TC is a rare malignant adnexal neoplasm originating from the outer root sheath, wherein cells demonstrate continuity with epidermis or follicular epithelium.5 Although the pathogenesis of TC has not been clearly identified, it appears to be related to actinic damage, long-term low-dose irradiation, or malignant transformation of trichilemmoma. Histological assessment of TC reveals varying degrees of local actinic damage. Oyama et al.7 reported trichilemmal carcinoma from long-standing seborrheic keratosis. In addition, other reports have described TC cases in patients with xeroderma pigmentosum or Cowden syndrome6, as well as in solid-organ transplant recipients. Some of the genetic alterations identified from our samples are in line with previous reports, enabling us to suggest possible pathogenetic mechanisms of TC.
Takata et al.8 reported that the total loss of the tumor suppressor gene TP53 may result in malignant transformation in TC. The complete loss of wild-type p53 due to the allelic loss of the short arm of chromosome 17p was suggested as a critical event. TP53 mutations were observed in two samples from our study, supporting the role of TP53 mutations in TC occurrence. PTEN deletion was identified in Case 2, but the patient did not have Cowden syndrome. Cowden syndrome, also called PTEN hamartoma syndrome, is caused by germline PTEN deletion. It develops multiple trichilemmoma, which is known to be a premalignant lesion of TC. O’Hare et al.6 reported a patient with Cowden disease who developed TC. Acquired PTEN deletion in Case 2 may have played an important role in the development of TC, possibly from trichilemmoma. TOP1 amplification was identified in one patient with underlying EV. EV is characterized by persistent human papillomavirus (HPV) infections, especially the ß-HPV subtype including HPV-5 and HPV-8. In addition, EV is prone to developing skin cancers,12 most of which contain HPV.13,14 The E6 protein from HPV-5 and HPV-8 decreases the level of ATM serine/threonine protein kinase through p300/ATR signaling, increasing carcinogenic potential after UV exposure.15,16 However, TOP1 inhibitor is reported to be effective for ATM mutations. We hypothesize that in this patient, decreased ATM caused by HPV resulted in the carcinogenesis of TC. In such circumstances, TOP1 inhibitor may be used as therapeutic agents. In addition, the association with HPV has been reported in other genetic alterations. FGFR3-TACC3 fusions, found in various solid tumors, were identified in tissues of head and neck squamous cell carcinoma that were also positive for HPV17. Moreover, PTEN deletion, identified in Case 2, frequently accompanied HPV-induced squamous cell carcinoma.18 These findings imply that HPV may play a leading role in TC carcinogenesis.
Our TC cases showed common genetic variants with skin cancers. TP53, NF1, and NRAS are frequently mutated genes identified in malignant melanoma.19 TP53 mutations in melanoma are correlated with sun exposure.20,21 In addition, TOP1 amplification was reported to be associated with more advanced and poor prognostic tumors in melanoma.22 Furthermore, TP53 and NRAS mutations have been identified in squamous cell carcinoma. Various skin cancers are reported to develop from seborrheic keratosis; one report reviewed a patient who developed TC in long-standing seborrheic keratosis.7 Given the shared risk factor, chronic sun exposure, TC may have similar molecular pathogenesis with skin cancers. Although controversial, some reports support the origin of basal cell carcinoma as outer root sheath of hair follicle, similar to TC. 23,24 This also suggests that carcinogenesis can be induced in outer root sheath cells by UV damage. Despite different cells of origin resulting in different types of cancer, a common carcinogen may cause the same genetic mutation for carcinogenesis.
ROS1-GOPC fusions are well known to be found in glioblastoma, cholangiocarcinoma, ovarian cancer, and non–small cell lung cancer. FGFR3-TACC3 fusions have been reported in glioblastoma, bladder urothelial tumors, nasopharyngeal carcinoma, head and neck cancer, and cervical cancer. Moreover, FGFR3-TACC325 and ROS1-GOPC fusions26 were reported in melanoma. For both cases, targeted agents showing therapeutic responses to the respective fusions were introduced. A recent report described a patient with metastatic melanoma harboring GOPC-ROS1 gene fusion who showed clinical response to entrectinib. Therapeutic responses can also be expected in TC cases after the identification of such fusion genes.
Our study nevertheless has several limitations. First, due to its retrospective design, the study was susceptible to selection bias. Second, we extracted DNA from FFPE tissue blocks, and fixation and storage process may have caused artifacts and low quality. Third, because we used a gene panel, genes not included in our panel could not be tested, possibly leading to significant implications. Fourth, we did not assess the functional consequences of the identified genetic mutations. Fifth, due to the limited sample size, statistical analysis was not possible. However, due to the rareness of the disease, our findings bear significance. Further studies are warranted to investigate certain molecular pathogenesis of TC and related therapeutic and prognostic implications.