Faculty Bibliography

BACKGROUND: Although standard colonoscopy is considered the optimal test to detect adenomas, it can have a significant adenoma miss rate. A major contributing factor to high miss rates is the inability to visualize adenomas behind haustral folds and at anatomic flexures. OBJECTIVE: To compare the diagnostic yield of balloon-assisted colonoscopy versus standard colonoscopy in the detection of simulated polyps in a colon model. DESIGN: Prospective, cohort study. SETTING: International gastroenterology meeting. SUBJECT: A colon model composed of elastic material, which mimics the flexible structure of haustral folds, allowing for dynamic responses to balloon inflation, with embedded simulated colon polyps (n = 12 silicone "polyps"). INTERVENTIONS: Fifty gastroenterologists were recruited to identify simulated colon polyps in a colon model, first using standard colonoscopy immediately followed by balloon-assisted colonoscopy. MAIN OUTCOME MEASUREMENTS: Detection of simulated polyps. RESULTS: The median polyp detection rate for all simulated polyps was significantly higher with balloon-assisted as compared with standard colonoscopy (91.7% vs 45.8%, respectively; P < .0001). The significantly higher simulated polyp detection rate with balloon-assisted versus standard colonoscopy was notable both for non-obscured polyps (100.0% vs 75.0%; P < .0001) and obscured polyps (88.0% vs 25.0%; P < .0001). LIMITATIONS: Non-randomized design, use of a colon model, and simulated colon polyps. CONCLUSION: As compared with standard colonoscopy, balloon-assisted colonoscopy detected significantly more obscured and non-obscured simulated polyps in a colon model. Clinical studies in human participants are being pursued to further evaluate this new colonoscopic technology.

BACKGROUND: Lymph nodes (LNs) echofeatures on endoscopic ultrasound (EUS) and concurrent fine needle aspiration (FNA) are alternatives to highly invasive approaches for etiologic diagnosis of mediastinal lymphadenopathy (MLAD). AIMS: To evaluate the efficacy of LNs echofeatures and FNA via EUS to distinguish benign LNs from LNs involved by sarcoidosis, lymphoma, and metastasis in non-lung cancer patients. METHODS: A retrospective review of patients who underwent EUS-FNA for MLAD was performed. Echofeatures of LNs including echogenicity, margins, shape, and LN size were recorded. Final diagnosis was made based on surgical sampling or clinical diagnosis with long-term follow-up. Only patients diagnosed as benign MLAD, sarcoidosis, lymphoma, and metastasis included. Diagnostic value of echofeatures and FNA was evaluated. RESULTS: Included were 162 patients with final diagnosis of benign (68), sarcoidosis (33), lymphoma (20), and metastasis (41). The median LN along axis in the benign group [20.5 mm (6-76)] was significantly shorter than in the metastasis [28 mm (9-82)] and sarcoidosis [27 mm (17-50)] groups (p < 0.05). The median LN short axis in the benign group [11 mm (2-50)] was significantly shorter than in the metastasis [17 mm (5-44)] and lymphoma [16 mm (7-47)] groups (p < 0.05). No other echofeatures showed a discriminant value among the groups. When performing FNA, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of EUS-FNA were 73.7, 100, 100, 72.2, and 84.4 %, respectively. CONCLUSION: Although benign MLAD tend to be smaller than other etiologies, echofeatures of LNs are not reliable etiologic diagnostic approach to MLAD. Therefore, FNA is suggested when feasible. However, due to relatively low sensitivity, LNs with benign FNA results should be subjected to further work-up if they are clinically suspicious.