We underestimated tumour extent at only 4% of margins. Tissue shrinkage was the most important factor affecting eventual histological margin. Our novel technique allowed us to demonstrate that this shrinkage is not uniform across the specimen, but is disproportionately high in the tumour-free margin. This suggests that previous estimates of margin shrinkage, based on whole-specimen contraction measurements, may have been erroneously low.

Prognosis has been shown to be worse for patients with multiple primary melanomas than those with a single melanoma. One recent retrospective study showed that older, white men were at higher risk of multiple primary melanomas. In our institution 2057 melanomas were diagnosed between January 1994 and March 2016. We identified 99 (4.8%) patients who had multiple primary melanomas. The average number of melanomas was 2.5 (range: 2-10). The site for first and second melanomas was similar in 30%. We found that subsequent melanomas were more likely to be in situ and thinner in terms of Breslow thickness and Clarks level. The commonest subtypes were superficial spreading and lentigo maligna. The commonest sites involved were the lower limbs. We found no significant difference in age between our total melanoma group and the multiple primary melanoma group (64 and 66 years, respectively). The average time of diagnosis between the first and second melanomas was 33.7 months. Additionally, 70% of second melanomas were diagnosed within 2 years of first diagnosis, highlighting the importance of ongoing skin surveillance in patients with a recent diagnosis of melanoma.

Understanding the events at a protein level that govern the progression from melanoma in situ to invasive melanoma are important areas of current research to be developed. Recent advances in the analysis of formalin-fixed, paraffin-embedded tissue by proteomics, particularly using the filter-aided sample preparation protocol, has opened up the possibility of studying vast archives of clinical material and associated medical records. In the present study, quantitative protein profiling was performed using tandem mass spectrometry, and the proteome differences between melanoma in situ and invasive melanoma were compared. Biological pathway analyses revealed several signalling pathways differing between melanoma in situ and invasive melanoma, including metabolic pathways and the phosphoinositide 3-kinase-Akt signalling pathway. Selected proteins of interest (14–3-3ε and fatty acid synthase) were subsequently investigated using immunohistochemical analysis of tissue microarrays. Identifying the key proteins that play significant roles in the establishment of a more invasive phenotype in melanoma may ultimately aid diagnosis and treatment decisions.