Alimentary Pharmacology and Therapeutics
Review article: advances in the management of patients with cirrhosis and portal hypertension-related renal dysfunction J. A. Leithead*,†, P. C. Hayes‡ & J. W. Ferguson*
*Liver Unit, Queen Elizabeth Hospital, Birmingham, UK. † NIHR Biomedical Research Unit and Centre for Liver Research, University of Birmingham, Birmingham, UK. ‡ Department of Hepatology, Royal Infirmary of Edinburgh, Edinburgh, UK.
Correspondence to: Dr J. A. Leithead, NIHR Biomedical Research Unit, Centre for Liver Research, Institute of Biomedical Research (5th floor), University of Birmingham, Edgbaston, Birmingham, UK. E-mail: [email protected]
Publication data Submitted 30 September 2013 First decision 12 October 2013 Resubmitted 30 December 2013 Accepted 19 January 2014 EV Pub Online 13 February 2014 This commissioned review article was subject to full peer-review and the authors received an honorarium from Wiley, on behalf of AP&T.
SUMMARY Background In cirrhosis, portal hypertension is associated with a spectrum of renal dysfunction that has significant implications for morbidity and mortality. Aim To discuss recent progress in the patho-physiological mechanisms and therapeutic options for portal hypertension-related renal dysfunction. Methods A literature search using Pubmed was performed. Results Portal hypertension-related renal dysfunction occurs in the setting of marked neuro-humoral and circulatory derangement. A systemic inflammatory response is a pathogenetic factor in advanced disease. Such physiological changes render the individual vulnerable to further deterioration of renal function. Patients are primed to develop acute kidney injury when exposed to additional ‘hits’, such as sepsis. Recent progress has been made regarding our understanding of the aetiopathogenesis. However, treatment options once hepatorenal syndrome develops are limited, and prognosis remains poor. Various strategies to prevent acute kidney injury are suggested. Conclusion Prevention of acute kidney injury in high risk patients with cirrhosis and portal hypertension-related renal dysfunction should be a clinical priority. Aliment Pharmacol Ther 2014; 39: 699–711
ª 2014 John Wiley & Sons Ltd doi:10.1111/apt.12653
699
J. A. Leithead et al. INTRODUCTION Renal dysfunction is a common complication of chronic liver disease (CLD) and has significant implications for patient morbidity and mortality. Portal hypertension is accompanied by progressive systemic circulatory derangement and a spectrum of renal dysfunction that evolves in parallel with advancing disease. This portal hypertension-related renal dysfunction manifests as tubular dysfunction in early compensated cirrhosis and, at the extreme end, hepatorenal syndrome (HRS) and renal failure.1 Importantly, if exposed to additional ‘hits’, the physiological changes of portal hypertension render the CLD patient vulnerable to further deterioration of renal function. Patients are primed to develop acute kidney injury (AKI), and when presenting with triggers such as sepsis and gastro-intestinal haemorrhage, or when undergoing large volume paracentesis, prevention of renal impairment should be a key focus of care.2, 3 Besides portal hypertension-related renal dysfunction, patients with CLD have a relatively high frequency of intrinsic chronic kidney disease reflecting the prevalence of comorbidities such as diabetes mellitus and hypertension, and glomerulonephritides associated with alcoholic liver disease, hepatitis B and hepatitis C.4, 5 The aetiology of renal dysfunction in this setting has important ramifications for prognosis. In portal hypertension-related renal dysfunction, the glomerular filtration rate (GFR) only falls appreciably once liver disease is advanced and circulatory derangement severe.1 Consequently, the 3-month probability of survival for hospitalised patients with HRS is only 15% compared with 73% for those with parenchymal nephropathy.2 In this review, we discuss recent progress in our understanding of the patho-physiological mechanisms and therapeutic options for portal hypertension-related renal dysfunction. Particular attention will be paid to the prevention of AKI and HRS in the primed patient. METHODS A literature search was conducted in 2013 on Pub-MED using the following search terms: HRS, liver cirrhosis, portal hypertension, renal dysfunction, AKI. After considering the titles and abstracts, the full-text articles that appeared to be relevant were examined in detail. The references of retrieved articles were also reviewed, and additional relevant articles were evaluated. The most scientifically robust and pertinent manuscripts were referenced in this comprehensive review. 700
Definitions and spectrum of portal hypertensionrelated renal dysfunction HRS is characterised by renal vasoconstriction and reduced GFR, and is a complication of advanced cirrhosis.1, 6 Central to the diagnostic criteria is the exclusion of alternative causes of renal dysfunction such as nephrotoxic drugs, diuretic-induced hypovolaemia and renal parenchymal disease (Table 1). The syndrome is divided into two distinct clinical subtypes. Type 1 HRS is defined as a doubling of the serum creatinine to a level >226 lmol/L (2.5 mg/dL) within a 2-week period, and has a clinical picture of profound circulatory dysfunction, intense sodium and water retention, and dilutional hyponatraemia.6–8 Median survival time without treatment has been estimated at 1.7 weeks, with a probability of survival of 25% at 1 month.7 On the other hand, type 2 HRS defines more modest renal impairment, as indicated by a serum creatinine >133 lmol/L (1.5 mg/dL).6 Renal impairment progresses slowly, the predominant clinical feature is of refractory ascites and the median survival time is reported to be 6 months.6, 9 Although the diagnostic criteria for HRS are widely employed, their dependence on a rigid cut-off value of serum creatinine has rendered them controversial.10, 11 Creatinine production is proportional to muscle mass, and is greater in men than women, in younger than older individuals and in black people than in Caucasians despite similar GFR. In cirrhosis specifically, reduced creatine production by the liver, muscle wasting and increased renal tubular secretion contribute to a falsely low serum creatinine.12 Furthermore, the poor prognosis of patients once the HRS criteria are met is consistent with extreme renal dysfunction that is rarely reversible even with optimal treatment. Earlier Table 1 | New diagnostic criteria for hepatorenal syndrome in patients with cirrhosis6 Diagnostic criteria for hepatorenal syndrome Cirrhosis with ascites Serum creatinine >133 lmol/L (1.5 mg/dL) No improvement of serum creatinine (decrease to level ≤133 lmol/L) after at least 2 days with diuretic withdrawal and volume expansion with albumin. The recommended dose of albumin is 1 g/kg of body weight per day up to a maximum of 100 g/day Absence of shock No current or recent treatment with nephrotoxic drugs Absence of parenchymal kidney disease as indicated by proteinuria >500 mg/day, microhaematuria (>50 red blood cells per high power field) and/or abnormal renal ultrasonography Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
Review: management of portal hypertension-related renal dysfunction identification of renal impairment and intervention is likely to be beneficial. Recent guidelines by Kidney Disease: Improving Global Outcomes (KDIGO) have advised that AKI is defined as an increase in serum creatinine ≥26.5 lmol/L (0.3 mg/dL) in 48 h, or a rise to ≥1.5 times baseline within 7 days.13 This classification has been shown to predict survival in patients with cirrhosis.14 Yet, opinion remains divided amongst Hepatologists as to the most appropriate marker of renal impairment.10, 11 We advocate that the change in serum creatinine from baseline is the ideal tool for real-life practise, and any observed rise is an early warning requiring action to prevent the development of HRS.10 In a patient hospitalised with acute alcoholic hepatitis, for example, an increase in serum creatinine from 20 to 40 lmol/L may be indicative of clinically relevant renal dysfunction irrespective of the absolute value remaining below the normal range of many laboratories. It is important to emphasise that the renal dysfunction of portal hypertension is a spectrum, not a black and white phenomenon, and that the fall in GFR that heralds HRS is a late sequelae. From early cirrhosis, there is evidence of increased tubular sodium reabsorption that is followed chronologically by reduced free water clearance and renal haemodynamic dysfunction.1 Non-ascitic cirrhotic patients have altered renal sodium metabolism, with decreased natriuresis in the standing position and following a saline load.15, 16 With disease progression, sodium retention becomes overt, there is a positive sodium balance and ascites develop. Once refractory ascites presents there is profound sodium retention, with patients excreting less than 10 mmol of sodium per day.17 Impaired free water clearance is a feature of decompensated cirrhosis, occurring after sodium retention, and evidenced by dilutional hyponatraemia despite the increased total body sodium.1 Fifty per cent of patients with ascites demonstrate hyponatraemia and the prevalence is greater in those with refractory ascites.18 Ascites and hyponatraemia should be considered clinical manifestations of pre-HRS portal hypertension-related renal dysfunction. Cirrhotic patients with these complications are at increased risk of HRS, and both are prognostic markers independent of serum creatinine.7, 19, 20
Incidence of portal hypertension-related renal dysfunction In view of the limitations of the diagnostic criteria and obvious practical difficulties of long-term follow-up Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
studies in patients with early cirrhosis, the incidence of portal hypertension-related renal dysfunction remains poorly described. In a population presenting with variceal haemorrhage, the 5-year probability of the development of ascites was reported to be 73%, and of the development of HRS was 21%.21 Gines et al. observed that in patients with ascites but maintained GFR, the probability of HRS occurrence was 18% and 39% by 1and 5-years respectively.7 In a more contemporary study, the 1-year probability of any form of functional renal impairment (including ‘pre-renal’, infection and HRS) after the first episode of ascites was 23.6%.22 When patients with refractory ascites were observed for a mean duration of 18 months, approximately half developed type 2 HRS.23
Pathophysiology of portal hypertension-related renal dysfunction An understanding of the underlying pathophysiology of portal hypertension-related renal dysfunction is crucial to appreciate the vulnerability of renal function in CLD patients. The driving force is profound systemic circulatory derangement, the so called ‘splanchnic steal phenomenon’.24 Pooling of blood within a dilated splanchnic vascular bed is accompanied by the activation of multiple homoeostatic neurohumoral systems, and secondary peripheral vasoconstriction with reduced tissue perfusion ensues. The resultant effects within the kidney are the tubular and haemodynamic changes that typify the renal dysfunction in this setting.25, 26 Nitric oxide appears to be of central importance in both instigation and potentiation of the splanchnic vasodilation.27–30 In early portal hypertension, vascular stretch within the intestinal microcirculation may trigger endothelial nitric oxide synthase (eNOS) upregulation, and it is speculated that this is the first step in the neurohumoral and circulatory cascade.31, 32 In later disease shear stress, endotoxaemia and inflammatory mediators contribute via increased eNOS as well as increased inducible nitric oxide synthase (iNOS) activity.33–42 With progressive splanchnic vasodilatation and intravascular underfilling, high pressure baroreceptors activate the renin-angiotensin-aldosterone system (RAAS) and the sympathetic nervous system (SNS), and stimulate nonosmotic hypersecretion of vasopressin.43–46 These systems act to maintain arterial blood pressure, lead to compensatory peripheral vasoconstriction, and are considered the principal mediators of the renal tubular and renal haemodynamic dysfunction.1, 8, 47–51 With progressive splanchnic vasodilatation, the circulation becomes 701
J. A. Leithead et al. increasingly hyperdynamic. A relatively low cardiac output state further exacerbates the arterial underfilling.52, 53 There may be a cirrhotic cardiomyopathy characterised by myocardial hypertrophy, diastolic dysfunction, and an abnormal inotropic and chronotropic response.54, 55 The clinical picture is one of reduced total systemic vascular resistance, increased heart rate and cardiac output, expanded blood volume, and eventually arterial hypotension.56–58 Exaggerated renal tubular sodium reabsorption is observed when renal blood flow and GFR are still within normal limits. At this point, sodium retention is solely at the tubular level.25, 59 Later, once the GFR drops, reduced filtered sodium contributes and exacerbates sodium retention.1 Patients with ascites frequently have increased circulating levels of plasma renin activity (PRA) and aldosterone, and the administration of the angiotensin II receptor antagonist losartan and the aldosterone antagonist spironolactone reduces sodium retention.60–63 Consequently, the RAAS is thought to play a key role in the increased tubular sodium reabsorption.60 Moderate SNS activity that may be insufficient to alter renal haemodynamics similarly increases sodium retention throughout the nephron via a direct tubular effect. Therefore, the SNS is thought to be an additional important stimulus for sodium reabsorption.64, 65 Vasopressin or anti-diuretic hormone is probably the chief mediator of impaired free water clearance.1 Vasopressin stimulates aquaporin-2 channel insertion into the epithelial cell membrane increasing the permeability of the distal nephron to water.66 Urinary aquaporin-2 excretion is markedly increased in patients with cirrhosis, rises in parallel with Child-Pugh class and demonstrates a significant negative correlation with spontaneous free water clearance.66 Moreover, vasopressin 2 receptor antagonists have been shown to have an aquaretic effect and to increase serum sodium in cirrhotic patients with hyponatraemia and ascites.67, 68 The hallmark of advanced portal hypertension-related renal dysfunction is renal haemodynamic dysfunction.1 With advancing circulatory derangement and neuro-humoral activation, there is progressive renal vasoconstriction and a fall in total renal blood flow. Intra-renal compensatory mechanisms initially maintain the GFR at normal or even elevated levels.69–71 With progressive arterial underfilling, the accompanying activity of the vasoconstricting systems becomes extreme, and the renal production of vasodilators eventually falls.72 Renal ischaemia stimulates additional intra-renal secretion of endothelin-1 and increased SNS 702
activity. A self-perpetuating cycle develops within the kidney.73–75 The balance is tipped in favour of reduced GFR, and HRS develops.
The role of the systemic inflammatory response in portal hypertension-related renal dysfunction The systemic inflammatory response is increasingly recognised as a pathogenetic factor in the circulatory dysfunction of advanced cirrhosis, in the absence of overt infection. Patients with Child-Pugh Class C cirrhosis demonstrate an increased frequency of bacterial translocation of enteric organisms to mesenteric lymph nodes.76 Bacterial DNA is present in the blood of approximately 40% of non-infected patients with ascites.77–79 In addition, the plasma levels of lipopolysaccharide binding protein (LBP), which is considered a better marker of transient endotoxaemia, given its longer half-life, are also elevated.80 In rats with cirrhosis, bacterial translocation is associated with nitric oxide overproduction in the mesenteric vasculature, which appears to aggravate arterial vasodilatation.37 Humans with advanced cirrhosis display increased mesenteric lymph node tumour necrosis factor–alpha (TNF-a) expression.81 Moreover, ascitic patients with high LBP levels have a more pronounced circulatory dysfunction.80 The exaggerated immunohaemodynamic derangement is reversed by the administration of norfloxacin, an effect that is not reproduced in cirrhotics with ascites and normal LBP levels.80, 82 Tying this together, in decompensated cirrhosis, bacterial translocation and the secondary systemic inflammatory response may result in increased nitric oxide generation and exacerbate the pre-existing portal hypertensive syndrome. Indeed, long-term prophylactic antibiotics in patients with advanced CLD reduce the incidence of HRS and improve survival, independent of the prevention of infection.83 The additional ‘hit’ theory Hospitalised cirrhotic patients have a high rate of renal dysfunction. Using nonstandardised definitions, the incidence of renal dysfunction has been reported to be approximately 25%.84 At time of admission to the Intensive Care Unit, the prevalence may be at least twice that.85 The majority of episodes of renal dysfunction are in the context of infection, or hypovolaemia secondary to bleeding, diuretics or gastro-intestinal fluid losses.2 Although the short-term prognosis of patients with such renal dysfunction is superior to patients with HRS, the 90-day survival remains only 31–46%.2 Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
Review: management of portal hypertension-related renal dysfunction The systemic haemodynamic and renal derangement of portal hypertension render the CLD patient vulnerable to AKI when further insults occur (Figure 1). The superimposed ‘hit’ most studied is sepsis. In addition to the baseline low grade systemic inflammatory response demonstrated by many patients, there is an exaggerated circulatory response to infection. Cirrhotics have a greater and more prolonged rise in circulating inflammatory mediators compared with patients without liver disease, and more marked haemodynamic changes.86, 87 Acute renal injury is common with a reported incidence as high as 45% in bacteraemia of unknown origin, 33% in spontaneous bacterial peritonitis and 29% in pneumonia.88, 89 In the Intensive Care Unit, cirrhosis is associated with a 2 times increased risk of renal failure.90 Likewise, cirrhotic patients hospitalised with gastrointestinal bleeding are more likely to develop renal injury when compared with noncirrhotic patients.91
When large volume paracentesis is performed without the administration of plasma expanders, 75% of patients will develop Paracentesis-Induced Circulatory Dysfunction (PICD), a condition characterised by marked activation of the RAAS and often associated with renal dysfunction.3 Mechanical decompression, shear stress-related eNOS upregulation within the splanchnic vasculature and ascites re-accumulation exacerbate the baseline neuro-humoral and circulatory derangement.3, 92, 93 Within 6 days of the procedure, patients demonstrate a significant increase in serum creatinine and fall in serum sodium levels.3, 92 Thereafter, there is faster re-accumulation of ascites, an increased diuretic requirement and impaired prognosis.3 Similar long-term consequences are observed in patients who have suffered a septic episode. After resolution of infection, cirrhotics with renal failure demonstrate a further deterioration in haemodynamics, with even greater over-activity of the
Diuretic controlled ascites
Drugs (NSAIDs, ACE-I)
Hyponatraemia
Large volume paracentesis
Infection Refractory ascites Gastrointestinal haemorrhage
Type 1 HRS
Type 2 HRS
Figure 1 | The additional ‘hit’ theory of portal hypertension-related renal dysfunction: patients with cirrhosis and portal hypertension are primed to develop acute kidney injury, and additional ‘hits’ may precipitate an irreversible deterioration in renal function with long-lasting implications. Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
703
J. A. Leithead et al. RAAS and SNS, aggravation of portal hypertension and increased mortality.53, 89 Therefore, it is speculated that additional ‘hits’ that would have transient effects only in other patient populations may precipitate an irreversible deterioration in cirrhotics with lasting prognostic implications.
Treatment of portal hypertension-related renal dysfunction Despite progress in our understanding of the mechanisms underlying the spectrum of portal hypertension-related renal dysfunction, treatment options remain limited. In the earlier phase, management is supportive, aiming to minimise patient morbidity as a result of sodium and water retention. In later phases, specific therapies are employed to increase the effective circulating volume, renal perfusion and GFR, both directly and via a reduction in endogenous vasoactive mediators. The onset of type 1 HRS should prompt a search for the precipitating event. In particular, infection should be considered and there should be a low threshold for antibiotics even in culture-negative patients. Intravascular volume status should be assessed and corrected as appropriate. Type 2 HRS signifies end-stage disease and therapy primarily focuses on management of refractory ascites and hyponatraemia. Nonsteroidal anti-inflammatory drugs, and angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor antagonists, are avoided because of the reno-protective effect of the prostaglandins and angiotensin II, respectively. Management of sodium and water retention. In patients with clinically apparent ascites, a salt-restricted diet (80 to 120 mmol sodium per day) is recommended and escalating doses of diuretics are employed.94 The simultaneous administration of an aldosterone antagonist and loop diuretic from the beginning of treatment may be more efficacious than sequential diuretic treatment alone.95 The addition of clonidine, a centrally acting a2-agonist with sympatholytic activity, to patients with increased SNS activity is associated with reduced diuretic requirements and slower ascites re-accumulation.96 Oral fluid restriction is poorly adhered to and seldom effective.94 Large volume paracentesis is first-line therapy for patients with grade 3 ascites prior to initiation of maintenance diuretics, or may be the main stay of treatment in refractory ascites.94 AKI is not a contraindication to paracentesis. In this setting, large volume paracentesis with albumin administration is accompanied by an increase in renal perfusion pressure, creatinine clearance 704
and fractional excretion of sodium, possibly as a result of the fall in intra-abdominal pressure.97 A meta-analysis has suggested that vasopressin 2 receptor antagonists reduce the time to first paracentesis, and increase the serum sodium. However, the authors concluded that the cost and modest benefit did not justify their routine clinical use.98
Therapies that increase the effective circulating volume, renal perfusion and glomerular filtration rate. Terlipressin is a vasopressin analogue with predominant vasopressin 1, although some vasopressin 2 receptor effects.99 Terlipressin is mainly thought to improve renal perfusion in portal hypertension-related renal dysfunction by reducing splanchnic vasodilatation and hence splanchnic steal.24 In patients with ascites without HRS, terlipressin resulted in improved systemic haemodynamics, a fall in circulating neurohumoral mediators, reduced renal arterial resistive index, improved GFR and increased urinary sodium excretion.100, 101 When ascites is nonrefractory the increase in GFR occurs in association with an increase in filtration fraction but no change in total renal blood flow, supporting the concept that some of the benefit may be mediated by post-glomerular vasoconstriction.101 Randomised controlled trials have demonstrated that terlipressin reverses type 1 HRS and results in a small reduction in short-term mortality.102 Nevertheless, prognosis remains poor with a reported 6-month transplant-free survival of only 13%.103, 104 In type 2 HRS, renal failure invariably recurs following treatment withdrawal.6 However, terlipressin may be useful as a bridge to liver transplantation in this cohort.105 Noradrenaline appears to be as effective as terlipressin for treatment of type 1 HRS, but has practical disadvantages and is generally used only in countries where terlipressin is not available.106–108 Intravenous human albumin has multiple beneficial properties in patients with portal hypertension. In addition to the volume loading provided by less concentrated preparations and increase in oncotic pressure, albumin binds endotoxin, has anti-inflammatory and antioxidant effects and, consequently, may impact on endothelial and cardiac dysfunction.109 Administration of albumin results in an increase in systemic vascular resistance and improvement in cardiac function, which is not seen following hydroxethyl starch.110, 111 In patients with ascites receiving diuretics, intravenous albumin therapy is associated with a faster rate of ascites mobilisation and a lower probability of ascites Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
Review: management of portal hypertension-related renal dysfunction reaccumulation.112 Furthermore, a daily albumin infusion has been shown to be superior to fluid restriction for correcting severe hyponatraemia.113 In those with HRS, the addition of albumin to terlipressin resulted in a three times increased occurrence of reversal of renal dysfunction.114 The main predictor of reversal of type 1 HRS with terlipressin is a lower baseline serum creatinine.104, 115 Fifty per cent of patients with a serum creatinine less than 265 lmol/L (3.5 mg/dL) demonstrate reversal, compared to 11% of patients with a serum creatinine greater than 442 lmol/L (5 mg/dL).115 This emphasises the importance of earlier intervention in such patients and suggests survival benefit with even lesser degrees of renal dysfunction. Our local practice is to administer 40–60 grams of albumin per day to patients with at least a 50% increase in serum creatinine from baseline (stage 1 AKI).13 Terlipressin is commenced in patients with no response after 24–48 h, in all patients with at least a two times increased serum creatinine (stage 2 AKI), and/or those patients fulfilling the diagnostic criteria for type 1 HRS.13
Nonpharmacological therapies. Transjugular intra-hepatic portosystemic shunt (TIPSS) has a role in the management of select patients with portal hypertension-related renal dysfunction. Although the immediate effect is an exacerbation of the hyperdynamic state as evidenced by an increase in cardiac output and reduction in systemic vascular resistance, in longer term, the fall in portal pressure is associated with improved circulatory and neuro-humoural derangement.116, 117 Patients demonstrate a marked and sustained fall in PRA, and plasma aldosterone and noradrenaline concentrations.118–121 In those with refractory ascites, and in HRS, TIPSS results in increased total renal blood flow and GFR, and less sodium retention.118–121 Meta-analyses of randomised controlled trials confirm that TIPSS in refractory ascites is associated with a reduced recurrence of ascites when compared with large volume paracentesis.122, 123 In this group, there may be a small but statistically significant positive effect on survival.124 Nevertheless, patients with poor synthetic function are unlikely to benefit and a serum bilirubin greater than 85 lmol/L is considered an absolute contra-indication.94, 124 The applicability of TIPSS once HRS develops is likely to be limited, given the high prevalence of jaundice and encephalopathy.94 Haemofiltration may be useful in some patients with AKI when there is a reversible precipitating factor such as infection.125 There remains a paucity of evidence to Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
support this recommendation. However, reversibility of type 1 HRS has been strongly correlated with resolution of infection.126 Moreover, in a single observational study, 30-day survival was reported to be 50% if renal replacement therapy was the only organ support.127 When mechanical ventilation was also instituted, no patients survived.127 Longer term prognosis is nevertheless extremely poor.127 Consequently, the Acute Dialysis Quality Initiative (ADQI) Group has advised that renal support is only helpful if the patient is a candidate for liver transplantation.125 Equally, in type 2 HRS, renal replacement therapy is not appropriate unless as a bridge to transplant. Albumin dialysis with the molecular adsorbent recirculating system (MARS) has been shown in a single small randomised controlled trial to improve renal function in type 1 HRS when compared with haemodiafiltration alone.128 Yet, whether the effects are sustained is unknown and MARS remains an experimental therapy only at present.94 Liver transplantation is the definitive treatment for portal hypertension-related renal dysfunction, offering a clear survival advantage.129 Most patients with HRS demonstrate an improvement in renal function, although post-transplant chronic kidney disease is more common in this group.129 Duration of reduced GFR is important in influencing outcomes, probably reflecting the development of secondary acute tubular necrosis.130, 131 Combined liver kidney transplantation is recommended for all patients with stage 3 AKI or a GFR of less than 25– 35 mL/min for more than 4 weeks.132 This population is considered to gain survival benefit from combined liver kidney transplantation over liver transplantation alone, although the supporting literature is flawed by the heterogeneity of the cohorts studied.132, 133 Terlipressin may have a role in improving survival to transplantation of patients with type 1 and type 2 HRS, and may reduce post-transplant renal disease and mortality.105 However, pre-transplant optimisation of renal function could disadvantage patients by prolonging the wait-list time. Current organ allocation policies are based on prognostic scores that are heavily weighted by serum creatinine. It has been suggested therefore that HRS is considered a priority criterion for organ allocation, and accompanied by, for example, MELD exception points.105
The prevention of hepatorenal syndrome Given that patients with CLD are primed to develop AKI even when the GFR is preserved, and the poor prognosis once HRS develops, prevention of HRS in high-risk individuals should be a clinical priority (Table 2). 705
J. A. Leithead et al. Early recognition of infection and prompt initiation of antibiotic therapy are imperative in patients presenting with decompensated cirrhosis. In those with spontaneous bacterial peritonitis, and possibly other infections, the addition of intravenous albumin results in a reduced incidence of renal impairment and death.110, 134, 135 Antibiotic prophylaxis in gastrointestinal bleeding has been shown to prevent bacterial infection and improve survival.136 Furthermore, long-term antibiotic administration to prevent recurrence of spontaneous bacterial peritonitis has also been advocated, although it remains controversial because of the possibility of resistant bacteria.137 Other particular subgroups in which the benefits of antibiotic prophylaxis may outweigh the risks are those with low ascitic fluid protein concentrations and high serum bilirubin levels.83, 137, 138 Pentoxifylline has been associated with a reduced risk of HRS and mortality in patients with acute alcoholic hepatitis, and in advanced cirrhosis resulted in a lower probability of bacterial infection and renal impairment despite no effect on survival.139, 140 The proposed mechanism of action is downregulation of TNF-a with dampening of the immuno-haemodynamic derangement.141 Moreover, pentoxifylline may reduce intestinal bacterial overgrowth and bacterial translocation directly.141 Nonselective beta-blockers may have a similar effect.142 It is noteworthy, however, that beta-blockade in patients with refractory ascites has been linked with increased
mortality, possibly via a negative effect on cardiac function.143, 144 Nevertheless, the suggestion that nonselective beta-blockers should be contraindicated in these patients requires corroboration, and has been widely challenged by the Hepatology community. In large volume paracentesis, intravenous albumin reduces the secondary systemic and renal dysfunction, with beneficial effects on long-term prognosis.145, 146 The administration of a vasoconstrictor (terlipressin or noradrenaline) may be equally advantageous to the administration of albumin.147, 148 Whether the combination of both has additional positive effects on preventing PICD is yet to be assessed.
CONCLUSION The spectrum of portal hypertension-related renal dysfunction first becomes evident in early, compensated cirrhosis and evolves in parallel with advancing disease. Initial renal tubular sodium retention progresses to impaired free water clearance. The GFR only falls appreciably when the circulatory derangement becomes extreme. The physiological changes of portal hypertension render the individual vulnerable to further deterioration of renal function. Patients are primed to develop AKI when exposed to additional ‘hits’, such as sepsis. Treatment options once HRS develops are limited, and prognosis remains poor. Thus, prevention of renal impairment in high risk patients should be a key focus of care.
Table 2 | Recommended measures to prevent the development of acute kidney injury in high-risk cirrhotic patients with portal hypertension. Preventative measures for particular at risk groups are presented At risk group
Preventative measures
Ascites
Use minimal doses of diuretic therapy necessary Early recognition of deteriorating renal function and modification of diuretic therapy Avoid nonsteroidal anti-inflammatory drugs and angiotensin converting enzyme (ACE) inhibitors/angiotensin II receptor antagonists Consider long-term antibiotic prophylaxis if low ascitic protein levels (
Review article: advances in the management of patients with cirrhosis and portal hypertension-related renal dysfunction J. A. Leithead*,†, P. C. Hayes‡ & J. W. Ferguson*
*Liver Unit, Queen Elizabeth Hospital, Birmingham, UK. † NIHR Biomedical Research Unit and Centre for Liver Research, University of Birmingham, Birmingham, UK. ‡ Department of Hepatology, Royal Infirmary of Edinburgh, Edinburgh, UK.
Correspondence to: Dr J. A. Leithead, NIHR Biomedical Research Unit, Centre for Liver Research, Institute of Biomedical Research (5th floor), University of Birmingham, Edgbaston, Birmingham, UK. E-mail: [email protected]
Publication data Submitted 30 September 2013 First decision 12 October 2013 Resubmitted 30 December 2013 Accepted 19 January 2014 EV Pub Online 13 February 2014 This commissioned review article was subject to full peer-review and the authors received an honorarium from Wiley, on behalf of AP&T.
SUMMARY Background In cirrhosis, portal hypertension is associated with a spectrum of renal dysfunction that has significant implications for morbidity and mortality. Aim To discuss recent progress in the patho-physiological mechanisms and therapeutic options for portal hypertension-related renal dysfunction. Methods A literature search using Pubmed was performed. Results Portal hypertension-related renal dysfunction occurs in the setting of marked neuro-humoral and circulatory derangement. A systemic inflammatory response is a pathogenetic factor in advanced disease. Such physiological changes render the individual vulnerable to further deterioration of renal function. Patients are primed to develop acute kidney injury when exposed to additional ‘hits’, such as sepsis. Recent progress has been made regarding our understanding of the aetiopathogenesis. However, treatment options once hepatorenal syndrome develops are limited, and prognosis remains poor. Various strategies to prevent acute kidney injury are suggested. Conclusion Prevention of acute kidney injury in high risk patients with cirrhosis and portal hypertension-related renal dysfunction should be a clinical priority. Aliment Pharmacol Ther 2014; 39: 699–711
ª 2014 John Wiley & Sons Ltd doi:10.1111/apt.12653
699
J. A. Leithead et al. INTRODUCTION Renal dysfunction is a common complication of chronic liver disease (CLD) and has significant implications for patient morbidity and mortality. Portal hypertension is accompanied by progressive systemic circulatory derangement and a spectrum of renal dysfunction that evolves in parallel with advancing disease. This portal hypertension-related renal dysfunction manifests as tubular dysfunction in early compensated cirrhosis and, at the extreme end, hepatorenal syndrome (HRS) and renal failure.1 Importantly, if exposed to additional ‘hits’, the physiological changes of portal hypertension render the CLD patient vulnerable to further deterioration of renal function. Patients are primed to develop acute kidney injury (AKI), and when presenting with triggers such as sepsis and gastro-intestinal haemorrhage, or when undergoing large volume paracentesis, prevention of renal impairment should be a key focus of care.2, 3 Besides portal hypertension-related renal dysfunction, patients with CLD have a relatively high frequency of intrinsic chronic kidney disease reflecting the prevalence of comorbidities such as diabetes mellitus and hypertension, and glomerulonephritides associated with alcoholic liver disease, hepatitis B and hepatitis C.4, 5 The aetiology of renal dysfunction in this setting has important ramifications for prognosis. In portal hypertension-related renal dysfunction, the glomerular filtration rate (GFR) only falls appreciably once liver disease is advanced and circulatory derangement severe.1 Consequently, the 3-month probability of survival for hospitalised patients with HRS is only 15% compared with 73% for those with parenchymal nephropathy.2 In this review, we discuss recent progress in our understanding of the patho-physiological mechanisms and therapeutic options for portal hypertension-related renal dysfunction. Particular attention will be paid to the prevention of AKI and HRS in the primed patient. METHODS A literature search was conducted in 2013 on Pub-MED using the following search terms: HRS, liver cirrhosis, portal hypertension, renal dysfunction, AKI. After considering the titles and abstracts, the full-text articles that appeared to be relevant were examined in detail. The references of retrieved articles were also reviewed, and additional relevant articles were evaluated. The most scientifically robust and pertinent manuscripts were referenced in this comprehensive review. 700
Definitions and spectrum of portal hypertensionrelated renal dysfunction HRS is characterised by renal vasoconstriction and reduced GFR, and is a complication of advanced cirrhosis.1, 6 Central to the diagnostic criteria is the exclusion of alternative causes of renal dysfunction such as nephrotoxic drugs, diuretic-induced hypovolaemia and renal parenchymal disease (Table 1). The syndrome is divided into two distinct clinical subtypes. Type 1 HRS is defined as a doubling of the serum creatinine to a level >226 lmol/L (2.5 mg/dL) within a 2-week period, and has a clinical picture of profound circulatory dysfunction, intense sodium and water retention, and dilutional hyponatraemia.6–8 Median survival time without treatment has been estimated at 1.7 weeks, with a probability of survival of 25% at 1 month.7 On the other hand, type 2 HRS defines more modest renal impairment, as indicated by a serum creatinine >133 lmol/L (1.5 mg/dL).6 Renal impairment progresses slowly, the predominant clinical feature is of refractory ascites and the median survival time is reported to be 6 months.6, 9 Although the diagnostic criteria for HRS are widely employed, their dependence on a rigid cut-off value of serum creatinine has rendered them controversial.10, 11 Creatinine production is proportional to muscle mass, and is greater in men than women, in younger than older individuals and in black people than in Caucasians despite similar GFR. In cirrhosis specifically, reduced creatine production by the liver, muscle wasting and increased renal tubular secretion contribute to a falsely low serum creatinine.12 Furthermore, the poor prognosis of patients once the HRS criteria are met is consistent with extreme renal dysfunction that is rarely reversible even with optimal treatment. Earlier Table 1 | New diagnostic criteria for hepatorenal syndrome in patients with cirrhosis6 Diagnostic criteria for hepatorenal syndrome Cirrhosis with ascites Serum creatinine >133 lmol/L (1.5 mg/dL) No improvement of serum creatinine (decrease to level ≤133 lmol/L) after at least 2 days with diuretic withdrawal and volume expansion with albumin. The recommended dose of albumin is 1 g/kg of body weight per day up to a maximum of 100 g/day Absence of shock No current or recent treatment with nephrotoxic drugs Absence of parenchymal kidney disease as indicated by proteinuria >500 mg/day, microhaematuria (>50 red blood cells per high power field) and/or abnormal renal ultrasonography Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
Review: management of portal hypertension-related renal dysfunction identification of renal impairment and intervention is likely to be beneficial. Recent guidelines by Kidney Disease: Improving Global Outcomes (KDIGO) have advised that AKI is defined as an increase in serum creatinine ≥26.5 lmol/L (0.3 mg/dL) in 48 h, or a rise to ≥1.5 times baseline within 7 days.13 This classification has been shown to predict survival in patients with cirrhosis.14 Yet, opinion remains divided amongst Hepatologists as to the most appropriate marker of renal impairment.10, 11 We advocate that the change in serum creatinine from baseline is the ideal tool for real-life practise, and any observed rise is an early warning requiring action to prevent the development of HRS.10 In a patient hospitalised with acute alcoholic hepatitis, for example, an increase in serum creatinine from 20 to 40 lmol/L may be indicative of clinically relevant renal dysfunction irrespective of the absolute value remaining below the normal range of many laboratories. It is important to emphasise that the renal dysfunction of portal hypertension is a spectrum, not a black and white phenomenon, and that the fall in GFR that heralds HRS is a late sequelae. From early cirrhosis, there is evidence of increased tubular sodium reabsorption that is followed chronologically by reduced free water clearance and renal haemodynamic dysfunction.1 Non-ascitic cirrhotic patients have altered renal sodium metabolism, with decreased natriuresis in the standing position and following a saline load.15, 16 With disease progression, sodium retention becomes overt, there is a positive sodium balance and ascites develop. Once refractory ascites presents there is profound sodium retention, with patients excreting less than 10 mmol of sodium per day.17 Impaired free water clearance is a feature of decompensated cirrhosis, occurring after sodium retention, and evidenced by dilutional hyponatraemia despite the increased total body sodium.1 Fifty per cent of patients with ascites demonstrate hyponatraemia and the prevalence is greater in those with refractory ascites.18 Ascites and hyponatraemia should be considered clinical manifestations of pre-HRS portal hypertension-related renal dysfunction. Cirrhotic patients with these complications are at increased risk of HRS, and both are prognostic markers independent of serum creatinine.7, 19, 20
Incidence of portal hypertension-related renal dysfunction In view of the limitations of the diagnostic criteria and obvious practical difficulties of long-term follow-up Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
studies in patients with early cirrhosis, the incidence of portal hypertension-related renal dysfunction remains poorly described. In a population presenting with variceal haemorrhage, the 5-year probability of the development of ascites was reported to be 73%, and of the development of HRS was 21%.21 Gines et al. observed that in patients with ascites but maintained GFR, the probability of HRS occurrence was 18% and 39% by 1and 5-years respectively.7 In a more contemporary study, the 1-year probability of any form of functional renal impairment (including ‘pre-renal’, infection and HRS) after the first episode of ascites was 23.6%.22 When patients with refractory ascites were observed for a mean duration of 18 months, approximately half developed type 2 HRS.23
Pathophysiology of portal hypertension-related renal dysfunction An understanding of the underlying pathophysiology of portal hypertension-related renal dysfunction is crucial to appreciate the vulnerability of renal function in CLD patients. The driving force is profound systemic circulatory derangement, the so called ‘splanchnic steal phenomenon’.24 Pooling of blood within a dilated splanchnic vascular bed is accompanied by the activation of multiple homoeostatic neurohumoral systems, and secondary peripheral vasoconstriction with reduced tissue perfusion ensues. The resultant effects within the kidney are the tubular and haemodynamic changes that typify the renal dysfunction in this setting.25, 26 Nitric oxide appears to be of central importance in both instigation and potentiation of the splanchnic vasodilation.27–30 In early portal hypertension, vascular stretch within the intestinal microcirculation may trigger endothelial nitric oxide synthase (eNOS) upregulation, and it is speculated that this is the first step in the neurohumoral and circulatory cascade.31, 32 In later disease shear stress, endotoxaemia and inflammatory mediators contribute via increased eNOS as well as increased inducible nitric oxide synthase (iNOS) activity.33–42 With progressive splanchnic vasodilatation and intravascular underfilling, high pressure baroreceptors activate the renin-angiotensin-aldosterone system (RAAS) and the sympathetic nervous system (SNS), and stimulate nonosmotic hypersecretion of vasopressin.43–46 These systems act to maintain arterial blood pressure, lead to compensatory peripheral vasoconstriction, and are considered the principal mediators of the renal tubular and renal haemodynamic dysfunction.1, 8, 47–51 With progressive splanchnic vasodilatation, the circulation becomes 701
J. A. Leithead et al. increasingly hyperdynamic. A relatively low cardiac output state further exacerbates the arterial underfilling.52, 53 There may be a cirrhotic cardiomyopathy characterised by myocardial hypertrophy, diastolic dysfunction, and an abnormal inotropic and chronotropic response.54, 55 The clinical picture is one of reduced total systemic vascular resistance, increased heart rate and cardiac output, expanded blood volume, and eventually arterial hypotension.56–58 Exaggerated renal tubular sodium reabsorption is observed when renal blood flow and GFR are still within normal limits. At this point, sodium retention is solely at the tubular level.25, 59 Later, once the GFR drops, reduced filtered sodium contributes and exacerbates sodium retention.1 Patients with ascites frequently have increased circulating levels of plasma renin activity (PRA) and aldosterone, and the administration of the angiotensin II receptor antagonist losartan and the aldosterone antagonist spironolactone reduces sodium retention.60–63 Consequently, the RAAS is thought to play a key role in the increased tubular sodium reabsorption.60 Moderate SNS activity that may be insufficient to alter renal haemodynamics similarly increases sodium retention throughout the nephron via a direct tubular effect. Therefore, the SNS is thought to be an additional important stimulus for sodium reabsorption.64, 65 Vasopressin or anti-diuretic hormone is probably the chief mediator of impaired free water clearance.1 Vasopressin stimulates aquaporin-2 channel insertion into the epithelial cell membrane increasing the permeability of the distal nephron to water.66 Urinary aquaporin-2 excretion is markedly increased in patients with cirrhosis, rises in parallel with Child-Pugh class and demonstrates a significant negative correlation with spontaneous free water clearance.66 Moreover, vasopressin 2 receptor antagonists have been shown to have an aquaretic effect and to increase serum sodium in cirrhotic patients with hyponatraemia and ascites.67, 68 The hallmark of advanced portal hypertension-related renal dysfunction is renal haemodynamic dysfunction.1 With advancing circulatory derangement and neuro-humoral activation, there is progressive renal vasoconstriction and a fall in total renal blood flow. Intra-renal compensatory mechanisms initially maintain the GFR at normal or even elevated levels.69–71 With progressive arterial underfilling, the accompanying activity of the vasoconstricting systems becomes extreme, and the renal production of vasodilators eventually falls.72 Renal ischaemia stimulates additional intra-renal secretion of endothelin-1 and increased SNS 702
activity. A self-perpetuating cycle develops within the kidney.73–75 The balance is tipped in favour of reduced GFR, and HRS develops.
The role of the systemic inflammatory response in portal hypertension-related renal dysfunction The systemic inflammatory response is increasingly recognised as a pathogenetic factor in the circulatory dysfunction of advanced cirrhosis, in the absence of overt infection. Patients with Child-Pugh Class C cirrhosis demonstrate an increased frequency of bacterial translocation of enteric organisms to mesenteric lymph nodes.76 Bacterial DNA is present in the blood of approximately 40% of non-infected patients with ascites.77–79 In addition, the plasma levels of lipopolysaccharide binding protein (LBP), which is considered a better marker of transient endotoxaemia, given its longer half-life, are also elevated.80 In rats with cirrhosis, bacterial translocation is associated with nitric oxide overproduction in the mesenteric vasculature, which appears to aggravate arterial vasodilatation.37 Humans with advanced cirrhosis display increased mesenteric lymph node tumour necrosis factor–alpha (TNF-a) expression.81 Moreover, ascitic patients with high LBP levels have a more pronounced circulatory dysfunction.80 The exaggerated immunohaemodynamic derangement is reversed by the administration of norfloxacin, an effect that is not reproduced in cirrhotics with ascites and normal LBP levels.80, 82 Tying this together, in decompensated cirrhosis, bacterial translocation and the secondary systemic inflammatory response may result in increased nitric oxide generation and exacerbate the pre-existing portal hypertensive syndrome. Indeed, long-term prophylactic antibiotics in patients with advanced CLD reduce the incidence of HRS and improve survival, independent of the prevention of infection.83 The additional ‘hit’ theory Hospitalised cirrhotic patients have a high rate of renal dysfunction. Using nonstandardised definitions, the incidence of renal dysfunction has been reported to be approximately 25%.84 At time of admission to the Intensive Care Unit, the prevalence may be at least twice that.85 The majority of episodes of renal dysfunction are in the context of infection, or hypovolaemia secondary to bleeding, diuretics or gastro-intestinal fluid losses.2 Although the short-term prognosis of patients with such renal dysfunction is superior to patients with HRS, the 90-day survival remains only 31–46%.2 Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
Review: management of portal hypertension-related renal dysfunction The systemic haemodynamic and renal derangement of portal hypertension render the CLD patient vulnerable to AKI when further insults occur (Figure 1). The superimposed ‘hit’ most studied is sepsis. In addition to the baseline low grade systemic inflammatory response demonstrated by many patients, there is an exaggerated circulatory response to infection. Cirrhotics have a greater and more prolonged rise in circulating inflammatory mediators compared with patients without liver disease, and more marked haemodynamic changes.86, 87 Acute renal injury is common with a reported incidence as high as 45% in bacteraemia of unknown origin, 33% in spontaneous bacterial peritonitis and 29% in pneumonia.88, 89 In the Intensive Care Unit, cirrhosis is associated with a 2 times increased risk of renal failure.90 Likewise, cirrhotic patients hospitalised with gastrointestinal bleeding are more likely to develop renal injury when compared with noncirrhotic patients.91
When large volume paracentesis is performed without the administration of plasma expanders, 75% of patients will develop Paracentesis-Induced Circulatory Dysfunction (PICD), a condition characterised by marked activation of the RAAS and often associated with renal dysfunction.3 Mechanical decompression, shear stress-related eNOS upregulation within the splanchnic vasculature and ascites re-accumulation exacerbate the baseline neuro-humoral and circulatory derangement.3, 92, 93 Within 6 days of the procedure, patients demonstrate a significant increase in serum creatinine and fall in serum sodium levels.3, 92 Thereafter, there is faster re-accumulation of ascites, an increased diuretic requirement and impaired prognosis.3 Similar long-term consequences are observed in patients who have suffered a septic episode. After resolution of infection, cirrhotics with renal failure demonstrate a further deterioration in haemodynamics, with even greater over-activity of the
Diuretic controlled ascites
Drugs (NSAIDs, ACE-I)
Hyponatraemia
Large volume paracentesis
Infection Refractory ascites Gastrointestinal haemorrhage
Type 1 HRS
Type 2 HRS
Figure 1 | The additional ‘hit’ theory of portal hypertension-related renal dysfunction: patients with cirrhosis and portal hypertension are primed to develop acute kidney injury, and additional ‘hits’ may precipitate an irreversible deterioration in renal function with long-lasting implications. Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
703
J. A. Leithead et al. RAAS and SNS, aggravation of portal hypertension and increased mortality.53, 89 Therefore, it is speculated that additional ‘hits’ that would have transient effects only in other patient populations may precipitate an irreversible deterioration in cirrhotics with lasting prognostic implications.
Treatment of portal hypertension-related renal dysfunction Despite progress in our understanding of the mechanisms underlying the spectrum of portal hypertension-related renal dysfunction, treatment options remain limited. In the earlier phase, management is supportive, aiming to minimise patient morbidity as a result of sodium and water retention. In later phases, specific therapies are employed to increase the effective circulating volume, renal perfusion and GFR, both directly and via a reduction in endogenous vasoactive mediators. The onset of type 1 HRS should prompt a search for the precipitating event. In particular, infection should be considered and there should be a low threshold for antibiotics even in culture-negative patients. Intravascular volume status should be assessed and corrected as appropriate. Type 2 HRS signifies end-stage disease and therapy primarily focuses on management of refractory ascites and hyponatraemia. Nonsteroidal anti-inflammatory drugs, and angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor antagonists, are avoided because of the reno-protective effect of the prostaglandins and angiotensin II, respectively. Management of sodium and water retention. In patients with clinically apparent ascites, a salt-restricted diet (80 to 120 mmol sodium per day) is recommended and escalating doses of diuretics are employed.94 The simultaneous administration of an aldosterone antagonist and loop diuretic from the beginning of treatment may be more efficacious than sequential diuretic treatment alone.95 The addition of clonidine, a centrally acting a2-agonist with sympatholytic activity, to patients with increased SNS activity is associated with reduced diuretic requirements and slower ascites re-accumulation.96 Oral fluid restriction is poorly adhered to and seldom effective.94 Large volume paracentesis is first-line therapy for patients with grade 3 ascites prior to initiation of maintenance diuretics, or may be the main stay of treatment in refractory ascites.94 AKI is not a contraindication to paracentesis. In this setting, large volume paracentesis with albumin administration is accompanied by an increase in renal perfusion pressure, creatinine clearance 704
and fractional excretion of sodium, possibly as a result of the fall in intra-abdominal pressure.97 A meta-analysis has suggested that vasopressin 2 receptor antagonists reduce the time to first paracentesis, and increase the serum sodium. However, the authors concluded that the cost and modest benefit did not justify their routine clinical use.98
Therapies that increase the effective circulating volume, renal perfusion and glomerular filtration rate. Terlipressin is a vasopressin analogue with predominant vasopressin 1, although some vasopressin 2 receptor effects.99 Terlipressin is mainly thought to improve renal perfusion in portal hypertension-related renal dysfunction by reducing splanchnic vasodilatation and hence splanchnic steal.24 In patients with ascites without HRS, terlipressin resulted in improved systemic haemodynamics, a fall in circulating neurohumoral mediators, reduced renal arterial resistive index, improved GFR and increased urinary sodium excretion.100, 101 When ascites is nonrefractory the increase in GFR occurs in association with an increase in filtration fraction but no change in total renal blood flow, supporting the concept that some of the benefit may be mediated by post-glomerular vasoconstriction.101 Randomised controlled trials have demonstrated that terlipressin reverses type 1 HRS and results in a small reduction in short-term mortality.102 Nevertheless, prognosis remains poor with a reported 6-month transplant-free survival of only 13%.103, 104 In type 2 HRS, renal failure invariably recurs following treatment withdrawal.6 However, terlipressin may be useful as a bridge to liver transplantation in this cohort.105 Noradrenaline appears to be as effective as terlipressin for treatment of type 1 HRS, but has practical disadvantages and is generally used only in countries where terlipressin is not available.106–108 Intravenous human albumin has multiple beneficial properties in patients with portal hypertension. In addition to the volume loading provided by less concentrated preparations and increase in oncotic pressure, albumin binds endotoxin, has anti-inflammatory and antioxidant effects and, consequently, may impact on endothelial and cardiac dysfunction.109 Administration of albumin results in an increase in systemic vascular resistance and improvement in cardiac function, which is not seen following hydroxethyl starch.110, 111 In patients with ascites receiving diuretics, intravenous albumin therapy is associated with a faster rate of ascites mobilisation and a lower probability of ascites Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
Review: management of portal hypertension-related renal dysfunction reaccumulation.112 Furthermore, a daily albumin infusion has been shown to be superior to fluid restriction for correcting severe hyponatraemia.113 In those with HRS, the addition of albumin to terlipressin resulted in a three times increased occurrence of reversal of renal dysfunction.114 The main predictor of reversal of type 1 HRS with terlipressin is a lower baseline serum creatinine.104, 115 Fifty per cent of patients with a serum creatinine less than 265 lmol/L (3.5 mg/dL) demonstrate reversal, compared to 11% of patients with a serum creatinine greater than 442 lmol/L (5 mg/dL).115 This emphasises the importance of earlier intervention in such patients and suggests survival benefit with even lesser degrees of renal dysfunction. Our local practice is to administer 40–60 grams of albumin per day to patients with at least a 50% increase in serum creatinine from baseline (stage 1 AKI).13 Terlipressin is commenced in patients with no response after 24–48 h, in all patients with at least a two times increased serum creatinine (stage 2 AKI), and/or those patients fulfilling the diagnostic criteria for type 1 HRS.13
Nonpharmacological therapies. Transjugular intra-hepatic portosystemic shunt (TIPSS) has a role in the management of select patients with portal hypertension-related renal dysfunction. Although the immediate effect is an exacerbation of the hyperdynamic state as evidenced by an increase in cardiac output and reduction in systemic vascular resistance, in longer term, the fall in portal pressure is associated with improved circulatory and neuro-humoural derangement.116, 117 Patients demonstrate a marked and sustained fall in PRA, and plasma aldosterone and noradrenaline concentrations.118–121 In those with refractory ascites, and in HRS, TIPSS results in increased total renal blood flow and GFR, and less sodium retention.118–121 Meta-analyses of randomised controlled trials confirm that TIPSS in refractory ascites is associated with a reduced recurrence of ascites when compared with large volume paracentesis.122, 123 In this group, there may be a small but statistically significant positive effect on survival.124 Nevertheless, patients with poor synthetic function are unlikely to benefit and a serum bilirubin greater than 85 lmol/L is considered an absolute contra-indication.94, 124 The applicability of TIPSS once HRS develops is likely to be limited, given the high prevalence of jaundice and encephalopathy.94 Haemofiltration may be useful in some patients with AKI when there is a reversible precipitating factor such as infection.125 There remains a paucity of evidence to Aliment Pharmacol Ther 2014; 39: 699-711 ª 2014 John Wiley & Sons Ltd
support this recommendation. However, reversibility of type 1 HRS has been strongly correlated with resolution of infection.126 Moreover, in a single observational study, 30-day survival was reported to be 50% if renal replacement therapy was the only organ support.127 When mechanical ventilation was also instituted, no patients survived.127 Longer term prognosis is nevertheless extremely poor.127 Consequently, the Acute Dialysis Quality Initiative (ADQI) Group has advised that renal support is only helpful if the patient is a candidate for liver transplantation.125 Equally, in type 2 HRS, renal replacement therapy is not appropriate unless as a bridge to transplant. Albumin dialysis with the molecular adsorbent recirculating system (MARS) has been shown in a single small randomised controlled trial to improve renal function in type 1 HRS when compared with haemodiafiltration alone.128 Yet, whether the effects are sustained is unknown and MARS remains an experimental therapy only at present.94 Liver transplantation is the definitive treatment for portal hypertension-related renal dysfunction, offering a clear survival advantage.129 Most patients with HRS demonstrate an improvement in renal function, although post-transplant chronic kidney disease is more common in this group.129 Duration of reduced GFR is important in influencing outcomes, probably reflecting the development of secondary acute tubular necrosis.130, 131 Combined liver kidney transplantation is recommended for all patients with stage 3 AKI or a GFR of less than 25– 35 mL/min for more than 4 weeks.132 This population is considered to gain survival benefit from combined liver kidney transplantation over liver transplantation alone, although the supporting literature is flawed by the heterogeneity of the cohorts studied.132, 133 Terlipressin may have a role in improving survival to transplantation of patients with type 1 and type 2 HRS, and may reduce post-transplant renal disease and mortality.105 However, pre-transplant optimisation of renal function could disadvantage patients by prolonging the wait-list time. Current organ allocation policies are based on prognostic scores that are heavily weighted by serum creatinine. It has been suggested therefore that HRS is considered a priority criterion for organ allocation, and accompanied by, for example, MELD exception points.105
The prevention of hepatorenal syndrome Given that patients with CLD are primed to develop AKI even when the GFR is preserved, and the poor prognosis once HRS develops, prevention of HRS in high-risk individuals should be a clinical priority (Table 2). 705
J. A. Leithead et al. Early recognition of infection and prompt initiation of antibiotic therapy are imperative in patients presenting with decompensated cirrhosis. In those with spontaneous bacterial peritonitis, and possibly other infections, the addition of intravenous albumin results in a reduced incidence of renal impairment and death.110, 134, 135 Antibiotic prophylaxis in gastrointestinal bleeding has been shown to prevent bacterial infection and improve survival.136 Furthermore, long-term antibiotic administration to prevent recurrence of spontaneous bacterial peritonitis has also been advocated, although it remains controversial because of the possibility of resistant bacteria.137 Other particular subgroups in which the benefits of antibiotic prophylaxis may outweigh the risks are those with low ascitic fluid protein concentrations and high serum bilirubin levels.83, 137, 138 Pentoxifylline has been associated with a reduced risk of HRS and mortality in patients with acute alcoholic hepatitis, and in advanced cirrhosis resulted in a lower probability of bacterial infection and renal impairment despite no effect on survival.139, 140 The proposed mechanism of action is downregulation of TNF-a with dampening of the immuno-haemodynamic derangement.141 Moreover, pentoxifylline may reduce intestinal bacterial overgrowth and bacterial translocation directly.141 Nonselective beta-blockers may have a similar effect.142 It is noteworthy, however, that beta-blockade in patients with refractory ascites has been linked with increased
mortality, possibly via a negative effect on cardiac function.143, 144 Nevertheless, the suggestion that nonselective beta-blockers should be contraindicated in these patients requires corroboration, and has been widely challenged by the Hepatology community. In large volume paracentesis, intravenous albumin reduces the secondary systemic and renal dysfunction, with beneficial effects on long-term prognosis.145, 146 The administration of a vasoconstrictor (terlipressin or noradrenaline) may be equally advantageous to the administration of albumin.147, 148 Whether the combination of both has additional positive effects on preventing PICD is yet to be assessed.
CONCLUSION The spectrum of portal hypertension-related renal dysfunction first becomes evident in early, compensated cirrhosis and evolves in parallel with advancing disease. Initial renal tubular sodium retention progresses to impaired free water clearance. The GFR only falls appreciably when the circulatory derangement becomes extreme. The physiological changes of portal hypertension render the individual vulnerable to further deterioration of renal function. Patients are primed to develop AKI when exposed to additional ‘hits’, such as sepsis. Treatment options once HRS develops are limited, and prognosis remains poor. Thus, prevention of renal impairment in high risk patients should be a key focus of care.
Table 2 | Recommended measures to prevent the development of acute kidney injury in high-risk cirrhotic patients with portal hypertension. Preventative measures for particular at risk groups are presented At risk group
Preventative measures
Ascites
Use minimal doses of diuretic therapy necessary Early recognition of deteriorating renal function and modification of diuretic therapy Avoid nonsteroidal anti-inflammatory drugs and angiotensin converting enzyme (ACE) inhibitors/angiotensin II receptor antagonists Consider long-term antibiotic prophylaxis if low ascitic protein levels (
