Review
Current Issues in Adjuvant Treatment of Stage II
Colon Cancer
Thierry Andre
´
, MD,
1,2,3
Daniel Sargent, PhD,
4
Josep Tabernero, MD,
5
Michael OÕConnell, MD,
6
Marc Buyse, ScD,
7
Alberto Sobrero, MD,
8
Jean-Louis Misset, MD,
9
Corrado Boni, MD,
10
and Aimery de Gramont, MD
2,3,11
1
Service d’Oncologie Me
´
dicale, H
^
opital Tenon, 4 Rue de la Chine, 75970 Paris Cedex 20, France
2
CancerEst, H
^
opital Saint Antoine, 184 Rue du Faubourg Saint Antoine, 75571 Paris Cedex 12, France
3
GERCOR (Franch Oncology Research Group), 22 Rue Malher, 75004 Paris, France
4
Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905
5
Vall d’Hebron University Hospital, P Vall d’Hebron 119-129, 08035 Barcelona, Spain
6
NSABP Foundation Research Program, 4 Allegheny Center, Pittsburgh, Pennsylvania 15212
7
IDDI, 430 Avenue Louise, BP 14, B 1050 Brussels, Belgium
8
Ospedale S. Martino, Largo Benzi 10, Genova 16132, Italy
9
H
^
opital Saint-Louis, 1 Avenue Claude Vellefaux, 75010 Paris, France
10
Arcispedale S. Maria Buova, Viale Resorgimento 80, Regio Emilia 42100, Italy
11
H
^
opital Saint-Antoine, 184 Rue du Faubourg Saint Antoine, 75571 Paris Cedex 12, France
Background: Adjuvant chemotherapy with 5-fluorouracil modulated by folinic acid, com-
bined with oxaliplatin, has now become an accepted standard of care for patients with stage
III colon cancer. In contrast, the use of adjuvant therapy for stage II patients remains con-
troversial, and the identification of reliable prognostic factors to aid therapeutic decision
making is crucial.
Methods: The authors critically review the results of clinical trials and meta-analyses
investigating the value of adjuvant chemotherapy for stage II patients, emphasizing the het-
erogeneous nature of this population and the difficulty of performing clinical trials with
sufficient power to reliably assess treatment efficacy. They also discuss the evidence concerning
potential prognostic factors, particularly molecular markers.
Results: Available clinical trial data do not support the routine use of adjuvant chemo-
therapy for all stage II patients but suggest that it should be considered, particularly for certain
high-risk patients. Recent guidelines advocate considering factors such as tumor differentia-
tion, tumor perforation, number of lymph nodes examined, and T stage when assessing the
likely benefit:risk ratio. Microsatellite instability and allelic imbalance seem to be strong pre-
dictors of good and poor prognosis, respectively, and in the near future, therapeutic decision-
making models are likely to be further refined by the inclusion of such molecular markers.
Conclusions: There is growing evidence that the prognosis of certain stage II patients with
unfavorable prognostic factors can be improved by adjuvant chemotherapy, and increasingly
refined tools are now available to define those most likely to benefit. Referral of stage II patients
for individual assessment is strongly recommended.
Received July 5, 2005; accepted November 15, 2005; published
online April 14, 2006.
Address correspondence and reprint requests to: Thierry Andre
´
,
MD, E-mail: thierry.andre@tnn.aphp.fr.
Published by Springer Science+Business Media, Inc. Ó 2006 The Society of
Surgical Oncology, Inc.
Annals of Surgical Oncology, 13(6): 887)898
DOI: 10.1245/ASO.2006.07.003
887
Key Words: Colon cancer, stage II—Adjuvant chemotherapy—5-Fluorouracil—Leucovorin—
Oxaliplatin.
Over the past years, considerable progress has been
made in the treatment of patients with colon cancer.
The use of optimized folinic acid (FA)-modulated 5-
fluorouracil (5-FU) regimens has increased the med-
ian survival of patients with metastatic (stage IV)
colon cancer from barely 6 months without treatment
to >12 months.
1
The addition of new chemothera-
peutic agents (e.g., oxaliplatin and irinotecan) to
these regimens has further extended the median sur-
vival to approximately 20 months.
2–7
The combina-
tion of newly approved biological agents, such as
bevacizumab and cetuximab, with optimal chemo-
therapy promises to prolong survival still more.
8,9
At diagnosis, approximately 14% of patients have
stage I disease, 28% have stage II, 37% have stage III,
and 21% have stage IV.
10
In the context of continu-
ally improving treatments for advanced disease, it is
not surprising that the investigation of newer
chemotherapies has been extended to patients with
stage II and III colon tumors after surgical resection.
In these patients, the prognosis varies widely
according to the disease stage and, in particular, the
extent of lymph node involvement and the extent to
which the tumor has invaded the bowel wall
(expressed by the T [tumor] and N [node] stage in the
tumor-node-metastasis classification). The 5-year
survival rate decreases from >90% for stage I tumors
to 50% for stage III tumors.
11
Patients with stage II
colon cancer constitute a particularly heterogeneous
population. Patients with tumors that are barely
starting to penetrate through the bowel wall have a
prognosis approaching that of patients with stage I
cancer. In contrast, similarly staged patients with
aggressive and extensive tumors showing extramural
venous spread and involvement of the serosa or
adjacent organs have a prognosis similar to that of
patients with stage III cancer.
12
In view of the generally good prognosis of pa-
tients with resected stage I tumors, clinical studies
on adjuvant chemotherapy have mainly focused
exclusively on patients with resected stage III tumors
or on patients with stage II or stage III tumors but
without differentiating between these two groups.
Trials performing separate analyses of the effect of
adjuvant therapy on stage II and stage III patients
tend to have insufficient power to assess treatment
efficacy adequately in stage II patients, because the
number of these patients is generally small and the
event rate is low. It has been estimated that a pro-
spective clinical trial of adjuvant chemotherapy
including a nontreatment control arm and designed
to detect a 4% survival benefit at 5 years among
stage II colon cancer patients with a baseline 5-year
survival prognosis of 75% would require at least
4700 patients.
13
No study of this size has yet been
published.
On the basis of published data on patients with
stage III cancer, 6 months of adjuvant chemotherapy
with 5-FU/FA became the standard of care for such
patients in the early 1990s.
14–18
More recently, the
addition of oxaliplatin to 5-FU/FA has further im-
proved patient outcomes, thus establishing this
combination as a new standard of care.
19,20
Although
the benefit of adjuvant chemotherapy for stage III
colon cancer is now acknowledged, the use of adju-
vant therapy in patients with stage II colon cancer
remains controversial. Increasing attention is being
focused on the identification of unfavorable prog-
nostic factors that could aid in making the decision
for or against treatment on the basis of the relative
benefits and risks for each individual patient. This
review presents an overview of the recent and exten-
sive published data concerning adjuvant chemother-
apy for stage II colon cancer.
EFFICACY OF ADJUVANT CHEMOTHERAPY
IN STAGE II COLON CANCER: ANALYSES,
META-ANALYSES, AND RECENT CLINICAL
STUDIES
Published clinical trials conducted until very re-
cently were not been large enough to demonstrate a
statistically significant advantage for chemotherapy
in stage II colon cancer. For this reason, several
analyses and meta-analyses have been performed of
clinical trials comparing adjuvant therapy with
observation in patients with stage II colon or colo-
rectal cancer.
The National Surgical Adjuvant Breast and Bowel
Project (NSABP) performed an analysis of the results
of its four trials on adjuvant chemotherapy in pa-
tients with colon cancer: two comparing surgery
alone with surgery plus adjuvant chemotherapy, and
two comparing different adjuvant chemotherapy
regimens.
21
A total of 3820 patients were available for
T. ANDRE
´
ET AL.888
Ann. Surg. Oncol. Vol. 13, No. 6, 2006
analysis: 41% had stage II and 59% had stage III
disease. Two pooled treatment groups were com-
pared within each stage of disease: one comprising
the less effective treatment group from each trial and
the other comprising the more effective group. The
relative reduction in risk in the more effective treat-
ment group was 30% for the 1565 patients with stage
II colon cancer (irrespective of the presence or ab-
sence of adverse prognostic factors), compared with
18% for the 2255 patients with stage III colon cancer.
In two of the trials, the risk reduction achieved in
stage II patients was statistically significant. How-
ever, this analysis has been criticized for pooling the
results from trials with different control and treat-
ment arms.
The International Multicentre Pooled Analysis of
B2 Colon Cancer Trials (IMPACT B2) included data
on 1016 patients with stage II colon cancer generated
in 5 randomized clinical trials comparing 5-FU–
containing adjuvant chemotherapy with no treat-
ment.
22
This analysis found 5-year overall survival
(OS) and disease-free survival (DFS) rates of 80%
and 73%, respectively, in untreated patients versus
82% and 76%, respectively, in treated patients, cor-
responding to relative risk reductions of 17% for DFS
and 14% for OS with adjuvant treatment. Although
these differences failed to reach statistical signifi-
cance, there was a trend toward a small benefit in
absolute survival with adjuvant therapy (2%; one si-
ded). In the multivariable Cox analysis, age and
tumor grade were the only independent predictors for
OS and DFS. In the population analyzed, 1.5% of the
patients had T4 tumors (defined as tumors directly
invading an adjacent organ or perforating the visceral
peritoneum), and 11% had poorly differentiated tu-
mors. The prognostic factors analyzed did not in-
clude perforation, bowel obstruction, venous
invasion, or the number of lymph nodes examined.
One of the major criticisms of this meta-analysis is
the relatively small number of cases included. An-
other meta-analysis of 3499 patients participating in
10 studies evaluating the efficacy of short continuous
portal vein infusion of chemotherapy demonstrated
that the relative risk benefit of this treatment on the
risk of death was nearly identical in stage III and in
stage I/II patients.
23
More recently, a pooled analysis of seven ran-
domized trials comparing adjuvant 5-FU/FA or
5-FU/levamisole with surgery alone in stage II and
III colon cancer (including the five IMPACT B2
trials) was conducted, with the aim of clarifying the
benefit of adjuvant therapy in specific subgroups of
patients.
24
This pooled analysis showed a significant
improvement in 5-year DFS in the 1440 patients
who presented with node-negative disease (76% vs.
72% for node-positive disease; P = .0490). How-
ever, the difference in 5-year OS (81% vs. 80%,
respectively) did not reach statistical significance. A
significant interaction between stage and treatment
was evident, and adjuvant chemotherapy benefited
stage III patients to a greater extent than stage II
patients. Significant prognostic factors in this anal-
ysis, besides nodal status, included histological tu-
mor grade and depth of tumor invasion into the
bowel wall. Other prognostic factors (e.g., perfora-
tion, bowel obstruction, venous invasion, and
number of lymph nodes examined) were not evalu-
ated in this study.
In an attempt to help physicians assess the bene-
fit:risk ratio of adjuvant chemotherapies for each
individual patient, tools have been developed that
incorporate the T and N stage with age, tumor dif-
ferentiation, and tumor perforation, in addition to
the number of lymph nodes retrieved.
24
A model
incorporating these factors is now available to pro-
vide physicians with tailored estimates of 5-year DFS
and OS probabilities with surgery alone and with
surgery plus 5-FU–based adjuvant chemotherapy.
25
The Cancer Care Ontario Program in Evidence-
Based Care conducted two systematic reviews of data
from clinical trials comparing adjuvant therapy with
observation in patients with resected stage II colon or
colorectal cancer.
26,27
The more recent of these re-
viewed 37 randomized controlled trials and 11 meta-
analyses that included a total of 20,317 patients
(7,803 with colon cancer and 12,514 with colorectal
cancer). The proportion of patients with stage II
disease varied from 23% to 100%.
27
A meta-analysis
was conducted on the data obtained on stage II pa-
tients whenever available (4187 patients in 18 trials).
Data were selected on the basis of stringent criteria
that required inclusion of a surgery-alone control arm
and 5-FU–based adjuvant chemotherapy (16 trials)
or immunotherapy alone (2 trials). The relative risk
reduction in mortality across the trials was .87
(hazard ratio [HR], .87; 95% confidence interval [CI],
.75–1.01; P = .07); however, this result should be
interpreted cautiously because the trials used a vari-
ety of adjuvant therapies.
Several studies have investigated adjuvant therapy
with oral fluoropyrimidines. The first clinical trials
comparing surgery followed by adjuvant therapy with
oral fluoropyrimidines versus surgery alone were
conducted in Japan, where the development of
adjuvant chemotherapy for colorectal cancer was
essentially based on these agents. A meta-analysis of
STAGE II COLON CANCER: ADJUVANT THERAPY 889
Ann. Surg. Oncol. Vol. 13, No. 6, 2006
three of these trials—all conducted by the Japanese
Foundation for the Multidisciplinary Treatment for
Cancer and using separate randomizations for
patients with colon and rectal cancer—was recently
published.
28
The total population analyzed com-
prised 5233 patients: 2848 (54%) presented with colon
cancer and 2385 (46%) with rectal cancer. Overall,
2295 patients (44%) had stage II disease, and 2348
(45%) had stage III disease. Both OS (HR, .89; P =
.04) and DFS (HR, .85; P < .001) were statistically
significantly increased by adjuvant chemotherapy in
the population as a whole, with a similar benefit
observed in patients with colon and rectal tumors.
Five-year survival was increased by 3.7%, 4.3%, and
2.4% in patients with stage I, II, and III tumors,
respectively.
The Netherlands Adjuvant Colorectal Cancer
Project trial reported an improved 5-year survival
rate in stage II colon and rectal cancer patients
receiving adjuvant 5-FU plus levamisole (78% vs.
70%) but did not specify the level of statistical sig-
nificance for the difference observed (no subgroup
analyses).
29
The recently completed Quick And Simple And
Reliable (QUASAR) trial, not included in either
published meta-analyses or the Cancer Care Ontario
Program in Evidence-Based Care review and so far
published only in abstract form, compared adjuvant
chemotherapy (5-FU modulated by FA with or
without levamisole) with observation in patients with
resected colon or rectal cancer.
30
A total of 3238
patients were enrolled, of whom 91% presented with
stage II disease (71% colon and 29% rectal cancer).
Updated results presented at the American Society of
Clinical Oncology (ASCO) 2004 meeting (median
follow-up, 4.6 years) showed a significantly improved
5-year OS (80.3% vs. 77.4%, respectively; P = .02)
and a significantly lower 5-year recurrence rate
(22.2% vs. 26.2%, respectively; P = .001) in the
adjuvant chemotherapy group compared with
observation. Separate analysis of stage II patients
showed a significant reduction in the number of
deaths with adjuvant chemotherapy (224 vs. 262,
respectively; P = .04). This is the largest single study
in the population of stage II patients, and, if these
preliminary results are confirmed in the subsequent
publication, it will be the first to demonstrate a sta-
tistically significant advantage of adjuvant chemo-
therapy for this group. The analysis of prognostic
variables in this study should help to identify which
subgroups of stage II patients are likely to benefit the
most from adjuvant chemotherapy.
An analysis conducted between 1991 and 1996 in
3151 US Medicare patients aged between 65 and 75
years with resected stage II colon cancer revealed that
27% received adjuvant chemotherapy, compared with
75% of stage III patients.
31
In that study, a younger age
at diagnosis, white race, low comorbidity, and poorly
differentiated tumor histological characteristics were
all associated with a greater likelihood of being treated,
whereas sex, socioeconomic status, and the number of
lymph nodes examined had no bearing on the treat-
ment rate. The 5-year survival rate was 78% for the
treated patients and 75% for those not treated. After
adjusting for known differences between the groups,
the HR for survival associated with adjuvant treatment
was .91 (95% CI, .77–1.09), which is consistent with the
clinical trial data discussed previously.
When viewed as a whole, the results of all these
studies (clinical studies, analyses, and meta-analyses)
demonstrate remarkable consistency. The repeated
finding is that 5-FU–based adjuvant therapy provides
a small but real benefit in terms of DFS and OS rates.
The physician and patient, when fully armed with this
knowledge, must then make a decision regarding the
relative importance of the very modest benefit in
prognosis when weighed against the risks, toxicity,
and inconvenience of therapy.
EFFECT OF NEW CHEMOTHERAPEUTIC
AGENTS ON STAGE III ONLY OR BOTH
STAGE II AND STAGE III COLON CANCER
Two studies that were not designed to test the value
of chemotherapy for patients with stage II colon cancer
compared oral fluoropyrimidines with a 5-FU/FA
bolus regimen. These two studies were presented at the
ASCO meeting in 2004. The first compared capecita-
bine with the Mayo clinic regimen (monthly bolus
regimen of 5-FU/FA) in patients with stage III cancer
only. The HR for DFS at 3 years was .87 (95% CI, .75–
1). The primary end point of this study was met, thus
demonstrating that capecitabine is not inferior to the
Mayo Clinic regimen (in fact, there was even a trend to
superiority for capecitabine).
32
The second study,
conducted by the NSABP (C06 study) in patients with
stage II and III colon cancer, compared oral uracil and
tegafur + FA to the Roswell Park regimen (weekly
bolus regimen of 5-FU/FA) and demonstrated the
perfect equality of uracil and tegafur + FA and the
Roswell Park regimen (in terms of DFS and OS at 5
years).
33
No analysis concerning only patients with
stage II colon cancer was performed.
33
T. ANDRE
´
ET AL.890
Ann. Surg. Oncol. Vol. 13, No. 6, 2006
The Multi-Center International Study of Oxalipl-
atin/5-Fluorouracil/Leucovorin in the Adjuvant
Treatment of Colon Cancer study compared the
benefit of standard 5-FU/FA adjuvant chemotherapy
(LV5FU2) with that of the same regimen but with the
addition of oxaliplatin (FOLFOX4) in 2246 patients
who had undergone curative resection for stage II or
III colon cancer.
19
After a median follow-up of 37.9
months, the 3-year DFS in the total population of
stage II and III patients was 72.9% in the LV5FU2
arm and 78.2% in the FOLFOX4 arm (HR, .77; 95%
CI, .65–.91; P = .002), corresponding to a 23%
reduction in the risk of relapse. Among the 899
patients with stage II disease, the addition of oxa-
liplatin resulted in a 20% decrease in risk of relapse,
with an increase in the DFS rate from 84.3% to 87%
after 3 years.
34
An analysis of stage II patients with at
least one adverse prognostic factor (T4 tumor, bowel
obstruction or tumor perforation, poor differentia-
tion, venous invasion, or <10 lymph nodes exam-
ined) was reported recently. In this subpopulation
(n = 576), FOLFOX4 patients had a 28% decrease in
the risk of relapse (HR, .72; 95% CI, .48–1.08) and a
3-year DFS of 84.9% vs. 79.8% for LVSFU2 patients.
These benefits were within the range of those ob-
served in stage III patients and were associated with
limited toxicity. No analysis of molecular markers
was performed, because no samples were collected
during the trial.
34
In the NSABP C07 study, 2407 patients with stage
II and III carcinoma of the colon (28.6% stage II and
71.4% stage III) were randomized to receive either
FULV (a 5-FU 500 mg/m
2
bolus weekly and FA
[leucovorin; LV] 500 mg/m
2
weekly, each for 6 weeks
of every 8-week cycle, for 3 cycles) or FLOX (same
FULV regimen with the addition of oxaliplatin 85
mg/m
2
on weeks 1, 3, and 5 of each 8-week cycle, for 3
cycles). The addition of oxaliplatin to the FULV
regimen significantly improved the 3-year DFS in
patients with stage II and III colon cancer (71.6% vs.
76.5%, respectively; P < .004; HR = .79; 95% CI,
.67–.93). No subgroup analysis was performed. The
global test for interaction between treatment and tu-
mor stage (II and III) was not significant (P = .70).
20
PROGNOSTIC FACTORS IN STAGE II COLON
CANCER
It is increasingly recognized that not all stage II
patients are likely to benefit equally from adjuvant
chemotherapy, and the decision to treat patients is
often made on the basis of perceived additional risk
factors that may outweigh the inconvenience and
likely toxicity of treatment. Despite the past uncer-
tainty concerning the benefit:risk ratio of adjuvant
chemotherapy in patients with stage II colon cancer,
many of these patients nevertheless receive such
treatment. Conversely, some centers deny chemo-
therapy to stage II patients because of the perceived
lack of convincing data.
Stage II Colon Cancer: A Heterogeneous Population
Stage II colon cancer, characterized by the absence
of lymph node involvement, covers a wide spectrum
of disease, ranging from tumors with little penetra-
tion through the bowel wall to aggressive and
extensive tumors with extramural venous spread and
involvement of the serosa or adjacent organs. It is
becoming increasingly evident that colon cancers are
also heterogeneous with respect to many other clini-
cal, pathologic, and biological factors and that
traditional pathologic staging systems may be insuf-
ficient to predict outcome accurately. The proven
value of adjuvant chemotherapy for patients with
stage III colorectal cancer and the controversy sur-
rounding its use in stage II patients have been a
particular spur to the investigation of potential
prognostic factors that enable identification of the
patients most likely to benefit from this treatment.
With regard to stage II cancers, a wide variety of
potential clinical and pathologic risk factors for
recurrence have been investigated. The most impor-
tant factors for predicting the risk of recurrence are
emergency presentation (bowel perforation or occlu-
sion), poorly differentiated tumor (histological
grade), depth of tumor invasion and adjacent organ
involvement (T4), extramural venous invasion, and
peritoneal involvement.
11,24,33
In the case of
involvement of the resection margins, the resection is
not curative. Age affects survival because of deaths
from other causes; however, a benefit of adjuvant
chemotherapy for both young and old patients was
demonstrated in a large pooled analysis.
35
In a study
that specifically excluded patients with lymph node
involvement (i.e., stage III patients), Petersen et al.
36
identified six factors by univariate analysis that sig-
nificantly influence prognosis (P < .01). These were
extent of tumor spread, peritoneal involvement,
margin involvement, venous invasion, tumor perfo-
ration, and adjacent organ involvement. Patient age,
sex, tumor site, and differentiation were not found to
be statistically significant prognostic factors. The
final validated Cox regression model combined four
independent prognostic factors: peritoneal involve-
STAGE II COLON CANCER: ADJUVANT THERAPY 891
Ann. Surg. Oncol. Vol. 13, No. 6, 2006
