SEARCH

SEARCH BY CITATION

Keywords:

  • Polyomavirus BK;
  • polyomavirus-associated nephropathy;
  • polyomavirus-associated haemorrhagic cyctitis;
  • kidney transplantation;
  • hematopoietic stem cell transplantation;
  • immunosuppression

Abstract

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

Polyomavirus BK (BKPyV) infects most people subclinically during childhood and establishes a lifelong infection in the renourinary tract. In most immunocompetent individuals, the infection is completely asymptomatic, despite frequent episodes of viral reactivation with shedding into the urine. In immunocompromised patients, reactivation followed by high-level viral replication can lead to severe disease: 1–10% of kidney transplant patients develop polyomavirus-associated nephropathy (PyVAN) and 5–15% of allogenic hematopoietic stem cell transplant patients develop polyomavirus-associated haemorrhagic cystitis (PyVHC). Other conditions such as ureteric stenosis, encephalitis, meningoencephalitis, pneumonia and vasculopathy have also been associated with BKPyV infection in immunocompromised individuals. Although BKPyV has been associated with cancer development, especially in the bladder, definitive evidence of a role in human malignancy is lacking. Diagnosis of PyVAN and PyVHC is mainly achieved by quantitative PCR of urine and plasma, but also by cytology, immunohistology and electron microscopy. Despite more than 40 years of research on BKPyV, there is still no effective antiviral therapy. The current treatment strategy for PyVAN is to allow reconstitution of immune function by reducing or changing the immunosuppressive medication. For PyVHC, treatment is purely supportive. Here, we present a summary of the accrued knowledge regarding BKPyV.

Abbeviations
BKPyV

polyomavirus BK

CNS

central nervous system

CSF

cerebrospinal fluid

CMV

cytomegalovirus

CPE

cytopathic effect

GEq

Genome Equivalents

HEL

human embryo lung fibroblasts

HIV

Human immunodeficiency virus

HPyV6

human polyomavirus 6

HPyV7

human polyomavirus 7

HPyV9

human polyomavirus 9

HPyV10

human polyomavirus 10

HPyV12

human polyomavirus 12

HSCT

hematopoietic stem cell transplant

JCPyV

polyomavirus JC

KIPyV

polyomavirus KI

LTag

large Tumour antigen

MCPyV

merkel cell polyomavirus

MK

monkey kidney

MWPyV

malawi polyomavirus

MXPyV

MX polyomavirus

NCCR

non-coding control region

OAT1

organic anion transporters

PML

progressive multifocal leucoencephalopathy

PyVAN

polyomavirus-associated nephropathy

PyVHC

polyomavirus-associated haemorrhagic cystitis

SV40

simian virus 40

sTag

small tumour antigen

STLPyV

STL polyomavirus

TruncTag

truncated tumour antigen

TsPyV

trichodysplasia spinulosa-associated polyomavirus

VP1

viral structural protein 1

VP2

viral structural protein 2

VP3

viral structural protein 3

VLPs

virus like particles

WUPyV

polyomavirus WU

Polyomavirus BK (BKPyV) belongs to the genus Orthopolyomavirus in the family Polyomaviridae [1]. Until 1999, BKPyV together with polyomaviruses JC (JCPyV) and simian virus 40 (SV40) were grouped with papillomaviruses and were called Papovaviruses, short for papilloma-polyoma-vaculating viruses [2].

Polyomavirus BK was isolated for the first time in 1970 [3, 4]. A midstream urine sample, from a patient who underwent kidney transplantation 3.5 months earlier, contained many inclusion-bearing epithelial cells. Dr. Anne Field at the Central Public Health Laboratory in London examined the sample by negative staining electron microscopy and found a large number of viral particles that were similar to the common wart virus. Electron microscopy of ultrathin section of cells in a urine sample taken 5 days later demonstrated many viral particles within enlarged cell nuclei. The first urine sample was inoculated on secondary Rhesus monkey kidney (MK) cells, African monkey kidney (Vero) cells and human embryo lung fibroblasts (HEL). Eighteen days post-inoculation, a cytopathic effect (CPE) was found in the MK cell culture and the cell culture supernatant was found to contain viral particles. In Vero cells, CPE was delayed until day 28 while no CPE was detected in HEL cells. The virus was named BK after the initials of the patient and the original isolate was called the Gardner strain [4].

At the same time, independently, JCPyV was cultivated for the first time from the brain tissue of a patient with Hodgkin's disease and progressive multifocal leukoencephalopathy (PML) [5]. For almost 40 years, these were the only known human polyomaviruses. In 2007 the human polyomaviruses KI (KIPyV) [6] and WU (WUPyV) [7] were identified by cloning and since then eight more human polyomavirus species have been detected using molecular techniques: Merkel cell polyomavirus (MCPyV) [8], human polyomaviruses 6 and 7 (HPyV6 and HPyV7) [9], Trichodysplasia spinulosa-associated polyomavirus (TSPyV) [10], human polyomavirus 9 (HPyV9) [11], human polyomavirus 10 (HPyV10) [12] and two variants of HPyV 10 denoted Malawi polyomavirus (MWPyV) [13] and MX polyomavirus (MXPyV) [14], STL polyomavirus (STLPyV) [15] and lastly human polyomavirus 12 (HPyV12) [16]. For the recently discovered polyomaviruses, an association with clinical disease has only been found for MCPyV, which is involved in Merkel cell carcinoma, and TSPyV, which is involved in Trichodysplasia spinulosa. Here, we review the accumulated knowledge on BKPyV, including structure, replication, epidemiology, pathogenesis and diagnostic- and treatment strategies.

Structure and Genome Organization

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

The non-enveloped, ichosahedral virion of BKPyV measures about 45 nm in diameter [17]. Its capsid is composed of 72 capsomers, each consisting of five copies of the major structural protein VP1 and one internal minor structural protein, either VP2 or VP3, linking the viral genome to the capsid [18, 19]. The BKPyV genome is a circular, double-stranded DNA molecule of about 5000 basepairs packed around the cellular histones H2A, H2B, H3 and H4, thereby forming a minichromosome [20]. On average, each genome contains about 20 nucleosomes [21].

The genome is divided into three functional regions [20] (Fig. 1)

image

Figure 1. Schematic illustration of the circular double-stranded DNA genome of Polyomavirus BK archetype strain showing the early region and the late region separated by the non-coding control region (NCCR). Transcription of the early and late genes proceeds in the direction of the arrows. Solid arrows indicate exons while dashed arrows in the early region indicate introns. The position of the microRNAs, which are processed from a late transcript, is shown. The numbers on the transcription factor binding sequence blocks in the NCCR indicate the size in base pairs.

Download figure to PowerPoint

(i) The early region on the proximal side of the origin of replication, encodes the regulatory proteins large tumour antigen (LTag, 80.5 kDa), small tumour antigen (sTag, 20.5 kDa) and the truncated tumour antigen (truncTag, 17 kDa) [22], which are expressed from three different mRNAs derived by alternative splicing of a single primary transcript. These three proteins have 81 amino acids in common at the N-terminal end. The multifunctional LTag is the major regulatory protein and is indispensable for BKPyV replication. As it shows 76% amino acid homology with SV40 LTag, it is thought to exert many of the same functions [23]. While the functions of sTag and truncTag during replication are still uncertain, they are likely to be important in pathogenesis, considering their wide cross species conservation [22, 24]. In addition, the early region encodes a pre-miRNA that generates two functional miRNAs [25]. The pre-miRNAs are actually made from late transcripts which continue into the early region of the genome. They are complementary to the LTag mRNA and posttranscriptionally down-regulate LTag expression, possibly to evade an immune response. Recently the 3′miRNA was found to target the mRNA of the cellular stress induced ligand ULBP3, an important component of the immune system, thereby further increasing the chances of escape from immune elimination [26].

(ii) The late region on the distal side of the origin of replication, encodes the three structural proteins VP1 (40.1 kDa), VP2 (38.3 kDa) and VP3 (26.7 kDa) and the non-structural agnoprotein (7.4 kDa) [27]. VP2 and VP3 are translated from the same transcript and VP3 is colinear to the C-terminal two-thirds of VP2. The unique 119 amino acid N-terminal end of VP2 contains a putative myristoylation site. Agnoprotein and VP1 are translated in a different reading frame than VP2 and VP3 from partly overlapping transcripts. The function of the small cytoplasmic agnoprotein, often found associated with lipid droplets, has been subject to much speculation but still remains elusive [28-30]. The late region contains some additional open reading frames, one of which may encode a protein similar to the VP4 protein of SV40 [31]. Apparently this protein triggers the lytic release of SV40 [32].

(iii) The non-coding control region (NCCR), also known as the regulatory region or the transcription control region, is a nuclesome-free area separating the early and late region. The NCCR contains the origin of replication and a bidirectional promoter-enhancer region containing several transcription factor binding sites. As such, this region directs early and late transcription and replication of the genome. The NCCR of the archetype strain, which is most frequently found in urine, has been arbitrarily divided into five sequence blocks denoted O143, P68, Q39, R63 and S63 where the numbers indicate the number of base pairs [33, 34]. The NCCR may have deletions, insertions or duplications of complete or partial blocks and is then generally referred to as a rearranged NCCR. Strains with rearranged NCCRs have been found in urine, plasma and biopsies of patients suffering from BKPyV diseases [35-37]. Experiments in human cell cultures and observations of samples from healthy individuals and kidney transplant patients suggest that archetype virus is the transmissible form and that rearranged strains are generated de novo during high-level viral replication [36, 38, 39], and are associated with disease progression in kidney transplant patients. However, this may not be the case for hematopoietic stem cell transplant (HSCT) patients [40, 41].

The nature of the molecular mechanisms leading to rearrangement of NCCRs is still unclear. Yoshiike and Takemoto [42] suggested a mechanism involving non-homologous recombination between two newly synthesized DNA segments. The resulting molecule would be a dimer composed of one molecule with deletion and the other with duplication, which could be converted to monomers by homologous recombination. Another mechanism suggested is an error during viral DNA replication shifting the replication point downstream. The latter mechanism is probably less likely, as BKPyV uses the accurate host replication system.

The Replication Cycle

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

The replication cycle of BKPyV is initiated by binding to a cellular receptor. N-linked glycoproteins with α2,3-linked sialic acids [43] and gangliosides GD1b and GT1b [44] have been found to act as receptors for BKPyV. After attachment, BKPyV is internalized via caveolae-mediated endocytosis, and is transported towards the endoplasmic reticulum via the cellular cytoskeleton [45] where complete or partial uncoating takes place. It is known from research on SV40 that nuclear localization signals of the minor capsid proteins VP2 and/or VP3 are required for the genome to use the nuclear import machinery of the host for nuclear entry [46]. In a study by Maraldi et al. [47], electron microscopy revealed BKPyV particles inside the nucleus only 12 h after infection thereby suggesting that complete or partly uncoated viral particles can enter the nucleus. Whether this is the dominant strategy or an aberration is unknown.

As soon as the genome enters the nucleus, the early BKPyV genes are expressed. LTag prepares the infected cell for viral replication by binding to members of the retinoblastoma (Rb) family of tumour suppressor proteins (pRb, p107, p130), thereby stimulating cell-cycle progression from G0 or G1 into the replicative S-phase. By binding to tumour suppressor p53, LTag also counteracts apoptosis [48] [reviewed in more detail in [49]]. This interference with normal cell-cycle progression gives BKPyV an oncogenic potential.

Besides making the cell environment optimal for DNA replication, LTag initiates the genome replication by binding to the origin of replication and locally unwinding the double-stranded DNA [50]. BKPyV does not encode a viral DNA polymerase or other replication factors, but LTag recruits cellular DNA polymerase, α-primase, topoisomerase I and replication protein A and by functioning like a helicase it facilitates the bidirectional DNA replication. After DNA replication, the late genes are expressed and virions are assembled inside the nucleus. The progeny in renal tubular epithelial cells is apparently released by cell lysis during necrosis [51-53] while release of progeny in primary human urothelial cells seems to be occur after host cell detachment [54].

The replication capacity is mainly decided by the NCCR although polymorphism in the early and late regions may also influence this in cell culture and probably also in vivo [55]. The archetype virus, which is shown to have a weak early promoter, and therefore produces only small amounts of LTag, replicates relatively slowly. Most rearranged strains, on the other hand, have strong early promoters, make larger amounts of LTag and replicate faster [36, 56, 57]. In line with this, archetype BKPyV, which usually replicates poorly in vitro [38], replicates efficiently in 293TT, a human embryonic kidney cell line with high expression of SV40 LTag [58]. This demonstrates that LTag from a different polyomavirus can support BKPyV infection, thereby suggesting that co-infections with other polyomaviruses may promote BKPyV replication. Also, other potentially co-infecting viruses like Human immunodeficiency virus (HIV) have been suggested to influence BKPyV replication [59].

Epidemiology

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

The natural transmission mode of BKPyV is still unknown, but likely involves the oral or respiratory route [60]. Epidemiological studies have shown that BKPyV infection typically occurs at a young age [61] and that the virus thereafter persists in epithelial cells of the renourinary tract [62]. It is unknown if this is a true latent infection, with no replication and limited or no viral gene expression, or if it is a persistent infection, with low-level virus replication. However, several independent studies have shown that BKPyV reactivates and is asymptomatically shed in the urine of immunocompetent individuals at low viral loads <5 log10 Genome Equivalents (GEq)/mL [39, 63, 64]. In immunocompromised individuals such as solid organ or HSCT patients, BKPyV shedding occurs more frequently and at higher viral loads (>7 log10 GEq/mL) [65]. Usually this results in decoy cells which can be observed by urine cytology [66] and free viral particles which can be observed by negative staining electron microscopy [67].

Previously BKPyV was divided into four serotypes [68]. Genetic heterogeneity in the epitope region of the VP1 gene has been used to divide BKPyV into four major genotypes I, II, III and IV, where genotype I is the most prevalent worldwide [69]. The postulated reason for this dominance, is an enhanced replicative ability [55]. Recently the genotypes were further divided into subgroups and these subgroups seem to correspond to the geographical distribution [70, 71]. Both healthy and diseased individuals may harbour BKPyV of more than one genotype [72].

The overall seroprevalence for BKPyV, determined by hemagglutination inhibition or ELISA based on viral like particles or capsomers, was found to be about 80% [39, 61, 73]. The number of seropositive people increases until the age of 40 years and then slightly decreases [61, 73]. Interestingly, this is in contrast with seroprevalence for JCPyV which increases throughout life [39, 61, 73].

Associated Diseases

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

Different disease processes have been linked to BKPyV infection, but the two most important diseases are polyomavirus-associated nephropathy affecting kidney transplant patients and polyomavirus-associated haemorrhagic cystitis affecting allogenic HSCT patients. Other pathological processes that have been associated with BKPyV are ureteric stenosis, encephalitis, meningoencephalitis, pneumonia, vasculopathy and bladder cancer.

Polyomavirus-associated nephropathy

The first reports describing disease resembling polyomavirus-associated nephropathy (PyVAN) in kidney transplant patients, came as early as 1978 [74, 75]. However, it was not until the late 1990s that PyVAN emerged [76] and BKPyV was found to be the aetiological agent [77, 78]. At that time, new and more potent immunosuppressive regimens were being taken into use. The new drugs decreased rejection rates at the expense of increased risk of opportunistic infections like BKPyV. PyVAN affects between 1 and 10% of kidney transplant patients during the first two years post-transplantation [66, 77, 79, 80]. PyVAN rarely affects patients other than kidney transplant recipients, and so far only 26 cases of biopsy confirmed PyVAN, in native kidneys of other immunocompromised patients, have been reported (reviewed in [81]. This clearly suggests that there must be risk factors associated with the allograft and not only with the suppressed immune status. Prognostication is complicated by the many putative risk factors, including HLA-mismatch, donor of female gender, high donor age, recipient of male gender, type and dose of immunosuppressive medication and transplantation events such as ureteric stents, steroid exposure and acute rejection [reviewed by [65, 82]].

The pathogenesis of PyVAN is characterized by high-level BKPyV replication in renal-tubular epithelial cells of the transplanted kidney (Fig. 2A) leading to cytopathic loss and thereby denudation of the epithelial monolayer in the allograft tubulus. As a consequence of this, virus leaks into the tissue and bloodstream and inflammatory cells infiltrate the interstitium leading to tubular atrophy and interstitial fibrosis [65, 83]. This reduces the graft function and increases the risk of graft loss. Of note, the urothelial cells may also play an important role in PyVAN. Mathematical modelling of BKPyV replication in kidney transplant patients with PyVAN suggests that viral replication starts in the renal tubular epithelial cells but is then carried to the urothelial cell compartment where more than 90% of urine BKPyV loads are generated [84]. This is supported by histopathological data revealing extensively infected urothelial cells in the bladder of patients with PyVAN [78] (Fig. 2B). In agreement with this, primary human urothelial cells from bladder were found to be very permissive to BKPyV infection [54] (Fig. 2D).

image

Figure 2. Polyomavirus BK (BKPyV) replication in different cells in vivo and in vitro. (A) Immunohistochemistry of a kidney allograft biopsy, from a kidney transplant patient with PyVAN, stained with anti-BKPyV agnoprotein serum (brown). Of note, this antiserum does not cross-react with JCPyV or SV40 thereby confirming that the cells are infected by BKPyV. The figure is reprinted from Ref. [184] with permission from the publisher. Magnification × 400. (B) Immunohistochemistry of an urothelium biopsy, from a kidney transplant patient with PyVAN, stained with an antibody directed against SV40 LTag but cross-reacting with BKPyV LTag (brown). A positive staining pattern is found in superficial transitional cells and occasionally in basal cells. There is no sign of inflammation in the lamina propria. The figure is reprinted from Ref. [78] with permission from the publisher. Magnification × 160. (C) Urine cytology from a kidney transplant patient with PyVAN, showing BKPyV-infected decoy cell with a ground-glass type intranuclear inclusion body (arrow). Papanicolaou stain was used. The figure is reprinted from Ref. [78] with permission from the publisher. Magnification × 400. (D) Electron microscopy of a BKPyV infected primary human bladder urothelial cell at 72 h after infection showing numerous progeny in the nucleus.

Download figure to PowerPoint

Polyomavirus-associated haemorrhagic cystitis

Five to 15% of allogenic HSCT patients are affected by polyomavirus-associated haemorrhagic cystitis (PyVHC), a late onset haemorrhagic cystitis usually starting about 50 days post-transplantation. The pathogenesis of PyVHC is not fully understood, but is suggested to result from a sequence of events [85, 86]. First the bladder mucosa is subclinically damaged by the toxic metabolite of cyclophosphamide which is used as a conditioning protocol prior to stem cell transplant. Next, during the aplastic phase, immunologically uncontrolled high-level BKPyV replication causes denudation of the damaged bladder mucosa. Finally inflammation occurs upon engraftment of the allogeneic stem cell graft. Though PyVHC usually affects allogenic HSCT patients, there have been some case reports on PyVHC in other immunocompromised patients [reviewed in [60, 87]].

Ureteric stenosis

The kidney transplant patient from whom BKPyV was initially isolated, was suffering from ureteric stenosis [4]. Since this first report, there have been several reports on BKPyV-associated ureteric stenosis, usually in kidney transplant patients, both paediatric [88] and adult [74, 89, 90], but also in allogenic HSCT patients [91, 92]. The pathogenesis of ureteric stenosis is still not completely resolved. As early as 1978, Coleman et al. [74] suggested that high-dose steroids, given to kidney transplant patients post-transplantation, permitted reactivation of BKPyV. The ureteric epithelium which was damaged by ischaemia or inflammation supported infection and was replaced with granulation tissue. In a study of kidney transplanted cynomolgus monkeys receiving immunosuppression, a reactivation of a new polyomavirus called cynomolgus polyomavirus occurred in 12 of 57 monkeys [93]. The virus was detected in the urothelium of graft ureters in association with inflammation and in smooth muscle cells of the ureteric wall showing signs of apoptosis. A significant incidence of late onset stenosis was seen. Apparently, ureteric stenosis is now less frequently reported which could be due to better surgical techniques and a decline in use of ureteral stents [94, 95].

Central nervous system involvement

There have been some case reports on BKPyV involvement in central nervous system (CNS) diseases of immunocompromised patients [96-103]. The first described case was a man with haemophilia type 2 and AIDS, who developed subacute meningoencephalitis and subsequently died [104]. BKPyV DNA and/or proteins were found in CNS and in cerebrospinal fluid (CSF) and in cells in the kidneys and lungs suggesting a disseminated BKPyV disease. The last reported case was a 6-year-old child on tacrolimus treatment due to a heart transplant [105, 106]. She first developed PyVAN which was unsuccessfully treated with bi-monthly intravenous cidofovir and she then developed BKPyV encephalitis which was confirmed by PCR analysis of brain parenchyma and CSF post-mortem. There have also been case reports suggesting PML like disease caused by BKPyV rather than JCPyV [107], but no confirmation in tissue has been presented. Recently, a case described as the first definitive case of BKPyV-associated PML was published [108]. The immunocompromised patient was referred for the investigation due to progressive apraxia and right-sided neglect. Magnetic resonance imaging of the brain demonstrated two areas of abnormality and somewhat surprisingly, quantitative real-time PCR of the CSF revealed BKPyV and not JCPyV DNA as expected. A brain biopsy was taken and immunohistochemistry with an antibody directed against BKPyV LTag and not cross-reacting with JCPyV LTag [109] positively stained intranuclear inclusions in scattered cells. In line with this, electron microscopy showed cells with enlarged nuclei containing viral particles. However, the value of the immunohistochemistry data may be questioned as the antibody used is known to cross-react with the large subunit of the cellular Ku autoantigen and therefore is unsuitable for studies in human cells [110].

Cancer

The prerequisite functions for the induction of tumours are encoded by BKPyV, and many reports on detection of BKPyV DNA and/or proteins in different tumours have been published as recently reviewed by Dalianis and Hirsch [111]. In particular, in some cases of bladder cancer, both the primary tumour and the metastasis have been found to stain positive when a SV40 LTag antibody cross-reacting with BKPyV LTag was used [90]. However, a definitive causal role for BKPyV in human malignancy is still lacking [reviewed by [111-113]]. Nevertheless, in 2012 BKPyV together with JCPyV were classified as ‘possibly carcinogenic to humans’ (Group 2B) by a WHO International Agency for Cancer Research Monograph Working Group [114]. In a recent study of seroresponses to BKPyV, JCPyV and MCPyV in 1135 patients with bladder cancer and in 982-matched controls, the same seroprevalence but a significantly higher seroreactivity was found for BKPyV and MCPyV in cancer patients than controls, proving a compelling association [115]. However, a possible explanation of this association could be that the patients with bladder cancer have more frequent BKPyV reactivation than other patients. Moreover, the finding of LTag in both primary and secondary tumour cells but not in adjacent non-neoplastic tissue may be due to actively growing cancer cells having enhanced susceptibility to BKPyV.

Rare diseases

One case of systemic vasculopathy with capillary leakage and multi organ failure has been reported in a renal transplant patient [116]. A child with allogenic HSCT initially developed symptoms of PyVHC. Two weeks later she developed numerous papules and vesicles on her hands and feet, and the vesicular fluid was found to be PCR positive for BKPyV [117]. An 8-month-old child received an unrelated umbilical cord transplant and subsequently developed PyVHC and severe interstitial pneumonia. No other pathogen than BKPyV was detected in the lung tissue during autopsy [118]. Autopsy of patients with HIV-AIDS has also revealed BKPyV pneumonia [60, 104].

Diagnosis

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

PyVAN

As the initial presentation of PyVAN is insidious, it is strongly recommended to screen kidney transplant patients regularly for early diagnosis [65]. Screening should be performed at least every 3 months the first 2 years after transplantation and then annually until the fifth year of post-transplantation. Screening is usually accomplished by quantitative PCR of urine and/or plasma for detection of high-level BKPyV viruria ≥7 log10 GEq/mL or viremia ≥4 log10 GEq/mL [119]. Alternatively, cytological examination of urine in search of decoy cells [75] (Fig 2C) or electron microscopy in search of viral aggregates [67] can be performed. Plasma PCR has a higher positive predictive value than urine PCR as episodic viruria is quite frequent in this patient group, while viremia is less common and usually precedes PyVAN [66]. In patients with sustained plasma BKPyV DNA load of ≥4 log10 GEq/mL or equivalent, the diagnosis presumptive PyVAN can be made. It is important to avoid using PCRs with primers and probes targeting the variable regions of the genome like the NCCR or VP1 region as this may give false negative results or incorrect viral loads when samples contain rare genotypes [120, 121]. Many labs use in-house PCR methods. This gives them a possibility to assure optimal performance by performing regular database analysis of primer and probe sequences. If small polymorphisms are detected, degenerated primers or probes may sometimes be a good option [122]. Usually the diagnosis of PyVAN requires a histological demonstration of BKPyV replication [76]. However, due to the focal nature of PyVAN, a negative biopsy result cannot rule out PyVAN [123]. Based on the histopathological changes introduced by viral replication and inflammation, the disease is graded into patterns A–C where pattern C is the most advanced stage [124].

PyVHC

To distinguish PyVHC from the more frequent early-onset haemorrhagic cystitis occurring prior to the engraftment due to urotoxic conditioning or total body irradiation, the triad of cystitis, hematuria (grade II or more) and high-level BKPyV replication with urine loads of ≥7 log10 GEq/mL, must be present. Other infectious and bleeding disorders should also be ruled out [83] and the plasma BKV load, which appears to be a marker of significant PyVHC, should be measured [66, 125-127]. The plasma BKV load has also been found to correlate better with clinical recovery than the urine BKV load [127].

Treatment

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

PyVAN

Safe and effective antiviral treatment of BKPyV diseases is still lacking [128]. In the management of kidney transplant patients with BKPyV viremia, regardless of the presence or absence of BKPyV-positive cells in a kidney biopsy, it is recommended to reduce, change or discontinue immunosuppressive drugs to allow for a BKPyV-specific antiviral immune response [reviewed by [65, 82]]. If the disease is detected at an early stage, a reduction in the immunosuppression is often enough to clear the BKPyV viremia and stop the progression of disease [129-131]. In a recent histopathological study of 35 patients with PyVAN treated by reduction in immunosuppression alone, 83% of the patients showed stable or improved allograft function 2 years later, while 2 patients (6%) experienced clinical rejection [132]. Interestingly, the morphological pattern of resolving PyVAN could not be distinguished from rejection and revealed intraepithelial lymphocytes and interstitial inflammation. However, as the majority of patients resolved their PyVAN, the inflammation was apparently beneficial.

A systematic analysis of all published treatments for PyVAN concluded with no graft survival benefit of treatment with cidofovir or leflunomide combined with reduction in immunosuppression compared to the reduction in immunosuppression alone [133]. There is no clinical evidence to support the use of any other drugs either. Nevertheless, cidofovir, leflunomide, fluroquinolones and intravenous immunoglobulins are sometimes given as adjunctive therapies.

Cidofovir, trade name Vistide (Gilead, Foster City, CA, USA), is an intravenously administered nucleoside analogue of deoxy cytidine monophosphate that is licenced by the U.S. Food & Drug Administration for the treatment of cytomegalovirus retinitis in AIDS patients. In 2000, a patient with PyVHC and simultaneous cytomegalovirus (CMV) infection was successfully treated with cidofovir [134] and shortly after it was reported to be useful in the treatment of one patient with PyVAN [135]. The drug is taken up via the organic anion transporters (OAT1), which are mainly expressed on the basolateral side of renal tubular epithelial cells. The drug becomes active after two phosphorylation steps performed by cellular enzymes [136, 137]. In vitro studies on cidofovir and CMV replication suggest that incorporation of a single molecule of cidofovir in the viral DNA results in reduced replication, while incorporation of two consecutive molecules efficiently terminates elongation [138]. In agreement with this, in vitro studies on cidofovir and BKPyV replication showed decreased intracellular and extracellular BKPyV loads but also reduced cellular DNA replication and metabolic activity [139, 140]. As BKPyV does not encode a viral polymerase, unlike CMV, the results suggest that the antiviral activity is caused by the cytostatic and cytotoxic effects of the drug. The cytostatic and cytotoxic effects found in vitro were not completely unexpected since cidofovir is known to be nephrotoxic [141]. While some studies report that cidofovir treatment stabilizes kidney function, others report no benefit [reviewed in [65]]. However, as immunosuppression is usually decreased simultaneously with cidofovir treatment, the effect is difficult to evaluate. A lipid conjugate of cidofovir, 1-O-hexadecyloxypropyl cidofovir (CMX001), was found to have a longer lasting, less toxic and 400 times stronger inhibitory effect on BKPyV replication in primary human kidney epithelial cells [142]. It has also been well tolerated in healthy volunteers [143] and is therefore a potential drug candidate for treatment of BKPyV infections. Currently clinical trials with CMX001 treatment of infections with double-stranded DNA viruses in immunocompromised patients are under way.

Leflunomide, trade name Arava (Sanofi, Bridgewater, NJ, USA) is an orally administered drug that was approved by the U.S. Food & Drug Administration in 1998 for the treatment of rheumatoid arthritis. Leflunomide is converted to the active metabolite, A771726, presumably within the mucosal interstitia and liver. A771726 inhibits the mitochondrial enzyme dihydroorotate dehydrogenase, which is responsible for de novo pyrimidine synthesis. The lack of pyrimidines affects proliferation of T- and B-lymphoytes in particular, thereby causing an immunosuppressive effect. Leflunomide is therefore administered to kidney and liver transplant patients when existing immunosuppression is inadequate [144]. Moreover, it has been used as an antiviral treatment against CMV and herpes simplex virus type 1 infections. The suggested mechanism for this antiviral effect is prevention of tegumentation and thereby, virion assembly [145, 146]. Despite the fact that BKPyV is a non-enveloped virus without tegument, leflunomide was used for the first time in 2003 to treat PyVAN in a kidney transplant patient [147, 148] and is still in use for this indication. Leflunomide is often preferred over cidofovir as it is not nephrotoxic. It may, however, cause other severe side effects such as hemolysis, thrombotic microangiopathy, bone marrow suppression and hepatitis [149]. In vitro studies examining BKPyV replication found that A771726 inhibited BKPyV replication, but that this was closely connected to the significant cytostatic effects triggered by pyrimidine depletion [139, 150]. There are several case reports presenting favourable outcomes in both adult [151-154] and paediatric kidney transplant patients [155], but also reports where some patients experienced graft function deterioration [156, 157]. Of note, no difference in clearance of viremia was found between patients maintained at a high (>40 μg/mL) or low dose (<40 μg/mL) [149]. Similar to the clinical studies of cidofovir, immunosuppression is usually simultaneously decreased, making evaluation of the effect difficult.

Fluoroquinolones are synthetic broad-spectrum antimicrobial agents targeting the bacterial enzymes topoisomerase II and IV [158] and are also suggested to interfere with the helicase activity of BKPyV LTag as described for SV40 LTag [159]. A modest inhibition of BKPyV replication in vitro has been observed in several studies [160-163]. One study on BKPyV replication in primary renal tubular epithelial cells supported that fluoroquinolones might interfere with LTag but also suggested inhibition of the cellular topoisomerase II enzyme as an explanation of the reduced cellular DNA replication found [164]. A prophylactic effect has been reported in some non-randomized studies of HSCT and kidney transplant patients [162, 165, 166], but the effect was either short lived [167] or not observed in all patients [168].

The known immunomodulatory activity [169] and potential anti-BKPyV effect of intravenous immunoglobulin (IVIG) has given rise to empiric use in patients with PyVAN. Commercial IVIG contains pooled immunoglobulin G (IgG) from the plasma of approximately 1000 blood donors. The IVIG is found to contain neutralizing antibodies against BKPyV, reducing in vitro infections by 90% [170]. While some report favourable responses in patients receiving IVIG [171-174], others find no response [175] or even an increased viral load after IVIG administration [176]. The fact that some kidney transplant patients with PyVAN already have high titres of BKPyV IgG raises the question if antibodies really give any protection [177]. As for the other adjunctive therapies, the evaluation of an effect is complicated by the simultaneous reduction in immunosuppression. Although most of the adverse effects of IVIG are mild and transient [178], IVIG is in short supply and is an expensive treatment.

PyVHC

Therapy for PyVHC is purely supportive, involving symptom relief by analgesia, hyperhydration to increase diuresis and continuous bladder irrigation to prevent clot formation and urinary tract obstruction [83]. Moreover, lost platelets and erythrocytes are substituted.

The same adjuvant treatments have been tried for PyVHC as for PyVAN again without any documented benefit. In a recent study of five paediatric allogenic HSCT patients treated with intravenous ciofovir, intravesical cidofovir or both, only patients mounting BKPyV-specific IgM and IgG responses showed clinical resolution [125]. The authors speculated that the patients with increasing BKPyV antibody titres also had residual cellular immune competence which facilitated the resolution of PyVHC under cidofovir treatment. They also discussed the possibility of neutralization or clearing effect, which would suggest a limited to non-existent effect of cidofovir. If intravenous cidofovir is given as treatment for PyVHC, it should be given together with probenecid which prevents the uptake of cidofovir into renal tubular epithelial cells and thereby prevents nephrotoxicity [179].

Hyperbaric oxygen is frequently used for radiation induced haemorrhagic cystitis and has also been reported to be of benefit for patients with PyVHC [180]. Apparently the hyperbaric oxygen stimulates mucosal repair in the urinary bladder. In a retrospective study, 16 patients with PyVHC received 100% oxygen in a hyperbaric chamber at 2.1 atmospheres for 90 min, 5 days per week for between 4 and 84 times and 15 patients showed complete resolution of hematuria [181]. There are also reports on intravesicular instillation of formalin and aluminium to stop bleeding [182, 183], but formalin is known to give different side effects like ureteric stenosis and renal parenchymal damage while aluminium can cause intoxication especially in children [182].

Conclusions

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References

Polyomavirus BK persistently infects the majority of people at an early age, usually without causing disease. Pathological consequences appear mainly in immunodeficient individuals, an expanding patient group due to transplantations, use of immunosuppressive therapy and the increasing average age of the population. Safe and effective antiviral treatment of BKPyV diseases is still lacking and for patients with PyVAN, the only treatment strategy with documented effect is reduction of immunosuppressive therapy.

More than 40 years after the discovery of our intimate relationship with BKPyV major questions remain unanswered: How do we get infected? Does BKPyV establish a truly latent infection and if so, which genes, if any, are expressed? How does the natural course of infection affect our health? How does the virus interact with the innate and adaptive immune systems and avoid detection? What triggers reactivation? What causes NCCR rearrangement in association with development of disease and why is only wild type BKPyV transmitted? How is BKPyV infection controlled immunologically? What causes haemorrhagic cystitis in some patients and interstitial nephritis in other patients? The answers to these basic questions could lead to better understanding of risk, pathogenesis and treatment approaches, and will ultimately help to improve patient care.

We thank Dr Ruomei Li for kindly providing the electron microscopy picture and Dr Hans H. Hirsch for helpful comments.

References

  1. Top of page
  2. Abstract
  3. Structure and Genome Organization
  4. The Replication Cycle
  5. Epidemiology
  6. Associated Diseases
  7. Diagnosis
  8. Treatment
  9. Conclusions
  10. References