BK Polyomavirus: Clinical Aspects, Immune Regulation, and Emerging Therapies

SUMMARY

BK polyomavirus (BKV) causes frequent infections during childhood and establishes persistent infections within renal tubular cells and the uroepithelium, with minimal clinical implications. However, reactivation of BKV in immunocompromised individuals following renal or hematopoietic stem cell transplantation may cause serious complications, including BKV-associated nephropathy (BKVAN), ureteric stenosis, or hemorrhagic cystitis. Implementation of more potent immunosuppression and increased posttransplant surveillance has resulted in a higher incidence of BKVAN. Antiviral immunity plays a crucial role in controlling BKV replication, and our increasing knowledge about host-virus interactions has led to the development of improved diagnostic tools and clinical management strategies. Currently, there are no effective antiviral agents for BKV infection, and the mainstay of managing reactivation is reduction of immunosuppression. Development of immune-based therapies to combat BKV may provide new and exciting opportunities for the successful treatment of BKV-associated complications.

INTRODUCTION

BK polyomavirus (BKV) is a member of the Polyomaviridae family of double-stranded DNA (dsDNA) viruses. Different members of the family infect different species of mammals, primates, rodents, and birds. The list of human polyomaviruses has evolved over the past 2 decades (Fig. 1). BKV was first isolated from an immunosuppressed renal transplant recipient with ureteric stenosis in 1971 (1) and is named after the initials of this individual (1, 2). BKV causes a common childhood infection without major clinical sequelae, and >80% of adults are seropositive for BKV (3, 4). After primary infection, BKV remains dormant and does not cause significant morbidity in healthy individuals (5). Clinically significant reactivation of latent BKV occurs in some immunosuppressed individuals, such as following HIV infection or transplantation (6). Kidney transplant recipients (KTRs) comprise the patient population that most frequently experiences complications of BKV reactivation, and a median of 19.5% of KTRs experience BK viremia posttransplantation. A proportion of these recipients will go on to develop BKV-associated nephropathy (BKVAN), which is associated with a significant risk of allograft loss (6–8). In addition, hemorrhagic cystitis (HC) is a well-recognized BKV-associated complication in hematopoietic stem cell transplant (HSCT) recipients (9).

FIG 1.

Phylogenetic tree showing the distances of relationships between the human polyomaviruses discovered to date.

EPIDEMIOLOGY AND TISSUE TROPISM

Primary infection with BKV occurs during early childhood, and studies have shown that 70% of children are infected with BKV by the age of 10 years; however, the route of transmission remains unclear (4). The primary mode of transmission is speculated to be via a respiratory route, as supported by evidence of BKV infection in the respiratory tract and tonsils of children; however, other routes of transmission are also proposed (Table 1). Primary infection is usually asymptomatic or, rarely, causes a mild respiratory illness (10). It has been suggested that BKV enters the bloodstream via infected tonsillar tissue, thereby infecting peripheral blood mononuclear cells and disseminating virus to secondary sites of infection, including the kidney (3). Following resolution of primary infection, BKV persists primarily in kidney epithelial cells for life, with occasional reactivation manifesting as asymptomatic viruria (11). BKV has also been found in leukocytes, the brain, and lymph nodes, as evidenced by the presence of BKV DNA (3, 12). Viral shedding in urine is more common for immunocompromised patients than for healthy individuals (13). The mechanism of viral persistence and the conditions that lead to viral reactivation upon immunosuppression are still unclear.

TABLE 1.

Route of transmission of BKV

Route of transmission	Source of BKV identification	Reference(s)
Respiratory	Upper respiratory tract	10
	Tonsils	10
	Waldeyer's ring	232
Fecal-oral	BKV DNA in sewage	233
Blood transfusion	Leukocytes	3
Transplacental	Fetus	234–236

VIRION STRUCTURE

Members of the Polyomaviridae family demonstrate structural similarity, with similar capsid sizes, high levels of genetic homology, and comparable genome sizes. BKV has a small, nonenveloped, icosahedral capsid with a diameter of 40 to 44 nm comprised of the virus-encoded capsid proteins VP1, VP2, and VP3 (14) (Fig. 2). The capsid proteins surround a single molecule of DNA complexed with histone proteins in the form of chromatin chains. The capsid proteins are arranged in a T=7 icosahedral structure containing 360 molecules of VP1 organized into 72 pentamers. Each pentamer is linked to a single moiety of the minor capsid proteins, VP2 and VP3, internally (Fig. 2). Therefore, VP1 is the only viral protein exposed on the outside of the virion and is responsible for attachment of the virus to host cell receptors that aid in entry of virus into the cell. In a recent study of BKV structure by use of cryo-electron microscopy, Hurdiss et al. showed that the minor capsid proteins (VP2 and VP3) have discrete points of contact with the histones and genomic DNA (Fig. 2) (15). They also proposed a model in which the minor capsid proteins could act as a bridge between VP1 and the chromatinized viral DNA; however, the significance of such contact with the genomic DNA has yet to be studied.

FIG 2.

Cryo-electron microscopy structure of BKV, showing an external view, minor capsid proteins, and genome organization. (A) External view of the virion at a contour level of 0.022. The electron density maps were sharpened using a negative B factor correction (B = −456 and −804 Ų). (B) View of a 40-Å-thick slab through the unsharpened/unmasked virion map, shown at a contour level of 0.0034. Pyramidal density below each VP1 penton and two shells of electron density adjacent to the inner capsid layer can be seen. The density within 6 Å of the fitted coordinates for SV40 VP1 is colored gray. The remaining density is colored in a radial color scheme. Densities for VP2 and VP3 are colored blue and green, and those for packaged dsDNA are yellow and pink. (C) Enlarged view of the pyramidal density beneath a single VP1 penton of the virion, shown at a contour level of 0.0032. Strands of dsDNA wrapped around a human histone octamer (PDB entry 1AOI) are shown, indicating that the two shells of density have comparable spacings. A discrete connective density between the pyramidal density and internal shells is also apparent. (Adapted from reference 15 [published under a Creative Commons {CC BY} license].)

MECHANISM OF CELL ENTRY

BKV entry into host cells is mediated via caveolae, unlike that of other polyomaviruses, which typically use clathrin-mediated entry (Fig. 3). VP1 has a cleft between the β-strand C1 (BC1) and BC2 loops that can bind to α2-8-linked disialic acid motifs on gangliosides GD1b and GT1b expressed on host cell membranes (16–19). Entry into the cell is then driven by a caveola-mediated endocytic pathway (16, 20). Following internalization, BKV is carried toward the endoplasmic reticulum by use of microtubules, from which it follows the classical endocytic pathway for capsid uncoating (21, 22). VP2 and VP3 mediate the entry of BKV into the nucleus via importin after the uncoating of VP1 (23–25). After entering the nucleus, the BKV genome remains episomal in human cells, in contrast to the integration of the BKV genome into host DNA of rodent cells, which leads to the development of tumors in rodents (26–28). While little is known about the mechanism of integration of BKV DNA into rodent genomic DNA leading to malignancy, it would be interesting to study the factors involved in such transformation. These may provide insights into the involvement of BKV in oncogenesis in humans, which is higly debated.

FIG 3.

Schematic outline of BKV entry into host cells. For caveola-mediated entry of BKV into host cells, attachment of VP1 to the ganglioside receptors GD1 and GT1 initializes the internalization of BKV through caveola formation. The caveola encapsulating the virus then translocates in the cytoplasm with the help of microtubules. The virus then fuses with the endoplasmic reticulum (ER), where the VP1 layer disassembles. The virus then enters the host nucleus and lies episomally. In the nucleus, the early coding regions are transcribed first, which regulates the transcription of late coding regions.

GENOMIC STRUCTURE AND REPLICATION

The BKV genome is approximately 5,300 bp long and contains genes coding for the structural proteins (VP1, VP2, and VP3) and the viral replication proteins (large T antigen [LTA], small T antigen [STA], and agnoprotein) and a noncoding control region (NCCR) (Fig. 4). The length of the genome varies in individual variants due to the alterations in the NCCR. The NCCR is a hypervariable region containing various binding sites for host cellular regulatory factors, and hence it is also referred to as the hypervariable regulatory region (HVRR) (26, 29, 30). The open reading frame is situated in the center of the genome such that replication can proceed in a bidirectional way (31) (Fig. 4). The early coding regions are transcribed before the start of replication, promoting the expression of LTA and STA. These antigens accumulate in the nucleus and help in the replication of viral DNA (32, 33). During the viral DNA replication process, LTA forms a multimeric complex which binds to the origin of replication and acts like a helicase to facilitate the transcription of late coding regions (34, 35). LTA is also a key regulatory molecule driving the host cell to S phase of the cell cycle by binding to the tumor suppressor proteins Rb, p107, p130, and p53 (30, 36–38). Based on studies with simian virus 40 (SV40), STA is involved in viral replication, cell cycle progression, and transformation (39–41). The late coding regions are expressed only after the onset of viral DNA replication, as they code for structural proteins involved in viral packaging as well as the agnoprotein (26). The viral capsid proteins, VP1, VP2, and VP3, are produced in the cytoplasm and recruited into the nucleus by use of the nuclear localization signals attached to the proteins. Once the capsid proteins enter the nucleus, viral assembly occurs and the viral progenies accumulate in the nucleus (42, 43).

FIG 4.

Genome stucture of BKV. The transcription of both early and late coding regions proceeds in a bidirectional way from the origin of replication (ORI) within the noncoding control region (NCCR). The transcriptional splicing regions are represented by dashed lines. The late coding regions encode structural proteins (VP1, VP2, and VP3), while the early coding regions transcribe the tumorigenic proteins (LTA and STA). The late coding regions also encode the nonimmunogenic agnoprotein. The expression of miRNA complementary to the 3′ end of LTA has been shown to be involved in replication control of BKV.

BKV GENOTYPES AND VARIANTS

BKV can be categorized into four genotypes based on sequence variation in the genomic region of VP1 (44, 45). Serological studies indicate that BKV genotype I has the highest prevalence in the human population, followed by genotype IV. Genotypes II and III are found to infect only a minority of adults. BKV genotype I is further classified into four subgroups: subgroups Ia, Ib1, Ib2, and Ic (46, 47). BKV genotype IV can be further subgrouped into subgroups IVa1, IVa2, IVb1, IVb2, IVc1, and IVc2 (48). The distribution pattern of each of the genotypes has been studied extensively. Genotype I viruses are present worldwide, while genotype IV is found in northeastern Asia and Europe (46–51). The clinical and immunological implications of infection with the different genotypes of BKV are still unknown. Apart from the genotypes based on the variation in the VP1 region, there are also two other forms of BKV due to the variation in the NCCR, namely, archetype (ww) and rearranged (rr) variants. Archetypal BKV contains a linear block of O-P-Q-R-S, where O represents the origin of replication and P-Q-R-S represents promoters and regulatory regions of early and late coding regions (52). Deletion and duplications in the NCCR pattern sequences occur due to the continuous replication of the viral DNA during reactivation, leading to the generation of the rearranged variant viruses (53). These genetic variants of BKV are commonly found in individuals with BKV-associated diseases along with the archetypal BKV. While preliminary studies have suggested that these variations and rearrangements are not involved in the development of BKV-associated diseases (52, 54), more detailed studies on the genetic variants of the NCCR in disease development are required. Furthermore, it is also important to appreciate that BKV can persist in many tissues throughout the body, and analysis of viral sequences from renal tissue and serum alone may miss potential sources of pathogenic virus within other tissue reservoirs (e.g., the central nervous system [CNS] and lymph nodes). Analysis of viral sequences from these tissues may reveal novel BKV genotypes and variants which correlate wth disease development.

VIRAL PERSISTENCE

BKV establishes lifelong persistent infection in the host. It is still unclear whether BKV stays latent in the host cell or maintains a low level of gene expression with persistent infection. The mechanism of latency is well studied for herpesviruses, which has highlighted the regulation of various genes controlling viral lytic replication (55). BKV encodes microRNAs (miRNAs) similar to those of herpesviruses, which act as regulators of viral replication. Recently, Broekema and Imperiale showed that a high level of expression of miRNA complementary to the 3′ end of the LTA mRNA is responsible for control of BKV replication (Fig. 4) (56). Further studies have shown that miRNA expression suppresses the autoregulation of viral replication (57). While no other mechanism of latency has yet been reported, the evidence regarding the association of miRNA with latency is promising and requires more detailed research.

IMMUNE RESPONSES TO BKV

Innate Immune Response

While a few studies have investigated the interaction of BKV and the innate immune system, there are few data regarding the role that innate immune mediators play in controlling BKV infection. Most studies have focused upon settings of viral reactivation and disease (BKVAN) in KTRs. Womer and colleagues showed a role for dendritic cells (DCs) in BKVAN, demonstrating a reduced number of DCs in the peripheral blood of renal transplant recipients with BKVAN. Further studies demonstrated that patients with fewer DCs before transplantation are at a higher risk of BKVAN (58). DCs also likely play a critical role in promoting the induction of the adaptive immune response, particularly the generation of BKV-specific T cells. Using a mouse model of BKVAN, Drake and colleagues showed that an increased number of DCs in vivo correlated with an enhanced magnitude of virus-specific CD8⁺ T cells (59, 60). These studies strongly suggest that DC activation likely plays an important role in priming protective immune responses against BKV; however, the mechanisms of DC activation following BKV encounter remain unclear. Unlike those of other polyomaviruses, the VP1 protein of BKV does not induce strong maturation of DCs (61).

Natural killer (NK) cells are also likely to play a role in controlling BKV infection in KTRs. Bohl and colleagues showed that the absence of HLA C7 from the donor or recipient was associated with an increased risk of viremia (62). While the precise mechanism for this association still remains unclear, more recent observations by Trydzenskaya et al. proposed a role for killer cell immunoglobulin-like receptors (KIR) in the immune defense against BKV. They found that patients with BKVAN had a small number of activating KIR, in particular those of the KIR3DS1 genotype, in NK cells compared to the control group (63). Recent studies showed that BKV immune evasion strategies may be mediated via the inhibition of NK cell recognition. Bauman demonstrated that BKV produces miRNAs which can suppress the expression of the stress-induced ligand ULBP3, which is recognized by the NKG2D receptor on NK cells, thereby avoiding NK cell-mediated cytotoxicity (64).

Other innate mediators have been shown to contribute to BKV control. Defensins are key mediators of innate immunity, exhibiting antimicrobial activity against a wide range of viruses, fungi, and bacteria. Dugan et al. showed that human α defensin 5 (HD5) inhibits binding of BKV to host cells by aggregating viruses, thereby blocking binding of viruses to the cells (65). HD5 is commonly found in the urogenital tract, which is also the resident site of BKV, and it was demonstrated that HD5 could neutralize BKV in a serum-independent manner, suggesting that it may play an important role in mediating asymptomatic reactivation of BKV in the urogenital tract.

Innate immune mediators also likely play a role in graft loss in KTRs with BKVAN. BKVAN is associated with renal inflammation, the increased expression of interleukin-6 (IL-6), IL-8/CXCL8, RANTES/CCL5, MCP-1/CCL2, and IP-10/CXCL10 in kidney biopsy specimens (66), and an upregulation of two proinflammatory genes, encoding pentraxin 3 (PTX3), a cytokine-inducible protein, and MICB, which interacts with NKG2D receptors on NK cells during BKV infection in kidney proximal tubular cells (67). Furthermore, it has been shown that the microenvironment due to BKV infection promotes graft fibrosis in BKVAN patients, which is evidenced by the increased expression of matrix collagens, transforming growth factor beta (TGF-β), and MMP2 and -9, as well as markers of epithelial-mesenchymal transformation (EMT), in BKVAN biopsy specimens (68).

Adaptive Immune Response

Humoral response.

Adaptive immune responses against viruses typically develop soon after initial exposure to viral antigens. BKV is usually encountered early in childhood, which induces BKV-specific neutralizing antibody responses. In a study of healthy individuals, Egli and coworkers found BKV-specific IgG antibodies in 87% (87 of 100 individuals) of younger individuals (aged 20 to 29 years) and 71% (71 of 100 individuals) of older individuals (aged 50 to 59 years) (69). Similarly, Randhawa and colleagues found that 80% (57 of 71 individuals) of kidney donors had IgG antibodies and 22% (16 of 71 individuals) had IgA antibodies specific to BKV (70). Reports suggest that the presence of BKV-specific antibodies alone does not provide protection against BKV reactivation and associated diseases (71).

In most viral infections, antibodies serve as neutralizing agents by various mechanisms, one of which is attachment of antibodies to viral receptors, thereby restricting further infection by the virus. Mutations in viral receptors can cause escape from antibody-mediated neutralization (72–74). Pastrana and colleagues recently showed that certain BKV genotypes can escape neutralization from antibody raised against another genotype due to variation in antibody receptor binding (75). The same group had earlier postulated that KTRs who lack antibodies capable of neutralizing a wider range of BKV serotypes are at greater risk of graft rejection (76). These results suggest that developing cross-neutralizing antibodies that can act on different BKV genotypes may be a potential strategy to restrict BKV infection before immunosuppression therapy in KTRs. Uncontrolled replication of BKV in the kidney during BKVAN results in an increase in the viral load in blood, which in turn can induce a higher humoral response. Although BKV-specific antibody responses likely play an important role in neutralizing circulating virus, antibody alone is unable to control persistent latent infection (71). Similar to all persistent viral infections, the efficient control of latent viral reactivation is most likely dependent upon the induction of stable antiviral memory T cell responses.

T cell responses.

Recent studies have begun to elucidate the important role of T cells in controlling BKV infection. It has been demonstrated that in KTRs, reconstitution of BKV-specific T cells is associated with better control of viruria and viremia (8, 71, 77–81). These studies indicate that immune control is dependent on both CD4⁺ and CD8⁺ T cells (82) and correlates with the frequency of polyfunctional BKV-specific T cells (83). This suggests that monitoring the frequency of BKV-specific T cells following transplantation may provide a strategy to predict the risk of viral reactivation and BKVAN, as shown previously for T cell responses to human cytomegalovirus (CMV) in transplant recipients (84). Analysis of the T cell responses from BKV-seropositive individuals has shown that BKV-specific T cell responses can be detected against all five BKV proteins in both seropositive healthy individuals and transplant patients (77, 79–81, 85–88). T cells specific for VP1 and LTA are most frequently detected in BKV-infected individuals (89, 90). While BKV-specific CD4⁺ and CD8⁺ T cells can be detected in healthy individuals and in KTRs, it has been suggested that the T cell response to BKV is mediated predominantly by CD4⁺ T cells (87, 88, 91). Recent reports have shown that CD4⁺ T cells likely have a direct role in controlling BKV infection, mediated through the expression of proinflammatory cytokines, including gamma interferon (IFN-γ) and tumor necrosis factor (TNF), and via the expression of the cytolytic molecule granzyme B (92). These studies also suggest that CD4⁺ T cells can control BKV reactivation even in the absence of CD8⁺ T cell immunity (92). More recently, the same group reported that the frequency of CD4⁺ T helper cells significantly increased before and after the clearance phase, while cytolytic CD4⁺ cells increased during the clearance phase (93). They also reported that immunosuppressant (IS) drugs used to prevent graft rejection in KT patients can cause a reduction in the expression of cytokines, such as IFN-γ and TNF, by BKV-specific T cells. Other studies of KTRs with active BKV replication showed that CD8⁺ T cells were predominantly LTA specific, while VP1 elicited a mainly CD4⁺ T cell response (81). Mueller and colleagues recently showed VP3 to be an antigenic target eliciting both CD4⁺ and CD8⁺ T cell responses in BKVAN patients (87).

Recent studies have also begun to investigate the presence of BKV-specific T cells in renal tissue and their association with BKVAN (94, 95). Zeng and colleagues tracked and quantitated BKV-specific T cells expanded from peripheral blood in renal allograft biopsy specimens with or without BKVAN (96). While T cell receptor analysis using next-generation sequencing revealed the presence of virus-specific T cells in biopsy specimens from patients with BK viremia or nephropathy, interestingly, these biopsy specimens contained 7 to 8 times more alloreactive T cell clones than virus-specific T cells. These observations suggest that the tissue injury during BKV-associated nephropathy is mediated primarily by the influx of bystander secondary T cells triggered by both alloreactive and virus-specific immunity.

One of the major roadblocks in studying BKV-specific T cells has been their low precursor frequency in the peripheral blood of both healthy volunteers and KTRs (79). Hence, most studies have used in vitro-expanded virus-specific T cells stimulated with antigenic peptide pools or BKV lysate. Typically, overlapping peptide pools (OPP) of BKV proteins have been used to stimulate T cells because BKV-specific T cell epitopes have not been characterized comprehensively (87, 91). Trydzenskaya et al. demonstrated that T cell stimulation using overlapping peptide pools of all five BKV antigens improved the sensitivity compared to that with single-antigen stimulation (91). This approach of using mixed peptide pools for BKV-specific T cell responses was also used by Mutlu et al. to attain maximum sensitivity for the quantitation of BKV-reactive CD4⁺ T cells prior to or following transplantation (97).

A number of studies have identified potential CD4⁺ and CD8⁺ T cell determinants from BKV, although thorough investigations in a large cohort of healthy donors and kidney patients have not been undertaken (86, 88, 98–103). These early studies identified BKV-specific CD8⁺ T cell epitopes restricted through HLA A1, A2, A3, A24, B7, and B8 (79, 86, 99, 101, 103, 104). A detailed list of CD8⁺ T cell epitopes mapped from BKV is presented in Table 2. More recently, Cioni and colleagues employed a combination of bioinformatics and an in vitro expansion strategy to predict 9-mer epitopes from the BKV early viral gene region (105). These epitopes were specifically mapped for 14 major HLA class I alleles prevalent in Europe and North America. Using this highly innovative approach, Cioni et al. successfully expanded the list of BKV T cell epitopes to at least 39, among which 21 epitopes were further confirmed by HLA peptide streptamer staining (105). Sahoo et al. assessed genetic variability in sections of the BKV genome coding for T cell epitopes, demonstrating very limited variability (less than 5%) in a cohort of 65 samples (106). Cioni and colleagues demonstrated that some of the epitopes they identified shared strong homology with JC polyomavirus (JCV) (105). Similar observations have been seen in other studies investigating an overlap between BKV and JCV immunity (90, 98, 99). These studies also imply that prior JCV infection may induce T cell responses that cross-protect from BKV-associated diseases, and vice versa, due to the high level of homology between both viruses. These observations have defined the conserved nature of BKV immunity and its potential overlap with JCV, providing a platform for the development of a standardized immune monitoring protocol for BKV that may potentially be extended to include JCV. While this research has led to the determination of a significant number of BKV-specific T cell epitopes, it is imperative to expand these analyses to comprehensively map T cell epitopes within all five antigenic proteins of BKV and to expand the HLA restriction to alleles which are common in ethnic groups other than European and North American groups. This will help in developing effective immune monitoring tools for BKV-associated diseases, particularly for regions with high rates of kidney transplantation outside Europe and North America.

TABLE 2.

List of HLA class I-restricted CD8⁺ T cell epitopes mapped from BKV

Epitope sequence	Antigen	HLA restriction	Amino acid positions	Tested in healthy controls	Tested in transplant recipients	Reference(s)
MLTERFNHIL	LTA	HLA-A*02:01	362–370	Yes	Yes	85
VIFDFLHCI	LTA	HLA-A*02:01	406–414	Yes	Yes	85
FLHCIVFNV	LTA	HLA-A*02:01	410–418	Yes	Yes	85
LLMWEAVTV	VP1	HLA-A*02:01	108–116	Yes	Yes	79
AITEVECFL	VP1	HLA-A*02:01	44–52	Yes	Yes	79
LLLIWFRPV	LTA	HLA-A*02:01	579–587	Yes	yes	99, 103
LPLMRKAYL	LTA	HLA-B*07:02/B*08	27–35	Yes	Yes	99, 103
CLLPKMDSV	LTA	HLA-A*0201	398–406	Yes	No	99, 103
KLCTFSFLI	LTA	HLA-A*0201	216–224	Yes	No	99, 103
RLDSEISMY	LTA	HLA-A*0101	604–612	Yes	Yes	104
VSWKLITEY	LTA	HLA-A*0101	270–278	Yes	Yes	104
YSALTRDPY	LTA	HLA-A*0101	235–243	Yes	Yes	104
WSSSEVPTY	LTA	HLA-A*0101	77–85	Yes	Yes	104
FLICKGVNK	LTA	HLA-A*0301	222–230	No	Yes	104
ILYKKLMEK	LTA	HLA-A*0301	172–180	Yes	Yes	104
SVKVNLEKK	LTA	HLA-A*0301	506–514	Yes	Yes	104
RTLACFAVY	LTA	HLA-A*0301	156–164	Yes	Yes	104
ACFAVYTTK	LTA	HLA-A*0301	159–167	Yes	Yes	104
GVNKEYLLY	LTA	HLA-A*0301	227–235	No	Yes	104
IVFNVPKRR	LTA	HLA-A*0301	414–422	Yes	Yes	104
SAINNFCQK	LTA	HLA-A*0301	208–216	Yes	Yes	104
AWLHCLLPK	LTA	HLA-A*0301	394–402	No	Yes	104
AYLRKCKEF	LTA	HLA-A*2402	33–41	Yes	Yes	104
PYHTIEESI	LTA	HLA-A*2402	242–250	Yes	Yes	104
QYMAGVAWL	LTA	HLA-A*2402	388–396	Yes	Yes	104
VFLLLGMYL	LTA	HLA-A*2402	287–295	No	Yes	104
RYWLFKGPI	LTA	HLA-A*2402	422–430	No	Yes	104
RTLACFAVY	LTA	HLA-A*0101	156–164	No	Yes	105
GVNKEYLLY	LTA	HLA-A*01/A*11	227–235	No	Yes	105
DVFLLLGMY	LTA	HLA-A*0101	286–294	No	Yes	105
KLCTFSFLI	LTA	HLA-A*0201	216–224	No	Yes	105
MYLEFQYNV	LTA	HLA-A*0201	293–301	No	Yes	105
TLAAGLLDL	LTA	HLA-A*0201	436–444	No	Yes	105
ILYKKLMEK	LTA	HLA-A*0301	172–180	No	Yes	105
LLLGMYLEF	LTA	HLA-A*24/B*51	289–297	No	Yes	105
LERAAWGNL	LTA	HLA-B*0702	19–27	No	Yes	105
FPSDLHQFL	LTA	HLA-B*0702	140–148	No	Yes	105
MLTERFNHIL	LTA	HLA-B*0801	362–370	No	Yes	105
LLLIWFRPV	LTA	HLA-B*0801	579–587	No	Yes	105

CLINICAL AND LABORATORY DIAGNOSIS OF BKV INFECTION

BKV predominantly causes disease in immunocompromised patients. The major clinical manifestations are BKV-associated nephropathy (BKVAN) or ureteric stenosis in KTRs and HC, which is typically seen only in HSCT recipients. Other, less common clinical manifestations that have been linked to BKV infection include interstitial pneumonitis and meningoencephalitis (107–116). BKV-associated central nervous system infections are predominantly seen in patients who have an underlying immunodeficiency or comorbid illness, including HSCT or HIV infection. Clinical manifestations in these patients include headache, confusion, ataxia, dizziness, paraplegia, and seizures (114).

PCR-based viral load quantitation in the plasma, urine, or cerebrospinal fluid (for CNS infection) is the standard clinical tool for monitoring BKV reactivation (117). Several reports have shown that sustained viremia in the plasma of renal transplant patients is associated with a higher risk of BKVAN development (118–121). Although qualitative PCR is highly sensitive for detecting active viral replication, it has a relatively low positive predictive value (PPV) (30 to 50%) for BKVAN (120, 121). Studies reporting quantitative BKV PCR results demonstrate a positive correlation between higher viral loads and an increased probability of developing BKVAN. For example, in a prospective study of KTRs, Hirsch et al. found that all recipients with BKVAN had viral loads of >7,700 copies/ml (121). Based on this, a threshold of 1 × 10⁴ copies/ml of BKV has been proposed as a threshold for improving the PPV, although the lack of a standardized protocol for BKV PCR assays has created difficulty in directly comparing viral loads between studies. However, the World Health Organization Expert Committee on Biological Standardization (ECBS) recently published an international standard for BKV PCR-based assays, which will hopefully lead to harmonization of BKV loads between laboratories (122).

More recently, measurement of BKV mRNA levels for the detection of active BKV replication was reported (123), and a noninvasive method to detect BKV mRNA levels in urine was shown to be highly specific and sensitive. The method showed 93.9% specificity using 6.5 × 10⁵ BKV VP1 mRNAs/ng RNA in urinary cells as a cutoff limit (123, 124). Though this mRNA-based method for detecting BKV replication is promising, this method requires further validation as a tool for predicting patients who will develop more severe features of BKV infection, such as BKVAN. Singh and colleagues recently described a polyomavirus Haufen test as a noninvasive biomarker to predict BKVAN (125, 126). Haufen is the term used to describe BKV aggregates in the urine; cast-like three-dimensional BKV aggregates are detected using electron microscopy. This method is reported to have a high positive predictive value (more than 90%) for BKVAN (125). However, because the method has been published only as a single-center report, it requires further validation, and the assay is not feasible for routine clinical practice due to the cost and limited availability of electron microscopy.

BKV REACTIVATION IN KIDNEY TRANSPLANT RECIPIENTS

Evidence of BKV reactivation is frequently detected in kidney transplant recipients receiving contemporary immunosuppressive regimens, as summarized in Table 3. Although less common, there have also been reports of BKV reactivation and BKVAN occurring in recipients of nonrenal solid organ transplants (127, 128). Viruria or detection of urinary decoy cells is the most sensitive marker of BKV reactivation, occurring in 23 to 73% of KTRs (Table 3). BK viremia affects 8 to 62% of KTRs, with the peak incidence occurring at 3 to 6 months posttransplantation (95, 129), and the incidence of BKVAN in the first year posttransplantation is reported to be in the range of 1 to 7% (Table 3). The wide range of reported frequencies of BKV reactivation may be due to the extensive variation in immunosuppressive regimens utilized at different transplant centers, as well as to differences in frequency of monitoring and assay sensitivity (130). Clinically, BKV reactivation is asymptomatic, and if BKVAN occurs, this manifests as deterioration in allograft function (118, 121, 131, 132). As a consequence, most clinicians now screen KTRs prospectively for BKV reactivation, either by monitoring for decoy cells in urine or by quantitative PCR analysis of urine or peripheral blood for BKV detection (Table 4) (133–135). A definitive diagnosis of BKVAN requires a kidney biopsy, which is typically performed when BK viremia persistently exceeds 1 × 10⁴ copies/ml or if there is evidence of allograft dysfunction. Histologically, viral cytopathic changes affecting tubular epithelial cells are the key manifestation of BKVAN, characterized by nuclear enlargement and basophilic viral inclusions (136). Typically, there is also significant tubular cell injury and associated interstitial inflammation potentially leading to misdiagnosis of acute cellular rejection, particularly during resolution of BKVAN, when viral inclusions are less prominent (137). Persistent BKVAN eventually leads to renal parenchymal scarring with progressive tubular atrophy and interstitial fibrosis (136, 138, 139). The presence of BKV within renal tissue can be confirmed via immunohistochemistry using antibodies reactive to the large T antigen of simian virus 40 (SV40), which cross-react with BKV (140). This helps to distinguish the interstitial inflammatory changes associated with BKVAN from acute cellular rejection and is pathognomic of BKV replication within the kidney. Classification of BKVAN into three stages based on the severity of the histological findings has been proposed, since more severe changes are associated with higher BKV viral loads and poorer allograft survival (94, 138).

TABLE 3.

Reported frequencies of BKV reactivation and BKVAN in kidney transplant recipients^a

Study author(s), yr (reference)	Initial immunosuppression regimen	% of recipients with:				No. of recipients
		BK viruria/decoy cells	BK viremia	Biopsy-confirmed BKVAN	Graft loss due to BKVAN
Hirsch et al., 2002 (121)	T, A, P or C, M, P	23	10	5	Nil	78
Brennan et al., 2005 (129, 161)	ATG, T/C, M/A, P	35	12	Nil	Nil	200
Bessollette-Bodin et al., 2005 (13)	ATG/IL-2, T/C, P	57	29	Nil	Nil	104
Drachenberg et al., 2007 (153)	T, M, P	73	62	6	Nil	103
Dadhania et al., 2008 (237)	ATG/IL-2, T, M, ±P	NR	31	7	Nil	120
Almeras et al., 2011 (159)	ATG/IL-2, T/C/S, P	NR	11	<1	<1	119
Chakera et al., 2011 (134)	IL-2, T, A/M, P or AZM, T, M	18	8	1	Nil	313
Sood et al., 2012 (238)	ATG/IL-2, T, M, P	17	27	2	Nil	240
Barbosa et al., 2013 (239)	AZM/IL-2, R, IVIG, T, M, P	NR	20	2	<1	187
	ATG/IL-2, T/C, M, P	NR	10	1	<1	284
Borni-Duval et al., 2013 (147)	T/C, M, P	40	20	7	NR	240
Hirsch et al., 2013 (94)	IL-2, T/C, M, P	40	23	NR	NR	629
Schaub et al., 2010 (162)	IL-2/ATG, T/C, M, P/S	NR	19	6	Nil	203
Theodoropoulos et al., 2013 (152)	AZM/IL-2, T, M, ±P	38	12	5	1	666
Knoll et al., 2014 (194)	IL-2/ATG, T/C, M, P	NR	31	Nil	Nil	154
3C Study Collaborative Group et al., 2014 (151)	AZM, T, M	NR	7	1	Nil	426
	IL-2, T, M, P	NR	3	2	Nil	426
Schwarz et al., 2016 (240)		40	29	10	1	214
Sawinksi et al., 2015 (241)	ATG, T, M, P	NR	17	2	Nil	785
Wunderink et al., 2017 (142)	AZM/IL-2, T/C, M, P	NR	27	3	Nil	407

^aT, tacrolimus; C, cyclosporine; A, azathioprine; M, mycophenolate; S, sirolimus; P, prednisolone; AZM, alemtuzumab; ATG, thymoglobulin; IL-2, basiliximab/daclizumab; R, rituximab; IVIG, intravenous immunoglobulin; NR, not reported.

TABLE 4.

Diagnostic testing and prognostic values for BKV infection and disease

Diagnostic method	Sensitivity for detection of BKV infection (%)	Specificity for detection of BKV infection (%)	Positive predictive value (PPV) for diagnosis of BKVAN (%)	Negative predictive value (NPV) for diagnosis of BKVAN (%)	Comment(s)	Reference(s)
Urine cytology for decoy cells	>80	70–84	20–35	>95	Useful for determining BKV reactivation but low PPV for BKVAN	120, 121, 242–244
Urine PCR	>98	78	30–40	>95	Effective at detecting possible BKVAN with BKV loads of >1 × 107 copies/ml	118, 120, 245, 246
Serum PCR	90–100	83–96	50–80	>95	Highly specific and sensitive for detecting BKV reactivation; PPV for BKVAN increases with a higher BKV load; a cutoff of 1 × 104 copies/ml has been suggested as a threshold for biopsy specimens to exclude BKVAN	118, 120, 121, 243, 245, 246
Haufen detection (electron microscopy)	100	>95	>95	100	Higher reported PPV than that for any other method, but the method is expensive	125, 126
Histopathology			>98	100	Kidney biopsy is the gold standard for determining disease progression	140
					Stage/class A: infection/cytopathic changes, <25%; interstitial inflammation/tubular atrophy/fibrosis, <10%	94
					Stage/class B: infection/cytopathic changes, 11–50%; interstitial inflammation/tubular atrophy/fibrosis, <50%
					Stage/class C: infection/cytopathic changes, >50%; interstitial inflammation/tubular atrophy/fibrosis, >50

RISK FACTORS FOR BKV REACTIVATION

The frequencies of BKV reactivation and BKVAN after kidney transplantation vary considerably in the published literature (Table 3). Several factors have been identified that modify the risk of BKV reactivation in KTRs, although the overall degree of immunosuppression is thought to be the single largest factor promoting BKV reactivation (Fig. 5). It has been shown that both donor BKV seropositivity and recipient BKV seronegativity increase the risk of developing BKVAN and graft rejection, with a 10-fold higher risk of BK viremia in KTRs in cases where the donor was seropositive and the recipient seronegative than in cases where both were seronegative (141, 142). Recently, a number of studies also showed that not just recipient seronegativity but also the differential titers of BKV-specific antibodies are associated with the risk of viremia and viruria in KTRs (141–144). Higher anti-BKV IgG titers in donors and lower anti-BKV IgG titers in recipients increase the risk of early BK viremia (142, 143). These reports suggest that increased vigilance for BKV in kidney donors may identify KTRs at greater risk for BKV reactivation posttransplantation. Furthermore, Verghese et al. recently showed that transplantation from donors with active BKV viruria increased the risk of BK viremia in the recipients (145).

FIG 5.

Identified risk factors for BKV infection and/or reactivation in transplant recipients. These risk factors are associated with either the immunological, age, sex, or metabolic status of the recipient and donor or the physiological/clinical profile of the transplanted organ. Each of these risk factors can directly or indirectly increase the risk of BKV infection and/or reactivation.

KTRs typically receive both induction and maintenance immunosuppression. Induction agents are potent immunosuppressive medications administered early after transplantation, when the risk of acute rejection is highest (146). Maintenance immunosuppression is required indefinitely after transplantation to prevent chronic rejection and promote long-term allograft survival. Induction therapy is typically a monoclonal or polyclonal antibody that depletes host lymphocytes (thymoglobulin/ATG or alemtuzumab) or that targets CD25, the high-affinity IL-2 receptor alpha chain (basiliximab and daclizumab). Administration of antithymocyte globulin is associated with a longer duration of BK viremia as well as a higher incidence of BKVAN than those with induction with an anti-CD25 monoclonal antibody (MAb) (147–149) or no induction (121). This is consistent with the view that the major driver of BKV reactivation is immunosuppression, since lymphocyte-depleting therapy is significantly more immunosuppressive than CD25 blockade, as reflected in a much lower incidence of acute rejection (150). Similar findings were observed in the large 3C trial, which reported a significantly higher rate of BK viremia in the first 6 months following kidney transplantation for recipients randomized to alemtuzumab induction than for those receiving basiliximab (hazard ratio, 1.92 [95% confidence interval, 1.06 to 3.45]; P = 0.03) (151). This is in contrast to a smaller single-center retrospective series that did not identify alemtuzumab induction as a risk factor after adjustment for other potential confounders (152).

Maintenance immunosuppression for most KTRs consists of three immunosuppressive medications, typically a calcineurin inhibitor (tacrolimus or cyclosporine), an antiproliferative agent (mycophenolate or azathioprine), and a corticosteroid. Tacrolimus is associated with lower acute rejection rates than those with cyclosporine and is therefore considered a more potent immunosuppressive agent. In keeping with this, several studies suggest that tacrolimus is associated with a greater risk of BK viremia than that with cyclosporine. The DIRECT trial was a prospective open-label randomized controlled trial comparing tacrolimus to cyclosporine; all recipients received mycophenolate and corticosteroids (94). The incidence of BK viremia at 6 months posttransplantation was 16.3% in the tacrolimus arm, compared to 10.6% in the cyclosporine arm. This positive association between tacrolimus and BK viremia has been observed in other studies (95, 121, 147), as well as in U.S. registry data (148). The combination of tacrolimus, mycophenolate, and corticosteroids appears to confer a particularly high risk of BK viremia compared to that with alternative regimens (147, 153, 154), and there is also a positive correlation between greater exposure to corticosteroids and BK viremia (94, 121, 155). In contrast, BK viremia and BKVAN appear to be less common in patients receiving regimens based on mTOR inhibitors (sirolimus or everolimus), which are considered less potent immunosuppressive agents than calcineurin inhibitors (148, 156, 157). Taken together, the available data suggest that the risk of BKV reactivation is associated with the net degree of immunosuppression rather than a specific effect of particular immunosuppressive agents on BKV replication. Several other risk factors for BKV reactivation have been reported (Fig. 5). Some of these (such as episodes of acute rejection) are likely to be surrogate markers for greater exposure to immunosuppression.

THERAPEUTIC INTERVENTIONS

There are few controlled studies available to guide the management of BK viremia and BKVAN in KTRs (158). The usual clinical approach upon identification of BK viremia or BKVAN is gradual reduction of immunosuppression, guided by serial monitoring of BK viremia by PCR (149, 159–162). Typical therapeutic strategies that have been reported are discussed in Table 5. A reduction in immunosuppression risks precipitating acute rejection, which can be challenging to distinguish histologically from resolving BKVAN, as there is often persistent interstitial inflammation, while viral inclusions and SV40 immunohistochemistry may become negative (137). As a result, a long-term consequence of modifying immunosuppression to treat BKV-associated disease may be a higher incidence of chronic rejection, as suggested by one study reporting an excess of de novo donor-specific anti-HLA antibody development in recipients with persistent BK viremia (140).

TABLE 5.

Therapeutic interventions for treatment of BKV infection/disease

Therapeutic intervention	Description	Reference	Intervention strategy	No. of patients studied (no. of patients with BKVAN)	Outcome and/or comments
Reduction in immunosuppression	Reduction of immunosuppression is first-line therapy for transplant recipients with significant BKV reactivation and may lead to resolution of BK viremia or BKVAN. Typically, the dose of antiproliferative agent (mycophenolate or azathioprine) is progressively decreased, with or without a reduction in calcineurin inhibitor (CNI; tacrolimus or cyclosporine) target levels.	238	30% to 50% reduction in MMFa and/or CNI	28 (5)	All patients cleared BKV. Early intervention was performed immediately after the detection of BK viremia and biopsy confirmation of BKVAN.
		162	Concomitant reduction of MMF (44%) and tacrolimus (41%)	24 (16)	23/24 patients have functioning graft. Reduction in immunosuppression was performed without any antiviral therapy soon after the detection of viremia.
		149	Reduction of CNI and/or MMF to 50% or withdrawal of MMF or CNI	910 (35)	19 patients retained graft function. Withdrawal of CNI was proved to increase the survival rate.
		247	30–50% reduction in tacrolimus and/or MMF	229 (7)	Successful resolution was seen in 30 BK viremia patients and 7 BKVAN patients.
Cidofovir	Cidofovir showed successful treatment of BKV infection in patients with BKVAN or HC.	248	13 doses of cidofovir with reduced immunosuppression	75	53/75 patients cleared BKV infection.
		163	4–10 doses of cidofovir in combination with reduction or removal of tacrolimus/MMF/cyclosporine	705 (21)	9/13 patients with no adjuvant cidofovir lost their grafts. Eight patients treated with cidofovir recovered, with no side effects.
Brincidofovir	Brincidofovir is a prodrug of cidofovir and also has in vitro activity against BKV (249), but it appears to be less nephrotoxic than cidofovir. Case reports have described the use of brincidofovir for BKVAN.	169	Brincidofovir for 10 weeks with no reduction in immunosuppressants	20 (3)	1 patient recovered completely. This was the first report of managing BKVAN with brincidofovir.
		168	36-week course with reduced immunosuppression	1 (1)	This study is a case report on one BKVAN patient who recovered with treatment.
Leflunomide	Leflunomide has immunosuppressant and antiviral properties. In several studies, leflunomide was introduced as a substitute for the antiproliferative agent in KTRs with BKVAN (171, 174, 176). These reports suggest potential benefits, though it is hard to differentiate direct antiviral benefits of leflunomide from overall changes in immunosuppression.	250	MMF reduced to 50%, with a 60-mg dose of leflunomide	28 (5)	71% of patients cleared BKV, with 39% showing side effects.
		174	MMF was replaced by 100 mg/day of leflunomide	12 (12)	42% of patients cleared viremia, with 17% showing side effects.
Intravenous immunoglobulin (IVIG)	Pooled human immunoglobulins contain neutralizing antibodies against BKV (178) and are immunomodulatory, potentially reducing the risk of acute allograft rejection in the setting of immunosuppression reduction.	184	Reduction in immunosuppression with intravenous administration of IVIG (1 g/kg of body weight) for patients with high viral loads	53 (20)	8 patients were treated with IVIG. After 1 year, 7 patients had functioning grafts.
		180	Reduction in immunosuppression with administration of IVIG (2 g/kg)	8 (8)	88% of patients had functioning grafts.
Fluoroquinolones	Ciprofloxacin and levofloxacin have been shown to inhibit BKV replication in vitro; however, the clinical effectiveness of fluoroquinolones appears to be low.	194	500 mg/day of levofloxacin for 3 mo	154 (76)	Levofloxacin did not significantly reduce BKV reactivation.
		193	500 mg/day of levofloxacin for 30 days, with reduction in overall immunosuppression	39	There was no effect on BKV reactivation.
		251	500 mg/day of ciprofloxacin	43	There was no effect on BKV reactivation.
Adoptive T cell therapy	Primed BKV-reactive T cells capable of controlling BKV-associated disease are adoptively transferred to patients.	207	0.5 × 107 to 2 × 107 cells/m2 were administered	11	Of the 7 patients with BKV reactivation, 5 achieved complete clearance, while 1 patient achieved partial clearance.
		252	2 × 107 cells/m2 were infused	13	4 patients had a complete response, while 9 patients showed a partial response.

^aMMF, mycophenolate mofetil.

Currently, there are no antiviral medications with strong evidence of clinical efficacy against BKV. Nevertheless, multiple reports have described the use of agents with potential anti-BKV activity in patients with BKVAN. In most cases, these agents have been combined with immunosuppression reduction and have been reported from uncontrolled retrospective observational studies, and therefore it is difficult to make firm conclusions about their therapeutic efficacy. Various single-center case reports and case series indicate a potential clinical benefit for cidofovir along with the reduction of immunosuppression, but no randomized trial has been reported so far (163–166). However, the clinical application of cidofovir is frequently limited by nephrotoxity. Brincidofovir (CMX001) is an orally administered prodrug of cidofovir currently undergoing phase III clinical trials and is reported to have a lower incidence of nephrotoxicity than that with cidofovir (167). Case reports have described successful outcomes for KTRs and HSCT patients with BKVAN after therapy with brincidofovir (168, 169); however, a clinical trial is needed to establish the efficacy and safety of this drug in treating BKV-associated disease.

Leflunomide, an immunosuppressant agent that also has antiviral properties against BKV in vitro (170), was utilized as a replacement agent in lieu of mycophenolate in several case series (171–176). These studies indicate that leflunomide is associated with a fall in BKV viral load, although it is unclear whether this reflects a reduction in overall immunosuppression or a direct antiviral effect. Leflunomide is associated with a number of significant adverse effects, including hepatitis, thrombotic microangiopathy, hemolysis, and bone marrow suppression (171). The active metabolite of leflunomide (known as teriflunomide or A771726) can be measured, and therapeutic monitoring of this has been proposed to aid in effective dosing of leflunomide, with the aim of minimizing toxicity (176). In a case series of 22 patients, achieving A771726 levels of 50 to 100 μg/ml was associated with reductions in BKV viral load during follow-up (175). Further prospective controlled studies of leflunomide are needed to confirm the efficacy and safety of this drug against BKV and associated disease (176, 177).

Intravenous immunoglobulin (IVIG) has been demonstrated to contain neutralizing antibodies against BKV (178). Several case reports and case series have described the use of IVIG as an adjunctive therapy for BKVAN (179–185); however, no controlled studies have been reported. IVIG therapy may be particularly beneficial in individuals with hypogammaglobulinemia, both with the aim of contributing passive anti-BKV immunity and also because IVIG is immunomodulatory (186) and may help to prevent allograft rejection in the context of reduced immunosuppression.

In vitro studies have demonstrated that fluoroquinolone antibiotics can inhibit BKV or SV40 polyomavirus replication in vitro, and fluoroquinolones have therefore been considered potential agents for controlling BKV replication (187–189). In vitro, the inhibitory effect of fluoroquinolones appears to be mediated via both reduction of large T antigen expression and inhibition of large T antigen helicase activity (187–189). Retrospective analysis of a trial in which a fluoroquinolone was administered as antibiotic prophylaxis at the time of kidney transplant found that fewer KTRs who received the fluoroquinolone developed BK viremia, further suggesting a possible benefit for these antibiotics in preventing BKV replication (190, 191). The combination of ciprofloxacin and leflunomide was also reported to be successful in controlling BKV replication in a single-center nonrandomized study (192). However, two subsequent prospective randomized controlled studies of levofloxacin in KTRs did not demonstrate a benefit with respect to reducing either the incidence or the level of BK viremia (193, 194). Overall, these data do not suggest that fluoroquinolones currently have a clinically significant role in the management of BKV-related disease (94).

BKV reactivation is associated with a reduction in allograft survival in KTRs. In a U.S. registry study of over 42,000 KTRs, 3-year allograft survival was 79% for recipients who required treatment for BKV, compared to 90% for BKV-negative patients (195). A systematic review of therapy for BKV-associated complications in KTRs was published in 2010 (158). The rate of death-censored allograft loss for immunosuppression reduction alone was 8/100 patient-years, and there was no evidence that addition of leflunomide or cidofovir improved allograft survival (158). The management of BKV-associated diseases varies from center to center, and there is a need for further randomized controlled trials to define the optimal treatment strategy for KTRs with BKV reactivation.

CELLULAR IMMUNOTHERAPY

Reactivation of latent BKV is exacerbated by immunosuppression, presumably following failure of BKV-reactive T cells to control viral replication (196). Therefore, adoptive transfer of primed BKV-reactive T cells may be an effective approach to controlling BKV-associated disease. The use of T cells as a cellular therapy to restore antiviral immunity in immunocompromised patients was first demonstrated by Riddell et al. in the early 1990s (197). Over the past 2 decades, various research groups have refined the process of adoptive T cell therapy for treating various chronic virus-associated diseases (198–202). For the successful generation of virus-specific T cells, it is necessary to identify the immunogenic antigens of the virus. The immunodominant epitopes of herpesviruses, such as CMV and Epstein-Barr virus (EBV), have been well defined, enabling researchers to successfully expand CMV- or EBV-specific T cells by using synthetic viral peptides or overlapping peptide pools (203). More recently, techniques have been developed to utilize major histocompatibility complex (MHC) multimer (199, 204, 205) or IFN-γ capture technology to allow rapid selection and enrichment of virus-specific T cells (201, 206). Immunotherapeutic approaches to treat BKV-associated diseases are still in their early stages. There are very limited data about the immunodominant BKV epitopes for T cell priming. Blyth et al. recently reported the use of overlapping peptide pools derived from all five BKV antigens to expand BKV-specific human T cells (89). Functional characterization of the expanded T cell population confirmed BKV reactivity, cytokine production, and in vitro cytotoxicity (89). A pilot study published in 2014 reported the in vitro expansion of virus-reactive T cells by use of overlapping peptide pools including antigens from EBV, CMV, adenovirus, BKV, and human herpesvirus 6 (207). An alternative approach to generating virus-specific T cell products, including BKV-reactive cells, is the use of an adenoviral vector to express multiple viral proteins (208). Cell products generated using this protocol were administered to 11 HSCT recipients, either prophylactically or in response to single or multiple viral infections. Several of the study participants had active BKV replication that improved following adoptive T cell transfer, with concurrent increases in the frequency of circulating BKV-reactive T cells detected using an enzyme-linked immunosorbent spot (ELISpot) assay. None of the participants experienced significant infusion reactions or other adverse safety concerns following adoptive transfer (207). These data suggest that adoptive transfer of BKV-reactive T cells has the potential to ameliorate BKV-associated pathology and may be a significant therapeutic advance in patients with BKVAN or HC. Further investigation into the comprehensive determination of immunogenic T cell epitopes for BKV antigens and the optimization of T cell expansion protocols should aid in the development of enhanced immunotherapeutic approaches to treat BKV-associated diseases.

HEMORRHAGIC CYSTITIS

Hemorrhagic cystitis (HC) is associated with high morbidity in HSCT patients. The association of BKV with HC was first detected in the 1980s, as evidenced by the presence of high loads of BKV in the urine samples of HSCT recipients (209, 210). The frequency of BKV-associated HC in HSCT recipients is about 10%, typically at approximately 2 weeks posttransplantation (211). BKV viruria is found in about 50% of bone marrow transplant (BMT) recipients, and high levels of viruria are associated with a higher risk of developing HC (212–214). The symptoms of HC include dysuria, urinary frequency, urinary urgency, suprapubic pain, and hematuria (211). A definitive diagnostic method for the detection of BKV-associated HC in transplant recipients has not been identified. The commonly used diagnostic approaches followed for BKV-associated diseases are listed in Table 4. While immunocompetent individuals also periodically shed BKV in urine, detection of urinary viral loads of 10⁶ to 10⁷ copies/ml and BK viremia of >10⁴ copies/ml is associated with a higher risk of HC in transplant recipients (214, 215). Therapeutic approaches to treat BKV-associated HC in HSCT and BMT recipients are not well established. Approaches similar to those for the management of BKVAN have been reported for patients with HC, including modification of immunosuppressive medications and the use of cidofovir, leflunomide, and fluoroquinolone antibiotics (164, 166, 175, 216, 217).

BKV AND CANCER

BKV has been linked to various cancers, including prostate cancer and urothelial tumors (218–220), although whether BKV has a causal role in the development of cancer is controversial. It has been suggested that BKV may be oncogenic due to the expression of the early coding region-encoded proteins LTA and STA, which can initiate or drive neoplastic transformation. T antigens are known to be prooncogenic due to their ability to inactivate tumor suppressor proteins, such as p53 and pRb, leading to increased cell proliferation. Therefore, binding of the polyomavirus LTA antigen to wild-type p53 in infected cells may lead to interference with the cell cycle and increase the risk of malignant transformation (221–223). In support of this hypothesis, BKV LTA-p53 protein complexes have been detected in the cytoplasm of prostate cancer tissue (224, 225). The presence of BKV LTA-p53 complexes is not sufficient to prove that LTA is oncogenic but raises suspicion that BKV may be a risk factor for prostate cancer development. In addition, STA has been shown to increase activation of the mitogen-activated protein (MAP) kinase pathway, which may also augment cell proliferation and transformation (226).

Various studies have shown that the BKV early coding regions can induce transformation in rodent and human cells. Trabanelli et al. showed that BKV T antigens caused cellular transformation via chromosomal alteration when human fibroblasts were transfected with BKV antigens (227). However, the mechanism through which BKV antigens mediate clastogenic effects has not yet been elucidated fully. Further evidence suggesting that BKV is oncogenic comes from animal studies in which inoculation of BKV resulted in the development of tumors in rodents (27, 228). Recently, Kenan et al. showed that oncogenesis requires integration of the BKV genome into the chromosome and demonstrated a possible mechanism of BKV DNA integration and lytic infection (229). This is the only study showing integration of BKV DNA in tumor cells and needs to be confirmed.

There have been reports of a higher prevalence of BKV DNA in precancerous and early-stage cancer tissues than in healthy control tissues, possibly suggesting a “hit and run” mechanism of carcinogenesis by BKV or that BKV infection can promote the early stages of tumorigenesis (224, 225). There are also contradictory reports about the presence of BKV DNA in prostate cancer specimens (230, 231). Novel approaches are required in order to prove the role of BKV in cancer development. Whether BKV plays an important role in human cancer development remains a topic of debate.

CONCLUSIONS

BKV infection is a common childhood infection that establishes permanent latency within renal tubular and uroepithelial cells. BKV reactivation is common in immunocompromised individuals and causes significant morbidity, in particular BKVAN in KTRs and hemorrhagic cystitis in hematopoietic stem cell recipients. No specific antiviral therapy is available, and management typically consists of reducing immunosuppression with the aim of reconstituting effective BKV immune responses. BKVAN in kidney transplant recipients is associated with poorer allograft survival, due either to progressive BKV-associated damage or to rejection precipitated by a reduction in immunosuppression. Studies of the immune response to BKV are limited but indicate that T cells play a vital role in controlling BKV replication. There is a need for high-quality controlled clinical studies to define the optimal treatment strategies following BKV reactivation. Adoptive cell therapy using ex vivo-expanded BKV-reactive T cells is a novel therapeutic approach that is currently in early clinical development.

Biographies

George R. Ambalathingal is a Ph.D. graduate student at the QIMR Berghofer Medical Research Institute/University of Queensland, Brisbane, Australia. He completed his bachelor's degree in technology at Anna University, India, followed by a master's degree in engineering in biotechnology from Satyabhama University, India. Before starting his Ph.D. program, he worked as a Scientist in the Discovery Biology Division of Anthem Biosciences Ltd. His current research focuses on characterizing the functions of BKV-specific T cells in comparison to other virus-specific T cells.

Ross S. Francis is a Consultant Nephrologist and Transplant Physician at Princess Alexandra Hospital in Brisbane, Australia. His clinical and research interests range from translation of novel treatments for immunological kidney disease and transplantation, including biological and cellular therapies, to improving transition care for young adults with kidney disease. He is a council member for the Australian and New Zealand Society of Nephrology (ANZSN) and a member of the Australasian Kidney Trials Network Transplantation Working Group and represents ANZSN on the Renal Transplant Advisory Committee of the Transplantation Society of Australia and New Zealand.

Mark J. Smyth is a Senior Scientist and Immunology Coordinator at the QIMR Berghofer Medical Research Institute. He completed his Ph.D. studies in 1988 and trained at the National Cancer Institute (NCI) (1988 to 1992) before commencing his independent research career in Australia. Over the last 15 years, he has rekindled worldwide interest in cancer immune surveillance and defined immune-mediated dormancy of cancer. Historically, his work on natural killer (NK) cells and immunoregulation has been extended to host antiviral responses, and more recently, he provided a new means of classifying NK cell subtypes.

Corey Smith completed his Ph.D. studies in 2004 at the University of Melbourne and then took up a research position at the QIMR Berghofer Medical Research Institute. His work focuses on the development of immunotherapeutic approaches to treat disease and cancer associated with persistent viral infection. This has led to the commencement of a number of adoptive immunotherapy studies. His work has also focused on understanding the mechanisms that influence the efficient induction of T cell responses to persistent human viral infections and the role that immune evasion strategies play in virus-associated cancer and disease.

Rajiv Khanna is a Senior Scientist and Senior Principal Research Fellow at the QIMR Berghofer Medical Research Institute. He has extensive expertise in immunotherapy clinical trials, cancer immunology, and vaccine development. Over the last 2 decades, his group has successfully translated his research towards the development of novel T cell-based immunotherapeutic strategies for the treatment of cancer patients and transplant recipients. Professor Khanna has been invited by the International Transplant Society to participate in the development of guidelines for the clinical management of cytomegalovirus infection in solid organ transplant patients.