Mo1667 WATER EXCHANGE ALONE (WE) AND WATER EXCHANGE WITH CHROMOENDOSCOPY (CWE) SIGNIFICANTLY ENHANCED DETECTION OF PROXIMAL COLON SESSILE SERRATED LESIONS (SSLS) COMPARED WITH AIR INSUFFLATION (AI) IN A RCT

Abstract

Proximal colon hyperplastic polyps (HP), with a different pathology compared to distal HP, are more closely related to sessile serrated adenoma/polyps (SSA/P) (CGH 2013;11:760). Collectively proximal sessile serrated lesions (SSLs: SSA/P and HP) are linked to proximal colon cancers through a different pathway than conventional adenoma-carcinoma sequence. Predominantly flat, proximal SSLs are more difficult to identify by endoscopy with AI; and as missed lesions, are believed to contribute to interval cancers in the proximal colon. We completed a RCT (NCT01607255) comparing sedated screening colonoscopy performed with chromoendoscopy and water exchange (CWE), water exchange alone (WE) and air insufflation (AI). In a 2017 ACG late breaking abstract we reported that there were no differences in patient demographics (age and BMI), cecal intubation success rate, overall and proximal colon adenoma detection rate (ADR) (primary outcome) amongst groups. Overall and proximal colon ADR were >60% and >50%, respectively, confirming high quality examinations. We have now completed further analysis of the secondary outcome of proximal colon SSLs and procedural variables. 480 patients were randomized to sedated screening colonoscopy performed with CWE (160), WE (161) and AI (159) method. AI was performed with conventional air insufflation. WE was performed with air button turned off and water was infused/suctioned (exchanged) to facilitate scope insertion until cecal intubation. Air insufflation was used on withdrawal to facilitate removal of lesions. CWE used a dilute (0.008%) solution of indigo carmine and the method was otherwise same as WE. Polyps were removed using biopsy forceps, cold or hot snares. Table 1 shows the proximal colon SSLs detection rate was significantly higher for CWE and WE than AI (26.3%, 23.6% and 11.3%, respectively, p=0.002). Compared to AI, insertion time was not prolonged, withdrawal time was shorter, and withdrawal bowel prep score was significantly better, with WE. Single operator and subjects limited to Veterans; SSLs are only surrogate markers of, and not interval cancers. Use of WE or CWE significantly and equally improved detection of SSLs in the proximal colon compared to AI. Since missed proximal colon SSLs may account for proximal colon interval cancers developed through an alternative pathway (DCR 2009;52:1535), the enhanced detection of proximal colon SSLs could reduce risks of interval cancers (AJG 2012;107:1315-30). Without prolonging insertion time, with shorter withdrawal time and significantly better withdrawal bowel prep score, WE alone is sufficient to produce a significant improvement in proximal colon SSLs compared to AI. WE could improve clinical outcome by reducing interval cancers. The prevention of interval cancers by WE deserves to be studied.Tabled 1Table 1. Comparison of proximal SSLs detection rateProcedure characteristicsOverall p-valuesTukey adjusted p-values for pairwise [email protected]CWE (N=160)WE (N=161)AI (N=159)CWE vs WECWE vs AIWE vs AICecal intubation time (SD) min.13.3 (7.8)12.0 (6.3)12.2 (7.8).22**nsnsnsWithdrawal time (SD) min. with adenoma26.2 (8.6)20.2 (8.6)21.8 (10.2)<.001**<.001.003.52Volume water infused for procedure (SD) ml655.5 (513.6)614.0 (478.9)479.2 (457.5).004**.08<.001<.001Overall preparation in2.45 (0.29)2.49 (0.24)2.48 (0.27).41**nsnsnsOverall preparation out#3.33 (0.39)3.48 (0.34)3.22 (0.36)<.001**<.001.03<.001Proximal colon SSLs (HP/ SSA/P) detection rate (%)26.323.611.3.002.58<.001.004CWE=chromoendoscopy + water exchange; WE=water exchange; AI=air insufflation; SSLs=sessile serrated lesions (sessile serrated adenoma/polyp + hyperplastic polyps); Proximal colon: splenic flexure to cecum; SD=standard deviation.*Chi square test and **Wilcoxon rank sum test, p<0.05 is significant; ns=not significant @ only pairwise comparisons for tests for which the omnibus test was significant are required. Open table in a new tab