Surgical adjuvant therapy of large bowel carcinoma an evaluation of levamisole and the combination of levamisole and fluorouracil


Laurie, J.A.; Moertel, C.G.; Fleming, T.R.; Wieand, H.S.; Leigh, J.E.; Rubin, J.; McCormack, G.W.; Gerstner, J.B.; Krook J.E.; Malliard, J.

Journal of Clinical Oncology 7(10): 1447-1456

1989


A total of 401 eligible patients with resected stages B and C colorectal carcinoma were randomly assigned to no-further therapy or to adjuvant treatment with either levamisole alone, 150 mg/d for 3 days every 2 weeks for 1 years, or levamisole plus fluorouracil (5-FU), 450 mg/m2/d intravenously (IV) for 5 days and beginning at 28 days, 450 mg/m2 weekly for 1 year. Levamisole plus 5-FU, and to a lesser extent levamisole alone, reduced cancer recurrence in comparison with no adjuvant therapy. These differences, after correction for imbalances in prognostic variables, were only suggestive for levamisole alone (P = .05) but quite significant for levamisole plus 5-FU (P = .003). Whereas both treatment regimens were associated with overall improvements in survival, these improvements reached boderline significance only for stage C patients treated with levamisole plus 5-FU (P = .03). Therapy was clinically tolerable with either regimen and severe toxicity was uncommon. These promising results have led to a large national intergroup confirmatory trial currently in progress.

Surgical
Adjuvant
Therapy
of
Large
-Bowel
Carcinoma:
An
Evaluation
of
Levamisole
and
the
Combination
of
Levamisole
and
Fluorouracil
By
John
A.
Laurie,
Charles
G.
Moertel,
Thomas
R.
Fleming,
Harry
S.
Wieand,
John
E.
Leigh,
Joseph
Rubin,
Greg
W.
McCormack,
James
B.
Gerstner,
James
E.
Krook,
James
Malliard,
Donald
I.
Twito,
Roscoe
F.
Morton,
Loren
K.
Tschetter,
and
John
F.
Barlow
for
the
North
Central
Cancer
Treatment
Group
and
the
Mayo
Clinic
A
total
of
401
eligible
patients
with
resected
stages
B
and
C
colorectal
carcinoma
were
randomly
assigned
to
no
-further
therapy
or
to
adjuvant
treatment
with
either
levamisole
alone,
150
mg/d
for
3
days
every
2
weeks
for
1
year,
or
levamisole
plus
fluorouracil
(5
-FU),
450
mg/m
2
/d
intravenously
(IV)
for
5
days
and
beginning
at
28
days,
450
mg/m
2
weekly
for
1
year.
Levamisole
plus
5
-FU,
and
to
a
lesser
extent
levami-
sole
alone,
reduced
cancer
recurrence
in
comparison
with
no
adjuvant
therapy.
These
differences,
after
correction
for
imbalances
in
prognostic
variables,
were
only
suggestive
for
levamisole
alone
(P
=
.05)
but
C
ARCINOMA
of
the
large
bowel,
the
colon,
and
the
rectum
is
one
of
the
most
common
malignant
diseases
encountered
in
Western
civi-
lization
and
one
of
the
most
frequent
causes
of
cancer
death.
Although
the
great
majority
of
these
patients
are
discovered
at
a
stage
of
disease
where
surgical
resection
with
curative
intent
is
possible,
national
end
-result
statistics
still
show
that
almost
half
of
the
patients
afflicted
with
large
-bowel
cancer
will
die
of
it.
Early
diagnosis
and
prevention
are
laudable
future
goals,
but
for
the
patient
today
and
in
the
years
immediately
ahead
the
most
realistic
hope
would
seem
to
lie
in
attempts
to
enhance
the
effectiveness
of
surgical
therapy.
Although
historically
controlled
studies
have
predictably
been
positive,
all
past
random-
ized
studies
of
colon
cancer
surgical
adjuvant
therapy
compared
with
untreated
controls
have
either
been
clearly
negative
or,
at
best,
highly
equivocal
in
results.
Notable
past
chemotherapy
failures
have
included
thiotepa,
fl
uorouracil
(5
-
FU)
by
a
variety
of
doses
and
schedules,
and
5-fluoro-2-deoxyuridine
(5-FUDR).
1
Even
when
results
of
a
selected
group
of
5
-FU
studies
involving
a
total
of
3,182
patients
were
sum-
mated
in
a
retrospective
meta
-analysis,
any
ther-
apeutic
advantage
was,
at
best,
marginal.'
This
led
the
authors
to
conclude
that
their
meta
-
analysis
"provides
only
moderate
evidence
that
adjuvant
chemotherapy
may
provide
an
overall
quite
significant
for
levamisole
plus
5
-FU
(P
=
.003).
Whereas
both
treatment
regimens
were
associated
with
overall
improvements
in
survival,
these
improve-
ments
reached
borderline
significance
only
for
stage
C
patients
treated
with
levamisole
plus
5
-FU
(P
=
.03).
Therapy
was
clinically
tolerable
with
either
regimen
and
severe
toxicity
was
uncommon.
These
promising
results
have
led
to
a
large
national
intergroup
confir-
matory
trial
currently
in
progress.
J
Clin
Onto!
7:1447-1456.
m
1989
by
American
Soci-
ety
of
Clinical
Oncology.
survival
benefit
but
strong
evidence
that
such
benefit,
if
it
indeed
exists,
would
likely
be
small:
the
effect
of
prolonged
fl
uorouracil-containing
chemotherapy
results
in
a
fi
ve-year
survival
ben-
efit
of
less
than
5%."
A
recent
study
of
the
combination
of
5
-FU,
lomustine
(methyl
CCNU),
and
vincristine
reported
a
survival
advantage
that
approached,
but
did
not
reach,
the
tradi-
tional
P
<
.05
level
of
significance.
3
These
results
have
not
been
confirmed
and
any
true
therapeu-
tic
benefit
must
be
questioned
because
three
From
the
Grand
Forks
Clinic,
Ltd,
Grand
Forks,
ND;
Mayo
Clinic,
Rochester,
MN;
Fred
Hutchinson
Cancer
Research
Center,
Seattle,
WA;
Fargo
Clinic
Community
Clinical
Oncology
Program,
Fargo,
ND;
Illinois
Oncology
Research
Association
Community
Clinical
Oncology
Pro-
gram,
Peoria,
IL;
The
Duluth
Clinic
Community
Clinical
Oncology
Program,
Duluth,
MN;
Creighton
University,
Omaha,
NE;
Billings
Clinic,
Billings,
MT;
Iowa
Oncology
Research
Association
Community
Clinical
Oncology
Pro-
gram,
Des
Moines,
IA;
Sioux
Falls
Community
Cancer
Consortium
Community
Clinical
Oncology
Program,
Sioux
Falls,
SD;
and
the
Clinical
Laboratory
of
the
Black
Hills,
Rapid
City,
SD.
Submitted
February
16,
1989;
accepted
May
26,
1989.
Supported
in
part
by
Public
Health
Service
Grants
No.
CA
-25224,
CA
-15083,
CA
-35101,
CA
-35103,
CA
-35113,
CA
-35269,
and
CA
-37417
from
the
National
Cancer
Insti-
tute,
Department
of
Health
and
Human
Services.
Address
reprint
requests
to
Charles
G.
Moertel,
MD,
Mayo
Clinic,
Rochester,
MN
55905.
CO
1989
by
American
Society
of
Clinical
Oncology.
0732-183X/89/0710-0021$03.00
Journal
of
Clinical
Oncology,
Vol
7,
No
10
(October),
1989:
pp
1447-1456
1447
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
1448
LAURIE
ET
AL
other
studies
have
failed
to
demonstrate
a
signifi-
cant
advantage
for
5
-FU
plus
methyl
CCNU
over
untreated
controls,
4
and
because
a
large
study
comparing
5-FU/methyl
CCNU
with
5
-
FU
alone
produced
essentially
identical
survival
experiences.
5
Similarly
unconvincing
have
been
a
smattering
of
immunotherapy
attempts
such
as
bacillus
Calmette-Guerin
(BCG),
3
'
6
MER-BCG,
7
and
C
parvum.
8
However,
there
have
been
two
positive
trials.
The
fi
rst
involved
portal
-vein
infu-
sion
of
5-FU,
9
a
method
that
is
currently
the
subject
of
confirmatory
studies.
The
second
was
a
small
trial
conducted
by
Verhaegen
et
al
in
which
patients
with
surgically
treated,
large
-bowel
can-
cer
were
assigned
after
surgery
to
no
-further
treatment
or
to
levamisole
therapy.
These
inves-
tigators
reported
a
striking
and
statistically
signif-
icant
survival
advantage
for
the
levamisole-
treated
group.
Levamisole
has
been
widely
used
as
an
antihel-
minthic
drug
for
over
20
years.
It
attracted
interest
for
possible
cancer
treatment
in
the
1970s
when
it
was
found
to
have
immunostimula-
tory
activity.
11,12
In
some
animal
tumor
models
it
showed
antineoplastic
effects
that
seemed
more
striking
against
metastasis.
There
was
also
im-
pressive
activity
in
some
animal
models
simulat-
ing
the
surgical
adjuvant
setting.
In
view
of
these
fi
ndings,
as
well
as
the
mild
toxicity
of
the
drug,
a
number
of
human
therapeutic
trials
were
initi-
ated,
several
of
which
have
claimed
positive
results.
We
felt
the
interesting
result
in
colorectal
cancer
reported
by
Verhaegen
et
al
merited
a
confirmatory
attempt.
In
addition,
on
an
empiric
basis,
we
elected
to
study
a
regimen
of
5
-FU
plus
levamisole.
For
the
5
-FU
regimen
in
this
combi-
nation
we
chose
a
5
-day
intensive
course
followed
by
weekly
maintenance
since
among
the
many
negative
5
-FU
colon
surgical
adjuvant
trials,
a
similar
schedule
used
by
the
Central
Oncology
Group
had
shown
a
suggestive
but
not
significant
therapeutic
advantage
for
5
-FU
(P
=
.21)."
It
was
our
judgment
that
the
inclusion
of
an
un-
treated
control
arm
was
mandatory.
To
avoid
the
data
quality
problems
produced
by
patient
re-
fusal
of
treatment
assignment,
we
also
elected
to
obtain
patient
informed
consent
prior
to
random-
ization.
We
presented
the
preliminary
results
of
this
trial
in
1986,"
and
this
is
our
fi
nal
report.
METHODS
Patient
Selection
All
patients
were
required
to
have
had
a
potentially
curative
resection
of
a
histologically
confirmed
adenocarci-
noma
of
the
colon
or
rectum.
There
could
be
neither
gross
nor
microscopic
evidence
of
residual
disease,
and
it
was
required
that
the
margins
of
resection
be
demonstrated
free
of
tumor.
It
was
also
required
that
the
resected
specimen
show
one
of
the
following
indicators
of
poor
prognosis:
invasion
of
serosa
or
pericolonic
fat,
invasion
of
adjacent
organs
by
direct
extension,
or
metastasis
to
regional
lymph
nodes.
No
patient
could
have
had
any
prior
radiation
therapy
to
the
lumbar
spine
or
pelvis
or
any
prior
5
-FU
therapy.
It
was
further
required
that
the
patient
reliably
tolerate
oral
medications
and
have
a
WBC
count
of
4000/AL
or
greater
and
a
platelet
count
of
130,000/µL
or
greater.
Patients
were
ineligible
if
they
had had
any
other
malignant
disease
within
the
previous
5
years
except
for
superficial
squamous
or
basal
cell
carci-
noma
of
the
skin
or
in
situ
carcinoma
of
the
cervix.
Patients
were
also
ineligible
if
they
had
any
evidence
of
distant
metastasis
or
any
regional
metastasis
that
could
not
be
resected
en
bloc
with
the
primary
lesion.
To
ensure
eligibility,
surgery
and
pathology
reports
were
reviewed
on
each
patient
and
slides
of
tissue
were
reviewed
by
the
North
Central
Cancer
Treatment
Group
(NCCTG)
Pathology
Committee.
Patients
were
allowed
entry
in
the
study
as
soon
during
the
postoperative
course
as
they
were
able
to
reliably
tolerate
oral
medication,
but
they
could
not
enter
any
later
than
5
weeks
after
surgery.
Stratification
and
Randomization
Procedures
A
signed
informed
consent
was
obtained
prior
to
study
entry.
Patients
were
then
stratified
according
to
stage
defined
as:
(1)
invasion
into
or
through
serosa
or
into
pericolonic
or
perirectal
fat,
or
involving
adjacent
organs
by
direct
extension;
no
lymph
node
metastasis
(stages
B2
and
B3);
(2)
metastasis
involving
one
to
four
regional
lymph
nodes
(stage
C1)
with
the
primary
tumor
not
invading
into
or
through
serosa;
(3)
metastasis
involving
one
to
four
regional
lymph
nodes
(stage
C1)
with
the
primary
tumor
extending
through
bowel
wall
and
invading
serosa,
pericolonic
or
perirectal
fat,
or
adjacent
organs;
and
(4)
metastasis
involving
more
than
four
regional
lymph
nodes
(stage
C2).
Patients
were
also
stratified
accord-
ing
to
the
following
anatomic
location
criteria:
(1)
cecum,
ascending
or
tranverse
colon;
(2)
descending,
sigmoid,
or
rectosigmoid
colon;
and
(3)
rectum
with
all
tumor
below
peritoneal
reflection.
Following
stratification,
patients
were
randomized
to
the
following
study
arms:
(1)
follow-up
without
adjuvant
therapy;
(2)
levamisole
therapy;
or
(3)
levamisole
plus
5
-FU
therapy.
For
patients
assigned
to
active
treatment,
it
was
required
that
this
treatment
be
initiated
within
6
days
of
randomization.
Protocol
Management
Within
72
hours
prior
to
randomization
it
was
required
that
the
patient
have
a
medical
history,
a
physical
examina-
tion,
a
hematology
grouping
including
hemoglobin,
leukocyte
count,
platelet
count,
and
differential
count,
a
blood
chemis-
try
panel,
and
a
chest
x-ray
if
one
was
not
obtained
preopera-
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
SURGICAL
ADJUVANT
THERAPY
OF
COLORECTAL
CANCER
1449
tively.
Management
after
randomization
is
described
below
according
to
treatment
arm.
No
adjuvant
therapy.
Patients
assigned
to
the
control
arm
were
observed
after
surgery
with
no
planned
treatment.
During
the
fi
rst
year
they
were
reevaluated
every
8
weeks,
during
the
second
year
every
4
months,
and
following
every
6
months
for
a
total
of
5
years.
These
evaluations
consisted
of
an
interim
history
and
physical
examination,
blood
counts,
a
blood
chemistry
panel,
and
a
chest
x-ray.
In
addition,
this
group
of
patients
had
either
a
proctoscopic
examination
and
colon
x-ray
(barium
enema)
or
a
colonoscopic
examination
every
6
months.
Follow-up
after
5
years
was
continued
but
without
formal
protocol
requirements
for
evaluation
proce-
dures.
Levamisole
alone.
Patients
randomized
to
this
arm
had
periodic
evaluations
at
the
same
time
as
patients
randomized
to
the
no
-adjuvant
-therapy
arm.
In
addition,
they
received
levamisole,
50
mg,
every
8
hours
over
3
days,
which
was
repeated
every
2
weeks
for
I
year.
During
the
period
of
treatment,
hematology
groups
and
blood
chemistry
groups
were
repeated
every
4
weeks.
Levamisole
plus
5
-FU.
Patients
randomized
to
this
arm
had
periodic
evaluations
at
the
same
time
as
patients
random-
ized
to
the
no
-adjuvant
-therapy
arm.
These
patients
were
given
levamisole
at
the
same
dose
and
schedule
as
described
above
for
patients
randomized
to
levamisole
alone.
In
addi-
tion,
they
received
5
-FU
at
a
dose
of
450
mg/m
2
/d
by
rapid
intravenous
(IV)
injection
for
5
consecutive
days.
Beginning
on
day
28
they
were
initiated
on
weekly
injections
of
5
-FU
at
a
dose
of
450
mg/m
2
.
5
-FU
therapy
was
also
administered
over
a
total
period
of
1
year.
Leukocyte
counts
were
obtained
before
each
weekly
dose
of
5
-FU.
If
the
patient
experienced
stomatitis,
diarrhea,
or
leukopenia
during
weekly
5
-FU
administration,
5
-FU
administration
was
deferred
until
these
side
effects
subsided.
If
these
side
effects
were
moderate
to
severe
in
intensity,
5
-FU
was
resumed
at
a
20%
reduction
in
dose.
Statistical
Methods
The
protocol
was
designed
to
assure
that
there
would
be
a
75%
probability
of
detecting
a
50%
increase
in
survival
for
a
pairwise
comparison
of
treatments
using
a
one-sided
log
-rank
test
at
significance
level
of
0.05.
Survival,
time
to
recurrence,
and
recurrence
rates
were
identified
as
primary
end
points.
Statistical
analyses
were
carried
out
using
Statistical
Application
System
(SAS)
procedures."
The
survival
curves
were
generated
using
the
Kaplan
-Meier
method.'
The
log
-
rank
statistic"
was
used
for
the
comparison
of
survival
distributions.
Since
our
study
was
originally
designed
for
all
P
values
for
treatment
comparisons
versus
control
to
be
one-
sided,
all
results
will
be
presented
in
this
manner.
We
chose
a
one-sided
test
because
we
had
a
"standard
therapy"
un-
treated
control
arm,
because
our
only
objective
and
interest
was
to
improve
therapeutic
results
in
comparison
with
con-
trol,
and
because
other
studies
with
levamisole
alone
as
surgical
adjuvant
therapy
or
in
combination
with
5
-FU
in
advanced
disease
made
it
unlikely
that
either
of
our
treat-
ment
arms
would
detract
from
recurrence
-free
time
or
survival.
We
recognize
and
respect
an
alternative
view
that
two-sided
tests
may
be
more
appropriate
under
these
circum-
stances,
and
they
certainly
represent
a
more
conservative
approach.
For
our
primary
end
results
(overall
intervals
to
progression
and
survivals)
we
have
therefore
provided
both
one-sided
and
two-sided
P
values.
The
Cox
proportional
hazards
model"
was
used
to
obtain
the
ratios
of
relapse
and
survival
rates
and
for
all
multivariate
analyses.
A
backward
regression
was
used
to
fi
nd
the
most
significant
factors,
variables
being
eliminated
based
on
the
maximum
partial
likelihood
estimate
(MLE)
statistics.
To
adjust
for
covariates
when
evaluating
treatments,
we
kept
treatment
in
the
model
and
used
the
backward
regression
for
other
covariates,
keeping
covariates
whose
MLE
statistics
satisfied
P
<
.05.
RESULTS
A
total
of
408
patients
were
entered
in
this
study.
Seven
patients
(1.7%)
were
declared
ineli-
gible,
primarily
because
of
inappropriate
stage
of
disease.
Because
eligibility
would
not
be
biased
by
treatment
assignment,
these
patients
were
excluded
from
the
analyses
to
follow.
Among
the
401
eligible
patients
there
were
four
(1%)
who
refused
their
randomized
assignment.
Because
this
decision
could
be
biased
by
treatment
assign-
ment,
these
patients
are
included
in
all
of
the
analyses
to
follow.
We
have
conducted
separate
analyses
including
the
ineligible
patients
or
ex-
cluding
the
cancelled
patients.
These
analyses
will
not
be
presented
since,
as
is
evident
by
their
small
numbers,
data
from
these
patients
had
no
substantive
influence
on
results.
Patient
Characteristics
Characteristics
of
our
401
eligible
patients
are
displayed
according
to
study
arm
assignment
in
Table
1.
Overall,
it
can
be
seen
that
there
was
essentially
an
equal
sex
distribution.
Mean
and
median
age
for
all
patients
was
61
years
with
a
range
from
22
to
83
years.
There
is
a
dispropor-
tionately
small
representation
of
tumors
primary
to
the
rectum,
no
doubt
related
to
the
fact
that
the
NCCTG
developed
an
alternative
protocol
for
primary
rectal
carcinoma
approximately
1
year
after
this
protocol
was
opened.
Two-thirds
of
all
patients
had
nodal
metastasis
(stage
C),
and
the
great
majority
of
these
had
from
one
to
four
nodes
involved
(stage
C
1).
A
total
of
only
26
patients
had
nodal
involvement
without
invasion
of
serosa
or
pericolonic
fat.
Most
patients
had
tumors
that
were
either
well
or
moderately
differ-
entiated
and
less
than
20%
had
a
highly
anaplas-
tic
histology.
As
is
evident
in
Table
1,
patient
characteristics
are
remarkably
well
distributed
between
the
three
study
arms.
The
only
imbal-
ances,
and
these
of
a
minor
degree,
are
a
slightly
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
1450
Table
1.
Patient
Characteristics
According
To
Treatment
Arm
Characteristics
Surgery
Only
Levamisole
Levomisole
(%)
Alone
(%)
Plus
5
-FU
(%)
(n
=
135)
(n
=
130)
(n
=
136)
Male
50
49
51
Age
Median
61 61 61
Range
27-81
23-83
22-82
Stage
B2
and
B3
36
35
33
Cl
(1-4
nodes,
no
serosal
invasion)
8
8
10
CI
(1-4
nodes,
serosal
invasion)
39
41
40
C2
(>
4
nodes)
17
17
17
Anatomic
site,
primary
Ascending
and
trans-
verse
47
48
.14
Descending
and
sigmoid
47
45
47
Rectum
6
6
9
Grade
of
anaplasia*
1-2
81
83
81
3-4
19
17
19
Performance
scoret
0
76
79
68
1
20
14
27
2-3
4
6
5
*Broders'
grade:
1,
well
-differentiated
to
4,
highly
undifferenti-
ated.
tEastern
Cooperative
Oncology
Group
score:
0,
fully
active
to
4,
totally
disabled.
larger
proportion
of
patients
with
nodal
involve-
ment,
with
a
rectal
primary,
and
with
impaired
performance
on
the
levamisole/5-FU
arm
when
compared
with
the
control
arm.
Therapeutic
Results
At
present,
this
study
is
very
mature.
The
median
follow-up
is
7
years
and
9
months,
and
the
minimum
follow-up
is
just
over
4
years.
No
patient
has
been
lost
to
follow-up.
Tumor
Recurrence
At
the
time
of
this
writing,
191
patients
have
experienced
recurrent
colorectal
carcinoma,
and
it
is
estimated
that
nearly
100%
of
all
anticipated
recurrences
have
been
documented.
The
esti-
mated
overall
reduction
in
recurrence
rate
(Cox
proportional
hazards
model)
for
levamisole
alone
is
27%
with
a
90%
confidence
interval
from
2%
to
45%,
and
for
levamisole
plus
5
-FU
it
is
31%
with
a
confidence
interval
from
8%
to
48%.
Plots
of
recurrence
-free
intervals
for
all
eligible
patients
are
displayed
in
Fig
1.
The
levamisole
plus
5
-FU
Recurrence
-free,
%
too
80
LAURIE
-
Lev
+
5
-FU
-
Lev
.....===•
Control
ET
AL
n
=
136
n
=130
n=
135
60
40
Log
rank
Cox
model
20
Control
vs.
lev
P
=0.04
P
=0.05
Control
vs.
lev
+
5
-FU
P
=0.02
P=0.003
0
0
2
3
4
Years
from
randomization
5
6
Fig
1.
Recurrence
-free
interval,
all
patients.
combination
shows
a
clear
advantage
over
con-
trol,
whereas
the
advantage
is
borderline
for
levamisole
alone
(respective
one-sided
P
values,
.02
and
.04;
two-sided,
.04
and
.08).
It
is
appar-
ent
that
the
combination
provides
not
only
fewer
recurrences
but
also
a
delay
for
the
recurrences
that
are
observed.
Figures
2
and
3
demonstrate
similar
relationships
between
treatment
arms
when
the
patient
populations
are
divided
into
subsets
according
to
stage.
The
advantage
for
levamisole
plus
5
-FU
is
significant
only
in
stage
C.
In
exploring
other
subsets
we
found
the
most
favorable
treatment
advantages
to
be
in
primary
tumors
of
the
ascending
and
transverse
colon
for
both
levamisole
and
levamisole
plus
5
-FU;
in
females,
for
levamisole
plus
5
-FU;
in
lesions
of
low
to
moderate
degrees
of
anaplasia,
for
5
-FU
plus
levamisole;
and
in
the
less
than
65
-year
age
group,
for
levamisole
plus
5
-FU.
However,
these
subset
analyses
must
be
interpreted
with
great
caution.
In
no
case
did
we
fi
nd
a
significant
interaction
of
treatment
with
a
covariate.
Table
2
displays
a
number
of
potential
prognos-
tic
determinants
for
recurrence.
Tumor
stage,
100
c*
80
aS
El'
60
40
CC
20
0
0
-
Lev
+
5
-FU
n=45
-Lev
n
=45
Control
n
=49
Control
vs.
lev
P=0.14
Control
vs.
lev
+
5
-FU
P
=0.10
2
3
4
Years
from
randomization
5
6
Fig
2.
Recurrence
-free
interval,
stage
B.
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
SURGICAL
ADJUVANT
THERAPY
OF
COLORECTAL
CANCER
1451
Recurrence
-free,
%
100
-Lev
+
5
-FU
n
=91
80
-Lev
n
=85
-
Control
n
=86
60
40
20
Control
vs.
lev
P
=0.06
Control
vs.
lev
+
5
-FU
P
=0.02
0
0
2
3
4
Years
from
randomization
5
6
Fig
3.
Recurrence
-free
interval,
stage
C.
primary
location,
and
grade
of
anaplasia
were
strongly
determinant
and
performance
score
was
of
borderline
significance,
whereas
age
and
sex
played
no
important
role.
Using
a
step
-wise
proportional
hazard
model
to
correct
for
the
influence
of
prognostic
variables,
levamisole
again
showed
a
borderline
advantage
over
control
with
a
one-sided
P
value
of
.05
(two-sided,
.10),
whereas
levamisole
plus
5
-FU
showed
a
very
significant
advantage
with
a
one-sided
P
value
of
Table
2.
Prognostic
Determinants
For
Recurrence
Determinant
No.
of
Patients
Percentage
Recurring
by
5
Years
Sex
Male
201
49
.19
Female
200
44
Age
1-64
yr
253
48
.21
a
65
yr
148
43
Location,
primary
Ascending
and
transverse
186
39
.0013
Descending
and
sigmoid
187
49
Rectum
28
75
Grade
of
anaplasiat
1-2
315
43
.0010
3-4
83
61
Stage
B2
and
B3
139
30
.0000
Cl
(1-4
nodes,
no
serosal
invasion)
35
32
Cl
(1-4
nodes,
serosal
invasion)
159
54
C2
(>
4
nodes)
68
69
Performance
score
0
297
44
.13
1-3
104
52
*Two-sided
log
-rank.
tThree
patients
had
an
unknown
tumor
grade.
.003
(two-sided,
.006).
In
the
multivariate
analy-
sis,
stage,
location,
grade,
and
performance
score
were
all
significant
(two-sided,
P
<
.05).
The
increased
significance
associated
with
levamisole
and
5
-FU
in
the
multivariate
analysis
appears
to
be
due
primarily
to
the
unfavorable
imbalance
in
performance
score
and,
to
a
lesser
extent,
to
the
imbalance
in
the
number
of
rectal
patients.
In
the
analyses
above,
patients
who
died
with-
out
recurrence
were
censored.
Analyses
were
also
performed
in
which
death
without
recurrence
was
considered
an
event
(ie,
progression
-free
survival).
These
analyses
produced
comparable
results
to
those
recorded
above.
Survival
At
the
time
of
this
writing,
195
patients
have
died
and
an
estimated
95%
of
the
deaths
within
5
years
of
study
entry
have
been
documented.
There
have
been
21
deaths
without
evidence
of
recurrence:
fi
ve
on
the
control
arm,
eight
on
levamisole
alone,
and
eight
on
levamisole
plus
5
-FU.
There
are
17
patients
who
have
had
known
recurrence
but
are
still
alive:
nine
on
the
control
arm,
fi
ve
on
levamisole
alone,
and
three
on
levamisole
plus
5
-FU.
These
data
make
it
likely
that
any
survival
advantage
for
levamisole
5
-FU
will
increase
in
the
future.
Plots
of
survival
for
all
eligible
patients
are
displayed
in
Fig
4.
Perhaps
because
of
dispropor-
tionate
deaths
without
recurrence
and
dispropor-
tionate
recurrences
without
deaths,
these
differ-
ences
are
not
as
striking
as
differences
in
recurrence
rates
(one-sided
P
value
for
levami-
sole
alone,
.13
and
for
5-FU/levamisole,
.22;
two-sided,
.26
and
.44).
The
estimated
overall
reduction
in
death
rate
(Cox
proportional
haz-
ards
model)
on
levamisole
alone
is
18%
with
a
100
80
a
C
I
60
-
Lev
I-
5
-FU
-
Lev
-
Control
n=136
n=130
n=135
to
4
°
Log
rank
Cox
model
20
Control
vs.
lev
P
=0.13
P
=0.12
Control
vs.
lev
+
5
-FU
P
=0.22
P
=0.09
0
2
3
4
Years
from
randomization
Fig
4.
Survival,
all
patients.
5
6
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
1452
LAURIE
ET
AL
90%
confidence
level
of
39%
to
10%,
and
for
levamisole
plus
5
-FU,
13%,
with
a
confidence
level
of
34%
to
16%.
Figure
5
shows
little
difference
in
survival
times
for
stage
B
patients.
Differences
are
more
striking
among
stage
C
patients
(Fig
6),
with
a
stronger
advantage
for
levamisole
plus
5
-FU.
No
meaningful
differences
were
found
in
exploration
of
other
subsets.
Table
3
displays
a
number
of
potential
prognos-
tic
determinants
for
survival.
As
was
true
for
recurrence,
stage
and
primary
location
and
grade
of
anaplasia
had
strong
and
independent
predic-
tive
values
for
survival,
whereas
performance
score
was
of
borderline
significance.
After
correc-
tion
for
the
influence
of
prognostic
variables,
both
levamisole
and
levamisole
plus
5
-FU
showed
only
suggestive
treatment
advantages
(one-sided
P
values,
.12
and
.09,
respectively;
two-sided,
.24
and
.18).
However,
within
the
stage
C
subset,
levamisole/5-FU
provided
an
advantage
over
control
after
correction
at
the
one-sided
P
=
.03
level
(two-sided,
P
=
.06).
As
in
recurrence,
the
increased
significance
associated
with
levamisole
and
5
-FU
seems
to
be
attributable
primarily
to
the
imbalance
in
performance
score
and
rectal
site
of
primary.
Toxicity
Toxic
reactions
according
to
treatment
arm
are
presented
in
Tables
4
and
5.
Reactions
to
levamisole
alone
were
infrequent,
consisting
pri-
marily
of
mild
nausea,
mild
diarrhea,
and
mild
leukopenia.
The
lowest
leukocyte
count
recorded
was
2,000/µL
and
the
lowest
platelet
count,
40,000/A.
One
patient
experienced
severe
der-
matitis
requiring
discontinuance
of
therapy.
A
frequent
minor
symptom
was
a
metallic
taste
in
the
mouth.
100
eo
-
Lev
+
5
-FU
n
=45
01
so
-
-Lev
n=45
:5
Control
n=49
z
40
-
20
Control
vs.
lev
P=0.43
Control
vs.
lev
+
5
-FU
P=0.26
0
0
2
3
4
5
6
Years
from
randomization
Fig
5.
Survival,
stage
B.
100
Log
rank
Control
vs.
lev
P=0.07
Control
vs.
lev
+
5
-FU
P=0.05
-
Lev
+
5
-FU
n
=91
-
Lev
n
=85
Control
n=86
Cox
model
P=0.06
P=0.03
2
3
4
Years
from
randomization
Fig
6.
Survival,
stage
C.
5
6
Toxic
reactions
to
the
combination
arm
were
consistent
with
what
might
be
anticipated
with
5
-FU
alone
with
the
exception
of
perhaps
a
bit
more
hematologic
depression
in
the
weekly
main-
tenance
phase.
Five
patients
(4%)
experienced
leukopenia
of
1,000/AL
or
less,
and
three
pa-
tients
experienced
sepsis.
There
were
no
treat-
ment
-related
deaths.
Treatment
Compliance
Treatment
compliance,
in
the
main,
was
excel-
lent.
Ninety-five
percent
of
eligible
patients
ran
-
Table
3.
Prognostic
Determinants
for
Survival
Determinant
No.
of
Eligible
Patients
Percentage
Surviving
at
5
Years
P*
Sex
Male
201
60
.84
Female
200
58
Age
1-64
yr
253
61
.19
65
yr
148
55
Location,
primary
Ascending
and
transverse
186
63
.033
Descending
and
sigmoid
187
58
Rectum
28
39
Grade
of
anaplasia
1-2
315
60
.023
3-4
83
52
Stage
B2
and
B3
139
78
.0000
C1
(1-4
nodes,
no
serosal
invasion)
35
63
Cl
(1-4
nodes,
serosal
invasion)
159
50
C2
(>
4
nodes)
68
37
Performance
score
0
297
61
.09
1-3
104
52
"Two-sided
log
-rank.
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
SURGICAL
ADJUVANT
THERAPY
OF
COLORECTAL
CANCER
1453
Table
4.
Toxicity
—Non
hematologic
Toxic
Reaction
Levamisole
Alone
(%)
(n
=
130)
Levamisole
plus
5
-FU
(%)
(n
=
133)
Nausea
18
68
Vomiting
5
23
Diarrhea
13
53
Stomatitis
2
28
Dermatitis
5
20
Alopecia
3
21
NOTE.
Cancelled
patients
excluded.
domized
to
the
levamisole-alone
arm
and
83%
of
those
randomized
to
levamisole
plus
5
-FU
were
totally
compliant,
continuing
therapy
for
1
year
or
until
required
by
protocol
to
discontinue
ther-
apy
because
of
toxicity
or
recurrence.
On
the
levamisole-alone
arm
seven
patients
withdrew
from
treatment
before
completion
in
the
absence
of
recurrence
or
severe
toxicity.
On
the
5-FU-plus-
levamisole
arm,
three
patients
refused
their
treat-
ment
assignment
and
20
patients
withdrew
from
treatment
before
completion
and
in
the
absence
of
recurrence
or
severe
toxicity.
Seven
patients
assigned
to
levamisole
alone
had
therapy
discon-
tinued
because
of
toxicity
or
other
medical
rea-
sons,
as
did
10
patients
assigned
to
levamisole
plus
5
-FU.
No
levamisole-alone
patient
had
ther-
apy
temporarily
interrupted
because
of
toxicity.
On
the
levamisole-plus-5-FU
arm
33
patients
had
a
dosage
reduction
or
temporary
treatment
interruption
because
of
toxicity.
Eight
patients
on
the
combination
arm
required
discontinuance
of
levamisole
but
continued
5
-FU
alone,
whereas
six
patients
required
discontinuance
of
5
-FU
but
continued
levamisole
alone.
Second
Primary
Malignant
Diseases
Twenty
patients
have
developed
second
pri-
mary
cancers
since
entry
into
this
study.
Four
of
these
were
on
the
control
arm,
seven
on
levami-
Table
5.
Toxicity
—Hematologic
Toxic
Reaction
Levamisole
Plus
5
-FU
Levamisole
Alone
Induction
Maintenance
(%) (%)
(
96
)
(n
=
124)
(n
=
128)
(n
=
118)
Leukopenia
<
4,000
a-
2,000
14
47
42
<
2,000
0
12
3
Thrombocytopenia
<
130,000
50,000
3
10
24
<
50,000
1
0
0
NOTE.
Excludes
cancelled
patients
and
patients
with
inade-
quate
counts
recorded.
sole
-alone
arm,
and
nine
on
levamisole-plus-5-
FU
arm.
Seven
patients
developed
a
second
primary
colorectal
cancer
that
was
clearly
sepa-
rate
from
the
previous
line
of
anastomosis.
The
other
second
primaries
were
three
gastric,
two
lung,
and
one
each
of
bile
duct,
bladder,
breast,
esophagus,
ovary,
prostate,
pancreas,
and
kid-
ney.
No
patient
has
developed
leukemia.
The
small
increases
in
second
primaries
on
the
levami-
sole
and
levamisole-plus-5-FU
arms
are
not
sta-
tistically
significant
and
are
not
likely
to
be
clinically
meaningful
in
view
of
the
large
number
of
controlled
studies
of
levamisole
conducted
in
other
diseases
without
documentation
of
a
simi-
lar
fi
nding.
DISCUSSION
This
study
indicates
the
likelihood
that
adju-
vant
therapy
with
levamisole
plus
5
-FU
substan-
tially
reduces
the
recurrence
rate
for
patients
with
resected
but
poor
-prognosis
large
-bowel
cancer.
It
also
indicates
the
possibility
that
a
similar
result
might
be
obtained
with
levamisole
alone.
Either
therapy
is
patient
-tolerable,
associ-
ated
with
a
minimal
risk,
and
could
easily
be
adopted
as
standard
practice.
However,
before
regarding
the
results
of
this
study
as
a
major
cancer
treatment
breakthough,
several
precau-
tionary
notes
should
be
sounded.
Whereas
we
have
documented
significant
reductions
in
cancer
recurrence,
we
have
documented
significant
sur-
vival
improvement
in
only
our
stage
C
subset.
Survival
is
the
only
hard
end
point
of
a
surgical
adjuvant
trial,
and
subset
analyses
are
notorious
for
their
misleading
results.
In
addition,
the
past
record
of
levamisole
in
human
therapeutic
trials
indicates
that
positive
results
can
be
very
fragile
and
evanescent.
Levamisole
is
a
drug
with
a
fascinating
his-
tory.
Globally,
it
has
been
an
agent
of
singular
importance,
due
to
some
extent
to
its
use
as
an
antihelminthic
agent
in
humans
in
Third
World
countries,
but
particularly
because
of
its
effective-
ness
in
treating
nematode
infections
in
farm
animals
where
it
has
had
a
major
effect
on
agricultural
economy.
10
It
has
periodically
been
claimed
to
have
effectiveness
for
a
variety
of
therapeutically
intractable,
disabling,
or
life
-
threatening
human
diseases
including
rheuma-
toid
arthritis,
ankylosing
spondylitis,
multiple
sclerosis,
chronic
hepatitis,
simple
and
genital
herpes,
Crohn's
disease,
ulcerative
colitis,
upper
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
1454
LAURIE
ET
AL
respiratory
infections,
aphthous
stomatitis,
warts,
and
leprosy.
However,
essentially
all
of
these
claims
for
effectiveness
have
been
disclaimed
by
later
and
more
scientifically
rigorous
studies.
Levamisole
has
been
extensively
evaluated
in
cancer
clinical
trials,
particularly
in
the
minimal
disease
and
surgical
adjuvant
settings.
However,
again
evidence
for
therapeutic
benefit
has
almost
invariably
failed
the
test
of
confirmation.
For
example,
in
postmenopausal,
stage
II
breast
cancer
a
randomized
Finnish
trial
claimed
that
levamisole
surgical
-adjuvant
therapy
produced
a
highly
significant
improvement
in
survival
(P
=
.003),
19
whereas
a
large
trial
from
The
Netherlands
showed
no
benefit"
and
an
even
larger
study
by
a
Danish
group
showed
levami-
sole
to
be
associated
with
a
less
favorable
survival
than
untreated
controls.
21
As
surgical
adjuvant
treatment
for
non
—small
-cell
lung
cancer,
a
ran-
domized
trial
by
Amery
et
al
claimed
that
levamisole
produced
a
significant
survival
advantage,
22
whereas
the
randomized
trial
by
Van
Houtte
et
al
showed
no
survival
advantage
at
all
for
levamisole,
23
and
in
stark
contrast,
the
trials
by
Anthony
et
a1
24
and
Herskovic
et
a1
25
actually
showed
the
survival
of
levamisole-
treated
patients
to
be
inferior
to
that
of
placebo
-
treated
patients.
In
malignant
melanoma
a
Cana-
dian
surgical
adjuvant
trial
showed
a
significant
advantage
for
levamisole
over
untreated
controls.
26
However,
in
this
same
study
when
levamisole
was
combined
with
BCG,
no
advan-
tage
whatsoever
was
obtained
and
in
a
compara-
ble
melanoma
adjuvant
trial,
Spittler
et
al
found
levamisole
therapy
to
produce
no
advantage
over
placebo
in
either
time
to
recurrence
or
survival.
27
The
same
confusion
of
conflicting
levamisole
results
is
seen
for
colorectal
carcinoma
specifi-
cally.
It
begins
in
the
basic
laboratory
setting
where
three
animal
model
colorectal
carcinoma
surgical
-adjuvant
studies
produced
two
results
that
were
strongly
positive
28
.
29
and
one
result
that
was
decisively
negative.
30
In
human
surgical
adjuvant
trials,
the
small
study
by
Verhaegen
et
a1
showed
a
significant
survival
advantage
for
levamisole-treated
patients.
Our
trial
shows
a
borderline
significant
advantage
in
tumor
recur-
rence
but
only
a
suggestion
of
advantage
in
survival.
A
small
trial
of
the
Western
Cancer
Study
Group
(only
26
patients
on
the
control
arm)
was
completely
negative.'
A
large
and
much
more
convincing
trial
of
the
European
Organization
for
Research
and
Treatment
of
Cancer
(EORTC)
compared
levamisole
with
placebo
therapy
in
297
surgically
treated
colon
cancer
patients.
They
reported
5
-year
survival
rates
at
51%
(SE,
5.4%)
with
levamisole
versus
39%
(SE,
5.9%)
with
placebo.
The
overall
sur-
vival
difference,
however,
was
not
significant
(P
=
0.35)
and
they
concluded
that
the
overall
effect
of
therapy,
if
any,
was
likely
to
be
small.
The
most
convincing
evidence
of
meaningful
therapeutic
activity
produced
by
our
study
is
the
combination
of
levamisole
and
5
-FU.
Here
we
have
a
significant
delay
in
tumor
recurrence,
a
significant
overall
reduction
of
tumor
recurrence,
and
a
suggestion
of
improved
survival
that
is
particularly
evident
in
our
stage
C
subset.
The
validity
of
these
observations
is
enhanced
by
the
recent
colon
surgical
adjuvant
trial
of
Windle
et
al,
who
randomized
patients
between
no
treat-
ment,
5
-FU
alone,
and
a
levamisole/5-FU
combination.
32
Although
this
study
was
very
small
(only
42
to
45
patients
per
treatment
arm),
it
did
demonstrate
a
highly
significant
survival
advantage
for
patients
treated
with
the
levami-
sole/5-FU
combination
in
comparison
with
both
no
treatment
and
5
-FU
alone.
However,
it
should
be
noted
that
there
were
very
important
differ-
ences
between
this
study
and
ours.
These
differ-
ences
include
a
large
proportion
of
rectal
cancer
patients
and
considerable
difference
in
regimen
for
both
levamisole
and
5
-FU.
It
is
interesting
to
speculate
why
the
addition
of
levamisole
to
5
-FU
may
be
producing
a
positive
colorectal
cancer
surgical
adjuvant
re-
sult
in
spite
of
the
fact
that
results
in
this
setting
with
levamisole
alone
must
be
considered
equivo-
cal,
and
numerous
studies
with
5
-FU
alone
have
failed
to
demonstrate
a
significant
survival
advan-
tage.
Certainly,
the
most
immediate
presumption
is
that
levamisole
is
acting
as
an
immunorestor-
ative
agent
in
patients
who
are
immunosup-
pressed
by
both
recent
surgery
and
subsequent
chemotherapy.
There
is
a
wealth
of
animal
-
model
evidence,
as
well
as
some
human
evidence,
that
levamisole
does
have
immunomodulatory
activity
93°
'
33
and
that
this
activity
is
particularly
exerted
in
the
host
with
suppressed
immune
mechanisms.
This
effect
is
probably
exerted
through
T
-cell
activation,
augmentation
of
mac-
rophage
activity,
and
an
increase
in
chemotactic
response
of
polymorphonuclear
cells
and
mono-
cytes.
In
some
animal
tumor
models
levamisole
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
SURGICAL
ADJUVANT
THERAPY
OF
COLORECTAL
CANCER
1455
has
clearly
demonstrated
an
antineoplastic
effect.
34
However,
with
all
this
it
is
disturbing
that
there
has
not
been
any
consistent
demonstra-
tion
of
antineoplastic
activity
in
the
human
setting.
An
alternative
postulate
is
that
levami-
sole
may
be
contributing
to
the
activity
of
5
-FU
in
a
manner
that
is
entirely
independent
of
immune
modulation.
Levamisole
has
a
broad
spectrum
of
additional
pharmacologic
activities.
It
has
cholinergic
properties,
it
possesses
some
mood
-elevating
effects
in
humans,
and
at
high
doses
it
acts
as
a
convulsant.
34
It
probably
influ-
ences
prostaglandin
activity,
it
is
a
specific
inhib-
itor
of
fumarate
reductase
in
several
nematodes,
and
it
is
a
very
potent
inhibitor
of
mammalian
alkaline
phosphatases.
In
a
number
of
animal
tumor
models
the
combinations
of
levamisole
with
cyclophosphamide
or
methyl
CCNU
or
carmustine
(BCNU)
have
produced
striking
im-
provements
in
animal
survival
when
compared
with
any
of
these
chemotherapeutic
agents
used
alone.
35
It
is
of
interest
that
in
most
of
these
animal
tumor
models
levamisole
alone
has
no
effect
whatsoever
on
survival.
Whereas
it
is
tempting
to
state
that
levamisole
is
acting
as
a
biochemical
modulator
of
5
-FU
activity,
at-
tempts
to
demonstrate
this
effect
clinically
have
predictably
met
with
conflicting
results,
at
least
in
the
advanced
disease
setting.
For
a
random-
ized
comparison
of
5
-FU
alone
versus
levamisole
plus
5
-FU,
Borden
et
al
gave
a
preliminary
report
suggesting
a
survival
advantage
for
patients
treated
with
the
combination.
35
In
contrast,
a
study
by
our
group
of
a
similar,
but
not
identical,
REFER
1.
Moertel
CG:
Trials,
errors
and
glimmers
of
success
in
the
surgical
adjuvant
treatment
of
colorectal
cancer,
Levin
B
(ed):
Gastrointestinal
Cancer.
Austin,
TX,
University
of
Texas
Press,
1988,
pp
3-17
2.
Buyse
M,
Zelenuich-Jacquotte
A,
Chalmers
TC:
Adju-
vant
therapy
of
colorectal
cancer.
Why
we
still
don't
know.
JAMA
259:3571-3578,
1988
3.
Wolmark
N,
Fisher
B,
Rockette
H,
et
al:
Postoperative
adjuvant
chemotherapy
or
BCG
for
colon
cancer.
J
Natl
Cancer
Inst
1:30-36,
1988
4.
Moertel
CG:
Surgical
adjuvant
treatment
of
large
bowel
cancer.
J
Clin
Oncol
6:934-939,
1988
5.
Mansour
E,
Ryan
L,
Lerner
H,
et
al:
Lack
of
effective-
ness
of
5
-FU
+
methyl
CCNU
as
compared
to
5
-FU
for
adjuvant
therapy
in
colon
cancer.
Proc
Am
Soc
Clin
Oncol
8:115,
1989
6.
Panettiere
FJ,
Goodman
PJ,
Costanzi
JJ,
et
al:
Adju-
vant
therapy
in
large
bowel
adenocarcinoma.
J
Clin
Oncol
6:947-954,
1988
levamisole/5-FU
combination
produced
a
sur-
vival
curve
that
completely
overlapped
that
of
5
-FU
alone.
36
Finally,
one
could
speculate
that
levamisole
had
no
role
at
all
and
that
our
study
simply
demonstrates
effectiveness
of
5
-FU
alone
as
displayed
by
modern
clinical
trial
methodol-
ogy.
Whereas
the
results
of
Windle
et
al
speak
against
this,
32
the
remote
possibility
cannot
be
denied.
Certainly,
the
checkered
history
of
levamisole
trials
in
human
cancer
speaks
strongly
against
overinterpretation
of
the
results
of
any
single
trial.
Nevertheless,
the
results
of
our
study
are
provocative
and
raise
the
hope
that
a
step
of
progress
may
have
been
taken.
Even
though
the
results
of
Windle
et
al
would
appear
to
give
confirmation,
32
this
study
involved
only
a
rela-
tively
few
patients
and
their
levamisole/5-FU
regimen
was
different
from
ours.
We
felt
very
strongly
that
further
confirmation
was
required.
A
national
intergroup
study
was
therefore
under-
taken
involving
the
Eastern
Cooperative
Oncol-
ogy
Group,
the
NCCTG,
and
the
Southwest
Oncology
Group.
This
trial,
which
essentially
duplicates
the
methods
of
the
study
reported
here,
has
enrolled
over
1,200
patients,
all
of
whom
completed
their
year
of
therapy
by
Octo-
ber
1987.
We
hope
that
preliminary
results
may
be
available
in
1
to
2
years.
ACKNOWLEDGMENT
Additional
participating
institutions
include
The
St
Cloud
Clinic
of
Internal
Medicine,
Ltd,
St
Cloud,
MN,
and
Quain
and
Ramstad
Clinic,
Bismarck,
ND.
ENCES
7.
Gastrointestinal
Tumor
Study
Group:
Adjuvant
ther-
apy
of
colon
cancer.
N
Engl
J
Med
310:737-743,
1984
8.
Souter
AG,
Gill
PG,
Morris
PJ:
A
trial
of
non-specific
immunotherapy
Using
systemic
C.
parvum
treated
patients
with
dukes
B
and
C
colorectal
cancer.
Br
J
Cancer
45:506-
512,
1982
9.
Taylor
J,
Machin
D,
Mulleet
M,
et
al:
A
randomized
controlled
study
of
adjuvant
portal
vein
cytotoxic
perfusion
in
colorectal
cancer.
Br
J
Surg
72:359-363,
1984
10.
Verhaegen
H,
DeCree
J,
DeCock
W,
et
al:
Levamisole
therapy
in
patients
with
colorectal
cancer,
in
Terry
WD,
Rosenberg
SA
(eds):
Immunotherapy
of
Human
Cancer,
New
York,
NY,
Elsevier,
1982,
pp
225-229
11.
Chirigos
MA,
Amery
WK:
Combined
levamisole
ther-
apy:
An
overview
of
its
protective
effects,
in
Immunotherapy
of
Human
Cancer.
New
York,
NY,
Raven,
1978,
pp
181-195
12.
Renoux
G:
The
general
immunopharmacology
of
le-
vamisole.
Drugs
20:89-99,
1980
13.
Grage
TB,
Moss
SE:
Adjuvant
chemotherapy
in
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.
1456
LAURIE
ET
AL
cancer
of
the
colon
and
rectum:
Demonstration
of
effective-
ness
of
prolonged
5
-FU
chemotherapy
in
a
prospectively
controlled
randomized
trial.
Surg
Clin
North
Am
61:1321-
1329,
1981
14.
Laurie
J,
Moertel
C,
Fleming
T,
et
al:
Surgical
adjuvant
therapy
of
poor
prognosis
colorectal
cancer
with
levamisole
alone
or
combined
levamisole
and
5-fluorouracil
(5
-FU).
Proc
Am
Soc
Clin
Oncol
5:316,
1986
(abstr)
15.
SAS
Program
Institute
Inc,
SUGI:
Supplemental
Library
Users'
Guide.
(ed
5).
Cary,
NC,
SAS
Institute,
1986,
pp
662
16.
Kaplan
EL,
Meier
P:
Nonparametric
estimation
from
incomplete
observations.
J
Am
Stat
Assoc
53:457-481,
1958
17.
Mantel
H,
Naenszel
W:
Statistical
aspects
of
the
analysis
of
data
from
retrospective
studies
of
disease.
J
Natl
Cancer
Inst
22:719-748,
1959
18.
Cox
DR:
Regression
models
and
life
tables.
J
Roy
Statis
Soc
B
34:187-202,
1972
19.
Klefstrom
P,
Holsti
P,
Grohn
P,
et
al:
Levamisole
in
the
treatment
of
stage
II
breast
cancer.
Cancer
55:2753-
2757,
1985
20.
Treurniet-Donker
AD,
Meiscke
de
Jongh
ML,
van
Puttens
WLJ:
Levamisole
as
adjuvant
immunotherapy
in
breast
cancer.
Cancer
59:1590-1593,
1987
21.
Danish
Breast
Cancer
Cooperative
Group:
Increased
breast
cancer
recurrence
rate
after
adjuvant
therapy
with
levamisole.
Lancet
2:824-827,
1980
22.
Amery
WK,
Cosemans
J,
Goozen
HC,
et
al:
Four
year
results
from
double-blind
study
of
adjuvant
levamisole
treat-
ment
in
resectable
lung
cancer,
in
Terry
WE,
Rosenberg
SA
(eds):
Immunotherapy
of
Human
Cancer.
New
York,
NY,
Elsevier,
1982,
pp
123-133
23.
Van
Houtte
P,
Bondue
H,
Rocmans
P,
et
al:
Adjuvant
immunotherapy
by
levamisole
in
resectable
lung
cancer.
Eur
J
Cancer
16:1597-1601,
1980
24.
Anthony
HM,
Mearns
AJ,
Masono
MK,
et
al:
Levami-
sole
and
surgery
in
bronchial
carcinoma
patients.
Thorax
34:4-12,
1979
25.
Herskovic
A,
Bauer
M,
Seydel
HG,
et
al:
Post-
operative
thoracic
irradiation
with
or
without
levamisole
in
non
-small
cell
lung
cancer.
Int
J
Radiat
Oncol
Biol
Phys
14:37-42,
1988
26.
Quirt
I,
Shelley
W,
Bodurtha
A,
et
al:
Adjuvant
levamisole
improves
survival
and
disease
free
survival
in
patients
with
poor
prognosis
malignant
melanoma.
Proc
Am
Soc
Clin
Oncol
5:130,
1986
(abstr)
27.
Spittler
LE,
Sagebeil
R,
Allen
R,
et
al:
Levamisole
in
the
treatment
of
melanoma,
in
Terry
WD,
Rosenberg
SA
(eds):
Immunotherapy
of
Human
Cancer.
New
York,
Elsevier,
1982,
pp
289-297
28.
House
AK,
Maley
MAL:
Clinical
and
in
vivo
response
following
surgery
or
surgery
plus
adjuvant
chemotheapy
or
immunotherapy
for
colorectal
carcinoma
in
a
rat
model.
J
Royal
Soc
Med
76:833-840,
1983
29.
Weese
JL,
Gilbertson
EM,
Syrjala
SE:
Prevention
of
rat
colon
cancer
metastases
by
perioperative
immunostimula-
tion.
Surgery
96:420-426,
1984
30.
Martin
MS,
Justrabo
E,
Martin
F,
et
al:
Ineffective-
ness
of
levamisole
as
adjuvant
to
surgery
with
two
lines
of
transplanted
rat
colonic
carcinoma.
Br
J
Cancer
44:464-466,
1981
31.
Chlebowski
RT,
Nystrom
S,
Reynolds
R,
et
al:
Long-
term
survival
following
levamisole
or
placebo
adjuvant
treat-
ment
of
colorectal
cancer.
Oncology
45:141-143,
1988
32.
Windle
R,
Bell
PRF,
Shaw
D:
Five
year
results
of
a
randomized
trial
of
adjuvant
5-fluorouracil
and
levamisole
in
colorectal
cancer.
Br
J
Surg
74:569-572,
1987
33.
Harris
BG:
Levamisole:
Mode
of
action.
Biochem
Soc
Trans
15:61,
1987
34.
Goldin
A,
Chirigos
MA,
Macdonald
JS,
et
al:
Biologic
response
modifiers
and
adjuvant
chemotherapy.
Recent
Re-
sults
Cancer
Res
80:351-356,
1982
35.
Borden
EC,
Davis
TE,
Crowley
JJ,
et
al:
Interim
analysis
of
a
trial
of
levamisole
and
5-fluorouracil
in
meta-
static
colorectal
carcinoma,
in
Terry
WD,
Rosenberg
SA
(eds):
Immunotherapy
of
Human
Cancer.
New
York,
NY,
Elsevier,
1982,
pp
231-235
36.
Buroker
TR,
Moertel
CG,
Fleming
TR,
et
al:
A
controlled
evaluation
of
recent
approaches
to
biochemical
modulation
or
enhancement
of
5-fluorouracil
therapy
in
colorectal
carcinoma.
J
Clin
Oncol
3:1624-1631,
1985
from
005.189.204.108
Copyright
©
2017
American
Society
of
Clinical
Oncology.
All
rights
reserved.