- Email: [email protected]
Endoscopic management of complications in digestive surgery
Journal of Visceral Surgery (2013) 150S, S3—S9
Available online at
www.sciencedirect.com
REVIEW
Endoscopic management of complications in digestive surgery X. Dray ∗, M. Camus , U. Chaput Université Paris Diderot, Paris 7, AP—HP, Service d’Hépatogastroentérologie, Hôpital Lariboisière, 2, rue Ambroise-Paré, 75010 Paris, France
KEYWORDS Therapeutic endoscopy; Stricture; Anastomotic fistula; Anastomotic leak; Postoperative complications
Summary Endoscopy has an ever-increasing role in the treatment of complications in digestive surgery. Endoscopic treatment is essentially used for (i) fistula or intra-abdominal collection secondary to anastomotic dehiscence and (ii) anastomotic stricture, especially esophagogastric, but also sometimes after colorectal surgery. First intention treatment of fistula following esophagogastric surgery is the insertion of an extractable self-expandable metallic stent (partially or entirely covered); this is supported by a low level of scientific evidence, but clinical experience has been satisfactory. Other techniques for treatment of anastomotic leak have also been reported anecdotally (clip placement, sealing with glue). There are few data available in the literature on endoscopic management (stents essentially) of postoperative colonic fistula. Whatever the approach chosen to treat a postoperative digestive tract fistula, management is medico-surgical and cannot be limited to simple closure of the digestive tube wall defect. Drainage of any collections by endoscopic, radiologic or surgical approach, systemic treatment of infection and nutritional support are essential adjuvant treatment modalities. Treatment of postoperative esophageal or colonic strictures is essentially endoscopic and is based on initial dilatation (endoscopic with hydrostatic balloon or bougie), and placement of extractable metallic stents (partially or entirely covered) in case of refractory outcome. © 2013 Elsevier Masson SAS. All rights reserved.
Introduction Endoscopy has an increasing role in the treatment of complications in digestive surgery. Endoscopic treatment is essentially used in case of: • fistula or intra-abdominal collection secondary to early or late (necrosis) anastomotic dehiscence, with fistula or intra-abdominal collection; • anastomotic stricture, especially esophagogastric, but also sometimes after colorectal or biliopancreatic surgery.
∗
Corresponding author. AP—HP, Service d’Hépatogastroentérologie, Pôle Digestif-Uro-Néphrologie-Gynécologie-Obstétrique, Hôpital Lariboisière, 2, rue Ambroise-Paré, 75010 Paris, France. Tel.: +33 1 49 95 25 45 or 78; fax: +33 1 49 95 25 77. E-mail address: [email protected] (X. Dray). 1878-7886/$ — see front matter © 2013 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.jviscsurg.2013.04.005
S4 The goal of this review is to expose the techniques and results of endoscopic interventional management and discuss the indications of endoscopic treatment of postoperative gastrointestinal tract complications. The treatment of anastomotic bleeding, rare and most often benign, calls for classical endoscopic hemostatic techniques and will not be dealt with here. Endoscopic management of biliopancreatic complications is discussed in another update of this thematic issue [1].
Endoscopic treatment of anastomotic fistula after esophagogastric surgery Therapeutic principles and level of evidence The anastomotic leak rate after esophagectomy is approximately 5% [2]. Mortality ranges from 20% to 25% when the leak is intrathoracic [2]. Progress has been made in endoscopic treatment giving it an ever-increasing role in temporary prosthetic management (removable stents) as a mini-invasive alternative to surgery in this setting when anastomotic dehiscences involve less than 70% of the circumference [3—6]. Other endoscopic treatments are occasionally used to close smaller caliber leaks (clips, glue,
X. Dray et al. endoscopic suture) and have a limited place in the treatment of postoperative fistula in the mediastinum (Fig. 1). Endoscopic treatment of anastomotic leak after esophagectomy should not be an isolated modality; it must be associated with drainage, broad-spectrum antibiotic therapy, nutritional support and surgery. The choice of medical treatment (including stent insertion) should be decided as early as possible, before massive mediastinal contamination or uncontrolled sepsis sets in. There are few data in the literature attesting to the efficacy of endoscopic treatment of esophageal fistula. Most publications are case reports (often selected on the basis of therapeutic success, and therefore biased); there are a few retrospective series with low numbers of patients, and more ideally, some prospective non-comparative studies, also with low numbers. A few principles emerge from these studies and can be applied to clinical practice, without waiting for controlled trials (which are unlikely to be performed) comparing endoscopic to surgical treatment (the indications are too different). Therefore, in general, non-surgical treatment, particularly endoscopic, should be of primary interest for: • contained (in the mediastinum and/or the neighboring pleura) fistulae and anastomotic dehiscences; • patients who can be managed early; • patients in good enough condition to undergo urgent surgery in case of unfavorable evolution.
Figure 1. Endoscopic treatment by fibrin glue, clips and covered stent for esophagopleural fistula after Lewis-Santy esophagectomy. Reduction of pleural drainage volume from 400 to 40 mL/d by 7 days after insertion of self-expandable completely covered metallic stent for anastomotic leakage. (A) Endoscopic view showing a migrated stent that no longer covers the perforation. After retrieval of the stent, a catheter is introduced into the primary orifice of the fistula (B) and shows contrast material in the pleural drain (C). Abrade the fistular tract with a biliary brushing catheter (D, E). After sealing the fistular tract (F, G), obstruction by glue of the primary orifice is visible (H). Placement of two clips on the primary orifice (I, J, K, L). Insertion of guide-wire (M, N) for a self-expanding completely covered metallic stent in the esophagus (O, P). Proximal release of the stent (Q, R). At the end of the procedure, the stent is in good position, without leakage of contrast material into the pleural space (S, T).
Endoscopic management of complications in digestive surgery
S5
Digestive tract endoprostheses or stents
Delivery technique
Material
Esophageal self-expanding stents are placed under endoscopic and radiologic guidance and usually require general anesthesia. In certain situations, nonetheless, they can be placed with only endoscopic or only radiologic guidance. The endoscopist must have at hand a full array of guide-wires, opacification catheters and stents. The initial step consists of visualizing and landmarking the fistula endoscopically and by contrast opacification, taking note of the exact location in order to best choose the length of the stent. Radio-opaque objects (e.g., paper clip or needle) are used for cutaneous landmarking if patient movements can be limited, or more rarely internal markings (submucosal injection of contrast or clip placement) are used. After placement of a guidewire the endoscope is withdrawn since the esophageal stent is too wide to be inserted through the endoscope-operating channel. The second step is the insertion and delivery of the stent, over the guide-wire. A rigid guide-wire (Savary type) is preferred. An endoscope can be inserted alongside the guide-wire for visualization. The distal end of the stent is placed in the stomach, gastroplasty or distal jejunum, according to the previous surgery. Small adjustments can be made at the beginning of the delivery thanks to a recapturing system within the sheath. Spontaneous expansion of the stent ensures immediate anchorage, but there is always a slight risk of early migration. Contact between the stent and the endoscope must be avoided immediately after the delivery (increased risk of migration). Certain authors feel that the proximal extremity of the stent should be fixed by hemostatic clips to the esophageal wall to prevent migration (however, the level of evidence for this manoeuvre is low) [15]. Opacification within the stent itself after stent deployment is performed to check that there is no extravasation of the contrast material.
Removable self-expanding stents are the material of choice in this setting. Historically, endoluminal treatment of esophageal fistula was developed using the only selfexpandable (plastic) stent available at that time (Polyflex® , Boston Scientific). A wide array of stents are currently available, with diameters ranging from 16 to 21 mm and lengths from 90 to 150 mm. The releasing system must be assembled by the operator. The stent is made of silicone, entirely covered with a polyester film and flared at the proximal end. Coating limits the risk of incarceration into the esophageal mucosa, and facilitates stent extraction, but enhances the risk of migration (approximately 30% in stricture and probably higher in fistula, vs. 3—12%, for metallic stents depending on whether they are partially or completely coated) [7—9]. Plastic stents are less expensive than the metallic models. The large diameter of these stents and their rigid character are risk factors for complications (perforation, hemorrhage): 9% vs. 3% for metallic stents [7]. Currently, removable self-expanding metal stents (partially or completely coated) are used in the treatment of complications of esophagogastric surgery. While their cost is clearly higher than the plastic stents, their utilization is easier (pre-assembled and more flexible), the risk of complications (perforation, hemorrhage, migration) is less elevated [10—12]. These stents are constituted of one or several braided strands of a metal with high shape memory (most often made of Nitinol). They are contained within a tight sheath placed on a carrier tube. The caliber of the assembled system ranges from 18 to 24 Fr, which is not compatible with the passage of the stent through the operating channel of an endoscope (over-the-wire, OTW, system). There are newer infrequently-used models with a thinner delivery diameter (10 Fr) that can be passed through a 3.7 mm operating channel (through-the-scope system, TTS). The entire delivery system (carrier tube, pleated stent and over catheter) is inserted over the rigid guide-wire, under radioscopic and/or endoscopic guidance, with the endoscope lying alongside the stent. The self-expandable metallic stents vary according to the alloy used, the type of latticework, the form, the length (70 to 170 cm and even longer), the diameter (18 to 23 mm or greater), and the delivery system as well as the coating material. The proximal, and sometime the distal, extremity are flared (approximately 5 mm larger than the shaft), in an attempt to limit the risk of migration. Depending on the model, the length of the delivered stents shrinks after expansion by 20% to 40% from the original length within the sheath. The most classic models are of the ‘‘distal delivery’’ type. Several options (anti-reflux valve, anti-migration systems) are available, but there has been no formal demonstration of their benefits at the present time. The choice of the stent is crucial in case of proximal esophageal fistula, especially after the Lewis-Santy operation. The low radial force of the Ultraflex® (Boston Scientific) stent in this indication is worthy of mention [13]. A stent with a short upper cuff (5 mm) was developed for this indication (M.I.Tech Co., Pyontack, Korea) [14]. Proximal delivery systems allow adjustment of the position of the stent between the high-riding fistula and upper esophageal sphincter. A proximal delivery stent with a short cuff has recently been commercialized (Taewoong Medical, Seoul, Korea).
Precautions necessary during the procedure Strict and prolonged fasting is necessary (with adequate nutritional support as needed) until formal and durable demonstration of the absence of perforation. This can be ensured by radiology (esophageal follow-through) or methylene blue. A thoracic X-ray on day 3 can help evaluate of the degree of expansion of the stent and the absence of early migration. Alimentation per mouth can be envisioned if the clinical evolution is satisfactory. Recommendations to avoid the risk of food impaction within the stent include: eating while in the seated position, avoiding thick, dense aliments (meat, for example), avoiding stringy aliments (leeks for example), avoiding karaya gum gastric demulcents, and drinking sparkling water at the end of the meal.
Extraction technique of esophageal stents Stent withdrawal can be envisioned 4 to 8 weeks after placement. Withdrawal can be done simply by pulling on the nylon string attached to the proximal end of the stent, either with a toothed forceps or polypectomy snare. Certain partially covered models (Ultraflex® ) can be withdrawn by traction on their distal extremity, which progressively turns the stent inside out. In more difficult cases, particularly extraction of partially covered metallic stents with epithelial ingrowth into the mesh, grasping the stent at two points via the two channels of a double-channel endoscope can be helpful. When the proximal extremity abuts the esophageal upper
S6 sphincter, placing an over-tube can be useful for extraction. The placement of a plastic stent within a partially covered metallic stent can promote pressure necrosis of the hyperplastic epithelium and thus facilitate retrieval as a two-step procedure [16].
Results of endoluminal stenting While early experience argues in favor of the use of selfexpandable plastic stents in this setting, there are no randomized trials or high-level evidence on the topic. Freeman et al. observed a successful outcome in 20 out of 21 patients despite a 25% migration rate [17]. Likewise, in 2007, Ott et al. evaluated the use of plastic stents (Polyflex® , Boston Scientific, USA) in a retrospective series of 35 patients, 12 of whom had an esophageal fistula. The results were considered very good in spite of the 37% migration rate [18]. Hunerbein et al. retrospectively compared the outcome of nine patients treated with a plastic expanding stent (PolyflexTM ) with that of 10 patients treated previously either by revisional surgery or with conservative measures. Stent placement resulted in immediate occlusion of the fistula in eight of nine patients and the stent could be removed at a mean interval of 29 days after insertion. By comparison with conventional treatment, patients were able to start eating earlier (11 days), and had shorter intensive care stay (25 days) and overall hospital stay (35 days), while mortality was nil (vs. 20% after conventional surgical or conservative treatment) [19]. The same team updated their experience with 22 patients; five cases of stent migration were treated endoscopically without compromising the final outcome. The patients were able to resume eating early, 4 days after the insertion of the stent [20]. There are less series with extractible self-expanding metallic stents. Most of the complications in these series concerned esophageal perforation [21]. In spite of a very limited evaluation, these stents now have a predominant place (more than plastic stents) in this indication.
Endoscopic clips Endoscopic clips were developed to ensure endoscopic hemostasis and not for sutures. There are different models, some that need to be assembled, others pre-assembled, repositionable or not, rotational or not. The spread of conventional clips in the open position (11 mm at the most) was designed for mucosa-to-mucosa approximation only. The endoscopic technique consists of approximation of the clips starting from one end of the opening to the other. This must be done with minimal carbon dioxide insufflation, to prevent enlargement of the opening and mediastinal contamination, as well as to avoid creating a tangential angle of the endoscope (and therefore of the clip) in relation to the esophageal wall. Hemostatic clips were initially proposed for early closure of acute perforations, before substantial contamination of the mediastinum had occurred. The success rate was 98% in a series of 117 gastric perforations after endoscopic submucosal dissection [22]. Treatment of fistula may require initial ablation of the re-epitheliized margins, for example by argon plasma coagulation. Certain authors have reported failures when the margins were not treated in this manner, but these failures could be transformed into successes if initial mucosal ablation was performed [23—25]. New generation macroclips (11 to 14 mm) (Over-the-scope clips, Ovesco, Tübingen, Germany) have been used with
X. Dray et al. success in a few isolated cases of fistulas after esophagogastric and/or esophago-jejunal anastomosis [26—28].
Endoscopic sealing Fibrin patches have been proposed to treat esophageal leaks [29]. This technique was originally designed to manage bleeding, particularly in the case of ruptured esogastric varices due to portal hypertension. A specific technical skill must be mastered, because the glue has to be injected through a sclerosing needle passed through the operating canal of the endoscope. Injection of a bolus of lipiodol, and especially, visual control help to avoid premature release of glue in the operating channel of the endoscope. Standardized endoscopic fibrin glue techniques have been proposed [30]. A few cases of endoscopic biological glue sealing of esophageal fistula and perforations have been published. Rabago et al. reported the efficacy of such a treatment in one case of acute esophageal perforation after endoscopic extraction of impacted food, and in two patients with postoperative esophageal leaks [31]. Fischer et al. completed clip closure of a large (40 mm) iatrogenic esophageal perforation (diagnostic endoscopy) with injection of glue [32]. In neither of these two cases, however, was it possible to deduce an added benefit of fibrin glue compared with management of perforations (conservative treatment in the first case, conservative treatment + clip closure in the second case).
Echoendoscopic and endoscopic drainage of mediastinal abscess Mediastinal abscesses are classically treated via a transthoracic approach. One team has reported their experience with echoendoscopic-guided transesophageal puncture and drainage [4]. Drainage begins with the introduction of a 19G needle under combined radiologic and echoendoscopic guidance. After fluid aspiration for cultures, a guide-wire is introduced and the needle puncture tract is enlarged (cystotome and/or hydrostatic dilatation), and then a double pig-tail catheter is inserted. In a series of eight patients undergoing transesophageal drainage of abscesses resulting from anastomotic leak into the mediastinum or pleura, seven patients had a favorable outcome (fistula closure in 6 to 41 days) and were able to return home [33]. After drain removal, the transmural esophageal tract is closed with an endoclip.
Specific management of anastomotic complications after bariatric surgery Dehiscence and anastomotic fistula are observed in 2 to 4% of morbidly obese patients who undergo gastric bypass [34]. A removable self-expanding (partially covered) metallic stent was used to treat the leaking anastomosis in a series of 21 complications (eight after gastric bypass, 12 after sleeve gastrectomy); fistula closure was observed in 17 cases (81%, 13 immediately, and four after additional clips or fibrin glue sealing). The four remaining patients underwent surgery with three postoperative deaths [35]. Isolated cases of fibrin glue management have also been reported [36].
Endoscopic management of complications in digestive surgery
Combined approach Amrani et al. in Marseille recently showed that it is possible to manage complete esophageal anastomotic disruption by an endoluminal stent, combined with drainage and biological glue closure through an echoendoscopic or endoscopic approach [37]. Although the level of evidence is still low, these indications constitute clear progress in the management of severe, potentially lethal and universally complex postoperative complications.
Endoscopic management of strictures after esophagogastric surgery Esophagogastric anastomotic strictures occur frequently (approximately 20% of complications after esophagectomy). In the absence of associated leak or fistula, their treatment is based essentially on dilatation (usually with a hydrostatic balloon, occasionally with a bougie). In refractory situations (usually after three failures at hydrostatic dilatation), or in case of associated fistula, the insertion of an extractible plastic or metallic self-expandable stent represents another solution.
Dilatation Bougienage without endoscopy and hydrostatic endoscopic balloon dilatation are performed using radiologic visualization under general anesthesia. Sessions are repeated until a clinically satisfactory and durable result is obtained, which often requires at least three sessions. There are commercially available disposable bougie sets with a diameter up to 18 mm that can be inserted over a guide-wire are now available [38]. The bougies are capable of transmitting both axial and radial stress. A maximum of three bougies of progressively increasing diameter may be used per session. Hydrostatic balloons can easily be passed through the operating channel of adult gastroscopes (TTS) with diameters up to 3.8 mm wide and often admit a guide-wire. Their usual length is 55 mm and their diameter ranges from 6 to 25 mm. The radial force transmitted is 3 to 8 atmospheres. The success rates of the two techniques are comparable: from 78 to 83% for the bougies, and from 83 to 100% for the balloons [39]. The perforation rates are also similar (1.7% for the bougies, 1.25% for the balloon) [40]. Bacteremia occurs frequently (approximately 20%) while bleeding is rare (< 1%). The choice between the two is often made according to the preferences and habits of the operator (hydrostatic dilatation is used more and more often).
Removable endostents According to the same general and technical principles as those described for the management of fistula, refractory anastomotic strictures can also be treated with removable self-expanding stents [41]. Clinical evaluation is still poor (only a few clinical series with small numbers of patients) [42—44]. The migration rate seems high (> 30%, especially for plastic stents) and there is a related risk of perforation. Nonetheless, this approach seems to provide adequate possibilities of long-term alimentation in 50% of patients [40]. A newer option for benign strictures is the insertion of biodegradable esophageal stents (therefore no extraction is necessary). In a Japanese series on patients with refractory strictures after Lewis-Santy esophagectomy who
S7 underwent dilatation, 13 patients were treated with this stent: notwithstanding a high migration rate of 77%, dysphagia was relieved in all patients, without further need for dilatation, and with a satisfactory follow-up (7 months to 2 years) [45].
Endoscopic treatment of anastomotic leaks after colonic surgery The therapeutic principles described for the endoscopic management of upper GI complications are sometimes applicable to complications after colonic surgery. Their evaluation is insufficient, since clinical series are rare and have small numbers of patients. The frequency of anastomotic leakage after total coloproctectomy is evaluated at 9% [34]. Endoscopic treatment can be envisioned preferentially for subperitoneal anastomotic leaks. The treatment is based on the same principles: clipping and sealing techniques for small-size primary defects [46], removable endostents for larger anastomotic dehiscences (with an inherent increased risk of stent migration) [37]. Encouraging results have been published for innovative endoscopic techniques (over-thescope clip and biodegradable stents) but again, the numbers were small and the studies were not controlled [47,48]. Hydrostatic balloon dilatation is the favored endoscopic treatment for rectocolic or ileocolic strictures; bougie dilatation should be proscribed because it carries a high risk of perforation [49]. Postoperative collections are frequent after colorectal surgery (up to 10—20% of cases after anterior resection of the rectum). Their treatment relies drainage guided by echoendoscopy, especially for posterior collections that are difficult to access with CT-guided needle aspiration. The technique is similar to that described for mediastinal collections, with a success rate of 80% described for selected patients treated in expert centers [50].
General conclusions Whatever the approach chosen for postoperative digestive tract fistula, management is medico-surgical and cannot be limited to simple closure of the breach in the digestive tube wall. Drainage of all collections, adequate systemic treatment of infection and nutritional support are mandatory adjuvants. Management of strictures is essentially endoscopic (first hydrostatic dilatations, then removable metallic stents for refractory cases). KEY-POINTS • Anastomotic fistula after esophagogastric surgery: ◦ Treatment combining antibiotic therapy, fasting, and nutritional support should be initiated as early as possible. ◦ Removable stents should be used for large anastomotic leaks (up to 70% of the circumference). ◦ Clip or glue can be envisioned for small perforations (or as complementary treatment along with stents). ◦ Efficacy is 70%. ◦ Echoendoscopic drainage of mediastinal collections: efficacy up to 80% in expert centers.
S8
X. Dray et al. • Esophagogastric or esophago-jejunal anastomotic stricture: ◦ Hydrostatic balloon dilatation (or bougie dilatation according to operator experience). ◦ Removable stents for refractory situations. ◦ Efficacy 85—100% of cases. • Colonic anastomotic fistula: ◦ Prevalence of anastomotic leakage after total coloproctectomy is 9%. ◦ Postoperative collections occur frequently after colorectal surgery. Treatment: CT-guided and echoendoscopic guided aspiration. ◦ Clipping, fibrin glue sealing for small-size primary orifices; removable stents for larger dehiscences. • Colorectal or ileocolic anastomotic strictures: ◦ Hydrostatic balloon dilatation initially. ◦ Removable stents for refractory situations.
Disclosure of interest The authors declare that they have no conflicts of interest concerning this article.
References [1] Pioche M, Ponchon T. Management of bile duct leaks. J Chir Visc 2013, http://dx.doi.org/10.1016/j.jviscsurg.2013.05.004. [2] Lozacc’h P, Vandenbroucke F, Timani A, et al. Le cancer de l’œsophage. Réflexions après 25 ans d’expérience et prise en charge de 1000 cas. Mem Acad Natl Chir 2006;5(3):31—6. [3] Langer FB, Wenzl E, Prager G, et al. Management of postoperative esophageal leaks with the Polyflex self-expanding covered plastic stent. Ann Thorac Surg 2005;79(2):398—403 [Discussion 4]. [4] Raju GS, Thompson C, Zwischenberger JB. Emerging endoscopic options in the management of esophageal leaks (videos). Gastrointest Endosc 2005;62(2):278—86. [5] Schubert D, Scheidbach H, Kuhn R, et al. Endoscopic treatment of thoracic esophageal anastomotic leaks by using siliconecovered, self-expanding polyester stents. Gastrointest Endosc 2005;61(7):891—6. [6] Siersema PD. Treatment of esophageal perforations and anastomotic leaks: the endoscopist is stepping into the arena. Gastrointest Endosc 2005;61(7):897—900. [7] Conio M, Repici A, Battaglia G, et al. A randomized prospective comparison of self-expandable plastic stents and partially covered self-expandable metal stents in the palliation of malignant esophageal dysphagia. Am J Gastroenterol 2007;102(12):2667—77. [8] Verschuur EM, Repici A, Kuipers EJ, Steyerberg EW, Siersema PD. New design esophageal stents for the palliation of dysphagia from esophageal or gastric cardia cancer: a randomized trial. Am J Gastroenterol 2008;103(2):304—12. [9] Yakoub D, Fahmy R, Athanasiou T, et al. Evidence-based choice of esophageal stent for the palliative management of malignant dysphagia. World J Surg 2008;32(9):1996—2009. [10] Brinster CJ, Singhal S, Lee L, Marshall MB, Kaiser LR, Kucharczuk JC. Evolving options in the management of esophageal perforation. Ann Thorac Surg 2004;77(4):1475—83. [11] Pasricha PJ, Fleischer DE, Kalloo AN. Endoscopic perforations of the upper digestive tract: a review of their pathogenesis, prevention, and management. Gastroenterology 1994;106(3):787—802.
[12] Wehrmann T, Stergiou N, Vogel B, Riphaus A, Kockerling F, Frenz MB. Endoscopic debridement of paraesophageal, mediastinal abscesses: a prospective case series. Gastrointest Endosc 2005;62(3):344—9. [13] Siersema PD. Esophageal cancer. Gastroenterol Clin North Am 2008;37(4):943—64 [x]. [14] Shim CS, Jung IS, Bhandari S, et al. Management of malignant strictures of the cervical esophagus with a newly-designed selfexpanding metal stent. Endoscopy 2004;36(6):554—7. [15] Vanbiervliet G, Filippi J, Karimdjee BS, et al. The role of clips in preventing migration of fully covered metallic esophageal stents: a pilot comparative study. Surg Endosc 2012;26(1):53—9. [16] Swinnen J, Eisendrath P, Rigaux J, et al. Self-expandable metal stents for the treatment of benign upper GI leaks and perforations. Gastrointest Endosc 2011;73(5):890—9. [17] Freeman RK, Ascioti AJ, Wozniak TC. Postoperative esophageal leak management with the Polyflex esophageal stent. J Thorac Cardiovasc Surg 2007;133(2):333—8. [18] Ott C, Ratiu N, Endlicher E, et al. Self-expanding Polyflex plastic stents in esophageal disease: various indications, complications, and outcomes. Surg Endosc 2007;21(6): 889—96. [19] Hunerbein M, Stroszczynski C, Moesta KT, Schlag PM. Treatment of thoracic anastomotic leaks after esophagectomy with selfexpanding plastic stents. Ann Surg 2004;240(5):801—7. [20] Dai YY, Gretschel S, Dudeck O, Rau B, Schlag PM, Hunerbein M. Treatment of oesophageal anastomotic leaks by temporary stenting with self-expanding plastic stents. Br J Surg 2009;96(8):887—91. [21] Johnsson E, Lundell L, Liedman B. Sealing of esophageal perforation or ruptures with expandable metallic stents: a prospective controlled study on treatment efficacy and limitations. Dis Esophagus 2005;18(4):262—6. [22] Minami S, Gotoda T, Ono H, Oda I, Hamanaka H. Complete endoscopic closure of gastric perforation induced by endoscopic resection of early gastric cancer using endoclips can prevent surgery (with video). Gastrointest Endosc 2006;63(4): 596—601. [23] Merrifield BF, Lautz D, Thompson CC. Endoscopic repair of gastric leaks after Roux-en-Y gastric bypass: a less invasive approach. Gastrointest Endosc 2006;63(4):710—4. [24] Raymer GS, Sadana A, Campbell DB, Rowe WA. Endoscopic clip application as an adjunct to closure of mature esophageal perforation with fistulae. Clin Gastroenterol Hepatol 2003;1(1):44—50. [25] van Bodegraven AA, Kuipers EJ, Bonenkamp HJ, Meuwissen SG. Esophagopleural fistula treated endoscopically with argon beam electrocoagulation and clips. Gastrointest Endosc 1999;50(3):407—9. [26] Albert JG, Friedrich-Rust M, Woeste G, et al. Benefit of a clipping device in use in intestinal bleeding and intestinal leakage. Gastrointest Endosc 2011;74(2):389—97. [27] Manta R, Manno M, Bertani H, et al. Endoscopic treatment of gastrointestinal fistulas using an over-the-scope clip (OTSC) device: case series from a tertiary referral center. Endoscopy 2011;43(6):545—8. [28] Surace M, Mercky P, Demarquay JF, et al. Endoscopic management of GI fistulae with the over-the-scope clip system (with video). Gastrointest Endosc 2011;74(6):1416—9. [29] Erdogan A, Gurses G, Keskin H, Demircan A. The sealing effect of a fibrin tissue patch on the esophageal perforation area in primary repair. World J Surg 2007;31(11):2199—203. [30] Seewald S, Ang TL, Imazu H, et al. A standardized injection technique and regimen ensures success and safety of N-butyl-2-cyanoacrylate injection for the treatment of gastric fundal varices (with videos). Gastrointest Endosc 2008;68(3): 447—54. [31] Rabago LR, Castro JL, Joya D, et al. Esophageal perforation and postoperative fistulae of the upper digestive tract treated endoscopically with the application of Tissucol. Gastroenterol Hepatol 2000;23(2):82—6.
Endoscopic management of complications in digestive surgery [32] Fischer A, Schrag HJ, Goos M, von Dobschuetz E, Hopt UT. Nonoperative treatment of four esophageal perforations with hemostatic clips. Dis Esophagus 2007;20(5):444—8. [33] Infante M, Valente M, Andreani S, et al. Conservative management of esophageal leaks by transluminal endoscopic drainage of the mediastinum or pleural space. Surgery 1996;119(1):46—50. [34] Raval MJ, Schnitzler M, O’Connor BI, Cohen Z, McLeod RS. Improved outcome due to increased experience and individualized management of leaks after ileal pouch-anal anastomosis. Ann Surg 2007;246(5):763—70. [35] Eisendrath P, Cremer M, Himpens J, Cadiere GB, Le Moine O, Deviere J. Endotherapy including temporary stenting of fistulas of the upper gastrointestinal tract after laparoscopic bariatric surgery. Endoscopy 2007;39(7):625—30. [36] Garcia-Caballero M, Carbajo M, Martinez-Moreno JM, Sarria M, Osorio D, Carmona JA. Drain erosion and gastro-jejunal fistula after one-anastomosis gastric bypass: endoscopic occlusion by fibrin sealant. Obes Surg 2005;15(5):719—22. [37] Amrani L, Menard C, Berdah S, et al. From iatrogenic digestive perforation to complete anastomotic disunion: endoscopic stenting as a new concept of ‘‘stent-guided regeneration and re-epithelialization’’. Gastrointest Endosc 2009;69(7):1282—7. [38] Wichern Jr WA. Perforation of the esophagus. Am J Surg 1970;119(5):534—6. [39] Cox JG, Winter RK, Maslin SC, et al. Balloon or bougie for dilatation of benign esophageal stricture? Dig Dis Sci 1994;39(4):776—81. [40] Kim HC, Shin JH, Song HY, et al. Fluoroscopically guided balloon dilation for benign anastomotic stricture after Ivor-Lewis esophagectomy: experience in 62 patients. J Vasc Interv Radiol 2005;16(12):1699—704. [41] de Wijkerslooth LR, Vleggaar FP, Siersema PD. Endoscopic management of difficult or recurrent esophageal strictures. Am J Gastroenterol 2011;106(12):2080—91 [Quiz 92].
S9 [42] Paul L, Pinto I, Gomez H, Fernandez-Lobato R, Moyano E. Metallic stents in the treatment of benign diseases of the colon: preliminary experience in 10 cases. Radiology 2002;223(3):715—22. [43] Repici A, Hassan C, Sharma P, Conio M, Siersema P. Systematic review: the role of self-expanding plastic stents for benign oesophageal strictures. Aliment Pharmacol Ther 2010;31(12):1268—75. [44] Thomas T, Abrams KR, Subramanian V, Mannath J, Ragunath K. Esophageal stents for benign refractory strictures: a meta-analysis. Endoscopy 2011;43(5): 386—93. [45] Saito Y, Tanaka T, Andoh A, et al. Usefulness of biodegradable stents constructed of poly-l-lactic acid monofilaments in patients with benign esophageal stenosis. World J Gastroenterol 2007;13(29):3977—80. [46] Del Rio P, Dell’Abate P, Soliani P, Ziegler S, Arcuri M, Sianesi M. Endoscopic treatment of esophageal and colorectal fistulas with fibrin glue. Acta Biomed 2005;76(2): 95—8. [47] Arezzo A, Verra M, Reddavid R, Cravero F, Bonino MA, Morino M. Efficacy of the over-the-scope clip (OTSC) for treatment of colorectal postsurgical leaks and fistulas. Surg Endosc 2012;26(11):3330—3. [48] Perez Roldan F, Gonzalez Carro P, Villafanez Garcia MC, et al. Usefulness of biodegradable polydioxanone stents in the treatment of postsurgical colorectal strictures and fistulas. Endoscopy 2012;44(3):297—300. [49] Suchan KL, Muldner A, Manegold BC. Endoscopic treatment of postoperative colorectal anastomotic strictures. Surg Endosc 2003;17(7):1110—3. [50] Giovannini M, Bories E, Moutardier V, et al. Drainage of deep pelvic abscesses using therapeutic echo endoscopy. Endoscopy 2003;35(6):511—4.
Available online at
www.sciencedirect.com
REVIEW
Endoscopic management of complications in digestive surgery X. Dray ∗, M. Camus , U. Chaput Université Paris Diderot, Paris 7, AP—HP, Service d’Hépatogastroentérologie, Hôpital Lariboisière, 2, rue Ambroise-Paré, 75010 Paris, France
KEYWORDS Therapeutic endoscopy; Stricture; Anastomotic fistula; Anastomotic leak; Postoperative complications
Summary Endoscopy has an ever-increasing role in the treatment of complications in digestive surgery. Endoscopic treatment is essentially used for (i) fistula or intra-abdominal collection secondary to anastomotic dehiscence and (ii) anastomotic stricture, especially esophagogastric, but also sometimes after colorectal surgery. First intention treatment of fistula following esophagogastric surgery is the insertion of an extractable self-expandable metallic stent (partially or entirely covered); this is supported by a low level of scientific evidence, but clinical experience has been satisfactory. Other techniques for treatment of anastomotic leak have also been reported anecdotally (clip placement, sealing with glue). There are few data available in the literature on endoscopic management (stents essentially) of postoperative colonic fistula. Whatever the approach chosen to treat a postoperative digestive tract fistula, management is medico-surgical and cannot be limited to simple closure of the digestive tube wall defect. Drainage of any collections by endoscopic, radiologic or surgical approach, systemic treatment of infection and nutritional support are essential adjuvant treatment modalities. Treatment of postoperative esophageal or colonic strictures is essentially endoscopic and is based on initial dilatation (endoscopic with hydrostatic balloon or bougie), and placement of extractable metallic stents (partially or entirely covered) in case of refractory outcome. © 2013 Elsevier Masson SAS. All rights reserved.
Introduction Endoscopy has an increasing role in the treatment of complications in digestive surgery. Endoscopic treatment is essentially used in case of: • fistula or intra-abdominal collection secondary to early or late (necrosis) anastomotic dehiscence, with fistula or intra-abdominal collection; • anastomotic stricture, especially esophagogastric, but also sometimes after colorectal or biliopancreatic surgery.
∗
Corresponding author. AP—HP, Service d’Hépatogastroentérologie, Pôle Digestif-Uro-Néphrologie-Gynécologie-Obstétrique, Hôpital Lariboisière, 2, rue Ambroise-Paré, 75010 Paris, France. Tel.: +33 1 49 95 25 45 or 78; fax: +33 1 49 95 25 77. E-mail address: [email protected] (X. Dray). 1878-7886/$ — see front matter © 2013 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.jviscsurg.2013.04.005
S4 The goal of this review is to expose the techniques and results of endoscopic interventional management and discuss the indications of endoscopic treatment of postoperative gastrointestinal tract complications. The treatment of anastomotic bleeding, rare and most often benign, calls for classical endoscopic hemostatic techniques and will not be dealt with here. Endoscopic management of biliopancreatic complications is discussed in another update of this thematic issue [1].
Endoscopic treatment of anastomotic fistula after esophagogastric surgery Therapeutic principles and level of evidence The anastomotic leak rate after esophagectomy is approximately 5% [2]. Mortality ranges from 20% to 25% when the leak is intrathoracic [2]. Progress has been made in endoscopic treatment giving it an ever-increasing role in temporary prosthetic management (removable stents) as a mini-invasive alternative to surgery in this setting when anastomotic dehiscences involve less than 70% of the circumference [3—6]. Other endoscopic treatments are occasionally used to close smaller caliber leaks (clips, glue,
X. Dray et al. endoscopic suture) and have a limited place in the treatment of postoperative fistula in the mediastinum (Fig. 1). Endoscopic treatment of anastomotic leak after esophagectomy should not be an isolated modality; it must be associated with drainage, broad-spectrum antibiotic therapy, nutritional support and surgery. The choice of medical treatment (including stent insertion) should be decided as early as possible, before massive mediastinal contamination or uncontrolled sepsis sets in. There are few data in the literature attesting to the efficacy of endoscopic treatment of esophageal fistula. Most publications are case reports (often selected on the basis of therapeutic success, and therefore biased); there are a few retrospective series with low numbers of patients, and more ideally, some prospective non-comparative studies, also with low numbers. A few principles emerge from these studies and can be applied to clinical practice, without waiting for controlled trials (which are unlikely to be performed) comparing endoscopic to surgical treatment (the indications are too different). Therefore, in general, non-surgical treatment, particularly endoscopic, should be of primary interest for: • contained (in the mediastinum and/or the neighboring pleura) fistulae and anastomotic dehiscences; • patients who can be managed early; • patients in good enough condition to undergo urgent surgery in case of unfavorable evolution.
Figure 1. Endoscopic treatment by fibrin glue, clips and covered stent for esophagopleural fistula after Lewis-Santy esophagectomy. Reduction of pleural drainage volume from 400 to 40 mL/d by 7 days after insertion of self-expandable completely covered metallic stent for anastomotic leakage. (A) Endoscopic view showing a migrated stent that no longer covers the perforation. After retrieval of the stent, a catheter is introduced into the primary orifice of the fistula (B) and shows contrast material in the pleural drain (C). Abrade the fistular tract with a biliary brushing catheter (D, E). After sealing the fistular tract (F, G), obstruction by glue of the primary orifice is visible (H). Placement of two clips on the primary orifice (I, J, K, L). Insertion of guide-wire (M, N) for a self-expanding completely covered metallic stent in the esophagus (O, P). Proximal release of the stent (Q, R). At the end of the procedure, the stent is in good position, without leakage of contrast material into the pleural space (S, T).
Endoscopic management of complications in digestive surgery
S5
Digestive tract endoprostheses or stents
Delivery technique
Material
Esophageal self-expanding stents are placed under endoscopic and radiologic guidance and usually require general anesthesia. In certain situations, nonetheless, they can be placed with only endoscopic or only radiologic guidance. The endoscopist must have at hand a full array of guide-wires, opacification catheters and stents. The initial step consists of visualizing and landmarking the fistula endoscopically and by contrast opacification, taking note of the exact location in order to best choose the length of the stent. Radio-opaque objects (e.g., paper clip or needle) are used for cutaneous landmarking if patient movements can be limited, or more rarely internal markings (submucosal injection of contrast or clip placement) are used. After placement of a guidewire the endoscope is withdrawn since the esophageal stent is too wide to be inserted through the endoscope-operating channel. The second step is the insertion and delivery of the stent, over the guide-wire. A rigid guide-wire (Savary type) is preferred. An endoscope can be inserted alongside the guide-wire for visualization. The distal end of the stent is placed in the stomach, gastroplasty or distal jejunum, according to the previous surgery. Small adjustments can be made at the beginning of the delivery thanks to a recapturing system within the sheath. Spontaneous expansion of the stent ensures immediate anchorage, but there is always a slight risk of early migration. Contact between the stent and the endoscope must be avoided immediately after the delivery (increased risk of migration). Certain authors feel that the proximal extremity of the stent should be fixed by hemostatic clips to the esophageal wall to prevent migration (however, the level of evidence for this manoeuvre is low) [15]. Opacification within the stent itself after stent deployment is performed to check that there is no extravasation of the contrast material.
Removable self-expanding stents are the material of choice in this setting. Historically, endoluminal treatment of esophageal fistula was developed using the only selfexpandable (plastic) stent available at that time (Polyflex® , Boston Scientific). A wide array of stents are currently available, with diameters ranging from 16 to 21 mm and lengths from 90 to 150 mm. The releasing system must be assembled by the operator. The stent is made of silicone, entirely covered with a polyester film and flared at the proximal end. Coating limits the risk of incarceration into the esophageal mucosa, and facilitates stent extraction, but enhances the risk of migration (approximately 30% in stricture and probably higher in fistula, vs. 3—12%, for metallic stents depending on whether they are partially or completely coated) [7—9]. Plastic stents are less expensive than the metallic models. The large diameter of these stents and their rigid character are risk factors for complications (perforation, hemorrhage): 9% vs. 3% for metallic stents [7]. Currently, removable self-expanding metal stents (partially or completely coated) are used in the treatment of complications of esophagogastric surgery. While their cost is clearly higher than the plastic stents, their utilization is easier (pre-assembled and more flexible), the risk of complications (perforation, hemorrhage, migration) is less elevated [10—12]. These stents are constituted of one or several braided strands of a metal with high shape memory (most often made of Nitinol). They are contained within a tight sheath placed on a carrier tube. The caliber of the assembled system ranges from 18 to 24 Fr, which is not compatible with the passage of the stent through the operating channel of an endoscope (over-the-wire, OTW, system). There are newer infrequently-used models with a thinner delivery diameter (10 Fr) that can be passed through a 3.7 mm operating channel (through-the-scope system, TTS). The entire delivery system (carrier tube, pleated stent and over catheter) is inserted over the rigid guide-wire, under radioscopic and/or endoscopic guidance, with the endoscope lying alongside the stent. The self-expandable metallic stents vary according to the alloy used, the type of latticework, the form, the length (70 to 170 cm and even longer), the diameter (18 to 23 mm or greater), and the delivery system as well as the coating material. The proximal, and sometime the distal, extremity are flared (approximately 5 mm larger than the shaft), in an attempt to limit the risk of migration. Depending on the model, the length of the delivered stents shrinks after expansion by 20% to 40% from the original length within the sheath. The most classic models are of the ‘‘distal delivery’’ type. Several options (anti-reflux valve, anti-migration systems) are available, but there has been no formal demonstration of their benefits at the present time. The choice of the stent is crucial in case of proximal esophageal fistula, especially after the Lewis-Santy operation. The low radial force of the Ultraflex® (Boston Scientific) stent in this indication is worthy of mention [13]. A stent with a short upper cuff (5 mm) was developed for this indication (M.I.Tech Co., Pyontack, Korea) [14]. Proximal delivery systems allow adjustment of the position of the stent between the high-riding fistula and upper esophageal sphincter. A proximal delivery stent with a short cuff has recently been commercialized (Taewoong Medical, Seoul, Korea).
Precautions necessary during the procedure Strict and prolonged fasting is necessary (with adequate nutritional support as needed) until formal and durable demonstration of the absence of perforation. This can be ensured by radiology (esophageal follow-through) or methylene blue. A thoracic X-ray on day 3 can help evaluate of the degree of expansion of the stent and the absence of early migration. Alimentation per mouth can be envisioned if the clinical evolution is satisfactory. Recommendations to avoid the risk of food impaction within the stent include: eating while in the seated position, avoiding thick, dense aliments (meat, for example), avoiding stringy aliments (leeks for example), avoiding karaya gum gastric demulcents, and drinking sparkling water at the end of the meal.
Extraction technique of esophageal stents Stent withdrawal can be envisioned 4 to 8 weeks after placement. Withdrawal can be done simply by pulling on the nylon string attached to the proximal end of the stent, either with a toothed forceps or polypectomy snare. Certain partially covered models (Ultraflex® ) can be withdrawn by traction on their distal extremity, which progressively turns the stent inside out. In more difficult cases, particularly extraction of partially covered metallic stents with epithelial ingrowth into the mesh, grasping the stent at two points via the two channels of a double-channel endoscope can be helpful. When the proximal extremity abuts the esophageal upper
S6 sphincter, placing an over-tube can be useful for extraction. The placement of a plastic stent within a partially covered metallic stent can promote pressure necrosis of the hyperplastic epithelium and thus facilitate retrieval as a two-step procedure [16].
Results of endoluminal stenting While early experience argues in favor of the use of selfexpandable plastic stents in this setting, there are no randomized trials or high-level evidence on the topic. Freeman et al. observed a successful outcome in 20 out of 21 patients despite a 25% migration rate [17]. Likewise, in 2007, Ott et al. evaluated the use of plastic stents (Polyflex® , Boston Scientific, USA) in a retrospective series of 35 patients, 12 of whom had an esophageal fistula. The results were considered very good in spite of the 37% migration rate [18]. Hunerbein et al. retrospectively compared the outcome of nine patients treated with a plastic expanding stent (PolyflexTM ) with that of 10 patients treated previously either by revisional surgery or with conservative measures. Stent placement resulted in immediate occlusion of the fistula in eight of nine patients and the stent could be removed at a mean interval of 29 days after insertion. By comparison with conventional treatment, patients were able to start eating earlier (11 days), and had shorter intensive care stay (25 days) and overall hospital stay (35 days), while mortality was nil (vs. 20% after conventional surgical or conservative treatment) [19]. The same team updated their experience with 22 patients; five cases of stent migration were treated endoscopically without compromising the final outcome. The patients were able to resume eating early, 4 days after the insertion of the stent [20]. There are less series with extractible self-expanding metallic stents. Most of the complications in these series concerned esophageal perforation [21]. In spite of a very limited evaluation, these stents now have a predominant place (more than plastic stents) in this indication.
Endoscopic clips Endoscopic clips were developed to ensure endoscopic hemostasis and not for sutures. There are different models, some that need to be assembled, others pre-assembled, repositionable or not, rotational or not. The spread of conventional clips in the open position (11 mm at the most) was designed for mucosa-to-mucosa approximation only. The endoscopic technique consists of approximation of the clips starting from one end of the opening to the other. This must be done with minimal carbon dioxide insufflation, to prevent enlargement of the opening and mediastinal contamination, as well as to avoid creating a tangential angle of the endoscope (and therefore of the clip) in relation to the esophageal wall. Hemostatic clips were initially proposed for early closure of acute perforations, before substantial contamination of the mediastinum had occurred. The success rate was 98% in a series of 117 gastric perforations after endoscopic submucosal dissection [22]. Treatment of fistula may require initial ablation of the re-epitheliized margins, for example by argon plasma coagulation. Certain authors have reported failures when the margins were not treated in this manner, but these failures could be transformed into successes if initial mucosal ablation was performed [23—25]. New generation macroclips (11 to 14 mm) (Over-the-scope clips, Ovesco, Tübingen, Germany) have been used with
X. Dray et al. success in a few isolated cases of fistulas after esophagogastric and/or esophago-jejunal anastomosis [26—28].
Endoscopic sealing Fibrin patches have been proposed to treat esophageal leaks [29]. This technique was originally designed to manage bleeding, particularly in the case of ruptured esogastric varices due to portal hypertension. A specific technical skill must be mastered, because the glue has to be injected through a sclerosing needle passed through the operating canal of the endoscope. Injection of a bolus of lipiodol, and especially, visual control help to avoid premature release of glue in the operating channel of the endoscope. Standardized endoscopic fibrin glue techniques have been proposed [30]. A few cases of endoscopic biological glue sealing of esophageal fistula and perforations have been published. Rabago et al. reported the efficacy of such a treatment in one case of acute esophageal perforation after endoscopic extraction of impacted food, and in two patients with postoperative esophageal leaks [31]. Fischer et al. completed clip closure of a large (40 mm) iatrogenic esophageal perforation (diagnostic endoscopy) with injection of glue [32]. In neither of these two cases, however, was it possible to deduce an added benefit of fibrin glue compared with management of perforations (conservative treatment in the first case, conservative treatment + clip closure in the second case).
Echoendoscopic and endoscopic drainage of mediastinal abscess Mediastinal abscesses are classically treated via a transthoracic approach. One team has reported their experience with echoendoscopic-guided transesophageal puncture and drainage [4]. Drainage begins with the introduction of a 19G needle under combined radiologic and echoendoscopic guidance. After fluid aspiration for cultures, a guide-wire is introduced and the needle puncture tract is enlarged (cystotome and/or hydrostatic dilatation), and then a double pig-tail catheter is inserted. In a series of eight patients undergoing transesophageal drainage of abscesses resulting from anastomotic leak into the mediastinum or pleura, seven patients had a favorable outcome (fistula closure in 6 to 41 days) and were able to return home [33]. After drain removal, the transmural esophageal tract is closed with an endoclip.
Specific management of anastomotic complications after bariatric surgery Dehiscence and anastomotic fistula are observed in 2 to 4% of morbidly obese patients who undergo gastric bypass [34]. A removable self-expanding (partially covered) metallic stent was used to treat the leaking anastomosis in a series of 21 complications (eight after gastric bypass, 12 after sleeve gastrectomy); fistula closure was observed in 17 cases (81%, 13 immediately, and four after additional clips or fibrin glue sealing). The four remaining patients underwent surgery with three postoperative deaths [35]. Isolated cases of fibrin glue management have also been reported [36].
Endoscopic management of complications in digestive surgery
Combined approach Amrani et al. in Marseille recently showed that it is possible to manage complete esophageal anastomotic disruption by an endoluminal stent, combined with drainage and biological glue closure through an echoendoscopic or endoscopic approach [37]. Although the level of evidence is still low, these indications constitute clear progress in the management of severe, potentially lethal and universally complex postoperative complications.
Endoscopic management of strictures after esophagogastric surgery Esophagogastric anastomotic strictures occur frequently (approximately 20% of complications after esophagectomy). In the absence of associated leak or fistula, their treatment is based essentially on dilatation (usually with a hydrostatic balloon, occasionally with a bougie). In refractory situations (usually after three failures at hydrostatic dilatation), or in case of associated fistula, the insertion of an extractible plastic or metallic self-expandable stent represents another solution.
Dilatation Bougienage without endoscopy and hydrostatic endoscopic balloon dilatation are performed using radiologic visualization under general anesthesia. Sessions are repeated until a clinically satisfactory and durable result is obtained, which often requires at least three sessions. There are commercially available disposable bougie sets with a diameter up to 18 mm that can be inserted over a guide-wire are now available [38]. The bougies are capable of transmitting both axial and radial stress. A maximum of three bougies of progressively increasing diameter may be used per session. Hydrostatic balloons can easily be passed through the operating channel of adult gastroscopes (TTS) with diameters up to 3.8 mm wide and often admit a guide-wire. Their usual length is 55 mm and their diameter ranges from 6 to 25 mm. The radial force transmitted is 3 to 8 atmospheres. The success rates of the two techniques are comparable: from 78 to 83% for the bougies, and from 83 to 100% for the balloons [39]. The perforation rates are also similar (1.7% for the bougies, 1.25% for the balloon) [40]. Bacteremia occurs frequently (approximately 20%) while bleeding is rare (< 1%). The choice between the two is often made according to the preferences and habits of the operator (hydrostatic dilatation is used more and more often).
Removable endostents According to the same general and technical principles as those described for the management of fistula, refractory anastomotic strictures can also be treated with removable self-expanding stents [41]. Clinical evaluation is still poor (only a few clinical series with small numbers of patients) [42—44]. The migration rate seems high (> 30%, especially for plastic stents) and there is a related risk of perforation. Nonetheless, this approach seems to provide adequate possibilities of long-term alimentation in 50% of patients [40]. A newer option for benign strictures is the insertion of biodegradable esophageal stents (therefore no extraction is necessary). In a Japanese series on patients with refractory strictures after Lewis-Santy esophagectomy who
S7 underwent dilatation, 13 patients were treated with this stent: notwithstanding a high migration rate of 77%, dysphagia was relieved in all patients, without further need for dilatation, and with a satisfactory follow-up (7 months to 2 years) [45].
Endoscopic treatment of anastomotic leaks after colonic surgery The therapeutic principles described for the endoscopic management of upper GI complications are sometimes applicable to complications after colonic surgery. Their evaluation is insufficient, since clinical series are rare and have small numbers of patients. The frequency of anastomotic leakage after total coloproctectomy is evaluated at 9% [34]. Endoscopic treatment can be envisioned preferentially for subperitoneal anastomotic leaks. The treatment is based on the same principles: clipping and sealing techniques for small-size primary defects [46], removable endostents for larger anastomotic dehiscences (with an inherent increased risk of stent migration) [37]. Encouraging results have been published for innovative endoscopic techniques (over-thescope clip and biodegradable stents) but again, the numbers were small and the studies were not controlled [47,48]. Hydrostatic balloon dilatation is the favored endoscopic treatment for rectocolic or ileocolic strictures; bougie dilatation should be proscribed because it carries a high risk of perforation [49]. Postoperative collections are frequent after colorectal surgery (up to 10—20% of cases after anterior resection of the rectum). Their treatment relies drainage guided by echoendoscopy, especially for posterior collections that are difficult to access with CT-guided needle aspiration. The technique is similar to that described for mediastinal collections, with a success rate of 80% described for selected patients treated in expert centers [50].
General conclusions Whatever the approach chosen for postoperative digestive tract fistula, management is medico-surgical and cannot be limited to simple closure of the breach in the digestive tube wall. Drainage of all collections, adequate systemic treatment of infection and nutritional support are mandatory adjuvants. Management of strictures is essentially endoscopic (first hydrostatic dilatations, then removable metallic stents for refractory cases). KEY-POINTS • Anastomotic fistula after esophagogastric surgery: ◦ Treatment combining antibiotic therapy, fasting, and nutritional support should be initiated as early as possible. ◦ Removable stents should be used for large anastomotic leaks (up to 70% of the circumference). ◦ Clip or glue can be envisioned for small perforations (or as complementary treatment along with stents). ◦ Efficacy is 70%. ◦ Echoendoscopic drainage of mediastinal collections: efficacy up to 80% in expert centers.
S8
X. Dray et al. • Esophagogastric or esophago-jejunal anastomotic stricture: ◦ Hydrostatic balloon dilatation (or bougie dilatation according to operator experience). ◦ Removable stents for refractory situations. ◦ Efficacy 85—100% of cases. • Colonic anastomotic fistula: ◦ Prevalence of anastomotic leakage after total coloproctectomy is 9%. ◦ Postoperative collections occur frequently after colorectal surgery. Treatment: CT-guided and echoendoscopic guided aspiration. ◦ Clipping, fibrin glue sealing for small-size primary orifices; removable stents for larger dehiscences. • Colorectal or ileocolic anastomotic strictures: ◦ Hydrostatic balloon dilatation initially. ◦ Removable stents for refractory situations.
Disclosure of interest The authors declare that they have no conflicts of interest concerning this article.
References [1] Pioche M, Ponchon T. Management of bile duct leaks. J Chir Visc 2013, http://dx.doi.org/10.1016/j.jviscsurg.2013.05.004. [2] Lozacc’h P, Vandenbroucke F, Timani A, et al. Le cancer de l’œsophage. Réflexions après 25 ans d’expérience et prise en charge de 1000 cas. Mem Acad Natl Chir 2006;5(3):31—6. [3] Langer FB, Wenzl E, Prager G, et al. Management of postoperative esophageal leaks with the Polyflex self-expanding covered plastic stent. Ann Thorac Surg 2005;79(2):398—403 [Discussion 4]. [4] Raju GS, Thompson C, Zwischenberger JB. Emerging endoscopic options in the management of esophageal leaks (videos). Gastrointest Endosc 2005;62(2):278—86. [5] Schubert D, Scheidbach H, Kuhn R, et al. Endoscopic treatment of thoracic esophageal anastomotic leaks by using siliconecovered, self-expanding polyester stents. Gastrointest Endosc 2005;61(7):891—6. [6] Siersema PD. Treatment of esophageal perforations and anastomotic leaks: the endoscopist is stepping into the arena. Gastrointest Endosc 2005;61(7):897—900. [7] Conio M, Repici A, Battaglia G, et al. A randomized prospective comparison of self-expandable plastic stents and partially covered self-expandable metal stents in the palliation of malignant esophageal dysphagia. Am J Gastroenterol 2007;102(12):2667—77. [8] Verschuur EM, Repici A, Kuipers EJ, Steyerberg EW, Siersema PD. New design esophageal stents for the palliation of dysphagia from esophageal or gastric cardia cancer: a randomized trial. Am J Gastroenterol 2008;103(2):304—12. [9] Yakoub D, Fahmy R, Athanasiou T, et al. Evidence-based choice of esophageal stent for the palliative management of malignant dysphagia. World J Surg 2008;32(9):1996—2009. [10] Brinster CJ, Singhal S, Lee L, Marshall MB, Kaiser LR, Kucharczuk JC. Evolving options in the management of esophageal perforation. Ann Thorac Surg 2004;77(4):1475—83. [11] Pasricha PJ, Fleischer DE, Kalloo AN. Endoscopic perforations of the upper digestive tract: a review of their pathogenesis, prevention, and management. Gastroenterology 1994;106(3):787—802.
[12] Wehrmann T, Stergiou N, Vogel B, Riphaus A, Kockerling F, Frenz MB. Endoscopic debridement of paraesophageal, mediastinal abscesses: a prospective case series. Gastrointest Endosc 2005;62(3):344—9. [13] Siersema PD. Esophageal cancer. Gastroenterol Clin North Am 2008;37(4):943—64 [x]. [14] Shim CS, Jung IS, Bhandari S, et al. Management of malignant strictures of the cervical esophagus with a newly-designed selfexpanding metal stent. Endoscopy 2004;36(6):554—7. [15] Vanbiervliet G, Filippi J, Karimdjee BS, et al. The role of clips in preventing migration of fully covered metallic esophageal stents: a pilot comparative study. Surg Endosc 2012;26(1):53—9. [16] Swinnen J, Eisendrath P, Rigaux J, et al. Self-expandable metal stents for the treatment of benign upper GI leaks and perforations. Gastrointest Endosc 2011;73(5):890—9. [17] Freeman RK, Ascioti AJ, Wozniak TC. Postoperative esophageal leak management with the Polyflex esophageal stent. J Thorac Cardiovasc Surg 2007;133(2):333—8. [18] Ott C, Ratiu N, Endlicher E, et al. Self-expanding Polyflex plastic stents in esophageal disease: various indications, complications, and outcomes. Surg Endosc 2007;21(6): 889—96. [19] Hunerbein M, Stroszczynski C, Moesta KT, Schlag PM. Treatment of thoracic anastomotic leaks after esophagectomy with selfexpanding plastic stents. Ann Surg 2004;240(5):801—7. [20] Dai YY, Gretschel S, Dudeck O, Rau B, Schlag PM, Hunerbein M. Treatment of oesophageal anastomotic leaks by temporary stenting with self-expanding plastic stents. Br J Surg 2009;96(8):887—91. [21] Johnsson E, Lundell L, Liedman B. Sealing of esophageal perforation or ruptures with expandable metallic stents: a prospective controlled study on treatment efficacy and limitations. Dis Esophagus 2005;18(4):262—6. [22] Minami S, Gotoda T, Ono H, Oda I, Hamanaka H. Complete endoscopic closure of gastric perforation induced by endoscopic resection of early gastric cancer using endoclips can prevent surgery (with video). Gastrointest Endosc 2006;63(4): 596—601. [23] Merrifield BF, Lautz D, Thompson CC. Endoscopic repair of gastric leaks after Roux-en-Y gastric bypass: a less invasive approach. Gastrointest Endosc 2006;63(4):710—4. [24] Raymer GS, Sadana A, Campbell DB, Rowe WA. Endoscopic clip application as an adjunct to closure of mature esophageal perforation with fistulae. Clin Gastroenterol Hepatol 2003;1(1):44—50. [25] van Bodegraven AA, Kuipers EJ, Bonenkamp HJ, Meuwissen SG. Esophagopleural fistula treated endoscopically with argon beam electrocoagulation and clips. Gastrointest Endosc 1999;50(3):407—9. [26] Albert JG, Friedrich-Rust M, Woeste G, et al. Benefit of a clipping device in use in intestinal bleeding and intestinal leakage. Gastrointest Endosc 2011;74(2):389—97. [27] Manta R, Manno M, Bertani H, et al. Endoscopic treatment of gastrointestinal fistulas using an over-the-scope clip (OTSC) device: case series from a tertiary referral center. Endoscopy 2011;43(6):545—8. [28] Surace M, Mercky P, Demarquay JF, et al. Endoscopic management of GI fistulae with the over-the-scope clip system (with video). Gastrointest Endosc 2011;74(6):1416—9. [29] Erdogan A, Gurses G, Keskin H, Demircan A. The sealing effect of a fibrin tissue patch on the esophageal perforation area in primary repair. World J Surg 2007;31(11):2199—203. [30] Seewald S, Ang TL, Imazu H, et al. A standardized injection technique and regimen ensures success and safety of N-butyl-2-cyanoacrylate injection for the treatment of gastric fundal varices (with videos). Gastrointest Endosc 2008;68(3): 447—54. [31] Rabago LR, Castro JL, Joya D, et al. Esophageal perforation and postoperative fistulae of the upper digestive tract treated endoscopically with the application of Tissucol. Gastroenterol Hepatol 2000;23(2):82—6.
Endoscopic management of complications in digestive surgery [32] Fischer A, Schrag HJ, Goos M, von Dobschuetz E, Hopt UT. Nonoperative treatment of four esophageal perforations with hemostatic clips. Dis Esophagus 2007;20(5):444—8. [33] Infante M, Valente M, Andreani S, et al. Conservative management of esophageal leaks by transluminal endoscopic drainage of the mediastinum or pleural space. Surgery 1996;119(1):46—50. [34] Raval MJ, Schnitzler M, O’Connor BI, Cohen Z, McLeod RS. Improved outcome due to increased experience and individualized management of leaks after ileal pouch-anal anastomosis. Ann Surg 2007;246(5):763—70. [35] Eisendrath P, Cremer M, Himpens J, Cadiere GB, Le Moine O, Deviere J. Endotherapy including temporary stenting of fistulas of the upper gastrointestinal tract after laparoscopic bariatric surgery. Endoscopy 2007;39(7):625—30. [36] Garcia-Caballero M, Carbajo M, Martinez-Moreno JM, Sarria M, Osorio D, Carmona JA. Drain erosion and gastro-jejunal fistula after one-anastomosis gastric bypass: endoscopic occlusion by fibrin sealant. Obes Surg 2005;15(5):719—22. [37] Amrani L, Menard C, Berdah S, et al. From iatrogenic digestive perforation to complete anastomotic disunion: endoscopic stenting as a new concept of ‘‘stent-guided regeneration and re-epithelialization’’. Gastrointest Endosc 2009;69(7):1282—7. [38] Wichern Jr WA. Perforation of the esophagus. Am J Surg 1970;119(5):534—6. [39] Cox JG, Winter RK, Maslin SC, et al. Balloon or bougie for dilatation of benign esophageal stricture? Dig Dis Sci 1994;39(4):776—81. [40] Kim HC, Shin JH, Song HY, et al. Fluoroscopically guided balloon dilation for benign anastomotic stricture after Ivor-Lewis esophagectomy: experience in 62 patients. J Vasc Interv Radiol 2005;16(12):1699—704. [41] de Wijkerslooth LR, Vleggaar FP, Siersema PD. Endoscopic management of difficult or recurrent esophageal strictures. Am J Gastroenterol 2011;106(12):2080—91 [Quiz 92].
S9 [42] Paul L, Pinto I, Gomez H, Fernandez-Lobato R, Moyano E. Metallic stents in the treatment of benign diseases of the colon: preliminary experience in 10 cases. Radiology 2002;223(3):715—22. [43] Repici A, Hassan C, Sharma P, Conio M, Siersema P. Systematic review: the role of self-expanding plastic stents for benign oesophageal strictures. Aliment Pharmacol Ther 2010;31(12):1268—75. [44] Thomas T, Abrams KR, Subramanian V, Mannath J, Ragunath K. Esophageal stents for benign refractory strictures: a meta-analysis. Endoscopy 2011;43(5): 386—93. [45] Saito Y, Tanaka T, Andoh A, et al. Usefulness of biodegradable stents constructed of poly-l-lactic acid monofilaments in patients with benign esophageal stenosis. World J Gastroenterol 2007;13(29):3977—80. [46] Del Rio P, Dell’Abate P, Soliani P, Ziegler S, Arcuri M, Sianesi M. Endoscopic treatment of esophageal and colorectal fistulas with fibrin glue. Acta Biomed 2005;76(2): 95—8. [47] Arezzo A, Verra M, Reddavid R, Cravero F, Bonino MA, Morino M. Efficacy of the over-the-scope clip (OTSC) for treatment of colorectal postsurgical leaks and fistulas. Surg Endosc 2012;26(11):3330—3. [48] Perez Roldan F, Gonzalez Carro P, Villafanez Garcia MC, et al. Usefulness of biodegradable polydioxanone stents in the treatment of postsurgical colorectal strictures and fistulas. Endoscopy 2012;44(3):297—300. [49] Suchan KL, Muldner A, Manegold BC. Endoscopic treatment of postoperative colorectal anastomotic strictures. Surg Endosc 2003;17(7):1110—3. [50] Giovannini M, Bories E, Moutardier V, et al. Drainage of deep pelvic abscesses using therapeutic echo endoscopy. Endoscopy 2003;35(6):511—4.