In patients with end-stage liver disease, portal hypertension frequently leads to the formation of spontaneous portosystemic shunts, including coronary, periumbilical, and splenorenal shunts (SRS). Among these, large SRSs can divert portal blood flow away from the liver graft after living donor liver transplantation (LDLT), resulting in portal steal syndrome. This condition may compromise graft regeneration and function, potentially leading to graft failure. Therefore, intraoperative interruption of large shunts is essential to ensure adequate portal inflow and graft survival [1–3].

The following two major surgical approaches are used to interrupt SRSs: Direct splenorenal shunt ligation (SRSL) and left renal vein ligation (LRVL) [4, 5]. SRSL involves anatomically precise identification and ligation of the shunt vessel itself, but often present technical challenges during the procedure (particularly in patients with severe adhesions or complex vascular anatomy) and carries a risk of bleeding from surrounding tissues or from the shunt itself. In contrast, LRVL achieves functional interruption of the SRS by ligating the left renal vein (LRV), and is often preferred due to its simplicity and ease of access.

Our institution previously reported the safety and feasibility of LRVL in LDLT, and several subsequent studies have confirmed its effectiveness in preventing portal steal syndrome and maintaining stable portal perfusion [1, 4]. However, concerns persist regarding the potential impact of LRVL on renal function. Because the LRV is the major venous drainage pathway for the left kidney, its ligation may lead to venous congestion and impaired renal perfusion. Although previous nontransplant studies have suggested that LRV ligation has minimal long-term effects on renal function—because collateral pathways such as the left gonadal vein can provide sufficient outflow—the applicability of these findings to transplant recipients remains to be evaluated [6–9].

To date, most studies on LRVL in LDLT have focused on surgical feasibility and short-term safety, with few addressing long-term renal outcomes. Moreover, there is a lack of direct comparisons between SRSL and LRVL in the transplant setting, especially using methodologies that adjust for baseline differences in renal function and comorbidities.

We aimed to compare the long-term outcomes of patients undergoing either LRVL or SRSL during LDLT, focusing on graft-related endpoints such as shunt recanalization, graft failure, and patient survival, as well as changes in renal function. To account for baseline differences, propensity score matching was applied, and renal function was assessed over a 10-year follow-up period in a single-center setting using a consistent immunosuppressive protocol.

Effective management of large spontaneous SRSs during LDLT is essential to ensure adequate portal inflow and prevent portal steal syndrome. Two intraoperative techniques are currently employed for this purpose: SRSL, which anatomically targets the shunt itself, and LRVL, which functionally interrupts shunt outflow by occluding the LRV. SRSL provides anatomically precise control but can be technically demanding in patients with dense adhesions or distorted venous anatomy. In contrast, LRVL is less technically challenging and is often preferred in hemodynamically unstable or surgically complex cases [1–3].

In this study, both techniques were effective in preventing portal steal and maintaining adequate graft perfusion. Patient and graft survival, re-transplantation rates, and shunt recanalization were comparable between groups, supporting LRVL as a feasible and effective alternative when direct shunt ligation is difficult or deemed unsafe.

Nevertheless, concerns remain regarding the potential long-term renal consequences of LRVL. Because the LRV is the primary venous drainage route of the left kidney, its ligation may predispose patients to impaired renal perfusion and progressive renal dysfunction, particularly in LDLT recipients who require lifelong exposure to nephrotoxic immunosuppressants such as tacrolimus. Although previous nontransplant studies have suggested minimal long-term renal impact after LRV ligation, these findings may not be fully applicable to liver transplant recipients under chronic immunosuppression. Beyond simple mechanical obstruction of venous outflow, progressive renal dysfunction after LRVL may involve multiple pathophysiological mechanisms, including microvascular injury, altered renal autoregulation, and synergistic nephrotoxicity associated with prolonged calcineurin inhibitor exposure. These mechanisms may partly explain the greater variability in creatinine levels and the exclusive occurrence of dialysis in the LRVL group during long-term follow-up.

To evaluate renal outcomes more rigorously, we performed 1:1 propensity score matching using the baseline values for relevant variables. After matching, long-term tacrolimus exposure was comparable between the groups, although the LRVL group demonstrated a tendency toward earlier tacrolimus tapering during the early postoperative period, likely reflecting concerns by the clinicians about renal function. Despite this early tapering, the LRVL group exhibited a sustained trend toward higher serum creatinine throughout follow-up, and all cases of renal failure requiring dialysis occurred exclusively in the LRVL groups between 5 and 10 years post-transplant. Although not definitive, these findings raise the possibility that LRVL may be associated with a subtle but progressive long-term renal burden.

These observations align with prior experience at our institution. In an early report, Lee et al. found no short-term deterioration in renal function, proteinuria, or hematuria following LRVL [4]. However, a 20-year follow-up study by Hwang et al. in the same patient cohort revealed that 9.1% of them eventually required dialysis, and more than half experienced a decline in the estimated glomerular filtration rate [12]. Although underlying comorbidities such as diabetes or glomerulonephritis likely contributed to this outcome, chronic venous congestion from LRV ligation may also have played a role in the progressive renal decline. These longer-term findings underscore the importance of careful patient selection and appropriate renal monitoring following LRVL.

Our study provides follow-up data up to 10 years after transplantation. While no statistically significant differences in renal failure were observed during this interval, the trends in creatinine elevation and the exclusive occurrence of cases requiring dialysis in the LRVL group suggest that more pronounced differences may emerge with a more extended follow-up, as suggested by the long-term data reported by a Hwang et al. [12] This highlights the need for continued surveillance and emphasizes the importance of choosing surgical strategies that minimize long-term renal risk.

In this context, the potential advantages of SRSL warrant renewed attention. SRSL preserves physiological renal venous drainage and avoids direct compromise of kidney perfusion. Although it may be technically challenging, particularly in the presence of dense adhesions, it provides targeted shunt interruption while minimizing collateral vessel injury. By avoiding alterations in systemic venous return, SRSL may offer the advantage of sparing the kidney, which is particularly important in patients with pre-existing renal risk factors.

At our institution, SRSL has consistently provided reliable portal decompression and stable long-term renal outcomes. Given the rising prevalence of chronic kidney disease after liver transplantation and the widespread use of tacrolimus, SRSL may represent the preferred SRS management strategy when anatomically feasible. Surgical planning should incorporate both anatomical considerations and long-term organ protection, particularly in recipients predisposed to renal dysfunction.

This study has several strengths, including a relatively large cohort of LDLT recipients with large SRS, long-term follow-up of up to 10 years, and rigorous propensity score matching to minimize selection bias. In addition, standardized immunosuppressive protocols and objective renal biomarkers further strengthened the validity of the comparative analyses. However, several limitations should be acknowledged. This was a retrospective single-center study, which limits generalizability, and residual confounding cannot be completely excluded. Procedural selection was influenced by intraoperative assessment, surgeon preference, and individual judgment, introducing potential selection bias. Although we attempted to mitigate this by using MELD score and RBC transfusion as surrogate markers in the propensity score model, unmeasured surgeon-related factors may still have contributed to residual confounding. Furthermore, technical difficulty in determining the ligation strategy was difficult to quantify objectively, and renal function was primarily assessed using only serum creatinine. and the limited number of patients requiring dialysis reduced the statistical power for detecting differences in rare outcomes. Another limitation is that detailed graft function parameters, including rejection and longitudinal liver function tests, were not systematically analyzed. Therefore, subtle differences in graft function between groups may not have been fully captured. These strengths and limitations should be considered when interpreting our findings, and further prospective multicenter studies are warranted to validate our results.

In conclusion, both SRSL and LRVL are effective strategies for managing SRS during LDLT. Although survival outcomes were comparable, LRVL may be associated with potential long-term renal risks, even when early tacrolimus tapering is implemented. When anatomically feasible, SRSL may be considered, whereas LRVL remains an important alternative in technically challenging or hemodynamically unstable cases. In patients requiring LRVL, strategies aimed at renal protection, including careful titration of nephrotoxic immunosuppressants, should be implemented, and long-term renal function should be closely monitored. Further prospective multicenter studies with extended follow-up are warranted to validate these findings and guide optimal surgical decision-making in the management of SRS during liver transplantation.