1. health and fitness
2. disease
3. cancer

A d j u v a n t C h e m o t h e r a p y U s e f o r M e d i c a r e B e n e fi c i a r i e s
With Stage II Colon Cancer
By Deborah Schrag, Sheryl Rifas-Shiman, Leonard Saltz, Peter B. Bach, and Colin B. Begg
Purpose: Clinical trials have not demonstrated that
adjuvant chemotherapy improves survival for patients
with resected stage II colon cancer. Nevertheless, patients may receive this treatment despite its uncertain
benefit. The objective of this study was to determine the
extent to which adjuvant chemotherapy is used for
patients with stage II colon cancer.
Patients and Methods: Using the Surveillance, Epidemiology, and End Results-Medicare linked database,
we identified 3,151 patients aged 65 to 75 with resected stage II colon cancer and no adverse prognostic
features. The primary outcome was chemotherapy use
within 3 months of surgery ascertained from claims
submitted to Medicare. Relationships between patient
characteristics and adjuvant chemotherapy use were
measured and their significance was assessed using
multivariable logistic regression. Survival for treated
and untreated patients was compared using a Cox
model.
Results: Twenty-seven percent of patients received
chemotherapy during the 3 postoperative months.
Younger age at diagnosis, white race, unfavorable
tumor grade, and low comorbidity were each associated with a greater likelihood of receiving treatment.
Sex, the number of examined lymph nodes in the tumor
specimen, the urgency of the surgical admission, and
median income was each unrelated to treatment. Fiveyear survival was 75% for untreated patients and 78%
for treated patients. After adjusting for known between-group differences, the hazard ratio for survival
associated with adjuvant treatment was 0.91 (95%
confidence interval, 0.77 to 1.09).
Conclusion: A substantial percentage of Medicare
beneficiaries with resected stage II colon cancer receive
adjuvant chemotherapy despite its uncertain benefit.
J Clin Oncol 20:3999-4005. © 2002 by American
Society of Clinical Oncology.
SE OF ADJUVANT chemotherapy remains controversial for patients with stage II (T3N0M0 Dukes’ B2)
colon cancer. Randomized trials have not demonstrated that
treatment improves overall survival.1-4 The most informative
source of data comes from the International Multicentre Pooled
Analysis of Colon Cancer Trials B2 investigators, who pooled
data from 1,016 stage II patients treated on five randomized
trials comparing fluorouracil-containing regimens to no treatment. They found that 5-year overall survival was 80% for
untreated and 82% for treated patients, a difference that was
not statistically significant.4 Unfortunately, no randomized
trials have been large enough to exclude the possibility that
adjuvant treatment provides a survival benefit in the singledigit range. A study would need to randomize 3,572 patients to
detect a 4% survival difference between chemotherapy recipients and nonrecipients at 5 years with 80% power. Despite the
persistent uncertainty regarding the benefit of adjuvant chemotherapy for stage II colon cancer, the clearly established
survival benefit for patients with stage III tumors and evidence
that treatment increases relapse-free survival have led some
physicians to recommend adjuvant chemotherapy for their
stage II patients. Clinical guidelines recommend that stage II
patients consider participation in clinical trials, but there is
continued debate regarding whether inclusion of a no-treatment control arm is appropriate.5-7 A no-treatment arm is not
included in a current cooperative group study.
To examine how physicians and cancer patients have
interpreted the controversy regarding optimal treatment of
stage II colon cancer, we measured chemotherapy use
among Medicare beneficiaries with this diagnosis made in a
Surveillance, Epidemiology, and End Results (SEER) region during the years 1991 to 1996. Our goal was to identify
the clinical and demographic characteristics predictive of
chemotherapy use and postoperative patterns of care. We
also sought to compare clinical outcomes for treated and
untreated patients, adjusting for known differences in these
groups to the extent possible in the nonrandomized setting.
U
PATIENTS AND METHODS
Data Sources
Data from the SEER cancer registries and the Medicare claims files
of the Health Care Financing Administration (HCFA) have been linked
in order to facilitate population-based studies of the medical and
From the Department of Epidemiology and Biostatistics and Department of Medicine, Memorial Sloan-Kettering Cancer Center, New
York, NY.
Submitted November 19, 2001; accepted June 17, 2002.
This study used the linked Surveillance, Epidemiology, and End
Results-Medicare database. The interpretation and reporting of these
data are the sole responsibility of the authors.
Address reprint requests to Deborah Schrag, MD, MPH, Health
Outcomes Research Group, Department of Epidemiology and Biostatistics, Memorial-Sloan Kettering Cancer Center, 1275 York Ave, New
York, NY 10021; email: [email protected].
© 2002 by American Society of Clinical Oncology.
0732-183X/02/2019-3999/$20.00
Journal of Clinical Oncology, Vol 20, No 19 (October 1), 2002: pp 3999-4005
DOI: 10.1200/JCO.2002.11.084
Downloaded from jco.ascopubs.org on September 9, 2014. For personal use only. No other uses without permission.
Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
3999
4000
SCHRAG ET AL
economic outcomes of cancer treatment. The SEER registries, sponsored by the National Cancer Institute, ascertain all incident cancer
cases diagnosed in five states and six United States metropolitan areas,
representing approximately 14% of the United States population.8 The
SEER program collects information on each incident cancer, including
the primary site, histology classified according to the International
Classification of Diseases for Oncology, 2nd edition schema (ICD-0 to
2),9 tumor stage at diagnosis, and patient demographics. Ninety-four
percent of patients in the SEER registry aged 65 years or older have
been successfully linked to their Medicare records.10
The Medicare program provides health insurance for 97% of the
United States population aged 65 and older. The Medicare Provider
Analysis and Review files provide details of all hospitalizations. For the
96% of Medicare beneficiaries who opt for part B coverage, claims for
care delivered in hospital outpatient departments and physician’s
offices are also recorded. In order to receive payment, health care
providers submit medical claims coding diagnoses and procedures
using the ICD, Ninth Revision (ICD-9) classification.
Cohort Definition
All patients aged 65 to 75 diagnosed with primary colon cancer in a
SEER area during the years 1991 to 1996 and enrolled in parts A and
B Medicare were potentially eligible for inclusion in our study. We
excluded the 16% of patients enrolled in Medicare-managed care plans
because these insurers were not required to submit detailed claims to
the HCFA. Patients over the age of 75 were excluded from our analysis
on the basis of a prior study indicating that few patients in this age
bracket receive chemotherapy even in the setting of stage III disease.11
Adenocarcinomas of the colon were defined using SEER codes for
cancer site (18.0 to 18.9) and histology (8140, 8210-11, 8220-21,
8260-63, 8470, 8480-81, and 8490). Diagnoses noted exclusively on
death certificates or at autopsy were excluded, as were those where the
month of diagnosis or number of examined lymph nodes was unknown.
We searched Medicare claims records for patients who had colon
cancer operations performed within 3 months of primary diagnosis and
operations consistent with definitive tumor resection, according to the
ICD-9, Clinical Modification classification (45.7x, 45.8, 48.4x, 48.5,
and 48.6x). We identified patients with stage II disease using information on tumor size, nodal involvement, and distant spread recorded in
the SEER database and classified according to the American Joint
Committee on Cancer staging schema. Our analysis was restricted to
patients who survived for at least 3 months after surgery because
adjuvant therapy is not relevant for those who die in the immediate
postoperative period. Patients diagnosed with a second malignancy
other than skin cancer or a second colon primary tumor during this
interval were also excluded.
Identification of Adjuvant Chemotherapy Use
Stage II patients who had at least one claim for chemotherapeutic
treatment, administration, or agents in Medicare files at any point
within the 3-month postoperative period were considered adjuvant
therapy recipients. Medicare claims for medical evaluation for chemotherapy (ICD-9 codes V58.1, V66.2, and V67.2), chemotherapy administration (ICD-9 code 99.25; Current Procedural Terminology codes
96408, 96410, 96412, 96414, 96520, 96530, and 96545; Health Care
Common Procedure Coding System codes Q0083 to Q0085; and
revenue center codes 0331, 0332, and 0335), and intravenous chemotherapy agents (Health Care Common Procedure Coding System codes
J9190, fluorouracil; J0640, leucovorin; and J9200, floxuridine) identified usage. Usage of experimental agents such as monoclonal antibody
17-1A could not be assessed because claims for experimental agents
are not consistently submitted to Medicare for reimbursement.
Patient Characteristics Associated With Adjuvant
Chemotherapy Use
Patient characteristics were evaluated on the basis of the information
recorded in SEER-Medicare and included sex, race, the number of
lymph nodes in the resected specimen, the year of diagnosis, the
urgency of hospital admission, and the reporting SEER registry.
Socioeconomic status was estimated on the basis of the median income
in the patient’s census tract of residence.
To examine the effect of patient comorbidity on treatment rates, we
used Romano’s modification of the Charlson comorbidity index.12,13
The diagnoses included in the Charlson-Romano comorbidity index
include myocardial infarction, diabetes, and moderate liver or renal
failure and thus capture many of the absolute or relative contraindications to administration of adjuvant chemotherapy. We examined
Medicare claims for the 15-month period extending from 12 months
before the index surgical admission until 3 months postoperatively and
assigned patients the maximal comorbidity observed.
Patients with adherent (T4) tumors, obstruction, or bowel perforation
have a prognosis that more closely resembles that of stage III patients and,
therefore, adjuvant chemotherapy use is often recommended.14-16 Thus,
we distinguished between patients with no adverse prognostic features and
those with T4N0M0 tumors, obstruction, or perforation on the basis of the
detailed staging information in the SEER registry (extent-of-disease codes
40, 45, 50, and 55) and diagnosis codes for obstruction (ICD codes 560.89
and 560.90) or perforation (ICD code 569.83) on Medicare claims. Only
those with no adverse features are included in our primary analyses.
To better understand factors influencing adjuvant treatment decisions, we also measured the frequency of medical evaluations during
the 3-month postoperative period using the specialty codes submitted
on provider claims. Claims submitted by physicians who administer
chemotherapy do not consistently identify their specialty as oncology
but may instead use the more general code for internal medicine.
Therefore, we distinguished between evaluations performed by any
physician with either primary care or cancer expertise versus those
performed by oncologists using unique physician identification numbers and HCFA specialty codes submitted on provider claims.
Statistical Analysis
We assessed the rates of chemotherapy use for patients grouped
according to each clinical or demographic characteristic. In order to
assess whether particular patient characteristics were independently
associated with treatment, we used multivariable logistic regression to
control for potential confounding by other variables. Characteristics
associated with either treatment or survival were included in the
regression model according to the categories listed in Table 1 with the
exception of tumor registry, which was not included because of the
multiple categories and small numbers of patients represented by some
registries. Odds ratios for the receipt of chemotherapy compared to a
referent category were calculated and adjusted for case mix.
To determine the association between adjuvant treatment and survival, we calculated risk ratios using a Cox proportional hazards model.
Adjusted risk ratios included variables associated with either survival
or treatment. We also used multiple logistic regression to construct
propensity scores, an index for each patient that represents the
probability of receipt of treatment, and then compared the association
between treatment and survival for patients in each propensity score
quintile. All P values are two-sided and SAS software (Version 8.0,
Downloaded from jco.ascopubs.org on September 9, 2014. For personal use only. No other uses without permission.
Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
4001
ADJUVANT CHEMOTHERAPY FOR COLON CANCER
Table 1.
The Clinical and Demographic Characteristics of the Cohort and the Percentage Treated with Adjuvant Chemotherapy
Patient Characteristics
Adjusted Odds Ratio for
Receipt of Adjuvant
Treatment*
95% CI
No. of Patients
%
% Treated with Chemotherapy
3,151
100
27
1,187
1,964
38
62
34
24
1.0
0.59
Referent
0.50-0.69
1,522
1,629
48
52
27
27
1.0
0.99
Referent
0.84-1.16
2,656
222
273
84
7
9
28
18
29
1.0
0.58
0.98
Referent
0.41-0.84
0.74-1.30
1,033
963
1,011
144
33
31
32
5
27
28
28
24
1.0
1.03
1.01
0.84
Referent
0.84-1.27
0.83-1.23
0.55-1.28
2,532
508
111
80
16
4
26
32
24
1.0
1.35
0.96
Referent
1.09-1.67
0.61-1.52
1,990
699
462
63
22
15
27
27
27
1.0
0.93
0.92
Referent
0.71-1.22
0.72-1.16
2,087
756
308
66
24
10
30
25
15
1.0
0.75
0.41
Referent
0.62-0.99
0.29-0.56
1,649
1,502
52
48
26
29
1.0
1.16
Referent
0.99-1.37
1,512
1,538
101
48
49
3
26
28
28
1.0
1.05
1.01
Referent
0.89-1.25
0.64-1.61
Total cohort
Age
65-69 years
70-75 years
Sex
Male
Female
Race
White
Black
Other
No. of examined lymph nodes
1-7
8-13
14-70
Unknown
Tumor differentiation
Well/moderate
Poor/undifferentiated
Unknown
Admission type
Elective
Urgent
Emergent
Comorbidity
0
1
2 or greater
Year of diagnosis
1991-1993
1994-1996
Median income in census tract
Lower half
Upper half
Unknown
Tumor registry
Utah
Connecticut
Hawaii
Iowa
New Mexico
San Francisco
San Jose
Los Angeles
Seattle
Atlanta
Detroit
88
437
67
390
108
158
95
215
230
138
365
16
21
28
29
18
24
28
33
32
26
32
N/A
*Odds ratios are adjusted for the variables shown in the table with the exception of tumor registry.
SAS Institute, Inc, Cary, NC) was used for all statistical analyses.
Overall survival was observed until death from any cause, for 6 years,
or until the censoring date of December 31, 1998, and is illustrated
using the Kaplan-Meier method (Fig 1).
RESULTS
Rates of Adjuvant Chemotherapy Use
Among the 3,151 stage II colon cancer patients who met
our study criteria, 860 (27%) received adjuvant chemotherapy. Among recipients, a claim specifying the drugs fluorouracil or leucovorin was available for 88%. Consistent
with the standard from clinical trials, both the median and
the mean time to initiation of adjuvant therapy was 5 to 6
weeks. The demographic and clinical characteristics of the
cohort and the percentage in each subgroup who received
adjuvant chemotherapy are listed in Table 1. Age at diagnosis was by far the strongest predictor of treatment (P ⬍
.001). White race, low comorbidity, and poorly differentiated tumor grade was each independently associated with
higher treatment rates. Sex, the number of examined lymph
nodes in the surgical specimen, the urgency of the hospital
admission, and socioeconomic status were not. There was a
Downloaded from jco.ascopubs.org on September 9, 2014. For personal use only. No other uses without permission.
Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
4002
SCHRAG ET AL
Fig 1. Illustration of the survival for the 860 cohort members who received chemotherapy within 3 months of surgical resection (blue line) and the 2,291
who did not (red line).
trend toward increasing use for patients diagnosed in more
recent years that did not achieve statistical significance.
Treatment rates varied from a low of 16% in Utah to a high
of 33% in Los Angeles and were generally higher for
patients diagnosed in urban rather than statewide registries
(Table 1).
For the 293 patients excluded from our cohort because of
the presence of a T4 tumor or other adverse prognostic
feature, the treatment rate was 33%. For patients excluded
because their age at diagnosis was over 75, treatment rates
were 16% for those aged 76 to 80 years and 7% for those
aged 81 to 85 years.
Postoperative Medical Evaluations
Ninety-one percent of cohort members had at least one
claim for an evaluation by a medical provider in the
postoperative period. This rate was 98% for chemotherapy
recipients and 90% for nonrecipients. Physician claims from
a medical oncologist were submitted for 63% of recipients
and 20% of nonrecipients. For the 994 cohort members who
had a claim for a postoperative medical oncology consultation, 54% received adjuvant treatment and 46% did not.
Association Between Treatment and Survival
The Kaplan-Meier survival curves for the patients in the
cohort are shown in Fig 1. Five-year survival was 75% for
the 2,291 untreated patients and 78% for the 860 treated
patients. The unadjusted risk ratio was 0.80 (95% confidence interval [CI], 0.68 to 0.95) but, after adjusting for
known differences between the groups, the risk ratio was
0.91 (95% CI, 0.77 to 1.09). Propensity score analysis
yielded similar results. To situate this risk ratio of 0.91 in
context, we also calculated adjusted risk ratios for the
association between other clinical characteristics and survival. Older age, male sex, black race, higher comorbidity,
and urgency of hospital admission were each independently
associated with worse survival. Neither chemotherapy treatment, nor the number of examined lymph nodes in the
surgical specimen, nor the tumor grade, nor socioeconomic
Downloaded from jco.ascopubs.org on September 9, 2014. For personal use only. No other uses without permission.
Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
4003
ADJUVANT CHEMOTHERAPY FOR COLON CANCER
Table 2.
Patient Characteristics
Risk Ratios for Mortality According to Patient Characteristics
Unadjusted Risk Ratio
95% CI
Adjusted Risk Ratio
95% CI
1.0
0.80
Referent
0.68-0.95
1.0
0.91
Referent
0.77-1.09
1.0
1.19
Referent
1.03-1.38
1.0
1.23
Referent
1.06-1.43
1.0
0.74
Referent
0.65-0.86
1.0
0.77
Referent
0.67-0.89
1.0
1.49
0.82
Referent
1.17-1.89
0.62-1.08
1.0
1.41
0.86
Referent
1.10-1.80
0.64-1.14
1.0
0.85
0.66
0.86
Referent
0.72-1.01
0.55-0.78
0.62-1.21
1.0
0.90
0.72
0.90
Referent
0.76-1.07
0.60-0.87
0.64-1.23
1.0
1.06
0.88
Referent
0.88-1.29
0.59-1.28
1.0
1.13
0.83
Referent
0.94-1.37
0.57-1.23
1.0
1.39
1.72
Referent
1.12-1.72
1.42-2.08
1.0
1.36
1.31
Referent
1.09-1.67
1.56-1.88
1.0
1.77
3.28
Referent
1.51-2.08
2.71-3.96
1.0
1.66
3.03
Referent
1.41-1.96
2.50-3.68
1.0
1.03
Referent
0.87-1.21
1.0
0.98
Referent
0.84-1.16
1.0
0.84
1.13
Referent
0.73-0.97
0.77-1.65
1.0
0.87
1.16
Referent
0.75-1.01
0.79-1.71
Adjuvant chemotherapy
No
Yes
Age
65-69 years
70-75 years
Sex
Male
Female
Race
White
Black
Other
No. of examined lymph nodes
1-7
8-13
14-70
Unknown
Tumor differentiation
Well/moderate
Poor/undifferentiated
Unknown
Admission type
Elective
Urgent
Emergent
Comorbidity
0
1
2 or greater
Year of diagnosis
1991-1993
1994-1996
Median income in census tract
Lower half
Upper half
Unknown
status, nor reporting registry was independently predictive
of survival (Table 2).
DISCUSSION
In a population-based sample of 65- to 75-year-old
Medicare enrollees diagnosed with stage II colon cancer
between 1991 and 1996, we found that 27% received
postoperative adjuvant chemotherapy. Clearly, persistent
uncertainty regarding the optimal management strategy for
stage II colon cancer has led to substantial variation in care.
Adjuvant fluorouracil is better tolerated than many other
chemotherapy regimens and there is no compelling evidence that the elderly tolerate treatment much less well than
younger patients.17 However, treatment requires 6 to 8
months of frequent outpatient medical visits and is associ-
ated with a 15% risk of grade 3 or higher diarrhea and an
8% or greater risk of leukopenia.17,18 Our analysis of
SEER-Medicare data indicates that, nevertheless, a substantial proportion of Medicare beneficiaries initiates this treatment. Therefore, close scrutiny of the physician-patient
decision-making process is warranted to determine whether
patients are cognizant of the lack of well-established survival benefit, an ongoing controversy, when they embark on
this treatment course and the threshold in survival benefit
they require to opt for therapy.
Our stage II treatment rate of 27% contrasts with the 75%
treatment rate for a parallel cohort of 65- to 75-year-olds
with stage III colon cancers.11 Similar to treatment patterns
for patients with stage III disease, we found that age at
diagnosis is strongly associated with treatment. Further-
Downloaded from jco.ascopubs.org on September 9, 2014. For personal use only. No other uses without permission.
Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
4004
SCHRAG ET AL
more, this strong age trend suggests that an even larger
proportion of patients under age 65 receive adjuvant chemotherapy. We were encouraged to note that sex and
socioeconomic status appear to play no role in treatment
decisions. As described for myriad medical conditions,
blacks received treatment at substantially lower rates than
whites.19 In contrast, treatment decisions do appear related
to basic clinical and tumor characteristics. For example,
higher levels of comorbidity and well-differentiated tumor
histology were associated with lower treatment rates.
Retrospective comparison of treated and untreated patients in the nonrandomized setting is potentially flawed as
a result of selection bias. Nonetheless, we proceeded to
compare outcomes for treated and untreated patients because this clinical dilemma is encountered by approximately
26,000 colon cancer patients in the United States each year,
and no randomized trial data are forthcoming over the short
term. We selected a homogeneous cohort of stage II patients
with similar age at diagnosis and no clearly identifiable
adverse prognostic features and used both Cox modeling
and propensity score analysis to minimize selection bias.
We found that the survival curves for chemotherapy-treated
and -untreated patients are superimposable for the first 3
postoperative years. Subsequently, survival is slightly superior for the chemotherapy-treated patients, with a 3%
increment in 5-year survival that was nonsignificant in
appropriately adjusted analyses. We emphasize the caveat
that no firm conclusion can be drawn from nonrandomized
data. Notably, our results are strikingly similar to results
from randomized trials reported by the International Multicentre Pooled Analysis of Colon Cancer Trials B2 investigators.4 In their analysis with a median age of 62 (range, 22
to 86 years), survival was 80% for untreated and 82% for
treated patients; whereas in our cohort with a median age of
69 (range, 65 to 75 years), survival was 75% for untreated
and 78% for treated patients. Thus, if indeed there is any
benefit to adjuvant chemotherapy for stage II disease, its
absolute magnitude is likely to be small, less than a 6%
absolute difference in 5-year survival. Because stage II
colon cancer poses a frequently encountered clinical conundrum and commands significant health care resources,
clinical trials that use molecular strategies in combination
with innovative chemotherapy regimens should be developed and should include patients with stage II disease.
However, we maintain that inclusion of a no-treatment arm
in these trials remains justified. Although thousands of
patients must be randomized and a long duration of follow-up is required to determine whether chemotherapy
confers a small survival benefit in stage II disease, this effort
would appear worthwhile, given that so many patients
currently receive treatment of uncertain benefit in the
off-protocol setting. Results of a clinical trial in stage II
colon cancer would also have relevance for patients in the
many countries where national health services effectively
dictate what treatment patients are permitted to receive.
We found that both chemotherapy recipients and nonrecipients generally had an evaluation by a general internist
and/or an oncologist in the immediate postoperative period
and that these visits usually occurred in the outpatient
setting. However, 63% of chemotherapy recipients and only
20% of nonrecipients had evidence of evaluation by an
oncology subspecialist. Some oncologists are characterized
according to their specialty (internal medicine) rather than
their subspecialty in the Medicare claims, and this explains
why 63% rather than 100% of chemotherapy recipients had
evidence of an oncology consultation. Nevertheless, this
pattern of postoperative evaluations yields several useful
insights. First, the majority of patients who opt not to
receive adjuvant therapy appear to make these decisions
without input from an oncologist. Whether this is driven by
surgeon or patient preferences cannot be ascertained from
our data. Second, 54% of patients who had an evaluation by
an oncologist received chemotherapy. It is noteworthy that,
subsequent to consultation with the most knowledgeable
subspecialists, so many elderly patients opt to receive an
arduous treatment that has no clearly established survival
benefit. This should lead us to carefully scrutinize the
information that is communicated in these encounters,
including such issues as whether survival benefit is addressed qualitatively or qualitatively. Although this high
treatment rate may be attributable to patient preferences
(patients who obtain oncology evaluations do so because
they desire adjuvant treatment), it is more plausible that
many United States oncologists actually endorse chemotherapy for their elderly stage II patients. This may be
because oncologists genuinely believe that treatment
confers a benefit. Economic motivations may also play a
role. When surveyed, older patients have indicated that
the primary determinant of their decisions regarding
chemotherapy is physician advice.20 However, ascertaining what really drives these decisions will require careful
research that directly observes the patient-physician decision-making process.
Several additional limitations of our analysis must be
noted. First, the potential for inaccurate coding exists for
any claims-based analysis, and clinical information available from billing records is not as detailed as that available
from chart review.21-23 Chemotherapy may be underreported in Medicare claims, compromising the sensitivity of
our approach to ascertaining treatment.10,24 Second, our
results may not generalize to the non-Medicare population
or to the approximately 16% of Medicare beneficiaries who
Downloaded from jco.ascopubs.org on September 9, 2014. For personal use only. No other uses without permission.
Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
4005
ADJUVANT CHEMOTHERAPY FOR COLON CANCER
receive care in a health maintenance organization setting,
where patterns of care may be different.25 However, because two thirds of colon cancer patients are over 65 years
of age and Medicare is the primary insurer for the vast
majority, our study represents care received by typical
United States patients.
Because fluorouracil treatment is generally well tolerated,
physicians may perceive that young patients have little to
lose by completing a course of adjuvant therapy and,
because of their long life expectancy, potentially significant
absolute gain from even a small incremental improvement
in cure rate. However, the unwillingness of some clinicians
to encourage their young patients to participate in a study
with a no-treatment control arm would not make a randomized trial unfeasible because colon cancer is primarily a
disease of the elderly. The major impediment to a trial
evaluating the benefit of adjuvant chemotherapy compared
to no treatment is practical. Accruing the necessary thousands of patients would require collaborative effort, consensus regarding the optimal treatment strategy, and long-term
follow-up, because survival differences may only become
manifest beyond 5 years. Despite these obstacles, the fact
that 27% of the United States Medicare population with
stage II colon cancer receives chemotherapy despite its
uncertain benefit suggests that we are already devoting
significant resources to this treatment strategy. Therefore,
despite the size, costs, and long duration required to yield
mature results, efforts to obtain a definitive answer to
whether or not adjuvant chemotherapy is beneficial in
stage II colon cancer are still warranted.
ACKNOWLEDGMENT
The acknowledgment is available online at www.jco.org.
REFERENCES
1. Moertel CG, Fleming TR, Macdonald JS, et al: Levamisole and
fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl
J Med 322:352-358, 1990
2. Laurie JA, Moertel CG, Fleming TR, et al: Surgical adjuvant
therapy of large-bowel carcinoma: An evaluation of levamisole and the
combination of levamisole and fluorouracil—The North Central Cancer
Treatment Group and the Mayo Clinic. J Clin Oncol 7:1447-1456, 1989
3. Efficacy of adjuvant fluorouracil and folinic acid in colon cancer:
International Multicentre Pooled Analysis of Colon Cancer Trials
(IMPACT) investigators. Lancet 345:939-944, 1995
4. Efficacy of adjuvant fluorouracil and folinic acid in B2 colon
cancer—International Multicentre Pooled Analysis of B2 Colon Cancer
Trials (IMPACT B2) investigators. J Clin Oncol 17:1356-1363, 1999
5. Mamounas E, Wieand S, Wolmark N, et al: Comparative efficacy
of adjuvant chemotherapy in patients with Dukes’ B versus Dukes’ C
colon cancer: Results from four National Surgical Adjuvant Breast and
Bowel Project adjuvant studies (C-01, C-02, C-03, and C-04). J Clin
Oncol 17:1349-1355, 1999
6. Harrington DP: The tea leaves of small trials. J Clin Oncol
17:1336-1338, 1999
7. Benson AB III, Choti MA, Cohen AM, et al: NCCN Practice
Guidelines for Colorectal Cancer. Oncology (Huntingt) 14:203-212, 2000
8. About SEER. Http://seer.cancer.gov/about (accessed verified August 13, 2002)
9. Percy C, Van Holton V, Muir C: International Classification of
Diseases for Oncology, 2nd ed. Geneva, World Health Organization, 1990
10. Potosky AL, Riley GF, Lubitz JD, Mentnech RM, Kessler LG:
Potential for cancer related health services research using a linked
Medicare-tumor registry database. Med Care 31:732-748, 1993
11. Schrag D, Cramer LD, Bach PB, Begg CB: Age and adjuvant
chemotherapy use after surgery for stage III colon cancer. J Natl Cancer
Inst 93:850-857, 2001
12. Romano PS, Roos LL, Jollis JG: Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: Differing
perspectives. J Clin Epidemiol 46:1075-1090, 1993
13. Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new
method of classifying prognostic comorbidity in longitudinal studies:
Development and validation. J Chronic Dis 40:373-383, 1987
14. Peeters M, Haller DG: Therapy for early-stage colorectal cancer.
Oncology (Huntingt) 13:307-321, 1999
15. Moore HC, Haller DG: Adjuvant therapy of colon cancer. Semin
Oncol 26:545-555, 1999
16. Fuchs CS, Mayer RJ: Adjuvant chemotherapy for colon and
rectal cancer. Semin Oncol 22:472-487, 1995
17. Sargent DJ, Goldberg RM, Jacobsen SD, et al: A pooled analysis
of adjuvant chemotherapy for resected colon cancer in elderly patients.
N Engl J Med 345:1091-1097, 2001
18. Rothenberg ML, Meropol NJ, Poplin EA, Van Cutsem E,
Wadler S: Mortality associated with irinotecan plus bolus fluorouracil/
leucovorin: Summary findings of an independent panel. J Clin Oncol
19:3801-3807, 2001
19. Bach PB, Cramer LD, Warren JL, Begg CB: Racial differences
in the treatment of early-stage lung cancer. N Engl J Med 341:11981205, 1999
20. Newcomb PA, Carbone PP: Cancer treatment and age: Patient
perspectives. J Natl Cancer Inst 85:1580-1584, 1993
21. Hsia DC, Krushat WM, Fagan AB, Tebbutt JA, Kusserow RP:
Accuracy of diagnostic coding for Medicare patients under the prospective-payment system. N Engl J Med 318:352-355, 1988
22. Iezzoni LI: Assessing quality using administrative data. Ann
Intern Med 127(8 Pt 2):666-674, 1997
23. Lloyd SS, Rissing JP: Physician and coding errors in patient
records. JAMA 54:1330-1336, 1985
24. Du X, Freeman JL, Warren JL, Nattinger AB, Zhang D,
Goodwin JS: Accuracy and completeness of Medicare claims data for
surgical treatment of breast cancer. Med Care 38:719-727, 2000
25. Riley GF, Potosky AL, Klabunde CN, Warren JL, BallardBarbash R: Stage at diagnosis and treatment patterns among older
women with breast cancer: An HMO and fee-for-service comparison.
JAMA 281:720-726, 1999
Downloaded from jco.ascopubs.org on September 9, 2014. For personal use only. No other uses without permission.
Copyright © 2002 American Society of Clinical Oncology. All rights reserved.