*Divisions of Dermatopathology and Dermatology, Department of Pathology Albany Medical College, Albany, NY; †Troy, NY; ‡Sikl's Department of Pathology, Charles University Medical Faculty Hospital, Pilsen, Czech Republic.
To the Editor:
Recently, Kazakov et al1 performed a clinicopathologic and genotypic analysis on a series of 16 patients suffering from multiple (familial) trichoepitheliomas (MFT: MIM 601606) and found mutations in the CYLD gene in 46%. Based on the frequent presence of CYLD mutations in MFT, it, along with familial cylindromatosis (FC: MIM 132700) are considered 2 phenotypic ends within the spectrum of Brooke-Spiegler syndrome (BSS: MIM 605041), whose patients develop cylindromas, spiradenomas, spiradenocylindromas, and trichoepitheliomas.2 Herein, we document another MFT kindred with a severe disfiguring phenotype associated with a recurrent CYLD germline nonsense mutation (c.1112 C>A/p.S371X) previously reported in 6 unrelated patients: 2 with MFT,1 3 with FC,3-5 and 1 whose phenotype was not reported.6
A 66-year-old African American woman presented to dermatology clinic requesting evaluation and treatment of numerous, confluent, disfiguring, skin to pink-colored papules and nodules, the largest measuring 2cm in diameter, located primarily over her central face and most prominently along the nasolabial folds, the nose, and the preauricular region (Fig. 1). Smaller papules, ≤3 mm, scattered over the cheek and ears were also found. Her scalp was clear. A family history consisted of similar lesions affecting her brother and father, but 6 sisters and 3 other brothers were not reported to have similar facial tumors. Compared with her affected family, she reports to have the largest tumors. Shave biopsy of 2 nodules revealed trichoepitheliomas: follicular germinative cells arranged in cribriform nests surrounded by follicular stroma. Stroma-stroma clefting and formation of papillary mesenchymal bodies was evident (Fig. 1). DNA was recovered from formalin-fixed paraffin-embedded tumoral tissue microdissected from both biopsies to analyze for CYLD germline and somatic mutations (methods previously described1). A single, base substitution, germline, CYLD nonsense mutation was detected, c.1112C>A/p.S371X. In one tumor, the somatic mutation represented loss of heterozygosity; however, in the second tumor, somatic CYLD defect was not elucidated. Treatment options were discussed with the patient, but she has not elected any therapy after 3 months of follow-up.
A total of 73 germline CYLD mutations have been reported to date in BSS inclusive of MFT and FC and occur in exons 9-20, in 116 families with diverse ethnic and racial backgrounds.1,2,6-13 Of these 73 mutations, about 84% are expected to lead to truncated proteins, mostly due to frame shift (approximately 42%) or nonsense (approximately 32%) mutations. The 9 missense mutations reported to date occur only within the ubiquitin (Ub)-specific protease domain of the CYLD protein, and most were associated with MFT.2,6,10,13,14 After the c.2806C>T/p.R936X mutation affecting 8 families,2 the second most common reported mutation in BSS is the nonsense c. 1112 C>A/p.S371X mutation affecting 7 families [3 MFT families, African American, Austria, and Ireland (this report1); 3 FC families, United States, United Kingdom, and Irish3-5; and 1 phenotype not characterized, presumptively of Dutch origin6].
Of the 2 MFT with this nonsense mutation reported by Kazakov et al,1 both patients were young and had small tumors along the nasolabial folds and on the nose. It is not known whether their tumors will enlarge with age and show the disfiguring phenotype found in our patient; however, based on her family history and other reports, documenting significantly less family members affected than would be expected for an autosomal dominant gene,11 nonpenetrance of the mutation, or mild incomplete condition that is underreported would account for the wide phenotypic variation seen in these kindred. Approximately 28%-54% of patients with MFT do not have a detectable CYLD mutation,1,2 indicating genetic heterogeneity in the pathogenesis of MFT, or the presence of large deletions, insertions, or complex rearrangements of CYLD that are not detectable by direct sequencing.2 Recently, van den Ouweland et al 6 reported a large rearrangement identified by quantitative polymerase chain reaction analysis in a patient with FC, highlighting that a combination of technologies is required to fully evaluate the presence of pathogenic mutations in CYLD.6
Functionally, CYLD acts as tumor suppressor gene and encodes for deubiquitinating (DUB) enzyme that negatively regulates the nuclear factor κB and c-Jun N-terminal kinase pathways.2 Therapeutic targeting of this signaling pathway may yield more effective prevention and management of the tumors in the spectrum of BSS.
Because of their superficial nature, trichoepithelioma is amendable to ablative methods such as carbon dioxide or erbium:yttrium-aluminum-garnet (Erb:YAG) Erb:YAG laser evaporation or electrosurgery.15 Positive and negative regulators of CYLD hold promise, but they have not been extensively investigated to date. For example, the use of antagonists of CYLD such as tumor necrosis factor alpha inhibitor coupled with salicylic acid (nuclear factor κB antagonist) resulted in a gradual improvement of multiple cylindromas.16 In addition, topical application of salicylic acid to multiple cylindromas resulted in partial to complete tumor remission in a minority.17 Contrarily, topical imiquimod, an agonist of CYLD function and stimulator of tumor necrosis factor alpha and interferon gamma combined with tretitoin cleared approximately 80% of trichoepitheliomas after 3 years of therapy.18
In conclusion, we report the third MFT family with the c.1112C>A/p.S371X nonsense CYLD mutation that was associated with a severe disfiguring phenotype. As 3 FC families have also been reported with this nonsense mutation,3-5 a direct genotype-phenotype correlation is not apparent, and other genetic factors combine with loss of CYLD function to produce multiple facial cutaneous adnexal tumors of varied type and size. The fact that CYLD function affects both morphogenesis of the folliculosebaceous-apocrine unit and inflammation may underlie these complex phenotypes.
Konstantinos Linos, MD*
Joseph Schwartz, MD†
Dmitry V. Kazakov, MD‡
Tomas Vanecek, PhD‡
J. Andrew Carlson, MD, FRCPC*
*Divisions of Dermatopathology and Dermatology, Department of Pathology Albany Medical College, Albany, NY; †Troy, NY; ‡Sikl's Department of Pathology, Charles University Medical Faculty Hospital, Pilsen, Czech Republic
1. Kazakov DV, Vanecek T, Zelge B, et al. Multiple (familial) trichoepitheliomas: a clinicopathological and molecular biological study, including CYLD and PTCH genes analysis, of a series of 16 patients. Am J Dermatopathol
. In press.
2. Blake PW, Toro JR. Update of cylindromatosis gene (CYLD) mutations in Brooke-Spiegler syndrome: novel insights into the role of deubiquitination in cell signaling. Hum Mutat
3. Bowen S, Gill M, Lee DA, et al. Mutations in the CYLD gene in Brooke-Spiegler syndrome, familial cylindromatosis, and multiple familial trichoepithelioma: lack of genotype-phenotype correlation. J Invest Dermatol
4. Bignell GR, Warren W, Seal S, et al. Identification of the familial cylindromatosis tumour-suppressor gene. Nat Genet
5. Saggar S, Chernoff KA, Lodha S, et al. CYLD mutations in familial skin appendage tumours. J Med Genet
6. Van den Ouweland AM, Elfferich P, Lamping R, et al. Identification of a large rearrangement in CYLD as a cause of familial cylindromatosis. Fam Cancer
. 2010. [Epub ahead of print].
7. Amaro C, Freitas I, Lamarao P, et al. Multiple trichoepitheliomas-a novel mutation in the CYLD gene. J Eur Acad Dermatol Venereol
8. Kazakov DV, Schaller J, Vanecek T, et al. Brooke-Spiegler syndrome: report of a case with a novel mutation in the CYLD gene and different types of somatic mutations in benign and malignant tumors. J Cutan Pathol
. 2010; 37:886-890.
9. Scholz IM, Numann A, Froster UG, et al. New mutation in the CYLD gene within a family with Brooke-Spiegler syndrome. J Dtsch Dermatol Ges
10. Wang FX, Yang LJ, Li M, et al. A novel missense mutation of CYLD gene in a Chinese family with multiple familial trichoepithelioma. Arch Dermatol Res
11. Rajan N, Trainer AH, Burn J, et al. Familial cylindromatosis and Brooke-Spiegler syndrome: a review of current therapeutic approaches and the surgical challenges posed by two affected families. Dermatol Surg
12. Kazakov DV, Thoma-Uszynski S, Vanecek T, et al. A case of Brooke-Spiegler syndrome with a novel germline deep intronic mutation in the CYLD gene leading to intronic exonization, diverse somatic mutations, and unusual histology. Am J Dermatopathol
13. Sima R, Vanecek T, Kacerovska D, et al. Brooke-Spiegler syndrome: report of 10 patients from 8 families with novel germline mutations: evidence of diverse somatic mutations in the same patient regardless of tumor type. Diagn Mol Pathol
14. Kazakov DV, Zelger B, Rutten A, et al. Morphologic diversity of malignant neoplasms arising in preexisting spiradenoma, cylindroma, and spiradenocylindroma based on the study of 24 cases, sporadic or occurring in the setting of Brooke-Spiegler syndrome. Am J Surg Pathol
15. Rallan D, Harland CC. Brooke-Spiegler syndrome: treatment with laser ablation. Clin Exp Dermatol
16. Fisher GH, Geronemus RG. Treatment of multiple familial trichoepitheliomas with a combination of aspirin and a neutralizing antibody to tumor necrosis factor alpha: a case report and hypothesis of mechanism. Arch Dermatol
17. Oosterkamp HM, Neering H, Nijman SM, et al. An evaluation of the efficacy of topical application of salicylic acid for the treatment of familial cylindromatosis. Br J Dermatol
18. Urquhart JL, Weston WL. Treatment of multiple trichoepitheliomas with topical imiquimod and tretinoin. Pediatr Dermatol