E McFarlane

Spectrum of liver histology in presumed decompensated alcoholic liver disease.

BACKGROUND:In presumed decompensated alcoholic liver disease (ALD; liver decompensation, heavy alcohol intake, and negative results of noninvasive screening for other causes), liver biopsy is often performed to assess severity of liver injury and to rule out other liver diseases.AIM:The aim of the study is to describe the spectrum of liver histology in such patients.METHODS:We reviewed all patients with presumed decompensated ALD seen between 1998 and 2004, in whom liver tissue was available for histology (N = 110).RESULTS:A total of 104 of the 110 patients had at least one of the histological features suggestive of ALD: fat, Mallorys hyalin, neutrophilic infiltrate, and hepatocyte ballooning. These features were more prevalent in tissue obtained within a month after presentation with decompensation than in that obtained before decompensation or more than 1 month after. These features were also associated with more severe liver dysfunction. Histology revealed a major additional diagnosis (Budd-Chiari syndrome) in only one case. In 41 patients biopsied within a month of first presentation with decompensation, Child score and Maddrey discriminant function (DF), but none of the histological features, were predictive of survival by Cox multivariate analysis. Of the 26 of these 41 patients with a Maddrey DF >32, 22 (85%) had alcoholic hepatitis.CONCLUSIONS:In patients with presumed decompensated ALD, other liver diseases are uncommon. Routine liver biopsy is of limited added value but biopsy should be considered in those in whom the noninvasive workup, or failure to recover despite abstinence, raises the possibility of other liver diseases.

Effectiveness of an upper-gastrointestinal haemorrhage unit: a prospective analysis of 900 consecutive cases using the Rockall score as a method of risk standardisation

Objectives To assess the effectiveness of a centralised upper-gastrointestinal haemorrhage (UGIH) unit. Methods The UK Audit of acute UGIH resulted in the formulation of a simple numerical scoring system. The Rockall score categorises patients by risk factors for death and allows case-mix comparisons. A total of 900 consecutive patients admitted to a UGIH unit between October 1995 and July 1998 were analysed prospectively. Patients were given an initial Rockall score and, if endoscopy was performed, a complete score. This method of risk stratification allowed the proportion of deaths (in our study) to be compared with the National Audit using risk standardised mortality ratios. Results The distribution of both initial and final Rockall scores was significantly higher in our study than in the National Audit. A total of 73 (8.1%) patients died, compared with the National Audit mortality of 14%. Risk-standardised mortality ratios using both initial and complete Rockall scores were significantly lower in our study when compared with those in the National Audit. Conclusion A specialised UGIH unit is associated with a lower proportion of deaths from UGIH, despite comprising a greater number of high-risk patients than the National Audit. This lower mortality therefore cannot be attributed to a more favourable case mix and demonstrates that further improvements in mortality for UGIH can be made.

Clinical significance of azathioprine metabolites for the maintenance of remission in autoimmune hepatitis

Azathioprine (AZA) is used to maintain remission in autoimmune hepatitis (AIH), but up to 18% of patients are unresponsive. AZA is a prodrug, and the formation of active thioguanine nucleotide (TGN) metabolites varies widely. We aimed to assess the relationship between AZA metabolite concentrations (i.e., TGNs and methylmercaptopurine nucleotides [MeMPNs]), thiopurine methyltransferase (TPMT) activity, therapeutic response, and toxicity in adult patients with AIH prescribed a stable dose of AZA for the maintenance of remission. Red blood cell (RBC) TGNs and MeMPNs were measured in serial blood samples over a 2‐year period. The average TGNs (avTGNs) and MeMPNs (avMeMPNs) concentrations for each patient were used for analysis. Therapeutic response was defined as the ability to maintain remission, defined as a normal serum alanine aminotransferase (ALT) level (ALT <33 IU/mL). Patients who maintained remission (n = 53), compared to those who did not (n = 17), tended to be on lower doses of AZA (1.7 versus 2.0 mg/kg/day; P = 0.08), but had significantly higher concentrations of avTGN (237 versus 177 pmol/8 × 108 RBCs; P = 0.025). There was no difference in MeMPN concentrations or TPMT activities between the two groups. There was a negative correlation between ALT and avTGN (rs = −0.32; P = 0.007). An avTGN concentration of >220 pmol/8 × 108 RBCs best predicted remission, with an odds ratio of 7.7 (P = 0.003). There was no association between TGN, MeMPN, or TPMT activity and the development of leucopenia. Two patients developed AZA‐induced cholestasis and the avMeMPN concentration was higher in those patients, compared to those who did not (14,277 versus 1,416 pmol/8 × 108 RBCs). Conclusion: TGN concentrations of >220 pmol/8 × 108 RBCs are associated with remission. TGN measurement may help identify inadequate immunosupression. AZA‐induced cholestasis was associated with increased MeMPN concentrations. (HEPATOLOGY 2012)

Long-Term Prognostic Significance of Persisting Histological Activity Despite Biochemical Remission in Autoimmune Hepatitis.

Objectives:Biochemical remission is widely considered a satisfactory treatment end point in autoimmune hepatitis (AIH). The significance of persisting histological activity despite biochemical remission is unknown. We aimed to assess the frequency and prognostic significance of persisting histological inflammation in patients with AIH who had achieved biochemical remission with treatment.Methods:We studied 120 patients (median age at diagnosis 57 years; 81% female) with AIH by International Criteria (59% definite), who received immunosuppressive treatment and underwent a follow-up liver biopsy after at least 6 months of sustained biochemical remission (defined as normal serum ALT and globulin).Results:Fifty-five patients (46%) had persisting histological activity (Ishak histological activity index (HAI) ≥4). These patients had higher serum ALT (24 vs. 18 IU/l, P=0.003) and AST (27 vs. 23 IU/l, P=0.03) at the time of follow-up biopsy, compared with patients who achieved histological remission (HAI ≤3). They had less frequent regression of fibrosis on follow-up biopsy compared with those achieving histological remission (32 vs. 60%, P=0.004) and had excess mortality (standardized mortality ratio 1.4 vs. 0.7, P<0.05). The excess mortality was due to liver disease. On multivariate analysis, persisting histological activity was independently associated with all-cause death/transplantation (HR 3.1 (95% CI 1.2–8.1); P=0.02); an association with liver-related death/transplantation fell short of significance (HR 9.7 (95% CI 0.84–111.6; P=0.07).CONCLUSIONS:Persisting histological activity, despite biochemical remission, is frequent in patients with treated AIH and is associated with lower rates of fibrosis regression and reduced long-term survival.

PTU-136 Does A Normal Igg Indicate Histological Remission In Autoimmune Hepatitis (aih)?

Introduction Response to immunosuppressive treatment in AIH is often monitored by measurement of serum immunoglobulin G (IgG) as well as ALT. It is commonly assumed that serum IgG level correlates with histological activity (or Ishak necroinflammatory score: NIS) on liver biopsy, the historical “gold standard”. However, only one group (Luth et al 2008; J of Clin 42(8):926–930.) have examined this relationship, finding that normalisation of both serum ALT and IgG reliably predicted a NIS of <6 but not a NIS of <4 (corresponding to minimal hepatitis, seen in less than half of the patients presumed to be in remission). Here, we aimed to reassess how well serum IgG correlated with NIS in treated patients with AIH undergoing follow-up biopsy for confirmation of disease remission. Methods We assessed 31 follow-up biopsies, performed to confirm histological remission in 28 patients with AIH (International Group criteria; Alvarez J Hepatol 1999; 31:929) on immunosuppressive treatment and an accompanying serum IgG (measured within 6 weeks of biopsy). Results For 29 of the 31 follow-up biopsies, accompanying serum IgG was in the normal range (≤16 gm/L). However, only 13 of these 31 biopsies showed NIS <4 (minimal hepatitis). On ROC analysis, area under the curve (AUC) for IgG in predicting a NIS of ≥4 (n = 31) was 0.596 (p = 0.368). Sensitivity and specificity of IgG (cut off >16 grm/L) in predicting a NIS of ≥4 was 5.5% and 57% respectively. Corresponding positive (PPV) and negative (NPV) predictive values were only 50% and 59% respectively. AUC for change in IgG (∆IgG: baseline values minus values accompanying follow up biopsy; n = 29) in predicting NIS >4 was 0.551 (p = 0.642). Defining histological remission instead as minimal or mild hepatitis- NIS <6 (as Luth’s group did because they found that such patients did not develop fibrosis progression), there were still 4/29 (14%) patients with normal serum IgG who were not in histological remission. On ROC analysis, AUC for IgG in prediction of NIS >6 on follow-up biopsy (n = 31) was 0.62 (p = 0.39). PPV and NPV for serum IgG (cut off >16 grm/L) in predicting of NIS ≥6) was 50% and 86% respectively. AUC for ∆IgG in prediction of NIS ≥6 was 0.608 (p = 0.453). Conclusion Normalisation of serum IgG values in patients with AIH following treatment is not predicative of histological remission. At present, this requires a liver biopsy. Disclosure of Interest None Declared.

PTU-085 Decompensated Alcoholic Liver Disease (ALD): High Long-Term Mortality Despite Initial Survival

Introduction Whilst early management and outcome of decompensated ALD has been extensively studied, there are few published data on long-term outcome. We have previously (McFarlane, Gut 2006; 55:A2 and 55:A36) reported on early and 4-yr mortality in 249 patients (163 men, age (mean(range)) 50(27–77) yr) admitted consecutively to our unit between 1/4/1998 and 31/12/2005 with first presentation of decompensated ALD (Child grade B or C).Here we aimed to assess long-term mortality and its associations in this cohort. Methods We reviewed available hospital records, and death certificates and contacted surviving patients and general practitioners to assess who had died, the causes of death and the patients’ overall alcohol drinking behaviour subsequent to the index hospital episode (classified as: abstinent, continued drinking but reduced, and did not reduce). Results 37 patients died during the index hospital episode, all because of liver disease. The other 212 patients (including one transplanted during the index episode) were followed up for 4.3 (0.03–13.0) years. Only one other patient was transplanted. 154 patients have subsequently died. Cause of death is known in 134 (87%) and was due to liver disease in 95 (71%) of these. Only 4 patients died of hepatocellular carcinoma. Overall 5- and 10- year total mortality rates were 52+(SEM)4% and 75+3% respectively; corresponding rates from causes known to be liver related were 41+5% and 51+4%. Patients who were abstinent (n = 52) had lower total and known liver-related mortality (61+9% and 20+6% after 10 yr) compared to those who continued but reduced (n = 105; 73+5% p = 0.122 and 53+6% p = 0.013) and to those who did not reduce (n = 53; 91+4% p < 0.001 and 71+7% p < 0.001). In Cox regression analysis, both total and known liver-related mortality were independent of age, gender and severity of liver dysfunction at index presentation (Child, MELD, Glasgow and Maddrey scores) but were strongly associated with subsequent drinking behaviour (both p < 0.001) and inversely associated with serum albumin at discharge following index hospital episode (p = 0.001 and 0.019). Conclusion Patients with decompensated ALD who survive their first hospital episode have high long-term mortality, mainly due to liver disease, which is reduced but not prevented by abstinence. Disclosure of Interest None Declared

Comments on ‘Early features of acute-on-chronic liver failure: a prospective cohort study’

We read with interest the study by Katoonizadeh et al 1 of acute-on-chronic (alcoholic) liver failure (ACLF) and the accompanying Editorial,2 which cites our report3 of liver histology in decompensated alcoholic liver disease (ALD). We found features of alcoholic hepatitis (intralobular neutrophils, Mallory bodies and ballooning) to be more frequent and more severe in patients biopsied within a month of first onset of decompensation, compared with those biopsied before decompensation or later, and argued for a causal relationship. We acknowledge the strengths of the prospective study of Katoonizadeh et al 1 but would interpret their results differently. Their definition of ACLF1 excludes patients with no prior liver disease and also those without cirrhosis on biopsy. It thus excludes many patients with first presentation …

Treatment of autoimmune hepatitis: what is the optimal end point on follow-up biopsy?

Introduction The authors have reported1 that following immunosuppressive treatment of Autoimmune Hepatitis (AIH), many patients fail to achieve histological remission (necro-inflammatory score (NIS) ≥3), despite attaining biochemical remission (normal serum ALT). A recent report2 suggests that, following treatment, a NIS≥5 is not associated with fibrosis progression and hence, may be an acceptable treatment outcome.2 The authors therefore aimed to assess the associations of NIS on follow-up biopsy within the range 0–5, with change in fibrosis and survival. Methods The authors studied 114 patients with AIH by IAIHG criteria (81 definite, 94 female, mean age 48.7±1.7 years), treated initially with reducing dose prednisolone and 1 mg/kg azathioprine, who had achieved normal serum ALT and a NIS between 0 and 5 on follow-up biopsy (performed at median (range) 2.16 (0.65–13.67) years) after diagnosis (paired diagnostic and follow-up biopsies available in 93 patients). Biopsies were graded using the Ishak system. Results Fibrosis score between baseline and follow-up biopsy decreased in patients with follow-up NIS 0–3 (mean 3.4±0.24 to 2.7±0.21, n=59 p=0.001) but was unchanged in those with follow-up NIS of 4 or 5 (3.5±0.3 to 3.4±0.3, n=34 p=0.846). Fibrosis score on follow-up biopsy was higher in patients with NIS 4–5 (n=40) than with NIS 0–3 (n=72) (mean 3.3±0.3 vs 2.4±0.2, p=0.014). Regression of fibrosis was independently associated with lower NIS (p=0.014) and less portal inflammation (p=0.006) on follow-up biopsy, and with longer interval between the two biopsies (p=0.001). All cause death/transplantation rate was higher in those with NIS of 4–5 than in those with NIS 0–3 (18% vs 6% and 64% vs 24% after 10 and 20 years respectively; p<0.001) and this categorisation was independently associated with survival (p=0.001). A similar trend for liver death/transplantation just failed to reach significance (p=0.07). Conclusion In patients with AIH treated with immunosupression, even mild histological activity (NIS 4 or 5) on follow-up biopsy is associated with more fibrosis and with reduced survival and hence may not be an optimal end point of treatment.

PWE-077 Long-term outcomes in autoimmune hepatitis: an update

Introduction Short-term outcome of treated AIH is good but liver disease may progress despite treatment. We previously reported on AIH patients presenting to our (non-transplant) centre between 1971–2007 (1). Case-capture was complete only after 1987, possibly resulting in underestimated mortality. We aim to assess outcome and temporal trends in patients presenting between 1/1/1987 and 31/12/2016. Method We assessed overall survival (life table analysis) and standardised mortality ratio (SMR; compared to regional population data, considering liver transplant as death). Patients presenting between 1987–2006 were compared with those since 2007. Results From 345 patients (275 female) there were 106 deaths (28 liver-related deaths) and 8 transplanted. Overall 10- and 20 year death/transplant rates were respectively (mean±SEM) 24%±3% and 53%±5% (all cause) and 10%±2% and 21%±4% (liver-related). SMR was (mean(95% CI)) 1.77 (1.4–2.1) overall and was 5.16 (1.7–8.6), 1.84 (1.34–2.5) and 1.51 (1.11–1.92) respectively in those presenting aged <45, 45–65 and >65 years. SMR was close to unity for non-liver deaths. Abstract PWE-077 Table 1 Year of presentation 1987–2006 2007–2016 1987–2016 Number 202 143 345 Follow-up time (yrs)^ 13 (0.0–28.8) 4.5 (0.0–10) 8 (0.0–29) Age at presentation^ 57 (2.5–87) 60 (14–85) 58.0 (2.5–87) Cirrhosis at presentation (%)* 64/198 (32) 27/140 (19) 91/338 (27) Decompensation at presentation (%) 33/198 (17) 17/138 (12) 50/236 (15) SMR over 1st decade (all cause death/transplant)* 2.28 (1.6–2.9) 1.27 (0.57–1.98) 1.77 (1.4–21) Total deaths or transplants 93 13 106 Number transplants/liver deaths 6/24 2/4 8/28 5 year death/transplant rate:All-cause (%(SEM))Liver (%(SEM)) 15(2)8 (2) 11(3)4 (2) 13 (2)7 (1) ^ median(range), *p<0.01 for 87–06 vs 07–16 All-cause death/transplant rate in those presenting since 2007 was lower compared to 1987–2006 (p=0.155 by Log-Rank test but 0.041 by Cox regression analysis, along with age and baseline cirrhosis and decompensation). SMR (over the first decade of follow up) was lower in those presenting since 2007 (Table 1). Corresponding liver death/transplant rates did not differ. Conclusion Overall survival and SMR are similar to those we previously reported. AIH patients still have excess long-term mortality due to their liver disease, including those presenting at age over 65 (new observation). Patients presenting since 2007 compared to those presenting earlier, are older, less likely to have cirrhosis and have a trend towards lower all-cause short-term mortality. Reference . Hoeroldt B, McFarlane E, Dube A, Basumani P, Karajeh M, Campbell J, Gleeson D. Long-term outcomes of patients with autoimmune hepatitis managed at a nontransplant centre. Gastroenterology2011; 140:1980 Disclosure of Interest None Declared

PTH-101 Accuracy of Transient Elastography in Predicting Histological Fibrosis Severity in Treated Autoimmune Hepatitis?

Introduction The role of transient elastography (TE) or Fibroscan in assessing severity of liver fibrosis has been validated against liver biopsy in chronic hepatitis B and C and in NAFLD. There are few studies assessing its role in autoimmune hepatitis (AIH). Four preliminary studies have reported data on the role of Fibroscan in AIH. In these, most patients had recently started treatment and thus, still had active disease. Minimal data on transaminases was provided. We aimed to assess the accuracy of Fibroscan in predicting histological fibrosis severity in patients who had achieved biochemical remission (normal transaminases) and were undergoing follow-up liver biopsy to confirm histological remission as per our Unit’s clinical policy. Methods Between 1/12/13 and 31/12/15 36 same-day Fibroscan and liver biopsy were performed in 32 patients with AIH (1999 International Group criteria; 25 female, age 56 (17–78) years who had achieved biochemical remission (normal serum ALT and globulins) after 2.7 (2.1–24.9) years treatments. No patient had ascites, extrahepatic cholestasis or congestive cardiac failure based on clinical and laboratory evaluation. Fibroscan was performed in fasted patients using the Echosens machine (M or XL probe used as needed) by trained operators. We assessed how accurately the liver stiffness evaluation (LSE) score on Fibroscan could predict Ishak fibrosis stage on biopsy (assessed independently by AKD). Results The ALT was 21 (9–99), normal value <33. 89% patients had normal serum ALT on the day. Ishak necroinflammatry score ≤3 (histological remission) was present in 12 biopsies (33%). Of the 36 Fibroscan’s carried out, 27 were valid. A valid scan defined by: ≥10 liver stiffness measurements, interquartile range (IQR)/median of <0.30 and a success rate ≥ 60%.Abstract PTH-101 Table 1 Accuracy of fibroscan in prediction of liver fibrosis Patients IshakFibrosisStage (n) AUROC Fibroscan cut-off score Sens.(%) Spec.(%) PPV NPV All (n = 36) 5 or 6 (n = 5) 0.78 11.0 0.60 0.80 0.33 0.92 4-6 (n = 11) 0.78 11.0 0.55 0.88 0.67 0.81 3-6 (n = 21) 0.62 7.0 0.52 0.66 0.69 0.50 Validscans(n = 27) 5 or 6 (n = 3) 0.97 11.0 1.00 0.83 0.43 1.00 4-6 (n = 6) 0.90 11.0 0.83 0.90 0.71 0.95 3-6 (n = 15) 0.72 7.0 0.46 0.83 0.78 0.55 Conclusion In this study, Fibroscan showed good accuracy in excluding, but lower accuracy in predicting Ishak fibrosis stage of 4 or more. Accuracy was improved if a valid scan was obtained. Fibroscan was less accurate in predicting lower fibrosis stages. A combination of methods of assessing liver fibrosis may be necessary in some patients. Disclosure of Interest None Declared