FUNCTIONAL BREAST IMAGING

Breast cancer is the most common invasive malignancy in women in the United States. Every year, more than 180,000 new breast cancer diagnoses are made (1). A rising rate of 3% has been reported in the new cases of breast cancer each year (2). Current statistics show that approximately one in nine women will develop breast cancer during her lifetime (3).

Screening mammography has proven to be an effective method for early breast cancer detection. Its wide availability and extensive use have resulted in earlier diagnosis, and up to 30% reduction in relative risk of dying form breast cancer in women over the age of 50 (4). Yet, screening mammography is associated with some important limitations. Mammography has a specificity of 87-97%, a sensitivity of 78-96%, and a positive predictive value of 15-30% (3). This technique is less reliable for detecting lesions in patients with dense breasts, breast implants, severe dysplastic disease, or for patients who have had prior breast surgery or radiotherapy. In addition, mammography does not always detect non-palpable lesions. Another limitation is that mammography cannot reliably differentiate between benign and malignant lesions. Consequently, many surgical breast biopsies are done on benign lesions which may result in unnecessary morbidity, cost and emotional disturbance. "Functional Breast Imaging" provides reliable complementary imaging techniques in certain clinical settings to aid in the diagnosis of breast cancer. These include scintimammography, PET scanning, breast MRI, and digital mammography.

RADIONUCLEIDE BREAST IMAGING

The use of radiopharmaceuticals in the detection of breast cancer has been documented as early as 1946 (5). Different radiopharmaceuticals have been used for planar or SPECT imaging. The most commonly used radiopharmaceuticals for the detection of breast cancer are the myocardial perfusion imaging agents, such as 201-Thallium, 99m-Sestamibi and 99m-Tetrofosmin. Other commonly used agents include 18F-FDG and 99m-MDP.

201-Thallium

Although 201-Thallium has demonstrated a high diagnostic accuracy in the detection of breast cancer, its use has been limited due to its relatively long half-life (73 hours) and low counting statistics. In addition, the normal uptake of 201-Thallium in the adjacent myocardium, liver and muscles may limit its use for breast cancer localization.

99m Tc-Sestamibi

99m Tc-Sestamibi offers physical properties that are superior to those of 201-thalium, namely a shorter half-life (6 hours) and a significantly higher target to non-target counting statistics. Also, experimental comparative culture cell studies suggest that tumoral uptake of 99m Tc-Sestamibi far exceeds that of 201-thallium (6). Tc99m Tc-sestamibi is sequestered within the cytoplasm and mitochondria of cultured mouse fibroblasts. Research suggests that the degree of MIBI uptake is multifactorial and related to the degree of cellular proliferation and desmoplastic activity, and to a lesser extent to neovascularity and mitochondrial density (7).

Patients are imaged in the prone position, which provides improved separation of the breast tissue from the myocardium and the liver. The prone position also allows evaluation of deep breast tissue from the myocardium, the liver, and the thoracic wall. Usually, a combination of prone and supine position is usually employed since the supine position is more useful for better localization of primary tumors in the inner quadrants, and also to visualize the axillae and possibly internal mammary lymph node involvement (8).

The sensitivity of 99mTc-sestamibi scintimammography in the detection of primary breast cancer varies between 80% to 90%, with an average of 85%(3). Fig 1.

The sensitivity for palpable abnormalities is significantly higher than that of non-palpable lesions. In addition, the sensitivity for lesions measuring greater than 1 cm is higher than for smaller lesions. The specificity of 99mTc-sestamibi scintimammography for malignant breast lesions is 74% (9). Fig 2. Fig 3.

Although 99mTc-sestamibi is mostly concentrated in breast cancers, increased uptake of the radiotracer can also be detected in various types of benign breast disease. False positives have been reported with some benign tumors such as juvenile adenomas, fibroadenomas, papilomas, abscesses, local inflammation, or hyperproliferative breast disease (10).

Limitation of Scintimammography

The major limitation of 99m Tc-sestamibi as well as other forms of radionucleide imaging is the lower resolution in imaging small lesions. Most of the studies have been performed in patients with palpable lesions, which generally have a diameter greater than 1 cm and are thus more likely to be detected (9). The sensitivity for T1a and T1b cancers with a size of less than 1 cm is lower. In order to improve the sensitivity of scintimammography, studies have been done using a breast-dedicated high-resolution gamma camera to evaluate high-resolution scintimammography (HRSM). Studies by Scopinaro et al. demonstrate an improved sensitivity of 81.25% with respect to T1a/T1b cancers using HRSM, thus stating the improved sensitivity of in tumors less than 1 cm without apparently reducing its specificity (11).

Having performed over 2000 breast scintimammograms at Harbor-UCLA Medical Center, we have found potential applications in the following clinical settings:

Patients with dense breast tissue on mammography. It is estimated that approximately 25% of women have dense breast on mammography (12). This is responsible for a lower sensitivity for mammography in patients 35-49 years old, since this subgroup tends to have denser breast tissue. 99mTc-sestamibi is independent of the breast density, and therefore associated with an improved sensitivity in women with dense breasts on mammography. In a study performed at Harbor-UCLA Medical center, scintimammograms were obtained in 80 patients who had grade 3 or 4 glandular density on mammograms and a palpable breast mass. Excisional biopsy or FNA biopsy was obtained in 68 lesions on 67 patients. Scintimammography resulted in a sensitivity of 95.6%, specificity of 91.1%, positive predictive value of 84.6%, and negative predictive value of 80% (13). The authors therefore demonstrated the higher sensitivity of scintimammography over mammography in patients with dense breasts.

Patients with architectural distortion of the breast from previous surgery, radiation therapy, chemotherapy or biopsy. Scarring due to these iatrogenic treatments make mammographic evaluation more difficult. Scintimammography can be more specific than mammography since it is not affected by disturbances in tissue attenuation in breasts with such architectural distortions. Scintimammography can be more specific than mammography in determining the presence of recurrent disease in these circumstances.

Patients with breast implants. Unlike mammography, scintimammography is not affected by the density of the breast implants. The implant is seen as a photopenic defect.

Evaluation of non-palpable breast lesions. Tolmos et al. have demonstrated that scintimammography may complement mammography by improving the low positive predictive value in the detection of early breast cancer in patients with non-palpable breast lesions (14). Scintimammography was performed preoperatively on 70 women with Birads category 4 and 5 non-palpable, abnormal mammograms. Prior to excisional biopsy, these patients underwent scintimammography. This group demonstrated a sensitivity, specificity, PPV and NPV of 56%, 87%, 38% and 93%, respectively (14). The sensitivity of scintimammography in this group of patients did not show a significant improvement over conventional mammography, and therefore more studies with dedicated breast imager are required to improve scintimammography’s sensitivity for non-palpable lesions.

Assessment of multifocal cancer of the breast. In those patients with normal or nondiagnostic mammography, scintimammography can be done to evaluate for the presence of multifocal disease. This issue has a major clinical significance since the treatment of multifocal breast cancer is mastectomy. Derebek et. al. (15) demonstrated in a case of a 54 year old woman with fibrocystic breast disease, and a single lesion in the right breast which was studies with scintimammography using 99mTc-sestamibi, evidence of multiple sites of multifocal infiltrating lobular carcinoma as well as infiltrating ductal carcinoma. In this case, MRI of the breast failed to demonstrate multifocality of this lesion.

Evaluation of tumor response to chemotherapy. Tc99m MIBI can provide functional imaging for the evaluation of the susceptibility of a breast cancer to chemotherapeutic agents, as well as predict the response to chemotherapeutic agents (16,17).

Evaluation of metastatic axillary lymph nodes. A sensitivity of 67% and a specificity of 80% have been reported with planar scintimammography in detecting axillary lymph node metastasis (18). Obviously, the clinical value of scintimammography for the detection of lymph node metastasis is highly questionable since more than 20% of patients with metastatic involvement would be missed.

18-FDG PET AND SPECT IMAGING

Currently, several applications have been evaluated in the use of FDG imaging in breast cancer patients. In the detection of primary breast tumors, in general, most lesions of sufficient size (>1 cm) can be detected by dedicated, PET imager, whereas FDG-SPECT accurately detects lesions greater than 2.4cm (19). The strength of PET imaging is in its relative specificity of radiotracer uptake in the target tissue as compared with background. In general, lesions with high FDG activity are most likely to be cancer. However, due to the small number of benign lesions that have been studied, the specificity and negative predictive value of FDG imaging in primary breast lesions has not been fully determined in an unbiased, prospective population. However, performing an FDG PET may still be valuable in patients with dense breasts or with silicone breast implants as well as monitoring response to neo-adjuvant chemotherapy.

Another potential application of FDG PET is in the detection of axillary node metastasis in patients with primary breast carcinoma. Various studies have shown a relatively high sensitivity and specificity using FDG PET (19). The first study to show that FDG has a high accumulation in lymph nodes was performed in animals by Wahl et al (20) who also demonstrated promising preliminary results in a study that detected 8/9 patients with axillary metastases.

In addition, detection of residual or metastatic disease after radiation or chemotherapy can be achieved with PET imaging. Whole body PET can detect metastatic breast cancer outside of the breast and axillary nodal basins, and thus aid in restaging and monitoring of these patients (21,22). Similar to scintimammography, PET can also monitor the effectiveness or ineffectiveness of various chemotherapy regimens in the research or clinical settings. Metabolic changes in breast cancer can be detected after 8 days of initiating treatment without any appreciable change in the anatomical tumor size (23). The true efficacy of PET breast imaging, however requires further investigation and demonstrating the cost-effectiveness of this technology.

DYNAMIC ENHANCED MRI

Breast MR imaging is extremely sensitive in the detection of breast cancer and has numerous advantages over diagnostic mammography, including improved anatomic detail, imaging in multiple planes, and the ability to image the chest wall. These advantages are especially important when assessing posterior breast lesions that may involve the underlying muscle. Because of superior anatomic representation, MR imaging is able to allow differentiation of the various muscles of the chest and can be used to assess which muscles of the chest wall are involved (24). This distinction is important for both staging and treatment planning. In addition, in women with palpable cancers obscured by dense tissue at mammography, MR was useful in mapping the extent of local disease (25,26,27).

Fischer et al (28) compared different imaging modalities for the breast, and report that MR imaging has the highest sensitivity in the detection of invasive breast cancer. The rates reported range form 94% to 99%. MR imaging can be a valuable preoperative tool as it may reveal unsuspected multifocal, multicentric, or contralateral breast carcinoma (28,29). MR imaging may therefore aid in preoperative planning and result in therapy changes. MR imaging of the breast is a helpful adjunct in the evaluation of selected cases in which mammographic findings were inconclusive for determining the need for biopsy or in those cases in which biopsy was desired but could not be performed because of difficulty in targeting the lesion. In these patients, a positive MR examination prompted the recommendation for biopsy and allowed for the timely detection of malignancies that might not otherwise have been diagnosed. In addition, a negative MR examination allowed treatment by surveillance mammography rather that biopsy for lesions that were thought to be benign (30).

The combination of contrast-enhanced dynamic and postcontrast-enhanced MR imaging provides accurate data for the diagnosis of malignant breast cancer. Four sequences have been used in the studies done on breast imaging (31). Fast spin-echo T2-weighted images, spin-echo T1-weighted images (SE-T1WI), dynamic MR imaging with acquisition of dynamically enhanced, fat-saturated fast multiplanar spiled gradient-echo (2D-FMPSPGR) imaging with gadopentetate dimeglumine, and fat suppressed postcontrast-enhanced images (CE-T1WI). The majority of benign lesions (83%) and malignant lesions (93%) are hypointense on SE-T1WI. Other lesions are isointense relative to the normal mammary tissue on both SE-T1WI and FSE-T2WI. The majority of benign tumors show a smooth tumor border and homogeneous enhancement, while malignant masses have an irregular tumor border and heterogeneous enhancement. In some cases, ring-like enhancement and spiculated appearance may be seen in malignant breast masses. The peak enhancement occurs earlier in the malignant lesions (mean 197 seconds) than in the benign lesions (mean 369 seconds) (31) and the washout of contrast material is more evident in the malignant lesions. Therefore these four parameters on dynamic MR images, namely the appearance of the tumor border followed by the washout ratio, peak time of contrast appearance, and internal architecture of the lesion, can correctly characterize the majority of breast lesions.

Leong et al. (32) have reported that adding three-dimensional spectral-spatial excitation magnetization transfer (3DSSMT) to a dynamic breast imaging protocol adds significant value to the breast MRI. 3DSSMT is a new pulse sequence with a water-selective excitation and on-resonance magnetization transfer pulse that is designed to optimize visualization of lesion morphology when performed after dynamic scanning. Water-selective (i.e., fat-nulled) imaging is used because of its ability to improve visualization of the margins of the lesions. Magnetization transfer is

used because of its ability to suppress signal from normal fibroglandular tissue and improve the contrast between enhancing breast lesions and background breast parenchyma. They report that when performed immediately following a dynamic breast MRI exam, this protocol adds significant value since it provides a way to characterize lesions not just by their enhancement kinetics on dynamic imaging, but by their morphology and enhancement pattern as well (32,33). Further, this group found that the presence of either skin thickening, or a combination of a spiculated or microlobulated border, with a rim, ductal, linear, or clumped enhancement pattern is 94% specific and 54% sensitive for malignancy. Conversely, the presence of either a perfectly smooth border, a well-defined margin, non-enhancing internal septations, or a microlobulated border was 97% specific and 35% sensitive for a benign diagnosis.

One possible problem with MR imaging includes the differentiation of possible malignancy from scar tissue. Scar tissue and postoperative changes can enhance for up to 6 months, and performing MR imaging before that amount of time has elapsed may not be helpful in differentiating scar from tumor. In addition, because of diffuse or patchy enhancement of breast tissue after radiation, it has been reported that MR imaging is not useful in the first 9 months after treatment and that enhancement of the lumpectomy site without the presence of tumor can persist up to 18 months (30).

DIGITAL MAMMOGRAPHY AND COMPUTER-ASSISTED MAMMOGRAPHIC ANALYSIS

Unlike traditional film-screen mammography, where the film serves as an image acquisition detector, a storage mechanism and a display device, each of these functions is performed by a separate component in digital mammography. Digital mammography therefore has several theoretical advantages over film-screen mammography (34). These include a greater image latitude, as well as greater contrast of structures such as masses and microcalcifications that are early signs of breast cancer. This potential advantage of digital mammography may be most useful in dense fibroglandular breasts. Also, with digital mammography, random fluctuation of image density (noise) may be reduced since film granularity is eliminated. X-ray dose to the breast tissue may be lower since the X-ray detection efficiency of digital mammography systems may be higher than that of film-screen mammography. In addition, digital mammography would more easily lend itself to the performance of several new functions and techniques, such as telemammography, computer-aided diagnosis, dual-energy mammography, post-acquisition image enhancement, and image archival and retrieval (35,36,37,38).

Computer-Aided Diagnosis (CAD) can be defined as the diagnosis made by a radiologist considering the computer results as a ‘second opinion’ on lesion characterization and diagnostic decision-making; it is thus a complement to the radiologist activity and does not replace it. As a rule, CAD needs image digitalization to permit computer database handling; image analysis (segmentation of the regions of interest); then characteristics of specific lesions. This double reading improves the detection rate (39). CAD applied to breast lesions demonstrates higher sensitivity (100 versus 89%) and higher positive predictive value (71% vs. 56%) than when read by the radiologist alone (36).

SENTINEL NODE LOCALIZATION IN BREAST CANCER

The age-adjusted breast cancer death rate in the United States dropped by 6.8% between 1989 and 1993 (40) This improvement has been documented to be continuing (41). This decline is likely explained in part by earlier stage of breast cancer at time of diagnosis A meta-analysis of randomized mammographic screening studies shows a 33% decrease in breast cancer mortality for women between 50 and 69 years of age (42). Also, the increasing popularity and technical improvements in mammography in recent years account for the increase in proportion of patients presenting with more curable stage 0 and stage 1 disease. This increase has been reported from 42.5% in 1985 to 56.2% in 1995 (43,44). Improvements in adjuvant treatment have also been shown to increase 10-year relative survival rates (45).

Axillary node metastasis is the strongest known prognostic variable for predicting both survival and recurrence in patients with early breast cancer (46). Knowledge of the tumor status of regional lymph nodes has become cornerstone information on which both prognosis and therapy of the tumors is based. Both recurrence and survival rates are related to the number of axillary nodes involved (47-49). Also, tumor size correlates with the incidence of axillary involvement as well as with the absolute number of involved axillary nodes (49,50). These features indicate that axillary nodal spread is a consistent, although not invariant, feature of the disease and that axillary involvement is often a sign of systemic disease.

In the absence of other coexistent morbid conditions, almost all node-positive patients now receive postoperative systemic therapy. In node-negative patients, systemic adjuvant therapy may or may not be felt to outweigh the morbidity of treatment. Knowledge of axillary nodal status remains essential for the management of early breast cancer, and level I and II axillary dissections have become standard in the initial staging of disease.

Morton et al. have established that the early lymphatic spread of malignant tumors occurs not just randomly to any nearby nodes, but instead via highly predictable routes (51). The status of regional lymph nodes has revolutionized the staging and management of many cancers, particularly melanoma and breast cancer, and has now become part of routine clinical management of patients with malignant melanoma. Sentinel node techniques are now being widely investigated and adopted into clinical protocols around the world.

HISTORY

Many investigators over the centuries have contributed to the understanding of the progression of tumor cells through the lymphatic system. One of the first reports of the use of lymphoscintigraphy to identify the paths of lymphatic drainage from the breast was published in 1972 by Vendrell-Torne and associates (52) as a study of lymphatic drainage pathways in the intact, normal breast. They established the presence of unexpected, sometimes multiple, pathways of drainage to axillary, internal mammary and supraclavicular nodal basins in a minority of their subjects. Studies done by Morton et al. (51) leading to the sentinel node hypothesis and the selective approach to lymphadenectomy were initiated over 20 years ago. In 1977, they described the use of cutaneous lymphoscintigraphy to identify the lymph basins at risk for metastasis from truncal primary melanomas (53). Around the same time, Cabanas (54) used the term sentinel node in penile cancer to indicate a node detected by lymphangiography.

CONCEPT OF SENTINEL LYMPH NODES AND CURRENT APPLICATIONS OF LYMPHOSCINTIGRAPHY IN BREAST CANCER

The sentinel node concept holds that the lymphatic effluent of a tumor drains initially to one or two lymph nodes before other nodes receive the tumoral drainage. Hence the sentinel node develops lymphatic metastases before other nodes. If the sentinel node is negative for tumor, then other nodes are not likely to contain metastases, and the patient can be spared the unnecessary morbidity (i.e., seroma, hematoma, and less common but more serious problems such as neural or vascular injury or lymphedema of the arm) and expense of a more extensive node dissection.

Moreover, in a series of patients evaluated at the John Wayne Cancer Institute, 45% of positive sentinel nodes demonstrated only micro metastases (<2mm in size) (55). With more careful scrutiny of nodes, micro metastases can be discovered that would remain undetected otherwise (55-57). Such focused scrutiny includes the examination of more sections per node as well as the use of special stains, e,g,, immunohistochemistry using antibodies to cytokeratin proteins. There is growing evidence indicating a decrease in disease-free survival in patients harboring micro metastases compared with patients with tumor-negative lymph nodes (58-62).

LYMPHATIC DRAINAGE OF THE BREAST

Lymphatics within the breast begin as small channels that lie very close to the duct wall (63). These channels coalesce into sequentially larger vessels that generally follow the vascular supply of the breast, namely the external mammary vessels, branches of the thoracoacromial artery, and intercostal branches from the internal mammary artery.

These three principal vascular routes are accompanied, in general, by three respective patterns of lymphatic drainage (64,65): to (1) axillary nodes, (2) juxtaclavicular, supraclavicular, or third echelon axillary nodes , and (3) internal mammary nodes. The pattern of drainage in any one patient is unpredictable, and is difficult to determine from first location of the primary tumor. Drainage patterns can also be altered by surgery, radiation, or by trauma to the breast parenchyma.

TECHNIQUE OF LYMPHOSCINTIGRAPHY

Optimal localization of the sentinel lymph node requires the use of both pre-operative lymphoscintigraphy and intra-operative radiosensitive probes. Intra-operative lymphatic mapping of the sentinel node is facilitated by vital blue dyes which provide a visual cue for identification of the sentinel node(s). Pre-operative lymphoscintigraphy can identify the location of the sentinel node, and intra-operative mapping with the gamma probe can provide an auditory signal that complements the visual guide provided by the blue dye.

Although no one radiopharmaceutical can satisfy all the requirements of an ideal radiopharmaceutical, several radiopharmaceuticals have been used successfully. Identification of the node should be accomplished quickly, requiring a tracer with rapid migration, yet retention within the sentinel node should be prolonged. The agent should be retained both by normal nodes as well as by adjacent nodes that are minimally or extensively involved with the tumor. Further, washout from the site of injection should be rapid to reduce local radiation doses, but the tracer should migrate no further than the sentinel node.

Animal studies have shown that the particle size of a radiopharmaceutical agent is a critical factor in determining the migration rate from the injection site and the rate of uptake in lymph nodes. The particles should be larger than 0.005 nm in size because smaller particles may penetrate or leak into the capillary membranes, and, therefore, become unavailable to migrate through the lymphatic channel (66). The particles move through the lymphatic system by rhythmical contractions and relaxations of the smooth muscles that surround the lymph vessels (67).

Au-198 colloid was the first agent that was widely used but it was rapidly replaced by other radionuclides and radiopharmaceuticals. The agents that are commonly used are Tc-99m antimony trisulfide colloid, Tc-99m nanocolloid, and Tc-99m sulfur colloid. In Europe, the predominant agent is Tc-99m nanocolloid (Eshima, 00). In the United States today, the only FDA-approved agent for lymphoscintigraphy is Tc-99m sulfur colloid (68).

Tc-99m Sulfur colloid, as prepared for imaging the reticuloendothelial system, features large particle sizes (100 nm to 1,000 nm) (69), and is considered by many to be a sub optimal radiopharmaceutical for lymphatic studies. The particle size of this preparation may be reduced by the use of shortened boiling time during preparation of the radiopharmaceuticals (70). Commonly, the particle size in the injectate is also reduced by filtration of the whole agent through a 100 or 200 nm Millipore filter prior to use for lymphatic studies (69,71).

Several different injection techniques for breast lymphoscintigraphy have been reported. Injection volumes vary, ranging from less than 1 ml (72) to 10 ml (73). Locations of injections also vary, from intratumoral (74), to peritumoral (56,75,72) to subdermal overlying the tumor (75). Intratumoral injections carry the potential of spreading tumor cells and of poor migration of tracer after injection into a solid tumor mass. Placement of the entire injectate in one location can result in local pooling with poor migration. The injectate should generate sufficient interstitial pressure to induce a lymphatic migration (76).

If the tumor is palpable, 6 peripheral injections are made at the depth of the tumor. With non-palpable lesions that have been only mammographically identified, injection is made circumferentially around the lesion, approximately 1 cm from the localization wire (77). The use of ultrasound is advocated during all injection procedures, but especially when the tumor (or seroma from previous excisional biopsy) is non-palpable (77). Ultrasound allows a precise injection at the correct depth and through the shortest distance of tissue. Ultrasound also minimizes the chances of creating a pneumothorax.

Much has been published about the benefit of performing breast massage, three to five minutes following the injection of the radiopharmaceutical (77,78). Although imaging can be done immediately , usually the lymph nodes will be more apparent following a delay as this allows for the radiopharmaceutical to traverse through the afferent Lymphatics (77). Therefore, if a SLN is not identified with immediate imaging a two to four hour delay should be used.

Imaging should begin within a few minutes after injection. Some patients demonstrate migration almost immediately after injection. Waiting several hours after injection before imaging may cause confusion with regard to identification of the true sentinel node (76). Positioning of the patient is important for successful identification of the sentinel node. The axillary nodes are most directly visualized with the patient in a semi recumbent position with the ipsilateral shoulder elevated the arm raised over the patient’s head. This position optimizes separation of the axillary nodes from residual activity at the injection site. The internal mammary nodes however, are best identified with the patient in the supine position. Imaging is performed immediately after injection and subsequently following a 2-4 hour delay, just before transfer to the surgical suite.

As previously mentioned, lymphatic migration of tracer form the injection site can be accelerated by heat, massage, and elevation of venous pressure. Exercise of the ipsilateral arm may hasten migration to the sentinel node after mammary injections.

The various dyes assessed as potential lymphatic mapping agents included methylene blue, isosulfan blue, patent blue-V, 0.05% disodium hydrogen phosphate, and flourescein (79). After general anesthesia or local anesthesia with heavy sedation, 3 to 5 ml of blue dye is injected around the periphery of the tumor or in the wall of the biopsy cavity (79). It is important not to inject the blue dye into the tumor, as this well prevent migration of dye into the Lymphatics.

SENTINEL NODE LOCALIZATION

Once the sentinel node(s) are visualized, their locations are marked on the patient’s skin with the aid of the scintillation camera. The position of skin mark overlying a node is subject to variation relative to angulation of the parallel hole collimator and patient position. Therefore, the location is best confirmed using a hand-held probe, while the patient is in the same position that will be used on the operating table.

Most experienced surgeons who do sentinel node procedures will use both blue dye as well as radioisotopes (80). This will give the benefits of both visual cues from the blue dye and count-rates from the radiosensitive probe. A transverse incision is made in the axilla and an attempt is made to identify a blue lymph node. If the blue lymphatic channel can be identified, it is traced directly to the blue sentinel node and the gamma probe is used to confirm that this node is indeed radioactive, both in vivo and ex vivo. Sentinel lymph nodes are defined as any blue node or any node with greater than or equal to three times in vivo background counts per minute (ex vivo, ten times the counts per minute of a non sentinel node). After initial excision, the gamma probe is returned to the axillary bed. Dissection is continued so long as the lymph node bed has more than 150% of background counts (77).

PATHOLOGIC EXAMINATION OF THE SENTINEL LYMPH NODE

After the surgeon excises the blue-colored and maximally radioactive sentinel nodes, the pathologist critically evaluates these by histology and immunohistochemistry for the presence of a metastatic tumor. If the sentinel node(s) are tumor free, no further nodal dissection in undertaken; if a tumor is present, a complete dissection of the nodal basin is performed (81).

Although frozen sections have been used in the past, currently there has been a general trend to prepare well-fixed permanent sections of the sentinel lymph nodes. In addition to conventional histology, all sentinel nodes must be examined by immunohistology (S-100 protein, HMB-45, or MART1), unless the node contains overt tumor macroscopically or microscopically (81).

DISCUSSION

In a study done by Veronessi et al. (82), the predictive value of the sentinel node or nodes was reported to be 97.5%, thereby concluding that axillary dissection is probably unnecessary for patients in whom the sentinel node is negative. Further, they report that in patients with small tumors (less than 1.5 cm), the status of the sentinel node predicted axillary-node involvement with 100% accuracy. Also, they discuss that multifocal tumors are likely to involve more than one lymphatic trunk from the mammary gland to the axillary nodes, which may give rise to skip metastases. Thus, they advise that the sentinel node method should not be used in cases of extensive multifocality, i.e., in which the edge-to-edge distance between foci is 3 cm or more. Table 1 demonstrates different level of sensitivities in sentinel node identification by different technique(s) as reported by Niewig et al. (83).

Clearly, using all available tools (blue dye, lymphoscintigraphic imaging before surgical incision, and the gamma-detecting probe) have been reported to result in the highest sensitivities for finding a sentinel lymph node.

Although experienced investigators at many institutions have proved that high rates of sentinel node identification are possible, each surgeon must ascend a learning curve to acquire technical expertise in lymphatic mapping and sentinel lymphadenectomy. Studies indicate that successful mapping regional lymphatic anatomy is directly related to the surgeon’s experience; the rate of sentinel node identification is highest in a surgeon’s most recent cases and highest for the surgeon who has performed the most mapping procedures (84,85,86). Regardless of the mapping agent used to identify the sentinel node, during the learning phase the surgeon should always perform complete lymphadenectomy after lymphatic mapping to monitor his/her rate of false-negative sentinel nodes (51).

Likewise, for the nuclear medicine physician, initial experience is needed to overcome the learning curve problems (68). This initial experience should include marking the sentinel nodes on the patient’s skin with the aid of the probe, and following up with the patient to the operating room to work with the surgeon in localizing the sentinel nodes intraoperatively. The counts detected by the probe are recorded and should be included in the nuclear medicine report to document significance of counts in excised nodes, which ideally in a sentinel node are greater than 10 times (ex vivo counts) the background counts (non sentinel node, ex vivo) (87). Lack of standardization of methodologies might also be contributing to the variations reported in finding sentinel nodes.

The primary reason for the current delay in accepting sentinel node staging of early breast cancer is that as yet, no consensus and no standardization of an optimum methodology for sentinel node identification exists. At different centers, in addition to the blue dye versus radiotracer approaches, sentinel lymph node identification using radiotracer is being performed differently. The major variables are the following: (1) the radiopharmaceutical used, (2) injections site, (3) injection volume, (4) injection activity, and (5) use of imaging.

SUMMARY

Sentinel node localization for breast cancer is rapidly gaining popularity. This is a minimally invasive technique, which has proven highly accurate, and beneficial at institutions with increasing experience, resulting is less radical surgery and less morbidity for node-negative patients. Sentinel node techniques should play an increasingly important role in the more refined staging and management of breast cancer in the future.