Baseline characteristics of study population

Table 1 presents a comprehensive comparison of various recipient, donor, immunosuppressant, and operative variables between OKT and RAKT both before and after PSM. Before matching, the mean ages of recipients were 50.51 years for the OKT group and 47.17 years for the RAKT group (p = 0.154, SMD = 0.3256). After matching, the mean ages were 48.56 years (OKT) and 47.96 years (RAKT) (p = 0.810, SMD = 0.0063). Gender distribution was 46.19% females (OKT) and 41.38% (RAKT) before matching (p = 0.626, SMD = 0.0977), and 48.9% vs. 42.3% after matching (p = 0.692, SMD = 0.0911). BMI averaged 23.24 kg/m² (OKT) and 24.23 kg/m² (RAKT) before matching (p = 0.300, SMD = 0.2038), and 23.66 kg/m² vs. 24.16 kg/m² after matching (p = 0.580, SMD = 0.0144). Dialysis duration showed a significant initial difference with medians of 5 months (OKT) vs. 1 month (RAKT) (p = 0.021, SMD = 0.1572), which reduced to 4 months vs. 1.5 months after matching (p = 0.152, SMD = 0.0225). Retransplantation rates were 8.57% (OKT) and 0% (RAKT) before matching (p = 0.140, SMD = 0.326), with alignment after matching. The number of HLA mismatches was similar between OKT and RAKT before matching, with a median of 4 (range 3–5) in both groups (p = 0.993, SMD = 0.0081). After matching, the median remained 4 (range 3–5) in the OKT group and 3 (range 3–5) in the RAKT group, with no significant difference (p = 0.705, SMD = 0.0992). IgM titer levels were comparable between the groups before matching (median 64, range 32–128 in OKT vs. median 32, range 16–64 in RAKT, p = 0.094, SMD = 0.9886) and after matching (median 64, range 32–128 in OKT vs. median 32, range 16–64 in RAKT, p = 0.209, SMD = 0.1660). The causes of ESRD were analyzed between patients undergoing OKT and RAKT before PSM. HTN caused ESRD in 9.5% of OKT patients and 10.3% of RAKT patients, while diabetes mellitus (DM) was more frequent in RAKT (34.5%) than in OKT (28.6%). Glomerulonephritis (GN) accounted for 9.5% of OKT cases and 13.8% in RAKT. IgA nephropathy was found in 18.6% of OKT and 20.7% of RAKT patients. Focal segmental glomerulosclerosis (FSGS) occurred in 2.9% of OKT and 3.5% of RAKT patients. Polycystic kidney disease (PCKD) was seen in 7.1% of OKT cases but absent in RAKT. Other causes such as Alport’s syndrome, systemic lupus nephritis, congetinal renal agenesis, nephronophthisis, obstructive uropathy, reflux nephropathy and horseshoe kidney accounted for 6.2% of OKT and 6.9% of RAKT cases. Overall, the distribution of causes of ESRD between OKT and RAKT patients did not show statistically significant differences (p = 0.763).

Table 1 Baseline characteristics before and after propensity score matching

The use of induction agents like basiliximab and thymoglobulin showed no significant difference before (p = 0.795, SMD = 0.0971) and after matching (p = 1, SMD = 0.0765), with 82.86% of OKT recipients and 86.21% of RAKT recipients initially receiving basiliximab, and 85.5% vs. 84.6% after matching. Calcineurin inhibitor use, mainly tacrolimus, was consistent before (p = 0.479, SMD = 0.0846) and after matching (p = 1, SMD = 0.0527), with nearly all recipients treated with tacrolimus (98.1% OKT vs. 96.55% RAKT before; 97.7% OKT vs. 96.2% RAKT after).

Donor mean age was 51.02 years for OKT and 49.14 years for RAKT before matching (p = 0.401, SMD = 0.1453), and 49.9 vs. 50 years after matching (p = 0.968, SMD = 0.0281). Gender distribution was 58.57% females (OKT) and 68.97% (RAKT) before matching (p = 0.285, SMD = 0.2247), and 56.5% vs. 65.4% after matching (p = 0.534, SMD = 0.0758). Donor BMI showed no significant difference before (24.40 kg/m² OKT vs. 23.89 kg/m² RAKT, p = 0.424, SMD = 0.1329) and after matching (24.23 kg/m² vs. 23.89 kg/m², p = 0.637, SMD = 0.0452). Other donor characteristics, including DM, HTN, and renal function (24-hour creatinine clearance and proteinuria), were balanced before and after matching.

Perioperative outcomes

Operative times were significantly longer for RAKT recipients (Table 1). Before matching, cold ischemic time was 48.57 min for OKT vs. 116.83 min for RAKT (p < 0.0001), and 47.22 min vs. 119.77 min after matching (p < 0.001). Anastomosis and total operative times were also longer for RAKT both before (21.09 vs. 45.41 min for anastomosis, p < 0.0001; 152.91 vs. 296.83 min for operative time, p < 0.0001) and after matching (29.49 vs. 45.35 min for anastomosis, p < 0.001; 150.81 vs. 291.42 min for operative time, p < 0.001). Hospitalization was shorter for RAKT recipients both before (median 6 days vs. 8 days, p < 0.0001) and after matching (median 6 days vs. 8 days, p < 0.001). Post-transplant renal function, measured by eGFR (CKD-EPI) one month after KT, was similar between the groups (68.23 ml/min/1.73 m² in OKT vs. 71.28 ml/min/1.73 m² in RAKT before matching, p = 0.426; 68.57 vs. 71.81 ml/min/1.73 m² after matching, p = 0.439). Notably, there were no cases of delayed graft function (DGF) observed in either the RAKT or OKT groups.

Univariate and multivariate analysis for a composite of BPAR, graft failure, and de novo DSA

In the univariate analysis with a composite outcome of BPAR, graft failure, and de novo DSA, the type of surgery did not significantly affect the composite outcome (Table 2). However, HLA-incompatible kidney transplants were significantly associated with the outcome, demonstrating a hazard ratio (HR) of 3.75 (95% CI: 1.744–8.074, p = 0.0007). The multivariable analysis indicates that this association might be significant, with an HR of 2.889 (95% CI: 0.974–8.564, p = 0.0557), suggesting a potential differential impact of HLA incompatibility across transplant groups. Pre-transplant DSA also displayed a significant association in the univariate analysis (HR 3.092, 95% CI: 1.475–6.482, p = 0.0028), but this significance was not maintained in the multivariate analysis (HR 2.161, 95% CI: 0.842–5.547, p = 0.1091). Additionally, thymoglobulin induction was significantly associated with the composite outcome in the univariate analysis (HR 2.527, 95% CI: 1.15–5.553, p = 0.021), but this association was not significant in the multivariate analysis, indicating that thymoglobulin’s effect might be influenced by additional variables. Lastly, no donor factors showed significant association with the composite outcome. Overall, these analyses suggest that factors such as HLA incompatibility and pre-transplant DSA have significant associations in univariate contexts, but their impacts become more nuanced in multivariate analyses.

Table 2 Univariate and multivariate analysis of factors associated with a composite of graft failure, BPAR and de novo DSAComparison of biopsy-proven acute rejection-free graft survival and renal function between RAKT and OKT recipients

To compare BPAR-free graft survival between RAKT and OKT recipients, Kaplan-Meier curves were used (Fig. 1). Both groups showed similar BPAR-free graft survival rates over the observation period. The RAKT group had a BPAR-free survival rate of 92.4% at 1 and 2 years, while the OKT group had rates of 93.1% at 1 year and 91.9% at 2 years, with no statistically significant difference (p = 0.99).

Fig. 1

Kaplan-Meir survival curve for biopsy-proven acute rejection free graft survival

We performed a PSM analysis to further assess the impact of RAKT versus OKT on BPAR-free survival and composite outcomes, including recipient and donor characteristics, immunosuppressant usage, and operative details for balanced comparisons.

Table 3 shows the adjusted relative hazard for BPAR-free survival and composite outcomes in both groups, before and after PSM. Initially, the hazard ratio (HR) for BPAR-free survival in the RAKT group compared to the OKT group was 1.007 (95% CI: 0.229, 4.431, p = 0.993), indicating no significant difference. After PSM, the HR for BPAR-free survival in the RAKT group was 0.858 (95% CI: 0.180, 4.096, p = 0.848), suggesting no significant difference between RAKT and OKT in terms of BPAR-free survival when matched for relevant covariates.

Table 3 The adjusted relative hazard for BPAR-free survival and composite graft failure in robot-assisted versus open surgery for ABO-incompatible kidney transplants

For the composite outcome of graft failure, BPAR, and de novo DSA-free survival, the HR for the RAKT group in the original unadjusted data was 0.982 (95% CI: 0.224, 4.307, p = 0.981), showing no significant difference. After matching, the HR was 0.858 (95% CI: 0.180, 4.096, p = 0.848), still indicating no significant difference between RAKT and OKT. These PSM results confirm that the risks of BPAR-free survival and composite outcomes are comparable between the groups, supporting the equivalence of these surgical techniques in post-transplant outcomes.

Figure 2 shows the overall trend in eGFR values determined via the CKD-EPI Creatinine Equation, with similar patterns between RAKT and OKT groups (p = 0.20). The mean eGFR values were 62.15 ml/min/1.73 m² for the RAKT group and 64.53 ml/min/1.73 m² for the OKT group.

Fig. 2

Trends of eGFR values over follow-up periods up to 1 year after transplantation in matched cohort

Table 4 provides a detailed comparison of post-transplant eGFR values at various time points between OKT and RAKT recipients before and after PSM. Before matching, the mean eGFR at discharge was 76.40 (± 20.79) ml/min/1.73 m² for OKT and 74.48 (± 18.91) ml/min/1.73 m² for RAKT (p = 0.638). At 1 month post-transplant, the mean eGFR values were 68.23 (± 19.55) ml/min/1.73 m² for OKT and 71.28 (± 16.95) ml/min/1.73 m² for RAKT (p = 0.426). At 6 months, the values were 62.91 (± 17.44) ml/min/1.73 m² for OKT and 62.03 (± 17.26) ml/min/1.73 m² for RAKT (p = 0.799). At 12 months, the mean eGFR values were 64.17 (± 19.61) ml/min/1.73 m² for OKT and 62.38 (± 14.85) ml/min/1.73 m² for RAKT (p = 0.636).

Table 4 Comparison of post-transplant estimated glomerular filtration rate (CKD-EPI) at various time points between OKT and RAKT before and after propensity score matching

After matching, the comparison remained consistent with no significant differences observed. At discharge, the mean eGFR was 74.50 (± 20.93) ml/min/1.73 m² for OKT and 74.12 (± 18.50) ml/min/1.73 m² for RAKT (p = 0.931). At 1 month, the mean eGFR was 68.57 (± 19.80) ml/min/1.73 m² for OKT and 71.81 (± 17.41) ml/min/1.73 m² for RAKT (p = 0.440). At 6 months, the values were 63.09 (± 18.37) ml/min/1.73 m² for OKT and 61.58 (± 18.15) ml/min/1.73 m² for RAKT (p = 0.701). At 12 months, the mean eGFR was 64.53 (± 20.01) ml/min/1.73 m² for OKT and 62.15 (± 15.33) ml/min/1.73 m² for RAKT (p = 0.568).

Infectious complications

We also performed a comparative analysis of the etiologies behind post-transplant infections necessitating hospital admission (see Additional Data 1 for details). For urinary tract infections, the OKT group had diverse organisms: E. coli (9.5%), K. pneumoniae (5.2%), Klebsiella aerogenes (2.4%), Enterococci (3.3%), and Proteus mirabilis (1%). The RAKT group had lower incidences: E. coli (6.9%) and K. pneumoniae (13.8%). Pneumonia from Covid-19 was noted in 2.4% of the OKT group. Cytomegalovirus syndrome was exclusive to the OKT group (1.4%). Bacteremia in the OKT group included E. coli (2.4%) and Bacteroides fragilis (0.5%); in the RAKT group, E. coli and K. pneumoniae (3.4%, each). Postoperative infections in the OKT group involved various organisms, including K. pneumoniae, Streptococcus parasanguinis, Staphylococcus aureus, Enterococcus faecalis, E. coli, B. fragilis, and Candida albicans (0.5–1% each). Fungal infections (candidiasis) were found in 1% of the OKT group. Gastrointestinal infections in the OKT group included Sapovirus (0.5%) and Clostridium difficile (1%).

The forest plot in Fig. 3 illustrated the risk reduction for various infections with RAKT compared to Open KT. For urinary tract infections, the risk reduction was − 2.12% (95% CI: -17.77, 13.54). Pneumonia had a risk reduction of 2.38% (95% CI: 0.32, 4.44), viral infections 1.43% (95% CI: -0.18, 3.03), and bacteremia of unknown origin − 0.59% (95% CI: -7.60, 6.42). Postsurgical and fungal infections had risk reductions of 0.84% (95% CI: -6.35, 8.02) and 0.95% (95% CI: -0.36, 2.27) respectively. Gastrointestinal infections had a risk reduction of 1.43% (95% CI: -0.18, 3.03). None of these differences were statistically significant, indicating that RAKT does not substantially change the risk of postoperative infections compared to OKT.

Fig. 3

Forest plot showing risk reduction in infectious complications for RAKT compared to OKT