosa over approximately 5 cm. RESULTS: Fibrotic submucosa was dissected within 90 seconds in 17 of 23 attempts. Isolating the ulcer scar from the muscularis with the SFC prior to balloon dissection and using a thinner balloon catheter both ensured a better dissection. CONCLUSIONS: The fibrotic submucosa underlying post-EMR scars can be dissected with the novel balloon dissector, although the technique is less effective in cases with no sign of lifting.

Assessing the invasiveness of NOTES perforated viscus repair: a comparative study of NOTES and laparoscopy.

BACKGROUND: Natural orifice transluminal endoscopic surgery (NOTES) repair of perforated peptic ulcers may decrease surgical invasiveness and improve patient outcomes. METHODS: Full thickness gastrotomy was created laparoscopically in swine followed by soilage time. Repair proceeded with a laparoscopic (n = 14) or the NOTES (n = 14) approach. For NOTES repair, the omentum was endoscopically pulled into the gastric lumen and clipped. Intraoperative and postoperative parameters were recorded, including arterial blood gas (ABG) analysis and serum samples for white blood cell (WBC), TNF-α, IL-1, and IL-6 analysis. RESULTS: Twenty-four of 28 animals thrived to study completion. NOTES repair could not be accomplished in one animal. At necropsy, all repairs were intact. Blood pressure was equivalent between groups. Pulse examined during the last 30 min of each procedure revealed a slightly higher mean pulse in the animals undergoing NOTES procedures (NOTES, 102 ± 28; laparoscopy, 83 ± 24). ABG obtained at the conclusion of the procedure revealed a pH of 7.47 in NOTES animals and 7.43 in the laparoscopy animals (p = 0.06), a change from baseline in both groups. The final pCO(2) was lower in the NOTES group (NOTES, 40.62; laparoscopy, 47.49, p = 0.03). WBC counts were comparable on postoperative day (POD) 1 (NOTES, 21.1; laparoscopy, 19.0; p = 0.49). Mean TNF-α serum levels were equivalent at all time points between groups; however, TNF-α varied significantly from baseline to POD 7 (p = 0.002). CONCLUSION: NOTES omental repair appears comparable to that of laparoscopy. The lower arterial pCO(2) at the conclusion of the NOTES procedure may be advantageous in critically ill patients.

Outcome of emergency ERCP in the intensive care unit.

There are limited data on the outcome of emergency endoscopic retrograde cholangiopancreatography (ERCP) performed in the intensive care unit (ICU). We sought to assess the frequency, indications, and clinical outcomes of ERCPs performed in ICU patients who were too unstable to be transported to the endoscopy unit. An electronic endoscopy database was used to identify the patients (n = 22) and to assess procedural success, complications, and mortality. The indications for ERCP included suspected biliary sepsis, suspected gallstone pancreatitis, and known choledocholithiasis with cholangitis. Biliary cannulation, which was attempted in all patients, was successful in 19 patients (86 %), and of these 18 (95 %) underwent a technically successful endoscopic therapy. There were no apparent endoscopic complications. Therefore, emergency bedside ERCP in ICU patients, which is primarily performed for the management of suspected biliary sepsis and gallstone pancreatitis, can achieve high technical success rates when performed by experienced endoscopists, although the 30-day mortality rate remains high due to multiorgan dysfunction.

The challenges with NOTES retroperitoneal access in humans.

BACKGROUND: Animal studies have supported natural orifice transluminal endoscopic surgery (NOTES) retroperitoneal access. NOTES also may offer unique retroperitoneal access in humans. OBJECTIVES: This study was designed to assess the feasibility of endoscopic transgastric and transrectal retroperitoneal access in a cadaver model using prone and supine positioning, and to compare NOTES retroperitoneal examination with endoscopic ultrasound. METHODS: Using a multidisciplinary team, this institutional review board-approved study evaluated transgastric and transrectal retroperitoneal examination in six cadavers (3 male, 3 female; body mass index range, 25-37 kg/m(2)). Endoscopic ultrasound retroperitoneal examination preceded NOTES access. Transgastric Access: Using a prototype dual channel endoscope, a needle knife gastrotomy was created on the preantral posterior gastric wall. Retroperitoneal examination specifically targeted the pancreas and surrounding structures with the cadaver supine and prone. Transrectal Access: Using the same endoscope, a posterior needle knife rectotomy distal to the upper valve of Houston provided extraluminal access. Retroperitoneal examination proceeded with the cadaver prone and supine. Open dissection followed procedure completion. RESULTS: Access into the retroperitoneum succeeded at all sites. Significant challenges locating identifiable landmarks were faced-mostly transrectal and improved transgastric prone. All cadavers, despite body mass index or sex, had significant retroperitoneal adipose tissue limiting the endoscopic view. CONCLUSIONS: Although porcine studies have highlighted successful NOTES retroperitoneal procedures, the abundant human retroperitoneal adipose tissue challenged the translation of porcine research to humans. Additionally, although access to the retroperitoneal space and dissection within this space were accomplished easily, the appearance of cadaveric tissue and lack of blood flow made confident landmark identification impossible. Further study should continue in this area and focus on confident landmark identification for directed dissection. In a cadaveric model, this would best be improved by pre-NOTES anatomic marking or active perfusion of vasculature along with consideration of direct entry into the retroperitoneum from a targeted intraperitoneal site in clinical patients.

Endoscopic retrograde cholangiopancreatography using a single-balloon enteroscope in patients with altered Roux-en-Y anatomy.

BACKGROUND AND STUDY AIMS: Endoscopic retrograde cholangiopancreatography (ERCP) is challenging to perform in patients with postsurgical gastrointestinal anatomy. We assessed the diagnostic and therapeutic success rates using single-balloon enteroscopy in patients with Roux-en-Y anastomosis. PATIENTS AND METHODS: Patients who underwent single-balloon ERCP between April 2008 and February 2010 were retrospectively identified using a computerized endoscopy database. Diagnostic success was defined as successful duct cannulation or securing the diagnosis, and therapeutic success was defined as the ability to successfully carry out endoscopic therapy. Complications of ERCP were defined according to standard criteria. RESULTS: A total of 50 patients (34-male, mean age 57 years, range 19 - 85 years) with Roux-en-Y anastomosis underwent ERCP using a single-balloon enteroscope on 56 occasions. Indications for ERCP were cholestasis, acute cholangitis, recurrent primary sclerosing cholangitis with strictures, and choledocholithiasis. Overall diagnostic success was achieved in 39 / 56 cases (70 %). Therapeutic success was achieved in 21/23 cases (91 %). In 16 cases therapeutic intervention was not required. Therapeutic interventions included balloon dilation of strictures (n = 14), retrieval of retained biliopancreatic stents (n = 5), biliary stone extraction (n = 2), insertion of biliopancreatic stents (n = 4), and biliary and pancreatic sphincterotomy (n = 5). No major complications occurred. Importantly, in 22 / 56 procedures (39 %) a prior attempt at ERCP failed using conventional colonoscopes; single-balloon ERCP was successful in 15 / 22 (68 %) of these cases. CONCLUSIONS: Single-balloon ERCP is feasible in patients with complex postsurgical Roux-en-Y anastomosis, allows diagnostic evaluation and therapeutic intervention in patients with pancreaticobiliary disease, and is a useful salvage technique in the majority of patients in whom ERCP using colonoscopies has failed.

Chemically assisted submucosal injection facilitates endoscopic submucosal dissection of gastric neoplasms.

BACKGROUND AND STUDY AIMS: A randomized in vivo animal study previously demonstrated that topical injection of mesna solution (sodium-2-mercaptoethanesulfonate) chemically softened submucosal connective tissues and facilitated mechanical dissection of the submucosal tissue plane. The present study evaluated the technical feasibility and safety of chemically assisted endoscopic submucosal dissection (CA-ESD) using mesna in 20 consecutive patients who underwent endoscopic excision of gastric neoplasm. MATERIALS AND METHODS: Following the margination of the lesion with a mucosal circumcision, 4 - 12 mL of 10 % mesna solution was injected into the submucosal layer. Mechanical submucosal dissection was then performed by bluntly cleaving the chemically treated submucosal layer with the tip of a cap-fitted gastroscope. The use of cautery was restricted to prophylactic hemostasis, dissection of the coagulated vessels and persistent submucosal tissues, and the final snare resection. Post-therapeutic ulceration repair and adverse events were followed up during a 1-week hospitalization and by repeat endoscopies at 1 day, 1 week, and 1 month after the procedure. RESULTS: Sixteen gastric cancers and four adenomas were treated in this study. The sampled tissue measured 38.25 +/- 14.53 mm, with an en bloc resection rate of 100 %. Mean operation time was 21.17 +/- 11.6 minutes. The time spent using cautery was limited to 26.1 % of the total submucosal dissection time. Ulcerations healed normally without complications. CONCLUSIONS: This preliminary study demonstrates that submucosal injection of mesna facilitates and expedites mechanical submucosal dissection. The major limitations in this study include the single-arm study design and a small patient population.

Functional and comparative evaluation of flexible monopolar endoscopic scissors.

BACKGROUND: The continued success of natural orifice translumenal endoscopic procedures requires reliable, accurate tissue dissection and suture cutting. This study aimed to evaluate a flexible endoscopic scissors prototype. METHODS: An acute study of two domestic swine was conducted. Laparoscopic access provided an overview and allowed comparison of the flexible endoscopic scissors with laparoscopic shears. The endoscopic cautery-compatible scissors consists of cutting blades with a flexible wire catheter assembly (outer diameter, 3.1 mm) for use within a therapeutic endoscopic channel. A dual-channel colonoscope was advanced through a gastrotomy created for peritoneal access. With the aide of a grasping forceps, multiple peritoneal biopsies were obtained using the endoscopic scissors. Cautery (Monopolar 18-30 W) then was attached to the scissors, and gallbladder dissection proceeded. The device then was used to perform a small bowel enterotomy. A timed comparison of this function with laparoscopic enterotomy was made. Finally, 3-0 Polyglactin 910 suture was cut using the endoscopic scissors. RESULTS: Peritoneal biopsies 2 cm(2) in size were obtained from multiple abdominal locations and endoscopic positions, including the retroflexed position. The scissors cut effectively and accurately. Cutting performance was enhanced with parallel countertraction provided by grasping forceps passed through the second endoscopic channel. The use of cautery with the scissors controlled small vessels (2-3 mm) and permitted partial dissection of the gallbladder from the hepatic bed. Small bowel enterotomy sufficiently large for stapler passage was created in 4 min and 54 s. Laparoscopically, this was completed in 1 min and 22 s. Suture was successfully cut at the first attempt in a controlled, reproducible fashion. CONCLUSIONS: Controlled tissue biopsy, dissection, enterotomy creation, and suture cutting can be performed with this endoscopic scissors. Endoscopic tissue dissection and enterotomy creation was completed effectively but less efficiently than with laparoscopy primarily due to parallel device use imposed by the dual-channel endoscope.

Risk of pancreatitis following endoscopically placed large-bore plastic biliary stents with and without biliary sphincterotomy for management of postoperative bile leaks.

BACKGROUND: Endoscopic transpapillary biliary stent placement is effective for closure of postoperative bile leaks. Large-bore stents (10 French) may transiently obstruct the adjacent pancreatic duct orifice causing acute pancreatitis. Endoscopic biliary sphincterotomy may reduce this risk, but it introduces separate risks of bleeding and perforation. The objective of this study was to compare complications after large-bore biliary stent placement (10 Fr) with and without sphincterotomy in patients with bile leaks. METHODS: The institutional endoscopy database was queried to identify patients who had undergone endoscopic retrograde cholangiopancreatogrpahy (ERCP) for bile leak between March 1996 and August 2006. Procedural reports were reviewed for evidence of biliary sphincterotomy, cholangiographic and pancreatographic findings, transpapillary stent placement, and procedural complications. Patients with prior biliary sphincterotomy, choledochoenteric anastomosis, placement of multiple biliary stents and expandable metal biliary stents, biliary stents smaller than 10 Fr, and patients in whom a stent was not placed were excluded. The chi-square test was used for categorical variables. Probability

Endoscopic transpapillary gallbladder drainage: 10-year single center experience.

AIM: A subset of patients with acute cholecystitis is severely ill and extremely high-risk to undergo cholecystectomy. Data on the use of endoscopic transpapillary gallbladder drainage (ETGBD) in the treatment of acute cholecystitis are limited. This article reviews the 10-year experience of ETGBD at Mayo Clinic and evaluated patient and procedure characteristics. METHODS: A retrospective review of the endoscopy database from 1998-2007 was performed to identify patients who had undergone ETGBD. Clinical information and procedure details were abstracted from the electronic medical record. RESULTS: Fifty one patients underwent ETGBD for acute cholecystitis between 1998 to July 2007. The mean age was 62+/-19 years and 67% of patients were males. The median number of comorbid medical conditions was two (range 0-5) and 27% had underlying diabetes mellitus. Acute calculous cholecystitis was the predominant indication for ETGBD (78%). A gallbladder stent was used in 33 (65%) patients, nasocholecystic drain in 14 (27%) patients, and both in four patients (8%). Bleeding (4%) and sedation-related complications (4%) were the most common complications noted. Among patients who underwent cholecystectomy, the majority (76%) needed an open procedure. The median time to cholecystectomy was 15 days (range 1-352 days). Four patients (8%) succumbed to septic shock during their hospitalization. CONCLUSIONS: ETGBD is a valuable alternative therapeutic modality for the treatment of patients with acute cholecystitis who are at high-risk for early cholecystectomy, and/or those who have contraindications to percutaneous gallbladder drainage.

Submucosal endoscopic esophageal myotomy: a novel experimental approach for the treatment of achalasia.

BACKGROUND AND STUDY AIMS: The most permanent method of treating achalasia is a surgical myotomy. Because of the requirement for a mucosal incision and the risk of perforation, this procedure has not generally been approached endoscopically. We hypothesized that we could perform a safe and robust myotomy by working in the submucosal space, accessed from the esophageal lumen. MATERIALS AND METHODS: Four pigs were used for this experiment. Baseline lower esophageal sphincter (LES) pressures were recorded and the pigs underwent upper endoscopy using a standard endoscope. A submucosal saline lift was created approximately 5 cm above the LES and a small nick was made in the mucosa in order to facilitate the introduction of a dilating balloon. After dilation, the scope was introduced over the balloon into the submucosal space and advanced toward the now visible fibers of the LES. The circular layer of muscle was then cleanly incised using an electrocautery knife in a distal-to-proximal fashion, without complications. The scope was then withdrawn back into the lumen and the mucosal defect was closed with endoscopically applied clips. The entire procedure took less than 15 minutes. Manometry was repeated on day 5 after the procedure and the animals were euthanized on day 7. RESULTS: LES pressures fell significantly from an average of 16.4 mm Hg to an average of 6.7 mm Hg after the myotomy. The necropsy examinations revealed no evidence of mediastinitis or peritonitis. CONCLUSIONS: Endoscopic submucosal esophageal myotomy is feasible, safe, and effective in the short term. It has the potential for being useful in patients with achalasia. The submucosal space is a novel and potentially important field of operation for endoscopic procedures.

Peroral transgastric endoscopic primary repair of a ventral hernia in a porcine model.

BACKGROUND AND STUDY AIMS: Multiple studies have demonstrated the feasibility of peroral transgastric endoscopic procedures in animal models. The aim of the study was to evaluate the feasibility of a peroral transgastric endoscopic approach to repair abdominal wall hernias. PATIENTS AND METHODS: We performed acute experiments under general anesthesia with endotracheal intubation using 50-kg pigs. Following peroral intubation an incision of the gastric wall was made and the endoscope was advanced into the peritoneal cavity. An internal anterior abdominal wall incision was performed with a needle knife to create an animal model of a ventral hernia. After hernia creation an endoscopic suturing device was used for primary repair of the hernia. After completion of the hernia repair the endoscope was withdrawn into the stomach and the gastric wall incision was closed with endoscopic clips. Then the animals were killed for necropsy. RESULTS: Two acute experiments were performed. Incision of the gastric wall was easily achieved with a needle knife and a pull-type sphincterotome. A large (3 x 2 cm) defect of the abdominal wall (ventral hernia model) was closed with five or six sutures using the endoscopic suturing device. Postmortem examination revealed complete closure of the hernia without any complications. CONCLUSIONS: Transgastric endoscopic primary repair of ventral hernias in a porcine model is feasible and may be technically simpler than laparoscopic surgery.

Preliminary pneumoperitoneum facilitates transgastric access into the peritoneal cavity for natural orifice transluminal endoscopic surgery: a pilot study in a live porcine model.

BACKGROUND AND STUDY AIMS: Safe entrance into the peritoneal cavity through the gastric wall is paramount for the successful clinical introduction of natural orifice transluminal endoscopic surgery (NOTES). The aim of the study was to develop alternative safe transgastric access to the peritoneal cavity. PATIENTS AND METHODS: We performed 11 survival experiments on 50-kg pigs. In sterile conditions, the abdominal wall was punctured with a Veress needle. The peritoneal cavity was insufflated with 2 L carbon dioxide (CO (2)). A sterile endoscope was introduced into the stomach through a sterile overtube; the gastric wall was punctured with a needle-knife; after balloon dilation of the puncture site, the endoscope was advanced into the peritoneal cavity. Peritoneoscopy with biopsies from abdominal wall, liver and omentum, was performed. The endoscope was withdrawn into the stomach. The animals were kept alive for 2 weeks and repeat endoscopy was followed by necropsy. RESULTS: The pneumoperitoneum, easily created with the Veress needle, lifted the abdominal wall and made a CO (2)-filled space between the stomach and adjacent organs, facilitating gastric wall puncture and advancement of the endoscope into the peritoneal cavity. There were no hemodynamic changes or immediate or delayed complications related to pneumoperitoneum, transgastric access, or intraperitoneal manipulations. Follow-up endoscopy and necropsy revealed no problems or complications inside the stomach or peritoneal cavity. CONCLUSIONS: Creation of a preliminary pneumoperitoneum with a Veress needle facilitates gastric wall puncture and entrance into the peritoneal cavity without injury to adjacent organs, and can improve the safety of NOTES.

A new stapler-based full-thickness transgastric access closure: results from an animal pilot trial.

BACKGROUND AND STUDY AIMS: Reliable closure of the transluminal incision is the crucial step for natural orifice transluminal endoscopic surgery (NOTES) procedures. The aim of this study was to evaluate the feasibility and effectiveness of transgastric access closure with a flexible stapling device in a porcine survival model. PATIENTS AND METHODS: We carried out four experiments (two sterile and two nonsterile) on 50 kg pigs. The endoscope was passed through a gastrotomy made with a needle knife and an 18-mm controlled radial expansion dilating balloon. After peritoneoscopy, a flexible linear stapling device (NOLC60, Power Medical Interventions, Langhorne, Pennsylvania, USA) was perorally advanced over a guide wire into the stomach, positioned under endoscopic guidance, and opened to include the site of gastrotomy between its two arms; four rows of staples were fired. One animal was sacrificed 24 hours after the procedure (progression of pre-existing pneumonia). The remaining animals were survived for 1 week and then underwent repeat endoscopy and postmortem examination. RESULTS: Peroral delivery and positioning of the stapling device involved some technical difficulties, mostly due to the short length (60 cm) of the stapling device. The stapler provided complete leak-resistant gastric closure in all pigs. None of the surviving animals had any clinical signs of infection. Necropsy demonstrated an intact staple line with full-thickness healing of the gastrotomy in all animals. Histologic examination confirmed healing, but also revealed intramural micro-abscesses within the gastric wall after nonsterile procedure. CONCLUSIONS: Gastrotomy closure with a perorally delivered flexible stapling device created a leak-resistant transmural line of staples followed by full-thickness healing of the gastric wall incision. Increasing the length of the instrument and adding device articulation will further facilitate its use for NOTES procedures.

Pancreatic stent prophylaxis of post endoscopic retrograde cholangiopancreatography pancreatitis: spontaneous migration rates and clinical outcomes.

AIM: Pancreatic duct (PD) stents diminish the risk of post endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) in high-risk patients; 3 Fr stents are reported to spontaneously migrate at a significantly higher rate than 5 Fr stents in a cohort of mostly sphincter of Oddi (SOD) patients. We sought to assess spontaneous migration rates of 5 Fr and 7 Fr stents and effectiveness in preventing PEP in a diverse group of high risk patients. METHODS: A total of 4,332 ERCP exams performed between January 2002 and August 2005 were reviewed to identify patients undergoing PD stent placement. Follow-up was obtained from electronic medical records and contact with referring MDs. Plain abdominal radiographs were used to document stent passage. RESULTS: PD stents for PEP prophylaxis were placed in 246 exams (232 patients) undergoing: PD (major or minor) sphincterotomy (84), ampullectomy (50), SOD (46), bile duct precut (35), papillary stenosis balloon dilation (9) and difficult cannulation (8). Stents placed: 218 5-Fr (140 were 3 cm long and 78 =or> 5 cm long) and 28 7-Fr (12 were 3 cm long, 16=or> 5cm long). Follow-up was available in 197 (171 5-Fr, 26 7-Fr) of 246 placements (80%). Twenty of 171 5-Fr stents were electively removed via EGD within=or< 24 h per endoscopist preference and were not included in analysis; 128 of the remaining 151 5-Fr stents (85%) spontaneously migrated by (or within) median of 8 days and 23 failed to pass and required EGD removal. Of 26 7-Fr stents one was electively removed =or< 24 h later; of the remaining 25, 15 (60%) spontaneously migrated by median of 16 days, 10 required EGD removal. The spontaneous migration rate of 5 Fr stents was: 1) significantly higher than 7 Fr stents; 2) significantly higher than the previously reported 67% passage rate of 5 Fr stents; and 3) similar to the previously reported 86% passage rate of 3 Fr stents. PEP occurred in 15% (n=36: 24 mild, 11 moderate, 1 severe). CONCLUSION: The spontaneous dislodgement rate of 5 Fr stents in patients where the indication is primarily non-SOD is approximately 85% - significantly higher than previously reported and similar to the reported rate of spontaneous dislodgement of 3 Fr stents in SOD patients; 5 Fr stents migrate spontaneously earlier and more frequently than 7 Fr stents.

Transgastric endoscopic splenectomy: is it possible?

BACKGROUND: We have previously reported the feasibility of diagnostic and therapeutic peritoneoscopy including liver biopsy, gastrojejunostomy, and tubal ligation by an oral transgastric approach. We present results of per-oral transgastric splenectomy in a porcine model. The goal of this study was to determine the technical feasibility of per-oral transgastric splenectomy using a flexible endoscope. METHODS: We performed acute experiments on 50-kg pigs. All animals were fed liquids for 3 days prior to procedure. The procedures were performed under general anesthesia with endotracheal intubation. The flexible endoscope was passed per orally into the stomach and puncture of the gastric wall was performed with a needle knife. The puncture was extended to create a 1.5-cm incision using a pull-type sphincterotome, and a double-channel endoscope was advanced into the peritoneal cavity. The peritoneal cavity was insufflated with air through the endoscope. The spleen was visualized. The splenic vessels were ligated with endoscopic loops and clips, and then mesentery was dissected using electrocautery. RESULTS: Endoscopic splenectomy was performed on six pigs. There were no complications during gastric incision and entrance into the peritoneal cavity. Visualization of the spleen and other intraperitoneal organs was very good. Ligation of the splenic vessels and mobilization of the spleen were achieved using commercially available devices and endoscopic accessories. CONCLUSIONS: Transgastric endoscopic splenectomy in a porcine model appears technically feasible. Additional long-term survival experiments are planned.

Utility of endoscopic therapy in the management of Boerhaave syndrome.

Background/aims: The optimal intervention for Boerhaave perforation has not been determined. Options include surgical repair with/without a pedicled muscle flap, T tube placement, esophageal resection or diversion, or an endoscopic approach. All management strategies require adequate drainage and nutritional support. Our aim was to evaluate outcomes following Boerhaave perforation treated with surgery, endoscopic therapy, or both. Patients and methods: We performed a 10-year review of our prospectively maintained databases of adult patients with Boerhaave perforations. We documented clinical presentation, extent of injury, primary intervention, "salvage" treatment (any treatment for persistent leak), and outcome. Results were analyzed using the Fisher's exact and Kruskal - Wallis tests. Results: Between October 2004 and October 2014, 235 patients presented with esophageal leak/fistula with 17 Boerhaave perforations. Median age was 68 years. Median length of perforation was 1.25 cm (range 0.8 - 5 cm). Four patients presented with systemic sepsis (two treated with palliative stent and two surgically). Primary endotherapy was performed for eight (50 %) and primary surgery for eight (50 %) patients. Two endotherapy patients required multiple stents. Median stent duration was 61 days (range 56 - 76). "Salvage" intervention was required in 2/8 (25 %) endotherapy patients and 1/8 (13 %) surgery patient (stent). All patients healed without resection/reconstruction. There were no deaths in the surgically treated group and two in the endotherapy group (stented with palliative intent due to poor systemic condition). Readmission within 30 days occurred in 3/6 of alive endotherapy patients (50 %) and 0/8 surgery patients. Re-intervention within 30 days was required for one endotherapy patient. Conclusion: Endoscopic repair of Boerhaave perforations can be useful in carefully selected patients without evidence of systemic sepsis. Endoscopic therapy such as stenting is particularly valuable as a "salvage" intervention. The benefits of endoscopic therapy and esophageal preservation are offset against an increased risk of readmission in patients primarily treated endoscopically.

Acute gastrointestinal bleeding--a common problem revisited.

Acute gastrointestinal bleeding, a common problem, becomes a serious situation in the 15% of afflicted patients in whom the bleeding does not stop spontaneously. The typical high-risk patient is elderly, has multisystem disease, has a hemorrhagic episode during hospitalization for other disorders, or is an outpatient with severe bleeding of short duration and symptoms of volume depletion. Endoscopy can be performed to identify the active site of hemorrhage and is associated with an overall diagnostic accuracy that exceeds 90% in patients with upper gastrointestinal bleeding. The rate and character of observed bleeding have substantial predictive value for continued bleeding. The mortality among high-risk patients with bleeding in either the upper or the lower gastrointestinal tract ranges from 17 to 44%. Therefore, an organized aggressive approach to management of acute gastrointestinal bleeding is necessary.