|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Natural History of More Than 20 Years of Node-Positive Primary Breast Carcinoma Treated With Cyclophosphamide, Methotrexate, and Fluorouracil–Based Adjuvant Chemotherapy: A Study by the Cancer and Leukemia Group B
From the Walter Reed Army Medical Center, and Georgetown University Medical Center, Washington, DC; Cancer and Leukemia Group B Statistical and Data Management Centers, Duke University, Durham, NC; Mt Sinai School of Medicine, and Memorial Sloan-Kettering Cancer Center, New York, NY; West Virginia University Medical Center, Morgantown, WV; University of California at San Francisco, San Francisco, CA; and University of Pretoria, Pretoria, South Africa. Address reprint requests to Raymond B. Weiss, MD, 15304 Narcissus Way, Rockville, MD 20853-1744; email: RayWeissMD{at}aol.com.
Purpose: Breast cancer heterogeneity dictates lengthy follow-up to assess outcomes. Efficacy differences for three regimens that are based on adjuvant cyclophosphamide, methotrexate, and fluorouracil (CMF) are presented in this article, but cancer recurrence sites, time of relapse, subsequent primary cancers, and causes of death in the natural history of node-positive breast cancer are emphasized. Patients and Methods: Beginning in 1975, 905 patients with node-positive cancer were randomly assigned to receive CMF or two regimens of CMF plus other agents. Median follow-up is 22.6 years. The natural-history analysis was performed on a subset of 814 patients. Results: Eighty percent of the 599 women known to have died, died of metastatic breast cancer. Only 8.5% of the deceased women died of a cause other than breast cancer, a second or third cancer, or adjuvant chemotherapy toxicity. One hundred five women (12.8%) developed other primary cancers, with 49 (46.6%) occurring in the contralateral breast. Therapeutic efficacy differences of the CMF regimens reported earlier have been maintained more than 20 years later. For certain subsets, the five-drug regimen had advantages over CMF. Bone was the most common recurrence site. The longest interval to relapse has been 23.5 years, and 18% of those who relapsed did so more than 10 years later. Conclusion: Despite adjuvant chemotherapy, a large majority (80%) of women with node-positive breast cancer die of the disease, and many recurrences develop more than 10 years later. CMF plus vincristine and prednisone provides a benefit compared with CMF, but the magnitude varies with the number of involved nodes. Outcome trends in earlier analyses of this study were maintained even years later.
BREAST CARCINOMA may have the most heterogeneous natural history of all human cancers. The disease can have a brief onset and a fatal outcome within months, or it can be appropriately diagnosed and treated initially only to reappear as metastatic disease even 2 or more decades later. Thus, the usual milestone of 5-year relapse-free survival that is used to denote a cure for many carcinomas is often meaningless in this disease. A truer picture of outcome for breast cancer requires information well beyond 10 years from diagnosis. For example, Rosen et al1 reported the 20-year follow-up results of 170 patients with node-positive, T1 breast cancer. None of these patients received adjuvant chemotherapy because they were diagnosed with breast cancer before the 1970s, when adjuvant chemotherapy was seldom used. Forty-eight percent of the patients developed recurrent breast cancer, with 12% to 16% of the recurrences occurring during the second decade of follow-up.1 In 1975, the Cancer and Leukemia Group B (CALGB) initiated a trial of adjuvant chemotherapy for node-positive, primary breast carcinoma. On a randomized basis, all patients received one of three chemotherapy regimens after mastectomy that contained at least cyclophosphamide, methotrexate, and fluorouracil (CMF). One group received CMF only, one group received CMF plus vincristine and prednisone (CMFVP), and one group received CMF plus methanol extraction residue of bacillus Calmette-Guérin (CMF-MER). The study results for patients with four or more positive axillary lymph nodes and one to three positive nodes were previously published in 19832 and 1987,3 respectively. Certain subgroups benefitted from the addition of VP to CMF,2 but no benefit was achieved from the addition of MER. Adjuvant chemotherapy has been demonstrated to provide disease-free survival (DFS) and overall survival (OS) benefits to patients with node-positive breast cancer in large meta-analyses performed by the Early Breast Cancer Trialists’ Collaborative Group4–6 and in many individual studies. Improvement in survival from adjuvant chemotherapy has also been shown in a population-based study.7 Long-term outcome results have been reported in some individual trials of adjuvant chemotherapy,8–11 but these reported results appropriately focus on the DFS and OS produced by one systemic treatment compared with a control treatment or no treatment. In this report, we focus not on which regimen produced a better result but on the natural history of node-positive breast carcinoma treated with CMF-based adjuvant chemotherapy. We emphasize the relapses and deaths and analyze the long-term survival of this breast cancer category after more than 20 years of follow-up. We also report the instances of second and third primary cancers.
Women with operable carcinoma confined to the breast who had undergone a mastectomy and axillary lymph-node dissection were registered onto this study. The patients were to have been screened for the presence of obvious metastases with physical examination, chest radiograph, liver function blood tests, and a radionuclide bone scan. The first patient was entered in June 1975 and the last patient was entered in January 1981. All patients were to have at least one positive axillary lymph node. Additional protocol eligibility criteria have been described previously.2 The study was conducted in accordance with precepts established by the Declaration of Helsinki in 1974. This study involved 905 patients from multiple institutions in the United States and five other countries (South Africa, Canada, Denmark, Switzerland, and France). During the last 25 years, some of these institutions and their investigators left CALGB and follow-up of these study patients was abandoned. By 1991, approximately 245 patients had been lost to follow-up or did not have recent follow-up. With such a large amount of missing follow-up information, a reliable long-term outcome analysis was impossible. Thus, the first author made a concerted effort in the last decade to locate all possible lost patients and to verify the date and cause of death or the current disease status of the living patients. A variety of resources outside CALGB were used for this purpose, including tumor registries, state vital statistics offices, the Social Security Administration, a credit reporting service, and various Internet sites. This task involved a great deal of detective work to locate patients whose last study contact had been 10 to 15 years ago. Once a patient was located who was still alive, follow-up continued to a point as close to the present time as possible. In addition to the effort to locate patients, the first author reviewed the CALGB paper records of all patients to establish accurately the initial disease status, the development of subsequent primary cancers, the site and date of relapse, and the cause of death when possible. The three chemotherapy regimens used have been described previously in detail.2 All patients began the CMF-based chemotherapy with a 6-week induction program of daily oral cyclophosphamide and weekly intravenous fluorouracil and methotrexate. Those patients assigned to CMFVP also received weekly vincristine. Oral prednisone was given for the first 21 days and then tapered to no drug by day 28. The CMF-MER group received MER subcutaneously on weeks 2, 3, and 5 of the induction therapy. Randomization to this treatment arm was discontinued in October 1978, when MER toxicity was observed to be significant without providing any benefit compared with CMF alone. The third group of patients received only CMF. After the initial 6-week induction schedule, patients were to receive CMF and their other assigned drugs in a 2-weeks-on, 2-weeks-off schedule, for a total treatment interval of 12 months. A second year of CMF-only therapy was to be administered to all patients in a 2-weeks-on, 4-weeks-off schedule. The original eligibility criteria included the requirements that the mastectomy be performed within 2 weeks after breast biopsy, no insulin-dependent diabetes mellitus could be present, the initial WBC count must have been greater than 4,000/µL, the serum hepatic enzymes must have been within normal limits, and the patient must have had unilateral breast carcinoma with at least one histologically positive axillary lymph node. Patients who did not meet these eligibility criteria were excluded from the analyses previously published.2,3 However, we have now included all patients in this long-term analysis provided they received at least one course of the study chemotherapy. Therefore, this analysis is a modified version of an intent-to-treat analysis. The only exclusions were 19 patients who were entered onto but were never treated on this study. Disease recurrence was diagnosed by biopsy and/or the appearance of typical abnormalities on a radiographic study. A new cancer developing in the contralateral breast was considered a second primary cancer and not a disease recurrence. A chest wall recurrence was defined as a tumor lesion in the mastectomy area. A lymph node recurrence was defined as relapse in an ipsilateral node in either the supraclavicular or axillary areas. Multiple site recurrences were defined as two or more tumor sites discovered within 1 month. Bone marrow metastases were considered to be bone relapses.
Statistical Analysis
For the intent-to-treat analysis, the outcome variables are DFS and OS. OS was measured from the study entry date to date of death. Survivors were censored at the last date they were known to be alive. DFS was measured from study entry date to date of relapse (local or distant) or death from any cause. Those without relapse were censored at the last date they were known to be relapse-free. We estimated survival distributions by the Kaplan-Meier product-limit method and compared two or more distributions using the log-rank test. Cox proportional hazards regression was used to model the relationship of DFS and OS with clinical variables. In addition to treatment arm, these variables included the number of positive nodes, menopausal status, and tumor size. Survival after relapse, or time from relapse to death, was also examined using Kaplan-Meier survival curves.
The patients who were entered onto this study before the early closure of the CMF-MER treatment arm in October 1978 were compared with the patients who were entered after this time point. No major differences in major prognostic variables or survival (Fig 1A
Long-Term Outcomes and Natural History of the Node-Positive Group Of the 814 patients eligible for the natural-history analysis, nine were permanently lost to follow-up (despite a maximal effort to locate them) or withdrew consent for follow-up. A total of 206 of the 814 patients were known to be alive as of March 2002, with no patient having her last documented follow-up earlier than August 1993. Of these 206 patients known to be alive at last follow-up, 189 have not had a breast cancer recurrence. Fourteen patients were alive with distant disease at last follow-up. Seven patients are alive who had local (chest wall) tumor relapses occurring a median of 11 years (range, 1 to 19 years) after mastectomy, and five of these patients remain free of metastases at present.
Of the 814 patients, 511 (63%) are known to have had a recurrence of their original breast cancer either locally or distantly; 491 of these patients (96%) are known to have died, and the survival status for one patient is unknown. An additional 108 patients died without a previous relapse, for a total of 619 DFS failures. There were 300 patients still at risk of DFS failure at 10 years. An additional 110 patients (18% of the 619 DFS failures and 13% of the original entrants) relapsed between 10 and 20 years after their mastectomy. The longest interval to relapse so far (without an intervening second primary breast cancer) has been 23.5 years and occurred in the liver. The most common site of initial metastatic relapse was in bone (with or without other concurrent metastatic sites), accounting for 29% of the relapses (Table 2
A total of 599 patients (73%) are known to be deceased. The causes of death are listed in Table 3  . The overwhelming cause (80% of 599 deaths) was metastatic breast cancer. Fifty-two deaths were a result of a condition unrelated to cancer and 46 deaths were a result of second or third primary cancers. Eleven patients died of unknown reasons.
Eight patients (1.3% of 596 deaths) died because of acute toxicity related to the chemotherapy (Table 4 ), and seven of these eight patients died within 1 year of study entry. Two of these events were a result of neutropenic sepsis, two events were a result of pulmonary insufficiency, and four events were a result of venous thromboembolism, presumably induced by the chemotherapy.12
Table 5 lists the causes of death for the 52 patients who died of reasons unrelated to cancer. Cardiovascular (usually an acute myocardial infarction) or cerebrovascular (a stroke) disease accounted for most of these deaths. The remaining causes cover a variety of conditions, with pneumonia and chronic obstructive pulmonary disease being the most prominent. These 52 deaths (of 599 known deaths) not caused by cancer mean that only 8.6% of those patients who have died did so from a condition not clearly related to the first or subsequent cancers or the adjuvant chemotherapy given. One patient committed suicide.
Table 6 lists subsequent malignancies that were documented. One hundred five patients developed other primary cancers, with breast cancer being the most common, occurring in 49 (6%) of the 814 patients. Of the 56 patients developing second cancers other than in the contralateral breast, 31 have subsequently died from this second cancer. Three patients developed acute myelogenous leukemia (AML) 6, 8, and 10.5 years, respectively, after study entry. Two of these patients died of the AML, and one patient died of both the AML and metastatic breast cancer for which she was being treated. Nine of the 105 patients developed a third cancer (Table 6) and two subsequently died of it. Four of the nine remain alive and free of recurrence of any cancer. Figure 2 shows the actuarial rate of subsequent malignancies from study entry. The ordinate in this figure is the cumulative probability of second primary cancers diagnosed by organ site (contralateral breast v nonbreast). These cancers were approximately evenly distributed among the treatment groups.
The most important prognostic factor for patients with node-positive breast cancer is the number of nodes. Figure 3A and 3B illustrate the DFS and OS on the basis of the number of involved nodes. The median DFS is 11.1 years for the patients with one to three nodes, 5.4 years for the patients with four to nine nodes, and 2.1 years for the patients with 10 or more positive nodes.
Table 7 lists the DFS for the 174 patients who had 10 or more positive nodes at the time of mastectomy. Of these 174 patients, 149 relapsed, and 25 (14.4%) did not relapse. Eleven patients relapsed within 6 months of study entry and an additional 27 patients relapsed by 12 months (a total of 22% of all the patients with 10 or more positive nodes who relapsed). Conversely, relapse occurred 5 or more years after study entry in 21 patients (12%), with five patients (3%) relapsing more than 10 years later.
A total of 146 of the 149 patients with 10 or more nodes who relapsed are known to have died. Eleven patients died without relapse, including five who developed and died of second primary cancers. Of the three patients who relapsed but are not known to have died, two relapsed within 2 years of study entry (one locally in the chest wall/skin and the other in the regional lymph nodes), but they were still alive more than 18 years later. Four of the 14 patients who are alive and relapse-free had 18 to 31 involved axillary nodes; the other 10 patients had 10 to 15 involved nodes.
Intent-to-Treat Analysis
Figure 4A and 4B show the DFS and OS for the individual drug regimens, comparing the CMFVP group versus the two CMF groups combined. There is a significant difference in DFS; the DFS in the CMFVP group is better than the DFS in the CMF groups (log-rank, P = .048). The OS comparison between treatment arms is not significant (log-rank, P = .26).
Table 10 lists the conditional risk of relapse given that the patient survives and is disease-free up to a particular time point. The average annual rate of relapse within the first 5 years after study entry is 13.0%. Given that the patient has remained disease-free and alive for 5 years, the average annual rate of relapse or death in the next 5-year interval is 5.8%. Given that a patient has reached 10 years and is still disease-free, the average annual rate of relapse or death in the next 5 years (10 to 15 years) is 4.8%. For patients who are disease-free at 15 years, the average annual rate of relapse or death within the next 5-year interval is 5.3%. Table 10 demonstrates that the highest rate of relapse occurs during the first 5 years after mastectomy. After this time, the rate decreases by approximately 60% and stays constant at 5% to 6%. Moreover, this long-term hazard rate depends little, if any, on a woman’s initial risk for relapse. As Table 10 shows, the rates of both groups with one to three or four or more positive nodes drop to approximately 5% after 5 years.
Multivariate Analyses With Outcome Table 11 lists correlations between pairs of variables. A correlation indicates both the direction and magnitude of the linear relationship between two variables. A correlation can range from -1.0 to +1.0. A positive correlation indicates the fact that as values of one variable increase, values of the other variable tend to increase. A negative correlation indicates that as values of one variable increase, values of the other variable tend to decrease. The closer the correlation is to +1.0 or -1.0, the stronger the relationship. The closer the correlation is to zero, the weaker the relationship.
There was a significant positive correlation between tumor size and the number of positive lymph nodes (r = 0.11; P = .0021); larger tumors were associated with more positive nodes. Hormone receptor status and menopausal status were correlated (r = 0.13; P = .0124), with younger women tending to have estrogen receptor–negative tumors. However, receptor status was available for only approximately 40% of the patients because this information was often not obtained during the 1970s when these patients had their primary cancers treated.
Tables 12
The relative risk (RR) quantifies the differential risk of DFS or OS failure between categories or values of the predictor variable. If both comparison categories are at equal risk, the RR equals 1.00. Some examples in the context of model A are (1) a patient with 10 positive nodes is more than twice as likely to relapse or die as a patient with only one positive node (RR, 2.14); (2) the risk of relapse or death for a postmenopausal patient is 47% higher (RR, 1.47) than the risk for a premenopausal patient; and (3) for a tumor size of 5 cm as opposed to a tumor of 2 cm, the risk of relapse or death increases by 34% (RR, 1.34).
Model A shows that treatment, number of positive nodes, menopausal condition, and tumor size are significantly related to DFS (P < .0093). There is an overall benefit of CMFVP compared with CMF, as is indicated in Fig 4A
Figures 5
The difference between treatment arms in Fig 5A  and 5B is striking. There is an apparent advantage of CMFVP compared with CMF only in patients with a large number of positive nodes. This difference is difficult to explain on a biological basis. A possible explanation for this observation is that the greater proportion of events in these groups of patients with a large number of positive nodes makes it easier to distinguish a treatment effect and to distinguish it earlier. To evaluate the plausibility of this explanation, we compared the two nodal groups (one to three nodes v  four nodes) at time points when they had comparable DFS probability and plotted their adjusted DFS curves in Fig 7B (raw data shown in Fig 7A). For CMF-treated patients with one to three positive nodes, the average DFS at 15 years is 42%. For the CMF-treated patients with four or more positive nodes, the time at which their DFS was 42% is 3.67 years. We, therefore, multiplied DFS by 15/3.67 for both treatment categories (CMFVP and combined CMF) in the group with four or more nodes, so the adjusted DFS for the CMF-treated group with four or more nodes agrees as closely as possible with the CMF-treated group with one to three nodes at 15 years from study entry. The adjusted DFS for the CMF-treated group with four or more nodes does not agree at earlier time points, because hazard rates adjusted by time do not themselves have a simple multiplicative relationship across the two groups. In particular, the hazard rate within each group is not constant over time, as is demonstrated in Table 10. Figure 7B shows that the CMFVP and the combined CMF groups have already diverged at 3.67 years among patients with four or more positive nodes, whereas they have not diverged at 15 years for patients with one to three nodes. This difference in DFS at 3.67 years in the group with four or more nodes (66% for CMFVP v 42% for CMF) is statistically significant (P = .0001). This same effect is manifested in the OS plots (data not shown).
Survival After Relapse The median time from relapse to death was approximately 18 months for both treatment regimens (Fig 8 ).
Breast cancer is a particularly heterogeneous cancer, even for a given tumor size and number of positive axillary lymph nodes. Although most relapses occur early in patients with positive nodes, disease in patients may continue to recur over many years after the initial diagnosis. With more than 22 years of follow-up, this study allows the assessment of long-term outcomes of the disease and its treatment. We found that 18% of the patients who relapsed (constituting 13% of those studied) did so 10 or more years after initial diagnosis. Fourteen patients are alive with systemic metastases 19 to 22 years after their original diagnosis. These observations reinforce the need for long follow-up of patients with breast cancer. Our results are consistent with those from a long-term natural-history study13 in which there were still a few deaths caused by the initial breast cancer occurring between 20 and 25 years after diagnosis. However, the trends of DFS and OS demonstrated during the analyses of this clinical trial reported in 19832 and 19873 have been maintained with this long-term follow-up. The survival advantages gained by the patients with four or more positive lymph nodes from the CMFVP regimen have been maintained, whereas there have been no late-occurring differences in treatment arms for the patients with one to three nodes. Bonadonna et al8 also saw persistence of the beneficial effects of adjuvant chemotherapy 20 years later compared with use of surgery alone, with no change in the conclusions drawn in their first report of the same study. This result is to be expected because when one treatment is associated with a longer survival than another after 10 years of follow-up, the earlier difference will be maintained if the treatment groups have the same survival distribution in subsequent years. Only if the second treatment group were to outperform the first in later years would the earlier difference be lost. No evidence of such a turnaround was seen in this current analysis of our study. Although cautions regarding changes in the results of breast cancer treatments with longer follow-up were expressed more than 25 years ago14 and more recently,15 the data herein demonstrate that trends established early do not necessarily change years later, even for a cancer with well-known tendencies for late recurrences and late deaths caused by metastases. Adjuvant chemotherapy has been shown to improve survival of patients with node-positive, primary breast cancer.4–10 Nonetheless, our study shows that the overwhelming majority (80%) of node-positive patients who die within the 26 years after diagnosis still die of breast cancer recurrence despite the use of CMF-based adjuvant chemotherapy. In addition, the DFS at 20 years is only 23%. This figure is slightly lower than that of the Milano group study,8 in which the 20-year, relapse-free survival was 32% for CMF-treated patients. However, this difference is largely a result of the inclusion in our study of deaths from causes in addition to breast cancer as DFS events (data not shown). In addition, the 20-year OS for our study of 28% is comparable with the 32% OS rate from the Milano study.8 Our figures are also comparable with the 13-year outcome results of the International (Ludwig) Breast Cancer Study Group in Trial II using a CMF regimen.9 The heterogeneity of breast cancer also is illustrated by the outcome of patients with 10 or more involved axillary nodes. Although 86% of the 174 patients in this subgroup relapsed despite receiving adjuvant chemotherapy, 14% did not. Twenty-two percent of patients in this group who relapsed did so within 12 months of study entry. If screening for occult metastases had been possible with currently available radiographic techniques, it is likely that many patients would already have been found to have stage IV disease at the time of primary tumor diagnosis. Even despite an early relapse, three patients are alive and disease-free more than 18 years later. The most common site of relapse was bone (alone or in conjunction with other sites). This observation has been made in other studies.16 This finding re-emphasizes a need for heightened clinical suspicion of developing bone metastases whenever a patient with node-positive breast cancer has any symptoms related to bone during follow-up. The fact that 31% of the relapsing patients had their initial recurrence on the chest wall or in regional nodes also emphasizes the need to investigate suspicious new physical findings in these anatomic areas during follow-up. Twelve (8%) of the 156 patients who had chest wall or nodal relapse are alive and disease-free up to 18 years from the relapse event. Even though 92% of the patients who had such local recurrence did not survive the relapse, it is evident that not everyone will succumb with distant disease later, which is just one more indicator of the heterogeneity of this cancer.
Table 4 Only 6.1% of the 814 patients in the natural-history analysis of this study died during the first 22 years of follow-up of a known cause other than relapse of the first cancer, a second or third cancer, or an effect of the chemotherapy. The fact that 80% of patients died of breast cancer demonstrates that this disease will be the overwhelming cause of death in the 20 years after diagnosis of node-positive breast cancer. Three patients (0.37% of the 814 evaluated patients) developed AML 6 to 10.5 years after receiving the adjuvant chemotherapy. One patient each developed and died of chronic myelogenous leukemia and myeloma, the only other marrow-related second cancers. One patient developed polycythemia vera. There were no instances of myelodysplastic syndrome. Only one of these six patients (one with AML) experienced a relapse of her breast cancer; the other five patients all died of the second disease. The single cases of myeloma and chronic myelogenous leukemia were probably coincidental and not related to the adjuvant chemotherapy. Whether the AML cases are random events or a result of the CMF-based chemotherapy is indeterminable, but the rarity of AML as a second malignancy in our study is consistent with the experience of other studies in which cyclophosphamide was the alkylating agent used. Valagussa et al18 observed no patients with acute leukemia in 845 patients who were observed for a median of 10 years after treatment with adjuvant CMF. Tallman et al19 found five patients with myelodysplastic syndrome or acute leukemia (0.19%) in 2,638 patients treated with CMF- or cyclophosphamide, fluorouracil, and doxorubicin–based drug regimens. A total of 100 patients developed second and third cancers other than bone marrow malignancies. The most common cancer was contralateral breast cancer; 52 of the patients (5.2%) developed second or third cancers involving the contralateral breast. Nineteen of these patients also developed distant metastases and, subsequently, died of breast cancer. It cannot be determined whether the first or second breast cancer was responsible for the metastases. Our figure of 5.2% is the same as the 5.2% figure of contralateral breast cancer observed in the series from the Milano group18 using CMF and similar to the rate of 8% observed in a series of node-positive patients who did not receive adjuvant chemotherapy.20 Such data once again emphasize the need of monitoring patients who have survived one breast cancer for evidence of a second cancer arising in the contralateral breast. It is of interest that our overall figure of 105 patients (12.9%) developing a second cancer of any type is higher than the 6% rate reported by Valagussa et al.18 Tamoxifen was not used in the adjuvant therapy in this study. This drug has been shown to increase the risk of developing uterine cancer.21 Only two patients developed a uterine cancer in this study.
Figure 3A
We previously reported for this study that the number of positive lymph nodes and menopausal status were significantly related to DFS and OS (P < .01), and that there were strong interactions of treatment with the number of involved nodes and menopausal status.2,3 These same findings hold true with this analysis. The interaction between nodes and treatment (illustrated in Fig 5A The multivariate Cox regression models showed the number of positive nodes, menopausal status, tumor size, and the two interactions involving treatment arms to be significant. Patients with fewer positive nodes have a significantly lower risk of both relapse and death than patients with more positive nodes. Premenopausal patients have lower risks than postmenopausal patients, as do patients with smaller tumor sizes. In summary, long-term follow-up of patients with node-positive breast cancer who received adjuvant chemotherapy continues to demonstrate relapses with metastases occurring 20 or more years later, but as many as 38% of such patients remain disease-free 10 years after initial diagnosis. Second cancers, which often cause death, occur in 13% of patients, and marrow malignancies (especially AML) are rare. Finally, despite the existence of late recurrences of, and deaths caused by, breast cancer, outcome trends observed in shorter follow-up intervals (eg, 5 years) are maintained 1 or 2 decades later. There are no late divergences of the DFS or OS curves, and the initial effect of the respective chemotherapy regimens is recognizable even 2 decades later.
We thank Patricia Wolfgang, New York State Cancer Registry; Darlene Maxwell, West Virginia Cancer Registry; Rene Jooste, University of Pretoria; and Marva Delapp, Brooklyn Hospital Center for their valuable assistance tracing patients on this study.
Supported in part by the following grant nos. CA 26806, CA 31946, CA 33601, CA 04457, CA 28562, CA 60138, CA 77597, and CA 77651 from the National Cancer Institute, Bethesda, MD, to the Cancer and Leukemia Group B.
1. Rosen PP, Groshen S, Saigo PE, et al: A long-term follow-up study of survival in stage I (T1 N0 M0) and stage II (T1 N1 M0) breast carcinoma. J Clin Oncol 7:355–366, 1989[Abstract] 2. Tormey DC, Weinberg VE, Holland JF, et al. A randomized trial of five and three drug chemotherapy and chemoimmunotherapy in women with operable node positive breast cancer. J Clin Oncol 1:138–145, 1983[Abstract] 3. Weiss RB, Korzun AH, Tormey DC, et al: Adjuvant chemotherapy for breast carcinoma using CMFVP versus CMF versus CMF-MER: The nine-year results for patients with 1–3 axillary nodes, in Salmon SE, Jones SE (eds): Adjuvant Therapy of Cancer V. Orlando, FL, Grune & Stratton, 1987, pp 327–336 4. Early Breast Cancer Trialists’ Collaborative Group: Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy: 133 Randomized trials involving 31,000 recurrences and 24,000 deaths among 75,000 women (part 1). Lancet 339:1–15, 1992[Medline] 5. Early Breast Cancer Trialists’ Collaborative Group: Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy: 133 randomized trials involving 31,000 recurrences and 24,000 deaths among 75,000 women (part 2). Lancet 339:71–85, 1992[Medline] 6. Early Breast Cancer Trialists’ Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomised trials. Lancet 352:930–942, 1998[CrossRef][Medline]
7. Olivotto IA, Bajdik CD, Plenderleith IH, et al: Adjuvant chemotherapy and survival after breast cancer. N Engl J Med 330:805–810, 1994
8. Bonadonna G, Valagussa P, Moliterni A, et al: Adjuvant cyclophosphamide, methotrexate, and fluorouracil in node-positive breast cancer: The results of 20 years of follow-up. N Engl J Med 332:901–906, 1995
9. Castiglione-Gertsch M, Johnsen C, Goldhirsch A, et al: The International (Ludwig) Breast Cancer Study Group trials I-IV: 15 years of follow-up. Ann Oncol 5:717–724, 1994
10. Senn HJ, Maiback R, Castiglione M, et al: Adjuvant chemotherapy in operable breast cancer: Cyclophosphamide, methotrexate, and fluorouracil versus chlorambucil, methotrexate, and fluorouracil—11-year results of Swiss Group for Clinical Cancer Research Trials SAKK 27/82. J Clin Oncol 15:2502–2509, 1997 11. Rivkin SE, Green S, Lew D, et al: Adjuvant CMFVP versus melphalan for operable breast cancer with positive axillary nodes: 23-year results of a Southwest Oncology Group study. Proc Am Soc Clin Oncol 18:69a, 1999 (abstr 259) 12. Weiss RB, Tormey DC, Holland JF, et al: Venous thrombosis during multimodal treatment of primary breast cancer. Cancer Treat Rep 65:677–679, 1981[Medline]
13. Joensuu H, Toikkanen S: Cured of breast cancer? J Clin Oncol 13:62–69, 1995 14. Costanza ME: Sounding board: The problem of breast-cancer prophylaxis. N Engl J Med 293:1095–1098, 1975[Medline]
15. Melnychuk D, Panasci LC: Adjuvant therapy for breast cancer. N Engl J Med 331:741–742, 1994 (letter) 16. Jain S, Fisher C, Smith P, et al: Patterns of metastatic breast cancer in relation to histologic type. Eur J Cancer 29A:2155–2157, 1993[CrossRef][Medline] 17. Colleoni M, Price KN, Castiglione-Gertsch M, et al: Mortality during adjuvant treatment of early breast cancer with cyclophosphamide, methotrexate, and fluorouracil. Lancet 354:130–131, 1999[CrossRef][Medline]
18. Valagussa P, Tancini G, Bonadonna G: Second malignancies after CMF for resectable breast cancer. J Clin Oncol 5:1138–1142, 1987 19. Tallman MS, Gray R, Bennett JM, et al: Leukemogenic potential of adjuvant chemotherapy for early-stage breast cancer: The Eastern Cooperative Oncology Group experience. J Clin Oncol 13:1557–1563, 1995 20. Rosen PP, Groshen S, Kinne D, et al: Contralateral breast carcinoma: An assessment of risk and prognosis in stage I (T1 N0 M0) and stage II (T1 N0 M0) patients with 20-year follow-up. Surgery 106:904–910, 1989[Medline]
21. Fisher B, Costantino JP, Wickerham DL, et al: Tamoxifen for prevention of breast cancer: Report of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J Natl Cancer Inst 90:1371–1388, 1998
22. Perloff M, Norton L, Korzun AH, et al: Postsurgical adjuvant chemotherapy of stage II breast carcinoma with or without crossover to a non-cross-resistant regimen: A Cancer and Leukemia Group B study. J Clin Oncol 14:1589–1598, 1996 Submitted September 9, 2002; accepted February 5, 2003.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
