Desmoplastic melanoma: a brief review and the efficacy of immunotherapy

KEYWORDS:

• Checkpoint inhibitors
• immunotherapy
• nivolumab
• renal cell carcinoma
• targeted therapy

1. Introduction

Desmoplastic melanoma (DM) is a rare histological variant of melanoma that was first described by Conley, Lattes and Orrin [1] in 1971 when they presented a detailed report of 7 cases of inconspicuous superficial melanotic lesions located in the head and neck regions that evolved into an aggressive, infiltrating, and potentially metastasizing tumor, characterized locally by large subcutaneous tumefactions of fibrous appearance. Outcomes were usually fatal, resulting from widespread dissemination

2. Definition

DM is considered a variant of spindle cell melanoma, arising from intra-epidermal melanocytic proliferation. Due to this histological presentation and clinical behavior, it has been proposed as a sarcomatoid variant of melanoma [2]. DM’s can occur as either a pure or mixed histology including features of non DM.

In 1979, neurotropic melanoma, a variant of DM was first described by Reed & Leonard. Neurotropic melanomas resemble neural tumors and can invade the epineurium and endoneurium of nerve fascicles [3].

3. Epidemiology

Overall, DM accounts for less than 4% of all primary cutaneous melanomas and recent data from the Surveillance, Epidemiology and End Results (SEER) program from the National Cancer Institute (NCI) revealed that it has a male to female ratio of approximately 2:1, the mean age at diagnosis is 66 years, and the incidence rate is 2 per million, with the incidence increasing steadily over the past 15 years [4].

4. Risk factors

The etiology of DM is uncertain, but it has been linked to chronic ultraviolet exposure as it often presents as an indurated lesion in chronically sun-damaged skin. It commonly presents in the head and neck area, extremities and trunk, but most sites of the body can be involved including oral, conjunctival and genital areas. They can present as a de novo lesion but can sometimes arise in the context of other types of melanoma such as lentigo maligna [5,6].

5. Clinical features

DM typically presents as an amelanotic, pale, fleshy nodule or plaque that can resemble a scar [1].

It is particularly prone to misdiagnosis since its presentation is often non-specific and it often lacks the typical clinical attributes of melanoma and can simulate various benign and malignant nonmelanocytic lesions.

DM is frequently associated with neurotropism, which can give rise to cranial neuropathies due to direct extension of the tumor. Neurotropic DM has been associated with worse survival [7]. Local recurrence rates have been reported to be higher when neurotropism was present [8].

6. Histopathologic features

DM is characterized by the association of invasive melanoma with abundant collagenous matrix that gives the tumor silhouette a pink scarlike appearance on H&E staining. Among DM, there is significant morphologic variability leading to further subclassification into pure DM when desmoplasia is prominent throughout the entire tumor and combined DM, when it only represents a portion of an otherwise nondesmoplastic melanoma [9].

Pure DM appears to have a distinct clinical biology and has been associated with a longer disease-free survival. Patients with combined DM have a 3.5-fold greater risk of death or metastasis than those with pure DM [9]. Nerve involvement is common and varies from focal perineural invasion to extensive neurotropism [3].

Immunophenotypically, DM is strongly and homogenously positive for S-100 protein and often negative or only focally positive for other melanocyte differentiation antigens such as tyrosinase, gp100, Melan-A and MITF [10].

7. Management of local DM

The initial management for cutaneous melanoma including DM is surgical excision with negative margins. DM has been generally managed with a wide local excision. Due to its microscopic and clinical behavior routine recommendations for surgical resection may not apply to DM [11] and aggressive resection with margins of at least 2 cm have been proposed when possible in order to ensure that negative margins are obtained [9].

DM is a distinct entity when compared to other cutaneous melanomas as the majority of these tumors tend to present with an increased Breslow thickness and at an advanced Clark level likely contributing to an increased propensity towards local recurrence [9,12], with rates reported as high as 40–60% [13] compared to 1–5% for other melanomas. When DM is located to the head and neck region, cosmetic and anatomic constraints can prevent excision with adequate margins.

Given the radial growth pattern of DM local excision can often times leave positive margins leaving the clinician with the dilemma to pursue additional surgery versus adjuvant radiation. Depending on the location of the tumor one modality may be favored over another.

The high local recurrence rates have been attributed to failure to completely excised tumor projections, inadequate surgical margins, and the neurotropic nature of infiltration [8,13,14].

Local recurrence is associated with an increased risk of developing systemic metastases [1] which has been reported to occur in 11–40% of cases [6,9], with lung being one of the most common sites.

Sentinel lymph node biopsy (SLNB) remains controversial since regional lymph node involvement at the time of diagnosis is less frequent than with other cutaneous melanomas and selective SLNB for patients with additional high-risk factors has been suggested [15]. Factors that do associate with a positive sentinel node include presence of ulceration, age<40, and a thick primary, and our group does consider and offer sentinel lymph node biopsies in select patients who meet the above criteria. (1)

The role of radiation therapy as an adjuvant to wide local excision remains uncertain. It has been suggested in patients with localized DM, wide local excision with careful attention to appropriate margins produces excellent local control rates without the need for adjuvant radiation [16]. However, several studies [17,18] suggest that radiation therapy provides superior local control compared to surgery alone and it might be of value in patients with narrow or positive margins, residual gross tumor or locally recurrent DM.

8. Management of metastatic or locally advanced DM

The management of locally advanced or metastatic DM is similar to cutaneous melanoma with some notable exceptions. Like cutaneous melanoma, patients with metastatic DM rarely benefit from cytotoxic chemotherapy [19]. Unlike cutaneous melanoma, actionable mutations such as BRAF are rarely found in DM tumors [20,21]. Early observations suggested that patients with DM respond better than expected to immunotherapy, prompting a more formal analysis [22]. Unlike cutaneous melanomas, DM’s are rarely sentinel lymph node positive and therefore patients rarely develop stage three disease and therefore patients rarely receive adjuvant anti-PD1 therapies which have demonstrated improvement in relapse free survival in patients with lymph node positive, stage III cutaneous melanoma.

Based on the observation that patients with DM respond well to immunotherapy, investigators recently performed a retrospective, multi-site review of 60 patients with advanced, unresectable DM who received PD-1 or PD-L1 blockade therapy [23]. In this cohort of patients, objective tumor responses rate (ORR) were observed in 70% of patients, one of the highest response rates to single agent PD-1 blockade therapy in any pathologically defined cancer. In addition to the high ORR, the investigators also noticed that 32% of patients in this study achieved a complete response. The most frequently administered drug was pembrolizumab (75% of patients), followed by nivolumab (13%), nivolumab or pembrolizumab combined with iplimumab (5%) and the anti-PD-L1 antibody BMS-936559 (5%).

In this cohort of patients, tumors from 17 patients were available for whole exome sequencing revealing the genetic signature of ultraviolet light-induced DNA damage. Mutations in NF1 and the absence of BRAF and NRAS mutations were the most common genetic event seen in 82.4% of samples. In addition, the mutational burden of these tumors was higher than typical cutaneous melanoma tumors that lack NF1 mutations and typically harbor mutations in BRAF and NRAS. Finally, when compared to non-DM melanomas, DM tumors were more likely to stain positive for PD-L1 by immunohistochemistry.

9. Conclusion

Desmoplastic melanoma is a relatively rare variant of melanoma with unique clinical and pathologic characteristics as well as genetic drivers. Unlike cutaneous melanoma, DM usually lack pigment and are characterized by dense spindle-shaped melanoma cells with abundant fibrous connective tissue. DM typically presents at a later stage and often times have a neurotropic pattern of spread likely influencing the high rate of local recurrence seen in this disease. While DM lacks actionable mutations, it often harbors a high mutational burden possibly due to the primary location of the head and neck and thus excess exposure to UV light induced DNA damage, providing the rationale behind the high response rate observed with immunotherapy.

Recently, investigators demonstrated that patients with locally advanced or metastatic DM experienced a response rate of 70% to immunotherapeutic agents such as anti-PD1 and/or anti-PD-L1. In a subset of these patients, the investigators then confirmed that the mutational burden and PD-L1 expression of these tumors was higher than non-DM melanoma likely explaining the high ORR.

10. Expert commentary

DM is a rare histologic variant of melanoma and given its rarity, there is a lack of prospective data to guide the clinical management of patients with DM. Given the lack of data, treatment decisions are based on less than ideal single armed studies and retrospective experiences. Recently, investigators demonstrated that patients with locally advanced or metastatic DM had an exceptionally high response rate of 70% to immunotherapy. While this study was retrospective in nature which lends itself to recall bias, the biologic explanation of the high response rate to immunotherapy provides further validation that patients with DM do in-fact have improved response rates to immunotherapy. Given the significant activity of immunotherapy and the few, if any, systemic treatment alternatives, immunotherapy should be considered the standard of care for patients with advanced DM.

While it is clear that immunotherapy should be considered the standard of care for patients with advanced DM, what remains unclear is the role of immunotherapy for patients with non-advanced disease. Given that DM’s most commonly occur in cosmetically important areas such as the head and neck, alternatives to potentially disfiguring surgery and/or adjuvant radiation for patients with DM would be welcomed. With that being said, we believe that a well-designed, well-controlled prospective study is necessary to confirm that this approach is viable for the majority of patients with DM. We anticipate within the next five years we will have more clinical experience with the use of immunotherapy in patients with DM. If the activity of immunotherapy in patients with DM mirrors the activity recently demonstrated, we predict that up-front immunotherapy will be incorporated as a possible first line option in patients who want to avoid surgery or adjuvant radiation.

Key issues

• Desmoplastic melanoma is clinically and genetically unique when compared to typical cutaneous melanoma.

• Desmoplastic melanomas have higher mutational tumor burden and are more likely to express PD-L1 than non-desmoplastic cutaneous melanomas.

• Immunotherapy with anti-PD1 or anti-PD-L1 agents appears to demonstrate excellent activity in patients with advanced desmoplastic melanoma.

• Further work is necessary to determine the use of immunotherapy earlier in the course of patients with desmoplastic melanoma.

Declaration of interest

RW Joseph is in the advisory/consulting boards for Bristol-Myers Squibb, Merck, Exelixis, Incyte and Novartis. He has performed clinical trials with Bristol-Myers Squibb, Merck, Roche, Amgen, X4P and Syndax. The authors have no other 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 apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.