Identification And Biologic Significance Of Micrometastases In Axillary Lymph Nodes In Patients With Invasive Breast Cancer

The status of the axillary lymph nodes, the most common sites of breast cancer metastases, is the most important factor in determining the prognosis of patients with invasive breast cancer.1,2

The goals of axillary lymph node dissection are to provide accurate staging information and local control of the disease.3,4 However, the procedure has many potential complications, including lymphedema, persistent seroma, shoulder dysfunction, and paresthesias. Fewer breast cancer patients have presented with axillary lymph node involvement during the past 2 decades because of the increased use of breast cancer screening programs and detection of the disease at an earlier stage. Lymphatic mapping with sentinel lymph node (SLN) biopsy for breast cancer was introduced in the early 1990s. The sentinel node is the first node to receive lymphatic drainage from the area of the primary tumor and provides accurate staging of the regional nodes with a significant reduction in surgical morbidity. On the basis of the results of large randomized trials that confirmed the efficacy and reduced morbidity of SLN biopsy, today in many institutions, SLN biopsy has been accepted as the standard of care for axillary evaluation in breast cancer patients with early-stage disease.5-8 Published data suggest that the absence of metastatic tumor cells in the SLNs accurately predicts the absence of metastasis in the remaining axillary nodes in 95% to 100% of cases.9-11 Additionally, SLN biopsy may be more sensitive to detect small metastases than axillary node dissection because the small number of lymph nodes removed during the SLN procedure can be subjected to a more detailed pathologic evaluation.11 Patients without any metastases in the SLNs are unlikely to have metastases in the remaining axillary lymph nodes, and, therefore, completion of the axillary lymph node dissection is considered unnecessary. However, if the SLN contains metastases, further treatment of the axilla is indicated.10-12

On the basis of presence or absence of metastatic carcinoma, lymph nodes can be classified as positive or negative. Although in most cases this is a straightforward diagnosis, in some cases there are only small clusters of tumor cells and establishing a diagnosis of metastasis is not easy. The terms micrometastases and isolated tumor cells (ITCs) have been defined by the American Joint Committee on Cancer (AJCC) in the most recent revision of the breast cancer staging system.13 The clinical significance and management of patients after diagnosis of micrometastases or ITCs in the SLNs have remained a point of controversy in the literature.14-17 Furthermore, there are no uniform histologic criteria to separate ITCs and micrometastases in cases with multiple clusters of cells in SLNs. The lack of uniform criteria and use of different terminology have led to the inconsistent interpretation and classification of SLNs.

In this review we discuss the controversies regarding the definition, identification and clinical significance of micrometastases, and histopathologic evaluation of SLNs.

DEFINITION

Any metastasis not found on gross inspection and palpation but identified on light-microscopic evaluation could be broadly defined as a micrometastasis. Traditionally, micrometastasis is often used to describe any small metastasis, but the definition of small varies from study to study. Originally, Huvos et al18 described micrometastasis as a single metastasis smaller than 2 mm. This definition has been used in the literature for a long time. During the years, improved surgical and pathologic techniques, especially introduction of SLN biopsy and focused and detailed pathologic evaluation, have resulted in a steady increase in identification of smaller and smaller metastases in lymph nodes. The current edition of the AJCC staging system, which was revised in 2002, includes findings in SLNs and use of adjunct tests including immunohistochemical staining and molecular testing in identification of metastases. This revision defines micrometastasis as a metastasis measuring from 0.2 mm to not more than 2.0 mm. Metastases larger than 2 mm are denoted as macrometastases. Tumor clusters measuring up to 0.2 mm are classified as ITCs or submicrometastases. In the TNM staging, N categories were revised to include pN1mi and pN0(i+) to indicate the presence of micrometastases and ITCs, respectively.19 According to the original 6th edition of the AJCC staging system, the identifier (i) indicated detection of carcinoma on immunohistochemical (IHC) staining in lymph nodes that were negative by hematoxylin- eosin (H&E) staining. The designation pN0(i+) was used to indicate the presence of ITCs identified by immunohistochemistry. A clarification of the AJCC staging system was published later, and the identifier (i) indicated presence of ITCs regardless of detection methods.20,21 All tumor clusters smaller than 0.2 mm, whether detected by H&E or immunohistochemistry, are classified as pN0(i+) (Table 1).

Some authors use the terms micrometastasis and occult metastasis interchangeably. Occult metastases are not observed during the initial routine histopathologic evaluation but become apparent at deeper levels of routine histologic sections or on IHC evaluation. Although occult does not necessarily denote a size, occult metastases are usually quite small. Friedman et al22 introduced a similar term, clandestine metastases, to describe small tumor emboli in the sinuses.

DETECTION METHODS

Routine histopathologic evaluation of axillary lymph nodes removed for staging of breast cancer consists of dividing the node into halves, and if the gross evaluation is negative for metastatic disease, submission of 1 representative section or the whole node into a paraffin block, and microscopic examination of a single H&E-stained section of each paraffin block. It has been long recognized that this evaluation method fails to identify all metastatic foci in lymph nodes and several methods have been proposed to increase the detection rate of axillary metastases.23-28 Some investigators have proposed serial subgrossing and histologic step-sectioning of each lymph node. This intensive evaluation has been shown to upstage nodal status in a significant number of patients. Other investigators have suggested using IHC assays to detect metastatic carcinoma cells.28-32 More recently, molecular biology techniques have been used to detect even smaller numbers of metastatic tumor cells.33-36 Unfortunately, in most clinical settings, such detailed histopathologic evaluations and the application of such techniques for the entire axilla are prohibitively expensive and the College of American Pathologists still endorses submission of the entire lymph node into paraffin blocks and examining 1 representative H&E section of each paraffin block. Because the SLN specimen consists of only a few lymph nodes and additional testing such as multiple-level H&E sectioning and IHC assay, this is feasible in most clinical settings.

Consensus meetings and statements from both the College of American Pathologists and the American Society of Clinical Oncology recommend that serial sectioning of each SLN into less than 2-mm intervals is necessary to reliably detect macrometastases.1,37 Cserni38 has suggested that sectioning of the SLNs into 1-mm parts would detect almost all macrometastases. The identification rate of micrometastases depends on the number of sections evaluated from each paraffin block and the interval between each section. There is no question that the examination of more levels at smaller intervals will increase the detection rate of micrometastases and ITCs. The addition of immunohistochemistry for cytokeratin may facilitate identification of micrometastases and ITCs. Immunohistochemistry is especially helpful when tumor cells do not form cohesive clusters in lymph nodes. This is a typical pattern of metastatic lobular carcinoma, and IHC staining high lights carcinoma cells that may be difficult to see on H&E sections. Although IHC staining for cytokeratin is not routinely advocated by any consensus recommendations, it is commonly used in routine practice in the United States and in many European countries.39-42 Sentinel lymph node studies using immunohistochemistry have lower falsenegative identification rates compared with studies that did not use immunohistochemistry. A number of different polyclonal and monoclonal antibodies against epithelial antigens (cytokeratins and other proteins such as epithelial membrane antigen) have been used. Metastases have been reported in 9% to 30% of breast cancer patients whose lymph nodes were believed to be negative on initial examination of H&E sections.43-45 The percentage is generally higher in studies in which immunohistochemistrywas performed on deeper levels instead of on destained original H&E slides. These reports of higher detection rates on immunohistochemistry performed on additional levels further illustrate the significance of serial sectioning. Unfortunately, there is no breast carcinoma specific marker; therefore, IHC staining for epithelial markers demonstrates the presence of epithelial elements. Benign glandular inclusions or benign transport of epithelial clusters in lymph nodes will have IHC positivity. The differential diagnosis of benign epithelial inclusions and metastatic carcinoma should be made based on morphologic features of the epithelial clusters and the primary tumor.

During the last decade, some innovative molecular diagnostic techniques have been used to increase the likelihood of detecting breast cancer metastasis. In reverse transcription- polymerase chain reaction (RT-PCR) messenger RNA related to breast carcinoma cells can be detected.46-50 RT-PCR can detect 1 metastatic cell among 106 normal lymphoid cells. The markers used for this purpose include cytokeratins, MUC-1, mammaglobin, and carcinoembryonic antigen. Depending on the markers used, 15% to 40% of lymph nodes believed to be negative on light-microscopic evaluation of H&E and IHC stains showed positive results with RT-PCR.48-50 Because RT-PCR is so sensitive, false positivity due to contamination from normal breast cells is a serious concern, and the specimen must be handled with special care. Cross-reactivity with homologous genes can also cause false-positive results. The biologic and clinical significance of RT-PCR-positive and light microscopy-negative lymph nodes is not known. Clearly, additional studies are needed to refine these techniques and determine the clinical significance of their results.

CLINICAL SIGNIFICANCE

Although patients with node-negative breast cancer have an excellent prognosis, up to 25% to 30% of these patients will develop local recurrences or distant metastases within 10 years.51,52 Studies have suggested that this unfavorable outcome may be due in part to undetected occult metastases in the lymph nodes.53-58

The prognostic significance of occult micrometastases in axillary lymph nodes has been a controversial topic for nearly half a century. Dowlatshahi et al59 reviewed all reported series evaluating the prognostic significance of occult metastases published from 1948 to 1996. This comprehensive review demonstrated that initial studies did not show survival differences in patients with occult metastases, whereas most studies published after the mid 1980s have consistently shown a prognostic disadvantage for patients with micrometastases. The authors have postulated that the lack of prognostic significance in earlier studies can be partially explained by relatively small patient population size, short length of follow-up, and different techniques that were used to evaluate lymph nodes. The larger studies with longer follow-ups tend to show more significant differences. However, there were some other published studies that included long follow-up and large populations that still failed to show survival differences among patients with and without occult metastases. 25 In a more recent study Kahn et al54 used the AJCC staging system to assess the prognostic significance of occult micrometastases in breast cancer patients and concluded that occult micrometastases are not associated with poor prognosis. Sakorafas et al60 published a meta-analysis of reported series between 1966 and 2003 and concluded that micrometastases are associated with poorer outcome and the presence of micrometastases should be taken into consideration for patient management.

A large study including 42 197 patients from the Surveillance, Epidemiology, and End Results database showed that patients with micrometastases in single or multiple lymph nodes had a slightly elevated risk of death due to breast cancer compared with patients with nodenegative breast cancer.61 Another study using the Surveillance, Epidemiology, and End Results database including more than 200 000 patients demonstrated that patients with micrometastases had a prognosis intermediate to patients with macrometastases and patients with node-negative disease.62 The tumor burden of the axillary lymph nodes appears to be inversely correlated with clinical outcome. In a retrospective review of 10 000 patients Kuijt et al63 reported that patients with solitary micrometastases had poorer clinical outcome if they did not receive adjuvant systemic therapy. In some studies subset analyses showed significant differences among patients with occult lymph node metastases. Cote et al58 reported significantly poorer disease-free survival and overall survival in postmenopausal, but not premenopausal, breast cancer patients with occult micrometastases identified by either deeper H&E sections or immunohistochemistry. Trojani et al56 and de Mascarel et al57 emphasized that the length of follow-up to assess the significance of occult micrometastases is very important. In many of these studies occult micrometastases were defined differently and the detection methods varied significantly.

In addition to these studies based on evaluation of complete axillary lymph node dissection specimens, there are several studies that evaluated prognostic significance of micrometastases in SLNs. Most published SLN studies have evaluated incidence of further metastases in axillary lymph nodes. Cserni et al64 reported a meta-analysis and found that the incidence of non-SLN metastases after detection of low-volume SLN involvement was 20% and if SLN involvement was identified by IHC examination the incidence was 9%. Viale et al65 studied the risk of additional axillary metastases in patients with positive SLNs in 1228 patients. Micrometastases and ITCs were identified in 26% and 9.5% of patients, respectively. Non-SLN metastases were identified in 15% of patients whose SLNs showed ITCs and 21% of patients whose SLNs had micrometastases. When SLNs were further stratified according to the size of the micrometastasis (1 mm vs 1-2 mm), the prevalence of non-SLN involvement was identified in 17% and 30%, respectively. The authors suggested that patients with a positive SLN could be stratified in 3 groups at significantly different risk for metastases to non-SLNs. Patients with ITCs only or micrometastases up to 1 mm had a significantly lower risk of additional metastases, compared with those with micrometastases 1 to 2 mm in size. Czerni et al66 reported SLN findings of 2929 patients with small (up to 15 mm) invasive carcinomas. The authors evaluated clinical and histopathologic variables associated with SLN and non-SLN involvement. Factors that influenced SLN involvement included size, multifocality, and grade of the primary tumor and age of the patient. Non- SLN involvement was associated with tumor grade, ratio of involved SLNs, and SLN involvement pattern (ITCs vs micrometastasis vs macrometastasis). Further axillary lymph node involvement was identified in 37% and 12% of patients with macrometastases and micrometastases in the SLNs, respectively. Less than half of the patients with SLNs showing ITCs had a completion axillary node dissection and none showed further disease in the non-SLNs. van Deurzen et al67 studied histopathologic findings of SLNs in relation to presence of metastases in second echelon lymph nodes. Frequencies of second echelon lymph node metastases were reported to be 13%, 20%, and 48% for patients who had ITCs, micrometastases, and macrometastases, respectively. All patients who had ITCs in SLNs and positive second echelon lymph nodes had only one additional lymph node with metastatic cells in second echelon lymph nodes.

Investigators have evaluated histopathologic parameters of primary tumors in search of reliable predictors of non- SLN metastases. In a meta-analysis Degnim et al68 reported 5 parameters with the highest likelihood of predicting non-SLN metastasis. These factors were SLN metastasis greater than 2 mm, extranodal extension in the SLNs, primary tumor size greater than 2 cm, more than 1 positive SLN, and lymphovascular invasion in the primary tumor. In addition to the number of positive SLNs, the proportion of positive SLNs has been found to correlate with non-SLN positivity in many studies.

Prognostic significance of microanatomical localization (whether they are subcapsular, parenchymal, or both) of occult metastatic deposits in axillary lymph nodes have been evaluated in the literature and results are conflicting. Some authors found no significant association between survival difference and localization of occult metastases in lymph nodes.22,32,45 In contrast, Reed et al30 reported subcapsular localization being predictive of unfavorable outcome. In addition to these historical studies, microanatomical localization of metastases has been evaluated in SLNs in relation to non-SLN involvement. Parenchymal localization and larger tumor penetration into the parenchyma have been found to be highly predictive for non-SLN involvement. 69-71

Although it is controversial in the literature, most studies suggest that extranodal tumor extension in axillary lymph nodes has prognostic value to predict local recurrence and overall survival. Goyal et al72 suggested that extranodal tumor extension in SLNs is a marker of lymphatic obstruction and is significantly associated with reduced radioisotope uptake. The extent of extranodal tumor extension increases with increasing size of metastases in SLNs. Most studies have found extranodal tumor extension in SLNs to be a strong predictor of non-SLN metastases. Recently, Cserni73 has described a distinctive pattern of SLN involvement with predominantly capsular and extranodal involvement and minimal nodal parenchymal involvement. In this report of 8 cases, micrometastases with extranodal tumor extension were seen predominantly in patients with well-differentiated carcinomas and tubular carcinomas. The biologic significance of this pattern of extranodal tumor involvement is currently unknown.

S-classification, which was originally introduced for staging of SLNs in patients with malignant melanoma,74 has been applied to breast cancer SLNs to predict the status of non-SLNs.75 The S-classification measures the maximum depth of tumor cells into lymph node parenchyma from the lymph node capsule. Although results of the initial studies are promising, they have to be confirmed in larger studies.

Investigators have developed nomograms to predict non-SLN lymph node involvement using multiple histopathologic features of the SLNs and primary tumor.76Multiple studies have evaluated the clinical utility and applicability of these nomograms. Although most studies in general show that these nomograms have good overall accuracy, they have limited usefulness in the clinically most important subgroup of patients-patients who have positive SLNs and a very low risk of non-SLN metastasis such that the completion of axillary lymph node dissection can be omitted.76-83 Therefore, in current practice these nomograms have limited value clinically.

In many studies guidelines and definitions of micrometastases have not been clarified. It is likely that different histopathologic criteria have been used for pathologic classification of small-volume disease in these published studies and outcome differences may partially be explained by variation in histopathologic categorization.84,85

There are 2 main challenges in histopathologic categorization of small-volume disease in lymph nodes. The first issue is microanatomic location of tumor cells in lymph nodes. When small clusters of tumor cells are present in the nodal parenchyma, most pathologists agree that these clusters should be called either micrometastases or ITCs based on the size of the tumor clusters (Figure 1). However, there are no histopathologic guidelines to classify tumor clusters when they are identified in the subcapsular sinus, in the capsule, or in perinodal tissues (Figure 2). Different pathologists report these cases differently. The second issue is measurement of tumor cells or clusters when they show a dispersed pattern in nodal sections (Figure 3). There are no clear guidelines on how to report tumor burden on these lymph nodes. Czerni et al86 conducted a reproducibility study to assess the pathology practice of reporting small-volume disease in SLNs in Europe. Digital images of lymph nodes with small-volume tumor were distributed to 24 participant pathologists who were asked to categorize the lesions as ITCs, micrometastasis, or something else and to classify each case into a pathologic lymph node category of the TNM staging system. κ statistics were used to assess interobserver variability. The values for the consistency of categorization and pN designation were 0.39 and 0.35, respectively. After clarification of definitions, participants reevaluated the cases and the values improved for both categorization into groups (0.49) and pN classification (0.44). The authors recommended refinement of histopathologic categorization criteria of small-volume disease in lymph nodes. They provided detailed histopathologic criteria and made illus trative images available in public Web sites. Turner et al87 reported that pathologic subcategorization of small-volume disease can vary significantly even among expert breast pathologists and high-level agreement can be achieved when pathologists are asked to apply a set of well-defined criteria. In this study the participating pathologists were asked to classify images of SLNs containing small-volume metastases before and after a training program. A significant improvement in agreement rate was achieved after training.

EVALUATION OF THE SLNs

The number of lymph nodes present and the size, color, and texture of each node in the specimen must be recorded. The presence of blue dye should be noted. Pathologists should also note the relation of the SLNs to the surrounding adipose tissue because extranodal tumor extension is an important prognostic factor in breast cancer. To assess the size of the extranodal extension, pathologists should include perinodal adipose tissue in the histologic evaluation. Lymph nodes should be sectioned serially at 1- to 2-mm intervals to reveal as much of the parenchymal surface as possible. Sections should be made with a sharp, clean scalpel to avoid artifact crushing and contamination. The cut surfaces must be carefully examined and described. Crushed lymphocytes or histiocytes may look like carcinoma cells on sections that are suboptimally fixed or processed. Contamination especially becomes a concern when tissue is processed for molecular assays. Even contamination with benign breast tissue may lead to falsepositive results.

After gross identification of lymph nodes, sections are taken for histologic evaluation. Identification of metastases especially small-volume metastases depends on the extent of histopathologic evaluation of the lymph nodes. It is clear that evaluation of SLNs may have false-negative and false-positive aspects. A false-negative result may be due to the size of metastases and suboptimal sampling of lymph node sections. Although it is clear that assessment of more histologic sections and use of immunohistochemistry will result in an increase in the detection rate of small-volume tumor deposits in SLNs, the precise size stratification of tumor deposits for prediction of clinical significance has not been determined. Until we have these data, routine clinical practice should include evaluation of SLNs to detect all metastases greater than 2 mm (pN1) and most micrometastases (pN1mi). This can be achieved by gross serial sectioning of SLNs as 1- to 2-mm parts and submission of all sections for histopathologic evaluation. Although IHC staining for cytokeratin is not recommended by many of the official guidelines, we routinely use it in our practice and find that immunohistochemistry is helpful to confirm the diagnosis of small-volume metastases (Figure 3). Detection of each possible ITC should not be the goal of routine histopathologic evaluation of SLNs. In addition to sampling factors, interpretational factors also play a role in false-negative assessment.88,89 Small-volume tumor can simply escape recognition by the pathologist screening H&E sections. Tools such as supervised automated microscopy and automated computer-assisted image analysis have been reported to increase effectiveness of conventional light-microscopic evaluation.90,91 However, these adjunctive methods are currently investigational. Morphologically unusual metastases may mimic nonneoplastic cells such as histiocytes or nevus cells and may cause false-negative interpretation (Figure 4). In clinical practice, this is a more significant issue with frozen section evaluations. Immunohistochemical staining for epithelial markers should be used to delineate epithelial clusters whenever there is a question on H&E sections.

False-positive identification in SLNs has not been extensively documented in the literature but occurs occasionally in clinical practice. Benign epithelial inclusions may mimic metastatic deposits in lymph nodes (Figure 4).92,93 The presence of myoepithelial cells around these benign inclusions is the most helpful diagnostic clue and can be demonstrated by immunohistochemistry for myoepithelial specific markers. Passive mechanical transport of either normal or neoplastic epithelium into lymph nodes following needle biopsy procedures on the breast has been reported. 94-96 Because histologic criteria to identify these cases are not well established the true incidence of this phenomenon is not clear. It is possible to speculate that some of the ITC cases actually may represent mechanical transport of tumor cells.

Intraoperative evaluation of SLNs should be performed only if the surgical approach to the axilla is going to be based on SLN status. Either frozen section or imprint cytology can be used and both methods have similar sensitivity and specificity. Sensitivity of both methods is quite high for detection of macrometastases and uniformly low for small-volume disease.97-102 The choice of method for intraoperative evaluation should be based on institutional practices and pathologists' preference.

REPORTING OF SLN HISTOPATHOLOGIC FINDINGS

All surgical pathology reports of SLNs should include the total number of lymph nodes received and the presence or absence of macroscopic tumor involvement.1,41,103 If there is a macroscopically involved node, a representative section will be sufficient to document metastases. The number of lymph nodes involved with metastatic carcinoma, the size of metastatic carcinoma, and presence or absence of extranodal tumor extension should be clearly documented (Table 2). If special techniques are used to detect metastases, this needs to be documented.

CONCLUSIONS AND PRACTICAL RECOMMENDATIONS

Reliable pathologic staging is essential to planning treatment for invasive breast cancer patients. Sentinal lymph node biopsy can be used to accurately detect axillary metastases. There are marked variations in histopathologic SLN processing among different institutions. Although micrometastatic disease is frequently detected, no conclusive data on clinical relevance of these smallvolume metastases yet exist. However, ITCs are now generally believed to be of limited prognostic significance and are categorized as node-negative disease. Therefore, we recommend using the current edition of the AJCC staging criteria to assign pathologic stage. Decisions about local and systemic therapy should be based on staging information and primary tumor characteristics (Table 3). Further refinement of the histopathologic features of the primary tumor and micrometastases may facilitate a more selective approach to local and systemic therapy options in various clinical settings.

(C) 2009 Archives of Pathology & Laboratory Medicine. via ProQuest Information and Learning Company; All Rights Reserved