Sentinel Lymph Node Biopsy in Patients With Melanoma 

Background

Melanoma is a lethal disease that continues to rise in incidence. Although it has been known since the time of Hippocrates, it was not described in western literature until John Hunter recognized it in 1787. However, Hunter’s specimen would not be examined under a microscope until 1968, when Dr. Bodenham confirmed that the “black cancer” described by Hunter was in fact melanoma.^[1]

In 2009, an estimated 68,720 new cases of cutaneous melanoma were diagnosed. Of those cases, 8,650 patients will have died of the disease in the United States.^[2] Melanoma is the fastest growing cancer in men and the second fastest growing cancer in women after lung cancer.^[3] Its median age of diagnosis is 59 years, and it is second only to adult leukemia in terms of loss of years of potential life per death.^[4] In addition, melanoma has surpassed cervical cancer as the most common diagnosis in women aged 20-29 years.

Risk factors for development of melanoma include multiple clinically atypical moles or dysplastic nevi, family history, prior melanoma, inherited genetic mutations, and sun exposure.^{[5, 6]} Although melanoma often arises in patients with fair skin and with significant sun exposure, it can arise in any ethnicity and in areas of the body without substantial sun exposure.^[3]

The theory of lymphatic spread of cancer has been known for quite some time. Accordingly, it was reasoned that removing the lymph nodes before melanoma could spread could prevent systemic disease. In 1898, Snow published his work on the concept of “anticipatory gland excision,” or what would be known today as elective lymph node dissection (ELND).^[7]

Although Snow’s reasoning was sound, multiple randomized controlled trials, short of subset analysis, showed no beneficial effect on survival or local control.^{[8, 9, 10]} These trials were performed mainly on intermediate-depth melanomas (depth 1-4 mm). In patients with thin melanomas (depth < 1 mm), the rate of nodal involvement seemed low enough that an ELND was not warranted.^[11] For thicker melanomas (depth >4 mm), the risk of systemic metastasis seemed to outweigh the potential benefit of ELND.^[12]

In 1955, Seaman and Powers used radiolabeled colloidal gold to describe the path of lymphatic channels followed by cancer.^[13] Studies by Gould et al then described the first node to be involved in lymphatic spread, which was then termed the “sentinel node” (SN) or “sentinel lymph node” (SLN).^[14] Finally, in 1992, Morton et al showed that the SLN could accurately predict the potential for cancer in the associated basin.^[15]

Although the complex anatomy and lymphatic drainage of the head and neck (see Anatomic Considerations) can make treatment of melanoma in this area quite challenging, sentinel lymph node dissection (SLND) can be performed with acceptable morbidity. SLND carries profound prognostic implications and dictates subsequent therapeutic decisions. The impact of further therapy in patients with a positive SLN (eg, completion lymph node dissection [CLND] and interferon) is yet to be determined.

For more information, see Malignant Melanoma, Head and Neck Mucosal Melanomas, and Malignant Melanoma Staging.

Indications

SLND should be offered to all patients with a clinically negative nodal basin and a primary melanoma greater than 1 mm in depth. As noted (see Background), evaluation of the nodal basins via ELND was most commonly limited to melanoma with a depth of 1-4 mm. SLND carries a lower risk of morbidity, which has expanded the indications for evaluating the nodal basins associated with melanoma.

Patients with melanomas less than or equal to 1.0 mm in thickness are at lower risk for nodal involvement (2-5%),^[16] and there is debate as to whether they would benefit from evaluation of their nodal basin. Currently, the National Comprehensive Cancer Network (NCCN) recommends against SLND for patients with melanoma in situ or melanoma less than 1.0 mm in thickness.

Patients with melanoma between 0.76 mm and 1.0 mm in thickness may be considered for SLND if they have adverse features such as positive deep margins, lymphovascular invasion, age less than 40 years, significant vertical growth face, increased mitotic rate, and Clark’s level IV or higher.^[3] Those with melanoma that exhibits regression may also be considered, but this is more controversial. Some recent studies have indicated that there may be no association between regression and positive SLN status.^{[17, 18, 19]}

SLND is also offered to patients with deep (>4 mm) melanoma and clinically negative nodes because it has proven to yield valuable prognostic information, with rates of positive SLNs ranging from 30-40%.^{[20, 21]}

Contraindications

In certain situations, SLND may not be warranted. It is unnecessary when a patient presents with systemic disease. Similarly, when a patient presents with a clinically evident node, the first step should be evaluation of the node with fine-needle aspiration as opposed to SLND.^[3]

Whether SLND should be performed after a wide local excision is somewhat controversial; there is concern that lymphatics may be altered. It appears, however, that if extensive reconstruction has not been done, SLND can be used in patients with a prior wide local excision.^{[22, 23]}

For patients who have previously undergone SLND, there is nothing in the melanoma literature to establish definitive recommendations. The breast cancer literature suggests that SLND may be feasible,^[24] but whether such limited data can be applied to melanoma patients is a matter of debate. Nevertheless, given the reliability of lymphatic mapping for the identification of nodal pathways, the authors feel that prior SLND is not a contraindication to repeat SLND.

Anatomic Considerations

Following the lymphatic channels below the neck is relatively straightforward. In the extremities, the lymphatics follow the path toward either the axilla or groin. Clinicians should be aware that whereas the majority of SLNs are found in the major nodal basins, some can also be found in the epitrochlear nodal basin in the upper extremities and the popliteal nodal basin in the lower extremities.

In the truncal region, lymphatics can go to multiple nodal basins. This has made finding SLNs even more valuable, in that doing eliminates the guesswork from efforts to determine where melanoma might spread. It is important, however, to keep in mind that nodes may be found anywhere along the line from the melanoma site to the nodal basin and that any of these nodes has the potential to harbor metastasis.^[25]

The lymphatic drainage of the head and neck is more extensive and variable than that of the trunk and extremities. The lymphatic drainage of the anterior neck is based on a system developed at Memorial Sloan Kettering in the 1930s. Seven levels of lymph nodes exist (see the image below).

Levels I-V describe the lateral neck lymph node basins, and levels VI and VII describe the central lymph node basins, as follows:

• Level I nodes are the submandibular nodes and lie between the mandible and the posterior and anterior bellies of the digastric.
• Level II nodes are the upper jugular nodes, which lie between the skull and the hyoid bone along the internal jugular vein.
• Level III nodes are the middle jugular nodes and lie between the hyoid bone and the omohyoid muscle along the internal jugular vein.
• Level IV nodes are the lower jugular nodes and lie between the omohyoid muscle and the clavicle along the internal jugular vein.
• Level V nodes are the posterior triangle nodes which lie between the trapezius and the sternocleidomastoid muscle above the clavicle.
• Level VI nodes are centrally located between the carotid sheath and above the suprasternal notch.
• Level VII nodes are the superior mediastinal nodes and lie below the suprasternal notch.

Traditionally, a line that bisected the ear caudad to cephalad marked the boundaries of the anterior and posterior drainage of the scalp (see the image below). Lesions anterior to this line would drain to the parotid, submandibular, and cervical lymph nodes (levels I-V). Lesions posterior to this line would drain to the posterior auricular, suboccipital, upper jugular, and posterior triangle lymph nodes. It was assumed that lymphatic drainage would not cross the midline.

The advent of lymphoscintigraphy, however, showed that these assumptions can be inaccurate. In 1994, Wells et al found that lymphoscintigraphy was discordant with clinical predictions of nodal drainage 84% of the time.^[26] Even after revision of the clinical parameters, a study by O’Brien the following year still found the rate of discordance to be 33%.^[27] Uren also reported discordance rates of 33% and determined that nodal basins draining across the midline in 10% of cases.^[28]

Staging

The staging of melanoma has undergone many changes. It is commonly done according to the tumor-node-metastasis (TNM) system of the American Joint Committee on Cancer (AJCC). Revisions made by the AJCC in 2009 included additions for ulceration in the T classification and micrometastasis versus macrometastasis in the N classification (see the tables below).

Table 1. AJCC TNM Classification of Melanoma.^[29] (Open Table in a new window)

Table 2. AJCC Clinical Staging of Melanoma.^[29] (Open Table in a new window)

Overall, rates of finding sentinel lymph nodes (SLNs) have been very good. Initially, Morton and colleagues used only blue dye and were able to identify SLNs in 194 of the 237 basins they examined (82%).^[15] The identification rates were improved by the use of radiolabeled isotopes. Gershenwald et al increased their identification rates of all sites from 87% to 99% when using^99m Tc-labeled sulfur colloid.^[30]

The identification rates for SLNs in the head and neck region are somewhat lower than those for other sites. Rates of successful SLN identification range from 90-96%.^[31] The main reason for the decreased identification rates is the complex lymphatic pathways mentioned earlier. The high density of lymphatic basins may cause significant background noise, which can interfere with use of the gamma probe.

Additionally, the use of a combination of blue dye and lymphoscintigraphy presents challenges in the head and neck region. On the face, blue dye may persist for several weeks, and many patients may not tolerate such a visible mark. In these cases, the authors will use blue dye as long as it is included in the area of resection. If the resection is a small area, the authors will use a smaller amount of dye to ensure that none is left behind.

Overview

After melanoma is diagnosed and it has been determined that no contraindications to sentinel lymph node dissection (SLND) are present, the patient may be scheduled to undergo wide local excision of the primary melanoma site and SLND. Some centers wait for 24 hours after injection of the radiotracer before performing SLND, but the authors have found it more practical to perform the injection on the day of surgery.

Lymphoscintigraphy is performed in the nuclear medicine department by injecting the radiotracer intradermally around the melanoma lesion. Lymphatic imaging is then performed after injection to confirm appropriate uptake of the radiotracer (see the image below). It may be beneficial to use a handheld gamma probe in the preoperative area to identify potential sites of sentinel lymph nodes (SLNs).

After the patient is transported to the operative suite, anesthesia is induced at the discretion of the anesthesiologist in consultation with the surgeon. Approximately 1 mL of blue dye is injected intradermally at the site of the lesion. Either isosulfan blue or methylene blue may be used, but the authors prefer methylene blue because of its decreased cost and equivalent efficacy. The lesion is then massaged for 4-5 minutes to enhance lymphatic drainage.

A handheld gamma probe is used to identify “hot spots,” which will indicate areas where there are SLNs. A small incision is then made overlying the hot spot. Clinicians should be aware that completion lymph node dissection (CLND; completion lymphadenectomy) may be necessary after pathologic examination and should plan their incisions accordingly.

One area deserves special mention. For nodes in the parotid region, the authors have rarely found it necessary to perform a parotidectomy and facial nerve dissection to complete a SLND and therefore recommend a preauricular incision for hot spots noted in the parotid region. With careful dissection parallel to the facial nerve branches, SLNs can usually be identified without a formal parotidectomy. The authors have found that repeat surgery for CLND can be performed with acceptable risk to the facial nerve; the literature supports this finding.^[32]

After the nodal basin is identified, blue lymphatics can help visually guide the surgeon to the blue nodes (see the first image below). The handheld gamma probe is used to identify hot nodes in the field. Any nodes with significant radiotracer activity are removed and their radioactivity counts measured ex vivo (see the second image below).

SLNs are then sent to pathology for appropriate staining. An SLN is defined as any of the following:

• Any grossly suspicious lymph node
• Any node that harbors blue dye
• Any node whose radioactivity count is greater than or equal to 10% of that of the hottest node removed

Dissection is continued until the nodal bed count is 10% of that of the hottest node removed.^[33] The literature has not established the maximum number of lymph nodes that can be harvested. The literature on breast cancer SLND suggests that the great majority of SLNs harboring malignancy will be among the first 4 nodes harvested^[34] ; any additional nodes will likely be negative, and their removal may in fact increase the potential morbidity of the procedure. Once SLND is complete, wide local excision of the primary melanoma is performed.

Pathologic examination

In sentinel lymph node dissection (SLND) for melanoma, unlike SLND for breast cancer, the utility of routine intraoperative pathologic sectioning is somewhat debatable. Concerns have been raised about the possibility that frozen sectioning may disrupt what few malignant cells may be present and prevent further analysis.^[35]

In 2000, Koopal et al found that intraoperative frozen sectioning carried a sensitivity of 34% and a false-negative rate of 12%.^[36] Tanis et al compared intraoperative frozen sectioning of SLND in melanoma patients with intraoperative frozen sectioning in breast cancer patients and found that the sensitivities were 47% and 74%, respectively.^[37]

Others have found that frozen sectioning can provide accurate results and minimize repeated trips to the operating room (OR), with decreased exposure to anesthetic.^{[38, 39, 40]} In 2008, Alkhatib et al reported a much higher sensitivity rate—91%—and argued that intraoperative frozen sectioning should be performed routinely.^[38]

With regard to head and neck melanoma, the authors’ position is that intraoperative frozen sectioning should not be performed. The reasoning is that conversion from SLND to completion lymph node dissection (CLND) in the axilla or femoral triangle is relatively straightforward. Contrary to this, a positive sentinel lymph node (SLN) in the neck may dictate a modified radical neck dissection with parotidectomy and facial nerve dissection.

The wide disparity between these 2 procedures makes it impractical to prepare a patient for both possible outcomes. Awaiting the final pathology results on a SLN permits discussion of additional surgery and seems like the most appropriate option.

For permanent sectioning, the authors’ institution performs hematoxylin-eosin (H&E) staining followed by staining with immunohistochemical (IHC) markers S-100, MART-1, and HMB-45. In addition to IHC evaluation, National Comprehensive Cancer Network (NCCN) guidelines also recommend that multiple sectioning of each node be performed to pick up potential micrometastatic deposits.^[3] Consideration can also be given to performing polymerase chain reaction (PCR) evaluation of the SLN.

Also see Sentinel Lymph Node Biopsy in Melanoma Pathology.

Expected Outcomes

Because the lymphatic system of the head and neck region is complex, some concern exists regarding the possibility of missing metastatic SLNs and the potential for recurrence in a nodal bed after a negative SLND. The literature is not unanimous, suggesting a failure rate anywhere from 0% to 25%.^{[27, 32, 41, 42, 43, 44]} In the authors’ experience, patients with head and neck melanoma have a higher rate of recurrence than those with melanoma at other sites.

A positive SLN carries extremely valuable prognostic information. Survival rates for patients with a negative SLN have been shown to be superior to those with a positive SLN. In addition, the method of detection has been to shown to be a predictor of survival as well. SLNs that are positive only on PCR are associated with a better survival rate than those detected with H&E staining (see the image below).^[45]

In patients with a positive SLN, current recommendations are to perform CLND and treat with adjuvant interferon. Ironically, elective lymph node dissection (ELND) failed to demonstrate a survival advantage when micrometastatic disease was found. Furthermore, the studies showing benefit of interferon included patients with a higher tumor burden, and the impact of interferon in the treatment of micrometastatic disease (ie, a positive SLN) is unknown.^[46]

Multiple prospective randomized controlled trials, such as the Sunbelt Melanoma Trial, have been designed to determine the impact of additional therapy in the setting of a positive SLN. However, because the data are not yet mature, the verdict is still out.

Complications

Although often fraught with peril, surgery in the head and neck area can generally be navigated with the help of experience and sound judgment. Aside from the usual complications of bleeding and infection, the most concerning complication in this area is injury to the facial nerve. Although rates of injury have not been reported, they are likely to be low.

The potential for injury to the facial nerve has led some to recommend a superficial parotidectomy as opposed to mapping the parotid gland.^[47] However, a study by Schmalbach et al on 80 patients with head and neck melanoma reported no dysfunction of the cranial nerves, including the facial nerve, and concluded that a superficial parotidectomy is unnecessary for evaluating SLNs in the parotid region.^[43] The authors have had similar experiences.

Long-Term Monitoring

The stage of disease dictates follow-up, as outlined by the NCCN. Patients with stage 0 or in situ melanoma should have annual skin examinations and monthly self examinations. Patients with stage I-IIA disease should undergo physical examinations with attention to skin and nodes every 3-12 months for 5 years and then annually as clinically indicated. Routine laboratory and radiologic testing for surveillance is not recommended.

Patients with stage IIB-IV disease require examinations every 3-6 months for 2 years, then every 3-12 months for 3 years, then yearly for the remainder of their lives. The use of screening chest radiography, computed tomography (CT), positron emission tomography (PET)-CT, or magnetic resonance imaging (MRI) is left to the discretion of the treating physician.^[3] Given the low yield, high false-positive rate, and cumulative radiation exposure, the authors do not endorse this practice.