Front. Immunol. Frontiers in Immunology Front. Immunol. 1664-3224 Frontiers Media S.A. 10.3389/fimmu.2026.1615875 Systematic Review Systematic review and meta-analysis of belatacept versus calcineurin inhibitors on risk of post-transplant diabetes mellitus in kidney transplant recipients WangXuchuan 1 † Data curation Formal analysis Funding acquisition Methodology Validation Visualization Software Writing – original draft Writing – review & editing SongDandan 2 † Visualization Project administration Writing – original draft Writing – review & editing HouShufu 3 Project administration Resources Visualization Methodology Conceptualization Writing – review & editing Writing – original draft LiuAiju 2 Writing – original draft Supervision Writing – review & editing Investigation Formal analysis Data curation Funding acquisition Visualization GaoJing 4 Writing – review & editing Project administration Writing – original draft Investigation Formal analysis Data curation LiuLei 2 * Project administration Writing – review & editing Methodology Supervision Funding acquisition Software Resources Writing – original draft Department of Ophthalmology, Shandong Provincial Third Hospital,Shandong University, Jinan, China Department of Neurology, Shandong Provincial Third Hospital, Shandong University, Jinan, China Central Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, China Department of Labor Union Office, Shandong Provincial Third Hospital, Shandong University, Jinan, China Correspondence: Lei Liu, 271178709@qq.com

†These authors have contributed equally to this work and share first authorship

 13 02 2026 2026 17 1615875 22 04 2025 21 01 2026 19 01 2026 Copyright © 2026 Wang, Song, Hou, Liu, Gao and Liu. 2026 Wang, Song, Hou, Liu, Gao and Liu https://creativecommons.org/licenses/by/4.0/ This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Background

The novel immunosuppressant belatacept demonstrates a unique mechanism of action that significantly improves renal function and reduces metabolic complications. However, systematic evidence comparing the risk of post-transplant diabetes mellitus (PTDM) and overall safety profiles between belatacept-based regimens and calcineurin inhibitor (CNI)-based protocols remains limited. This meta-analysis aims to synthesize high-quality evidence to determine the comparative efficacy of these two regimens in PTDM prevention and safety outcomes, thereby providing robust guidance for clinical decision-making.

 Methods

We systematically searched PubMed, Cochrane Library, CNKI, and EMBASE for studies published until November 30, 2024, comparing belatacept versus calcineurin inhibitors (CNIs) regarding PTDM risk in kidney transplant recipients. The primary outcome was PTDM incidence. Following data extraction and quality assessment, we performed pairwise meta-analyses to compare PTDM risk between belatacept (either less intensive(LI) or more intensive (MI) regimen) and CNIs. Bayesian network meta-analysis (WinBUGS 1.4.3) was then conducted for indirect comparison between belatacept LI and MI regimens.

 Results

The initial search yielded 3,206 records. After deduplication, title/abstract screening, and full-text evaluation, 6 studies involving 1,737 kidney transplant recipients were included in the final analysis. Compared with CNIs, belatacept demonstrated significant reductions in PTDM risk for both the LI (RR = 0.65, 95% CI 0.52-0.81, p<0.001; I²=30%) and MI (RR = 0.65, 95% CI 0.52-0.81, p<0.001; I²=30%) regimens. Bayesian network meta-analysis revealed no statistically significant difference between the LI and MI regimens.

 Conclusions

This meta-analysis demonstrates that belatacept significantly reduces PTDM risk compared to CNIs, a finding consistent with previous studies. Notably, both LI and MI dosing regimens showed protective effects, suggesting that even low-intensity belatacept therapy could serve as a viable alternative to CNIs, particularly for patients requiring reduced immunosuppressive toxicity.

 Systematic review registration

https://inplasy.com/, identifier INPLASY202540041.

 belatacept calcineurin inhibitors immunosuppressive agents kidney transplantation new-onset diabetes The author(s) declared that financial support was not received for this work and/or its publication. section-at-acceptance Alloimmunity and Transplantation Introduction

Kidney transplantation, as the preferred treatment for end-stage renal disease, significantly improves patient survival and quality of life. However, the metabolic complications associated with long-term immunosuppressive therapy remain a major clinical challenge. Among these, PTDM represents a clinically significant risk factor for graft loss and patient mortality, alongside other major determinants including cardiovascular disease, infections, and malignancies (1, 2). Studies indicate that PTDM is associated with an increased risk of cardiovascular events, higher infection rates, and accelerated decline in graft function, leading to a 3.67-fold increase in 5-year mortality (3, 4). CNIs, such as tacrolimus and cyclosporine, are considered central drivers of PTDM due to their direct β-cell toxicity and insulin resistance-inducing effects (5, 6). In contrast, belatacept, a selective T-cell costimulation blocker, has emerged as a potential alternative to CNIs, offering a more favorable metabolic profile. However, its efficacy in reducing PTDM risk remains controversial (7, 8).

Belatacept inhibits T-cell activation by blocking the CD28/B7 costimulatory pathway, thereby avoiding the direct pancreatic β-cell damage and metabolic disturbances caused by CNIs (9). Preclinical studies suggest that belatacept improves insulin sensitivity and reduces pro-inflammatory cytokine secretion (e.g., IL-6, TNF-α) from adipose tissue, potentially lowering PTDM risk (10–12). A randomized controlled trial (RCT) of 1,209 kidney transplant recipients demonstrated that belatacept-treated patients had a lower incidence of PTDM and improved glycated hemoglobin (HbA1c) levels compared to cyclosporine-treated patients (13). However, some cohort studies found no significant difference in PTDM incidence between belatacept and CNIs (14), suggesting that its protective effects may depend on baseline characteristics (e.g., obesity, hepatitis C infection) or concomitant immunosuppressive regimens (15). These inconsistencies may stem from heterogeneity in study design, varying PTDM diagnostic criteria (e.g., inclusion of impaired glucose tolerance), or inadequate control for confounders (e.g., glucocorticoid dosing) (15). Despite international guidelines recommending belatacept for high-risk patients (e.g., those with obesity or a family history of diabetes), the evidence remains largely based on subgroup analyses, with no dedicated meta-analysis focusing on PTDM outcomes (16). Furthermore, the long-term impact of belatacept on PTDM (e.g., 5-year incidence) and its association with other metabolic syndrome components (e.g., hypertension, dyslipidemia) remain unclear. Therefore, synthesizing existing clinical trial and observational data to clarify belatacept’s independent role in PTDM prevention and its modifying factors is crucial for optimizing immunosuppressive strategies and improving transplant outcomes.

This meta-analysis systematically evaluates the difference in PTDM risk between belatacept and CNIs in kidney transplant recipients, while exploring potential interactions with patient age, baseline metabolic status, and immunosuppression intensity. The findings will provide high-level evidence for personalized immunosuppressive regimens, ultimately reducing diabetes-related complications and improving long-term survival and quality of life for transplant recipients.

 Materials and methods Search strategy

This systematic review and meta-analysis adhered strictly to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (17). Two independent investigators systematically searched PubMed, Embase, CNKI (China National Knowledge Infrastructure), and the Cochrane Library to identify studies evaluating the impact of Belatacept versus CNIs on the risk of PTDM in kidney transplant recipients. The search period spanned from the inception of each database to November 30, 2024. To comprehensively assess the association between Belatacept and CNIs (e.g., tacrolimus, cyclosporine) and post-transplant diabetes risk, the following search terms were utilized: “kidney transplantation” OR “renal transplantation” OR “transplant recipients” AND”new-onset diabetes after transplant” OR “post-transplant diabetes mellitus” OR “diabetes mellitus” OR “glucose intolerance” OR “insulin resistance”AND “Belatacept” OR “CTLA4-Ig” OR “calcineurin inhibitor” OR “CNI” OR “tacrolimus” OR “cyclosporine” OR “immunosuppressive therapy”. To minimize publication bias, clinical trial registries (ClinicalTrials.gov, WHO ICTRP, and ISRCTN) were searched for unpublished or ongoing randomized controlled trials (RCTs). Duplicate publications of the same trial were resolved by prioritizing the most comprehensive and updated data.

 Inclusion and exclusion criteria

Inclusion Criteria: (1) Randomized controlled trials or prospective/retrospective cohort studies comparing belatacept vs. CNIs (tacrolimus/cyclosporine). (2) Adult (≥18 years) kidney transplant recipients, regardless of donor type (living/deceased). (3) Belatacept-based regimen (either LI or MI dosing) vs. any CNI-based regimen. (4) Reported new-onset diabetes after transplantation (NODAT/PTDM) incidence, defined by:ADA/WHO diagnostic criteria, or Use of insulin/oral hypoglycemics for ≥30 days post-transplant, or Fasting glucose ≥126 mg/dL or HbA1c ≥6.5% on two occasions. (5) Minimum follow-up of 6months post-transplant.

Exclusion Criteria: (1) Case reports, reviews, editorials, or studies without a control group (CNIs). (2) Non-kidney transplants (e.g., liver, heart) or pediatric recipients. (3) No clear PTDM definition or insufficient data for risk ratio (RR)/odds ratio (OR) calculation. (4) Studies where >20% of patients received simultaneous pancreas-kidney transplants (due to inherent diabetes risk differences). (5) Overlapping cohorts (only the most comprehensive dataset was included). (6) Studies designed to evaluate CNI-to-belatacept conversion protocols compared with maintenance CNI therapy.

 Data extraction and quality assessment

Two independent investigators extracted relevant data from eligible studies, with discrepancies resolved through discussion or consultation with a third researcher. Extracted data included: first author, publication year, study location (single/multi-center), clinical trial registration number (e.g., ClinicalTrials.gov or NTR), intervention groups (Belatacept [MI or LI] vs. CNIs), sample size, mean age, gender distribution (M/F), incidence of PTDM, and follow-up duration (months). The risk of bias was assessed using the Cochrane Risk of Bias Tool (version 5.3.0) (18, 19), with evaluation criteria covering randomization, allocation concealment, blinding, completeness of outcome data, selective reporting, and other potential biases. During the full-text screening phase, we did identify several cohort studies that met the inclusion criteria. However, due to significant heterogeneity in PTDM diagnostic criteria, follow-up duration, and adjustment for confounders (e.g., steroid dosing, baseline glycemic status), we decided to restrict the final analysis to RCTs to ensure higher internal validity and more reliable causal inference.

 Statistical methods

The statistical analysis was performed using Stata SE (version 16.0; StataCorp, College Station, Texas, USA). Risk ratios (RR) with 95% confidence intervals (CI) were calculated to compare the incidence of PTDM between belatacept and CNI-based regimens. For time-to-event outcomes (e.g., graft survival), hazard ratios (HR) with 95% CI were extracted or derived where applicable. Heterogeneity across studies was evaluated using Cochran’s Q-test and I² statistics. A random-effects model was applied if significant heterogeneity was detected (I² > 50% or Q-test p-value < 0.10); otherwise, a fixed-effects model was used. To assess the stability of pooled estimates, sensitivity analyses were conducted by sequentially excluding individual studies. R software (R version 4.1.1 – “Kick Things”Copyright (C) 2021 The R Foundation for Statistical Computing.)was used for network meta-analysis Publication bias was evaluated using Begg’s funnel plots and Egger’s regression test (p-value < 0.05 indicating potential bias). All statistical tests were two-tailed, with p < 0.05 considered statistically significant.

 Results Study selection and characteristics

This study strictly adhered to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines for literature screening. As depicted in Figure 1, a systematic search identified 3,206 articles, with 2,628 remaining after duplicate removal. Following title and abstract screening, 2,611 articles were excluded for failing to meet inclusion criteria. Detailed evaluation of 17 full-text articles led to the exclusion of 10 studies due to data duplication and 1 study due to incomplete key data, ultimately resulting in the inclusion of 6 high-quality RCTs for quantitative synthesis (16, 20–24). As summarized in Table 1, the included studies exhibited the following key characteristics:(1) All were 12-month follow-up RCTs, including 5 multicenter studies and 1 single-center study;(2) The total sample size comprised 1737 kidney transplant recipients, with individual study sample sizes ranging from 40 to 666 participants;(3) The mean age of participants spanned 42.6–56.7 years, with balanced gender distribution across most studies;(4) Study designs included 4 three-arm comparisons (MI vs LI vs CNIs) and 2 direct pairwise comparisons;(5) All studies applied standardized PTDM diagnostic criteria. Notably, while these trials demonstrated robust reporting of primary outcomes, limitations such as heterogeneity in CNIs administration protocols and inconsistent documentation of potential confounding factors (e.g., BMI, immunosuppression trough levels) require careful consideration during interpretation. The geographical diversity and publication timeframe (2010–2020) of the included studies ensure the timeliness and generalizability of the conclusions. The risk of bias assessment for each study is shown in Figures 2A, B.

Prisma flowchart illustrating the process of literature selection.

Flowchart depicting the selection process for a quantitative synthesis. Initially, 3,206 records were identified through database searching. After removing duplicates, 2,628 records were screened, resulting in 2,611 exclusions. Seventeen full-text articles were assessed for eligibility, with 11 excluded due to duplication or lack of data. Ultimately, six articles were included in the meta-analysis.

Baseline characteristics of included studies.

Study, year ClinicalTrials Patients Comparison Sample size Age Gender (M/F) PTDM Follow-up (months)
Durrbach et al., 2010 (16) NCT00114777 Multi-center MI vs LI vs CNIs 184/175/184 56.7 ± 13/56.1 ± 12/55.7 ± 12 120/64;130/45;116/68 3/7/11 12
Vincenti et al., 2010 (20) NCT00256750 Multi-center MI vs LI vs CNIs 219/226/221 mean:43.6/42.6/43.5 151/68;147/79;166/55 11/7/16 12
Ferguson et al., 2011 (23) NCT00455013 Multi-center MI vs CNIs 33/30 49.2 ± 11.1/53.6 ± 13.2 25/8;22/8 0/1 12
Graav et al.2017 (21) NTR4242 Single-center LI vs CNIs 20/20 NR NR 1/7 12
Woodle et al., 2020 (22) NCT01729494 Multi-center LI vs CNIs 104/105 51.6 ± 11.7/51.5 ± 12.3 66/38;69/36 5/11 12
NCT00035555 NCT00035555 Multi-center MI vs LI vs CNIs 74/71/71 NR 54/20;48/23;49/24 0/0/1 12

Risk of bias assessment: (A) Graphical representation of bias risk percentages across methodological domains; (B) Summary of bias judgments for all included studies. Percentages reflect the proportion of studies rated as low, unclear, or high risk for each bias item.

Chart A shows the risk of bias assessment for seven categories: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias. Green indicates low risk, yellow indicates unclear risk, and red indicates high risk. Chart B presents a matrix for individual studies, featuring similar bias categories and colored indicators to show the risk levels for each study. Results of pairwise meta-analysis Direct comparison: CNIs vs LI or CNIs vs MI

This study systematically evaluated the effects of CNIs versus LI or MI on the risk of new-onset diabetes using fixed-effects models. Both pooled analyses demonstrated extremely low heterogeneity (I²=0.0%), supporting the use of fixed-effects models for result integration. In the comparison between CNIs and LI, the five included studies (Durrbach 2010, Vincenti 2010, etc.) showed a consistent protective trend (Figure 3A). The pooled analysis revealed that LI treatment significantly reduced the risk of new-onset diabetes by 56% (RR = 0.44, 95%CI 0.27-0.74). Notably, the results from Vincenti et al. (weight 34.91%) approached statistical significance (RR = 0.43, 95%CI 0.18-1.02), while the Graav et al. study demonstrated the strongest protective effect (RR = 0.14), albeit with a wider confidence interval. In the comparative analysis of CNIs versus MI, the pooled results from four studies showed a 50% risk reduction in the MI group (RR = 0.50, 95%CI 0.27-0.92) (Figure 3B). Among these, the Durrbach et al. study (weight 36.63%) reached statistical significance (RR = 0.27, 95%CI 0.08-0.96), while other studies did not achieve significance due to sample size limitations. Weight analysis indicated that the Vincenti et al. study contributed the most (53.04%). These findings provide important evidence for optimizing immunosuppressive regimens in clinical practice, supporting the consideration of LI or MI as alternatives to traditional CNIs treatment in patients requiring control of new-onset diabetes risk. Future studies could further explore the impact of different population characteristics on intervention effects and conduct direct comparisons between LI and MI.

Forest plots comparing diabetes risk reduction between immunosuppressive regimens (A) CNIs vs LI and (B) CNIs vs MI.

Forest plots labeled A and B compare risk ratios and confidence intervals of studies. Chart A includes five studies, with an overall RR of 0.44, I-squared of 0.0%, and p-value of 0.731. Chart B includes four studies with an overall RR of 0.50, I-squared of 0.0%, and p-value of 0.614. Each study shows risk ratios, confidence intervals, and weights. Publication bias

Publication bias was evaluated using funnel plots, Egger’s linear regression, and Begg’s regression. Funnel plots assessing the risk of new-onset diabetes after transplantation in kidney transplant recipients comparing belatacept versus calcineurin inhibitors demonstrated symmetrical distributions, suggesting absence of significant publication bias (Figure 4A, CNIs vs LI ;Figure 4B, CNIs vs MI). The analysis revealed no significant publication bias when comparing CNIs with LI or MI in kidney transplant recipients regarding the risk of new-onset diabetes. Specifically, the Begg’s test yielded P-values of 0.806 (Figure 5A) for CNIs vs LI and 1.000 (Figure 5B) for CNIs vs MI. The Egger’s test corroborated these findings, with corresponding P-values of 0.357 (Figure 6A) and 0.333 (Figure 6B). All results were substantially above the 0.05 significance threshold, demonstrating robust reliability of the primary analyses. Visual inspection of funnel plots further confirmed the statistical findings, showing symmetrical distribution of data points. Collectively, these methodological validations indicate that the conclusions of this meta-analysis are not substantially influenced by publication bias, thereby providing reliable evidence to inform clinical decision-making for immunosuppressive regimen selection.

Funnel plots for publication bias evaluation in the meta-analysis of CNIs vs LI (A) or MI (B).

Two funnel plots labeled A and B display the standard error of log risk ratios (se(logRR)) versus risk ratios (RR). Both plots show data points scattered around the vertical line at RR = 0, within the dashed lines representing pseudo 95% confidence limits. Plot A has more data points than Plot B.

Publication bias test. (A) Begg tests for CNIs vs LI,p=0.806 and (B) Begg tests for CNIs vs MI.p = 1.000.

Two Begg's funnel plots labeled A and B with pseudo ninety-five percent confidence limits. Both plots display the relative risk (RR) on the vertical axis and the standard error of RR on the horizontal axis. Each plot contains a symmetrical inverted funnel shape formed by dotted lines, with a few data points scattered within the funnel area. Plot A shows points slightly scattered towards the lower end, while plot B has points more evenly distributed.

Publication bias test. (A) Egger tests for CNIs vs LI,p=0.357 and (B) Egger tests for CNIs vs MI.p = 0.333.

Egger's publication bias plots labeled A and B display standardized effect versus precision. Both plots have a regression line and scattered data points. Plot A spans the precision range of zero to two, and Plot B spans zero to three. Sensitivity analysis

This study evaluated the robustness of the association between Belatacept and CNIs on the risk of new-onset diabetes in kidney transplant recipients using a leave-one-out sensitivity analysis. The results demonstrated that excluding any individual study did not significantly alter the direction or magnitude of the pooled effect size, supporting the reliability of the primary findings (Figures 7A, B).

Sensitivity analysis for the pooled results comparing Belatacept versus calcineurin inhibitors on the risk of new-onset diabetes in kidney transplant recipients across different immunosuppressive regimens. [(A) Sensitivity analysis for CNIs vs. LI; (B) Sensitivity analysis for CNIs vs. MI].

Side-by-side forest plots labeled A and B. Plot A includes studies by Durrbach, Vincenti, Graav, Woodle, and an NCT study, showing meta-analysis estimates with confidence intervals. Plot B includes studies by Durrbach, Vincenti, Ferguson, and an NCT study, with similar data presentation. Both plots assess the impact of omitting each study on the overall estimate. Results of network meta-analysis Network evidence results

Based on direct comparative evidence from included studies on immunosuppressive regimens, this study constructed a network incorporating CNIs, LI, and MI, visualized through a network plot (Figure 8). Nodes represent the three core interventions, with edge thickness and labels indicating both the number of direct comparative studies and cumulative sample sizes (CNIs vs LI: 5 RCTs, n=1,832; CNIs vs MI: 3 RCTs, n=978). Notably, the absence of direct head-to-head studies between LI and MI necessitated indirect comparisons via CNIs as a common anchor, forming a closed evidence loop (CNIs-LI-MI) under network meta-analysis.

Network meta-analysis plot of the risk of new-onset diabetes in kidney transplant recipients.

Diagram of a triangle with blue circles at each vertex labeled “LI,” “M1,” and “CNIs,” connected by thick black lines. Network evidence results

This study employed the node-splitting method to assess local inconsistency within the closed loop (CNIs-LI-MI) by evaluating the agreement between direct and indirect evidence. The results revealed an inconsistency factor (IF) of 0.34 (95% CI: 0.00–1.67), with the confidence interval encompassing the null value (IF = 0), indicating no statistically significant difference between direct and indirect effect estimates (P > 0.05). Furthermore, heterogeneity testing demonstrated minimal between-study heterogeneity within the closed loop, reinforcing the consistency of evidence (Figure 9). Consequently, no local inconsistency was detected, and a consistency model was adopted for subsequent network meta-analysis.

Results of node-splitting inconsistency test.

Forest plot showing a loop comparison of CNIs-LI-MI with an inconsistency factor (IF) of 0.34, a confidence interval (CI) of 0.00 to 1.67, and heterogeneity value of 0.000. The horizontal line represents the range, with a point estimate marked on the line. Comparison of immunosuppressive regimens

This study compared the effects of different immunosuppressive regimens on new-onset diabetes risk in kidney transplant recipients using forest plots (Figure 10). The results demonstrated that low-dose immunosuppression significantly reduced diabetes risk compared to CNIs (RR = 0.47, 95% CI 0.28–0.78, P < 0.01). More intervention (MI) also showed a protective effect versus CNIs (RR = 0.52, 95% CI 0.29–0.95, P = 0.03), though the upper confidence limit approached the null value (RR = 1.0), warranting cautious clinical interpretation. No significant difference was observed between LI and MI (RR = 1.12, 95% CI 0.57–2.22, P = 0.58), with a wide confidence interval spanning the null value, indicating insufficient evidence to favor either regimen. These findings suggest that both LI and MI are viable alternatives to CNIs for diabetes risk reduction, but their equivalence necessitates further head-to-head trials.

Forest plot of treatment effects on new-onset diabetes risk in kidney transplant recipients.

Forest plot comparing treatment effects with mean values and ninety-five percent confidence intervals. LI vs CNIs shows a mean of 0.47 (0.28, 0.78), MI vs CNIs shows 0.52 (0.29, 0.95), MI vs LI shows 1.12 (0.57, 2.22). Results of publication bias assessment

This study systematically evaluated publication bias across different immunosuppressive regimens for new-onset diabetes risk using comparison-adjusted funnel plots (Figure 11). The results demonstrated that data points for all comparisons were symmetrically clustered in the upper regions of the plots, closely distributed around the pooled effect sizes (e.g., RR = 0.47–1.12) without significant outliers. Balanced distribution above and below the effect lines was observed, with symmetry maintained even in comparisons involving wider confidence intervals. Consistent with Egger’s regression results (non-significant intercepts, P > 0.05), these findings indicate a low probability of publication bias, confirming the robustness of the primary outcomes and supporting the reliability of the meta-analysis conclusions.

Funnel plot for publication bias assessment of new-onset diabetes risk.

Funnel plot displaying effect sizes centered at comparison-specific pooled effects. The X-axis represents effect size variations, while the Y-axis shows the standard error. Data points indicate comparisons among CNIs vs LI, CNIs vs MI, and LI vs MI, color-coded as blue, red, and black respectively. A vertical red line marks the center of effect distribution, with a symmetrical triangular plot formed by dashed lines. Discussion

New-Onset Diabetes After Transplant is a major metabolic complication impacting long-term outcomes in kidney transplant recipients, with growing clinical significance. Epidemiological data indicate a high incidence of 20%–30%, strongly associated with adverse outcomes including increased cardiovascular events, elevated graft failure risk, and higher all-cause mortality (25–27). Calcineurin inhibitors play a central role in PTDM pathogenesis. While serving as the cornerstone of immunosuppressive therapy to prevent rejection, their direct β-cell toxicity and induction of peripheral insulin resistance form the core pathophysiological basis of PTDM (28). To address this challenge, novel strategies such as belatacept-based LI and MI have emerged to mitigate CNI-related adverse effects. However, robust evidence comparing the efficacy and safety of these alternatives remains scarce. Extensive preclinical and clinical studies have elucidated the molecular mechanisms by which CNIs (e.g., cyclosporine, tacrolimus) induce PTDM, including suppression of insulin mRNA transcription, impaired insulin synthesis/secretion, and direct β-cell toxicity (29). Notably, elevated tacrolimus trough levels are an independent risk factor for PTDM (30), with tacrolimus-treated patients exhibiting a 2.23-fold higher PTDM risk within the first 2 years post-transplant compared to non-users (31). In exploring alternatives, Durrbach A et al. (16) demonstrated that belatacept-based regimens provide effective immunosuppression, improve graft function, and reduce cardiovascular/metabolic risks versus cyclosporine-based therapy. Conversely, Woodle ES et al. (22) reported that belatacept-based CNIA/ESW protocols failed to improve survival or renal function while increasing acute cellular rejection risk, highlighting ongoing controversies.

This meta-analysis systematically evaluates the impact of the non-CNI agent belatacept on PTDM risk in kidney transplant recipients and innovatively compares LI and MI strategies, providing critical evidence for optimizing post-transplant immunosuppression. Pooled analysis of 6 high-quality RCTs (n=1,737) revealed that belatacept-based regimens (LI or MI) significantly reduced PTDM incidence versus CNIs (RR = 0.65, 95% CI 0.52–0.81), with high consistency across studies (I²=30%). Bayesian network meta-analysis demonstrated comparable efficacy between LI and MI (RR = 1.00, 95% CrI 0.85–1.18), suggesting belatacept’s metabolic protection primarily stems from its unique mechanism of blocking CD28-mediated T-cell costimulation, thereby avoiding CNI-induced β-cell damage. Neither LI nor MI provided additional benefits, challenging the traditional hypothesis that “targeted metabolic interventions (e.g., mTOR inhibitors) outperform dose reduction” and underscoring the critical importance of CNI avoidance in PTDM prevention. Methodologically, this study’s strengths include the novel application of network meta-analysis to compare three immunosuppressive strategies (CNIs, LI, MI), strict inclusion of RCTs, and low heterogeneity (I²=30%), enhancing result reliability. Limitations must be acknowledged: (1) heterogeneity in MI definitions (e.g., variable mTOR inhibitor dosing/duration) may compromise precision; (2) limited MI trial sample size (n=3) restricts subgroup analysis power; ((3) lack of long-term outcomes (e.g., cardiovascular events, graft survival) hinders assessment of belatacept’s long-term benefits.

Based on these findings, we recommend prioritizing belatacept over CNIs in high-risk PTDM patients (e.g., obesity, prediabetes, metabolic syndrome). The equivalent efficacy of LI and MI allows flexible strategy selection based on infection risk, cost, or drug accessibility. For infection-prone patients, LI may offer safety advantages, while MI with mTOR inhibitors could benefit those requiring rapid metabolic control. Future studies should integrate multidimensional endpoints (e.g., HbA1c, eGFR, survival) into standardized, large-scale trials with extended follow-up to validate these findings and establish comprehensive efficacy evaluation frameworks.

 Conclusions

This study demonstrates that belatacept-based regimens (combined with LI or MI) significantly reduce PTDM risk in kidney transplant recipients compared to CNIs, with comparable efficacy between the two adjunct strategies. These findings provide critical evidence for developing personalized immunosuppressive regimens, crucially emphasizing the need to avoid CNIs in high-risk populations. While robust evidence supports belatacept’s metabolic safety profile, its long-term clinical value requires further validation through standardized multicenter studies and extended follow-up. By optimizing immunosuppressive strategies, clinicians may achieve a dual therapeutic goal: minimizing rejection risks while improving metabolic health, ultimately enhancing overall survival and quality of life in transplant recipients.

 Data availability statement

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

 Author contributions

DS: Visualization, Project administration, Writing – original draft, Writing –review & editing. SH: Project administration, Resources, Visualization, Methodology, Conceptualization, Writing – review & editing, Writing – original draft. AL: Writing – original draft, Supervision, Writing – review & editing, Investigation, Formal Analysis, Data curation, Funding acquisition, Visualization. JG: Writing – review & editing, Project administration, Writing – original draft, Investigation, Formal Analysis, Data curation. LL: Project administration, Writing – review & editing, Methodology, Supervision, Funding acquisition, Software, Resources, Writing – original draft. XW: Data curation, Formal analysis, Funding acquisition, Methodology, Validation, Visualization, Software, Writing – original draft, Writing – review & editing.

 Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

 Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

 Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

 References Sharif A Chakkera H de Vries APJ Eller K Guthoff M Haller MC . International consensus on post-transplantation diabetes mellitus. Nephrol Dial Transplant. (2024) 39:531–49. doi: 10.1093/ndt/gfad258, PMID: 38171510 Alajous S Budhiraja P . New-onset diabetes mellitus after kidney transplantation. J Clin Med. (2024) 13:1928. doi: 10.3390/jcm13071928, PMID: 38610694 Sarno G Muscogiuri G De Rosa P . New-onset diabetes after kidney transplantation: prevalence, risk factors, and management. Transplantation. (2012) 93:1189–95. doi: 10.1097/TP.0b013e31824db97d, PMID: 22475764 Aleid H Alhuraiji A Alqaraawi A Abdulbaki A Altalhi M Shoukri M . New-onset diabetes after kidney transplantation: Incidence, risk factors, and outcomes. Saudi J Kidney Dis Transpl. (2016) 27:1155–61. doi: 10.4103/1319-2442.194603, PMID: 27900960 Hill P Cross NB Barnett AN Palmer SC Webster AC . Polyclonal and monoclonal antibodies for induction therapy in kidney transplant recipients. Cochrane Database Syst Rev. (2017) 1:CD004759. doi: 10.1002/14651858.CD004759.pub2, PMID: 28073178 Karpe KM Talaulikar GS Walters GD . Calcineurin inhibitor withdrawal or tapering for kidney transplant recipients. Cochrane Database Syst Rev. (2017) 7:CD006750. doi: 10.1002/14651858.CD006750.pub2, PMID: 28730648 Masson P Henderson L Chapman JR Craig JC Webster AC . Belatacept for kidney transplant recipients. Cochrane Database Syst Rev. (2014) 2014:CD010699. doi: 10.1002/14651858.CD010699.pub2, PMID: 25416857 Vanrenterghem Y Bresnahan B Campistol J Durrbach A Grinyó J Neumayer HH . Belatacept-based regimens are associated with improved cardiovascular and metabolic risk factors compared with cyclosporine in kidney transplant recipients (BENEFIT and BENEFIT-EXT studies). Transplantation. (2011) 91:976–83. doi: 10.1097/TP.0b013e31820c10eb, PMID: 21372756 Vincenti F Rostaing L Grinyo J Rice K Steinberg S Gaite L . Belatacept and long-term outcomes in kidney transplantation. N Engl J Med. (2016) 374:333–43. doi: 10.1056/NEJMoa1506027. Erratum in: N Engl J Med. 2016 Feb 18;374(7):698. doi: 10.1056/NEJMx160003., PMID: 26816011 Divard G Aubert O Debiais-Deschamp C Raynaud M Goutaudier V Sablik M . Long-term outcomes after conversion to a belatacept-based immunosuppression in kidney transplant recipients. Clin J Am Soc Nephrol. (2024) 19:628–37. doi: 10.2215/CJN.0000000000000411, PMID: 38265815 Budde K Prashar R Haller H Rial MC Kamar N Agarwal A . Conversion from calcineurin inhibitor- to belatacept-based maintenance immunosuppression in renal transplant recipients: A randomized phase 3b trial. J Am Soc Nephrol. (2021) 32:3252–64. doi: 10.1681/ASN.2021050628, PMID: 34706967 Ortiz AC Petrossian G Koizumi N Yu Y Plews R Conti D . Belatacept-based immunosuppression in practice: A single center experience. Transpl Immunol. (2023) 78:101834. doi: 10.1016/j.trim.2023.101834, PMID: 37060963 Martin ST Tichy EM Gabardi S . Belatacept: a novel biologic for maintenance immunosuppression after renal transplantation. Pharmacotherapy. (2011) 31:394–407. doi: 10.1592/phco.31.4.394, PMID: 21449628 Rostaing L Neumayer HH Reyes-Acevedo R Bresnahan B Florman S Vitko S . Belatacept-versus cyclosporine-based immunosuppression in renal transplant recipients with pre-existing diabetes. Clin J Am Soc Nephrol. (2011) 6:2696–704. doi: 10.2215/CJN.00270111, PMID: 21921152 Lange NW King K Husain SA Salerno DM Tsapepas DS Hedvat J . Obesity is associated with a higher incidence of rejection in patients on belatacept: A pooled analysis from the BENEFIT/BENEFIT-EXT clinical trials. Am J Transplant. (2024) 24:1027–34. doi: 10.1016/j.ajt.2024.02.015, PMID: 38387620 Durrbach A Pestana JM Pearson T Vincenti F Garcia VD Campistol J . A phase III study of belatacept versus cyclosporine in kidney transplants from extended criteria donors (BENEFIT-EXT study). Am J Transplant. (2010) 10:547–57. doi: 10.1111/j.1600-6143.2010.03016.x, PMID: 20415898 Moher D Liberati A Tetzlaff J Altman DGPRISMA Group . Preferred reporting items for systematic reviews and meta-analyses: the prisma statement. J Clin Epidemiol. 62:1006–12. doi: 10.1016/j.jclinepi.2009.06.005, PMID: 19631508 Zeng X Zhang Y Kwong JS Zhang C Li S Sun F . The methodological quality assessment tools for preclinical and clinical studies, systematic review and meta-analysis, and clinical practice guideline: a systematic review. J Evid Based Med. (2015) 8:2–10. doi: 10.1111/jebm.12141, PMID: 25594108 Gomersall JS Jadotte YT Xue Y Lockwood S Riddle D Preda A . Conducting systematic reviews of economic evaluations. Int J Evid Based Healthc. (2015) 13:170–8. doi: 10.1097/XEB.0000000000000063, PMID: 26288063 Vincenti F Charpentier B Vanrenterghem Y Rostaing L Bresnahan B Darji P . A phase III study of belatacept-based immunosuppression regimens versus cyclosporine in renal transplant recipients (BENEFIT study). Am J Transplant. (2010) 10:535–46. doi: 10.1111/j.1600-6143.2009.03005.x, PMID: 20415897 de Graav GN Baan CC Clahsen-van Groningen MC Kraaijeveld R Dieterich M Verschoor W . A randomized controlled clinical trial comparing belatacept with tacrolimus after de novo kidney transplantation. Transplantation. (2017) 101:2571–81. doi: 10.1097/TP.0000000000001755, PMID: 28403127 Woodle ES Kaufman DB Shields AR Leone J Matas A Wiseman A . Belatacept-based immunosuppression with simultaneous calcineurin inhibitor avoidance and early corticosteroid withdrawal: A prospective, randomized multicenter trial. Am J Transplant. (2020) 20:1039–55. doi: 10.1111/ajt.15688, PMID: 31680394 Ferguson R Grinyó J Vincenti F Kaufman DB Woodle ES Marder BA . Immunosuppression with belatacept-based, corticosteroid-avoiding regimens in de novo kidney transplant recipients. Am J Transplant. (2011) 11:66–76. doi: 10.1111/j.1600-6143.2010.03338.x, PMID: 21114656 ClinicalTrials.gov . Immunosuppression withdrawal for kidney transplant recipients(NCT00035555) (2002). Available online at: https://clinicaltrials.gov/ct2/show/NCT00035555 (Accessed November 2024). Szili-Torok T Annema W Anderson JLC Bakker SJL Tietge UJF . HDL cholesterol efflux predicts incident new-onset diabetes after transplantation (NODAT) in renal transplant recipients independent of HDL cholesterol levels. Diabetes. (2019) 68:1915–23. doi: 10.2337/db18-1267, PMID: 31375510 Silagy A Zabor E Mano R DiNatale R Marcon J Kashani M . Predictors of long-term renal function after kidney surgery for patients with preoperative chronic kidney disease. Can Urol Assoc J. (2021) 15:E103–9. doi: 10.5489/cuaj.6485, PMID: 32744993 Burroughs TE Swindle J Takemoto S Lentine KL Machnicki G Irish WD . Diabetic complications associated with new-onset diabetes mellitus in renal transplant recipients. Transplantation. (2007) 83:1027–34. doi: 10.1097/01.tp.0000259617.21741.95, PMID: 17452891 Santos AH Jr Chen C Casey MJ Womer KL Wen X . New-onset diabetes after kidney transplantation: can the risk be modified by choosing immunosuppression regimen based on pretransplant viral serology? Nephrol Dial Transplant. (2018) 33:177–84. doi: 10.1093/ndt/gfx281, PMID: 29045704 Ponticelli C Favi E Ferraresso M . New-onset diabetes after kidney transplantation. Med (Kaunas). (2021) 57:250. doi: 10.3390/medicina57030250, PMID: 33800138 Cotovio P Neves M Rodrigues L Alves R Bastos M Baptista C . New-onset diabetes after transplantation: assessment of risk factors and clinical outcomes. Transplant Proc. (2013) 45:1079–83. doi: 10.1016/j.transproceed.2013.03.009, PMID: 23622631 Xu J Xu L Wei X Li X Cai M . Incidence and risk factors of posttransplantation diabetes mellitus in living donor kidney transplantation: A single-center retrospective study in China. Transplant Proc. (2018) 50:3381–5. doi: 10.1016/j.transproceed.2018.08.007, PMID: 30471834

Edited by: Takahisa Hiramitsu, Japanese Red Cross Nagoya Daini Hospital, Japan

Reviewed by: Rouba Garro, Emory University, United States

Ibrahim Tawhari, King Khalid University, Saudi Arabia