The development of kidney dysfunction after liver transplantation (LT) is associated with an increased risk of recipient morbidity and mortality.1-5 Kidney disease is frequently seen in patients with end stage liver disease, both before and after LT. Factors previously found to be associated with posttransplant kidney dysfunction include: the extent of pretransplant renal disease, the presence of hepatorenal syndrome (HRS), diabetes, hypertension, dyslipidemia, intraoperative hypovolemia, posttransplant sepsis, and immunosuppression-induced nephrotoxicity.2,3,6-8 Since the adoption of organ allocation by use of the model for end-stage liver disease (MELD) score, patients with greater kidney dysfunction are listed with a higher priority for transplantation. Simultaneously, a greater number of elderly patients are undergoing LT, so that an increasing number of patients with marginal or poor renal function are undergoing LT.9,10 The severity of the kidney disease, and the extent to which it is reversible, is difficult to determine in the complex liver failure patient.
There are several recent studies which examine the incidence, clinical outcomes, and risk factors for chronic kidney disease (CKD) in post-LT patients.2,3,6-14 There is considerable variation in the definition of kidney disease in published studies, making it difficult to quantitate the frequency and severity of renal dysfunction, and to compare study protocols and outcomes. The U.S. National Kidney Foundation has published guidelines requiring that the assessment of renal function be calculated through a mathematical formula, preferentially using the modification of diet in renal disease (MDRD) score (glomerular filtration rate [GFR] (mL/min per 1.73 m2) = 186 × (Pcr)−1.154 × (age)−0.203 × (0.742 if female) × (1.210 if African American).15 The serum creatinine alone is not considered to be an accurate indicator of renal disease and is not part of the guidelines, though it is 1 component of the MDRD equation. It has been suggested that the MDRD equation is the most accurate in assessing renal function in patients with liver disease and is more accurate than the serum creatinine, 24-hour creatinine clearance or calculation of the GFR using the Cockcroft and Gault calculation.16 A 2004 paper by Gonwa et al16 found this equation to most accurately reflect actual GFR as measured by I125-iothalamate clearance. This study examines the change in post-LT GFR calculated using the MDRD equation in a large number of consecutive LT patients for the time period from pretransplant baseline to 1 year after transplantation to determine if there is a recovery of kidney function after LT.
MATERIALS AND METHODS
The medical records of all deceased donor LTs performed at a single center between July 1, 2001, and June 30, 2011, were reviewed. Extracted data came from the comprehensive transplant recipient registry at our center and individual written and electronic medical records. Recipient inclusion criteria included all LT recipients 18 years or older receiving a deceased donor, whole organ liver allograft. Patient survival data were collected from the transplant database. This database also includes all laboratory values measured at any time point before and after transplantation for all recipients. There were 90 patients who underwent combined liver and kidney transplantation during the study period, and these patients were included in the study but analyzed separately. There were 132 deaths in the first year (12%), and renal function for these patients was included in the analysis up until the time of their death. Pretransplant recipient history of hypertension and diabetes was extracted from the original medical records completed during the transplant evaluation. New posttransplant hypertension and diabetes were not extracted from the medical record and are not included in this analysis.
Serum creatinine levels were obtained from the patient database at baseline and on postoperative days 1, 7, 14, 30, 90, 180, and 365. Baseline creatinine was defined as the most recent value immediately prior to transplantation. The GFR was calculated at each time point by using the 4-variable MDRD equation which consists of serum creatinine, sex, age, and race, as previously mentioned. Patients undergoing dialysis were calculated as having a GFR of 10 mL/min during the period in which they were on dialysis, and the use of median values in the analysis assured that these patients were always in the lowest subgroup for renal function at these time points. Dialysis data were collected from hospital charge listings for the transplant hospitalization. Additionally, the ongoing patient surveillance notes by the individual transplant coordinators in the transplant center database were reviewed to identify any patient receiving dialysis at any time point in the first year after LT. Finally, the GFR values for each patient were reviewed to find outlying values that could indicate that a patient had progressed to dialysis during a particular time period; these periods were then investigated individually to assure the accuracy of dialysis data.
All recipients were listed for transplantation according to standardized criteria as established at our center and by the United Network for Organ Sharing. In patients receiving retransplantation within 1 year of the original transplant, the analysis included only data for the first transplant. The immunosuppression protocol has been described previously and includes induction therapy with rabbit antithymocyte globulin (total dose, 6 mg/kg) and steroids (3 bolus doses total), followed by initiation of maintenance prograf monotherapy started between postoperative days 2 and 4. Goal tacrolimus levels were 7 to 9 ng/dL in the first month and 5 to 8 ng/dL, thereafter.17 Some patients were started on additional immunosuppressant agents after the first 3 month posttransplant period when they returned to their primary hepatologist. Agents included mycophenolate mofetil, azathioprine, and sirolimus, but the use of these was limited to cases of refractory renal insufficiency and accounted for less than 5% of the total population.
Donor livers were recovered using standard procurement techniques including aortic flushing and cold storage as has been described previously.18 The overall median cold ischemia time was 7 hours with median warm ischemia time of 27 minutes. Ninety-five percent of all transplants during the study period were performed using the piggyback hepatectomy technique which has been described previously and was technically equivalent for all participating surgeons.19 Overall median anesthesia time was 4 hours and 40 minutes. There was no use of venovenous bypass or portocaval shunting in any case.
Primary study variables included absolute change in GFR, and percent change in GFR, at time points up to and including 1 year after transplantation. Demographic and clinical variables included recipient age, sex, race, body mass index ([BMI], kg/m2), MELD score, pretransplant GFR, clinical history of hypertension and diabetes, and graft warm and cold ischemia time. Donor variables included age, race, sex, BMI, and hepatic steatosis as measured by reperfusion biopsy. All patients with available data for each measured time point were included in the analysis for that time point, but may have been excluded at subsequent time points for death, incomplete follow-up or loss to follow-up.
Nonparametric testing (χ2) and analysis of variance was used in the bivariate analysis to assess for relationships between demographic and clinical variables and the primary outcome variables. Retrospective analysis of data for liver transplant patients at our center was reviewed and approved by the institutional review board of the Indiana University School of Medicine. Statistical testing was performed on statistical software for the social sciences (IBM SPSS Statistics, Version 22, IBM Corporation, 2014).
Patient demographics are shown in Table 1. There were 1185 adult, deceased donor whole single organ LTs performed between 2001 and 2011. Complete data were available for 100% of patients at transplant and 88% of patients at 1 year. Minimum follow-up was 12 months. Data for the intervening time periods were included as available. There were 90 combined liver and kidney transplants during this period. The overall median recipient age at transplant was 53 years, with 68% men, 88% white, and a median BMI of 27.6. The mean MELD score at transplant was 19 (range, 6‐40). Median baseline serum creatinine level was 1.0 mg/dL (mean, 1.1; range, 0.4-7.5 mg/dL), and median baseline GFR was 78 mL/min per 1.73 m2 (mean, 83; range, 6‐326 mL/min per 1.73 m2). The prevalence of pretransplant recipient hypertension and diabetes was 31% and 32%, respectively, with 14% having both hypertension and diabetes. Donor characteristics included a median age of 42 years, 56% men, 81% white, median BMI of 25.9, with stroke and trauma accounting for 80% of donor deaths. Cold and warm ischemia times were 6.6 hours and 26 minutes, respectively.
Table 2 lists the percentage of patients with each stage of CKD at baseline and then at 1, 3, 6, and 12 months after transplantation. There is a clear downward shift from stages 0 and I to stages II and III. The percentage of patients in stage IV and V remains consistent. The greatest change is for those with previously normal renal function (GFR > 90). At baseline, 42% of patients have GFR greater than 90, but this decreases to only 18% at 1 year. Patient survival was similar for all recipients with GFR greater than 30 (86%‐90%, 1-year survival), but was much lower for those with GFR less than 30 at transplant (67%‐72%, 1-year survival) (not shown). Table 3 lists the absolute 1-year change in GFR for various groupings of pretransplant MELD score. All 3 groups with MELD less than 26 experience a significant decline in GFR, whereas the higher MELD groups and the liver/kidney group experience improvement. At 1 year, all groups have very similar renal function (GFR range, 55‐68; P value, 0.12).
Figure 1 (graphs A–F) provides a graphical illustration of 1-year absolute change in GFR stratified by baseline GFR, recipient age group, MELD at transplant grouped, recipient sex and race, and pretransplant history of hypertension and diabetes. The recipient subgroups with baseline GFR less than 30 mL/min per 1.73 m2 (Figure 1, graph A) were the only groups to experience improvement in renal function in the first year. Patients with GFR 30 to 59 mL/min per 1.73 m2 (stage III CKD) had no change, and those with GFR greater than 60 mL/min per 1.73 m2 experienced a large decrease. When stratified by recipient age (Figure 1, graph B), all groups experienced a decline in GFR. For recipients with MELD greater than 30 (Figure 1, graph C), the GFR improved markedly, whereas it was essentially unchanged for MELD 26 to 30. All patient groups with MELD of 25 and less experienced a worsening of renal function. Males and females experienced a similar decrease in GFR, with males having a higher GFR at all time points (Figure 1, graph D). Similarly, blacks and non-blacks experienced a similar decrease in GFR at all time points, with blacks having a higher GFR (Figure 1, graph E). Finally, patients with hypertension, diabetes, both hypertension and diabetes, or no hypertension or diabetes all experienced a decrease in GFR. The disease free patients have the highest GFR at all time points, and those with both diseases have the lowest GFR at all time points (Figure 1, graph F). Review of these graphs demonstrates a consistent decrease in GFR in the first 30 days after transplantation. Thereafter, there is a gradual improvement or stabilization up to 12 months after transplantation, but this improvement does not return to the baseline GFR. Of particular interest is the fact that nearly all groups tend toward a GFR of 60 to 70 mL/min per 1.73 m2 by 1 year after transplantation. Cox proportional hazards modeling of 10-year posttransplant patient survival demonstrates an incremental worsening of survival with a lower baseline GFR (Figure 2). The group with GFR greater than 90 mL/min per 1.73 m2 has an improved survival that reaches statistical significance (P = 0.05) (not shown).
Absolute and percent changes in GFR at 1 year after transplantation are stratified by recipient and donor factors (Table 4). Absolute change in GFR was divided into 4 groups based on statistical and clinical distributions: (1) increase in GFR greater than 5 mL/min per 1.73 m2 (22%), (2) no change in GFR ± 5 mL/min per 1.73 m2 (14%), (3) decrease in GFR from 6 to 20 mL/min per 1.73 m2 (22%), and (4) a decrease in GFR greater than 20 mL/min per 1.73 m2 (42%). These 3 groups did not differ in with regard to age, sex, race, history of hypertension or diabetes, or BMI. The patients transplanted with higher MELD scores, higher baseline serum creatinine levels, and a lower baseline GFR were the most likely to experience improvement in GFR at 1 year. Data for percent change in GFR at 1 year after transplantation show similar results, with the addition of hypertension and diabetes being associated with variable changes in GFR.
Of 1032 patients with complete 1-year follow-up, there were 203 (20%) who experienced some improvement in GFR. Analysis of this subgroup demonstrated that many patients experienced an acute decline in their renal function before the time of transplant. Data taken for this group at 3 months before transplant found that 76% of these patients had a GFR greater than 60 mL/min per 1.73 m2 (median, 76 mL/min per 1.73 m2), with 31% having a GRF greater than 90 mL/min per 1.73 m2. The GFR for this group declined from 76 mL/min per 1.73 m2 (at 3 months before transplantation) to a median of 55 mL/min per 1.73 m2 at the time of transplant. This included 42% with GFR greater than 60 mL/min per 1.73 m2 and 11% greater than 90 mL/min per 1.73 m2. Among this subgroup, there is a clear trend of steadily worsening GFR measured at 90, 60, and 30 days before transplantation. This supports the current opinion that patients with an acute decline in renal function immediately before transplantation have an increased chance for recovery of lost function.
The presence of kidney disease after LT appears to be pervasive, with greater than 80% of patients in this study having at least stage II CKD 1 year after transplantation. Additionally, 39% have at least stage III CKD. These numbers only account for the patients alive at 1 year (12% death rate), so that these results likely underrepresent the severity of CKD in this population. Results from this analysis support previous research in the area of postliver transplant renal function demonstrating an early and marked decline in renal function in the first year after transplantation.20-22 The etiology of this decline is likely multifactoral and may include factors, such as toxicity of calcineurin inhibitors, unrecognized severity of pretransplant renal dysfunction, inaccurate measures of renal function in liver transplant patients, and an aging transplant population. Previous research has demonstrated that 1 year after successful LT, 40% of patients have stage 3 CKD and 20% develop end-stage renal disease at 5 years.23,24 With current 5-year liver transplant patient survival of 70%, this epidemic of kidney disease results in an enormous number of dialysis-dependent patients, many of whom will qualify for listing for renal transplantation and put additional pressures on the limited resource of deceased donor kidneys.
One of the primary contributors to the epidemic of renal dysfunction in cirrhotic patients is the inaccurate measure of renal function. The primary predictor of posttransplant CKD is the presence of pretransplant CKD.25,26 Serum creatinine is the measure of renal function most commonly followed by clinicians, both before and after transplantation, and is a heavily weighted component of the MELD score. Creatinine is a byproduct of the metabolism of creatine which is stored in the muscles. Patients with liver failure have a marked decline in creatine, and the expected baseline serum creatinine in liver failure patients drops to 0.4 and 0.6 mg/dL for cirrhotic women and men, respectively.27,28 This compares to an upper limit of normal for serum creatinine of 1.2 and 1.5 mg/dL for women and men in the general population. Therefore, when a patient with end-stage liver disease meets the criteria for HRS with a serum creatinine greater than 1.5 mg/dL, they have lost 50% to 75% of their baseline renal function. Unfortunately, the commonly used measures of renal function, including serum creatinine and 24-hour creatinine clearance, are poor markers for renal dysfunction in liver failure patients. The currently recommended measure of GFR in the cirrhotic patient is the MDRD equation which incorporates the serum creatinine, but adds additional factors for age, sex, and race.16
Despite the routine use of antibody-based immunosuppression induction at our center, coupled with the delayed use of low-dose calcineurin inhibitors, there is a marked decrease in GFR in the first 30 days after LT. Though there is some recovery of renal function after the perioperative period, 64% of recipients have a reduced GFR at 1 year, which includes 42% who experience a decrease of at least 20 mL/min per 1.73 m2 compared to baseline. There are 63% of patients with a decrease in GFR greater than 10% from baseline, including 39% with a decrease greater than 25%. These results are particularly concerning given the relatively low goal tacrolimus levels at our center of 7 to 9 ng/dL in the first month and 5 to 8 ng/dL, thereafter. We do not routinely use dual agent or sirolimus-based immunosuppression, so it is unclear what effect these differences would have on renal outcomes.
The only patient subgroups with demonstrated posttransplant improvement in renal function are those with severe renal dysfunction at baseline, as demonstrated by a GFR less than 50 mL/min per 1.73 m2. Though HRS as a component of renal dysfunction cannot easily be quantitated, clearly the patients with the lowest GFR and highest MELD scores have a substantial component of HRS and experience some recovery of renal function with transplantation. All other subgroups do not appear to have the same degree of recoverable HRS and experience a decrease in GFR. The 2 subgroups with baseline GFR less than 30 mL/min per 1.73 m2 do demonstrate a recovery of renal function at 1 year, but the absolute GFR remains at 55 and 40 mL/min per 1.73 m2, for the 2 groups. This suggests that the posttransplant improvement in renal dysfunction related to pretransplant HRS does little to compensate for the absolute loss in GFR related to prolonged liver failure. The change in GFR by MELD score (Figure 1, graph C) follows a similar pattern for the high MELD patients, though all groups are more tightly clustered and have a similar GFR by 90 days after transplantation and thereafter. This finding suggests that the high calculated MELD score is driven by poor renal function, which improves with transplantation.
Several recent studies have examined the incidence, clinical outcome, and prognostic indicators of renal insufficiency in the postliver transplant period.9-14 Results from the present study support previous research suggesting a poor prognosis for renal function for most patients after LT. Those patients with lower MELD scores, and higher baseline GFR, experience a significant loss of renal capacity after transplantation. Conversely, patients with higher MELD scores and a lower baseline GFR experience improvement in renal function after transplantation. Unfortunately, a significant proportion of the lost renal function in all patients appears to be permanent. Clearly, a component of renal function can be recovered after transplantation, but the absolute GFR recovered may be variable and may not approach the GFR maintained by recipients who never experienced clinically significant loss in function. From a long-term perspective, allograft and patient life span is likely optimized with transplantation of patients at lower MELD scores before they have experienced a significant deterioration of clinical status and renal function. Finally, this paper does not account for new onset diabetes and hypertension because these data points were not collected in this study. Minimization of these 2 disease processes after transplantation may offer a means to improve posttransplant renal function.
Results from this study have relevance to clinical management of the patient with liver failure. Patients listed for liver transplant should not have an unrealistic expectation of recovery of renal function after transplantation. This paper clearly demonstrates that most patients do not have a recovery of decreased renal function after LT. And, for those that do have improvement, the actual positive change in GFR is likely not clinically relevant. And, finally, the improvement does not correlate with an improvement in survival. Therefore, the threshold for combined liver and kidney may need to be higher in these patients. Or, perhaps they should be excluded from liver transplant for poor posttransplant prognosis. The data inform the ongoing discussion regarding the increased use of kidney grafts in the liver failure population and may contribute to the ongoing discussion regarding calcineurin inhibitor minimization after solid organ transplant. Clinically, strict use of GFR, rather than serum creatinine, as a measure of renal function may assist the clinician in recognizing changes in renal function, prompting them to more aggressively address this issue. For each patient visit, the pretransplant GFR should be noted to determine the current change from previous best function. Such measures will serve to focus the clinician on preservation of renal function as a priority.
1. Lebron Gallardo M, Herrera-Gutierrez ME, Seller-Perez G, et al. Risk factors for renal dysfunction in the postoperative course of liver transplant. Liver Transpl
. 2004; 10 (11): 1379–1385.
2. Ojo AO, et al. Chronic renal failure after transplantation of a nonrenal organ.[see comment]. N Engl J Med
. 2003; 349 (10): 931–940.
3. Pawarode A, Fine DM, Thuluvath PJ. Independent risk factors and natural history of renal dysfunction in liver transplant recipients. Liver Transpl
. 2003; 9 (7): 741–747.
4. Durand F, Valla D. Assessment of prognosis of cirrhosis. Semin Liver Dis
. 2008; 28 (1): 110–122.
5. Bloom RD, Reese PP. Chronic kidney disease after nonrenal solid-organ transplantation. J Am Soc Nephrol
. 2007; 18 (12): 3031–3041.
6. Bahirwani R, et al. Transplantation: impact of pretransplant renal insufficiency. Liver Transpl
. 2008; 14 (5): 665–671.
7. Campbell MS, et al. Renal function after orthotopic liver transplantation is predicted by duration of pretransplantation creatinine elevation. Liver Transpl
. 2005; 11 (9): 1048–1055.
8. Wilkinson A, Pham PT. Kidney dysfunction in the recipients of liver transplants. Liver Transpl
. 2005; 11 (11 Suppl 2): S47–S51.
9. Israni AK, et al. Predicting end-stage renal disease after liver transplant. Am J Transplant
. 2013; 13 (7): 1782–1792.
10. Gonwa TA, et al. Kidney transplantation after previous liver transplantation: analysis of the organ procurement transplant network database. Transplantation
. 2011; 92 (1): 31–35.
11. Leithead JA, et al. Chronic kidney disease after liver transplantation for acute liver failure is not associated with perioperative renal dysfunction. Am J Transplant
. 2011; 11 (9): 1905–1915.
12. Rodriguez-Peralvarez M, Germani G, Darius T, et al. Tacrolimus trough levels, rejection and renal impairment in liver transplantation: a systematic review and meta-analysis. Am J Transplant
. 2012; 12 (10): 2797–2814.
13. Ruebner R, Goldberg D, Abt PL, et al. Risk of end-stage renal disease among liver transplant recipients with pre transplant renal dysfunction. Am J Transplant
. 2012; 12: 2958–2965.
14. Sharma P, et al. Impact of MELD-based allocation on end-stage renal disease after liver transplantation. Am J Transplant
. 2011; 11 (11): 2372–2378.
15. Kidney Disease Outcomes Quality Initiative (KDOQI). Clinical practice guidelines and clinical practice recommendations for diabetes and chronic kidney disease. Am J Kidney Dis
. 2007; 49 (2 Suppl 2): S12–S154.
16. Gonwa TA, et al. Estimation of glomerular filtration rates before and after orthotopic liver transplantation: evaluation of current equations. Liver Transpl
. 2004; 10 (2): 301–309.
17. Mangus RS, Fridell JA, Vianna RM, et al. Immunosuppression induction with rabbit antithymocyte globulin (rATG)± rituximab in 1000 liver transplants with long-term follow-up. Liver Transpl
. 2012; 18: 786–795.
18. Mangus RS, et al. Comparison of histidine-tryptophan-ketoglutarate solution (HTK) and University of Wisconsin solution (UW) in adult liver transplantation.[see comment]. Liver Transpl
. 2006; 12 (2): 226–230.
19. Mangus RS, et al. Use of the piggyback hepatectomy technique in liver transplant recipients with hepatocellular carcinoma. Transplantation
. 2008; 85 (10): 1496–1499.
20. Fabrizi F, et al. Chronic kidney disease after liver transplantation: recent evidence. Int J Artif Organs
. 2010; 33 (11): 803–811.
21. Bahirwani R, Reddy KR. Outcomes after liver transplantation: chronic kidney disease. Liver Transpl
. 2009; (15 Suppl 2): S70–S74.
22. Cabezuelo JB, et al. Risk factors of acute renal failure after liver transplantation. Kidney Int
. 2006; 69 (6): 1073–1080.
23. Yalavarthy R, Edelstein CL, Teitelbaum I. Acute renal failure and chronic kidney disease following liver transplantation. Hemodial Int
. 2007; (11 Suppl 3): S7–S12.
24. Pham PT, Pham PC, Wilkinson AH. Renal function outcomes following liver transplantation and combined liver-kidney transplantation. Nat Clin Pract Nephrol
. 2007; 3 (9): 507–514.
25. Jeyarajah DR, et al. Hepatorenal syndrome: combined liver kidney transplants versus isolated liver transplant. Transplantation
. 1997; 64 (12): 1760–1765.
26. Seu P, et al. The hepatorenal syndrome in liver transplant recipients. Am Surg
. 1991; 57 (12): 806–809.
27. Takabatake T, et al. Low serum creatinine levels in severe hepatic disease. Arch Intern Med
. 1988; 148 (6): 1313–1315.
28. Spencer K. Analytical reviews in clinical biochemistry: the estimation of creatinine. Ann Clin Biochem
. 1986; 23 (Pt 1): 1–25.