- Email: [email protected]
225 THE PROTEIN FINGERPRINT TECHNOLOGY REFLECTS LIVER FUNCTION AND DETECTS CLINICALLY SIGNIFICANT PORTAL HYPERTENSION IN PATIENTS WITH ALCOHOLIC LIVER CIRRHOSIS
POSTERS to decompensated form. PAS test highlighted a complete sensitivity (100%) in showing the presence of infection in a very short time (15 min), confirmed by the results of positive blood cultures. A greater number of patients is necessary to confirm these data. 223 COST-EFFECTIVENESS OF A NEW MODEL OF SPECIALISTIC CAREGIVING FOR OUTPATIENTS WITH CIRRHOSIS AND ASCITES; A PROSPECTIVE STUDY F. Morando1 , G. Maresio1 , S.S. Piano1 , S. Fasolato1 , M. Cavallin1 , A. Romano1 , S. Rosi1 , E. Gola1 , A. Sticca1 , A.C. Frigo2 , M. Stanco1 , C. Destro3 , G. Rupolo3 , D. Mantoan4 , P. Angeli5 , A. Gatta1 . 1 Department of Medicine, 2 Department of Cardiac, Thoracic and Vascular Sciences, University of Padova, 3 Medical Direction, University and General Hospital of Padova, 4 Direction of Health System of Veneto Region, 5 Department of Medicine, Unit of Hepatic Emergencies and Liver Transplantation, University of Padova, Padova, Italy E-mail: fi[email protected] Background and Aims: The development of ascites in patients with cirrhosis is associated with a high rate of health care utilization. New models of specialistic caregiving support are necessary to optimize its management. The aim of the study was to evaluate the “Day management check-up” as a new model of specialistic caregiving support based on a series of diagnostic facilities performed in real time and on the integrated activity of hepatologists, dedicated nurses and primary physicians, opposed to standard care in outpatients with cirrhosis and ascites. Methods: 100 cirrhotic patients recovered in our hospital for complication of their liver disease were allocated, after dimission, to the “Day management check-up” group (40 patients, group 1), or to the “Standard outpatient care” group (60 patients, group 2), and followed prospectively as outpatients up to death or for at least 12 months. Patients of the two groups could also access to “Day hospital” when an invasive procedure was required. In group 1 the “Day management check-up” and the “Day hospital” taken together defined the “Day management program”. Results: Twelve-month mortality was higher in group 2 than in group 1 (45.7% vs 23.1%, p < 0.025). The rates of 30-day readmission and 12-month readmission were also higher in group 2 (42.4% vs 15.4%, p < 0.01 and 71.2% vs 46.2%, p < 0.025 respectively). The global cost attributable to the management per patient-month of life was lower (1479.19±2184.43 Euros) in group 1 than (2816.13±3893.03 Euros) in group 2 (p < 0.05). Conclusions: The study suggests that this new model of specialist caregiving, based on real time diagnostic facilities and on an integrated team activity reduces 12-month mortality in patients with cirrhosis and ascites. The favorable cost profile, due to a more rational use of hospital services, provides further evidence of the potential value of this model. 224 NON-INVASIVE PREDICTION OF UPPER GASTROINTESTINAL BLEEDING BY RUPTURE OF ESOPHAGEAL VARICES (ROV) USING LIVER AND SPLEEN STIFFNESS (AIXPLORER), LIVER STIFFNESS BY FIBROSCAN AND FIBROTEST M. Rouyer1 , L. Royer2 , H. Perazzo3 , M. Munteanu4 , Y. Ngo4 , E. Luckina5 , L. Fedchuck3 , F. Sattonet3 , P. Lebray3 , M. Rudler3 , D. Thabut3 , V. Ratziu3 , T. Poynard3 . 1 FMPMC, UPMC APHP Liver Center, GH Piti´e Salpˆetri`ere, Paris Cedex, 2 FMPMC, 3 UPMC APHP Liver Center, GH Piti´e Salpˆetri`ere, 4 Hepatology Research Unit, BioPredictive, 5 ANRS, GH Piti´e Salpˆetri`ere, Paris, France E-mail: [email protected] Background: Cirrhotics frequently undergo endoscopic screening for oesophageal varices (OV) to establish primary prevention for ROV. FibroTest (Biopredictive) and the liver stiffness measure (LSM) by Fibroscan (Echosens) have a good NPV for excluding S96
the diagnosis of OV grade II-III (LiverInternational-2006, JHepatol2009). The spleen stiffness by Fibroscan could predict the OV presence but not the grade (JGastroenterolHepatol-2011). Aims: To evaluate the diagnostic values for OV presence and for bleeding by ROV of the liver-stiffness (SWE) and spleen-stiffness (SWE-S) by Aixplorer (Supersonic Imagine) compared to Fibrotest and LSM. Methods: Chronic liver disease patients have undergone prospectively LSM (M and XL-probes), Fibrotest, SWE, SWE-S and LSM. Applicability (App) was determined: for LSM, excluding failures and unreliable LSM (success rate <60%, <10LSM, ratio IQR/median-LSM>=30%); for Fibrotest, according to the manufacturer’s recommendations. Non-App minimum, maximum and mean SWE and SWE-S by Aixplorer, were considered in case of doubt, failures and minimum stiffness=0 kPa. Diagnostic values were expressed by AUROCs. Results: 71-patients had OV-screening in a time interval of 0.1 months (0–1.2 years) from the non-invasive assessment of fibrosis. Non-App were excluded: Fibrotest 3/71 (4.2%), LSM (M/XL) 7/71 (9.9%), SWE-S 4/71 (5.6%) and SWE 10/71 (14%). 57-patients all App were included: age 56 years, 81%men, cirrhosis as per Fibrotest 38/57 (67%) and LSM 40/57 (70%). 26-patients had OV grade-III or grade-II with red signs (II-RS) and 16-patients were admitted for ROV. Median (range) stiffness by Aixplorer were: SWE 12.4 kPa (1.5–146), SWE-Max 23.3 kPa (4–300), SWE-S 29.4 kPa (0.4–149) and SWE-S-Max 76 kPa (1.2–179). Only LSM (p = 0.0003) and Fibrotest (p = 0.04) excluded OV (absence vs OV grade-I or higher), AUROCs: 0.76 vs 0.65 (p = 0.17) with PPV 0.97 for LSM (cutoff 20 kPa) and Fibrotest 0.94 (cut-off 0.80). For excluding OV, both LSM and Fibrotest had higher AUROCs than SWE (0.62), SWE-Max (0.62), SWE-S (0.54) and SWE-S-Max (0.53), all p < 0.05. For OV grade-III and grade II-RS AUROCs were: SWE-Max 0.77, SWE 0.76 and LSM 0.73, all higher than the SWE-S-Max, SWE-S and FT (all p < 0.05). In 10 cirrhotics with OV grade-III, only SWE-S-Max was predictive for ROV (AUROC 0.84, p = 0.01). Conclusion: Previously validated LSM and Fibrotest had the best PPV to exclude OV. For OV grade-III and grade-II-RS liver-stiffness was better than the spleen-stiffness. Maximum spleen stiffness by Aixplorer was associated with ROV in cirrhotics with OV grade-III. 225 THE PROTEIN FINGERPRINT TECHNOLOGY REFLECTS LIVER FUNCTION AND DETECTS CLINICALLY SIGNIFICANT PORTAL HYPERTENSION IN PATIENTS WITH ALCOHOLIC LIVER CIRRHOSIS M.J. Nielsen1,2 , D.J. Leeming1 , M.A. Karsdal1 , J. Trebicka3 , I. Byrjalsen1 , F. Bendtsen2 , S. Møller4 , A. Krag2 . 1 Fibrosis Biology and Biomarkers, Nordic Bioscience A/S, Herlev, 2 Department of Gastroenterology, Hvidovre Hospital, Faculty of Health Sciences, Universiity of Copenhagen, Copenhagen, Denmark; 3 Department of Internal Medicine I, University of Bonn, Bonn, Germany; 4 Department of Clinical Physiology, Hvidovre Hospital, Faculty of Health Sciences, Universiity of Copenhagen, Copenhagen, Denmark E-mail: [email protected] Introduction: Liver fibrosis increases intrahepatic resistance, a crucial step in development of portal hypertension. Fibrosis is a dynamic process, which requires permanent remodeling of extracellular matrix (ECM). Small fragments of ECM, the Protein Fingerprint markers, generated during this remodeling and released into the circulation might reflect portal hypertension. This study investigates the relationship between a panel of these markers and clinical data, and its ability to predict the hepatic venous pressure gradient (HVPG). Methods: In plasma from 94 cirrhotic patients and 20 controls without liver disease ECM degradation markers (C1M (type I collagen), C3M (type III collagen), C4M (type IV collagen), C5M
Journal of Hepatology 2013 vol. 58 | S63–S227
POSTERS (type V collagen), C6M (type VI collagen), BGM (Biglycan), ELM (Elastin), and CRPM (CRP)), and ECM formation markers (P3NP (type III collagen) and P4NP7S (type IV collagen)) were measured. Results: The markers measured in hepatic venous blood correlated to the level measured in arterial blood (R=0.89–0.98; p < 0.0001). In hepatic venous blood, all markers correlated directly to Child Score and MELD, and inversely to ICG clearance and serum albumin, strongest with P3NP (R=0.46, 0.48, −0.53, and −0.46; respectively, p < 0.0001). All markers except ELM were inversely correlated to hematocrit and to hemoglobin, especially C4M and C1M showed strongest correlations (R = −0.31; −0.40; respectively, p = 0.001p < 0.0001). In both arterial and hepatic venous blood all markers except CRPM correlated to HVPG (e.g. P3NP for both sites R = 0.47, p < 0.0001). A multiple regression analysis including P3NP, C6M and MELD improved the correlation (R = 0.62, p < 0.0001). P3NP could clearly separate controls from HVPG levels <10 mmHg (p < 0.01) and those with HVPG<10 mmHg from those with HVPG levels ≥10 mmHg (p < 0.0001). C4M, C5M, and ELM were all significantly higher in patients with HVPG levels ≥12 mmHg compared to lower HVPG levels (p < 0.01–0.0001). Conclusion: The Protein Fingerprint markers reflect liver function and stage of disease. A strong correlation between the markers measured in hepatic venous and arterial blood was observed. A multimarker model in combination with clinical scores predicted HVPG and separated clinical relevant HVPG thresholds noninvasively. Therefore these markers and their models are suitable for non-invasive evaluation of portal hypertension in cirrhosis. 226 LONG-TERM ANTIBIOTIC AND PROTON-PUMP INHIBITOR USE ARE STRONG PREDICTORS OF RECURRENT INFECTIONS IN CIRRHOSIS: A PROSPECTIVE MULTI-CENTER STUDY FROM NACSELD J.G. O’Leary1 , K.R. Reddy2 , F. Wong3 , P.S. Kamath4 , H.M. Patton5 , S.W. Biggins6 , M.B. Fallon7 , G. Garcia-Tsao8 , R.M. Subramanian9 , G. Brown10 , R. Malik11 , L.R. Thacker12 , J.S. Bajaj12 , NACSELD. 1 Hepatology, Baylor University Medical Center, Dallas, TX, 2 Medicine, University of Pennsylvania, Philadelphia, PA, USA; 3 Medicine, University of Toronto, Toronto, ON, Canada; 4 Medicine, Mayo Clinic School of Medicine, Rochester, MN, 5 Medicine, University of California, San Diego, San Diego, CA, 6 Medicine, University of Colorado, Denver, CO, 7 Medicine, University of Texas Health Science Center, Houston, TX, 8 Medicine, Yale University School of Medicine, New Haven, CT, 9 Medicine, Emory University, Atlanta, GA, 10 Medicine, University of Texas Southwestern, Dallas, TX, 11 Medicine, Beth Isreal Deaconess, Boston, MA, 12 Medicine, Virginia Commonwealth University, Richmond, VA, USA E-mail: [email protected] Infections in cirrhosis negatively impact prognosis and transplant candidacy. However, in survivors, factors associated with risk of recurrent infections remain unclear. Aim: To determine risk factors for recurrent infections in cirrhotic patients following resolution of initial infection. Methods: NACSELD comprises North-American tertiary-care hepatology centers that prospectively enter data on hospitalized cirrhotics into a centralized database. Data collected include demographics, co-morbidities, infections, and medications. This report focuses on 6-month post hospital discharge outcomes of infected cirrhotic patients. Descriptive statistics and logistic regression modeling using recurrent infection as the outcome were performed. Results: 377 patients from 14 sites were evaluated; during the index hospitalization 17 (5%) were transplanted and 77 (20%) died, leaving 283 eligible for post-discharge outcomes. 193 patients reached the 6-month follow-up goal: 21 (11%) were transplanted, 56 (29%) died without transplant, and 116 (60%) remain alive without transplant. Following discharge, 175 patients had infection
outcomes reported. In a median follow-up of 4.5 months, 82 (47%) had recurrent infections. 67% of the 175 cirrhotics were on protonpump inhibitors (PPI), 47% on rifaximin and 45% on SBP prophylaxis. Patients with recurrent infections were older (58 vs. 55 yrs, p = 0.04) with similar Child (9.6 vs. 10.4, p = 0.10) and MELD scores (19 vs. 19, p = 0.12). There were significantly higher rates of PPI and antibiotic use in patients with recurrent infections (PPI: 84% of patients with recurrent infection vs. 52% uninfected, p < 0.0001, rifaximin: 67% recurrent infection vs. 30% uninfected, p < 0.0001 and SBP prophylaxis: 61% recurrent infection vs. 31% uninfected, p < 0.0001). There were no significant differences in MELD/Child scores between patients on PPI, rifaximin or SBP prophylaxis and the rest. Logistic regression (variables included were Child score, medications, age, gender, prior infections) showed only rifaximin (OR = 3.93, CI 1.97– 7.84), PPI (OR = 3.67, CI 1.65–8.17), SBP prophylaxis (OR = 2.34, CI 1.16–4.71), and age (OR = 1.04, CI 1–1.08) as predictors of recurrent infections; Child and MELD score were not significant. In conclusion, even after resolution of the index infection, cirrhotic patients remain at high risk for recurrent infections (47%). Patients at highest risk for recurrent infections are previously infected older patients, and those on rifaximin, PPI, and SBP prophylaxis. Primary and secondary infection prevention strategies are needed to improve patient outcomes. 227 EVALUATION OF ADRENAL INSUFFICIENCY IN CIRRHOTIC PATIENTS IN KOREA S.R. Shin, H. Yoo, I.H. Moh, S.H. Park, M.S. Lee. Department of Internal Medicine, Kangnam Sacred Heart Hospital, Hallym University, Seoul, Republic of Korea E-mail: [email protected] Background and Aims: Adrenal insufficiency (AI) is frequent in cirrhotic patients with severe infection or hemodynamic instability. However, there is limited data about the prevalence of AI in stable cirrhotic patients and the impact of etiology of cirrhosis on AI. This study prospectively evaluated prevalence and characteristic of AI in stable cirrhotic using the conventional short synacthen, a synthetic analogue of the naturally-occurring adrenocorticotrophic hormone. Methods: From July 2011 to Aug 2012, a total of 50 cirrhotic patients without infections or hemodynamic instability admitted to two university-affiliated hospitals were enrolled. AI was defined by a total serum cortisol level <18 mg/dL at 30 or 60 min after injection of 250 mg of synacthen.
Results: Thirty-eight patients were men and age was ranged 29–89 years. Etiology of cirrhosis was alcohol/viral hepatitis/both/others in 36/14/9/6 patients. AI was present in 22 patients (44%). No significant difference was observed regarding age, gender, mean arterial pressures, and heart rates between patients with and without AI. The etiology of cirrhosis and degree of alcohol consumption did not affect presence of AI or basal and peak serum cortisol levels either. Total bilirubin and prothrombin time were higher, while albumin and cholesterol levels were lower in patients with AI than those without AI. However, in multivariate analysis, there was no independent predictor of AI. The prevalence
Journal of Hepatology 2013 vol. 58 | S63–S227
S97
the diagnosis of OV grade II-III (LiverInternational-2006, JHepatol2009). The spleen stiffness by Fibroscan could predict the OV presence but not the grade (JGastroenterolHepatol-2011). Aims: To evaluate the diagnostic values for OV presence and for bleeding by ROV of the liver-stiffness (SWE) and spleen-stiffness (SWE-S) by Aixplorer (Supersonic Imagine) compared to Fibrotest and LSM. Methods: Chronic liver disease patients have undergone prospectively LSM (M and XL-probes), Fibrotest, SWE, SWE-S and LSM. Applicability (App) was determined: for LSM, excluding failures and unreliable LSM (success rate <60%, <10LSM, ratio IQR/median-LSM>=30%); for Fibrotest, according to the manufacturer’s recommendations. Non-App minimum, maximum and mean SWE and SWE-S by Aixplorer, were considered in case of doubt, failures and minimum stiffness=0 kPa. Diagnostic values were expressed by AUROCs. Results: 71-patients had OV-screening in a time interval of 0.1 months (0–1.2 years) from the non-invasive assessment of fibrosis. Non-App were excluded: Fibrotest 3/71 (4.2%), LSM (M/XL) 7/71 (9.9%), SWE-S 4/71 (5.6%) and SWE 10/71 (14%). 57-patients all App were included: age 56 years, 81%men, cirrhosis as per Fibrotest 38/57 (67%) and LSM 40/57 (70%). 26-patients had OV grade-III or grade-II with red signs (II-RS) and 16-patients were admitted for ROV. Median (range) stiffness by Aixplorer were: SWE 12.4 kPa (1.5–146), SWE-Max 23.3 kPa (4–300), SWE-S 29.4 kPa (0.4–149) and SWE-S-Max 76 kPa (1.2–179). Only LSM (p = 0.0003) and Fibrotest (p = 0.04) excluded OV (absence vs OV grade-I or higher), AUROCs: 0.76 vs 0.65 (p = 0.17) with PPV 0.97 for LSM (cutoff 20 kPa) and Fibrotest 0.94 (cut-off 0.80). For excluding OV, both LSM and Fibrotest had higher AUROCs than SWE (0.62), SWE-Max (0.62), SWE-S (0.54) and SWE-S-Max (0.53), all p < 0.05. For OV grade-III and grade II-RS AUROCs were: SWE-Max 0.77, SWE 0.76 and LSM 0.73, all higher than the SWE-S-Max, SWE-S and FT (all p < 0.05). In 10 cirrhotics with OV grade-III, only SWE-S-Max was predictive for ROV (AUROC 0.84, p = 0.01). Conclusion: Previously validated LSM and Fibrotest had the best PPV to exclude OV. For OV grade-III and grade-II-RS liver-stiffness was better than the spleen-stiffness. Maximum spleen stiffness by Aixplorer was associated with ROV in cirrhotics with OV grade-III. 225 THE PROTEIN FINGERPRINT TECHNOLOGY REFLECTS LIVER FUNCTION AND DETECTS CLINICALLY SIGNIFICANT PORTAL HYPERTENSION IN PATIENTS WITH ALCOHOLIC LIVER CIRRHOSIS M.J. Nielsen1,2 , D.J. Leeming1 , M.A. Karsdal1 , J. Trebicka3 , I. Byrjalsen1 , F. Bendtsen2 , S. Møller4 , A. Krag2 . 1 Fibrosis Biology and Biomarkers, Nordic Bioscience A/S, Herlev, 2 Department of Gastroenterology, Hvidovre Hospital, Faculty of Health Sciences, Universiity of Copenhagen, Copenhagen, Denmark; 3 Department of Internal Medicine I, University of Bonn, Bonn, Germany; 4 Department of Clinical Physiology, Hvidovre Hospital, Faculty of Health Sciences, Universiity of Copenhagen, Copenhagen, Denmark E-mail: [email protected] Introduction: Liver fibrosis increases intrahepatic resistance, a crucial step in development of portal hypertension. Fibrosis is a dynamic process, which requires permanent remodeling of extracellular matrix (ECM). Small fragments of ECM, the Protein Fingerprint markers, generated during this remodeling and released into the circulation might reflect portal hypertension. This study investigates the relationship between a panel of these markers and clinical data, and its ability to predict the hepatic venous pressure gradient (HVPG). Methods: In plasma from 94 cirrhotic patients and 20 controls without liver disease ECM degradation markers (C1M (type I collagen), C3M (type III collagen), C4M (type IV collagen), C5M
Journal of Hepatology 2013 vol. 58 | S63–S227
POSTERS (type V collagen), C6M (type VI collagen), BGM (Biglycan), ELM (Elastin), and CRPM (CRP)), and ECM formation markers (P3NP (type III collagen) and P4NP7S (type IV collagen)) were measured. Results: The markers measured in hepatic venous blood correlated to the level measured in arterial blood (R=0.89–0.98; p < 0.0001). In hepatic venous blood, all markers correlated directly to Child Score and MELD, and inversely to ICG clearance and serum albumin, strongest with P3NP (R=0.46, 0.48, −0.53, and −0.46; respectively, p < 0.0001). All markers except ELM were inversely correlated to hematocrit and to hemoglobin, especially C4M and C1M showed strongest correlations (R = −0.31; −0.40; respectively, p = 0.001p < 0.0001). In both arterial and hepatic venous blood all markers except CRPM correlated to HVPG (e.g. P3NP for both sites R = 0.47, p < 0.0001). A multiple regression analysis including P3NP, C6M and MELD improved the correlation (R = 0.62, p < 0.0001). P3NP could clearly separate controls from HVPG levels <10 mmHg (p < 0.01) and those with HVPG<10 mmHg from those with HVPG levels ≥10 mmHg (p < 0.0001). C4M, C5M, and ELM were all significantly higher in patients with HVPG levels ≥12 mmHg compared to lower HVPG levels (p < 0.01–0.0001). Conclusion: The Protein Fingerprint markers reflect liver function and stage of disease. A strong correlation between the markers measured in hepatic venous and arterial blood was observed. A multimarker model in combination with clinical scores predicted HVPG and separated clinical relevant HVPG thresholds noninvasively. Therefore these markers and their models are suitable for non-invasive evaluation of portal hypertension in cirrhosis. 226 LONG-TERM ANTIBIOTIC AND PROTON-PUMP INHIBITOR USE ARE STRONG PREDICTORS OF RECURRENT INFECTIONS IN CIRRHOSIS: A PROSPECTIVE MULTI-CENTER STUDY FROM NACSELD J.G. O’Leary1 , K.R. Reddy2 , F. Wong3 , P.S. Kamath4 , H.M. Patton5 , S.W. Biggins6 , M.B. Fallon7 , G. Garcia-Tsao8 , R.M. Subramanian9 , G. Brown10 , R. Malik11 , L.R. Thacker12 , J.S. Bajaj12 , NACSELD. 1 Hepatology, Baylor University Medical Center, Dallas, TX, 2 Medicine, University of Pennsylvania, Philadelphia, PA, USA; 3 Medicine, University of Toronto, Toronto, ON, Canada; 4 Medicine, Mayo Clinic School of Medicine, Rochester, MN, 5 Medicine, University of California, San Diego, San Diego, CA, 6 Medicine, University of Colorado, Denver, CO, 7 Medicine, University of Texas Health Science Center, Houston, TX, 8 Medicine, Yale University School of Medicine, New Haven, CT, 9 Medicine, Emory University, Atlanta, GA, 10 Medicine, University of Texas Southwestern, Dallas, TX, 11 Medicine, Beth Isreal Deaconess, Boston, MA, 12 Medicine, Virginia Commonwealth University, Richmond, VA, USA E-mail: [email protected] Infections in cirrhosis negatively impact prognosis and transplant candidacy. However, in survivors, factors associated with risk of recurrent infections remain unclear. Aim: To determine risk factors for recurrent infections in cirrhotic patients following resolution of initial infection. Methods: NACSELD comprises North-American tertiary-care hepatology centers that prospectively enter data on hospitalized cirrhotics into a centralized database. Data collected include demographics, co-morbidities, infections, and medications. This report focuses on 6-month post hospital discharge outcomes of infected cirrhotic patients. Descriptive statistics and logistic regression modeling using recurrent infection as the outcome were performed. Results: 377 patients from 14 sites were evaluated; during the index hospitalization 17 (5%) were transplanted and 77 (20%) died, leaving 283 eligible for post-discharge outcomes. 193 patients reached the 6-month follow-up goal: 21 (11%) were transplanted, 56 (29%) died without transplant, and 116 (60%) remain alive without transplant. Following discharge, 175 patients had infection
outcomes reported. In a median follow-up of 4.5 months, 82 (47%) had recurrent infections. 67% of the 175 cirrhotics were on protonpump inhibitors (PPI), 47% on rifaximin and 45% on SBP prophylaxis. Patients with recurrent infections were older (58 vs. 55 yrs, p = 0.04) with similar Child (9.6 vs. 10.4, p = 0.10) and MELD scores (19 vs. 19, p = 0.12). There were significantly higher rates of PPI and antibiotic use in patients with recurrent infections (PPI: 84% of patients with recurrent infection vs. 52% uninfected, p < 0.0001, rifaximin: 67% recurrent infection vs. 30% uninfected, p < 0.0001 and SBP prophylaxis: 61% recurrent infection vs. 31% uninfected, p < 0.0001). There were no significant differences in MELD/Child scores between patients on PPI, rifaximin or SBP prophylaxis and the rest. Logistic regression (variables included were Child score, medications, age, gender, prior infections) showed only rifaximin (OR = 3.93, CI 1.97– 7.84), PPI (OR = 3.67, CI 1.65–8.17), SBP prophylaxis (OR = 2.34, CI 1.16–4.71), and age (OR = 1.04, CI 1–1.08) as predictors of recurrent infections; Child and MELD score were not significant. In conclusion, even after resolution of the index infection, cirrhotic patients remain at high risk for recurrent infections (47%). Patients at highest risk for recurrent infections are previously infected older patients, and those on rifaximin, PPI, and SBP prophylaxis. Primary and secondary infection prevention strategies are needed to improve patient outcomes. 227 EVALUATION OF ADRENAL INSUFFICIENCY IN CIRRHOTIC PATIENTS IN KOREA S.R. Shin, H. Yoo, I.H. Moh, S.H. Park, M.S. Lee. Department of Internal Medicine, Kangnam Sacred Heart Hospital, Hallym University, Seoul, Republic of Korea E-mail: [email protected] Background and Aims: Adrenal insufficiency (AI) is frequent in cirrhotic patients with severe infection or hemodynamic instability. However, there is limited data about the prevalence of AI in stable cirrhotic patients and the impact of etiology of cirrhosis on AI. This study prospectively evaluated prevalence and characteristic of AI in stable cirrhotic using the conventional short synacthen, a synthetic analogue of the naturally-occurring adrenocorticotrophic hormone. Methods: From July 2011 to Aug 2012, a total of 50 cirrhotic patients without infections or hemodynamic instability admitted to two university-affiliated hospitals were enrolled. AI was defined by a total serum cortisol level <18 mg/dL at 30 or 60 min after injection of 250 mg of synacthen.
Results: Thirty-eight patients were men and age was ranged 29–89 years. Etiology of cirrhosis was alcohol/viral hepatitis/both/others in 36/14/9/6 patients. AI was present in 22 patients (44%). No significant difference was observed regarding age, gender, mean arterial pressures, and heart rates between patients with and without AI. The etiology of cirrhosis and degree of alcohol consumption did not affect presence of AI or basal and peak serum cortisol levels either. Total bilirubin and prothrombin time were higher, while albumin and cholesterol levels were lower in patients with AI than those without AI. However, in multivariate analysis, there was no independent predictor of AI. The prevalence
Journal of Hepatology 2013 vol. 58 | S63–S227
S97