Low-dose cidofovir and conversion to mTOR-based immunosuppression in polyomavirus-associated nephropathy

Background: Polyomavirus-associated nephropathy (PVAN) remains a relevant com-plication following kidney transplantation with allograft loss rates of up to 50%. Reduction in overall immunosuppression is a cornerstone of therapy, whereas no specific antiviral regimen has shown conclusive benefit to date. The present case series demonstrates the efficacy of a dual therapeutic approach with low-dose cidofovir and conversion to mTOR-based immunosuppression in PVAN. Methods: Patients with biopsy-proven PVAN having received low-dose cidofovir (0.25 mg/kg) according to the Tübingen Cidofovir Protocol and been converted to mTOR-based immunosuppression were analyzed retrospectively. Results: Twenty-three patients with a median follow-up of 2.24 [IQR 1.55-5.01] years were included in the analysis. Median time to PVAN diagnosis was 268 [IQR 153-869] days after transplantation. Polyomavirus clearance from plasma was achieved in 78% of patients after a median of 118 [IQR 76-293] days. Of the 23 patients, nine patients (39%) lost their allograft function during follow-up, but only three of these (13%) due to PVAN. Fourteen patients (61%) stabilized or improved allograft function. The cido fovir protocol allowed for specific antiviral therapy without adverse nephrotoxicity, even in patients with low allograft function. Conclusions: Low-dose cidofovir and conversion to mTOR-based immunosuppres sion allow for effective virus clearance and preservation of allograft function in a high proportion of patients with PVAN and progressive allograft dysfunction and may prolong allograft survival in these patients.

will develop progressive allograft dysfunction and subsequent allograft loss when treated with reduction in immunosuppression alone. 3,4 Moreover, reduction in the overall immunosuppression carries the risk of acute or chronic rejection. Therefore, other treatment options including specific antiviral agents as well as modification of the immunosuppressive regimen rather than reduction in the overall immunosuppression have been investigated.
Although in vitro drug effects on the course of BK virus (BKV) replication have been demonstrated for leflunomide, fluoroquinolones and intravenous immunoglobulins, [5][6][7] clinical data have not shown convincing results for a broad implementation of these strategies. The nucleotide analog cidofovir has also demonstrated in vitro activity against polyomavirus, whereas various clinical case series and retrospective studies have shown inconsistent results following administration of cidofovir at variable doses (0.25-1 mg/kg), duration and treatment cycles. [8][9][10][11] The rationale for the use of cidofovir is based on the intratubular uptake and activity at the site of viral replication, however, this fact being also the cause of the most common side effect, nephrotoxicity.

A number of studies have shown benefits of an mTOR inhib-
itor-based immunosuppression regarding the risk of BK viremia compared to the standard of care tacrolimus-based regimen 12,13 ; however, limited data exist on effectiveness of this strategy in case of a biopsy-proven polyomavirus-associated nephropathy. The rationale for this treatment approach relies on the interaction of BK virus replication with components of the mTOR pathway as well as data showing an increased differentiation of virus-specific CD8 + memory T cells [14][15][16] following mTOR inhibition.
We now present data from our kidney transplant center focusing on probability and time course of BK virus clearance as well as allograft function in kidney transplant recipients with biopsy-proven PVAN, treated with a novel dual therapeutic approach consisting of low-dose cidofovir together with conversion to mTOR-based immunosuppression.

| MATERIAL S AND ME THODS
We performed a retrospective analysis of the center protocol in kidney transplant recipients from the Tübingen Collaborative Transplant Center between 04/2006 and 03/2018 with biopsyproven PVAN. The analysis was conducted in accordance with the Declaration of Helsinki and approved by the institutional review board (651/2016BO2). As a retrospective analysis, no written informed consent was required. All patients had shown BK viremia measured by quantitative PCR (qPCR) as well as deterioration of renal function prior to confirmation of the diagnosis PVAN in kidney biopsy. Screening for BKV viremia was historically a clinical decision, mostly incidence-based upon worsening of allograft function. With the advent of routine monitoring, a surveillance protocol was implemented for a standardized approach starting at the first visit after transplantation, 3, 6, 9 and 12 months post-transplantation, as well as incidence-based. PCR analysis was performed using the LightMix Kit Polyomaviruses JC and BK and was replaced by the Realstar BKV PCR Kit 1.0 in December 2014. Patients with BK viremia without histological evidence of PVAN were not included in the analysis.
The application of low-dose cidofovir was performed in an inpatient clinic according to the Tübingen Cidofovir Protocol, developed to effectively deliver therapeutic drug concentrations at limited nephrotoxicity (Table 1). Adequate intravenous hydration aiming at a urine output >100 mL/h is used in order to achieve a calculated tubular passage time for reduction in toxicity. No probenecid is administered to allow for tubular uptake of cidofovir via the basolateral membrane human renal organic anion transporter 1 (hOAT1) and obtain therapeutic drug concentrations at the site of viral replication. A single cidofovir application was performed. The necessity of subsequent applications was evaluated clinically and based on the course of BKV replication rate and allograft function.
In most cases, conversion to mTOR-based immunosuppression was performed via reduction in calcineurin inhibitor (CNI) trough levels and stopping as soon as targeted mTOR inhibitor trough levels were achieved. In few patients with high immunological risk, the CNI was continued with cessation of mycophenolate mofetil (MMF).
Data are given as median [interquartile range]. Comparison between groups was tested using Wilcoxon-Test. Results with twosided P ≤ .05 were considered statistically significant. The JMP (version 14.0; SAS Institute) statistical software package was used.

| BKV clearance and allograft function
Clearance of BK viremia, defined as a reduction in BKV-DNA levels below the threshold of 1000 copies/mL plasma in qPCR, was achieved in 18 (78%) patients. and death with functioning graft (n = 1), whereas in two patients the reason for allograft loss was unknown. that intrarenal concentrations of radioactive-labeled cidofovir were about 10 times higher than in plasma. 18 Therefore, we do not think measurement of plasma levels is helpful, but-in our opinion-achieving sufficient intratubular cidofovir concentrations accomplished by omission of probenecid is essential for the antiviral efficacy of cidofovir. In fact, 57% of our patients achieved clearance of BK viremia after a single cidofovir application. Our finding that patients with higher eGFR at diagnosis had a lower probability of clearance of BK viremia has led to the notion, that in patients with higher eGFR, repetitive cidofovir applications may be beneficial. We can only speculate that, having a higher eGFR, more cidofovir is lost in urine due to better glomerular filtration and saturated tubular uptake.

| D ISCUSS I ON
Even though studies have consistently shown lower rates of BKV infection following an mTOR-based immunosuppression, 2,12,13,19 results on the course of PVAN following conversion to an mTOR-based immunosuppression in manifest disease are lacking. In vitro data have shown a reduction in BKV replication with mTOR inhibition, which is explained by dependency of early BKV replication on the mTOR pathway, 14 as well as by an improved immune response through regulation of the differentiation of memory CD8 + T cells. 16  from an ongoing protocol, strictly and consistently followed in all patients. Patients underwent two simultaneous interventions, precluding a distinction whether antiviral therapy or conversion to mTOR-based immunosuppression plays the leading role. Nonetheless, we do believe that the combination of these two strategies with the intratubular, antiviral effects of cidofovir together with the beneficial effects on the antiviral immune response allowed by mTOR inhibition is key. The strength of our study is the standardized approach to patients with biopsy-proven PVAN as well as the long follow-up time.
In conclusion, we demonstrate that our novel dual therapeutic approach to PVAN with application of low-dose cidofovir and conversion to an mTOR-based immunosuppression allows for clearance of BK viremia and preservation of allograft function in a high proportion of patients. Randomized, prospective studies on this issue are highly warranted.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest associated with this manuscript.

AUTH O R S' CO NTR I B UTI O N S
TM designed the work, acquired data, interpreted results, drafted