Lymphatic Mapping and Sentinel Lymphadenectomy for Breast Cancer

History of Sentinel Node Concept in Breast Cancer

Surgeons, radiologists, and pathologists have had a long-standing interest in the draining lymphatics of malignant disease. Tumors have unique lymphatic drainage patterns (e.g., Virchow node for gastric cancer, Delphian node for thyroid cancer). The urologist Ramon Cabanas coined the term sentinel node (SN/SLN) as a specific group of lymph nodes associated with the superficial epigastric vein for penile carcinoma. Preoperative lymphangiograms were used to direct the anatomic dissection to resect this specific lymph node center. In his series, a positive lymph node center was identified in 15 patients, and complete inguino-femoro-iliac dissection demonstrated no additional positive nodes in 80% (12 of 15). Penile carcinoma was staged with preoperative lymphangiograms to identify the specific nodal center at risk for metastases, and bilateral biopsy of the sentinel node (SN) center with inguino-femoro-iliac dissection was performed only when tumor cells were discovered in the SN center. On the basis of these results, when the sentinel nodal center was negative, no further nodal dissection was indicated.

The SN concept is founded on the principle that the afferent lymphatic channel draining a primary tumor courses to the first, “sentinel,” lymph node in that specific regional lymphatic basin ( Fig. 33.1 ). Morton and colleagues at the John Wayne Cancer Institute (JWCI) first tested the SN hypothesis in an animal model and then validated it in patients with melanoma.

Investigators next asked if this minimally invasive technique could be applied to staging of breast cancer. Radiographic imaging to stage the axilla had been attempted to identify nodal metastases in breast cancer, but no predictable preoperative technique had emerged with sufficient sensitivity and specificity to replace formal ALND. Interestingly, Kett and Lukacs had reported that the first regional lymph node, the “Sorgius node,” could be identified in breast cancer using direct mammalymphography. Contrast material, Lipiodol Ultra-Fluide, was injected into a lymphatic collecting channel that had been visualized after intradermal injection of patent blue violet into the areola. Benign lymph nodes could be distinguished from malignant nodes by the pattern of distribution of contrast material in the node. This technique did not attain general acceptance for diagnosing malignant nodes in breast disease. It was labor intensive, time consuming, and required serial roentgenography preoperatively for at least 24 hours. In addition, an ALND had to be performed regardless of the preoperative imaging studies, to isolate the suspected lymph nodes. Postoperative roentgenograms of the excised axillary nodes were necessary to distinguish nodes that contained metastatic deposits from those that did not. Simplification of the technique with indirect lymphography was attempted but still required multiple projections of the axilla taken at regular intervals, followed by surgery 24 to 48 hours later. The correlation of pathologically positive nodes with radiographically opacified nodes was poor, and the technique was abandoned.

Armed with this historical information and the success of intraoperative lymphatic mapping for melanoma, three technical approaches for SN identification in breast cancer evolved: dye-directed lymphatic mapping, isotope-based radiolocalization, and the combination of vital dye and isotope techniques.

Evolution of Dye-Directed Sentinel Lymphadenectomy for Breast Cancer

In October 1991 the feasibility of lymphatic mapping and sentinel lymphadenectomy in breast cancer was investigated in the hope of developing a more accurate staging procedure with lower morbidity than ALND. Several hypotheses for SLND in breast cancer were proposed and tested in a prospective manner. The discoveries made by answering each hypothesis have brought us to the current acceptance of observation of the axilla in SN-negative patients. Although these studies established a technique and showed the feasibility of SLND, new questions emerged from these initial studies that were the basis of national clinical trials and rigorous scientific research (see later discussion).

In the absence of any prior experience with intraoperative lymphatic mapping for breast cancer, a learning period was required to define the technical aspects of the procedure. Several factors were identified that determined its ultimate success. These included patient selection, injection method, dissection technique, and histopathologic evaluation of the SN.

Complete Nonsentinel Node Staging and Proof of Principle

The SN hypothesis was not immediately accepted but challenged as an artifact of increased SN histopathologic processing and not a reflection of the status of the axilla. This challenge led to the initiation of the pivotal study of lymphatic mapping in breast cancer and proof of principle that the SN is the first regional node draining the primary tumor.

Complete histopathologic evaluation of SNs and non-SNs was performed with serial sectioning, H&E, and IHC in all cases of H&E-negative SNs. H&E staining identified a tumor-bearing SN in 32% of patients, and IHC upstaged an additional 14.3% ( Table 33.4 ). In 60 patients whose SNs were negative by H&E and IHC, 1087 non-SNs were examined at two levels by IHC. Only one additional tumor-positive node was identified. The SN was sufficient for staging and therapy in 57.3% of patients because they had no additional nodal disease ( Table 33.5 ).

This study provided evidence that the SN is the first draining lymph node from the primary tumor and is not an artifact of histopathologic evaluation. The SN hypothesis was validated and has been confirmed by independent investigators.

Prospective Study of Sentinel Lymphadenectomy alone for a Tumor-Free Sentinel Node

Standardization of the SLND technique, validation with completion ALND, and complete histopathologic assessment of the SN and non-SNs confirmed the accuracy and prognostic value of SLND. The natural segue in the incorporation of SLND was the study to determine whether ALND could be avoided when the SN was negative. The purpose of the next study was to determine the complication rate and the local recurrence rate in women who had a tumor-free SN who did not undergo ALND. This was the first study to eliminate ALND in SN-negative patients.

Women with tumors of 4 cm or less underwent SLND for staging as the only axillary treatment if the SN contained no tumor cells. Completion ALND was performed when the SN contained metastatic cells. Patients did not receive axillary irradiation. SLND identification rate was 99% ( Table 33.6 ). Complications occurred in 20 patients (35%) undergoing ALND after SLND, but in only two patients (3%) undergoing SLND alone ( P = 0.001). The complications in both groups were minimal, primarily wound seromas. There were no locoregional recurrences at 39 months. Subsequent studies support these findings. The axillary recurrence rates are under 1% in the majority of large studies.

SLND can eliminate the need for ALND in patients who have negative SNs, because removal of negative axillary lymph nodes does not alter disease outcome. The role of SLND alone for SN-negative breast cancer patients has been investigated in large national and international trials (see later discussion).

Identification of the Sentinel Node in Breast Cancer by Radiolocalization

During the 1970s and 1980s, radiocolloid lymphoscintigraphy was shown to be a versatile and simpler technique than mammalymphography, but the prognostic value of axillary lymphoscintigraphy alone was insufficient to avoid axillary dissection for staging. In the early 1990s, experimentation with radiolocalization of the SN on the skin surface that could direct surgical removal with a gamma detection probe was initiated in an animal model. Gamma probe–guided SN identification had some theoretical advantages: the location of the SN could be identified, surgery could be confined to a small area in a nodal basin, the SN could be distinguished from non-SNs by quantitative counts, and extraaxillary SNs could be detected.

In 1993 Krag and colleagues reported a pilot study in 22 patients that established intraoperative radiolocalization of an SN in breast cancer. Unfiltered technetium-99m–labeled ( ^99m Tc) human sulfur colloid (0.5 mCi suspended in 0.5 mL saline) was injected in a 180-degree arc oriented toward the axilla, successfully identified a radioactive node in 82%, and reflected the status of the axilla in all these cases.

Identification of the Sentinel Node with Preoperative Lymphoscintigraphy and Intraoperative Radioguided Surgery

In 1997 Veronesi and colleagues reported a large series of patients who had subdermal injection of ^99m Tc-human colloidal albumin, preoperative lymphoscintigraphy, and intraoperative gamma probe detection of the SN. The SN was identified in 98% of cases and accurately predicted the status of the axilla in 97.5%. In the large majority of early-stage breast cancer patients, the combination of preoperative lymphoscintigraphy with intraoperative radiolocalization could locate an SN in the axilla with a high degree of accuracy.

Combined Technique of Vital Dye and Radioisotope

The group from the Moffitt Cancer Center adapted the discoveries made from blue dye lymphatic mapping and isotope radiolocalization and reported on the combined technique in 62 women. The authors predicted that the ability to identify the SN by combined technique would be greater. In fact, in their hands, the SN was localized successfully in 92% of patients. A hot spot could be identified on the skin of the axilla and therefore limit the axillary exploration, minimizing tissue disruption. Radiocolloid identified more SNs, but in no case did a “hot” nonblue node contain micrometastatic disease. No skip metastases were identified in this small series.

Global Experience With Sentinel Lymphadenectomy in Breast Cancer

Following these lead studies of blue dye alone, radioisotope alone with or without preoperative lymphoscintigraphy, or combined dye and nuclide, investigators from many large academic centers and smaller community hospitals in the United States and throughout the world rapidly experimented with variations on the techniques. These studies demonstrate a high accuracy rate and a low false-negative rate ( Table 33.7 ). The value of the technique is inherent to the successful identification of the correct SN that reflects the status of the axilla.

Multicenter Lymphatic Mapping Trials

In 1995 the first multicenter trial was initiated to evaluate the ability of surgeons with varied experience in SLND to identify and to accurately stage the axilla using the radioisotope technique. Eleven surgeons entered 443 patients into the study. The SN was identified using a gamma detection probe. A completion level I and II ALND was performed in all cases. A level III dissection was performed for suspicious nodes. Failure to identify a hot spot before excision of the SN was associated with a previous excisional biopsy, age 50 years or older, and medial tumor location. The SN was identified in 88.6% of cases. The accuracy was 96.8%, the negative predictive value was 95.7%, and the sensitivity was 89%. The false-negative rate ranged from 0% to 28.6%. All 13 false-negative SNs occurred with medial hemisphere tumors. The SN was located in a region outside the axilla in 8% of cases. Variations in success were attributed to both surgeon experience and patient characteristics.

This study was followed by the 42-center Department of Defense Multicenter Trial of Breast Lymphatic Mapping, which required surgeons to participate in a 2-day course using an intraparenchymal injection of blue dye and radioisotope. Patients were randomized to SLND alone if the SN was negative or SLND followed by ALND. Preoperative lymphoscintigraphy was performed in 84% of cases and demonstrated at least one SN in 66%. The nodal drainage pattern was to the axilla in 99.9%, axilla only in 78%, internal mammary in 14%, internal mammary only in 2%, and other sites in 5%. The success rate for identifying the SN was 92% at the Moffitt Cancer Center, 91.4% at other university centers, and 85.2% in community/regional centers. The SN was positive in 32% of patients, and the only site of metastases in 63%. Skip metastases were identified for a false-negative rate of 4%. There were no axillary recurrences in the SLND-alone group at 16 months. This multicenter trial demonstrated successful mapping, a low skip metastasis rate, and a reliable performance rate for both university and nonuniversity institutions, thus confirming not only the accuracy of SLND but its widespread feasibility.

In another multicenter trial, surgeons from academic institutions and the private sector attended a formal lymphatic mapping course with hands-on experience, in a training laboratory facility first. The surgeons achieved a 90% identification rate and a 4.3% false-negative rate when they had performed more than 30 cases. Multivariate analysis demonstrated the poorest success rate in older patients (≥50 years) and by inexperienced surgeons (≤10 cases). Identification was not affected by type of prior surgery (fine-needle aspiration, core biopsy, excision), filtered or unfiltered ^99m Tc, time from injection to surgery, tumor size, or tumor location. The false-negative rate was worse for central lesions ( P < 0.02).

In one of the largest multicenter studies evaluating the optimal SLND technique, 99 surgeons enrolled in the University of Louisville Breast Cancer Sentinel Lymph Node Study and performed SLND in 806 patients. The surgeons were provided flexibility in choosing the technique. Each case was evaluated for patient and tumor characteristics, identification rate, false-negative rate, and technique used. There was no difference in SN identification for dye alone or dye plus isotope (86% vs. 90%, respectively). The false-negative rate was significantly lower with the dual agents (11.8% vs. 5.8%, P < 0.05). The isotope-dye combination resulted in more SNs removed (2.1 vs. 1.5, P < 0.0001). In this large study, the optimal technique with the lowest false-negative rate was dual-agent SLND. With this technique the false-negative rate is low enough to be considered a suitable alternative to ALND in routine surgical practice.

The prospective Swedish Multicenter Cohort Study in 3354 patients from 26 hospitals treated by 131 surgeons demonstrated a negative SN in 2246 cases, and no further ALND was performed. Peritumoral, subcutaneous, or intracutaneous injection of radioisotope was followed with lymphoscintigraphy. SNs were identified intraoperatively with a handheld gamma detection probe. At 37 months, nodal failure was 1.2%. The overall survival was 91.6%, and the disease-free survival (DFS) was 92.1%.

Clinical, Pathologic, and Technical Aspects of Sentinel Lymphadenectomy