LAUSR

Melanoma incidence has risen rapidly over the past decades. Although melanoma is the most lethal skin cancer, of the 106 100 people with invasive melanomas diagnosed annually in the USA, over 93 3% will survive at least 5 years. While people diagnosed with bulky, thick melanomas are at obvious risk of melanoma metastasis and mortality, many patients who ultimately succumb to melanoma are initially diagnosed with thin, clinically and histologically less concerning lesions. In the absence of efficacious treatment, melanoma prognostic factors only predicted the inevitable. With the advent of targeted melanoma therapies and immunotherapy, predicting melanoma outcomes and treatment response became essential. Although several novel prognostic factors and predictive tools have been described, one of the most critical melanoma prognostic factors remains ulceration. Ulceration is a time-tested poor prognostic factor in thin and thick primary melanomas, which retains prognostic significance in stage III melanomas. Ulceration is also one of the few prognostic factors with established therapeutic relevance. Ulceration was once viewed as a function of fast growth and potential trauma. However, biological differences in ulcerated vs. nonulcerated melanomas were identified, suggesting that ulceration is more than simple mechanical damage to the epidermal barrier. Moreover, besides ulceration, other epidermal changes with diagnostic and prognostic significance were also described in melanoma. Dermoepidermal cleft formation and consumption of the epidermis were recognized as epidermal features with a diagnostic and potentially prognostic value in melanoma. In addition, AMBRA1 (autophagy and beclin 1 regulator 1) and loricrin expression loss in the melanoma-associated epidermis were shown to signal poor outcomes. All of these epidermal features were significantly associated with ulceration but were also found in nonulcerated melanomas. Thus, these epidermal features were thought to represent a preulceration phenotype. Although the prognostic role of these epidermal features suggested the existence of a possible interplay between melanoma cells and the epidermis, the basis of the interplay between malignant melanocytes and the overlying epidermis has until now remained elusive. In the current issue of the BJD, Cosgarea et al. have published groundbreaking findings suggesting a link between epidermal changes in melanoma and melanoma–epidermal paracrine signalling. They showed that increased transforming growth factor (TGF)-b2 in the tumour is associated with loss of epidermal AMBRA1 and loricrin, and increased metastasis risk. Moreover, they found that TGF-b2 can directly inhibit expression of AMBRA1 in its role as a keratinocyte autophagy regulatory protein and consequently cause keratinocyte loricrin and claudin-1 downregulation. These findings are intriguing because TGF-b in invasive melanoma is an established driver of melanoma genesis and metastatic spread, and a potential target for melanoma therapy. Therefore, a paracrine link between TGF-b2 signalling and features of epidermal barrier deterioration indicates that these features, similar to ulceration, may bear prognostic and therapeutic significance. Furthermore, the findings highlight the complexity of TGF-b signalling in ulceration by showing that TGF-b signalling, a mediator of wound healing in ulceration, serves as a potential initiator of ulceration in melanoma, thus broadening our understanding of the complex role of inflammatory mediators in melanoma progression. The suggested TGF-b2-based melanoma–epidermal paracrine signalling establishes for the first time a possible biological link between the previously recognized features of AMBRA1 and loricrin loss, dermoepidermal clefting, and consumption of the epidermis and progression of the underlying invasive melanoma. More importantly, the findings also indicate that these epidermal features of preulceration may represent steps on a ladder towards clinically apparent macroscopic ulceration. A better understanding of the uncovered melanoma–epidermal paracrine interplay is expected to help develop markers to better identify individuals at high risk for melanoma progression who may benefit from early adjuvant therapy.