- Email: [email protected]
W1673 Advanced Disease Not a Risk Factor for Delayed Gastric Emptying After Pancreaticoduodenectomy (Whipple Procedure)
or 16% of all the patients experienced intraoperative blood loss of more than 1500ml. During operation, 16 patients (7%) received red cell blood transfusion. Two patients (0.9%) were returned to the operating room for postoperative hemorrhage. Univariate analysis identified body mass index (BMI, p=0.001), tumor size (p=0.002), resected liver volume (p<0.001), serum total bilirubin (p¬=0.053), serum prothrombin time (p=0.044), serum glutamate pyruvate transaminase (p=0.041), major hepatectomy (p<0.001), wide incision with right thoracotomy (p=0.085), and additional operative procedure (p=0.034) as risk factors for massive intraoperative blood loss; multivariate analysis identified major hepatectomy (p<0.001) and BMI (p=0.005) as independent risk factors of blood loss of more than 1500 ml. Conclusions: Autologous blood storage might be indicated in patients with these predictive factors. Laparoscopic hepatectomy should not be considered in these patients, because of increased intraoperative blood loss.
W1671 Prognostic Impact of Dihydropyrimidine Dehydrogenase Expression in Adjuvant Gemcitabine Plus S-1 Chemotherapy After Surgical Resection for Pancreatic Adenocarcinoma Naru Kondo, Yoshiaki Murakami, Kenichiro Uemura, Yasuo Hayashidani, Takeshi Sudo, Yasushi Hashimoto, Hiroki Ohge, Taijiro Sueda Objective: Although the prognosis in patients with pancreatic adenocarcinoma remains poor, adjuvant gemcitabine plus S-1 chemotherapy (GEM + S-1) after surgical resection for pancreatic adenocarcinoma has been shown to improve survival. S-1 is a novel oral fluoropyrimidine combination including tegafur (a prodrug of 5-fluorouracil; 5-FU), dihydropyrimidine dehydrogenase (DPD) inhibitor (5-chloro-2,4-dihydroxypyrimidine), and orotate phosphoribosyltransferase (OPRT) inhibitor (potassium oxonate). To clarify the relationship between expression of intratumoral enzymes related to the metabolism of 5FU and its derivatives and response to adjuvant chemotherapy with GEM + S-1 for pancreatic adenocarcinoma, we evaluated thymidylate synthase (TS), DPD, and OPRT expression immunohistochemically in resected pancreatic adenocarcinoma tissues. Methods: Polyclonal antibodies were used to immunostain sections of 106 formalin-fixed paraffin-embedded specimens of pancreatic adenocarcinoma resected between 1998 and 2009. The relationship between intratumoral TS, DPD, and OPRT expression and prognosis was evaluated statistically. Results: Out of 106 patients, 75 (70.1%) received adjuvant GEM + S-1 chemotherapy. High intratumoral TS, DPD, and OPRT expression was present in 68 (64.1%), 39 (36.8%), and 70 (66.0%) cases, respectively. Comparison of overall survival between High and Low intratumoral TS or OPRT expression revealed no significant difference regardless the application of adjuvant GEM + S-1 chemotherapy. In the GEM + S-1 (+) group, overall survival was significantly longer in the Low DPD subgroup than in the High DPD subgroup (hazard ratio [HR], 0.661: 95% confidence interval [CI], 0.447 - 0.978; P = 0.031), whereas in the GEM + S-1 (-) group, there was no significant difference between the High DPD and Low DPD subgroups. Moreover, in the High DPD group, there was no significant difference in overall survival between the GEM + S-1 (+) and GEM + S-1 (-) subgroups, whereas in the Low DPD group overall survival was significantly higher in the GEM + S-1 (+) subgroup by univariate analysis (HR, 0.456; 95% CI, 0.295 - 0.683; P < 0.001). Conclusion: Low intratumoral DPD expression was associated with increased overall survival in patients with pancreatic adenocarcinoma who received adjuvant GEM + S-1 chemotherapy. DPD is a relevant predictive marker of benefit from adjuvant GEM + S-1 chemotherapy in patients with resected pancreatic adenocarcinoma.
W1668 The use of Intraoperative Ablation Extends the Limits of Potentially Curative Treatment for Recurrent Colorectal Liver Metastases Anand Govindarajan, Dean Arnaoutakis, Michael D'Angelica, Peter J. Allen, Ronald P. DeMatteo, Leslie H. Blumgart, William R. Jarnagin, Yuman Fong
SSAT Abstracts
Introduction: Most patients undergoing liver resection for colorectal liver metastases develop recurrent hepatic disease. In this setting, complete resection is often technically difficult or impossible. This study evaluates the role of intraoperative ablation as an adjunct to resection in patients with recurrent colorectal liver metastases. Methods: A retrospective cohort study was performed using a prospectively collected database. The study population included patients who underwent curative reoperative surgery for recurrent colorectal liver metastases from 1991-2009. Patients were categorized based on whether they were treated with resection alone (RES) or in combination with ablation (COMB). The groups were compared with respect to tumor factors, perioperative factors and overall survival. Multivariable Cox regression models were used to examine overall survival while adjusting for confounding variables. Results: A total of 234 operations were performed on 104 patients. The median number of operations performed per patient was 2 (range 2-4). There was an increase in the proportion of patients treated with resection and ablation from 0% (1991-1994) to 48% (2007-2009). There was also an increase in the proportion of patients undergoing surgery with abnormal liver parenchyma (steatosis, fibrosis or inflammation) from 37.5% (1992-1994) to 52.2% (2007-2009). Patients undergoing resection and ablation had a greater tumor burden (median 3 metastatic lesions vs. 1, p<0.0001) and higher baseline clinical risk scores (median 3 vs. 2, p=0.041) than patients undergoing resection alone. Patients undergoing resection and ablation had lower intraoperative blood loss compared to patients undergoing resection alone (312±48 mL vs. 860±99mL, p=0.0060). At a median followup of 54 months, 5-year overall survival from the time of first liver metastasis was 47.4%, with no significant difference between the RES and COMB groups in univariate (5-year survival 46.6% RES vs. 59.6% COMB, p=0.86) or multivariable analysis (HR: 0.94, 95%CI: 0.30 to 2.93, p=0.91). Conclusion: In patients with recurrent colorectal liver metastases, the combination of intraoperative ablation with resection extends the limits of potentially curative therapy to include patients with advanced disease that might otherwise not be amenable to surgical intervention. The combination of resection and ablation yields survival results that are equivalent to that of resection alone but with reduced surgical morbidity.
W1672 Interventional Radiology Assisted Endoscopic Transgastric Peripancreatic Fluid Collection Drainage Abdulrahim Alawashez, James Ellsmere BACKGROUND: Many centers advocate the use of EUS to perform transgastric drainage of peripancreatic fluid collections (PFCs). However most endoscopists do not perform EUS. We report our experience using interventional radiology (IR) placed transgastric drains as a first stage to a subsequent endoscopic transgastric PFC drainage. OBJECTIVE: To evaluate the technical and treatment success and safety profile of IR assisted endoscopic transgastric PFC drainage. PATIENTS: This study involved 11 consecutive patients referred for endoscopic drainage of PFCs over a 2-year period. INTERVENTIONS: The first stage is performed by IR under CT guidance. A suitable window to the PFC through the stomach is determined. An 18-gauge trochar needle is then advanced through the anterior and posterior wall of the stomach into the PFC. Following removal of the inner stylet, a stiff Amplatz wire is advanced. The tract is dilated and over the wire, an 8.5 F multipurpose cathether is positioned in the PFC. The second stage is performed under endoscopic and fluoroscopic guidance. The drain is removed. The posterior gastrotomy is cannulated using a wire guided sphincterotome. The tract is dilated to 15mm using an endoscopic dilating balloon. Two 10F x 5cm double pigtail stents are left across the posterior gastrostomy. The anterior gastrostomy is closed with endoscopic clips. RESULTS: 11 patients (6 males and 5 females) with mean age of 61 (range 38- 79 years) underwent IR assisted endoscopic transgastric PFC drainage (7 pseudocyst, 4 abscess) over a 2-year period. The PFCs were gallstone related in 7 patients, post surgical in 2 patients and idiopathic in 2 patients. The mean size of the PFCs was 12 cm (range 6-19 cm) in its largest dimension. The procedures were technically successful in all 11 patients. There were 3 complications (pneumothorax that was treated with a chest tube, uncomplicated pneumoperitoneum that was confirmed by diagnostic laparoscopy, delayed bleeding from posterior gastrostomy requiring endoscopic hemostasis). All patients had a successful resolution of their PFCs on follow-up imaging. The median length of hospital stay was 11.5 days (range 1-86 days). At a mean follow-up period of 14 months (range 2-29 months), all patients were doing well. CONCLUSIONS: IR assisted endoscopic transgastric PFC drainage is technically feasible and safe and is associated with favourable clinical outcomes.
W1670 Underutilization of Surgical Treatment of Hepatocellular Carcinoma in the Medicare Population Shimul A. Shah, YouFu Li, Sing Chau Ng, Andrew Burr, Jennifer F. Tseng The incidence of hepatocellular carcinoma (HCC) is increasing in the United States. Due to factors such as portal hypertension, tumor biology and a donor organ shortage the care of these patients remain highly specialized and complex. Multiple treatment options are available for HCC but their use and utility remains unknown. Methods: Using SEER-Medicare linked data, we identified 8570 patients diagnosed with HCC between 1991 and 2007. Treatments included resection, liver transplantation (LT), ablation or transarterial chemoembolization (TACE). Patients who received no or palliative-only treatment were grouped (NoTx). Demographic, clinical and tumor factors were examined as determinants of therapy. Univariate and multivariate analyses were performed to determine predictors of overall survival. Results: Median age was 74.9 years. HCC therapies included resection (n=745; 8.7%), LT (n=120; 1.4%), ablation (n=312; 3.6%), TACE (n=1214; 14.2%) and NoTx (n=6179; 72.1%). Of the 8570 patients, 68% were coded as White, 11.3% Asian, 3.8% Hispanic, and 8.4% Black. Only 13.7% of patients underwent localized therapy that was potentially curable (TX group; resection, ablation or LT). If tumors were <5cm in largest dimension, then 64% of patients underwent TX. In the NoTx group, 49% were not cirrhotic, 36% had tumors < 5cm and 21% were tumor stage I or II. Median survival for all HCC patients increased over time (2002-2005: 116 months, 1999-2001: 96 months, 1991-1998: 87 months). In multivariate regression analysis, patients who received any modality of treatment achieved some benefit compared to NoTx. Specifically, adjusted overall survival was greatest in the patients in the TX group: LT (HR 0.35; 95% CI 0.26-0.47), resection (HR 0.33, 95% CI 0.29-0.37) and ablation (HR 0.51, 95% CI 0.42-0.62). Conclusion: In the Medicare population, HCC patients who receive potentially curative therapy such as resection, LT, and ablation experienced a substantial survival advantage over their non-operative peers (NoTx). Despite evidence that many patients had favorable biological characteristics, less than 15% of patients diagnosed with HCC received any treatment. Barriers to treatment and its underutilization must be identified to improve survival in patients diagnosed with HCC in the United States.
SSAT Abstracts
W1673 Advanced Disease Not a Risk Factor for Delayed Gastric Emptying After Pancreaticoduodenectomy (Whipple Procedure) Ronald J. Shonkwiler, Isadora C. Botwinick, John Steele, Gary Yu, John A. Chabot BACKGROUND: Delayed gastric emptying (DGE) is a common cause of morbidity after pancreaticoduodenectomy (PD). DGE can be a source of distress and discomfort to patients, resulting in prolonged hospital stays and subsequent drain on staff and resources. AIMS: We aimed to investigate whether any preoperative or intraoperative variables predict risk of DGE. Ability to predict DGE would potentially enable therapeutic interventions such as insertion of a feeding jejunostomy at the time of PD in high-risk patients. METHODS: We performed a retrospective chart review of 523 patients who underwent PD at our institution between 2000 and 2009; 384 resections were performed since 2005. Data collected included: patient demographics, preoperative symptoms, preoperative chemoradiotherapy, type of
S-900
resection performed (pylorus-preserving PD or standard PD, with or without vascular resection) OR time, estimated blood loss and transfusions, tumor size and histology, lymph node status, postoperative hematocrit and leukocyte count and postoperative NSAID use. We defined DGE as failure to tolerate a regular diet by postoperative day seven. Data was analyzed with alpha=0.05 using Fisher exact test for categorical variables and Mann-Whitney U test for continuous variables. RESULTS: A total of 114 patients with DGE were compared with 134 patients without DGE. Patients who suffered DGE were significantly more likely to have smaller tumors (p=0.015) and fewer positive lymph nodes (p=0.025), while patients without DGE were more likely to have received preoperative chemotherapy (p= 0.005) or preoperative radiation (p= 0.010). A possible explanation for these results is that patients with malignant disease have less normal functioning pancreatic parenchyma and are consequently less at risk for digestive symptoms resulting from pancreatic injury and inflammation. In addition, DGE was associated with an infection during hospitalization (p=0.008.) Hematocrit on the third postoperative day was significantly lower (p=0.019) for patients who experienced DGE , excluding patients who experienced post-operative bleeding. We hypothesize that DGE patients experienced more postoperative third space sequestration and bowel edema, resulting in greater hemodilution when sequestered fluid redistributed back into the vasculature. CONCLUSIONS: Contrary to our expectations, patients with smaller tumors and benign disease were more likely to experience DGE than patients with more advanced, malignant disease. As our ability to diagnose premalignant lesions improves, there will likely be more PD performed for benign disease, with a subsequent increase in the frequency of patients with DGE.
initial stent placement and stent replacement was 70.8 days (± 32). Median survival time for pat. treated only with endoscopic stent implantation (Group 1) was significantly (p<0.001) shorter than that of pat. who were first stented and subsequently treated with hepaticojejunostomy (Group 2) (5.1 vs. 9.4 months). None of the pat. who underwent surgery were rehospitalized for recurrent jaundice. In conclusion, we found that surgical hepaticojejunostomy can be performed with adequate results and an acceptable complications. Considering that biliary stents need to be replaced (mean time to stent replacement 70 days),management by hepaticojejunostomy may be superior to endoscopic stenting, especially for pat. with a life expectancy of greater than 6 months. The operative approach should be especially favored in patients whose disease is first found to be unresectable in the intraoperative setting. W1676 The Role of Procalcitonin in the Early Diagnosis of Postoperative Pancreatic Fistula After Pancreatic Resection Dominique Suelberg, Franziska Beuschel, Christiane Kotschenreuther, Torsten Herzog, Ansgar M. Chromik, Waldemar Uhl Background: Postoperative pancreatic fistula (PF) are frequent major complications after pancreatic surgery. Detection of amylase in drainage-fluid and serum-CRP are well established diagnostic parameters. The importance of Procalcitonin (PCT) in the diagnosis of PF remains to be elucidated. Aim of the study was therefore, to evaluate of PCT in the diagnostic of PF and major complications following pancreatic surgery (PS). Methods: All patients with PS were prospectively collected from Jan. 2009 - Sept. 2009. Serum-PCT, -CRP as well as amylase in drainage-fluid were measured every second postoperative day. Following parameters were analyzed: diagnosis, age, operation, complications (major, minor, secondary), PF (grade A, B and C; according to the ISGPF-definition). Results: During the observation period, n = 107 pat. received PS and developed n = 15 minor, n = 12 major and n = 3 secondary complications. Among major complications, n = 9 PF (8.9%) were diagnosed in the drainage-fluid, thereof n = 3 grade A, n = 2 grade B and n = 4 grade C. CRP and PCT were significantly increased in pat. with PF and other major complications after the 10th postop. day (p ≤ 0.001; Anova). In the detection of PF, the ROC-Analysis for PCT revealed an AUC of 0.736 with a sensitivity of 83% and a specificity of 72% for the cut-off-value of 17.8ng/ml. For CRP the AUC was 0.855 with a sensitivity of 86% and a specificity of 83% for the cut-off-value of 77.1mg/l. For the detection of major complications, the ROC-Analysis of PCT resulted in an AUC of 0.977 with a sensitivity of 100% and a specificity of 93% for the cut-off-value of 20.9 ng/l. For CRP, the AUC was 0.770 with a sensitivity of 73% and a specificity of 84% for the cut-off-value of 77.1 mg/l. Conclusion: Our results demonstrate that PCT is equivalent with CRP in the early diagnosis of postoperative pancreatic fistula, however superior in the diagnosis of all major complications following PS. Therefore we advocate the routine use of PCT and CRP following PS.
W1674 Diagnostic Laparoscopy for Pancreatic Cancer in an MRI Driven Practice: What's it Worth? Elliot Tapper, Bobby Kalb, Diego R. Martin, David Kooby, N. Volkan Adsay, Juan M. Sarmiento Introduction: For many patients with pancreatic cancer, CT is inadequate in determining unresectability; 10-48% of patients deemed resectable receive an unnecessary laparotomy. Accordingly, many groups have studied the role for diagnostic laparoscopy (DL) though none have evaluated it in an MRI driven practice. Methods: All MRI's administered for suspected pancreatic cancer between December 2004 and 2008 were evaluated. Radiographic diagnoses were prospective judged resectability based on the presence of metastases and relationship of the tumor with the surrounding vasculature. Unresectable disease received endoscopic biliary and duodenal stenting. Resectable and borderline disease received Whipples and double bypasses if unresectable intraoperatively. We performed a decision analysis for the cost-effectiveness of incorporating DL. We queried our billing database to render average costs for all inpatients with pancreatic cancer who received Whipples, doublebypasses and double-stenting procedures. We did not include professional fees. The marginal cost of DL was derived from the itemized costs of the materials, space and ancillary staff, presuming routine utilization, no missed metastases, and no complications. Results: Preoperative MRI deemed 94 patients' tumors resectable; 86 agreed to a laparotomy. Six patients were found to have metastases intraoperatively and 15 patients had unresectable disease (vascular involvement or benign pancreatitis) and thus received double-bypass procedures for which the average total cost of the hospitalization was $21,957.18. Whipples were provided to 65 patients at an average cost of $26,122.43. DL would thus be offered to 86 patients. For the 6 patients with metastases, it would be the only operation ($3604.07). This would be added to the cost of endoscopic stenting procedures, which results in an average total cost of hospitalization of $18,451.41. For the patients without metastases, the marginal cost of DL before a laparotomy would be $2651.71, which we added to the total costs above. Conclusions: For DL to be cost-effective, it would have to increase the rate at which we diverted patients to the GI lab for palliative stenting. In our model, DL would increase our costs by $191,072.18, equivalent to the total cost of hospitalization for treating 7 patients with Whipple procedures. Given our rate of missed metastases - 6% - and presuming perfect yield from DL, 15 patients would have unnecessary DL for every patient with occult metastases. For DL to be cost-effective, its intraoperative yield would have to be 70%.
W1677
Background: Determination of fecal elastase-1 is well established in the diagnosis of exocrine pancreatic insufficiency (EPI). However, accelerated gastro-intestinal passage could result in false positive results. Serum-beta-Carotin is an indirect parameter to measure the malassimilation of lipids, but is not determined in the diagnostic of EPI yet. Aim of the study was to evaluate the validity of Serum-beta-Carotin in the EPI-diagnostic pre- and postoperatively. Methods: All patients underwent pancreatic surgery were prospectively collected from Jan.2009-Sept. 2009. Fecal Elastase-1, serum-beta-Carotin and clinical features were analyzed preoperatively and 1 week, 3 and 6 month postoperatively together with surgical procedure and diagnosis. Ergebnisse: During the observation period 125 patients underwent pancreatic surgery. The results of preoperative fecal elastase-1 and serum-beta-Carotin showed significant lower values in pat. with chronic pancreatitis (CP) in contrast to pat. with benign and malign tumors (p≤0.005; Anova). One week postoperatively, serum-beta-Carotin was significantly decreased in pat. after pancreatic head resection (PHR) in contrast to pat. after left pancreatic resection (LPR) or bypass-procedures (BP) (p≤0.05; Anova). In contrast, one week postoperatively fecal elastase-1 was decreased in all groups regardless of the surgical procedure without any differences in the clinical features. The follow-up revealed reduced values for both parameters in pat. with PHR compared to pat. with LPR or BP. Overall, postoperative serum-beta-Carotin was more frequently in normal range than the values of fecal elastase-1 in patients without clinical signs of exocrine insufficiency. Conclusion: Our data demonstrate that serum-beta-Carotin is equivalent with fecal elastase-1 in the diagnosis of exocrine pancreatic insufficiency following pancreatic surgery. Moreover, the validity of serum-beta-Carotin seems to be superior to fecal Elastase-1.
W1675 Palliative Treatment of Obstructive Jaundice in Patients With Carcinoma of the Pancreatic Head or Distal Biliary Tree: Endoscopic Stent Placement vs. Hepaticojejunostomy Marius Distler, Stephan Kersting, Felix Rückert, Frank Dobrowolski, Stephan Miehlke, Robert Grützmann, Hans D. Saeger Only 20-30% of patients with malignant neoplasms of the pancreatic head or distal biliary tree can be cured by surgery. Palliative procedures play an important role in their management. We sought to determine if surgical or non-surgical management was the most appropriate therapy for the treatment of obstructive jaundice in the palliative setting. We retrospectively analyzed the outcomes of our pat. with regard to method of treatment and complications as well as the survival times.The pat. were divided into three groups based on treatment. Pat. in Group 1 underwent endoscopic bile duct stent placement.Group 2 underwent preoperative stenting followed by laparotomy with curative intent. When pat. were found to have unresectable or metastatic disease intraoperatively, a palliative hepaticojejunostomy was performed. Moreover,hepaticojejunostomy was performed in pat. with stent failure and unresectable disease.Group 3 underwent hepaticojejunostomy without preoperative stenting. For pat. in Group 1, we determined the frequency of rehospitalization for recurrent jaundice. In these pat., a new endoprosthesis was placed when possible.Over an 8-year period, 342 pat. (151women, 191men) with adenocarcinoma of the pancreatic head or distal biliary tree received palliative treatment. Median age was 63ys (range 36-89.Symptoms of obstructive jaundice were exhibited by 26 pat.(76%). In 14 pat., none of the aforementioned procedures were performed. The remaining 247 pat. were divided into the groups described: Group 1(n=138,56%);Group 2(n= 68,28%);Group 3(n=41,16%).The 30-day mortality rates for pat. in Groups 1,2,and 3 were 2.2%,0%,and 2.4%. The morbidity rates were 6.5%,19.1%,and 14.6%, respectively. For the pat. treated with endoscopic stenting, the mean interval between
W1678 Will Ki-67 Predict Lymph Node Status in Pancreatic Endocrine Tumors? Derick Christian, Amit S. Khithani, A. Joe Saad, Joshua G. Barton, Jeffrey D. Linder, Prabhleen Chahal, Riteshkumar Patel, D. Rohan Jeyarajah Background: Pancreatic endocrine tumors (PET) are rare and exhibit an uncertain biological behavior. Many studies have suggested that certain tumor markers and factors such as size and mitotic count may predict the behavior of PET. However the role of Ki-67 proliferative index is much debated. Aims and objectives: To evaluate the significance of tumor size and Ki-67 proliferative index with lymph node status in PET. Methods: Records of patients who underwent surgical resection for PET between September 2005 to September 2009 by a single surgeon were reviewed. Pathologic variables were tumor size, lymph node status and ki-67 proliferative index. Results: Of the 248 pancreatic resections performed at our Non University Tertiary care center (NUTCC), a total of 24 pancreatic resections were performed for PET. Of these, 18 (75%) underwent distal pancreatectomy and 6 (24%) underwent pancreatoduodenectomy. PET was limited to the pancreas or regional lymph nodes in 23
S-901
SSAT Abstracts
SSAT Abstracts
Diagnostic of Exocrine Pancreatic Insufficiency - Fecal Elastase-1 vs. serumbeta-Carotin Dominique Suelberg, Johanna Krause, Orlin Belyaev, Ansgar M. Chromik, Waldemar Uhl
W1671 Prognostic Impact of Dihydropyrimidine Dehydrogenase Expression in Adjuvant Gemcitabine Plus S-1 Chemotherapy After Surgical Resection for Pancreatic Adenocarcinoma Naru Kondo, Yoshiaki Murakami, Kenichiro Uemura, Yasuo Hayashidani, Takeshi Sudo, Yasushi Hashimoto, Hiroki Ohge, Taijiro Sueda Objective: Although the prognosis in patients with pancreatic adenocarcinoma remains poor, adjuvant gemcitabine plus S-1 chemotherapy (GEM + S-1) after surgical resection for pancreatic adenocarcinoma has been shown to improve survival. S-1 is a novel oral fluoropyrimidine combination including tegafur (a prodrug of 5-fluorouracil; 5-FU), dihydropyrimidine dehydrogenase (DPD) inhibitor (5-chloro-2,4-dihydroxypyrimidine), and orotate phosphoribosyltransferase (OPRT) inhibitor (potassium oxonate). To clarify the relationship between expression of intratumoral enzymes related to the metabolism of 5FU and its derivatives and response to adjuvant chemotherapy with GEM + S-1 for pancreatic adenocarcinoma, we evaluated thymidylate synthase (TS), DPD, and OPRT expression immunohistochemically in resected pancreatic adenocarcinoma tissues. Methods: Polyclonal antibodies were used to immunostain sections of 106 formalin-fixed paraffin-embedded specimens of pancreatic adenocarcinoma resected between 1998 and 2009. The relationship between intratumoral TS, DPD, and OPRT expression and prognosis was evaluated statistically. Results: Out of 106 patients, 75 (70.1%) received adjuvant GEM + S-1 chemotherapy. High intratumoral TS, DPD, and OPRT expression was present in 68 (64.1%), 39 (36.8%), and 70 (66.0%) cases, respectively. Comparison of overall survival between High and Low intratumoral TS or OPRT expression revealed no significant difference regardless the application of adjuvant GEM + S-1 chemotherapy. In the GEM + S-1 (+) group, overall survival was significantly longer in the Low DPD subgroup than in the High DPD subgroup (hazard ratio [HR], 0.661: 95% confidence interval [CI], 0.447 - 0.978; P = 0.031), whereas in the GEM + S-1 (-) group, there was no significant difference between the High DPD and Low DPD subgroups. Moreover, in the High DPD group, there was no significant difference in overall survival between the GEM + S-1 (+) and GEM + S-1 (-) subgroups, whereas in the Low DPD group overall survival was significantly higher in the GEM + S-1 (+) subgroup by univariate analysis (HR, 0.456; 95% CI, 0.295 - 0.683; P < 0.001). Conclusion: Low intratumoral DPD expression was associated with increased overall survival in patients with pancreatic adenocarcinoma who received adjuvant GEM + S-1 chemotherapy. DPD is a relevant predictive marker of benefit from adjuvant GEM + S-1 chemotherapy in patients with resected pancreatic adenocarcinoma.
W1668 The use of Intraoperative Ablation Extends the Limits of Potentially Curative Treatment for Recurrent Colorectal Liver Metastases Anand Govindarajan, Dean Arnaoutakis, Michael D'Angelica, Peter J. Allen, Ronald P. DeMatteo, Leslie H. Blumgart, William R. Jarnagin, Yuman Fong
SSAT Abstracts
Introduction: Most patients undergoing liver resection for colorectal liver metastases develop recurrent hepatic disease. In this setting, complete resection is often technically difficult or impossible. This study evaluates the role of intraoperative ablation as an adjunct to resection in patients with recurrent colorectal liver metastases. Methods: A retrospective cohort study was performed using a prospectively collected database. The study population included patients who underwent curative reoperative surgery for recurrent colorectal liver metastases from 1991-2009. Patients were categorized based on whether they were treated with resection alone (RES) or in combination with ablation (COMB). The groups were compared with respect to tumor factors, perioperative factors and overall survival. Multivariable Cox regression models were used to examine overall survival while adjusting for confounding variables. Results: A total of 234 operations were performed on 104 patients. The median number of operations performed per patient was 2 (range 2-4). There was an increase in the proportion of patients treated with resection and ablation from 0% (1991-1994) to 48% (2007-2009). There was also an increase in the proportion of patients undergoing surgery with abnormal liver parenchyma (steatosis, fibrosis or inflammation) from 37.5% (1992-1994) to 52.2% (2007-2009). Patients undergoing resection and ablation had a greater tumor burden (median 3 metastatic lesions vs. 1, p<0.0001) and higher baseline clinical risk scores (median 3 vs. 2, p=0.041) than patients undergoing resection alone. Patients undergoing resection and ablation had lower intraoperative blood loss compared to patients undergoing resection alone (312±48 mL vs. 860±99mL, p=0.0060). At a median followup of 54 months, 5-year overall survival from the time of first liver metastasis was 47.4%, with no significant difference between the RES and COMB groups in univariate (5-year survival 46.6% RES vs. 59.6% COMB, p=0.86) or multivariable analysis (HR: 0.94, 95%CI: 0.30 to 2.93, p=0.91). Conclusion: In patients with recurrent colorectal liver metastases, the combination of intraoperative ablation with resection extends the limits of potentially curative therapy to include patients with advanced disease that might otherwise not be amenable to surgical intervention. The combination of resection and ablation yields survival results that are equivalent to that of resection alone but with reduced surgical morbidity.
W1672 Interventional Radiology Assisted Endoscopic Transgastric Peripancreatic Fluid Collection Drainage Abdulrahim Alawashez, James Ellsmere BACKGROUND: Many centers advocate the use of EUS to perform transgastric drainage of peripancreatic fluid collections (PFCs). However most endoscopists do not perform EUS. We report our experience using interventional radiology (IR) placed transgastric drains as a first stage to a subsequent endoscopic transgastric PFC drainage. OBJECTIVE: To evaluate the technical and treatment success and safety profile of IR assisted endoscopic transgastric PFC drainage. PATIENTS: This study involved 11 consecutive patients referred for endoscopic drainage of PFCs over a 2-year period. INTERVENTIONS: The first stage is performed by IR under CT guidance. A suitable window to the PFC through the stomach is determined. An 18-gauge trochar needle is then advanced through the anterior and posterior wall of the stomach into the PFC. Following removal of the inner stylet, a stiff Amplatz wire is advanced. The tract is dilated and over the wire, an 8.5 F multipurpose cathether is positioned in the PFC. The second stage is performed under endoscopic and fluoroscopic guidance. The drain is removed. The posterior gastrotomy is cannulated using a wire guided sphincterotome. The tract is dilated to 15mm using an endoscopic dilating balloon. Two 10F x 5cm double pigtail stents are left across the posterior gastrostomy. The anterior gastrostomy is closed with endoscopic clips. RESULTS: 11 patients (6 males and 5 females) with mean age of 61 (range 38- 79 years) underwent IR assisted endoscopic transgastric PFC drainage (7 pseudocyst, 4 abscess) over a 2-year period. The PFCs were gallstone related in 7 patients, post surgical in 2 patients and idiopathic in 2 patients. The mean size of the PFCs was 12 cm (range 6-19 cm) in its largest dimension. The procedures were technically successful in all 11 patients. There were 3 complications (pneumothorax that was treated with a chest tube, uncomplicated pneumoperitoneum that was confirmed by diagnostic laparoscopy, delayed bleeding from posterior gastrostomy requiring endoscopic hemostasis). All patients had a successful resolution of their PFCs on follow-up imaging. The median length of hospital stay was 11.5 days (range 1-86 days). At a mean follow-up period of 14 months (range 2-29 months), all patients were doing well. CONCLUSIONS: IR assisted endoscopic transgastric PFC drainage is technically feasible and safe and is associated with favourable clinical outcomes.
W1670 Underutilization of Surgical Treatment of Hepatocellular Carcinoma in the Medicare Population Shimul A. Shah, YouFu Li, Sing Chau Ng, Andrew Burr, Jennifer F. Tseng The incidence of hepatocellular carcinoma (HCC) is increasing in the United States. Due to factors such as portal hypertension, tumor biology and a donor organ shortage the care of these patients remain highly specialized and complex. Multiple treatment options are available for HCC but their use and utility remains unknown. Methods: Using SEER-Medicare linked data, we identified 8570 patients diagnosed with HCC between 1991 and 2007. Treatments included resection, liver transplantation (LT), ablation or transarterial chemoembolization (TACE). Patients who received no or palliative-only treatment were grouped (NoTx). Demographic, clinical and tumor factors were examined as determinants of therapy. Univariate and multivariate analyses were performed to determine predictors of overall survival. Results: Median age was 74.9 years. HCC therapies included resection (n=745; 8.7%), LT (n=120; 1.4%), ablation (n=312; 3.6%), TACE (n=1214; 14.2%) and NoTx (n=6179; 72.1%). Of the 8570 patients, 68% were coded as White, 11.3% Asian, 3.8% Hispanic, and 8.4% Black. Only 13.7% of patients underwent localized therapy that was potentially curable (TX group; resection, ablation or LT). If tumors were <5cm in largest dimension, then 64% of patients underwent TX. In the NoTx group, 49% were not cirrhotic, 36% had tumors < 5cm and 21% were tumor stage I or II. Median survival for all HCC patients increased over time (2002-2005: 116 months, 1999-2001: 96 months, 1991-1998: 87 months). In multivariate regression analysis, patients who received any modality of treatment achieved some benefit compared to NoTx. Specifically, adjusted overall survival was greatest in the patients in the TX group: LT (HR 0.35; 95% CI 0.26-0.47), resection (HR 0.33, 95% CI 0.29-0.37) and ablation (HR 0.51, 95% CI 0.42-0.62). Conclusion: In the Medicare population, HCC patients who receive potentially curative therapy such as resection, LT, and ablation experienced a substantial survival advantage over their non-operative peers (NoTx). Despite evidence that many patients had favorable biological characteristics, less than 15% of patients diagnosed with HCC received any treatment. Barriers to treatment and its underutilization must be identified to improve survival in patients diagnosed with HCC in the United States.
SSAT Abstracts
W1673 Advanced Disease Not a Risk Factor for Delayed Gastric Emptying After Pancreaticoduodenectomy (Whipple Procedure) Ronald J. Shonkwiler, Isadora C. Botwinick, John Steele, Gary Yu, John A. Chabot BACKGROUND: Delayed gastric emptying (DGE) is a common cause of morbidity after pancreaticoduodenectomy (PD). DGE can be a source of distress and discomfort to patients, resulting in prolonged hospital stays and subsequent drain on staff and resources. AIMS: We aimed to investigate whether any preoperative or intraoperative variables predict risk of DGE. Ability to predict DGE would potentially enable therapeutic interventions such as insertion of a feeding jejunostomy at the time of PD in high-risk patients. METHODS: We performed a retrospective chart review of 523 patients who underwent PD at our institution between 2000 and 2009; 384 resections were performed since 2005. Data collected included: patient demographics, preoperative symptoms, preoperative chemoradiotherapy, type of
S-900
resection performed (pylorus-preserving PD or standard PD, with or without vascular resection) OR time, estimated blood loss and transfusions, tumor size and histology, lymph node status, postoperative hematocrit and leukocyte count and postoperative NSAID use. We defined DGE as failure to tolerate a regular diet by postoperative day seven. Data was analyzed with alpha=0.05 using Fisher exact test for categorical variables and Mann-Whitney U test for continuous variables. RESULTS: A total of 114 patients with DGE were compared with 134 patients without DGE. Patients who suffered DGE were significantly more likely to have smaller tumors (p=0.015) and fewer positive lymph nodes (p=0.025), while patients without DGE were more likely to have received preoperative chemotherapy (p= 0.005) or preoperative radiation (p= 0.010). A possible explanation for these results is that patients with malignant disease have less normal functioning pancreatic parenchyma and are consequently less at risk for digestive symptoms resulting from pancreatic injury and inflammation. In addition, DGE was associated with an infection during hospitalization (p=0.008.) Hematocrit on the third postoperative day was significantly lower (p=0.019) for patients who experienced DGE , excluding patients who experienced post-operative bleeding. We hypothesize that DGE patients experienced more postoperative third space sequestration and bowel edema, resulting in greater hemodilution when sequestered fluid redistributed back into the vasculature. CONCLUSIONS: Contrary to our expectations, patients with smaller tumors and benign disease were more likely to experience DGE than patients with more advanced, malignant disease. As our ability to diagnose premalignant lesions improves, there will likely be more PD performed for benign disease, with a subsequent increase in the frequency of patients with DGE.
initial stent placement and stent replacement was 70.8 days (± 32). Median survival time for pat. treated only with endoscopic stent implantation (Group 1) was significantly (p<0.001) shorter than that of pat. who were first stented and subsequently treated with hepaticojejunostomy (Group 2) (5.1 vs. 9.4 months). None of the pat. who underwent surgery were rehospitalized for recurrent jaundice. In conclusion, we found that surgical hepaticojejunostomy can be performed with adequate results and an acceptable complications. Considering that biliary stents need to be replaced (mean time to stent replacement 70 days),management by hepaticojejunostomy may be superior to endoscopic stenting, especially for pat. with a life expectancy of greater than 6 months. The operative approach should be especially favored in patients whose disease is first found to be unresectable in the intraoperative setting. W1676 The Role of Procalcitonin in the Early Diagnosis of Postoperative Pancreatic Fistula After Pancreatic Resection Dominique Suelberg, Franziska Beuschel, Christiane Kotschenreuther, Torsten Herzog, Ansgar M. Chromik, Waldemar Uhl Background: Postoperative pancreatic fistula (PF) are frequent major complications after pancreatic surgery. Detection of amylase in drainage-fluid and serum-CRP are well established diagnostic parameters. The importance of Procalcitonin (PCT) in the diagnosis of PF remains to be elucidated. Aim of the study was therefore, to evaluate of PCT in the diagnostic of PF and major complications following pancreatic surgery (PS). Methods: All patients with PS were prospectively collected from Jan. 2009 - Sept. 2009. Serum-PCT, -CRP as well as amylase in drainage-fluid were measured every second postoperative day. Following parameters were analyzed: diagnosis, age, operation, complications (major, minor, secondary), PF (grade A, B and C; according to the ISGPF-definition). Results: During the observation period, n = 107 pat. received PS and developed n = 15 minor, n = 12 major and n = 3 secondary complications. Among major complications, n = 9 PF (8.9%) were diagnosed in the drainage-fluid, thereof n = 3 grade A, n = 2 grade B and n = 4 grade C. CRP and PCT were significantly increased in pat. with PF and other major complications after the 10th postop. day (p ≤ 0.001; Anova). In the detection of PF, the ROC-Analysis for PCT revealed an AUC of 0.736 with a sensitivity of 83% and a specificity of 72% for the cut-off-value of 17.8ng/ml. For CRP the AUC was 0.855 with a sensitivity of 86% and a specificity of 83% for the cut-off-value of 77.1mg/l. For the detection of major complications, the ROC-Analysis of PCT resulted in an AUC of 0.977 with a sensitivity of 100% and a specificity of 93% for the cut-off-value of 20.9 ng/l. For CRP, the AUC was 0.770 with a sensitivity of 73% and a specificity of 84% for the cut-off-value of 77.1 mg/l. Conclusion: Our results demonstrate that PCT is equivalent with CRP in the early diagnosis of postoperative pancreatic fistula, however superior in the diagnosis of all major complications following PS. Therefore we advocate the routine use of PCT and CRP following PS.
W1674 Diagnostic Laparoscopy for Pancreatic Cancer in an MRI Driven Practice: What's it Worth? Elliot Tapper, Bobby Kalb, Diego R. Martin, David Kooby, N. Volkan Adsay, Juan M. Sarmiento Introduction: For many patients with pancreatic cancer, CT is inadequate in determining unresectability; 10-48% of patients deemed resectable receive an unnecessary laparotomy. Accordingly, many groups have studied the role for diagnostic laparoscopy (DL) though none have evaluated it in an MRI driven practice. Methods: All MRI's administered for suspected pancreatic cancer between December 2004 and 2008 were evaluated. Radiographic diagnoses were prospective judged resectability based on the presence of metastases and relationship of the tumor with the surrounding vasculature. Unresectable disease received endoscopic biliary and duodenal stenting. Resectable and borderline disease received Whipples and double bypasses if unresectable intraoperatively. We performed a decision analysis for the cost-effectiveness of incorporating DL. We queried our billing database to render average costs for all inpatients with pancreatic cancer who received Whipples, doublebypasses and double-stenting procedures. We did not include professional fees. The marginal cost of DL was derived from the itemized costs of the materials, space and ancillary staff, presuming routine utilization, no missed metastases, and no complications. Results: Preoperative MRI deemed 94 patients' tumors resectable; 86 agreed to a laparotomy. Six patients were found to have metastases intraoperatively and 15 patients had unresectable disease (vascular involvement or benign pancreatitis) and thus received double-bypass procedures for which the average total cost of the hospitalization was $21,957.18. Whipples were provided to 65 patients at an average cost of $26,122.43. DL would thus be offered to 86 patients. For the 6 patients with metastases, it would be the only operation ($3604.07). This would be added to the cost of endoscopic stenting procedures, which results in an average total cost of hospitalization of $18,451.41. For the patients without metastases, the marginal cost of DL before a laparotomy would be $2651.71, which we added to the total costs above. Conclusions: For DL to be cost-effective, it would have to increase the rate at which we diverted patients to the GI lab for palliative stenting. In our model, DL would increase our costs by $191,072.18, equivalent to the total cost of hospitalization for treating 7 patients with Whipple procedures. Given our rate of missed metastases - 6% - and presuming perfect yield from DL, 15 patients would have unnecessary DL for every patient with occult metastases. For DL to be cost-effective, its intraoperative yield would have to be 70%.
W1677
Background: Determination of fecal elastase-1 is well established in the diagnosis of exocrine pancreatic insufficiency (EPI). However, accelerated gastro-intestinal passage could result in false positive results. Serum-beta-Carotin is an indirect parameter to measure the malassimilation of lipids, but is not determined in the diagnostic of EPI yet. Aim of the study was to evaluate the validity of Serum-beta-Carotin in the EPI-diagnostic pre- and postoperatively. Methods: All patients underwent pancreatic surgery were prospectively collected from Jan.2009-Sept. 2009. Fecal Elastase-1, serum-beta-Carotin and clinical features were analyzed preoperatively and 1 week, 3 and 6 month postoperatively together with surgical procedure and diagnosis. Ergebnisse: During the observation period 125 patients underwent pancreatic surgery. The results of preoperative fecal elastase-1 and serum-beta-Carotin showed significant lower values in pat. with chronic pancreatitis (CP) in contrast to pat. with benign and malign tumors (p≤0.005; Anova). One week postoperatively, serum-beta-Carotin was significantly decreased in pat. after pancreatic head resection (PHR) in contrast to pat. after left pancreatic resection (LPR) or bypass-procedures (BP) (p≤0.05; Anova). In contrast, one week postoperatively fecal elastase-1 was decreased in all groups regardless of the surgical procedure without any differences in the clinical features. The follow-up revealed reduced values for both parameters in pat. with PHR compared to pat. with LPR or BP. Overall, postoperative serum-beta-Carotin was more frequently in normal range than the values of fecal elastase-1 in patients without clinical signs of exocrine insufficiency. Conclusion: Our data demonstrate that serum-beta-Carotin is equivalent with fecal elastase-1 in the diagnosis of exocrine pancreatic insufficiency following pancreatic surgery. Moreover, the validity of serum-beta-Carotin seems to be superior to fecal Elastase-1.
W1675 Palliative Treatment of Obstructive Jaundice in Patients With Carcinoma of the Pancreatic Head or Distal Biliary Tree: Endoscopic Stent Placement vs. Hepaticojejunostomy Marius Distler, Stephan Kersting, Felix Rückert, Frank Dobrowolski, Stephan Miehlke, Robert Grützmann, Hans D. Saeger Only 20-30% of patients with malignant neoplasms of the pancreatic head or distal biliary tree can be cured by surgery. Palliative procedures play an important role in their management. We sought to determine if surgical or non-surgical management was the most appropriate therapy for the treatment of obstructive jaundice in the palliative setting. We retrospectively analyzed the outcomes of our pat. with regard to method of treatment and complications as well as the survival times.The pat. were divided into three groups based on treatment. Pat. in Group 1 underwent endoscopic bile duct stent placement.Group 2 underwent preoperative stenting followed by laparotomy with curative intent. When pat. were found to have unresectable or metastatic disease intraoperatively, a palliative hepaticojejunostomy was performed. Moreover,hepaticojejunostomy was performed in pat. with stent failure and unresectable disease.Group 3 underwent hepaticojejunostomy without preoperative stenting. For pat. in Group 1, we determined the frequency of rehospitalization for recurrent jaundice. In these pat., a new endoprosthesis was placed when possible.Over an 8-year period, 342 pat. (151women, 191men) with adenocarcinoma of the pancreatic head or distal biliary tree received palliative treatment. Median age was 63ys (range 36-89.Symptoms of obstructive jaundice were exhibited by 26 pat.(76%). In 14 pat., none of the aforementioned procedures were performed. The remaining 247 pat. were divided into the groups described: Group 1(n=138,56%);Group 2(n= 68,28%);Group 3(n=41,16%).The 30-day mortality rates for pat. in Groups 1,2,and 3 were 2.2%,0%,and 2.4%. The morbidity rates were 6.5%,19.1%,and 14.6%, respectively. For the pat. treated with endoscopic stenting, the mean interval between
W1678 Will Ki-67 Predict Lymph Node Status in Pancreatic Endocrine Tumors? Derick Christian, Amit S. Khithani, A. Joe Saad, Joshua G. Barton, Jeffrey D. Linder, Prabhleen Chahal, Riteshkumar Patel, D. Rohan Jeyarajah Background: Pancreatic endocrine tumors (PET) are rare and exhibit an uncertain biological behavior. Many studies have suggested that certain tumor markers and factors such as size and mitotic count may predict the behavior of PET. However the role of Ki-67 proliferative index is much debated. Aims and objectives: To evaluate the significance of tumor size and Ki-67 proliferative index with lymph node status in PET. Methods: Records of patients who underwent surgical resection for PET between September 2005 to September 2009 by a single surgeon were reviewed. Pathologic variables were tumor size, lymph node status and ki-67 proliferative index. Results: Of the 248 pancreatic resections performed at our Non University Tertiary care center (NUTCC), a total of 24 pancreatic resections were performed for PET. Of these, 18 (75%) underwent distal pancreatectomy and 6 (24%) underwent pancreatoduodenectomy. PET was limited to the pancreas or regional lymph nodes in 23
S-901
SSAT Abstracts
SSAT Abstracts
Diagnostic of Exocrine Pancreatic Insufficiency - Fecal Elastase-1 vs. serumbeta-Carotin Dominique Suelberg, Johanna Krause, Orlin Belyaev, Ansgar M. Chromik, Waldemar Uhl