Prosthetic mesh contamination during NOTES(®) transgastric hernia repair: a randomized controlled trial with swine explants.

PURPOSE: Natural Orifice Translumenal Endoscopic Surgery (NOTES(®)) is a developing field in minimally invasive surgery that has been applied across a wide range of procedures; however, infectious concerns remain. Most of the applications have been for extraction, rather than reconstructive procedures. Prosthetic hernia repair, is a constructive procedure, has the unique challenge of avoiding contamination and infection of a permanent implant. Utilizing a novel device, we hypothesize that we can significantly reduce or eliminate prosthetic contamination during a transgastric approach for delivery of a clinically relevant, permanent, synthetic prosthetic. METHODS: 20 swine explants of stomach with attached esophagus were prepared by placing an ultraviolet (UV) light sensitive gel within the lumen of the stomach. Each stomach then underwent endoscopic gastrotomy utilizing a needle, wire guide, and 18-mm balloon dilator. A 10 × 15 cm polypropylene prosthetic was rolled and tied with a 2-0 silk suture, and delivered with one of two methods. Group A (control) utilized a snare to grasp the prosthetic adjacent to the endoscope, which was used to drag it through the gastrotomy. Group B (device) utilized a modified esophageal stent delivery system to deliver the prosthetic through the gastrotomy. Each prosthetic was then digitally photographed with UV illumination, with the contaminated areas illuminating brightly. Software analysis was performed on the photographs to quantify areas of contamination for each group. Statistical analysis was performed using a two-tailed t test with unequal variance. RESULTS: Group A demonstrated a mean of 57 % of the surface area of the prosthetic contaminated with UV light sensitive gel. Group B (experimental group) showed a mean of 0.01 % of the surface area contaminated (p < 0.0001). 95 % confidence intervals indicated that the unprotected delivery technique exposes approximately 6,000 times more of the surface area to contamination than the delivery device. CONCLUSION: Use of this modified stent delivery system can nearly eliminate prosthetic contamination when placed via a transgastric approach in a swine explants model. Theoretically, the reduced inoculum size would reduce or eliminate clinical infection. Since the inoculum size required for clinical prosthetic infection for intraperitoneal mesh is unknown, further study is warranted to test the ability to eliminate clinical infection related to prosthetic delivery with this technique.

NOTES transgastric abdominal wall hernia repair in a porcine model.

INTRODUCTION: With approximately 1 million ventral and inguinal hernia repairs performed in the United States each year, even small rates of complications translate into large numbers of patients. Less invasive approaches that potentially lower morbidity deserve consideration, recognizing there are many technical considerations that currently limit their use. We describe a reproducible technique and lessons learned in our laboratory that answer some existing questions with regards to the use of NOTES for hernia repair. METHODS: A non-survival porcine model with general anesthesia was utilized in all cases. Each animal underwent transgastric peritoneal access with a percutaneous endoscopic gastrostomy (PEG) technique, and the gastrotomy was dilated with a wire-guided balloon dilatation catheter. An Esophageal Z-stent delivery device (Cook Medical, Winston-Salem, NC) was modified ex-vivo to allow us to introduce and protect a 10 x 15 cm lightweight polypropylene hernia prosthetic with pre-placed sutures. Once deployed, the sutures were pulled through the abdominal wall using a looped spinal needle technique in combination with the flexible endoscope. After the four anchoring sutures were tied, proprietary endoscopically placed tacks (Cook Medical) were placed at regular intervals between the sutures to secure the edges of the prosthetic. RESULTS: Hernia repairs were performed on five animals. In each case, we successfully completed prosthetic delivery and deployment into the peritoneal cavity, anchoring to the abdominal wall with full-thickness abdominal wall sutures, and endoscopically placed nitinol tacks. All prosthetics were deployed flat against the anterior abdominal wall. Operative times ranged from 65 to 120 min. CONCLUSION: Transgastric abdominal wall hernia repair is feasible, consistent, and reproducible. In particular, the delivery system can successfully deliver the prosthetic across the gastric wall via a transoral route. Survival animal experiments investigating outcomes related to quality of repair, microbiology, adhesions, and visceral closure need to be done. Human studies are not recommended until these issues are formally investigated.

Natural orifice transluminal endoscopic surgery gastrotomy closure in porcine explants with the Padlock-G clip using the Lock-It system.

BACKGROUND AND STUDY AIMS: The success of transgastric surgery depends on reliable, secure closure of the gastrotomy. Few tests of the integrity of these closures have been published. This study aimed to determine whether a gastrotomy suitable for a NOTES procedure can be closed safely and effectively from within the stomach using a novel endoscopically placed device, the Padlock-G with the Lock-It delivery system. METHODS: In a series of eight consecutive porcine gastric explants gastrotomy was performed in an ex vivo animal laboratory, the gastrotomy being closed with the Padlock-G followed by burst pressure testing after completion of the procedure. Gastrotomies were made in porcine explants. T-tags were placed on either side of the gastrotomy, and, with the T-tags pulled into an endoscopic cap, the Padlock-G was deployed. Gastric transmural pressure gradients at bursting of these closures were measured during insufflation of the explanted stomachs with a high-pressure insufflator. RESULTS: The mean burst pressure of the gastrotomy closures was 68.0 mm Hg (range: 45 - 107 mm Hg). All of the stomachs ultimately ruptured at the closure sites, with the exception of the stomach that ruptured at the highest value (107 mm Hg), which ruptured at a site approximately 5 cm away from the closure site. All of the closures were accomplished in 30 minutes or less. CONCLUSIONS: The Padlock-G clip provides a secure gastric closure for natural-orifice surgery.

Endoscopic management of tumors of the major duodenal papilla: Refined techniques to improve outcome and avoid complications.

BACKGROUND: Adenomas of the major duodenal papilla have malignant potential and are traditionally treated by pancreaticoduodenectomy. This is a report of our experience with endoscopic management and a description of techniques for decreasing complications and enhancing efficacy. METHODS: Forty-one patients were referred for endoscopic management of papillary tumors. If there was no duct invasion and the appearance suggested a benign lesion, biductal sphincterotomy with pancreatic duct stent placement was performed. If the lesion could be elevated by injection of an epinephrine solution, piecemeal resection was performed. The base of the lesion was thermally ablated as needed. Resection/ablation together with stent removal was performed 1 month later. RESULTS: Nine patients (22%) had lesions other than papillary adenoma or cancer. Malignant appearance, ductal stricturing, or extension into the ducts was found in 16 of 41 patients (39%) in whom biopsy specimens alone were obtained. Three patients with adenomas (7%) did not undergo endoscopic resection (because of extremely large lesions and/or comorbid illnesses). Thirteen patients with adenomas (32%) had endoscopic resection; 12 (92%) were lesion-free after 32 ERCPs (mean 2.7). Endoscopic management was unsuccessful in 1 patient (8%). Pancreatitis developed in 1 patient. CONCLUSIONS: Endoscopically treatable papillary neoplasms can be identified on the basis of endoscopic, radiographic, and biopsy features. Preresection sphincterotomy, stent placement, elevation by epinephrine injection, and piecemeal resection may reduce complications and permit more aggressive treatment.

Cholestatic jaundice associated with the use of metformin.

We report a patient who developed cholestatic jaundice shortly after initiation of treatment with metformin hydrochloride. Ultrasound of the liver and abdominal CT were normal. An ERCP showed normal biliary anatomy. A percutaneous liver biopsy was obtained showing marked cholestasis, with portal edema, ductular proliferation, and acute inflammation. Metformin hydrochloride was discontinued, and the patient's jaundice resolved slowly over a period of several months. Given the onset of his jaundice 2 wk after the initiation of metformin, we believe that this case represents an example of metformin-associated hepatotoxicity, the first such case reported.

Lectin binding to human colonocytes is predictive of colonic neoplasia.

OBJECTIVE: The aim of this study was to determine whether lectin binding to exfoliated human colonocytes could be used as a noninvasive test for colorectal polyps or cancer. METHODS: Colonocytes were harvested from 31 patients (10 controls, 10 with adenomatous polyps, and 11 with cancer), incubated with a panel of fluorescent-labeled lectins, and assayed by flow cytometry. RESULTS: The lectins jacalin (JAC) and wheat germ agglutinin (WGA) were useful in predicting the presence of a colorectal neoplasm (p = 0.0018 for JAC and p = 0.0099 for WGA). For JAC, sensitivity reached 81% with a specificity of 80%, and for WGA the sensitivity and specificity were both 75%. CONCLUSIONS: Lectin binding to human colonocytes can predict the presence of malignant and premalignant lesions of the colon, and has potential as a noninvasive screening tool for colorectal neoplasms.

Malignant jugulotympanic paraganglioma.

Jugulotympanic paragangliomas are neoplasms that arise in the jugular bulb (glomus jugulare tumor) or temporal bone (glomus tympanicum tumor). They are histologically similar to carotid body tumors; both are derived from extra-adrenal paraganglia. Of only 20 metastases reported, five had biopsy-proved bone involvement. We present a review of the literature of metastatic jugulotympanic paragangliomas and report a case that recurred locally with regional metastases 10 years after initial presentation and metastasized to spine 13 years later. Histologic data are reviewed, and potential treatment regimens are contrasted.