proven
taken
london
counted
metastasis
fh
lesions
option
death
prostatectomy
appropriate
rate
cutoff
recurrence
recognised
eastham5
outcome
valente
show
eschenbach
dissection
cancerous
final
1983
panel
already
knowledge
112
registry
nicemedia
bergera
232
radical
references
stained
quantitatively
biopsies
sampled
melia
freeman
both
rev
breference
hanks
regular
jw
41c
urologic
mortality
discriminating
disease
androgen
genetics
age
cure
748
brandt
results
741
validate
zagars
tmprss2
kudahetti1
kinukawa
necessary
extent
moss
four
investigated
gerald
individual
tonsil
erg
routine
laboratories
elucidated
multivariate
microarray
pathological
t1
studies
4department
fraction
widespread
utility
2133
college
provide
124
identified
cardinal
dearnaley
shows
dublino
685
rearrangement
351
series
conservative
toda
complaining
fearn
md
kattan7
original
increase
aggressiveness
doi
embedded
published
months
widely
citric
semi
khor
ra
past
specimen
reuter2
biological
gleason47
under
officers
0003
earlier
bartholomew
fk
263
accepted
medical
records
magnification
purely
without
bettencourt
kamoto
follow
practice
order
basal
ae
growing
ncri
variables
1194
assessment
multiples
centre
org
limitations
berney
148
scoring
bone
2009
1990
overnight
compared
relationship
did
garcia
method
any
more
patterns
class
thousands
measurement
dichotomous
tively
iii
left
seen
only
abel
recorded
jama
119
fulfilled
screen
0012
immunostained
cf
numerous
associated
file
fontana
almost
value
mr
biopsy
264
situ
some
549
1361
feasibility
basis
symptomatically
improve
6king
t3
detected
dm
fusion
levels
resection
intense
mla1
univariate
specific
pc
shipley
term
parkinson
aggressive
view
powerful
manner
186
guide
troncoso
o10
choice
undetectable
staining
recent
predictor
followed
ojea
luukkaala
cohort
hospital
marker
dw2
give
figure
node
being
scanned
december
http
1999
pdf
may
different
feneley
blocks
ag
venkatesan
int
number
bartsandthelondon
relevance
population
test
produce
cellularity
therefore
dations
asymptomatic
carried
also
tissue
meistrich
mostofi
denominator
cowen
sloan
phases
mcleod
managing
undertaken
include
papers
stronger
ml
comparisons
pain
th
unknownc
assessed
0920
time
2140
human
complex
pa
cell
largest
rs
has
proliferation
2007
receptor
103
kettering
ablation
useful
urology
retrospective
moul
rodriguez
confidence
therapy
distant
x300
rep
antigen
corre
420
established
perform
range
moved
effect
tumors
organs
unscientific
localised
positivity
fisher3
immunobiochemical
materials
verified
kwast
breast
supported
there
uncertainties
tmpss
kinoshita
york
adjunct
usefulness
interest
frequently
biomarker
men
waiting
clinical
index
laurila
additional
p99
radiotherapy
kingdom
numerator
gerald2
fisher
cervino
findings
cooper
artefact
interval
performed
whereas
antibody
center
1990s
incubation
indicated
power
important
samples
all
british
total
ambroisine3
development
add
307
surgery
factor
turp
control
including
sciences
552
299
carpinteria
mw
cases
grade
reid
methods
identifies
892
details
significant
characterization
hammond
predict
easily
comes
2095
atlantic
micro
monoclonal
oh
showing
staff
amounts
protein
symptoms
every
intermediate
national
135
145
care
stenning
sequences
measured
study
standards
cooper8
defined
hormonal
1148
pretreatment
jp
wolkov
together
ib
histologically
rationale
normal
200
regulating
spore
monitoring
curative
departments
hypothesis
histopathological
independently
tapg
collection
wohlenberg
large
attempt
cc
grand
treatment
1038
long
1064
vs
measure
initial
category
tion
p53
albertsen
baseline
cd
2005
orchid
mitotic
october
nottingham
reported
dako
indicator
barros
whether
moeller6
was
assess
countries
ff
identification
half
wolfson
progression
freemasons
biologic
oliver
available
block
however
patients
incorrect
predicts
array
decision
year
909
paris
preventive
laboratory
prised
among
450
grignon
heydon
roach
porpiglia
still
pg
will
future
cause
marsden
nice
auvinen
160
popular
410
eh
189
107
then
markers
thermal
independent
urethral
thus
differentiate
similar
liverpool
bely
lemke
bott
given
ensure
indicators
fletcher
symptom
serum
parts
clin
4124
pathologists
profiles
biomarkers
140
www
diseases
entry
programmes
problems
240
115
must
needle
characteristics
sections
single
bcl
clinicians
clearly
1798
colombo
derivation
limited
dentistry
within
estima
wu
enhanced
alteration
select
because
unproven
wg
maattanen
spite
done
management
hanley
attard
vary
survival
used
oncogene
when
predicting
gj
incidence
incorporating
grouped
base
berian
535
disadvantaged
institute
ca
312
mary
adjacent
beneficial
tl
0005
promising
point
dysfunction
williams
small
bju
spinal
limits
haematoxylin
low
gene
urol
developed
invasive
suffer
tonita
paraffin
gerlach
pelvic
immuno
pathol
ray
review
sanchez
values
finnish
informative
mibi
bauer
discriminatory
cleveland
693
modern
rosell
limbs
unlike
foster4
169
majority
simply
mri
steenbrugge
dealing
existing
protocol
mouse
172
pathologic
conservatively
mp
chinyama
subgroups
ascore
requests
overall
2003
benavente
medicine
shown
excluded
schwab
me
histology
parker
interesting
flad
2002
1068
giving
eur
polo
schlueter
importance
identify
treated
local
0001
eligible
1991
opas
dj
serious
better
fracture
germinal
foster
viable
diagnosed
schroder
1656
fj
desilvio
continues
table
359
1001
vast
142
cg58fullguideline
gleason
counting
kirby
deviation
could
expression
rates
logical
research
ja
application
425
decisions
would
assembly
440
academic
bjc
sk
approach
mcloughlin
absence
rather
158
sexual
correlate
1368
isola
van
key
area
die
preliminary
natural
undergo
who
failure
factors
calculation
assembled
many
deprivation
jm
planned
ki67
139
gleasono7
uro
quick
patient
presented
po0
variety
account
alone
little
beyond
psa
score
von
suggested
canada
possible
length
650
1154
szende
molecular
our
its
compression
group
microarrayed
utilises
293
dp
great
than
233
2000
cells
following
data
after
strongest
quantitative
difference
new
evaluation
barts
reviewed
states
respectively
needed
183
407
kovacs
variable
actas
253
sanz
049
predictors
prognosis
trend
intent
extensive
material
mathematics
groups
detailed
most
offering
had
information
prostate
saleem
cs
again
ct
much
337
fatal
watchful
lawton
assessing
means
crush
strong
430
into
advanced
making
max
khoo
cuzick3
1993
percentage
analysis
fixed
paper
received
oncol
gerdes
epidemiology
influencing
choose
immunohistochemically
cohen
middle
eastham
applied
671
function
advantages
t2
techniques
strangely
non
immunohistochemical
needs
missing
clinico
methodologies
ability
likely
102
undoubtedly
positive
politically
thames
com
visakorpi
hybridisation
2101
why
know
dquantitative
mosteiro
behaviour
doubling
other
united
metastatic
significance
university
ungradeable
another
125
2008
clinically
ge
lopez
soft
morbidity
szegedi
wildly
localized
inadequate
well
potential
cellular
none
156
greatest
early
translocation
zudaire
examined
various
derived
possibly
presently
trained
requested
mismatched
managed
100
their
included
tapia
171
far
required
distribution
cancer
but
over
eligibility
criterion
inoue
receive
parameters
especially
evidence
easy
europe
caused
based
rolle
wl
assigned
stahmer
martikainen
ogawa
made
estimation
such
heterogeneity
degree
current
computer
ti
ment
5department
routinely
individually
scientific
showed
represent
dying
assessments
reason
production
kloth
trends
use
despite
ny
queen
arise
appeal
cancers
rises
essential
stages
consuming
angiogenesis
mean
tumours
ambroisine
surveillance
306
search
166
major
absent
acknowledgements
ki
trans
reuter
tumour
153
cores
activity
res
presence
type
histochemistry
specimens
reactive
sandler
duplication
recommen
slides
marked
116
confirmed
few
price
001
calculated
selected
796
fine
mildh
6604951
888
treatments
bjcancer
awareness
p63
moller
known
scardino5
immunochemistry
correspondence
660
robles
biochemical
plus
scorea
centres
need
does
process
hr
young
cylinders
difficult
statistics
although
2006
systems
drastically
predictions
stein
js
unless
malignancy
mj
ia
890
pathology
negative
867
during
equally
recently
revised
eeles
unnecessary
relative
1186
forms
precluded
284
808
they
diagnosis
memorial
validation
fulton
incontinence
aa
clinic
cycle
transatlantic
options
predicted
lovasz
genitourinary
je
radiation
mail
dilution
date
segawa
tested
hd
system
model
appears
risk
retrieved
cuzick
similarly
benefit
clark
outcomes
case
deaths
effectively
fluorescence
shiraishi
1987
399
gopalan2
continuous
esp
161
charity
variability
active
urgently
predictive
db
currently
proteins
detection
added
remains
jhavar
provided
japanese
relatively
der
2department
nuclear
3cancer
sesterhenn
become
muston
891
history
discovery
careful
several
pollack
hazard
comparable
might
maximal
stable
treat
corresponding
nomogram
discussion
immunostaining
moderate
scardino
second
174
rebiopsy
sponding
188
conducted
three
histopathology
settings
two
dominguez
count
each
technique
203
progress
pt
indications
rocena
koch
unscientifically
873
doubt
eosin
testing
gonzalez
one
hardly
estimated
547
usa
informa
cthese
164
january
fasolis
889
mitosis
average
8institute
carcinoma
very
yamada
status
prognostic
ref
gk
locally
journal
ratio
torda
formalin
pd
reserved
urinary
ets
keywords
ci
due
argued
475
138
staging
adds
sr
evaluated
areas
described
instituted
royal
expressed
yet
303
between
kattan
jj
insignificant
though
347
lymph
macroscopically
trial
proliferative
flohr
td
verify
public
pattern
fully
diez
ac
using
unspecifiedc
bono
since
advantage
1996
rearrangements
ingly
scan
nhs
dna
appeared
mc
313
ly
0007
2004
unnecessarily
rights
formalised
subsets
level
supplemented
centrally
4118
173
uk
bellina
guidelines
ck5
scores
indicates
challenge
418
2333
ruutu
glands
malignant
further
contrasting
dg
centralised
suitability
unknown
oncology
hb
diagnostic
found
excellence
grading
yoshida
clear
sutton
screening
conserva
zd
alonso
school
resting
maintained
hospitals
prostatic
utilised
translated
strategy
332
7department
intervention
analysed
common
health
maximum
romics
oomens
out
tammela
core
stage
thompson
simplest
mcdavid
893
3959
duchrow
acid
short
apoptosis
hm
raises
noted
around
binding
not
years
foundation
remained
lee
scientists
standard
community
before
high
demonstrated
first
candidate
etv1