© 2002 AlphaMed Press
Preoperative Therapy in Breast Cancer: Lessons from the Treatment of Locally Advanced DiseaseThe Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins and The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA Correspondence: Antonio C. Wolff, M.D., The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Bunting-Blaustein Cancer Research Building, Room 189, 1650 Orleans Street, Baltimore, Maryland 21231-1000, USA. Telephone: 410-614-4192; Fax: 410-955-0125; e-mail: awolff{at}jhmi.edu
Access and take the CME test online and receive one hour of AMA PRA category 1 credit at CME.TheOncologist.com
The greater use of screening has changed the stage distribution of breast cancer, and an increasing number of patients are diagnosed with earlier stages of the disease. Still, locally advanced breast cancer (LABC) remains a major clinical problem in the United States and a common presentation in many parts of the world. There is no standard definition of LABC. One commonly used includes patients with large primary tumors greater than 5 cm (T3) or with skin/chest wall involvement (T4), and/or fixed axillary (N2) or ipsilateral internal mammary (N3) lymph node involvement. According to the tumor node metastasis staging, these usually include stage IIIa (T0-2N2 or T3N1-2) and stage IIIb (T4Nx or TxN3) disease. Inflammatory breast cancer (T4d) is included in most classifications despite its distinct clinical behavior and worse prognosis overall, but it serves as an example of combined modality intervention. Historically, the term LABC has been applied to those clinical presentations where the disease is considered inoperable. However, these therapeutic principles (including preoperative or primary systemic therapy [PST]) are increasingly being applied to patients presenting with tumors greater than 5 cm and negative lymph nodes (stage IIbT3N0) or even smaller tumors, who are considered to have operable disease and a better outcome than those traditionally classified as having LABC. PST is increasingly being used in otherwise operable stage I and II patients aiming at greater rates of breast conservation and earlier efficacy assessment. This article reviews many of these issues and ongoing research questions. Key Words. Preoperative therapy • Neoadjuvant therapy • Locally advanced disease
Until the middle of the 20th century, the primary treatment of locally advanced breast cancer (LABC) was radical mastectomy [1]. This did not change until Haagensen and Stout identified markers of poor outcome, such as skin ulceration, edema, and tumor fixation [2]. Radiation therapy alone or followed by surgery became the primary treatment for these patients with a better locoregional control [3], but with major complications, such as brachial plexopathy, skin ulceration, and rib necrosis and minimal impact on overall survival [4]. The Oxford Overview confirmed that the use of adjuvant systemic therapy to treat micrometastatic disease in the adjuvant setting significantly reduces the risk of breast cancer recurrence and death [5,6]. Some of the trials included patients with operable stage III disease, and these therapeutic principles have also been applied to the subgroup of patients with LABC. As a result, systemic therapy was added to the locoregional treatment of these patients with the goal of improving tumor resectability, maintaining locoregional control, and eradicating micrometastatic systemic disease. For the most part, a multidisciplinary approach with combined modality therapy (radiation, surgery, and systemic therapy) became the standard approach to most patients with LABC. Beginning in 1988, patients with disease extending to the supraclavicular nodes were reclassified from N3 (stage IIIb) to M1 (stage IV) disease [7]. However, these patients with no evidence of other systemic disease are often rendered disease free with the use of combined modality therapy, and large databases indicate that these patients have an outcome quite similar to other patients with stage III disease [8]. Most of the experience in the treatment of LABC has been reported as small trials or case series from single-institutions, with limited data available from multi-institution, randomized trials. As a result, treatment guidelines are mostly derived from level III ("well-designed, quasi-experimental studies such as nonrandomized, controlled, single group, pre-post, cohort, and time or matched case-control series") or level V ("case reports") types of evidence. In preparing this article, we also tried to review the limited available data from randomized trials (level II evidence). However, many of these trials do not clearly differentiate between operable and inoperable LABC.
The Cancer and Leukemia Group B compared chemotherapy followed by surgery with chemotherapy followed by radiation therapy (RT) in patients with inoperable disease [9]. One hundred and thirteen patients (67% with stage IIIb disease) received 3 months of induction chemotherapy with CAFVP (cyclophosphamide, doxorubicin, 5-fluorouracil, vincristine, and prednisone). Eighty-one percent were deemed operable and randomized to either surgery or radiation, followed by an additional 2 years of CAFVP. At a median follow-up of 37 months, there was no significant difference in either disease-free survival (DFS) (29.2 versus 24.4 months) or overall survival (OS) (39.3 versus 39 months) between the two treatment groups [9]. Investigators in Milan treated 132 patients with three cycles of systemic therapy (doxorubicin and vincristine) followed by surgery or radiation therapy, then an additional seven cycles of the same chemotherapy [10]. Although the group randomized to surgery following initial chemotherapy had better initial control than those treated with radiation therapy, the short- and long-term outcomes in both groups were essentially the same after completion of the additional cycles of systemic therapy given to both groups.
Three studies have documented the potential benefit of systemic therapy in addition to RT in LABC. The European Organization for Research and Treatment of Cancer showed the added benefit of chemotherapy (12 cycles of cyclophosphamide, methotrexate, and 5-fluorouracil [CMF]) and hormonal therapy (ovarian ablation or tamoxifen) compared with RT alone in 410 patients, most of whom had stage IIIb disease [11]. A smaller study, from The Netherlands, showed a possible benefit of adding chemo/hormonal therapy to RT [12], but another small trial from Edinburgh did not show any advantage of adding CHOP-based chemotherapy to RT in patients with T4 lesions [13].
The Nottingham group compared induction chemotherapy, surgery, RT, and tamoxifen against tamoxifen alone with further therapy at the time of relapse in 108 patients with LABC [14]. Despite better local control at 6 months (57% versus 36%), overall survival at 30 months was no different, suggesting the possibility that micro/macrometastatic disease was present at diagnosis in these patients [14].
The impact of RT, when added to surgery and chemotherapy in operable breast cancer, has been shown by several trials. Overgaard et al. published data on two Danish trials: trial 82b in 1,708 premenopausal women with stage II/III breast cancer treated with mastectomy and CMF for eight cycles with or without RT [15]; and trial 82c in 1,375 postmenopausal women with similar stage disease treated with mastectomy and tamoxifen for 1 year with or without RT [16]. A survival benefit was seen in both studies with a median follow-up of 10 years. A third study, done by the Eastern Cooperative Oncology Group, examined the addition of RT to mastectomy and 6 months of CAFTH (cyclophosphamide, doxorubicin, 5-fluorouracil, tamoxifen, and aminoglutethimide) chemotherapy in 332 patients with T4N0-2, T3N1-2, and T1-2N2 disease [17]. In that study, patients treated with RT had a better local control with worse short-term systemic control, but similar overall survival.
Lessons from trials in LABC have generated substantial interest in the potential use of preoperative or primary systemic therapy in earlier stage patients with operable disease. The long-term success of adjuvant systemic therapy can only be confirmed by default when patients reach arbitrary milestones with no evidence of disease recurrence, as there have been no good surrogate markers for disease-free survival or overall survival. The existing data on the reduction in the risk of death or breast cancer recurrence with adjuvant systemic therapy apply to a whole population at risk, but do not help us with the identification of specific patients at risk. Preclinical data suggested a potential deleterious effect from the traditional approach of delaying systemic therapy until after resection of the primary tumor. This was based on the assumption that early exposure to systemic therapy would lead to less favorable growth kinetics for micrometastases and minimize the development of drug resistance [18,19]. PST also offers the potential to carry out pathologic and clinical assessment of tumor response, which could lead to the identification of those patients still at risk and who might (or not) benefit from additional therapy. Consequently, the high rates of response observed with systemic therapy in LABC led investigators to consider its use in operable patients with stage I and II disease prior to any locoregional intervention with the goal of improving rates of breast conservation and overall survival. Several nonrandomized series confirmed the significant activity of PST in operable breast cancer with overall response rates ranging from 69%-100%, pathologic complete response (pCR) around 12%, and high rates of breast conservation [2023]. Although these results were observed with various chemotherapy regimens, recent data appear to indicate that the choice of systemic therapy may affect the rate of pCR. For instance, Green et al. described a higher rate of pCR following preoperative therapy using paclitaxel on a weekly schedule rather than every 21 days [24]. Antiestrogens are an effective alternative to chemotherapy in patients with hormone receptor-positive LABC. Hoff et al. described the significant benefits from primary use of tamoxifen in women older than age 75 or those with major comorbid conditions [25]. The response rate was 47%, and ultimately, 62% were rendered disease free with surgery. Ellis et al. compared the aromatase inhibitor letrozole with tamoxifen in postmenopausal patients who were ineligible for breast-conserving surgery [26]. Both the response (55% versus 36%) and breast conservation rates (45% versus 35%) favored the letrozole group. Several randomized trials have compared preoperative with postoperative chemotherapy-based regimens, with standard regimens given either before (PST) or after surgery [2733]. The largest of all studies comparing primary with postoperative therapy was the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-18 trial [34,35]. In that study, 1,523 women with palpable, operable breast cancers (T1-3 N0-1 M0) diagnosed by fineneedle aspiration (FNA) or core biopsy were randomized to receive four cycles of doxorubicin and cyclophosphamide (AC) given either before or after surgery. The overall clinical response rate in the breast after primary therapy was 80% (clinical complete response [cCR] 36%, pCR 9%) [35]. Tumor size and clinical nodal status were independent predictors of cCR. Eighty-nine percent of patients with palpable axillary adenopathy at diagnosis had a nodal response (cCR 73% and pCR 32%), which represented a 37% increase in the number of patients with pathologically negative nodesa 17% absolute increase in the percentage of patients with pathologically negative lymph nodes from 43%-60%. Prior to randomization, lumpectomy was considered possible for 86% of women with T1 tumors, 70% with T2 tumors, and 3% with T3 tumors. Patients in the PST group had a slightly higher rate of treatment with lumpectomy (67% versus 60% in the standard postoperative group, p = 0.002) [35]. Survival is ultimately the most important outcome but, thus far, no studies have shown a survival benefit of PST over standard postoperative therapy. In NSABP B-18, the 5-year DFS (67.3% versus 66.7%), distant disease-free survival ([DDFS] , 73.2% versus 73.3%), and OS (80% versus 79.6%) rates between the two groups were essentially the same [34]. There were some concerns about the risk of ipsilateral breast tumor recurrence (IBTR) in the group of patients initially felt to be candidates for mastectomy, but who had breast conservation after PST. Among the 60% (435 patients) of those randomized to the control arm (surgery followed by AC) and treated with lumpectomy, 6.9% (30/435) had IBTR. The rate of IBTR was higher in the investigational arm (AC followed by surgery). Sixty-seven percent of patients in this arm ultimately had breast conservation, including 69 among 256 patients who had been considered mastectomy candidates at registration, for a conversion rate of 27%. Among these 69 patients, 10 (14.5%) had IBTR. When all patients in both arms are included, the overall rate of IBTR was essentially the same in the two arms (standard arm 7.1%, PST arm 9.9%) and did not impact the OS of the group treated with PST. Perhaps the most important finding in this trial was the correlation between pathologic response to primary therapy and OS [34]. Those patients who had a cCR had better 5-year results than all those with any lesser response in terms of DFS (cCR 75.9% versus clinical partial response [cPR] 63.5% versus clinical no response [cNR] 60.3%, p = 0.0014), relapse-free survival (RFS) (cCR 80.0% versus cPR 68.1% versus cNR 63.9, p = 0.0001), and DDFS (cCR 79.3% versus cPR 72.5% versus cNR 65.0%, p = 0.005). However, cCR did not correlate with better 5-year OS (cCR 81.5% versus cPR 78.4% versus cNR 76.9%, p = 0.19). A major survival impact was only observed in those patients who had a pCR at surgery, with a survival rate better than even patients with a cCR and just microscopic residual disease. This correlation between pCR and clinical outcome is of interest, particularly if these parameters could be used as surrogate or intermediate biomarkers for evaluating the effect of chemotherapy on systemic micrometastases. NSABP recently completed accrual to the largest of all trials attempting to address these questions [36]. In NASBP B-27, 2,411 women with palpable, operable breast cancer were randomized to four cycles of AC followed by surgery (arm 1), four cycles of AC followed by four cycles of docetaxel followed by surgery (arm 2), or four cycles of AC followed by surgery followed by four cycles of docetaxel (arm 3). All patients aged 50 years and older were also started on 5 years of tamoxifen. The primary end point of this study is whether the addition of docetaxel can improve DFS and OS compared with treatment with AC alone. Secondary goals are to increase locoregional response and pCR rates, improve lymph node downstaging, and increase the number of patients treated with breast conservation. Preliminary results discussing the pathology findings after PST (ACarms 1 and 3 with 1,606 patients, or AC followed by docetaxelarm 2 with 805 patients) were recently presented [36]. At presentation, 45% of the patients had tumors >4.0 cm, 30% had clinically palpable axillary lymph nodes, and 50% were considered breast conservation candidates. Clinical response (85% versus 91%, p <0.001) and pCR (13.7% versus 25.6%) favored the AC-docetaxel group. The observed rates of residual ductal carcinoma in situ (3.9% and 6.9%)included in the pCR figuresand the percentage of patients with uninvolved lymph nodes after PST (51.5% versus 59.5%) also favored the AC-docetaxel group. There was a correlation between breast and axillary response. Among the patients with a pCR in the breast, 13% had pathologic disease in the axilla at surgery. This compares with 45% in those with a cCR but residual pathologic disease in the breast, and 55%-61% in the others with less than cCR in the breast. The increase in lumpectomy rate was similar in both groups (from 50% to 61% in the AC group, and from 50% to 63% in the AC-docetaxel group), but with a significant increase in grade 3-4 toxicities (23% versus 40%, respectively) [36]. One question raised by this study is the issue of duration. Patients in arms 1 and 3 received four cycles of chemotherapy before surgery (AC), while patients in arm 2 received twice as many cycles (AC followed by docetaxel). Thus, it is difficult to ascertain whether the higher pCR observed with the addition of docetaxel is a result of longer duration of therapy, the addition of the taxane particularly, or a combination of the two factors. In addition, longer follow-up is needed to ascertain any survival benefit. In the meantime, investigators in Scotland have examined the activity of an additional four cycles of preoperative CVAP or docetaxel in 97 (67%) of 145 patients with large tumors (T >3 cm or LABCT3, T4, TxN2 disease) who had an objective response after an initial four cycles of preoperative CVAP [37]. In this small trial, both cCR (34% versus 62%) and pCR (18% versus 34%) at the time of surgery following eight cycles of chemotherapy favored the patients randomized to the docetaxel arm. These data do not address the issue of duration of PST, but indicate the lack of cross-resistance between an anthracycline-based regimen and docetaxel. Preliminary results appear to indicate a survival advantage among the women who initially responded to CVAP and were randomized to docetaxel versus continuing on CVAP before surgery.
Important questions on laboratory correlate studies are being addressed by NSABP B-27 [38]. NSABP B-27.1 is assessing the impact of therapy on serum levels of HER2/neu extracellular domain and anti-HER2/neu antibodies over time, as well as the prognostic and predictive value of these circulating markers. NSABP B-27.2 is evaluating if tumor biomarkers obtained at diagnosis (FNA) and at surgical resection can predict response to primary chemotherapy, long-term outcome, and how levels of expression are affected by primary chemotherapy. These assays include estrogen receptor/progesterone receptor status, nuclear grade, levels of expression of p53 and HER2, and markers for cell proliferation, multidrug resistance, and apoptosis. Successful completion of this trial and its ancillary studies should provide much needed information on the value of clinical response and several serum and tissue markers as predictors of clinical outcome.
Aside from the data discussed above from trials B-18 and B-27 in operable breast cancer, large datasets on the treatment of LABC from institutions such as the M. D. Anderson Cancer Center (MDACC; Houston, TX) provide important additional insight. Kuerer et al. showed that, among 30 of 163 patients (19%) treated with PST using FAC (5-fluorouracil, cyclophosphamide, and doxorubicin), 19 (63%) also had pathologically negative axilla compared with 33% or less of those with less than pCR in the breast [39]. They also showed that in those with a positive axillary FNA at diagnosis, 23% were found at surgery after PST to have pCR in the axilla, but the negative predictive value of clinical exam plus ultrasonography after PST was quite limited (
The MDACC group is also addressing the role of sentinel lymph node (SLN) biopsy following PST, which some consider a relative contraindication for this form of axillary staging [42]. Breslin et al. successfully mapped the SLN in 43 patients among 51 stage II and III patients [43]. Twenty-two of 43 patients (51%) had involved SLNs, and the SLN was the only site of axillary disease in 10 of those 22 patients (45%). All patients had an immediate axillary dissection, and three patients with a negative SLN procedure were found to have disease in the axilla for an overall sensitivity of 88%. Concerns about adequate local control have led to the often routine recommendation for mastectomy without immediate reconstruction as the primary surgical approach in patients with LABC who are ultimately deemed operable. However, single-institution data suggest that a less stringent approach might be possible. Singletary et al. showed that breast conservation is possible in up to 25% of patients with LABC following PST [44]. Factors that appear to identify those individuals include resolution of skin edema, lack of multicentric disease, and no pathologic evidence of lymphatic/vascular involvement. Newman et al. also showed that immediate reconstruction is not associated with either a significant delay in subsequent treatment or an increased risk of local systemic relapse [45]. Despite the preliminary data on SLN biopsy in LABC, further data are required to indicate whether PST can safely precede axillary node assessment. It is unclear whether the proper timing of axillary staging should be before PST, due to concerns about uneven response within the axilla in a given patient. Many institutions who adopted SLN biopsy as a standard procedure suggest that it be done immediately after the diagnosis and before PST. Others would argue that the axillary status will not influence treatment decisions in these patients, who are likely being treated with combined modality therapy. An important unanswered question is whether axillary RT can substitute for routine lymph node dissection in patients with positive SLN after PST.
Largely based on single-institution trials and a few randomized studies, multimodality therapy has become the standard approach for patients with LABC. It is important to realize that there are few data to support aggressive multimodality therapy in those patients with evidence of metastatic disease at presentation. An exception appears to be those patients whose only evidence of systemic disease is the presence of supraclavicular lymph node involvement, as their survival appears to be no different than that observed among other patients with stage III disease1. While there is indirect evidence suggesting that perhaps a small minority of patients with so-called oligometastatic disease (e.g., isolated site of visceral metastasis) could have a prolonged DFS (or perhaps even be cured) if aggressively treated with combined multimodality therapy with curative intent [46], only properly conducted randomized trials can put this issue to rest and address whether such approach can offer greater long-term benefit when compared with therapy with palliative intent. In the meantime, in vivo assessment of tumor response (tissue and/or imaging) is likely to improve clinical decision making and outcome, though its potential has yet to be fulfilled. As a result, PST should not routinely replace standard adjuvant approaches in operable breast cancer. At present, it allows a higher rate of breast conservation and is an acceptable alternative to women with palpable tumors. Perhaps more important, it offers an excellent arena for clinical trials and for the testing of novel interventions and compounds in the treatment of operable breast cancer.
Presented in part at the Mayo Clinic's 11th Annual Hematology/Oncology Reviews, State-of-the-Art Treatment of Hematology/Oncology Treatments, Amelia Island, Florida, August 16-19, 2001.
1 The AJCC released in May 2002 the 6th edition of the AJCC Cancer Staging Manual with a revised TNM staging, in which patients with ipsilateral supraclavicular node metastases and no other distant metastases will once again be classified as N3 (stage IIIb) rather than M1 disease.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||