Kidney transplantation (KT) is a vital treatment option for patients with end-stage renal disease, significantly improving both survival rates and quality of life [1, 2]. Despite its many advantages, post-transplant complications continue to pose challenges to graft longevity and patient outcomes [3, 4]. Among these complications, BK polyomavirus (BKPyV) DNAemia is recognized as a major concern affecting post-transplant outcomes [5–7]. The BKPyV, a member of the polyomavirus family, typically remains latent in renal tissue [8]. However, under conditions of immunosuppression, which are necessary to prevent graft rejection, the virus can reactivate [9]. This reactivation may lead to BK virus-associated nephropathy (BKVN), which is a leading cause of graft dysfunction.

The management of immunosuppression in KT recipients presents a critical clinical dilemma. Immunosuppressants, particularly calcineurin inhibitors (CNI) such as tacrolimus and mycophenolic acid (MPA), are essential for preventing organ rejection [10]. However, these same medications may inadvertently promote viral reactivation [11]. The challenge lies in reducing immunosuppression to mitigate the risk of BKPyV-DNAemia while simultaneously maintaining adequate immunosuppression to prevent rejection. Previous research has underscored the importance of maintaining optimal tacrolimus levels to maximize graft survival [12]. The present study builds upon this foundational work by offering a detailed analysis of risk factors, refining tacrolimus thresholds, and evaluating the efficacy of alternative immunosuppressive strategies that can minimize complications related to the BKPyV.

As the number of immunologically high-risk KT recipients continues to rise, BKPyV-DNAemia has become an increasingly critical concern for graft survival [13]. However, large-scale, multicenter studies addressing this issue are limited, and there is a notable lack of research on the relationship between CNI concentration and BKPyV-DNAemia outcomes. By leveraging clinical data from a large multicenter cohort, our study aims to establish the most effective immunosuppressive management following BKPyV-DNAemia onset by defining appropriate CNI trough levels and assessing the impact of different immunosuppressive regimens—such as leflunomide and sirolimus—on viral control, rejection risk, and long-term graft survival. Additionally, we seek to identify significant predictors and risk factors for BKPyV-DNAemia, enabling early detection and targeted intervention.

Our multicenter retrospective study underscores the critical importance of personalized immunosuppressive strategies for managing BKPyV-DNAemia in KT recipients. Key risk factors for BKPyV-DNAemia included older age, induction therapy with ATG, and elevated tacrolimus levels, which should be considered for risk stratification and targeted surveillance. Notably, the balance between preventing rejection and minimizing BKPyV-DNAemia heavily depends on maintaining optimal concentrations of CNI. Maintaining tacrolimus TDM levels at or above 5 ng/mL was associated with a lower risk of graft failure. In contrast, CNI withdrawal, even with the use of sirolimus as an alternative, showed a trend toward increased graft failure. These findings suggest that adequate CNI exposure is crucial for balancing viral control and immune suppression. Substituting MPA with leflunomide effectively reduced BKPyV load but was associated with a higher risk of rejection and inferior long-term graft survival. These findings suggest that prolonged maintenance of leflunomide instead of MPA may increase the risk of acute rejection and compromise graft survival.

Building upon existing literature on tacrolimus TDM [12], we found that tacrolimus trough levels between 5 and <7 ng/mL may represent the “optimal range” to mitigate the risk of BKPyV-DNAemia while maintaining sufficient immunosuppression to prevent rejection. In line with our findings, Schaub et al. demonstrated the effectiveness of a CNI-focused strategy for managing BKV infection in KT recipients by prioritizing tacrolimus reduction. Tacrolimus levels were reduced in a stepwise manner, with adjustments to MMF considered only after achieving sufficient CNI reduction [17]. This approach, supported by evidence of tacrolimus’s inhibitory effect on BKPyV-specific T cells, achieved a 92% clearance rate of BKPyV-DNAemia while maintaining stable allograft function over a median follow-up of 34 months [18, 19]. Moreover, the low clinical rejection rate of 8.6% and stable graft function despite subclinical inflammation further highlight the effectiveness of this strategy. The Kidney Disease Improving Global Outcomes (KDIGO) guidelines recommend reducing MPA first, followed by a reduction in CNI dosage, while the Second International Consensus Guideline presented both antimetabolite-first and CNI-first strategies as viable options. Additionally, KDIGO suggests a general 50% reduction in CNI dosage, whereas the International Consensus Guideline recommends target trough levels of tacrolimus (3–5 ng/mL) and cyclosporine (75–125 ng/mL) [20, 21]. Brennan et al. suggested that reducing antimetabolites before CNI reduction yields similar outcomes in BKPyV-DNAemia clearance compared to direct CNI reduction strategies [22]. These findings indicate that both approaches may be viable, emphasizing the need for individualized adjustments in immunosuppression. Furthermore, based on our study results, minimizing the duration of tacrolimus exposure below 5 ng/mL appears to be the most critical factor in optimizing post- BKPyV-DNAemia outcomes. Notably, large-scale studies stratifying outcomes by CNI levels are lacking, highlighting the significance of our findings in guiding immunosuppressive management.

A recent study from the Swiss Transplant Cohort proposed a five-group classification of BKPyV-DNAemia trajectories based on onset, duration, and clearance patterns [23]. This categorization demonstrated clinical relevance, showing that sustained or recurrent viremia, particularly among early-onset cases, was associated with higher rates of persistent replication and impaired graft function, whereas transient early-onset viremia correlated with more favorable outcomes. To explore this further, we performed a subgroup analysis among patients with early-onset BKPyV-DNAemia (≤90 days post-transplant), stratifying them by whether viremia resolved within 6 months or persisted thereafter. As shown in Supplementary Figure S3, the early-persistent subgroup exhibited a trend toward lower graft survival, although the difference did not reach statistical significance (log-rank p = 0.26). This suggests that duration of viremia may be a more critical determinant of outcome than timing of onset alone. Although we did not formally apply the full trajectory model used in the Swiss study, our findings support the clinical utility of integrating both onset and clearance patterns in future risk stratification frameworks.

Reduction or discontinuation of MPA in the treatment of BKPyV-DNAemia in KT recipients carries a risk of allograft rejection, even after achieving BKPyV-DNAemia clearance. The incidence of acute rejection in patients treated with immunosuppression reduction for BKPyV-DNAemia has been reported to be approximately 10%–30%, with a higher risk observed in patients undergoing more aggressive reductions or conversions to alternative immunosuppressive regimens. [24–26]. In our study, the rejection rates for patients treated with MPA, sirolimus, or leflunomide were 10.9%, 34.4%, and 22.5%, respectively. Notably, the rejection rate in the MPA group (10.9%) was consistent with prior studies that reported rates of 9%–12% for tacrolimus-based regimens combined with MPA or azathioprine [24, 27]. This suggests that reducing immunosuppression in the context of MPA-based regimens can effectively mitigate the risk of rejection while maintaining control of BKPyV-DNAemia The rejection rate in the sirolimus group (34.4%) was significantly higher than that in the MPA group (P < 0.001). This discrepancy may be attributed to the elevated initial BKPyV PCR levels in the sirolimus group and the treatment strategy employed at one participating center, where sirolimus was initiated when BKPyV PCR levels reached ≥4, accompanied by the withdrawal of CNI. Consequently, many patients in the sirolimus group underwent CNI withdrawal, which likely contributed to the higher rejection rate. Notably, in the multivariate analysis for acute rejection, sirolimus itself was not identified as a significant risk factor. The leflunomide group exhibited the most effective response to BKPyV-DNAemia treatment, despite having the highest maximum BKPyV-DNAemia PCR levels. However, the risk of acute rejection within 1 year after the onset of BKPyV-DNAemia was 2.1 times higher compared to the MPA group, suggesting that substituting leflunomide may have a similar immunosuppressive effect as withdrawing MPA. These findings indicate that transitioning from MPA to leflunomide can be a highly effective treatment strategy for patients with elevated BKPyV PCR levels. Nevertheless, based on our results, reintroducing a low dose of MPA or maintaining appropriate CNI levels as BKPyV PCR stabilizes may be advisable to minimize the risk of rejection.

Leflunomide is an immunomodulatory and antiviral agent that inhibits dihydroorotate dehydrogenase, thereby suppressing BKPyV replication and lymphocyte proliferation. Its antiviral effects are particularly pronounced in renal tubular epithelial cells, where it reduces the replication of BKPyV by inhibiting DNA synthesis [28, 29]. This dual mechanism allows for a reduction in the use of immunosuppressive drugs without increasing the risk of rejection. While effective in high-risk cases, its use is still associated with rejection and graft dysfunction, and the absence of a clear correlation between serum levels and efficacy complicates its clinical application. Our subgroup analysis suggests that leflunomide is a promising option for managing BKPyV-DNAemia in KT recipients, especially in high-risk cases. Similarly, a study by Aldieri et al. reported a BKPyV-DNAemia clearance rate of 91.4%, including viral eradication in 8 of 11 patients with biopsy-proven BKVN, when leflunomide was used as an adjunct to reduced immunosuppression rather than complete discontinuation of antiproliferative agents [30]. Further evidence from a multicenter study [31] demonstrated a 76% BKPyV-DNAemia clearance rate in KT recipients treated with leflunomide after failing prior therapies. However, 11 patients experienced graft loss, with 9 of these cases attributed to BKVN; rejection episodes occurred in 33% of patients, emphasizing the challenges of balancing immunosuppression and antiviral efficacy. A systematic review [32] corroborated these findings, reporting BKPyV-DNAemia clearance rates ranging from 33% to 92%, although significant heterogeneity in dosing regimens and pharmacokinetics complicated the interpretation of results. Notably, adverse events such as hemolytic anemia and thrombotic microangiopathy were observed, highlighting the importance of monitoring during treatment. Smaller prospective studies further support the efficacy of leflunomide. Faguer et al. reported that 42% of KT recipients with BKVN achieved viral clearance, and 66.6% maintained stable or improved graft function after switching from MMF to leflunomide [33]. Our study suggests a potential role for leflunomide in BKPyV suppression while underscoring the need for careful monitoring to balance efficacy and safety. Study by Bischof et al. summarize contemporary treatment options and emphasize the importance of tailoring immunosuppressive reduction based on viral dynamics histologic severity, and graft function [34]. Their study highlights the limitations of a one-size-fits-all approach and outlines the variable efficacy of adjunctive therapies, such as leflunomide and immunoglobulin, especially in the absence of randomized controlled trials. Our findings align with this perspective, suggesting that while immunosuppressive modulation remains the cornerstone, its optimization requires greater clinical granularity and prospective validation.

The role of sirolimus in managing BKPyV-DNAemia has been highlighted in several studies, demonstrating both antiviral effects and potential benefits in specific patient populations. The TRANSFORM study, a randomized, multicenter trial, evaluated everolimus with reduced exposure to CNIs compared to MPA with standard CNI exposure in de novo KT recipients. While not primarily designed to assess BK virus infection, the study reported a significantly lower incidence of BKV replication, based on center-reported data, in the everolimus group compared to the MPA group (8.8% vs. 14.8%, p < 0.001). [35]. Similarly, a retrospective study by Tohme et al. demonstrated a lower incidence of BKPyV-DNAemia in patients converted to sirolimus-based regimens, with clinically significant BKPyV-DNAemia observed in only 4.3% of the sirolimus group compared to 17.9% in the tacrolimus group [36]. These findings suggest a potential role for sirolimus in reducing BKPyV replication, particularly in low-risk populations. The recent BKEver study further supports the effectiveness of early reduction of both MPA and tacrolimus as a first-line approach for managing new-onset BKPyV-DNAemia [37]. In this prospective multicenter cohort, 81.3% of kidney transplant recipients achieved viral clearance within 6 months without an increased incidence of acute rejection. Notably, patients converted to everolimus had a lower clearance rate of 55.7%, suggesting that mTOR inhibitor conversion may be less effective as an initial strategy. These findings are consistent with our results and reinforce the value of a measured, stepwise reduction in immunosuppression for achieving viral control while minimizing rejection risk.

Several in vitro studies further support the antiviral properties of sirolimus. One study reported that sirolimus inhibits v replication by impairing mTOR-SP6-kinase activation and suppressing the expression of the BKPyV large T antigen in renal epithelial cells [38]. The inhibitory effects of sirolimus on BKPyV replication were most effective within 24 h of infection, particularly during early viral gene expression, but diminished during the late phase. These findings underscore a potential therapeutic window for sirolimus in the management of BKPyV-DNAemia In contrast, tacrolimus has been shown to activate BK viral replication via the same FKBP-12 pathway, highlighting a mechanistic divergence that could inform tailored immunosuppressive strategies. Moreover, sirolimus may modulate the immune response to the BKPyV through its effects on T-cell function. A study by Araki et al. demonstrated that rapamycin (sirolimus) enhances the formation of memory CD8⁺ T cells, which exhibit superior antiviral functionality, higher expression of markers associated with long-lived immunity (e.g., CD127, CD62L, Bcl-2), and reduced expression of senescence markers such as KLRG-1 [39]. These findings suggest that sirolimus may augment the antiviral immune response while providing essential immunosuppression for transplant recipients. However, complete withdrawal of CNIs when using sirolimus, particularly in immunologically high-risk patients, may increase the risk of acute rejection and negatively impact long-term graft survival.

Our study also highlights the wide variability in BKPyV DNAemia dynamics. The median duration of DNAemia exceeded 1.5 years in our cohort, with some patients experiencing persistence for more than 4 years. The initial and peak viral loads were notably higher in patients requiring alternative immunosuppressive regimens such as sirolimus or leflunomide. These findings reinforce the notion that viral kinetics—not just presence or absence of viremia—may influence both treatment decisions and graft outcomes, and should therefore be considered in future prospective stratification models.

Interestingly, our study found that female sex was associated with a lower risk of developing BK viremia, which complements prior observations identifying male sex as a potential risk factor, such as those noted in the The Transplantation Society (TTS) guidelines [21]. Although the underlying mechanisms remain unclear, pharmacokinetic studies have reported that female recipients tend to exhibit higher tacrolimus exposure and slower clearance, which may affect immunosuppressive intensity and susceptibility to viral reactivation [40, 41]. Additionally, sex-based differences in antiviral immunity have been described. These biological factors may contribute to the observed association, though further investigation is warranted to clarify causality.

Prolonged pre-transplant dialysis duration may, in part, reflect underlying immunologic barriers—such as HLA or ABO incompatibility—that delay transplantation and potentially influence post-transplant infection risk. However, in our cohort, there was no statistically significant association between pre-transplant dialysis and the need for desensitization (Pearson χ² = 1.684, P = 0.194), suggesting that pre-dialysis status was not primarily driven by immunologic risk factors.

The newly published consensus standard for BKPyV-associated nephropathy recommends not only timely reduction of immunosuppression upon BKPyV-DNAemia detection but also careful re-escalation of maintenance immunosuppression once viral clearance is achieved [21, 42]. In our retrospective cohort, data on post-clearance immunosuppressive intensification—including MPA reintroduction or increased tacrolimus dosing—were not consistently recorded. Moreover, substitution of MPA with leflunomide—a less potent immunosuppressant—may leave patients functionally under-immunosuppressed, potentially contributing to late acute rejection and graft loss. This observation highlights the need for closure of the immunosuppressive gap following viral clearance, ideally in line with expert guideline recommendations.

Despite its large sample size and extended follow-up, this study has several limitations. Its retrospective design and focus on Korean transplant centers may limit the generalizability of the findings to other populations. Normalization of BKV PCR values, such as using fold-change relative to the assay’s lower limit of detection, can enhance cross-center comparability in multicenter studies. In our cohort, however, the distribution of viral load values was empirically consistent across institutions, supporting the validity of using absolute values for analysis without additional transformation. Additionally, transplant practices evolved over the 15-year study period, potentially introducing unmeasured confounders. Variability in BKPyV detection and management protocols across centers may have resulted in selection bias, particularly in the sirolimus group, where one center exclusively implemented MPA discontinuation, sirolimus initiation, and complete CNI withdrawal for patients with BKPyV PCR levels greater than 4. This approach likely resulted in more severe or refractory BKPyV infections at baseline, influencing treatment outcomes despite multivariate adjustments. We therefore attempted to address these biases through comprehensive multivariate analyses to ensure robust findings. Moreover, our cohort included a relatively high proportion of immunologically high-risk patients, which may have impacted both BKPyV-DNAemia incidence and rejection patterns compared to lower-risk populations. These factors highlight the complexity of immunosuppressive modifications in BKPyV-DNAemia management and underscore the need for individualized treatment strategies based on patient-specific risk profiles. We acknowledge that immune reconstitution after immunosuppression reduction may lead to antiviral inflammatory infiltrates that mimic T-cell mediated rejection (TCMR). In our study, rejection diagnoses were based on local Banff assessments without centralized or molecular review, limiting our ability to distinguish true TCMR from beneficial antiviral responses. This represents a limitation and highlights the need for more refined biopsy evaluation in future studies. Lastly, corticosteroid exposure, including pulse therapy for acute rejection, was not uniformly documented across centers and could not be systematically analyzed. While most centers followed standard protocols, the lack of detailed data on cumulative steroid burden is a limitation that future studies should address.

Another limitation of our study is the composition of the sirolimus group. The vast majority (92.3%) of patients receiving sirolimus were managed in a CNI-withdrawal setting, with only 10 patients receiving sirolimus in combination with tacrolimus. No patients received cyclosporine plus sirolimus. Although the sirolimus + TAC subgroup showed a significantly higher rate of BKPyV treatment failure (40.0% vs. 8.3%, p = 0.002) but a lower rate of 1-year rejection (20.0% vs. 47.5%, p = 0.086) compared to the CNI-free sirolimus group, as shown in Supplementary Table S1, these results should be interpreted with caution due to the small sample size and the relatively high tacrolimus trough levels (mean TDM 7.6 ng/mL). Therefore, our study is not adequately powered to assess the effects of standard low-dose CNI + mTORi regimens and may not reflect their clinical efficacy.

This multicenter retrospective study highlights key risk factors for BKV and offers guidance on immunosuppressive strategies in kidney transplant recipients. Maintaining tacrolimus trough levels between 5 and 7 ng/mL balances BKPyV-DNAemia control and rejection risk. Adjusting or replacing MPA, including with leflunomide, may aid BKPyV-DNAemia management but carries long-term immunologic considerations. While CNI withdrawal may promote viral clearance, it raises rejection risk. We recommend individualized adjustment of immunosuppression based on BKPyV PCR trends. These findings support more personalized management approaches to improve long-term outcomes. Future prospective studies and incorporation of molecular diagnostics may enhance risk prediction and treatment optimization.