39 12 1861
39 12 1861
39 12 1861
median of 10 days (range, 1–25 days) before the event. Of these only factors associated with BK viremia. However, we were
3 patients, 1 patient had PCR performed for the detection of unable to demonstrate that the presence of CMV DNA in
BK virus in urine samples at the time of the event, and the plasma or pp65 antigenemia after transplantation was associ-
results were positive. Urine samples from all 3 patients were ated with BK viremia; thus, the correlation between CMV sero-
tested for adenovirus, and the results were negative. Univariate status and BK viremia does not have an obvious biological
time-dependent analysis showed a strong association between explanation, and it requires further validation.
BK viremia and postengraftment hemorrhagic cystitis (P p In contrast to the results of studies of other viral infections
.004). The maximum levels of BK viremia in patients with among HSCT recipients, factors affecting cellular immunity,
postengraftment hemorrhagic cystitis (median, 9.7 ⫻ 10 3 cop- such as lymphopenia, the use of corticosteroids, and graft-
ies/mL; range, 1.8 ⫻ 10 3 to 2.4 ⫻ 10 5 copies/mL) were higher versus-host disease, were not significantly associated with BK
than those in patients without hemorrhagic cystitis (median, 0 viremia in the present study. This finding suggests that, in
copies/mL; range, 0 to 3.7 ⫻ 10 6 copies/mL; P p .008) and addition to impairment of the immune system, other—as yet
those in patients with preengraftment hemorrhagic cystitis (me- unknown—factors contribute to the dissemination of BK virus
dian, 0 copies/mL; range, 0 to 2.7 ⫻ 10 4 copies/mL; P p infection.
.0006). An association between hemorrhagic cystitis in HSCT recip-
Discussion. The present large longitudinal study of 132 ients and the presence of BK virus DNA in plasma has been
patients showed that 33% of patients had BK virus that was examined by Leung et al. [1]. In their study, BK viremia was
quantifiable in plasma. Univariate analysis of risk factors for not consistently associated with hemorrhagic cystitis. Hemor-
any occurrence of BK viremia or for persistent BK viremia rhagic cystitis was associated with low platelet counts; therefore,
identified CMV seropositivity and the underlying disease as the in the present study, we focused our analysis on postengraft-
Persistent
a
BK viremia BK viremia
(44 events) (20 events)
Risk factor HR (95% CI) Pb HR (95% CI) Pb
CMV serostatus
Of patient .03 .06
Negative (n p 75) 1.0 (Referent) 1.0
Positive (n p 57) 1.9 (1.1–3.4) 2.3 (1.0–5.7)
Of donor and recipient .007 .03
NOTE. Only results with a P value of !.1 are shown. Additional risk factors that were tested but
were not found to be significant included the sex of the transplant recipient, the source of the cells
(peripheral blood stem cells vs. bone marrow), the donor type (autologous vs. allogeneic), the use of
antithymocyte globulin, acute graft-versus-host disease, plasma cytomegalovirus viremia (pp65 110
leukocytes and 150 leukocytes), or positive PCR results (11000 copies/mL and 110,000 copies/mL) for
blood samples, or the lack of lymphocyte engraftment. CML-CP, chronic myeloid leukemia–chronic
phase; D, donor; HD, Hodgkin disease; HHV-6, human herpesvirus 6; HR, hazard ratio; MM, multiple
myeloma; NHL, non-Hodgkin lymphoma; R, recipient; TBI, total body irradiation.
a
Analysis was performed for patients who had at least 3 plasma samples available and for whom
⭓50% of the PCR results for those samples were positive.
b
Overall test of association for each category of risk factor.
c
Based on average.
d
Includes scleroderma, solid tumor, myelodysplastic syndrome, aplastic anemia, and multiple
myeloma.
ment hemorrhagic cystitis to avoid the potential confounding viremia. We speculate that BK virus reactivation originates in
role of thrombocytopenia in the occurrence of hemorrhagic the urinary tract, leading to secondary BK viremia.
cystitis. BK viremia was highly correlated with hemorrhagic The present study has strengths and limitations. Its
cystitis in patients with sustained platelet engraftment; however, strengths include the large sample size, which represents the
we were unable to adjust for others factors because of the small largest cohort to date; the unselected and consecutive nature
number of cases. Although the difference in viral loads between of the patient cohort; weekly sampling; and the use of a val-
patients with postengraftment hemorrhagic cystitis and patients idated quantitative assay. Its limitations include the retro-
with preengraftment hemorrhagic cystitis or patients without spective determination of hemorrhagic cystitis; the lack of
hemorrhagic cystitis was highly significant, the wide range of detection of BK virus in the urine samples obtained from
viral loads in each group limited the clinical usefulness of this some patients; and the small number of cases of postengraft-
measurement. ment hemorrhagic cystitis, which prevented us from per-
Further studies are needed to determine the origin of BK forming multivariate statistical modeling.