Background
BK polyomavirus-associated nephropathy (BKVAN) affects up to 15% of renal transplant recipients and is an important cause of graft failure, due to insidious inflammatory destruction of the renal tissue [
1‐
5]. Monitoring of BK polyomavirus (BKV) infection in these patients is required for early detection of reactivation. It can be performed by detection of decoy cells (DC) in urine or detection of virus in plasma and urine using polymerase chain reaction (PCR) tests [
5‐
10]. BKVAN diagnosis depends on specific morphological findings in allograft biopsy or detection of small and cohesive aggregates of polyomavirus called Haufen-polyomavirus in ultrastructural urine tests [
11‐
13].
Many risk factors are involved in BKV reactivation [
13‐
15], but the identification of patients who have high risk to develop BKVAN remains a challenge [
16]. Specific antiviral treatments for BKV are not available [
14,
17]. The absence of a standard protocol for BKV infection treatment makes clinical management of these patients difficult.
We present two cases of BKVAN. Patient 1 was not monitored for BKV reactivation, while patient 2 was regularly monitored for BKV by urine testing for the presence of DC, which led to early diagnosis of viral reactivation. Both patients lost graft function due to BKVAN approximately a year after transplantation.
Discussion
Effective and safe antiviral therapies for BKVAN are not available. The management of patients with BKV reactivation is a challenge because there is not a reliable and universally accepted protocol to follow.
The pathogenesis of BKVAN is multifactorial, with several known risk factors: immunosuppression (a general prerequisite); “high dose” of new drugs such as tacrolimus; pre-transplant use of antilymphocyte therapy and MMF use at baseline; tubular injury/regeneration and/or ischemia/reperfusion in allograft (since native kidneys are generally not involved); human leukocyte antigen (HLA) mismatches; recipient diabetes; previous acute rejection; recipient age >55 years; recipient race (white); and recipient gender (male) [
4,
5,
14,
15,
19‐
21]. Both cases discussed here are male, presenting tacrolimus nephrotoxicity before BKVAN development. Higher blood levels of tacrolimus are associated with an increased incidence of BK viremia, which supports the notion that immunosuppression increases the risk of BKVAN [
14]. Epithelial cell proliferative state in response to different forms of injury may increase BKV replication [
15]. When their charts were reviewed, both patients presented previous tubular damage (multifocal epithelial necrosis) in their pre-implantation donor kidney biopsies.
Once BKVAN is diagnosed, definition of the presence of concurrent rejection can be very difficult to establish, because some morphological aspects may be shared by both conditions and they may co-occur [
15,
22]. Some clues for BKVAN diagnosis are: a heterogeneous inflammatory reaction, sometimes minimal, present especially in the medulla and composed of mononuclear cells; polymorphonuclear leukocytes, which can be seen in response to urinary leakage from damaged tubules; and inconspicuous tubulitis and viral lesions restricted to the medulla [
19]. On the other hand, in concurrent rejection there is abundant tubulitis, cortical inflammatory infiltrates (more pronounced in areas without viral inclusions), transplant endarteritis, glomerulitis, glomerulopathy, sclerosing vasculopathy, and C4d deposition along the peritubular capillaries [
19,
23]. Patient 1 exhibited interstitial edema and hemorrhagic foci. Although no morphological criteria for rejection (Banff classification) were present in this sample, it would be very difficult to definitely rule out this association.
It is difficult to make predictions about the progression of BKVAN. The risk of graft loss function in stage A is <10%, in stage B close to 50%, and in stage C >80% [
5]. Both patients were diagnosed at stage B.
Tacrolimus is a calcineurin inhibitor and some histological lesions have been associated with its chronic use, such as striped interstitial fibrosis, tubular atrophy, medial arteriolar hyalinosis, and tubular microcalcification [
24]. Tacrolimus may have played a role in the fibrosis observed in the graft biopsies of both patients.
Patient 2 presented early DC shedding. Considering that he had two renal transplants in 4 years, we wondered if BKV reactivation could be linked to his first kidney transplant. A previous transplant, in general, has not been confirmed as a risk factor for BKVAN [
21]. However, retransplantation after a graft loss due to BKVAN may be an important factor [
5]. Unfortunately, the reason for the loss of his first graft is unknown.
PCR tests for BKV in urine and plasma have higher positive predictive value for BKVAN than urinary cytology. However, urinary cytology is a low cost and simple method for BKV screening, with negative predictive value for BKVAN of 100% [
5,
25]. Some authors value the number of DC (>10/HPF), presence of a necro-inflammatory background in urine containing DC, persistent DC shedding (over 6 weeks), and detection of DC casts to identify patients with possible active BKVAN [
1,
26,
27]. Patient 2 presented all of these characteristics, but we only had an unequivocal criterion to perform graft biopsy when a clear drop in renal function was detected. The choice of the ideal moment to perform a renal allograft biopsy in BKV infection is a matter of controversy. It is an invasive procedure, which requires strict indication. On the other hand, initial BKVAN can have an indolent presentation, with no clear allograft dysfunction [
5]. The associated tacrolimus nephrotoxicity may have masked the real importance of BKV reactivation in the clinical scenario of patient 2.
Both patients were diagnosed with BKVAN at the end of the first year post-transplant. The diagnosis of viral reactivation was conceivably earlier in patient 2, who was systematically monitored. In spite of that, the course of BKVAN in the two cases was similar. An early diagnosis can be very important to preserve the tissue from inflammation and fibrosis. It is unclear if the availability of an effective antiviral drug would have made a difference in the clinical course of these patients.
Conclusions
Urinary monitoring for DC is a simple and efficient strategy for routine screening of BKV reactivation. Early detection of BKV infection in patients who have undergone a renal transplant is crucial to identify patients demanding closer clinical supervision. The presence of massive and persistent DC shedding can indicate a high risk for BKVAN development, even if renal function is normal.
Tacrolimus nephrotoxicity is a common complication in patients who have undergone a renal transplant and may mask the real importance of BKV reactivation.
Acknowledgements
This work was performed in the Department of Clinical Medicine and Department of Pathology, Universidade Federal Fluminense. We thank the Biology Institute of Universidade Federal Fluminense for allowing us to use the transmission electron microscope (JEM 1011) and the technical support biologist Alan Moraes for acquiring images. We also thank Anna Karoline Fausto da Silva for immunohistochemical stains of SV40 T-ag and Dr Paulo Alexandre Menezes, who was the attending physician of the cases.