Current status of sentinel lymph node biopsy in patients with melanoma

Summary

The sentinel node is defined as a lymph node on a direct lymphatic drainage pathway from the primary tumour. The sentinel node biopsy is an invasive and potentially complex diagnostic test that requires considerable skill and experience from nuclear medicine physicians, surgeons and pathologists involved in the process. The procedure provides important staging information. The tumour bearing status of a regional lymph node basin is the most important prognostic factor for patients with a clinically localised melanoma. The recently published final report of the first Multicenter Selective Lymphadenectomy Trial (MSLT-1) shows that sentinel node biopsy with subsequent regional node dissection improves the survival rate in patients with nodal metastases from an intermediate Breslow thickness melanoma (1.2−3.5 mm). The false negative rate is substantial but can be limited by experience, a meticulous technique and close cooperation of the specialists involved. The surgical procurement of the sentinel nodes is associated with minor morbidity.

Sentinel node biopsy is now part of the standard work-up in patients with intermediate thickness melanoma and can be considered in patients with a thinner or thicker melanoma. New lymph node tracers and innovative imaging techniques are likely to improve the sensitivity of the procedure further. Now that the results of MSLT-1 are available, the next question is whether sentinel node-positive patients require further surgery. This issue is addressed by EORTC’s Minitub study and the MSLT-2. Until these studies yield results, regional node dissection is recommended for patients with sentinel node metastases.

Key words:
melanoma − review − sentinel lymph node biopsy – staging − survival

Introduction

Approximately one in 50 individuals in Europe developed melanoma. Every year 4% more patients are diagnosed with the disease and this makes melanoma the cancer with the greatest increase in incidence [1]. At the time of the diagnosis there is usually no evidence of dissemination to other sites, but one in five patients who present with a primary melanoma that has a Breslow thickness of more than 1 mm have occult involvement of their regional lymph nodes. These metastases are too small to palpate and they usually elude ultrasound screening at this stage. If left undisturbed, such metastases are detected later when they grow and become palpable. At that time, they have often generated metastases at distant sites and these are mostly incurable. Prophylactic regional node dissection has been the standard management of the regional lymph nodes for many years until randomised studies revealed that there is no survival benefit from this procedure [2−5]. However, an improved survival was shown if one considers the patients in whom prophylactic node dissection did reveal metastases, as was also demonstrated in other studies [6−10]. This survival advantage is particularly evident in patients with melanomas of intermediate thickness. These lesions have a substantial risk of lymph node involvement, while the risk of fatal blood-borne metastases to vital organs is still limited. Sentinel node biopsy was developed as a diagnostic technique to detect lymph node metastases at this early stage so that regional node dissection could be carried out to gain this potential survival benefit [11].

The sentinel node concept of orderly progression of lymph node metastases is exciting and appeals to the surgical frame of mind. Its momentum caused people to get carried away by their enthusiasm. This is why the procedure was immediately embraced and made into standard management in analogy to many other new surgical procedures in the past, which is without much scientific evidence. Since then, a lot of research has been done. The purpose of this article is to present a balanced summary of the current status of lymphatic mapping in melanoma and to speculate about its future. The emphasis is on the data gathered by the dedicated team at The Netherlands Cancer Institute since 1993.

Concept, technique and definition

Morton et al introduced the sentinel node procedure based on the concept that lymph fluid from the primary tumour area carrying detached malignant cells drains upon a particular lymph node first and subsequently travels to nodes further downstream and distant sites [11]. Sentinel node biopsy requires a joint effort from nuclear medicine physician, surgeon and pathologist. The best techniques to find the sentinel node were quickly established [12]. Lymphoscintigraphy is important to identify a lymph node on a direct drainage pathway and learn its location. The subsequent operation requires both a blue dye and a gamma ray detection probe. Using this combined technique, surgeons can identify sentinel nodes in almost all patients [13,14]. Pathologists obtain multiple sections and use sensitive and specific immunohistochemistry staining techniques.

A sentinel node was initially defined as “the initial lymph node upon which the primary tumour drains”[11]. The word “initial” can be interpreted in different ways and may lead to confusion. For instance, multiple lymph vessels originating in the primary tumour may drain upon different nodes that may not necessarily be visualised simultaneously, yet they all may harbour tumour cells. In the following years, this definition was challenged [15−18]. Physicians with different backgrounds are involved and tend to consider the concept from their own perspective. The lymphoscintigrams often show multiple nodes and the order of lymph drainage may not be clear-cut. Also, the surgical procedure may be challenging and requires considerable expertise because the surgeon needs to translate the two-dimensional scintigrams and the one-dimensional probe readings into a three dimensional image in his mind. Furthermore, the fragile blue lymphatic vessel that directs the surgeon to the sentinel node is easily damaged and can then no longer be traced. Considerable experience and finesse are required to dissect this delicate structure through a small incision in a narrow space in a lymph node basin that may be quite deep. For these reasons, different definitions have been proposed (Tab. 1) [18−22]. Apparently, one man’s sentinel node is not necessarily somebody else’s. Although these alternative definitions point to the node(s) directly at risk most of the time, they are based neither on the physiology of lymph drainage, nor on the biology of the disease [16]. These alternative definitions may cause the surgeon to refrain from removing a node on a direct lymphatic drainage pathway, which results in a false negative procedure, and they often lead to unnecessary removal of nodes that are not directly at risk of containing malignant cells. To resolve the confusion from multiple definitions, the original definition was slightly modified to “a sentinel node is any lymph node on the direct drainage pathway from the primary tumour”[18]. This definition reflects the physiology of lymph drainage and the concept of step-wise spread of cancer through the lymphatic system, and this is the definition most experts adhere to.

Results

The sentinel node biopsy is a staging procedure. Staging helps to select the best therapeutic option for an individual patient and provides prognostic information. The tumour status of the sentinel node proved to be the most important prognostic factor [23,24]. Based on these findings, the American Joint Committee on Cancer and the Union Internationale Contre le Cancer incorporated the tumour status of the sentinel node in their melanoma staging classification [25]. The early detection and removal of lymph node metastases obviously reduces the incidence of relapse in the nodal basin and the anxiety associated with a recurrence is a very traumatizing experience and reduces the quality of life [26]. Also, the rates of recurrence in a regional nodal basin range from 2% to 10% after treatment with formal lymphadenectomy performed because of a positive sentinel node, but range from 20% to 50% among patients who undergo lymphadenectomy for palpable disease [27−30]. The extent of a neck dissection or groin dissection is often greater in case of recurrence with palpable disease resulting in a longer hospital stay and more oedema [31].

The obvious benefits of the procedure need to be weighed against the disadvantages. A single-day admission with general anaesthesia is customary. Lymphoscintigraphy necessitates a minor exposure to ionising radiation. The morbidity of sentinel node biopsy is generally described as limited, yet at The Netherlands Cancer Institute postoperative complications were seen in 9% of the patients [14]. These were all minor and resolved completely. Long-term complications after sentinel node biopsy were seen in 18% with (slight) oedema of the leg being the most frequent and serious. Our initial impression was that sentinel node biopsy could lead to an increase in the occurrence of in transit metastases [32]. Fortunately, a subsequent study disproved this notion, as in-transit metastases were seen in 15% of the sentinel node-positive patients, similar to the 14% in patients with palpable node metastases [33].

Many investigators look favourably upon the ability of lymphatic mapping to detect nodal metastases, but the follow-up duration of many studies is too short for all nodal recurrences to become evident after removal of a non-metastatic sentinel node. In a study with a mean follow up duration of 54 months, the mean interval to discovery of the missed metastases in the regional lymph node field was 28 months with a range of 4.6−106 months [34]. Also, many investigators calculate the rate of false-negative procedures over the entire group of patients or over the group of sentinel node-negative patients [35]. However, one cannot miss an involved node in a patient who does not have involved nodes [36]. As such, the sensitivity of a diagnostic test to detect disease should be determined in a population in which this abnormality is present. Early publications from reputable institutions and cooperative groups around the world showed false-negative rates ranging from 16% to 38%, when recalculated in the above recommended fashion [35]. This is a lot worse than the false negative rate in breast cancer. A systematic literature review and meta-analysis of the published studies in that disease revealed false negative results between 0% and 3% with a weighed combined sensitivity of 100% [37].

These poor results were analysed and led to improvement. At The Netherlands Cancer Institute, we adhered to the correct definition of a sentinel node and we meticulously used the combined detection technique. All sentinel node procedures were done by one of two surgeons and subsequently evaluated in a multidisciplinary meeting. The results of 708 consecutive patients with a median follow-up of 54 months were analysed. Sentinel node biopsy was positive in 164 (23%) of the patients and false negative in ten (1.4%), which results in a false negative rate of 5.7%. Five of the ten failures occurred in the first year after the sentinel node biopsy was introduced, indicating a learning phase with a false negative rate of 29%. In the subsequent years, the false negative rate was 3.0%. The mishaps could be attributed once to the nuclear medicine physician, once to the surgeon and twice to the pathologist. No cause could be identified in the remaining six patients.

Survival

After its early and rapid embracement as a routine procedure, sentinel node biopsy became the subject of a fierce debate [38−40]. Opposing views as to its impact on survival were vigorously defended. Whether the procedure leads to an improvement in the chances of melanoma patients to survive is indeed the most important question that needed to be answered. Randomised studies of diagnostic techniques are unusual but we felt this approach was justified because of the clinical importance of the matter and the invasive nature of the procedure. The first Multicenter Selective Lymphadenectomy Trial (MSLT-1) was initiated in 1994. This was the largest and longest running surgical melanoma trial ever. Patients with a melanoma of at least 1 mm Breslow thickness or Clark level IV were randomised to either sentinel node biopsy with subsequent completion node dissection if a metastasis was found or to observation with node dissection when a nodal metastasis became evident later on. All patients were followed for ten years.

The end results of MSLT-1 have recently been published and show that 21% of the entire population of 2001 patients had lymph node metastases [41]. The incidence of nodal metastases was similar in the two arms of the study. Sentinel nodes were found in 99.4% of the individuals who were randomised to undergo the procedure. The most notable results concern the 1270 patients with a melanoma of intermediate Breslow thickness, defined as a Breslow thickness of 1.2−3.5 mm. Ten-year melanoma specific survival in the target population of node-positive patients increased from 41.5% when regional node dissection was performed for palpable nodal disease to 62.1% when early dissection was performed on the basis of an involved sentinel node (p=0.006). There was no survival difference in patients with thicker melanomas and between the two entire randomised groups. The number of patients with thin melanomas was not sufficient for a meaningful analysis. Nodal metastases were detected during observation in 31 of 643 patients with negative sentinel nodes, resulting in a false negative percentage of 20.3.

Reviewing these results, one cannot but be excited about the 20.6% improvement in survival, which is exceptional in oncology. Opponents emphasise the lack of a survival difference between the two entire randomised groups [38]. The reason is that the survival gain in the node-positive 21% target proportion is diluted by the other 79% who have no metastases and, therefore, cannot benefit from the procedure. After all, a diagnostic test that improves survival sounds too good to be true.

Current clinical practice

Patients with a clinically localised melanoma with a Breslow thickness of 1.2−3.5 mm are candidates for the sentinel node procedure. The risk of nodal recurrence is reduced substantially and node-positive patients have a 20.6% better chance to survive.

How should we manage patients with a melanoma outside the intermediate thickness range? The number of patients with a melanoma thinner than 1.2 mm in MSLT-1 was too small to allow proper analysis and no survival benefit was established if the melanoma was thicker than 3.5 mm. Only 5% of the patients with a thin melanoma have lymph node metastases. Thick melanomas have often already generated occult metastases to visceral organs and are associated with a poor prognosis. Early treatment of only their lymph node metastases will not improve their chance to survive but will improve regional tumour. Sentinel node biopsy can be discussed with patients with thin or thick melanomas for staging and for providing prognostic information. At the Melanoma Institute Australia, sentinel node biopsy is considered in patients with a thin melanoma in the presence of unfavourable features such as young age, ulceration, high-mitotic rate, microsatellite metastases or lymphovascular invasion.

Future

Although generally successful, the current techniques used for the sentinel node procedure have limitations. In the past decade several new tracers and imaging techniques have been introduced to improve sentinel node retrieval. Single photon emission computed tomography with radiographic computed tomography (SPECT/CT) combines the physiologic information on lymph drainage with visualization of the surrounding anatomy and enables more accurate planning of the operation [42−44]. A portable gamma camera for use in the operating theatre can visualise radioactive hot spots directly in the operative field [43].

A novel hybrid tracer was developed combining the fluorescent indocyanine green with ^99mTc-nanocolloid [45]. This tracer allows preoperative lymphoscintigraphy, probe detection and adds the option of intraoperative fluorescence imaging [46,47]. A near-infrared camera can visualize the sentinel node through a thin layer of tissue.

Freehand SPECT is another innovative intra-operative imaging technique [48]. This technique synchronises information on position, orientation and readings of the gamma ray detection probe. Intra-operatively, a three dimensional real-time image is constructed that guides the surgeon to the sentinel node [49]. These new technologies may be useful in the difficult cases. They may make the procedure easier and faster in the routine cases but they provide no additional benefit in the routine cases. The main potential of these innovative diagnostic approaches is that they may open up other cancer types to sentinel node biopsy, cancers in deeper locations, with a more intricate anatomy and cancer types with more complex lymph drainage than melanoma.

Now that the results of MSLT-1 are available, the next question is whether sentinel node-positive patients require further surgery. Completion node dissection yields more nodal metastases in 12% of such patients, but half of these patients harbour occult distant disease that leads to their demise [14]. A third of the remaining patients can still be cured when the node dissection is carried out at the time that the nodes become palpable and not all minimal metastases may necessarily progress to relevant disease. Observational studies yielded different outcomes, some indicating the need for completion node dissection, others suggesting that additional surgery is not routinely required [50,51]. The question is particularly relevant in patients with limited involvement of the sentinel node [52]. I collected 31 published factors that are claimed to suggest whether other nodes are involved, but none of these are adequate. One observational prospective study (Minitub) and one randomised investigation (MSLT-2) are currently addressing this issue. Until these studies yield results, the practice at the Melanoma Institute Australia is to subject all patients with sentinel node metastases to regional node dissection [53].

There are now a few new drugs that improve survival in patients with inoperable regional and visceral melanoma metastases. Next, the value of such systemic therapies will be examined in earlier stages of the disease. Sentinel node biopsy will be needed to identify patients who may benefit from these drugs when they prove to be effective as adjuvant treatment for operable lymph node involvement. The first results of a randomised study of adjuvant immunotherapy with ipilimumab have recently been presented [54]. An improved recurrence-free survival was established in patients with lymph node metastases greater than 1 mm, but overall survival data are not known at this time. The morbidity of this treatment is substantial and five patients (1.1%) died from the treatment. In my view, these preliminary results do not warrant sentinel node-positive patients to be treated with ipilimumab routinely. Other such studies have been initiated with even newer and more promising drugs and these trials deserve our surgical support.

Concluding remarks

Lymphatic mapping has evolved since its introduction in 1992. It is now clear that a node on a direct lymph pathway from the melanoma should be called sentinel lymph node. The nuclear medicine and surgical technique have evolved to the level that the node can be visualised and harvested in virtually every patient. With experience and dedication, sentinel node biopsy is an accurate staging procedure with a low false negative rate. The tumour status of the node is the most powerful prognostic factor. And last but not least, it identifies patients whose chance to survive improves by 20.6% with a subsequent regional node dissection. Sentinel node biopsy is evolving further and when adequate systemic therapy becomes available the indication range will expand to enhance its importance even further.

Omgo E. Nieweg, MD, PhD

Melanoma Institute Australia

40 Rocklands Road 40

North Sydney, NSW 2060

e-mail: omgo.nieweg@melanoma.org.au