Hepatic steatosis is associated with increased complications following major liver resection for cancer but does not impact survival

NS398 and U0126 in Hep3B cells resulted in a synergistic increase in apoptosis (7X control). Relative apoptosis in HepG2 cells was increased with U0126 alone or in combination with NS398 (9-10X control). Relative apoptosis in both cell lines was strongly correlated with changes in the expressionof the anti-apoptotic protein Bcl-xL Cellular growth was assessed by colorimetric proliferation assay and confirmed by cell counts (trypan blue exclusion). HepG2 and Hep3B cells had concentration-dependent inhibition of cell growth with NS398 or U0126 treatment alone. The combination of NS398 and U0126 resulted in additive inhibitory effects on growth. Growth inhibitory effects in HepG2 and Hep38 cells appear to be in part secondary to the induction of GO/G1 and G2/M cell cycle arrest respectively, as determined by flow cytometry. Despite differential signaling in HepG2 and Hep3B cells, the sum effect of combining a COX-2 inhibitor and a MEK inhibitor results in enhanced anti-tumor actions. This novel combination may be useful for treating patients with HCC.

8O2 Clinical Risk Score Correlates with Yield of PET Scan in Patients with Colorectal Hepatic Metastases Chris Schussler-Fiorenza, David M. Mahvi, John Niederhuber, Layton F. Rikkers, Sharon M. Weber Introduction: Positron emission tomography (PET) detects occult metastatic disease in up to 20% of patients with isolated hepatic colorectal metastases. However, the majority of patients do not benefit from PET scanning. We sought to evaluate the utility of PET scanning in patients with isolated colorectal hepatic metastases and correlate results with a clinical risk score (CRS). Methods: Patients with isolated colorectal hepatic metastases on conventional axial imaging who were imaged with PET scan between 1998-2002 were identified from a prospective PET database. A retrospective analysis of clinical parameters, radiological imaging results, intraoperative findings, and outcome was performed, All patients were assigned a clinical risk score, with one point added for each of five preoperative factors (disease-free interval <1 year, tumor size > 5 cm, tumor number >1, CEA >200, node positive primary). Results: PET scanning was utilized in the preoperative evaluation of 85 patients. Radiological evaluation included contrast-enhanced abdomen/pelvic CT in 98%, chest CT in 38%, and abdominal MRI in 13%. In the majority of patients (53%) PET provided no additional information over conventional imaging. Overall, PET resulted in either detection of occult extrahepatic disease or confirmation of questionable findings on conventional imaging in 20% of patients, whereas PET readings were inaccurate in 27% (median F/U, 12 months). In order to determine if patient selection for PET could be enhanced to increase the yield, we evaluated the correlation of PET scan readings with CRS. There was a significant association of CRS with the yield on PET scan, such that in patients with a CRS of O, 50% had false positive readings and no patient had extrahepatic disease detected with PET, while in patients with a CRS of -> 1, there were no false positives and 10/77 (13%) were found to have extrahepatic disease detected only with PET scanning (Table, p<0.001, Fisher exact test). Conclusion: The clinical risk score correlated with the yield of PET, with a greater proportion of patients at low risk (CRS = 0) having false positive readings on PET scan. In addition, no patient with a CRS of 0 was found to have occult extrahepatic disease on PET scan. Patients with isolated hepatic colorectal metastases and a CRS of 0 should undergo conventional imaging alone prior to surgical exploration.

733 Transcystic Biliary Manometry to Evaluate Sphincter of Oddi Dysfunction in Chronic Acalculous Chnlecystitis Kirpal Singh, David J. Soto, Mark A Ingram, Maurice E. Arregni Introduction: Sphincter of oddi dysfunction (SOD) may have a role in chronic acalculous cholecystitis (CAC). The purpose of this study is to 1) evaluate incidence of SOD in patients with CAC 2) to correlate SO pressures with patient outcome and 3) to correlate intraoperative manometry with postoperative ERCP/manometry. Method: Transcystic biliary manometry was attempted in one hundred and twenty nine patients with CAC during laparoscopic cholecystectomy between Aug/94 and Sept/02. Diagnosis of SOD was made if the average basal pressure was greater than 40 mm Hg. The biliary manometry data was correlated to the patients' outcome using Fisher's exact and Student's t-test. Patients with persistent or recurrent symptoms underwent postoperative ERCP/manometry/sphincterotomy. This data was then correlated with operative manometry and their outcomes followed. Results: Intraoperative manometry was completed in 91 patients but seven were lost to follow-up. Folhiw-up was a mean of 32 months (0.25-99mo). Fifty-five (65%) had SOD. Forty-three (78%) of these had resolution or significant improvement and twelve (22%) had persistence of symptoms. Twenty-nine (35%) patients had normal SO function. Twentyfour (83%) had resolution or significant improvement and five (17%) had persistence of symptoms. Outcomes were similar (p =0.62) regardless of the SO pressure. Pain score (O10) was obtainable from 49% of patients with resolution/improvement, and went from 8.5 to 1.1 (p<0.0001). Five (12%) patients with SOD recurred after the initial operation. Mean time to recurrence was 23mo (11-40mo). All improved after endoscopic sphincterotomy. Patients with persistent symptoms also had a trend towards improvement with postoperative endoscopic sphincterotomy. Patients who had postoperative ERCP/manometry, intraoperative findings were confirmed. Conclusion: SOD is a significant factor in CAC. Most with CAC improve after cholecystectomy regardless of the SO pressures. There is a good concordance between intra-operative and post-operative manometry. Prebminary data suggests a correlation between SO pressures and resolution/improvement following endoscopic sphincterotomy

PET ,ScanFindings False PoMflve False Nega6ve O~lcted ExtmheNtlr ~ ConflrrnM Quea~onable Find~g= on CT scan

Cystic Lesions of the Pancreas: Selection Criteria for Operative and NonOperative Management in 209 Patients Peter J. Allen, David P. Jaques, Michael D'Angelica, Wilbur B. Bowne, Kevin C. Conlon, Murray F. Brennan Introduction: Because of the inability to determine benign from malignant, many have recommended that all cystic lesions of the pancreas be resected. Recent studies however, have suggested that some patients may be safely followed as current imaging techniques have improved the ability to characterize these lesions. Methods: Inpatient and outpatient physician billing data was obtained, and patients evaluated between Jan 1995 and Dec 2000 with the ICD-9 diagnosis of pancreatic cyst (577.2) were reviewed. Patients presumed to have a pancreatic pseudocyst, intraductal papillary mucinous neoplasm, or cystic islet cell tumor were excluded. Patient, cyst, and treatment characteristics were recorded. Comparisons were made between patients who underwent non-operative and operative management. Patients managed non-operatively were typically followed with either high quality C.T scanning, or M R.C.P., on an annual or biannual basis. Results: Over the five year period, 209 patients were evaluated with the pre-treatment ICD-9 diagnosis of pancreatic cyst. Nonoperative treatment was chosen for 138 patients (66%), and within this group the median radiographic follow-up was 31 months. The average initial cyst diameter was 2.5 cm (range 0.5 - 13.0), and the median change in cyst diameter during follow-up was 0 cm (range 1.5 - 4.0). Cyst growth warranting resection occurred in six patients (4%); however none of these patients had a malignant diagnosis. Operative treatment was chosen for 71 of the 209 patients (34%). A malignant diagnosis was found in five of the operative patients (5/ 71, 7%), and 39 patients (55%) had a serous cystadenoma. All five of the patients with a malignant cyst had septated cysts that were symptomatic, with an average cyst diameter of 7.0 cm (range 5.0 - 10.0). Differences between operative and non-operative groups are depicted in the table. Conclusions: Selected patients with cystic lesions of the pancreas may be safely followed radiographically. Selection criteria identified in this study (cyst size, solid component, septations, symptoms), and the utilization of new imaging techniques, allow the creation of logical treatment paradigms for these patients.

Hepatic Steatosis Is Associated with Increased Complications Following Major Liver Resection for Cancer but Does Not Impact Survival David Kooby, Yuman Fong, Mithat Gonen, Arief Suriawinata, Peter Allen, David Klimsta, Ronald Dematteo, Michael D'Angelica, Leslie Blumgart, William Jarnagin Purpose: Fatty accumulation m the liver, or steatosis, is common after systemic chemotherapy, and has been considered a major risk factor for adverse outcome after hepatic resection, but its true impact is unclear. This study evaluates effects of hepatic steatosis on outcome after liver resection for malignancy in a large group of patients. Methods: All patients with underlying steatosis who underwent hepatic resection for cancer at Memorial Hospital from 12/91-9/01 (N = 319) were identified from a prospective database. Histology was re-reviewed by a pathologist, and non-tumor bearing liver was stratified by degree of fatty change according to a 4-poim scale: 1, focal steatosis (N= 139); 2, diffuse but mild (N=85); 3, diffuse (N = 59); 4~ severe (N = 36). Cases were further stratified as follows: mild steatosis, 1-2 (N =224, 70%); marked steatosis, 3-4 (N=95, 30%). Clinical outcome was correlated with histology and results were compared to a cohort of patients with normal livers (N = 160) matched statistically by age, diagnosis and extent of resection. Results: Operative procedures consisted of 164 trisegmentectomies, 134 lobectomies and 181 resections of less than one lobe. Median operative time was 250 rain (range, 38-555) and median blood loss was 640 cc (range, 20-6000). Morbidity was 49% and operative mortality was 3.5%. On multivariate analysis, marked steatosis was associated with increased complications but did not impact on perioperative mortality (Table). Morbidity and mortality were unaffected by mild steatosis. Additionally, overall survival of patients with normal parenchyma (43 months) was no different from that of patients with marked steatosis (44 months, p = 0 . 8 by log-rank test). Conclusions: The presence of marked steatosis was associated with increased complications after hepatic resection, but had no impact on operative mortality or long-term survival. Steatosis alone, therefore, should not preclude a potentially curative hepatic resection, if clinically warranted.

Comparh~nof cystcharestedstleafor pallentsundecgolagoflecaUveand non-operativemanagenwnt Factor Cyst DkmMw (r Solid ComponKit SeplaUont Symptomatic

Multivariate analysis of facton; associated with con~ncatlon| and poltqNraUve mortalRy

SSAT

Abstracts

N (%) 218 (46) 249 (52) 147 (31) 54 (11) 95 (20) 320 (67) 42 (9)

Contp/Icatlonl Percent p value 54 NS 54 NS 57 NS 57 NS 60 < 0.01 60 < 0.01 66 < 0.01

CRS = I 0 25% 13% 8%

803

734

Factor (N=479 paCaate) Age > 65 Male gender Operative tim > 5 hours Diabetes MeN#us Marked e t e ~ l e Relec6on of a lobe or more Blondlons > 2 liters

CR$ 0 50% 0 0 13%

Ota4ratlveMo~allty Percent p value 3 NS 2 NS 2 NS 2 NS 3 NS 5 0.01 4 0~05

A-794

O~e (mean) Yes Yes Yes

(n,,7t) 5.8 35 (49%) 51 (72%) 40 (56%}

I ~ e

(n=138) p-value 2.5 0.01 7 (5%) <0,001 39 (28%) <0.001 11 (8%) <0.001