64 96
Priority
Focus
–
Editorial
Referring
to
the
article
published
on
pp.
54–63
of
this
issue
Inflammation: A
Significant Contributor
to Upper-tract Urothelial
Carcinoma
Prognosis?
Brant
Inman
* ,Kae
Jack
Tay
Division
of Urology, Duke University Medical
Center, Durham, NC, USA
Upper-tract
urothelial
carcinoma
(UTUC)
is
a
rare
disease
that
tends
to present
late and at a muscle-invasive
stage
in
comparison
to
its
lower-tract
counterpart
[1].
To
com-
pound matters,
patients with
UTUC
are
usually
smokers
and older, and havemultiple comorbidities, often
including
significant
renal
impairment. While
immediate
postoper-
ative
instillation and neoadjuvant
systemic
chemotherapy
have become
the standard of care
for
lower-tract urothelial
cancers,
these
remain
difficult
to
apply
in
patients with
UTUC
[2].
Furthermore,
the
development
of
prognostic
tools
that
could
help
in
selecting
patients
for
these
therapies
has
been
challenging.
Thus,
it
is
timely
that
Tanaka
and
colleagues
[3],
through
a multicenter
collabo-
rative
effort,
propose
a
new marker
score
as
a
prognostic
tool
for UTUC.
Virchow
first noted
the
connection between
inflamma-
tion
and
cancer
when
he
observed
the
presence
of
a
‘‘lymphoreticular
infiltrate’’
in
cancers
arising
at
sites
of
chronic
inflammation
[4]. The concept of
inflammation as a
critical
component
of
tumor
progression
has
since
been
expanded
with
the
recognition
of
inflammatory
cells
as
orchestrators
of
the
tumor microenvironment
[5].
In
an
attempt
to
elucidate
the
role
of
inflammation
in
UTUC
outcomes,
the
authors
have
laudably
built
on
previous
work
identifying
C-reactive
protein
(CRP)
as
a
prognostic
marker,
and
combine
those
findings with
the
neutrophil/
lymphocyte
ratio
(NLR)
and
plasma
fibrinogen
levels,
markers
identified
in
European
patients
[6,7].
Although
this
is
a
step
in
the
right direction,
the
scientific
approach
used
in
the
study
warrants
discussion
and
the
results
should
interpreted
in
context.
1.
Model
development
The proposed
cutoff points
for NLR, plasma
fibrinogen,
and
CRP
were
developed
from
previous
retrospective
studies
and
their combination
into a single tool
is a
logical next step.
However,
the
population
in
which
CRP
was
identified
originated
from
the
same
institutions
as
the
current
population
being
tested.
The
high
likelihood
of
overlap
between
the
development
(prior manuscript)
and
valida-
tion
(current
manuscript)
populations
raises
concern
regarding
significant
confirmation
bias,
whereby
the
previous
statistically
significant
results
are merely
being
replicated
in
the
same
population.
Generalizability
to
an
independent non-Asian cohort
is necessary
for validation of
the model.
Second,
the
cutoff
values
for
the markers were
adopted
on
the
basis
of
findings
in
their
individual
discovery
cohorts.
These
cutoffs
may
be
different
when
all
three
markers
are
combined.
Continuous
rather
than
binary
assessment
of
the
variables might
have
facilitated
the
construction of
a model more
informative
in determin-
ing
a
specific
pattern
predicting
recurrence
and
death.
Indeed, haphazard dichotomization of continuous variables
is undesirable because
information and statistical power are
usually
lost
during
this
process.
A
risk-score
logistic
regression
approach
would
probably
have
produced
a
different
and
possibly more
accurate
predictor.
Third,
the
authors
reported
a
high
correlation
between
plasma
fibrinogen
and
CRP,
and
moderate
correlations
among
NLR,
plasma
fibrinogen,
and
CRP.
This
is
not
surprising
because
all
three
markers
are
part
of
the
inflammatory
cascade. The exact
relationships merit
further
investigation,
E U R O P E A N U R O L O G Y F O C U S 1 ( 2 0 1 5 ) 6 4 – 6 5available
at
www.sciencedirect.comjournal
homepage:
www.europeanurology.com* Corresponding
author. Division
of Urology, Duke University Medical
Center, DUMC
2812, Durham, NC
27710, USA.
Tel.
+1
919
6818760;
Fax:
+1
919
6845220.
address:
brant.inman@duke.edu(B.
Inman).
http://dx.doi.org/10.1016/j.euf.2015.02.0042405-4569/
#
2015
European
Association
of Urology.
Published
by
Elsevier
B.V.
All
rights
reserved.