Extracorporeal Liver Support: A Bridge to Somewhere

Abstract

Liver failure (LF) is associated with significant morbidity and mortality and can occur in the presence or absence of underlying chronic liver disease. ACLF generally presents after an acute decompensation in a patient with preexisting liver disease or cirrhosis. In contrast, acute liver failure (ALF) arises in the absence of underlying chronic liver disease, and in Western countries is most commonly caused by acetaminophen toxicity. The underlying mechanisms involved in LF are multifactorial and are associated with the accumulation of hepatotoxins, including proinflammatory cytokines, vasoactive toxins, and endotoxins released from intestinal flora. Impaired hepatic clearance of these toxins leads to a cascade of biochemical processes, including oxidative stress, increased capillary permeability, immune dysregulation, and cellular damage. ECLS devices were developed with the intention of removing circulating toxins, preventing further damage, and stabilizing liver function while allowing for recovery of native hepatic function or bridging to liver transplantation.1, 2 ECLS devices can be separated into two broad categories: artificial (A-ECLS) and bioartificial ECLS (B-ECLS)2 (Table 1). A-ECLS devices use endogenous or exogenous albumin molecules to bind and scavenge hepatotoxins followed by membrane-based filtration. B-ECLS devices instead use a cell-based approach as part of a bioactive system to support hepatic synthetic function and detoxification capacity. For the purposes of this review, we will focus on the mechanisms and utility of A-ECLS devices. Currently, there are three commercially available A-ECLS devices: (1) molecular adsorbent recirculating system (MARS; Gambro, Lund, Sweden), (2) fractionated plasma separation and adsorption device (Prometheus), and (3) single-pass albumin dialysis (SPAD). MARS functions to remove circulating molecules by perfusing a patient’s blood through an albumin circuit consisting of an albumin-impregnated membrane with adsorbent columns and a circulating 20% human albumin dialysate (Fig. 1A). This circuit allows for free passage of molecules <50 kDa, but molecules >50 kDa bind to the albumin membrane and are filtered out. This albumin circuit is coupled with a standard continuous renal replacement therapy (CRRT) circuit to remove water-soluble toxins. In the Prometheus (Fresenius Medical Care, Bad Homburg, Germany) system, a 250-kDa albumin-permeable filter is used to fractionate and separate circulating plasma from the blood (Fig. 1B). Subsequently, the separated plasma is purified through an anion-exchange column and passed through a standard CRRT circuit before being returned to the patient. The third system, SPAD, uses a 4% albumin dialysate in a standard CRRT circuit (Fig. 1C). In 2013, a randomized controlled trial (RCT) conducted in 16 French liver transplantation centers failed to demonstrate a mortality benefit of MARS therapy compared with conventional treatment in ALF.3 A major limitation of this study was the short time from treatment initiation to liver transplantation among all patients (median, 16.2 hours), which may not reflect the availability of liver allografts in other countries. Despite the lack of mortality benefit, the use of MARS remains appealing given the consistent improvements seen in hepatic encephalopathy (HE) and hemodynamics.4, 5 The beneficial effect on hemodynamics was initially hypothesized to be related to the elimination of proinflammatory cytokines and vasoactive toxins; however, the RELIEF study did not demonstrate a decrease in circulating cytokine levels with MARS therapy.6 Nevertheless, significant improvements in mean arterial pressures, systemic vascular resistance, and oxygen delivery have been noted even with a single MARS session,7, 8 which may be important in slowing the progression of organ failure and allowing time for native liver recovery. Similar to MARS, Prometheus has not demonstrated a survival benefit in primary endpoint analyses in patients with LF. In the HELIOS study of 145 patients with ACLF randomized to standard medical therapy (SMT) with or without Prometheus, the Prometheus group demonstrated a significantly decreased bilirubin level compared with the SMT-alone group.9 Although there was no difference in survival at day 28 or 90 between the two groups, a subgroup analysis of patients with type 1 hepatorenal syndrome (HRS; 28-day survival probability, 62% versus 39%; 90-day survival probability, 42% versus 6%, respectively; log-rank test, P = 0.04) or those with a Model for End-Stage Liver Disease (MELD) score >30 (28-day survival probability: 57% versus 42%, respectively; 90-day survival probability: 48% versus 9%, respectively; log-rank test, P = 0.02) demonstrated a statistically significant survival benefit. Prospective studies of Prometheus in ALF are lacking. One study did compare the efficacy of MARS versus Prometheus in a small cohort of 18 patients with alcoholic cirrhosis and superimposed alcoholic hepatitis.4 Patients were randomized to SMT alone, SMT + MARS, or SMT + Prometheus. Both Prometheus and MARS treatment were associated with significant reductions in bilirubin levels; however, Prometheus demonstrated greater reductions than MARS. MARS, however, was the only treatment associated with improvements in mean arterial pressure and systemic vascular resistance. Despite the consistent effects in bilirubin levels and improvements in hemodynamics (in the case of MARS), neither MARS nor Prometheus has shown a consistent survival benefit over SMT. One hypothesis for this is that both systems are predicated on albumin binding of circulating toxins. In the setting of LF, both the quantity and binding capacity of albumin may be compromised and therefore may hamper the detoxification ability of these devices. SPAD was compared with MARS in a small RCT of 32 patients with LF.5 Although both systems resulted in significantly reduced plasma bilirubin levels, only MARS functioned to reduce levels of bile acids, creatinine, and urea. In contrast, SPAD was associated with increased metabolic derangements and electrolyte disturbances. Despite advances in intensive care management, mortality in the setting of severe LF remains high. Liver transplantation remains the only curative option for patients with irreversible LF. Various case series have demonstrated promise in the use of ECLS devices in bridging patients with severe LF of varying causes, including ALF, ACLD, and severe liver trauma.10 In a retrospective review of 27 patients undergoing MARS therapy for severe ALF, 13 patients underwent MARS therapy with the intention of bridging to liver transplantation.10 Overall, the mean MELD score in this cohort of patients was 40, and all patients were listed as status 1A on the transplant wait list. Patients underwent a mean of 2.5 MARS sessions (range 1-6). Four patients did not undergo liver transplant, and their mortality rate was 100%. Of the remaining nine patients who did undergo transplantation, 7/9 (77.8%) survived with normal hepatic function, representing a significant improvement in survival (0% versus 77.8%; P < 0.01). MARS therapy in patients successfully bridged to transplantation was associated with improvement in various laboratory parameters, including aspartate aminotransferase, alanine aminotransferase, and creatinine. Although these results are promising, at this time definitive statements regarding the efficacy of ECLS devices in bridging patients to transplantation are limited by a paucity of randomized data11 (Table 2). LF presents a therapeutic challenge given the various etiologies, severity of illness at presentation, and often multiorgan system involvement. Despite improvements in the critical care of patients with LF, as well as advances in liver transplantation, morbidity and mortality remain high. ECLS devices provide an opportunity to stabilize liver function while allowing for native liver recovery or as a bridge to transplantation. Prior to routine implementation of ECLS devices, larger prospective RCTs are necessary to define the optimal modality, indications, and patient population that would most benefit.

Topics & Concepts