Since the first liver transplant in the early 1960s, great progress has been made in both the surgical and medical management of liver transplant recipients. Major strides have been made in overcoming obstacles such as acute and chronic rejection, and once dreaded infectious complications such as cytomegalovirus and hepatitis B infections are largely manageable. The associated surgical procedures have become almost routine at many centers because of advances in surgical techniques, organ preservation, and anesthetic management. What most significantly affects the long-term outcomes of liver transplant recipients are the medical complications of liver transplantation, which include recurrent hepatitis C,1 diabetes mellitus,2 malignancies,3 and renal dysfunction.4
The development of renal dysfunction before or after liver transplantation remains a complicated, multifaceted, and critical issue that adversely affects a host of outcomes, which range from increased costs of care to inferior grafts and decreased patient survival. The treatment of specific renal diseases in this setting varies with the etiology and is not addressed in this article. Rather, the landscape of renal dysfunction in the setting of liver transplantation is evaluated in this review of the literature, which focuses on the following critical issues: (1) the measurement and definition of renal dysfunction, (2) the evaluation of renal dysfunction, (3) the prevalence and etiology of renal dysfunction, (4) posttransplant immunosuppression, and (5) the allocation of kidneys for transplantation.
It is clear that serum creatinine (SCr)–based estimates of renal function are not accurate in patients with cirrhosis, and a rise in SCr is often a late indicator of kidney injury.5, 6 Patients with cirrhosis are known to have low SCr levels, which are related to lower muscle mass, decreased production of creatine by the liver, and potentially increased tubular secretion of creatinine related to medications commonly prescribed in this setting.5, 7, 8 However, in clinical practice, SCr testing is widely available and relatively inexpensive. Despite the lack of a consistent correlation between SCr and the glomerular filtration rate (GFR) in the setting of cirrhosis, it is a key factor in calculating the Model for End-Stage Liver Disease (MELD) score. Formulas that use SCr to determine renal function, such as the Modification of Diet in Renal Disease and Cockcroft-Gault formulas, overestimate GFR in patients with cirrhosis.9, 10 The measurement of 24-hour creatinine clearance (CrCl) in patients with cirrhosis also overestimates GFR in comparison with inulin clearance, particularly in patients with lower GFRs.11 Measuring GFR on the basis of the clearance of exogenous markers such as inulin, iohexol, and iothalamate in a steady state is generally the most accurate way of assessing renal function.12-14 However, many patients with cirrhosis have actively fluctuating renal function, and this makes the interpretation of exogenous marker clearance difficult. Furthermore, in patients with ascites, edema, and pleural effusions, the clearance of exogenous markers can be altered by the abnormal distribution volume. Cystatin C is a small peptide produced by all nucleated cells, and it has many features of an ideal marker for kidney function in the setting of cirrhosis. It is freely filtered by the glomerulus, is not secreted by tubules, and is essentially unaffected by diet, muscle mass, or inflammation.15 Several studies have shown cystatin C to be superior to SCr-based methods in the setting of cirrhosis.16-20 Although cystatin C appears to be more accurate than SCr, it is also more expensive and is not universally available; this makes it less practical for daily monitoring.
Niemann et al.21 prospectively evaluated the utility of serum neutrophil gelatinase–associated lipocalin (NGAL), a marker of early kidney injury, as a surrogate for post–liver transplant acute kidney injury (AKI) in 59 patients at 2 institutions. In their study, 33% of liver transplant recipients developed AKI, which was defined according to the Risk, Injury, Failure, Loss, and End-Stage Kidney Disease (RIFLE) criteria as a doubling of the baseline SCr level within the first 48 hours after surgery. A single NGAL value (≥139 ng/mL) after liver reperfusion predicted AKI in all liver transplant recipients with pre–liver transplant SCr values < 1.5 mg/dL. The difference between the baseline and liver reperfusion NGAL values predicted AKI in all patients regardless of the pre–liver transplant SCr levels. Other biomarkers of kidney injury such as kidney injury molecule 1 are currently being studied, and they may allow earlier detection and prompter intervention.22 In the near future, it is anticipated that the utilization of a panel of predictive novel and traditional biomarkers (NGAL, kidney injury molecule 1, and others) will aid immensely in the earlier identification of renal dysfunction in the setting of cirrhosis.
No single, currently available method of assessing renal dysfunction in the setting of cirrhosis is sufficient or practical; therefore, a logical combination of approaches is required as outlined below.
Evaluating Renal Dysfunction in Patients With Cirrhosis
When patients with cirrhosis are being evaluated for renal dysfunction, it is reasonable to start the assessment with SCr (with an understanding of the limitations), urine analysis (UA; including microscopy), a quantitative assessment of urine protein, and renal ultrasound (US). If there is a change from the baseline or a strong suspicion that the SCr value represents a chronically low GFR, a single cystatin C measurement can be obtained (daily serial measures are currently impractical), or GFR can be assessed with an exogenous marker. This additional method of assessing renal function allows a more accurate understanding of the true GFR; it should be kept in mind that SCr tends to lead to overestimations of GFR in patients with cirrhosis. Any acute rise in SCr should be further evaluated in all patients. As for ongoing monitoring of renal function, SCr is appropriate in the setting of AKI in patients with cirrhosis. Electrolytes need to be followed on a regular basis because metabolic disturbances are common in these patients. In patients with cirrhosis and renal dysfunction, it is imperative to rule out hypovolemia as the etiology before further expensive and potentially harmful evaluations (Fig. 1).
Beyond attempts to assess GFR, a further evaluation is important when renal dysfunction is being assessed in the setting of cirrhosis. Renal US is helpful for ruling out obstructions, but it may also indicate the presence of a chronic process such as cortical thinning or small kidneys. Urinalysis is important in all patients with renal dysfunction. Proteinuria greater than 500 mg/24 hours and abnormal urine sediment are both suggestive of renal parenchymal disease. If renal parenchymal disease is suspected, kidney biopsy may be helpful in determining not only the etiology of dysfunction but also the chronicity. Furthermore, kidney biopsy may be helpful in determining whether or not a liver transplant candidate would be better served by a simultaneous liver-kidney (SLK) transplant. Importantly, the risks associated with kidney biopsy, such as the increased risk of bleeding in this coagulopathic patient population, need to be considered. Patients with cirrhosis are at increased risk for major complications due to coagulopathy. Overly aggressive attempts to reverse the coagulopathy associated with cirrhosis may result in volume overload and thus pulmonary edema. Percutaneous kidney biopsy was performed 44 times in a study evaluating the ways in which kidney biopsy could aid in kidney allocation in liver transplant candidates.23 In that study, there were 13 biopsy complications, 5 of which were severe enough to require intervention. Another approach that is often used when coagulopathies exist is transjugular kidney biopsy. Sam et al.24 reported 29 patients with cirrhosis who underwent transjugular kidney biopsy. Eight of these patients subsequently required blood transfusions, and 23% were found to have perirenal hematomas. Jouet et al.25 reported 70 patients with cirrhosis and clotting disorders who underwent transjugular kidney biopsy. Adequate tissue was obtained from 55 of the 70 patients. Persistent hematuria was reported in 4 cases, 2 cases required transfusions, and perirenal hematoma was reported in 4 cases. Both percutaneous kidney biopsy and transjugular kidney biopsy are important to consider when renal dysfunction is being evaluated in patients with cirrhosis and especially when SLK transplantation is being considered.
The evaluation of post–liver transplant renal dysfunction with kidney biopsy was assessed by O'Riordan et al.26 Major bleeding, which was defined as bleeding requiring a transfusion or embolization, occurred in 17% of the cases. Kidney biopsy in high-risk patients before or after liver transplantation requires a careful consideration of risks and benefits. If kidney biopsy is pursued, clear protocols should be established and strictly followed. In those patients deemed to be at highest risk, it is reasonable to proceed with transjugular kidney biopsy. However, the drawback of transjugular kidney biopsy is the higher probability of not obtaining adequate renal cortical sampling.
Defining Renal Dysfunction
Our understanding of the scope of renal dysfunction at the time of liver transplantation is often clouded by a lack of accepted definitions, as evidenced by the many studies that do not differentiate between AKI, chronic kidney disease (CKD), and AKI superimposed on CKD. Although it is often clinically obvious when renal dysfunction occurs, there is currently no consensus on the definition of AKI in the setting of liver transplantation, and this makes the existing literature difficult to compare. O'Riordan et al.27 proposed using the RIFLE criteria as a standard definition of perioperative AKI in the setting of liver transplantation (Table 1); however, this has not been widely accepted. Recently, a working group composed of representatives from the International Ascites Club and the Acute Dialysis Quality Initiative met to redefine the definitions of AKI and CKD in patients with cirrhosis29 (Table 2). This group proposed broadening the scope of renal dysfunction in the setting of cirrhosis to include both AKI and CKD not caused by hepatorenal syndrome (HRS). They defined AKI as an acute increase in SCr ≥ 50% from the baseline or a rise in SCr > 0.3 mg/dL in less than 48 hours. CKD was proposed to be defined as an estimated glomerular filtration rate (eGFR) < 60 mL/minute for more than 3 months [calculated with the 6-variable Modification of Diet in Renal Disease (MDRD6) formula]. Acute-on-chronic renal dysfunction was proposed to be a combination of these 2 definitions. This classification system is appealing because both acute kidney disease and CKD are considered and because CKD is defined according to the proposal from the National Kidney Foundation. However, the limitations of SCr-based measurements of renal function must be acknowledged.
Table 1. Comparison of the RIFLE and AKIN Classification Systems27, 28
1.5-fold increase in SCr or GFR decrease > 25%
UO < 0.5 mL/kg/hour for 6 hours
Increase in SCR ≥ 0.3 mg/dL or increase to ≥150%-200% of the baseline
2-fold increase in SCr or GFR decrease > 50%
UO < 0.5 mL/kg/hour for 12 hours
Increase in SCr to >200%-300% of the baseline
3-fold increase in SCr or SCr ≥ 4 mg/dL with an acute rise > 0.5 mg/dL or GFR decrease > 75%
UO < 0.3 mL/kg/hour for 24 hours or anuria for 12 hours
Increase in SCr to >300% of the baseline or SCr ≥ 4.0 mg/dL with an acute increase of at least 0.5 mg/dL
Persistent AKI (complete loss of renal function for >4 weeks)
End-Stage Kidney Disease
End-stage kidney disease for >3 months
Table 2. Working Party Proposal for a Revised Classification System for Renal Dysfunction in Patients With Cirrhosis
NOTE: This table has been adapted with permission from Gut.29
Rise in SCr ≥ 50% from the baseline or rise in SCr ≥ 0.3 mg/dL in <48 hours. HRS type 1 is a specific form of AKI.
GFR < 60 mL/minute for >3 months (calculated with the MDRD6 formula). HRS type 2 is a specific form of CKD.
Acute-on-chronic kidney disease
Rise in SCr ≥ 50% from the baseline or rise in SCr ≥ 0.3 mg/dL in <48 hours in a patient with cirrhosis whose GFR is <60 mL/minute for >3 month (calculated with the MDRD6 formula).
Prevalence of Renal Dysfunction at the Time of Liver Transplantation
Despite a lack of standardized definitions, renal dysfunction before liver transplantation is common and may be due to CKD, AKI, or their combination. Sharma et al.30 demonstrated that at the time of liver transplantation, 51% of their patients had an eGFR < 60 mL/minute, and 6.3% were receiving dialysis.
Nair et al.30 assessed renal dysfunction in nearly 20,000 patients undergoing liver transplantation between 1988 and 1996. Renal function was determined with the calculated CrCl value. At the time of liver transplantation, 67% of the patients had normal renal function (mean CrCl = 118 ± 50 mL/minute), and 33% of the patients were classified as having renal dysfunction. Twenty-two percent of the patients with renal dysfunction had mild renal dysfunction (mean CrCl = 56 ± 8.5 mL/minute), 8% had moderate renal dysfunction (mean CrCl = 30 ± 5.7 mL/minute), and 3% had severe renal dysfunction (mean CrCl = 14 ± 3.6 mL/minute). This study again did not differentiate between CKD and AKI. An analysis of Organ Procurement and Transplantation Network/United Network for Organ Sharing data showed that the proportion of liver transplant recipients with SCr values ≥ 2.0 mg/dL or on dialysis increased after 2002.37 A subsequent analysis of the Scientific Registry of Transplant Recipients showed that the proportion of liver transplant recipients with renal dysfunction, which was defined as an SCr level ≥ 1.5 mg/dL at the time of liver transplantation, increased from 26.1% in 2002 to 32.5% in 2005, but it remained essentially stable from 2005 to 2008.33 These observations are likely a result of the implementation of the MELD system, which is further discussed later. Overall, it is clear that renal dysfunction is grossly underrecognized in patients awaiting liver transplantation and likely affects well over half of these patients to some degree.
Etiology of Renal Dysfunction in Patients With Cirrhosis
The pathophysiology of renal dysfunction in many patients with cirrhosis is due to a reduction of the effective circulating volume related to systemic vasodilation, a sodium-avid state, and intrarenal vasoconstriction leading to what is called HRS (Fig. 2). HRS may be spontaneous or may be precipitated by factors leading to decreased renal perfusion, such as infections and large-volume paracentesis.34 There are 2 types of HRS. Type 1 is characterized by a rapid deterioration of renal function, whereas type 2 has a more indolent course. The diagnosis of HRS is common in patients with cirrhosis and ascites and occurs in roughly 18% of nonazotemic patients with cirrhosis at 1 year and in 39% at 5 years, with the greatest risk in patients with hyponatremia and high renin activity.33 HRS is commonly incriminated as the cause of renal dysfunction at the time of liver transplantation; however, the etiology of renal dysfunction in this setting is hardly limited to HRS. Fraley et al.35 found that 19% of liver transplant recipients developed pretransplant AKI, which was defined as a doubling of the SCr level over a period of 24 hours or the onset of oliguria requiring dialysis; prerenal azotemia was not defined as AKI. The most common cause was acute tubular necrosis (ATN), which was followed by HRS. Clearly, etiologies other than HRS should be investigated in patients presenting with AKI before liver transplantation because the prognosis and therapies differ considerably.
Perhaps the best evidence describing kidney pathology at the time of liver transplantation was published in a case series report of patients with hepatitis C.36 McGuire et al.36 described the spectrum of kidney histological findings in 30 patients with hepatitis C who underwent kidney biopsy at the time of liver engraftment. The most common histological finding was membranoproliferative glomerulonephritis (MPGN) type 1 (n = 12). Additionally, immunoglobulin A (IgA) nephropathy (n = 7) and mesangial glomerulonephritis (n = 6) were found. However, none of these patients were found to have cryoglobulins (common mediators of renal dysfunction in patients with hepatitis C) in the blood or kidneys. Many of these patients with documented immune complex glomerulonephritis had what were classified as normal SCr values, and this highlights the limitations of SCr in detecting renal dysfunction in this population.
Consequences of Renal Dysfunction Before Liver Transplantation
Preexisting renal dysfunction is associated with many adverse outcomes after liver transplantation, which include inferior short- and long-term patient survival,16, 35, 37-42 increased costs,36, 43 posttransplant sepsis and longer intensive care unit (ICU) stays,36, 44 and the need for dialysis.36, 45 Using the liver transplantation database of the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), Brown et al.43 reviewed survival and other outcomes of patients with renal dysfunction undergoing liver transplantation alone versus SLK transplantation. Renal dysfunction was defined as an SCr level ≥ 1.6 mg/dL or the need for dialysis. Renal dysfunction in patients with cirrhosis without the need for dialysis did not significantly affect survival except for those who underwent SLK transplantation. Patient and graft survival rates at 1 year were 65% and 60%, respectively, for patients on dialysis and 74% and 70%, respectively, for patients not on dialysis. Furthermore, patients requiring dialysis experienced a 30% increase in costs for renal dysfunction, a 75% increase in costs with the need for dialysis, and a 150% increase in costs with the need for SLK transplantation. In contrast, Gonwa et al.30 found that preoperative renal dysfunction was an independent risk factor for survival in liver transplant recipients but not in SLK transplant recipients. In comparison with patients with an SCr level < 1 mg/dL, patients with an SCr level > 2 mg/dL had a 58% greater risk of death, and patients on dialysis had a 77% greater risk of death.
The impact of pretransplant renal function on long-term patient survival after liver transplantation was studied by Gonwa et al.44 in a different set of 569 consecutive liver transplant recipients. The 5-year survival rate was 60% for patients with HRS and 68% for patients without HRS (P < 0.03). Furthermore, pretransplant renal dysfunction other than HRS had no impact on survival after liver transplantation. In contrast, the presence of HRS did not influence the results of Lafayette et al.,37 who reviewed the impact of preoperative renal function on the posttransplant course of 115 liver transplant recipients. AKI requiring dialysis occurred in 23% of the patients and was associated with prolonged ICU stays, more hospital-acquired infections, and increased costs. Most importantly, patients with AKI requiring dialysis were associated with a higher mortality rate than patients without AKI (46% ± 11% versus 9% ± 3%). Patients with a pretransplant SCr level > 1.0 mg/dL had longer ICU stays (18 ± 3 versus 10 ± 2 days) and a higher mortality rate (32% ± 1% versus 6% ± 1%) than patients with pretransplant SCr values ≤ 1.0 mg/dL. In fact, an elevated pretransplant SCr value was the strongest predictor of both the development of AKI requiring dialysis and subsequent death. Dialysis may be a stronger predictor of death than the MELD score in liver transplant recipients with preexisting renal dysfunction. Narayanan Menon et al.45 showed that patients requiring dialysis before liver transplantation had a 5-fold increased risk of death. It is unclear from their study what percentage of the patients had AKI, CKD, or AKI superimposed on CKD. In the group of patients surviving the first year after liver transplantation, pretransplant dialysis and the MELD score were no longer risk factors for death.
In all, renal dysfunction before liver transplantation appears to be associated with increased morbidity, costs, and mortality. The greater number of patients with any renal dysfunction, including patients who require dialysis at the time of liver transplantation, is likely a reflection of the use of SCr in the MELD score, which is the basis for liver allocation. It should be noted that in the development and validation of the MELD score, patients with established renal disease were excluded.47, 48 Furthermore, the MELD system does not differentiate between AKI and CKD. Still, it appears that renal dysfunction at the time of liver transplantation generally affects survival for patients after liver transplantation. Because liver transplantation is acutely lifesaving, the implementation of the MELD system may have resulted in an overall transition of mortality from the period before liver transplantation to the period after liver transplantation. Therefore, it has been proposed that a survival benefit–based allocation system replace the current urgency-based MELD allocation system.49
RENAL DYSFUNCTION AFTER LIVER TRANSPLANTATION
AKI After Liver Transplantation
Renal dysfunction after liver transplantation may be due to perioperative AKI, preexistent CKD, or CKD that develops after liver transplantation. AKI after liver transplantation is unfortunately a frequent complication with a reported incidence of 12% to 95%.4, 26, 28, 35, 39, 46, 50, 51 In a study by Fraley et al.,35 30% of the patients developed AKI, which was defined as a doubling of the SCr level over a period of 24 hours or as the onset of oliguria requiring dialysis. In their study, ATN was the cause of 70% of AKI cases, and it should be noted that prerenal azotemia was not defined as AKI. Zhu et al.51 applied a modified RIFLE/Acute Kidney Injury Network (AKIN) definition to look at the development of AKI in the postoperative period after liver transplantation, and they found that 60.1% of liver transplant recipients developed AKI. Clearly, AKI is common in the perioperative period after liver transplantation. Significant nonhemodynamic AKI probably complicates approximately half of liver transplants in the immediate posttransplant period. The predominant etiology of AKI after liver transplantation is ATN (Fig. 2).35
Iglesias et al.53 performed a retrospective cohort study of 916 patients enrolled in the NIDDK liver transplantation database to evaluate preexisting factors that lead to AKI after liver transplantation. Six hundred eighty-eight patients with AKI (defined with the AKIN definition28; see Table 1) were included in the analysis. Unexpectedly, a univariate analysis showed an inverse relationship between improvements in pre–liver transplant renal function and the development of AKI after liver transplantation. It was postulated that this may have been due to a dilutional decline in eGFR in the setting of pre–liver transplant AKI or a loss of lean body mass (which may have predisposed the patients to post–liver transplant AKI), or pre–liver transplant declines in eGFR may actually have protected the patients against post–liver transplant AKI through ischemia preconditioning; this phenomenon has been described in other organ systems, including the kidneys.54-56 A number of pre–liver transplant variables were actually associated with the development of post–liver transplant AKI: an increased body mass index (P < 0.001), a decreased serum sodium level (P = 0.0005), a decreased platelet count (P < 0.00001), an elevated partial thromboplastin time (P = 0.003), an elevated prothrombin time (P < 0.00001), and a decreased serum albumin level (P < 0.00007). These numerous factors were presumably reflections of the severity of the underlying liver disease. Cabezuelo et al.57 reviewed 184 liver transplant cases to determine the risk factors for the development of AKI after liver transplantation. Patients in their study were classified as having early-onset AKI (first week after liver transplantation) or late-onset AKI (second to fourth week after liver transplantation). In contrast to Iglesias et al., the main risk factors for early-onset AKI were pretransplant AKI [odds ratio (OR) = 10.2, P = 0.02] and grade II to IV liver graft dysfunction (OR = 5.6, P = 0.002); risk factors for late-onset AKI were reoperation (OR = 3.1, P = 0.01) and bacterial infections (OR = 2.9, P = 0.017).
It is unclear whether pre–liver transplant AKI is a predictor of post–liver transplant AKI. The observation that AKI before liver transplantation may actually be protective against posttransplant AKI is unexpected. Differences between studies may be related to differing definitions and etiologies of AKI. Liver-related factors such as poor synthetic function leading to transplantation and the degree of graft dysfunction appear to be consistent predictors of AKI after liver transplantation.
CKD After Liver Transplantation
Liver transplant recipients develop CKD after transplantation at higher rates than heart and lung transplant recipients, with a 5-year cumulative incidence of 22% (CKD is defined as an eGFR < 30 mL/minute/1.73 m2 persisting for more than 3 months, the initiation of dialysis, or listing for kidney transplantation).30, 58 In a pre–MELD era cohort study, Gonwa et al.59 defined CKD as an SCr level ≥ 2.5 mg/dL, and they found that the 5-year prevalence of CKD after liver transplantation was only 4.3%, but it was 18% 13 years after transplantation. However, pre–MELD era reports of CKD prevalence are highly variable because of the lack of a standardized definition. Additionally, liver transplant recipients typically have lower muscle mass than other patients with CKD; therefore, SCr almost certainly overestimates GFR (as discussed for pre–liver transplant patients).
The second pre–MELD era cohort study used data from the Scientific Registry of Transplant Recipients and included 36,849 adult liver recipients between 1990 and 2000.58 This study defined CKD as an eGFR < 30 mL/minute (the same definition used in the MELD era study by Sharma et al.30). This study found the pre–MELD era CKD prevalence to be 18% at 5 years and 26% at 10 years. With the same definition of CKD, the pre–MELD era and MELD era prevalences of CKD were similar despite a higher prevalence of pretransplant renal dysfunction in the MELD era. Potential reasons for the differences in the findings include the following: (1) the study by Sharma et al. was simply too small, (2) the follow-up time was too short (2.6 years), and (3) much of the preexisting renal dysfunction was hemodynamic in nature and did not represent true structural injury. More frequent use of calcineurin inhibitor (CNI)–sparing immunosuppression protocols may be responsible; however, the use of CNIs was not a predictor of post–liver transplant CKD in the study by Sharma et al.
In a pediatric population, Campbell et al.60 measured GFR in 117 liver transplant recipients with an average time of 7.6 ± 3.4 years from transplant. The authors found that renal dysfunction was present in 32% of these patients, and this confirmed that CKD after liver transplantation is also a critical complication for children.
The development of CKD after liver transplantation is common in adults and children, and the risk appears to compound over time. Ten years after transplantation, patients have an approximately 30% to 50% risk of developing CKD; this takes into account the inherent overestimation with SCr-based measurements and the available data on GFR measurements.
Etiology of Renal Dysfunction After Liver Transplantation
The etiology of CKD after liver transplantation or late AKI is varied. In a review of 2100 adults who underwent liver transplantation, approximately 3% had subsequent kidney biopsy with a mean posttransplant renal referral time of 5.3 ± 4.6 years.26 The etiology of renal dysfunction was found to be attributable to 1 or more of the following etiologies: apparent CNI toxicity (48%), hypertensive vascular changes (44%), MPGN (17%), IgA nephropathy (9%), diabetic nephropathy (9%), proliferative glomerulonephritis with crescents (4%), and ATN (4%). Fisher et al. found vascular obliteration, tubular atrophy, interstitial fibrosis, and glomerular sclerosis in 10 of 13 patients categorized as having CNI toxicity. Two patients were found to have hemolytic uremic syndrome that was thought to be cyclosporine-related. One patient was found to have IgA nephropathy.61
CKD after liver transplantation is often attributed to CNI toxicity (discussed later), but other etiologies are also important to consider (Fig. 2).
Consequences of Renal Dysfunction After Liver Transplantation
United Network for Organ Sharing data show 1-, 5-, and 10-year liver transplant recipient survival rates of 88%, 74%, and 60%, respectively.33 The NIDDK long-term follow-up study looked at risk factors for mortality in liver transplant recipients.62 The mean age of the recipients was 49.4 years, and hepatitis C infection was the most common cause of liver disease (24%). Recipient survival rates 1, 3, 5, and 10 years after transplantation were 87.0%, 78.6%, 74.9%, and 59.4% respectively. As for the cause of death, 63.3% were non–liver-related, and 6.8% were directly attributed to kidney failure. Renal dysfunction before or after liver transplantation modified the risk for overall death more than 1 year after transplantation [hazard ratio (HR) = 3.59] and for hepatic failure–associated death (HR = 5.1). Renal dysfunction before or after liver transplantation conveyed an HR for death of 2.66. Furthermore, the timing of renal dysfunction predicted the risk of mortality more than 1 year after liver transplantation. In comparison with normal pretransplant renal function, renal dysfunction developing less than 1 year after transplantation resulted in an HR for death of 2.41; renal dysfunction 1 to 5 years after transplantation resulted in an HR of 6.58; and the development of renal dysfunction more than 5 years after transplantation afforded an HR for death of 7.49.
The consequences of post–liver transplant AKI for mortality were evaluated by Zhu et al.52 The 28-day and 1-year mortality rates for liver transplant recipients who did not develop AKI were 0% and 3.9%, respectively. However, in those who developed AKI, the mortality rates were markedly increased to 15.5% and 25.9%, respectively.
As mentioned earlier, Sharma et al.30 evaluated the outcomes of 221 liver transplant recipients in the MELD era and found that a decline in eGFR over time was associated with a decline in survival after liver transplantation. Furthermore, post–liver transplant CKD (eGFR < 30 mL/minute) had an HR for death of 3.2 in comparison with an eGFR > 60 mL/minute (P = 0.02).
The development of renal dysfunction (both AKI and CKD) after liver transplantation appears to be associated with a profound mortality risk. It is unclear whether AKI after liver transplantation is the primary driver of worse mortality outcomes or is merely associative. There is a strong association between the development of CKD and increased mortality after liver transplantation, but as with AKI, causality is unclear. Regardless, renal dysfunction after liver transplantation is a marker of poor outcomes, and prevention should be a major focus of posttransplant care.
IMMUNOSUPPRESSION AND RENAL DYSFUNCTION AFTER LIVER TRANSPLANTATION
Improved early survival after liver transplantation has generated interest in factors leading to late CKD. Countless observations have incriminated CNI use with the development of renal dysfunction after liver transplantation. Fisher et al.40 examined 883 consecutive liver transplant recipients undergoing liver transplantation for the first time between 1982 and 1996 at a single center. They found that a higher daily cyclosporine dose at annual intervals was a risk factor for late-onset renal dysfunction (median daily dosage at 3 years: 7.2 versus 5.0 mg/kg, P = 0.02; median daily dosage at 5 years: 6.1 versus 4.0 mg/kg, P = 0.01), and so was a higher cumulative dose at 5 years (13.6 versus 10.0 g/kg, P = 0.02). However, the cumulative cyclosporine level at 5 years was not a risk factor for late-onset renal dysfunction in this study. Furthermore, the investigators could not determine that reducing the CNI dose improved renal function. Similarly, Sharma et al.30 were unable to show that CNI use was a predictor of post–liver transplant CKD.
The case for CNI use being a major cause of renal dysfunction after liver or other organ transplantation may be overestimated. Many of these recipients have preexisting renal dysfunction and are at higher risk for AKI and diabetes (particularly recipients with hepatitis C).63 A retrospective study of 688 patients in the first year after liver transplantation compared CNI use to the use of a mammalian target of rapamycin inhibitor (sirolimus) and found that there was no difference in the decline in renal function.64 Short of performing serial kidney biopsies in liver transplant recipients, the case for chronic CNI nephrotoxicity will remain unproven.
SIMULTANEOUS LIVER-KIDNEY TRANSPLANTATION
Because of the high prevalence of pre–liver transplant renal dysfunction and its apparent impact on the risk for post–liver transplant dialysis and survival, it is important to understand when one should proceed with SLK transplantation. The number of SLK transplants increased from 135 in 2001 to 439 in 2007, whereas patient survival and kidney graft survival declined.59, 62 Gonwa et al.32 showed that SLK transplantation does not necessarily provide a survival advantage in comparison with liver transplantation alone.31 Data from the United Network for Organ Sharing were analyzed and showed that there was no difference in 3-year survival between liver transplant–alone recipients with an SCr level > 2 mg/dL who were not on dialysis and similar SLK recipients (69.8% versus 69.9%, P = 0.18). This study also showed that almost 30% of SLK transplant recipients were not on dialysis at the time of transplantation. Furthermore, nearly 15% underwent SLK transplantation with an SCr level < 2 mg/dL. This study suggests that the long-term survival benefit of SLK transplantation is limited to patients who have an SCr level > 2 mg/dL and are started on dialysis before transplantation, and many patients probably undergo SLK transplantation without a survival benefit.
To evaluate the utility of SLK transplantation for patients with chronic primary disease of both organs, Ruiz et al.65 evaluated 98 patients over a 16-year period. Seventy-six SLK recipients were determined to have primary renal disease, and 22 were thought to have HRS. In their study, the liver graft survival rates were 70%, 65%, and 65% at 1, 3, and 5 years, respectively; the kidney graft survival rates were 76%, 72%, and 70%, respectively; and the recipient survival rates were 76%, 72%, and 70%, respectively. SLK transplantation did not result in a significant survival advantage for patients with HRS at 1 year (72% versus 66%, P = 0.88). However, the authors suggested that a subgroup of patients with HRS (those requiring more than 8 weeks of dialysis) would benefit from SLK transplantation. Importantly, there were only 8 recipients in this subgroup; however, data on these patients were not reported.
In trying to understand the etiology of renal dysfunction leading up to SLK, Gonwa et al.32 found that only 2% of patients carried a diagnosis of AKI or HRS, and 8.3% had previously undergone kidney transplantation. Surprisingly, 0.7% had no renal diagnosis yet still underwent SLK transplantation. There were also combined transplants for HRS and AKI patients, who presumably would recover at fairly high rates. Diabetic nephropathy (16.5%) and glomerular disease (14%) were also common causes of renal dysfunction in this study. The primary cause of renal dysfunction before SLK transplantation reported by Thuluvath et al.33 was not specified and was classified as other in 24% to 45% of cases (depending on the year). Glomerular disease (6.7%-17%), diabetes (11%-18.3%), polycystic kidney disease (4.4%-16%), and tubulointerstitial disease (10.3%-22%) were other common causes of renal dysfunction.
The usefulness of kidney histology in directing kidney allocation in liver transplant candidates was evaluated by Wadei et al.23 Percutaneous kidney biopsy was performed in 44 liver transplant candidates who were on dialysis (n = 7) or had renal dysfunction of an unclear etiology and an iothalamate GFR < 40 mL/minute (n = 37). Patients with ≥30% interstitial fibrosis, ≥40% global glomerulosclerosis, or diffuse glomerulonephritis were approved for SLK transplantation. In this study, GFRs, urine sodium levels, dialysis requirements, and kidney sizes were similar for the 27 liver transplant–alone candidates and the 17 SLK transplant candidates; these indices did not correlate with the biopsy diagnosis. After nearly 80 days, 16 patients underwent liver transplantation alone, and 8 underwent SLK transplantation. The 5 liver transplant–alone recipients on dialysis recovered some degree of kidney function after transplantation, and this resulted in similar SCr values for liver transplant–alone recipients and SLK recipients at the last follow-up.
The results from a consensus conference on SLK transplantation were published in 2008.66 The consensus was that SLK transplantation is appropriate for patients with end-stage kidney disease needing liver transplantation, patients with AKI (SCr > 2 mg/dL at any point) and presently on dialysis for 8 or more weeks, and patients with CKD and a kidney biopsy sample demonstrating >30% interstitial fibrosis or >30% global glomerulosclerosis. This group also recommended that a regional review board should determine appropriateness for SLK transplantation.
Allocating kidneys to SLK transplant candidates takes kidneys away from patients who are waiting for deceased donor kidney transplants while they are on dialysis. The benefit of allocating kidneys to SLK transplant candidates must be weighed against the loss of kidney grafts that would clearly benefit patients on dialysis. Furthermore, kidney graft survival differences between SLK transplant recipients and deceased donor kidney transplant recipients should also be considered. The decline in patient and kidney graft survival is almost certainly due to transplantation for sicker patients, which is an important issue in itself. Who should undergo SLK transplantation remains an unanswered question, and this can best be addressed with a well-designed randomized controlled trial. Presently, there is a lack of hard evidence to guide us, so it is important to approach this issue in a thoughtful and systematic way; an example is outlined in Fig. 3.
Renal dysfunction in the setting of liver transplantation is common and has an adverse impact on a multitude of outcomes, including short- and long-term graft and patient survival. However, in order to better understand the scope of this problem, both AKI and CKD in this setting must be clearly defined. Solid definitions have been outlined by the International Ascites Club/Acute Dialysis Quality Initiative work group. As the number of liver transplant candidates with renal dysfunction has increased, so has the number of SLK transplants. There needs to be a better understanding of outcomes for this population, and uniform criteria (such as those outlined in Fig. 3) need to be in place before kidney grafts are allocated to liver transplant recipients. Late CKD after liver transplantation has developed into a significant clinical problem despite improvements in the management of acute posttransplant complications. CNIs have been implicated in this late development and progression of CKD after liver transplantation; however, the relationship may not be as clear-cut as assumed. Resolving these critical issues that surround renal dysfunction in liver transplant recipients is essential to reducing costs of care, improving organ allocation, and reducing patient morbidity and mortality.
The authors acknowledge Matthew Romo, Pharm.D., for providing editorial support with funding from Bristol-Myers Squibb.