REVIEW URRENT C OPINION

Update in intensive care medicine: acute liver failure. Initial management, supportive treatment and who to transplant Chris Willars

Purpose of review Acute liver failure (ALF) is associated with significant mortality. Although specific therapies may be available, the evidence base for these and for many aspects of supportive therapy has been slow to emerge. Liver transplantation continues to be a cornerstone of treatment, and the management of ALF, therefore, remains the domain of the specialist ICU. The purpose of this review is to identify and critically appraise the recent evidence and to inspire those who strive to provide excellent care for a difficult patient cohort. Recent findings Effective vaccination programmes have reduced the incidence of viral hepatitis in Europe and the USA. Spontaneous survival has improved in causes such as acetaminophen toxicity. Early recognition and proactive intensive management have reduced the incidence of early neurological death. The use of artificial liver assist devices and therapeutic plasma exchange is controversial, yet intriguing, with some early evidence of efficacy. Summary Increasingly sophisticated prognostication tools are evolving, which have the potential to transform clinical decision-making. A review of the indications for transplantation in acetaminophen toxicity is overdue. The use of therapeutic plasma exchange and extracorporeal liver support in ALF requires further investigation. Keywords acetaminophen toxicity, acute liver failure, artificial liver support, plasma exchange, prognostication

INTRODUCTION Acute liver failure (ALF) is a clinical entity which is separate and distinct from other forms of liver disease. It is rapidly progressive, life-threatening and occurs when there is massive liver injury. ALF is characterized by coagulopathy and encephalopathy, which occurs within days or weeks. Liver cell death is accompanied by an inflammatory response which has the potential to propagate multiorgan failure. Within the specialist ICU, the prognosis of ALF has been transformed over the last 4 decades. Hospital survival in ALF was as low as 17% in 1973–1978, rising to 62% in 2004–2008 [1 ]. This phenomenon is not restricted to those undergoing emergent liver transplantation. Nontransplanted survival has also risen from 17 to 48% over the same period [1 ], as increasingly sophisticated strategies for intensive support have been developed. Acetaminophen toxicity remains a common cause of ALF, although deaths and transplant registrations (but not numbers of actual transplants

performed) because of acetaminophen toxicity have diminished significantly in the UK after legislation was introduced in 1998 (following recommendations by the UK Medicines Control Agency) restricting the volume of acetaminophen available for purchase to 32  500 mg packs in pharmacy and 16  500 mg in nonpharmacy sales [2]. Acetaminophen poisoning has risen inexorably in both France [3] and the United States [4] during the same period. However, data from the European Transplant register demonstrates that acetaminophen poisoning is responsible for only 12% of transplants for

&&

&&

www.co-criticalcare.com

Liver Intensive Care Unit, King’s College Hospital, London, UK Correspondence to Chris Willars, MBBS, FRCA, FFICM, Institute of Liver Studies, 3rd Floor Cheyne Wing, King’s College Hospital, Denmark Hill, London, SE5 9RS, UK. Tel: +44 203 299 9000; fax: +44 203 299 3899; e-mail: [email protected]; Web: www.liverintensive care.co.uk Curr Opin Crit Care 2014, 20:202–209 DOI:10.1097/MCC.0000000000000073 Volume 20  Number 2  April 2014

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Update in intensive care medicine Willars

KEY POINTS  The efficacy of liver transplantation in the context of acute liver failure is well established.  Improved spontaneous survival with causes such as acetaminophen toxicity may force clinicians to rethink the tools used for prognostication.  Increasingly sophisticated prognostication tools are evolving, which have the potential to transform clinical decision-making.  A review of the indications for transplantation in acetaminophen toxicity is overdue.  Given the high prevalence of drug-induced and viral causes worldwide, public health measures will be key in reducing the global incidence of disease of young adults which carries a high mortality.  The efficacy of therapeutic plasma exchange and extracorporeal liver support requires further investigation.

DEFINITIONS The term ‘acute liver failure’ is used inconsistently within the medical literature [8] and multiple definitions have been proposed. The essence of ALF is a severe hepatic dysfunction which manifests within 6 months of the onset of symptoms. The diagnosis of ALF is dependent on the presence of hepatic encephalopathy. If there is cirrhosis or evolution is over a period exceeding 6 months, then chronic liver disease is said to be present [9]. Acute-on-chronic liver failure (ACLF) is said to be present when chronic liver disease is associated with organ failure, and is most often secondary to spontaneous bacterial peritonitis, variceal bleeding and other infective processes. The terms ‘acute hepatitis’ or ‘acute liver injury’ describes deranged liver function evolving within a 6-month period with reduced synthetic function [manifested by coagulopathy, with international normalized ratio (INR) 1.5].

PATHOPHYSIOLOGY Cellular injury and death is because of necrosis and early and transient apoptosis. Apoptotic protein array profiling demonstrates that early elevation in serum cell death markers is predictive of poor outcome in ALF because of acetaminophen toxicity [10,11]. Following liver cell death, innate immune responses trigger massive inflammation which drives vasoplegic circulatory shock. Monocytes traffic to the liver and differentiate into macrophages. Pro-inflammatory and anti-inflammatory cytokines drive the systemic inflammatory response (SIRS) and compensatory anti-inflammatory response (CARS).

(b) 16 14 12 10 8 6 4 2 0

Incidence (per 100 000 people per year)

(a)

Incidence (per 100 000 people per year)

ALF [5] – approximately three times lower than the UK, where the proportion of ALF transplants related to acetaminophen has remained 36–39% [6]. Hepatitis B virus-induced and hepatitis A virus (HAV)-induced ALF have declined in incidence throughout Europe and the USA [7]. Effective vaccines have been available for hepatitis B since 1981 and hepatitis A since 1995 (Fig. 1). Viral cause remains the most common cause of ALF worldwide. Seronegative (indeterminate hepatitis) is commonly associated with subacute presentation, which is associated with a poor outcome in the absence of transplantation.

16 14 12 10 8 6 4 2 0 1982

1985

1988

1991

1994

1997

2000

2003

2006

Year

FIGURE 1. Incidence of hepatitis A virus (a) and hepatitis B virus (b) per 100 000 people in the USA, 1982–2007. Data from Daniels et al. [7]. 1070-5295 ß 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

www.co-criticalcare.com

203

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Gastrointestinal system

Monocyte deactivation, recurrent sepsis and multiple organ dysfunction result from an imbalance in this system [12]. Monocyte deactivation (as evidenced by reduced HLA-DR) appears to be a powerful prognostic marker of outcome in ALF [13].

PROGNOSTICATION Prognostic models facilitate the early identification of patients in whom spontaneous recovery is less likely, and are therefore unlikely to survive in the absence of liver transplantation. The most enduring models are the King’s College criteria (Fig. 2) [14], Clichy criteria and the Model for End-Stage Liver Disease (MELD). These models remain imperfect, with a number of studies reporting inadequate sensitivity and specificity and the search for tools with greater discriminative ability continues. The consequences of failure of any prediction model are severe – individuals who would otherwise survive without transplantation are subjected to major surgery, with its incumbent morbidity and mortality and to a lifetime of immunosuppression. A graft is lost from the donor pool, disadvantaging another patient. Conversely, failure of a prognostic tool to identify a patient who will go on to die without transplantation results in a death which might otherwise have been avoided. A recent systematic review of prediction models [15] reported on 20 studies, of which 10 studies compared newly developed models to existing models. Although these studies all reported superiority of the new models over old, methodological flaws and reporting limitations were identified in all of the studies, and the predictive performance of any new model has yet to be evaluated prospectively in a large cohort of patients.

The ALF early dynamic (ALFED) model proposed by Kumar et al. [16 ] is a prediction model based on the early dynamicity of four variables: arterial ammonia, INR, serum bilirubin and hepatic encephalopathy. The derivation (n ¼ 244) and validation cohorts (n ¼ 136) consisted largely of patients with viral and indeterminate (seronegative) hepatitis. The authors report that the ALFED model was superior to the King’s College criteria and the MELD score, even when their 3-day serial values were taken into consideration (area under the receiver operator curve 0.91 and 0.92, respectively). It might be argued that as consistently poor outcomes are reported in patients with subacute presentations (often indeterminate and seronegative cause) in the absence of transplantation, these are the patients who will derive most benefit from emergent transplantation [17 ]. Spontaneous (untransplanted) survival in acetaminophen overdose, however, has improved over the last 2 decades [1 ], but there has been no evolution in the selection criteria to reflect this. A high-performance prediction model specific to acetaminophen toxicity might be a useful adjunct to clinical decision-making. In the absence of such a tool, a practical aid to decision-making has been presented (Fig. 3) [17 ]. &&

&

&&

&

Proceed • Persistence of accepted criteria associated with a poor prognosis • Absence of co-morbidity independent of acute liver failure that would impact on survival • Absence of complications of acute liver failure associated with reduced survival • Absence of psychosocial profile suggestive of poor adaptation of rigour of post-transplant survival

Wait • Patients showing sustained evidence of improvement of prognostic criteria in the absence of clinical deterioration • Paracetamol induced acute liver failure patients who do not have grade 3 or 4 encephalopathy irrespective of severity of coagulopathy • Patients with paracetamol induced acute liver failure and severe acidosis or elevated serum lactate that responds rapidly to resuscitative measures • Most patients when the liver allocated is marginal, especially steatotic, non-ABO identical or split, ABOincompatible or the donor is aged over 60 years

Paracetamol Arterial pH less than 7.3 following adequate volume resuscitation, or combination of encephalopathy grade 3 or more, creatinine 300 µmol/L or more, and international normalised ratio more than 6.5 Non-Paracetamol Any grade encephalopathy and international normalised ratio 6.5 or more, or any three of: international normalised more than 3.5, bilirubin 300 mmol/L or more, age less than 10 or more than 40 years, unfavourable cause (drug-induced liver injury, seronegative disease)

FIGURE 2. Modified King’s College criteria for ALF. ALF, acute liver failure. Data from Bernal et al. [14]. 204

www.co-criticalcare.com

Stop • Evidence of compromised brainstem function, especially fixed and dilated pupils • Invasive fungal infection • Rapidly escalating inotrope requirements • Severe pancreatitis [usually in paracetamol-related ALF]

FIGURE 3. Decision-making triggers in patients with acute liver failure. Data from O’Grady [17 ]. &

Volume 20  Number 2  April 2014

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Update in intensive care medicine Willars

EARLY SUPPORTIVE MANAGEMENT Early management of the patient with ALF includes (1) (2) (3) (4) (5)

recognition and referral; investigation and diagnosis; resuscitative and supportive measures; prognostication; management of organ failure and intracranial hypertension; (6) specialist intervention, including liver transplantation. Although evidence-based protocols for patient management have been proposed [18,19], specialist input from hepatology, transplant surgery and radiology complements the principles of intensive support and early transfer to a high-volume specialist liver centre is essential [20].

(2)

(3)

(4)

Investigation and diagnosis The clinical classification described by O’Grady et al. [21] over 2 decades ago (Table 1) has endured. The time elapsed between the development of clinical jaundice and the onset of encephalopathy informs the clinician not only of likely cause, but also of clinical progression. Hyperacute liver failure is associated with profound coagulopathy, high-grade encephalopathy, severe organ dysfunction and a higher incidence of intracranial hypertension, but confers higher rates of spontaneous survival. Subacute liver failure has a more insidious onset, but outcomes tend to be worse in the absence of transplantation. Determination of the cause of ALF (see list below) is also important in terms of prognostication and also because specific therapies (Table 2) [22–26] are available, albeit that randomized controlled trial (RCT) evidence of efficacy for a majority of specific therapies is lacking. Causes of ALF are as follows:

(5)

(6)

(b) Herpes simplex, cytomegalovirus, chickenpox – usually limited to immunocompromised hosts; drug related: (a) acetaminophen; (b) antituberculous drugs; (c) recreational drugs (ecstasy and cocaine); (d) idiosyncratic reactions [anticonvulsants, antibiotics and nonsteroidal anti-inflammatory drugs (NSAIDs)]; (e) aspirin in children may lead to Reye’s syndrome; (f) kava kava; toxins: (a) carbontetrachloride, phosphorus, Amanita phalloides, alcohol; vascular events: (a) ischaemia, veno-occlusive disease and Budd–Chiari syndrome (hepatic vein thrombosis); (b) hyperthermic liver injury; pregnancy: (a) acute fatty liver of pregnancy, haemolysis, elevated liver enzymes and low platelets syndrome and liver rupture; other: (a) Wilson’s disease, autoimmune, lymphoma, carcinoma, haemophagocytic syndrome and trauma.

Clinical consistency may be achieved through a standardized investigative pathway. In addition to laboratory testing (see list below), Doppler ultrasound, axial computed tomography imaging and echocardiography may provide useful information. Liver biopsy is rarely undertaken in the acute setting because of the high risk of bleeding. Nonspecific confluent necrosis is the most common histological finding and does not contribute to diagnosis. Table 2. Specific therapies in ALF

(1) viral hepatitis: (a) hepatitis A, B, D, E, seronegative hepatitis; Table 1. Classification of ALF Definition

Time

Common causes

Cause

Intervention (s)

RCT

HBV [22]

Lamivudine/adefovir

Yes (nonfulminant; n ¼ 35)

HSV [23]

Aciclovir/foscarnet/TPE

No

Autoimmunea

Immunosuppression

No

Hyperacute

Update in intensive care medicine: acute liver failure. Initial management, supportive treatment and who to transplant.

Acute liver failure (ALF) is associated with significant mortality. Although specific therapies may be available, the evidence base for these and for ...
419KB Sizes 1 Downloads 3 Views