Ottar M. Bergmann

Incidence, Presentation, and Outcomes in Patients With Drug-Induced Liver Injury in the General Population of Iceland

BACKGROUND & AIMS Little is known about the incidence of drug-induced liver injury (DILI) in the general population. We investigated the incidence and the quantitative risk of DILI in a population-based cohort. METHODS We performed a prospective study and collected data from 96 individuals diagnosed with DILI in Iceland from 2010 through 2011 (54 women; median age, 55 y). Liver injury was defined based on levels of alanine aminotransferase that were more than 3-fold the upper limit of normal and/or alkaline phosphatase levels more than 2-fold the upper limit of normal. Patients with acetaminophen toxicity were excluded. Drug history and clinical outcome were analyzed. Causality was assessed using the Roussel Uclaf Causality Assessment Method. The patients were registered in prescription databases for outpatients and inpatients. RESULTS The crude annual incidence rate of DILI was 19.1 (95% confidence interval [CI], 15.4-23.3) cases per 100,000 inhabitants. DILI was caused by a single prescription medication in 75% of cases, by dietary supplements in 16% of cases, and by multiple agents in 9% of cases. The most commonly implicated drugs were amoxicillin-clavulanate (21 of 96; 22%), diclofenac (6%), azathioprine (4%), infliximab (4%), and nitrofurantoin (4%). The median duration of therapy was 20 days (range, 8-77 days); 26 patients had jaundice (27%) and 22 patients were hospitalized (23%) for a median of 5 days (range, 2-8 days). Overall 35,252 patients received amoxicillin-clavulanate as outpatients, and DILI occurred in 1 of 2350 (43 of 100,000; 95% CI, 24-70). DILI also occurred in 1 of 9480 patients taking diclofenac (11 of 100,000; 95% CI, 4-24), 1 of 133 patients taking azathioprine (752 of 100,000; 95% CI, 205-1914), 1 of 148 patients taking infliximab (675 of 100,000; 95% CI, 184-718), and 1 of 1369 patients taking nitrofurantoin (73 of 100,000; 95% CI, 20-187). CONCLUSIONS In a population-based study in Iceland, the incidence of DILI was the highest reported to date. Amoxicillin-clavulanate was the most commonly implicated agent. The highest risk of hepatotoxicity was associated with azathioprine and infliximab, but the actual number of cases attributed to these agents was small.

The epidemiology and natural history of primary biliary cirrhosis: a nationwide population-based study.

Background and aims Very few population-based studies exist on the epidemiology of primary biliary cirrhosis (PBC), and none have been conducted in the last decade. We aimed to determine the epidemiology and prognosis of PBC over the past two decades. Methods Patients were identified by multiple case finding strategies, covering the total population of Iceland. A search was conducted in the centralized database of antimitochondrial antibody (AMA) measurements and computerized diagnosis and pathological registries. All AMA measurements taken in Iceland between 1991 and 2010 were analyzed. Relevant clinical information was gathered from medical records, pathology reports, and death certificates. Incidence was compared between two periods, 1991–2000 versus 2001–2010. Results A total of 168 patients were identified, of which 138 were female (82%), with a median age 62 years (range 13–92). Prevalence at the end of the study period was 38.3 cases per 100 000 person-years. Age-standardized incidence for female patients during the first period was 3.4 versus 4.1 during the second (NS) and that for male patients was 0.6 during the first period versus 1.0 per 100 000 during the second (NS). Overall incidence in the first period was 2.0 and that in the second was 2.5 per 100 000 (NS). Stage III–IV liver fibrosis was present in 28% of patients at diagnosis with no significant differences between the two decades. Median survival after diagnosis was 15 years. Five patients underwent liver transplantation. Conclusion The incidence and prevalence figures of PBC in Iceland are among the highest reported and have been stable over the last two decades. The prognosis of patients in this population-based cohort is better than that previously reported.

Paracetamol intoxications: a retrospective population-based study in Iceland

Abstract Objective. Paracetamol is the most common cause of acute liver failure (ALF) in many countries. Much data on paracetamol toxicity originate from liver transplant centers and tertiary referral institutions. The authors analyzed the population-based annual incidence of paracetamol overdoses and ALF, and described the risk factors for hepatotoxicity. Methods. A search was undertaken for the diagnosis of paracetamol overdoses in the diagnoses registry of the National University Hospital of Iceland from 2004 to 2009 serving a population of 219,249 inhabitants. Relevant information was collected from medical records. Results. A total of 1913 drug-related poisoning episodes were identified and reviewed, 352 (18%) involved paracetamol overdoses. The annual incidence of paracetamol overdoses declined from 30.0 (2004) to 16.0/100,000 per year (2009) (p < 0.05). The female/male ratio was 3.0 and the largest age group was 16–25 years. After the initial examination, 26% were discharged home. Hospitalized index visits were 182 with accidental overdoses constituting 16 (9%) with no gender difference. Compared with intentional overdose the accidental group had higher aminotransferases (p < 0.005). ALF occurred in 3.8% (7/182) of the index visits and the incidence was 0.7/100,000 per year. In the intentional group, 1.2% (2/163) developed ALF versus 25% (4/16) of the accidental group (p = 0.001). Only one patient died from ALF and none underwent liver transplantation. Conclusion. The annual incidence of paracetamol overdoses was high in this population-based study but declined. Young females with intentional overdose accounted for most of the cases, whereas accidental overdoses were more common in older patients. The occurrence of ALF was low and mostly associated with accidental overdose.

Secondary sclerosing cholangitis in patients with drug-induced liver injury

BACKGROUND Secondary sclerosing cholangitis has clinical features similar to primary sclerosing cholangitis but originates from a known pathological entity. Secondary sclerosing cholangitis has not been investigated in patients with drug-induced liver injury. METHODS Overall 102 patients diagnosed with drug-induced liver injury were identified and magnetic resonance cholangiopancreatography images of 25 patients were reviewed. RESULTS Ten patients (all females) out of 102 had confirmed features of secondary sclerosing cholangitis on biliary imaging. Overall 70% of patients with sclerosing cholangitis had jaundice vs. 25% without sclerosing cholangitis (p<0.01). All sclerosing cholangitis patients had cholestatic/mixed type of liver injury and compared with patients with cholestatic/mixed liver injury without confirmed abnormal MRCP (n=52), they also had more frequently jaundice, 70% vs. 23% (p=0.0065), higher peak alkaline phosphatase 551 (352-716) vs. 329 (202-543) (p=0.055) and longer time to resolution of liver injury 152 days (123-353) vs. 62 days (36-91) than patients without confirmed sclerosing cholangitis (p<0.0009). CONCLUSIONS Our results indicate that drugs can lead to bile duct injury visualized on imaging. This should be a part of the differential diagnoses of secondary sclerosing cholangitis. These patients were more likely to present with jaundice and longer recovery of liver injury than other patients with cholestatic/mixed type of drug-induced liver injury.

Jaundice Due to Suspected Statin Hepatotoxicity: A Case Series

Statin drugs are widely used worldwide and are generally considered safe and well tolerated. Only small proportion of patients receiving statins develop elevations of liver enzymes and an even smaller proportion will have clinically significant hepatitis induced by statins. We describe four patients with jaundice caused by drug-induced liver injury, where the most likely agent was a statin drug, over a period of approximately three year in Iceland. We calculate the risk of jaundice caused by statin drugs, from sale in the whole country of Iceland, to be one in 17,434 users a year. This is a higher risk than has previously been estimated and we challenge the current opinion that statins rarely cause clinically significant drug-induced liver injury and encourage alertness when managing patients with statins with regard to clinical signs of hepatitis before jaundice occurs.

Drug-Induced Autoimmune Hepatitis: Response to Corticosteroids and Lack of Relapse After Cessation of Steroids

Most patients had been prior participants in previous studies. Inclusion criteria were a drug reaction with well-documented cause of DIAIH, with positive antinuclear antibodies and/or smooth muscle antibodies and/ or elevated IgG; 10 times above the upper limit of normal elevation in alanine aminotransferase with a hepatocellular or mixed pattern; or requirement of corticosteroids in patients with hepatocellular type of injury, the latter defined as no improvement in alanine aminotransferase and aspartate aminotransferase after discontinuation of the implicated agent. Roussel Uclaf Causality Assessment Method was used for causality assessment of suspected DILI. The new simplified criteria for autoimmune hepatitis (AIH) were calculated.

A prospective study on the causes of notably raised alanine aminotransferase (ALT)

Abstract Objective High levels of alanine aminotransferase (ALT) can be a marker of severe liver disease with variable aetiologies and prognosis. Very few prospective studies have been undertaken on the aetiology and prognosis of patients with high ALT levels. No population-based prospective study has systematically evaluated drug-induced liver injury (DILI) among these patients. The objective was to determine the aetiology and prognosis of patients with high ALT. Materials and methods In a catchment area of 160,000 inhabitants, a population-based prospective study identified all adult patients with serum level of ALT >500 U/L during a 12-month period. All underwent thorough diagnostic work-up and follow-up. In suspected DILI, causality was assessed with Roussel Uclaf Causality Assessment Method. Results A total of 155 patients were identified with ALT >500 U/L, 12 children and one with ALT of non-liver-related origin, leaving 142 patients for the analysis: 73 (51%) males, median age 52 (IQR 36–68, range 19–89 years). The most common causes were choledocholithiasis 48/142 (34%), ischaemic hepatitis 26 (18%), viral hepatitis 16 (11%) and DILI 15 (11%), hepatobiliary malignancy (n = 6), surgery/interventions (n = 8) and other aetiologies (n = 23). No specific aetiology was found in 6% of cases. In the total study cohort 99 (70%) required hospitalisation, 78 (55%) had jaundice and 22 (16%) died, liver-related death in 10%, 35% in IH and 7% in DILI. Conclusions The most common cause of notably high ALT was choledocholithiasis. Ischaemic hepatitis was a common aetiology with approximately 35% liver-related mortality. Viral hepatitis and DILI were important aetiologies among these patients.

Using the Icelandic genealogical database to define the familial risk of primary biliary cholangitis

Hereditary factors in primary biliary cholangitis (PBC) have been well defined in genome‐wide association studies, but there are few direct data available that define the relative risk (RR) for family members with an affected proband. An increased risk in first‐degree relatives has been demonstrated in a variety of studies, but data have been lacking on further detailed associations for subsequent generations. The objective of this study was to use the unique Icelandic genealogical database to study the familiality of PBC. All patients with positive antimitochondrial antibody measurements in Iceland during the period 1991‐2015 who fulfilled diagnostic criteria for PBC were included. The Icelandic genealogical database was used to assess familial relations. For each case of PBC, 10,000 control subjects matched for age, sex, and number of known relatives were randomly chosen from this database to calculate the familial RR of PBC. The average kinship coefficient (KC) of the patients was calculated and compared with the average KC of controls. Overall, 222 PBC patients were identified (182 females, 40 males; median age, 62 years). First‐, second‐ and third‐degree relatives of the PBC patients had a high RR of the disease: 9.13 (P < 0.0001), 3.61 (P = 0.014) and 2.59 (P = 0.008), respectively. In fourth‐ and fifth‐degree relatives, the RR was also increased to 1.66 (P = 0.08) and 1.42 (P = 0.08), respectively. The average KC of the patients was also higher than that of the control subjects, with 21.34 × 10−5 versus 9.56 × 10−5 (P < 0.0001). Conclusion: Relatives of PBC patients had markedly higher risk for development of the disease compared with controls and importantly our data demonstrate that the risk was significantly increased even in second‐ and third‐degree relatives. (Hepatology 2018;68:166‐171).

Lifetime drinking history in patients with alcoholic liver disease and patients with alcohol use disorder without liver disease

Abstract Objective: To determine the differences in lifetime alcohol intake (LAI) and drinking patterns between patients with alcoholic liver disease (ALD) and alcohol use disorder (AUD) without notable liver injury and between males and females with ALD. Methods: Alcohol drinking patterns were assessed using the Lifetime Drinking History (LDH) a validated questionnaire, during an outpatient visit. Patients with AUD, currently in addiction treatment, were matched for gender and age (±5 years) with the ALD group. Results: A total of 39 patients with ALD (26 males and 13 females; median age 58) and equal number of AUD patients were included (median age 56 years). The onset age for alcohol drinking and duration of alcohol consumption was similar in ALD and AUD. The number of drinking days was higher in women with ALD than in women with AUD: 4075 [(3224–6504) versus 2092 (1296–3661), p = .0253]. The LAI and drinks per drinking day (DDD) were not significantly different between patients with ALD and AUD. Females with ALD had lower LAI than males with ALD: 32,934 (3224–6504) versus 50,923 (30,360–82,195), p = .0385, fewer DDD (p = .0112), and lower proportion of binge drinking as compared to males with ALD (p = .0274). Conclusions: The total LAI was similar in patients with ALD and AUD. The number of drinking days over the lifetime was associated with the development of ALD in females. Females with ALD had significantly lower alcohol consumption than men with ALD despite similar duration in years of alcohol intake which supports the concept of female propensity of ALD.

Meðferð lifrarbólgu C með peg-interferóni og ríbavíríni á Íslandi 2002-2012

BACKGROUND/AIMS Hepatitis C is a major cause of chronic liver disease and cirrhosis in Western countries. Its treatment aims at eradicating the virus and patients are considered cured if the virus is undetectable by PCR in blood 12-24 weeks after end of treatment (sustained virological response, SVR). The aim of this study is to investigate the results of treating hepatitis C in Iceland during the period 2002-2012. MATERIALS AND METHODS Retrospective study including all patients with hepatitis C receiving treatment with peginterferone and ribavirin at Landspitali University hospital during the period 2002-2012. Patients who had been treated previously were excluded. Information was obtained from medical records and the hospital pharmacy. RESULTS A total of 207 patients were included, 136 (66%) males and 71 (34%) females. Mean age was 38 years (range 17-66). Genotyping revealed that 71 (34%) patients had genotype 1, 135 (65%) genotype 3 and one genotype 2. A total of 147 (71%) patients achieved SVR. The rate of SVR was 77.8% for genotype 3 and 57.7% for genotype 1. 9 patients (4%) had cirrhosis and 3 of them had SVR. Of 161 patients who finished treatment per protocol, 87.5% and 77.1% with genotypes 3 and 1 respectively had SVR. CONCLUSIONS The study demonstrates higher rates of SVR in clinical practice in Iceland compared to controlled clinical trials. The improved effectiveness may be explained by younger patient population, low rate of cirrhosis and close follow-up of patients. Key words: Hepatitis C, peg-interferon, sustained virological response. Correspondence: Sigurdur Olafsson, sigurdol@landspitali.is.