Treatment of Head and Neck Melanoma, Lentigo Maligna Subtype: Title and subTitle BreakA Practical Surgical Technique

The incidence of melanoma continues to rise faster than that of any other human cancer. It is predicted that 1 in 74 people born in the United States in the year 2000 will develop a melanoma during their lifetime. ^Article Genetic and environmental factors are causative. The most significant environmental factor is UV light. Blistering and/or peeling sunburns in childhood, teenage, and young adulthood years are a universal risk factor for melanoma. Four main histological pattern subtypes of melanoma are generally recognized: superficial spreading, nodular, acraloral lentiginous, and lentigo maligna (LM). The LM subtype is characterized by development within a background of chronically sun-damaged skin and is the most common melanoma subtype located on the head and neck. ^Article It is an in situ melanoma that usually grows slowly and proliferates until it eventually invades the dermis, heralding invasive lentigo maligna melanoma (LMM). Melanoma subtype by itself does not influence prognosis after correction for tumor thickness and other prognostic variables. However, the LM/LMM pattern is associated with greater and variable subclinical extension, often making complete excision more difficult, with a higher risk of local recurrence, using standard surgical excision methods and standard bread-loaf histological margin control. ^Article Johnson et al ^Article first reported an alternative staged excision approach with total histological margin evaluation, the "square" procedure, in 1997 based on their initial experience from 1994 to 1996. The procedure is reviewed herein, updated with modifications and current information, and based primarily on experience in our patient population from 1997 to present. Its usefulness for treating lesions on the head and neck is stressed.

Several treatment approaches have been recommended for the treatment of head and neck LM and LMM. Destructive techniques reported for LM (not LMM) include cryosurgery, radiotherapy, electrodesiccation and curettage, laser surgery, and therapy with topical agents. Local recurrence rates with superficial destructive modalities for LM are relatively high owing to the failure to treat the atypical junctional melanocytes that often percolate deeper along adnexal epithelium and the inability to identify potentially extensive superficial silent subclinical peripheral extension. ^Article Destructive approaches may be useful in the setting of high comorbidity. Any destructive technique used for treatment needs to extend at least to the base of the adnexal structures in the middle to deep dermis to obtain high local control rates. Furthermore, cryosurgery and laser surgery may destroy the melanocyte's ability to make pigment, making a local recurrence amelanotic and therefore difficult to follow up and/or detect early. Complete surgical excision with total histopathologic margin evaluation results in the highest long-term local control and cure rate for both LM and LMM and is the treatment of choice. ^Article

The current recommendations for surgical margins for melanoma are based on 2 randomized prospective trials, nonrandomized clinical series, expert consensus groups, and even editorials. ^{Article - Article} Surgical margins of 0.5 cm for in situ melanoma and 1.0 to 2.0 cm for invasive melanoma are currently recommended, depending primarily on tumor thickness and the presence or absence of ulceration. To date, all evidence-based randomized prospective trials evaluating margins have excluded melanoma on the head and neck. The present recommended margins are often inadequate for LM/LMM subtypes, particularly on the head and neck, owing to the characteristic silent extension of lesional atypical junctional melanocytic hyperplasia (AJMH). ^{Article - Article} A study using Mohs surgery with immunostaining for melanoma demonstrated that for head and neck LM (in situ) the average margin required for clearance was 0.8 cm and that 1.5-cm margins were required to clear 96% of these tumors. For LMM (invasive), the average margin required for clearance was 1.1 cm, and 2.6-cm margins were required to clear 95% of these tumors. ^Article Another study in which standard Mohs surgery was used found that the margin required for complete excision was most closely dependent on lesion diameter. ^{Article - Article} One hundred six in situ melanomas and 125 invasive melanomas on the head and neck were included in this series of 553 melanomas. For these 231 melanomas, a margin of 1.5 cm was required to clear 97% of tumors and 2.5 cm for lesions greater than 3 cm in diameter.

The goal of surgery is to excise the entire lesion using the fundamental oncological principle: tumor clearance first, reconstruction second. Minimal tissue resection of uninvolved skin on the head and neck is desirable owing to the cosmetic significance of the face and the presence of vital structures. In efforts to preserve vital functional and cosmetic features of the face, alternatives to standard surgical excision with wide margins have been explored, including Mohs micrographic surgery with and without immunostaining and Mohs surgery with rush permanent-section processing. ^{Article - Article} These surgical techniques are associated with local control rates of 90% to 99%. Mohs surgery requires considerable training and relies on frozen-section interpretation of margins. The ability to accurately interpret melanoma margins using Mohs frozen sections is based on a steep learning curve, with considerable experience and frozen sections of exceptional quality. Still, it can be done, and the addition of a labor-intensive process with immunostaining may improve accuracy. Few authors would argue, however, that permanent formalin-fixed sections are easier to process and interpret and that, in most laboratories, they are superior in quality to frozen sections for melanoma. ^Article In general, obtaining standard frozen sections for melanoma margins in the general operating room arena is of no value and should not be done. Artifacts from quick-frozen sections routinely obtained in the operating room make evaluation for lesional AJMH almost impossible. Unresected AJMH may present as an in situ local recurrence and, less frequently, as an invasive local recurrence with metastatic potential ( Article ). Also, amelanotic desmoplastic invasive melanoma, the most difficult type to detect early, most commonly arises in the LM/LMM subtype.

The typical histological appearance of LM shows areas of markedly atypical melanocytic hyperplasia in a background of significant long-term solar damage. ^Article The evaluation of the peripheral margins of LM/LMM is particularly difficult, as the AJMH progresses from marked to moderate to slight atypia. In fact, sporadic AJMH with slight atypia may be an incidental finding in some severely sun-damaged skin. The distinction between incidental and lesional AJMH with slight atypia can be particularly difficult and requires clinicopathologic correlation, dermatopathologic expertise, and high-quality histological material.

The LM/LMM subtypes treated at our institution with standard Mohs and surgical excision in the 1980s and early 1990s were associated with a significant number of local recurrences, often at 5 to 10 years after initial excision. In an attempt to develop a better approach, we used a staged procedure, first achieving total (100%) peripheral margin tumor clearance with permanent formalin-fixed histological sections (slides), followed by excision of the central lesion and reconstruction of the surgical defect after peripheral margins are clear. A multidisciplinary team approach is used, with the participation of experienced dermatopathologists and surgeons.

The clinical lesion is usually identified with the aid of a Wood lamp. The Wood lamp (UV-A blacklight) may be helpful to identify subclinical lesions with visual delineation of subclinical pigment. A surgical excisional margin of 0.5 to 1.0 cm is drawn with straight lines and sharp-angled corners. Thus, a shape such as a square, diamond, or rectangle is usually drawn around the lesion. Configuration with straight lines and sharp-angled corners is necessary to facilitate tissue processing for total (100%) peripheral margin interpretation. If circular or rounded margins are used, complete evaluation of the peripheral margins is not possible without cutting deep into the tissue block (false peripheral margins).

Local anesthesia is infiltrated in the skin and superficial adipose tissue along the planned lines of excision. Using a 2-bladed scalpel with 2 to 4 1-mm spacers ( Article ), a peripheral strip of tissue, 2 to 4 mm wide, is excised along the margin lines en bloc to the underlying adipose tissue. The outer blade corresponds to the peripheral margin. A second single-blade scalpel is used to cut the peripheral tip corners. The tissue that contains 100% of the peripheral margin is tagged with a suture for orientation. The specimen is sent to pathology and pinned to Styrofoam to prevent rolling of the edges. The pinned specimen is fixed in 10% buffered formalin for 24 hours. The specimen is paraffin embedded, and routine vertical sections containing 100% of the peripheral margins are processed. A 2-bladed scalpel is not necessary for the procedure, but makes excision of a thin strip of tissue easier. A simple cartoon map is included with the specimen for precise anatomical orientation of the peripheral margins. Careful orientation is maintained with the map and tissue inks. The excisional strip wound is simply sutured with a running nonabsorbable suture. The procedure is usually performed in a relatively short time in a treatment room with the patient under local anesthesia. The central island of tissue containing the lesion is left intact while the margins are assessed over 2 to 3 days.

The dermatopathologist examines the histological slides and notes any areas that are postive for lesional AJMH. Using the map, any corresponding areas of positivity are again excised with straight lines and sharp-angled corners in a subsequent stage(s) until all peripheral margins are free of AJMH. After the entire peripheral margin is interpreted as being free of lesional AJMH, the central island(s) of tissue is excised within the defined clear peripheral margins and sent for serial sectioning to assess for any invasive component. Reconstruction is performed with the confidence that 100% of the peripheral margin has been evaluated and interpreted as being free of pathological AJMH. A representative case illustrating the technique is demonstrated in Article and Article .

Processing and evaluation of the peripheral margins using permanent sections usually takes 2 to 3 days. In addition, more time is needed if multiple stages are required to fully clear the periphery of lesional AJMH. Therefore, in patients with LMM, a central incision of areas in which any invasive component of the lesion is suspected is usually performed during the first stage of the square resection to reduce the risk of further invasion and metastasis. For patients with invasive melanoma that is usually 1.0 mm or more in thickness who are undergoing sentinel node mapping and biopsy, the lesion is excised most often with 1.0-cm margins, with primary closure at the time of sentinel lymphadenectomy. The peripheral margins are definitively cleared using the square procedure at a later date if necessary (eg, positive for AJMH or close to the margins).

In cases in which a delayed full-thickness skin graft is anticipated, the entire lesion is excised during the first stage. This is most useful for lesions on the nose. The peripheral margins are evaluated as previously described. The center of the specimen is serially sectioned. The wound granulates for 1 to 4 weeks to allow formation of a richly vascularized recipient graft bed, which enhances graft survival and prevents permanent contour deformities.

The square procedure represents a practical staged excision technique with formalin-fixed, vertical-section, peripheral-margin control. Definitive excision of the central lesion is delayed until 100% of the peripheral margins have been examined and are interpreted as being free of lesional AJMH. The final stage involves excision of the central lesion and reconstruction of the defect in 1 surgical procedure after peripheral margins are free of disease. The square procedure has the benefit of visualization of 100% of the peripheral margin, with tissue conservation, which may be particularly important for lesions on the head and neck. Also, it provides tissue that is technically easier to process and interpret using standard vertical sectioning. However, it cannot be performed without a dermatopathologist with considerable expertise. The differentiation between lesional atypical melanocytic hyperplasia, scattered incidental atypical melanocytes, and benign melanocytic hyperplasia, which is commonly encountered in chronically sun-damaged skin, is challenging.

We have used the square procedure in approximately 150 cases, from 1997 to 2000, with 1 short-term (<5 years) local recurrence. Longer follow-up periods of 5 and 10 years are needed to definitively determine the therapeutic local control rate with the square procedure using a prospective database. This technique is a simple, labor efficient, multidisciplinary method that results in high local control rates for LM/LMM of the head and neck. It also allows the facial plastic reconstructive surgeon to proceed with flap reconstruction with the knowledge that 100% of the peripheral margins have been examined and that the risk of local recurrence is low.

Accepted for publication February 20, 2001.

Corresponding author and reprints: Timothy M. Johnson, MD, Department of Dermatology, University of Michigan, 1910 Taubman Center, Ann Arbor, MI 48109-0314 (e-mail: timjohn@med.umich.edu).