Adjuvant chemotherapy improves survival in patients with American Joint Committee on Cancer stage II colon cancer


  • Presented in part at the Annual Scientific Meeting of the American Society of Colon and Rectal Surgeons; May 15-19, 2010; Minneapolis, Minnesota.



It is unclear whether the administration of adjuvant chemotherapy improves survival in patients with American Joint Committee on Cancer (AJCC) stage II colon cancer.


The authors used the State of California Cancer Surveillance Program (CSP) to assess patients ages 18 to 80 years with AJCC stage II colon cancer (ie, T3 or T4 and N0) who underwent surgical resection during 1991 and 2006. Patients who had rectal and rectosigmoid cancers were excluded. The cohort was stratified according to the receipt of adjuvant chemotherapy, and clinical and pathologic characteristics and outcomes were assessed.


From the CSP data, 3716 patients were identified who underwent curative-intent surgical resection for stage II colon cancer. When the 2 treatment groups (surgery plus adjuvant chemotherapy [n = 916] and surgery alone [n = 2800]) were compared, patients who received adjuvant chemotherapy were more likely to be younger and to have left-sided lesions with ≥12 lymph nodes examined. There was no difference in sex or tumor differentiation between the 2 groups. According to a Kaplan-Meier analysis, patients who received adjuvant chemotherapy had improved overall survival compared with patients who underwent surgery alone (median survival, 12 years vs 9.2 years, respectively; P < .001). In multivariate analysis, adjuvant chemotherapy was identified as an independent predictor of improved survival (hazard ratio, 0.88; 95% confidence interval, 0.78-0.99; P = .031).


To the authors' knowledge, this is the first population-based analysis to identify a survival advantage for adjuvant chemotherapy in patients with AJCC stage II colon cancer. On the basis of the current findings, the authors concluded that the administration of adjuvant chemotherapy improves survival in select patients with stage II disease. Cancer 2011;. © 2011 American Cancer Society.

In the United States, colon cancer occurs in approximately 106,000 patients each year and accounts for >50,000 deaths annually.1 Among those with newly diagnosed disease, 25% to 40% of patients will present with disease that extends beyond the muscularis propria but without lymph node involvement or metastatic disease.2, 3 These patients with American Joint Committee on Cancer (AJCC) stage II disease have a good prognosis with 5-year survival rates that range from 72% to 85%.3 Nevertheless, approximately 25% of patients with stage II colon cancer ultimately develop recurrent disease, and most die of their disease.2, 3

Although adjuvant chemotherapy has become the standard of care for patients who have lymph node-positive colon cancer (ie, AJCC stage III), the role of adjuvant chemotherapy for patients with stage II disease remains controversial. Previous randomized controlled trials, pooled analyses, meta-analyses, and population-based analyses of patients with stage II disease were underpowered or improperly designed, leading to an inadequate evaluation of adjuvant chemotherapy for these patients.4-10 Indeed, the evidence to support adjuvant chemotherapy for patients with stage II disease has been drawn from clinical trials that evaluated patients with stage II and III disease.5, 11, 12 Other indirect support for adjuvant chemotherapy derives from studies that evaluated select clinicopathologic characteristics, including tumor location, tumor differentiation, lymphovascular or perineural invasion, obstruction, perforation, carcinoembryonic antigen levels, and extent of lymph node retrieval.2, 13-18 Aside from those reports, there is no direct evidence that adjuvant chemotherapy provides a survival benefit for patients who have stage II colon cancer with defined high-risk features.7

The objective of the current study was to evaluate the role of adjuvant chemotherapy in patients with AJCC stage II colon cancer within a large, heterogeneous population. By using a population-based cancer registry, we sought to overcome the limitations of previous smaller cohort studies to determine whether adjuvant chemotherapy provided a survival benefit in these patients.


Cancer Surveillance Program for Los Angeles County

The California Cancer Registry's Cancer Surveillance Program (CSP) for Los Angeles County is a population-based cancer registry. CSP captures data on nearly all cancer diagnoses, and the collected data are focused on cancer diagnosis and treatment with the stated aim of improving cancer care in the State of California. This focus on comprehensive data for cancer therapy (ie, chemotherapy, radiotherapy, and intent-to-cure surgery) distinguishes CSP from many other regional and national databases.

CSP reporting for colon cancer location, histology, staging, and differentiation are based on the International Classification of Diseases for Oncology (ICD-O). The ICD-O colon adenocarcinoma histology codes used in our study included codes 8140 through 8147, 8210 through 8211, 8220 through 8221, 8260 through 8263, 8480 through 8481, and 8570 through 8576. CSP staging of disease was reported as localized, regional (extension only, lymph nodes only, extension and lymph nodes, not otherwise specified [NOS]), and distant. Tumor grade was categorized by CSP as well differentiated, moderately differentiated, poorly differentiated, or undifferentiated. Topography codes were reported as ascending colon (C18.0-C18.3), transverse colon (C18.4), and descending and sigmoid colon (C18.5-C18.7). Overlapping (C18.8), NOS (C18.9), and rectosigmoid tumors (C19.9) were excluded from this analysis.

Cancer-specific therapeutic data have been available from CSP since 1988; but the first year of our investigation was 1991, when adjuvant chemotherapy initially was recommended.19 Chemotherapy was classified as negative (none, recommended, not given, or refused) or positive (single agent, multiple agent, or NOS). Patients who received radiation therapy (beam, implants, isotopes, combination, or NOS) were excluded from this investigation. All patients underwent curative-intent surgical therapy defined by CSP codes 30-80 (partial colectomy, hemicolectomy, subtotal colectomy, resection with continuous organ, colectomy with en bloc resection of other organs and pelvic exenteration, total colectomy, total proctocolectomy, and colectomy NOS). Patients with postoperative mortality within 30 days of the operation were excluded from analysis.

Statistical Analysis

The primary endpoint was to determine whether the addition of chemotherapy to curative surgical resection conferred a survival benefit to patients with AJCC stage II disease. Only patients who were diagnosed between 1991 and 2006 who were diagnosed with AJCC stage II disease (ie, tumor classification 3 [T3] or T4 and negative lymph node status [N0]) were included for analysis. These patients were reported to CSP as having regional disease with extension only, negative lymph nodes (N0), and without metastatic disease (M0). Patients were grouped by receipt of adjuvant chemotherapy, and characteristics were compared using chi-square analyses. Analyzed factors included age (ages ≤50 years, 51-64 years, or ≥65 years), sex (man or woman), tumor location (ascending, transverse, or descending/sigmoid colon), grade (well, moderately, or poorly differentiated or undifferentiated), and the number of lymph nodes examined (1-11 lymph nodes or ≥12 lymph nodes). Overall survival was assessed for the 2 groups using the Kaplan-Meier method and then compared using the log-rank test. Logistic regression was performed with the Cox proportional hazards model and was reported using hazard ratios with 95% confidence intervals to assess the receipt of chemotherapy with overall survival while controlling for the other factors listed above. All reported P values were 2-sided, and P < .05 was considered statistically significant.


Patient Characteristics

During the study period, 3716 patients underwent curative-intent colon resection for pathologically confirmed stage II colon cancer. The majority of patients were aged ≥65 years (66%) and had moderately differentiated tumors (73%). Most patients (75%) did not receive adjuvant chemotherapy (Table 1). When grouped according to the receipt of adjuvant chemotherapy, patients who received chemotherapy were disproportionately younger and had larger tumors (>5 cm), and their tumors were located more frequently in the descending or sigmoid colon (Table 2). It is noteworthy that patients who received adjuvant chemotherapy were more likely to have ≥12 lymph nodes examined.

Table 1. Characteristics of Patients With American Joint Committee on Cancer Stage II Colon Cancer
CharacteristicNo. of Patients (%)
Age, y 
 ≤50370 (10)
 51-64898 (24.2)
 ≥652448 (65.9)
 Men1908 (51.3)
 Women1808 (48.7)
Tumor location 
 Ascending colon1815 (48.8)
 Transverse colon428 (11.5)
 Descending/sigmoid colon1473 (39.6)
Grade of tumor differentiation 
 Well224 (6)
 Moderate2720 (73.2)
 Poor692 (18.6)
 Undifferentiated21 (0.6)
 Unknown59 (1.6)
Tumor size, cm 
 ≤51834 (49.4)
 >51558 (41.9)
 Unknown324 (8.7)
No. of lymph nodes examined 
 1-112048 (55.1)
 ≥121668 (44.9)
 No2800 (75.3)
 Yes916 (24.7)
Table 2. Comparison of Patient Characteristics Stratified According to the Receipt of Adjuvant Chemotherapy
 No. of Patients (%) 
CharacteristicNo Chemotherapy, N=2800Chemotherapy, N=916P
Age, y   
 ≤50210 (7.5)160 (17.5)<.001
 51-64591(21.1)307 (33.5) 
 ≥651999 (71.4)449 (49) 
 Men1441 (51.5)467 (51).800
 Women1359 (48.5)449 (49) 
Tumor location   
 Ascending colon1408 (50.3)407 (44.4).002
 Transverse colon326 (11.6)102 (11.1) 
 Descending/sigmoid colon1066 (38.1)407 (44.4) 
Grade of tumor differentiation   
 Well174 (6.2)50 (5.5).259
 Moderate2067 (73.8)653 (71.3) 
 Poor502 (17.9)190 (20.7) 
 Undifferentiated16 (0.6)5 (0.5) 
 Unknown41 (1.5)18 (2) 
Tumor size, cm   
 ≤51421 (50.8)413 (45.1).007
 >51134 (40.5)424 (46.3) 
 Unknown245 (8.8)79 (8.6) 
No. of lymph nodes examined   
 1-111579 (56.4)469 (51.2).006
 ≥121221 (43.6)447 (48.8) 

Survival Analysis

When the 2 experimental cohorts were compared, the administration of adjuvant chemotherapy was associated with a significant improvement in overall survival (median survival [MS], 12 years vs 9.2 years; P < .001; 5-year survival rate, 74% vs 66%; P < .001) (Fig. 1). Univariate analysis was then performed to identify predictors of survival in the entire cohort (Table 3). Advancing age was associated with poorer survival; whereas ≥12 lymph nodes examined and receipt of adjuvant chemotherapy were associated with improved survival. Next, multivariate analysis was performed to identify independent predictors of survival. Our analysis demonstrated that the receipt of adjuvant chemotherapy was a significant prognostic factor for improved survival (hazard ratio, 0.88; 95% confidence interval, 0.78-0.99; P = .031). Age, sex, and the number of lymph nodes examined also were identified as independent predictors of improved survival.

Figure 1.

This chart illustrates the overall survival of patients with American Joint Committee on Cancer stage II colon cancer who underwent surgical resection. Improved survival was observed for patients who received chemotherapy compared with those who received no chemotherapy (median survival, 12 years vs 9.2 years, respectively; P < .001).

Table 3. Univariate and Multivariate Analysis for the Entire Cohort
   Univariate AnalysisMultivariate Analysis
VariableNo. of Patients (%)Median Survival, yHR95% CIPHR95% CIP
  1. Abbreviations: CI, confidence interval; HR, hazard ratio; LNs, lymph nodes; NR, not reported.

Age, y        
 ≤50370 (10)NR1.01.0
 51-64898 (24)<.0011.861.43-2.43<.001
 ≥652448 (66)<.0013.672.87-4.70<.001
 Men1908 (51)
 Women1808 (49)
Tumor location        
 Ascending colon1815 (49)9.41.0   
 Transverse colon428 (12)   
 Descending/sigmoid colon1473 (40)   
Grade of tumor differentiation        
 Well224 (6)
 Moderate2720 (73)
 Poor692 (19)
 Undifferentiated21 (0.6)
 Unknown59 (1.6)
Tumor size, cm        
 ≤51834 (49)9.71.0   
 >51558 (42)   
 Unknown324 (8.7)   
No. of LNs examined        
 1-112048 (55)
 ≥121668 (45)<.0010.760.69-0.84<.001
 No2800 (75)
 Yes916 (25)<.0010.880.78-0.99.031


The excellent overall prognosis of patients with stage II colon cancer has made the evaluation of adjuvant therapy for this stage of disease difficult. The Adjuvant Colon Cancer Endpoints (ACCENT) registry, a pooled patient registry of 18 adjuvant colon cancer trials, evaluated the follow-up necessary to determine a clinically relevant survival benefit for patients with stage II colon cancer.20, 21 In the analysis, de Gramont et al concluded that future trials evaluating adjuvant chemotherapy for stage II colon cancer should incorporate a minimum follow-up of 6 to 7 years.21 Evaluating our large Los Angeles County cohort over a 15-year study period, we observed a significant improvement in overall survival in favor of adjuvant chemotherapy.

It is not without surprise that previous trials have had difficulty identifying an improvement in overall survival in favor of adjuvant chemotherapy. In fact, current investigations regarding the role of chemotherapy in stage II disease have come primarily from either subset analyses of randomized trials that included both stage II and stage III disease or pooled analyses of those trials. The National Surgical Adjuvant Breast and Bowel Project (NSABP) has conducted 7 randomized trials and at least 2 pooled analyses of these trials to determine the role of adjuvant chemotherapy for stage II colon cancer.5, 11, 12, 22-24 Each of those trials enrolled both stage II patients and stage III patients and evaluated different chemotherapeutic regimens. Several of the trials did not contain a surgery-alone cohort, and none were powered adequately to perform a subset analysis of stage II patients. Although the pooled analyses have suggested that the relative benefit of chemotherapy is similar between patients with stage II disease and stage III disease,5, 11, 12 the lack of a clear survival advantage for the stage II cohorts along with the heterogeneity of the trial designs has tempered the enthusiasm for the pooled results. Indeed, only 1 randomized controlled trial to date that included both stage II and stage III patients has reported a significant improvement in overall survival for stage II patients (n = 468). In that study, Taal et al reported a significant 8% improvement in overall survival for patients with stage II disease.25 However, the study has been criticized for including patients with rectal cancer and for closing before reaching accrual.

There have been 3 randomized controlled trials evaluating adjuvant chemotherapy exclusive to patients with stage II disease: the Austrian Breast and Colorectal Cancer Study Group (ABCCSG) trial,9 the Intergroup 0035 trial,6 and the QUASAR Collaborative Group trial.8 Unfortunately, those trials were either powered insufficiently to detect an improvement in 5-year overall survival or they suffered from inconsistencies in study design that weakened the analyses. For example, the ABCCSG trial was designed to detect a 10% difference in overall survival; and was not powered to detect smaller improvements in survival.9 The Intergroup 0035 trial reported improved recurrence-free-survival but failed to identify a significant benefit in 5-year overall survival.6 The largest of these studies, the Quasar study, enrolled 2963 patients but was limited by the inclusion of patients who had rectal cancer with a relatively short median follow-up (<6 years).8 Although there was a 5-year survival benefit of 3.6%, it was exclusive to the rectal cancer cohort and failed to reach significance in the subgroup analysis of patients with colon cancer.

On the basis of these 3 trials and similar results from recent meta-analyses, notably, the International Multicenter Pooled Analysis of B2 Colon Cancer Trials (IMPACT B2),4, 16, 26 the American Society of Clinical Oncology (ASCO) and the National Comprehensive Cancer Network (NCCN) both have recommended against the routine use of chemotherapy for patients with stage II colon cancer.27, 28 Despite this paucity of data, both ASCO and the NCCN nevertheless recommend that patients who have high-risk disease features be considered for adjuvant chemotherapy and that a detailed discussion be held with the patient regarding the potential risks of adjuvant chemotherapy with uncertain clinical benefits. Our population-based analysis supports the use of chemotherapy in patients with stage II disease without consideration for high-risk disease characteristics. The NCCN and ASCO guidelines define high-risk features as obstruction, perforation, lymphovascular invasion, poor differentiation, and T4 disease.27, 28 These guidelines, however, do not incorporate other prognostic factors, such as perineural invasion, tumor location, or carcinoembryonic antigen.2, 13-18

Our investigation has the major strengths of evaluating an adequate number of patients over a 15-year period to adequately perform this analysis. This 15-year time period of our study occurred primarily during an era when the predominant systemic therapy was 5-fluorouracil and levamisole or leucovorin. We postulate that the observed survival difference between the 2 study cohorts could have been more pronounced if contemporary systemic chemotherapy agents had been used. In contrast, there are limitations to our study that warrant reader consideration. Our study was a retrospective analysis susceptible to selection bias. For example, there were differences between the 2 study cohorts, particularly the disparities in age, suggesting that patients with better performance status were offered chemotherapy. However, the current guidelines for adjuvant chemotherapy in stage II disease are inherently biased, because the recommendations are given only for patients with high-risk disease. Unfortunately, we could not address the frequency of all high-risk or favorable prognostic factors (eg, microsatellite instability) in our study cohort. Many patients in our cohort had inadequate lymph node retrieval, as recommended under current national guidelines,27, 28 but the identification of high-risk factors, including low lymph node count, may not be necessary to direct therapy, as demonstrated by a recent pooled analysis of NSABP studies C-01 through C-05.11

Pending large, multicenter, randomized controlled trials with adequate follow-up, the decision to offer adjuvant chemotherapy to patients with stage II disease relies on the evidence from pooled or population-based analyses. Our investigation of adjuvant chemotherapy in the large, heterogeneous Los Angeles County population supports the use of adjuvant chemotherapy to improve overall survival in select patients with stage II colon cancer.


No specific funding was disclosed.


The authors made no disclosures.