Dr. DeAngelis is an Associate Professor of Mammography, Dr. Gizienski is a Fellow in Mammography, and Dr. Moore is Assistant Professor at the University of Virginia Health Sciences Center in Charlottesville, VA.

T he surgical treatment of breast carcinoma is becoming less invasive. Lumpectomy or simple mastectomy with axillary node sampling or modified radical mastectomy has replaced radical mastectomy as the standard treatment for invasive breast cancer, except in cancers with chest wall invasion. Because axillary status is the single most important predictor of systemic recurrence, axillary dissection has become a standard part of surgery for invasive breast cancer to determine the need for systemic (chemotherapy or endocrine) therapy for many patients. However, unresolved in the surgical literature is whether axillary dissection itself also provides some patients with local control of axillary metastases.

In general, earlier diagnosis of breast cancers has resulted in higher proportions of patients found to have low-stage disease and, thus, there has been a gradual decrease in the proportion of patients with positive axillary lymph nodes. With most patients having one or no positive axillary nodes, attention has focused on the recent development of techniques to minimize the extent (and, hence, the morbidity) of the axillary dissection technique. A less invasive means to determine axillary lymph node status is imperative.

Beginning with studies of melanoma patients, surgeons have developed techniques to identify the first (or sentinel) node draining the area of a tumor. ¹ Giuliano ² found that identification and examination of the sentinel lymph node is feasible and accurately reflects the status of the remaining axillary lymph nodes, with false negatives in the 5 to 15% range. The American College of Surgeons recently began a national trial to confirm that sentinel node surgery is a feasible alternate to the standard Level I and II axillary dissection. If accuracy is determined to be acceptable through clinical trials, sentinel node sampling is likely to reduce the need for axillary dissection for axillary lymph node staging.

Axillary nodal levels

Axillary surgery ranges from a complete dissection of all nodal tissue to axillary sampling or removal of several nodes in the lower axilla without necessarily conforming to defined anatomic levels. The number of nodes found in this region is highly variable. Much of this variability is patient-dependent, though there are several published studies that indicate that the number of nodes found in a surgical specimen (which consists of nodes embedded in axillary fat) is also highly dependent on pathologist technique.

The extent of axillary surgery is correlated with the extent of lymph node dissection from lateral to medial within the axillary triangle. This triangle is bordered by the axillary vein superiorly, latissimus dorsi laterally, and the serratus anterior medially. Lymph nodes within the triangle are grouped as level I, II, and III nodes: Level I (proximal) nodes are lateral to the pectoralis minor muscle and often are removed with the breast tissue as part of a simple mastectomy specimen; level II (medial) nodes are located posterior to the pectoralis minor; and level III (axillary) nodes are medial to the pectoralis minor against the chest wall (figure 1). Metastases to supraclavicular and internal mammary lymph nodes are considered distant metastases.

Metastases almost always first involve level I and II nodes before involving nodes in the upper axilla. Skip metastases to level III are estimated to occur between 1.5 and 5%. ^3-6 Thus, the axillary dissection currently performed by most surgeons is limited to a level I and II dissection because this technique has been found to accurately stage the patient, and there is a decreased risk of lymphedema.

Indications for sentinel node biopsy and axillary dissection

The size of the primary tumor and the presence and number of positive axillary lymph nodes are the most important factors in predicting distant metastases in patients with breast cancer. Presence of nodal metastases decreases the 5-year survival rate by 40%. ² If a diagnostic axillary dissection reveals positive nodes, the addition of adjuvant radiation or chemotherapy or hormonal therapy decreases the likelihood of systemic recurrence. Furthermore, there is new evidence that axillary radiation may decrease systemic recurrence rates in pre-menopausal patients with two or more positive lymph nodes. ^7,8 This new information contradicts the widely believed paradigm that radiation therapy is effective for control of local disease in breast cancer but has minimal impact on systemic disease or survival rates.

Indications for axillary dissection in breast cancer are controversial. Most believe that axillary dissection is not indicated where the likelihood of metastasis is negligible, as in patients who have ductal carcinoma-in-situ (DCIS) or a small, well-differentiated tubular carcinoma.

On the other hand, there is evidence that micrometastases do have prognostic importance. ^9,10 Twenty to thirty percent of patients with negative axillary nodes eventually develop metastases ¹¹ because of early systemic spread or because undetected metastases did indeed occur to the axillary lymph nodes or involved other chains, such as the internal mammary chain. Greater sectioning and immunohistochemical staining in
sentinel node biopsy (SNB), as shown in some preliminary work done at The Moffitt Cancer Center, may prove to be more effective at detecting clinically significant metastases from invasive breast carcinomas, and may alter practice in that SNB is performed in patients with even minimal disease. The Moffitt Cancer Center also has reported that serial sectioning and immunohistochemical staining of sentinel nodes in DCIS yielded a 5% positive nodal rate. ¹² Because systemic failure is rare with DCIS (less than 1%), SNB and immunohistochemical techniques may, in at least some instances, be too sensitive to be clinically relevant.

Controversy exists for axillary dissection in patients with small invasive cancers which have favorable prognostic features. Cox found that 16% of women with tumors between 0.1 mm and 1 cm in size, and 33% of those with tumors between 1 and 2 cm had positive SNBs. ¹²

Contraindications to SNB should include prior major breast or axillary surgery which could interfere with identifying lymphatic drainage pathways in SNB, palpable axillary nodes, and cases of multifocal breast cancer. ² Finally, sentinel node biopsy is not indicated for those women with clinically palpable axillary nodes and for those patients who, because of comorbidities, will not be candidates for infusional chemotherapy. Many patients with larger, higher grade tumors are undergoing standard axillary dissections at present because high dose chemotherapy with stem cell rescue protocols require documentation of the number (often 4 or more) of positive nodes in the axilla.

Advantages of sentinel node biopsy

Morbidity of axillary node sampling includes axillary paresthesias or numbness associated with damage to the sensory intercostal brachial nerve, seroma, infection, arm lymphedema (found in up to 9% of all patients), and decreased shoulder mobility or pain, and is usually greater than that of resection of the primary breast tumor itself. Additionally, SNB is strictly an outpatient procedure and does not require the use of a postoperative drain.

Detection of axillary metastases is related to the extent of dissection and of pathological analysis. ¹³ SNB may be more sensitive for detecting axillary lymph node metastases because examination of the single sentinel node can result in a more focused and extensive pathological examination. In one study, the only positive node was the sentinel node in 38% of the 85 women with axillary metastases. ²² Because there are many nodes to examine, node analysis in axillary dissection includes only a few slices per node, without immunohistochemical analysis. In SNB, the single lymph node may be subjected
to additional sections and immunohistochemical analysis; the latter is extremely sensitive in detecting micrometastases. Multiple studies have showed an increased yield of 25 to 50% in detecting micrometastases when multiple sections through the lymph nodes were obtained or when immunohistochemical staining was used. ^14-19

Sentinel node identification using a radiocolloid may identify drainage outside the axilla, including drainage to the internal mammary, supraclavicular, and infraclavicular nodes. Lymph drainage of outer quadrant tumors was traditionally believed to be to the axillary nodes, and drainage of inner quadrant tumors was believed to be to the internal mammary nodes. Most series find a 2 to 4% rate of aberrant drainage patterns (figure 2). Identification of unexpected pathways for drainage may lead to revision of traditional node dissections or may lead to adjunct therapy--such as radiation therapy to mediastinal fields. At M.D. Anderson Cancer Center, if a patient's tumor has the appropriate size and grade and if preoperative lymphoscintigraphy shows drainage to the internal mammary chain, the patient receives radiation therapy to mediastinal fields in spite of negative axillary nodes. This is routine in melanoma therapy, where planned operative nodal dissections were changed in 47% because lymphoscintigraphy demonstrated drainage to non-classical nodal basins, drainage to multiple basins, or lack of a predominant drainage basin. ²⁰

Disadvantages of sentinel node biopsy

At the present time, the majority of surgeons performing sentinel node surgery will use a standard axillary dissection if the SNB is positive. However, false-negative results of a frozen section of the SNB is approximately 17%. Thus, some patients will need to return for a second procedure in the situation where the final evaluation of the SNB reveals disease not discovered during the frozen section evaluation. This may no longer be an issue if axillary radiation is automatically performed after a positive SNB.

The greater sensitivity of immunohistochemical and serial sectioning is also a strength of the SNB technique. However, the greater duration of time of analysis requires that the patient return for an axillary dissection at another time. Thus, it may be more beneficial to perform SNB in patients with a low risk of axillary nodal metastases who are, therefore, unlikely to have to return for a second surgery.

Sensitivity of SNB

Several studies have demonstrated a very high sensitivity (97 to 100%) of SNB biopsy. ^21,22 However, note was made in the New England Journal of Medicine that false-negative rates may seem low because so many women are node negative to begin with. ² False-negative rates may be higher if true positives were used as the denominator. For example, if 75 patients are negative and 25 are positive, 3 false negatives using 100 as a denominator gives a F(-) rate of 3%; using 25 as a denominator gives a F(-) rate of 12%. The tolerated F(-) rate may therefore be higher for women with a low risk of axillary lymph nodes but lower in the opposite situation.

Technique: melanoma precedent

The precedent for sentinel lymph node identification began several years ago with its use in melanoma staging. Lymph node staging is important for melanoma because the FDA approved interferon alfa-2b adjuvant therapy for high-risk melanomas and because there is a survival benefit in patients 60 or younger after therapeutic lymphadenectomy.

Investigators have demonstrated an orderly progression of nodal metastases in melanoma rather than a random pattern. ²³ If the first or sentinel node was identified and was negative, disease in other nodes was essentially excluded. Because of the high negative predictive value of the sentinel node in melanoma, patients with a negative sentinel node can now be spared a complete node dissection. ²⁴

Injection of blue dye around a melanoma initially was used in this technique. Later, injection of a radioactive tracer adjacent to the melanoma and identification of the sentinal node using a hand-held gamma probe considerably improved sentinel node identification and decreased operative morbidity as it allowed identification on SNB under the skin and enabled smaller incisions. The use of both blue dye and radiocolloid has enabled identification of the sentinel node in 98 to 99% of patients. ²⁵

Melanoma SNB is relatively easy to perform because of the visibility of the superficially located tumor and because the lymphatic pathways are superficial and easy to identify. In the breast, however, lymphatic drainage is not as great as drainage from the skin and, thus, the flow of tracer is not as rapid. Also, most breast cancers are in the upper outer quadrant and thus may be close to the involved lymph nodes, making it more difficult to separate tracer from tumor.

Technique of breast SNB

Techniques of sentinel node biopsy remain relatively variable as the technology evolves. SNB may be performed after an excisional biopsy or when tracer is injected around the intact tumor. The latter may be more optimal.

Lymphazurin blue dye (5 ml of 1% isosulfan blue vital dye) was first utilized as a carryover from the melanoma experience. Three to 5 ml of 1% isosulfan blue vital dye is injected into the breast parenchyma surrounding the primary tumor or into the wall of the biopsy cavity 10 to 15 minutes prior to surgery. The surgeon utilizes a small, low axillary incision that typically reveals a blue-stained lymphatic channel just beneath the pectoralis fascia and traces the lymphatic proximally and distally to the blue-stained node. Disadvantages to using the dye include blue tattoos in the breast. Additionally, difficulties can arise with visualization of the node, as significant dissection may be needed before visually detecting a blue node. ²²

Preoperative lymphoscintigraphy using such agents as technetium-99m-labelled sulfur colloid, Tc-99m serum albumin, and Tc-99m dextran has been researched. The sulfur colloid tends to be more rapidly and efficiently taken up by the lymphatics and better retained by the sentinel node. ²⁶ However, the particle size needs to be made sufficiently small; this can be achieved by reducing the heating time or by using a micropore filter.

Surgeons using both blue dye and Tc-99m sulfur colloid have reported higher success rates than by either blue dye or radioisotopes alone, and the combination may be beneficial, especially in a surgeon's early experience with the technique. ²⁷ At the present time, most investigators use both blue dye and radioisotopes or radioisotopes alone. ^18,19

If a radioisotope is used, it can be injected one day prior to surgery; however, it is often injected 2 to 4 hours prior to surgery. Injecting the radioisotope into the site 2 to 4 hours prior to surgery may be more convenient for the patient, as repeated visits to the hospital are minimized. The timing of the injection must be coordinated between the nuclear medicine department and the operating room. If the patient is injected a day prior to surgery, he or she simply returns for the localization by the nuclear medicine physician prior to going to the operating room. Loss of radioactivity due to decay from the SN may be a concern in these patients secondary to length of time before imaging; however, early experience with sulfur colloid has not shown this to be true. Sulfur colloid is retained by the lymphatics for a considerable time after the injection.

Injection at four sites into the breast tissue, around the tumor or biopsy site using approximately 1 mCi of radioisotope in 4 to 6 cc of volume was the standard procedure in a multicenter trial in the United States. ²⁸ In Europe, approximately 0.25 cc of 0.2 mCi of smaller particle size radioisotope is injected subdermally. Stereotactic localization is performed for non-palpable lesions. Imaging by a gamma camera usually is done within 15 minutes after the injection and can be performed at short intervals until a sentinel node is identified. After a SN is found, the site is marked on the patient's skin to aid intraoperative localization (figure 3). It is often helpful to obtain oblique views for better visualization and subsequent localization of the SN. If a SN is not seen, gentle massage at the injection site may help. Patients in whom a sentinel node is not found require an axillary dissection. Intraoperative gamma probes are subsequently used by the surgeon to identify the exact location of the SN, which is then dissected and removed. X-rays and electric cautery may interfere with an accurate gamma probe reading. AR