Pathologic examination of sentinel lymph nodes from melanoma patients

Seminars in Diagnostic Pathology (2008) 25, 100-111

Pathologic examination of sentinel lymph nodes from melanoma patients Richard A. Scolyer, MD, FRCPA, FRCPath,a,b,c Rajmohan Murali, MBBS, FRCPA,a,b,c Stanley W. McCarthy, MBBS, FRCPA,a,b,c John F. Thompson, MD, FRACS, FACSa,d From the aSydney Melanoma Unit, Royal Prince Alfred Hospital, Camperdown, NSW, Australia; b Department of Anatomical Pathology, Royal Prince Alfred Hospital, Camperdown, NSW, Australia; c Discipline of Pathology, The University of Sydney, Sydney, NSW, Australia; and the d Discipline of Surgery, The University of Sydney, Sydney, NSW, Australia. KEYWORDS Sentinel node; Melanoma; Pathology; Protocol; Accuracy; False negative

In melanoma patients, the sentinel node biopsy (SNB) procedure is a highly accurate staging method, and the tumor-harboring status of the sentinel node (SN) is the most important prognostic factor for patients with early stage disease. For the SN to provide accurate prognostic information, however, it is essential that all “true” SNs are removed and examined diligently. Pathologists should examine multiple hematoxylin– eosin and immunohistochemically stained sections from each SN, but it is unclear from the currently available evidence what is the most appropriate sectioning and staining protocol. Relevant factors to consider include the accuracy of the procedure, the time, labor, and costs involved, and clinical follow-up data which are likely to vary between institutions; hence, individual protocols should be developed locally by pathologists in consultation with their surgical colleagues. At the Sydney Melanoma Unit, four sequential sections of both halves of each SN are examined. The first and fourth sections are stained with hematoxylin– eosin, the second section is stained for S-100 protein, and the third section is stained for HMB-45. Pathologists should not only identify the presence of melanoma metastases within the SN, but also record the size of the largest metastatic focus, tumor penetrative depth (measured from the inner margin of the node capsule to the deepest tumor cell within the SN), and the percentage nodal cross-sectional area involved (as measured on the slides). Potential diagnostic pitfalls in SN evaluation include the misinterpretation of nevus cells, macrophages, or antigenpresenting interdigitating dendritic cells as melanoma. Careful assessment of the morphologic characteristics of the cells and their immunohistochemical profile should prevent misdiagnosis. Routine frozen section examination of SNs from melanoma patients is not recommended. The utility of ultrasound to detect SN metastases (confirmed by fine needle biopsy) is currently being investigated. Whereas potentially this may avoid the need for formal sentinel lymphadenectomy and histopathologic evaluation in some patients, the lack of sensitivity of currently available ultrasound technologies to detect the small micrometastases (⬍2 mm in diameter), that are typically present in most melanoma patients with a positive SN, limits its current role. In the future, other techniques, such as the use of carbon particles or antimony analysis, may better localize the site of metastases within SNs and permit more focused and efficient pathologic examination of SNs. At present, the role of nonhistopathologic methods of SN evaluation, such as reverse transcription polymerase chain reaction (RT-PCR) and magnetic resonance spectroscopy, remains unclear, and these techniques require further evaluation. © 2008 Elsevier Inc. All rights reserved.

Address reprint requests and correspondence: Richard A. Scolyer, MD, FRCPA, FRCPath, Department of Anatomical Pathology, Royal Prince Alfred Hospital, Missenden Road, Camperdown, NSW 2050, Australia. E-mail: [email protected].

0740-2570/$ -see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1053/j.semdp.2008.04.002

Scolyer et al.

Histopathologic Examination of Sentinel Lymph Nodes

The sentinel node biopsy (SNB) procedure in melanoma patients is a highly accurate staging method, and the tumorharboring status of the sentinel node (SN) represents with early stage disease the most important prognostic factor for melanoma patients.1–11 Before the modern SNB era pioneered by Drs. Morton, Cochran, and colleagues in the 1980s and early 1990s,12 patients with clinically localized melanoma considered at immediate or high risk of metastases were offered a “prophylactic” regional node field clearance in many melanoma treatment centers, with the aim of removing occult metastatic disease.13 However, as only about 20% of these patients had regional node field involvement, the remaining 80% of patients were placed at unnecessary risk of anesthetic complications and surgical morbidity. By identifying, removing, and carefully examining only those lymph nodes receiving direct drainage from the primary site (ie, the SNs), it became possible to accurately determine the tumor-harboring status of the entire node field with a minimally invasive technique.8,9,12 In the absence of clinically detected metastatic disease in regional nodes, complete lymph node dissection (CLND) is now restricted in most melanoma treatment centers to those patients with demonstrated metastatic disease in SNs, hence sparing other patients this major surgery and its inherent risks and complications. There is preliminary clinical trial evidence suggesting that early CLND in patients who are SN-positive improves their survival outcome.14 Determination by pathologists of the presence or absence of metastatic melanoma in SNs is critical to providing patients with an accurate estimate of prognosis and identifying individuals who may benefit from immediate CLND. To provide such information, it is therefore essential that all “true” SNs are removed and examined thoroughly. The methods and protocols for receiving, handling, assessing, examining, and reporting SNs and potential pitfalls in pathologic evaluation are therefore pertinent issues for pathologists to consider and are the subject of this article.

Preparing sentinel nodes for pathologic examination: what the pathologist should request of the surgeon SNs should be removed intact, preferably with a thin rim of surrounding adipose tissue and be devoid of crush or diathermy artifacts that may complicate pathologic assessment. Surgeons may apply marking sutures to the SN to indicate the site of entry of afferent lymphatic channels (because this region of the SN is the most likely site to harbor a metastasis)15 or to highlight particular areas of concern. The pathology request form should indicate the number of removed SNs and their anatomic locations, as well as the significance and orientation of any marking sutures. Any “second tier” lymph nodes or non-SNs that have also been removed should be indicated as such on the request form and the specimens clearly labeled.

101

In most melanoma patients, only one SN is present; there are sometimes two and occasionally three, but rarely more. Surgeons should be made aware that the strength of the SNB technique is that it allows focused and detailed pathologic examination of the most metastasis-susceptible lymph nodes (ie, the SNs). By harvesting multiple lymph nodes, some of which are not true SNs, and requesting that their pathology colleagues perform the special and detailed techniques of SN examination on them, some surgeons are placing an inappropriate and unnecessary workload and financial burden on pathologists and pathology departments. SNs should be sent to the pathology laboratory either fresh or in an appropriate fixative (such as 10% buffered formalin). Use of tissue for research purposes should not be allowed to compromise subsequent pathologic assessment of the SN. Frozen section evaluation of SNs from melanoma patients is not recommended for several reasons: such evaluation may reduce the accuracy of the procedure because of loss of diagnostic tissue as a result of the technical manipulation associated with frozen section examination and because of difficulty interpreting tissues that have been previously frozen; the poor sensitivity (29-82%) of frozen section examination; inefficiency in terms of time (operating theater, pathologist, and laboratory); and local logistical issues.16 The topic of frozen section examination of SNs is discussed in another article in this issue of the journal.

Macroscopic examination of sentinel nodes Following 12 to 24 hours of fixation, each SN specimen is dissected. The size of the specimen, the number of lymph nodes present, and the location and orientation of any marking sutures should be recorded. Usually there is only one lymph node in each specimen; however, if more than more one putative SN is present, each should be handled separately as a SN. Excessive adipose tissue surrounding the SN that may hinder optimal tissue processing should be removed, although preservation of a small amount of fat allows assessment of perinodal lymphatic vessels that may contain tumor metastases. The presence or absence of blue coloration of the node is noted, and the dimensions of the SN are recorded. The location of any marking suture should be inked to allow identification of this region of the node microscopically. The levels of radioactivity in SNs are well below recommended safety limits at the time of SNB and are even lower when SNs are examined in the pathology laboratory.17–19 Hence, examination of SN specimens poses no significant radiation risk to pathology personnel. Most authorities recommend that SNs from melanoma patients be initially bisected through their longest meridian in the hilar plane (Figure 1). For larger nodes (⬎10 mm in thickness), additional parallel slices may be made. As well as allowing efficient examination of the largest portion of the SN, this recommendation is supported by the results of studies by Cochran and colleagues, which suggested that most SN me-

102

Figure 1 A sentinel node bisected through the central meridian in the hilar plane.

tastases are small and subcapsular and are located close to the central meridian.20 However, some more recent studies have highlighted the occurrence of nodal metastases in other locations within SNs (see below).21–27 Macroscopic examination of the cut surfaces of the SN should be performed carefully looking for evidence of tumor, and any abnormality identified should be described in the pathology report.

Pathologic sampling of sentinel nodes for microscopic examination Pathologists should examine multiple sections from each half of the SN, including sections stained with hematoxylin– eosin (HE) and sections stained immunohistochemically for melanoma-associated antigens. It is unclear from the currently available evidence what is the most appropriate number of sections to examine, at what levels these should be cut from the tissue blocks, and the most appropriate number and combinations of immunostains that should be assessed. Various histopathologic sectioning protocols have been suggested for optimal examination of SNs in melanoma patients. These have ranged from the examination of a single HE-stained slide from each paraffin tissue block of the SN to the examination of serial sections through the entire SN.22,24 –29 Many studies have shown that immunohistochemistry can increase the detection rate of melanoma cells in SNs.22,23,25,26 What are the practical implications of serially sectioning and examining an entire SN? If it is assumed that each paraffin block includes a piece of tissue that is 3 mm in thickness (an accurate estimate in our experience) and that the paraffin-embedded tissue is serially sectioned into tissue slices 5 ␮m in thickness (the standard thickness for cutting sections from paraffin blocks for histopathologic examination), every paraffin block of each SN from the patient will

Seminars in Diagnostic Pathology, Vol 25, No 2, May 2008 yield 600 sections for microscopic examination. If it takes the pathologist 1 minute to examine each slide (a conservative estimate) and there is a mean of 2 paraffin blocks per SN and a mean of 2 SNs per patient (accurate assumptions based on our experience), then it would take 40 hours (600 ⫻ 1 ⫻ 2 ⫻ 2 minute ⫽ 2400 minutes) to examine the SNs from each melanoma patient. Thus, it is clearly impractical and prohibitively expensive to serially section and examine SNs in their entirety. The challenge is to identify that histopathologic examination protocol which optimally balances the accuracy and clinical value of the result against the labor, time, and costs involved. When considering histopathologic sampling of SNs and attempting to determine the optimal sectioning protocol, there are a number of questions that must be considered. These include: 1. Does increased sampling of SNs detect more melanoma metastases? 2. What is the biologic significance and clinical relevance of metastatic deposits detected by more extensive sampling protocols? 3. Are there nonhistopathologic causes for false-negative SNs? 4. Can techniques be used to better localize the site of metastases within the SN, thereby allowing the development of more efficient methods for the detection of SN metastases? 5. What is the role of nonhistopathologic methods in SN evaluation?

Sentinel node sampling The most commonly used method for pathologic examination of SNs is that described by Cochran and colleagues.29 This involves bisecting SNs through their longest dimension in the central meridian through the hilar plane and cutting 10 full face sections from each face of the bivalved SN following “facing up” (ie, trimming of the paraffin block until a section that includes all areas of tissue is obtained). At the Sydney Melanoma Unit (SMU), the pathologic protocol used for the examination of SNs is a modification of the Cochran method. All SNs are sliced initially through the central meridian and hilar plane (and subsequently into 3-mm slices if the halves are ⬎5 mm in thickness), then embedded in their entirety for pathologic examination. Following “facing up,” 4 sequential sections are cut from each block of every SN. The first and fourth sections are stained with HE, the second section is stained for S-100 protein, and the third section is stained for HMB-45 (Gp100). At the discretion of the reporting pathologist, the procedure may be repeated (a) in melanoma patients with thick primary tumors in whom the risk of melanoma metastasis in the SN is high, (b) if there are changes seen in the original sections which are difficult to interpret and it is uncertain whether the SN contains melanoma metastasis, or (c) if the original

Scolyer et al.

Histopathologic Examination of Sentinel Lymph Nodes

sections are suboptimal. This method has been used for examination of SNs in melanoma patients since SNB was introduced in our institution in the early 1990s. Despite the use of a more limited histopathologic sampling protocol compared with that adopted in some other centers, our results using this technique compare very favorably with those of others. It has been shown clearly that the SN tumor-harboring status is the most important prognostic factor for early stage melanoma patients. Patients with a negative SNB have a disease-specific 5-year survival of 90% compared with 56% for those patients with a positive SN.30 The failure rate of the SNB technique for melanoma patients is 2.6% in our hands, a rate which compares favorably with that of other series which have reported a median failure rate of 4% (range 0% to 6%) (Table 1). Similarly, our false-negative rate of 13.2% compares favorably with a median of 13.6% reported in 18 other series (range 7.0% to 24.8%).14,21–27,31–34 This is despite a median follow-up period of 42 months in our series compared with relatively shorter follow-up (median 32 months) in other reported series. It should be noted that the false-negative rate is calculated by dividing the number of false negatives by the false negatives plus the true positives. In contrast, the failure rate is the number of false negatives divided by the sum of the false and true negatives.

pling protocols. Most of these studies have shown that increased histopathologic sampling of SNs from melanoma patients detects more melanoma, although the extent of the increase in detection rates reported has differed in each of the studies (Table 2).24 In 2003, Cook and coworkers25 reported the detection rate of melanoma metastases within SNs using 3 different histopathologic sampling protocols and compared this with the rate of positivity determined by reverse transcription polymerase chain reaction (RT-PCR) using primers for tyrosinase and Mart-1. Each of the protocols involved bivalving the SN along its longitudinal axis. In the first protocol, 8 serial sections were examined (4 stained with HE, 3 for S-100, and 1 for HMB-45). The authors reported a SN positivity rate of 17.8% (positive SNs were identified in 74 of 416 patients; the mean* (*, median not reported) Breslow thickness of the primary tumors was 2.03 mm). When they prospectively examined their SNs using an extension of this protocol (protocol 2), where an additional 2 levels were cut 50 ␮m apart and 2 sections stained from each level (with HE and for S-100 protein), a SN positivity rate of 25.2% (26 of 103 patients; mean* Breslow thickness of primary tumors 2.16 mm) was detected. Cook and coworkers then examined a third protocol which involved cutting 20 sections and staining 14 of them. The first full-face section was obtained, and then 5 step sections 50 ␮m apart were examined. Fourteen of the sections were stained (6 with HE, 6 for S-100 protein, 1 for HMB-45, and 1 for “pan melanoma”). With this protocol, a SN positivity rate of 33.8% was detected (positive SNs were identified in 25 of 74 patients with a mean* Breslow thickness of 1.77 mm). The authors noted that this approached the RT-PCR positivity rate of

Increased histopathologic sampling of sentinel nodes detects more melanoma metastases A number of studies have assessed the detection rate of metastatic melanoma in SNs using different pathologic samTable 1

103

Failure and false-negative rates after negative sentinel node biopsy in published series

Author

Year

n

Median follow-up (months)

Nodal recurrence*

Failure rate (%)†

False negative rate (%)‡

Gershenwald Essner Gadd Clary Cascinelli Statius-Muller Jansen Harlow Doting Vidal-Sicart Chao Morton Nowecki Yee Morton Nowecki Van Akkooi Caraco Roulin

1998 1999 1999 2000 2000 2000 2000 2001 2002 2002 2002 2003 2003 2005 2006 2006 2006 2007 2008

243 225 89 121 646 204 151 297 150 363 950 1277 579 836 642 979 185 650 253

35 45 23 24 29 42 32 36 47 26 16 § 34 42 60 36 23 43 33

10 11 7 7 40 3 6 10 6 7 14 33 27 22 26 57 6 34 7

4.1 4.8 7.9 5.8 6.2 1.5 4.0 3.4 4.0 1.9 1.5 2.6 4.7 2.6 3.4 5.8 3.2 5.2 2.8

16.0 20.7 § 18.4 24.8 7.0 11.0 20.4 10.0 9.1 7.1 9.2 13.6 13.2 17.6 20.0 7.2 18.4 8.6

*First recurrence only. †Failure rate ⫽ 100 ⫻ false negatives/(false negatives ⫹ true negatives). ‡False-negative rate ⫽ 100 ⫻ false negatives/(false negatives ⫹ true positives). §Not stated.

104

Seminars in Diagnostic Pathology, Vol 25, No 2, May 2008

Table 2 Studies reporting the detection rate of metastatic melanoma in sentinel nodes with differing pathologic sampling protocols are used Year published

Author 24

Cook, et al

2003

Gietema, et al25 Abrahamsen, et al26 Spanknebel, et al27

2004 2004 2005

Protocol no.

No. of cases

No. of sections examined (HE/IHC)

1 2 3

416 103 74 189 100 49

8 (4/4) 12 (6/6) 14 (6/8) 18 (5/13) 20** (8/12) 62 (22/40)

% of patients with ⫹SN

Median Breslow thickness of primary melanomas

17.8 25.2 33.8 15.3 28 61

2.03* 2.16* 1.77* 2.78* 1.56 3.0

Abbreviations: No., number; HE, hematoxylin and eosin; IHC, immunohistochemical; ⫹, positive; *, mean Breslow thickness (median not reported); **, median number of sections; SN, sentinel node.

39.3% (in 28 patients). A modified version of the third protocol has since been adopted by the European Organization for Research and Treatment of Cancer (EORTC) as necessary for EORTC clinical trials involving SNB in melanoma patients. In 2004, Gietema and coworkers26 assessed the detection rate of their histopathologic protocol for examination of SNs in melanoma patients. SNs less than 5 mm in diameter were embedded whole, nodes 5-10 mm in size were bivalved, and nodes greater than 10 mm were sliced into two 5-mm slices. Multiple sections were examined at each of 5 levels, 50 ␮m apart. Four sections were cut at each level and stained with HE, and for S-100, HMB-45, and Melan-A, ie, 18 sections (5⫻ HE and 13⫻ immunohistochemistry). The SN positivity rate was 15.3% (positive SNs were identified in 29 of 189 patients, mean* Breslow thickness 2.78 mm). They noted that 83% of their ultimately SN-positive patients were identified in the sections examined at level 1, 86% when level 2 was included, 90% when level 3 was included, and 97% when level 4 was included. Also in 2004, Abrahamsen and coworkers27 reported the results of their SN sectioning protocol. All SNs were cut in a longitudinal axis and one-half was used for histopathologic examination. The nodes were sliced into 2- to 3-mm slices, and each slice was placed in a separate paraffin tissue block. Each block was cut at 250-␮m intervals to extinction. One HE-stained section was examined at every level, and 3 immunohistochemically stained sections were examined from every second level. The authors examined a median number of 20 sections per SN (range 10 to 59). With this sectioning protocol, a SN positivity rate of 28% was obtained (positive SNs were identified in 28 of 100 patients; median Breslow thickness 1.56 mm). They noted a cumulative increase in their SN positivity ranging from 49% with the sections examined at level 1 to 56% at level 2, to 79% at level 3, to 83% at level 4, to 89% at level 5, to 95% at level 8, and 100% at level 11. Recently, authors from the same institution reported that, by limiting such detailed pathologic analysis to the SNs with the highest gamma counts, workload could be markedly reduced with only a slight increase in the false-negative rate.35

In 2005, Spanknebel and coworkers36 reported the results of their histopathologic protocol. The SNs were sliced in the longitudinal plane, and a single HE-stained slide and a frozen section were cut from each block. Subsequently, the SNs were re-examined. Twenty levels were cut at 50-␮m intervals from each tissue block. Three sections were examined at each level (with HE and for S-100 and HMB45), ie, 62 sections. In this small series of 49 patients, the authors identified metastases within the SNs in 33% of their patients using their initial examination protocol (1⫻ HE and 1⫻ frozen section). With their extended protocol, positive SNs were identified in a remarkable 61% of the patients (30 of 49 patients; median Breslow thickness 3.0 mm). The authors noted an increase in the cumulative percentage of patients with positive SNs from 42% with the first 3 levels, to 75% with the first 10 levels, to 90% with the first 15 levels, to 100% with all levels. In 20 of the patients, the authors mapped out the distribution of the SN metastases and attempted to determine projected rates of detection for different sectioning strategies (varying the distance between sections and number of levels examined). They noted that 71% of the metastases within the SNs of these 20 patients would be detected by examining 3 levels 250 ␮m apart with 3 sections examined at each level (stained with HE and for S-100 and HMB-45 at each level). The authors estimated the cost of this protocol would be $1050 (US dollars) per SN. From the results of these studies, it appears clear that increased histopathologic sampling of SNs increases the rate of detection of melanoma metastases within SNs, although both the detection rate and number of metastases detected by more extensive protocols have varied in the studies performed.

Practical issues of relevance to increased sampling protocols In addition to detection rates, time and economic considerations are of great relevance when considering the adoption of a more extensive sampling and histopathologic examina-

Scolyer et al.

Histopathologic Examination of Sentinel Lymph Nodes

tion protocol for SNs. What are the practical implications of examining 20 sections per block of each SN? At our institution, we perform an average of 4 SN procedures per day on melanoma patients. If 20 sections were examined per block (rather than the 4 sections that are currently examined), using the assumptions referred to previously (1 minute to examine each slide, a mean of 2 paraffin blocks per SN, and a mean of 2 SNs per patient), this would require an additional 64 minutes of pathologist’s time per SN procedure, ie, about 4 additional hours of pathologist’s time per day. Furthermore, laboratory expenses, such as the technicians’ time to cut and stain the sections and the technical costs of materials and reagents, must be considered.37 The current worldwide shortage of pathologists is another issue that could limit the capacity of the pathology workforce to cope with the additional demands of an exhaustive SN examination protocol.

Microscopic examination of sentinel nodes It is a pathologist’s responsibility to examine all sections cut and stained on each SN. We recommend that each slide be scanned at 100⫻ magnification (10⫻ objective and 10⫻ eyepiece). Particular attention should be paid to the subcapsular sinus region of the SN because this is the most likely site to harbor tiny micrometastases.32 Other regions of the SN and perinodal tissue should also be assessed, including the paracortical region, deeper parenchyma, and perinodal lymphatics. High-power magnification (400⫻) is used to confirm the nature of tumor cells. Extranodal spread of tumor is infrequent but should be recorded in the pathology report. Immunohistochemistry is an important aid to histopathologic examination of SNs and, in fact, many pathologists examine the sections stained immunohistochemically before the HE-stained sections. Depending on the methodology used in sectioning, immunostaining and reviewing negative SNs examined initially with single HE slides only by immunohistology, the reported rate of detecting micrometastatic melanoma has varied from 12% (in a study where all negative SNs were reviewed regardless of the follow-up outcome), to a rate of 80% in studies where only patients with subsequent recurrence in that regional node field were evaluated.21,22,38 The most commonly used immunostains for the purpose of detecting melanoma micrometastases are those for S-100 protein and HMB45 (Gp100). S-100 is a more sensitive but less specific marker of melanoma cells than HMB45 or Melan-A/MART-1. Although S-100 stains virtually all melanomas, it is also positive in paracortical antigen-presenting interdigitating cells, macrophages, nevus cells, nerves, and fat, all of which may be present in SNs. HMB45 and Melan-A are more specific markers but may be negative in up to 25% of melanomas. There are a number of potential pitfalls in the microscopic examination of SNs that can usually be resolved

105

through assessment of morphology and immunohistochemistry. Apart from melanoma cells, other cells that may be found within lymph nodes and which are positive for S-100 include macrophages, antigen presenting dendritic cells, nevus cells, and nerves. These can usually be distinguished from melanoma cells on the basis of their size, nuclear and cytoplasmic characteristics, and distribution within the node.37 Nevus cells are found within the capsule or trabecular connective tissue septa of one or more regional lymph nodes in up to 25% of melanoma patients (Figure 2). Melanoma cells are usually larger than nevus cells and are usually located within the subcapsular sinus and lymphoid tissues of the node (Table 3). Morphologic features that may be useful in distinguishing melanoma cells from nevus cells, histiocytes, or interdigitating cells include large cell size, prominent nucleoli, high nuclear to cytoplasmic ratio, mitotic figures, and fine melanin granules (Figure 3). Coarse melanin granules, by comparison, are usually found in macrophages (Figure 4). Immunohistochemical stains also show differences between melanoma cells and other cells that may be confused with them. Nevus cells show positive nuclear and cytoplasmic staining for S-100 and cytoplasmic staining for Melan-A; however, they are usually negative for HMB45 and the proliferation marker Ki67. Interdigitating and nerve cells stain positively for S-100 but are negative for other melanocytic markers. The use of alkaline phosphatase (red) rather than diaminobenzidine (brown) as the chromogen for antibody visualization can reduce confusion between melanin-containing macrophages and melanoma cells. Macrophages are usually negative for HMB45, although macrophages which show positive staining for HMB45 are occasionally observed. Careful attention to morphological details will usually make it possible to distinguish melanoma metastases from other cells within lymph nodes.

Melanoma tumor burden and location in sentinel nodes A number of recent studies have shown that the location and extent of tumor deposits within SNs (Figure 5) correlates with the presence of positive non-SNs in CLND specimens and provides important prognostic information. If there are only a few metastatic cells in the subcapsular sinus of the SN, the prognosis is good and the chance of finding additional metastatic disease in a CLND specimen is extremely small. If, on the other hand, there are multiple large foci of tumor which extend deeply into the central part of the SN, the prognosis is much worse, and the chance of finding metastases in non-SNs in a CLND specimen is high. However, it remains unclear which particular method of assessment of SN tumor burden or other micromorphometric parameters of melanoma deposits in SNs accurately predict the probability of non-SN involvement and prognosis.

106

Figure 2

Seminars in Diagnostic Pathology, Vol 25, No 2, May 2008

Nevus cells within the capsule of a sentinel node (A and B, HE; C, HMB45).

Micromorphometric parameters of SN metastases that have been assessed include the size of metastases, maximum subcapsular depth (also known as centripetal thickness or tumor penetrative depth of the deposits) (Figure 6), the microanatomic location of SN tumor deposits, the percentage cross-sectional area of the SN involved, and the presence of extracapsular spread.39 – 46 However, assessing, classifying, and measuring these parameters can be difficult. For example, the assessment of tumor burden has been performed on the basis of the maximum subcapsular depth, the largest size of the deposits, the cross-sectional area of the deposit(s), or the tumor volume in the SNs, and these values have been estimated by pathologists, using calculations based on micrometer measurements, or calculated with greater precision using computer-assisted morphometry in various studies. Tumor deposits are often irregularly shaped, the limits of tumor deposits can be difficult to discern, and tumor burden is to some degree dependent on sectioning protocols, as more extensive sectioning may reveal additional tumor deposits or demonstrate a greater dimension of deposit(s) in the deeper sections. Yet, although the methods and likely level of precision of mea-

surement vary widely in reported studies, they all show a positive correlation between the SN tumor burden and non-SN positivity/clinical outcome, suggesting that this parameter is an important predictor of these end-points, whose predictive value holds whether it is assessed with crude and imprecise methods or using more sophisticated and precise means. The microanatomic location of tumor in subcapsular locations in SNs is less often associated with non-SN involvement than location in nonsubcapsular locations.40 However, assessment of the precise localization of deposits can be difficult in routine histologic sections, in which the architecture of the SN is not clearly apparent. In a recent multi-institution study evaluating interobserver reproducibility of the assessment of micromorphometric parameters of melanoma deposits in SNs, it was shown that assessment of quantifiable variables (maximum subcapsular depth and the percent cross-sectional area of SN involved by tumor) were highly reproducible.47 However, assessment of microanatomic location and extracapsular spread was less reproducible, casting some doubt on the reproducibility of their predictive value. Clearer definitions and consensus criteria for the evaluation of these

Scolyer et al.

Histopathologic Examination of Sentinel Lymph Nodes

Table 3 Morphologic features useful in distinguishing between nodal nevus cells and metastatic melanoma Feature

Nevus

Melanoma

Location in lymph node

Intracapsular or in intranodal fibrous trabeculae Smaller Condensed Inconspicuous Sparse Rare

Subcapsular sinus or parenchyma

Cell size Nuclear chromatin Nucleoli Cytoplasm Fine intracytoplasmic melanin granules Mitotic figures Immunohistochemistry S-100 HMB45 Melan-A Ki-67 index

Absent ⫹⫹⫹ ⫹⫹⫹ 0%

Larger Often vesicular Prominent Variable Sometimes present Sometimes present ⫹⫹⫹ ⫹-⫹⫹⫹ ⫹⫹⫹ ⬎2%

Abbreviations: ⫹, weakly positive; ⫹⫹⫹, strongly positive; -, negative.

parameters may improve the reproducibility of their assessment.

Are all melanoma metastases within sentinel nodes clinically relevant? Some have recently suggested that very small tumor deposits within the subcapsular sinuses of SNs may not be clinically relevant.48,49 Careful, prolonged follow-up of such cases in prospective clinical trials will be necessary to answer this question. Although some studies have been interpreted as suggesting that very small melanoma metastases within SNs may not be clinically significant, these

Figure 3

107

studies have had short follow-up periods, are likely to be confounded by lead time bias, and furthermore, appear to have discounted the very real possibility of a potential therapeutic benefit resulting from the SNB itself and the subsequent CLND. Further evidence comes from studies in the ipsilateral nodal basin reviewing false-negative SNs following clinical recurrence. Although these studies are again limited by lack of long-term follow-up, the number of clinical recurrences following a negative SNB is much lower than might be expected if all metastases detected by more extensive sectioning protocols of SNs (up to 61% of SNs in one study8) were clinically relevant.50,51 In addition, results of interim analyses of the first Multicenter Selective Lymphadenectomy trial (MSLT-1)14 show that the percentage of patients who had a positive SN (16.0%) was very similar to the percentage of patients who developed nodal recurrence following a wide excision only of the primary melanoma site (15.6%), suggesting that clinically relevant metastases were usually detected using the protocols adopted in this study.64

Nonhistopathologic causes of false-negative sentinel nodes Regional node field recurrence of melanoma following a reportedly negative SNB sometimes occurs (failure rates of up to 7.9% have been reported in some series; see Table 1) as a consequence of errors in lymphatic mapping, the surgical procedure of sentinel lymphadenectomy, and histopathologic assessment. At the SMU, measurement of antimony levels in SNs has been used to identify surgical causes of false-negative SNB. In Australia, technetium99m-labeled antimony sulfide colloid is used for preoperative lymphoscintigraphy. The technetium-99m decays before SNs are examined pathologically. However, antimony is a heavy metal that remains in tissue blocks of SNs

Metastatic melanoma cells within the subcapsular sinus of a sentinel node (A, HE; B, HMB-45).

108

Seminars in Diagnostic Pathology, Vol 25, No 2, May 2008

Nonhistopathologic methods of sentinel node evaluation in melanoma patients

Figure 4 Pigment-laden macrophages in a sentinel node. The presence of coarse melanin pigment granules and nuclear characteristics help to distinguish such cells from metastatic melanoma.

indefinitely. We have found that antimony concentrations in tissue blocks can be measured accurately by inductively coupled plasma mass spectroscopy (ICP-MS), allowing distinction of true SNs from non-SNs.52 When antimony concentrations were measured in 20 false-negative SNs, in 5 cases, the antimony levels were very low, suggesting that the removed nodes were not “true” SNs.37,52 Comprehensive analysis of our false-negative SNs, including more detailed pathologic examination and review of lymphoscintigraphy, suggested that the cause of the falsely negative SN was due (in approximately equal measure) to a deficiency in nuclear medicine, surgery or pathology.29,53–55

Other methods that are potentially more cost efficient and at least as accurate as histopathologic examination for the evaluation of SNs have been studied. Molecular methods (including RT-PCR to detect mRNA for melanoma-associated markers) suggest that submicroscopic disease may be present in 28% to 41% of patients with pathologically negative SNs, but there is continuing debate about which markers should be used. Early studies used tyrosinase, but it now appears that this is not a good marker, and in any case, multiple positive markers are probably required before a patient is categorized as “SN positive” on the basis of an RT-PCR result. It has been shown that ultra-staging of SNs using RT-PCR correlates with patient outcome.59 However, the role of RT-PCR in routine SN evaluation remains unclear. The clinical relevance of histopathologically negative but RT-PCR-positive SNs (as assessed by 5 different markers) is being evaluated in a prospective randomized clinical trial, the second Multicenter Selective Lymphadenectomy Trial (MSLT-2). Studies at the SMU using magnetic resonance spectroscopy (MRS) analysis of fine needle biopsy (FNB) specimens obtained from a SN have shown that this technique can distinguish melanoma-containing SNs from negative SNs. A particularly useful clinical application of MRS would be to examine FNBs of SNs that remain in situ (after they have been identified by lymphoscintigraphy and accurately localized with high resolution ultrasound so that there can be certainty that the correct node is being biopsied). It has also been demonstrated that it is possible to use ICP-MS on the same specimens to identify the antimony used for the preoperative lymphoscintigram, and thereby confirm that

Techniques to better localize sites of melanoma metastases within sentinel nodes If it was possible to better localize the anatomic site of micrometastases within SNs, more concentrated examination and sectioning could be focused on the area most likely to harbor metastases, thereby improving the efficiency of pathologic examination of SNs. Investigators from the John Wayne Cancer Institute have studied the use of carbon particles injected with the blue dye into the primary melanoma site immediately before the SNB procedure. They found that the carbon particles entered the SN via the afferent lymphatic and were deposited in this region of the node. As predicted, tumor deposits were located in the same region of the SN.56 –58 Difficulty in obtaining suitable carbon particles has to date limited the examination of this technique on a larger scale (personal communication from Dr. Donald L. Morton). At the SMU, we are investigating whether antimony within SNs can be used in a similar fashion to localize the site of SN metastases.

Figure 5 The tumor penetrative depth of a deposit of metastatic melanoma in a sentinel node is measured from the inner margin of the nodal capsule to the deepest cell within the node. This and other surrogate measurements of tumor burden within sentinel nodes correlate with the risk of patients having positive nonsentinel nodes and their clinical outcome.

Scolyer et al.

Figure 6

Histopathologic Examination of Sentinel Lymph Nodes

109

A sentinel node extensively replaced by metastatic melanoma (A, HE; B, S-100).

the correct (sentinel) node has been assessed.60 Ultimately, the objective would be to have a completely noninvasive system for SN assessment in vivo using MRS and ICP-MS. The advent of sophisticated magnetic resonance imaging machines with 3 Tesla magnets and appropriately designed surface coils makes this a realistic possibility. It would be extremely useful to obtain the necessary information from SNs without performing an invasive operative procedure that is sometimes associated with troublesome morbidity, such as seroma development, wound infection, and even persistent lymphedema of a limb. As well, the considerable cost of a surgical operation could be avoided and operating room use reduced, with further economic benefits. Recently, the use of preoperative ultrasound (US) examination of SNs for the detection of melanoma metastases has been investigated. The aim of this approach is to confirm US-detected positive SNs by FNB, thereby allowing these patients to proceed directly to CLND without the need for a SNB. The success of this approach is dependent on the sensitivity of US in detecting melanoma nodal micrometastases. Starritt and coworkers described the SMU experience of the use of high-resolution US in the evaluation of SNs, and found that SN metastases ⬍4.5 mm in size were not usually detectable by US.61 Rossi and coworkers reported a sensitivity of 39% and a specificity of 100% for preoperative US combined with FNB for detecting SN metastases.62 However, they failed to detect metastases less than 2 mm in diameter. Voit and colleagues reported a sensitivity of 65% and a specificity of 99% for the US identification of involved SNs. The identification rate using SN and FNB ranged from 40% in patients with primary melanomas ⱕ1 mm thick to 79% in patients with primaries ⬎4 mm in thickness.63 Remarkably, the claimed sensitivity of USFNB for the detection of tumors in SNs where the deposits were between 0.50 mm and 1.99 mm in size was 52%. Although some of these results suggest a promising role for routine use of US and FNB in the evaluation of SNs, this approach appears unlikely to replace SNB in all melanoma

patients in the immediate future because most melanoma micrometastases are very small and therefore beyond the detection limits of currently available US technology.

Conclusions The tumor-harboring status of SNs is the most important prognostic factor for early stage melanoma patients, and pathologists are therefore obliged to make a diligent assessment of those nodes. Determination of the most appropriate

Table 4 An example of a template for the microscopic pathology reporting of positive sentinel nodes in melanoma patients Feature Anatomic site of sentinel node No. of tumor foci Intranodal compartment(s) involved by tumor Max. dimension of largest deposit (mm) Max. tumor penetrative depth (mm) % cross-sectional area of SN involved by tumor Extranodal spread Immunophenotype of tumor S-100 HMB-45 MelanA Nodal nevus cells Other comments

Right axilla 3 Subcapsular sinus and parenchyma 3.25 0.68 10 Absent ⫹⫹⫹ (nuclear and cytoplasmic) ⫹⫹ (patchy) ⫹⫹⫹ (strong and diffuse) Present (capsular) Nil

Abbreviations: No., number; Max., maximum; mm, millimeters; ⫹⫹⫹, strong positivity; ⫹⫹, moderate positivity.

110

Seminars in Diagnostic Pathology, Vol 25, No 2, May 2008

histopathologic sampling, sectioning, and staining protocol for SNs must take into account the accuracy of the procedure, the time, labor, and costs involved, and clinical follow-up data where it is available. The balance between these competing factors will vary between institutions, and hence a decision should probably be made locally. In addition to determining whether a SN contains metastatic melanoma, pathologists should report the location and size of the metastatic deposits within positive SNs as this provides valuable prognostic information (Table 4). In the future, it is likely that methods for routine SN evaluation will change as new techniques to localize the site of metastases within the SN are developed and refined and nonhistopathologic methods of SN evaluation become increasingly available.

16. Scolyer RA, Thompson JF, McCarthy SW, et al: Intraoperative frozensection evaluation can reduce accuracy of pathologic assessment of sentinel nodes in melanoma patients.[comment]. J Am Coll Surg 201:821-823, author reply 823-824,2005 17. Klausen TL, Chakera AH, Friis E, et al: Radiation doses to staff involved in sentinel node operations for breast cancer. Clin Physiol Funct Imaging 25:196-202, 2005 18. Miner TJ, Shriver CD, Flicek PR, et al: Guidelines for the safe use of radioactive materials during localization and resection of the sentinel lymph node. Ann Surg Oncol 6:75-82, 1999 19. Nejc D, Wrzesien M, Piekarski J, et al: Sentinel node biopsy in skin melanoma patients: measurements of absorbed doses of radiation to the hands of medical staff. J Surg Oncol 93:355-361, 2006 20. Cochran AJ, Wen DR, Morton DL: Occult tumor cells in the lymph nodes of patients with pathological stage I malignant melanoma. An immunohistological study. Am J Surg Pathol 12:612-618, 1988 21. Gershenwald JE, Colome MI, Lee JE, et al: Patterns of recurrence following a negative sentinel lymph node biopsy in 243 patients with stage I or II melanoma. J Clin Oncol 16:2253-2260, 1998 22. Yu LL, Flotte TJ, Tanabe KK, et al: Detection of microscopic melanoma metastases in sentinel lymph nodes. Cancer 86:617-627, 1999 23. Baisden BL, Askin FB, Lange JR, et al: HMB-45 immunohistochemical staining of sentinel lymph nodes: a specific method for enhancing detection of micrometastases in patients with melanoma. Am J Surg Pathol 24:1140-1146, 2000 24. Cook MG, Green MA, Anderson B, et al: The development of optimal pathological assessment of sentinel lymph nodes for melanoma. J Pathol 200:314-319, 2003 25. Gietema HA, Vuylsteke RJ, de Jonge IA, et al: Sentinel lymph node investigation in melanoma: detailed analysis of the yield from step sectioning and immunohistochemistry. J Clin Pathol 57:618-620, 2004 26. Abrahamsen HN, Hamilton-Dutoit SJ, Larsen J, et al: Sentinel lymph nodes in malignant melanoma: extended histopathologic evaluation improves diagnostic precision. Cancer 100:1683-1691, 2004 27. Spanknebel K, Coit DG, Bieligk SC, et al: Characterization of micrometastatic disease in melanoma sentinel lymph nodes by enhanced pathology: recommendations for standardizing pathologic analysis. Am J Surg Pathol 29:305-317, 2005. 28. Cochran AJ: Surgical pathology remains pivotal in the evaluation of ‘sentinel’ lymph nodes. Am J Surg Pathol 23:1169-1172, 1999 29. Cochran AJ, Roberts A, Wen DR, et al: Update on lymphatic mapping and sentinel node biopsy in the management of patients with melanocytic tumours. Pathology 36:478-484, 2004 30. Yee VS, Thompson JF, McKinnon JG, et al: Outcome in 846 cutaneous melanoma patients from a single center after a negative sentinel node biopsy. Ann Surg Oncol 12:429-439, 2005 [Published erratum appears in Ann Surg Oncol 12:843, 2005] 31. Nowecki ZI, Rutkowski P, Nasierowska-Guttmejer A, et al: Survival analysis and clinicopathological factors associated with false-negative sentinel lymph node biopsy findings in patients with cutaneous melanoma. Ann Surg Oncol 13:1655-1663, 2006 32. van Akkooi AC, de Wilt JH, Verhoef C, et al: High positive sentinel node identification rate by EORTC melanoma group protocol. Prognostic indicators of metastatic patterns after sentinel node biopsy in melanoma. Eur J Cancer 42:372-380, 2006 33. Caraco C, Marone U, Celentano E, et al: Impact of false-negative sentinel lymph node biopsy on survival in patients with cutaneous melanoma. Ann Surg Oncol 14:2662-2667, 2007 34. Roulin D, Matter M, Bady P, et al: Prognostic value of sentinel node biopsy in 327 prospective melanoma patients from a single institution. Eur J Surg Oncol 34:673-679, 2008 35. Riber-Hansen R, Sjoegren P, Hamilton-Dutoit SJ, et al: Extensive pathological analysis of selected melanoma sentinel lymph nodes: high metastasis detection rates at reduced workload. Ann Surg Oncol 15: 1492-1501, 2008 36. Scolyer RA, Thompson JF, McCarthy SW: Sentinel lymph nodes in malignant melanoma: extended histopathologic evaluation improves

References 1. Cascinelli N, Belli F, Santinami M, et al: Sentinel lymph node biopsy in cutaneous melanoma: the WHO Melanoma Program experience. Ann Surg Oncol 7:469-474, 2000 2. Thompson JF, Scolyer RA, Kefford RF: Cutaneous melanoma.[see comment]. Lancet 365:687-701, 2005 3. Clary BM, Brady MS, Lewis JJ, et al: Sentinel lymph node biopsy in the management of patients with primary cutaneous melanoma: review of a large single-institutional experience with an emphasis on recurrence. Ann Surg 233:250-258, 2001 4. Gershenwald JE, Thompson W, Mansfield PF, et al: Multi-institutional melanoma lymphatic mapping experience: the prognostic value of sentinel lymph node status in 612 stage I or II melanoma patients. J Clin Oncol 17:976-983, 1999 5. Jansen L, Nieweg OE, Peterse JL, et al: Reliability of sentinel lymph node biopsy for staging melanoma. Br J Surg 87:484-489, 2000 6. Krag DN, Meijer SJ, Weaver DL, et al: Minimal-access surgery for staging of malignant melanoma. Arch Surg 130:654-658, discussion 659-660, 1995 7. Morton DL, Thompson JF, Essner R, et al: Validation of the accuracy of intraoperative lymphatic mapping and sentinel lymphadenectomy for early-stage melanoma: a multicenter trial. Multicenter Selective Lymphadenectomy Trial Group. Ann Surg 230:453-463, discussion 463-465, 1999 8. Reintgen D, Cruse CW, Wells K, et al: The orderly progression of melanoma nodal metastases. Ann Surg 220:759-767, 1994 9. Thompson JF, McCarthy WH, Bosch CM, et al: Sentinel lymph node status as an indicator of the presence of metastatic melanoma in regional lymph nodes. Melanoma Res 5:255-260, 1995 10. Thompson JF: The Sydney Melanoma Unit experience of sentinel lymphadenectomy for melanoma. Ann Surg Oncol 8:44S-47S, 2001 (suppl) 11. Balch CM, Soong SJ, Gershenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 19: 3622-3634, 2001 12. Morton DL, Wen DR, Wong JH, et al: Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 127:392-399, 1992 13. McCarthy WH, Shaw HM, Cascinelli N, et al: Elective lymph node dissection for melanoma: two perspectives. World J Surg 16:203-213, 1992 14. Morton DL, Thompson JF, Cochran AJ, et al: Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med 355:1307-1317, 2006 15. Diaz LK, Hunt K, Ames F, et al: Histologic localization of sentinel lymph node metastases in breast cancer. Am J Surg Pathol 27:385-389, 2003

Scolyer et al.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

Histopathologic Examination of Sentinel Lymph Nodes

diagnostic precision [comment]. Cancer 101:2141-2142, author reply 2142-2143, 2004 Li LX, Scolyer RA, Ka VS, et al: Pathologic review of negative sentinel lymph nodes in melanoma patients with regional recurrence: a clinicopathologic study of 1152 patients undergoing sentinel lymph node biopsy. Am J Surg Pathol 27:1197-1202, 2003 Gadd MA, Cosimi AB, Yu J, et al: Outcome of patients with melanoma and histologically negative sentinel lymph nodes. Arch Surg 134:381-387, 1999 Cochran AJ, Wen DR, Huang RR, et al: Prediction of metastatic melanoma in nonsentinel nodes and clinical outcome based on the primary melanoma and the sentinel node. Mod Pathol 17:747-755, 2004 Dewar DJ, Newell B, Green MA, et al: The microanatomic location of metastatic melanoma in sentinel lymph nodes predicts nonsentinel lymph node involvement. J Clin Oncol 22:3345-3349, 2004 Scolyer RA, Li LX, McCarthy SW, et al: Micromorphometric features of positive sentinel lymph nodes predict involvement of nonsentinel nodes in patients with melanoma. Am J Clin Pathol 122:532-539, 2004 Starz H, Balda BR, Kramer KU, et al: A micromorphometry-based concept for routine classification of sentinel lymph node metastases and its clinical relevance for patients with melanoma. Cancer 91:21102121, 2001 Pearlman NW, McCarter MD, Frank M, et al: Size of sentinel node metastases predicts other nodal disease and survival in malignant melanoma. Am J Surg 192:878-881, 2006 Debarbieux S, Duru G, Dalle S, et al: Sentinel lymph node biopsy in melanoma: a micromorphometric study relating to prognosis and completion lymph node dissection. Br J Dermatol 157:58-67, 2007 Govindarajan A, Ghazarian DM, McCready DR, et al: Histological features of melanoma sentinel lymph node metastases associated with status of the completion lymphadenectomy and rate of subsequent relapse. Ann Surg Oncol 14:906-912, 2007 Sabel MS, Griffith K, Sondak VK, et al: Predictors of nonsentinel lymph node positivity in patients with a positive sentinel node for melanoma. J Am Coll Surg 201:37-47, 2005 Murali R, Cochran AJ, Cook MG, et al: Interobserver reproducibility of the assessment of morphometric features of sentinel lymph nodes containing metastatic melanoma. Ann Surg Oncol 15:29-30, 2008 (suppl 1) van Akkooi AC, de Wilt JH, Verhoef C, et al: Clinical relevance of melanoma micrometastases (⬍0.1 mm) in sentinel nodes: are these nodes to be considered negative? Ann Oncol 17:1578-1585, 2006 Satzger I, Volker B, Meier A, et al: Prognostic significance of isolated HMB45 or Melan A positive cells in Melanoma sentinel lymph nodes. Am J Surg Pathol 31:1175-1180, 2007 Dawson M, Doble P, Beavis A, et al: Antimony by ICP-MS as a marker for sentinel lymph nodes in melanoma patients. Analyst 128: 217-219, 2003

111

51. Scolyer RA, Thompson JF, Li LX, et al: Antimony concentrations in nodal tissue can confirm sentinel node identity. Mod Pathol 17:11911197, 2004 52. Scolyer RA, Thompson JF, Li LX, et al: Failure to remove true sentinel nodes can cause failure of the sentinel node biopsy technique: evidence from antimony concentrations in false-negative sentinel nodes from melanoma patients. Ann Surg Oncol 11:174S-178S, 2004 (suppl) 53. Morton DL, Hoon DS, Cochran AJ, et al: Lymphatic mapping and sentinel lymphadenectomy for early-stage melanoma: therapeutic utility and implications of nodal microanatomy and molecular staging for improving the accuracy of detection of nodal micrometastases. Ann Surg 238:538-549, discussion 549-550, 2003 54. Haigh PI, Lucci A, Turner RR, et al: Carbon dye histologically confirms the identity of sentinel lymph nodes in cutaneous melanoma. Cancer 92:535-541, 2001 55. Karim RZ, Scolyer RA, Li W, et al: False negative sentinel lymph node biopsies in melanoma may result from deficiencies in nuclear medicine, surgery and/or pathology. Ann Surg 247:1003-1010, 2008 56. Giese T, Engstner M, Mansmann U, et al: Quantification of melanoma micrometastases in sentinel lymph nodes using real-time RT-PCR. J Invest Dermatol 124(3):633-7, 2005 57. Blaheta HJ, Ellwanger U, Schittek B, et al: Examination of regional lymph nodes by sentinel node biopsy and molecular analysis provides new staging facilities in primary cutaneous melanoma. J Invest Dermatol 114(4):637-42, 2000 58. Bostick PJ, Morton DL, Turner RR, et al: Prognostic significance of occult metastases detected by sentinel lymphadenectomy and reverse transcriptase-polymerase chain reaction in early-stage melanoma patients. J Clin Oncol 17(10):3238-44, 1999 59. Stretch JR, Somorjai R, Bourne R, et al: Melanoma metastases in regional lymph nodes are accurately detected by proton magnetic resonance spectroscopy of fine-needle aspirate biopsy samples. Ann Surg Oncol 12(11):943-9, 2005 60. Beavis A, Dawson M, Doble P, et al: Sentinel lymph node identity confirmed by analysis of fine needle biopsy samples using inductively coupled plasma-mass spectrometry. Ann Surg Oncol 2007:in press 61. Starritt EC, Uren RF, Scolyer RA, et al: Ultrasound examination of sentinel nodes in the initial assessment of patients with primary cutaneous melanoma. Ann Surg Oncol 12(1):18-23, 2005 62. Rossi CR, Mocellin S, Scagnet B, et al: The role of preoperative ultrasound scan in detecting lymph node metastasis before sentinel node biopsy in melanoma patients. J Surg Oncol 83(2):80-4, 2003 63. Voit C, Van Akkooi AC, Schafer-Hesterberg G, et al: Role of ultrasound (US) and US-guided fine needle aspiration cytology (USFNAC) prior to sentinel lymph node biopsy (SLNB) in 500 melanoma patients: Reduction of need for SNLB by high US-FNAC SN positive identification rate. J Clin Oncol 25(18S):8512, 2007 64. Morton DL, Thompson JF, Cochran AJ: Sentinel-node biopsy in melanoma. N Engl J Med 356:419-421, 2007