Journal of the Egyptian Nat. Cancer Inst., Vol. 18, No. 3, September: 203-208, 2006
Bone Metastases in Muscle-Invasive Bladder Cancer
AZZA N. TAHER, M.D.* and MAGDY H. KOTB, M.D.**
The Departments of Radiation Oncology* and Nuclear Medicine**, National Cancer Institute, Cairo University.
ABSTRACT Conclusion: Meticulous search for bone metastasis
alone or as a component of distant failure in the newly
Purpose: To address the necessity of incorporating diagnosed bladder cancer patients is crucial to offer them
isotopic bone scan in the routine staging work-up of the proper management and avoid undue radical surgical
muscle invasive bladder cancer patients, we analyzed the procedures. Thus bone scan is suggested to be performed
data in our files to determine the incidence of bone me- routinely in patients with evidence of muscle invasion.
tastasis in such patients. The rate of subsequent develop-
ment of bone metastasis along the natural history of the Key Words: Bladder cancer - Bone metastasis - Isotopic
disease was also investigated. bone scan.
Patients and Methods: A total of 179 files of consec- INTRODUCTION
utive bladder cancer patients who presented to the National
Cancer Institute, Cairo University, between January 2000 Identifying the metastatic status is an essen-
and December 2001 were reviewed to check the percentage tial determinant of prognosis in patients with
of positive bone scans on presentation and to check the
subsequent development of distant metastases and bone
muscle-invasive bladder cancer treated by cys-
metastasis in these patients’ records. tectomy, and preoperative metastases detection
is crucial for treatment selection in these pa-
Results: Amongst the 179 patients, 26 (14.5%) had tients.
bone metastasis on presentation, a finding that showed a
statistically significant correlation with the increasing Although the exact rate and pattern of me-
depth of muscle invasion; 61.5% of the metastatic cases
had deep muscle invasion,19.2% had superficial muscle tastases are not well documented, there is a
invasion and there was no muscle invasion in 7.7% rough estimate that 30% of the patients have
(p=0.000). undetected distant metastases at the time of
treatment of the primary tumor .
Transitional cell carcinoma was the pathology in
92.3% of those patients, while only 7.7% had squamous PATIENTS AND METHODS
cell carcinoma (p=0.036).
The cumulative 3-year incidence of bone-metastasis A retrospective review was performed of
in the non metastatic patients after treatment mounted to the records of 179 patients with histologically
19.4±4.4%. The cumulative 3-year bone metastasis inci- proven bladder cancer who presented to the
dence in the 153 patients was higher with increasing National Cancer Institute, Cairo University,
clinical stage; 8.4±8% for c-stage 2 and 49.1±18.5% for between January 2000 up till December 2001.
c-stage 4 (p=0.046). As for the p-category of the tumor
in the 130 patients who underwent operation, the incidence Patients who had no initial bone scan on pre-
increased with higher p-stages (p=0.006). Though pelvic sentation were excluded. Our patients comprised
nodal involvement was not associated with statistically those who underwent radical cystectomy or
significant increase in the incidence of bone metastases, anterior pelvic exentration with pelvic lym-
yet when incorporated as one of the 3 risk factors (grade>3, phadenectomy plus urinary diversion, those
p≥4a and LN positive at surgery) according to which
patients were grouped, there was a statistically significant who were medically unfit for surgery, inopera-
difference in the incidence between patients with no risk ble, or metastatic at presentation.
factors, only 1 and 2 or more factors (p=0.021).
The tumors were staged according to the
TNM system of the American Joint Committee
Correspondence: Dr Azza N. Taher, 66 Manial St., Cairo on Cancer (AJCC) . Positive bone scans at
firstname.lastname@example.org presentation in non-symptomatic patients were
204 Bone Metastases in Muscle-Invasive Bladder Cancer
coupled with plain X-rays of the areas of in- pT2 disease. One hundred and five patients
creased uptake in bone scan and in case of a (58.7%) had grade II tumors, while 72 (40.2%)
negative X-ray a C.T. or M.R.I. was performed had grade III and only 2 (1.1%) had grade I
to verify or exclude any bone metastasis. All tumors. Pelvic nodal involvement was docu-
patients had cystoscopic examination with a mented in 64 patients (49.2%) with a comparable
biopsy performed to verify the pathological number of 66 patients (50.8%) as node negative.
diagnosis, alongside the verification of presence
or absence of muscle invasion and its depth. Transitional cell carcinoma was reported in
Only 2 pathological reports had no mention of 121 patients (67.6%), squamous cell in 52
the muscle invasion status. (29.1%), adenocarcinoma in 4 (2.2%), while 2
(1.1%) had undifferentiated carcinoma (Table
Statistical analysis: 1).
Comparisons between different percentages
and frequencies were performed using the stan- On analyzing the 26 patients who were met-
dard Student t-test, F-test, or Pearson’s chi- astatic on presentation, 14 (54%) of them were
square test and p≤0.05 was considered signifi- clinically stage 3, 9 (34.6%) were stage 4, while
cant. The 3-year cumulative bone metastasis 3 (11.4%) were stage 2 (p=0.449).
incidences were measured using the Kaplan-
Meier product limit method . The period of Looking at the depth of muscle invasion
freedom from distant metastasis or bone me- detected histopathologically from the cystoscop-
tastasis was defined as the period from the date ic biopsy specimen, 16 (61.5%) of the metastatic
of radical cystectomy or end of radical radio- cases had deep muscle invasion, 5 (19.2%) had
therapy to the date of occurrence of the first superficial muscle invasion, there was no muscle
site of distant failure and/or bone metastasis or invasion in 2 (7.7%) cases, while 2 cases had
to the date of evaluation. The log-rank test was no comment on the muscle invasion (p=0.000).
used for comparison between survival curves
. The histopathological subtypes also showed
a statistically significant difference concerning
RESULTS the incidence of bone metastasis; 92.3% (24
patients) had transitional cell carcinoma, while
This is a retrospective analysis of 179 pa- only 7.7% (2 patients) had squamous cell car-
tients with histologically proven urinary bladder cinoma (p=0.036).
carcinoma. The male to female ratio was 3.5:1.
Their age ranged from 35 to 82 years, with a Regarding the site distribution of those bony
mean of 57.4±8.9 and a median of 58.0 years. lesions; 8 patients (30.8%) demonstrated a single
bone metastasis, while 18 had 2 or more sites
The majority of the patients (116 = 64.8%) (Table 2).
were clinically stage 3, while 47 (26.3%) were
stage 4, and only 16 (8.9%) of them were As regards the 153 patients with no metasta-
stage 2. sis, the cumulative 3-year distant failure inci-
dence amounted to 29.6±4.9%. The 30 patients
Twenty six patients (14.5%) had bone me-
out of the 153 who experienced distant failure
tastasis on presentation as proved by bone scan.
were distributed as follows; 12 patients had
Though 153 patients were not metastatic at isolated bone metastasis (40%), 7 (23%) had
presentation, only 130 had undergone radical bone metastasis plus visceral metastasis (4 lung,
cystectomy or anterior pelvic exentration with 2 lymph nodes, 1 liver), 5 (17%) had only lung
pelvic lymphadenectomy plus urinary diversion, metastasis, 3 (10%) had liver while only 2 had
since 5 patients were medically unfit for surgical isolated supraclavicular lymph nodal metastasis
intervention, 5 others refused surgery, while 13 and 1 had intestinal metastasis. Whereas, the
patients were judged to be inoperable by the cumulative 3-year bone-metastasis incidence
surgeon. was 19.4±4.4% for the same group Fig. (1).
A total of 99 patients (76.2%) had pT3 dis- As for the cumulative bone metastasis inci-
ease, 22 (16.9%) had pT4, and 9 (6.9%) had dence according to the clinical staging of the
Azza N. Taher & Magdy H. Kotb 205
153 patients, the incidence was 8.4±8% for The same cumulative incidences were 0%
stage 2, 15.4±4.6% & 49.1±18.5% for stage 3 for grade 1, 18.3±5.8% & 25.9±8.5% for grade
& 4 respectively (p=0.046) (Fig. 2). As for the 2 and 3 respectively, p=0.39 (Fig. 4).
p-category of the tumor in the 130 patients who
underwent surgery the incidence was 10±9.5% On stratifying the patients according to the
for p2, 15.8±4.7% for p3, and 53.3±17.1% for presence or absence of pathological risk factors,
p4, these differences were statistically signifi- the 3-year cumulative bone metastasis incidence
cant at a p-value of 0.006 (Fig. 3). for patients with no risk factors (G<3, p<4a and
LN negative at surgery) was 12.9±6.0%. The
When the incidence was stratified according incidence was 10.2±4.3% for patients who had
to the lymph node status (LN) it was 17.9±5.8% only one of those risk factors, while the inci-
for the LN negative cases (66 patients) and dence went up to 48.2±13.6% for patients with
25.4±8.0% for the LN positive cases (64 pa- more than one risk factor. These differences
tients) (p=0.42). were statistically significant (p=0.021) (Fig. 5).
0 10 20 30 40 0 10 20 30 40
Time (months) Time (months)
c-stage 4 c-stage 3 c-stage 2
4-censored 3-censored 2-censored
Fig. (2): Cumulative 3-year metastasis incidence according
Fig. (1): Cumulative 3-year bone metastasis incidence in
to the clinical stage of tumor.
0 10 20 30 40 -10 0 10 20 30 40
Time (months) Time (months)
Path tumor stage Grade
Stage p4 Stage p3 Stage p2 Grade 3 Grade 2 Grade 1
p4-censored p3-censored p2-censored G3-censored G2-censored G1-cenosred
Fig. (3): Cumulative 3-year bone metastasis incidence Fig. (4): Cumulative 3-year bone metastasis incidence
according to the tumor’s pathological stage. according to tumor’s grade.
206 Bone Metastases in Muscle-Invasive Bladder Cancer
0.5 Table (2): Anatomical distribution of 54 bone metastases
in 26 patients.
Sites No. (%)
Pelvis 17 (32)
0.2 Spine 11 (20)
Femur 8 (15)
0.1 Ribs 5 (9)
Skull 4 (7)
0.0 Humerus 4 (7)
Clavicle 3 (6)
0 10 20 30 40 Sternum 2 (4)
Total No. (%) 54 (100)
>1 Risk factor 1 Risk factor No risk factor
>1-censored 1-censored 0-censored DISCUSSION
Fig. (5): Cumulative 3-year bone metastasis incidence
There is no consistent data or any systematic
according to risk factors stratification. review addressing the rate of bone metastasis
in newly diagnosed bladder cancer cases. Rather,
all published data looked at the rate of distant
Table (1): Patients’ characteristics. failure along the natural history of the disease,
Category Number of patients % whether in a retrospective analysis or more so
as part of autopsy studies that would thus be
Gender: assessing metastases at the end point of those
Male 140 78 cases. Yet dating since 1979 there has been a
Female 39 22
theory suggesting that bladder cancer may be
Clinical stage: a systemic disease from the start, where 30%
II 16 8.9 of the study group developed distant metastasis
III 116 64.8
IV 47 26.3
ñmainly to the lungs and bones- where 80% of
them were detected within 1 year from cystec-
Bone scan: tomy suggesting that the metastases must have
+ve 26 14.5
–ve 153 85.5 been present at cystectomy or as a result of it
 . On the other hand, the more acceptable
Stage: explanation is that most cancer cells spread by
pT2 9/130 6.9
pT2a 1 embolization through lymphatics and blood
pT2b 8 vessels, a theory supported by the identification
pT3 99/130 76.2 of 9 patients with superficial papillary tumors
pT3a 18 with no evidence of muscle invasion in whom
pT3b 81 distant metastases developed; 5 cases had bone
pT4 22/130 16.9
pT4a 20 metastases, lung in 3 and liver in 1 . Two out
pT4b 2 of the 26 patients who presented with bone
metastasis at presentation in our study (7.7%)
I 2 1.1 had no muscle invasion at TUR.
II 105 58.7
III 72 40.2 Reviewing one autopsy series of untreated
patients, the incidence of distant metastases
–ve 66/130 50.8 was 65%  . Another autopsy study on 367
+ve 64/130 49.2 patients with muscle invasive tumors (pT2-4)
found bone metastases in 32% of the cases
Transitional 121 67.6 (ranking third after liver and lung metastases)
Squamous 52 29.1 and the frequency of distant metastases in-
Adenoca. 4 2.2 creased with local tumor extension; and was
Undiff. 2 1.1 slightly higher in patients treated by cystectomy
Azza N. Taher & Magdy H. Kotb 207
(metastases in 45% if pT2 and 89% if pT4 In our present study, patients with transitional
tumors) than in patients without cystectomy cell carcinoma had a statistically significant
(36% of pT2 and 79% of pT4 tumors) implying (p=0.036) higher incidence of bone metastasis
that metastasis often occurs before the time of on presentation than those with squamous cell.
diagnosis . On the other hand, Zaghloul’s study found that
squamous cell carcinoma had a 5-year cumula-
A retrospective study of 145 muscle invasive
tive distant metastasis rate of 15% (CI 13-17%),
bladder patients reported distant metastasis as
which was lower than the incidence in transi-
the only site of failure in 23.4% of the cases
tional cell carcinoma being 39% (CI 33-45%).
and associated with local recurrence in 13.1%
This difference was statistically significant in
of the cases .
univariate analysis. Upon multivariate analysis,
Our results showed a 29.6±4.9% 3-year he found that this effect was not an independent
cumulative incidence of distant failure, 63% of prognostic factor and it was dependent upon
which were detected in bone, thus being in other factors (tumor stage, grade and nodal
concordance with the study of distant metastasis involvement) .
in bilharzial bladder cancer which reported a
rate of 23%. 70% of these distant failures were Looking at the pathological prognostic fac-
detected in bone, mainly the pelvis and the tors affecting the incidence of bone metastasis
spine . after treatment, we identified the higher clinical
stage (p=0.046), higher pathological tumor stage
Bone metastases were radiographically di- (p=0.006) and more than one pathological risk
agnosed in 24/86 (28%) patients via bone scan factor (p=0.021) as having statistically signifi-
and/or bone surveys. Again, the spine and pelvic cant higher incidence of bone metastasis during
bones were the most common sites involved. the course of follow-up of our cases. This is
Bony metastases ranked second in frequency similar to the findings of Zaghloulís retrospec-
after lung metastases . The pelvic bone and tive study that identified 3 risk factors on mul-
spine represented the more frequent sites affect- tivariate analysis of 357 patients who were
ed in the skeleton in the present study. They treated with cystectomy alone or in addition to
were 32% and 20% respectively (Table 2). preoperative or postoperative irradiation. The
Similar results were reported by Sengelov et 3 independent risk factors that significantly
al. . affected distant metastasis were pathologic stage
The incidence of bone metastases upon first (p=0.04), histopathologic grade (p=0.05) and
presentation was found to be 14.5% of our study pelvic nodal involvement (p=0.005). The author
group, that was much higher than the data pub- grouped the patients into 3 risk groups; those
lished. The only retrospective study looking at who had no risk factors (G<3, p<4a and LN
patients with recurrent or metastatic urothelial negative at surgery), those with only one risk
cancer detected a 4% incidence of metastases factor and those with 2 or more. The 5-year
diagnosed concurrently with the primary tu- distant metastases rates were 11±3%, 29±4%
mor.Whereas bone metastases was the most and 51±5% for the 3 risk groups, respectively
frequent site of distant metastases occurring in . Applying the same grouping to the patients
35% of the cases post treatment, most cases in the present study revealed also a statistically
being in the spine and pelvis . However it significant difference between the three groups
is worth mentioning that most of the patients (p=0.021).
in the study of Sengelov et al., were in the
superficial bladder cancer category with no Conclusion:
muscle invasion. Pelvic bones followed by the Bone metastasis as a component of distant
spine were also our most frequent anatomical failure is rather a substantial element. Having
sites involved (52% of cases). An important the same importance is the accurate and system-
finding in our study identified the depth of atic detection of bone metastasis in the newly
muscle invasion as having a statistically signif- diagnosed cases, where a bone scan should be
icant effect on the rate of bone metastasis on performed to identify such metastasis in cases
presentation (p=0.000), which we can only who prove to have muscle invasion on TUR, or
compare to data from other studies yet related advanced clinical stage, to offer such patients
to post treatment distant failure . the proper management and avoid undue radical
208 Bone Metastases in Muscle-Invasive Bladder Cancer
surgical procedures with all its negative effects 5- Matthews PN, Madden M, Bidgood KA, Fisher C.
on the quality of life of the patient plus an The clinicopathological features of metastatic super-
ficial papillary bladder cancer. J Urol. 1984, 132 (5):
unjustified financial burden. Isotopic bone scan 904-6.
is still an excellent tool to detect bone metastasis
during follow up of bladder cancer patients after 6- Friedell GH, Mcauley RL. Untreated bladder cancer:
31 autopsy cases. J Urol. 1978, 100: 293-6.
7- Wallmeroth A, Wagner U, Moch H, Gasser TC, Sauter
REFERENCES G, Mibatsch MJ. Pattern of metastasis in muscle-
invasive bladder cancer (pT2-4): An autopsy study
1- Prout GR Jr, Griffin PP, Shipley WU. Bladder carci- on 367 patients. Urol Int. 1999, 62 (2): 69-75.
noma as a systemic disease. Cancer. 1979, 43 (6):
2532-9. 8- Honma I, Masumori N, Sato E, Takayanaqi A, Taka-
hashi A. Local recurrence after radical cystectomy
2- Fleming ID, Cooper JS, Henson DE. TNM staging of for invasive bladder cancer: An analysis of predictive
the urinary bladder. AJCC Cancer Staging Manual, factors. Urology. 2004, 64 (4): 744-8.
5th ed. Philadelphia: Lippincot-Raven. 1997, pp. 241-
3. 9- Zaghloul MS. Distant metastasis from bilharzial
bladder cancer. Cancer. 1996, 77: 743-9.
3- Kaplan EL, Meier P. Nonparametric estimation from
incomplete observations. J Am Stat Assoc. 1958, 53: 10- Stanford MG, Arturo AF, Rogelio CN, John EM.
457-61. Metatatic transitional cell carcinoma from the bladder:
Radiographic manifestions. AJR. 1979, 132: 419-25.
4- Peto R, Pike MS, Armitage P, Breslow NW, Cox DR,
Howard SV, et al. Design and analysis of randomized 11- Sengelov L, Kamby C, Masse H. Pattern of metastases
clinical trials requiring prolonged observation of each in relation to characteristics of primary tumor and
patient. II. Analysis and examples. Br J Cancer. 1977, treatment in patients with disseminated urothelial
35: 1-39. carcinoma. J Urol. 1996, 155: 111-4.