© 1993 International Medical Society of Paraplegia
Paraplegia 31 (1993) 105-110
28 K in squamous metaplasia of the bladder in patients with spinal cord
injury
J Z
C
1
J
Montgomerie MB ChB FRACP,l H
Bennett MD,4 D G Schick BSI
J
Holshuh DVM? A
J
Keyser PhD?
Department of Infectious Diseases, 2Comparative Medical and Veterinary Services,
4Urology Service and Departments of Medicine and Surgery, Rancho Los Amigos
Medical Center, 7601 E Imperial Highway, Downey, California 90242; 3Departments of
Chemical Pathology and Surgery, University of Southern California School of Medicine,
Los Angeles, California
90033, USA.
Antibody to 28 K was used to examine sections of bladder biopsies obtained by
cystoscopy from 14 patients with spinal cord injury (SCI). Most of the biopsies
were obtained from patients with indwelling catheters during the investigation
for possible malignancy. Sections of bladder were stained by the streptavidin
procedure.
The 28 K in the normal transitional epithelium of the bladder was in the
superficial cells (umbrella cells). All the biopsies from patients with indwelling
urethral catheters showed areas of squamous metaplasia usually associated with
evidence of chronic inflammation. Cystitis cystica glandularis was also seen in
one patient. Staining was most marked in the areas of squamous metaplasia with
intracellular granular staining. The basal layers were not well stained. With
marked squamous metaplasia, there was a superficial hyperkeratotic layer that
stained variably and often did not stain at all. Staining was less marked in areas
of hyperplasia, regenerating urothelium, and cystitis cystica glandularis.
These findings raise the possibility that the presence of 28 K glycoprotein in
the tissues or released into the urine may be used as an indicator of squamous
metaplasia and chronic inflammation of the bladder.
Keywords:
spinal cord injury; urinary bladder metaplasia; urinary bladder
malignancy; 28 K glycoproteins.
Introduction
Methods
In a previous study of human urothelium,
we identified a component of the umbrella
cells of normal human urinary bladder
epithelium that was also present in squam­
ous metaplastic cells in the bladder. 1 West­
ern immunoblotting and tissue immuno­
chemistry identified a 28 Kd molecule which
was present within the surface cells of the
urothelium (umbrella cells) of the human
bladder. Antiserum to 28 Kd also stained
areas of squamous metaplasia in the blad­
der. In the present study we examined the
distribution of staining in biopsies of the
bladder of patients with longstanding spinal
cord injury.
Specimens were obtained from the patho­
logy laboratory at Rancho Los Amigos
Medical Center (RLAMC). These were
routinely processed biopsies of the bladder
from 1 4 patients with spinal cord injury who
had undergone cystoscopy. The main indic­
ation for cystoscopy and biopsy had been to
examine the bladder for evidence of neopla­
sia in persons who had used long term
indwelling catheterization. Three patients
with SCI who had developed hematuria and
did not have indwelling catheters were also
included. Biopsies had been obtained from
any abnormal areas. In addition, biopsies
were taken from the right and left bladder
106
Montgomerie et al
walls, the dome of the bladder, and the
trigone in most patients. All biopsies had
been fixed in formalin and embedded in
paraffin blocks.
The methods for the isolation of 28 K
bladder surface component and the produc­
tion of antibody were described previously. 1
Briefly, hyperimmune antiserum, obtained
from rabbits immunized with partially puri­
fied scrapings from the luminal surface of
human urinary bladders, was absorbed with
IgG-agarose and human type 0 blood. This
resulted in specificity to the 28 K antigen.
Absorbed antibody in TRIS buffered saline
pH 7.5 was applied to the sections for 2
hours at 37°C. Slides were washed and
incubated with biotinylated antibody against
rabbit IgG (link antibody) for 30 minutes at
37°C. Slides were then washed, incubated
with streptavidin conjugated horseradish
peroxidase for 30 minutes, and washed
again. The substrate, 3-amino-9-ethylcarba­
zole (AEC), was used to visualize bound
antibody and the slides counterstained with
Mayer's hematoxylin. Dilutions of the ab­
sorbed antibody were selected after titration
of the antibody using the index case of
squamous cell carcinoma (RIV). Negative
controls included tissue sections with sub­
stitution of the primary antibody with
normal rabbit serum.
Areas of squamous metaplasia, keratin­
ization, cystitis cystica glandularis, hyper­
plasia, epithelial regeneration and normal
urothelium were identified in each biopsy.
For the purpose of this study, squamous
metaplasia was defined as urothelium in
which the first few outer layers of the
urothelial cells and nuclei become increas­
ingly flattened. The layers of cells from
basal cells to surface cells are nucleated,
partially nucleated and finally non nucle­
ated. The degree of staining with the 28 K
antibody was recorded in each of these areas
and graded 0 to + + +. The presence of
inflammation was also recorded.
Results
The results of the study are shown in Table I.
Acute and chronic inflammatory cells
(including neutrophils, lymphocytes, occa­
sional mast cells and monocytes) with fibro-
Paraplegia 31 (1993) 105-110
sis were observed in almost all of the
biopsies. Squamous metaplasia of the blad­
der was found in one or more biopsies of all
patients that had had catheters in place
(range 9-36 years) and stained + to + + +
in all instances. The staining of squamous
metaplasia (Fig 1) was usually throughout
the thickness of the epithelium but often the
basal cells did not stain. Superficial keratin
stained variably and often remained un­
stained (Fig 2). Cystitis cystica glandularis
was present in multiple biopsies of the
bladder of one patient with squamous meta­
plasia being present only at the urethra.
Areas of cystitis cystica glandularis did not
stain.
One patient showed a significant degree
of epithelial dysplasia, basal hyperplasia and
disorderly maturation of the urothelium.
This included mitotic figures in occasional
midzonal cells displaying early keratiniza­
tion as well as lobular downgrowth of the
proliferating epithelium. There was signific­
ant anisokaryosis but no evidence of pene­
tration of the urothelial basement mem­
brane or other features needed to confirm
carcinoma-in-situ. Two malignancies stu­
died offered contrasting patterns. The trans­
itional cell carcinoma (TCC) grade 1 was
non reactive, while a case of squamous cell
carcinoma was intensely stained, the dis­
tribution of the 28 K antigen appearing to be
directly related to the degree to which areas
within the malignancy were differentiated.
One of three patients cystoscoped for
hematuria but without indwelling urethral
catheter had squamous metaplasia - one
other patient had TCC grade 1.
Areas of hyperplastic epithelium and re­
generating urothelium stained poorly.
Hyperplastic epithelium, while generally
staining poorly, showed immunoreactivity
in the cells closer to the basal layer. Areas of
normal urothelium were found in one or
more biopsies of 5 patients. The staining of
these areas was comparable to that seen in a
normal subject with staining localized to the
luminal side of the cell (unpublished stu­
dies).
Discussion
In this study we found squamous metaplasia
Duration of
catheter (yrs)
Sex/age
""
Reaction w!Antibody to 28 K
SM
NEp
Bladder site
Other
Other
...
...
Patients with indweIling catheters
9
M
55
++
28
17
11
13
M
F
F
F
M
36
34
33
F
M
F
51
33
36
54
41
56
49
71
+
0
0
C cyst
L
C cyst + reg ep
Dome
waIl
Trigone
......
0
v.
I
......
......
0
Dome
R waIl
L
Bladder calculi
waIl
Dome
++
+
+
++
++
++
++
++
Dysplasia
Trigone
Dysplasia
R waIl
Dysplasia
L
Dysplasia
Dome
Bladder calculi
waIl
I/)
..t:l
:;;:
�
Trigone
+
�
R wall
L
\:)
:;::
'"
wall
Dome
+
+
++
++
Urethra
Urethra fistula
Trigone
suprapubic catheter
L
wall
R wall
+++
+++
Not specified
+
+
+
Trigone
+
++
++
+
Bladder calculi
�
R wall
++
+++
+
++
++
++
+
Bladder calculi
R waIl
C cyst
+
+
+
......
'0
'0
�
Urethra
+
+
;;-
"E-
Not specified
L
wall
R wall
+
+
Dome
Trigone
+
Hyperplasia
R wall
L
wall
Posterior wall
�
'"
.§
is'
'"
;:;5;;.
'"
<::l"
is'
i::l..
�
....
5-
1/)
Q
......
0
-.I
o
00
I-'
Table I
(cant)
Duration of
catheter ( yrs)
Sex/age
�
Reaction w/Antibody to 28 K
SM
NEp
Other
Bladder site
�
Other
aQ
<::>
3
'"
....
Patients with indwelling catheters
21
F
65
24
F
45
10
M
54
++
SCC
+
++
+
+
Not specified
Squamous cell carcinoma
R wall
L
(i;.
�
12..
wall
Dome
Posterior wall
+
+
Trigone
R wall
+
Hyperplasia
Dome
Patients without catheters
M
34
0
TCC
++
++
38
F
44
17 years quadriplegic (C5)
R wall
Hematuria
L
++
M
Trigone
reg ep
+++
wall
Posterior wall
Not specified
22 years paraplegic ( T7)
Not specified
16 years paraplegic
Bladder calculi
+
reg ep
(L2)
Hematuria
Alkaline encrusting cystitis
Corynebacterium sp
SM
=
squamous metaplasia
NEp
=
normal epithelium
C cyst
=
cystitis cystica glandularis
TCC
=
transitional cell carcinoma
SCC
=
squamous cell carcinoma
reg ep
=
regenerating epithelium
=
changes not present in biopsy
=
degree of reaction with antibody
0- + + +
�
....
{5
�
is·
w
...
�
f-'
'-D
'-D
�
o
Vl
I
f-'
.....
f-'
o
Paraplegia 31
Squamous metaplasia in the bladder in SCI
(1993) 105-110
Figure 1 Squamous metaplasia of the human
urinary bladder showing reaction with antibody
to 28 K.
of the bladder in all patients with indwelling
catheters in whom the catheter had been
present for 9-36 years. One patient had
cystitis cystica glandularis in multiple biop-
109
sies with squamous metaplasia only in the
urethra.
This incidence of squamous metaplasia in
these patients with spinal cord injury is
higher than previous studies. Broeker et al2
studied bladder biopsies of 50 patients with
indwelling catheters for > 10 years, and 31
patients with external urinary catheter
drainage for> 1 5 years. Squamous metapla­
sia occurred in 11 patients (28%) with
indwelling catheters and 4 patients (8%)
with external catheter drainage. Squamous
metaplasia may occur with any form of
bladder management. Polsky3 et al used
cytology and bladder biopsy from 36 pa­
tients to demonstrate an 86% incidence of
squamous metaplasia of the bladder among
patients with chronic infection, and a 64%
incidence in patients with bladder diversion.
The reason for the higher incidence in the
present study is not clear. Biopsies obtained
from multiple sites may have been import­
ant, although in most patients squamous
metaplasia was seen in all biopsies.
Squamous metaplasia of the urinary blad­
der has been associated with chronic inflam­
matory processes, especially those seen in
conjunction with longstanding indwelling
catheters, infection and lithiasis. Squamous
metaplasia was found in the biopsy of one of
the patients without a catheter. This patient
also had bladder calculi. We did not study
patients who had been catheterized for <9
years. The presence of squamous metaplasia
may be suggestive of a variety of potentially
Figure 2 Squamous metaplasia of the bladder
stained with antibody to 28 K showing marked
staining of the midzone, less staining of the
basal layer and little or no staining of the
superficial keratin.
110
Montgomerie et al
serious problems that may develop within
the bladder. The patient with significant
urothelial dysplasia was demonstrating early
development of a staining pattern seen in
squamous cell carcinoma. There is con­
tinuing debate as to whether the condition
may be considered precancerous. Many
factor" are apparently capable of stimulating
urothclial squamous metaplasia including
potent carcinogens such as nitrosamines
(which may be produced in the urine of SCI
patients), physical agents such as catheters,
or even the installation of substances as
benign as sterile water. 4 The progression to
neoplasia appears to depend on the nature
of the agent involved and the duration of the
exposure.
In previous studies the distribution of the
Paraplegia 31 (1993) 105-110
stammg of the 28 K suggested that it was
being produced in the more superficial
layers of the urothelium. Staining was most
marked in the umbrella cells and its pres­
ence suggested that it may be being se­
creted.1 The absence of 28 K in cystitis
cystica and cystitis glandularis might suggest
that this substance is not part of 'mucin' .
28 K may be a useful marker for squam­
ous metaplasia. The consequences of
squamous metaplasia are not well under­
stood. This change may represent only a
reaction to foreign bodies and other irrit­
ants, but there is also concern that the
change is precancerous. This marker has
potential as a noninvasive means of invest­
igating squamous metaplasia.
References
1 Montgomerie JZ, Keyser AJ, Holshuh HJ. Schick DG (1992) 28 K antigen in the urothelium. JUral 147:
1388-1390.
2 Broecker BH, Klein FA. Hackler RH (1981) Cancer of the bladder in spinal cord injury patients. JUral
139: 428-432.
3 Polsky MS, Weber CH Jr, Williams JE III, Nikolewski RF, Barr MT, Ball TP Jr (1976) Chronically infected
and postdiversionary bladders. Cytologic and histopathologic study. Urology 7(5): 531.
4 Murphy WM, Blatnik AF, Shelton TB, Soloway MS (1986) Carcinogenesis in mammalian urothelium:
Changes induced by non-carcinogenic substances and chronic indwelling catheters. JUral 135: 840.