Biol Blood Marrow Transplant 20 (2014) 564e570

BK Virus Disease after Allogeneic Stem Cell

Transplantation: A Cohort Analysis ASBMT
American Society for Blood
and Marrow Transplantation
Nienke M.G. Rorije 1, 2, Margaret M. Shea 1,
Gowri Satyanarayana 1, Sarah P. Hammond 1, 3,
Vincent T. Ho 3, Lindsey R. Baden 1, 3, Joseph H. Antin 3,
Robert J. Soiffer 3, Francisco M. Marty 1, 3, *
1
Division of Infectious Diseases, Brigham and Women’s Hospital, Boston, Massachusetts
2
University of Groningen Medical School, Groningen, The Netherlands
3
Division of Hematologic Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts

Article history:
Received 18 November 2013
Accepted 15 January 2014
Key Words:
BK virus
Allogeneic stem cell
transplantation
Cord blood transplantation
Graft-versus-host disease
Human BK polyomavirus
a b s t r a c t
The clinical epidemiology of BK virus (BKV) disease after allogeneic hematopoietic stem cell transplantation
(HSCT) is not well defined. We evaluated 491 patients transplanted from January 2010 to December 2011 at a
single transplant center to assess incidence, severity, and risk factors for BKV disease after HSCT. BKV disease
was defined as BKV detection in urine by PCR testing in association with genitourinary symptoms without
other concurrent genitourinary conditions. BKV disease occurred in 78 patients (15.9%), for an incidence rate
of .47/1000 patient-days (95% confidence interval [CI], .37 to .59); BKV disease was considered severe in 27
patients (5.5%). In multivariate Cox modeling, time-dependent acute graft-versus-host disease (aGVHD)
grades II to IV (adjusted hazard ratio [aHR] 4.25; 95% CI, 2.51 to 7.21), cord blood HSCT (aHR 2.28; 95% CI, 1.01
to 5.15), post-transplant mycophenolate use (aHR 3.31; 95% CI, 1.83 to 5.99), and high-dose cyclophosphamide
conditioning (aHR 2.34, 95% CI 1.45 to 3.77) were significant predictors of BKV disease. Time-dependent
aGVHD grades III to IV (aHR 10.5; 95% CI, 4.44 to 25.0) and cord blood HSCT (aHR 5.40; 95% CI, 1.94 to
15.0) were independent risk factors for severe BKV disease. BKV disease is common and is associated with
significant and prolonged morbidity after HSCT. Prospective studies are needed to better define the morbidity
of post-HSCT BKV disease and inform the design of prophylaxis and treatment trials.
Ó 2014 American Society for Blood and Marrow Transplantation.

INTRODUCTION associated with BKV reactivation ranging from cystitis

Immunoincompetence associated with allogeneic he-
matopoietic stem cell transplantation (HSCT) causes
heightened susceptibility to infection [1,2]. BK virus (BKV) is
a human polyomavirus [3] that is typically acquired in early
childhood and becomes latent in urothelial cells of the kid-
ney and urinary tract [4]. BKV reactivation after allogeneic
HSCT is associated with manifestations ranging from
asymptomatic viruria to severe hemorrhagic cystitis (HC) [5].
HC occurs in 5% to 68% of HSCT recipients [6,7] and may
result in hospitalization, renal dysfunction [8,9], and
increased mortality [10,11]. BKV loads can be measured in
urine and serum by quantification of viral DNA with PCR
testing [12]. BKV viruria is associated with HC [6,13-17] but
also occurs in asymptomatic patients [7,13]. Various studies
have defined risk factors for development of HC [17-31], BKV
viruria [14], and BKV viremia [14,16] after HSCT, but these
studies did not distinguish asymptomatic BKV viruria or
viremia from clinical disease other than HC.
Some clinicians equate BKV PCR detection with disease
when in practice HSCT patients may develop asymptomatic
BKV viruria without clinical consequences. Furthermore,
previous case-control studies have focused on HC, but in
clinical practice a broader spectrum of urinary tract disease is
Presented in abstract form (abstract 1927) at the 54th annual meeting of
the American Society of Hematology, Atlanta, Georgia, December 8, 2012.
Financial disclosure: See Acknowledgments on page 569.
* Correspondence and reprint requests: Francisco M. Marty, MD, Division
of Infectious Diseases, Brigham & Women’s Hospital, 75 Francis Street, PBB-
A4, Boston, MA 02115.
E-mail address: fmarty@partners.org (F.M. Marty).
1083-8791/$ e see front matter Ó 2014 American Society for Blood and Marrow Transplantation.
http://dx.doi.org/10.1016/j.bbmt.2014.01.014
without hematuria to renal failure. Better understanding of
the spectrum and natural history of clinical BKV disease is
critical for the design of clinical trials and drug development
in this area. We assess and quantify the incidence, severity,
risk factors, and outcome of clinical BKV disease after allo-
geneic HSCT in a contemporary cohort.
METHODS
Patients
All patients who underwent an initial allogeneic HSCT between January
1, 2010 and December 31, 2011 at Dana-Farber Cancer Institute/Brigham and
Women’s hospital (DFCI/BWH) were identified through the data repository
of the DFCI/BWH HSCT program. The Office for Human Research Studies of
Dana-Farber/Harvard Cancer Center approved this retrospective study.
Time-at-risk was censored at time of death, second HSCT, or on July 1, 2012.
Covariates and Definitions
Data on covariates and outcomes of interest were identified through the
data repository of the DFCI/BWH HSCT program, queries of the Partners
HealthCare Research Patient Data Registry, and review of electronic medical
records. Collected data included age at time of transplantation, gender, race,
primary disease that required HSCT, transplant conditioning regimen, donor
relatedness, HLA donor matching, stem cell source, development and
maximal grade of acute graft-versus-host-disease (aGVHD), aGVHD pro-
phylaxis regimens, recipient cytomegalovirus seropositivity before trans-
plant, and days to engraftment.
Transplantation conditioning regimens were classified as either mye-
loablative or reduced-intensity conditioning based on conditioning agents
and dose used. Myeloablative conditioning consisted of different combina-
tions of chemotherapeutic agents with or without total body irradiation
(TBI). Most patients received cyclophosphamide (1800 mg/m2
2 days) and
TBI with a total dose of 1400 cGy delivered in 7 fractions. Other chemo-
therapeutic agents used for myeloablative conditioning were high-dose i.v.
busulfan and fludarabine [32]. Reduced-intensity conditioning mainly con-
sisted of fludarabine with low-dose i.v. busulfan. Combinations of fludar-
abine and melphalan or fludarabine and cyclophosphamide were also used
 N.M.G. Rorije et al. / Biol Blood Marrow Transplant 20 (2014) 564e570 565

for reduced-intensity conditioning, sometimes in combination with a single
dose of TBI of 200 cGy [32].
Additional rabbit or horse antithymocyte globulin (ATG) was primarily
used in patients who received cord blood HSCT or in patients with aplastic
anemia as primary disease and in some unrelated donor transplantation
patients with acute myelogenous leukemia participating in a study protocol
[33,34]. Hyperhydration was used for prevention of cyclophosphamide-
associated HC in patients receiving high-dose cyclophosphamide [35].
Patients received 1 of several GVHD prophylactic regimens, including
glucocorticoids, methotrexate, mycophenolate mofetil (MMF), sirolimus,
and tacrolimus. GVHD prophylaxis mainly consisted of a combination of
tacrolimus with methotrexate and/or sirolimus in marrow or peripheral
HSCT and tacrolimus/sirolimus in cord blood HSCT. Methotrexate was
administered on days 1, 3, 6, and 11 after transplantation [36,37]. MMF was
used to substitute for tacrolimus or methotrexate at the discretion of the
physician if severe mucositis, hepatic or renal dysfunction, or thrombotic
microangiopathy occurred early post-transplantation. GVHD prophylactic
medications were generally tapered starting between day þ60 and
day þ100 and were discontinued by 6 to 9 months after HSCT in the absence
of GVHD.
All GVHD prophylactic regimens were included individually in the
analysis and modeled as baseline risks considering the exposure pattern
outlined earlier [38]. aGVHD was graded according to the consensus system
[39]. Conditioning regimens and aGVHD prophylaxis were administered
either in the setting of a single-arm or randomized study protocols or at the
discretion of the treating transplant physician [38,40]. Engraftment day was
defined as the first of 3 consecutive days of an absolute neutrophil count >
500 cells/mL.
BKV Testing
BKV testing was performed at the discretion of the treating clinicians
usually for dysuria or hematuria. In some cases conditions such as fever and
change of mental status led to BKV testing. Inpatient BKV DNA PCR testing
was performed at the BWH clinical microbiology laboratory using a real-
time PCR assay (Roche Diagnostics, Indianapolis, IN) with a quantifiable
range of 500 to 5
108 copies/mL. Outpatient DFCI BKV DNA testing of
plasma and urine was performed at Viracor-IBT laboratories (Lee’s Summit,
MO) using a real-time quantitative PCR assay: urine range of 500 to 1
1010
copies/mL and plasma range of 100 to 1
1010 copies/mL. For certain ana-
lyses in which BKV load was used as a continuous covariate, detectable BK
PCR virus loads below the lower limit of quantification were assigned values
of half the lower limit of quantification; BKV loads above the upper limit of
quantification were assigned values of twice the upper limit of quantifica-
tion to acquire standardized numbers for analysis.

Table 1
Characteristics of Allogeneic HSCT Cohort, Patients with and without BKV Disease

Characteristics Total (N ¼ 491) No BKV Disease (n ¼ 413) BKV Disease (n ¼ 78) P

Male 282 (57.4) 232 (56.2) 50 (64.1) .21
Median age at transplant, yr (range) 54 (19-73) 55 (19-73) 52 (19-70) .02
Survival as of July 1, 2012 326 (66.4) 277 (67.1) 49 (62.8) .51
White race 454 (92.5) 383 (92.7) 71 (91.0) .64
Primary disease
Acute myeloid leukemia 181 (36.9) 165 (40.0) 16 (20.5) .0009
Non-Hodgkin lymphoma 90 (18.3) 67 (16.2) 23 (29.5) .01
Myelodysplastic syndrome 54 (11.0) 43 (10.4) 11 (14.1) .33
CLL, SLL, PLL 48 (9.8) 41 (9.9) 7 (9.0) 1.00
Acute lymphoblastic leukemia 38 (7.7) 30 (7.3) 8 (10.3) .36
Myeloproliferative disorder 19 (3.9) 16 (3.9) 3 (3.9) 1.00
Chronic myeloid leukemia 17 (3.5) 15 (3.6) 2 (2.5) 1.00
Multiple myeloma/plasma cell dyscrasia 15 (3.1) 15 (3.6) 0 (.0) .14
Hodgkin disease 12 (2.4) 9 (2.2) 2 (2.5) .69
Anemia/red cell disorder 11 (2.2) 7 (1.7) 5 (6.4) .03
Other 6 (1.2) 5 (1.2) 1 (1.3) 1.00
HLA match and donor relatedness
Matched related 153 (31.2) 132 (31.9) 21 (26.9) .43
Matched unrelated 267 (54.4) 227 (55.0) 40 (51.3) .62
Mismatched 71 (14.4) 54 (13.1) 17 (21.8) .05
Stem cell source .004
Peripheral blood 428 (87.2) 369 (89.3) 59 (75.6) d
Umbilical cord blood 34 (6.9) 23 (5.6) 11 (14.1) d
Bone marrow 29 (6.9) 21 (5.1) 8 (10.3)
Conditioning regimen .005
Reduced intensity 312 (63.5) 274 (66.3) 38 (48.7) d
Myeloablative 179 (36.5) 139 (33.7) 40 (51.3) d
Conditioning agents
Fludarabine 336 (68.4) 293 (71.0) 43 (55.1) .008
i.v. busulfan 282 (57.4) 252 (61.0) 30 (38.5) .0003
Cyclophosphamide 173 (35.2) 130 (31.5) 43 (55.1) <.0001
TBI 156 (31.8) 119 (28.8) 37 (47.4) .002
ATG 51 (10.4) 38 (9.2) 13 (16.7) .07
Melphalan 40 (8.2) 32 (7.8) 8 (10.3) .50
CMV-seropositive recipient 299 (60.9) 250 (60.5) 49 (62.8) .80
aGVHD <.0001
Grades 0-I 338 (68.8) 301 (72.9) 37 (47.4) d
Grades II-IV 153 (31.2) 112 (27.1) 41 (52.6) d
aGVHD prophylactic regimen
Tacrolimus 483 (98.4) 405 (98.1) 78 (100.0) .37
Methotrexate 340 (69.3) 292 (70.7) 48 (61.5) .11
Sirolimus 298 (60.7) 258 (62.5) 40 (51.3) .08
MMF 37 (7.5) 23 (5.6) 14 (18.0) .0006
Corticosteroids 4 (.8) 4 (1.0) 0 (.0) 1.00
Neutrophil counts
Neutropenia 405 (82.5) 335 (81.1) 70 (89.7) .07
Did not nadir 86 (17.5) 78 (18.9) 8 (10.2)
Days to engraftment, median (range, IQR) 13 (0-114, 11-15) 13 (0-114, 10-15) 13.5 (0-49, 12-16) .01

CLL indicates chronic lymphocytic leukemia; SLL, small lymphocytic leukemia; PLL, prolymphocytic leukemia.
Values are number of cases, with percents in parentheses, unless otherwise specified.
Dash indicates that values were not calculated.
566 N.M.G. Rorije et al. / Biol Blood Marrow Transplant 20 (2014) 564e570
Case Definition
BKV disease was defined as detection of BKV by PCR testing in associ-
ation with genitourinary symptoms along with well-documented clinical
course in absence of concurrent genitourinary conditions that could influ-
ence symptom presentation. Time to BKV disease was defined as time be-
tween date of HSCT and date of first detected BKV PCR. Time to first
symptoms was captured as time between date of HSCT and date of first
reported genitourinary symptoms. End of symptoms was captured as
documented in the medical record. Time between onset of symptoms and
end of symptoms during BKV PCR positivity was considered as a BKV disease
episode.
For patients who underwent BKV testing, we recorded reasons for
testing, clinical course, imaging results, pathology results, treatments,
concomitant conditions, and other urinary tract infections. Symptoms were
categorized into 7 groups: bladder spasms, dysuria, flank pain, frequency,
hematuria, urgency, and other. Other reasons for testing that did not fit into
1 of the 7 groups included fever and mental status changes. Hematuria was
graded on a scale of 0 to 4, according to established criteria [6,7]. Imaging
results were reviewed for signs compatible with BKV disease and catego-
rized into 5 groups: bladder wall thickening, urinary clots and/or hemor-
rhage, hydronephrosis, perivesical stranding, and ureteral thickening [41].
Autopsy, urine cytology, and bladder biopsy results were also reviewed for
evidence of BKV disease when available.
Treatments for BKV disease were captured during BKV disease episodes,
including fluoroquinolone use even if BKV disease was not the main indi-
cation, because of putative prophylactic efficacy [42,43]. Foley catheter
placement, bladder irrigation, intravesicular use of cidofovir or alum,
cystoscopy, and percutaneous nephrostomy placement were considered
urologic interventions. Antispasmodics, analgesics, renal replacement
therapy including hemodialysis, continuous venovenous hemofiltration,
and use of finasteride and tamsulosin were captured.
BKV disease was defined as severe when patients had at least 1 of the
following: grade 3 or 4 hematuria, imaging with findings compatible with
genitourinary tract inflammation or intraluminal clot formation with or
without associated hydronephrosis, need for invasive genitourinary in-
terventions, or hospitalization for BKV disease management. Patients were
considered to have a second episode of BKV disease if symptoms recurred
and BKV was documented more than 30 days after symptom resolution.
Figure 1. Flow chart: Overview of patients. Numerals indicate the number of patients. GU indicates genitourinary; w/o, without; UTI, urinary tract infection; AKI,
acute kidney injury. Other diagnosis includes bladder spasms, fever, neurological disease/symptoms.
Statistical Analysis
Baseline characteristics were compared using 2-sided Fisher’s exact test,
Wilcoxon rank-sum test, or chi-square test where appropriate. Kaplan-
Meier curves were generated for survival and time-to-event analyses.
Characteristics associated with BKV disease or severe BKV disease were
evaluated using Cox modeling. aGVHD and neutropenia were modeled as
time-dependent covariates. Candidate covariates were included in the
multivariate models if they were associated with BKV disease or severe BKV
disease in the univariate analysis or to explore potential confounders of
characteristics associated with BKV disease. SAS version 9.2 (SAS Institute
Inc., Cary, NC) was used for these analyses. Incidence rates of BKV disease
and severe BKV disease and their 95% confidence intervals (CI) were
calculated by Fisher’s method using OpenEpi version 2.3.1 (http://www.o-
penepi.com, Atlanta, GA). Descriptive statistical analyses of clinical features
of BKV disease were performed by JMP Pro 10.0 2012 (SAS Institute Inc.,
Cary, NC). P < .05 were considered statistically significant.
RESULTS
During the study period, 491 patients underwent a first
allogeneic HSCT at DFCI/BWH, and 168 (34.2%) patients were
tested for BKV. Median time to testing was 54.5 days (range,
0 to 559; interquartile range [IQR], 14.5 to 137.5 days).
Seventy-eight patients (15.9%; 46.4% of patients tested) met
criteria for clinical BKV disease for an overall incidence rate
of .47/1000 patient-days (95% CI, .37 to .59). Ten additional
patients had documented BKV viruria but were not consid-
ered to have BKV disease due to concomitant conditions that
could explain genitourinary symptoms (bacterial urinary
tract infections in 2, adenovirus disease in 1) or due to no
well-documented symptoms (7 patients). Median time to
diagnosis was 58.5 days (range, 2 to 663; IQR, 19 to 111 days).
Baseline characteristics of the HSCT cohort and patients with
and without BKV disease are presented in Table 1.
 N.M.G. Rorije et al. / Biol Blood Marrow Transplant 20 (2014) 564e570 567

Clinical Features bladder of 5 patients (6.4%). Viral cytopathic changes

Among the 78 patients with BKV disease, the most com-
mon documented symptoms were dysuria (56, 71.8%), he-
maturia (40, 51.3%), and urinary frequency (39, 50.0%).
Among 66 patients with documented symptom onset and
resolution, median symptom duration during the BKV dis-
ease episode was 33.5 days (range, 2 to 385; IQR, 11 to
70 days). Forty-six patients with BKV disease (10.4% of
cohort) met criteria for diagnosis of HC (ie, clinical symptoms
of cystitis with grade 1 hematuria or higher). Hematuria with
clot formation was seen in 16 of 78 patients (20.5%). Man-
agement of BKV associated symptoms was the reason for
hospitalization in 4 patients (5.1%).
Imaging was performed in 39 patients (50.0%), and results
compatible with BKV disease were seen in 14 patients
(17.9%), including bladder wall (12, 15.4%) or ureteral thick-
ening (5, 6.4%), clot formation/hemorrhage (5, 6.4%), peri-
vesicular stranding (5, 6.4%), and hydronephrosis (3, 3.8%). In
5 patients (6.4%), pathological evaluation of urine cytology or
bladder biopsy was done. In 2 patients (2.6%), evidence of
human polyomavirus was seen in urine cytology or bladder
biopsy. Invasive urologic interventions were performed in 14
patients (18.0%), including Foley catheter placement in 9,
bladder irrigation in 7, intravesicular administration of
cidofovir in 3, alum instillation in 1, cystoscopy in 2, and
bilateral percutaneous nephrostomy tube placement in 1.
Medical treatments prescribed during BKV disease epi-
sodes included quinolones (35, 44.9%), antispasmodics (29,
37.2%), i.v. immunoglobulin (19, 24.4%), pain medications (17,
21.8%), and drugs for lower urinary tract symptoms (3, 3.9%).
Intravenous cidofovir was used in 5 patients (6.4%) and
leflunomide in 8 (10.3%). Four patients (5.1%) required renal
replacement therapy. Transfusion of platelets was needed in
10 patients (12.8%), and 4 patients (5.1%) received packed
RBC transfusion for hematuria.
During the study period, 29 of 78 patients with BKV dis-
ease died (37.2%). Autopsies were performed in 12 patients
and showed hemorrhagic and inflammatory changes in the
consistent with BKV infection were documented in micro-
scopic examination of the bladder of 1 patient (1.3%). No
cases of BKV nephropathy were identified on autopsy. One
patient died during hospitalization for BKV disease man-
agement due to hemorrhagic complications associated with
BKV HC. Seventeen patients (21.8%) experienced recurrent
BKV disease during the study period but were not analyzed
further. Among 80 patients without BKV viruria, 8 patients
had a bacterial urinary tract infection, 17 had acute kidney
injury, and 21 had hematuria. Figure 1 summarizes the
clinical features.
Risk Factors
Univariate and multivariate hazard ratios (HRs) for each
covariate associated with BKV disease in the initial analysis
(P < .2) were calculated as shown in Table 2. Kaplan-Meier
curves of BKV disease stratified by aGVHD grade indicated
that aGVHD grades II to IV were associated with develop-
ment of BKV disease. BKV disease (HR 1.73; 95% CI, 1.17 to
2.60, P ¼ .007) and time-dependent aGVHD grades III to IV
(HR 2.78; 95% CI, 1.93 to 3.99, P < .0001) were associated with
increased risk of death. When adjusted to aGVHD grades III to
IV, the HR of BKV disease on survival declined and was no
longer significant (HR 1.45; 95% CI, .96 to 2.19, P ¼ .08).
The increased HR of myeloablative conditioning was
mainly due to cyclophosphamide use, observed in multi-
variable modeling of these 2 covariates. A similar pattern was
seen with TBI use; when adjusted to cyclophosphamide use
in a multivariate model, the increased HR of TBI was no
longer significantly increased. Myeloablative conditioning
regimens were more frequently used in younger patients;
thus, age at HSCT as a risk for BKV disease was confounded by
indication. Neutropenia as a BKV disease risk factor was
confounded by cord blood HSCT; these patients had signifi-
cantly (P < .0001) longer times to engraftment (median,
18.5 days; range, 14 to 49; IQR, 16 to 27 days) when compared

Table 2
Proportional Hazards Modeling of Risk of BKV Disease (n ¼ 78) after Allogeneic HSCT

Characteristics Univariate HR (95% CI) P Multivariate aHR (95% CI) P

Age at transplant .98 (.97-1.00) .03 d d
Primary disease
Acute myeloid leukemia .42 (.24-.73) .002 d d
Non-Hodgkin lymphoma 1.97 (1.21-3.20) .007 d d
Aplastic anemia/red cell disorder 4.35 (1.76-10.8) .002 d d
Stem cell source
Peripheral blood .36 (.21-.60) .0001 d d
Cord blood 2.86 (1.51-5.42) .001 2.28 (1.01-5.15) .047
HLA match and donor relatedness
Mismatched donor 1.91 (1.11-3.27) .02 d d
Conditioning regimen
Myeloablative conditioning 2.01 (1.29-3.13) .002 d d
Conditioning agents
Fludarabine .51 (.33-.80) .003 d d
i.v. busulfan .39 (.25-.61) <.0001 d d
Cyclophosphamide 2.64 (1.69-4.13) <.0001 2.34 (1.45-3.77) .0005
TBI 2.20 (1.41-3.43) .0005 d d
ATG 1.94 (1.07-3.51) .03 2.14 (.99-4.60) .05
aGVHD prophylactic agents
Methotrexate .64 (.41-1.01) .06 d d
Sirolimus .64 (.41-1.00) .05 d d
MMF 3.61 (2.02-6.44) <.0001 3.31 (1.83-5.99) <.0001
aGVHD
Grades II-IV 4.68 (2.81-7.79) <.0001 4.25 (2.51-7.21) <.0001
Neutropenia 2.18 (.77-6.19) .14 d d

aHR indicates adjusted HR, adjusted for the 5 variables included in the multivariate Cox model.
568 N.M.G. Rorije et al. / Biol Blood Marrow Transplant 20 (2014) 564e570
with adult donor HSCT (median, 13 days; range, 0 to 114; IQR,
10 to 14 days).
The final multivariable model that included cyclophos-
phamide, ATG, cord blood HSCT, MMF aGVHD prophylaxis,
and aGVHD grades II to IV demonstrated that MMF use for
aGVHD prophylaxis (P < .0001), aGVHD grades II to IV
(P < .0001), cyclophosphamide use (P ¼ .03), and cord blood
HSCT (P ¼ .047) remained stable and significant risk factors
for BKV disease. ATG use (P ¼ .15) was not significant in the
multivariate model.
BKV Virus Loads
Among patients with BKV disease, median initial urine
BKV load was 5.1
108 copies/mL (range 600, >1.00
1010;
IQR, 6.99
105, >1.00
1010 copies/mL). Median peak urine
BKV load during BKV disease episode was 1.5
109 copies/
mL (range 600, >1.00
1010; IQR, 1.50
106, >1.00
1010
copies/mL). Blood BKV loads were obtained in 30 patients
with median initial blood BKV load of 850 copies/mL (range
0, 1.52
106; IQR 0, 4,620 copies/mL) and median peak BKV
load of 1705 copies/mL (range 0, 1.52
106; IQR 0, 1.49
103
copies/mL).
Severe BKV Disease
Twenty-seven patients met criteria for severe BKV disease
(34.6%, 5.5% of the cohort): 16 patients had hematuria with
clot formation, 14 patients demonstrated imaging results
compatible with BKV disease, 4 patients were hospitalized
for management of BKV diseaseeassociated symptoms, and
14 patients required invasive urologic interventions. The
incidence rate of severe BKV disease was .15/1000 patient-
days (95% CI, .10 to .22). Detection of BKV viremia was 75%
sensitive and 36% specific for severe BKV disease. Kaplan-
Meier curves for severe BKV disease by aGVHD grade indi-
cated that aGVHD grades III to IV but not grade II aGVHD
were associated with development of severe BKV disease.
Significant risk factors for severe BKV disease are shown
in Table 3. Increased HR of neutropenia was mainly driven by
cord blood HSCT. In multivariable modeling, time-dependent
aGVHD grades III to IV (adjusted HR 10.5; 95% CI, 4.44 to 25.0)
and cord blood HSCT (adjusted HR 5.40; 95% CI, 1.94 to 15.0)
remained independent risk factors for severe BKV disease.
Sensitivity Analysis
We analyzed the cohort based on a less-specific BKV
disease definition that included all 88 patients with BKV
viruria, including 10 patients with undocumented symp-
toms or concomitant genitourinary diseases. Incidence
Table 3
Proportional Hazards Modeling of Risk of Severe BKV Disease (n ¼ 27) after Allogeneic HSCT
Characteristics Univariate HR (95% CI)
Age at transplant .97 (.95-1.00)
HLA match and donor relatedness
Mismatched donor 3.39 (1.52-7.57)
Stem cell source
Cord blood HSCT 3.78 (1.43-9.99)
Conditioning regimen
Myeloablative conditioning 2.29 (1.07-4.88)
Conditioning agents
Cyclophosphamide 2.48 (1.16-5.29)
TBI 2.11 (.99-4.48)
aGVHD
Grades III-IV 11.9 (5.13-27.5)
Neutropenia 8.25 (1.46-46.6)
aHR indicates adjusted HR, adjusted for the 3 variables included in the multivariate Cox model.
rates and characteristics of BKV viruria were similar to
those of BKV disease. The final multivariable model of risk
factors for BKV viruria was similar to the final model of BKV
disease. However, in this group only 3 risk factors remained
significant: MMF aGVHD prophylaxis (P < .0001), aGVHD
grades II to IV (P < .0001), and cyclophosphamide use
(P ¼ .03).
DISCUSSION
This study focused not only on occurrence and risk factors
of HC but determined also the broader clinical spectrum of
clinical BKV disease and associated symptoms in a 2-year
contemporary cohort of patients who underwent allogeneic
HSCT. By expanding the view beyond HC and using a cohort
analysis design, this study provides a more comprehensive
insight into the burden of BKV disease. We also propose a
new definition of severe BKV disease that includes additional
aspects of severity besides the degree of hematuria. A
strength of this study is that it included a cohort of 491 adults
who underwent different modalities of allogeneic HSCT from
2010 to 2011; this is the largest adult study population to
study this problem in the last decade.
The overall cumulative incidence of BKV disease indicates
that it is a common complication after allogeneic HSCT, more
frequent than other infectious disease complications such as
invasive fungal disease [2,44]. BKV disease was associated
with increased risk of death (HR 1.74; 95% CI, 1.18 to 2.57,
P ¼ .005) in this cohort, although this risk was not significant
when adjusted to grades III to IV aGVHD. Besides significant
hematuria, patients with severe BKV disease had important
radiological findings, hospitalization for BKV disease man-
agement, and need for invasive urologic interventions. We
found that BKV loads in urine and blood were not accurate
predictors of BKV disease severity.
We also found that aGVHD grades II to IV, cord blood
HSCT, MMF, and cyclophosphamide use were significant risk
factors for BKV disease on multivariable analysis. ATG use
was not significant in this model but remained associated
with a higher risk for development of BKV disease. For severe
BKV disease, time-dependent aGVHD grades III to IV and
cord blood HSCT were significant and independent risk fac-
tors in multivariate modeling. These results suggest that
occurrence of BKV disease is multifactorial and is associated
with both immunological and cytotoxic events that take
place during or after HSCT.
Leung et al. [45] proposed a 3-phase etiological model of
HC after allogeneic HSCT that may be relevant for patients
who received myeloablative conditioning but not for patients
P Multivariate aHR (95% CI) P
.03 d d
.003 d d
.007 4.47 (1.65-12.1) .003
.03 d d
.02 2.03 (.90-4.58) .09
.05 d d
<.0001 10.5 (4.44-25.0) <.0001
.02 d d
 N.M.G. Rorije et al. / Biol Blood Marrow Transplant 20 (2014) 564e570
who received reduced-intensity conditioning regimens. They
proposed that in a first phase the uroepithelium is damaged
due to conditioning regimens and subsequent uroepithelial
regeneration provides an appropriate milieu for BKV repli-
cation. It is well known that cyclophosphamide causes
uroepithelial damage by excretion of the metabolite acrolein
[19]. Although effective prophylaxis with forced diuresis
helps prevent cyclophosphamide-induced HC [35], it is
possible that cyclophosphamide still damages the uroepi-
thelium, making these tissues more vulnerable to BKV
reactivation. Cord blood transplantation is also generally
associated with increased risk of infection due to slow he-
matological and immunological reconstitution [1]. One study
proposed umbilical cord blood as a risk factor for develop-
ment of HC, but umbilical cord blood transplants were
grouped with haploidentical transplants for analysis, making
a comparison with our results difficult [29]. It is noteworthy
that patients who underwent cord blood HSCT may develop
clinical BKV disease before engraftment. Cord blood lacks
passively transferred immunity to BKV, suggesting that
donor immunity may be relevant to minimize BKV disease.
These observations do not support the proposed model of
Leung et al. [45], which suggests that donor lymphocytes
cause mucosal damage and thus contribute to development
of BKV-associated HC.
The presence and treatment of aGVHD are generally
considered risk factors for infection, and a few studies also
suggest an association of aGVHD with HC [20,21,26,30,31].
Our results confirm these findings and also emphasize a
strong association between aGVHD and development of BKV
disease. In this setting, it seems that reactivation of BKV in
the urothelium is facilitated by GVHD and its treatment,
rather than BKV disease being a manifestation of bladder
GVHD [45]. As suggested by Leung et al. [45], clinical BKV
disease could then develop in the setting of tapering
immunosuppression as an immune reconstitution phenom-
enon. Use of MMF as aGVHD prophylaxis has not been pre-
viously reported as a risk factor for HC. El-Zimaity et al. [24]
proposed a possible interaction between degree of pharma-
cological immunosuppression, donor type, and development
of HC. However, in our cohort patients did not receive MMF
as part of a standard aGVHD prophylaxis regimen, but MMF
was used to substitute for tacrolimus or methotrexate in
patients who had baseline or early organ dysfunction.
Therefore, these results may not be representative for pa-
tients who receive tacrolimus/MMF as planned aGVHD
prophylaxis.
The retrospective study design made it difficult to
retrieve more detailed information about clinical course of
BKV disease. There was likely under-reporting of symptoms
and duration of BKV disease, especially for milder cases.
We were unable to systematically capture symptoms
commonly mentioned by patients who experience BKV
disease, such as the degree urinary frequency experienced,
multiple times per hour at times, and sleep deprivation
because of urinary frequency or bladder spasms, which are
associated with a poor quality of life [46]. Prospective
studies that capture details of symptoms using standard-
ized forms or questionnaires would increase the precision
of these observations. Another limitation of this study is
that there was no systematic BKV screening of urine and
blooddthe trigger for BKV testing was symptom-based.
Therefore, under-reporting is possible due to lack of infor-
mation available on patients without clinical suspicion of
BKV disease.
569
In summary, BKV disease is a common complication of
HSCT, associated with significant and prolonged morbidity,
especially in the setting of aGVHD, cord blood HSCT, cyclo-
phosphamide use, and probably MMF use. BKV disease is
multifactorial in origin and is associated with immunological
and cytotoxic events that take place during or after alloge-
neic HSCT. Prospective studies are needed to better under-
stand the physiopathology of BKV disease, further define its
morbidity, and inform the design of prophylaxis and treat-
ment trials for this common condition after HSCT.
ACKNOWLEDGMENTS
Financial disclosure: The Dutch Kidney Foundation, the
Dutch Cancer Society, the Marco Polo fund, and the J.K. de
Cock foundation provided financial support for N.M.G.R.
Conflict of interest statement: F.M.M. has received consul-
ting honoraria from Chimerix and Vertex and research
funding from Chimerix. All other authors declare no
competing financial interests.
Authorship statement: F.M.M. N.M.G.R., and S.P.H,
conceived of and designed the study. N.M.G.R., M.M.S., G.S.,
V.T.H., and F.M.M. collected and assembled the data. N.M.G.R.
and F.M.M. performed statistical analysis. N.M.G.R. prepared
the manuscript. M.M.S., G.S., S.P.H., V.T.H., L.R.B., J.H.A., R.J.S.,
and F.M.M. reviewed and critically revised the manuscript.
REFERENCES
1. Copelan EA. Hematopoietic stem-cell transplantation. N Engl J Med.
2006;354:1813-1826.
2. Pasquini M, Wang Z. Current use and outcome of hematopoietic stem
cell transplantation. CIBMTR Summ. Slides 2011. Available at: http://
www.cibmtr.org. Accessed April 21, 2012.
3. Gardner S, Field A, Coleman D, Hulme B. New papovavirus (B.K.) iso-
lated from urine after renal transplantion. Lancet. 1971;1:253-257.
4. Chesters PM, Heritage J, McCance DJ. Persistence of DNA sequences of
BK virus and JC virus in normal persistence human tissues and in
diseased tissues. J Infect Dis. 1983;147:676-684.
5. Reploeg MD, Storch GA, Clifford DB. Bk virus: a clinical review. Clin
Infect Dis. 2001;33:191-202.
6. Arthur RR, Shah KV, Baust SJ, Santos GW, Saral R. Association of BK
viruria with hemorrhagic cystitis in recipients of bone marrow trans-
plants. N Engl J Med. 1986;315:230-234.
7. Bedi A, Miller CB, Hanson JL, et al. Association of BK virus with failure of
prophylaxis against hemorrhagic cystitis following bone marrow
transplantation. J Clin Oncol. 1995;13:1103-1109.
8. Lekakis LJ, Macrinici V, Baraboutis IG, et al. BK virus nephropathy after
allogeneic stem cell transplantation: a case report and literature re-
view. Am J Hematol. 2009;84:243-246.
9. Limaye AP, Smith KD, Cook L, et al. Polyomavirus nephropathy in native
kidneys of non-renal transplant recipients. Am J Transplant. 2005;5:
614-620.
10. Dalianis T, Ljungman P. Full myeloablative conditioning and an unre-
lated HLA mismatched donor increase the risk for BK virus-positive
hemorrhagic cystitis in allogeneic hematopoetic stem cell trans-
planted patients. Anticancer Res. 2011;31:939-944.
11. Dropulic LK, Jones RJ. Polyomavirus BK infection in blood and marrow
transplant recipients. Bone Marrow Transplant. 2008;41:11-18.
12. Bennett SM, Broekema NM, Imperiale MJ. BK polyomavirus: Emerging
pathogen. Microbes Infect. 2012;14:672-683.
13. Leung AYH, Suen C, Lie A, et al. Quantification of polyoma BK viruria in
hemorrhagic cystitis complicating bone marrow transplantation. Blood.
2001;98:1971-1978.
14. O’Donnell PH, Swanson K, Josephson MA, et al. BK virus infection is
associated with hematuria and renal impairment in recipients of
allogeneic hematopoietic stem cell transplants. Biol. Blood Marrow
Transplant. 2009;15:1038-1048.
15. Apperley JF, Rice SJ, Bishop JA, et al. Late-onset hemorrhagic cystitis
associated with urinary excretion of polyomaviruses after bone
marrow transplantation. Transplantation. 1987;43:108-112.
16. Erard V, Storer B, Corey L, et al. BK virus infection in hematopoietic
stem cell transplant recipients: frequency, risk factors, and association
with postengraftment hemorrhagic cystitis. Clin Infect Dis. 2004;39:
1861-1865.
17. Erard V, Kim HW, Corey L, et al. BK DNA viral load in plasma: evidence
for an association with hemorrhagic cystitis in allogeneic hematopoi-
etic cell transplant recipients. Blood. 2005;106:1130-1132.
570 N.M.G. Rorije et al. / Biol Blood Marrow Transplant 20 (2014) 564e570
18. Sencer SF, Haake RJ, Weisdorf DJ. Hemorrhagic cystitis after bone
marrow transplantation. Transplantation. 1993;56:875-879.
19. Seber A, Shu XO, Defor T, et al. Risk factors for severe hemorrhagic
cystitis following BMT. Bone Marrow Transplant. 1999;23:35-40.
20. Leung AYH, Mak R, Lie AKW, et al. Clinicopathological features and risk
factors of clinically overt haemorrhagic cystitis complicating bone
marrow transplantation. Bone Marrow Transplant. 2002;29:509-513.
21. Asano Y, Kanda Y, Ogawa N, et al. Male predominance among Japanese
adult patients with late-onset hemorrhagic cystitis after hematopoietic
stem cell transplantation. Bone Marrow Transplant. 2003;32:
1175-1179.
22. Tsuboi K, Kishi K, Ohmachi K, et al. Multivariate analysis of risk factors
for hemorrhagic cystitis after hematopoietic stem cell transplantation.
Bone Marrow Transplant. 2003;32:903-907.
23. Yamamoto R, Kusumi E, Kami M, et al. Late hemorrhagic cystitis after
reduced-intensity hematopoietic stem cell transplantation (RIST). Bone
Marrow Transplant. 2003;32:1089-1095.
24. El-Zimaity M, Saliba R, Chan K, et al. Hemorrhagic cystitis after allo-
geneic hematopoietic stem cell transplantation: donor type matters.
Blood. 2004;103:4674-4680.
25. Giraud G, Bogdanovic G, Priftakis P, et al. The incidence of hemorrhagic
cystitis and BK-viruria in allogeneic hematopoietic stem cell recipients
according to intensity of the conditioning regimen. Haematologica.
2006;91:401-404.
26. Wong ASY, Chan K-H, Cheng VCC, et al. Relationship of pre-
transplantation polyoma BK virus serologic findings and BK viral
reactivation after hematopoietic stem cell transplantation. Clin Infect
Dis. 2007;44:830-837.
27. Giraud G, Priftakis P, Bogdanovic G, et al. BK-viruria and haemorrhagic
cystitis are more frequent in allogeneic haematopoietic stem cell
transplant patients receiving full conditioning and unrelated-HLA-
mismatched grafts. Bone Marrow Transplant. 2008;41:737-742.
28. Tanaka K, Hori T, Hatakeyama N, et al. Quantification of BK polyoma
viruria in Japanese children and adults with hemorrhagic cystitis
complicating stem cell transplantation. J Med Virol. 2008;80:
2108-2112.
29. De Padua Silva L, Patah P, Saliba RM, et al. Hemorrhagic cystitis after
allogeneic hematopoietic stem cell transplants is the complex result of
BK virus infection, preparative regimen intensity and donor type.
Haematologica. 2010;95:1183-1190.
30. Arai Y, Maeda T, Sugiura H, et al. Risk factors for and prognosis of
hemorrhagic cystitis after allogeneic stem cell transplantation: Retro-
spective analysis in a single institution. Hematology. 2012;17:207-214.
31. Mori Y, Miyamoto T, Kato K, et al. Different risk factors related to
adenovirus- or BK virus-associated hemorrhagic cystitis following
allogeneic stem cell transplantation. Biol Blood Marrow Transplant.
2012;18:458-465.
32. Alyea EP, Kim HT, Ho V, et al. Comparative outcome of non-
myeloablative and myeloablative allogeneic hematopoietic cell trans-
plantation for patients older than 50 years of age. Blood. 2005;105:
1810-1814.
33. Mohty M, Gaugler B. Advances in umbilical cord transplantation: the
role of thymoglobulin/ATG in cord blood transplantation. Best Pract Res
Clin Haematol. 2010;23:275-282.
34. Cutler C, Stevenson K, Kim HT, et al. Double umbilical cord blood
transplantation with reduced intensity conditioning and sirolimus-
based GVHD prophylaxis. Bone Marrow Transplant. 2011;46:659-667.
35. Shepherd BJD, Pringle LE, Barnett MJ, et al. Mesna versus hyper-
hydration for the prevention of cyclophosphamide-induced hemor-
rhagic cystitis in bone marrow transplantation. J Clin Oncol. 1991;9:
2016-2020.
36. Ho VT, Aldridge J, Kim HT, et al. Comparison of tacrolimus and siroli-
mus (Tac/Sir) versus tacrolimus, sirolimus, and mini-methotrexate
(Tac/Sir/MTX) as acute graft-versus-host disease prophylaxis after
reduced-intensity conditioning allogeneic peripheral blood stem cell
transplantation. Biol Blood Marrow Transplant. 2009;15:844-850.
37. Cutler C, Li S, Ho VT, et al. Extended follow-up of methotrexate-free
immunosuppression using sirolimus and tacrolimus in related and
unrelated donor peripheral blood stem cell transplantation. Blood.
2007;109:3108-3114.
38. Marty FM, Bryar J, Browne SK, et al. Sirolimus-based graft-versus-host
disease prophylaxis protects against cytomegalovirus reactivation after
allogeneic hematopoietic stem cell transplantation: a cohort analysis.
Blood. 2007;110:2-4.
39. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference
on Acute GVHD Grading. Bone Marrow Transplant. 1995;15:825-828.
40. Hill JA, Koo S, Guzman Suarez BB, et al. Cord-blood hematopoietic stem
cell transplant confers an increased risk for human herpesvirus-6-
associated acute limbic encephalitis: a cohort analysis. Biol. Blood
Marrow Transplant. 2012;18:1638-1648.
41. Schulze M, Beck R, Igney A, et al. Computed tomography findings of
human polyomavirus BK (BKV)-associated cystitis in allogeneic he-
matopoietic stem cell transplant recipients. Acta Radiol. 2008;49:
1187-1194.
42. Leung AYH, Chan MTL, Yuen K, et al. Ciprofloxacin decreased polyoma
BK virus load in patients who underwent allogeneic hematopoietic
stem cell transplantation. Clin Infect Dis. 2005;40:528-537.
43. Miller AN, Glode A, Hogan KR, et al. Efficacy and safety of ciprofloxacin
for prophylaxis of polyomavirus BK virus-associated hemorrhagic
cystitis in allogeneic hematopoietic stem cell transplantation re-
cipients. Biol Blood Marrow Transplant. 2011;17:1176-1181.
44. Marty FM, Lee SJ, Fahey MM, et al. Infliximab use in patients with
severe graft-versus-host disease and other emerging risk factors of
non-Candida invasive fungal infections in allogeneic hematopoietic
stem cell transplant recipients: a cohort study. Blood. 2003;102:
2768-2776.
45. Leung AYH, Yuen K-Y, Kwong Y-L. Polyoma BK virus and haemorrhagic
cystitis in haematopoietic stem cell transplantation: a changing para-
digm. Bone Marrow Transplant. 2005;36:929-937.
46. Syrjala KL, Langer SL, Abrams JR, et al. Late effects of hematopoietic cell
transplantation among 10-year adult survivors compared with case-
matched controls. J Clin Oncol. 2005;23:6596-6606.