Fax: (011) 34-93-2607727
Primary chemotherapy with cyclophosphamide, methotrexate, and 5-fluorouracil in operable breast carcinoma
Version of Record online: 12 JAN 2005
Copyright © 2005 American Cancer Society
Volume 103, Issue 4, pages 657–663, 15 February 2005
How to Cite
Falo, C., Moreno, A., Benito, E., Lloveras, B., Varela, M., Serra, J. M., Prieto, L., Azpeitia, D. and Escobedo, A. (2005), Primary chemotherapy with cyclophosphamide, methotrexate, and 5-fluorouracil in operable breast carcinoma. Cancer, 103: 657–663. doi: 10.1002/cncr.20829
- Issue online: 3 FEB 2005
- Version of Record online: 12 JAN 2005
- Manuscript Accepted: 19 OCT 2004
- Manuscript Revised: 14 OCT 2004
- Manuscript Received: 29 MAR 2004
- breast carcinoma;
- primary chemotherapy;
- Cyclophosphamide/methotrexate/5-fluorouracil regimen;
Primary chemotherapy (PC) is becoming an accepted practice to treat large tumors to avoid mastectomies and as a surrogate of outcome.
A series of 305 patients with tumors > 3 cm with T2–3N0–1M0 classification were treated with a multimodal approach that consisted of 3 courses of primary cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) followed by appropriate local treatment and 3 more courses of CMF or 4 courses of doxorubicin. Response was assessed by mammography.
The overall response rate was 48% (a 3% pathologic complete response rate). Conservative surgery was achieved in 79.64% of the patients with a low rate of local disease recurrences (5%). Toxicity was minimal. With a median follow-up of 104 months, the 8-year disease-free survival (DFS) rate was 57.63% and the 8-year overall survival (OS) was 67.65%. The DFS and OS rates for patients with a clinical response were significantly longer, i.e., 70% (P = 0.0048) and 90% (P = 0.0042), respectively.
PC with CMF was feasible. A high rate of breast-conservative surgery was achieved. The current results stressed the value of PC to increase conservative surgery and as a predictor of outcome. Cancer 2005. © 2005 American Cancer Society.
Most of the initial reports of primary systemic therapy included small and heterogeneous series of patients with locally advanced Stage III and/or IV disease. These initial studies showed considerable rates of clinical tumor response, low rates of pathologic complete response (pCR), increased rates of breast-conserving procedures, and no apparent detrimental effect on survival.1, 2 One of the first reports of primary chemotherapy (PC) that included women with operable breast disease was made by Jacquillat et al. in the early 1990s.3 This group achieved a rate of breast preservation as high as 94%. The authors assured that radical local treatment (mastectomy) was not mandatory in most patients with large tumors whose main risk is metastasis. The Milan experience from two consecutive nonrandomized trials4, 5 showed that the histologic status of the axilla and the response were the variables that influenced significantly the relapse-free survival rate at 8 years. This relation between pathologic response and a better outcome was confirmed by Kuerer et al.6 However, nonrandomized studies could not address the relative efficacy of neoadjuvant versus adjuvant chemotherapy on disease-free (DFS) and overall survival (OS) rates.7 The National Surgical Adjuvant Breast and Bowel Project (NSABP) B-18 trial has been the largest randomized study to compare primary and postoperative therapy.8, 9 Patients who received PC achieved a greater number of breast-conserving surgical procedures, i.e., 67% versus 60%, and pathologic-negative lymphadenectomies. The 5-year results indicate no OS difference between postoperative and preoperative chemotherapy. Those data were confirmed further by other randomized studies.10–13
More recent studies are focused on improving the survival rates by using more intensive schemas by combining antracyclines and taxanes.14–18 These schemas have shown further increases in pCR rates, but with higher myelotoxicity. The most noteworhty studies include the NSABP B-2719 and Aberdeen trials.20 The preliminary results of the NSABP B-27 study have shown a significant benefit in favor of the docetaxel branch, such as an improved clinical complete response rate, overall clinical response, pCR, and the proportion of patients with negative lymph nodes.19 In the Aberdeen trial, the addition of sequential docetaxel to cyclophosphamide, vincristine, adriamycin, prednisone (CVAP) neoadjuvant chemotherapy resulted in a significantly enhanced clinical response rate and a substantially increased complete histopathologic response rate when compared with patients receiving CVAP alone. Furthermore, patients receiving docetaxel had an increased breast conservation rate and an increased survival rate at a median follow-up of 3 years. It is important to note that this was a small study, and the survival results should be interpreted with caution.20
The aim of our study was to evaluate the efficacy of cyclophosphamide, methotrexate, and 5-fluorouracil (5-FU; CMF) as primary systemic chemotherapy (PC) in a series of 305 patients with operable breast tumors > 3 cm classified as T2–3N0–1M0.
MATERIALS AND METHODS
From January 18, 1991 to December 16, 1999, the current study enrolled 305 women with operable breast tumors > 3 cm classified as T2–3N0–1M0. Palpable tumors should be evaluable by mammography in two dimensions. Informed consent was obtained from all patients before initiating the proposed treatment strategy, which had been approved by our local research ethics committee (Ciutat Sanitaria I Universitaria de Bellvitge [CSUB] number 3490). The criteria for exclusion were locally advanced or metastatic breast carcinoma, multicentricity, age lt; 18 or > 65 years, pregnancy, history of previous malignancies, or severe concomitant systemic disease.
Pathologic diagnosis was confirmed by fine-needle aspiration biopsy (FNAB) cytology. All tumors were palpable to ensure the infiltrating nature of the tumor. Seldom, a carcinoma in situ is presented as a palpable mass. Further staging was based on roentgenograms of the chest, skull, spine, and pelvis and on bone scans. Complete hemograma, blood biochemistry evaluations with liver function tests, and electrocardiograms were performed in all patients. A tattoo was used to mark the position of the tumor in the breast before starting chemotherapy.
At the time of the FNAB cytology, nuclear grade was assessed, as well as hormonal receptors using immunohistochemistry.
Patients were treated with CMF. The CMF regimen consisted of 600 mg/m2 cyclophosphamide, 40 mg/m2 methotrexate, and 600 mg/m2 5-FU administered intravenously on Days 1 and 8 of each treatment cycle every 28 days for 3 courses. PC was discontinued if patients had tumor progression.
Reduction of any of the drugs was permitted only after febrile neutropenia occurred, at which time a 15% reduction in the 3 drugs was required. On Day 1, treatment was delayed by 1 week if the leukocyte count was < 3000 cells per square meter, the neutrophil count was < 1500 cells per square meter, and/or the platelet count was < 100,000 cells per square meter. On Day 8, the CMF dose was reduced by 30% when the leukocyte count decreased from 3000 to 2500 cells per square meter and/or the platelet count decreased from 100,000 to 75,000 cells per square meter, and by 50% when the leukocyte count decreased from 2500 to 2000 cells per square meter and/or the platelet count decreased from 75,000 to 50,000 cells per square meter.
Assessment of Response
The size of primary tumor was measured by palpation on the first day of each treatment cycle and before surgery. Mammography was repeated within 3 weeks after the last dose of chemotherapy was received. Clinical and radiologic assessment of tumor size was accomplished by measuring the reduction in the product of the two largest tumor diameter size. The response was classified as clinical complete (cCR) in the absence of clinical evidence of tumor in the breast and the axilla, and as clinical partial (cPR) when the reduction in the clinical size of the breast tumor was ≥ 50%. Tumor progression was defined as an increase of ≥ 25% in the product of the 2 largest diameters.
Surgery was planned 3–4 weeks after the third course of CMF was delivered. Conservative surgery with wide tumorectomy was performed when radical and aesthetic criteria allowed it. Selected patients with a great response received radiologic-guided surgery. Occasionally, patients were submitted to conservative surgery with plastic surgery. The remaining patients underwent a modified radical mastectomy. In all patients, a three-level axillary dissection was performed.
Radiotherapy was delivered to all patients who received breast-conserving surgery, to patients classified as T3 or T4 who received a mastectomy, and to patients with > 4 metastasic axillary lymph nodes. Patients with surgical margins < 10 mm received a boost of 20 gray (Gy) instead of 10 Gy to the tumor bed. Radiotherapy started 3–4 weeks after adjuvant chemotherapy was completed.
Adjuvant treatment started 15 days after surgery. Patients with a response (cCR or cPR) received three more courses of CMF. Patients without a response were treated with doxorubicin at doses of 75 mg/m2 per 4 courses. Tamoxifen was not the standard treatment during our study period.
Changes secondary to chemotherapy21 include distinct loose fibrosis with capillary neoformation, periductal inflammatory infiltrate (ductulitis), stromal calcifications, and cumulus of foamy macrophages. Pathologic response was defined as a replacement of ≥ 10% of the tumor mass. A pCR suggests the absence of malignancy both in the tumor bed and in the axilla.
After the completion of the treatment program, a physical examination, hematologic tests, and blood biochemistry evaluations were performed every 3 months for the first 3 years, every 4 months during Years 4 and 5, and every 6 months thereafter. Mammography was performed once a year starting 6 months after the end of breast irradiation therapy. Other studies were performed annually and included chest and bone X-rays and bone scans. After the 10th year, controls became annual.
The time since the start of PC to the first documented disease recurrence was the DFS interval. Death from any causes was evaluated for OS, which also was measured from the date PC was started. Survival interval was calculated with Kaplan–Meier survival curves and log-rank tests. Survival was calculated separately for responders and nonresponders and compared with each other.
Of 305 patients, 1 was not evaluable because the patient died after the first cycle. Table 1 lists patient and tumor characteristics. The median age of the patients was 50.5 years (range, 26–65 years). Tumors were classified as T2N0 (n = 126), T2N1 (n = 81), T3N0 (n = 51), and T3N1 (n = 45). At palpation, the median tumor diameter was 42 mm (range, 30–90 mm). According to nuclear grade evaluated by cytology, 66.4% of the tumor specimens were Grade II, 20.1% were Grade III, and 6.9% were Grade I. Hormonal receptors were assayed for 97% of the specimens: 61.9% and 53.2% stained positive for estrogen and progesterone, respectively.
|Characteristics||No. of patients||(%)|
|Age||50 (26–65 yrs)|
Of the 304 evaluable patients, 168 presented with clinical responses (32 had cCR and 136 had cPR). Disease progression was documented in eight patients.
Radiologic response was observed in 148 patients (odds ratio, 48.7%). As shown in Table 2, there were 21 complete responses (10.5%) and 136 partial responses (44.7%). Disease progression was documented in 8 patients (2.6%).
|Response||No. of patients (%)|
|Complete disease remission||21||(6.9)|
|Partial disease remission||127||(41.8)|
|Pathologic complete disease remission||9||(2.9)|
Findings on mammograms correlated with clinical findings in 89% of patients. Seven patients who were considered to be nonresponders had radiologic evidence of response, whereas 26 patients who were considered to be responders did not have radiologic evidence of response.
After surgery, the mean pathologic tumor size was 23 mm (range, 0 [pCR]–95 mm). Metastatic deposits in the axillary lymph nodes were detected in 146 specimens (47.8%). The median number of dissected lymph nodes was 21 (range, 8–49 lymph nodes). Margins inferior to 10 mm were documented in 35 specimens.
Changes secondary to chemotherapy and unrelated to tumor stroma or fibrocystic disease were identified in 113 specimens (43.9%) and in 8.9% of the lymph nodal metastases. pCR at microscopic examination was documented in 9 patients (2.9 %). In two additional specimens, only carcinoma in situ was detected. Total replacement of lymph node metastases by fibrosis occurred in one patient.
In 216 specimens (71.3%), a good concordance between radiologic response and pathologic findings was found. However, in 87 specimens, we found either pathologic changes without radiologic evidence of tumor reduction or, more frequently, radiologic response without pathologic changes.
Primary CMF chemotherapy was well tolerated. On the planned day of drug administration, the frequency of myelosupression was minimal. Seventeen patients had febrile neutropenia that required hospitalization. This excellent drug tolerance also was reflected in the high dose intensity (DI) delivered: 1 (range, 0.58–1.15). Grade III–IV nonhematologic toxicity was not observed. Amenorrhea was observed in 87 patients (55% of premenopausal women), which was reversible in 20 patients (28%). A few patients had deep venous thromboses that were resolved with correct anticoagulant therapy. So far, there is no evidence of second malignancies associated with the CMF therapy.
There was one death due to toxicity after the first course of CMF. The necropsy was not authorized, it was considered a possible decay of dihydropyrimidine dehydrogenase.
Chemotherapy delivered after surgery also was well tolerated. Patients treated with doxorubicin developed alopecia.
Surgery was performed 3 weeks after the last course of chemotherapy was delivered (range, 3–5 weeks).
Breast-conserving surgery was achieved in 241 patients (79.64%). As shown in Table 3, 27 patients received breast-conserving surgery with the aid of plastic surgery (e.g., dorsal flap, bilateral breast reduction, remodeling)..
|Surgery||No. of patients (%)|
|Lumpectomy and ALND||214||(70.4)|
|Conservative and PS||27||(8.8)|
Morbidity did not increase with PC.
Adjuvant chemotherapy was delivered 2 weeks after surgery. CMF was delivered to 146 patients with a DI of 1 (range, 0.64–1). Doxorubicin was delivered to 149 patients with a DI of 0.97 (range, 0.58–1.1).
Tamoxifen (20 mg per day during 5 years) was given to 9 patients who were hormone receptor positive. These nine patients refused to continue adjuvant chemotherapy.
Radiotherapy was delivered to all patients, except to 35 patients.
After a median follow-up of 104 months (range, ≥ 51 to ≥ 161), 108 patients had disease recurrences (15 patients had local disease recurrence, 74 patients had systemic disease recurrence, and 19 patients had both local and systemic disease recurrence; Table 4). The median survival time was ≥ 158.16 months. The DFS rate at 8 years was 57.63% (Fig. 1).
|Events||Responders (%)||Nonresponders (%)|
|Local disease recurrence||11 (3.6)||4 (1.3)|
|Systemic disease recurrence||24 (7.9)||50 (16.4)|
|Local and systemic||7 (2.3)||12 (3.9)|
|Total disease recurrences||42 (13.8)||66 (21.7)|
|Deaths||25 (8.2)||46 (15.1)|
As shown in Figure 2, the DFS rate according to response was 70% (42 disease recurrences) for responders and 54% (66 disease recurrences) for nonresponders (P = 0.0048).
Local disease recurrences
Local disease recurrence as the first event occurred in 15 patients (5%). The median time between surgery and local disease recurrence was 28 months (range, 7–92 months).
Local disease recurrence occurred in 12 of 242 (4.95%) patients who received breast-conserving surgery and in 3 of 62 patients (4.83%) who received mastectomy. Of the 35 patients who had affected margins, only 2 developed local disease recurrence.
Local disease recurrences occurred in 11 responders (7.43%) compared with 4 (2.5%) nonresponders.
Distant disease recurrences
Distant disease recurrences were observed in 74 patients (24%). The median time between primary treatment and distant disease recurrence was 34 months (range, 7–82 months). The most frequent sites of disease recurrence were bone (n = 30), lung (n = 13), pleura (n = 6), liver (n = 6), and brain (n = 6). Thirteen patients had multiple sites of disease recurrences at the time of systemic failure.
Distant disease recurrences occurred in 24 responders (16.21%) compared with 50 nonresponders (31.44%).
Local and distant disease recurrences
Local and distant disease recurrences occurred concomitantly in 19 patients (6.2%). Local and systemic disease recurrences occurred in 7 responders (4.7%) compared with 12 nonresponders (7.5%).
At the time of the current report, 71 women have died of progressive breast carcinoma. Three other patients have died of unrelated causes. The median survival time was ≥ 158.16 months. The OS rate at 8 years was 67.65% (Fig. 3).
Patients without invasion of the lymph nodes had a median survival time of 129.43 months (range, 120.56–138.3 months), whereas patients with affected lymph nodes had a median survival time of 78.3 months (range, 46.23–110.36 months; P < 0.001). patients with tumors > 5 cm had a worse median survival: 160.53 months (range, 107.96–123.56 months) versus 64.4 months (range, 31–97.8 months). Survival was independent of nuclear grade and hormonal receptor status.
According to response, there were 25 deaths among responders versus 46 among nonresponders. The OS rate for responders was 90% versus 80% for nonresponders (P = 0.0042; Fig. 4).
PC was proposed initially for large tumors to avoid the need for mastectomies.22, 23 This practice is accepted widely.24 In contrast to Skipper's theory,25 which advocated a longer DFS through early eradication of distant metastases by delivering systemic treatment, randomized studies have not shown a benefit of PC on survival rates.8–13 The mounting interest for this approach remains not only in improving breast-conserving procedures, but also in predicting survival for patients with a response.
In the adjuvant setting, antracyclines have improved CMF results.26 The triplet of docetaxel, doxorubicin, and cyclophosphamide has emerged as an alternative chemotherapy regimen for adjuvant management of patients with lymph node-positive breast carcinoma. The most significant benefit from the use of docetaxel was observed among women with 1–3 positive lymph nodes (docetaxel reduced the risk of disease recurrence by 50%, as well as the mortality rate by 54% among these women). These results were supported by the findings of Henderson et al.27
In our series, patients were treated with CMF, the standard treatment in the 1990s. Breast-conserving rates were high, clinical reponse was moderate, and the 8-year survival rate was comparable to that of the Milan study,5 which comprised patients with similar characteristics (i.e., patients with large operable breast-infiltrating carcinomas).
Breast-conserving surgery was performed in 79.64% of the patients. The direct and continuous information between the clinician, radiologist, surgeon, and pathologist facilitates successful conservative procedures as indicated in a series of patients with large tumors like ours. It is remarkable that this high rate has been achieved with a low percentage of local disease recurrences (5%). Local disease recurrence rates were similar for patients receiving breast-conserving surgery and mastectomy (4.83% vs. 4.95%). Local disease recurrence was independent of margins and was lower in nonresponders (2.5% vs. 7.43%). This low disease recurrence rate among nonresponders can be explained because these patients present with distant disease recurrences before they develop local disease recurrences, whereas the local disease recurrence rate in responders is approximately the same as that described in the literature, i.e., a 1% annual rate in a series of patients who received breast-conserving surgery.
The overall response in our series is lower than that described in the literature. However, both the pCR and survival rates are similar. It is important to note that the clinical response was evaluated by the same clinician in all the cases according to the criteria described previously. This fact could explain the good correlation between the clinical and the radiologic assessment of response (89.1%). In addition, selection of patients according to response correlated significantly with outcome.
Survival rates were calculated after a median follow-up of 104 months (range, ≥ 51 to ≥ 161 months). The main clinical feature correlated with survival was the pathologic state of the axilla (P < 0.0001). Patient response provided additional information on outcome that was missing otherwise. Patients with response to chemotherapy had a significantly longer DFS and OS (P = 0.0048 and P = 0.0042, respectively).
In patients to be treated with PC, the evaluation of response is crucial because of its prognostic consequences. The more carefully patients with responding tumors are selected, the better we will be able optimize intense schemas of chemotherapy. A question to be answered is this: can pCR be achieved by prolonging treatment or by adding new drugs? Survival will be better in patients with a rapid response to treatment. For this reason, we should try to use the most effective drugs at the beginning of treatment.
Currently, CMF is an active and well tolerated treatment regimen that can still be offered to patients who cannot receive antracyclines or taxanes for any reason.
- 19The effect on tumor response of adding sequential preoperative docetaxel to preoperative doxorubicin and cyclophosphamide: preliminary results from National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol. 2003; 21: 4165–4174., , , et al.