| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
1 Sydney Melanoma Unit, Sydney Cancer Centre and Melanoma and Skin Cancer Research Institute, Royal Prince Alfred Hospital, Camperdown, New South Wales, 2050, Australia
2 Nuclear Medicine and Diagnostic Ultrasound, RPAH Medical Centre, Newtown, New South Wales, 2042, Australia
3 Department of Medicine, University of Sydney, New South Wales, 2006, Australia
4 Department of Anatomical Pathology, Royal Prince Alfred Hospital, Camperdown, New South Wales, 2050, Australia
5 Department of Surgery, University of Sydney, New South Wales, 2006, Australia
Correspondence: Address correspondence and reprint requests to: John F. Thompson, MD, Sydney Melanoma Unit, Sydney Cancer Centre, Royal Prince Alfred Hospital, Gloucester House, Camperdown, New South Wales, 2050, Australia; E-mail: thompson{at}smu.org.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: A total of 304 patients presenting with primary cutaneous melanomas had SLNs identified by lymphoscintigraphy and then examined in situ by the same physician with high-resolution US. Within 24 hours, the SLNs were removed for histopathologic assessment of sections stained conventionally and with immunohistochemical markers for S100 protein and HMB45 antigen.
Results: Metastatic disease was present in SLNs from 33 node fields in 31 patients. The US results in seven of these cases were suggestive of metastatic disease. Twenty-six node fields contained positive nodes not detected by US. Undetected deposits had diameters <4.5 mm.
Conclusions: These results suggest that a targeted US examination of SLNs can detect metastatic melanoma deposits down to approximately 4.5 mm in diameter. However, most metastatic melanoma deposits in SLNs are considerably smaller than this at the time of initial staging, and US therefore cannot be considered cost-effective in this setting.
Key Words: Ultrasound • Sentinel lymph node • Outcomes • Melanoma • Metastasis • Staging • Following
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
High-resolution ultrasound (US) has been shown to be superior to physical examination for the early detection of lymph node metastases in the follow-up of patients previously treated for primary cutaneous melanoma.2–6 However, the potential role of targeted US examination of sentinel lymph nodes (SLNs) in the initial staging of patients who present with primary cutaneous melanoma has not yet been fully evaluated. Moreover, the minimum size of metastatic melanoma deposits able to be detected in lymph nodes by US examination has not been determined; this has important implications for follow-up protocols. The aim of this study was to assess the potential value of targeted US examination as a staging tool and for follow-up investigation, by determining the threshold size of metastatic melanoma deposits that could be detected.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In each of the patients, both the lymphoscintigram and the high-resolution US examination were performed and interpreted by the same nuclear medicine/US physician on the same day. Lymphoscintigraphy was performed first, and the exact surface location of each SLN was marked on the skin. The depth of the SLN from the skin mark was also determined by measurement on an orthogonal view. The physician then performed a carefully targeted US examination of that same SLN, knowing exactly where it was to be found.
The US examinations were performed with an ATL Ultramark 9 HDI diagnostic US system (Advanced Technology Laboratories Australia, Pty, Ltd, Sydney, NSW, Australia) and a linear array L10-5 transducer (Advanced Technology Laboratories Australia) at a frequency of 5 to 10 MHz.
Lymphoscintigraphy was performed with 99mTc-antimony sulfide colloid tracer. The tracer was injected intradermally around the primary melanoma or on both sides of the central part of the excision-biopsy scar. The specific activity of the tracer used was 100 MBq/mL, and the volume of each injection was .05 mL. Most patients required four injections, giving a total activity of 20 MBq. Immediately after tracer injection, dynamic imaging was commenced. The initial 10 frames were acquired at a rate of 1 per minute and then less often as the lymphatic channels were followed until they reached the draining node fields. The usual duration of this phase of the lymphoscintigraphy examination was 20 minutes. Delayed static images of all node fields potentially receiving tracer were then obtained 150 minutes after tracer injection to ensure that all SLNs were detected. SLNs identified in each node field were marked by placing a pinpoint tattoo of carbon black ink and a small cross of indelible ink on the overlying skin surface. The depth of each SLN was determined by obtaining an orthogonal image with a small radioactive point source placed on the skin surface mark. The position of the patient during marking was replicated both during the subsequent US examination and during surgery.
At the time of each US study, the SLNs were classified as (1) no evidence of metastatic disease, (2) suggestive of metastatic disease, or (3) probably containing metastatic disease. This opinion was based on standard US teaching and the experience of the physician in examining benign and malignant lymph nodes with US. In general, features on US considered suggestive of metastatic disease were node thickness greater than two-thirds node length, areas of low-level echoes, and increased vascular signature on color Doppler imaging and color power angiography.7–9 The combination of low-level echoes in a node with a thickness greater than two thirds of its length (Fig. 1
) was considered an indication of probable metastatic disease.8 The absolute size of the lymph node was not included as an indication of metastatic disease, because lymph node enlargement is common in benign conditions.7,10 In some cases, the size of a suspected metastatic deposit could be estimated from the US image. When this was possible, it was expressed as the greatest diameter of the largest deposit.
|
). The sections were then examined microscopically, initially at a scanning magnification of x100. When metastatic melanoma deposits were identified, the maximum diameter of the largest deposit was measured in each SLN. When more than three positive SLNs were recovered from one node field, the maximum diameters of the two largest deposits from two different nodes were reported.
|
The association between the maximum diameter of the largest metastatic tumor deposit and the Breslow thickness of the primary melanoma was assessed with the Pearson correlation coefficient. The significance of the proportion of node fields containing SLNs with deposits reported as "suspicious" or "probable" on US being associated with primary tumors 
4.0 mm thick (American Joint Committee on Cancer [AJCC] stage IIB or above) was tested with a 
2 analysis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Metastatic disease was present in 33 individual lymph node fields sampled from the 31 patients. In six of these cases, the US was reported as "suspicious" of metastatic melanoma; in one case, the conclusion was "probable" metastatic melanoma.
|
In the 31 patients, there were 26 node fields containing positive nodes not detected by US. Most undetected melanoma deposits had maximum diameters < 4.0 mm. The only exceptions were three axillary nodes containing undetected deposits measuring 4.1, 4.5, and 4.5 mm in diameter. One of these axillary deposits was of desmoplastic melanoma, noted histopathologically to have been very difficult to measure accurately even in the fixed and stained tissue sections. The two 4.5-mm diameter deposits were in axillary SLNs located more than 2 cm from the skin surface.
The maximum diameter of the largest metastatic tumor deposit from a given node field was significantly correlated with the Breslow thickness of the primary tumor (r = .4; P < .05). Six of the seven node fields with SLNs containing metastatic deposits reported as "suspicious" or "probable" on US were associated with primary tumors with a Breslow thickness of 4.0 mm. This proportion was significant (P < .001).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Over the past decade, the technique of lymphatic mapping and SLN biopsy has become the preferred method for regional lymph node staging in patients with newly diagnosed cutaneous melanomas. However, although it is possible that this procedure, with subsequent complete regional node dissection when metastases are present, may offer patients a therapeutic advantage,12 this has not yet been validated by prospective multicenter randomized controlled trials. US may offer a means for preliminary staging in situations where there is no immediate access to the SLN biopsy procedure and, perhaps even more importantly, may be a cost-effective means of restaging patients during follow-up after initial staging and treatment.
Several previous studies have demonstrated that high-resolution US is a more sensitive and specific alternative to physical examination in the detection of lymph node metastases.2–6 However, despite the high sensitivities and specificities reported for US in these studies, the true sensitivity and specificity of US in detecting lymph node metastases is unknown. In most of these studies, the histopathologic status of the lymph nodes examined with US and palpation was determined only if findings from US or palpation were suspicious.2,4–6 In another study, the histopathologic status of the lymph nodes examined with both methods was sometimes determined up to 6 months after a reported examination.3 Not only do the true sensitivities and specificities of each method remain unknown, but the sensitivity of US in identifying clinically undetectable tumor deposits is also unknown.
The aim of this study was to define the limit of a high-resolution targeted US examination in detecting metastatic melanoma deposits in SLNs by determining the threshold size of deposits that could be detected with this technique. The results of this study of 33 lymph node fields that were histopathologically positive for metastatic disease suggest that US can detect SLN deposits of metastatic melanoma >4.0 mm in diameter in SLNs from regions other than the axilla and >4.5 mm in diameter in axillary SLNs. The lower sensitivity of US for axillary nodes seems likely to be because they are usually at a greater depth from the skin surface in this area, making imaging less satisfactory. There is no doubt that axillary SLNs are more difficult to visualize with US than those in the groin or neck.
In a recent study by Rossi et al.,13 the role of a nontargeted preoperative US examination in 125 patients with cutaneous melanomas >1 mm thick was evaluated. These authors reported that no nodes with tumor deposits <2 mm were detected with US and that the spatial resolution of US was, therefore, 2 mm. They also reported a very high sensitivity (100%) and a specificity of 39%. These data seem encouraging. However, because these authors did not use a targeted US examination, there must be doubt about whether the nodes examined with US were actually the same nodes later identified with lymphoscintigraphy and examined histopathologically as the SLNs. The fact that 140 node fields were evaluated before surgery with US and 178 node fields were then identified with lymphoscintigraphy supports this concern. For this reason alone the reliability of the data must be questioned. In addition, the proportion of groin node fields examined in the study (52%) was considerably greater than in the present study (30%), which should have improved sensitivity because these nodes usually lie closer to the skin.
In our study, node fields containing tumor deposits detected by US were associated with primary tumors with a Breslow thickness of at least 4.0 mm in all but one case. This corresponds to an AJCC stage designation of at least IIB at the time of initial treatment. In a recent study by Garbe et al.,14 the ability of US to detect melanoma recurrence in lymph nodes during the follow-up of treated patients was greatest in patients designated stage IIB or higher at the time of treatment. This suggests that the sizes of the deposits able to be detected in their study were similar to those detected in our study. In view of their findings and the results of our study, a strong case could be made for performing US of the SLNs in the subset of patients who present with melanomas thicker than 4 mm. Those found to have metastatic disease in one or more SLNs could then have an immediate complete regional lymph node dissection and thereby avoid the delay, inconvenience, and potential morbidity associated with a preliminary SLN biopsy procedure.
Features on US examination considered suggestive of metastatic disease in SLNs in our study were node thickness greater than two thirds of node length, areas of low-level echoes, and increased vascular signature on color Doppler imaging and color power angiography. These features have been well described previously.7–9 In a recent study by Kahle et al.,15 a specific asymmetrical cortical thickening phenomenon causing a "cap"-like sonomorphological pattern to occur on US in the SLNs of melanoma patients was described. The locations of any lymph nodes found to show this cap phenomenon on US were compared with the location of nodes subsequently determined to be the SLNs by lymphoscintigraphy. These authors found 85% agreement between the two sites and concluded that this sonomorphological pattern is characteristic of SLNs. Histopathologic examination of these SLNs revealed that 13.4% were positive for metastatic melanoma. It is noteworthy that we did not observe this phenomenon in our study, even with the advantage of performing the US examination after the SLNs had been identified by lymphoscintigraphy, and so cannot confirm the proposal that demonstration of a cap on US examination indicates that it is an SLN.
A recent study by Carlson et al.16 examining the amount of metastatic disease in the SLN as a prognostic factor in melanoma patients demonstrated a significant survival advantage for patients with SLN metastases < 2 mm in diameter compared with those who have metastases > 2 mm in diameter. Certainly, the ability to detect metastatic deposits < 2 mm in diameter would be likely to improve patient management on presentation and probably also during follow-up. However, the results of our study suggest that current US technology and methods do not permit the detection of SLN metastases < 2 mm in diameter.
In conclusion, because the great majority of metastatic melanoma deposits in SLNs at the time of initial staging are < 4.5 mm in diameter, the threshold of detection in this study, preoperative US will provide useful information in only a few patients. Targeted preoperative high-resolution US examination of SLNs therefore cannot be recommended for all patients with primary cutaneous melanomas as a cost-effective initial staging tool. However, in the subset of patients with AJCC stage IIb disease (melanomas thicker than 4 mm), pre-operative US will provide evidence of metastatic disease in an SLN more frequently and should be considered. With better appreciation of the capabilities and limitations of the technique, US may also serve as an important adjunct to routine physical examination for follow-up assessment, after initial staging and treatment, in patients with cutaneous melanoma. Efforts to develop US technology and scanning techniques, such as the targeted examination used in this study, should continue, in the hope that it will be possible to detect smaller tumor deposits in lymph nodes in the future.
Received for publication March 15, 2004. Accepted for publication August 30, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M G Cook and S Di Palma Pathology of sentinel lymph nodes for melanoma J. Clin. Pathol., August 1, 2008; 61(8): 897 - 902. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Koskivuo, L. Talve, P. Vihinen, M. Maki, T. Vahlberg, and E. Suominen Author Reply: Sentinel Lymph Node Biopsy in Cutaneous Melanoma Ann. Surg. Oncol., June 1, 2008; 15(6): 1810 - 1811. [Full Text] [PDF]
|
|
|
|
|
|
M. C. van Rijk, H. J. Teertstra, J. L. Peterse, O. E. Nieweg, R. A. V. Olmos, C. A. Hoefnagel, and B. B. R. Kroon Ultrasonography and Fine-needle Aspiration Cytology in the Preoperative Evaluation of Melanoma Patients Eligible for Sentinel Node Biopsy Ann. Surg. Oncol., November 1, 2006; 13(11): 1511 - 1516. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. F. Uren Sentinel Lymph Node Biopsy in Melanoma J. Nucl. Med., February 1, 2006; 47(2): 191 - 195. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
