- Email: [email protected]
Etiologies and Outcomes of Acute Liver Failure in Germany
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2012;10:664 – 669
Etiologies and Outcomes of Acute Liver Failure in Germany JOHANNES HADEM,* FRANK TACKE,‡ TONY BRUNS,§ JULIA LANGGARTNER,储 PAVEL STRNAD,¶ GERALD U. DENK,# PANAGIOTIS FIKATAS,** MICHAEL P. MANNS,* WOLF P. HOFMANN,‡‡ GUIDO GERKEN,§§ FRANK GRÜNHAGE,储 储 ANDREAS UMGELTER,¶¶ CHRISTIAN TRAUTWEIN,‡ and ALI CANBAY,§§ for the Acute Liver Failure Study Group Germany *Department of Gastroenterology, Hepatology & Endocrinology, Hannover Medical School, Hannover; ‡Department of Gastroenterology, Metabolic Diseases & Intensive Care Medicine, University Hospital Aachen, Aachen; §Department of Gastroenterology, Hepatology & Infectious Diseases, University Hospital Jena, Jena; 储 Department of Gastroenterology, Endocrinology, Rheumatology, Infectious Diseases & Emergency Medicine, University Hospital Regensburg, Regensburg; ¶ Department of Internal Medicine I, University Hospital Ulm, Ulm; #Medical Department II, Liver Centre Munich, University Hospital München-Großhadern, Munich; **Department of General Surgery, Visceral Surgery & Transplant Surgery, Virchow-Klinikum, Charite, Berlin; ‡‡Department of Gastroenterology, Hepatology, Pulmonary Medicine, Endocrinology & Nutrition, University Hospital Frankfurt, Johann-Wolfgang-Goethe-University, Frankfurt; §§Department of Gastroenterology & Hepatology, University Hospital Essen, Essen; 储 储Department of Medicine II, Saarland University Hospital, Homburg; and ¶¶Department of Gastroenterology, Hepatology, Endocrinology, Infectious Diseases & Intensive Care Medicine, Klinikum rechts der Isar, Technical University München, Munich, Germany
BACKGROUND & AIMS: Acute liver failure (ALF) is a severe form of acute liver injury that can progress to multiple organ failure. We investigated causes and outcomes of ALF. METHODS: Eleven university medical centers in Germany were asked to report patients with (primary) severe acute liver injury (sALI) (international normalized ratio [INR] ⬎1.5 but no hepatic encephalopathy) and primary ALF (INR ⬎1.5 with overt hepatic encephalopathy) treated from 2008 to 2009. Data were analyzed from 46 patients with sALI and 109 patients with ALF. RESULTS: The most frequent etiologies of primary ALF were non-acetaminophen drug-induced (32%), indeterminate (24%), and viral (21%); acetaminophen ingestion was the cause of ALF in only 9% of patients. The support of a ventilator was required by 44% of patients with ALF, vasopressors by 38%, and renal replacement by 36%. Seventy-nine patients with ALF (72%) survived until hospital discharge, 38 (35%) survived without emergency liver transplantation (ELT), and 51 received ELT (47%); 80% of patients who received ELT survived until discharge from the hospital. CONCLUSIONS: In Germany, drug toxicity, indeterminate etiology, and viral hepatitis appear to be the major causes of primary ALF, which has high mortality. Patients with ALF are at great risk of progressing to multiple organ failure, but 80% of patients who receive ELT survive until discharge from the hospital. Keywords: Hepatic Failure; Acute Hepatitis; King’s College Criteria; Liver Transplantation.
A
ccording to the first description by Trey and Davidson,1 acute liver failure (ALF) is an acute deterioration of liver function resulting in hepatic encephalopathy (HE) within 8 weeks of the onset of symptoms in a patient with a previously healthy liver. More recently, the American Association for the Study of Liver Diseases suggested a modified definition of ALF comprising a decline in liver function of less than 26-week duration, absence of cirrhosis, evidence of coagulation abnormality (usually indicated by an international normalized ratio [INR] ⱖ1.5), and any degree of HE.2 In clinical practice, however, considerable uncertainty exists when a patient should be considered as having ALF. This is because (1) there is an ongoing debate on whether HE should remain an obligatory element of ALF definition, and (2) standardization of neurologic scales for HE can still be regarded as preliminary.3,4
An important aspect of ALF that needs further clarification is the varying composition of ALF causes among different geographic regions.5–9 Two retrospective studies from Germany summarized 236 ALF cases from a metropolitan western area and a rural area in northern Germany. Drug toxicity, viral hepatitis, and indeterminate ALF were the most prevalent causes of ALF.10,11 Although these studies provided new data on management and outcome of ALF in Germany, the gained insight is limited because of a single center experience. In this situation, detailed knowledge on the current predominant etiologies and outcomes of ALF might help to prevent ALF and allow more selective treatment options. The Acute Liver Failure Study Group Germany was founded in June 2009 to improve the management of ALF patients in Germany by developing guidelines and establishing a nationwide ALF registry as a basis for multicenter research projects. This retrospective survey was regarded as a first approach to the epidemiology of ALF in Germany.
Methods Enrollment of Patients and Definitions of Severe Acute Liver Injury and Acute Liver Failure In 2010, all university clinics participating in the Acute Liver Failure Study Group Germany (including all major liver transplant centers) were asked to retrospectively identify patients with imminent and overt primary ALF treated at their clinics during a 2-year period (2008 –2009). Early recognition of ALF is important and has been declared as a main task of our study group. We therefore intended to Abbreviations used in this paper: ALF, acute liver failure; AUC, area under the curve; BiLE score, bilirubin-lactate-etiology score; CI, confidence interval; CMV, cytomegalovirus; DILI, drug-induced liver disease; EBV, Epstein–Barr virus; ELT, emergency liver transplantation; HAV, hepatitis A virus; HBV, hepatitis B virus; HE, hepatic encephalopathy; INR, international normalized ratio; IQR, interquartile range; KCC, King’s College criteria; MELD score, Model of End-Stage Liver Disease score; ROC, receiver operating characteristic; sALI, severe acute liver injury. © 2012 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2012.02.016
June 2012
include a wider range of disease severity than just those patients incurring encephalopathy, which is in accordance with the European Acute Liver Failure Registry. Imminent ALF or severe acute liver injury (sALI) was defined as acute (⬍12 weeks) liver dysfunction leading to an INR ⬎1.5, but without documented level of HE. Overt ALF was defined as acute (⬍12 weeks) liver dysfunction leading to an INR ⬎1.5 plus any documented level of HE. Two hundred seventy-two patients were reported by the centers and assumed to have ALF. Careful consideration of the clinical setting and time course of deteriorating events disclosed that 80 of these patients had preexisting or secondary liver failure (31.3% on top of a chronic liver disease, 35.0% in the setting of sepsis with multiple organ failure, and 32.5% caused by ischemia/shock), necessitating their exclusion. Another 37 patients were excluded because of only mild liver dysfunction (INR ⬍1.5). The remaining 46 patients with primary sALI and 109 patients with primary ALF were included in the analysis (Supplementary Figure 1).
Patient Collective and Parameter Evaluation ALF etiologies were based on accepted diagnostic criteria.2,5,6,11–15 All parameters except the maximum HE grade were documented at intensive care unit admission. HE was graded on a standard scale of 1– 4 as described previously.12 King’s College criteria (KCC) were determined as described.16 The calculation of labMELD score (below called Model for EndStage Liver Disease [MELD] score) was performed according to the equation: MELD score ⫽ [9.57 ⫻ loge creatinine (mg/dL) ⫹ 3.78 ⫻ loge bilirubin (mg/dL) ⫹ 11.20 ⫻ loge INR ⫹ 6.43]. If the individual values were ⬍1.0, they were set to 1.0; serum creatinine was adjusted to 4.0 mg/dL if the patient underwent dialysis during 2 weeks before assessment or the serum creatinine was ⬎4.0 mg/dL. MELD scores ⬎40 were set to 40.13,17 Clichy criteria could not be calculated because factor V was not available in most of the cases. Calculation of bilirubin-lactateetiology (BiLE) score was as follows: BiLE score ⫽ (baseline bilirubin [mol/L]/100) ⫹ baseline lactate [mmol/L] ⫹ 4 [in case of indeterminate ALF, Budd–Chiari syndrome, or phenprocoumon toxicity] – 2 [in case of acetaminophen toxicity] ⫹ 0 [in case of any other ALF etiology].11 Cutoff values of MELD and BiLE scores were adapted to available data.11,17,18 These parameters were then compared between the cohort of transplant-free survivors and patients who received emergency liver transplantation (ELT) or died. The study was performed in accordance with the ethical guidelines of the 1975 Declaration of Helsinki and was a priori approved by the institutional review board (#4373). Because the study relied on measurements taken as part of routine care, the Ethical Committee of Hannover Medical School waived the need for informed consent.
Statistical Analysis Data are presented as percentages or medians with corresponding 25th and 75th percentiles (interquartile range [IQR]). Comparison of baseline parameters between patient groups was performed with Student t test and U test for continuous and with 2 test for categorial parameters. Kaplan–Meier plots were used to illustrate the timing of events during follow-up in relation to baseline KCC, MELD score, or BiLE score, and statistical assessment was performed by the log-rank test. Receiver operating characteristic (ROC)
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
665
curve analysis was used for comparison of prognostic accuracy of different outcome prediction models. Tests were two-sided, and significance was accepted at P ⬍ .05. Data analysis was performed by using SPSS software (SPSS, Inc, Chicago, IL).
Results Distribution of Etiology and Outcome in Severe Acute Liver Injury Forty-six patients with sALI (imminent ALF) were reported by the participating centers. Seventeen patients (37.0%) developed sALI caused by non-acetaminophen drug toxicity, 8 (17.4%) had acute viral hepatitis, 7 (15.2%) had ingested acetaminophen, 6 (13.0%) were reported to have sALI of indeterminate cause, and 8 (17.4%) had other causes. Two of 4 sALI patients with phenprocoumon toxicity (50%) and 3 of 6 sALI patients with indeterminate cause (50%) died or underwent ELT in the absence of documented HE. Patients with other sALI etiologies exhibited high survival rates.
Distribution of Etiology and Outcome in Acute Liver Failure The etiologic distribution of the 109 patients with ALF differed significantly from those without HE (ie, sALI). ALF causes and outcomes are shown in Figure 1. Thirty-five patients (32.1%) developed ALF because of non-acetaminophen drug toxicity (8 of them because of phenprocoumon), 26 (23.9%) were reported to have ALF of indeterminate cause, 23 (21.1%) had acute viral hepatitis, and 10 (9.2%) had ingested acetaminophen. Acute viral hepatitis was due to hepatitis B virus (HBV) in 11 patients, hepatitis A virus (HAV) in 4, hepatitis E virus in 4, adenovirus in 2, Epstein–Barr virus (EBV) in 1, and cytomegalovirus (CMV) in another patient. In several patients hepatotoxicity developed because of multiple causative agents (eg, ingestion of numerous drugs). The chance of ELT-free survival was highest in Amanita ingestion (100%), acetaminophen-induced ALF (70%), and ALF related to HAV infection (75%). Conversely, risk of ELT/death appeared to be particularly high in ALF caused by malignancy, Wilson’s disease, and Budd–Chiari syndrome (100% each), indeterminate ALF (81%), viral hepatitis (65%), and non-acetaminophen drug toxicity (54% overall, but 63% in phenprocoumon toxicity).
Baseline Clinical and Laboratory Parameters Table 1 provides an overview of relevant baseline parameters. Patients with sALI were slightly younger (median age, 34 years; IQR, 17– 81) in comparison with ALF patients (46 years; IQR, 15–75). However, these differences did not reach statistical significance (P ⫽ .092, U test). No difference could be detected with respect to patients’ sex (63% female in ALF vs 59% in ALI, not significant, 2 test). Median baseline parameters of the 109 patients with ALF were INR 2.5 (1.8 – 4.1), lactate 3.0 (1.8 – 6.0) mmol/L, bilirubin 255 (94 –378) mol/L, and HE grade 1 (1–2). HE peaked at grade 1/2 in 40 of 89 (45%) and at grade 3/4 in 49 of 89 (55%).
Frequency of Organ Failure HE was, per definition, absent in all patients with sALI but present in all patients with ALF. Whereas extrahepatic
666
HADEM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 6
Figure 1. Causative etiologies and outcomes in ALF. Etiology and outcome of 109 patients with primary ALF. Patients who developed ALF due to ⬎1 causative agent appear in ⬎1 bar of the diagram. ACM, acetaminophen.
organ dysfunction appeared to be a rare event in patients with sALI, the frequency dramatically increased in patients with ALF. Ventilatory support was required in 48 ALF patients (44%), vasopressors in 41 (38%), and renal replacement therapy in 39 (36%) during intensive care unit stay (Supplementary Figure 2).
Outcomes in Severe Acute Liver Injury and Acute Liver Failure Of 46 patients with sALI, 37 (80.4%) survived spontaneously without ELT, 5 (10.9%) survived after ELT, 2 (4.3%)
underwent ELT but died postoperatively within 14 days after ELT, and 2 (4.3%) died without ELT. Overall survival to hospital discharge among sALI patients was 91.3%. In contrast, of the 109 patients with ALF, only 38 (34.9%) survived spontaneously without ELT, 41 (37.6%) survived after ELT, 10 (9.2%) underwent ELT but died after surgery, and 20 (18.3%) died without ELT (Figure 2). Patients on the waiting list for ELT received a graft within 5 (2– 6) days from admission. Overall, 79 ALF patients (72.3%) could be discharged from the hospital alive.
Table 1. Baseline Parameters in Patients With ALF (n ⫽ 109) Parameter
Overall (n ⫽ 109)
Age (y) Women, n (%) Encephalopathy grade at baseline Maximum encephalopathy grade (available in n ⫽ 89, 81.7%) 1/2 (n, %) 3/4 (n, %) INR at baseline Maximum INR WBC (per L) Creatinine (mol/L) Bilirubin (mol/L) Lactate (mmol/L) MELD score MELD score ⱖ32, n (%) KCC fulfilled (available in n ⫽ 93), n (%) BiLE score (available in n ⫽ 71), n (%) BiLE score ⱖ6.9, n (%)
46 (33–56) 69 (63.3) 1 (1–2) 3 (1–4) 40 (44.9) 49 (55.1) 2.5 (1.9–4.1) 3.6 (2.3–5.1) 10.7 (7.2–18.3) 92 (68–190) 255 (94–378) 3.0 (1.8–6.0) 29 (24–36) 44 (40.4) 40 (36.7) 8.1 (4.9–11.7) 42 (38.5)
NOTE. Data presented as median (IQR) unless otherwise stated. WBC, white blood cell count.
ELT-free survival (n ⫽ 38)
ELT and/or death (n ⫽ 71)
P value
42 (27–54) 22 (57.9) 1 (1–1) 1 (1–2)
47 (36–56) 47 (66.2) 1 (1–2) 3 (2–4)
.12 — .20 ⬍.0001
25 (80.6) 6 (19.4) 2.1 (1.7–3.2) 2.3 (1.8–4.0) 8.6 (6.3–14.1) 88 (66–173) 104 (53–214) 2.2 (1.3–4.0) 26 (20–31) 8 (21.1) 9 (28.1) 4.8 (1.6–8.1) 8 (33.3)
15 (25.9) 43 (74.1) 2.7 (1.9–4.2) 4.1 (3.1–5.3) 12.2 (7.7–19.2) 96 (69–196) 337 (189–474) 3.8 (2.0–7.2) 32 (26–37) 36 (50.7) 31 (50.8) 9.8 (6.5–13.1) 34 (72.3)
— — .13 .004 .54 .78 ⬍.0001 .02 .0007 — — .003 —
June 2012
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
Figure 2. Outcome analysis of ALF patients.
Patients progressing to ELT and/or death had a higher probability of progression to HE stage 3/4 and displayed higher baseline levels of bilirubin and lactate as well as higher maximum INR levels. An overview of those 20 patients who died without ELT is given in Supplementary Table 1.
Prediction of Emergency Liver Transplantation and/or Death in Acute Liver Failure KCC were available in 93 ALF patients. KCC were more likely to be fulfilled in patients progressing to ELT and/or death, as depicted by Kaplan–Meier analysis (log-rank test log rank 5.30, P ⫽ .0213; Figure 3A). KCC by itself achieved a sensitivity and specificity of 51% and 72%, respectively, for accurately predicting outcome. MELD score was available in all 109 patients and was significantly elevated in the patient group progressing to ELT and/or death (Table 1). With a cutoff value of 32, its sensitivity and specificity were 44% and 82%, respectively, and this cutoff also provided significant discrimination in Kaplan–Meier survival curve analysis (log-rank 8.84, P ⫽ .0030; Figure 3B). BiLE score was available in 71 ALF patients and differed significantly between the 2 prognostic groups, showing an improved discriminative power by Kaplan–Meier analysis (log rank 9.26, P ⫽ .0023; Figure 3C). With a cutoff value of 6.9, its sensitivity and specificity were 72% and 67%, respectively. In direct comparison by ROC curve analysis, BiLE score (area under the curve [AUC], 0.777; 95% confidence interval [CI], 0.652– 0.901) had better prognostic accuracy to predict outcome (transplant-free survival vs ELT or death) than the MELD score (AUC, 0.707; 95% CI, 0.570 – 0.844) or KCC (AUC, 0.583; 95% CI, 0.436 – 0.731; Figure 3D).
Discussion The main results of this multicenter retrospective analysis are (1) outcome of ALF patients is strongly associated with the underlying etiology, (2) development of HE is significantly associated with reduced chance of ELT-free survival (87% vs 35%), (3) ELT is performed in nearly half of ALF patients with an overall survival rate of 72%, and (4) KCC are widely used to
667
predict outcome of ALF but lack sensitivity in predicting a fatal outcome that can possibly be improved by BiLE score. The most common etiologies of sALI and ALF were drugs, indeterminate cause, and acute viral hepatitis. This is in accordance with the above-mentioned retrospective German singlecenter studies that analyzed patients with pediatric ALF before 2008.10,11 Interestingly, acetaminophen seems to be responsible for only about 25% of drug-induced hepatotoxicity in sALI as well as in ALF patients, which is in contrast to recent data from the United States and United Kingdom.6,19 This might be due to underestimation of acetaminophen as a cause of ALF because of (1) possible inaccurate retrospective data acquisition, (2) exclusion of acetaminophen ALF cases that were seen in non-university hospitals and therefore did not come to our attention, and (3) lack of proper history-taking concerning acetaminophen ingestion. It has been shown that half of acetaminophen overdoses are apparently unintentional,14,20 and that acetaminophen toxicity is in fact causing many cases of ALF not attributed to the drug without specific testing.21 Because low prevalences of acetaminophen-induced ALF have also been shown in the preceding German studies, the present data might just reflect a declining acetaminophen intake in our country, which would be in contrast to the United States where the annual percentage of acetaminophen-induced ALF rose from 28% in 1998 to 51% in 2003.22 Another surprising finding is that non-acetaminophen drugrelated sALI/ALF was diagnosed commonly (32.1%) in our series. A lower prevalence of non-acetaminophen drug toxicity had been noted in the 2 above-mentioned German single-center studies (23%,10 12%11). A recent retrospective German singlecenter investigation of 129 patients who underwent ELT for ALF revealed that 8% of them were drug-related.23 The US ALF study group recently published results from 1198 ALF patients enrolled during a 10.5-year period. Only 133 of these prospectively documented patients (11.1%) were deemed by expert opinion to have drug-induced liver injury (DILI).15 The present study might therefore overestimate drug-related causalities that are difficult to evaluate retrospectively. The prevalence of phenprocoumon (7.3%), a Coumadin derivative with long half-life, is likely to be underestimated. Phenprocoumon has been associated with hepatitis and subacute liver failure (2% and 0.2%, respectively, of phenprocoumonassociated adverse events in Germany).11,24 Hepatotoxicity of phenprocoumon is often detected late because monitoring of aminotransferases is not routinely performed during most cardiologic follow-ups. Our study emphasizes that better awareness of this disease entity is urgently needed. The present data are particularly interesting in face of the ongoing debate on whether HE should possibly be eliminated from the ALF definition as is already the case in pediatric ALF.25 The comparison between sALI and ALF in adult patients clearly shows that the occurrence of HE is associated with the development of extrahepatic organ failure. As a marker of imminent multiorgan dysfunction, HE is likely to be of value to discriminate patients with unfavorable prognosis. However, discrepancies in the judgment of HE between different observers are common,4 proper documentation of neurologic abnormalities is often neglected, and performing a retrospective analysis on possibly subtle neurologic alterations is even more difficult. The fact that 9 patients classified as sALI because of lack of docu-
668
HADEM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 6
Figure 3. Prediction of emergency liver transplantation and/or death by different prognostic models. Kaplan–Meier survival curve analysis of 109 patients with primary ALF comparing those who meet with those who do not meet (A) KCC, (B) MELD score, and (C) BiLE score. (D) ROC curve analysis to evaluate prognostic accuracy of KCC, MELD, and BiLE. AUC (confidence intervals) were 0.777 (0.652– 0.901) for BiLE, 0.707 (0.570 – 0.844) for MELD, and 0.583 (0.436 – 0.731) for KCC, respectively.
mented HE finally faced ELT and/or death is probably due to this problem, although KCC can theoretically predict a fatal outcome and push physicians toward ELT without HE under certain circumstances.16 ELT was obviously performed more often (47% of ALF patients) than reported from the United States (29%) or United Kingdom (37%) ALF patients, although these patients were similar with regard to age, maximum HE, and baseline INR levels.6,19 One explanation is again the much lower prevalence of acetaminophen-induced ALF (9% vs 39% and 57%, respectively)
in our cohort. Only the frequent performance of ELT in our ALF patients, of whom 56% had non-acetaminophen druginduced or indeterminate ALF (both with unfavorable prognosis), probably enabled an overall survival that was slightly higher than previously observed.6,19 The analysis of the predictive values of KCC, MELD, and BiLE scores is certainly limited by the retrospective design of this study. In addition, one might argue that defining poor outcome as ELT and/or death might have implicated a bias related to the decision as to whether to perform transplantation
June 2012
in an individual patient. However, patients who underwent ELT accounted for nearly half of the whole ALF cohort, and the high incidence of failure of a second extrahepatic vital organ system within this patient group clearly indicates an unfavorable natural course of disease. KCC are widely used as selection criteria for ELT, and their specificity has been confirmed in several large studies.26 However, its sensitivity was again limited in our patient cohort. This was also true for MELD score, which has been suggested as a prognostic tool in acetaminophen-induced ALF.18 Recently, a modification of classic MELD by cytokeratin-18 significantly improved the positive predictive value in a cohort of mainly acetaminophen-induced ALF.27 Our study indicates that the BiLE score could fill this gap. This score was empirically developed from a retrospective German single-center ALF cohort, where it achieved a promising prediction of poor outcome in 102 ALF patients.11 Although it performed slightly inferior to KCC in an external validation in 422 patients with predominating acetaminophen-induced ALF,19 our nationwide multicenter trial supports the further evaluation of the BiLE score for ALF in future prospective studies. We believe that the results of this survey are representative of the current situation of ALF in Germany. In our experience, ALF patients are transferred to one of the liver transplant centers promptly as soon as the diagnosis of ALF has been established. Only a minority of the 35 university hospitals in Germany has a liver transplant program and is involved in the care of ALF patients. The 11 participating hepatology centers are distributed all over the country and received referrals from different regions of Germany. Among them are those transplant centers with the largest volume of liver transplantations. In conclusion, we present data on the epidemiology and outcome of patients with severe acute hepatitis and ALF in Germany that should be regarded as a first step toward a prospective ALF database within Germany as well as a contribution to the existing European Acute Liver Failure Registry, which will build the basis for further improvements in ALF patient care.
Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org, and at doi:10.1016/ j.cgh.2012.02.016. References 1. Trey C, Davidson CS. The management of fulminant hepatic failure. In: Popper H, Schaffner F (eds). Progress in liver diseases. New York: Grune and Stratton, 1970:282–298. 2. Polson J, Lee WM, American Association for the Study of Liver Disease. AASLD position paper: the management of acute liver failure. Hepatology 2005;41:1179 –1197. 3. Mullen KD. Review of the final report of the 1998 working party on definition, nomenclature and diagnosis of hepatic encephalopathy. Aliment Pharmacol Ther 2007;25:11–16. 4. Córdoba J. New assessment of hepatic encephalopathy. J Hepatol 2011;54:1030 –1040. 5. O’Grady JG, Portmann B, Williams R. Fulminant hepatic failure. In: Schiff L, Schiff R (eds). Diseases of the liver. 7th ed. Philadelphia: Lippincott, 1993:852– 876. 6. Ostapowicz G, Fontana RJ, Schiødt FV, et al. Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med 2002;137:947–954.
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
669
7. Bernuau J, Rueff B, Benhamou JP. Fulminant and subfulminant liver failure: definitions and causes. Semin Liver Dis 1986;6:97–106. 8. Acharya SK, Bhatia V, Sreenivas V, et al. Efficacy of L-ornithine L-aspartate in acute liver failure: a double-blind, randomized, placebo-controlled study. Gastroenterology 2009;136:2159 –2168. 9. Bernal W, Schiødt FV, Hamid S, et al. Etiologies and outcomes for acute liver failure at 6 sites around the world. Dig Dis Sci 2005; 24:A318. 10. Canbay A, Jochum C, Bechmann LP, et al. Acute liver failure in a metropolitan area in Germany: a retrospective study (20022008). Z Gastroenterol 2009;47:807– 813. 11. Hadem J, Stiefel P, Bahr MJ, et al. Prognostic implications of lactate, bilirubin and etiology in German patients with acute liver failure. Clin Gastroenterol Hepatol 2008;6:339 –345. 12. Lee WM. Acute liver failure. N Engl J Med 1993;329:1862–1872. 13. Bundesärztekammer. Richtlinien zur Organtransplantation gem. Deutsches ärzteblatt International 2011;12:A662–A673. 14. Bernal W, Auzinger G, Dhawan A, et al. Acute liver failure. Lancet 2010;376:190 –201. 15. Reuben A, Koch DG, Lee WM, et al. Drug-induced acute liver failure: results of a U.S. multicenter, prospective study. Hepatology 2010;52:2065–2076. 16. O’Grady JG, Alexander GJ, Hayllar KM, et al. Early indicators of prognosis in fulminant hepatic failure. Gastroenterology 1989; 97:439 – 455. 17. Yantorno SE, Kremers WK, Ruf AE, et al. MELD is superior to King’s College and Clichy’s criteria to assess prognosis in fulminant hepatic failure. Liver Transplant 2007;13:822– 828. 18. Schmidt LE, Larsen FS. MELD score as a predictor of liver failure and death in patients with acetaminophen-induced liver injury. Hepatology 2007;45:789 –796. 19. Bernal W, Auzinger G, Wendon J. Prognostic utility of the bilirubin lactate and etiology score. Clin Gastroenterol Hepatol 2009;7:249. 20. Jalan R, Williams R, Bernuau J. Paracetamol: are therapeutic doses entirely safe? Lancet 2006;368:2195–2196. 21. Davern TJ 2nd, James LP, Hinson JA, et al. Measurement of serum-acetaminophen protein adducts in patients with acute liver failure. Gastroenterology 2006;130:687– 694. 22. Larson AM, Polson J, Fontana RJ, et al. Acetaminophen-induced acute liver failure: results of a United States multicenter prospective study. Hepatology 2005;42:1364 –1372. 23. Fikatas P, Ulrich F, Lee JE, et al. The APACHE III score as preoperative indicator of patient outcome in liver transplantation after fulminant hepatic failure. Ann Transplant 2011;16:18 –25. 24. Schimanski CC, Burg J, Möhler M, et al. Phenprocoumon-induced liver disease ranges from mild acute hepatitis to (sub-) acute liver failure. J Hepatol 2004;41:67–74. 25. Squires RH Jr, Shneider BL, Bucuvalas J, et al. Acute liver failure in children: the first 348 patients in the pediatric acute liver failure study group. J Pediatr 2006;148:652– 658. 26. Bailey B, Amre DK, Gaudreault P. Fulminant hepatic failure secondary to acetaminophen poisoning: a systematic review and meta-analysis of prognostic criteria determining the need for liver transplantation. Crit Care Med 2003;31:299 –305. 27. Bechmann LP, Jochum C, Kocabayoglu P, et al. Cytokeratin-18 based modification of the MELD score improves prediction of spontaneous survival after acute liver failure. J Hepatol 2010;53:639–647.
Reprint requests Address requests for reprints to: Johannes Hadem, MD, Department of Gastroenterology, Hepatology, and Endocrinology, Hannover Medical School, Carl-Neuberg-Strasse 1, 30625 Hannover, Germany. e-mail: [email protected]; fax: ⴙ49-511-532-2144. Conflicts of interest The authors disclose no conflicts.
June 2012
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
669.e1
Supplementary Figure 2. Frequency of extrahepatic organ failure in 46 patients with sALI and 109 patients with ALF.
Supplementary Figure 1. Enrollment of study patients.
669.e2
HADEM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 6
Supplementary Table 1. Characteristics of Subjects Who Died After ALF Recognition Patient number
Age (y)/sex
(Suspected) ALF etiology
Significant comorbidities
1 2 3 4 5 6 7 8 9
55/F 63/M 54/M 56/F 51/F 53/F 47/M 25/M 59/M
Phenprocoumon, clindamycin Acute hepatitis B Reactivation of latent hepatitis B Indeterminate Indeterminate Acute hepatitis B Non-ACM drug-induced Non-ACM drug-induced Ischemic hepatitis
— —
10 11
66/F 55/M
12 13
45/F 56/F
Phenprocoumon Viral hepatitis (CMV and adenovirus) EBV hepatitis Phenprocoumon
14 15 16 17 18 19 20 21
70/F 75/M 65/F 38/M 53/M 39/F 58/F 62/F
22 23 24 25 26
43/F 54/F 57/M 66/M 56/F
27
62/M
28 29
56/F 49/F
30
69/M
Indeterminate Indeterminate NASH Non-ACM drug-induced Vanishing bile duct syndrome Indeterminate Non-ACM drug-induced (Valproat) Non-ACM drug-induced (ibuprofen) Indeterminate Indeterminate ACM Non-ACM drug-induced Portal vein thrombosis due to malignancy Budd–Chiari syndrome due to malignancy Indeterminate Non-ACM drug-induced (possibly sertraline) Indeterminate
B-cell lymphoma — Cerebral herniation — Alcohol abuse — Injury of portal vein and hepatic artery during laparoscopic cholecystectomy Sepsis due to bowel ischemia Allogenic stem cell transplantation 6 mo ago — Muscular dystrophy, ICD implantation for cardiac arrhythmia Metabolic syndrome — ANA 1:16.000 Chemotherapy for lymphoma — — — — — — Alcohol abuse — Pancreatic cancer Colorectal carcinoma with hepatic metastases NAFLD Depression Metabolic syndrome, CAD, suspected bronchial carcinoma
ELT — — — Day 3 — Day 7 Day 1 — —
Days between ELT and death
Days between referral and death
— — — 4 — 5 0 — —
3 7 9 5 4 12 1 10 2
— —
— —
4 3
— —
— —
60 6
— Day 23 Day 3 — — Day 2 Day 6 Day 5
— 52 15 — — 1 122 51
3 75 17 8 45 2 128 56
Day 11 Day 2 — — —
0 1 — — —
11 2 35 11 2
—
—
1
— —
— —
1 29
—
—
48
ACM, acetaminophen; ANA, antinuclear antibody; CAD, coronary artery disease; ICD, implantable cardioverter defibrillator; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis.
Etiologies and Outcomes of Acute Liver Failure in Germany JOHANNES HADEM,* FRANK TACKE,‡ TONY BRUNS,§ JULIA LANGGARTNER,储 PAVEL STRNAD,¶ GERALD U. DENK,# PANAGIOTIS FIKATAS,** MICHAEL P. MANNS,* WOLF P. HOFMANN,‡‡ GUIDO GERKEN,§§ FRANK GRÜNHAGE,储 储 ANDREAS UMGELTER,¶¶ CHRISTIAN TRAUTWEIN,‡ and ALI CANBAY,§§ for the Acute Liver Failure Study Group Germany *Department of Gastroenterology, Hepatology & Endocrinology, Hannover Medical School, Hannover; ‡Department of Gastroenterology, Metabolic Diseases & Intensive Care Medicine, University Hospital Aachen, Aachen; §Department of Gastroenterology, Hepatology & Infectious Diseases, University Hospital Jena, Jena; 储 Department of Gastroenterology, Endocrinology, Rheumatology, Infectious Diseases & Emergency Medicine, University Hospital Regensburg, Regensburg; ¶ Department of Internal Medicine I, University Hospital Ulm, Ulm; #Medical Department II, Liver Centre Munich, University Hospital München-Großhadern, Munich; **Department of General Surgery, Visceral Surgery & Transplant Surgery, Virchow-Klinikum, Charite, Berlin; ‡‡Department of Gastroenterology, Hepatology, Pulmonary Medicine, Endocrinology & Nutrition, University Hospital Frankfurt, Johann-Wolfgang-Goethe-University, Frankfurt; §§Department of Gastroenterology & Hepatology, University Hospital Essen, Essen; 储 储Department of Medicine II, Saarland University Hospital, Homburg; and ¶¶Department of Gastroenterology, Hepatology, Endocrinology, Infectious Diseases & Intensive Care Medicine, Klinikum rechts der Isar, Technical University München, Munich, Germany
BACKGROUND & AIMS: Acute liver failure (ALF) is a severe form of acute liver injury that can progress to multiple organ failure. We investigated causes and outcomes of ALF. METHODS: Eleven university medical centers in Germany were asked to report patients with (primary) severe acute liver injury (sALI) (international normalized ratio [INR] ⬎1.5 but no hepatic encephalopathy) and primary ALF (INR ⬎1.5 with overt hepatic encephalopathy) treated from 2008 to 2009. Data were analyzed from 46 patients with sALI and 109 patients with ALF. RESULTS: The most frequent etiologies of primary ALF were non-acetaminophen drug-induced (32%), indeterminate (24%), and viral (21%); acetaminophen ingestion was the cause of ALF in only 9% of patients. The support of a ventilator was required by 44% of patients with ALF, vasopressors by 38%, and renal replacement by 36%. Seventy-nine patients with ALF (72%) survived until hospital discharge, 38 (35%) survived without emergency liver transplantation (ELT), and 51 received ELT (47%); 80% of patients who received ELT survived until discharge from the hospital. CONCLUSIONS: In Germany, drug toxicity, indeterminate etiology, and viral hepatitis appear to be the major causes of primary ALF, which has high mortality. Patients with ALF are at great risk of progressing to multiple organ failure, but 80% of patients who receive ELT survive until discharge from the hospital. Keywords: Hepatic Failure; Acute Hepatitis; King’s College Criteria; Liver Transplantation.
A
ccording to the first description by Trey and Davidson,1 acute liver failure (ALF) is an acute deterioration of liver function resulting in hepatic encephalopathy (HE) within 8 weeks of the onset of symptoms in a patient with a previously healthy liver. More recently, the American Association for the Study of Liver Diseases suggested a modified definition of ALF comprising a decline in liver function of less than 26-week duration, absence of cirrhosis, evidence of coagulation abnormality (usually indicated by an international normalized ratio [INR] ⱖ1.5), and any degree of HE.2 In clinical practice, however, considerable uncertainty exists when a patient should be considered as having ALF. This is because (1) there is an ongoing debate on whether HE should remain an obligatory element of ALF definition, and (2) standardization of neurologic scales for HE can still be regarded as preliminary.3,4
An important aspect of ALF that needs further clarification is the varying composition of ALF causes among different geographic regions.5–9 Two retrospective studies from Germany summarized 236 ALF cases from a metropolitan western area and a rural area in northern Germany. Drug toxicity, viral hepatitis, and indeterminate ALF were the most prevalent causes of ALF.10,11 Although these studies provided new data on management and outcome of ALF in Germany, the gained insight is limited because of a single center experience. In this situation, detailed knowledge on the current predominant etiologies and outcomes of ALF might help to prevent ALF and allow more selective treatment options. The Acute Liver Failure Study Group Germany was founded in June 2009 to improve the management of ALF patients in Germany by developing guidelines and establishing a nationwide ALF registry as a basis for multicenter research projects. This retrospective survey was regarded as a first approach to the epidemiology of ALF in Germany.
Methods Enrollment of Patients and Definitions of Severe Acute Liver Injury and Acute Liver Failure In 2010, all university clinics participating in the Acute Liver Failure Study Group Germany (including all major liver transplant centers) were asked to retrospectively identify patients with imminent and overt primary ALF treated at their clinics during a 2-year period (2008 –2009). Early recognition of ALF is important and has been declared as a main task of our study group. We therefore intended to Abbreviations used in this paper: ALF, acute liver failure; AUC, area under the curve; BiLE score, bilirubin-lactate-etiology score; CI, confidence interval; CMV, cytomegalovirus; DILI, drug-induced liver disease; EBV, Epstein–Barr virus; ELT, emergency liver transplantation; HAV, hepatitis A virus; HBV, hepatitis B virus; HE, hepatic encephalopathy; INR, international normalized ratio; IQR, interquartile range; KCC, King’s College criteria; MELD score, Model of End-Stage Liver Disease score; ROC, receiver operating characteristic; sALI, severe acute liver injury. © 2012 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2012.02.016
June 2012
include a wider range of disease severity than just those patients incurring encephalopathy, which is in accordance with the European Acute Liver Failure Registry. Imminent ALF or severe acute liver injury (sALI) was defined as acute (⬍12 weeks) liver dysfunction leading to an INR ⬎1.5, but without documented level of HE. Overt ALF was defined as acute (⬍12 weeks) liver dysfunction leading to an INR ⬎1.5 plus any documented level of HE. Two hundred seventy-two patients were reported by the centers and assumed to have ALF. Careful consideration of the clinical setting and time course of deteriorating events disclosed that 80 of these patients had preexisting or secondary liver failure (31.3% on top of a chronic liver disease, 35.0% in the setting of sepsis with multiple organ failure, and 32.5% caused by ischemia/shock), necessitating their exclusion. Another 37 patients were excluded because of only mild liver dysfunction (INR ⬍1.5). The remaining 46 patients with primary sALI and 109 patients with primary ALF were included in the analysis (Supplementary Figure 1).
Patient Collective and Parameter Evaluation ALF etiologies were based on accepted diagnostic criteria.2,5,6,11–15 All parameters except the maximum HE grade were documented at intensive care unit admission. HE was graded on a standard scale of 1– 4 as described previously.12 King’s College criteria (KCC) were determined as described.16 The calculation of labMELD score (below called Model for EndStage Liver Disease [MELD] score) was performed according to the equation: MELD score ⫽ [9.57 ⫻ loge creatinine (mg/dL) ⫹ 3.78 ⫻ loge bilirubin (mg/dL) ⫹ 11.20 ⫻ loge INR ⫹ 6.43]. If the individual values were ⬍1.0, they were set to 1.0; serum creatinine was adjusted to 4.0 mg/dL if the patient underwent dialysis during 2 weeks before assessment or the serum creatinine was ⬎4.0 mg/dL. MELD scores ⬎40 were set to 40.13,17 Clichy criteria could not be calculated because factor V was not available in most of the cases. Calculation of bilirubin-lactateetiology (BiLE) score was as follows: BiLE score ⫽ (baseline bilirubin [mol/L]/100) ⫹ baseline lactate [mmol/L] ⫹ 4 [in case of indeterminate ALF, Budd–Chiari syndrome, or phenprocoumon toxicity] – 2 [in case of acetaminophen toxicity] ⫹ 0 [in case of any other ALF etiology].11 Cutoff values of MELD and BiLE scores were adapted to available data.11,17,18 These parameters were then compared between the cohort of transplant-free survivors and patients who received emergency liver transplantation (ELT) or died. The study was performed in accordance with the ethical guidelines of the 1975 Declaration of Helsinki and was a priori approved by the institutional review board (#4373). Because the study relied on measurements taken as part of routine care, the Ethical Committee of Hannover Medical School waived the need for informed consent.
Statistical Analysis Data are presented as percentages or medians with corresponding 25th and 75th percentiles (interquartile range [IQR]). Comparison of baseline parameters between patient groups was performed with Student t test and U test for continuous and with 2 test for categorial parameters. Kaplan–Meier plots were used to illustrate the timing of events during follow-up in relation to baseline KCC, MELD score, or BiLE score, and statistical assessment was performed by the log-rank test. Receiver operating characteristic (ROC)
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
665
curve analysis was used for comparison of prognostic accuracy of different outcome prediction models. Tests were two-sided, and significance was accepted at P ⬍ .05. Data analysis was performed by using SPSS software (SPSS, Inc, Chicago, IL).
Results Distribution of Etiology and Outcome in Severe Acute Liver Injury Forty-six patients with sALI (imminent ALF) were reported by the participating centers. Seventeen patients (37.0%) developed sALI caused by non-acetaminophen drug toxicity, 8 (17.4%) had acute viral hepatitis, 7 (15.2%) had ingested acetaminophen, 6 (13.0%) were reported to have sALI of indeterminate cause, and 8 (17.4%) had other causes. Two of 4 sALI patients with phenprocoumon toxicity (50%) and 3 of 6 sALI patients with indeterminate cause (50%) died or underwent ELT in the absence of documented HE. Patients with other sALI etiologies exhibited high survival rates.
Distribution of Etiology and Outcome in Acute Liver Failure The etiologic distribution of the 109 patients with ALF differed significantly from those without HE (ie, sALI). ALF causes and outcomes are shown in Figure 1. Thirty-five patients (32.1%) developed ALF because of non-acetaminophen drug toxicity (8 of them because of phenprocoumon), 26 (23.9%) were reported to have ALF of indeterminate cause, 23 (21.1%) had acute viral hepatitis, and 10 (9.2%) had ingested acetaminophen. Acute viral hepatitis was due to hepatitis B virus (HBV) in 11 patients, hepatitis A virus (HAV) in 4, hepatitis E virus in 4, adenovirus in 2, Epstein–Barr virus (EBV) in 1, and cytomegalovirus (CMV) in another patient. In several patients hepatotoxicity developed because of multiple causative agents (eg, ingestion of numerous drugs). The chance of ELT-free survival was highest in Amanita ingestion (100%), acetaminophen-induced ALF (70%), and ALF related to HAV infection (75%). Conversely, risk of ELT/death appeared to be particularly high in ALF caused by malignancy, Wilson’s disease, and Budd–Chiari syndrome (100% each), indeterminate ALF (81%), viral hepatitis (65%), and non-acetaminophen drug toxicity (54% overall, but 63% in phenprocoumon toxicity).
Baseline Clinical and Laboratory Parameters Table 1 provides an overview of relevant baseline parameters. Patients with sALI were slightly younger (median age, 34 years; IQR, 17– 81) in comparison with ALF patients (46 years; IQR, 15–75). However, these differences did not reach statistical significance (P ⫽ .092, U test). No difference could be detected with respect to patients’ sex (63% female in ALF vs 59% in ALI, not significant, 2 test). Median baseline parameters of the 109 patients with ALF were INR 2.5 (1.8 – 4.1), lactate 3.0 (1.8 – 6.0) mmol/L, bilirubin 255 (94 –378) mol/L, and HE grade 1 (1–2). HE peaked at grade 1/2 in 40 of 89 (45%) and at grade 3/4 in 49 of 89 (55%).
Frequency of Organ Failure HE was, per definition, absent in all patients with sALI but present in all patients with ALF. Whereas extrahepatic
666
HADEM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 6
Figure 1. Causative etiologies and outcomes in ALF. Etiology and outcome of 109 patients with primary ALF. Patients who developed ALF due to ⬎1 causative agent appear in ⬎1 bar of the diagram. ACM, acetaminophen.
organ dysfunction appeared to be a rare event in patients with sALI, the frequency dramatically increased in patients with ALF. Ventilatory support was required in 48 ALF patients (44%), vasopressors in 41 (38%), and renal replacement therapy in 39 (36%) during intensive care unit stay (Supplementary Figure 2).
Outcomes in Severe Acute Liver Injury and Acute Liver Failure Of 46 patients with sALI, 37 (80.4%) survived spontaneously without ELT, 5 (10.9%) survived after ELT, 2 (4.3%)
underwent ELT but died postoperatively within 14 days after ELT, and 2 (4.3%) died without ELT. Overall survival to hospital discharge among sALI patients was 91.3%. In contrast, of the 109 patients with ALF, only 38 (34.9%) survived spontaneously without ELT, 41 (37.6%) survived after ELT, 10 (9.2%) underwent ELT but died after surgery, and 20 (18.3%) died without ELT (Figure 2). Patients on the waiting list for ELT received a graft within 5 (2– 6) days from admission. Overall, 79 ALF patients (72.3%) could be discharged from the hospital alive.
Table 1. Baseline Parameters in Patients With ALF (n ⫽ 109) Parameter
Overall (n ⫽ 109)
Age (y) Women, n (%) Encephalopathy grade at baseline Maximum encephalopathy grade (available in n ⫽ 89, 81.7%) 1/2 (n, %) 3/4 (n, %) INR at baseline Maximum INR WBC (per L) Creatinine (mol/L) Bilirubin (mol/L) Lactate (mmol/L) MELD score MELD score ⱖ32, n (%) KCC fulfilled (available in n ⫽ 93), n (%) BiLE score (available in n ⫽ 71), n (%) BiLE score ⱖ6.9, n (%)
46 (33–56) 69 (63.3) 1 (1–2) 3 (1–4) 40 (44.9) 49 (55.1) 2.5 (1.9–4.1) 3.6 (2.3–5.1) 10.7 (7.2–18.3) 92 (68–190) 255 (94–378) 3.0 (1.8–6.0) 29 (24–36) 44 (40.4) 40 (36.7) 8.1 (4.9–11.7) 42 (38.5)
NOTE. Data presented as median (IQR) unless otherwise stated. WBC, white blood cell count.
ELT-free survival (n ⫽ 38)
ELT and/or death (n ⫽ 71)
P value
42 (27–54) 22 (57.9) 1 (1–1) 1 (1–2)
47 (36–56) 47 (66.2) 1 (1–2) 3 (2–4)
.12 — .20 ⬍.0001
25 (80.6) 6 (19.4) 2.1 (1.7–3.2) 2.3 (1.8–4.0) 8.6 (6.3–14.1) 88 (66–173) 104 (53–214) 2.2 (1.3–4.0) 26 (20–31) 8 (21.1) 9 (28.1) 4.8 (1.6–8.1) 8 (33.3)
15 (25.9) 43 (74.1) 2.7 (1.9–4.2) 4.1 (3.1–5.3) 12.2 (7.7–19.2) 96 (69–196) 337 (189–474) 3.8 (2.0–7.2) 32 (26–37) 36 (50.7) 31 (50.8) 9.8 (6.5–13.1) 34 (72.3)
— — .13 .004 .54 .78 ⬍.0001 .02 .0007 — — .003 —
June 2012
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
Figure 2. Outcome analysis of ALF patients.
Patients progressing to ELT and/or death had a higher probability of progression to HE stage 3/4 and displayed higher baseline levels of bilirubin and lactate as well as higher maximum INR levels. An overview of those 20 patients who died without ELT is given in Supplementary Table 1.
Prediction of Emergency Liver Transplantation and/or Death in Acute Liver Failure KCC were available in 93 ALF patients. KCC were more likely to be fulfilled in patients progressing to ELT and/or death, as depicted by Kaplan–Meier analysis (log-rank test log rank 5.30, P ⫽ .0213; Figure 3A). KCC by itself achieved a sensitivity and specificity of 51% and 72%, respectively, for accurately predicting outcome. MELD score was available in all 109 patients and was significantly elevated in the patient group progressing to ELT and/or death (Table 1). With a cutoff value of 32, its sensitivity and specificity were 44% and 82%, respectively, and this cutoff also provided significant discrimination in Kaplan–Meier survival curve analysis (log-rank 8.84, P ⫽ .0030; Figure 3B). BiLE score was available in 71 ALF patients and differed significantly between the 2 prognostic groups, showing an improved discriminative power by Kaplan–Meier analysis (log rank 9.26, P ⫽ .0023; Figure 3C). With a cutoff value of 6.9, its sensitivity and specificity were 72% and 67%, respectively. In direct comparison by ROC curve analysis, BiLE score (area under the curve [AUC], 0.777; 95% confidence interval [CI], 0.652– 0.901) had better prognostic accuracy to predict outcome (transplant-free survival vs ELT or death) than the MELD score (AUC, 0.707; 95% CI, 0.570 – 0.844) or KCC (AUC, 0.583; 95% CI, 0.436 – 0.731; Figure 3D).
Discussion The main results of this multicenter retrospective analysis are (1) outcome of ALF patients is strongly associated with the underlying etiology, (2) development of HE is significantly associated with reduced chance of ELT-free survival (87% vs 35%), (3) ELT is performed in nearly half of ALF patients with an overall survival rate of 72%, and (4) KCC are widely used to
667
predict outcome of ALF but lack sensitivity in predicting a fatal outcome that can possibly be improved by BiLE score. The most common etiologies of sALI and ALF were drugs, indeterminate cause, and acute viral hepatitis. This is in accordance with the above-mentioned retrospective German singlecenter studies that analyzed patients with pediatric ALF before 2008.10,11 Interestingly, acetaminophen seems to be responsible for only about 25% of drug-induced hepatotoxicity in sALI as well as in ALF patients, which is in contrast to recent data from the United States and United Kingdom.6,19 This might be due to underestimation of acetaminophen as a cause of ALF because of (1) possible inaccurate retrospective data acquisition, (2) exclusion of acetaminophen ALF cases that were seen in non-university hospitals and therefore did not come to our attention, and (3) lack of proper history-taking concerning acetaminophen ingestion. It has been shown that half of acetaminophen overdoses are apparently unintentional,14,20 and that acetaminophen toxicity is in fact causing many cases of ALF not attributed to the drug without specific testing.21 Because low prevalences of acetaminophen-induced ALF have also been shown in the preceding German studies, the present data might just reflect a declining acetaminophen intake in our country, which would be in contrast to the United States where the annual percentage of acetaminophen-induced ALF rose from 28% in 1998 to 51% in 2003.22 Another surprising finding is that non-acetaminophen drugrelated sALI/ALF was diagnosed commonly (32.1%) in our series. A lower prevalence of non-acetaminophen drug toxicity had been noted in the 2 above-mentioned German single-center studies (23%,10 12%11). A recent retrospective German singlecenter investigation of 129 patients who underwent ELT for ALF revealed that 8% of them were drug-related.23 The US ALF study group recently published results from 1198 ALF patients enrolled during a 10.5-year period. Only 133 of these prospectively documented patients (11.1%) were deemed by expert opinion to have drug-induced liver injury (DILI).15 The present study might therefore overestimate drug-related causalities that are difficult to evaluate retrospectively. The prevalence of phenprocoumon (7.3%), a Coumadin derivative with long half-life, is likely to be underestimated. Phenprocoumon has been associated with hepatitis and subacute liver failure (2% and 0.2%, respectively, of phenprocoumonassociated adverse events in Germany).11,24 Hepatotoxicity of phenprocoumon is often detected late because monitoring of aminotransferases is not routinely performed during most cardiologic follow-ups. Our study emphasizes that better awareness of this disease entity is urgently needed. The present data are particularly interesting in face of the ongoing debate on whether HE should possibly be eliminated from the ALF definition as is already the case in pediatric ALF.25 The comparison between sALI and ALF in adult patients clearly shows that the occurrence of HE is associated with the development of extrahepatic organ failure. As a marker of imminent multiorgan dysfunction, HE is likely to be of value to discriminate patients with unfavorable prognosis. However, discrepancies in the judgment of HE between different observers are common,4 proper documentation of neurologic abnormalities is often neglected, and performing a retrospective analysis on possibly subtle neurologic alterations is even more difficult. The fact that 9 patients classified as sALI because of lack of docu-
668
HADEM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 6
Figure 3. Prediction of emergency liver transplantation and/or death by different prognostic models. Kaplan–Meier survival curve analysis of 109 patients with primary ALF comparing those who meet with those who do not meet (A) KCC, (B) MELD score, and (C) BiLE score. (D) ROC curve analysis to evaluate prognostic accuracy of KCC, MELD, and BiLE. AUC (confidence intervals) were 0.777 (0.652– 0.901) for BiLE, 0.707 (0.570 – 0.844) for MELD, and 0.583 (0.436 – 0.731) for KCC, respectively.
mented HE finally faced ELT and/or death is probably due to this problem, although KCC can theoretically predict a fatal outcome and push physicians toward ELT without HE under certain circumstances.16 ELT was obviously performed more often (47% of ALF patients) than reported from the United States (29%) or United Kingdom (37%) ALF patients, although these patients were similar with regard to age, maximum HE, and baseline INR levels.6,19 One explanation is again the much lower prevalence of acetaminophen-induced ALF (9% vs 39% and 57%, respectively)
in our cohort. Only the frequent performance of ELT in our ALF patients, of whom 56% had non-acetaminophen druginduced or indeterminate ALF (both with unfavorable prognosis), probably enabled an overall survival that was slightly higher than previously observed.6,19 The analysis of the predictive values of KCC, MELD, and BiLE scores is certainly limited by the retrospective design of this study. In addition, one might argue that defining poor outcome as ELT and/or death might have implicated a bias related to the decision as to whether to perform transplantation
June 2012
in an individual patient. However, patients who underwent ELT accounted for nearly half of the whole ALF cohort, and the high incidence of failure of a second extrahepatic vital organ system within this patient group clearly indicates an unfavorable natural course of disease. KCC are widely used as selection criteria for ELT, and their specificity has been confirmed in several large studies.26 However, its sensitivity was again limited in our patient cohort. This was also true for MELD score, which has been suggested as a prognostic tool in acetaminophen-induced ALF.18 Recently, a modification of classic MELD by cytokeratin-18 significantly improved the positive predictive value in a cohort of mainly acetaminophen-induced ALF.27 Our study indicates that the BiLE score could fill this gap. This score was empirically developed from a retrospective German single-center ALF cohort, where it achieved a promising prediction of poor outcome in 102 ALF patients.11 Although it performed slightly inferior to KCC in an external validation in 422 patients with predominating acetaminophen-induced ALF,19 our nationwide multicenter trial supports the further evaluation of the BiLE score for ALF in future prospective studies. We believe that the results of this survey are representative of the current situation of ALF in Germany. In our experience, ALF patients are transferred to one of the liver transplant centers promptly as soon as the diagnosis of ALF has been established. Only a minority of the 35 university hospitals in Germany has a liver transplant program and is involved in the care of ALF patients. The 11 participating hepatology centers are distributed all over the country and received referrals from different regions of Germany. Among them are those transplant centers with the largest volume of liver transplantations. In conclusion, we present data on the epidemiology and outcome of patients with severe acute hepatitis and ALF in Germany that should be regarded as a first step toward a prospective ALF database within Germany as well as a contribution to the existing European Acute Liver Failure Registry, which will build the basis for further improvements in ALF patient care.
Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org, and at doi:10.1016/ j.cgh.2012.02.016. References 1. Trey C, Davidson CS. The management of fulminant hepatic failure. In: Popper H, Schaffner F (eds). Progress in liver diseases. New York: Grune and Stratton, 1970:282–298. 2. Polson J, Lee WM, American Association for the Study of Liver Disease. AASLD position paper: the management of acute liver failure. Hepatology 2005;41:1179 –1197. 3. Mullen KD. Review of the final report of the 1998 working party on definition, nomenclature and diagnosis of hepatic encephalopathy. Aliment Pharmacol Ther 2007;25:11–16. 4. Córdoba J. New assessment of hepatic encephalopathy. J Hepatol 2011;54:1030 –1040. 5. O’Grady JG, Portmann B, Williams R. Fulminant hepatic failure. In: Schiff L, Schiff R (eds). Diseases of the liver. 7th ed. Philadelphia: Lippincott, 1993:852– 876. 6. Ostapowicz G, Fontana RJ, Schiødt FV, et al. Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med 2002;137:947–954.
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
669
7. Bernuau J, Rueff B, Benhamou JP. Fulminant and subfulminant liver failure: definitions and causes. Semin Liver Dis 1986;6:97–106. 8. Acharya SK, Bhatia V, Sreenivas V, et al. Efficacy of L-ornithine L-aspartate in acute liver failure: a double-blind, randomized, placebo-controlled study. Gastroenterology 2009;136:2159 –2168. 9. Bernal W, Schiødt FV, Hamid S, et al. Etiologies and outcomes for acute liver failure at 6 sites around the world. Dig Dis Sci 2005; 24:A318. 10. Canbay A, Jochum C, Bechmann LP, et al. Acute liver failure in a metropolitan area in Germany: a retrospective study (20022008). Z Gastroenterol 2009;47:807– 813. 11. Hadem J, Stiefel P, Bahr MJ, et al. Prognostic implications of lactate, bilirubin and etiology in German patients with acute liver failure. Clin Gastroenterol Hepatol 2008;6:339 –345. 12. Lee WM. Acute liver failure. N Engl J Med 1993;329:1862–1872. 13. Bundesärztekammer. Richtlinien zur Organtransplantation gem. Deutsches ärzteblatt International 2011;12:A662–A673. 14. Bernal W, Auzinger G, Dhawan A, et al. Acute liver failure. Lancet 2010;376:190 –201. 15. Reuben A, Koch DG, Lee WM, et al. Drug-induced acute liver failure: results of a U.S. multicenter, prospective study. Hepatology 2010;52:2065–2076. 16. O’Grady JG, Alexander GJ, Hayllar KM, et al. Early indicators of prognosis in fulminant hepatic failure. Gastroenterology 1989; 97:439 – 455. 17. Yantorno SE, Kremers WK, Ruf AE, et al. MELD is superior to King’s College and Clichy’s criteria to assess prognosis in fulminant hepatic failure. Liver Transplant 2007;13:822– 828. 18. Schmidt LE, Larsen FS. MELD score as a predictor of liver failure and death in patients with acetaminophen-induced liver injury. Hepatology 2007;45:789 –796. 19. Bernal W, Auzinger G, Wendon J. Prognostic utility of the bilirubin lactate and etiology score. Clin Gastroenterol Hepatol 2009;7:249. 20. Jalan R, Williams R, Bernuau J. Paracetamol: are therapeutic doses entirely safe? Lancet 2006;368:2195–2196. 21. Davern TJ 2nd, James LP, Hinson JA, et al. Measurement of serum-acetaminophen protein adducts in patients with acute liver failure. Gastroenterology 2006;130:687– 694. 22. Larson AM, Polson J, Fontana RJ, et al. Acetaminophen-induced acute liver failure: results of a United States multicenter prospective study. Hepatology 2005;42:1364 –1372. 23. Fikatas P, Ulrich F, Lee JE, et al. The APACHE III score as preoperative indicator of patient outcome in liver transplantation after fulminant hepatic failure. Ann Transplant 2011;16:18 –25. 24. Schimanski CC, Burg J, Möhler M, et al. Phenprocoumon-induced liver disease ranges from mild acute hepatitis to (sub-) acute liver failure. J Hepatol 2004;41:67–74. 25. Squires RH Jr, Shneider BL, Bucuvalas J, et al. Acute liver failure in children: the first 348 patients in the pediatric acute liver failure study group. J Pediatr 2006;148:652– 658. 26. Bailey B, Amre DK, Gaudreault P. Fulminant hepatic failure secondary to acetaminophen poisoning: a systematic review and meta-analysis of prognostic criteria determining the need for liver transplantation. Crit Care Med 2003;31:299 –305. 27. Bechmann LP, Jochum C, Kocabayoglu P, et al. Cytokeratin-18 based modification of the MELD score improves prediction of spontaneous survival after acute liver failure. J Hepatol 2010;53:639–647.
Reprint requests Address requests for reprints to: Johannes Hadem, MD, Department of Gastroenterology, Hepatology, and Endocrinology, Hannover Medical School, Carl-Neuberg-Strasse 1, 30625 Hannover, Germany. e-mail: [email protected]; fax: ⴙ49-511-532-2144. Conflicts of interest The authors disclose no conflicts.
June 2012
ETIOLOGIES AND OUTCOMES OF ALF IN GERMANY
669.e1
Supplementary Figure 2. Frequency of extrahepatic organ failure in 46 patients with sALI and 109 patients with ALF.
Supplementary Figure 1. Enrollment of study patients.
669.e2
HADEM ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 10, No. 6
Supplementary Table 1. Characteristics of Subjects Who Died After ALF Recognition Patient number
Age (y)/sex
(Suspected) ALF etiology
Significant comorbidities
1 2 3 4 5 6 7 8 9
55/F 63/M 54/M 56/F 51/F 53/F 47/M 25/M 59/M
Phenprocoumon, clindamycin Acute hepatitis B Reactivation of latent hepatitis B Indeterminate Indeterminate Acute hepatitis B Non-ACM drug-induced Non-ACM drug-induced Ischemic hepatitis
— —
10 11
66/F 55/M
12 13
45/F 56/F
Phenprocoumon Viral hepatitis (CMV and adenovirus) EBV hepatitis Phenprocoumon
14 15 16 17 18 19 20 21
70/F 75/M 65/F 38/M 53/M 39/F 58/F 62/F
22 23 24 25 26
43/F 54/F 57/M 66/M 56/F
27
62/M
28 29
56/F 49/F
30
69/M
Indeterminate Indeterminate NASH Non-ACM drug-induced Vanishing bile duct syndrome Indeterminate Non-ACM drug-induced (Valproat) Non-ACM drug-induced (ibuprofen) Indeterminate Indeterminate ACM Non-ACM drug-induced Portal vein thrombosis due to malignancy Budd–Chiari syndrome due to malignancy Indeterminate Non-ACM drug-induced (possibly sertraline) Indeterminate
B-cell lymphoma — Cerebral herniation — Alcohol abuse — Injury of portal vein and hepatic artery during laparoscopic cholecystectomy Sepsis due to bowel ischemia Allogenic stem cell transplantation 6 mo ago — Muscular dystrophy, ICD implantation for cardiac arrhythmia Metabolic syndrome — ANA 1:16.000 Chemotherapy for lymphoma — — — — — — Alcohol abuse — Pancreatic cancer Colorectal carcinoma with hepatic metastases NAFLD Depression Metabolic syndrome, CAD, suspected bronchial carcinoma
ELT — — — Day 3 — Day 7 Day 1 — —
Days between ELT and death
Days between referral and death
— — — 4 — 5 0 — —
3 7 9 5 4 12 1 10 2
— —
— —
4 3
— —
— —
60 6
— Day 23 Day 3 — — Day 2 Day 6 Day 5
— 52 15 — — 1 122 51
3 75 17 8 45 2 128 56
Day 11 Day 2 — — —
0 1 — — —
11 2 35 11 2
—
—
1
— —
— —
1 29
—
—
48
ACM, acetaminophen; ANA, antinuclear antibody; CAD, coronary artery disease; ICD, implantable cardioverter defibrillator; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis.