Surg Endosc DOI 10.1007/s00464-017-5922-1
and Other Interventional Techniques
Endoscopic transmural management of abdominal fluid collection following gastrointestinal, bariatric, and hepato-bilio-pancreatic surgery Gianfranco Donatelli1 · David Fuks2 · Fabrizio Cereatti3 · Guillaume Pourcher2 · Thierry Perniceni2 · Jean‑Loup Dumont1 · Thierry Tuszynski1 · Bertrand Marie Vergeau1 · Bruno Meduri1 · Brice Gayet2
Received: 3 March 2017 / Accepted: 8 October 2017 © Springer Science+Business Media, LLC 2017
Abstract Background Post-operative collections are a recognized source of morbidity after abdominal surgery. Percutaneous drainage is currently considered the standard treatment but not all collections are accessible using this method. Since the adoption of EUS, endoscopic transmural drainage has become an attractive option in the management of such complications. The present study aimed to assess the efficacy, safety and modalities of endoscopic transmural drainage in the treatment of post-operative collections. Methods Data of all patients referred to our dedicated multidisciplinary facility from 2014 to 2017 for endoscopic drainage of symptomatic post-operative collections after failure of percutaneous drainage or when it was deemed impossible, were retrospectively analyzed. Results Thirty-two patients (17 males and 15 females) with a median age of 53 years old (range 31–74) were included. Collections resulted from pancreatic (n = 10), colorectal (n = 6), bariatric (n = 5), and other type of surgery (n = 11). Collection size was less than 5 cm in diameter in 10 (31%), between 5 and 10 cm in 17 (53%) ,and more than 10 cm in 5 (16%) patients. The median time from surgery to endoscopic drainage was 38 days (range 6–360). Eight (25%) patients underwent endoscopic guided drainage whereas 24 (75%) * Gianfranco Donatelli
[email protected] 1
Unité d’Endoscopie Interventionnelle, Ramsay Générale de Santé, Hôpital Privé des Peupliers, 8 Place de l’Abbé G. Hénocque, 75013 Paris, France
2
Department of Digestive Surgery, Institut Mutualiste Montsouris, 75014 Paris, France
3
Digestive Endoscopy and Gastroenterology Unit, A.O. Istituti Ospitalieri di Cremona, Cremona, Italy
patients underwent EUS-guided drainage. Technical success was 100% and clinical success was achieved in 30 (93.4%) after a mean follow-up of 13.5 months (1.2–24.8). Overall complication was 12.5% including four patients who bled following trans-gastric drainage treated with conservative therapy. Conclusions The present series suggests that endoscopic transmural drainage represents an interesting alternative in the treatment of post-operative collection when percutaneous drainage is not possible or fails. Keywords Fluid collection · EUS drainage · Surgical complication · Double pigtail plastic stent · Internal drainage · Intra-abdominal collection · Pseudocyst · Wall off necrosis Post-operative collections are a recognized source of morbidity and mortality after abdominal surgery. Traditionally, such collections have been drained by re-exploration [1], thus adding a significant morbidity. In the last few decades, Image-guided percutaneous drainage has become the common practice for the management of intra-abdominal fluid collections. It has been proved to be a safe and effective option for the treatment of such complications. The advantages of image-guided percutaneous drainage include avoidance of general anesthesia, reduced hospitalization, and limited trauma to the surrounding tissue [2]. However, percutaneous drainage is not always feasible. EUS became the standard treatment for pancreatic pseudocysts because of very good proximity between the stomach and the collections [3]. Recently, several reports have described a variety of drainage procedures involving liver [4], subphrenic [5], and pelvic abscesses [6] as well as bilomas [7] and infected gallbladders [8]. EUS has also been successfully used for
13
Vol.:(0123456789)
the drainage of peripancreatic fluid collections after distal pancreatectomy [9, 10]. The advantages of EUS-guided drainage include obviating the need for an external drain, minimizing the risk of pancreatic fistulas, and preventing fluid and electrolyte losses [11]. Nevertheless, optimal timing and accurate indications of endoscopic drainage have not yet been widely investigated. In this series, we assessed the role of endoscopic and EUS-guided drainage in the management of post-operative fluid collections after abdominal surgery.
Patients and methods Population This is a retrospective single center review of prospectively collected data including all consecutive patients referred to our tertiary care hospital for transmural endoscopic drainage of post-operative collection from January 2014 to January 2017. All patients were referred to our unit from other hospitals with a clinically evident infected collection (fever, increase of leukocytosis, and CRP) or for abdominal pain, dysphagia, or vomiting. Fluid collections resulted from different type of upper and lower gastrointestinal surgery and gynecological procedures. CT scan was routinely carried out before the procedure in order to adequately evaluate the localization, extent, characteristics (fluid versus necrotic versus hematoma and well encapsulated versus not encapsulated), and anatomical relationship with adjacent organs and vascular structures of the collection (Fig. 1A–C). Transmural endoscopic drainage was performed only for intra-abdominal encapsulated fluid collections adjacent or close (maximum distance of 10 mm) to the gastrointestinal (GI) wall with a minimum diameter of 4 cm. In case of not capsulated collection, diffuse peritonitis surgery was preferred.
Surg Endosc
If cross-sectional imaging demonstrated bulging of the collection on the GI wall, a direct endoscopic approach was attempted, at the endoscopist’s discretion, whereas EUSguided approach was performed for non-bulging abdominal collections. Informed consent, including the possible need for multiple endoscopic sessions, was obtained from all patients. The study was approved by Institutional Review Board for Human Research. Endoscopic procedures All procedures were carried out under general anesthesia and oro-tracheal intubation with the patient in supine position. A standard duodenoscope (TJF160VR, TJF180, Olympus®, Tokyo, Japan) or a therapeutic linear array echoendoscope (UCT180, O lympus®, Tokyo, Japan) was used with CO2 insufflation. Endoscopic puncture of the bulging collection was performed using a triple-lumen needle knife cientific®) under direct view of the (MicroKnife®, Boston S gastric, duodenal, or rectal wall. Puncture of the collection was confirmed by fluid flowing back in the catheter lumen and by its opacification under fluoroscopic control. Samples of the fluid were collected for bacteriological, biochemical, and cytological analysis. Under fluoroscopy guidance, a long guidewire was advanced into the cavity and looped in the collection. Following that a fistula between the GI lumen and the fluid collection wall was created using a 10 FR cystotome (Gflex® Europe, Nivelles-Belgium) introduced over the guidewire. Endocut current was applied and one or two 10 or 7Fr, 3 or 5 cm long double pigtail plastic stents were deployed (Visio® Gflex Europe, Nivelles-Belgium -Advanix®, Boston Scientific®) according to the size of collection. EUS-guided drainage localization and evaluation of the fluid collection was performed with a linear echoendoscope. Puncture site was evaluated according to the distance of the collection from the GI wall, its relationship with surrounding structures, and the absence
Fig. 1 A–C CT scan showing an infected intra-abdominal fluid collection (red arrow) bigger than 5 cm in diameter following IPMN surgical enucleation. The collection is close to the gallbladder inducing compression of the duodenum. (Color figure online)
13
Surg Endosc
Fig. 2 A EUS evaluation of a periduodenal infected fluid collection. Color Doppler mode shows no interposition of vessels during EUS puncture. B 19 Gauge needle puncture of the collection
of intervening vessels (Fig. 2A, B). The collection was entered by a 19 G Needle (19A C ook ®, 19Flex Boston Scientific ®, Massachusetts, USA). After fluid sampling contrast opacification and angled guidewire (450 mm, Hydra®, Boston Scientific®, Massachusetts, USA) insertion, the 19 G needle was removed. A fistula was created by endoflex 6 or 8 Fr ( GFLEX®, Gflex Europe, NivellesBelgium), followed by hydrostatic dilation of the fistula tract ( Hurricaine ®, Boston S cientific ®, Massachusetts, USA). A second guidewire was easily inserted inside the cavity, if needed. At first a 7 Fr double pigtail was deployed (maximum diameter accepted in the EUS working channel of 3.8, with two guidewire in place) and only after removing the first guidewire a 10 Fr stent was delivered (Fig. 3A–C). The patients were discharged the same day of the procedure and oral diet was re-started the day after the procedure.
Studied criteria and post‑operative outcomes Patient features, collection characteristics, operative details, technical, and clinical success and complications rate were assessed. All patients underwent CT scan evaluation 3 weeks after index endoscopy whereas EUS was scheduled, if no complications occurred, after 3 months. Technical success was defined as successful deployment of double pigtail stents inside the collection without intraoperative or early complications (
