Safety and Efficacy of I tte-
Interferon �-2b Therapy Blad
in Patients With Prosthetic Devices
He
an
Fro
A
a
A
B
C
N
b
prov
Univ
City
FAX
uiow
with
cals
trum
Ente
19
nry M. Rosevear, Andrew J. Lightfoot, Kenneth G. Nepple
d Michael A. O’Donnell*,†
m the Department of Urology, University of Iowa, Iowa City, Iowa
bbreviations
nd Acronyms
HV� artificial heart valve
CG� bacillus Calmette-Guérin
IS� carcinoma in situ
MIBC� nonmuscle invasive
ladder cancer
Submitted for publication March 8, 2010.
Study received institutional review board ap-
al.
* Correspondence: Department of Urology,
ersity of Iowa, 200 Hawkins Dr., 3 RCP, Iowa
, Iowa 52242-1089 (telephone: 319-384-6981;
: 319-356-3900; e-mail: michael-odonnell@
a.edu).
† Financial interest and/or other relationship
Abbott Laboratories, Alynlam Pharmaceuti-
, Viventia, Anadys Pharmaceuticals, Spec-
, Loras, Endo Pharmaceuticals and Medical
rprises.
Purpose: Patients with bladder cancer who have prosthetic devices, such as a
cardiac pacemaker, artificial heart valve or orthopedic hardware, and who un-
dergo intravesical bacillus Calmette-Guérin therapy are theoretically at higher
risk for complications, including bacterial seeding of pacemaker wires or ortho-
pedic hardware, and at further risk for infective endocarditis. We assessed the
safety and efficacy of bacillus Calmette-Guérin plus interferon �-2b therapy in
patients with nonmuscle invasive bladder cancer and a pacemaker, artificial
heart valve or orthopedic hardware.
Materials and Methods: We evaluated 1,045 patients with nonmuscle invasive
bladder cancer enrolled in a multicenter American phase II trial of bacillus
Calmette-Guérin plus interferon �-2b therapy, including 143 with a prosthetic
device (pacemaker in 87, artificial heart valve in 13 and orthopedic hardware in
43). Weekly physician toxicity assessments and standard adverse effect reporting
were done.
Results: No patient had infective endocarditis or hardware infection. One patient
with a pacemaker, 2 with orthopedic hardware and none with an artificial heart
valve required treatment cessation for fever greater than 102.5F. All defervesced
within 24 hours and had no long-term sequelae. Due to intolerable, nonlife
threatening side effects 12 patients with a pacemaker, 2 with orthopedic hard-
ware and 1 with an artificial heart valve stopped treatment. Of the remaining
patients with a prosthesis 99 and 24 stopped treatment due to intolerable, nonlife
threatening and serious side effects, respectively.
Conclusions: Patients with a pacemaker, artificial heart valve or orthopedic
hardware were no more likely than the general population to have infection or
fever, or discontinue treatment due to side effects. These patients should not be
excluded from intravesical bacillus Calmette-Guérin plus interferon �-2b therapy
for nonmuscle invasive bladder cancer.
Key Words: bladder, bladder neoplasms, BCG vaccine, interferons,
prostheses and implants
ACCORDING to the most recent American
Urological Association guidelines BCG
is the preferred adjuvant therapy for
carcinoma in situ and high risk non-
muscle invasive bladder cancer.1 Inter-
feron-� combined with BCG has been
investigated to improve the overall out-
come, and allow BCG dose reduction
during maintenance and even during
induction in relapsing cases.2–4
BCG is a live attenuated strain of
Mycobacterium bovis originally devel-
20 www.jurology.com
0022-5347/10/1845-1920/0 Vol. 184, 1920-1924, November 2010
THE JOURNAL OF UROLOGY® Printed in U.S.A.
© 2010 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC. DOI:10.1016/j.juro.2010.06.149
ntravesical Bacillus Calme
for Nonmuscle Invasive
Guérin Plus
der Cancer
op
ex
for
a
tre
pr
Th
cli
th
blo
to
pa
co
wi
pl
fil
m
BC
ca
th
pa
th
m
(A
TI
tie
pa
fu
ex
pe
W
in
bin
NM
ha
PA
A
inc
ga
tre
BC
Fe
th
ter
su
pa
cyc
a t
th
sta
ac
his
ur
cop
pr
th
fir
ren
tio
dia
ex
AH
giv
we
irr
tit
ar
de
wh
We
sta
de
gra
viv
aft
tes
ing
an
fec
do
an
RE
Of
(9
or
(1
m
in
(se
or
an
stu
en
sta
th
5
th
me
re
po
pa
(4
wi
th
tre
en
BACILLUS CALMETTE-GUÉRIN IN PATIENTS WITH PROSTHETIC DEVICES 1921
ed in 1921 as a tuberculosis vaccine.5 While the
act mechanism of action of BCG is unknown, ef-
ts to use heat treated BCG have failed to produce
similar therapeutic effect, presumably since heat
ated BCG is not internalized by bladder cells,
eventing initiation of the inflammatory cascade.
us, the need to use a live bacterium creates a
nical challenge since in a small percent of patients
e serious complication of BCG infection in the
odstream (BCG sepsis) may develop.6
BCG can disrupt urothelial cells with the ability
gain access to the lymphatics and blood. Certain
tient populations are at theoretically increased
mplication risk. Immunocompromised patients
th a history of chronic steroid use or organ trans-
antation tolerate BCG treatment with toxicity pro-
es similar to those of the general population but
any clinicians are still reluctant to treat them with
G.7,8 Patients at theoretically increased compli-
tion risk due to intravesical BCG therapy are
ose with a history of prosthetic surgery, including
tients with a cardiac pacemaker, an AHV or or-
opedic hardware. Manufacturers of the most com-
only used strains of BCG, including Connaught
ventis-Pasteur, Swiftwater, Pennsylvania) and
CE®, recommend against using BCG in these pa-
nts. The concern is a risk of bacterial seeding of
cemaker wires or orthopedic hardware and the
rther risk of infective endocarditis. Case reports
ist describing pacemaker wire seeding and ortho-
dic hardware infection after BCG treatment.9–12
e analyzed the experience in a national, multi-
stitutional, phase II trial of treatment with com-
ed BCG plus interferon �-2b in patients with
IBC and a pacemaker, an AHV or orthopedic
rdware to evaluate BCG safety and efficacy.
TIENTS AND METHODS
total of 1,106 patients with NMIBC (CIS, T1 or Ta1),
luding those with primary and recurrent disease re-
rdless of previous intravesical chemotherapy or BCG
atment, were enrolled in a phase II study of combined
G plus interferon �-2b therapy between May 1999 and
bruary 2001.4 Briefly, all patients underwent transure-
ral resection, followed by an intravesical BCG plus in-
feron �-2b induction cycle tailored to prior BCG expo-
re with dose reduction to a minimum of a third in
tients previously treated with BCG. BCG maintenance
les were automatically dose reduced, usually as a third,
enth and a tenth during each maintenance cycle, but
e interferon �-2b dose of 50 million U remained con-
nt. TICE or Connaught strain was used at all centers
cording to physician practice preference.
Tumor recurrence was defined as visible tumor unless
tologically confirmed to be benign, definitive positive
ine cytology, positive biopsy even with negative cystos-
y and any transitional cell carcinoma regardless of
esenting site, including the upper tract, prostate, ure-
du
ha
ra or metastasis. Time to recurrence was indexed to the
st intravesical treatment date. Patients in whom recur-
ce was not observed were treated as censored observa-
ns on survival analysis. Patients with a history of a car-
c pacemaker, an AHV or orthopedic hardware were not
cluded by study protocol. Data on bovine vs mechanical
Vs were not available. Prophylactic antibiotics were not
en before intravesical BCG therapy according to protocol.
Each patient completed a daily questionnaire for 1
ek after each BCG plus interferon �-2b cycle, describing
itative symptoms such as frequency, urgency and cys-
is, hematuria, flu-like symptoms, fevers, arthralgia or
thritic symptoms and any other health related inci-
nts. Before subsequent cycles patients were asked
ether they had required hospitalization in the interim.
ekly physician toxicity assessments were done and
ndard serious adverse events were reported. Fever was
fined as temperature greater than 102.5F since low
de fever has been reported as a common side effect.
We used the Kaplan-Meier method to construct sur-
al curves and estimate cancer-free survival 3 years
er the initiation of intravesical therapy. The log rank
t for trend was used to test for recurrence rate order-
across age categories. We applied Cox univariate
d multivariate regression models to estimate the ef-
t of age on time to cancer recurrence. All tests were
ne with p �0.05 considered significant using SAS®
d GraphPad® Prism®.
SULTS
the original 1,106 clinical trial patients 1,045
5%) provided information on pacemakers, AHVs or
thopedic hardware. Of the 1,045 patients 143
3.7%) had a prosthetic device, including a pace-
aker in 87, an AHV in 13 and orthopedic hardware
43, including an artificial joint or Harrington rod
e table). Patients with an AHV, a pacemaker and
thopedic hardware tended to be older (ages 76, 75
d 73 years, respectively, vs 68 in the remaining
dy population) but this was not significantly differ-
t. Previous analysis of this study suggested that
ge (T1 vs Ta), tumor size (less than 1 vs greater
an 5 cm), tumor multifocality (greater than 5 vs 2 to
vs solitary) and prior BCG (greater than 2 vs less
an 1) were significant predictors of successful treat-
nt.4 However, study groups were similar in these
spects to the overall study population.
BCG dosing varied based on patient previous ex-
sure to BCG. Of the patients 49 of 87 (56%) with a
cemaker, 7 of 13 (54%) with an AHV, 19 of 43
4%) with orthopedic hardware and 554 of 902
thout hardware (61%) received full dose BCG
erapy. The remainder received reduced dose BCG
atment at a third or a tenth dose. These differ-
ces were not significantly different.
Figure 1 shows the Kaplan-Meier curve for the
rable disease-free rate. The AHV and the non-
rdware cohorts tended to have higher recurrence-
fre
th
(42
no
tie
wi
fol
m
af
wi
pa
to
tie
ph
(8
in
43
ce
wi
tie
ca
cia
wi
tio
fev
(2.
di
th
(2
di
fec
or
sim
DI
Pa
ha
ar
op
co
ra
wh
26
de
in
se
BC
ar
tio
m
tre
an
ar
op
ye
pla
tiv
be
dia
sin
Stu
Me
%
%
T
A
C
%
L
I
H
No.
L
1
G
%
No.
tu
L
5
%
S
2
G
%
* p
† M
BACILLUS CALMETTE-GUÉRIN IN PATIENTS WITH PROSTHETIC DEVICES1922
e survival (57% and 55% disease-free at 3 years)
an the pacemaker and orthopedic hardware groups
% and 37%, respectively) but this difference was
t statistically significant.
Regarding treatment safety and toxicity, no pa-
nt in any cohort had endocarditis, or pacemaker
re, joint or hardware infection during 3 years of
lowup. One of the 87 patients (1%) with a pace-
aker required cessation of treatment due to fever
ter treatment 6, which resolved within 24 hours
th no sequelae. Another 12 patients (14%) with a
cemaker requested treatment cessation due to in-
lerable, nonlife threatening side effects. No pa-
nt with an AHV had treatment stopped by the
ysician due to significant side effects while 1 of 13
%) stopped treatment at patient request due to
tolerable, nonlife threatening side effects. Two of
patients (5%) with orthopedic hardware required
ssation of treatment due to fever, which defervesced
thin 24 hours with no sequelae. An additional pa-
nt (2%) with orthopedic hardware showed signifi-
nt irritative voiding symptoms, for which the physi-
n discontinued treatment. Another 2 patients (5%)
th orthopedic hardware requested treatment cessa-
n due to nonsignificant side effects but neither had
er. Overall in the prosthetic group 3 of 143 patients
1%) had fever greater than 102.5F and 15 (10.5%)
dy cohort demographic data
Parameter* Nonprosthetic AHV Pacemaker
Orthopedic
Hardware
an age (median) 68 (70) 76 (77) 75 (76) 73 (75)
Male/female† 75/25 77/23 75/25 67/33
Stage:†
a only 48 38 41 53
ny T1 27 31 34 23
IS (any/isolated) 36 31 36 20
Grade:†
ow 17 23 16 21
ntermediate 28 15 25 26
igh 42 31 44 51
size (cm):
ess than 1 32 38 37 33
–5 43 46 40 47
reater than 5 7 15 10 11
Prior chemotherapy† 23 23 23 20
transurethral bladder
mor resection:
ess than 5 84 85 80 79
or Greater 14 15 20 19
Multifocal:†
olitary 34 23 26 35
–5 38 38 38 23
reater than 5 18 23 17 17
Primary/recurrent† 28/68 23/54 28/70 19/77
�0.05.
ay not total 100% since some parameters were not specified in some patients.
scontinued treatment due to minor side effects. Of
e remaining 902 patients with no prosthesis 20
.2%) had fever and 99 (11.0%) requested treatment
Fig
(p
scontinuation due to nonlife threatening side ef-
ts. Overall patients with a pacemaker, an AHV or
thopedic hardware had an overall toxicity profile
ilar to that of the remaining cohort (fig. 2).
SCUSSION
tients with bladder CIS or high risk NMIBC who
ve a pacemaker, an AHV or orthopedic hardware
e a clinical dilemma for practicing urologists since
timal therapy includes intravesical BCG, which is
ntraindicated in the drug package insert. A real but
re complication of BCG therapy is BCG bacteremia,
ich was reported by Lamm to develop in 0.4% of
02 treated patients.6 Populations with a prosthetic
vice would theoretically be at risk for the infection of
dwelling artificial prosthetic material.
Case reports exist describing pacemaker wire
eding and orthopedic hardware infection after
G treatment.9–12 In a report of loose total hip
throplasty revision done 10 years after implanta-
n watery pus was found that grew BCG 10
onths after the start of induction intravesical
atment that also included maintenance.11 Segal
d Krauss reported a case of BCG infected total hip
throplasty with an iliopsoas abscess that devel-
ed more than 10 years after implantation and 4
ars after intravesical treatment.10 Hip arthro-
sty osteomyelitis, vertebral osteomyelitis and reac-
e polyarthritis after intravesical BCG have also
en reported in rare cases.10,13 BCG infection of car-
c valves or leads are even less commonly reported
ce our literature review only revealed 1 reported
ure 1. Kaplan-Meier analysis of durable disease-free rate
�0.05).
ca
wi
ha
wa
ca
re
th
Re
th
gu
so
for
in
co
Gu
cle
tio
ica
an
pl
m
20
ge
th
ro
pl
th
Fig . C, pa
1 t
BACILLUS CALMETTE-GUÉRIN IN PATIENTS WITH PROSTHETIC DEVICES 1923
se of defibrillator BCG infection.9 BCG infection
th mycotic aortic aneurysms and an aortic graft
ve been reported but not in relation to an AHV.14–16
Based on the anecdotally reported risk of hard-
re or pacemaker wire infection, or infective endo-
rditis and the lack of safety data the previous
commendation was to withhold BCG therapy in
is patient population and err on the side of caution.
cently the AmericanHeart Association Taskforce on
e Prevention of Infective Endocarditis changed its
idelines, stating that “administration of antibiotics
lely to prevent endocarditis is not recommended
patients who undergo a genitourinary or gastro-
testinal tract procedure,” citing no evidence-based
nclusive link to an increased endocarditis risk.17
idelines for prior orthopedic hardware are less
ar since the 2008 American Urological Associa-
n guidelines recommend antibiotics before urolog-
l procedures only if there is a risk of bacteremia
d a risk factor in the patient, including joint re-
acement within 2 years, prior joint infection, im-
unocompromise, diabetes or malignancy.18 In
09 the American Academy of Orthopaedic Sur-
ons expanded recommendations to “recommend
at clinicians consider antibiotic prophylaxis (cip-
ure 2. Cohort toxicity profiles. A, AHVs. B, orthopedic hardware
o 6, respectively, of weekly treatment cycle.
floxacin dose 1 hour prior) for all total joint re-
acement patients before any invasive procedure
at may cause bacteremia,” especially in those with
ha
lat
te
k factors, while excluding nonsynovial hard-
re.19
Our retrospective review suggests that BCG plus
terferon �-2b therapy does not increase the risk of
stemic or local side effects in patients with a pace-
ker, an AHV or orthopedic hardware compared to
at in the general population without hardware. The
k of bacteremia due to catheterization and intraves-
l treatment is presumed to be quite low but it was
t specifically studied in our patient population.
In 2007 Agrawal et al reported a randomized trial
ggesting that BCG dose reduction decreases the
nical toxicity of treatment without affecting effi-
cy.20 This observation allows continued treatment
th BCG dose reduction in patients who have mild
nical side effects. The statistical similarity of our
tient cohorts supports full dose BCG treatment as
st line therapy in those with a prosthetic device
th dose reduction according to standard regimen.
Intravesical therapy in this population is not
thout risk. BCG is a live bacterium and the risks
herent to this treatment are well reported, includ-
g the ability to gain access to the lymphatics and
od. Rather, our study shows that clinical risks in
tients with a pacemaker, an AHV or orthopedic
cemakers. D, nonprosthetic cases. TR-I1 to TR-I6, treatments
ris
wa
in
sy
ma
th
ris
ica
no
su
cli
ca
wi
cli
pa
fir
wi
wi
in
in
blo
pa
rdware are no greater than in the general popu-
ion. We acknowledge that the implications of bac-
remia in patients with a prosthetic device are
cle shou
to loo
clu r tr
re cal
ac ma
iza
s (1
ca pedi
wh m a
tra ncer
kn ll p
bla e re
th nt i
po diti
wa he k
fit
thi
tiv ple
low spe
sig acy
BC pat
pa ha
su thes
pl o ou
sent
se u
BC
nts
sses
foll
rowt
to de
sho
CG
d to
s.
plus
acem
rs a
on.
y in
nt
e of
inf
ins
ion
RE
1.
for the management of nonmuscle invasive blad-
2.
3.
4.
5.
6.
7.
bacillus Calmette-Guerin for the treatment of
ncer in r
n 2003;
II and Kl
on of an
esical th
nfect Dis
S: Infect
al bacill
y 2007;
ers P: A
mycobac
nt with
ncer. J
efauvea
f BCG th
’Keefe
litis inv
lar baci
in Infect
an E, M
and ve
bacillu
oentgen
therapy for bladder cancer. South Med J 1995;
BACILLUS CALMETTE-GUÉRIN IN PATIENTS WITH PROSTHETIC DEVICES1924
der cancer (stages Ta, T1, and Tis): 2007 update.
J Urol 2007; 178: 2314.
Belldegrun AS, Franklin JR, O’Donnell MA et al:
Superficial bladder cancer: the role of interferon-
alpha. J Urol 1998; 159: 1793.
Weiss GR, O’Donnell MA, Loughlin K et al: Phase
1 study of the intravesical administration of re-
combinant human interleukin-12 in patients with
recurrent superficial transitional cell carcinoma of
the bladder. J Immunother 2003; 26: 343.
Joudi FN, Smith BJ and O’Donnell MA: Final results
from a national multicenter phase II trial of combi-
nation bacillus Calmette-Guerin plus interferon al-
pha-2B for reducing recurrence of superficial blad-
der cancer. Urol Oncol 2006; 24: 344.
Morales A, Eidinger D and Bruce AW: Intracavi-
tary bacillus Calmette-Guerin in the treatment of
superficial bladder tumors. J Urol 1976; 116: 180.
Lamm DL: Efficacy and safety of bacille Calmette-
Guerin immunotherapy in superficial bladder can-
cer. Clin Infect Dis, suppl., 2000; 31: S86.
Yossepowitch O, Eggener SE, Bochner BH et al:
Safety and efficacy of intravesical bacillus
Calmette-Guerin instillations in steroid treated
superficial bladder ca
tients. Transplantatio
9. Stone DR, Estes NA I
bacterium bovis infecti
lator following intrav
Calmette-Guerin. Clin I
10. Segal A and Krauss E
plasty after intravesic
therapy. J Arthroplast
11. Reigstad O and Siew
ment infected with
intravesicular treatme
Guerin for bladder ca
2008; 90: 225.
12. Clavel G, Grados F, L
ticular side effects o
Spine 2006; 73: 24.
13. Guerra CE, Betts RF, O
terium bovis osteomye
plasty after intravesicu
for bladder cancer. Cl
14. Rozenblit A, Wasserm
fected aortic aneurysm
tis after intravesical
therapy. AJR Am J R
and immunocompromised patients. J Urol 2006;
176: 482.
15. Hellinger WC, Oldenburg W
Vascular and other serious inf
enal transplant pa-
76: 1514.
empner MS: Myco-
implantable defibril-
erapy with bacille
1993; 16: 825.
ed total hip arthro-
us Calmette-Guerin
22: 759.
total hip replace-
terium bovis after
Bacille-Calmette-
Bone Joint Surg Br
u P et al: Osteoar-
erapy. Joint Bone
RJ et al: Mycobac-
olving a hip arthro-
lle Calmette-Guerin
Dis 1998; 27: 639.
arin ML et al: In-
rtebral osteomyeli-
s Calmette-Guerin
ol 1996;