Su1362 LONG-TERM OUTCOMES OF ENDOSCOPIC ULTRASOUND-GUIDED GASTROENTEROSTOMY USING LUMEN-APPOSING METAL STENTS FOR GASTRIC OUTLET OBSTRUCTION: A 4-YEAR COHORT

Abstract

EUS-guided gastroenterostomy (EUS-GE) using lumen-apposing metal stents (LAMS) is a promising procedure for the treatment of patients with gastric outlet obstruction (GOO). Reported short-term outcomes on EUS-GE include longer patency compared to enteral stents and less peri-procedural adverse events (AE) compared to surgical GE. However, little is known regarding mid- and long-term outcomes. To study long-term favorable outcomes after EUS-GE defined as a composite of technical and clinical success without occurrence of AEs or need for re-interventions for recurrent GOO at 6 and 12 months post-procedure. Patients who underwent EUS-GE for both benign and malignant GOO between 3/2014 and 10/2017 were included. Technical success is defined as successful placement of LAMS to create GE. Clinical success is defined as the ability to tolerate at least a full fluid diet. AE, re-interventions and outcomes after re-interventions were recorded. There were 37 patients with GOO, of which 30 had malignant GOO (81.1%), 18 were males (48.6%), with a median age of 63 years (IQR 53-71). Patients had a median follow-up time of 231 days (IQR 61-471). Median procedural time was 50 minutes (IQR 35.5-78). Those with balloon-assisted EUS-GE technique (n=10, 27%) required longer procedural time compared to those with direct EUS-guided technique (median 92 vs 41 minutes, p= 0.02). Technical success was achieved in 35/37 patients (94.6%). Clinical success was achieved in 30/35 patients (85.7%). Re-interventions were required in 5/30 patients (16.7%) who initially achieved clinical success. Technical failure, clinical failure, re-interventions, and AE are detailed in Table 1. There were 5 LAMS misdeployments (2 patients with technical failure and defects closed by over-the-scope clips; 3 patients had successful rescue EUS-GE during the same session). Among 37 patients who underwent attempted EUS-GE, there was one AE (2.7%) described as gastrocolonic fistula due to delayed LAMS migration. The majority of patients experienced short-term favorable outcomes (25/37 patients, 67.6%), and this remained at 6 months (82.3%) and 1 year (77.8%) (Figure 1). After EUS-GE, GOO score improved from baseline median GOO score of 0 (IQR 0-1) to 3 (IQR 2-4) (P < 0.0001). There were 14 deaths (none related to EUS-GE), all with malignant GOO, with a median time from index EUS-GE of 49.5 days (IQR 29-157). On univariate analysis, there was no particular patient factor, etiology of GOO, site of obstruction or procedural factor that was associated with favorable outcome at 6 months and 1 year. EUS-GE using LAMS for GOO resulted in favorable outcomes in two-thirds of patients at 6 and 12 months after index EUS-GE. Stent misdeployment may occur but can be salvaged in the majority of cases.Tabled 1Table 1: Patients with technical failure, clinical failure, and/or recurrence of gastric outlet obstruction requiring re-interventions.Age/genderEtiology of GOOSite of obstructionFollow-up time (days)InterventionOutcomes of re-interventionsPrimary technical failures81/maleGastric adenocarcinomaEfferent limb536Extent from the stomach to the efferent limb was >10 mm. Over-the-scope clip placed to gastric defect and an uncovered metal stent placed in the efferent limb.Deceased at POD 536 with no reintervention.66/femaleSevere peptic strictureSecond part of the duodenum47815mm balloon dilation of the stricture.No recurrence.Clinical failure with persistence of GOO symptoms after index EUS-GE72/maleDuodenal adenocarcinomaSecond part of the duodenum28Symptoms persisted. POD 19 with new multiple downstream jejunal mass. No interventions performed.Deceased on POD 28 from malignancy.18/femaleSMA syndromeThird part of the duodenum177Symptoms persisted. Received an endoscopic gastrojejunal tube placement on POD 99 after EUS-GE.Dependent on tube feeding.53/femaleMetastatic colon cancerThird part of the duodenum45Symptoms persisted. CT showed no obstruction. No further intervention.Deceased on POD 45 from malignancy.61/maleGallbladder cancerDuodenal sweep1100Symptoms persisted. Improved with endoscopic PEGJ on POD 170.Dependent on tube feeding.84/femalePancreatic ductal adenocarcinomaThird part of the duodenum130Symptoms persisted. Two EGDs on POD 14 and POD 30 showed patent LAMS.Deceased on POD 130 from malignancy.Reinterventions for recurrence of GOO symptoms78/maleChronic peptic duodenitisDuodenal bulb231POD17: Endoscopic placement of gastrostomy for palliation.No improvement. Bacteremia due to parenteral nutrition. Deceased POD 231.68/maleCholangiocarcinomaSecond part of the duodenum144POD 23: Disimpaction of food debris and a placement of a fully covered esophageal metal stent via the lumen of the LAMS for stent occlusion. The coaxial esophageal stent was removed on POD 29.Improved GOOS from 1 to 4. Deceased POD 144.53/maleSMA syndromeThird part of the duodenum1101POD 14: Unsuccessful endoscopic gastrostomy placement.Later diagnosed with rectal cancer. Dependent on parenteral nutrition.62/femaleMetastatic ovarian cancerPrepyloric area547POD 158: LAMS was removed due to clinical improvement. Symptoms recurred requiring fluoroscopic-guided gastrojejunal tube placement on POD 401.Dependent on tube feeding.Adverse event based on ASGE lexicon criteria68/femalePancreatic ductal adenocarcinomaDuodenal bulb843GOO symptoms improved after index EUS-GE. POD 317: CT showed distal migration of LAMS causing gastrocolonic fistula. Treatment for asymptomatic gastrocolonic fistula included stent removal, APC and endosuturing. Graded as moderate severity of AE based on lexicon ASGE criteria.No recurrence of GOO throughout POD 843. Open table in a new tab