Seborrheic keratoses and epidermal nevi: new pathogenetic insights and therapeutic implications

Figure 1. (A) Seborrheic keratoses and (B) epidermal nevus.

Fibroblast growth factor receptor 3 mutations in benign acanthotic skin tumors

Seborrheic keratoses (SK) and epidermal nevi (EN) are common benign skin tumors, which are well known among dermatologists. SK and EN share many histological features, such as acanthosis, papillomatosis, hyperkeratosis and hyperpigmentation. Furthermore, neither lesion is associated with a malignant potential. However, SK often show a large number of lesions in affected individuals (Figure 1a) and the incidence increases with age. EN are present at birth or develop in the first years of life and affected patients can display a widespread (systematized) involvement of the skin (Figure 1b). Thus, both skin tumors can cause huge disfigurement in affected individuals and often require surgical treatment despite their benign character.

Although SK represent one of the most frequent skin tumors in humans, their genetic basis has remained elusive and the number of research studies dealing with SK is rather limited, considering the high frequency of these lesions. In 2005, the presence of activating fibroblast growth factor receptor (FGFR) 3 mutations in human SK was first reported [1]. Transgenic mice expressing a mutant FGFR3 under the control of the keratin 5 promotor developed verrucous skin lesions closely resembling human SK histologically, therefore suggesting a causative role for those mutations in the pathogenesis of SK. Subsequently, activating FGFR3 mutations were also identified in common nonorganoid, nonepidermolytic EN [2]. These lesions follow the lines of Blaschko and it had been suggested that a mutational mosaicism of an unknown gene resulting from a postzygotic mutation may be the underlying genetic basis. It is now clear that FGFR3 represents the unknown gene, at least in a subset of EN, and the presence of the FGFR3 mutations only in lesional skin but not in unaffected adjacent skin, supports the mosaicism hypothesis. Organoid EN, such as nevi sebacei, did not reveal FGFR3 mutations, indicating that these mutations are specific only for a subgroup of EN. It is of interest that the same FGFR3 mutational hot spots identified in human SK and EN are found also in several cancers, such as urothelial carcinoma [3], as well as in severe skeletal dysplasia syndromes, such as thanatophoric dysplasia and severe achondroplasia with developmental delay and acanthosis nigricans (SADDAN) syndrome [4]. These mutations are usually lethal for affected individuals in the germline. However, in skin, these lethal mutations can obviously survive as a mosaicism, as proposed previously [5].

Many open questions remain regarding the role of FGFR3 mutations in benign acanthotic skin tumors. Although it is already known that these FGFR3 mutations are activating mutations resulting in a constitutive ligand-independent phosphorylation of the receptor tyrosine kinase [6], details of the pathological signaling in FGFR3 mutant cells are only partly known. Furthermore, previous results and clinical observations suggest that the functional effects of FGFR3 mutations may be rather cell-type specific [7]. For example, in chondrocytes, the FGFR3 mutations result in growth inhibition of the epiphyseal plate, causing dwarfism and skeletal dysplasia, such as thanatophoric dysplasia. By contrast, FGFR3 mutations in epithelial tissues, such as urothelium and epidermis, lead to the growth of tumors. Furthermore, the tumors demonstrating FGFR3 mutations are benign in skin and reveal a good prognosis in urothelial cancer but are associated with a worse prognosis in multiple myeloma. Since the knowledge regarding FGFR3 mutations in skin is brand-new and, therefore, limited, further studies are necessary to identify possible risk factors that cause the mutations and to determine the functional consequences of mutant FGFR3 signaling in keratinocytes. The mechanisms mediating the growth of acanthotic tumors in FGFR3 mutant skin are currently unknown. First, it seems plausible that a gain of function mutation of a growth factor receptor may lead to an enhanced proliferation of the cells. The proportion of basal cells undergoing proliferation was indeed significantly higher in transgenic mouse epidermis carrying a FGFR3 mutation than in control skin [1]. Conversely, enhanced immunohistochemical expression of bcl-2 and p16 in SK suggest that anti-apoptotic mechanisms and induction of senescence may also be involved in the pathogenesis of FGFR3 mutant skin tumors.

Another important question is which genetic alterations are involved in skin tumors without FGFR3 mutations. The frequency of FGFR3 mutations ranges from 39 to 85% in SK [1,8] and, in EN, 33% of patients revealed these mutations [2]. The patients without FGFR3 mutations at the investigated loci may show mutations in other regions of the FGFR3 gene that have not been analyzed thus far. Other promising candidates are signaling molecules downstream of FGFR3, such as phosphoinositide 3 kinase (PIK3CA). In superficial papillary bladder tumors, PIK3CA mutations are associated with FGFR3 mutations [9]. Further studies have to investigate a possible similar role of PIK3CA mutations in acanthotic skin tumors. Since FGFR2 is homologous to FGFR3 and some skeletal dysplasia syndromes demonstrate locus heterogeneity for both receptors, FGFR2 represents another promising candidate gene.

Therapeutical implications

Although a lot of open questions have to be addressed for the understanding of mutant FGFR3 signaling in SK and EN, the new insights in the pathogenesis of these skin tumors will open the gate for possible new therapeutic strategies. Small specific molecule tyrosine kinase inhibitors of FGFR3, such as SU5402, PD173074 and PKC412, are already available and some are being evaluated currently in Phase II trials for the systemic treatment of hematopoietic malignancies (e.g., PKC412) [10]. In vitro experiments demonstrated that these molecules effectively inhibit the growth of FGFR3 mutant tumor cells [11]. Therefore, a broader use of these drugs in other disorders associated with an increased activity of FGFR3 seems possible.

SK and EN of the human skin represent promising targets for FGFR3 inhibitors, since FGFR3 mutations can be involved causally in their pathogenesis. Furthermore, these skin lesions are accessible to external therapies, which may reduce side effects compared with systemic treatment. Provided that these FGFR3 inhibiting drugs can be applied in a topical formulation and will penetrate through the corneal layer, a couple of dermatological applications appear to be possible.

EN usually present very early in childhood. Since these often widespread lesions can be very disfiguring, the parents of the patients are often concerned despite the benign character of EN and ask for treatment of these lesions. Surgery or ablative laser treatment are then the main options for the treatment of EN. However, these procedures may be complicated by significant (irreversible) scar formation, which is a particular problem in exposed areas, such as the face. Furthermore, in many cases, the lesions are too extensive for an excision strategy. These invasive procedures also require local or general anesthesia, which may not be desirable in very young children. Likewise, adults with a high number of SK disseminated over the body frequently visit a dermatologist and wish for removal of the benign but cosmetically disturbing lesions. Since the prevalence of SK increases with age and the proportion of elderly people in the population of Western countries is growing, it is expected that the number of patients with multiple disfiguring SK will also increase over the next few decades. The standard therapy for SK is curettage but this therapy is sometimes limited owing to numerous lesions and the risk of scar formation.

All patients presenting with disfiguring EN or SK may benefit from a possible noninvasive therapy with topical FGFR3 inhibitors in the future but, currently, this idea is hypothetical because of many unanswered questions. It is expected that the application of drugs has to be repeated periodically because the inhibitors will cause a transitory block of the constitutive FGFR3 activation in the mutant keratinocytes rather than an elimination of the mutant cell clones. Animal models are required to prove the efficacy of FGFR3 inhibitors for the treatment of acanthotic skin tumors and also to clarify treatment-related questions, such as the concentration of the agent and the frequency of the drug application, which is necessary for a successful treatment. The involvement of other genes in the pathogenesis of EN and SK would limit a successful therapeutical application of FGFR3 inhibitors to a subset of FGFR3 mutant skin tumors. However, in this case, other inhibitors acting downstream of FGFR3 may be useful since various mutations could affect the same pathway.

Conclusion

Activating FGFR3 mutations in benign acanthotic skin tumors represent a very exciting and new field in dermatological research and represent a promising new target for future noninvasive topical treatment strategies of these very common skin lesions. However, further studies are required to improve understanding of the role of mutant FGFR3 in benign acanthotic skin tumors prior to possible therapeutical application.