Chronic kidney disease (CKD) is a well-known complication in solid-organ transplant recipients with a frequency ranging between 10 and 90%.1,2 This large variability has been related to numerous factors including heterogeneity in the definition of post-transplant kidney disease, various follow-up lengths, methods of measurement of eGFR and differences in the types of transplantation studied.1,2 CKD after the transplantation of a non-renal organ leads to increased morbidity and mortality.1,2

LT recipients are an important group of non-renal solid-organ recipients: long-term survival of LT recipients is currently longer in comparison with the past due to better immunosuppressive therapies, better selection criteria and surgical procedures.3–14 The overall 1-year and 5-year patient survival is 90% and 75%, respectively. As survival time lengthened after liver transplant, CKD has emerged as a major long-term complication after LT.3–14 CKD is independently associated with poor survival among LT recipients.15,16

Some information in the medical literature regarding the occurrence of CKD after LT already exists. Many demographic, clinical and biochemical factors have been shown to play a role in the development of CKD after LT including arterial hypertension, immunosuppression, diabetes mellitus, metabolic syndrome, dyslipidemia, pre-operative GFR and hepatitis C.3–14 The impact of perioperative management factors, which are potentially modifiable, on the pathogenesis of progressive CKD is under active investigation.17

Since the implementation of MELD in 2002, the number of patients with impaired kidneys who develop CKD after LT has increased and will continue to increase, as the number of patients transplanted with MELD of 40 or greater is also increasing, at least in part because more patients will have renal dysfunction before LT.18,19 CKD has become currently one of the leading reasons for morbidity and death rate after LT. We have performed a retrospective study to assess frequency and risk factors for CKD in a large cohort of LT recipients followed up to 10 years at our institution. In addition, we have addressed incidence and pathogenesis of AKI in this population.

The mean follow-up was 53.2±32.6 months. 410 patients were enrolled in the study, and the descriptive analysis is reported in Tables 1 and 2. The median serum creatinine was 0.90mg/dL (interquartile range, 0.71; 1.0) in the whole group at baseline. Some patients underwent re-transplant (n=24), one of them underwent three LT transplants. All recipients received livers from deceased donors. The course of serum creatinine over the follow-up in our cohort is reported in Fig. 1.

Fig. 1.

Serum creatinine (mean±standard deviation) during the post-transplant follow-up.

(0.05MB).

Prior to LT, 39 (9.5%) had a eGFR <60mL/min/1.73m2 and had CKD, while 371 (90.5%) had eGFR >60mL/min/1.73m2. In the subset of patients with baseline CKD, 2 patients had CKD stage 3 and two with CKD stage 4. There were three patients with cryoglobulinemic GN, one had diabetic nephropathy, one had post-surgery single kidney, one renal tubular acidosis and 4 hepatorenal syndrome. Also, 14 patients had arterial hypertension, and 18 diabetes mellitus.

The characteristics of study patients with CKD versus non-CKD (at the time of LT) are shown in Table 3. The comparison between patients with or without eGFR <60mL/min/1.73m2 showed difference between the two groups with regard to arterial hypertension (P=0.026), insulin dependent DM (P=0.0039), and raised levels of serum uric acid (P<0.0001) (Table 3). No difference occurred in the frequency of anti-HCV positive, HBsAg positive and HBV/HDV positive patients between the two groups (Table 3). According to logistic regression analysis, age at LT (P=0.044), increased values of serum uric acid (P<0.0001), and insulin dependent DM (P=0.0034) were independently associated with CKD at baseline (Table 4).

A group of patients developed CKD over the follow-up (n=95, 25.6%). The baseline characteristics of patients who developed de novo CKD after LT and those who did not are reported in Table 5. There was difference with regard to non insulin-dependent DM (P<0.01), raised uric acid levels (P<0.05), and serum creatinine (P=0.0002) (Table 5). Serum uric acid levels were greater in patients with de novo CKD compared with those without, 8.5±2.1 vs. 6.7±1.98mg/dL, P<0.04. Multivariate Cox regression model showed that three covariates were independently linked to incident CKD-age at LT (P<0.0001), early post-LT AKI (P=0.007), and serum creatinine at LT (P=0.0002) (Table 6).

Patients who developed early post-transplant AKI (n=95) were categorized in AKI stage 1 (n=35), AKI stage 2 (n=37) and 3 (n=23). There was difference between patients who developed de novo CKD and those who did not with regard to early post-transplant AKI (P<0.0001) (Table 5). No difference concerning the frequency of viral hepatitis was recorded between the two groups (Table 5). The characteristics of patients (at the time of LT) with early post-LT AKI and those with peri-operative normal kidneys are shown in Table 7. Dialysis dependent AKI was more common in patients who developed de novo CKD in comparison with those who did not (P<0.001) (Table 5). Logistic regression analysis reported that serum creatinine at baseline (P<0.0154) and MMF use (P<0.04) were associated with the occurrence of early post-transplant AKI.

Thirty patients developed diabetes mellitus after LT (PTDM), in many of them non-insulin dependent diabetes occurred (Table 8). In the group of LT recipients having viral hepatitis (n=264), no patients with HBV-related and a few (n=12) with HCV-related cryoglobulinemic glomerular disease were recorded.

At the end of follow-up, there were 116 LT recipients with CKD; 95 patients had de novo and 21 baseline CKD. 109 (93.9%) LT recipients had CKD stage 3, four (3.5%) and three (2.6%) had stage 4 and ESRD, respectively. Two LT recipients are undergoing regular dialysis. 368 patients were followed up at the end of the study period.

In this retrospective cohort study of stable LT recipients followed at our institution, we found that a large (n=95, 25.6%) number of LT recipients developed CKD over the observation period. At the end of the follow-up, there were 116 LT recipients with CKD. An overwhelming majority of liver recipients did not have advanced CKD – two LT recipients undergo regular dialysis at the last follow-up.

Post-transplant CKD is a major public health problem among all non-renal solid organ transplant recipients. Previous studies reported on the occurrence of CKD after LT, with an incidence ranging between 20% and 80%.3–14 In addition to the evidence reported above, this wide range of results is related to other factors such as patient selection, or differences in managing patients, among others.3–14 We addressed the incidence of CKD among LT recipients in the MELD era; the current MELD system gives consistent weight to serum creatinine and this translated into greater incidence of post-transplant CKD.26,27 We adopted the CKD EPI equation to estimate GFR. According to a meta-analysis of serum creatinine based equations that estimated GFR among solid organ transplant recipients (around 40% LT recipients), CKD EPI equation was better at higher GFR compared to others.28

No role of viral hepatitis in the development of post-transplant CKD in our cohort was reported, unlike what others reported.3,4 Various factors could explain this – we made diagnosis of viral hepatitis by serologic assays instead of nucleic acid testing (NAT), HBV- or HCV-related cryoglobulinemic glomerular disease was uncommon, antiviral treatment with DAAs or other agents (data not shown) could be have improved renal outcomes.

Although the number of LT recipients who underwent long-term dialysis at the last follow-up was extremely small, the high frequency of CKD highlights the burden of cardiovascular risk in this population. It is well known that reductions in estimated GFR predict the development of fatal and non-fatal cardiovascular events, regardless of traditional CV risk factors (blood pressure, smoke, cholesterol, age, gender, among others), in the general population and in high-risk cohorts.29,30

A consistent group of patients developed de novo CKD post-LT in our study, and the frequency of AKI in the early postoperative period was independently associated with de novo CKD. The link between AKI and CKD after LT remains controversial31 – in vitro studies highlighted the occurrence of permanent kidney damage following AKI.32 It has been suggested that patients who experience AKI, even those who showed complete recovery from AKI, remain at risk for CKD.33,34 Careful monitoring of kidney function is needed in these patients post-LT. According to our regression logistic analysis, we found that the most important predictors for early post-transplant AKI were baseline serum creatinine and therapy with MMF. MMF was adopted more frequently by LT recipients with early post transplant AKI compared to those without. Patients with early post-transplant AKI received immunosuppression without CNI (or reduced dose CNI) to preserve kidneys and consequently adopted immunosuppressive therapy with MMF.

Our study involved the assessment of the etiology of chronic kidney disease after LT. Unlike other studies, we did not find a role of dyslipidemia, or extended criteria grafts.35 Metabolic syndrome at baseline or blood pressure and CNI dosing 1 year post-transplant were not evaluated.36 Lack of knowledge of CKD among LT recipients is a barrier to patient engagement and self-management of chronic disease risk factors and has been associated with progression of CKD post-transplant.37,38

Despite the large cohort of transplant recipients, our single-center study had some limitations. First is the retrospective design, which hampered the analysis of treatment changes over time by the physicians, including detailed changes in the immunosuppression. Second, we were not able to collect data on pre-operative proteinuria. Limited evidence exists on the impact of proteinuria on patient/kidney survival following LT39–41; unfortunately, dipstick urine analysis is made in selected circumstances in LT candidates at our institution. Third, the design of our study did not allow to understand fully the etiology of early post-transplant AKI as we did not include in our model intra-operative factors (i.e., surgical techniques, intra-operative bleeding, hemodynamic instability, or volume of transfused blood products) or donor factors. Finally, our analysis may have been biased toward the selection of healthier patients with better long-term survival, which may have decreased the impact of covariates such as pre-transplant kidney dysfunction.42

In conclusion, we found that a good number of long-term liver transplant survivors developed CKD after transplant. Viral hepatitis had no role in the pathogenesis of CKD. New onset CKD was associated with early post-transplant AKI, according to our multivariate analysis. The timely management of post-transplant AKI may potentially improve patient survival and decrease post-transplant death risk.

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

None declared.