Ultrasound seeks out metastatic melanoma

Cutaneous malignant melanoma is a relatively common malignancy, and its incidence is increasing worldwide, especially in the Caucasian population. The clinical course of melanoma depends on tumor thickness, localization, ulceration, growth rate, and histology of the primary tumor. Metastatic spread may arise from very small masses.1

Cutaneous melanoma metastases develop via locoregional (two-thirds of cases) or distant dissemination (one-third).2 Locoregional dissemination through the lymphatic system presents as satellite and in-transit metastases, and metastases to lymph nodes. About 50% of metastases are associated with regional lymph nodes, 20% are satellite or in-transit metastases, and 30% are direct distant metastases.3

The development of different metastatic pathways appears to be related to the primary tumor location. Cutaneous melanomas on the trunk or upper extremities more frequently develop direct distant metastases. Tumor localization in the lower extremities or head and neck region leads primarily to satellite or in-transit metastases.4

Satellite and in-transit metastases are typical of melanoma. They develop between the site of the primary tumor and regional lymph nodes in the skin lymphatics and subcutaneous tissue. Metastases within 2 cm of the primary lesion are known as satellites. Those in the cutaneous or subcutaneous tissue and more than 2 cm from the primary tumor are referred to as in-transit metastases. Either can be responsible for the tendency of thicker melanomas to show local recurrence (Figure 1).

Lymph node metastases frequently precede hematogenic metastasis. Metastatic nodes are present at the time of initial diagnosis in 5% of cases. Most involvement will be linked to the regional lymph node group associated with the primary melanoma. The presence of affected lymph nodes can help determine the chance of further recurrences and likely survival.1 Detection of lymph node metastases is thus of special value when classifying patients with cutaneous melanoma.5

Melanomas can be grouped into five stages, according to a simplified tumor/ nodule/metastasis (TNM) classification:

• Stage 0: primary melanoma affecting epidermis alone (melanoma in situ);
• Stage I: T < 2-mm thick or < 1-mm thick and ulcerated, N0, M0;
• Stage II: T > 2-mm thick or >1-mm thick and ulcerated, N0, M0;
• Stage III: T = any thickness, N1, M0; and
• Stage IV: All M1 tumors.

KEY FEATURES

Characteristics of both regional lymph node and in-transit and satellite metastases can be apparent on ultrasound.

• Regional lymph nodes. Histologic changes seen in lymph nodes in patients with nodal disease have been used to explain morphologic changes shown on gray-scale ultrasound. The generally diffuse nature of the process in benign reactive nodes usually preserves the nodal shape and echogenic hilar structure. In malignant nodal disease, infiltration of the node by malignant tissue changes the shape and echogenicity of the node on gray-scale ultrasound and deforms the normal echogenic hilus.6

Morphologic criteria have been used to differentiate benign from malignant lesions in superficial lymphadenopathy. These criteria consist of an altered lymph node shape from flat oval to broad oval or round (longitudinal-transverse diameter ratio [L/T] <2), loss of the central echogenic hilus, and eccentric cortical widening.7 These criteria apply to nodal metastases from malignant melanoma. Differentiation between different types of lymph node malignancies is generally impossible, however.7,8

Ultrasound will show a solid, round, hypoechogenic nodule with no hyperechogenic medulla when metastases infiltrate the entire lymph node. If the metastases' dimensions are inferior to the thickness of the lymph node parenchyma, however, the node will still appear normal on ultrasound. Differentiation of micrometastases from the lymph node thus requires needle biopsy.9

Several studies using these criteria have reported that ultrasound of the lymphatic drainage areas and regional lymph nodes is more effective than clinical examination for detecting and diagnosing metastatic disease of cutaneous melanoma. Sensitivity and specificity for ultrasound are reported as 89.2% to 99.2 % and 98.2% to 99.7%, respectively. Comparative values for physical examination are 25.2% to 71.4% (sensitivity) and 90.9% to 99.7% (specificity).10-12

The introduction of color Doppler ultrasound has made it possible to study the vascular structure of lymph nodes. Histology has shown that lymph nodes have arterial and venous systems. The main arteries enter the lymph node at the hilum and spread via arterioles that run parallel to the node's long axis. The lymph node cortex is fed by sinuous capillaries originating from these hilar and medullary vessels. The veins generally run parallel to the arteries. This normal pattern of nodal vascularity appears as a radial and longitudinal configuration from the hilus on color Doppler ultrasound, with symmetric distribution (Figure 2).

In reactive nodal disease, the diffuse nature of the histologic process is more likely to preserve the normal pattern of vascularity (Figure 3). Changes in nodal architecture observed in malignant disease are dependent on the extent of tumor infiltration.

The original lymph node architecture is rarely altered during the early stages of microinfiltration. These malignant lymphadenopathies may have a small L/T ratio and hilar type of vascular pattern. As malignancy advances, the node will increase in size and adopt a more spherical shape as a result of neoplastic colonization, necrosis, or desmoplatic reaction. Vascularity may increase as a result of angiogenesis, or compression by neoplastic tissue or vascular enhancement may cause flow to decrease. Capsular vessels may be recruited in this case to supply the tumor with aberrant feeding vessels in the periphery. Corresponding vascular presentations show a spotted, peripheral, or mixed pattern.13,14 Hilus vessels are commonly absent, and perfusion is reduced in lymph node metastases of cutaneous melanoma.15 Needle biopsy is required in lymph nodes with a normal appearance.

Power Doppler ultrasound is less directionally dependent than color Doppler, has higher sensitivity, and shows the vascular contours with better contrast. These advantages make it a useful tool for studies of tumor vascularity. Color Doppler criteria have been adapted for power Doppler studies, and the reported results are similar (Figure 4).16,17

Contrast-enhanced ultrasound increases the amplitude of signals and improves visualization of vessel topography in lymph nodes.18 D-galactose-based agents such as Levovist provide additional information on vascularity, which can improve differentiation in small lymph nodes, in cases where gray-scale and color Doppler imaging gives ambiguous results.19

The use of second-generation contrast agents based on sulfur(Drug information on sulfur) hexafluoride microbubbles (SonoVue), along with a transducer with a low mechanical index (i.e., tissue is exposed to low pressure), allows the bubbles to emit harmonics that can be detected by gray-scale harmonic imaging techniques.20 A detailed visualization of the vascularity of lymph nodes is then possible.

Harmonic imaging allows real-time analysis of all vascular phases, including the parenchymal phase, even in normal-sized nodes. Intranodal focal "avascular" areas, representing neoplastic cells or necrosis, can also be seen. Preliminary data suggest that this novel imaging modality can improve the differentiation of malignant lymph nodes from reactive nodes and fine-tune the selection of nodes submitted for fine-needle aspiration biopsy.8

• In-transit and satellite metastases. Many initial recurrences of cutaneous melanoma are locoregional metastases in superficial soft tissues that are discovered on physical examination. Many patients with locoregional recurrence can be cured with aggressive resection. Soft-tissue metastases may be difficult to feel, however, if they are small, deep below the skin's surface, or located near a surgical scar.

High-frequency ultrasound can detect impalpable metastases and help differentiate between benign and malignant palpable lesions. Superficial melanoma metastases are seen on ultrasound as well-defined hypoechoic lesions with smooth or lobulated contours, mild to moderate heterogeneity, and increased acoustic enhancement.

Because melanoma metastases tend to be cellular, they have fewer acoustic interfaces than surrounding soft tissue. This explains the hypoechoic appearance, relative homogeneity, and increased acoustic trough transmission that is most often seen. The moderate or marked heterogeneity seen in some lesions could be related to necrotic areas or fibrous tissue.21,22

Color Doppler often reveals internal flow, given that melanoma incites angiogenesis (Figure 5). Internal flow may not be seen in cases of necrosis, undetectably slow flow, or very small tumor vessels that are unable to produce signal of sufficient amplitude. Metastatic lesions larger than 7 mm show internal vascularity on power Doppler imaging.23

EXAMINATION DETAILS

Ultrasound examination of subcutaneous lesions and lymph node basins must be performed on high-frequency scanners (~7.5 to 15 MHz) with color functionality to display intranodal or intratumoral vessels. Color Doppler will provide information on flow direction for these vessels. Power Doppler is more sensitive but does not offer information on flow direction.

Tissue harmonic imaging is a relatively new technique that can improve imaging resolution by moving to higher frequency signals. Harmonic imaging relies on the frequency change that ultrasound pulses undergo as they propagate through target media. This change is due to the normal resistance of tissue. The resultant signal change (harmonic) is essentially a doubling of the transmit frequency.

The examination procedure must include careful scanning of the palpable mass or surgical scar and a circular area of 10-cm radius around them. Scanning should be performed consecutively along the in-transit route from the primary tumor location to the regional lymph node chain:

• ipsilateral inguinal nodes for primary lesions in the lower extremities;
• ipsilateral axillary nodes for primary lesions in the upper extremities
• bilateral cervical nodes for primary lesions in the head or neck; and
• bilateral axillary and inguinal lymph nodes for primary melanomas of the trunk.

We use a high-frequency 6.6 to 9-MHz harmonic linear-array transducer (Aplio, Toshiba Medical Systems) with color/power Doppler parameters optimized for superficial lesions and slow flow.

A study of 19 patients revealed 51 locoregional metastases that were subsequently correlated with histology. Nine of these were lymph node metastases, ranging in size from 13 to 23 mm (median: 14.76 mm). The remainder (42) were soft-tissue metastases, ranging from 3 to 27 mm (median: 9.62 mm). These measurements refer to the longest length of the lesions.

The features observed on ultrasound agreed with those predicted in the literature. Metastatic lymph nodes demonstrated typical morphologic and vascular changes. Metastatic extension was limited in two cases, and the L/T ratio was > 2. The echogenic hilus was conserved in one of these cases, but an intranodal hypoechogenic lesion was identified with power Doppler.

Satellite and in-transit metastases were found mostly in subcutaneous tissue. They were generally round or oval in shape, sometimes lobulated, with well-defined contours and displayed low echogenicity. Approximately half were homogeneous in texture, and half were heterogeneous. Color Doppler signals were present in many lesions, and power Doppler signals were detectable in all lesions, including the smallest ( < 5 mm). Most of these lesions were not detectable clinically.

RIVAL OPTIONS

Early detection of locoregional metastases may be of great value when treating patients with cutaneous melanoma. Physical examination alone has low sensitivity to metastases and frequently detects only gross nodal involvement.11 Ultrasound is more effective than clinical examination in the detection and diagnosis of locoregional metastatic disease, but it cannot detect micrometastases.

Examination of lymphatic drainage should be included in the follow-up of all melanoma patients, particularly those at high risk of developing metastatic spread. An increasing number of sites now accept sentinel lymph node biopsy as a standard tool for initial disease classification. Detection of regional lymph node micrometastases using this technique has proven to be a good predictive marker for disease progression.24 Sentinel lymph node biopsy is costly, however, and has a high postoperative complication rate (10% to 30%).25

The number of biopsies required may be reduced by incorporating ultrasound into the preoperative staging workup. Lymph node ultrasound scanning is a noninvasive and relatively inexpensive technique. Detection of lymph node macrometastases on ultrasound avoids the need for sentinel node biopsy in approximately 10% of patients and reduces healthcare costs.25

Application of FDG-PET shows great promise in the detection of metastatic cutaneous melanoma, particularly in distant metastases.26 This technique is expensive, however, and has limited availability in many countries. FDG-PET can also generate false-negative results when micrometastases or lesions smaller than 10 mm are actually present. It is thus of limited use in patients with early-stage disease and cannot replace sentinel node biopsy, which is much more sensitive in detecting microscopic lymph node metastases.27 Superficial lesions smaller than 10 mm are easily detected by ultrasound, which may be a complementary option.

High-frequency ultrasound is a useful technique in the initial evaluation and subsequent follow-up of patients with malignant melanoma. This is particularly true for systems that offer harmonic and Doppler imaging. Ultrasound can detect lymph node macrometastases and the smallest ( < 5 mm) in-transit metastases. It is a fast and inexpensive technique that should be performed regularly to detect locoregional melanoma recurrence during routine follow-up visits.

DR. CALVO LOPEZ is head of the abdominal radiology section, and DR. VALLEJOS ROCA is a radiologist in this section, both in the department of radiology, Puerta del Mar University Hospital, Cadiz, Spain.