Expanding the boundaries of kidney replacement therapy in patients with liver failure

Abstract

Acute kidney injury (AKI) is common in patients with liver failure, and for a significant subset it is severe enough to require kidney replacement therapy (KRT). Patients with liver failure have distinct clinical characteristics (e.g., cardio-circulatory dysfunction and a tendency to bleed) that mandate customization of their overall care including KRT. Herein, we provide an overview of AKI in liver failure, discuss the basic pathophysiology of hepatorenal syndrome, including the often-underemphasized role of the heart in its clinical manifestations, and the current therapies afforded to these patients. We also discuss the general aspects of KRT and how they apply to patients with liver failure (e.g., preference for continuous renal replacement therapy and the need for regional, instead of systemic, anticoagulation). Moreover, we discuss hyperammonemia, an emerging non-renal indication of KRT in this patient population, and provide recommendations on how this therapy may be applied in this setting.

Key Words: Hepatorenal syndrome; Cirrhotic cardiomyopathy; Continuous renal replacement therapy; Liver transplantation; Hyperammonemia

Core Tip: Acute kidney injury (AKI) is common in patients with liver failure, and a significant subset requires kidney replacement therapy (KRT). These patients have distinct clinical characteristics and needs that mandate customization of their overall care, including KRT. This general overview is meant for critical care providers to familiarize themselves with the nuances of KRT in patients with concomitant liver failure and AKI.

INTRODUCTION

Acute kidney injury (AKI) is common in acute liver failure (ALF). In a retrospective analysis of data compiled by the United States ALF Study Group, approximately 70% of patients with ALF were found to have had an episode of AKI[1], with approximately 30% of those requiring kidney replacement therapy (KRT). ALF is defined by the presence of an acute onset of liver function deterioration, elevated international normalized ratio of 1.5 or greater, and hepatic encephalopathy (HE) in patients without known pre-existing liver disease[2]. There is regional variance, but overall, the ALF is relatively uncommon in the Western world, with a calculated incidence of around 1.13-1.61/100000 person-years[3-5]. Overall mortality for ALF is estimated on the order of 30%, with the estimated mortality without a liver transplant (LT) being as high as 50%, driven by multiple mechanisms, including shock and multi-organ failure[6,7]. As can be seen, a significant subset of these patients with ALF are critically ill. The presence of AKI, especially one severe enough to require KRT, portends worse short and long-term outcomes similar to other indications for intensive care[1,8].

A similar pattern is present in patients with cirrhosis as well, with an estimated incidence of AKI of 29% with a 6-fold increase of in-hospital mortality in cirrhotic patients with AKI[9,10]. The presence of AKI in patients with cirrhosis is associated with increased mortality when compared to an AKI on chronic kidney disease (CKD) in patients with cirrhosis[11]. Additionally, the severity of AKI in a decompensated cirrhotic prior to transplant contributes to the likelihood of renal recovery 4-6 weeks post-transplant, thus lending consideration for simultaneous liver-kidney transplantation[12]. The indications for KRT in these patients are fairly similar to the other critically ill patients, including uremic complications, fluid overload with oliguria or anuria, medically refractory hyperkalemia, or metabolic acidosis[13,14]. It should be noted however, that the presentation may be different due to additional factors in patients with cirrhosis, such as the common use of mineralocorticoid receptor antagonists, and the tendency for acidosis and intravascular volume depletion due to frequent diuretic and laxative use. Further considerations for KRT, which are distinct for patients with liver failure, include severe hyperammonemia with progressive HE[15]. Continuous renal replacement therapy (CRRT) is often the preferred modality of KRT in these patients due to their tendency for circulatory dysfunction and hemodynamic instability that may contribute to compromised cerebral perfusion even in the absence of sepsis[16]. Moreover, patients with liver failure often have subclinical cardiac function derangements that can be underappreciated in their management (see below)[17]. To summarize, the factors associated with increased mortality in liver failure-induced AKI include having high rates of intensive care unit admissions, having an AKI without a concomitant known history of CKD, worse liver function parameters, those with higher peak serum creatinine (Cr) values and rates of requiring KRT, and those diagnosed with acute on chronic liver failure[11,18].

AKI IN LIVER FAILURE AND HEPATORENAL SYNDROME

The criteria to diagnose AKI in patients with liver disease and cirrhosis, endorsed by the International Club of Ascites, is based on the Kidney Disease Improving Global Outcomes definition of a serum Cr increase by 0.3 mg/dL within 48 hours or an increase in Cr 50% above baseline Cr within 48 hours, or urine volume below 0.5 mL/kg/h for 6 hours[19]. In liver failure with portal hypertension, there is a special entity, hepatorenal syndrome (HRS) associated AKI (HRS-AKI), that is refractory to volume resuscitation and is due to reduced renal perfusion due to exaggerated splanchnic vasodilation with maladaptive hyperactivity of the neurohormonal axis[9,20]. This HRS-AKI can exist in addition to the common pre-renal, intra-renal, and post-renal etiologies[9,17]. The exact mechanism of this systemic vasodilation is not precise. Still, specific molecules such as the vasodilator nitric oxide (NO) and its distribution have been implicated in playing a key role in the development of portal hypertension, with it being high in the splanchnic circulation while being reduced in the intrahepatic microcirculation[21]. In the late stages of cirrhosis, the maintenance of arterial pressure is thought to be largely dependent on the activation of vasoconstrictor systems such as the renin-angiotensin-aldosterone system (RAAS) and the non-osmotic increased secretion of arginine vasopressin thus leading to ascites/edema and, due to subsequent intense intrarenal vasoconstriction and hypoperfusion, renal failure[22-24]. Recent data suggest that this circulatory dysfunction on display in HRS may overlap with a syndrome of cardiac dysfunction in cirrhosis that plays a role in kidney dysfunction in HRS[17,25]. In response to the reduction in systemic vascular resistance, there is an initial compensatory increase in cardiac output to maintain effective circulatory blood volume that eventually exhausts the cardiac reserve[17]. Cirrhotic cardiomyopathy is defined by cardiac dysfunction in cirrhosis suggested by electromechanical abnormalities, enlarged left atrium, and humoral changes such as an elevated brain natriuretic peptide[26]. This cardiomyopathy is defined by systolic dysfunction characterized by blunted responsiveness of cardiac output to volume and postural or pharmacological challenges that can often identified by advanced diagnostic techniques[27-29]. The systolic dysfunction is usually combined with impaired diastolic relaxation detected by echocardiography of E to A ratio less than one, prolonged deceleration time greater than 200 ms, and prolonged isovolumetric relaxation time greater than 80 ms[26,30]. As in HRS, the pathophysiology of cirrhotic cardiomyopathy is thought to be at least partly due to the pro-inflammatory state of cirrhosis and its corresponding increased activation of cellular signaling pathways due to increased circulating levels of c-reactive proteins and cytokines, such as tumor necrosis factor (TNF)-α and interleukin 6 (IL-6) alongside NO[24]. Based on animal models, these biomarkers could contribute to reduced cardiac contractility and abnormal function of the cardiac myofilaments[30,31]. This mechanism of liver failure-induced AKI is depicted graphically in Figure 1.

Figure 1 Pathophysiologic pathways linking the liver, heart, and kidney in patients with advanced cirrhosis[25]. The orange circle depicts the conventional concept of hepatorenal syndrome, where renal dysfunction is mediated through splanchnic vasodilation and consequent neurohormonal activation. The green circle highlights the hepatocardiac pathways through which cirrhosis leads to cardiac dysfunction (i.e., cirrhotic cardiomyopathy). The yellow circle summarizes the bidirectional mechanisms linking cardiac dysfunction to aberrant renal function (i.e., cardiorenal syndrome). SNS: Sympathetic nervous system; AVP: Arginine vasopressin; GFR: Glomerular filtration rate; LVEDP: Left ventricular end-diastolic pressure. Citation: Kazory A, Ronco C. Hepatorenal Syndrome or Hepatocardiorenal Syndrome: Revisiting Basic Concepts in View of Emerging Data. Cardiorenal Med 2019; 9: 1-7 Copyright ©Silverchair Publisher 2018. Published by Karger Publishers (Supplementary material).

While in a recently published retrospective study of hospitalized patients with cirrhosis with AKI, the AKI etiology was not independently associated with an increased risk of 90-day mortality, it is crucial for a provider to distinguish AKI in liver failure as HRS-AKI and non-HRS-AKI given differences in its management[32]. Acute tubular necrosis was the most common etiology of non-HRS-AKI among patients hospitalized with cirrhosis[32,33]. As is often the case in medicine, prevention is much more cost-effective than treatment; thus, for general measures, it is recommended to avoid hypovolemia, hypotension, known nephrotoxins and to provide prompt treatment of infections in cirrhosis[34].

There are specific implications to the management of HRS with this schema that HRS-AKI is borne out of maladaptive RAAS activation, inflammatory mediators involved in oxidative stress, and the knowledge that the heart plays a key role in HRS. For example, the use of albumin along with a vasoconstrictor has long been considered the cornerstone of therapy for HRS. However, in addition to volume expansion and resultant RAAS activation, albumin administration may also provide antioxidant and anti-inflammatory effects, modulating mediators like nuclear factor kappa-light-chain-enhancer activated B cells and NO synthase in cardiac tissue, thus improving cardiac contractility and enhancement of organ perfusion[17,35]. In addition to albumin, the use of terlipressin, the synthetic vasopressin analog derived from the natural hormone lysine-vasopressin, has been extensively studied in HRS-AKI. The investigators have reported that it results in splanchnic vasoconstriction, shunting of blood to the systemic circulation, and reducing activation of RAAS and catecholamine release, thus blunting the release of arginine vasopressin and enhancing renal perfusion[36-38]. The recently published CONFIRM (A Multi-Center, Randomized, Placebo-Controlled Double-Blind Study to Confirm Efficacy and Safety of Terlipressin in Subjects With HRS Type 1) trial showed that terlipressin was able to reverse HRS-AKI[39]. Of note, there were significant respiratory events in the CONFIRM trial that may be modulated by optimizing patients’ intravascular volume status and avoiding excessive albumin administration[36,39]. It is within this context that point-of-care ultrasonography may play a role in the evaluation and management of cirrhotic patients with AKI[40]. It should be noted that with the increasing awareness of the key role of the heart in the development and progression of HRS (i.e. considering HRS as a hepatic form of cardiorenal syndrome), there may be implications for change in the current management of these patients[17]. For example, more recent guidelines have emphasized on the fact that not only albumin administration is not necessary in all cases of HRS-AKI, but it may be useful to use diuretics to decongest these patients[9,17].

KRT for patients with liver failure

There are several points that need to be taken into account when deciding to initiate KRT for patients with advanced liver disease and cirrhosis, particularly given that intradialytic hypotension, which is associated with increased mortality, can rapidly occur in these patients with baseline cardio-circulatory dysfunction[14,41]. Studies have established that continuous modalities offer advantages related to hemodynamic stability when compared against intermittent hemodialysis (iHD) or intermittent hemodiafiltration in patients with HRS-AKI requiring RRT to survive to transplantation[42,43]. However, it is reasonable to use intermittent modalities for patients with advanced CKD and cirrhosis who are hemodynamically stable as a bridge to simultaneous liver and kidney transplant while utilizing midodrine to help reduce intradialytic hypotension and decrease the need for frequent drainage of ascitic fluid[16,44]. There are no worldwide standardized allocation criteria for simultaneous kidney transplantation with LTs, and the need for and timing of kidney transplantation is based on guidance from the Organ Procurement Transplant Network (OPTN) and United Network of Organ Sharing (UNOS) policy statement in the United States[45]. Briefly, OPTN/UNOS guidance advises simultaneous listing for a kidney transplant from the same deceased donor if at the time of a candidate being listed for a LT should they also have confirmed CKD and glomerular filtration rate (GFR) less than 30 mL/min or requiring dialysis for confirmed end-stage kidney disease, a sustained AKI with a GFR less than 25 mL/min or AKI requiring dialysis for 6 weeks prior to listing, or confirmed diagnosis of qualifying metabolic disease like hyperoxaluria[46]. Should the candidate not meet these criteria and the LT proceeds alone with sustained impaired advanced kidney dysfunction post-LT, they may be considered for a kidney after the LT[46].

In addition to mitigation of intra-treatment hypotension, CRRT also provides more effective solute clearance as well as the possibility of correcting hyponatremia in a controlled fashion. This is important prior to a potential LT given that cirrhotic patients may be severely hyponatremic, and physiologic changes from the sudden increase in plasma osmolality due to rapid correction of hyponatremia after LT can exacerbate intracranial hypertension or precipitate pontine demyelination[16]. Additionally, patients with ALF are at increased risk of cerebral edema; iHD is known to increase intracranial pressure, leading to patients’ neurologic deterioration[47]. Though continuous, the instantaneous clearance of CRRT is lower than iHD. Therefore, to maintain similar efficiency as iHD, every effort must be made to optimize circuit uptime and reduce circuit clotting, hence leading to the need for anticoagulation[48]. Given the well-recognized propensity for patients with liver failure to bleed, it can be counterintuitive that they can also be more prothrombotic, necessitating anticoagulation, and while “low-dose heparin” can be used[49], the effect of heparin can be unpredictable in these patients due to reduced antithrombin[16,50]. Regional anticoagulation with citrate would be a suitable option for CRRT in liver failure. However, due to the impairment of hepatic citrate metabolism in liver failure, the half-life is increased, leading to an increased risk of citrate toxicity[50]. Thus, if citrate is used, it should be started at about half the usual dose and then titrated to the lowest dose necessary to maintain a therapeutic post-filter ionized calcium while monitoring closely for evidence of citrate toxicity (e.g., high anion gap metabolic acidosis associated with a total calcium/ionized calcium ratio of greater than 2.5). The effluent dose can be augmented to enhance citrate removal if needed[51].

The conventional low dose of 25 mL/kg/h for CRRT is often utilized as a high dose CRRT of 40 mL/kg/h has not yet been established to confer superior benefits in terms of mortality reduction[8,52]. However, a higher dose of CRRT has been suggested to improve ammonia clearance[53]. Serum arterial ammonia is believed to be the main driver for cerebral edema in ALF, with higher levels associated with the onset of severe HE[6,54]. However, there is literature implicating serum endotoxin and inflammatory mediators (TNF, IL-6, IL-18) in HE[55]. For patients with AKI and ALF or cirrhosis managed with CRRT, our group tends to use higher clearances to potentially remove ammonia and other inflammatory mediators. A few hours prior to a LT, we lower serum potassium targets. This is due to the high concentration of potassium in common LT preservation solutions, which can lead to potential hyperkalemia upon reperfusion of the transplanted liver during surgery[56]. Moreover, we practice intra-operative CRRT to more efficiently manage fluid and electrolytes during the LT procedure.

The optimal timing for initiation of KRT in patients with decompensated cirrhosis is not known. In general, early KRT initiation has not been associated with better outcomes in critically ill patients[57]. Decompensated cirrhotic patients with AKI requiring KRT that are transplant ineligible have dismal prognoses, experience very short-term survival, and require intensive end-of-life care[58]. The decision for KRT initiation for patients with decompensated cirrhosis with AKI where LT is not an option or unlikely should be made with shared decision-making with patients and families to permit goals of care discussions among multidisciplinary care teams[14,59].

Expanding KRT beyond renal indications in liver failure

While higher levels of blood ammonia have been implicated in HE in ALF by causing cytotoxic cerebral edema and neuronal oxidative stress[60], inconsistencies in the direct correlation between ammonia concentration and severity of HE in cirrhosis suggest it may not be the sole responsible factor[61]. Rather, it is suspected that inflammation, infection, and ammonia have a complex synergistic relationship in bringing about HE, so there is insufficient data supporting KRT use in chronic liver dysfunction for hyperammonemia[62]. There is a dearth of large cohort studies on adults with chronic liver failure who received KRT for hyperammonemia, with much of the data on the use of KRT for elevated ammonia based on case studies and studies on neonates with inborn errors of metabolism and urea cycle disorders[62]. iHD is the modality that affords the most rapid removal of ammonia, given that it is a small molecule that is not significantly protein-bound. However, there is a significant rebound effect within a few hours of treatment discontinuation, and therefore CRRT has been used in this setting as, in addition to its advantage in patients with hemodynamic instability, it allows continuous removal of ammonia[63,64]. For this reason, it is recommended that iHD be considered as the initial KRT of choice for stable patients with severe hyperammonemia with concern for immediate cerebral herniation with a plan to transition patients to CRRT to prevent a rebound in ammonia level and use of CRRT in hemodynamically unstable patients with hyperammonemia[65-67]. In our institution, for hemodynamically stable patients with ALF and severe hyperammonemia, we perform an extended iHD followed by high-dose CRRT that is meant to reduce the rebound phenomenon. Figure 2 offers a practical approach to choosing modes of KRT in the setting of liver failure.

Figure 2 Practical approach to kidney replacement therapy in liver failure. This figure depicts a practical approach to initiation of kidney replacement therapy in the setting of liver failure. CRRT: Continuous renal replacement therapy; MARS: Molecular adsorbent recycling system.

Molecular adsorbent recycling system

While a complete review of the liver support therapies is beyond the scope of this review, it suffices to mention that it is a form of extracorporeal therapy in which albumin is used within the Molecular Adsorbent Recycling System (MARS)[65]. Briefly, it is a nonbiologic liver support system that reduces levels of protein-bound and water-soluble toxins such as ammonia, aromatic amino acids, and several inflammatory cytokines[68,69]. MARS is composed of 2 separate dialysis circuits, with the first containing human serum albumin that is in contact with the patient’s blood and allows toxins to pass through a semipermeable membrane. Then, this toxin-laden albumin solution is sent through another dialyzer using a conventional CRRT machine to remove water-soluble toxins with counter-current bicarbonate-based dialysate, and the protein-bound toxins are removed by adsorbent cartridges containing an anion exchanger and activated charcoal[68,70]. There also exists the fractionated plasma separation and adsorption (Prometheus) system[68]. This Prometheus system is also beyond the scope of this review, so briefly speaking, it is a type of albumin dialysis that utilizes a 250-kilo Dalton semipermeable membrane that generates an albumin-containing plasma-like solution that is then adsorbed on two albumin-detoxifying columns upon which hemodialysis is performed[71].

CONCLUSION

AKI and the need for KRT in liver failure portends a poor prognosis for patients. CRRT is often preferred given the advantages of cardiovascular stability and the possibility of gradual fluid extraction. Hyperammonemia is an emerging non-renal indication for initiation of KRT in this population. Although KRT in this setting is biologically relevant and its role in the management of hyperammonemia in children has been well studied, there remains a relative dearth of high-quality large studies in adults with liver failure to determine the optimal timing of initiation, dose, and the impact on the outcomes. Based on the data available it appears that high-dose CRRT provides more effective ammonia clearance than standard dose. However there does not yet appear to be literature to support improved mortality with high-dose CRRT. Regenerative medicine and gene therapy are among emerging fields that have been expanding exponentially to include several pathologic circumstances including AKI. Whether they will have a role in this specific form of AKI remains to be elucidated by future studies.