In a single generation the treatment of most women with early-stage breast cancer has changed dramatically. Clinical trials with more than 20 years of follow-up have documented that for appropriately selected patients breast conserving surgery, followed by whole breast irradiation, has an outcome equivalent to mastectomy. The appropriate selection of patients remains controversial. Questions remain regarding the role of new diagnostic imaging modalities, radiation therapy techniques, achieving optimal cosmesis, the choice and the timing of adjuvant chemotherapy, etc. To address these issues, the fifth international consensus conference of the Breast Health Institute, cosponsored by the European Institute of Oncology, was convened in Milan, April 29 through May 1, 2005. This article is a summary of the proceedings.
In a single generation the treatment of most women with early-stage breast cancer has changed dramatically. Clinical trials with more than 20 years of follow-up have documented that, for appropriately selected patients, breast conserving surgery followed by whole breast irradiation has an outcome equivalent to mastectomy. The appropriate selection of patients remains controversial. Questions remain regarding the role of new diagnostic imaging modalities, radiation therapy techniques, achieving optimal cosmesis, the choice and the timing of adjuvant chemotherapy, etc.
To address these issues, the fifth international consensus conference of the Breast Health Institute, co-sponsored by the European Institute of Oncology, was convened in Milan, April 29 through May 1, 2005, inviting a group of experts in breast cancer and breast conservation, representing each of the disciplines involved in the care of these patients. From recorded transcripts of the discussion, these proceedings were written to summarize the opinions and conclusions of the entire group. The conference dealt only with invasive cancer; ductal carcinoma in situ was intentionally not discussed.
Breast conservation (BCT) implies the complete removal of the breast tumor with a concentric margin of surrounding healthy tissue, performed in a cosmetically acceptable manner (“lumpectomy”), usually followed by radiation therapy. Surgical evaluation of the axillary lymph nodes is customarily a part of BCT; the panelists agreed that the treatment of the breast and the axilla should be considered separately; the presence of pathologically positive axillary lymph nodes is not a contraindication to BCT.
OUTCOME AFTER BREAST CONSERVATION
BCT and mastectomy offer equivalent long-term survival for appropriately selected patients. Breast conservation does imply risk for an “in-breast” recurrence or a new primary cancer in the same breast. Most panelists agreed that, unlike local recurrence after mastectomy, usually a forerunner of systemic disease, in-breast recurrence after BCT may be an isolated event treated successfully by additional surgery, usually mastectomy. Nevertheless, there was unanimous support for every effort being made to achieve optimal local control.
The goal of BCT should be a 10-year local recurrence rate between 5% and 10%—less than 1%/yr. The demographics of the treated population might produce variations in these observations, but the group unanimously supported these 10-year benchmarks.
Treatment of an in-breast recurrence is traditionally by mastectomy, with or without reconstruction. Several panelists championed local excision alone for recurrence as ductal carcinoma in situ (DCIS) or for small invasive tumors but acknowledged that a second attempt at BCT is associated with an increased risk of additional recurrence. The use of accelerated partial breast irradiation (APBI) (discussed later) to treat an in-breast recurrence was discussed; there was no agreement on the appropriate indications for such an approach. All agreed that a short interval between initial treatment and recurrence (<2 years) implies radioresistance, and APBI would be inappropriate in those patients. All concurred that a second course of whole breast irradiation is not appropriate, irrespective of the time between events.
The occurrence and treatment of axillary recurrence was addressed, although this event after a well-performed axillary dissection is negligible. With respect to sentinel node biopsy (SLNB), available data fail to document a greater frequency of axillary recurrence after SLNB performed by experienced surgical teams. The treatment of axillary recurrence must be individualized to the patient circumstances; if axillary recurrence were to follow SLNB, completion axillary dissection was advocated along with any other treatment recommendations.
Can a second SLNB be performed for an in-breast recurrence? The surgeons on the panel, all experienced in SLNB, agreed that a second SLNB is possible, and when successful, could be considered accurate. However, there is little evidence-based data to support this conclusion. All agreed that, if the axilla had not been completely dissected at the time of initial diagnosis, an axillary staging procedure was indicated.
The Primary Lesion
Wide local excision of the primary lesion is crucial. There was no controversy about the need to achieve clear microscopic margins around the boundaries of the tumor even if it means an additional procedure to do so. There was, however, considerable debate regarding the width of the clear margin required.
The initial diagnosis of breast cancer is optimally confirmed by “minimally invasive” core biopsy. Fine-needle aspiration (FNA) biopsy may be used, but only in close liaison with an experienced cytopathologist. After the diagnosis is established, therapeutic options should be discussed. Currently, the majority of women with T-1 and small T-2 (<3 cm) cancers are suitable candidates for BCT. Selected larger tumors may be treated by BCT if the primary tumor can be excised adequately with clear margins and acceptable cosmesis. Oncoplastic techniques may be warranted to allow wide excision with clear margins without compromising cosmesis. Oncoplastic surgery, combining sound principles of surgical oncology with those of plastic surgery, may extend BCT possibilities. When employed, breast reshaping techniques should be performed at the time of the initial surgical excision. It is difficult to correct a poor cosmetic result secondarily, after radiation therapy.
The first invasive surgical step after the diagnosis of invasive cancer by the initial core biopsy should be the attempt at complete excision (usually including whatever axillary procedure has been recommended). Some women remain in favor of mastectomy; their wishes should be respected. When there is a relative contraindication to radiation therapy (discussed later), mastectomy remains appropriate.
The panelists agreed to disagree with the width of the margins required to justify BCT after local excision of the primary tumor; all concurred that there should be no evidence of tumor at the transected edge of the excision, per National Surgical Adjuvant Breast and Bowel Project (NSABP) guidelines. Most panelists agreed that a wider margin implies a lower likelihood of recurrence, despite the lack of evidence-based data supporting any specific margin width. Most radiation oncologists were comfortable with a margin of at least 1–2 mm. All the panelists favoring wide margins insisted that this could be achieved in most patients, with a minimal negative effect on cosmesis. The technique of shaved margins after the wide excision of the lesion was also considered acceptable, since this technique implies at least a 10-mm free margin.
The group recognized the dichotomy between the extent of local excision and the cosmetic result. The surgeons skilled in oncoplastic techniques championed this approach to extend the scope of BCT to include patients with tumors 3–5 cm in diameter. The panelists endorsed the teaching of oncoplastic techniques as part of surgical training for specialists in breast cancer care.
Other previously controversial issues about BCT include what has been called an extensive in situ (ductal) component (EIC) within the excised tumor, and the presence of lobular carcinoma in situ (LCIS) at the margins. There was agreement that there was no volume of DCIS within the primary lesion that would preclude BCT. Likewise, the presence of LCIS or atypical ductal or atypical lobular hyperplasia at the margins was not a factor that would either require re-excision or interdict BCT. Marking the boundaries of the tumor site with radioopaque clips was encouraged by both surgeons and radiation oncologists.
An additional question was the treatment of central lesions. If the nipple–areolar complex is directly involved with tumor, then it must be removed with the local excision, and the patient must decide whether the sacrifice of this tissue is enough to tip the scales for her toward mastectomy, although the center of the breast can be subsequently reconstructed if desired. The wide excision of central lesions is often an indication for oncoplastic techniques to avoid flattening of the breast and to allow optimal reshaping of the breast contour. In patients with lesions close to but not involving the nipple–areolar complex, BCT without removing the nipple and areola is an option, so long as a clear plane can be developed between the base of the nipple and the lesion.
The localization of nonpalpable lesions (after initial core biopsy) by the traditional procedure of needle-guided localization most commonly used in North America has been replaced, in part, by the radioguided occult lesion localization (“ROLL”) technique in some centers, using a radioactive tracer injected under radiographic control to localize both the lesion and the sentinel nodes (SNs). Whatever the technique, the satisfactory identification of the lesion, proof of its excision, and clear margins are the criteria of appropriate surgery.
The contemporary treatment of carcinoma of the breast includes the axillary lymph nodes. In elderly patients, those with intercurrent disease or small, especially favorable cancers, such treatment may be debatable; nevertheless, the group unanimously endorsed axillary staging if axillary node status would influence a therapeutic decision.
Axillary sentinel lymph node biopsy (SLNB) has replaced traditional axillary dissection in women with clinically negative axillary findings (N0) and was unanimously endorsed. Whether internal mammary nodes were to be considered part of the sentinel node (SN) discussion was controversial. The majority of the panelists do not currently try to identify or dissect internal mammary nodes. Which technique of SLNB was best was not stressed, so long as the SNs could be identified and were accurate predictors in >90% of patients. Most clinical trials support the use of both radiocolloid and blue dye to achieve the highest rates of SN identification. (Radiocolloid is the only technique possible for the identification of internal mammary nodes.)
When SNs are positive for metastatic disease, whether the remainder of the axilla should be dissected was a minor point of contention. If the SNs demonstrate macrometastasis on frozen section, a majority agreed that completion dissection of the axilla was appropriate. Several panelists thought that radiation to the axilla was a suitable substitute for further surgery. Traditional breast radiation fields do not completely encompass the axillary node sites, and special attention to the anatomy of the fields would be necessary to ensure that Levels I and II would be included with the whole breast radiation. There is a current European Organization for Research and Treatment of Cancer (EORTC) clinical trial comparing completion axillary dissection and radiation therapy to the axilla for confirmed positive SNs that may help answer this question. A small minority of the panel questioned whether either further surgery or radiation was required, since these patients would be undergoing adjuvant chemotherapy.
When a single SN shows micrometastasis, the panel was evenly divided about the need for additional axillary surgery or axillary radiation because the majority of patients with micrometastasis do not have additional node metastasis. Immunohistochemical markers are being used commonly to identify submicroscopic metastasis, single cells or small groups of cells. Because their clinical significance is uncertain, the presence of immunohistochemically positive cells is not a current criterion per se for completion of axillary dissection or for adjuvant therapy.
The extent of axillary dissection when indicated by the presence of positive SNs was debated. North American surgeons favored level I and II axillary dissection. A few European surgeons favored including level III. Most panelists felt that the additional dissection did not improve survival and increases the risk of lymphedema. All agreed that a precise anatomical dissection removing at least 10 nodes was essential.
SELECTION OF PATIENTS FOR BREAST CONSERVATION
Nonpalpable tumors detected by initial imaging are usually suitable for BCT, but there is theoretically no size limitation for BCT so long as local excision with clear margins can be achieved with an acceptable cosmetic result. Tumors up to 5.0 cm in diameter have been successfully treated by BCT in both randomized trials and retrospective studies. There may be a role for oncoplastic techniques in such situations.
Detailed imaging should be performed prior to definitive surgery, to exclude the presence of multiple primary cancers, an occult contralateral breast cancer, and diffuse calcifications that may indicate widespread additional DCIS.
Multicentricity is the occurrence of 2 separate cancers in different quadrants of the same breast, far enough apart that they cannot be excised through a single incision. The panelists agreed that if 2 separate cancers were close enough to be excised as a single specimen, with clear margins and acceptable cosmesis, BCT could still be used. An additional concern with multicentric tumors is the complexity of adding a radiation boost to more than 1 site in the same breast, which is currently not recommended.
The panelists agreed that the younger the patient, the greater the likelihood of a local recurrence after BCT. Nevertheless, all concurred that this does not preclude the use of BCT in these patients so long as physicians and patients accept this increased risk. After controlling for other factors, there is no worse long-term survival after BCT when compared with that after mastectomy in young patients. Most women younger than 50 years should have a boost dose of radiation to improve local control.
Whether radiation therapy could be avoided in older women who would be undergoing adjuvant hormonal therapy after local excision was debated. All concurred that all groups of patients benefit by the addition of radiation therapy to local excision, but the gain in older women may be small, and the routine use of radiation treatment (XRT) in this group was not resolved. Individualizing the decision about XRT in older women based on tumor characteristics and physiologic rather than chronologic age was encouraged. The role of accelerated partial breast radiation in this group of patients was discussed (discussed later).
Ultrasound and Magnetic Resonance Imaging
Evidence for the use of ultrasound (US) as a supplement to mammography (MG) for screening does not exist, but US may help characterize masses, architectural distortions, or areas of asymmetry. It is not usually helpful in the differentiation of calcifications. Ultrasound may aid in evaluating the mammographically dense breast to avoid overlooking additional suspicious lesions and may help guide the percutaneous biopsy of palpable or image-detected lesions.
Because the sensitivity and specificity of US are so operator- and technique dependent, the practice guidelines for breast US promoted by the American College of Radiology (ACR) should be observed.1 Mammography and US are complementary, and so the group emphasized a preference for MG and US (and magnetic resonance imaging [MRI], if used), to be correlated and interpreted together by the same radiologist(s). Whether US of the axilla is valuable to identify metastatic nodes was controversial; US was considered helpful if it could identify and prove a positive node by percutaneous FNA or core biopsy prior to treatment planning.
The addition of MRI to MG and US has increased the diagnostic accuracy of these modalities when used together. MRI is highly sensitive but less specific; its use in staging breast cancer patients is probably helpful but still debatable. There is a need for agreement on which enhancing lesions require biopsy and how far to pursue so-called suspicious lesions before considering BCT safe. The routine use of MRI has not yet been shown to be valuable in a randomized trial, but the advocates of MRI argued that it might affect treatment recommendations in up to 20% of patients.
MRI may be used in patients with large lesions to aid in planning surgical excision or for patients with suspected second lesions on MG or US. In patients with dense breasts or with lesions difficult to evaluate by MG, such as invasive lobular carcinomas, MRI might help guide the extent of surgery required. However, US may be an easier and less costly technique to use in these situations. The panelists agreed that MRI requires considerable technical and interpretation skills; all agreed that MRI facilities should have dedicated breast coils with minimum field strength of 1.5 T.
The panelists agreed that breast MRI is a promising tool; its potential has not yet been realized, and its use remains controversial. Probably its best use is in patients who present with axillary metastasis and whose MG and US show no abnormalities, and for screening women who are Breast Cancer Genes 1 and 2 (BRCA1-2) positive or those younger than 40 years. The group agreed that MRI in premenopausal women should be performed between days 7–17 of the menstrual cycle because of parenchymal enhancement seen at other times of the cycle. There was agreement that all imaging procedures should be completed before any surgical procedure (other than biopsy) is performed, and that MRI should be performed by facilities that have the capability of working up the findings to completion, i.e., MRI-guided biopsy. MRI is helpful in establishing the precise size of the cancer, the presence of multifocality, and/or the presence of an additional cancer in the same breast (multicentricity). MRI, when used, should be performed before or as soon as possible after a core biopsy. A core biopsy may obscure neighboring findings or precise size measurements if accompanied by a significant hematoma, but the ability of MRI to exclude a second lesion will usually not be impaired.
Another situation in which MRI might be helpful is for cancers that present with nipple discharge. These cancers are often widespread within the ductal system and difficult to excise. MRI may define the tumor boundaries more clearly and make the margins easier to clear widely if its findings guide the dissection.
Finally, the group endorsed clinical trials that use MRI findings to modify treatment but concurred that the routine use of MRI prior to BCT was not yet confirmed. Its role in selected subgroups of patients, as described earlier, is promising. An MRI diagnosis of a second lesion should never influence a treatment decision without histologic proof (biopsy). Mastectomy should not be performed solely on the basis of MRI findings.
The panelists agreed that the handling and interpretation of breast biopsies and other breast specimens should follow standardized protocols that are regularly updated to incorporate modifications necessitated by advances in new immunohistochemical and subcellular techniques and surgical innovations. The College of American Pathologists (CAP) provides a set of guidelines specifying the current basic information required in pathology reports.2 This protocol should be followed; each pathology laboratory may make minor adjustments to the CAP guidelines to fit its own needs and requirements.
The accuracy of margin assessment and overall pathological interpretation of the surgical specimen are dependent upon the cooperation of the surgeon, the radiologist, and the pathologist. The surgeon should appropriately mark and orient the specimen. The margins of the specimen should be inked if submitted as a single tissue excision, or shaved and submitted in separately marked containers if that technique is used. Submitting a radiograph of the specimen for comparison with earlier imaging studies is helpful in locating the lesion within the specimen and assessing its extent. In patients undergoing BCT, it is preferable to submit the entire specimen after sectioning at 3–5 mm intervals. If the specimen is too large to submit in its entirety, then all the grossly and radiologically suspicious areas should be submitted. A specimen radiograph may also aid the identification of the closest margins to the tumor. In all cases in which specimen radiography is obtained, the radiographs should be available for the pathologist for radiographic–pathologic correlation.
To assess margin status, pathologists usually evaluate 2 types of specimens. Either multiple shaved margins are submitted as separate specimens by the surgeon, or the tumor is excised and submitted as 1 single specimen with specific markings regarding its orientation provided by the surgeon. Both of these techniques were considered appropriate.
Emerging technologies in evaluation of tumor characteristics may impact either outcome or therapeutic recommendations and should be adopted as available. Analysis of gene profiles in breast tumors may be a valuable tool with significant potential impact, but issues that should be resolved before gene profiling is routinely applied to patient care include quality control, statistical analysis of the data, and establishment of expression thresholds. All of these subjects were deemed suitable for research trials but are not ready for use in current clinical practice.
WHO TREATS BREAST CANCER?
The skill of the treatment team is crucial to a successful outcome. A dedicated, interdisciplinary breast center as a care-model for breast cancer diagnosis and treatment is optimal. In Europe, such centers are common, and it has been generally accepted that an institution must treat at least 150 patients with breast cancer per year to be considered a breast cancer specialty center.
Because breast cancer patients are usually referred first to a surgeon, European guidelines recommend that breast surgeons work in close collaboration with radiologists, pathologists, and medical and radiation oncologists. The European panelists also agreed that breast surgeons should treat a minimum number of breast cancer patients annually. The current European minimum requirement for a specialized center is 50 cases per surgeon per year. In favor of the establishment of such quotas and building specialized centers are the observations that the use of BCT is more common, the rate of post-BCT recurrence is lower, and the survival of women with breast cancer is higher when they are treated in specialized breast centers. In North America, where patterns of referral are not mandated, and geographical distances may preclude patients from traveling far from home, such guidelines may prove difficult to adopt.
The consensus panel discussion included technical aspects of XRT. The goal of XRT is to treat the target volume while minimizing dose to surrounding normal tissues. The conventional target volume initially includes the whole ipsilateral breast. The whole breast dose is generally 45–50.4 Gy delivered over 4.5–5 weeks. Typical daily fraction size is 1.8 or 2.0 Gy delivered 5 days per week. Megavoltage photons, most commonly 6 MV, treat the breast via tangent fields. There were no specific guidelines regarding the acceptable volume of lung and heart (for left-sided breast cancers) that could be treated, except that the treatment of these organs should be minimized while still treating the target volume adequately.
Historically, XRT planning was simple and used fluoroscopic simulation, single plane dosimetry, simple wedges, no lung correction, and no three-dimensional (3-D) treatment planning. Long-term outcome data for BCT treatment have been obtained from patients treated using these techniques.
More recently, CT-based simulation and treatment planning have become widely available and are now preferred by many radiation oncologists. Intensity-modulated radiation treatment (IMRT) is an even more sophisticated method of dose delivery; treatment planning studies have shown that IMRT can improve dose homogeneity within the target volume as well as reduce the dose to normal tissues, including ipsilateral lung and heart (for left-sided lesions). The panelists endorsed the potential value of more sophisticated XRT planning and dose delivery, but await long-term outcome data to support this practice change.
Randomized trials have shown an improvement in local control when a boost dose of XRT is delivered to the primary tumor site after the initial whole breast radiation, including the tumor site and a 1- to 2-cm margin of normal tissue. The absolute benefit from the boost is proportional to the baseline risk, and therefore, younger patients have the greatest absolute benefit from adding the boost. The tumor bed dose from combined whole breast and boost XRT is typically 60–66 Gy. Boost XRT most commonly uses electrons, but various methods have been described, including 3-D photon beams and brachytherapy. Some radiation oncologists use a boost dose for all patients; others selectively add a boost dose for higher risk patients or when the margins of resections are not confirmed as pathologically negative. The majority of North American radiation oncologists use a boost for virtually all patients with invasive cancers.
When to radiate regional lymph nodes (axilla, supraclavicular fossa, and/or internal mammary lymph nodes [IMNs]) remains controversial. No consensus could be reached regarding the indications for adding or modifying nodal radiation in the presence of extracapsular extension or involved lymph nodes, and this was an area of accepted lack of agreement among the panelists. The lower axilla is typically included within standard tangential fields, but the risk of arm lymphedema increases with the addition of direct nodal fields to the supraclavicular fossa and axilla. A randomized clinical trial (National Cancer Institute of Canada-Clinical Trials Group study MA20) is accruing patients in North America and Australia, comparing whole breast XRT to whole breast and nodal XRT.
Another area of disagreement was the management of the IMNs, especially when this area is the site of drainage on sentinel lymph node mapping. A minority of the panelists recommended biopsy of the specific IMN, restricting IMN radiation to those with positive biopsies. A large EORTC trial comparing IMN radiation to no IMN radiation has completed accrual, but results from this trial await analysis.
In selecting patients for treatment, advanced age and significant intercurrent disease were discussed as relative contraindications to XRT (discussed later). It was accepted that there is no group of patients for whom XRT does not further decrease in-breast recurrence when compared with local excision alone. The panelists agreed that individual patient circumstances might dictate the substitution of hormonal therapy for XRT in some elderly, informed patients with completely removed, estrogen receptor (ER) positive disease. In patients with ER-negative disease, hormonal therapy is not an option, and so the panelists recommended XRT for all but the most feeble patients.
Another group of patients for which XRT might be avoided is those with especially favorable tumors, e.g., small tubular or mucinous/colloid carcinomas. Again the same caveats apply. The panelists agreed that there may be patients with such favorable tumors who do not need XRT, but a precise algorithm to identify them does not exist.
CONTRAINDICATIONS TO WHOLE BREAST XRT
Pregnancy is currently an absolute contraindication to XRT. However, breast cancer occurring in pregnancy need not imply mastectomy. It is possible to perform surgery and then wait until after delivery to start the radiation therapy. The breast cancer care should be coordinated with a high-risk obstetrics specialist, since delivery may be induced or accomplished by cesarean section at an earlier than usual gestational age. A delay of up to 3 months between surgery and XRT was not considered harmful.
A relative contraindication to XRT is a past history of mantle radiation for Hodgkin disease (HD) because the mantle technique partially radiates both breasts. Whether there is a certain time from the treatment for HD and the diagnosis of breast cancer, beyond which the tissues will tolerate whole breast XRT, is controversial, and there are not much data extant to make such a recommendation. A small number of cases have been treated with BCT and whole breast XRT, with acceptable outcomes. A past history of mantle XRT for HD is one of the possible indications for APBI (discussed later) if the prior treatment fields do not overlap the excision site. There are no data to support or refute this suggestion.
A personal history of scleroderma, systemic or discoid lupus, or dermatomyositis was considered a relative contraindication to XRT; none of the radiation oncologists on the panel was eager to treat these patients. Rheumatoid arthritis was not considered a contraindication.
Another special circumstance was the presentation of breast cancer as axillary node metastases without a demonstrable primary lesion in the breast by clinical examination or by breast imaging with MG and US. In such situations, breast MRI should be performed, because MRI will detect the primary tumor in 80%–90% of patients whose MG and US and physical examination showed no abnormalities. Presuming that the diagnosis of breast cancer is a reasonable certainty based upon microscopic and immunohistochemical characteristics of the axillary node(s) and appropriate workup to rule out another primary site, many of these patients will be treated by neoadjuvant chemotherapy. Even if not, the group felt that axillary surgery, followed by XRT to the breast, was an alternative to mastectomy, even if the primary site could not be found. The need for postoperative radiation to the node-bearing areas after axillary dissection remained a controversial topic, with most panelists reserving judgment based upon the microscopic findings in the axilla.
A corollary of the treatment of the “occult” cancer is the fortuitous finding of cancer in a breast specimen removed during the course of a breast reduction or mastopexy. In such situations, if margin assessment is considered accurate, i.e. the tumor has not been transected, and there is no tumor thought to be within the remaining breast, BCT is reasonable. Surgical attention to the axilla, followed by XRT to the breast, is acceptable treatment. If there is a suspicion of macroscopic residual cancer in the breast, mastectomy may be the only tenable option.
Prior Breast Augmentation
The group felt strongly that a comment was necessary about the occurrence of carcinoma in patients who have previously undergone augmentation mammoplasty. The presence of an implant, silicone or saline, does not influence the decision for BCT and does not interfere with the technical delivery of XRT. There is no reason to remove the implant, assuming that the tumor can be successfully excised. Radiation therapy to that breast will produce some fibrosis and increase the risk of capsular contraction. So long as the patient understands that her breast may not be the same as it was before treatment, BCT is an appropriate option.
ACCELERATED PARTIAL BREAST IRRADIATION
Recent interest has focused on APBI. Accelerated partial breast irradiation is the delivery of radiation to a limited target volume—the surgical cavity and a 1- to 2-cm margin—generally in a single treatment or over 1 week. This contrasts with conventional whole breast XRT, which generally takes 5–7 weeks. APBI has been designed for patient convenience; it is not expected to improve local control or survival. Since its introduction in the 1990s, APBI has been used in multiple phase I-II clinical trials, with short-term (5 year) local failure rates that are similar to those for whole breast radiation. Support for APBI has been based upon observations of in-breast failures after traditional radiation therapy because recurrence is seen most frequently (70%–80%) at or near the site of the original tumor.
The optimal APBI technique has not been determined. At least 4 methods of APBI have been de scribed: 1) brachytherapy implant; 2) balloon brachytherapy (MammoSite®); 3) external beam XRT, with or without IMRT; 4) intraoperative radiation treatment, using electrons (ELIOT [electron intraoperative treatment]) or an orthovoltage source (TARGIT [targeted intraoperative radiotherapy]). Each of these APBI methods has potential advantages and disadvantages, and none has been proved superior to another.
Several Phase III randomized trials comparing APBI to conventional whole breast radiation are accruing patients. Each trial is evaluating a different APBI technique, but each trial uses whole breast photon XRT as the control arm. The panelists overwhelmingly endorsed entering patients into these randomized trials so that the relative worth of APBI in comparison with standard, whole breast XRT can be established. Long-term outcomes for these randomized trials will be necessary to judge the differences between these 2 forms of XRTs for the multiple end points of survival, local control, cosmesis, and complications.
Since there are now multiple open randomized trials, the panelists agreed that APBI should not be used outside of a protocol study except in exceptional circumstances, which might include women who live at a considerable distance from an XRT facility who cannot undertake a lengthy period of treatment, or elderly patients with intercurrent diseases that may already limit long-term survival. When APBI is used, the criteria for the surgical procedures and techniques and the pathologist's handling of the specimens are the same as when conventional whole breast radiation therapy is used.
As the indications for adjuvant chemotherapy have expanded, more patients must plan to undergo both radiation and chemotherapy. The correct sequence of these treatments has been questioned, and many institutions have adopted one approach or another. The panelists agreed that there are no current data to support which should be undertaken first, radiation or chemotherapy. All agreed that radiation and chemotherapy should not routinely be given concurrently. A majority felt that there was no advantage to XRT or chemotherapy given first for node-negative patients; patient choice and local traditions could influence this decision. For patients with node-positive disease, a majority agreed that chemotherapy should be given before XRT. There was consensus that hormonal agents should not be given concurrently with chemotherapy, but there was no contraindication to their use during XRT. The optimal choice and duration of chemotherapy and hormonal therapy were not part of the discussion.
INITIAL PATIENT WORK-UP AND FOLLOW-UP
For those patients with Stage I cancers, pretreatment evaluation includes the appropriate evaluation of the breast alone. There is no advantage in performing additional blood tests, bone scans, or chest X-rays either before or after treatment except in the presence of symptoms. For patients with Stage II or Stage III cancer, a pretreatment metastatic evaluation with liver enzymes, a bone scan, and chest radiograph is appropriate. The addition of chest, abdominal, and pelvic CT, and/or MRI scans and PET scans should be individualized and depends upon the presenting situation. Whether these studies are repeated serially after treatment should be an individual rather than a universal decision.
The timing of the first mammogram after treatment is controversial. The panel unanimously agreed that it should not be performed before 6 months have elapsed after completion of radiation therapy, and many thought a year would be preferable. The tradition of 6-month mammograms of the treated breast for 2 years after treatment has become popular but has not been shown to be superior to annual MG. Both breasts must be imaged at least annually, with other imaging studies, e.g., breast US and MRI, added as indicated by individual circumstances. Whether serial (annual) MRI examinations might add to the accuracy of observing the treated breast was debated; no data exist to support or challenge this suggestion. The cost may be prohibitive. The value of observing the treated breast by MRI was considered a suitable subject for a clinical trial.
The panelists were aware and supportive of the increasingly valuable information from molecular and genomic profiling of breast malignancies and their enormous potential. Such information is already having a major impact on reclassifying and managing breast malignancies, particularly for using systemic therapies. There is an increasing number of newer chemotherapeutic and biologic agents, some of which have already demonstrated substantial impact in reducing the number of breast cancer deaths in subsets of patients and possibly even reducing the rate of local recurrence. Although each of these issues is recognized, their direct impact on BCT treatment remains to be established by future investigation.
Newer imaging techniques, including functional imaging, are rapidly undergoing development. These newer imaging techniques hold great promise relative to BCT treatment. Such imaging advances will likely impact on many diverse aspects of BCT treatment, although this impact remains to be demonstrated in clinical practice.
As the conference concluded, the panelists largely concurred on standards for the selection of patients for BCT, agreeing that tumor characteristics, such as histology, multifocality, and presence of EIC, do not affect the suitability for treatment by BCT so long as the tumor can be completely excised with clear margins. The major disagreement was the width of the clear margin required around the tumor, ranging from no tumor at the transected edge of the excision to free margins of 10 mm circumferentially around the tumor.
The emerging role of MRI in the selection of patients for BCT was appreciated by all the participants. The lack of specificity of MRI requires continuing clinical trials to unravel this problem before MRI can be used to define patient selection more precisely.
The faculty of this conference unanimously agreed with the concept of multidisciplinary treatment centers as the most efficient and effective manner in which breast cancer patients should be treated. As the science of breast cancer expands exponentially, patient benefits depend upon the continuing collaboration and cooperation of multiple disciplines.
All of the panelists agreed that the major issues in BCT had been addressed in enough detail to permit the development of these formalized guidelines for patient care. The group recognized, however, that these published proceedings are editorial opinions of the group and not strictly evidence-based. Therefore, they must not be construed as establishing a standard of care to which all treating physicians should adhere. Individual patient management must be determined by unique patient circumstances and health providers' capabilities. Treatment recommendations are the responsibility of individual physician(s) with the full participation of the patient.