regardless of advances in impedance technology, as the pH probe has an unmatched ability to measure acid exposure duration (Am J Gastroenterol 2004;99:1037–1043). The availability of catheter- based technology that can estimate the refluxate—acidic or non- acidic—already provides a refined understanding of the mechanisms for GER symptoms resistant to standard antireflux therapy (Gastro- enterology 2001;120:1599 –1606). However, there remains a basic dilemma for the clinician: deter- mining whether abnormal impedance parameters will predict if a patient’s (refractory) symptom will improve with antireflux surgery, a potential mechanical solution to persisting symptomatic reflux. Si- frim et al attempt to address this issue by using a statistical reflux- associated symptom probability (RASP) test with data obtained from impedance testing (Gut 2005;54:449 – 454). The authors assessed if episodes of cough and reflux events detected by impedance testing could have a cause-and-effect relationship in 22 patients with unex- plained cough thought to be caused by GER. The Wuesten method of symptom association probability was used to determine for each 2-minute segment of the study whether there was evidence for symptoms, reflux, both, or neither and to generate a statistical rep- resentation of whether cough and reflux could have co-occurred just by chance. The use of RASP and impedance testing altered the final conclusion in a third of the patients, refuting a reflux-mediated cause for cough in 3 patients with abnormal conventional pH analysis or with typical GERD symptoms and establishing a reflux-cough rela- tionship in 4 patients with normal conventional pH analysis. The etiology of cough could be designated as reflux related in 45% of their study cohort. Although outcome of further therapy based on this etiologic determination was not addressed, escalation of antireflux therapy would seem appropriate in these subjects. RASP tests augment conventional parameters assessed during an ambulatory pH monitoring study by ascribing a statistical probabil- ity to cause-and-effect associations between symptoms and reflux events (Am J Gastroenterol 2006;101:in press; Clin Gastroenterol Hepatol 2005;3:329 –334). If the probability that symptoms and reflux events could have occurred by chance were to be low, the likelihood of symptoms improving after aggressive management of GERD (including antireflux therapy) would be high. There are lim- ited data suggesting that symptomatic outcome after antireflux sur- gery is better when RASP tests are positive, independent of other parameters of esophageal physiologic testing, in patients undergoing antireflux surgery (Gastroenterology 2002;122:A75). Associating symptoms to reflux events gains further importance in patients with atypical GERD symptoms as alternate causes can contribute to the symptomatic state, resulting in poor symptom relief after antireflux surgery if a conclusive association cannot be documented. Since impedance testing is now being used for evaluation of atypical symptoms refractory to conventional acid suppressive therapy, it seems appropriate that RASP tests be studied to determine if reflux by impedance parameters correlates with symptoms. Using impedance testing, administration of baclofen has been demonstrated to suppress inappropriate lower esophageal sphinc- ter relaxations, and hence reflux events (Aliment Pharmacol Ther 2002;17:1–9). In addition, there may be potential for impedance testing to direct invasive surgical techniques in improving reflux- related symptoms. There are limited data suggesting augmenta- tion of the gastroesophageal junction with fundoplication or en- doscopic gastroplication decreases GER estimated by impedance testing, both acid and nonacid (Thorax 2005;60:521–523, Gas- troenterology 2003;124:A97). However, correlation between re- flux events and symptoms is vital to establish causality before any attempt can be made to predict symptom control with antireflux surgery, particularly in patients with atypical symptoms. Further, the study of impedance testing needs to include other supra- esophageal associations of GER, especially ear, nose, and throat (ENT) manifestations, to determine if aggressive invasive tech- niques to augment the gastroesophageal junction will help prevent and heal these findings attributed to GER. Prospective studies are needed to address symptomatic outcome after antireflux therapy directed by impedance testing before the true value of this new addition to our diagnostic armamentarium can be assessed. CHANDRA PRAKASH, MD SREENIVASA JONNALAGADDA, MD CHROMOENDOSCOPY: A NEW VISION FOR COLONOSCOPIC SURVEILLANCE IN IBD Hurlstone DP, Sanders DS, Lobo AJ, McAlindon ME, Cross SS. (Gastroenterology and Liver Unit, The Royal Hallamshire Hospital, Sheffield, United Kingdom). Indigo carmine-as- sisted high-magnification chromoscopic colonoscopy for the detection and characterisation of intraepithelial neoplasia in ulcerative colitis: a prospective evaluation. Endoscopy 2005; 37:1186 –1192. Patients with long-standing ulcerative colitis are at an increased risk for developing dysplasia and colorectal carci- noma (CRC). This risk approaches 8% by 20 years and 18% by 30 years (Aliment Pharmacol Ther 2004;20(Suppl 4):24 –30, Gut 2001;48:526 –535). Moreover, CRC accounts for as many as 1 in 6 deaths in patients with ulcerative colitis. Patients with extensive Crohn’s colitis are also at increased risk for developing CRC and should undergo surveillance colonoscopy (Gastroenterology 2001;120:820 – 826). At present, despite a lack of evidence from randomized controlled trials (Cochrane Database Syst Rev 2004;2:CD000279), surveillance colonos- copy is the most widely used method to detect dysplasia and cancer in inflammatory bowel disease (IBD) patients (Inflamm Bowel Dis 2005;11:314 –321, Gastroenterology 2004;126: 1634 –1648, Gut 2002;51(Suppl 5):V10 –V12, Gastroenterol- ogy 2003;124:544 –560). However, there are several limita- tions to surveillance colonoscopy, and colonoscopy practices are not uniform (Am J Gastroenterol 1995;90:2106 –2114, Gas- trointest Endosc 2000;51:123–128). Surveillance colonoscopy requires multiple random biopsy specimens be taken, which is tedious, costly, and time consuming. It has been estimated that 33 biopsies are needed to achieve 90% confidence to detect dysplasia if it is present (Gastroenterology 1992;103: 1611–1620). Given these limitations, novel methods to detect dysplasia and cancer in patients with IBD have been sought. Accumulating data, including the recent study by Hurlstone et al, suggest that chromoendoscopy may improve the detec- tion of dysplasia and cancer during colonoscopic surveillance in IBD. In this study, Hurlstone and colleagues examined whether targeted chromoscopy could increase the total number of in- traepithelial neoplastic lesions detected compared with con- July 2006 SELECTED SUMMARIES 323