Fax: (713) 794-4200
Combined-modality treatment for isolated recurrences of breast carcinoma
Update on 30 years of experience at the University of Texas M. D. Anderson Cancer Center and assessment of prognostic factors
Article first published online: 26 JUL 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 6, pages 1158–1171, 15 September 2005
How to Cite
Hanrahan, E. O., Broglio, K. R., Buzdar, A. U., Theriault, R. L., Valero, V., Cristofanilli, M., Yin, G., Kau, S.-W. C., Hortobagyi, G. N. and Rivera, E. (2005), Combined-modality treatment for isolated recurrences of breast carcinoma. Cancer, 104: 1158–1171. doi: 10.1002/cncr.21305
- Issue published online: 31 AUG 2005
- Article first published online: 26 JUL 2005
- Manuscript Accepted: 23 MAY 2005
- Manuscript Revised: 25 APR 2005
- Manuscript Received: 18 APR 2005
- Aventis Pharmaceuticals
- breast carcinoma;
- isolated recurrence;
- Stage IV-NED;
- combined modality
In three prospective, single-arm studies, the authors previously showed an improved outcome for anthracycline-naïve patients with isolated sites of recurrent breast carcinoma (BC) who were treated with doxorubicin-based chemotherapy after local therapy (surgery and/or radiotherapy). In the current report, the initial results are presented from a Phase II trial of docetaxel (100 mg/m2 every 21 days for 6 cycles) given after local therapy for recurrent BC (Stage IV BC with no evidence of clinically measurable disease) in patients who received prior adjuvant anthracycline-based chemotherapy, and the authors provide an update of the 3 previous studies. An analysis of prognostic factors for these patients also is presented.
Eligibility criteria for all studies included histologic proof of recurrent BC that had been resected and/or irradiated with curative intent. Survival was calculated using the Kaplan–Meier method. Univariate survival analyses were performed to test for associations between patient characteristics and outcome (log-rank test). Cox proportional hazards models were used to determine the multivariable correlations between patient characteristics and outcome.
The median follow-up for the docetaxel-based trial (n = 26 patients) was 45 months. Early outcomes for this study are promising. The median disease-free survival (DFS) was 44 months, and the 3-year DFS and overall survival (OS) rates were 58% and 87%, respectively. In the 3 doxorubicin-based studies, the median follow-up was 121.5 months for all living patients, and the estimated 20-year DFS and OS rates were both 26%. On multivariable analysis of patients from all 4 studies, the only significant prognostic factor for DFS and OS (P = 0.0006) was the number of involved axillary lymph nodes at initial diagnosis.
A proportion of patients with isolated BC recurrences achieved prolonged DFS with combined-modality treatment. Patients who receive anthracycline-based chemotherapy at primary diagnosis may benefit from local treatment followed by docetaxel-based chemotherapy for isolated recurrences. The only significant independent prognostic factor was the number of involved axillary lymph nodes at initial diagnosis. Cancer 2005. © 2005 American Cancer Society.
Recurrent breast carcinoma at an isolated locoregional or distant site may be amenable to surgical resection and/or irradiation with curative intent. This is feasible in 1–10% of patients with newly diagnosed Stage IV breast carcinoma.1 After such treatment, patients are considered to have Stage IV breast carcinoma with no evidence of clinically measurable disease (Stage IV-NED). To our knowledge, data concerning the natural history of Stage IV-NED breast carcinoma are limited. Most researchers have focused on patient outcome after local treatment for locoregional recurrence. After local treatment only for locoregional recurrence, 30–40% of patients develop widely metastatic disease within 3 months, and 50–80% develop metastatic disease within 2 years.1–8 The 5-year survival rate reported in retrospective studies ranges from 4% to 36%.2, 4, 7, 9
To our knowledge, few studies published to date have considered the role of adjuvant systemic treatment in patients with recurrent breast carcinoma who have been rendered Stage IV-NED with surgery and/or radiotherapy. We previously reported the results from three Phase II studies of doxorubicin-based chemotherapy for Stage IV-NED breast carcinoma in anthracycline-naïve patients that were conducted between 1974 and 1992.1, 10–12 Significant improvements in the median disease-free survival (DFS) and overall survival (OS) were observed in all three studies among the patients who received doxorubicin-based chemotherapy after local therapy compared with a historic control group that was treated between 1967 and 1976 with local therapy only at our institution by the same group of physicians. Univariate analyses found that patients with negative axillary lymph nodes at initial breast carcinoma diagnosis had significantly greater OS and DFS rates compared with patients who had involved lymph nodes.11 No survival differences were found based on estrogen receptor (ER) status or primary tumor size. A multivariable analysis for prognostic factors was not performed. We and other groups have attempted to complete a randomized trial to demonstrate conclusively the contribution of cytotoxic chemotherapy to patient outcome in Stage IV-NED breast carcinoma. However, these studies failed to accrue adequate numbers of patients.
By the late 1990s, many patients with Stage IV-NED breast carcinoma had been treated previously for their primary tumor with adjuvant anthracycline-based chemotherapy. Because taxanes are active in metastatic breast carcinoma, including anthracycline-resistant disease, we assessed the efficacy of docetaxel in patients with Stage IV-NED breast carcinoma who had prior anthracycline exposure.13–17 The objective of this report was to present the initial survival results for this most recent study, to update the survival data for the three doxorubicin-based studies, and to provide a further analysis of prognostic factors in these patients.
MATERIALS AND METHODS
The three doxorubicin-based trials have been described in detail elsewhere.1, 10–12 Common inclusion criteria were histologic proof of recurrent breast carcinoma that had been resected surgically and/or irradiated with curative intent, and no clinical or radiologic evidence of disease at study entry (Fig. 1). Prior anthracycline exposure was not allowed.
Study participants were women with Stage IV-NED breast carcinoma who were seen at The University of Texas M. D. Anderson Cancer Center from 1998 to 2004. The protocol was approved by the Institutional Review Board. All participating patients signed written informed-consent forms before study entry. Eligible patients had histologic proof of recurrent breast carcinoma that had been resected and/or irradiated with curative intent, no evidence of disease on study entry, and a history of prior anthracycline-based adjuvant therapy (Fig. 1). Patients with resected brain metastases who had not received consolidation radiotherapy before study entry were required to do so on completion of chemotherapy. Prior taxane-exposure was an exclusion criterion. Other eligibility criteria included: age older 18 years; life expectancy > 3 months; Zubrod performance status < 2; adequate bone marrow, renal, and liver function; no concomitant malignancies or other serious illnesses; and no peripheral neuropathy > Grade 2 (National Cancer Institute Common Toxicity Criteria, version 2.0).18
Before the initiation of chemotherapy, all patients underwent a full evaluation with complete medical history, physical examination, radiologic assessment (chest X-ray, isotope bone scan, and computed tomography scan of abdomen and other areas if appropriate), and baseline blood tests for serum chemistry (including renal and liver profiles), complete blood count, and serum tumor markers (CA 27.29 and carcinoembryonic antigen). Women of childbearing potential had a serum pregnancy test.
Patients were to receive 6 cycles of 100 mg/m2 of docetaxel infused intravenously over 1 hour every 21 days or as soon as blood counts had recovered and/or all other toxicities had resolved to Grade 1 (National Cancer Institute Common Toxicity Criteria). After completion of chemotherapy, patients with hormone receptor-positive tumors or tumors of unknown hormone receptor status who had not previously demonstrated resistance to tamoxifen were to be given oral tamoxifen at a dose of 20 mg once daily for 5 years (resistance was defined as recurrence on or within 6 months of withdrawing tamoxifen). The protocol was modified later to allow postmenopausal women to receive an aromatase inhibitor. Patients were to be reviewed at 3 months, 6 months, 9 months, 12 months, 24 months, 36 months, 48 months, and 60 months after their final treatment and were assessed for recurrence according to the American Society of Clinical Oncology breast carcinoma surveillance guidelines.19
The primary objective of this docetaxel-based trial was to evaluate the efficacy of 6 cycles of docetaxel (either alone or followed by 5 years of hormonal therapy) in patients with Stage IV-NED breast carcinoma who failed on anthracycline-based adjuvant chemotherapy. The primary endpoint was survival of patients from the time of local treatment for their recurrence.
We estimated that patients who had Stage IV-NED breast carcinoma that had recurred after anthracycline-based adjuvant chemotherapy would have a 5-year survival of approximately 30%. We determined that the accrual of 45 evaluable patients would allow us to determine whether the combination of local therapy and docetaxel would result in a 50% OS rate at 5 years, a 20% improvement over the baseline expectation, with an α value of 0.05 and a power of 80%. The maximum width of the confidence interval would be + 14% from the observed survival rate at 5 years.
Patient characteristics were tabulated separately by protocol and for the three doxorubicin-based protocols combined. Continuous variables were described by their range and median. The median follow-up was calculated as the median observation time of all patients. The median follow-up of patients who remained alive was calculated as the median observation time of all patients who were alive at the last follow-up. OS was calculated from the date of local treatment for isolated recurrence to the date of death or last follow-up. DFS was calculated from the date of local treatment to the date of second disease recurrence or last follow-up. Patients who died without evidence of disease were considered censored at the date of death, and patients whose disease status was unknown at the time of death were considered censored at the date they were last known to be disease-free.
Breast carcinoma-specific survival (BCSS) was calculated for the patients on the doxorubicin-based studies. Because the exact cause of death was not known for all patients who died, it was assumed that all those who were known to have had a second disease recurrence had died from breast carcinoma, and it was assumed that patients with an unknown disease status at death also had died from breast carcinoma. It was expected that these assumptions would result in the most conservative estimate of BCSS for these patients. Patients who died without having experienced a disease recurrence were considered censored for BCSS at their date of death.
DFS, OS, and BCSS rates were determined by using the Kaplan–Meier product-limit method.20 Univariate survival analyses were performed for the docetaxel-based trial alone and then for all four trials combined. The log-rank test was used to test the associations between DFS and OS rates and patient characteristics. A multivariable analysis of all four studies combined was performed. Cox proportional hazards models were used to determine the multivariable correlations between patient characteristics and both DFS and OS rates.21 The fit of the model and the proportional hazards assumption was assessed visually with residual plots. A P value of 0.05 was considered statistically significant.
Two hundred eighty-five patients with Stage IV-NED breast carcinoma were treated on 1 of the 4 consecutive protocols between 1974 and 2004. Two hundred fifty-nine patients were treated on the 3 doxorubicin-based trials between 1974 and 1992, 26 patients were recruited to the docetaxel-based trial between January 1998 and December 2004. The docetaxel-based study is now closed due to slow accrual. Unlike the patients on the docetaxel-based study, all patients on the three earlier studies were anthracycline-naïve at the time of their Stage IV diagnosis, and only eight of these patients had received prior nonanthracycline-based adjuvant chemotherapy. Therefore, a separate survival analysis was performed for the patients who were treated on the docetaxel-based trial.
Table 1 summarizes the characteristics of the 26 patients. With regard to the staging of their primary breast carcinoma, 76% of patients had T1 tumors (≤ 2.0 cm in greatest dimension) or T2 tumors (> 2.0 cm but ≤ 5.0 cm in greatest dimension), and 46% of patients had no known metastatic involvement of the axillary lymph nodes (reasons for unknown lymph node status in 5 patients are provided in the footnote to Table 1).22 The median disease-free interval from the diagnosis of original breast carcinoma to the first recurrence was only 28.5 months. At the time of first recurrence, 54% of patients had a metastasis at a distant site.
|Patient characteristic||No. of patientsa||Proportion of patients (%)b||Patient characteristic||No. of patientsa||Proportion of patients (%)b|
|Total||26||—||First disease-free interval in (mos)|
|Age in (yrs)||Range||0–149||—|
|Median||51||—||Local therapy for recurrence|
|T classification||No. of docetaxel cycles received|
|Lymph node status of primary tumor||Hormonal therapy for Stage IV-NED|
|No. of positive lymph nodesc||Tamoxifen||5||—|
|Median||0||—||Status at last follow-up|
|Estrogen receptor status||A/NED||12||46|
|Negative||14||54||Sites of second recurrence|
|Site of first recurrence||Different||7||27|
The median follow-up for these patients was 45 months (range, 13–77 mos). The estimated 3-year and 5-year DFS rates were 58% (95% confidence interval [95% CI], 40.5–82%) and 34% (95% CI, 18–64%), respectively (Fig. 2). Fourteen patients had a second disease recurrence. Of these, seven patients developed recurrences at the same site or within the same organ system as the first recurrence; four patients had initial bone metastases, one patient had an initial chest wall recurrence, and two patients had initial lung metastases.
The median OS had not been reached at the time of last follow-up (Fig. 3). The estimated 3-year and 5-year OS rates were 87% (95% CI, 74–100%) and 59% (95% CI, 35–99%), respectively. Only six patients had died at the time of last follow-up, all due to a second disease recurrence. These six patients had their initial recurrences at distant sites, including the liver in two patients, bone in two patients, the brain in one patient, and the lung in one patient.
In the univariate analysis, factors that were found to be associated with a longer DFS were no axillary lymph node involvement at initial diagnosis (P = 0.04), age younger than 50 years at study entry (P = 0.02), premenopausal status at study entry (P = 0.049), and locoregional site of first recurrence (P = 0.05) (Table 2). The only factor that was associated with significantly longer OS was locoregional recurrence as the first site of recurrence (P = 0.02). Due to the small number of patients in this study, a multivariable analysis was not performed.
|Patient characteristic||No.||Overall survival||Disease-free survival|
|No. of events||Median (mos)||3-yr estimate (%)||95% CI (%)||P value||No. of events||Median (mos)||3-yr estimate (%)||95% CI (%)||P value|
|< 50 yrs||13||2||—||100||—||5||—||73||51–100|
|≥ 50 yrs||13||4||—||73||52–100||0.11||9||17||40||19–82||0.02|
|Lymph node statusa|
|Estrogen receptor status|
|< 24 mos||10||4||59||79||56–100||6||40||60||36–100|
|≥ 24 mos||16||2||—||92||79–100||0.20||8||45||55||34–91||0.83|
|Hormone therapy for Stage IV-NED|
Twenty-two patients (85%) completed 6 cycles of treatment. Four patients did not receive all cycles: one patient due to disease progression and three patients due to Grade 3 sensory neuropathy. There were no deaths due to toxicity. Only one patient experienced Grade 4 toxicity (rash). Seven patients (27%) had at least 1 episode of febrile neutropenia. Grade 2 or 3 sensory neuropathy occurred in 7 patients (27%).
Table 3 outlines the characteristics of patients enrolled on the 3 doxorubicin-based studies separately and combined. At the time they underwent surgery for their primary breast carcinoma, 77% of patients had T1 or T2 tumors, and 59% of patients had no metastatic involvement of axillary lymph nodes. The first isolated recurrence for 80% of patients was locoregional. The median time from primary breast carcinoma diagnosis to the first recurrence was 24 months. The ER status was known for only 125 patients (for nearly all patients on the more recent 2 studies but for only 18 patients on the earliest study), and 72 of those 125 patients (58%) had ER-positive tumors. Patients received hormonal therapy only in the two more recent studies.
|Characteristic||No. of patients (%)a|
|All 3 studies combined (1974–1992)||Study 1 (1974–1982)||Study 2 (1982–1988)||Study 3 (1988–1992)|
|Total no. of patients||259||134||80||45|
|Age in yrs|
|Postmenopausal||152 (59)||79 (59)||45 (56)||28 (62)|
|Premenopausal||102 (39)||50 (37)||35 (44)||17 (38)|
|T1||105 (41)||58 (43)||24 (30)||23 (51)|
|T2||93 (36)||58 (43)||17 (21)||18 (40)|
|T3||8 (3)||2 (1)||5 (6)||1 (2)|
|T4||12 (5)||11 (8)||1 (1)||0 (0)|
|Lymph node status of primary tumor|
|Negative||153 (59)||70 (52)||53 (66)||30 (67)|
|Positive||96 (37)||59 (44)||24 (30)||13 (29)|
|No. of positive lymph nodes|
|Estrogen receptor status|
|Negative||53 (20)||13 (3)||27 (34)||13 (29)|
|Positive||72 (28)||5 (2)||40 (50)||27 (60)|
|Distant||53 (20)||29 (22)||20 (25)||4 (9)|
|Bone||16 (6)||8 (6)||7 (9)||1 (2)|
|Viscera||31 (12)||18 (13)||11 (14)||2 (5)|
|Brain||6 (2)||3 (2)||2 (3)||1 (2)|
|LR||206 (80)||105 (78)||60 (75)||41 (91)|
|First DFI (mos)|
|Hormonal therapy for Stage IV-NED|
|No||56 (22)||—||43 (54)||13 (29)|
|Yes||69 (27)||1 (0.5)||37 (46)||31 (69)|
|Status at last follow-up|
|A/NED||44 (17)||14 (10)||13 (16)||17 (38)|
|A/WD||18 (7)||3 (2)||10 (13)||5 (11)|
|A/UNK||2 (1)||1 (1)||1 (1)||0 (0)|
|D/NED||23 (9)||16 (12)||5 (6)||2 (4)|
|D/WD||164 (63)||95 (71)||50 (63)||19 (42)|
|D/UNK||8 (3)||5 (4)||1 (1)||2 (4)|
|Sites of second recurrence|
The median follow-up for all patients was 87 months and, for the patients who alive at last follow-up, it was 212.5 months (range, 76–352 mos). In total, 182 patients (70%) have had a second disease recurrence. The median DFS was 42 months. The estimated DFS rates at 3 years, 5 years, 10 years, and 20 years were 57% (95% CI, 51–64%), 41% (95% CI, 36–48%), 34% (95% CI, 28–40%), and 26% (95% CI, 21–32%), respectively (Fig. 4). Among the patients who had a second recurrence, 145 patients had recurrences at a site that differed from that of the initial recurrence. Only 4 new recurrences have been recorded in the past 5 years, and 28 patients have remained free of disease for > 20 years since the local therapy for their first recurrence.
Of the 53 patients who had an initial distant site of recurrence, 12 patients (23%) had not developed a second recurrence at the time of last follow-up. The median DFS for these 12 patients was 207.5 months, and their initial sites of recurrence were bone (n = 1 patient), brain (n = 3 patients), extradural spinal deposit (n = 1 patient), and lung (n = 7 patients).
One hundred ninety-five patients (75%) had died, and the estimate of median survival was 87 months. The estimated OS rates at 3 years, 5 years, 10 years, and 20 years were 75% (95% CI, 70–81%), 56% (95% CI, 51–63%), 42% (95% CI, 36–48%), and 26% (95% CI, 21–33%), respectively (Fig. 5). The estimated BCSS rates at 10 years and 20 years are 45% (95% CI, 39–51%) and 31% (95% CI, 26–38%), respectively (Fig. 6).
In the univariate analysis of the 4 studies combined, the factors that were found to be associated significantly with longer OS were negative axillary lymph nodes at initial diagnosis (P = 0.002) and premenopausal status at study entry (P = 0.01) (Table 4). The factors that were found to be associated significantly with longer DFS were T1 disease at initial diagnosis (P = 0.04) compared with T2 disease or greater, negative axillary lymph nodes at initial diagnosis (P < 0.0001), and premenopausal status at study entry (P = 0.049). Patients with initial locoregional rather than distant sites of recurrence generally had longer DFS (P = 0.053) and OS (P = 0.051). ER status and hormonal therapy were not found to be associated significantly with outcome.
|Variable||No.||Overall survival||Disease-free survival|
|No. of events||Median (mos)||3-yr estimate (%)||5-yr estimate (%)||10-yr estimate (%)||15-yr estimate (%)||20-yr estimate (%)||P value||No. of events||Median (mos)||3-yr estimate (%)||5-yr estimate (%)||10-yr estimate (%)||15-yr estimate (%)||20-yr estimate (%)||P value|
|Lymph node status of primary tumor|
|Site of first recurrence|
|< 24 mos||135||95||90||71||56||42||35||27||90||48||55||43||36||34||28|
|≥ 24 mos||150||106||87||81||58||42||31||26||0.94||106||42||59||39||31||23||23||0.81|
|Hormone therapy for Stage IV-NED|
Table 5 shows the results from the Cox proportional hazards model. Both age at diagnosis and first disease-free interval were excluded as variables, because they were not significant in the univariate analyses. Furthermore, we believed that age at diagnosis may have been redundant with menopausal status and that menopausal status was more meaningful clinically. ER status and hormonal therapy were not included due to the large amount of missing data associated with these two variables. T classification was included as a dichotomous variable (T1 vs. T2–T4), because the dichotomous form of these variables better satisfied the proportional hazards assumption. Although clinical trial protocol was included in the models to account for unmeasurable patient characteristics, differences between patient populations over time suggest that treatment comparisons should not be made. Only the number of positive lymph nodes remained associated significantly with either OS or DFS (P = 0.0003 and P = 0.0001, respectively).
|HR||95% CI||P value||HR||95% CI||P value|
|Doxorubicin-based Trial 1||Reference||Reference|
|Doxorubicin-based Trial 2||0.70||0.46–1.07||0.1||0.83||0.55–1.26||0.38|
|Doxorubicin-based Trial 3||0.64||0.4–1.04||0.07||0.66||0.41–1.05||0.08|
|No. of positive lymph nodesa||1.07||1.03–1.1||0.0001||1.06||1.03–1.09||0.0003|
Patients with isolated recurrences of breast carcinoma may achieve prolonged DFS and OS with combined-modality treatment. According to the current results, the only significant independent prognostic factor is the number of involved axillary lymph nodes at initial diagnosis.
Patients with breast carcinoma who develop an isolated locoregional recurrence generally have an unfavorable prognosis, because the majority of these patients eventually die of distant metastatic disease. Surgery and/or radiotherapy have been regarded as the standard treatment for local control. However, local treatment with these modalities does not improve DFS for this patient population.4, 8, 23–25 This suggests that many of these patients have micrometastases at the time of their first recurrence. Consequently, studies were initiated to assess whether adequate local treatment to achieve Stage IV-NED followed by systemic therapy would improve long-term outcome for these patients.
We have reported the results of three Phase II studies of local therapy followed by doxorubicin-based chemotherapy in patients with Stage IV-NED breast carcinoma.1, 10, 11 We also compared the results in those patients with the results from a historic control group that was treated with local therapy only at our institution by the same group of physicians. The last update, in 2002, showed that the estimated 5-year OS rate was 36% for the control group and ranged from 48% to 67% for the patients on the 3 doxorubicin-based studies.11 Similarly, the estimated 5-year DFS rate was 7% for the control group but ranged from 36% to 59% for the patients on the doxorubicin-based studies. These single-arm studies from our institution have been criticized for the use of a historic control group that was treated from 1967 to 1976, because patients who were treated in earlier years may have had undetected distant metastases given the less sophisticated staging modalities available. However, as discussed earlier, to our knowledge to date, no group has successfully completed a Phase III trial of cytotoxic chemotherapy versus observation in patients with Stage IV-NED breast carcinoma. To our knowledge, the only completed Phase III randomized trial of systemic therapy for this disease stage is by Borner et al.26, 27 In their study, tamoxifen therapy was compared with observation alone after local treatment in 167 patients who had a locoregional recurrence of breast carcinoma. Those investigators found that tamoxifen significantly improved median DFS, but not OS.
To the best of our knowledge, the patients with Stage IV-NED breast carcinoma in our doxorubicin-based studies constitute the largest series of such patients to date with the longest follow-up. The current update of all patients from the 3 doxorubicin-based studies together demonstrates estimated 20-year DFS and OS rates of 26%, with only 4 new recurrences reported to have occurred in the past 5 years. Therefore, it is likely that many of these patients with long second disease-free intervals have been cured.
More recently, Nieto et al.28 conducted a Phase II prospective trial of high-dose chemotherapy and autologous stem cell transplantation for patients with Stage IV oligometastatic breast carcinoma. Fifty-nine of 60 patients in that trial received some form of local treatment for their metastases, 32 of whom were rendered Stage IV-NED prior to receiving the high-dose chemotherapy. At a median follow-up of 62 months, the median recurrence-free survival (RFS) and OS were 52 months and 80 months, respectively. The 5-year RFS and OS rates were 52% and 62%, respectively. Although theirs was a somewhat different patient population compared with our patients on the doxorubicin-based studies, the reported outcomes are similar.
To our knowledge, the results of our docetaxel-based trial are the first data to be presented regarding the use of taxanes in patients with Stage IV-NED breast carcinoma. Although the number of patients in the current study is small, their outcome at a median follow-up of 45 months is very good, particularly when one considers that these patients all received prior anthracycline-based therapy for their primary disease. These promising early results for docetaxel in patients with Stage IV-NED breast carcinoma who had recurrent disease after prior anthracycline-based adjuvant chemotherapy are in keeping with the findings of 3 randomized Phase III clinical trials of single-agent docetaxel in patients with anthracycline-resistant, metastatic breast carcinoma.16, 17, 29 In the adjuvant setting, docetaxel given in combination or in sequence with doxorubicin and cyclophosphamide for lymph node-positive primary breast carcinoma also has improved patient outcome.30, 31 The use of an aromatase inhibitor in approximately 25% of patients in the docetaxel-based study also may have had a positive influence on the outcome of the patient group, given the known advantage of aromatase inhibitors over tamoxifen in terms of DFS for the adjuvant management of primary breast carcinoma in postmenopausal women.32–35 It is interesting to note that the patients on the docetaxel-based study, despite having developed their first recurrence after adjuvant anthracycline-based chemotherapy, had a median DFS and 3-year DFS and OS rates after local treatment for first recurrence that were similar to those of the patients on the three doxorubicin-based studies, all of whom had not received anthracyclines previously. Of course, direct comparisons of outcome between the patients on the docetaxel-based study and patients on the three doxorubicin-based studies is not appropriate because of differences in modalities of disease staging between studies, differences in follow-up surveillance, and differences in the patient populations, most notably the higher proportions of patients with unknown ER status and with locoregional disease as the first site of recurrence in the doxorubicin-based studies. In addition, a greater proportion of patients in the docetaxel-based study received hormonal therapy. Nevertheless, these early results from the docetaxel-based study are encouraging considering the long-term OS and DFS rates observed in the doxorubicin-based studies, and it will be intriguing to learn how this patient group fares with longer follow-up.
In our docetaxel-based trial, the only significant prognostic factor for DFS and OS shown by univariate analysis was the site of initial recurrence; patients with locoregional recurrences had a more favorable outcome compared with patients with distant recurrences. In addition, the absence of axillary lymph node involvement at the time of initial diagnosis and an age younger than 50 years at study entry were found to be associated significantly with an improved DFS. These factors were not associated with an improved OS, however, but this may be due to the small number of patients, the relatively short follow-up, and the few deaths. This may also explain why we did not find that the use of hormonal therapy or HER-2/neu status was associated significantly with outcome. In contrast, Nieto et al. found that HER-2/neu overexpression was an independent, adverse risk factor for RFS and OS in their study group.26 In our univariate analysis of all four studies combined, the only significant prognostic factors for both DFS and OS were menopausal status and lymph node status at primary diagnosis. In our multivariable analysis of all four studies, the number of involved axillary lymph nodes at the time of primary breast carcinoma diagnosis was found to be the only significant independent prognostic factor for DFS and OS. HER-2/neu status was not available for the patients on the doxorubicin-based studies and therefore could not be included in the combined analyses of all four studies.
This importance of the lymph node status of the primary breast carcinoma as a prognostic factor after recurrence is consistent with the findings of a number of other studies.28, 36–46 For example, Juan et al. retrospectively analyzed 96 patients with Stage IV-NED breast carcinoma and found that both the absence of axillary lymph node involvement at the time of surgery for the primary tumor and the administration of systemic therapy after local treatment for isolated recurrence were associated with improved DFS and OS.44
We did not find that ER status was a significant prognostic factor for DFS or OS in the univariate analysis of all four studies, in keeping with the earlier results of Rivera et al.11 Because many patients in our earlier studies were treated before hormone receptor status was assessed routinely, ER status was available only for 151 patients. Nearly all patients with missing ER data were from our first doxorubicin-based study. Our analysis may be biased by the relatively small number of patients with known ER status and by the fact that those with known ER status are the more recently treated patients. However, it is interesting to note that, as shown in Table 4, the patients with ER-positive disease had a trend toward improved DFS in the first 5 years after local treatment for their first recurrence, and up to 10 years for OS, but not thereafter. This is consistent with the previously reported, time-dependent prognostic significance of ER status in early-stage breast carcinoma. A number of researchers have shown that patients with ER-positive primary breast carcinoma have an initially improved prognosis after treatment but that this advantage gradually disappears with increasing interval after primary treatment.47–50
Whether hormonal therapy was received was documented for only 53% of patients, mainly those on the 3 most recent studies in which such treatment was part of the study protocol. (Although it is likely that most of the patients on the earliest study did not receive tamoxifen, in the absence of definite documentation, we did not wish to assume that this was the case.) Among the patients for whom the use or nonuse of hormonal therapy was documented, we did not find that it was a significant prognostic factor in our univariate analyses. This is contrary to our expectations and the findings of Borner et al.26, 27 One possible explanation for this finding, as suggested by Rivera et al., is that only 69 patients on the doxorubicin-based studies are known to have received tamoxifen, and only 31 of those patients planned to take it for 5 years.11 Five years of adjuvant tamoxifen has been established as the optimal duration of treatment for early-stage breast carcinoma.50 Based on the study by Borner et al.26, 27 and the proven efficacy of hormonal therapy in the treatment of both primary and metastatic, hormone receptor-positive breast carcinoma, we believe that, despite our current findings, hormonal therapy should be considered in the treatment of all patients with hormone receptor-positive, Stage IV-NED breast carcinoma.
The International Breast Cancer Study Group (IBSCG) and the National Surgical Adjuvant Breast and Bowel Project (NSABP) currently are recruiting patients to a Phase III randomized study of adjuvant chemotherapy in women with resected locoregional recurrence of breast carcinoma (IBCSG 27-02, BIG 01-02, NSABP B37).52, 53 Although their study considers only a subset of patients with Stage IV-NED breast carcinoma, if it successfully accrues adequate patient numbers, then it may provide some conclusive evidence for the use of chemotherapy in patients with Stage IV-NED breast carcinoma. At a time when all but the very lowest-risk patients receive adjuvant chemotherapy for primary breast carcinoma in the hope of eradicating possible micrometastatic disease, it will continue to be difficult to accrue patients to Phase III studies of chemotherapy in Stage IV-NED breast carcinoma.54
Despite the current absence of data from a completed Phase III trial, we believe that systemic chemotherapy may play an important role in the management of patients with Stage IV-NED breast carcinoma. The choice of a specific chemotherapy regimen for a given patient should be made on the basis of any previous chemotherapy that the patient has received. As with our patients in the docetaxel-based trial, cytotoxic agents that are not cross-resistant with previously administered therapies should be used. Inspiration can be drawn from currently used regimens for patients with early stage and metastatic breast carcinoma.29–31, 54–58 Trastuzumab-containing regimens may be appropriate if patients have HER-2/neu-positive tumors.57, 58 Hormonal therapy either after chemotherapy or alone, if the patient's performance status is poor, also should be used for patients with hormone receptor-positive, Stage IV-NED breast carcinoma.
It is likely that data from a successfully completed Phase III trial of chemotherapy versus observation in patients with Stage IV-NED breast carcinoma never will be obtained. Perhaps the real issues for future research in this area are how to identify those patients with Stage IV-NED breast carcinoma who are most likely to benefit from systemic therapy and which agents are optimal for each patient. Molecular or genetic profiling technology may provide the answers.
- 10Combined modality approach for patients with isolated recurrences of breast cancer (IV-NED): the M. D. Anderson Cancer Center experience. Breast Dis. 1994; 7: 7–20., , , et al.
- 17Docetaxel compared with sequential methotrexate and 5-fluorouracil in patients with advanced breast cancer after anthracycline failure: a randomized Phase III study with crossover on progression by the Scandinavian Breast Group. Eur J Cancer. 1999; 35: 1194–1201., , , et al.
- 18National Cancer Institute. Common Toxicity Criteria of the National Cancer Institute, version 2.0. Available at URL: http://ctep.info.nih.gov/reporting/CTC-3.html [accessed February 7, 2005].
- 21Regression models and life tables (with discussion). J R Stat Soc B. 1972; 34: 187–220..
- 22GreeneFL, PageDL, FlemingID, et al., editors. AJCC cancer staging handbook. TNM classification of malignant tumors, 6th ed. New York: John Wiley & Sons, 2002.
- 30Five years analysis of the PACS 01 trial: 6 cycles of FEC100 versus 3 cycles of FEC100 followed by 3 cycles of docetaxel for the adjuvant treatment of node positive breast cancer [abstract]. Breast Cancer Res Treat. 2004; 88(Suppl 1): S16–S17., , , et al.
- 33ATAC Trialists' Group. The ATAC (Arimidex, Tamoxifen, Alone or in Combination) trial in postmenopausal women with early breast cancer—updated efficacy results based on a median follow-up of 5 years [abstract]. Breast Cancer Res Treat. 2004; 88(Suppl 1): S7., on behalf of the
- 52International Breast Cancer Study Group. Available at URL: http://www.ibcsg.org/public/general_pages/trials/open/pub_trials_open.shtml [accessed February 7, 2005].
- 53National Surgical Adjuvant Breast and Bowel Project. Available at URL: http://www.nsabp.pitt.edu/B-37.htm [accessed February 7, 2005].
- 54National Comprehensive Cancer Network. Clinical practice guidelines in oncology, version 1, 2005: breast cancer. Available at URL: http://www.nccn.org/professionals/physician_gls/PDF/breast.pdf [accessed February 7, 2005].
- 56Paclitaxel following doxorubicin/cyclophosphamide as adjuvant chemotherapy for node-positive breast cancer [abstract]. Proc Am Soc Clin Oncol. 2003; 22: 12a., , , et al.
- 57Chemotherapy guidelines for metastatic breast cancer. In: SingletarySE, RobbGL, HortobagyiGN, editors. Advanced therapy of breast disease, 2nd ed. Hamilton, Ontario: B.C. Decker Inc., 2004: 557–566., .