New-Onset Diabetes After Liver Transplantation: An SRTR Database Study of Incidence and Risk Factors

Introduction

New-onset diabetes following transplantation (NODAT) is a prevalent long-term issue that occurs after liver transplant surgery [1] and has been shown to adversely affect patient outcomes after liver transplant procedures [2]. Given that advances in transplantation and immunosuppression have improved survival rates, the number of patients at risk for long-term complications like NODAT is increasing, making it a growing clinical and public health concern [3,4]. Studies have shown that the incidence of NODAT following liver transplant is around 30%, although this varies widely depending on diagnostic criteria, follow-up duration, and study population [3,5].

A previous evaluation of data from the Organ Procurement and Transplant Network (OPTN)/United Network for Organ Sharing (UNOS) registry identified several critical factors associated with the occurrence of NODAT, including advanced recipient age, Black race, elevated BMI, presence of cirrhosis, increased donor age, donor diabetes history, immunosuppression based on tacrolimus usage, corticosteroid usage, and absence of induction therapy [5]. Our prior research also investigated the risk factors for NODAT among liver transplant patients with hepatitis C virus (HCV) and found that HCV-positive recipients had a higher incidence rate and more contributing factors for NODAT than HCV-negative recipients, with most cases occurring within 2 years after transplant [6]. In addition, a single-center study reported that 15% of recipients undergoing living donor liver transplantation developed NODAT, with 64% of cases occurring within the first 3 months following transplant. This study also identified advanced recipient age and higher BMI as potential contributing factors [7]. Further, a multi-center study from China found that while the occurrence of NODAT was comparable to its rates in Western nations, complications related to NODAT were more prevalent and serious in the Chinese population [8]. In terms of its prognosis, PTDM patients showed lower survival (P=0.029), with increased risks of sepsis (OR=3.1), fungal infection (OR=3.0), kidney disease (OR=3.3) and biliary complications (OR=3.5), as well as worse tumor-free survival in HCC patients (P=0.049). Thus, NODAT is a serious condition that requires particular attention in the Chinese population.

The studies discussed above were conducted over a decade ago. Significant advances have occurred in recent years in knowledge of transplant immunology, management of post-transplant metabolic complications, and transplant techniques. As advances in transplantation methods and immunosuppressive drugs have led to an increase in transplant volumes and improved recipient survival, the occurrence of NODAT has been rising in recent years [4]. However, there is a knowledge gap regarding the contemporary incidence of NODAT and its impact on graft survival in the modern era. This necessitates further investigation to elucidate the long-term impact of NODAT on transplant surgery outcomes. Therefore, this retrospective study sought to fill this research gap by using the Scientific Registry of Transplant Recipients (SRTR) database to assess the influence of NODAT on post-transplant outcomes (perioperative complications and survival) over the past decade and to identify potential prognostic contributing factors among liver transplant recipients.

Material and Methods

DATA SOURCE:

This study was conducted using information obtained from the SRTR database, which contains data on all donors, wait-listed candidates, and transplant recipients in the United States submitted by members of OPTN. The Health Resources and Services Administration, U.S. Department of Health and Human Services, oversees the activities of OPTN and SRTR contractors [9]. The study protocol was reviewed and approved by the Ethics Committee of the First Affiliated Hospital, College of Medicine, Zhejiang University, China.

STUDY COHORT:

We included adult liver transplant recipients (age ≥18 years) who received their primary liver graft between January 1, 2013, and December 31, 2022, and had at least 1 post-transplant follow-up report containing data on their diabetes status following transplantation.

The exclusion criteria were re-transplantation (n=3146), multi-organ transplantation (n=6448), pediatric liver transplantation (n=4152), pre-existing diabetes mellitus (n=15 612), unknown diabetes status (n=163), and perioperative death within 90 days (n=1439). These criteria were chosen to create a homogenous cohort of primary adult liver transplant recipients and to ensure that the diagnosis of NODAT was unambiguous, excluding those with pre-existing diabetes or confounding multi-organ procedures. After these criteria were applied, 39 828 liver transplant recipients were ultimately included, as shown in Figure 1.

DEFINITION OF NODAT:

NODAT was defined as the development of diabetes in non-diabetic liver transplant recipients based on the availability of at least 1 record of diabetes mellitus during the post-transplantation follow-up period. Consistent with the criteria of the American Diabetes Association [10], NODAT was characterized by fasting glucose ≥126 mg/dL (7 mmol/L) or non-fasting glucose ≥200 mg/dL (11.1 mmol/L) as documented on at least 2 separate occasions, or the need for antidiabetic medication that persisted beyond the first month after transplantation. The presence of diabetes was recorded as “yes” or “no” in the follow-up records of the SRTR dataset. We considered the NODAT onset date to be the earliest documentation of diabetes reported after liver transplantation.

COVARIATES:

We collected data on demographic, clinicopathologic, and transplant-specific variables for donors and recipients. Covariates were selected based on their established or plausible association with NODAT and survival outcomes reported in the literature [5,11]. In addition, data were collected on the following covariates that were considered to potentially affect NODAT and patient/graft survival: The recipient covariates were age; sex; BMI; race; etiology of liver disease; Model for End-Stage Liver Disease(MELD) score; pre-transplant medical status (ICU admission, hospitalization without ICU admission, or no hospitalization); pre-transplant clinical history (hypertension, peripheral vascular disease, coronary artery disease, previous cancer, pulmonary embolism, and/or chronic obstructive pulmonary disease); liver-related complications (spontaneous bacterial peritonitis, portal vein thrombosis, transjugular intrahepatic portosystemic stent shunt [TIPSS], and previous upper-abdominal surgery); HLA mismatch; length of hospital stay; acute rejection episodes; induction immunosuppression; discharge immunosuppression therapy with tacrolimus, cyclosporine, mycophenolate mofetil, sirolimus, and steroids; cytomegalovirus (CMV) infection; and Ebola virus (EBV) infection. The donor factors included age; BMI; sex; living donor liver transplantation (LDLT)/deceased donor liver transplantation (DDLT); race; cause of death; history of diabetes and hypertension; cold ischemia time (CIT); warm ischemia time (WIT); most recent aspartate aminotransferase, alanine transaminase, total bilirubin, international normalized ratio, serum sodium, and serum creatinine values; and donor risk index (DRI). DRI was calculated based on the model developed by Feng et al [12].

OUTCOME MEASUREMENTS:

We compared patient survival and graft survival between recipients with and without NODAT, and also evaluated the differences in the cause of death and graft failure across the 2 groups. The outcome measures included current status and time to event. We defined patient follow-up as the time from transplantation until death or last known follow-up date. Transplant centers reported data on the occurrence and date of death, supplemented with information from the U.S. Social Security Administration and OPTN. We conducted death-uncensored graft survival analysis from the transplant date until liver graft failure or patient death. Graft failure was defined in the SRTR as the date of being relisted for liver transplantation or patient death. Graft loss was analyzed on an event-time basis, with patients having a viable graft at their most recent post-transplant follow-up being censored.

For recipients experiencing graft failure or death, we obtained information on the cause of death from the SRTR dataset. The causes of graft loss were primary graft failure, vascular thrombosis, hepatitis, rejection, recurrent disease, death with a functioning graft, and unknown cause. The causes of death were graft failure, cardiovascular events, organ failure, infection, malignancy, other causes (eg, hemorrhaging and trauma), and unknown causes.

STATISTICAL ANALYSIS:

Continuous variables were compared using Student’s t-tests, and categorical variables were compared using chi-square tests. Survival rates were estimated using the Kaplan-Meier method, with log-rank tests for comparison between groups. Logistic regression was used to identify risk factors for NODAT. Cox proportional hazards models were used to assess the impact of NODAT on graft and patient survival. As a limitation of using a registry database, cases with missing data on specific covariates were excluded from the multivariable analyses in a complete-case analysis approach. A 2-tailed P value <0.05 was considered to indicate statistical significance. Analyses were performed using SPSS version 25.0 (IBM Corporation, Armonk, NY, USA).

Results

POPULATION CHARACTERISTICS:

Of the 39 828 liver transplant recipients who were included, 2973 (7.5%) developed NODAT. Table 1 presents the recipient characteristics. The recipients who developed NODAT were significantly older (55±11 vs 53±12 years, P<0.001) and had significantly higher BMI (28.9±5.9 vs 28.4±6.0 kg/m2, P<0.001). Moreover, a significantly higher number of those in the NODAT group were male (68.0% vs 64.2%, P<0.001). Recipients in the NODAT group also exhibited higher rates of hypertension (9.0% vs 5.1%, P<0.001), TIPSS (8.0% vs 6.4%, P<0.001), and CMV infection (64.3% vs 59.9%, P<0.001). At discharge, compared to the recipients who did not develop NODAT, a significantly higher proportion of recipients with NODAT were prescribed steroids (84.9% vs 75.0%, P<0.001) and tacrolimus (95.7% vs 93.6%, P<0.001), but a significantly lower proportion of those with NODAT were prescribed mycophenolate mofetil (81.8% vs 85.5%, P<0.001).

The donor characteristics are presented in Table 2. Compared to the non-NODAT group, the NODAT group had significantly longer WIT (20.2±23.8 vs 12.7±20.1 min, P<0.001) and CIT (6.0±2.3 vs 5.7±2.4 h, P<0.001). A significantly higher percentage of recipients with NODAT than those without NODAT received grafts from donors with diabetes (13.5% vs 11.8%, P=0.007) and hypertension (35.5% vs 33.6%, P=0.031). Additionally, the NODAT group also had a significantly higher proportion of transplants from deceased donors (96.0% vs 94.4%, P<0.001) and donors who had died from cerebrovascular accidents/stroke (30.5% vs 27.6%, P=0.001). The DRI was slightly, but significantly, lower in the NODAT group than in the non-NODAT group (1.41±0.36 vs 1.44±0.36, P=0.016).

All patients included in the study had at least 1 follow-up record in the database. We evaluated the incidence of NODAT at 6 months after transplant and annually for the subsequent 5 years. Our analysis showed that the incidence of NODAT was the highest in the early post-transplant period, with peak rates observed at the 6-month and 1-year follow-up evaluations, underscoring this timeframe as a critical-risk phase for its development (Figure 2).

RISK FACTORS FOR NODAT:

Univariate analysis identified recipient age, sex, BMI, CMV infection, steroid and tacrolimus use, LDLT, CIT, WIT, and donor diabetes history as factors associated with NODAT development. Multivariate analysis revealed several independent risk factors. Specifically, a recipient age of ≤50 years (OR=0.78, 95% CI=0.69–0.89, P<0.001) and a recipient BMI of ≤25 kg/m2 (OR=0.84, 95% CI=0.75–0.94, P=0.003) were associated with a significantly lower risk of developing NODAT. Conversely, male sex was associated with an increased risk (OR=1.21, 95% CI=1.07–1.34, P=0.002), as were CMV infection (OR=1.13, 95% CI=1.01–1.26, P=0.033), steroid use (OR=1.81, 95% CI=1.55–2.12, P<0.001), tacrolimus use (OR=1.47, 95% CI=1.21–1.77, P<0.001), DDLT (OR=0.80, 95% CI=0.65–0.99, P=0.035), prolonged WIT (OR=1.01, 95% CI=1.01–1.02, P<0.001), and donor diabetes history (OR=1.22, 95% CI=1.05–1.42, P=0.012) (Table 3).

IMPACT OF NODAT ON PATIENT AND GRAFT SURVIVAL:

Kaplan-Meier analysis showed there was no significant difference in patient survival between the NODAT and non-NODAT group (P=0.273, log-rank test, Figure 3A). However, NODAT was associated with decreased graft survival (P=0.005, log-rank test, Figure 3B).

According to multivariable Cox regression analysis (Table 4), NODAT was an independent predictor of decreased graft survival (HR=1.28, 95% CI=1.10–1.48, P<0.001). The results indicated that NODAT increases the risk of graft failure by 28%.

Some other significant recipient factors associated with decreased graft survival were male sex (HR=1.13, 95% CI=1.02–1.24, P=0.014), higher MELD scores (HR=0.99, 95% CI=0.98–0.99, P=0.012), and prolonged hospital stay (HR=1.01, 95% CI=1.01–1.02, P<0.001). This implies that a complex post-transplant disease course negatively affects graft outcomes. Male donors (HR=1.10, 95% CI=1.01–1.20, P=0.032) and higher DRI (HR=1.66, 95% CI=1.45–1.90, P<0.001) were also associated with decreased graft survival, underscoring the critical importance of donor selection. Interestingly, the use of tacrolimus at discharge was identified as a factor that promotes graft survival (HR=0.61, 95% CI=0.52–0.71, P<0.001), despite its known role as a risk factor for development of NODAT.

CAUSES OF GRAFT FAILURE AND PATIENT DEATH:

Comparison of causes of graft failure (Table 5) showed that the NODAT group had significantly higher rates of vascular thrombosis (10.5% vs 7.7%, P=0.038) and death with a functioning graft (43.6% vs 36.7%, P=0.004) than the non-NODAT group. There were no significant differences in the rates of primary non-function, recurrent primary disease, rejection, or infection as causes of graft failure between the 2 groups.

Analysis of the causes of patient death revealed that there were no significant differences between the NODAT and non-NODAT groups in terms of primary non-function, cardiovascular/cerebrovascular disease, multi-organ failure, infection, cancer, or other etiologies as causes of death (Table 5).

Discussion

This extensive nationwide cohort analysis of a U.S. database spanning a decade revealed that 7.5% of liver transplant recipients developed NODAT. We also identified several risk factors for NODAT development, such as advanced recipient age, male sex, elevated BMI, CMV infection, corticosteroid and tacrolimus use, transplant from a deceased donor, longer WIT, and donor diabetes history. Importantly, NODAT was associated with decreased transplant survival but did not significantly impact overall recipient survival. In particular, recipients who developed NODAT experienced higher rates of transplant failure attributed to vascular thrombosis. Overall, we believe that these observations reflect recent trends in transplant techniques and management and present an updated picture of the current risk factors for NODAT occurrence and its outcomes.

The incidence of NODAT in our cohort was lower than the previously reported rates of 10–30% [3,4]. This finding may reflect advances in immunosuppression protocols, perioperative management, and patient selection that have occurred over the past decade. In particular, the use of personalized immunosuppression regimens, such as steroid-sparing and tacrolimus minimization protocols, has been shown to reduce the risk of NODAT [13,14]. Moreover, the implementation of proactive lifestyle modifications and strict glucose monitoring has contributed to improved metabolic outcomes in liver transplant recipients [15]. Further, recent studies have emphasized the importance of early identification and management of post-transplant hyperglycemia in optimizing long-term outcomes [16,17]. Frequent monitoring and careful management in the peri-and postoperative period of liver transplant based on recent advances and recommendations may have contributed to the decreased incidence of NODAT observed in this study. Another crucial factor may be the enhanced and more rigorous screening for diabetes during the pre-transplant evaluation phase in recent years [18]. This practice ensures that patients with pre-existing, undiagnosed diabetes are correctly identified before surgery, and thus are not misclassified as NODAT cases after transplantation, which could have inflated incidence rates in older studies. However, the incidence could be underestimated, to some extent, given that the data were retrospectively collected from a registry and the presence of NODAT was determined based on the criteria applied by the database (and other relevant criteria may have not have been considered). Nonetheless, there is a notable difference in comparison with previous studies that warrants consideration and further exploration.

Our study identified several risk factors for NODAT that are consistent with previous literature [5,11,19]. According to our results, advanced age, male sex, and higher BMI, which are recognized risk factors for type 2 diabetes in the general public [20], seem to have a comparable impact on the development of NODAT. Notably, in line with reports from other solid-organ (heart and kidney) transplant settings, CMV infection was identified as an independent factor contributing to NODAT [21,22]. CMV has been suggested to contribute to diabetes development through various mechanisms, including direct pancreatic β-cell damage, increased insulin resistance, and altered glucose metabolism [23]. Thus, identifying and monitoring patients at risk (based on sex, age, and BMI) and diagnosing and treating CMV early may help reduce the risk of NODAT.

Corticosteroid usage and the use of tacrolimus for maintenance immunosuppression were linked to a heightened risk of NODAT in our analysis, consistent with previous reports [24]. Corticosteroids are known to induce insulin resistance and impair insulin secretion [25], while tacrolimus has been shown to exert direct pancreatic β-cell toxicity [26]. Our findings support the growing trend towards steroid-sparing and tacrolimus minimization strategies in liver transplantation [27].

Interestingly, while tacrolimus has previously been reported to contribute to the risk of NODAT, our analysis determined it to be a beneficial factor for transplant survival. This seemingly paradoxical effect may have several explanations. First, the superior immunosuppressive efficacy of tacrolimus might outweigh its diabetogenic effects in terms of overall graft protection, particularly when NODAT is well-managed [27–29]. Second, the impact of tacrolimus on graft survival could be dose-dependent, with lower doses providing immunosuppression benefits while minimizing metabolic side effects [27,30]. Future investigations need to explore this relationship between tacrolimus dose and its effects of metabolic pathways and the onset of post-transplant diabetes to determine the optimal dosing regimen. It is also important to address the counterintuitive finding of an HR of 0.99 for the MELD score, which would suggest a protective effect. Given that higher MELD scores are well-established predictors of poor outcomes [31], this result is likely an artifact of the complex multivariable analysis in a large database and should be interpreted with caution as it may be statistical noise rather than a true clinical effect.

The association between longer WIT and NODAT is a novel finding of this study. Although the observed effect size of WIT on NODAT development was modest (OR=1.01), its significance in our large cohort underscores the importance of minimizing ischemia time. As a modifiable factor directly influenced by the surgical process, any reduction in WIT could contribute to a lower incidence of NODAT and its associated complications – a crucial consideration in optimizing transplant outcomes [32]. Prolonged warm ischemia may exacerbate ischemia-reperfusion injury, leading to oxidative stress and inflammation that could, in turn, contribute to insulin resistance and β-cell dysfunction [32]. This finding highlights the importance of minimizing WIT during transplantation to decrease the likelihood of NODAT and related complications.

We found that NODAT was correlated with reduced transplant survival, even though it had no impact on overall patient survival. This observation is in line with some previous reports [24,33], while others have found no association between NODAT and graft survival [34,35]. However, the mechanisms underlying the link between NODAT and graft failure remain unclear and warrant further investigation. One possibility is that the metabolic derangements associated with NODAT, such as insulin resistance and hyperglycemia, may contribute to endothelial dysfunction and increased risk of vascular complications in the graft [36], as evidenced by the higher percentage of patients with graft failure caused by vascular thrombosis in the NODAT group in the present study. Future studies should aim to elucidate the specific pathways involved and identify strategies to mitigate the adverse impact of NODAT on transplant outcomes.

Our investigation has some inherent limitations. Firstly, the retrospective nature and reliance on registry-based information may have led to underestimation of the incidence of NODAT and the effects of potential confounders (as data on confounders were missing). Second, we lacked information on specific diabetes treatments and glycemic control, which could have impacted outcomes. Third, the definition of NODAT according to the database may not capture all cases, particularly those with more subtle glucose abnormalities. This may also have contributed to the underestimation of NODAT. Finally, the study duration may not be long enough to fully capture the long-term ramifications of NODAT on post-transplant clinical trajectories.

Conclusions

In summary, this large-scale analysis of contemporary data reveals that NODAT affects a significant portion (7.5%) of liver transplant recipients and stands out as an independent risk factor for decreased graft survival. Our findings highlight that immunosuppression strategies, such as the use of steroids and tacrolimus, along with operative factors like warm ischemia time, are significant contributors to NODAT risk. While NODAT did not affect patient survival within the study’s timeframe, its link to graft failure, particularly through vascular thrombosis, underscores the need for vigilant postoperative metabolic management. Proactive identification of high-risk patients and optimization of modifiable risk factors are crucial to improving long-term graft outcomes.