Lack of SF3B1 R625 mutations in cutaneous melanoma

Malignant melanoma is a devastating disease worldwide [1, 2]. Curative management of melanoma is limited to the stage of localized disease. Once metastatic spread has occurred, prognosis of patients is poor. However, a number of promising new treatment regimens have been introduced recently, showing for the first time a therapy induced increase in overall survival [3, 4].

Over the last couple of decades a number of genetic alterations have been identified in melanoma. Activating driver mutations in genes such as NRAS[5] and BRAF[6] were identified in cutaneous melanoma. Losses of tumor suppressors such as CDKN2A and PTEN have been well documented [7]. In uveal melanoma a different set of genes shows recurrent mutations, including GNAQ and GNA11[8, 9], with activating mutations as well as in BAP1[10] showing inactivating mutations. The distinct mutation profiles of cutaneous and uveal melanoma are striking and support a model of different developmental pathways. However there is some overlap in tumor biology as ~80% of blue nevi, which are benign melanocytic tumors of the skin, also harbor GNAQ or GNA11 mutations, [8] and BAP1 mutations can be found in both cutaneous nevi and cutaneous melanoma [11‐14].

Both genetic and immunohistological assays are becoming more and more relevant in determining the dignity and prognosis of melanocytic neoplasms [15‐18]. Further refining which biomarkers are relevant in which settings should allow pathologists and clinicians to make more detailed diagnostic calls, leading to appropriate follow-up and treatment decisions.

Recently a recurrent mutation hotspot in SF3B1 affecting codon 625 was found in 18.6% of uveal melanoma [19]. SF3B1 mutations had been previously detected in myeloid malignancies such as CLL (chronic lymphoid leukemia) and MDS (myelodysplastic syndrome) [20, 21] and also reported in breast cancer [22]. SF3B1 is a splice factor, with mutations expected to result in altered pre mRNA splicing. However the exact target of altered splicing is unknown and might be cell type dependent [22].

Genetic classification of different melanoma subtypes has become very important, especially with the introduction of effective therapies targeting genetic alterations such as BRAF[3, 4] and KIT mutations [25]. A detailed understanding of the genetic events occurring in different tumors will most likely prove critical to further improving the therapeutic modalities for metastasized melanoma patients.

The distribution of activating oncogene mutations in BRAF and NRAS in our cohort is comparable to those reported elsewhere [7]. Overall 65% of melanoma had a BRAF or NRAS mutation in a mutually exclusive pattern. Of the three melanoma subtypes analyzed in considerable numbers (SSM, NM, ALM), percentages of BRAF and NRAS mutations combined were highest in SSM reaching 82%, lower in NM with 67% and lowest in ALM with 46%. The KIT mutation/variant identified in an ALM sample led to a p.N505V change. This is not reported to be a frequent mutation in cutaneous melanoma [27]. However p.N505H (c.1513A > C) is listed as a “variant of unknown origin” in a gastrointestinal stromal tumor in the COSMIC database [28]. The cutaneous ALM sample lacked mutations in BRAF or NRAS which could support a potential relevance, as typically KIT mutations are found to be mutually exclusive with BRAF and NRAS mutations [26]. The p.N505H (c.1513A > C) change could however also represent a rare germ-line variant, which we could not check as corresponding normal DNA was not available.

We obtained high quality sequencing results allowing analysis of exon 14 and in particular codon 625 of SF3B1 in 81 samples and found no mutations. This argues against a major role for SF3B1 in tumorigenesis or progression of cutaneous melanoma. In uveal melanomas, mutations were primarily found in tumors with a favorable prognosis [19]. Future studies could analyze if SF3B1 mutations occur in benign cutaneous melanocytic tumors (nevi) or potentially in sites other than in codon 603–641 of exon 14 of SF3B1.

In recent years genetic analyses identified a number of key genes involved in melanoma formation or progression. Interestingly, almost all of those described in cutaneous melanoma are not known to be relevant in uveal melanoma [29, 30]. In contrast, genetic alterations in uveal melanoma such as GNAQ and GNA11 mutations were also found in selected cases of cutaneous melanoma and are frequently found in blue nevi (benign cutaneous melanocytic tumors) [9]. BAP1 inactivating mutations are found in cutaneous nevi and melanoma, although considerably less frequently than in uveal melanoma [12]. Our current study would signify that SF3B1 mutations, occurring in almost 20% of uveal melanoma, [19] do not play a major role in cutaneous melanoma. We believe this highlights once more the genetic differences between uveal and cutaneous melanoma and the need for development of melanoma subtype specific therapies.