Proceedings of the Consensus Conference on Neoadjuvant Chemotherapy in Carcinoma of the Breast, April 26–28, 2003, Philadelphia, Pennsylvania


  • Gordon F. Schwartz M.D., M.B.A.,

    Corresponding author
    1. Department of Surgery, Jefferson Medical College, Philadelphia, Pennsylvania
    • Department of Surgery, Jefferson Medical College, 1015 Chestnut Street, Suite 510, Philadelphia, PA 19107-4305
    Search for more papers by this author
    • Fax: (215) 627-3925

  • Gabriel N. Hortobagyi M.D.,

    1. Department of Breast Medical Oncology, M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Consensus Conference Committee

    Search for more papers by this author
    • The following were participants in the Consensus Conference on Neoadjuvant Chemotherapy in Carcinoma of the Breast: Co-Chairmen: Dr. Gordon F. Schwartz (Jefferson Medical College, Philadelphia, PA) and Dr. Gabriel N. Hortobagyi (M.D. Anderson Cancer Center, Houston, TX). Surgeons: Dr. Blake Cady (Brown University School of Medicine, Providence, RI), Dr. Krishna B. Clough (Institute Curie, Paris, France), Dr. Domenico M. D'Ugo (Universita Cattolica del Sacro Cuore, Rome, Italy), Dr. Laura J. Esserman (University of California San Francisco, San Francisco, CA), Professor Ian S. Fentiman (Guy's Hospital, London, UK), Dr. Oreste Gentilini (European Institute of Oncology, Milan, Italy), Dr. Armando E. Giuliano (John Wayne Cancer Institute, Santa Monica, CA), Dr. Terry P. Mamounas (Aultman Center Center, Canton, OH), Dr. Peter I. Pressman (Weill Medical College of Cornell University, New York, NY), Dr. Emiel J. Th. Rutgers (Netherlands Cancer Institute, Amsterdam, The Netherlands), Dr. Gordon F. Schwartz (Jefferson Medical College, Philadelphia, PA), Dr. Daniela Terribile (Universita Cattolica del Sacro Cuore, Rome, Italy), Dr. Theodore N. Tsangaris (The Johns Hopkins Hospital, Baltimore MD), Dr. Michael Untch (University of Munich Klinikum Grosshadern, Munich, Germany), and Professor Cornelis J. H. van de Velde (Leiden University Medical Center, Leiden, The Netherlands). Medical Oncologists: Dr. Aman U. Buzdar (M. D. Anderson Cancer Center, Houston, TX), Dr. Daniel F. Hayes (University of Michigan Medical School, Ann Arbor, MI), Dr. Gabriel N. Hortobagyi (M. D. Anderson Cancer Center, Houston, TX), Dr. Clifford Hudis (Memorial Sloan-Kettering Cancer Center, New York, NY), Dr. Olivier Rixe (Salpetriere Hospital, Paris, France), Professor Ian Edward Smith (The Royal Marsden NHS Trust, London, UK), and Dr. Sandra M. Swain (Cancer Therapeutics Branch, National Cancer Institute, Bethesda, MD). Radiation Oncologists: Dr. Thomas A. Buchholz (M. D. Anderson Cancer Center, Houston, TX), Dr. Lydia T. Komarnicky (Drexel University College of Medicine, Philadelphia, PA), Dr. Robert R. Kuske, Jr. (Arizona Oncology Services, Scottsdale, AZ), Dr. Beryl McCormick (Memorial Sloan-Kettering Cancer Center, New York, NY), and Dr. Lawrence J. Solin (University of Pennsylvania School of Medicine, Philadelphia, PA). Pathologists: Dr. Fred Gorstein (Jefferson Medical College, Philadelphia, PA), Dr. Roland Holland (University Hospital Nijmegen, Nijmegen, The Netherlands), Dr. Shahla Masood (University of Florida Health Science Center, Jacksonville, FL), Dr. David L. Page (Vanderbilt University Medical Center, Nashville, TN), and Dr. Juan P. Palazzo (Jefferson Medical College, Philadelphia, PA). Breast Imagers (Radiologists): Dr. R. James Brenner (John Wayne Cancer Institute, Santa Monica, CA), Dr. Stephen A. Feig (Mt. Sinai Medical Center, New York, NY), and Dr. Daniel B. Kopans (Massachusetts General Hospital, Boston, MA).


A consensus conference regarding the role of neoadjuvant chemotherapy in patients with carcinoma of the breast was held in April 2003 in Philadelphia to review the evidence in support of this form of treatment, discuss the issues related to its routine application, and establish more formal guidelines for its use in patients with all stages of breast carcinoma.

Despite the advances of the past generation in the earlier detection of breast carcinoma, all too often women still present with what is generally termed locally advanced disease. Previously, depending on the staging system employed, these women were considered “inoperable” or, at the very least, not curable by any of the existing surgical techniques. Palliative therapy programs generally were employed, recognizing that the outcomes for these patients (however radically they might be treated), although not entirely futile, were successful only infrequently, as measured by long-term disease-free or overall survival. As techniques of chemotherapy and radiation therapy expanded in the late 1970s, instead of relegating these women to palliative treatments alone, so-called multimodal therapy combining surgery, chemotherapy, and radiation evolved to offer this group of patients a greater degree of hope than ever before. Presurgical chemotherapy was employed, with the expectation that large, bulky tumors, often accompanied by matted involved axillary lymph nodes, might regress enough to permit mastectomy with a reasonable degree of long-term survival and a low incidence of local recurrence. Inflammatory carcinoma was added to the list of indications for such treatment, with anecdotal success being recorded even in this group of patients. In this past generation, the use of “chemo first” in such situations has become more prevalent, so much so that abundant controversy also has evolved concerning everything from the correct nomenclature for this treatment to the consideration of its use in a majority of women with breast carcinoma, in much earlier stages of disease than its initial proponents would have foreseen.

Neoadjuvant chemotherapy represents the use of cytotoxic therapy as the first modality of treatment for a primary malignant tumor. This strategy also has been referred to as preoperative chemotherapy, induction chemotherapy, initial chemotherapy, upfront chemotherapy, or primary chemotherapy. Although to our knowledge there is no consensus regarding the optimal nomenclature, there is increasing interest in the application of this strategy in the management of common human solid tumors, among them breast carcinoma.

Historical Perspective

The two major conceptual changes in the management of breast carcinoma over the past 50 years that have done the most to improve survival are 1) the systemic use of screening mammography to detect breast carcinoma in earlier, more curable stages (AJCC Stages I-II)1, 2 and 2) the use of adjuvant systemic therapies that reduce the odds of recurrence and death when administered after optimal locoregional treatments.3–5 Adjuvant chemotherapy developed in earnest in the early part of the 1970s.6, 7 In parallel with postoperative adjuvant chemotherapy, exploration of neoadjuvant chemotherapy also was initiated in the same decade, although this research focused on a different group of patients, those with locally advanced breast carcinoma.8–10 This entity has been defined in different ways, but in general, it included patients with large primary tumors, often with involvement of the overlying breast skin or fixation to the underlying chest wall and/or the presence of matted or fixed ipsilateral axillary lymph nodes or subclavicular or supraclavicular lymphadenopathy. The initial attempts with neoadjuvant chemotherapy were limited to patients with advanced, inoperable breast carcinoma, as defined by Haagensen and Stout.11, 12 Although some of these tumors were technically resectable, Haagensen and Stout documented the futility of surgical removal in view of the overwhelming risk of systemic metastases despite local elimination of the tumor.11, 12 Before the introduction of neoadjuvant chemotherapy, these patients were treated almost exclusively with palliative intent. The initial experience with neoadjuvant chemotherapy in this group of patients coincided with the development of anthracycline-based combination chemotherapy, clearly the most effective systemic intervention available before the development of the taxanes.13 These regimens produced very high rates of objective response, reducing by at least 50% the extent of primary tumor and regional lymph node metastases. This reduction in tumor volume converted many technically inoperable tumors to candidates for mastectomy and paved the way for the extension of this strategy to earlier stages of breast carcinoma and the use of breast conservation.14 This extension occurred in stages, first with the inclusion of patients with operable Stage III breast carcinoma, and subsequently with the development of clinical trials that included patients with all stages of operable breast carcinoma.15–18 The development of neoadjuvant chemotherapy also coincided with technical improvements in radiation therapy and increasing interest in less extensive surgical procedures. It was during the 1970s that the comparisons of radical mastectomy with extended radical mastectomy failed to demonstrate an advantage for the latter procedure, and the initial demonstration of the efficacy of local excision, axillary lymph node dissection, and radiation as alternatives to mastectomy was reported.19–23 These changes in surgical emphasis were facilitated by the development of systemic therapies, especially neoadjuvant chemotherapy. Multimodal therapy combining surgery, chemotherapy, and radiation evolved to offer this group of patients a greater degree of hope than ever before. Inflammatory carcinoma was, until the introduction of neoadjuvant chemotherapy, reported to be the most lethal form of breast carcinoma.14 Anecdotal success with neoadjuvant chemotherapy was reported in this group of patients in the mid-1970s.8, 24 Although combined modality therapy based on neoadjuvant chemotherapy was widely adopted during the past 20 years for patients with locally advanced breast carcinoma and inflammatory breast carcinoma, its indications for patients with earlier, operable breast carcinoma remain a matter of controversy.


Recognizing this burgeoning interest in neoadjuvant chemotherapy, this consensus conference was convened in Philadelphia in April 2003 and sponsored by the Breast Health Institute and the Fashion Group International–Philadelphia to review the evidence in support of this form of treatment, discuss the issues related to its routine application, and establish more formal guidelines for its use in patients with all stages of breast carcinoma. The participants, representing the disciplines of surgical oncology, surgical pathology, breast imaging (radiology), radiation oncology, and medical oncology, were renowned breast carcinoma experts as well as experienced specialists in this particular treatment strategy. Consensus conference participants were selected on the basis of their contributions to the literature regarding breast carcinoma in general or publications related to the subject of neoadjuvant chemotherapy.

An agenda, including a list of important questions related to neoadjuvant chemotherapy, was circulated to panelists prior to the conference, along with a list of references that included the most notable and influential publications in this area over the past 30 years. The conference syllabus included hard copies of all publications listed. Because the panelists were familiar with the basic background information concerning the rationale, history, and existing experience with neoadjuvant chemotherapy, individual presentations were brief and directly related to a specific issue under discussion at the time. Some presentations were based on recent clinical trials, whereas others were based on personal or institutional experience. As a colloquium rather than a series of formal presentations, vigorous dialogue among the panelists was encouraged as each topic was discussed at length, with combined opinions ranging from unanimous to those areas that defied consensus. The panelists attempted to separate their observations and conclusions into two basic categories, namely, those that were evidence-based and have passed the test of clinical trials and should be incorporated into current practice, and those that must still be considered investigational, requiring additional evaluation and validation. Much of the initial work with neoadjuvant chemotherapy was based on single-arm, Phase II studies or reviews of single-institution experience.14 There were few randomized trials on which to base decisions, and several randomized trials had design flaws. Therefore, the results of the clinical trials already completed were not universally accepted as defining standards of care, so that the opinions of the panelists were based as much on empiric experience as on evidence-based medicine.


The group recognized that the term “neoadjuvant” was not etymologically or technically correct and that other terms such as “primary,” “ preoperative,” “ initial,” “induction, ” or “upfront” chemotherapy were perhaps more accurate descriptions of the use of chemotherapy in specific contexts. However, the more commonly used term was “neoadjuvant ” and it was decided to retain this term for this discussion. For the remainder of these proceedings, the term “neoadjuvant chemotherapy, ” therefore, defines the use of cytotoxic chemotherapy, and “neoadjuvant hormonal therapy, ” that of endocrine therapy after the diagnosis of breast carcinoma and prior to any other treatment, either radiation therapy or surgery. For purposes of this article, we will use the term “neoadjuvant chemotherapy” (NACT) to refer to cytotoxic chemotherapy. Because hormonal therapy also has been evaluated as initial treatment for nonmetastatic breast carcinoma, we will refer to that approach as “neoadjuvant hormonal therapy” (NAHT). NACT and NAHT are integral components of combined modality treatment; this strategy utilizes both locoregional treatments to achieve optimal local control and systemic interventions to reduce the risk of systemic recurrences or metastases.

Potential Advantages and Disadvantages of NACT and NAHT

NACT might have at least three potential advantages over classic, postoperative adjuvant systemic therapy. First, NACT might either decrease local recurrences using the originally planned surgery and radiation therapy and/or it might permit less disfiguring surgery and/or radiation therapy than originally planned. Reduction in the volume of the primary tumor facilitates surgical intervention, reduces the extent of necessary surgical resection in many patients, and extends the indications of breast-conserving therapy to a larger group of patients.17, 25–29 Second, NACT might reduce distant metastasis compared with classic adjuvant systemic therapy. Finally, NACT might assist in the selection of “individualized”“ therapy through the early identification of treatment failures.

The most apparent advantage of NACT is the rapid reduction in tumor volume observed at the site of the primary tumor and enlarged regional lymphadenopathy. The observation that there is a close correlation between pathologic response of the primary tumor and long-term survival would suggest that the effect of NACT on the primary tumor would parallel the effect of NACT on distant micrometastases.14, 30–33 However, it is important to note that final confirmation of this correlation still is under evaluation in ongoing randomized trials.

Another important advantage of NACT is the early introduction of systemic therapy in combined modality treatment.34, 35 The life-threatening nature of breast carcinoma is the consequence of its ability to invade, metastasize, and cause vital organ dysfunction. The early introduction of systemic therapy was hypothesized to provide an advantage over later, postoperative initiation of systemic therapy. To our knowledge, results from clinical trials have not yet confirmed a survival advantage for the early, preoperative introduction of systemic therapy. Preclinical experiments on animal models suggested that, after resection of the primary tumor, a substantial increase in the growth rate of residual micrometastases occurs.36–38 This growth spurt can be completely abrogated by preoperative (or preradiation therapy) treatment with chemotherapy or hormonal therapy. Whether this observation is relevant to the clinical situation remains to be demonstrated.

Another potential advantage of NACT is the ability of the treating physician to determine the tumor's sensitivity to the systemic therapy being administered. In the presence of progressive growth of the primary tumor, it stands to reason that the systemic therapy being administered is ineffective; the administration of this group of drugs can be interrupted to avoid producing unnecessary toxicity and, if possible, to permit the introduction of a different noncross-resistant regimen to enhance the probability of therapeutic success. Early results of ongoing clinical trials suggest that both responders and nonresponders to the initial NACT will benefit from crossover to a second, noncross-resistant regimen,39, 40 although confirmation through long-term outcome (disease-free survival or overall survival) is awaited. In contrast, postoperative adjuvant chemotherapy is a “blind” procedure; because the only markers of tumor activity (the primary tumor and regional lymph node metastases) have been removed, the treating physician has no proven method with which to monitor the efficacy of adjuvant systemic therapy.

A further potential advantage of NACT that is of great importance to future clinical research in breast carcinoma is the potential for establishing short-term surrogate markers (clinical, pathologic, or molecular) during the administration of NACT to predict for long-term outcome in adjuvant therapy trials. This would greatly improve the rate at which new therapies could be assessed in patients with early breast carcinoma. Finally, administering systemic therapy while the primary tumor still is in place provides an excellent experimental model with which to assess the molecular effects of treatment over the tumor. Serial biopsies are possible before definitive local therapy eradicates the primary tumor, and these biopsies can be assayed for biologic changes temporally associated with NACT.41–44

NACT also might have some potential disadvantages. First, for patients with tumors resistant to NACT, appropriate local treatment is delayed for several weeks or months, potentially compromising local control. Outright resistance to NACT is rare, and progressive growth of the tumor during NACT is reported to occur in < 5% of patients so treated.14, 33, 45 Second, NACT might increase the rate or severity of surgical or radiation therapy-related complications. Extensive experience with NACT over the past 30 years has negated this hypothesis. Third, because NACT profoundly alters the primary tumor and regional lymph node metastases, it alters our ability to obtain an accurate baseline histopathologic evaluation.46, 47 It also has been stated that NACT alters the prognostic value of axillary lymph node involvement. Histopathologic evaluation of the tumor and lymph nodes may significantly change after NACT. However, the presence and number of positive lymph nodes after NACT provide important prognostic information.18, 48 Because the pretreatment clinical assessment of breast carcinoma is somewhat inaccurate, there is concern that NACT might lead to overtreatment of small primary tumors or undertreatment of others after a marked reduction in the extent of primary and regional disease.29 This concern is real; it is imperative to establish before NACT that invasive breast carcinoma is present, and that the risk of recurrence warrants the introduction of chemotherapy into the combined modality treatment of a patient with a small primary tumor. It also is essential to determine in future clinical trials the criteria with which to select patients for postmastectomy radiation therapy after NACT and NAHT.

Indications for Neoadjuvant Chemotherapy

Locally advanced breast carcinoma (Stage III)

The initial goal of NACT was to convert “inoperable” breast tumors (those with classic locally advanced disease, yet without overt evidence of systemic metastasis beyond breast and regional lymph nodes) into “operable” tumors. The term “inoperable” initially included patients with technically unresectable advanced tumors, as well as those with characteristics identified earlier by Haagensen and Stout as indicating an extremely high risk of metastases and death despite an initial surgical resection (grave signs).11, 12 These characteristics include T4 tumors by virtue of fixation to the chest wall, skin edema, ulceration, or the presence of satellite nodules or matted or fixed axillary lymph nodes, parasternal or supraclavicular lymph node involvement, or ipsilateral arm edema. Included also within this category were patients with inflammatory carcinoma.49 The intent of treatment was to reduce tumor extent in the breast and the regional lymph nodes into an operable entity. The surgical procedure usually performed in the early days of NACT was the classic radical mastectomy, often requiring skin grafts to cover the chest wall defect.50 The boundaries of the skin sacrificed over the residual tumor were measured based on the original size of the tumor, not the changes produced by the NACT.

As experience was gained with NACT, modified radical mastectomy replaced radical mastectomy, and the skin over the tumor that was sacrificed was measured by the size of the residual rather than the size of the original tumor. Starting in the 1980s, as breast conservation became accepted for early-stage breast carcinoma, the indications for NACT were extended to convert patients with large but still “operable” tumors into candidates for breast-conserving surgery.17 Because size is only one of several tumor or patient characteristics that determine eligibility for breast-conserving therapy, reduction in tumor volume with NACT does not convert all patients with large tumors into candidates for breast conservation.51 Patients with widespread areas of malignant calcifications on mammograms, those with multicentric primary tumors, and those with collagen vascular syndromes should be treated by mastectomy.

As experience with NACT increased, major clinical responses were documented in the majority of patients with locally advanced breast carcinoma, with 3–10% of patients achieving a pathologic complete response (pCR), with no evidence of invasive carcinoma remaining in the breast and/or lymph nodes at the time of surgery.14, 30, 31, 52–57 Analysis of long-term outcome (disease-free survival and overall survival) by clinical and pathologic response to NACT demonstrated that complete response, especially pCR, was associated with significantly higher disease-free survival and overall survival rates compared with lesser or no responses to therapy.30, 32, 33 Residual foci of ductal carcinoma in situ (DCIS) present at the tumor site did not appear to alter the prognostic value of pCR. Some groups defined pCR as no residual invasive tumor in the breast and the regional lymph nodes30, 32; others interpreted pCR as no residual invasive tumor in the breast33. The former is a more stringent definition and defines the overall response to NACT. It is hypothesized that response to NACT within the primary tumor and regional lymph nodes is a marker of response in distant, occult micrometastases. The correlation of pCR with long-term survival would be based on this hypothesis. This would indicate that the degree of pathologic response could be used as an early surrogate marker of the efficacy of NACT. Further validation of this concept would result in the more efficient and faster assessment of new systemic therapies and expedited integration of new drugs and biologic agents into the treatment of primary breast carcinoma.

Whether improvement in pathologic tumor response by the introduction of a more effective NACT regimen would result in a commensurate improvement in both disease-free and overall survival remains to be demonstrated and currently is under evaluation.40 Some authors have used the term pCR when there has been “minimal” residual tumor in either breast or lymph nodes. To our knowledge, this definition of pCR has not been validated and its prognostic significance remains unproven. Residual DCIS is found in 2–3% of patients treated with NACT. To our knowledge, there is no evidence to date that this finding confers a higher risk of distant recurrence or different long-term outcome to patients compared with the total absence of invasive and noninvasive tumor. However, it is unknown whether residual DCIS after NACT and lumpectomy influences local control rates. Residual DCIS alone is an uncommon finding, and to our knowledge none of the studies reported to date have the statistical power to rule out even moderate differences between these two groups of patients. On the basis of existing information, the panel accepted the absence of invasive disease in the breast and axilla as the definition of a pCR, ignoring the presence of DCIS.

The panel agreed with the concept that combined modality therapy based on NACT represents the standard of care for patients with locally advanced and inflammatory breast carcinoma. In this patient population, NACT contributes to improved operability and enhanced local control and makes breast-conserving surgery possible for some patients.

Operable breast carcinoma (Stages I and II)

In patients with so-called operable breast carcinoma, multiple therapeutic strategies can be used. Initial surgery followed by adjuvant systemic therapy and/or radiation therapy represents the most commonly used strategy used around the world for patients with Stage I and Stage II breast carcinoma.58 However, over the past 10–15 years, several Phase II trials and a few randomized trials have demonstrated the therapeutic equivalence of an alternate strategy based on NACT followed by surgical resection and radiation therapy, if appropriate.45, 59 It was noted in these studies that the ability to offer breast-conserving surgery increased in those patients who received NACT, although not all patients (not even all responders to NACT) became candidates for breast-conserving therapy. This is due, in part, to the insufficient reduction in tumor size after NACT observed in some patients, the presence of additional factors that preclude the use of breast-conserving surgery, or patient preference.

Some members of the panel advocated NACT for all patients with invasive breast carcinoma who would, by virtue of pretreatment evaluation (e.g., by tumor size or regional lymph node findings) be candidates for (postoperative) adjuvant chemotherapy. This might include many patients with Stage II (T1N1, T2N0, T2 N1, as well as T3, N0 disease) and some with Stage I (T1c, N0) tumors. This indication for NACT engendered significant debate because of current differences in opinion regarding which patients with Stage I (T1c) or small Stage II breast tumors (measuring 2.1–3.0 cm) will require adjuvant cytotoxic chemotherapy and if so, what type of chemotherapy. Some members of the panel were especially concerned about this selection criterion because adjuvant hormone therapy alone without cytotoxic chemotherapy is commonly employed for earlier stages of breast carcinoma.60 There was greater consensus with regard to the use of NACT for patients with Stage IIB or Stage IIIA breast carcinoma. All members of the panel agreed that using NACT in all situations in which cytotoxic chemotherapy would almost certainly be used postoperatively is promising but should be studied further by clinical trials. The panel summarized the potential indications of NACT for patients with operable breast carcinoma as follows.

  • 1Reduction of tumor extent for patients with large primary breast tumors to enhance the possibility of breast-conserving surgery. There is ample experience for this indication from clinical trials. However, to our knowledge there was concern regarding indications in some trials that breast-conserving surgery after NACT might be associated with a higher local recurrence rate.45, 59, 61
  • 2To determine response or resistance to NACT based on the serial evaluation of measurements of the primary tumor and use this information to influence continuing therapy. Although the majority of the panel agreed that this was an established indication, some panelists believed that there was insufficient evidence regarding the correlation between response to NACT in the primary tumor and the response of possible distant micrometastases. This correlation begs for a well designed clinical trial.
  • 3To determine response to therapy, especially pCR, as an early surrogate marker of long-term prognosis. It was agreed that this indication was still under investigation, and additional validation was necessary.
  • 4To test the activity of new drugs and expedite their integration into the therapy of primary breast carcinoma. This indication with potentially exciting ramifications for drug development was considered to be experimental, requiring additional study and validation.

Menopausal status does not appear to affect the indications for NACT. The treatment is equally well suited to both premenopausal and postmenopausal patients.

An additional relative indication for NACT is in the treatment of pregnant patients during the second or third trimester. Hormonal therapy is contraindicated during pregnancy; therefore, if the indications for systemic therapy exist, NACT would be the treatment of choice. There is some modest experience, based on retrospective analysis and at least one prospective clinical trial, that NACT can be employed through delivery, with a high degree of effectiveness and acceptable safety for the patient and the child.62 Tumors arising during pregnancy occur in younger women and tend to be more aggressive, so that adjuvant chemotherapy would almost always be used. Furthermore, tumors diagnosed during pregnancy are often larger or more advanced; therefore, combined modality therapy would be appropriate. Chemotherapy may be used during the latter part of the second trimester and in the third trimester without apparent deleterious effects on the fetus.

The randomized clinical trials that have been completed to date have shown at least equivalent survival for patients treated with NACT followed by locoregional therapy compared with groups treated with locoregional therapy followed by adjuvant therapy. Patients who have demonstrated a pCR (i.e., no residual invasive carcinoma in either the breast or axillary lymph nodes) have been shown to have improved disease-free survival and overall survival compared with patients with lesser or no response to NACT.17, 18, 30, 32, 33, 59 These observations have been reproducible in both randomized and nonrandomized trials.

Advanced age is not considered a contraindication to NACT. The presence of significant intercurrent disease would influence decisions regarding NACT in the same manner as it would influence decisions concerning postoperative adjuvant chemotherapy. If the comorbid condition limits life expectancy or substantially increases the risks associated with chemotherapy, then it might represent a contraindication to NACT. The panelists strongly believed that the same criteria should be used for NACT as for postoperative adjuvant chemotherapy, taking intercurrent disease into consideration in the same way. For those patients for whom NACT is contraindicated for medical reasons, hormonal therapy with tamoxifen or an aromatase inhibitor (AI) would be appropriate treatment if their tumors express positive hormone receptors.


Hormone therapy has been used successfully in the neoadjuvant setting.63–70 Most of the investigators have used tamoxifen, although over the past 5 years several clinical trials have been designed using a selective AI. NAHT with tamoxifen reduces overall tumor volume in approximately half of breast carcinoma cases. Responses tend to occur over several months and, for patients with receptor-positive breast carcinoma, treatment for 3–6 months in the absence of progressive disease is a reasonable approach. To our knowledge, the frequency of pCR after NAHT has not been well defined, although it appears to be low, around 5% after tamoxifen.71 Limited experience from clinical trials suggests that selective AIs are more effective than tamoxifen in the neoadjuvant setting, although to our knowledge the pCR rate after AI therapy is uncertain.68, 72 The correlation between response to NAHT and long-term prognosis has not been established in prospective clinical trials, nor the degree to which breast-conserving surgery is improved after NAHT. Several clinical trials have demonstrated that NAHT with tamoxifen followed by surgical resection of the primary breast tumor results in superior local control compared with tamoxifen alone.73–76 Patients with hormone receptor-positive locally advanced breast carcinoma and most patients with larger, operable breast tumors are best treated with both chemotherapy and hormonal therapy in the context of combined modality treatment. Additional clinical trials are needed to determine the optimal sequence of treatments for patients with early and locally advanced hormone receptor-positive breast carcinoma.

Diagnosis and Pretreatment Evaluation

Patients with locally advanced breast carcinoma have a greater risk of existing metastases at the time of diagnosis; therefore, their initial evaluation demands a greater degree of pretreatment scrutiny to rule out the presence of metastatic disease than usually advocated for patients with earlier tumors. The diagnosis of invasive (ductal or lobular) carcinoma must be established before NACT or NAHT is considered. Although areas of DCIS can sometimes reach a large size, DCIS is not treated with NACT or NAHT, irrespective of size. The diagnosis of invasive carcinoma should be made by core needle biopsy whenever possible, guided by palpation, ultrasonography, or stereotactic radiography. However, depending on local facilities and expertise, incisional biopsy can be used to establish the diagnosis of invasive carcinoma and to determine important biologic markers. The emphasis should be on the quality of the specimen, not on the technique required to obtain it. The microscopic report of the biopsy specimen should note the invasive character of the tumor; histologic type (ductal, lobular origin, or other); its nuclear and histologic grade and mitotic index; and assessment of estrogen receptor (ER), progesterone receptor (PR), and HER-2 content.77 The relative merits of fine-needle aspiration (FNA) biopsy alone and core needle biopsy were discussed at length by the panel. FNA cannot differentiate between invasive and noninvasive breast carcinoma.78 In the presence of palpable regional lymph nodes, the identification of malignant cells in a lymph node by cytology on an FNA specimen infers the presence of invasive breast carcinoma because, by definition, only invasive tumors can metastasize. In many centers, ER, PR, and HER-2 status can be determined by immunohistochemistry on FNA specimens. FNA, however, requires an expert cytopathologist.

The initial biopsy specimen must provide a definitive diagnosis of invasive breast carcinoma and determination of biologic markers because if NACT is successful there may be no residual tumor at the time of definitive surgical resection and these markers will never be determined. Information regarding ER and PR status also is important because it can predict response to NACT and NAHT; tumors with negative hormonal receptors respond better to NACT, whereas tumors with positive hormone receptors respond well to hormonal therapy.

The optimal pretreatment assessment of primary breast carcinoma before NACT or NAHT is no different from that performed before primary surgical therapy and has been summarized previously in other consensus documents.79 Additional immunohistochemical markers (such as Ki-67, p53, etc.) can be performed on the core needle biopsy specimen as well, although to our knowledge their current roles in determining treatment and assessing outcome are poorly understood. Because most tumors respond well to NACT, the issue of tumor localization after NACT or NAHT was discussed at length. The panel agreed with the recommendation to mark the location of the tumor using an externally visible or a radiopaque marker.80–82 In some centers, a radiopaque clip or coil is placed in or around the tumor site to note its location in the event that the response to NACT is complete or nearly so. These objects can be placed through the core needle biopsy devices after the acquisition of the biopsy material. New devices permit the placement of a permanent radiopaque marker at the site irrespective of the caliber of the core biopsy needle.83 Another method with which to accomplish the same objective is the pretreatment outlining of the tumor directly on the breast with a skin-marking pen and then photographing this tumor outline for posterity; thus, even if the tumor shrinks significantly, the initial area of involvement will be available for local treatment planning. Other panelists advocated tattooing the tumor margins on the skin of the breast, using four points to outline the tumor, in the same manner as radiation oncologists tattoo the radiation fields onto the skin. Whatever the technique employed, the group unanimously endorsed documenting the initial size and location of the tumor before embarking on NACT or NAHT to be certain the site can be identified after NACT by both the surgeon and pathologist, especially if there is a clinical CR.

The minimum workup to screen patients for metastasis will vary depending on the clinical stage of disease at diagnosis. Guidelines for initial staging workup have been promulgated by professional societies.79 For patients with Stages 0-II disease, initial evaluation includes physical examination, bilateral mammography, and chest radiograph. For patients with more advanced disease, a complete blood count, liver function studies, abdominal imaging (computed tomography or magnetic resonance imaging [MRI]) and bone scan usually are performed in addition to the physical examination, bilateral mammography, and chest radiograph. Nuclear medicine breast scans (sestamibi) were not considered valuable or indicated. The role of other imaging studies, such as positron emission tomography (PET), was considered investigational because their clinical value has yet to be defined. Several members of the panel championed ultrasound of the breast and axilla to assess the extent of disease and guide core needle or FNA biopsy of suspicious lymph nodes.84–86 However, other members of the panel pointed out that this approach required the availability of expert breast sonographers and standardized techniques, neither of which is available in some centers.

The value of breast MRI was discussed as part of the initial and subsequent evaluation of the extent of the tumor in the affected breast. MRI is more sensitive than other diagnostic techniques; however, it also is less specific, thus identifying abnormalities that may or may not be of a malignant nature but will certainly require additional interventions, even biopsy, to obtain an accurate diagnosis.87–91 Breast MRI has not been standardized and different institutions use different diagnostic protocols. Therefore, an MRI performed in one institution may not be comparable to one performed at another institution. Although some institutions report that breast MRI is a sensitive and reliable diagnostic tool in their hands, others have reported much less favorable experience and have come to opposite conclusions. In institutions with an expert breast MRI group and with positive experience in the use of this diagnostic approach, MRI is of value in monitoring tumor dimensions during NACT and providing quantitative measures of tumor response after NACT.89–91 MRI performed at time of diagnosis may identify those patients with more than one primary carcinoma in the breast as well as invasion of the underlying chest wall. Repeat MRI at the end of NACT, just prior to the time for surgical recommendations, may provide an accurate assessment of the magnitude of response to treatment, whether there are likely to be clear dissection planes behind the lesion, and whether the residual tumor is of limited extent, so that breast conservation can be considered. Members of the panel with extensive experience with breast MRI were enthusiastic about its use to help measure the extent and geography of the tumor and its response to chemotherapy. There was consensus with regard to breast MRI being an exciting diagnostic prospect and regarding the need to perform additional and carefully designed research studies to standardize the procedure and its interpretation and enhance its diagnostic accuracy, especially by reducing the rate of false-positive findings.

The Role of Axillary Lymph Node Staging and Lymphatic Mapping in Patients Treated with NACT or NAHT

Although there initially were theoretic concerns that NACT might vitiate the prognostic value of axillary lymph node involvement, data from clinical trials suggest that the presence and extent of axillary lymph node involvement after NACT have even greater prognostic value. Patients with negative lymph nodes after NACT fare better than those with positive lymph nodes, and the risk of recurrence and death increase in proportion to the number of positive lymph nodes.31, 46, 47, 92 To our knowledge, it is not known whether patients who demonstrate a response to NACT by virtue of having fewer or negative lymph nodes after NACT should receive less local or systemic therapy than they would have received if their therapy had been sequenced conventionally.

In recent years there has been much enthusiasm for the integration of sentinel lymph node mapping and biopsy into the management of patients with primary, clinically lymph node-negative breast carcinoma. Although this procedure still is undergoing evaluation in formal clinical trials, its use outside a clinical trial has expanded dramatically.

All the surgical members of the panel already have adopted axillary lymphatic mapping and biopsy of the sentinel axillary lymph node(s) (SLNB) as a replacement for traditional Level I and II axillary lymph node dissection in patients with Stages I and II carcinoma and clinically negative axillary lymph nodes. Only when the sentinel lymph node(s) are positive for metastasis or if they cannot be identified do they perform axillary lymph node dissection. Therefore, SLNB in patients about to undergo NACT or performed after NACT became a topic of significant debate. To our knowledge, there is no indication for this procedure in patients with clinically positive lymph nodes, so this controversy is largely irrelevant for the majority of patients with locally advanced breast carcinoma. As NACT becomes used more frequently in patients with smaller tumors (e.g., T2 tumors [measuring ≥ 3 cm]) some panelists advocated assessing axillary lymph node status microscopically by ultrasound-guided FNA or core needle biopsy prior to the initiation of NACT. In patients with small tumors that are likely to have negative lymph nodes, some panelists indicated that they would use the results of SLNB before NACT to determine what type of chemotherapy to use in the management of these patients. Some would reserve the use of taxanes and dose-dense regimens for patients with lymph node-positive breast carcinoma. Another reason advocated to perform SLNB before NACT was the strong impact of the number of positive lymph nodes on the decision to administer postoperative radiation therapy. Although there is agreement regarding the need for postmastectomy radiation therapy for patients with four or more positive lymph nodes, to our knowledge there is no such consensus regarding radiation therapy for patients with fewer than four lymph nodes. For example, if a patient has two positive lymph nodes after NACT, should that patient forego radiation therapy or should it be assumed that she had multiple positive lymph nodes before NACT that were sterilized by NACT? In this latter case, radiation therapy would be strongly considered. In addition, the number of positive lymph nodes present might change the radiation therapy field design (e.g., the use and/or extent of regional lymph node radiation therapy). Although these considerations had less relevance to patients with locally advanced breast carcinoma or large (> 5 cm) operable tumors, they elicited vigorous discussion that precluded consensus regarding the groups of patients with smaller or clinically lymph node-negative breast carcinoma. A third important consideration concerned the feasibility, accuracy, and reliability of SLNB after NACT. Some of the earliest reports of SLNB after NACT suggested that the false-negative rate of SLNB was higher in patients who had received NACT than in those who did not.93–98 More recent, larger series documented a learning curve and demonstrated that, with increasing experience, the rate of false-negative SLNB results decreased to levels similar to those reported in early breast carcinoma and in the absence of NACT.96, 99, 100 Several panelists with experience using SLNB after NACT believed that sentinel lymph nodes could be identified after NACT as well as if NACT had not been used (> 95%); these panelists indicated that the information obtained from examination of the sentinel lymph node(s) in these patients was equally accurate in predicting axillary lymph node status. Other panelists indicated that false-negative rates after NACT appear to be slightly higher than in those patients who have not undergone NACT in some, but not all, of the small studies reported. The panel recommended the same criteria for the use of SLNB after NACT as if it were being considered as part of the initial treatment (i.e., a clinically negative axilla (N0)). It was considered that if the axilla had been suspicious before NACT (N1 or N2) and “downstaged” to N0 by the chemotherapy, SLNB was an available option when used by surgeons with extensive experience in SLNB. However, additional evaluation of SLNB in this setting clearly is warranted. Experience with SLNB in patients whose axillary lymph nodes remain clinically suspicious after NACT was even more limited and therefore SLNB is not recommended in these circumstances.

The clinical assessment of the axilla is more accurate when the lymph nodes are considered positive for metastasis, but as many as 35% of patients with clinical N0 axillae will be found to be positive when the lymph nodes are examined microscopically. Several members of the group endorsed ultrasound examination of the axilla and FNA biopsy of any enlarged lymph nodes to confirm the presence of metastasis microscopically. Other panelists believed this procedure was unnecessary and had the same limitations discussed previously in relation to sonography and FNA. Histologic proof of axillary lymph node metastasis is not a requirement for NACT for patients with larger tumors because chemotherapy would be recommended on the basis of the size of the primary lesion. Patients with a T1a or T1b carcinoma with suspicious axillary lymph nodes would have different recommendations regarding adjuvant chemotherapy depending on histologic demonstration of axillary lymph node involvement; some panelists would prefer to treat patients with such small tumors with surgery first whereas others considered that NACT would be appropriate if axillary lymph node involvement could be documented before NACT.101 In these cases, image-guided FNA was a diagnostic option to consider.

There was extensive discussion regarding the indication and timing of SLNB in patients with initial clinical N1 or N2 axilla who were considered to be N0 after the completion of NACT. Another group discussed was the patients with larger tumors whose axilla was clinically N0 before NACT was administered (e.g., T3, N0). Would the opportunity to know pretreatment lymph node status precisely be affected and if so, was this significant? Part of the panel considered that the pre-NACT status of the axilla provided critical prognostic information that would be altered after NACT, and therefore axillary lymph node staging should be performed before NACT. Other panelists considered that the prognostic value of axillary lymph node staging after NACT was equivalent or superior to that obtained prior to NACT because it already included the contribution of chemotherapy and reflected the presence or absence of resistance to treatment. Both arguments carried some weight. In the absence of definitive information, it is recommended that these issues be carefully assessed in ongoing and future trials that include NACT.

For this reason, some panelists promoted the performance of SLNB as part of the pretreatment (pre-NACT) plan in patients with N0 or NX axillary lymph node status undergoing NACT. In this scenario, the results of SLNB performed before NACT was initiated would not affect NACT, but if the lymph node(s) were negative, axillary lymph node dissection after NACT would not be performed whereas if the sentinel lymph nodes were positive before NACT was initiated, traditional axillary lymph node dissection would be part of the post-NACT plan, regardless of whether breast conservation or mastectomy was recommended subsequently. Other panelists indicated that this strategy would lead to full axillary lymph node dissections after SLNB in all patients with initially positive lymph nodes. Because NACT may “sterilize” axillary lymph node metastases in approximately 20% of patients, performing the SLNB after NACT (as part of the definitive treatment after NACT) would avoid axillary lymph node dissection as part of subsequent treatment in that group of patients in whom lymph nodes were sterilized by NACT. These approaches to the timing of SLNB need to be examined in clinical trials.

Optimal Neoadjuvant Chemotherapy Regimens

Because NACT was introduced in the management of patients with locally advanced breast carcinoma, a high-risk group that requires the most effective available therapy, most of the experience with NACT is based on anthracycline-containing regimens. To our knowledge, there are few randomized trials in the area of NACT, and only some of the most recent trials asked chemotherapy-related questions. The panel discussed questions related to chemotherapy: which drugs to use, in what sequence or combination, and for how long? The panelists were unanimous in their choice of established, validated combination chemotherapy regimens that contained an anthracycline (doxorubicin or epirubicin) and cyclophosphamide as the initial therapy of choice. Although the doxorubicin and cyclophosphamide (AC) regimen is commonly employed, some of the panelists considered that 5-fluorouracil (5-FU) should be added to this regimen (5-FU, doxorubicin, and cyclophosphamide [FAC] or 5-FU, epirubicin, and cyclophosphamide [FEC]) because these three-drug combinations have performed best when compared with cyclophosphamide, methotrexate, and 5-FU (CMF)-type regimens in randomized clinical trials.102 The recommended doses and schedules were the same as those used in the adjuvant settings.

Since the introduction of taxanes, it has been established that these agents add substantial efficacy to treatment. The two most frequent methods to incorporate taxanes into NACT are: 1) As a cross-over regimen, after (or preceding) four cycles of an established anthracycline-containing regimen.103–105 In this method, four cycles of paclitaxel or docetaxel follow three to four cycles of AC, FAC, epirubicin and cyclophosphamide (EC), or FEC, or the same drugs are given by weekly administration for an equivalent duration of time. 2) As part of established combinations, substituting for an existing older drug (e.g., docetaxel, doxorubicin, and cyclophosphamide [TAC] for FAC).106 In this case, the combination would be administered for four to six cycles. It is uncertain whether the sequential or simultaneous administration of anthracyclines and taxanes is more effective. This issue currently is the subject of ongoing clinical trials.

There are universally accepted criteria for response to systemic therapy. The World Health Organization (WHO)/International Union Against Cancer (UICC) criteria have been used for the past 20 years. A clinical CR is defined as the complete disappearance of all clinically detectable disease and a partial response requires a > 50% reduction in the sum of the products of the 2 longest perpendicular dimensions of measurable tumor deposits, whereas progressive disease is defined as a 25% increase in the sum of the products of the 2 longest perpendicular dimensions of measurable tumor deposits or the appearance of new lesions.107 More recently, the Response Evaluation Criteria in Solid Tumor (RECIST) criteria were adopted in the evaluation of systemic anticancer treatments.108, 109 According to these criteria, the definition of a CR remains unchanged, but a partial response is now defined as a 30% decrease in the longest dimension of each measurable tumor deposit. These same criteria should be followed for the determination of the response to NACT or NAHT. In the absence of progressive disease, at least two but preferably four cycles of NACT should be given before patients are termed “nonresponders.” Similarly, in the absence of progressive disease, NAHT should be continued for at least 3 months but preferably 6 months to provide an adequate therapeutic trial. Limited information from clinical trials suggests that the pace of response varies between patients, and that to capture the maximal response to NACT for every patient, four to six cycles of therapy need to be administered. Most NACT programs recommend four cycles of a single regimen. Until the appearance of the taxanes, surgery was instituted after three to four cycles of NACT, often after even more cycles of the same treatment, using a “plateau” of response to determine the precise number of cycles. Treatment was continued until no further change in the tumor was clinically apparent from one cycle to the next (or the tumor had responded completely). Recently reported clinical trials suggest that both responders and nonresponders to an anthracycline-based chemotherapy regimen benefit further from crossover to an alternate, noncross-resistant therapy, most frequently a taxane.39, 40, 110, 111 As many as 90% of women who are administered NACT will manifest a clinical response as defined earlier, either to the first courses of chemotherapy or to a change in the agents used if no significant response is noted after the first three cycles.

Patients with large or locally advanced breast carcinoma who do not demonstrate a response to NACT tended to develop recurrent or metastatic disease rapidly, so that failure to respond to NACT is considered a grave prognostic sign.112 Unless all macroscopic evidence of tumor in the breast (including the skin of the breast) and in axillary lymph nodes can be resected, there is little benefit to surgery. Rather than commit these “nonresponders”“ to what would almost certainly prove to be a palliative mastectomy, with a high likelihood of local recurrence, preoperative or high-dose “definitive” radiation therapy without removing the primary tumor would be considered suitable therapy, as would have been used for inoperable carcinoma before the advent of NACT. For patients with hormone receptor-positive breast carcinoma, adjuvant hormonal therapy also should be instituted at this point. Disease progression while the patient is undergoing NACT occurs rarely, and is reported in 3–7% of patients with locally advanced breast carcinoma and an even smaller proportion of patients with earlier, operable breast carcinoma.14, 33, 45 However, when disease progression does occur, it is an especially ominous sign. Switching to a noncross-resistant regimen of NACT constitutes a reasonable alternative if the disease is inoperable at that point. Definitive radiation therapy should be employed for patients whose tumor remains unresectable after NACT.113

For patients with early-stage operable breast carcinoma who do not respond to NACT or who progress during NACT, surgical removal usually is the treatment of choice; on the basis of the full pathologic report, subsequent therapy might include radiation therapy and hormonal treatment. If progression of disease made the primary tumor unresectable, preoperative radiation therapy or exclusive radiation therapy might be employed. The radiation approach in nonresponders should be a course of 45–50 grays (Gy) over 4.5–5 weeks to the breast and peripheral lymphatics, with some radiation oncologists also favoring an additional boost of 10 Gy in 1 week to the site of the macroscopic tumor.112, 113 This technique can convert some inoperable tumors into candidates for mastectomy. If the tumor remains unresectable, additional radiation using brachytherapy or shrinking fields to bring the total tumor dose to 70–90 Gy may be indicated to achieve whatever disease regression might be possible. Some investigators have reported better results with hyperfractionated courses of radiation therapy, particularly in patients with chemotherapy-resistant disease or inflammatory breast carcinoma.114 To our knowledge, there are no comparative trials with which to establish the relative value of these techniques. Which regional lymph node-bearing areas need to be irradiated remains uncertain. Complete axillary fields, supraclavicular fields, and internal mammary fields normally would be included, along with the areas of macroscopic disease, but there is a lack of agreement regarding this in the radiation oncology community. There will be a small group of these patients who might achieve long-term survival using this approach, but it generally is accepted that the overall outlook for most of these women is grim, and the cosmetic appearance of the breast after this intensive treatment is generally poor. For women with smaller operable breast tumors, the fields or radiation therapy will be individualized based on the size and extent of the primary tumor as well as the number of axillary lymph nodes involved.

How long should NACT continue in women who respond? To our knowledge, the duration of NACT has never been tested in randomized trials. Earlier clinical experience in patients with locally advanced breast carcinoma suggested that the median number of cycles of an anthracycline-containing regimen to achieve the best response was 5 (lasting approximately 5 months) and was somewhat longer to achieve a CR. However, the panel considered that NACT should be used in the same manner as adjuvant chemotherapy (i.e., four cycles of the initial treatment [e.g., AC] with assessment of response at the end of that period). If there has been a complete or nearly complete clinical response as of the end of the fourth cycle, definitive treatment now would be appropriate. If the response has been less than that, the panel favored an additional four cycles of a taxane (i.e., a noncross-resistant drug or drugs, or vice versa), if a taxane had been used first. Recent clinical trial data suggested that the addition of a taxane after AC or FAC-type regimens improved outcome in both responders and nonresponders to initial NACT.39, 40, 115 Currently, all eight cycles of NACT generally are administered before definitive locoregional treatment in ongoing clinical trials. The addition of four cycles of a taxane has been demonstrated to increase the pCR rate significantly, even doubling it in some clinical trials. Whether this increase in pCR will translate to improvement in survival is currently under evaluation in clinical trials.

For patients treated with NAHT, a response to hormonal therapy is achieved more slowly. In the absence of progressive disease, response should be assessed at the end of 3 months, and if there is evidence of stability or a modest response, NAHT should be continued for a total of at least 6 months. The timing of definitive locoregional therapy for these patients should be individualized so that when a plateau of response has been achieved (with no observed further regression in the breast and/or lymph nodes), local treatment is indicated and should be based on the same principles as for locoregional treatment after chemotherapy.116 Tamoxifen is the hormonal agent that has been employed most often and for which the most information is available. However, it is not unreasonable to expect that the AIs, if used in postmenopausal women as an alternative to tamoxifen, will be as, if not more, effective.

Choice of Locoregional Treatment after NACT or NAHT

After the completion of the planned cycles of NACT (usually four or eight cycles), the assessment of tumor response will drive the locoregional treatment recommendations. A careful physical examination with measurements to document extent of residual disease in the breast and regional lymph nodes should be accompanied by repeat imaging studies, including mammography, ultrasonography, and MRI (if it had been used before NACT). The combined clinical and imaging evaluations may indicate the extent of and site(s) of the residual tumor. Unfortunately, breast carcinoma does not always undergo concentric regression, and determining the geography of any residual tumor is essential to make the appropriate recommendation for surgery. For example, tumors characterized by areas of malignant-appearing calcifications may still have large areas of calcifications after NACT, and breast conservation for these women would not be appropriate. MRI often is helpful in determining whether the residual tumor is of limited extent and is amenable to wide excision.89, 91 The caveats expressed previously concerning the routine use of breast MRI for this purpose apply here as well.

If there are questions regarding the distribution or extent of residual tumor within the breast, multiple core needle biopsies prior to a final recommendation may be helpful, especially if the patient strongly prefers breast conservation. The same criteria for the selection of candidates for breast conservation are appropriate in patients after NACT as for patients with tumors of lesser extent for whom NACT is not used, namely, the absence of multicentric tumor; the absence of widespread, malignant-appearing calcifications; and the ability to excise the residual tumor completely with clear margins and a suitable cosmetic result.117–119 Patients with locally advanced breast carcinoma must have responded to the extent that skin involvement has regressed and chest wall fixation, if initially present, has disappeared. If a clinical CR has been achieved, resection of an area surrounding the marker placed at the beginning of NACT is recommended to ensure that no microscopic residual disease remains. If a clinical CR has been achieved, breast conservation is nearly always possible. Limited data based on retrospective analysis suggest that the treatment of patients with a clinical CR exclusively by radiation therapy without surgery, even to reexcise the primary tumor site, is associated with a higher incidence of local recurrence, although to our knowledge this event has not to date altered overall survival.120, 121 For responding patients with clinically or radiographically documented residual tumors, all residual tumors should be resected with clear margins. The surgical intervention should be based on the clinical and radiographic characteristics of the primary tumor at the time of the evaluation after NACT and not the extent and characteristics at the time of the initial diagnosis.81 The choice between breast-conserving therapy or mastectomy should weigh the various properties of the tumor, its geography within the breast, and the likely cosmetic outcome after treatment. The site of the initial lesion must be excised and radiation therapy should follow.

With respect to inflammatory carcinoma, although members of the panel reported anecdotal, limited experience with breast conservation in these patients, the opinion of the group was that, except under unusual circumstances, breast conservation is not indicated in patients who present with inflammatory carcinoma, irrespective of the apparent improvement in the clinical findings after NACT. The clinical findings in the breast, even when significant improvement is achieved, often do not mirror the residual microscopic findings in the breast (and lymph nodes).

The magnitude of the surgical procedure (i.e., volume of breast to remove) is often a challenge because there may be no macroscopic evidence of disease to guide the surgeon. Therefore, this decision must combine the initial documentation of the tumor size and location and the change documented by careful physical examination and breast imaging. Detailed description of baseline physical findings and imaging results (with measurements of the primary tumor and involved regional lymph nodes; carefully dictated surgical notes; and mammographic, sonographic, and MRI studies) are all helpful in this regard. Neglecting the surgical excision of the site because of the apparent clinical regression of the lesion is followed by a greater incidence of local recurrence than if this site is excised.122, 123 Clinical CR does not imply complete pathologic disease regression! Between 30–50% of patients with a clinical CR will have residual breast carcinoma detected by the pathologist in the surgical specimen.14, 124 Conversely, some patients with clinically suspected residual disease after NACT will have a pCR on microscopic examination of the surgical specimen.

Appropriate timing of surgery is vital. Restoration of leukocyte counts to near normal (> 2500–3000 cells/μl) is important for wound healing. The hemoglobin and hematocrit usually decrease during chemotherapy, and these should be in acceptable ranges (> 10 g and > 35%, respectively) before surgery. Most patients experience complete hematologic recovery 3–4 weeks after the last chemotherapy cycle. However, for a minority of patients, it might take as long as 4–6 weeks from the date of the last chemotherapy before the patient's blood counts and physical status in general are optimal. A longer wait before surgery theoretically could jeopardize the effects of NACT if a pCR has not been achieved, but to our knowledge this has not been reported to date. Hematologic recovery can be accelerated with hematopoietic growth factors.

Breast conservation is possible after NACT in many patients with locally advanced disease.26, 45, 81, 125, 126 The proportion of patients with large tumors who become candidates for breast-conserving therapy varies according to the initial extent of disease, response to therapy, the criteria for selection of patients in the treating center, and the definition of acceptable cosmetic results, as well as the experience of the surgical team. Thus, for patients diagnosed with Stage IIIB breast carcinoma, the breast conservation rate after NACT varies from as low as 17% to as high as 85%.16, 17, 29, 80, 81 If criteria for patient selection are carefully applied and high doses of radiation therapy are avoided, most likely as much as 40% of patients with T4 primary tumors become candidates for breast-conserving therapy. Up to 50% of patients with T3 lesions and 70–80% of those with large T2 primary tumors become candidates for breast-conserving surgery after NACT. Whether the local excision of the residual disease is performed as a separate procedure to prove that breast conservation is feasible before addressing the axilla, or local excision and axillary lymph node surgery are combined into a single attempt, are not important provided the local excision is successful and the margins are clear of both in situ and invasive carcinoma. The optimal width of a negative surgical margin in patients who undergo breast conservation after NACT is controversial. Panelists' opinions ranged from 1 mm to ≥ 10 mm from margin to nearest evidence of tumor. Much of the data regarding wider margins was extrapolated from the DCIS data, in which margins ≥ 10 mm usually are recommended. It is unclear whether data extrapolated from clinical trials focusing on DCIS are really relevant to patients with advanced breast carcinoma. Within the limits of aesthetics, there was consensus that wider margins are always better, if achievable. The balance between the optimal oncologic resection and desirable aesthetic appearance of the breast after local excision is usually the surgeon's judgment call, with the patient's input, but should tilt toward the former and not the latter objective. The surgeons on the panel all endorsed the widest possible margins, up to 10 mm. This is one relative indication to separate the local excision from the axillary lymph node procedure. If the patient is unhappy about the cosmetic results so that mastectomy becomes preferable, or the tumor is more extensive than initially considered, the axillary lymph node procedure can be performed along with the mastectomy. Also, if the margins are not clear enough (i.e., clear but close) or only focally positive, and it is likely that a local procedure alone can still successfully clear them satisfactorily, reexcision of the margins and SLNB/axillary lymph node dissection can be performed together. In summary, the criteria for breast conservation after NACT are the same as for patients treated traditionally.

Despite the success of NACT in achieving breast conservation in a majority of patients, many women still require mastectomy for the reasons previously cited. When NACT was in its infancy, prior to the adoption of breast conservation as an alternative to mastectomy in any patient, the initial surgical approach used the original measurements of tumor size and extent to plan the procedure, regardless of the observed clinical response. Thus, many women underwent radical mastectomy with the sacrifice of most of the skin over the breast. Skin grafts often were required to cover the chest wall defect. As experience was gained, the need to use the pretreatment measurements was challenged, so that now it is the measurements of size and extent as well as the radiographic image of the breast following NACT that dictate the scope of the surgical procedure. Skin-sparing mastectomy usually is not indicated for patients with initial skin involvement, but might be an appropriate choice for those with T2 or even T3 tumors after an excellent response to NACT. Full axillary lymph node dissection (Levels I, II, and III) rarely is performed today; Levels I and II usually are dissected for patients with positive lymph nodes. This statement does not imply that full lymph node dissection is not indicated. When the axilla is noted to be clinically positive intraoperatively, there is no better way of avoiding axillary lymph node recurrence than meticulous complete axillary lymph node dissection. Sacrifice of the pectoralis minor muscle depends on the clinical judgment of the oncologic team. If there is concern regarding involvement of the interpectoral lymph nodes, the pectoralis minor should be sacrificed. Because mastectomy is likely to be used most often in patients with significant residual disease, when breast conservation cannot be employed, surgical compromises should not be made. The best postoperative radiation therapy is not a substitute for meticulous surgery.

Breast Reconstruction after NCAT or NCHT

All women who require or prefer a total mastectomy after NACT are candidates for breast reconstruction. Immediate (as opposed to delayed) breast reconstruction has become a popular choice for women requiring mastectomy; it may be offered to women undergoing mastectomy after NACT, but with several qualifiers. Many women undergoing mastectomy also will require postoperative radiation therapy to the chest wall, often to the regional lymph node-bearing areas; some women will require additional chemotherapy. Neither of these is an absolute contraindication to immediate reconstruction, but they may influence the technique chosen by the plastic surgeon. For example, if wound healing is threatened by trying to accommodate a saline-filled expander under a tenuously viable skin flap, it is preferable to postpone the reconstruction. When postoperative radiation therapy is likely to be required, autologous flap reconstructions (e.g., transverse rectus abdominus myocutaneous [TRAM] or latissimus reconstructions) may be preferred. However, these decisions should be made based on discussion between patient and plastic surgeon, with the input of the surgical oncologist and radiation oncologist. Because both breast reconstruction and NACT have become commonly performed procedures, it was the preference of most panelists to perform delayed reconstruction if radiation therapy to the chest wall was virtually mandated because of the preoperative findings. This delay would enhance the viability and the ultimate cosmetic results of the reconstructed breast.

Handling of the Specimen by the Surgical Pathologist

The most objective information concerning the effectiveness of NACT is provided by the microscopic assessment of the extent and type of residual tumor in the posttreatment surgical specimen. To aid the pathologist, the surgical specimen should be appropriately marked and oriented by the surgeon. The margins of the specimen should be inked or marked in another reproducible way. It is helpful if a specimen radiograph and/or ultrasound can be performed to compare with prior imaging studies. Imaging the specimen also is helpful in locating the lesion within the specimen and assessing its extent. In patients undergoing breast conservation, the entire specimen should be submitted after sectioning at intervals of 3–5 mm. If the specimen is too large to submit in its entirety in this manner, the macroscopically suspicious and the radiographically suspicious areas should be submitted. The step sections also aid in the semiquantitative estimation of residual tumor size based on the number of positive histologic slides.

When there has been a pCR (i.e., the absence of invasive tumor), the “tumor bed,” the pre-NACT tumor site, should be examined carefully. Characteristic findings at this site include fibrotic “scar” tissue with patchy lymphocytic infiltration, groups of foam cells, and a lack of glandular tissue.127, 128 Recognition of this “tumor bed” documents the validity of the pCR. It is mandatory for the pathologist to make every effort to find residual tumor. If not found on initial sampling, more extensive examination should be performed before a pCR can be confirmed.

The written pathology report should provide information concerning the following items: 1) the presence or absence of invasive and in situ carcinoma; 2) the histology, nuclear grade, histologic grade, and mitotic index of residual invasive carcinoma, and the histologic type and nuclear grade of residual DCIS; 3) the extent of residual invasive carcinoma (largest dimensions of invasive area[s]); 4) the total extent of both invasive and in situ carcinoma; 5) involvement of margins by invasive or in situ carcinoma (when positive, whether involvement is focal or extensive; if clear, distance from closest tumor focus to nearest margin, with identification of the location of the involved margin); and 6) the results of additional immunohistochemical studies and, eventually, gene profiling of the tumors.

ER, PR, and HER-2 assessments also are recommended. If ER/PR and HER-2 assessments were performed on the initial diagnostic specimen, it is not necessary to repeat them because they usually do not change after NACT. If it is convenient to do so (and permitted by the patient's insurance carrier), most of the panelists would, however, also repeat the ER, PR, and HER-2 assessments at the time of the definitive surgery.

The participants unanimously concurred that the pathologist's role in the treatment and probable outcome of the patient is critical and too often overlooked. Meticulous attention to the items outlined earlier is imperative to optimize patient care, especially for this group of women with locally advanced disease.

Indications for Postmastectomy Radiation Therapy

All the panelists, including the radiation oncologists, concurred that the role of postmastectomy radiation therapy after NACT is still evolving.129 Further research is indicated to identify appropriate candidates, optimal radiation therapy field arrangements, and optimal radiation doses to be used. However, several recommendations were acceptable to the majority of the group. Obviously, if the appropriate post-NACT surgical intervention is breast conservation, then radiation therapy is an integral part of the treatment regardless of the initial or post-NACT stage of disease. Radiation therapy is indicated to the intact breast with a boost to the primary site. Recommended radiation doses are 45–50 Gy over 4.5–5.0 weeks to the whole breast, followed by a boost to the primary tumor site, to a total dose of 60–66 Gy. Regional lymph node radiation also may be indicated in some patients, but these criteria are not well established because of the substantial variation in the clinical presentation of patients. The delivery of radiation in this setting is based largely on extrapolation from the setting of primary breast conservation treatment (without NACT) and is another appropriate topic for a clinical trial.

The tentative recommendations concerning postmastectomy radiation therapy will differ depending on the initial and post-NACT stage of the tumor. Therefore, patients with locally advanced breast carcinoma (Stages IIIA and IIIB) and inflammatory breast carcinoma (Stage IIIC) should receive comprehensive radiation therapy to the chest wall and supraclavicular fossa and some radiation oncologists also would consider adding comprehensive lymph node radiation. Recommended radiation doses would be the same as for postmastectomy radiation therapy in the absence of NACT.

Unlike patients who are treated conventionally (i.e., without NACT), for whom the pathology findings usually dictate the need for radiation therapy, it is necessary to consider the pretreatment clinical stage of disease and the pathologic extent after NACT to define criteria for radiation therapy of the chest wall and regional lymph nodes. To our knowledge, there currently are insufficient data to determine whether patients with clinical Stage I or Stage II breast carcinoma with fewer than three positive lymph nodes after NACT should receive radiation therapy to the axillary apex and supraclavicular fossa. However, patients with more advanced clinical stage disease or the presence of extracapsular extension of disease and those with four or more positive lymph nodes should receive radiation to both the breast/chest wall and the regional lymphatics.129, 130

For patients with Stage II breast carcinoma who are treated with a mastectomy after NACT, chest wall radiation therapy is indicated if four or more positive axillary lymph nodes are documented, similar to the indication for patients who have not undergone NACT.82, 131–133 As noted earlier, the indications for postmastectomy radiation therapy for patients with Stage II breast carcinoma with three or fewer positive axillary lymph nodes or those with negative lymph nodes after NACT are unknown because of lack of information. Because the number of positive lymph nodes after NACT is not necessarily the same as the number involved at the time of the initial diagnosis, there was again controversy regarding the need to perform axillary lymph node staging before NACT for patients with clinical Stage I or Stage II breast carcinoma to document lymph node status before it is altered by NACT. However, no consensus was reached on this subject. There was, however, broad consensus regarding the need to perform the appropriate clinical trials to answer these important questions concerning optimal locoregional control after NACT and NAHT.

Additional Chemotherapy after Definitive Treatment

Evidence from clinical trials provides support for the use of several NACT programs. Most of the existing long-term information is based on the use of three to six cycles of regimens such as FAC, FEC, CMF, or AC before definitive local therapy.14 The most recent generation of trials indicates that there is a definite advantage to adding a taxane, either simultaneously (TAC or epirubicin and a taxane [ET]) or sequentially (AC →taxane or taxane →FAC) to NACT programs.39, 40, 105, 106, 115, 134 In some recent clinical trials, four cycles of NACT are administered before surgery with an additional four cycles administered postoperatively, whereas in others, all eight cycles of NACT are given prior to the surgical intervention. It also is recommended that all patients with ER-positive and/or PR-positive tumors receive appropriate adjuvant endocrine therapy. It has been established by multiple reports of clinical trials that the presence of extensive residual invasive carcinoma within the breast or multiple positive axillary lymph nodes is associated with poor prognosis, regardless of the initial stage of the disease at the time of diagnosis. However, to our knowledge, there have been no clinical trials published to date that explored the administration of additional chemotherapy after the successful completion of a validated anthracycline-based and taxane-containing regimen. Therefore, it was the view of the majority that outside of a clinical trial a total of eight cycles of NACT should be offered (four of an anthracycline-containing regimen and four of a taxane-based treatment), and if all eight had been administered preoperatively, there was no demonstrated gain from additional chemotherapy. However, if only four cycles had been employed before locoregional treatment, an additional four cycles of an anthracycline or a taxane (depending on which had been used previously) would be indicated, regardless of the patient's initial response to therapy. Although the panelists believed that there are theoretic considerations that make 3-week cycles of treatment preferable to the newer, “dose-dense” 2-week cycle adjuvant chemotherapy regimens, whether postsurgical adjuvant therapy could be administered in 2-week cycles and whether subgroups of patients would be optimally treated with less or more therapy remains to be demonstrated by additional research.

Outcome after Neoadjuvant Chemotherapy

Clinical trials have documented that the integration of systemic therapies (chemotherapy and hormonal therapy) into the management of early and locally advanced primary breast carcinoma improves local control as well as recurrence-free and overall survival rates and decreases the risk of second primary breast tumors.3–5, 135 Randomized clinical trials have demonstrated that NACT and postoperative adjuvant chemotherapy are associated with at least identical recurrence-free and overall survival rates if the same regimens are used in a different sequence.33, 45 Based on these results, NACT is a reasonable alternative to traditional treatment strategies in patients with operable Stage II and Stage III breast carcinoma. Patients undergoing NACT should be aware of the reported data and ongoing controversies regarding the subject and agree to NACT knowing which information we have that is evidence-based, which is based on consensus, and that which is “probable” but not yet proven.

Clinical trial data have documented disease-free and overall survival benefits from NACT in patients who have achieved a pCR. Many publications of prospective but nonrandomized trials suggest improvements in both disease-free survival and overall survival in patients who have achieved significant tumor regression, even if not a pCR. The survival benefit associated with objective responses short of a pCR is difficult to document. Ongoing investigations are attempting to identify tumor-related or treatment-related factors that would successfully predict pathologic response to NACT.136

Future Directions

Much progress has been made in our understanding of the relative advantages and disadvantages of combined modality therapies based on NACT or NAHT. The efficacy of these regimens has been unquestionably established; NACT and NAHT produce major objective responses in the majority of patients with locally advanced and operable primary breast carcinoma. In 12–25% of patients, a pCR is achieved after state-of-the-art NACT regimens. Such objective responses often result in converting unresectable tumors into resectable tumors, and candidates for mastectomy only into candidates for breast-conserving procedures. We have identified the prognostic value of achieving a pCR and currently are attempting to validate its usefulness as a surrogate endpoint for survival. We understand better the surgical issues related to extent of resection, clear margins, and axillary lymph node staging and have defined those areas in which additional investigation is needed. The role of radiation therapy in patients treated with breast-conserving surgery after NACT is well defined. Similarly, there is general agreement regarding the value of postmastectomy radiation therapy for those patients with multiple positive lymph nodes or Stage III breast carcinoma at the time of diagnosis. Additional research is needed to determine the role of and indications for radiation therapy in all other subgroups treated with NACT. The role of and indications for NAHT currently are under evaluation. Although adjuvant hormonal therapy is well established, there is uncertainty regarding the role of NAHT. This is particularly challenging because pCRs have been uncommonly documented after NAHT.71 It has been suggested that NACT is more effective in patients with hormone receptor-negative tumors, and it has been demonstrated that NAHT is effective only in women with hormone receptor-positive tumors. The interaction of hormonal receptors with other molecular markers such as HER-2, p53, or bcl-2 remains to be explored. Similarly, whether all patients with hormone receptor-positive breast tumors larger than T1N0 are best treated with a combination of chemotherapy and hormonal therapy or whether there are subgroups who are optimally treated with hormonal therapy in the absence of chemotherapy is currently under intense investigation.

Additional research should determine the optimal sequence of treatments, both local and systemic. Because systemic therapy followed by locoregional therapy appears at least equivalent to the reverse sequence, how should we determine the optimal treatment strategy for individual patients? Should all chemotherapy be administered preoperatively? Should some be given as NACT and the remainder as postoperative adjuvant treatment? In this latter case, should chemotherapy precede or follow radiation therapy? There are multiple opinions influencing this decision without compelling data to support either approach.

At the conclusion of the conference, the panelists agreed that the questions for which they did not have answers far outnumbered those for which solid data were available. However, the group agreed that neoadjuvant chemotherapy was the appropriate treatment for women with locally advanced Stage III breast carcinoma (IIIA, IIIB, or IIIC) and was worthy of consideration in women with Stage IIA, T2N0, T1N1, or Stage IIB, T2N1, or T3N0 tumors, irrespective of tumor size (according to the TNM classification). Whether NACT should be offered to all women with breast carcinoma who would be chemotherapy candidates after treatment based on pretreatment criteria alone, irrespective of subsequent surgical pathology findings, was not determined to be a therapeutic imperative but was considered a reasonable treatment alternative.

One exciting aspect of the strategy based on NACT is the ability to monitor the effects of therapeutic interventions by obtaining serial biopsies of the primary tumor or regional lymph nodes while they are still in place, or monitoring other biologic tissues (plasma, serum, urine, or saliva) in an attempt to determine molecular profiles associated with prognosis, response to therapy, or predictors of resistance to specific interventions. These research approaches are quite limited or not possible if all evidence of tumor is removed by implementing definitive locoregional therapy as the first intervention.

There also exists the possibility of assessing serial functional imaging with PET or magnetic resonance spectroscopy for the early determination of response or resistance to NACT or NAHT. As clinical trials continue to answer the multiple questions identified during this consensus conference, these research approaches should be incorporated into their design to maximize the benefits of the research process and accelerate progress in the management of patients with primary breast carcinoma.