Controversies in the management of lentiginous melanoma in situ

In this issue of Expert Review of Dermatology, Abdelmalek et al. review the technique of staged excision (SE) for lentiginous melanoma in situ (LMIS) [1]. The topic is of great clinical interest, given the increasing incidence of this cutaneous malignancy and the multitude of management options available to the practicing clinician. Indeed, recent data suggest that lentiginous melanoma is now the most common form of melanoma [2]. Randomized controlled trials are lacking with regard to the treatment of LMIS, so it is not surprising that a number of controversies have arisen in the management of this disease. Some of the more notable issues are discussed here.

What is the appropriate terminology for this disease?

The term ‘lentiginous melanoma’ is gaining acceptance as a distinct subtype of melanoma [3]. This term encompasses lentigo maligna, lentigo maligna melanoma, as well as melanomas on sun-damaged skin with other histologic features. The term LMIS has several advantages over the term lentigo maligna. While dermatologists and dermatopathologists are familiar with the meaning of the term lentigo maligna, other medical practitioners may have less familiarity; as such, the term may be misinterpreted to represent either a benign solar lentigo or, alternatively, an invasive melanoma. This potential confusion can have treatment implications. LMIS has the advantage of specifically defining the tumor as a form of melanoma. Second, the term ‘lentiginous’ will alert the surgeon to the likelihood of subclinical extension of the lesion and the potential that a standard surgical margin may be inadequate.

Does LMIS require treatment?

In reviewing the various treatment options, patients and physicians alike may ask whether any treatment is, in fact, necessary. LMIS is an epidermal tumor with a growth pattern that is typically slow and peripheral. In a survey of 597 board-certified US dermatologists, over 10% of respondents hold the opinion that melanoma in situ(MIS) is a premalignant condition [4]. In lieu of the potential adversities of active treatment, close clinical observation for this disease has been advocated in the literature [5]. However, this approach seems unwise for two primary reasons. First, what appears to represent in situ melanoma upon biopsy of a portion of the lesion (even the most suspicious-appearing portion) is not infrequently (10–32% in most studies) found to be invasive melanoma when the entire lesion is excised [6–8]. Second, some cases of in situ melanoma will certainly progress over time to invasive disease, with consequent mortality risk [9,10]. A risk of progression of 5% is a frequently quoted (and probably conservative) estimate [7,11]. Although clinicians may recognize characteristics of high-risk tumors, it is currently impossible to predict which cases will progress to invasive disease. Similarly, even with the most advanced dermatoscopic and in vivo imaging techniques available, it is still impossible to ascertain with certainty via clinical examination whether invasive disease is concealed within a given tumor. Indeed, many surgeons can cite memorable examples from their clinical practice of a seemingly adequately treated case of melanoma, even MIS, which presented with lymph node metastasis years later. While this scenario can never be completely eliminated, it is certainly minimized with therapy. For these reasons, LMIS warrants treatment.

What is the most appropriate treatment?

Treatments for LMIS can be broadly classified into surgical and nonsurgical treatment. Nonsurgical treatments include topical therapies, radiotherapy and cryotherapy. Surgical treatments include excision with margin analysis and destructive treatments without margin analysis. With regard to increasing the chance of tumor eradication while minimizing the chance of recurrence, surgical treatments with margin analysis outperform nonsurgical or destructive treatment. However, not every case can be managed surgically and not every patient will make the informed decision to consent to surgery after being presented with the options.

Superficial destructive techniques include cryosurgery, curettage and electrodesiccation, radiotherapy and laser therapy. Advocates of these techniques report excellent results, with low rates of long-term recurrence [12–15]. These results reflect the well-honed technical expertise of the authors. Disadvantages of nonsurgical approaches include uncertainty of margin control, the possibility of not diagnosing unsuspected foci of invasive lentigo maligna melanoma and the risk of progression to invasive lentigo maligna melanoma in incompletely treated lentigo maligna [16]. In a large retrospective study, the recurrence rate of MIS managed with destructive therapies was 31%, compared with a recurrence rate of 7% for surgical excision [17].

Immunotherapy with imiquimod has generated interest, although long-term follow-up data is needed [18,19]. In a recent review of four clinical trials and 11 case reports of patients treated with topical imiquimod for lentigo maligna, 59 out of 67 (88%) patients showed at least a partial response to this immunotherapy over short follow-up intervals [19]. However, the development of invasive lentigo maligna melanoma in two patients is concerning [20,21]. The concept of treatment with imiquimod prior to SE [22] is intriguing and may prove to be a valuable approach to managing anatomically challenging cases.

Is melanoma a candidate disease for tissue sparing surgery?

To date, wider margins have not shown a significant survival benefit compared with narrow margins for invasive melanoma [23]. Thus, many experts have argued that the appropriate margin is that which simply removes the entire tumor. However, the concept of leaving no malignant cells behind is not straightforward in the management of melanoma. Nodular basal cell carcinoma, for example, tends to grow in a contiguous and histologically predictable pattern. Melanoma is more likely than other forms of cutaneous malignancy to display complex 3D growth patterns, showing discontinuous growth with skip areas of tumor advancement. Thus, while sparing normal tissue is a laudable goal, it seems prudent to have at least a small normal margin to minimize the chance of missing stray tumor nests.

What is the ideal surgical treatment?

The ideal surgical treatment is one that reliably removes the entire tumor, minimizes the chance of recurrence, minimizes loss of normal tissue and is performed in an environment that maximizes patient safety, comfort and convenience. These criteria are best fulfilled by total margin-control techniques, which are reviewed in this issue.

What is the appropriate depth of excision?

While excision to fascia is the commonly accepted depth for invasive melanoma, there are little data to support the necessary depth of excision for in situ disease. In the above-referenced survey, approximately 10% of dermatologists prefer to excise MIS to fascia, while 3% are satisfied with an intradermal depth; approximately two-thirds prefer a depth within the subcutaneous fat [4]. Since adnexal structures arising from the deep dermis often contain malignant melanocytes, my preference is a full-thickness excision to superficial fat.

What is the appropriate peripheral margin of excision?

The recommended excision margin of 5 mm for lentigo maligna is often inadequate, frequently leaving tumor at the margin. This probably contributes to the observed recurrence rate of up to 20% after standard excision [24–26]. In reviewing previous literature discussing necessary clinical margins for lentigo maligna by various techniques offering histologic margin control, 45% of tumors (151 out of 333) required over a 5-mm margin [8,27–29]. A recent study of 76 cases reported a mean surgical margin of 7.1 mm [30]. Without margin control, it is not possible to identify what portion of the tumor requires a wide or narrow margin. Thus, surgical techniques enabling total margin control are useful.

What is the best option for margin control?

Margin-control surgery, including Mohs micrographic surgery (MMS) and SE procedures, offers the best option for curative therapy [1,27–33]. MMS is, without question, the ideal margin-control option for keratinocyte malignancies, such as basal cell carcinoma and squamous cell carcinoma, but is controversial in the management of melanocytic lesions. As reviewed by Abdelmalek et al. in this issue [1], various SE procedures have been introduced in an effort to capture the tissue-sparing, efficiency and complete margin analysis of MMS, while avoiding the difficulties inherent in frozen-section interpretation of melanocytic neoplasia. My colleagues and I recently reported a comparative study of MMS versus SE with overnight preparation of paraffin sections for the treatment of lentigo maligna and lentigo maligna melanoma [34]. We found a significantly lower recurrence rate in patients treated with SE, with a mean follow-up duration exceeding 8 years [34]. Importantly, by comparing the areas of the preoperative lesion and the postoperative defect, we also found SE to be equally tissue-sparing compared with MMS [34].

A recent review of MMS for 116 cases of lentigo maligna found that frozen sections deemed clear and were subsequently thawed and paraffin embedded were verified as clear in 95.7% of cases [35]. This may be a reliable estimate of the upper-limit accuracy achievable with frozen sections when performed by an experienced Mohs surgeon. Practitioners will need to take this into consideration in deciding whether the advantages of MMS (primarily the chance to have the procedure and repair completed in a single day) outweigh the chance that perhaps one in 25 cases will be falsely clear on frozen sections.

Staged procedures require a close working relationship with a dermatopathology laboratory that is able to accurately and rapidly (ideally overnight) process the specimens. Close communication with a dermatopathologist who is experienced in the technique and its demands is also necessary. If either of these elements are lacking, a different management option might be more appropriate.

Are frozen sections equal to paraffin sections?

Frozen sections allow rapid determination of margin status, but the slides are of a lower technical quality. Paraffin sections are the gold standard (if not standard of care) for histologically diagnosing cutaneous melanoma. Accurate visualization of atypical melanocytes on frozen sections is difficult, particularly on sun-damaged skin where prominent single melanocytes may reside in the absence of disease. Reported sensitivities and specificities for interpreting lentigo maligna and lentigo maligna melanoma on frozen sections ranged from 59 to 73% and 68 to 81% in two studies, respectively [8,35], although others have found frozen sections more reliable [36].

This variability certainly indicates that some practitioners have an easier time with frozen sections than others, and also points out interobserver variability and the impact of human judgement when interpreting sections, even when identical histologic criteria are recognized. Prieto et al. reported poor agreement among dermatopathologists in correlating frozen sections and paraffin-embedded sections; discrepancies most often involved a false-negative margin on frozen sections [37]. This may potentially correlate with the finding in several studies of increased recurrence of LMIS in cases involving frozen-section histology.

How should the excised tissue be oriented prior to embedding?

Excisions processed by conventional serial cross-section (‘bread-loaf’) techniques are embedded ‘up and down’, with the prepared slide containing a vertical section through the lesion so that the superficial and deep tissues are visible in the same cut. Mohs surgery differs from this by embedding the excised tissue horizontally, or en face, such that the entire peripheral margin is analyzed in the same cut; a beveled peripheral margin allows the peripheral and deep margin to be visualized in the same histologic plane. As mentioned previously, this embedding technique is ideal for tumors with a contiguous growth pattern.

En face embedding enables the examination of 100% of the margin. Vertical embedding allows examination of only intermittent snapshots of the margin. Kimyai-Asadi et al. estimated that serial cross-sections at standard 2-mm intervals would have a 37% chance of detecting positive margins, and that 1-mm intervals would have a 58% chance of detecting positive margins [33]. Total margin analysis via vertical sections would essentially require sectioning through the entire block and, thus, an impossibly voluminous task for the pathologist. Vertical embedding does have the advantage of allowing better visualization of the tumor in the vertical plane and a better estimate of tumor depth in the case of unsuspected invasive disease.

Combining these methods, with vertical processing of the debulked tumor and en face processing of the margin, have been described in several variations of the SE technique. However, it is possible that this added complexity of tissue handling may introduce the opportunity for error and loss of topographical precision when the slides are read and correlated with the map and surgical defect.

What are the histologic criteria for diagnosing LMIS?

There are currently no uniform histologic criteria for diagnosing LMIS [38]. Features associated with LMIS include confluent atypical melanocytes, nesting of atypical melanocytes and presence of melanocytes above the dermal–epidermal junction (pagetoid spread). In our study, lentigo maligna was defined as MIS arising in actinically damaged skin, with continuous proliferation of atypical melanocytes, as single confluent cells or in nests along the basal layer of the epidermis and adnexae, without the requirement for pagetoid spread [34]. We did not consider the finding of single scattered atypical melanocytes sufficient for the diagnosis of lentigo maligna. Increased melanocyte density, confluence of up to six melanocytes and follicular penetration have been noted in ‘normal’ sun-damaged skin [39]. Of course, the biologic potential of a single atypical melanocyte is unknown.

Should immunohistochemical stains be used?

In an effort to overcome some of the difficulties in interpreting frozen (as well as permanent) sections, a number of immuno­histo­chemistry stains that highlight melanocytes have been used, including S-100, MART-1/MelanA, HMB-45 and Mel-5 [39–45]. However, these stains may highlight melanocyte hyperplasia in the absence of melanoma [35]. Some investigators have found that these immunostains may highlight pigmented actinic keratoses [46], although others have reported higher specificity for melanocytes [47,48]. Thus, these stains may help, but they certainly offer no guarantees in differentiating benign from malignant melanocytes. Another drawback to immunostains is the increased time required to process the tissue. While currently available protocols may add less than an hour to the processing time [40,43,44], this still doubles the time to produce sections without immunostains. The increased steps and potential for technicians running multiple staining protocols on the same day may also increase the potential for technical errors. Moreover, few Mohs labs (<15% according to a recent survey) currently utilize immunostains [49].

How should we evaluate the effectiveness of surgical therapy for LMIS?

Assuming treatment-specific safety and an acceptable functional and cosmetic outcome in a disease with relatively low mortality, local recurrence after therapy is probably the most clinically meaningful evaluation of the overall success of treating LMIS. Recurrence is also the most common means of comparing different treatment modalities. It is apparent that the biologic nature of LMIS renders it notoriously prone to recurrence after relatively prolonged disease-free intervals. In a retrospective analysis of nearly 2000 cases of lentigo maligna/lentigo maligna melanoma, almost 20% of recurrences occurred over 5 years after initial treatment [50]. Recurrence after seemingly adequate treatment is frequently noted, even beyond 10 years. One notable case in the literature reports recurrence of lentigo maligna (as invasive lentigo maligna melanoma) 30 years after initial treatment [51]. In our study comparing MMS and SE, nine recurrences were observed after a mean interval of 44 months [34]. The mean time to recurrence in another study with long-term follow-up was 45.6 months [12]. Informally surveying the 20 most recently published case reports and case series discussing LMIS recurrence after treatment showed that only four studies (20%) have mean follow-up intervals over 5 years [12,13,31,34]. Of these 20 reports, seven (35%) offer mean follow-up intervals of 2 years or less. While it is understandable that authors wish to share their results as soon as possible, especially when the findings relate to a new or improved modality, it would appear that meaningful conclusions regarding recurrence of this disease can only be drawn with a follow-up interval of 5 years at the very least.

In conclusion, these are some of the controversies in the management of LMIS. In the absence of a standard of care determined by randomized controlled trials, clinicians must weigh the advantages and disadvantages discussed previously, in the context of the resources available in their communities, to determine the most appropriate way to manage a given tumor in a particular patient.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.