Faculty Bibliography

BACKGROUND AND AIMS: The need for transpapillary drainage (TP) in patients undergoing transmural drainage (TM) of pancreatic fluid collections (PFC) remains unclear. The aims of this study were to compare treatment outcomes between patients with pancreatic pseudocysts undergoing TM versus combined (TM and TP) drainage (CD) and to identify predictors of symptomatic and radiologic resolution. METHODS: Retrospective review of 375 consecutive patients with PFC who underwent endoscopic ultrasound (EUS)-guided TM from 2008 to 2014 at 15 academic centers in the United States. Main outcome measures included TM and CD technical success, treatment outcomes (symptomatic and radiologic resolution) at follow-up, and predictors of treatment outcomes on logistic regression. RESULTS: A total of 375 patients underwent EUS-guided TM drainage of a PFC, of which 174 were pseudocysts. TM alone was performed in 95 (55%) and CD in 79 (45%) pseudocysts. Technical success: TM 92(97%) versus CD 35(44%) (P=0.0001). There was no difference in adverse events between the TM (15%) and CD (14%) cohorts (P=0.23). Median long-term (LT) follow-up after transmural stent removal was 324 days (interquartile range, 72-493 days) for TM and 201 days (interquartile range: 150-493 days) (P=0.37). There was no difference in LT symptomatic resolution (TM: 69% vs CD: 62%, P=0.61) or LT radiologic resolution (TM: 71% vs CD: 67%, P=0.79). TP attempt was negatively associated with LT radiologic resolution of pseudocyst (OR, 0.11; 95% CI, 0.02-0.8; P=0.03). CONCLUSIONS: TP has no benefit on treatment outcomes in patients undergoing EUS-guided TM of pancreatic pseudocysts and negatively impacts long-term resolution of PFC.

Introduction: EUS-guided transmural drainage (TM) of pancreatic pseudocysts is well established. It . remains unclear if TM alone is sufficient treatment in pseudocysts containing solid debris. Aim: To compare clinical outcomes in patients who undergo TM of pancreatic pseudocysts that are debris-free (PDF) vs those that contain solid debris (PSD). Methods: Multi-center retrospective cohort study across 15 centers of patients who underwent TM of pancreatic pseudocysts during a 6-year period (2008-2014). Outcome measures included technical success, symptom resolution, radiologic resolution, endoscopic re-intervention and adverse events (AE). PSD was defined as a pseudocyst containing solid material at time of index EUS TM drainage. Patients with obvious walled-offnecrosis on index imaging were excluded. Technical success was defined as drainage with successful placement of > of 1 transmural stent. Short-term (ST) followup was designated as > 2 weeks following TM, but before stent removal. Long-term (LT) follow-up was defined as > 2 weeks following stent removal. Symptom resolution was defined as the complete absence of symptoms at follow-up. Radiologic resolution was defined as the absence of the pseudocyst on repeat imaging. Re-intervention was defined as the need to repeat endoscopic drainage within a 4-week period. Results: A total of 142 patients underwent TM: 90 PDF and 52 PSD. There were no differences in patient characteristics, indication for drainage, size and location of pseudocyst between the two groups (Table 1). Technical success: PDF 87/90 (97%) vs PSD 51/52 (98%, p=1.00). Endoscopic re-intervention rate was similar between the two groups: PDF 5/90 (6%) vs PSD 6/52 (12%, p=0.21). One patient with PSD underwent endoscopic necrosectomy following initial TM. Median ST and LT follow-up were similar between the two cohorts (Table 2). There was no statistically significant difference in ST symptom resolution rate and LT outcomes in the two groups. ST radiologic resolution was higher in PDF 45/68(66%) compared to PSD 21/41(51%; p=0.009). Overall, procedure-related AE occurred in 13 (14%) in the PDF and 9 (17%) in the PSD group (P=0.64). Conclusion: There is no difference in technical success, ST or LT clinical outcomes in performing TM of PSD, when compared to patients with PDF. Thus, TM alone can be considered a suitable initial intervention in patients with PSD, without increased risk of infectious complications or need for endoscopic debridement. (Table Presented)

AIM: To ascertain fine needle aspiration (FNA) techniques by endosonographers with varying levels of experience and environments. METHODS: A survey study was performed on United States based endosonographers. The subjects completed an anonymous online electronic survey. The main outcome measurements were differences in needle choice, FNA technique, and clinical decision making among endosonographers and how this relates to years in practice, volume of EUS-FNA procedures, and practice environment. RESULTS: A total of 210 (30.8%) endosonographers completed the survey. Just over half (51.4%) identified themselves as academic/university-based practitioners. The vast majority of respondents (77.1%) identified themselves as high-volume endoscopic ultrasound (EUS) (> 150 EUS/year) and high-volume FNA (> 75 FNA/year) performers (73.3). If final cytology is non-diagnostic, high-volume EUS physicians were more likely than low volume physicians to repeat FNA with a core needle (60.5% vs 31.2%; P = 0.0004), and low volume physicians were more likely to refer patients for either surgical or percutaneous biopsy, (33.4% vs 4.9%, P < 0.0001). Academic physicians were more likely to repeat FNA with a core needle (66.7%) compared to community physicians (40.2%, P < 0.001). CONCLUSION: There is significant variation in EUS-FNA practices among United States endosonographers. Differences appear to be related to EUS volume and practice environment.

Objectives: Colonoscopy is the gold standard for colon cancer screening and prevention. Several new technologies have recently emerged in order to better assist gastroenterologists in their evaluation of the colonic mucosa to find and remove polyps during screening and surveillance colonoscopies. The ARC EndoCuff is a disposable plastic device with soft, hair-like projections that is placed at the tip of the colonoscope prior to the procedure (Image 1.). The device helps to stabilize the tip of the colonoscope and assists in flattening the folds of the colonic mucosa in order to better evaluate proximal colonic folds. We performed a retrospective chart review of patients who underwent screening and surveillance colonoscopy with and without EndoCuff. The aim of our study was to assess whether EndoCuff was more effective at detecting polyps and adenomas. Methods: Patients were retrospectively analyzed at a tertiary care medical center, urban and suburban outpatient practices from September 2013 thru November 2013. Only screening and surveillance colonoscopies were included. Based on previous learning curve results, the first 4 EndoCuff procedures for each operator were excluded. A comparison group was matched to Endocuff cases based on indication. Colonoscopy reports were retrospectively reviewed to obtain demographic information, total polyps and adenomas removed. This was used to calculate overall ADR and ADR by colon location (right vs. left). Paired t-tests were performed to assess for significance between EndoCuff and comparison group with respect to total polyps and adenomas removed and ADR. Results: 165 patients received EndoCuff and 153 patients underwent standard colonoscopy. Average polyps detected per patient in the EndoCuff group was 1.31 vs. 0.82 in standard colonoscopy (p=<0.001). Similar results were also observed in average adenoma per patient 0.8 vs 0.38 (p=<0.001). The ADR was highest amongst patients who underwent EndoCuff compared to standard colonoscopy (46.6% vs. 30.0%, p=0.!

Objectives: Colonoscopy is the gold standard for colon cancer screening, by detecting and removing adenomatous polyps. However, polyps can be missed in the proximal mucosal folds with traditional forward viewing colonoscopy. The ARC EndoCuff is a disposable attachment placed on the tip of the colonoscope. With soft, hair-like projections, the EndoCuff (Image 1.) helps to flatten colon mucosal folds during scope withdrawal allowing for increased mucosal inspection. A new technique often has to allow for proficiency and the aim of our study is to assess the learning curve for EndoCuff by comparing adenoma detection rate (ADR) with successive weeks of operator experience. Methods: We retrospectively analyzed patients who underwent colonoscopy with Endocuff for any indication at an outpatient urban practice over a 3-week period. ADR was calculated for each week and weeks 2 and 3 were compared to week 1, which served as our control. Total number of EndoCuff procedures in week 1 were divided by total number of operators to determine the average procedures needed to overcome the learning curve. Colonoscopy reports were retrospectively reviewed and paired t-tests were performed to assess for significance between ADR for each week. Results: A total of 58 patients underwent EndoCuff during the 3-week period. A total of 15, 21, and 22 patients had procedure performed in weeks 1,2, and 3, respectively (Table 1.). A total of 4 operators participated in week 1 and number of procedures ranged from 2 to 5 with a mean of 3.75. ADR for week 1 was 20% which increased to 54.5% in week 2 (p=0.03) and reached its peak at 63.6% in week 3 (p=0.004). Given the significant improvement in week 2 compared to week 1, learning curve was approximated to be 4 procedures based on a total of 15 procedures performed in week 1 divided amongst four operators. Conclusions: Our results suggest that a learning curve does exists for EndoCuff and that ADR significantly increases with operator experience after 4 procedures. Overall, End!