§Wilcoxon-Mann Whitney test; *Chi-square test; BE = barium esophagram; LES = lower esophageal sphincter; reflux score normal <14.7 Continuous variables are expressed as mean ± standard deviation. 672 A Novel Method for the Diagnosis of Reflux-Related Respiratory Symptoms: Normalization of Reflux-Associated Oxygen Desaturations Following Nissen Fundoplication and Establishment of Normal Values Candice L. Wilshire, Renato Salvador, Boris Sepesi, Stefan Niebisch, Thomas J. Watson, Virginia R. Litle, Christian G. Peyre, Carolyn E. Jones, Jeffrey H. Peters Background: Current diagnostic techniques aimed at establishing gastroesophageal reflux disease (GERD) as the underlying cause in patients presenting with respiratory symptoms are poor. We previously reported preliminary data suggesting that quantifying the association between reflux events and oxygen desaturation may be a useful discriminatory test. The aim of this study was to further refine the proof of principle by assessing whether antireflux surgery normalizes reflux-associated desaturations and to establish normal values. Methods: Forty seven patients with GERD-related respiratory symptoms, 10 with typical symptoms, and 11 normal subjects underwent simultaneous 24-hour multichannel intraluminal impedance (MII)-pH and pulse-oximetry monitoring. Eight patients returned for post-Nissen studies. Acid reflux episodes were defined as pH<4 5cm (distal) or 20cm (proximal) above LES and non-acid episodes as a drop ≥50% from baseline in impedance 3, 5, 7 or 9cm above LES (distal) and 15 or 17cm above LES (proximal). Oxygen (O2) desaturation events were defined as a drop in O2 saturation <90%, or a decrease ≥6%. Reflux-associated desaturation (RAD) was defined as the first O2 desaturation event occurring within a 5-minute interval following a reflux episode. Values are expressed as median ±IQR. Results: Patients with typical symptoms had a median of 65, those with respiratory symptoms 64 and normal subjects 26 distal reflux events/24 hours. There was no significant difference in reflux events extending proximally in the 3 groups; 47% (380/804) typical, 45% (1411/3166) respiratory and 41% (114/276) in normal subjects. The number of distal reflux events associated with O2 desaturation was significantly greater in patients with respiratory symptoms 14 (9-20) than those with typical symptoms 5 (1-6; p<0.001) or normal subjects 2 (1-5; p<0.001). This was also true for the number of proximal RADs: 7 (4-13) in patients with respiratory symptoms versus 2 (0-3; p<0.001) with typical symptoms and 1 (0-2; p<0.001) in normals. Repeat study in 8 post-Nissen patients showed marked improvement with RADs approaching those of normal subjects in 6/8; 20 (9-20) distal pre-operative versus 3 (2-5; p=0.05) postoperative; and 12 (2-15) proximal pre-operative versus 2 (0-2) post-operative. Two postoperative patients were found to have recurrent GERD; minimal improvement and/or worsening in the number of RADs were identified in each patient. Using a threshold of 95th %tile of normal subjects, the number of RADs equaled or exceeded normal in 81% (38/47) of patients with respiratory symptoms. Conclusions: These data provide further proof of principle that measurement of the association between reflux events and oxygen desaturation may be a useful discriminatory test in GERD patients presenting with primary respiratory symptoms, and may predict response to antireflux surgery.
* Significant factors BDI = bile duct injury Distribution of Injuries
670a Gastroesophageal Reflux Disease (GERD) and Antireflux Surgery (ARS). What is the Proper Preoperative Work-up? Brian L. Bello, Marco Zoccali, Roberto Gullo, Arunas E. Gasparaitis, Mustafa Hussain, Fernando A. Herbella, Marco G. Patti Background: Many surgeons feel comfortable performing ARS on the basis of symptomatic evaluation, endoscopy and esophageal manometry, while a pH monitoring is seldom obtained. Aims: To analyze the sensitivity and specificity of symptoms, barium esophagogram, endoscopy and manometry as compared to pH monitoring in the preoperative evaluation of patients for ARS. Patients and Methods: 134 patients referred for ARS with a diagnosis of GERD based on symptoms, endoscopy, barium esophagogram and manometry. Ambulatory 24 hour pH monitoring was performed preoperatively in all of them. Results: Based on the presence or absence of GERD on pH monitoring, patients were divided into two groups: GERD+ (n = 78) and GERD- (n = 56). The groups were compared with respect to the incidence of symptoms, presence of reflux and hiatal hernia on esophagogram, endoscopic findings, and esophageal motility. There was no difference in the incidence of symptoms between the two groups (p=NS). Within the GERD+ group, 37 patients (47%) had reflux at the esophagogram and 41 (53 %) had no reflux. Among the GERD- patients, 17 (30%) had reflux and 39 (70%) had no reflux. Therefore, the sensitivity of esophagogram was 47% and the specificity was 70%. A hiatal hernia was present in 40% and 32% of patients respectively. Esophagitis was found at endoscopy in 16% of GERD+ patients and in 20% of GERD- patients, accounting for a sensitivity of 16% and a specificity of 80%. Esophageal manometry showed no difference in the pressure of the lower esophageal sphincter or the quality of peristalsis. Ambulatory 24-hour pH monitoring clearly separated the 2 groups (Table). Conclusions: The results of this study showed that: (a) symptoms were unreliable in diagnosing GERD; (b) the presence of reflux or hernia on esophagogram did not correlate with reflux on pH monitoring; (c) endoscopy had low sensitivity and specificity; and (d) manometry was mostly useful for positioning the pH probe and rule out achalasia. We conclude that ambulatory pH monitoring should be routinely performed in the preoperative work-up of patients suspected of having GERD in order to avoid useless ARS.
Staging and Survival of Resected Hilar Cholangiocarcinoma: an Analysis of 80 Consecutive Patients Victor M. Zaydfudim, Clancy J. Clark, Michael L. Kendrick, Florencia G. Que, Kaye M. Reid Lombardo, John H. Donohue, Michael B. Farnell, David M. Nagorney Introduction: Predicting long-term survival in hilar cholangiocarcinoma is difficult. The revised AJCC staging system has not been extensively evaluated and may not correlate with clinical outcomes. An alternative staging system which incorporates factors related to local tumor extent, including portal vein invasion and lobar atrophy, has been proposed. The aim of this study was to evaluate current staging systems for hilar cholangiocarcinoma and identify clinical factors associated with improved survival. Methods: In this retrospective cohort study, clinical and pathologic characteristics were obtained for all resected patients with Bismuth-Corlette Type IIIa and IIIb hilar cholangiocarcinoma from 1993 to 2011. Patients were stratified by the 7th edition AJCC TNM staging parameters and by the modified Blumgart staging system which includes portal vein invasion and presence of lobar atrophy. Univariate and multivariate analyses were used to test effects of clinicopathologic factors and staging systems on overall survival. Results: Eighty consecutive patients (median age 64 years (range 36-82), 64% male) underwent an anatomic hepatectomy with a bile duct resection and reconstruction for Bismuth-Corlette IIIa (51%) and Bismuth-Corlette IIIb (49%) cholangiocarcinoma. Margin negative resection was achieved in 94% of resections; 30-day mortality was 10%. Median follow-up was 26 months (range 0-181 months) with overall median survival of 34 months. Twenty-three percent of the patients had welldifferentiated cholangiocarcinoma. The AJCC staging system stratified patients into following groups: T1-26%, T2-58%, T3-16%; N0-61%, N1-39%; Stage I-20%, Stage II-30%, Stage III-50%. None of the patients had distant metastases at the time of resection. Kaplan-Meier estimates did not demonstrate an association between survival and AJCC staging parameters (all p≥0.121). Blumgart staging system stratified patients into following groups: Blumgart T1 - 58%, Blumgart T2 - 41%; one patient had a Blumgart T3 cholangiocarcinoma with invasion into portal vein bifurcation. Univariate analyses demonstrated an association of survival with tumor grade (p=0.033) and Blumgart T-stage (p=0.010). One- and five-year survival estimates for Blumgart T1 and Blumgart T2/T3 were 86% and 47% vs. 74% and 17% (p=0.010). After adjusting for tumor grade, Blumgart T2/T3 stage correlated with increased likelihood of mortality (HR=1.93, 95%CI: 1.09-3.42, p=0.024) Conclusions: While the current AJCC TNM staging system did not predict survival in the current study, the Blumgart staging system which emphasizes portal vein invasion and lobar atrophy predicted overall survival independent of other clinical and pathologic factors. Inclusion of lobar atrophy into the T classification might improve accuracy of the AJCC system, and help define prognosis in patients with hilar cholangiocarcinoma.