Sentinel Lymph Node Biopsy in Melanoma Pathology

Updated: Nov 18, 2016
  • Author: Victor G Prieto, MD, PhD; Chief Editor: Jon A Reed, MD, MS  more...
  • Print


A sentinel lymph node (SLN) is the first lymph node or group of nodes to receive lymphatic drainage from a tumor at a particular anatomic area. Thus, an SLN is most likely to contain any metastatic deposits.

Within the past 15 years, evaluation of SLNs has become the most popular method for early staging of several malignancies, including breast carcinoma, gastrointestinal carcinoma, and melanoma. The main advantages SLN biopsy has over complete lymphadenectomy are: (1) the lower number and lesser degree of side effects and (2) the selection of a small number of lymph nodes that can then be extensively analyzed by the pathologist (through examination of hematoxylin and eosin [H&E] sections and immunohistochemistry [IHC]).

Approximately 20% of patients with primary cutaneous melanoma will have metastatic melanoma in their SLN. According to most published data, patients in whom the SLN contains metastatic melanoma have a poorer prognosis. These patients are then offered a completion lymphadenectomy to remove the remainder of the lymph nodes in that anatomic area. Current studies are analyzing the possible therapeutic effect of removing positive SLNs. [1, 2]

See also the following:


Clinical Features

By definition, sentinel lymph nodes (SLNs) are analyzed in patients who do not have a clinically positive lymph node. Therefore, an SLN should be asymptomatic. Most protocols recommend SLN examination in melanoma cases with any of the following characteristics (from the College of American Pathologists CAP and the American Joint Committee on Cancer AJCC)  [3] :

  • Breslow thickness 1 mm or greater
  • For cases thinner than 1 mm in Breslow thickness:
    • Mitotic figures in the dermal melanoma cells (stage pTIb)
    • Ulceration (stage pT1b)

Other features that some institutions recommend to use as indication for SLN examination are Clark level IV (if quantification of mitotic figures is not available), satellitosis, regression, and vascular invasion. Interestingly, some recent studies have shown better prognosis and lower rate for positivity of SLN in those patients with regression in their melanoma lesions. [4]

On the other hand, regarding histologic subtypes, the current trend is not to recommend SLN examination for pure desmoplastic melanomas, regardless of tumor thickness. [5, 6]


Differential Diagnosis

Macrophages, pigmented or not, may resemble metastatic melanoma in the sentinel lymph nodes (SLN). In such cases, immunohistochemistry (IHC) is very helpful, because melanoma cells usually react with a panmelanocytic cocktail (see Immunohistochemistry) However, because this cocktail contains antibody to melanoma antigen recognized by T cells (MART)-1 (anti-MART-1) and some macrophages may express MART-1, [7] it may be necessary to use other, more specific markers, such as gp100 (with HMB-45), microphthalmia transcription factor (MiTF), or SOX10.

As many as 20% of lymphadenectomies from the axilla or groin contain benign collections of melanocytes known as capsular or nodal nevi. [8] The characteristic capsular location of these nevus deposits and the lack of significant cytologic atypia distinguishes them from the common subcapsular location of metastatic melanoma. Since analysis of the cytomorphologic features of such nevus cells in the lymph node is so important, it is always recommended to compare these cells with the melanoma cells in the primary lesion.

A potential diagnostic pitfall is the presence of vascular metastasis in the intracapsular lymphatic vessels of the node, which mimics capsular nevi. (In the authors' experiences, it is extremely rare to detect nevus aggregates within the vessels of the SLN capsule.) Use of IHC to label endothelial cells (CD31, CD34, or D2-40) may be helpful in detecting the rim of endothelial cells, thereby confirming the intravascular location of the cells and supporting a diagnosis of metastatic melanoma.

Capsular nevi may rarely extend into the underlying node parenchyma. Those nevus cells lack gp100 expression (with HMB-45) and show very low Ki67 expression (features characteristic of benign melanocytes). [9, 10]

The authors developed a homegrown cocktail that includes both anti-MART-1 and the MIB1 antibody (against Ki67). Because Ki67 is expressed in the nucleus and MART-1 in the cytoplasm, it is relatively easy to determine how many of the melanocytes (ie, cells expressing MART-1) are proliferating (ie, also expressing Ki67). Nevus cells should have a very low proliferation rate (< 1%).


SLN Identification and Localization

Identification of the sentinel lymph node (SLN) is usually achieved by means of both lymphoscintigraphy and the injection of a colored dye.

On the day of the operation, before the procedure is begun, a radioactive tracer (usually labeled with technetium-99m [99m Tc]) is injected in the area of the lesion; it is then transported by the lymphatic system to a nearby lymphatic basin, where it can be detected with a Geiger counter placed above the skin. In this way, the surgeon can determine which basin or basins may harbor metastatic melanoma.

During the procedure, a blue dye is also injected in the same area. The surgeon uses the Geiger counter to identify the area in the basin or basins with the highest count and then locates any SLNs that are labeled with the blue dye.

The surgeon removes any blue SLN found and then rescans the area with the Geiger counter to determine if there are other lymph nodes (secondary SLNs) that are also radioactive. If secondary SLNs are found, they are removed as well. The threshold of radioactivity at which a node is considered positive may vary among institutions. For example, some use a threshold of 10 times the radioactivity of the background skin, and others use a threshold of 10% of the count of the "hottest" SLN.

Lymphoscintigraphy is also useful for obtaining a precise determination of the pattern of lymphatic drainage. SLNs can be located not only in the main lymphatic basins (axilla, groin, and neck) but also, depending on the location of the primary lesion, in the elbow (epitrochlear), the knee (popliteal), the face (parotid gland), and other sites. Furthermore, some areas may drain to more than 1 basin—for example, a lesion on the midback may drain simultaneously to 1-4 different basins (axillae or groins).



Imaging techniques include ultrasonography to determine if any of the lymph nodes in the basin draining the region where the melanoma is located is likely to contain metastatic deposits. Suggestive findings include enlargement and an irregular image. Sonography is becoming so precise that some protocols relay on this technique for clinical follow-up of patients with cutaneous melanoma.



As mentioned earlier, the most commonly used modality is lymphoscintigraphy, in which a radioactive tracer is injected in the area of the melanoma. The tracer travels through the lymphatic system, and shortly afterward, with the aid of a Geiger counter, the area with the highest count is marked. The SLN usually underlies this area.

To improve the detection of the SLN, a blue dye is also usually injected in the skin where the melanoma is. This dye will also be transferred to the nearest basin, thereby facilitating identification of the SLN (by turning the SLN blue). In general, the lymph node with the highest count will also contain blue dye, although sometimes this is not the case.


Gross Features

The sentinel lymph node (SLN) may or may not be blue, depending on how much of the injected blue dye it contains. It should be labeled by the surgeon to indicate: (1) the anatomic source (basin), (2) whether it was "blue" or "not blue," and (3) the radioactivity count. In general, the amount of radioactivity contained in the SLN is minimal by the time it reaches the pathology laboratory. However, the institution's safety department should ensure that the technique employed permits safe processing of the SLN is safe.

Frozen sections and touch preparation cytology

Most protocols discourage the use of frozen sections to evaluate SLNs. Although frozen sections may be common in SLN biopsy for other malignancies (eg, breast carcinoma), the authors do not recommend them in SLN biopsy for melanoma, for several reasons.

Frozen sections provide a suboptimal morphology and may lack the subcapsular region of the lymph node (which is the area most commonly involved in melanoma metastases). Furthermore, because processing of the frozen tissue requires embedding it in paraffin, facing off the block, and performing new sectioning of the tissue, some tissue will inevitably be discarded; any micrometastases in those discarded, unexamined sections will be missed. [11]

A possible alternative technique is touch preparation cytology. [12, 13] Preparation cytology is not widely used in this setting; however, because evaluation of the specimens requires expertise in cytologic smears (melanoma cells and pigmented macrophages may be very difficult to distinguish).

Other techniques for SLN processing

There is no universally accepted protocol for processing SLNs. It is obvious, however, that the classic processing used with non-SLNs—that is, bivalving the node and examining a single routine hematoxylin and eosin (H&E) slide—misses a number of small metastases. [11, 14] Consequently, other methods have been proposed, all of them involving more than one H&E slide and the possible addition of immunohistochemistry (IHC). [15] (See Immunohistochemistry).

The original method described by Cochran et al calls for bivalving the SLN along the long axis, ideally through the hilum of the lymph node if possible. [16] In other institutions, including the authors', the SLN is breadloafed (ie, serially sectioned along the short axis) to avoid the problem of having to locate the hilum and additionally to increase the amount of the subcapsular region examined in the histologic sections. [17]

Some clinical trials call for preserving a portion of the node for polymerase chain reaction (PCR) analysis to try to detect mRNA associated with melanocytic differentiation (melanoma antigen recognized by T cells [MART]-1, gp100, etc). A study from 2016 indicates that patients whose sentinel lymph nodes are positive only by PCR do not benefit from completion lymph node dissection (CLND) or CLND plus high-dose interferon therapy. [18]  


Microscopic Features

The authors' protocol calls for 1 initial hematoxylin and eosin (H&E) slide of the breadloafed sentinel lymph node (SLN). If this slide is positive, it is reported as such; if not, the block is resubmitted to the laboratory to obtain a new H&E section slide (~200 microns deeper in the block) and 2 unstained slides.

Most studies report that approximately 20% of patients with cutaneous melanoma show deposits of melanoma cells in their SLN. Metastatic melanoma cells may be epithelioid or spindled, pigmented or amelanotic, but they most commonly resemble the cells in the primary lesion.

Therefore, in examining an SLN, as mentioned above, it may be very important to study the original melanoma so that the morphologic features can be compared. Such comparison is particularly essential in cases of nevoid melanoma (in which melanoma cells are small and may resemble nevus cells) and spindle cell melanoma (which may be negative for some immunohistochemical markers). It may be difficult to distinguish pigmented melanoma cells from melanophages; however, pigment granules are usually coarser and larger in macrophages than in melanoma cells.

Usually, melanoma cells in the SLN are located in the subcapsular sinus as single cells, small nests, or large clusters (see the images below). Less frequently, metastases are intraparenchymal. Very rarely do melanoma cells involve the fibrous capsule; in most such cases, the phenomenon is likely secondary to involvement of intracapsular lymphatic vessels. There may be extracapsular extension into the perinodal fibroadipose tissues (< 5% of cases).

A, B: Lymph node with nests of melanocytes (*) in A, B: Lymph node with nests of melanocytes (*) in capsule (C) and subcapsular regions (arrows). C, D: Nevus cells in capsule and subcapsular regions are positive for melanoma antigen recognized by T cells (MART)-1.
A: Nests of melanocytes in the capsule of a lymph A: Nests of melanocytes in the capsule of a lymph node with banal morphologic features. Nevus cells demonstrate smooth nuclear membrane and inconspicuous nucleoli (insert). B: Primary invasive melanoma with atypia of melanocytes. Melanoma cells have irregular nuclear borders and prominent nucleoli (insert). C: Evaluation of nests of melanocytes in a lymph node with double immunohistochemical slide with melanoma antigen recognized by T cells (MART)-1/Ki67. Nests of melanocytes are positive for MART-1 (red chromogen) with labeling in cytoplasm (arrows and insert). In contrast, internal control lymphocytes from germinal center show nuclear reactivity with anti-Ki67 (brown chromogen, insert). D: Nevus cells in lymph node are negative with HMB-45.


Approximately 20% of patients have a positive sentinel lymph node (SLN). In our experience, in 16% of patients the positive SLNs are detected in the initial hematoxylin and eosin (H&E) slide; in the remaining 4%, the positive SLNs are detected with serial sections or immunoperoxidase. [17] S100 is the most sensitive marker; however, it labels lymph node dendritic cells in addition to melanocytes, and as a result, single melanoma cells may be hard to distinguish from a background of dendritic cells. Therefore, other markers may be useful. [19, 20, 21]

The authors recommend a panmelanocytic cocktail that includes HMB-45, antibody to melanoma antigen recognized by T cells (MART)-1 (anti-MART-1), and anti-tyrosinase. [22] In addition, because MART-1 can be expressed by macrophages, [7] the authors sometimes use HMB-45 by itself to help differentiate between macrophages and melanoma cells (HMB-45 usually does not label macrophages) or similarly anti-MITF or anti-SOX10, since they are relatively very specific for melanocytic differentiation.

Additionally, in cases of spindle cell melanoma in which the tumor cells do not express MART-1 or gp100 (with HMB-45), the authors use anti-S100 or anti SOX10. As mentioned before, such cases will benefit by comparing the morphology of the cells in the SLN with those in the primary lesion.


Tumor Spread and Staging

The majority of cutaneous melanomas spread via lymphatic vessels through the lymphatic basins. Examination of sentinel lymph nodes (SLNs) in patients with melanoma is rapidly becoming an essential component of tumor staging; as such, it is included in the current classification of the American Joint Committee on Cancer (AJCC). [3] Furthermore, a number of current melanoma therapeutic protocols require information on SLN status to correctly stratify patients' chances of recurrence or survival.


Prognosis and Predictive Factors

Sentinel lymph node (SLN) positivity is associated with decreased survival in patients with cutaneous melanoma [23, 24, 25, 26, 27] and provides significant prognostic information on both univariate and multivariate analysis. Melanoma metastasis size in SLNs correlates with subsequent involvement of non-SLNs from the same anatomic region [28, 29, 30, 31, 32, 33] and with overall prognosis. [29, 34, 35, 36, 37, 38, 39]  Although some of the original studies suggested that examination of SLN is useful only in thin or intermediate melanomas, it actually provides clinically significant prognostic information also in thick melanomas (T4 lesions) [40, 41]  

Also controversial is the value of SLN in pediatric melanoma patients. Although a study Paradela et al did not detect differences between positive and negative SLN in patients younger than 18 years, [42] , at least one study indicates impaired prognosis in patients with positive SLN. [43] SLN provides a prognostic measure for pediatric melanoma-specific survival (MSS), but it does not appear to confer any survival benefit in children. In an analysis of data from 310 pediatric patients, those who underwent SLNB (n = 261) and had a positive finding had a worse MSS than those with negative results. However, the SLNB-positive children had more favorable survival when compared to patients older than 20 years. [43]

Some authors recommend measuring in the SLN the distance between the capsule and the most deeply located deposit of melanoma cells. [15] The authors measure the tumor burden [39] in the SLN as determined by the size of the largest tumor deposit (in mm, 2 dimensions), the location (subcapsular vs intraparenchymal), and the presence or absence of extracapsular extension.

A majority of the responders to a survey in Europe also reported the size of the largest tumor deposit in the SLN, [15] but some studies have not detected an association between tumor location (subcapsular vs intraparenchymal) and survival. [44]

A large, single institution, retrospective analysis of the histologic features of melanoma in 1029 SLN from 409 patients with a single primary cutaneous melanoma was completed. It demonstrated that size of tumor deposit in SLN (size >2.0 mm), extranodal extension (ENS), perilymphatic invasion (PLI), and positive nonsentinel lymph nodes in complete lymph node dissection (CLND) specimens were all independent predictors of poor disease-free survival (DFS). When controlled for CLND status, SLN tumor deposit (size >10.0 mm) and ENS were independent predictors of distant metastasis-free survival (DMFS). SLN with ENS and PLI were independent predictors of melanoma specific survival (MSS).

In addition, the authors noted cutaneous site of melanoma on the trunk, presence of ulceration, and satellites were independent poor predictors of DFS. Male gender, ulceration, satellites, and absence of TILs were independent predictors of DMFS. Ulceration, age older than 50 years, and satellites were independent predictors of MSS when controlled for CLND status. [45]

In a separate study of 1417 patients who had 237 positive SLNs, [44] preliminary data [46] suggest a stratification into 3 groups with progressively worsening prognoses, as follows:

  1. One or 2 positive SLNs and metastasis size 2 mm or smaller, without ulceration of the primary lesion
  2. Ulceration of the primary lesion or any metastatic nest larger than 2 mm
  3. Involvement of 3 or more SLNs or ulceration of the primary lesion and any metastatic nest larger than 2 mm

The authors are currently analyzing a larger series to determine whether such a stratification scheme provides clinically significant prognostic information.

Increased angiogenesis in the SLN-positive patient appears to be associated with a higher incidence of additional positive lymph nodes. [47]  With regard to the relation between tumor size and prognosis, patients with smaller melanoma metastases to the SLN have a better prognosis than those with larger metastases; however, most data indicate that there is not a definite cut-off in the size of the metastatic deposits. Furthermore, the authors have seen rare cases in which only a single melanoma cell was identified in the SLN, but multiple distant metastases subsequently developed within 4 years of diagnosis.

In addition to SLN positivity and nodal status (N category) in the TNM staging system for melanoma, lymph node ratio (LNR) has been reported to be an independent predictor of disease-specific survival. [48] LNR or N-ratio is defined as the ratio between metastatic and examined lymph nodes from a radical lymph node dissection (RLND). The authors retrospectively analyzed 8,177 patients with cutaneous melanoma who had a minimum of 5 lymph nodes histologically examined. The risk of melanoma-related death (hazard ratio) was found to increase by 2% for each percentage unit of LNR increase.

Furthermore, when LNR values were categorized into intervals of 10 percentage units, there was a 10-fold increased risk of melanoma-related deaths. The prognostic value of LNR was further demonstrated in a subgroup of patients with 10 or more lymph nodes examined, as well as in patients with stage III disease. Since LNR is dependent on the number of lymph nodes examined, the authors demonstrated LNR with ≥ 5 and ≥ 10 lymph nodes. Both appear to have equal predictive power of disease-specific specific survival. LNR may provide additional information for risk stratification in patients with cutaneous melanoma.

Another study identified 9 factors predictive of non-SLN metastases that should be evaluated and recorded. It concluded that patients with low SLN tumor burden could probably be spared CLND, but further research is needed to confirm this finding. [49]  A large multiinstitutional study indicates that there may not be prognostic advantage for CLND in patients with lymph node metastases smaller than 1 mm. [50]

Some studies have indicated that detection of melanocytic mRNA by polymerase chain reaction (PCR) analysis in SLNs correlates with decreased survival [51, 52, 53] ; however, not all authors agree with this finding. [54, 55]  A possible explanation for this discrepancy may be the presence of nevus cells in some SLNs; clearly, those cells also contain melanocytic mRNA and thus would result in positive PCR results. Therefore, to institute PCR analysis for SLNs in melanoma, it may be necessary to discover mRNA specific for melanoma cells.

A retrospective review by Scheier et al found that patient management is not altered by preoperative PET/CT staging in sentinel lymph node-positive cutaneous malignant melanoma since it has a high false-positive rate. [56, 57]