The Role of Sentinel Node Biopsy in Skin Cancer Treatment & Management

The use of sentinel lymph node biopsy and selective lymphadenectomy in patients with malignant melanoma with clinically negative regional lymph node basins has been widely adopted. Current NCCN guidelines recommend this as a component of treatment for clinically node-negative melanoma patients with tumors thicker than or equal to 1 mm and patients with T1b melanomas. With this approach, a complete regional lymph node dissection is performed only in patients with evidence of metastatic melanoma in the sentinel node(s). This technique is also being used for other skin cancers (eg, neuroendocrine carcinoma of the skin [Merkel cell carcinoma]) that have a propensity for regional lymphatic spread.

Despite this, current compliance with these guidelines is only approximately 50% in the United States and is lower in Europe.^{[36, 37]} Single-institution overall compliance data are better (74-84%) and appear to be improving over time, but data are lowest for those patients with thicker (T2 to T4) melanomas (47-74%) who stand to benefit the most.^{[38, 39]} The technical details of sentinel lymph node biopsy for malignant melanoma are described below.

Multidisciplinary approach

Sentinel lymph node biopsy with selective lymphadenectomy is a team effort and requires close collaboration by dermatologists, surgical oncologists, nuclear medicine physicians, and pathologists. Adjuvant therapy may also involve medical and radiation oncologists. A multidisciplinary programmatic approach provides optimal patient care.^[40]

Injection technique

In the nuclear medicine department, 2-16 hours before the operation, the patient is injected intradermally around the site of the primary tumor or biopsy scar with radiocolloid, usually with a dose of 0.5-0.8 mCi. The area is massaged for 2-5 minutes.

Choice of isotope

While a variety of radiocolloids have been used successfully, unfiltered^99m Tc sulfur colloid diffuses out of the sentinel lymph nodes (to second echelon lymph nodes) less rapidly than tracers with smaller particles. In the United States,^99m Tc-labeled albumin colloid and human serum albumin also may be used. In Australia, colloidal antimony sulfide is used, and in Europe, human albumin nanocolloid is used.

Lymphoscintigraphy

Lymphoscintigraphy is helpful in defining at-risk nodal basins, especially for truncal melanomas and melanomas of the head and neck. Lymphoscintigraphy may help define the number of sentinel lymph nodes by showing more than one afferent lymphatic channel leading from the primary tumor site to a regional nodal basin. It can also help define the location of the sentinel lymph node within the lymphatic basin and can identify in-transit or interval lymph nodes that may contain metastatic disease.

Lymphoscintigraphy demonstrates drainage to more than one nodal basin in 10-15% of patients and drainage to interval nodes in 3-5%.^[34] Metastatic disease is found at the same frequency in these interval sentinel lymph nodes as within sentinel lymph nodes from conventional lymph node basins, and the status of one basin does not predict the status of the other.^{[34, 41]} The sentinel node is not always found in the closest nodal basin. Accurate lymphoscintigraphy helps identify all nodes, regardless of their location, receiving direct lymphatic drainage from a primary tumor site. See the image below.

While not all surgeons advocate the use of preoperative lymphoscintigraphy in all melanoma patients, most agree lymphoscintigraphy is necessary to accurately identify the sentinel node(s) in patients with head and neck melanoma. Drainage to nodal basins not predicted by standard anatomic considerations is observed with lymphoscintigraphy in as many as 60% of head and neck and 30% of truncal primary melanomas.^[42]

Dynamic scintigraphy can be performed most conveniently immediately after injection of radiocolloid in the nuclear medicine department on the morning of the operation. Usually, the sentinel lymph node(s) is identified within 30 minutes of injection. The site of each hot sentinel lymph node is tattooed on the patient's skin for identification in the operating room. Two-view lymphoscintigrams are obtained for the surgeon. Delayed 2-hour images may be taken. Delayed images are helpful in detecting sentinel lymph nodes close to the primary site, which may be obscured on the initial views, and to detect drainage to multiple nodal basins.^[43]

After radioisotope is injected in the nuclear medicine suite and lymphoscintigraphy is performed, the patient is moved to the operating room. Vital blue dye (1-2 mL) is injected intradermally around the primary tumor or biopsy site after the patient is prepared and draped using sterile technique. Some clinicians use isosulfan blue dye (Lymphazurin 1%, Zenith Parenterals; Rosemont, Ill). The site is massaged gently for 5 minutes. A handheld gamma probe (eg, C-Trak probe, Care Wise Medical Products; Morgan Hill, Calif]) is used to identify hot spots in the identified regional lymph node basin or basins. If in-transit or interval nodes are identified on preoperative lymphoscintigrams, these should also be sought. Interval nodes contain metastatic disease as often as the conventional nodal basins do and may represent the only site of metastatic disease.

See the images below.

The blue dye may also be seen coursing through the dermal lymphatics en route to the regional lymph node basin(s). See the image below.

A small incision is made over the hot spot. A blue-stained afferent lymphatic vessel is sought, and a combination of this visual cue and the gamma probe leads to the identification of a hot and/or blue sentinel lymph node or nodes. Care should be taken to minimize flap placement because this may interrupt the afferent lymphatics and make sentinel node identification more difficult. More than one sentinel lymph node (with a separate draining afferent lymphatic vessel) is identified in more than 50% of patients. See the image below.

After the sentinel lymph nodes are removed, the lymph node basin is scanned with the gamma probe for residual activity. Absent–to–low-level radioactivity confirms that the sentinel nodes have been removed. The ratio of activity of the sentinel lymph node to nonsentinel lymph nodes should be at least 3:1 in vivo. Generally, the ratio of activity of the ex vivo sentinel lymph node to nonsentinel lymph nodes should be greater than 10:1. At operation, a sentinel lymph node is any blue-stained node, any node with a blue-stained afferent lymphatic, or any node with greater than 10% of the radioactivity ex vivo of the hottest node.

The "10% rule" was generated from the Sunbelt Melanoma Trial experience and suggests that any lymph node with greater than 10% of the ex vivo radioactivity of the hottest lymph node should be removed.^[34] Ex vivo counts eliminate the contribution of shine-through from the primary site or from sentinel nodes remaining in the lymph node basin. This method is reportedly associated with a detection failure rate of approximately 2%. Any clinically suspicious, enlarged, firm, or pigmented lymph nodes should also be removed because tumor cells blocking afferent lymphatics may prevent uptake of radioactive tracer and/or blue dye into a tumor-containing sentinel lymph node.

If, despite these techniques, no sentinel lymph node is identified, either dissection of the lymph node basin or termination of the procedure and close clinical follow-up of the patient may be chosen. This decision should be based on a preoperative discussion with the patient and the metastatic risk associated with the primary tumor. Sentinel node identification rates can approach 99% when the procedure is performed by a trained multidisciplinary team.

Histopathology

Processing the sentinel lymph node involves taking step-sections at multiple levels and performing immunohistochemical staining with S-100 protein and homatropine methylbromide (HMB-45) to identify micrometastatic disease.^[10] If any question exists about abnormal cells on these first sections, additional sections of the lymph node are taken. Immunohistochemistry identifies an additional 10-20% of patients with positive sentinel lymph nodes, in whom micrometastases are not seen on routine sections stained with hematoxylin and eosin (H&E). Melanoma antigen–recognizing T cells 1 (MART-1 or Melan A) antibody staining also may be performed to help identify occult micrometastatic disease in a sentinel lymph node. At least some of the increased rate of detection of micrometastatic disease is attributable to step-sectioning at multiple levels.

See the image below.

An international consensus group advocates a more formal and detailed sampling of the sentinel lymph node.^[40] Their approach involves bivalving the lymph node, fixing the 2 halves face down, and subsequently sectioning each half into 10 sections, which alternately are used for H&E staining, immunohistochemistry, and molecular staging. The Sydney Melanoma Unit performs 2 H&E sections (1 and 4) and 2 with S-100 and HMB-45 immunohistochemistry (2 and 3) on 4 sequential 5-mm thick sections taken from each block of 3-mm longitudinal slices of the sentinel lymph node. They cite a detection failure rate of less than 1% and an acceptable cost-to-benefit ratio with this method. Recommendations for optimal enhanced pathologic analysis vary, but all include step-sectioning and immunohistochemistry.^{[10, 44]}

Another technique described to increase the accuracy of pathologic examination of the sentinel lymph node is carbon dye mapping of the microanatomy of the lymph node to direct the pathologist to the portion of the node most likely to contain metastatic disease.^[45] The sensitivity of intraoperative frozen-section examination of the sentinel lymph node is disappointingly low (< 50%), although false-positive results are almost never reported. Because of the risk of tissue loss and damage during the frozen-section procedure, most melanoma centers eschew frozen sectioning and rely on permanent sections, except to confirm grossly suggestive metastatic disease.^{[10, 40]}

The potential for understaging patients because of tissue loss after frozen sectioning is underscored by reports of lower than expected rates of sentinel node metastases from institutions where frozen section examinations are routinely performed. Intraoperative imprint cytology has been reported to have a sensitivity of approximately 40% for melanoma patients and better preserves the lymph node for permanent section evaluation.^[46]

Similar protocols are followed for the evaluation of sentinel lymph nodes from other skin cancers. For neuroendocrine carcinoma of the skin, the pancytokeratin antibody AE1/AE3, cytokeratin 20, and chromogranin A immunostaining on permanent sections helps identify micrometastatic disease in the sentinel lymph node.

Reverse-transcriptase polymerase chain reaction

Currently, in clinical studies, molecular staging is being applied to sentinel lymph nodes removed from melanoma patients in an effort to improve the sensitivity of detecting occult metastatic disease. This method uses reverse-transcriptase polymerase chain reaction (RT-PCR) to detect messenger RNA for markers frequently expressed by melanoma cells. The sensitivity of RT-PCR is estimated to be an order of magnitude greater than that of immunohistochemistry. The marker used most commonly is tyrosinase, the enzyme that governs tyrosine metabolism in pigment cells. RT-PCR can also be used to detect other melanoma surface antigens such as Gp-100, MART-1, melanoma-associated antigen 3, and melanoma inhibitory activity.

Studies suggest that the 2-year recurrence rate for patients with a histologically (including immunohistochemistry) negative and PCR-negative sentinel lymph node is 2-9%. The 2-year recurrence rate is 13-30% for patients with a histologically negative yet PCR-positive lymph node and 60-67% for patients with lymph nodes positive by both methods.^{[47, 48]} These data suggest that lymph node metastases missed by pathologic assessment are clinically relevant. Other studies have not found molecular staging of sentinel lymph nodes to have prognostic significance.^[49] The ongoing Sunbelt Melanoma Trial and Florida Melanoma Trial both use this technology and should provide useful information on the clinical relevance of molecular staging.

Completion regional lymphadenectomy

Complete regional lymphadenectomy should be performed in all patients with positive sentinel lymph node results, even when the volume of disease is quite small, in the absence of participation in a randomized clinical trial. As yet, no method reliably predicts which patients will have residual metastatic disease in other nonsentinel lymph nodes in the regional basin. While, in most series, 70-80% of patients (range 58-92%) have no further disease identified, the pathologic examination after regional lymph node dissection is, of necessity, less rigorous than that used for evaluation of the sentinel lymph node; therefore, these numbers probably underestimate the true incidence of additional nodal disease.

Several investigators have identified some correlation between various assessments of tumor burden (size, number, and/or location) within the sentinel lymph node or nodes and the likelihood of nonsentinel lymph node metastases on completion lymph node dissection.^{[10, 50, 51]} Others studies have not found such an association,^{[52, 53]} and numerous studies have failed to show any reliable and reproducible way of prospectively identifying those patients at an acceptably low risk of harboring residual disease in the nodal basin so as to forgo completion lymphadenectomy.^[10]

One population-based data analysis (National Cancer Institute's Surveillance, Epidemiology and End Results [SEER] database 1998 to 2001) showed that less than 70% of US melanoma patients with a positive sentinel lymph node receive a completion lymph node dissection.^[36] Another, using National Cancer Data Base data from 2004-2005, found that only 50% of melanoma patients with a positive sentinel lymph node biopsy underwent completion lymph node dissection.^[54] Because melanoma remains a disease in which appropriate surgical treatment is of paramount importance (ie, because of the paucity of highly effective systemic therapy options), the consequences of inadequate treatment may be grave.

A complete axillary lymph node dissection should include all level I, II, and III nodes and should remove all lymph nodes and fatty and areolar tissue from within the anatomic boundaries of the axilla. Usually, these specimens include a minimum of 15-20 lymph nodes. This is not the operation commonly performed for breast cancer staging and underscores the importance of including a trained surgical oncologist with experience in the treatment of melanoma as a member of the multidisciplinary team.

The appropriate extent of groin dissection remains controversial. Isolated pelvic, obturator, or iliac nodal disease may be seen without involvement of (superficial) inguinal nodes, but this is uncommon. After a positive inguinal sentinel lymph biopsy result, the approach may be to perform only a superficial groin dissection at all times, a radical groin dissection at all times, or the deep groin dissection selectively, depending on the number of involved superficial nodes, the status of the node of Cloquet, or the site of the primary tumor (extremity vs trunk).

Two studies, the Multicenter Sentinel Lymphadenectomy Trial (MSLT) II and part II of the Florida Melanoma Trial are currently enrolling subjects after a positive sentinel lymph node biopsy result to be randomized to completion lymphadenectomy or no further surgical treatment. The MSLT-II aims to accrue 4200 subjects with primary melanomas greater or equal to 1.2 mm (or Clark level IV/V or ulceration) to undergo lymphatic mapping (with blue dye, carbon dye, and radiocolloid) with sentinel lymphadenectomy. Those with a positive sentinel lymph node (by H&E, immunohistochemistry, or RT-PCR) will be randomized to completion lymph node dissection or observation (including follow-up nodal ultrasonography) with therapeutic lymph node dissection for relapse.

Part II of the Florida Melanoma Trial seeks to randomize subjects with positive sentinel nodes by H&E histology, immunohistochemistry, or PCR to completion lymph node dissection plus adjuvant therapy with interferon alfa-2b or interferon alfa-2b alone. The study will obtain candidates for randomization from a planned accrual of 3200 clinical stage I and II patients with melanomas thicker than 0.75 mm undergoing lymphoscintigraphy, lymphatic mapping, and sentinel lymph node biopsy from 10 institutions (part I of the study).

Whether or not small foci of disease in a sentinel lymph node (variably defined as < 0.2 mm, < 0.1 mm, or isolated tumor cells) is clinically significant and should be treated the same as other positive sentinel lymph nodes is a subject of great current interest. In some studies, isolated tumor cells in a sentinel lymph node have prognostic significance and are associated with additional melanoma-positive lymph nodes upon completion lymph node dissection in greater than 10% of patients, while other investigations have found no prognostic benefit for sentinel lymph node evaluation beyond routine H & E techniques.^{[49, 55]} A European Organisation for Research and Treatment of Cancer (EORTC) Melanoma Group trial scheduled to be activated in 2009, the MINITUB trial, will evaluate the role of completion lymph node dissection for melanoma patients with "sub-micrometastatic" disease (defined as < 0.1 mm) in the sentinel lymph node.

Postoperative care

In general, patients who undergo sentinel lymph node biopsy alone require no special postoperative care. They may be discharged the day of the procedure, depending on the nature of the treatment of the primary melanoma. Instructions and written materials on lymphedema prevention and precautions should be provided to patients. Extremity elevation and ACE bandage wraps may be used to prevent lymphedema in the initial treatment of patients who undergo lower extremity sentinel lymph node biopsy. These patients are usually fitted with a compression stocking 4-6 weeks postoperatively.

Adjuvant therapy

Systemic treatment with high-dose interferon alfa-2b or participation in a clinical trial is generally offered to stage III patients. Adjuvant radiotherapy is recommended for most patients with multiple cervical lymph node metastases and for those patients with bulky nodal metastases or extracapsular extension at any site.

Appropriate follow-up monitoring of patients with intermediate-thickness and deep primary cutaneous melanoma after sentinel lymph node biopsy, and treatment in general, remains controversial. Advantages to a regular follow-up program include a detailed history and physical examination (with total skin and regional lymph node examinations), with particular focus on the site of the primary melanoma and regional lymph node basin(s). Follow-up monitoring may detect second primary melanomas (approximately 5% of patients at 5 y) and recurrent disease at an early stage, when it may be treated more easily. Follow-up care also may include psychosocial support, identification of familial melanoma, and ongoing patient education.

Typically, examinations are performed at 3- to 6-month intervals for 5 years and annually thereafter. Usually, chest radiography and liver function tests with a lactate dehydrogenase level are performed annually, although most clinicians recognize that these tests are insensitive. Lifelong surveillance is recommended, not only to detect new primary skin cancers, but also to detect late recurrences (>10 y post diagnosis). These recommendations are typical for stage I or II melanomas thicker than 1 mm or for Clark level IV or V melanomas and for patients with stage III melanoma. Similar guidelines are used for follow-up monitoring of patients with neuroendocrine carcinoma of the skin and other types of skin cancer with some propensity for local, regional, and distant recurrence.

For excellent patient education resources, visit eMedicine's Cancer and Tumors Center. Also, see eMedicine's patient education articles Skin Cancer and Skin Biopsy.

Although sentinel lymph node dissection with selective lymphadenectomy is less morbid than elective lymph node dissection, as with any invasive procedure, complications do occur. In the Sunbelt Melanoma Trial, the overall complication rate was less than 5% for sentinel lymph node biopsy alone and 23% for sentinel lymph node biopsy with completion lymph node dissection for metastatic disease. The mortality rate was 0% for both procedures.^[34] Complications occur more frequently in patients with comorbid habits or illness, such as cardiac disease, obesity, diabetes mellitus, and cigarette smoking. The complication rate has also been reported to increase with the number of sentinel nodes removed.

Immediate complications of the procedure include failure to identify a sentinel lymph node (generally, < 5%), anaphylaxis or other allergic reactions to the intradermal injection of blue dye (< 1%), and bleeding. The immediate false-negative rate is defined as the proportion of nodal basins with positive nonsentinel lymph node results after negative results are obtained from sentinel lymph node biopsy immediately followed by elective completion lymph node dissection. In published series that include initial cases early on in the learning curve, the immediate false-negative rate ranged from 4.7-8.3%, with an associated accuracy of 98-99%.^{[14, 56, 57]} Further studies of this nature are unlikely, given the acceptance of sentinel lymph node biopsy with selective lymph node dissection.

Short-term postoperative complications of the procedure include hematoma, wound infection, seroma, and flap necrosis. In the Sunbelt Melanoma Trial, hematoma or seroma (2%) and wound infection (1%) were the most common complications of sentinel lymph node biopsy.

Long-term postoperative sequelae may include persistent blue discoloration of the skin at the injection site (for >30 d in < 10% of patients), lymphatic fistulae, lymphocele, lymphedema (generally < 2% and 0.66% in 1 large trial), and neurologic complications, including transient or persistent neurapraxia, cutaneous anesthesia, paresthesias, and neuropathy. Sentinel lymph node biopsy does not increase the risk of in-transit metastases.

Reports on the risk of locoregional recurrence with follow-up ranging from 13-60 months in 14 published series show that the risk of isolated nodal relapse in the mapped basin after a negative sentinel lymph node biopsy is 1-6%, while the locoregional relapse rate ranges from 5-10%. These data include varying numbers of patients with thin or otherwise low-risk melanomas expected to have minimal risk of relapse. A 2008 study found that the rate of locoregional disease alone as the first site of relapse was 33% in patients treated in a lymph node–noncompliant fashion versus 6% in patients who appropriately underwent sentinel lymph node biopsy followed by completion lymph node dissection for a positive sentinel lymph node.^[38] Historic (presentinel lymph node biopsy era) rates of melanoma relapse in a previously dissected nodal basin (including cases with clinically apparent lymph node metastases) are in the range of 3-5%.^[58]

Contemporary series of sentinel lymph node–negative melanoma patients with greater than or equal to 500 such patients followed for a median of 16-60 months report a failure rate (nodal recurrence in previously mapped basin in the absence of new in-transit or systemic disease or first recurrence in any regional nodal basin) of 1.5-6%. This corresponds to a false-negative rate (false-negative cases divided by true-positive plus false-negative cases) of 7-24%.^{[7, 9, 59, 45, 60, 61, 62, 63]} Data from the MSLT I trial suggest that the false-negative rate declines after 25 cases at an institution.^[64]

Potential technical reasons for failure include errors in surgical technique; failure of nuclear medicine mapping; alterations in lymphatic drainage by inflammation, infection, or previous surgery; and errors in the pathologic examination of a correctly identified true sentinel node.^[65] Reexamination of negative findings from sentinel lymph nodes from patients with regional nodal relapse using careful step-sectioning and immunohistochemistry has been demonstrated to detect micrometastatic disease in 40-80% of patients. Conversely, nodal relapse may result from newly developed nodal metastases arising from clinically occult or newly apparent local or in-transit disease. Most (30-60%) relapses in sentinel node–negative patients are distant metastasis without any evidence of local, in-transit, or regional disease.

The now-widespread use of sentinel lymph node biopsy to more accurately stage melanoma has provided valuable information toward a better understanding of the natural history of malignant melanoma. Significant stage migration has been seen as more patients previously classified as node-negative are reclassified as having stage III disease.

The latest AJCC staging system for melanoma, adopted officially in 2003, arose from the work of the AJCC Melanoma Staging Committee and was developed after evaluation of 17,600 patients from merged prospective databases from 13 cancer centers.^[11] The importance of microstaging using sentinel lymph node biopsy with selective lymph node dissection is emphasized by its inclusion in the pathological N staging of melanoma patients. From this data set, 5-year survival rates for stage III patients stratified by number and type (microscopic, ie, identified by sentinel lymph node biopsy or elective lymph node dissection, without extracapsular extension, and clinically and/or radiologically occult versus macroscopic) and primary tumor ulceration range from 69% (nonulcerated primary tumor, one microscopic positive lymph node) to 13% (ulcerated primary tumor, >3 macroscopic positive lymph nodes). This compares favorably with historic studies.

In a large series of patients who underwent regional lymph node dissection without sentinel lymph node biopsy, 5-year survival estimates ranged from 46-60% for 1 positive lymph node, 33-54% for 2-3 positive lymph nodes, and 21-44% for 4 or more positive lymph nodes, depending on the extent and completeness of the regional lymph node dissection.^[66] In other series, the 10-year survival rate is estimated to be as high as 60-70% for patients with nonpalpable micrometastatic disease in just one lymph node, while the 10-year survival rate for patients with 2-3 occultly positive lymph nodes is approximately the same as in patients with one macroscopically positive lymph node.^[67]

A 2005 population-based study of 41,417 patients from the National Cancer Institute's SEER cancer registry validated the AJCC melanoma staging system, but it revealed more favorable survival rates.^[23] In this study, the 5-year survival rate for N2b (2-3 macroscopically positive nodes) with a nonulcerated primary tumor was 50%, which dropped to 27% for patients with more than 3 macroscopically positive nodes with an ulcerated primary tumor. The SEER data for microscopically node–positive patients are, as yet, immature, but the 2-year survival rate for N1a and N2a patients ranged from 75-90%. Further data analysis, with long-term follow-up monitoring of patients undergoing selective lymph node dissection after sentinel lymph node biopsy with positive results, and data from ongoing clinical trials undoubtedly will provide further prognostic information over the next few years.

Interim results of the MSLT I trial show that while overall survival for intermediate-thickness patients randomized to wide local excision and sentinel lymph node biopsy with selective lymph node dissection versus observation is not yet statistically significant (93% vs 90% at 5 y), disease-free survival was significantly better for patients undergoing sentinel lymph node biopsy with selective lymph node dissection (78% vs 73% at 5 y, P = .009).^[9] In addition, 5-year overall survival was significantly better for patients with positive sentinel lymph nodes who underwent immediate-completion lymph node dissection versus those who underwent therapeutic lymph node dissection for recurrence after wide excision alone (72% vs 52%, P = .004). Patients in the latter group had a greater number of involved lymph nodes (3.3 vs 1.4 for the sentinel lymph node–positive group) and 26% had N3 disease.

In addition, sentinel lymph node status had prognostic significance, with 90% 5-year survival for sentinel node–negative patients versus 72% for sentinel node–positive patients (P < .001). Overall survival was identical for patients with a negative sentinel lymph node biopsy result and patients treated by wide excision and observation without recurrence. The proportion of sentinel lymph node–positive patients was 19.8%, and the proportion with nodal recurrence after wide excision alone was 20.3% at 7 years, further validating the sentinel lymph node concept. The data analysis includes the false-negative sentinel node patients. Sentinel lymph node biopsy did not increase the risk of developing in-transit disease.

The current standard of care is to offer adjuvant treatment with high-dose interferon alfa-2b to stage III melanoma patients or to offer them participation in a clinical trial. The Eastern Cooperative Oncology Group (ECOG) Trial 1684 compared high-dose adjuvant interferon alfa-2b therapy for 1 year with observation in patients with lymph node–positive results (89%) or thick primary (>4 mm, 11%) malignant melanoma. A survival benefit was seen in the treated patients. This regimen subsequently received US Food and Drug Administration approval as adjuvant therapy for high-risk patients with melanoma.

Subsequently, a pooled analysis of ECOG trials of adjuvant high-dose interferon alfa-2b for stage IIB and III melanoma patients has shown consistent improvement in disease-free survival but minimal improvement in overall survival in only 2 of 4 studies.

Further analysis of the data with longer-term follow-up observation is anticipated. Identification of patients who will benefit from adjuvant interferon alfa-2b therapy remains an active area of investigation.^[68] For example, prolonged adjuvant treatment with pegylated interferon appears to be of greatest benefit to patients with micrometastatic melanoma.^[69] Proper stratification of patients within clinical trials, using accurate staging of disease after sentinel lymph node biopsy with selective lymphadenectomy, should help determine which patients derive the greatest benefit from adjuvant treatment, regardless of whether they receive interferon alfa-2b or melanoma vaccines, with or without additional biologic response modifiers or chemotherapy.

While several studies suggest a benefit to early eradication of clinically occult regional lymph node metastases from malignant melanoma, no proven therapeutic benefit for sentinel lymph node biopsy with selective lymphadenectomy, in terms of overall survival for all such treated patients, has been documented. In addition, while the short-term risk of regional recurrence after negative biopsy results from a sentinel lymph node appears to be low, long-term data are lacking. Two randomized clinical trials, MSLT I and the Sunbelt Melanoma Trial, may resolve these issues within the next few years.

The Sunbelt Melanoma Trial will also examine the usefulness of RT-PCR detection of tyrosinase and 1 of 3 other markers in the sentinel lymph node as prognostic markers, including treatment stratification for patients with RT-PCR–positive results in whom lymph node biopsy results appear negative after routine histology and immunohistochemistry. Two newer trials, MSLT II and part II of the Florida Melanoma Trial (mentioned in greater detail above) are currently enrolling patients after a positive sentinel lymph node biopsy result, including those found to be node positive by PCR only, to be randomized to completion lymphadenectomy or no further surgical treatment. The results of these trials should be decisive in determining the proper role of lymphatic mapping and sentinel lymphadenectomy and the value of molecular staging of sentinel lymph nodes.

Directions for the future

In addition to the questions that will be answered by the trials referenced above, other areas of investigation are ongoing in the area of melanoma progression, staging, and treatment.

High-frequency ultrasonography, with FNA of suggestive lymph nodes, is being investigated as a tool to identify or exclude nodal disease in melanoma patients. To date, the sensitivity and positive predictive value is low, but these may improve with technological advances. Ultrasonography is also being investigated as a tool to identify sentinel lymph nodes and to follow patients postoperatively for the development of new nodal disease.^[70] Fluorodeoxyglucose positron-emission tomography scanning and positron-emission tomography scanning are being used increasingly in melanoma patients, but these are generally insensitive for detecting early nodal disease. False-positive results have been reported. Further technological refinements may decrease the size of the current detection limit of approximately 5 mm.

Proton magnetic resonance spectroscopy of FNA samples of sentinel lymph nodes has been reported.^[71] Evaluation of sentinel lymph nodes removed surgically had a sensitivity of 88%, specificity of 90%, and accuracy of 89% in predicting sentinel lymph node status. These investigators propose a nonsurgical sentinel lymph node evaluation using ultrasound-guided FNA with this technique, without the need for histopathological evaluation.

Biologic and genetic alterations in primary melanomas and lymph node metastases continue to be studied. B-raf and N-ras mutations; differential integrin expression; and the roles of other angiogenesis, cell motility, and invasion genes are under active investigation. This may prove to have clinical relevance (eg, specific antibody therapy) in the future. Preoperative molecular characterization of melanomas may lead to even less invasive or noninvasive staging and treatment in selected patients.

Conclusions

The use of sentinel lymph node biopsy with selective lymphadenectomy has increased the identification rate of micrometastatic disease and decreased the morbidity of proper treatment of melanoma patients. Sentinel lymph node status has been shown repeatedly to be the most powerful indicator of prognosis. Sentinel lymph node biopsy identifies patients for adjuvant therapy and stratifies patients into more homogeneous groups for inclusion in clinical trials. Sentinel lymph node biopsy identifies patients who need early therapeutic lymph node dissection.

Patients who undergo sentinel lymph node biopsy with selective lymph node dissection have an improved disease free-survival rate and improved regional disease control compared with those who undergo wide local excision of the primary tumor and observation alone. This technique may also result in future refinements of the staging system for malignant melanoma and greater understanding of the natural history and progression of melanoma. It also may be proven to result in an overall survival benefit for patients undergoing the procedure.