- Email: [email protected]
S-adenosylmethionine (SAMe) modulates IL-10 and TNF production by monocytes, macophages, and kupffer cells
338
340
MELO Scoring System Not Predictive Of Mortality In Patients Waiting For Liver
0nco-wnakly Recombinant Human EWthropnatin (Epontin Alla) Facilitates Optimal Ribavirin (RBV) Dosing in HepatHis C Virus (HCV)-intected Patients Receiving Interferan-a-2b (IFN)/ROV Combination Therapy Douglas T. Oietedch, Cabrini Medical Ctr, New York, NY; Run Wasserman, infectious Diseases Doctors Medical Group, Walnut Creek, CA; Norbert 6rau, Bronx VAMC, Bronx, NY; Tarek I. Hassanein, UCSD Liver Ctr, San Diego, CA; Edmund J. Bini, VA New York Harbor Heaithcare System, New York, NY; Mark Sulkowski, Johns Hopkins Univ Sch of Medicine, Baltimore, MD
Transplantation Jesus E. Carale, Timothy M. McCashland, Elizabeth Lyden, Daniel F. Schafer, Michael F. Sorrell, Alan Langnas, UNMC, Omaha, NE Background: The Mayo Clinic End-StageLiver Disease(MELD) model developed by Malinchoc et al. was found to accurately predict mortality following elective TIPS. (Hepatology 2000;31:864-871). This same model was also applied to non-liver transplant cirrhotic patients and accurately predicted mortality over 6 months. (GE 2000, A972) Recently, UNOS has proposed the MELD model to prioritize transplant candidates for liver allocation by predicting the risk of a patient s mo~lity on the waiting list. Aim: To determine the validity of the MELD model in predicting mortality in patients listed for liver transplantation. Methods: Seventy patients that were listed for liver transplantation from 1/1/99 to 1011100were reviewed at the University of Nebraska Medical Center (UNMC). Patients risk scores were calculated for predicted survival at 3, 6, and 12 months utilizing the MELD model. The patients actual mortality was compared to the predicted mortality. The date listed for transplantation was used as the start date. Cox proportional-hazard regression was used to evaluate model variables. Results: The patients were separated into two groups, those with a high risk score >1.8 (n=9) and those with low risk score <1.8 (n=61). Eight of nine patients died in the high-risk group, 8/61 died in the low risk group. Four patients were transplanted in the low risk group and none in the high-risk group. Actual and predicted mortality for the 2 risk groups showed that the predictive model is dissimilar to the observed mortality in both risk groups (p =0.0028 high-risk group, p =0.0002 low risk group). In the high-risk patients, the predictive model under estimated mortality, whereas in the low risk group, the predictive over estimated mortality. The only significant predictor of mortality for the model variables was log of creetinine (p =0.0061). Log INR, log bilirubin, nor cause of liver disease was predictive of outcome. Observed verses the predicted survival is illustrated below. Conclusion: The MELD model did not accurately predict mortality when applied to patients who were listed for liver transplantation at UNMC. Validity of this model to stratify patients listed for liver transplantation needs to be studied further prior to implementation.
Background: RBV dose reduction or discontinuation has beenthe standard approachto treating anemia due to IFN/RBV therapy Jn HCV-infected patients. A recent retrospective analysis suggested that maintaining RBV doses greater than 10.6 mg/kg per day of the intended dose was aseecbted with higher sustained virologic response rates. (McHutchison et al. AASLD. 2000:223A. Abstract 247) We conducted an open-label, randomized, muiticenter study to assess the efficacy and tolerability of once-weekly (qw) epoetin afro therapy for hemoglobin (Hb) decreases associated with IFN/RBV therapy and to evaluate the impact of epoetin affa therapy on RBV dosing. Methods: HCV-infected patients who developed Hb levels -< 12 g/ dL during the first 24 weeks of combination IFN/RBVtherapy (n = 60) were randomly assigned to treatment with epoetin aita (40 000 units) subcutaneously qw for up to 36 weeks or standard of care (SOC). The primary and secondary efficacy end points (Week 16) were changes from baseline in Hb level and RBV dose, respectively. Interim results are reported; final results will be presented. Results: At baseline, mean Hb levels and RBV doses were similar between the epoetin aita and SOC groups (11.0 g/dL for both, and 998 vs 986 mg/ day, respectively). At the last available assessment (LAST), the mean changes from baseline Hb levels were +2.7 vs +0.4 g/dL for epoetin alfavs SOC groups (P<.05), respectively; and the mean changes in RBV doses were -25 and -190 mg/day, respectively (P<.O5). The LAST mean Hb level in the epoetln alfa group (13.8 g/alL) was significantly greater (P<.05) than that of the SOC group (11.3 O/dL). Based on LAST, 77% vs 50% (P<.01) of patients in the epoctin alia ve SOC groups, respectively, were treated with RBV doses greater than 10.6 mo/kg per day. Conclusions: Epoetln alfa (40 000 units qw) was well tolerated and effectively increased Hb levels in HCV-infected patients treated with IFN/RBV. Epostin affa increased the proportion of patients receiving RBV doses greater than 10.6 mg/kg per day, suggested as the minimum dose to increase the likelihood of sustained virologic response.
/ 341 Prlinam Hepatitis B Vines Canyiep Lamivodina Resistant Mutations Are Not AnsnstakNI with ~ Replication Fitness and Are Sensitive to Adotovir Robert Ym Cbsn, Paul V. Decmond, St Viocent's Hasp, Melbourne Australia; William E. Debney, Danni L. Coitodge, Rosalind M. Edwards, Victorian Infectious Diseases Reference Lab, Melbourne Australia; Harriet C. Isom, Penn State Coil of Medicine, Hershey, PA; Stephen/L Locamini, Victorian Infectious Diseases Reference Lab, Melbourne Australia
339
IntTodoction: Hepatitis Be antigen (HBeAg) negative chronic hepatitis 8 (CHB) is associated with the G1896A (PC) mutation in the HBV precore region. Prevalence of HBeAg-negative CHB is higher than that of HBeAg positive CHB in some countries. Lamivudine resistant mutations develop in PC mutants at a similar rate during lamivudine therapy as in wild type (WT) HBV. Lamivodine resistant mutations in WT HBV cause a marked decrease in replication fitness, it is unclear if similar impairments occur in PC related lamivudine resistant mutants. In addition, drug sensitivity of lamivudine resistant mutants concurrently carrying the PC mutation are unclear. Aim: To examine the replication fitness of iamivudine resistant variants of precore mutant HBV exhibiting the G1896A using the recently described recombinant HBV paculovirus system (Deianey and Isom. Hepatology 1998;28:1134). In addition, the antiviral drug sensitivity of these mutants was evaluated against lamivudine, edefovir and penciclovir. Methods: The G1896A mutation was introduced into a 1.3 unit length infectious cDNA construct of WT HBV. Further concurrent lamivudine resistant mutations (M5521, M552V/M528M) were introduced, producing clones of PC, PC/MS521, PC/M552V/L528M, WT, WT/M5521, WT/ M552V/L528M Each infectious clone was then used to form recombinant HBV-boculovirus and then used to transduce HepG2 cells. The IC5ofor Penciclovir, Lamivudine and Adetovir were then individually measured using standard dose-response experiments (Colledge et al. 2000;44:551) against intracellular HBV DNA markers as measured by Southern blot. Results: WT and PC were found to have similar replication fitness as described in other studies. Replication fitness of WT/M5501 and W1"/M552V/L528M were much reduced compared with WT. However, PC/M5521and PC/552V/L528M were found to have similar replication fitness as PC and WT. Adetovir was active against PC/M5521and PC/M552V/L528M mutants, indicating no cross resistance between lamivudine and adefovir even in the setting of the PC mutation. The IC50 for PC/MS521and PC/M550V/L528M were similar to that of non drug-resistant PC. Lamivudine and penciclovir were inactive against lamivudine resistant mutants in either WT or PC. Conclusion: Lamivudine resistant mutants are not associated with decrease replication fitness within the framework of PC mutant. HBV with concurrent precore and lamivodine resistant mutations are sensitive to adefovir.
Bile Duct injury in CF Knockout Mice With ColHis: Further Evidence Of An Association Between CFTR Mutations And Primary Sclernsing Chotnagltis Munir M. Zaman, Paola G. Blanco, Rhonda K. Yantiss, Imed A. Nasser, Ignacio Calvo, Sunil G. Sheth, Steven D. Freedman, Beth Israel Deaconess Medical Ctr, Boston, MA We have recently demonstrated a strong association between the occurrence of primary sclerosing choiangitis (PSC) and mutations in one allele of the cystic fibrosis transmembrane conductance regulator (CFTR) gena in patients with inflammatory bowel disease (IBD). This suggests that the carrier state for CFTR mutations may predispose to the development of PSC in patients with IBD. Based on this, we hypothesize that the induction of colitis in homozygous (cftr -~-) or heterozygous (cftr +'-) CF knockout mice would lead to bile duct injury (BDI) and perhaps represent a mouse model of PSC. AIM: To determine whether cftr -/- and cffr +/- mice develop BDI in the setting of colitis. METHODS:Cftr ~1~.~- mice were bred and offspring weaned at 21 days of age on to a complete liquid diet (Peptamen, Nestle Clin.lnc.). On day 40, mice were fed Peptamen with or without 125 mg dextran sodium sulfate (DSS)/day for 10 days to induce colitis. Mice were then sacrificed and serum alkaline phosphatase (AP) was measured. Liver and colonic tissues were submitted for blinded histologic scoring by 2 hepatopathologists. BDI was categorized by the following features: epithelial injury, proliferation, anguiation, and fibrosis. Each of these four features were scored as 0 (absent or mild) or 1 (moderate or severe), then summated for a maximum score of 4. RESULTS:All mice on DSS developed bloody diarrhea and histologic evidence of colitis, most severe in wild type (WT) mice. The control mice [WT no DSS in=6); WT with DSS colitis in=6); and cftr -/- no DSS (n=3)] had no evidence of BDI with a score of O. In contrast, cftr -/- (n=14) and cftr +/- in=5) mice with DSS colitis had mean scores (_+S.E.) of 1.9+-0.4 and 1.4_+0.5,respectively (p
342 S-adennsyimothionlne (SAMe) Modulates IL-10 and TNF Production by Monocytns, Macophaons, and Kuplfer Cells Craig J. McClain, Univ of Louisville, Louisville, KY; Rajender Chawia, Weody Nelson, Waiter Watson, Daniel Skufca, Sankar Boss, Daniell 8. Hill, Univ of Kentucky, Lexington, KY; Shirish Barve, Univ of Louisville, Louisville, KY Background: Use of over the counter complimentary and alternative medicine (CAM) is prevalent in patients with chronic diseases such as liver disease. Indeed, studies indicate that over 40% of patients with chronic liver disease take some form of nutritional supplement or herbal medication. S-adenosylmethlonine (SAMe) is one such one CAM product that has substantial scientific rationale for the treatment of alcoholic liver disease (ALD). There is a well-documented detect in the hepatic transsuffuration pathway in ALD, with a buildup of methionine and a decrease in SAMe. This is due, inpart, to decreased activity of the enzyme
A-64
methionine adenosyltransferase. Many of the metabolic abnormalities and the liver injury in ALD are thought to be due to overproduction of the proflammatory cytokine tumor necrosis factor (TNF), with concomitant inadequate levels of the antiinflammatory cytokine [L-IO. Methods: In these studies we evaluated the role of SAMe in modulating TNF and IL-IO production. We first used a murine macrophagecell line raw 264.7 and incubated cells with increasing concentrationsof SAMeor ethionine (which inducesa functional SAMedeficiency). Compared to control cells, SAMe supplementedcells had a dose dependent increase in IL10 and a dose dependentdecrease in TNF concentrations. In ethionine treated cells, a dose dependent decrease in IL-IO and an increase in TNF was observed. Parallel changes were seen in TNF and IL-IO mRNA. Next, we demonstrated that SAMe was effective at down regulatingTNFproduction in peripheralblood monocytesfrom patientsfrom alcoholic hepatitis. Lastly, we isolated kupffer cells from rats fed a methyl-restricted choline deficient diet, which produces hepaticSAMe deficiency. Isolated kupfter cells spontaneouslyproduced large amounts of TNF and producedsignificantly more TNF in responseto a LPS stimulus. Results: In summary, peripheral blood monocytes from patients with alcoholic hepatitis and kupffer cells from rats with SAMe deficiency both produce significantly increased amounts of TNF in the basaland LPS stimulated state,and this excessTNFproduction could be down regulated with SAMe. In a murine macrophage cell line we documented that SAMe not only down regulated TNF, but also up regulatedthe critical antiinflammatory cytokine IL-IO. This may have major implications for therapeutic intervention in ALD and other forms of chronic liver disease.
345 incidence and Survival of Colorectal Cancer in the US: 1889-189"/ Linda Rabeneck,Hashes B. EI-Serag, Baylor Coil of Medicine, Houston, "IX; Robert S. Sandier, Univ of North Carolina, Chapel Hill, NC Background:The past decadehas witnessedimportant advancesin colorectal (CRC)screening and treatment. Research has shown that in average-risk individuals screening reduces CRC mortality. In addition, randomized controlled trials in CRC patients have shown improved survival with adjuvanttherapies. However,the impact of these advanceson the incidenceand survival of CRC in clinical practice remains unknown. Objectives:To examinetime trends in the incidenceand survival of CRCin the U.S.Methods: Registriesof the Surveillance,Epidemiology, and End Results (SEER) databaseof the National Cancer Institute account for 14% of the U.S. population. Using the SEERdatabasewe examinedtemporal changes in the incidence and survival of CRC by gender, race, site, and stage. Patientswith histologically proven CRC between 1989 and 1997 were identified and analyzedin 3-yr time periods (89-91; 92-94; 9597). We calculated the age-adjusted incidence rates and median, 1-, 2-, and 5-yr observed and relative survival rates. Results: Between 1989 and 1997 we identified 108,057 patients with CRC, 48.6% were women, 84.6% white, and 8.3% black. For all sites the incidencerates were higher in men than women during all time periods. For left and right CRC,the incidence rates were higher in blacks than whites during all time periods. For rectosigmoid CRC, the incidence rates tell over time. For example, from 89-91 to 95-97 the incidence rate fell in white men from 18.1 per 100,000 (17.6-18.6) to 15.1 per 100,000 (14.7-15.6); and in white women from 10.2 (9.8-10.5) to 9.0 (8.7-9.3). Lesser declines were seen in black men and women. Similarly, for left CRC the incidence rates fell over time. For right CRC,the incidence rates remained unchanged over time. For all sites, the incidence rates of metastatic CRC remained unchanged over time. For all sites, the median survival was higher in whites than blacks during all time periods. For all sites there was no change in the 1-yr, 5-yr, or median survival overall, or for subgroups based on gender, race, or stage over time. For example, for rectosigmoid CRC the overall 1-yr observed survival rates were 84.1%, 83.8%, and 83.0% during 89-91,92-94, and 95-97, respectively.Conclusions:The decreasein incidence, particularly for rectosigmoid CRC is consistent with wider use of screening and removal of adenomas, The stable incidence rates for metastatic CRC raise the possibility of a subset of cancers with rapid growth potential that resist early detection and account for the stable survival rates over the past decade.
343 Hepatic Cholesterol and Bile Acid Metabolism and Intestinal Cholesterol Absorption in Scavenger Receptor Class B Type I (SR-BI)-deficient Mice Pablo Mardones, Veronica Ouinones, Ludwig Amigo, Mauricio Moreno, Juan F. Miquel, Pontificia Univ Catolica, Santiago Chile; Margrit Schwarz, Univ of Texas Southwestern Medical Ctr, Dallas, TX; Helena E. Miettinen, Bernardo Trigatti, Monty Krieger, MA Institute of Technology, Cambridge, MA; Sonya Vanpatten, David E. Cohen, Albert Einstein Coil of Medicine, New York, NY; Attilio Rigotti, Pontiticia Univ Catolica, Santiago Chile BACKGROUND:The scavengerreceptor class B type I (SR-BI), which is expressedin the liver and intestine, plays a critical role in cholesterol metabolism in rodents. While hepatic SR-BI expression regulates high density lipoprotein (HDL) cholesterol metabolism, intestinal SR-BI has been proposedto facilitate cholesterol absorption. AIM: To further evaluatethe relevance of SR-BI in the enterohepaticcirculation of cholesterol and bile salts, we studied plasma HDL cholesterol clearance,hepatic cholesteroland bile acid metabolism, biliary lipid secretion, and intestinal cholesterolabsorption in SR-BI knockout mice. RESULTS:SR-BI deficiencyimpaired HDL cholesteryl ester clearanceand HDL cholesterol transfer from plasma into the bile. SRBI knockout mice exhibited a selective decrease in biliary cholesterol secretion, without concomitant changes in either biliary bile acid or phospholipid secretion. Hepatic total and unesterifled cholesterol contents were slightly increasedin SR-BI-deficient mice, while sterol synthesis and low density lipoprotein receptor expressionwere not significantly changed. Bile acid pool size and composition as well as fecal bile acid excretion were not altered in SR-BI knockout mice. Intestinal cholesterol absorption was increasedand fecal neutral sterol excretion was slightly decreasedin SR-BI KO mice relativeto controls. CONCLUSION:Thesefindings establish the critical role of hepatic SR-BI expression in selectively controlling the utilization of HDL cholesterol for biliary secretion. In contrast, SR-BI is not essential for intestinal cholesterol absorption.As a consequenceof its key role the last step of HDL-mediatedreverse cholesterol transport, hepatic SR-81 activity may explain the anti-atherogenicactivity of this lipoprotein receptor in vivo.
346 Complications of Screening Sigmoidoscopy Theodore R. Levin, Kaiser Permaneote,Walnut Creek, CA; Carol Conell, Kaiser Permanente Div of Research, Oakland,CA; Jean A. Shapiro, Centers for Disease Control and Prevention, Atlanta, GA; Shella G. Chazan,Kaiser Permanente Div of Research, Oakland, CA; Marion Nadel, Centers for Disease Control and Prevention,Atlanta, GA; Joe V. Selby, Kaiser Permanente Div of Research, Oakland, CA Purpose: To determine the incidence of serious adverse events following flexible sigmoidoscopy (FS) in a large scale colorectal cancer screening program. Methods: The automated databasesof Northem CaliforniaKaiser Permanentewere accessedto identify hospitalizations within and outside the health care program fogowing 109,534 screening FS procedures conducted on average-risk subjects aged 50 and over during the years 1994-96 as part of the ColorectalCancerPrevention(CoCaP)Program. Charts of patients hospitalizedduring the 4 weeks following FS were reviewed to confirm diagnoses and the reason for FS. Serious complications were defined as perforations, lower gastrointestinal bleeding requiring transfusion, unspecifiedcolitis and diverticulitis requiring surgery. Non-serious hospitalizationswere for fever, abdominalpain and lower GI bleedingnot requiringtransfusion. Automateddatabases were also reviewed to identify hospitalizations for myocardial infarction (MI) and unstable angina, with chart review confirmation. The rate of MI in in these subjects in the first day, first week and first 4 weeks after FS were compared to the rate of MI occurring in the remainderof the 52 week period after FS. Resuits: Nearlyhalf of subjects were female (48,8%), and 74% of exams were performed by non-gastroenterologistphysicians or nurses. Overall, 24 subjects were hospitalizedfor a complication, 7 of which were serious (incidencerate: 21.9/ 100,000 overall,6.4/100,000 for serious complications).The ratesfor perforation, diverticulitis requiring surgery and bleeding requiring transfusion were each 1.81100,000; colitis occurred in 0.9/100,000. In a multivariate logistic regression model, men were more likely than women to experiencea complication, and complicationswere more common in proceduresassociated with biopsy or polypectomy. Hospitalizationfor MI occurred in 28 subjects within 4 weeks of FS, including 8 in the first 7 days. In multivariate models, subjects were no more likely to experienceMI in the first day, week, or month after FS comparedto later periods. Conclusions: FS, administeredas part of a large scale screening program with a diverse group of examiners, is associatedwith a modest rate of complications.While MIs occur in screeningaged subjects following FS, there does not appear to be a measurable increase in risk for MI in the period immediately following FS, compared with later periods.
344 Nonalcoholic Fatty Liver Disease: The Most Common Cause ol Abnormal Liver Enzymes in the U.S. Population Jeanne M. Clark, Frederick L. 8rancati, Anna Mae E. Diehl, Johns Hopkins Univ Sch of Medicine, Baltimore, MD Background: There is increasing evidencethat nonalcoholic fatty liver disease (NAFLO) can progress to fibrosis and eventuallyto cirrhosis. In Europeanand Asian series the prevalence of NAFLD is estimated to be 14-21% and is more common among adults with obesity and/ or diabetes. However, the prevalence of NAFLD in the U.S. population is unknown. Our objective was to estimatethe prevalenceof NAFLDin the United Statesand identify associated clinical factors. Methods: Using data from Third National Health and Nutrition Examination Survey (1988-1994), we examinedthe prevalenceof NAFLD among adults, ages 17 and older (N=13,500). We defined NAFLD using criteria that Ratziu et al. validated by liver biopsy (Gastroenterology2000); i.e. an elevationabove normal of AST, ALT, or GGT (liver enzymes) and negative Hepatitis B and C serulogies, transferrin saturation less than 50%, and average daily alcohol intake < 2 drinks for women and < 3 drinks for men. We used bivariate and multivariate logistic regression to identify factors associated with NAFLD. Analyses were weighted according to the sampling design in order to reflect the U.S. population. Results: The prevalenceof individuals with 1 or more abnormal liver enzymes in the general U.S. population is 27.7%. Only 4.2% of these abnormalities were due to chronic viral hepatitis, alcoholic liver disease, and/or iron overload; hence, the prevalence of NAFLO in the U.S. population is 23.5% (p<.O01). NAFLD is more common in men (33% vs. 17%, p<.O01) in all age groups; howeverthis tendency is most pronounced in individuals under age 30 (25% vs. 9%). Nrican Americans have higher rates of NAFLDthan whites (33% vs. 23%, p<.O01). This is true in both men and women, and in all age groups. The prevalenceof NAFLDincreases with increasing age (OR: 1.01,95%01: 1.01-1.01, p<.O01), BMI (OR 1.09, 95%C1: 1.08-1.10, p<.O01), with increasinglevelsof triglycerides (OR: 1,01,95%C1:1.00-1.01, p<.001) and with diabetes(OR: 2.44, 95%01:1.99-2.99, p<.O01). After controlling for age,BMI, triglyceridesand diabetes, men remain more likely to have NAFLD (OR: 2.50, 95%01: 2.14-2.93, p<.O01), and African Americans are more likely to have NAFLDthan whites (OR: 1.98, 95%C1: 1.72-2.29, p<.O01). Conclusions: NAFLD is the most common cause of abnormal liver enzymes in the general U.S. population. Contrary to popular belief, NAFLD is more common among men than women and among Nrican Americansthan whites. Future research is neededto explore these racial and gender differences.
3M
90 Polyps Judged By EndescopistsTo De Hyperplastic (HP) Require Pathological Examination? Wendy S. Atkin, Imperial Cancer Research Fund, Harrow United Kingdom; Andrew Hart, West Norwich Hosp, Norwich United Kingdom; Christopher R. Cardwell, Claire F. Cook, Rekha Patel, Kevin Kavanagh,Imperial Cancer Research Fund, Harrow United Kingdom; lan C. Talbot, St Mark's Hosp, Harrow United Kingdom BACKGROUNDEndoscopic screening for colorectal cancer (CRC) places a huge burden on pathologists because of high polyp detection rates. Pathological examination of removed polyps is of value only if clinical management is affected eg if CRC or high-risk adenomas (AP) are detected(-> 3, villous, severelydysplastic). Polypsof size -> 1 cm requirecolonoscopy irrespective of pathology. In the UK Flexible Sigmoidoscopy (FS) Screening trial, 40,674 people aged 55-64 underwent screeningand 19,305 distal polyps were removedfrom 10,060 people. AIM To determine the accuracy of endoscopic diagnosis of HPs detected at FS.
A-65
340
MELO Scoring System Not Predictive Of Mortality In Patients Waiting For Liver
0nco-wnakly Recombinant Human EWthropnatin (Epontin Alla) Facilitates Optimal Ribavirin (RBV) Dosing in HepatHis C Virus (HCV)-intected Patients Receiving Interferan-a-2b (IFN)/ROV Combination Therapy Douglas T. Oietedch, Cabrini Medical Ctr, New York, NY; Run Wasserman, infectious Diseases Doctors Medical Group, Walnut Creek, CA; Norbert 6rau, Bronx VAMC, Bronx, NY; Tarek I. Hassanein, UCSD Liver Ctr, San Diego, CA; Edmund J. Bini, VA New York Harbor Heaithcare System, New York, NY; Mark Sulkowski, Johns Hopkins Univ Sch of Medicine, Baltimore, MD
Transplantation Jesus E. Carale, Timothy M. McCashland, Elizabeth Lyden, Daniel F. Schafer, Michael F. Sorrell, Alan Langnas, UNMC, Omaha, NE Background: The Mayo Clinic End-StageLiver Disease(MELD) model developed by Malinchoc et al. was found to accurately predict mortality following elective TIPS. (Hepatology 2000;31:864-871). This same model was also applied to non-liver transplant cirrhotic patients and accurately predicted mortality over 6 months. (GE 2000, A972) Recently, UNOS has proposed the MELD model to prioritize transplant candidates for liver allocation by predicting the risk of a patient s mo~lity on the waiting list. Aim: To determine the validity of the MELD model in predicting mortality in patients listed for liver transplantation. Methods: Seventy patients that were listed for liver transplantation from 1/1/99 to 1011100were reviewed at the University of Nebraska Medical Center (UNMC). Patients risk scores were calculated for predicted survival at 3, 6, and 12 months utilizing the MELD model. The patients actual mortality was compared to the predicted mortality. The date listed for transplantation was used as the start date. Cox proportional-hazard regression was used to evaluate model variables. Results: The patients were separated into two groups, those with a high risk score >1.8 (n=9) and those with low risk score <1.8 (n=61). Eight of nine patients died in the high-risk group, 8/61 died in the low risk group. Four patients were transplanted in the low risk group and none in the high-risk group. Actual and predicted mortality for the 2 risk groups showed that the predictive model is dissimilar to the observed mortality in both risk groups (p =0.0028 high-risk group, p =0.0002 low risk group). In the high-risk patients, the predictive model under estimated mortality, whereas in the low risk group, the predictive over estimated mortality. The only significant predictor of mortality for the model variables was log of creetinine (p =0.0061). Log INR, log bilirubin, nor cause of liver disease was predictive of outcome. Observed verses the predicted survival is illustrated below. Conclusion: The MELD model did not accurately predict mortality when applied to patients who were listed for liver transplantation at UNMC. Validity of this model to stratify patients listed for liver transplantation needs to be studied further prior to implementation.
Background: RBV dose reduction or discontinuation has beenthe standard approachto treating anemia due to IFN/RBV therapy Jn HCV-infected patients. A recent retrospective analysis suggested that maintaining RBV doses greater than 10.6 mg/kg per day of the intended dose was aseecbted with higher sustained virologic response rates. (McHutchison et al. AASLD. 2000:223A. Abstract 247) We conducted an open-label, randomized, muiticenter study to assess the efficacy and tolerability of once-weekly (qw) epoetin afro therapy for hemoglobin (Hb) decreases associated with IFN/RBV therapy and to evaluate the impact of epoetin affa therapy on RBV dosing. Methods: HCV-infected patients who developed Hb levels -< 12 g/ dL during the first 24 weeks of combination IFN/RBVtherapy (n = 60) were randomly assigned to treatment with epoetin aita (40 000 units) subcutaneously qw for up to 36 weeks or standard of care (SOC). The primary and secondary efficacy end points (Week 16) were changes from baseline in Hb level and RBV dose, respectively. Interim results are reported; final results will be presented. Results: At baseline, mean Hb levels and RBV doses were similar between the epoetin aita and SOC groups (11.0 g/dL for both, and 998 vs 986 mg/ day, respectively). At the last available assessment (LAST), the mean changes from baseline Hb levels were +2.7 vs +0.4 g/dL for epoetin alfavs SOC groups (P<.05), respectively; and the mean changes in RBV doses were -25 and -190 mg/day, respectively (P<.O5). The LAST mean Hb level in the epoetln alfa group (13.8 g/alL) was significantly greater (P<.05) than that of the SOC group (11.3 O/dL). Based on LAST, 77% vs 50% (P<.01) of patients in the epoctin alia ve SOC groups, respectively, were treated with RBV doses greater than 10.6 mo/kg per day. Conclusions: Epoetln alfa (40 000 units qw) was well tolerated and effectively increased Hb levels in HCV-infected patients treated with IFN/RBV. Epostin affa increased the proportion of patients receiving RBV doses greater than 10.6 mg/kg per day, suggested as the minimum dose to increase the likelihood of sustained virologic response.
/ 341 Prlinam Hepatitis B Vines Canyiep Lamivodina Resistant Mutations Are Not AnsnstakNI with ~ Replication Fitness and Are Sensitive to Adotovir Robert Ym Cbsn, Paul V. Decmond, St Viocent's Hasp, Melbourne Australia; William E. Debney, Danni L. Coitodge, Rosalind M. Edwards, Victorian Infectious Diseases Reference Lab, Melbourne Australia; Harriet C. Isom, Penn State Coil of Medicine, Hershey, PA; Stephen/L Locamini, Victorian Infectious Diseases Reference Lab, Melbourne Australia
339
IntTodoction: Hepatitis Be antigen (HBeAg) negative chronic hepatitis 8 (CHB) is associated with the G1896A (PC) mutation in the HBV precore region. Prevalence of HBeAg-negative CHB is higher than that of HBeAg positive CHB in some countries. Lamivudine resistant mutations develop in PC mutants at a similar rate during lamivudine therapy as in wild type (WT) HBV. Lamivodine resistant mutations in WT HBV cause a marked decrease in replication fitness, it is unclear if similar impairments occur in PC related lamivudine resistant mutants. In addition, drug sensitivity of lamivudine resistant mutants concurrently carrying the PC mutation are unclear. Aim: To examine the replication fitness of iamivudine resistant variants of precore mutant HBV exhibiting the G1896A using the recently described recombinant HBV paculovirus system (Deianey and Isom. Hepatology 1998;28:1134). In addition, the antiviral drug sensitivity of these mutants was evaluated against lamivudine, edefovir and penciclovir. Methods: The G1896A mutation was introduced into a 1.3 unit length infectious cDNA construct of WT HBV. Further concurrent lamivudine resistant mutations (M5521, M552V/M528M) were introduced, producing clones of PC, PC/MS521, PC/M552V/L528M, WT, WT/M5521, WT/ M552V/L528M Each infectious clone was then used to form recombinant HBV-boculovirus and then used to transduce HepG2 cells. The IC5ofor Penciclovir, Lamivudine and Adetovir were then individually measured using standard dose-response experiments (Colledge et al. 2000;44:551) against intracellular HBV DNA markers as measured by Southern blot. Results: WT and PC were found to have similar replication fitness as described in other studies. Replication fitness of WT/M5501 and W1"/M552V/L528M were much reduced compared with WT. However, PC/M5521and PC/552V/L528M were found to have similar replication fitness as PC and WT. Adetovir was active against PC/M5521and PC/M552V/L528M mutants, indicating no cross resistance between lamivudine and adefovir even in the setting of the PC mutation. The IC50 for PC/MS521and PC/M550V/L528M were similar to that of non drug-resistant PC. Lamivudine and penciclovir were inactive against lamivudine resistant mutants in either WT or PC. Conclusion: Lamivudine resistant mutants are not associated with decrease replication fitness within the framework of PC mutant. HBV with concurrent precore and lamivodine resistant mutations are sensitive to adefovir.
Bile Duct injury in CF Knockout Mice With ColHis: Further Evidence Of An Association Between CFTR Mutations And Primary Sclernsing Chotnagltis Munir M. Zaman, Paola G. Blanco, Rhonda K. Yantiss, Imed A. Nasser, Ignacio Calvo, Sunil G. Sheth, Steven D. Freedman, Beth Israel Deaconess Medical Ctr, Boston, MA We have recently demonstrated a strong association between the occurrence of primary sclerosing choiangitis (PSC) and mutations in one allele of the cystic fibrosis transmembrane conductance regulator (CFTR) gena in patients with inflammatory bowel disease (IBD). This suggests that the carrier state for CFTR mutations may predispose to the development of PSC in patients with IBD. Based on this, we hypothesize that the induction of colitis in homozygous (cftr -~-) or heterozygous (cftr +'-) CF knockout mice would lead to bile duct injury (BDI) and perhaps represent a mouse model of PSC. AIM: To determine whether cftr -/- and cffr +/- mice develop BDI in the setting of colitis. METHODS:Cftr ~1~.~- mice were bred and offspring weaned at 21 days of age on to a complete liquid diet (Peptamen, Nestle Clin.lnc.). On day 40, mice were fed Peptamen with or without 125 mg dextran sodium sulfate (DSS)/day for 10 days to induce colitis. Mice were then sacrificed and serum alkaline phosphatase (AP) was measured. Liver and colonic tissues were submitted for blinded histologic scoring by 2 hepatopathologists. BDI was categorized by the following features: epithelial injury, proliferation, anguiation, and fibrosis. Each of these four features were scored as 0 (absent or mild) or 1 (moderate or severe), then summated for a maximum score of 4. RESULTS:All mice on DSS developed bloody diarrhea and histologic evidence of colitis, most severe in wild type (WT) mice. The control mice [WT no DSS in=6); WT with DSS colitis in=6); and cftr -/- no DSS (n=3)] had no evidence of BDI with a score of O. In contrast, cftr -/- (n=14) and cftr +/- in=5) mice with DSS colitis had mean scores (_+S.E.) of 1.9+-0.4 and 1.4_+0.5,respectively (p
342 S-adennsyimothionlne (SAMe) Modulates IL-10 and TNF Production by Monocytns, Macophaons, and Kuplfer Cells Craig J. McClain, Univ of Louisville, Louisville, KY; Rajender Chawia, Weody Nelson, Waiter Watson, Daniel Skufca, Sankar Boss, Daniell 8. Hill, Univ of Kentucky, Lexington, KY; Shirish Barve, Univ of Louisville, Louisville, KY Background: Use of over the counter complimentary and alternative medicine (CAM) is prevalent in patients with chronic diseases such as liver disease. Indeed, studies indicate that over 40% of patients with chronic liver disease take some form of nutritional supplement or herbal medication. S-adenosylmethlonine (SAMe) is one such one CAM product that has substantial scientific rationale for the treatment of alcoholic liver disease (ALD). There is a well-documented detect in the hepatic transsuffuration pathway in ALD, with a buildup of methionine and a decrease in SAMe. This is due, inpart, to decreased activity of the enzyme
A-64
methionine adenosyltransferase. Many of the metabolic abnormalities and the liver injury in ALD are thought to be due to overproduction of the proflammatory cytokine tumor necrosis factor (TNF), with concomitant inadequate levels of the antiinflammatory cytokine [L-IO. Methods: In these studies we evaluated the role of SAMe in modulating TNF and IL-IO production. We first used a murine macrophagecell line raw 264.7 and incubated cells with increasing concentrationsof SAMeor ethionine (which inducesa functional SAMedeficiency). Compared to control cells, SAMe supplementedcells had a dose dependent increase in IL10 and a dose dependentdecrease in TNF concentrations. In ethionine treated cells, a dose dependent decrease in IL-IO and an increase in TNF was observed. Parallel changes were seen in TNF and IL-IO mRNA. Next, we demonstrated that SAMe was effective at down regulatingTNFproduction in peripheralblood monocytesfrom patientsfrom alcoholic hepatitis. Lastly, we isolated kupffer cells from rats fed a methyl-restricted choline deficient diet, which produces hepaticSAMe deficiency. Isolated kupfter cells spontaneouslyproduced large amounts of TNF and producedsignificantly more TNF in responseto a LPS stimulus. Results: In summary, peripheral blood monocytes from patients with alcoholic hepatitis and kupffer cells from rats with SAMe deficiency both produce significantly increased amounts of TNF in the basaland LPS stimulated state,and this excessTNFproduction could be down regulated with SAMe. In a murine macrophage cell line we documented that SAMe not only down regulated TNF, but also up regulatedthe critical antiinflammatory cytokine IL-IO. This may have major implications for therapeutic intervention in ALD and other forms of chronic liver disease.
345 incidence and Survival of Colorectal Cancer in the US: 1889-189"/ Linda Rabeneck,Hashes B. EI-Serag, Baylor Coil of Medicine, Houston, "IX; Robert S. Sandier, Univ of North Carolina, Chapel Hill, NC Background:The past decadehas witnessedimportant advancesin colorectal (CRC)screening and treatment. Research has shown that in average-risk individuals screening reduces CRC mortality. In addition, randomized controlled trials in CRC patients have shown improved survival with adjuvanttherapies. However,the impact of these advanceson the incidenceand survival of CRC in clinical practice remains unknown. Objectives:To examinetime trends in the incidenceand survival of CRCin the U.S.Methods: Registriesof the Surveillance,Epidemiology, and End Results (SEER) databaseof the National Cancer Institute account for 14% of the U.S. population. Using the SEERdatabasewe examinedtemporal changes in the incidence and survival of CRC by gender, race, site, and stage. Patientswith histologically proven CRC between 1989 and 1997 were identified and analyzedin 3-yr time periods (89-91; 92-94; 9597). We calculated the age-adjusted incidence rates and median, 1-, 2-, and 5-yr observed and relative survival rates. Results: Between 1989 and 1997 we identified 108,057 patients with CRC, 48.6% were women, 84.6% white, and 8.3% black. For all sites the incidencerates were higher in men than women during all time periods. For left and right CRC,the incidence rates were higher in blacks than whites during all time periods. For rectosigmoid CRC, the incidence rates tell over time. For example, from 89-91 to 95-97 the incidence rate fell in white men from 18.1 per 100,000 (17.6-18.6) to 15.1 per 100,000 (14.7-15.6); and in white women from 10.2 (9.8-10.5) to 9.0 (8.7-9.3). Lesser declines were seen in black men and women. Similarly, for left CRC the incidence rates fell over time. For right CRC,the incidence rates remained unchanged over time. For all sites, the incidence rates of metastatic CRC remained unchanged over time. For all sites, the median survival was higher in whites than blacks during all time periods. For all sites there was no change in the 1-yr, 5-yr, or median survival overall, or for subgroups based on gender, race, or stage over time. For example, for rectosigmoid CRC the overall 1-yr observed survival rates were 84.1%, 83.8%, and 83.0% during 89-91,92-94, and 95-97, respectively.Conclusions:The decreasein incidence, particularly for rectosigmoid CRC is consistent with wider use of screening and removal of adenomas, The stable incidence rates for metastatic CRC raise the possibility of a subset of cancers with rapid growth potential that resist early detection and account for the stable survival rates over the past decade.
343 Hepatic Cholesterol and Bile Acid Metabolism and Intestinal Cholesterol Absorption in Scavenger Receptor Class B Type I (SR-BI)-deficient Mice Pablo Mardones, Veronica Ouinones, Ludwig Amigo, Mauricio Moreno, Juan F. Miquel, Pontificia Univ Catolica, Santiago Chile; Margrit Schwarz, Univ of Texas Southwestern Medical Ctr, Dallas, TX; Helena E. Miettinen, Bernardo Trigatti, Monty Krieger, MA Institute of Technology, Cambridge, MA; Sonya Vanpatten, David E. Cohen, Albert Einstein Coil of Medicine, New York, NY; Attilio Rigotti, Pontiticia Univ Catolica, Santiago Chile BACKGROUND:The scavengerreceptor class B type I (SR-BI), which is expressedin the liver and intestine, plays a critical role in cholesterol metabolism in rodents. While hepatic SR-BI expression regulates high density lipoprotein (HDL) cholesterol metabolism, intestinal SR-BI has been proposedto facilitate cholesterol absorption. AIM: To further evaluatethe relevance of SR-BI in the enterohepaticcirculation of cholesterol and bile salts, we studied plasma HDL cholesterol clearance,hepatic cholesteroland bile acid metabolism, biliary lipid secretion, and intestinal cholesterolabsorption in SR-BI knockout mice. RESULTS:SR-BI deficiencyimpaired HDL cholesteryl ester clearanceand HDL cholesterol transfer from plasma into the bile. SRBI knockout mice exhibited a selective decrease in biliary cholesterol secretion, without concomitant changes in either biliary bile acid or phospholipid secretion. Hepatic total and unesterifled cholesterol contents were slightly increasedin SR-BI-deficient mice, while sterol synthesis and low density lipoprotein receptor expressionwere not significantly changed. Bile acid pool size and composition as well as fecal bile acid excretion were not altered in SR-BI knockout mice. Intestinal cholesterol absorption was increasedand fecal neutral sterol excretion was slightly decreasedin SR-BI KO mice relativeto controls. CONCLUSION:Thesefindings establish the critical role of hepatic SR-BI expression in selectively controlling the utilization of HDL cholesterol for biliary secretion. In contrast, SR-BI is not essential for intestinal cholesterol absorption.As a consequenceof its key role the last step of HDL-mediatedreverse cholesterol transport, hepatic SR-81 activity may explain the anti-atherogenicactivity of this lipoprotein receptor in vivo.
346 Complications of Screening Sigmoidoscopy Theodore R. Levin, Kaiser Permaneote,Walnut Creek, CA; Carol Conell, Kaiser Permanente Div of Research, Oakland,CA; Jean A. Shapiro, Centers for Disease Control and Prevention, Atlanta, GA; Shella G. Chazan,Kaiser Permanente Div of Research, Oakland, CA; Marion Nadel, Centers for Disease Control and Prevention,Atlanta, GA; Joe V. Selby, Kaiser Permanente Div of Research, Oakland, CA Purpose: To determine the incidence of serious adverse events following flexible sigmoidoscopy (FS) in a large scale colorectal cancer screening program. Methods: The automated databasesof Northem CaliforniaKaiser Permanentewere accessedto identify hospitalizations within and outside the health care program fogowing 109,534 screening FS procedures conducted on average-risk subjects aged 50 and over during the years 1994-96 as part of the ColorectalCancerPrevention(CoCaP)Program. Charts of patients hospitalizedduring the 4 weeks following FS were reviewed to confirm diagnoses and the reason for FS. Serious complications were defined as perforations, lower gastrointestinal bleeding requiring transfusion, unspecifiedcolitis and diverticulitis requiring surgery. Non-serious hospitalizationswere for fever, abdominalpain and lower GI bleedingnot requiringtransfusion. Automateddatabases were also reviewed to identify hospitalizations for myocardial infarction (MI) and unstable angina, with chart review confirmation. The rate of MI in in these subjects in the first day, first week and first 4 weeks after FS were compared to the rate of MI occurring in the remainderof the 52 week period after FS. Resuits: Nearlyhalf of subjects were female (48,8%), and 74% of exams were performed by non-gastroenterologistphysicians or nurses. Overall, 24 subjects were hospitalizedfor a complication, 7 of which were serious (incidencerate: 21.9/ 100,000 overall,6.4/100,000 for serious complications).The ratesfor perforation, diverticulitis requiring surgery and bleeding requiring transfusion were each 1.81100,000; colitis occurred in 0.9/100,000. In a multivariate logistic regression model, men were more likely than women to experiencea complication, and complicationswere more common in proceduresassociated with biopsy or polypectomy. Hospitalizationfor MI occurred in 28 subjects within 4 weeks of FS, including 8 in the first 7 days. In multivariate models, subjects were no more likely to experienceMI in the first day, week, or month after FS comparedto later periods. Conclusions: FS, administeredas part of a large scale screening program with a diverse group of examiners, is associatedwith a modest rate of complications.While MIs occur in screeningaged subjects following FS, there does not appear to be a measurable increase in risk for MI in the period immediately following FS, compared with later periods.
344 Nonalcoholic Fatty Liver Disease: The Most Common Cause ol Abnormal Liver Enzymes in the U.S. Population Jeanne M. Clark, Frederick L. 8rancati, Anna Mae E. Diehl, Johns Hopkins Univ Sch of Medicine, Baltimore, MD Background: There is increasing evidencethat nonalcoholic fatty liver disease (NAFLO) can progress to fibrosis and eventuallyto cirrhosis. In Europeanand Asian series the prevalence of NAFLD is estimated to be 14-21% and is more common among adults with obesity and/ or diabetes. However, the prevalence of NAFLD in the U.S. population is unknown. Our objective was to estimatethe prevalenceof NAFLDin the United Statesand identify associated clinical factors. Methods: Using data from Third National Health and Nutrition Examination Survey (1988-1994), we examinedthe prevalenceof NAFLD among adults, ages 17 and older (N=13,500). We defined NAFLD using criteria that Ratziu et al. validated by liver biopsy (Gastroenterology2000); i.e. an elevationabove normal of AST, ALT, or GGT (liver enzymes) and negative Hepatitis B and C serulogies, transferrin saturation less than 50%, and average daily alcohol intake < 2 drinks for women and < 3 drinks for men. We used bivariate and multivariate logistic regression to identify factors associated with NAFLD. Analyses were weighted according to the sampling design in order to reflect the U.S. population. Results: The prevalenceof individuals with 1 or more abnormal liver enzymes in the general U.S. population is 27.7%. Only 4.2% of these abnormalities were due to chronic viral hepatitis, alcoholic liver disease, and/or iron overload; hence, the prevalence of NAFLO in the U.S. population is 23.5% (p<.O01). NAFLD is more common in men (33% vs. 17%, p<.O01) in all age groups; howeverthis tendency is most pronounced in individuals under age 30 (25% vs. 9%). Nrican Americans have higher rates of NAFLDthan whites (33% vs. 23%, p<.O01). This is true in both men and women, and in all age groups. The prevalenceof NAFLDincreases with increasing age (OR: 1.01,95%01: 1.01-1.01, p<.O01), BMI (OR 1.09, 95%C1: 1.08-1.10, p<.O01), with increasinglevelsof triglycerides (OR: 1,01,95%C1:1.00-1.01, p<.001) and with diabetes(OR: 2.44, 95%01:1.99-2.99, p<.O01). After controlling for age,BMI, triglyceridesand diabetes, men remain more likely to have NAFLD (OR: 2.50, 95%01: 2.14-2.93, p<.O01), and African Americans are more likely to have NAFLDthan whites (OR: 1.98, 95%C1: 1.72-2.29, p<.O01). Conclusions: NAFLD is the most common cause of abnormal liver enzymes in the general U.S. population. Contrary to popular belief, NAFLD is more common among men than women and among Nrican Americansthan whites. Future research is neededto explore these racial and gender differences.
3M
90 Polyps Judged By EndescopistsTo De Hyperplastic (HP) Require Pathological Examination? Wendy S. Atkin, Imperial Cancer Research Fund, Harrow United Kingdom; Andrew Hart, West Norwich Hosp, Norwich United Kingdom; Christopher R. Cardwell, Claire F. Cook, Rekha Patel, Kevin Kavanagh,Imperial Cancer Research Fund, Harrow United Kingdom; lan C. Talbot, St Mark's Hosp, Harrow United Kingdom BACKGROUNDEndoscopic screening for colorectal cancer (CRC) places a huge burden on pathologists because of high polyp detection rates. Pathological examination of removed polyps is of value only if clinical management is affected eg if CRC or high-risk adenomas (AP) are detected(-> 3, villous, severelydysplastic). Polypsof size -> 1 cm requirecolonoscopy irrespective of pathology. In the UK Flexible Sigmoidoscopy (FS) Screening trial, 40,674 people aged 55-64 underwent screeningand 19,305 distal polyps were removedfrom 10,060 people. AIM To determine the accuracy of endoscopic diagnosis of HPs detected at FS.
A-65