The Long Look: 10-Year Follow-Up of Adjuvant Colon Trial NSABP C-01

Abstract & Commentary

Synopsis: The 10-year disease-free and overall survival rate is presented from a trial of adjuvant therapy (chemo-, or immunotherapy) for Dukes’ B and C colon cancer conducted by the National Surgical Adjuvant Breast and Bowel cooperative group. The disease-free and overall survival advantage for the chemotherapy arm, apparent at 5 years, was no longer evident by 10 years. Although new, likely more effective regimens are currently used in this setting, the report raises caution concerning the need for extended follow-up and the use of surrogate markers such as disease-free survival in predicting cures.

Source: Smith RE, et al. J Natl Cancer Inst. 2004;96:1128-1132.

In 1988, wolmark and colleagues reported the 5-year data from the National Adjuvant Breast and Bowel Project C-01.1 In that early trial, conducted between November, 1977 through February 1983, 1166 patients with resected Dukes’ stage B and C adenocarcinoma of the colon were randomly assigned to receive no further therapy (surgery alone; 394 patients), adjuvant chemotherapy (MOF [methyl CCNU, vincristine and 5-fluorouracil]; 379 patients), or adjuvant immunotherapy (bacillus Calmette-Guerin [BCG]; 393 patients). The adjuvant chemotherapy consisted of eight 10-week cycles of drug, whereas the immunotherapy (BCG) was administered weekly for a total of 12 injections. The results at 5 years1 indicated that MOF adjuvant chemotherapy was associated with a statistically significant improvement in disease-free survival (58% vs 51%; P = 0.02) and overall survival (67% vs 59%; P = 0.05) when compared with surgery alone. Adjuvant BCG therapy was associated with a trend toward improvement in disease-free survival (56% vs 51%; P = 0.09) and a statistically significant overall survival advantage (67% vs 59%; P = 0.03). However, when deaths with no evidence of tumor recurrence were eliminated, BCG therapy had no statistically significant benefit in either disease-free survival or overall survival, whereas benefit from chemotherapy was retained.

In the current manuscript, complete survival data at 10 years is presented. At this point, MOF adjuvant chemotherapy showed no benefit compared with surgery alone in terms of disease-free survival, relapse-free survival, or overall survival. With regard to immunotherapy with BCG, tumor relapses were not prevented, but there remained a statistically significant improvement in overall survival (53% vs 47%).

Thus, the disease-free and overall survival benefit associated with adjuvant chemotherapy (with MOF) was found to be of limited duration—apparent at 5 years but gone at 10.

Comment by William B. Ershler, MD

Without doubt, the results are a bit discouraging. However, it must be recalled that neither of the experimental arms examined (MOF chemotherapy or BCG immunotherapy) are currently used in this setting; the field having advanced somewhat with the introduction of more active chemotherapies, particularly irinotecan and oxaliplatin.2 However, certain points are worth noting. First, it is apparent that the drugs undertaken in this early trial delayed recurrences, but did not eradicate disease. This is made evident by the apparent advantages in overall survival at 5 years, but the disappearance of this difference by 10. Although adjuvant chemotherapy is generally prescribed with curative intention, a delay in tumor recurrence would also be considered a favorable outcome, particularly if the treatment is not associated with substantial toxicity. Another point is that clinical investigators are inclined to consider surrogates for overall survival, such as the recently proposed 3-year disease-free interval.3,4 Indeed, the 3-year disease-free interval may well predict 5 year overall survival, but continued surveillance would seem warranted, based upon this cooperative trial experience presented herein.

The long-term follow-up of the BCG-treated patients is also worthy of note. Although tumor recurrences were not avoided, overall survival at 10 years was better than the control group. One factor is that the BCG group experienced one half the frequency of life-shortening second malignancies (9 vs 18). Although this is a relatively small number of cases, it is tempting to speculate that this brief course of non-specific immunostimulation, somehow resulted in more effective tumor surveillance.


1. Wolmark RE, et al. J Natl Cancer Inst. 1988;80:30-36.

2. Andre T, et al. Clin Colorectal Cancer. 2004; 4(Suppl 1):S22-S28.

3. Sargent DJ, et al. Proc ASCO. 2004;23 (abstract 3502).

4. Grem J. J Natl Cancer Inst. 2004;96:1116-1117.

William B. Ershler, MD, INOVA Fairfax Hospital Cancer Center, Fairfax, VA; Director, Institute for Advanced Studies in Aging, Washington, DC is Editor of Clinical Oncology Alert.