Fax: (570) 271 6542
Efficacy and toxicity of adjuvant chemotherapy in elderly patients with colon carcinoma
A 10-year experience of the Geisinger Medical Center
Article first published online: 28 MAR 2002
Copyright © 2002 American Cancer Society
Volume 94, Issue 7, pages 1931–1938, 1 April 2002
How to Cite
Fata, F., Mirza, A., Craig Wood, G., Nair, S., Law, A., Gallagher, J., Ellison, N. and Bernath, A. (2002), Efficacy and toxicity of adjuvant chemotherapy in elderly patients with colon carcinoma. Cancer, 94: 1931–1938. doi: 10.1002/cncr.10430
- Issue published online: 28 MAR 2002
- Article first published online: 28 MAR 2002
- Manuscript Accepted: 29 OCT 2001
- Manuscript Revised: 28 OCT 2001
- Manuscript Received: 8 MAY 2001
- adjuvant chemotherapy;
- colon carcinoma;
- elderly patients;
Although the benefit from adjuvant chemotherapy has been established clearly in patients with Stage III colon carcinoma, the degree to which elderly patients with colon carcinoma can tolerate such therapy generally has remained unknown.
The authors reviewed all patients in their Tumor Registry with Stage II and Stage III adenocarcinoma of the colon who underwent potentially curative resection for their disease at the Geisinger Medical Center between January 1990 and September 2000. One hundred twenty patients underwent complete resection of their colon carcinoma and received 5-fluorouracil-based (5-FU) adjuvant chemotherapy.
The 5-year disease free survival rate for patients age ≥ 65 years (Group A) was 70% compared with 56% for patients age < 65 years (Group B) (P = 0.085). The 5-year overall survival rate for patients in Group B was 77% compared with 62% for the patients in Group A (P = 0.143). In a Cox regression model, age was not a predictor of disease free survival (P = 0.633) or overall survival (P = 0.900) when it was analyzed as a continuous variable. Only 19 patients were age > 75 years, and the disease free and overall survival rates for this group were similar but were underpowered compared with the rates for the patients ages between 65–75 years. When gender and disease stage were included in the model, age remained a nonsignificant variable (P = 0.400 for disease free survival; P = 0.615 for overall survival). Nine of 56 patients in Group A (16%) experienced Grade 3–4 toxicity compared with 14 of 64 patients in Group B (22%) (P = 0.420). The lack of a correlation between toxicity and age was maintained after controlling for disease stage and patient gender (P = 0.343). There were no correlations between preoperative carcinoembryonic antigen level, tumor grade, or lymph node involvement and patient age (P = 0.258, P = 0.256, and P = 0.519, respectively).
Elderly patients with Stage II and Stage III colon carcinoma benefit from 5-FU-based adjuvant therapy without a significant increase in toxicity compared with their younger counterparts. Adjuvant chemotherapy should be presented to elderly patients with high-risk, resected colon carcinoma. The data regarding age cannot be generalized to patients age > 75 years. Cancer 2002;94:1931–8. © 2002 American Cancer Society.
The median age of patients at the onset of colon carcinoma is 60–65 years, although the disease is seen increasingly in patients age ≥ 65 years.1 It has been demonstrated that adjuvant chemotherapy can reduce the risk of disease recurrence and death in patients with Stage III colon carcinoma by approximately 40% and 33%, respectively.2 Conversely, if there is a benefit from the currently used adjuvant chemotherapy in patients with Stage II colon carcinoma, then it is small in absolute magnitude.3 Although adjuvant chemotherapy is considered the standard of care for patients with high-risk, resected colon carcinoma, there are concerns regarding its relative efficacy and toxicity when it is used as a treatment for the elderly.
In a 1993 review,4 the calculated early dose intensity of chemotherapy for patients with colon carcinoma was lower in patients age ≥ 75 years compared with patients age < 75 years. In a more recent report, patients with colon carcinoma age < 50 years were more likely to receive recommended adjuvant chemotherapy compared with patients age > 75 years.5 Among all patients with colon carcinoma in the National Cancer Data Base,6 the use of surgery plus chemotherapy declined with age: 40% of patients age < 50 years received both treatments, compared with 20% of patients age 70–79 years. Comparable results have been reported elsewhere.7–9 One study that evaluated why adjuvant chemotherapy may not be offered to elderly patients with colon carcinoma suggested that many patients and their physicians simply did not support this treatment for the elderly.10
Other studies have suggested that elderly patients often do not achieve the intended target dose of adjuvant chemotherapy4 and that adjuvant chemotherapy for colon carcinoma in elderly patients is not considered a viable option.11 There also are concerns of increased toxicity from adjuvant chemotherapy in elderly patients with colon carcinoma.12 In this article, we report the results of a retrospective review of our experience with the safety and efficacy of adjuvant chemotherapy in elderly patients with colon carcinoma.
MATERIALS AND METHODS
We reviewed our Tumor Registry to identify all patients with Stage II and Stage III adenocarcinoma of the colon who underwent curative resection at the Geisinger Medical Center between January 1990 and September 2000. Chart audits were performed on 120 patients who underwent potentially curative resection for colon carcinoma and received 5-fluorouracil (5-FU)-based adjuvant chemotherapy. The study was reviewed and approved by the Geisinger Institutional Research Review Board.
Patient characteristics are listed in Table 1. Twenty-nine patients had Stage II colon carcinoma, and 91 patients had Stage III colon carcinoma. Fifty-six patients were age < 64 years, and 63 patients were age ≥ 65 years. There were 71 men and 49 women. Data on type of surgery, degree of tumor differentiation, status of lymph node metastasis, and preoperative carcinoembryonic antigen (CEA) levels were collected to determine their association with patient age and survival. Approximately 50% of patients had right-sided colon carcinoma and, thus, underwent right hemicolectomy. The majority of patients had intermediate grade tumors, and one-third of patients diagnosed with high-grade tumors.
|Patient characteristics||Stage II (n = 29 patients)||Stage III (n = 91 patients)|
|Pre-op CEA (ng/mL)|
|Lymph node involvement (no.)|
|5-FU + LV||15||38|
|5-FU + LEV||11||29|
|5-FU + LV + LEV||1||19|
Fifty-three patients received the Mayo Clinic 5-FU plus leucovorin daily regimen for 6 months; 40 patients received 5-FU plus levamisole for 1 year; 20 patients received 5-FU plus leucovorin and levamisole according to the Intergroup Study INT-0089 protocol; and 7 patients received other combinations from research protocols, including 5-FU plus irinotecan or oxaliplatin. In the INT-0089 study protocol, the inclusion criteria were the same for both younger patients and older patients, including patients with histologically confirmed adenocarcinoma of the colon who underwent curative resection. Patients with Stage II (Duke Stage B2) or Stage III (Duke Stage C1 or C2) chemotherapy-naïve colon carcinoma were eligible. Patients with multiple primary colorectal tumors were entered on the study and were stratified according to their most advanced stage tumor. The toxicity of chemotherapy, including bone marrow suppression, diarrhea, mucositis, and cutaneous toxicity, was recorded using the Common Toxicity Criteria of the National Cancer Institute. Toxicities from chemotherapy are listed in Table 2. More than 20% dose reductions of the chemotherapy also were reported. All patients were followed at the Geisinger Medical Center, and dates of disease recurrence and death were recorded to determine the disease free and overall survival rates.
|Stage||Gender||Age < 65 yrs||Age ≥ 65 yrs||P value|
|No. of patients (%)||5-yr survival (%)||No. of patients (%)||5-yr survival (%)|
|II||M&F||3 of 11 (27)||90||2 of 18 (11)||88||0.912|
|III||M&F||18 of 45 (40)||54||11 of 46 (24)||74||0.146|
|II–III||M||10 of 31 (32)||78||8 of 39 (21)||75||0.892|
|II–III||F||11 of 25 (44)||35||5 of 25 (20)||83||0.050|
|II–III||M&F||21 of 56 (38)||62||13 of 64 (20)||77||0.143|
|Disease free survival|
|II||M&F||3 of 11 (27)||80||2 of 18 (11)||88||0.606|
|III||M&F||20 of 45 (44)||50||13 of 46 (28)||65||0.115|
|II–III||M||11 of 31 (35)||68||9 of 39 (23)||72||0.668|
|II–III||F||12 of 25 (27)||37||6 of 25 (24)||61||0.042|
|II–III||M&F||23 of 56 (41)||56||15 of 64 (23)||70||0.085|
Study Objectives and Statistical Analysis
The primary objective of the study was to compare the toxicity levels, disease free survival rate, and overall survival rate of patients age ≥ 65 years with Stage II and III colon carcinoma who received adjuvant chemotherapy compared with their younger counterparts (age < 65 years). Maximum toxicity was defined as the highest toxicity grade for the different toxicity variables (gastrointestinal, mucositis, hematologic, cutaneous, and others). Toxicity levels were compared between patients age < 65 years and patients age ≥ 65 years using a Cochran–Armitage exact trend test. A secondary analysis compared the percent of patients who required dose reductions in the two age groups. This outcome was analyzed using a chi-square test.
Survival was estimated by using the Kaplan–Meier product-limit method. Statistical differences between survival curves were assessed with a log-rank test. Disease stage and gender also were considered when examining the correlations between survival and age.
The secondary objective of the study was to compare preoperative CEA level, tumor grade, and lymph node status between patients in the two age groups. These variables were compared between patients age < 65 patients with patients age ≥ 65 years using the Fisher exact test (tumor grade) and a chi-square test (preoperative CEA level and lymph node status).
Disease Free and Overall Survival
There was a persistent trend, which was not always statistically significant, among patients age ≥ 65 to have improved disease free survival and overall survival. This correlation was strongest for patients with Stage III disease and for female patients. Table 2 contains a summary of comparisons by age group of disease free and overall survival rates by stage and gender. The 5-year survival rate was determined by using Kaplan–Meier survival curves. The 5-year disease free and overall survival rates for elderly patients (age ≥ 65 years) with Stage II and III colon carcinoma were not statistically different from their younger counterparts (age < 65 years).
In a Cox regression model, age was not a predictor of disease free survival (P = 0.633) or overall survival (P = 0.900) when it was analyzed as a continuous variable. When gender and stage were included in the model, age remained a nonsignificant variable (P = 0.400 for disease free survival; P = 0.615 for overall survival). Gender was not a predictor of clinical outcome (P = 0.176 for disease free survival; P = 0.149 for overall survival); however, as expected, disease free survival and overall survival were improved slightly in patients with Stage II colon carcinoma compared with patients who had Stage III disease (P = 0.070 and P = 0.126, respectively). Figures 1–3 show the disease free and overall survival curves for patients with Stage II and/or Stage III colon carcinoma age ≥ 65 years compared with their younger counterparts age < 65 years.
Nine of 56 patients (16%) age < 65 years experienced Grade 3 or 4 toxicity compared with 14 of 64 patients (22%) age ≥ 65 years (P = 0.420). Twelve of 56 patients (21%) age < 65 years had no toxicity (Grade 0 for all toxicity types) compared with 13 of 64 patients (20%) age ≥ 65 years. Table 3 shows the distribution and types of toxicity experienced for this population by age group. Both age groups experienced similar toxicity.
|Toxicity type||Age < 65 yrs (%) (n = 56 patients)||Age ≥ 65 yrs (%) (n = 64 patients)|
|Grade 0||39 (70)||36 (56)|
|Grade 1–2||14 (25)||25 (39)|
|Grade 3–4||3 (5)||3 (5)|
|Grade 0||24 (44)||23 (36)|
|Grade 1–2||27 (49)||39 (61)|
|Grade 3–4||4 (7)||2 (3)|
|Grade 0||23 (41)||22 (34)|
|Grade 1–2||29 (52)||34 (53)|
|Grade 3–4||4 (7)||8 (13)|
|Grade 0||42 (75)||53 (83)|
|Grade 1–2||12 (21)||10 (16)|
|Grade 3–4||2 (4)||1 (2)|
|Grade 0||50 (89)||57 (89)|
|Grade 1–2||0 (0)||1 (2)|
|Grade 3–4||0 (0)||2 (3)|
|Unknown||6 (11)||4 (6)|
Table 4 displays maximum toxicity grade and age, which were not related statistically (P = 0.446; Cochran–Armitage exact trend test). The lack of a correlation was maintained after controlling for disease stage and patient gender (P = 0.343).
|Maximum toxicity grade||Age||Total|
|< 65 yrs||≥ 65 yrs|
Chemotherapy Dose Reductions and Dose Intensity
Seventeen of 56 patients (31%) age < 65 years underwent at least one level of dose reduction. Twenty of 64 patients (31%) age ≥ 65 years also underwent at least one level of dose reduction. This correlation was not significant statistically (P = 0.968). Once again, controlling for gender and disease stage did not affect this correlation. In summary, these data do not support the existence of a correlation between toxicity and age.
The older and the younger age groups did not differ in the median number of chemotherapy cycles or dose intensity. Both groups received a median of 6 cycles, with a minimum of 1 cycle and a maximum of 12 cycles (P = 0.896; Wilcoxon rank-sum test). Fifty-nine percent of patients in the younger age group and 60% of patients in the older age group had a dose intensity < 100%. The chemotherapy dose intensity was defined as the ratio of the dose planned to the dose delivered. Both groups had a median dose intensity of 100% and a minimum of dose intensity 16% (P = 0.842; Wilcoxon rank-sum test).
Correlation of Preoperative CEA Level, Tumor Grade, and Lymph Node Status with Patient Age
Preoperative CEA levels were available for 35 patients age < 65 years and for 31 patients age ≥ 65 years. Eight patients (42%) in the younger cohort had preoperative CEA levels > 5 ng/mL, and 11 patients (58%) in the older cohort had preoperative CEA levels > 5 ng/mL. The correlation between preoperative CEA level and age was not significant statistically (P = 0.258). Controlling for gender and disease stage did not affect this correlation. These data did not support a correlation between preoperative CEA level and patient age.
Among the patients with Stage III colon carcinoma age < 65 years, 18% had more than 4 lymph nodes involved with metastatic disease (Duke Stage C2). Conversely, 24% of patients with Stage III colon carcinoma age ≥ 65 years had > 4 lymph nodes involved. This correlation was not significant statistically (P = 0.519; chi-square test). Of 42 patients age < 65 years with valid tumor grade values, 1 patient (2%) had a low-grade tumor, 32 patients (76%) had intermediate grade tumors, and 9 patients (21%) had high-grade tumors. Of 59 patients age ≥ 65 years with valid tumor grade values, 0 patients (0%) had low-grade tumors, 40 patients (68%) had intermediate grade tumors, and 19 patients (32%) had high-grade tumors. This correlation was not significant statistically (P = 0.256; Fischer exact test). These data did not support a correlation between tumor grade and patient age or between lymph node status and patient age.
Over the past 2 decades, most studies of adjuvant chemotherapy in patients with Stage II or III colon carcinoma stratified patients by age using a cut-off age between 60 years and 70 years (Table 5). Patient age at the time of diagnosis was related inversely to survival in patients with colon carcinoma. Patients in younger age groups had a more favorable prognosis,13 and gender had no appreciable effect on survival. In our review, approximately 50% of patients were age ≥ 65 years, and, as expected, 42% of the patients were female.
|Study||No. of patients||Cut-off age (yrs)||Chemotherapy||Age association with survival (P value)|
|Wolmark et al.18 (NSABP C-01)||1166||< 60 vs. ≥ 60||BCG vs. MOF||NS|
|GITSG17||621||< 50 vs. 51–70 vs. > 70||BCG vs. FU and semustine||NS|
|Laurie et al.20 (NCCTG)||401||< 65 vs. ≥ 65||FU/LEV vs. LEV||0.21|
|Wolmark et al.21 (NSABP-C03)||1081||> 60 vs. ≥ 60||MOF vs. FU/LV||0.92|
|Moertel et al.2 (INT-0035)||929||< 61 vs. ≥ 61||FU/LEV vs. LEV vs. OBS||0.065|
|IMPACT25||1526||< 65 vs. ≥ 65||FU/LV||0.039|
|Newland et al.26 (Australia)||579||< 75 vs. ≥ 75||FU/LV||< 0.001|
|O'Connell et al.22 (NCCTG)||317||< 64 vs. ≥ 64||FU/LV vs. OBS||NS|
|Wang et al.23 (Taiwan)||218||< 60 vs. ≥ 60||FU||0.979|
|Wolmark et al.24 (NSABP-C04)||2151||< 60 vs. > 60||FU/LV vs. FU/LEV vs. FU/LV/LEV||< 0.05|
|IMPACT B23||1016||< 60 vs. ≥ 60||FU/LV vs. OBS||0.12|
|Elsaleh et al.27 (Australia)||656||< 60 vs. ≥ 61||FU/LEV||NS|
|QUASAR Collaborative Group28||4927||< 50 vs. 50–69 vs. ≥ 70||FU/LV||0.07|
In the current study, there were several reasons for choosing the cut-off age of 65 years for statistical analysis between the younger and older populations. Previous studies, as indicated above (Table 5), selected a cut-off age in this range, and the mean age of our patients was 65 years. An additional analysis could have consisted of comparing three age groups: < 65 years, 65–75 years, and > 75 years. However, only 19 patients were older than age 75 years, which would make statistical comparisons in this group underpowered and unreliable. In addition, disease free survival, overall survival, toxicity, CEA levels, disease stage, surgery type, histologic tumor differentiation, lymph node status, and chemotherapy type all were similar in patients age 65–75 years compared with patients age > 75 years (data not shown), and patient age was not a predictor of clinical outcome as a continuous variable.
Due in part to the increasing age of the population and also to increasing acceptance of the benefits of chemotherapy for the aged, older patients now are more likely to be offered adjuvant chemotherapy for resected colon carcinoma. Recently, there has been controversy regarding the relative toxicity of 5-FU-based therapy for the elderly. In a European study,14 patients age > 70 years with colon carcinoma experienced increased Grade 3–4 leukopenia and mucositis from 5-FU plus leucovorin therapy compared with a younger patient group. Differences in toxicity by age group also were noted in the recent Intergroup adjuvant trial INT-0089, which compared different schedules of 5-FU with leucovorin or levamisole in patients with Stage II and III colon carcinoma.15 In that study, 909 of 3682 enrolled patients (24.7%) were age > 70 years, and 80% of patients had Stage III disease. When they were stratified by age (< 40 years, 40–70 years, and > 70 years), there was a trend toward increased stomatitis (4% vs. 10% vs. 16%, respectively; P < 0.001) and leukopenia (2% vs. 8% vs. 14%, respectively; P < 0.001) with advancing age. The incidence of diarrhea did not vary by age. Conversely, a recent adjuvant study with 5-FU-based therapy16 that compared a population of patients age ≥ 70 years with patients age < 70 years demonstrated no increase in the toxicity rates of chemotherapy in the older patients but similar benefits regarding disease free survival.
The association of gender and toxicity with 5-FU-based adjuvant chemotherapy for patients with colon carcinoma, especially in older patients, has been controversial. The North Central Cancer Treatment Group (NCCTG) recently reported a meta-analysis of 6 NCCTG cancer-control trials involving 786 patients receiving 5-FU-based chemotherapy for colorectal carcinoma.17 Including the effect of age in a logistic regression analysis showed that women who received 5-FU based therapy experienced greater severe toxicity than men. Six toxicities were assessed in the analysis, including mucositis, leukopenia, alopecia, nausea, emesis, and diarrhea. Similar results were found in the INT-0089, in which older female patients with colon carcinoma experienced a greater severity of toxicity from adjuvant 5-FU-based therapy.15 In our study, statistically significant differences in Grade 3 and Grade 4 toxicity were not present by age (< 65 years vs. ≥ 65 years) or gender (female vs. male). All patient cohorts, older patients compared with their younger counterparts and male patients compared with female patients, experienced similar patterns and severity of toxicity from the chemotherapy.
Table 5 shows that there is controversy over whether age predicts overall survival in patients with stage II or III colon carcinoma who receive adjuvant chemotherapy. Ten prospective adjuvant trials2, 3, 13, 18–21, 22, 24 have shown that age is not associated significantly with survival. Conversely, three studies showed a correlation with age as predictor of overall survival in patients with colon carcinoma who received adjuvant, 5-FU-based therapy.21, 25, 26 Our data demonstrate that the overall survival for young and older populations are similar to those reported in randomized clinical trials.2, 22, 24 In the current study, although we included death from all causes, we found that the main cause of death was recurrent colon carcinoma. In both the younger and older populations, the liver was the most common site of recurrence, and 25% of recurrences included at least one site that was considered an extra-abdominal metastasis (data not shown). Although the disease free and overall survival rates for both younger and older populations were not statistically significant, there was a trend in improved clinical outcome for the elderly. Moertel et al. found that older patients with colon carcinoma had a greater survival benefit from 5-FU-based adjuvant therapy compared with younger patients.2 We do not assume that the trend in improved survival seen in elderly patients is due to age as independent factor; rather, it may be due to the chemotherapy effect in selected elderly patients. A selection bias may exist regarding the choice of patients who receive chemotherapy. This selection bias is apparent when the mean age of elderly patients who received chemotherapy is compared with the mean age of elderly patients who did not receive chemotherapy (mean age ± standard deviation: with chemotherapy, 72.0 ± 5.0 years; without chemotherapy, 78.0 ± 6.9 years; P = 0.001). It is unclear how this potential bias, which is a common artifact in retrospective studies, may have influenced the results of this study.
Randomized trials of adjuvant chemotherapy in patients with Stage II and III colon carcinoma have shown that preoperative CEA level,2, 23, 25 tumor grade,2, 24 and lymph node status2, 19, 20, 23, 25, 26 are independent predictors of tumor recurrence and overall survival. In our review, we did not find a correlation between tumor grade and patient age or between lymph node status and patient age. We found that 42% of patients age < 65 years had preoperative CEA levels > 5 ng/mL compared with 58% of patients in the cohort of older patients. Although it was not significant statistically, this difference was moderately large, and it is possible that there was no actual difference because of the lack of statistical power. However, we conclude that there was no correlation between any of the three prognostic parameters and patient age.
We conclude that adjuvant, 5-FU-based therapy should be presented as an option to older patients who are at high risk for recurrent colon carcinoma. The data regarding age cannot be generalized to patients age > 75 years. 5-FU-based adjuvant therapy is tolerated relatively well in the elderly and should be considered appropriate treatment in this population.
- 1National Institutes of Health. Annual cancer statistics review, including cancer trends, 1950-1985, ed 88. Bethesda, MD: National Institutes of Health, 1988.
- 4Adjuvant chemotherapy for colorectal cancer in the elderly: population-based experience [abstract]. Am Soc Clin Oncol Proc. 1993; 12: 195., , , , .
- 14Tomudex International Study Group. Haematological and non-haematological toxicity after 5-fluorouracil and leucovorin in patients with advanced colorectal cancer is significantly associated with gender, increasing age and cycle number. Eur J Cancer. 1998; 34: 1871–1875., , , , for the
- 15Fluorouracil (FU), leucovorin (LV) and levamisole (LEV) adjuvant therapy for colon cancer: preliminary results of INT-0089. Proc Am Soc Clin Oncol. 1996; 15: 486., , , .
- 16Adjuvant and palliative chemotherapy for colorectal cancer in patients aged 70 years or older. Am Soc Clin Oncol Proc. 1998; 17: 278a., , , , .
- 17Women experience greater toxicity with 5-FU based chemotherapy for colorectal cancer: a North Central Cancer Treatment Group (NCCTG) meta-analysis. Proc Am Soc Clin Oncol. 2000; 19: 2359., , , et al.
- 24Clinical trial to access the relative efficacy of fluorouracil and leucovorin, fluorouracil and levamisole, and fluorouracil, leucovorin, and levamisole in patients with Dukes' B and C carcinoma of the colon: results from National Surgical Adjuvant Breast and Bowel Project C-04. J Clin Oncol. 1999; 17: 3553–3559., , , et al.