A Trichoadenoma/Trichoepithelioma/Trichoblastoma-like Lesion in the Uterine Cervix Focally Mimics an Adenoid Basal Tumor

Abstract

The presence of ectodermal adnexal structures in the uterine cervix, including sebaceous glands, hair follicles, and sweat glands, has been well documented in the literature. In theory, there exists the possibility of developing cutaneous-type lesions from the ectopic ectodermal structures in this location. Here we report the first case of cervical hair follicle-derived proliferations reminiscent of trichoadenoma, trichoepithelioma, and trichoblastoma (TA/TE/TB) in a 52-year-old woman who underwent a prophylactic hysterectomy due to a germline microphthalmia-associated transcription factor (MITF) gene mutation. The lesion was an incidental finding in the cervix, exhibiting a spectrum of morphologic features ranging from germinative TB with basaloid cells, to TE with some degree of infundibulocystic differentiation, to well-differentiated TA. In some areas, hair follicle-like structures were associated with sebaceous glands, forming pilosebaceous units. The proliferations in the TB-like area resembled adenoid basal epithelioma/carcinoma; however, ancillary studies, particularly patchy p16 expression and non-detection of HPV, argued against this diagnosis. Similar to adenoid basal tumors, the TB-like lesion focally expressed NKX3.1, suggesting that it might be related to ectopic prostatic tissue or exhibit prostatic-lineage differentiation. While the theory of misplaced embryonal tissue, or an acquired metaplastic process, has been discussed, the histopathologic origin of these lesions remains largely unknown.

First described by Nikolowsky in 1958 as “organoid follicular hamartoma”, trichoadenoma (TA) is a rare, benign, slow-growing tumor of hair follicle, originating from the infundibular portion of the pilosebaceous unit with differentiation towards this lineage.¹ The tumor is characterized by multiple round and oval infundibulocysts and epithelial cords or islands throughout the dermis, with no evidence of continuity with the overlying epidermis.^2,3 Trichoepithelioma (TE) is a benign skin adnexal tumor originating from the outer root sheath of a hair follicle. TE is thought to be a superficial variant of trichoblastoma (TB) with prominent infundibular differentiation.³ However, this tumor is less mature than TA and is characterized by nests composed of germinative cells arranged in nodular, trabecular, and cribriform patterns in varying proportions.³ Both TA and TE usually occur on the skin of head and neck but can be found in other locations, including the trunk, buttocks, and vulva.

Although rare, ectodermal skin-type adnexal structures, including sebaceous glands, hair follicles, and sweat glands, have been reported in the vagina and uterine cervix.^4–9 These structures are thought to represent either misplaced embryonal/congenital tissue or an acquired metaplastic change secondary to chronic irritation. Notably, sebaceous hyperplasia, sebaceous carcinoma, and squamous cell carcinoma with sebaceous differentiation have been described in the cervix.^10–12 In theory, there exists the possibility of developing cutaneous-type lesions from these ectopic ectodermal structures in these locations. Here we present an exceptionally rare case of a TA/TE/TB-like lesion in the uterine cervix that focally mimics an adenoid basal tumor.

CASE STUDY

A 52-year-old woman had a remote history of estrogen receptor-positive breast carcinoma and a germline microphthalmia-associated transcription factor (MITF) gene mutation. Due to the MITF mutation, the patient underwent a prophylactic hysterectomy with bilateral salpingo-oophorectomy. Grossly, the ectocervical mucosa appeared gray-tan with a focal disruption along the anterior lip. Within the cervical stroma, approximately between the 3:00 and 10:00 positions, there was a cystic structure measuring 3.0×1.1 cm, partially filled with soft yellow-white material. No other discrete cervical lesions were noted. The endometrium appeared thin purple-gray, with a possible partial fibrous obliteration in the cavity. There was marked thickening of the uterine wall, consisting of a few whorled white-tan intramural nodules. No lesions were identified in the ovaries and fallopian tubes.

Microscopically, the cervical lesion exhibited various morphologic features. Close to, but without continuity with, the overlying squamous epithelium, there were multiple round to oval cysts of variable size, lined by stratified squamous epithelium and containing laminated keratin, indicating features of cutaneous TA (Fig. 1A–C). Some areas also exhibited morphology consistent with TE characterized by well-defined infundibular cysts intimately associated with nests of basaloid germinative cells arranged in nodular or trabecular pattern, along with keratin horn cysts (Fig. 1D, E). Leaf-like or frond-like architectural patterns were not evident. Some nests contained central mature sebocytes surrounded by a peripheral layer of flattened basophilic cells, consistent with a basal layer. Rare hair follicle-like structures were associated with sebaceous glands, forming pilosebaceous units (Fig. 1F).

FIGURE 1.

Multiple round to oval cysts in the uterine cervix were lined by stratified squamous epithelium and containing laminated keratin, indicating features of cutaneous trichoadenoma (A–C). Some areas exhibited morphology of trichoepithelioma characterized by well-defined infundibular cysts intimately associated with nests of basaloid germinative cells arranged in a nodular or trabecular pattern, along with keratin horn cysts (D, E). Rare hair follicle-like structures were associated with sebaceous glands, forming pilosebaceous units (F). The lesions were diffusely positive for p63 (G) and focally positive for CK7 (H). p16 displayed a patchy staining pattern, with some areas showing extensive-expression but lacking a “block” pattern of staining (I).

In other areas, the proliferation comprised small rounded nests of basaloid cells with scant cytoplasm, suggesting TB (Fig. 2A). The lesional cells displayed mild cytologic atypia, lacked evident mitosis, and formed rounded nests or short cords with peripheral palisading, which were embedded in the cervical stroma (Fig. 2B, C). These morphologic features mimic an adenoid basal tumor. A few nests exhibited pseudoglandular, cribriformlike spaces containing basement membrane-like material, resembling an adenoid cystic-like lesion (Fig. 2D, E). The surface squamous epithelium of the cervix showed some degree of reactive changes but lacked evidence of a squamous intraepithelial lesion. All pericervical soft tissue margins were negative for the lesions.

FIGURE 2.

In some areas, the proliferation comprised small rounded nests of basaloid cells with features suggestive of trichoblastoma (A). The lesional cells displayed mild cytologic atypia, lacked evident mitosis, and formed rounded nests with peripheral palisading which were embedded in the cervical stroma (B, C), mimicking an adenoid basal tumor. A few nests exhibited pseudoglandular, cribriform-like spaces containing basement membrane-like material, resembling an adenoid cystic-like lesion (D, E). p16 displayed a patchy staining pattern (F).

Immunohistochemically, all components were diffusely positive for p63 (Fig. 1G). CK7 immunostain showed focal positivity (Fig. 1H). p16 displayed a patchy staining pattern, with some areas showing extensive-expression but lacking a “block” pattern of staining (Figs. 1I and 2F). Scattered NKX3.1-positive cells were present in some nests (Fig. 3A–C), while PSA and C-Kit were negative. In situ hybridizations (ISH) were performed on 3 blocks using a high-risk human papillomavirus (HPV) RNA cocktail probe solution (HPV types 16, 18, 26, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, 82) and a type-specific probe for low-risk HPV6/11; no HPV was detected. Nested polymerase chain reaction (PCR) using the MY09/MY11 and GP05/ GP06 primer pairs¹³, followed by Sanger sequencing, also failed to detect any HPV. After a hysterectomy, the patient had no evidence of disease during the 14-month follow-up.

FIGURE 3.

A few nests (A) contained scattered NKX3.1-positive cells (B, C).

DISCUSSION

While the presence of ectodermal adnexal structures in the uterine cervix, including sebaceous glands, hair follicles, and sweat glands, has been well documented in the literature^4–10, our case represents, to the best of our knowledge, the first case of a TA/TE/TB-like lesion in this location. The lesion we described, similar to most reported extracutaneous ectodermal structures, was an incidental finding in a patient with a germline MITF gene mutation who underwent a prophylactic hysterectomy. Unlike those cases without gross findings, our case revealed a cystic area in the cervical stroma measuring 3.0×1.1 cm, partially filled with soft yellow-white material. Microscopically, the proliferation exhibited a spectrum of morphologic features ranging from germinative TB with basaloid cells to TE with some degree of infundibulocystic differentiation to well-differentiated TA.

The most intriguing aspects of this case include the lesion’s origin and its focal resemblance to adenoid basal tumor (epithelioma/carcinoma). Typically discovered as an incidental finding in patients treated for a high-grade squamous intraepithelial lesion (HSIL), adenoid basal tumor is a rare cervical lesion that usually occurs in women older than 50 years.^14,15 These tumors are composed of nests of bland cells with basaloid morphology and commonly demonstrate central squamous or glandular differentiation, some resembling basal cell carcinoma of the skin. Adenoid basal tumors share many morphologic features with adenoid cystic carcinoma (ACC) of the uterine cervix.^16,17 In fact, these 2 types of tumors were once regarded as a single entity and often coexis.^16,18 For this reason, some pathologists believe that ACC is not a distinct entity in the cervix, which has led to its removal from the WHO 2020 female genital tumor classification system.¹⁹ In our case, some areas of the lesion displayed overlapping features with adenoid basal tumor, and interestingly, several nests exhibited ACC-like morphology. Adenoid basal tumors have been shown to be etiologically related to high-risk HPV, particularly HPV 16, and are commonly associated with HSIL.¹⁴ Our case demonstrated a patchy p16 staining pattern, which is different from a “block” staining seen in high-risk HPV-related tumors. The overlying squamous epithelium exhibited reactive changes without evidence of HSIL. Consistently, ISHs for high-risk HPV (18 types) and low-risk HPV 6/11 were performed in multiple blocks but failed to detect any HPV; nor did PCR-based analysis. The ancillary studies demonstrated that this cervical lesion is not HPV-driven, arguing against an adenoid basal epithelioma/carcinoma. Instead, the more immature areas were in keeping with a TB-like lesion given the context of coexistence with TA-like and TE-like morphology.

As benign hair follicle tumors, TA, TE, and TB share a common histogenesis and display morphologic features recapitulating different stages of normal hair follicle development. In the cervix, these lesions are thought to originate from ectopic hair follicles, as found in our case. The next question naturally arises as to where these ectopic hair follicles originate. Whether these extracutaneous adnexal structures within the cervix represent congenital misplacement of embryonal tissue (true heterotopia) or acquired metaplastic changes secondary to chronic irritation remains a topic of debate. While some authors speculate both theories might be correct⁸, others believe the metaplastic hypothesis is the most common explanation for the origin of these ectopic elements.^5–7

Mostly located in the transformation zone of the cervix, reserve cells are a type of pluripotent progenitor cells that have the capacity to differentiate into either columnar or metaplastic squamous epithelium. Upon high-risk HPV infection, these cells can theoretically give rise to adenoid basal epithelioma/carcinoma, ACC-like carcinoma, squamous cell carcinoma, and adenosquamous carcinoma. It has been postulated that, in addition to their capacity of squamous differentiation, these cells may also retain the potential to give rise to extracutaneous appendages, including sebaceous glands, hair follicles, and sweat glands, through a metaplastic process.¹² Although not specific, focal CK7 positivity in the lesional cells supports the reserve cell hypothesis. Not surprisingly, adenoid basal tumors and TB, as seen in our case, can display overlapping morphology since they may be derived from the same progenitor cells.

Another unexpected observation in our case is the presence of rare NKX3.1-positive cells in a few basaloid nests. NKX3.1 is a relatively sensitive and specific marker of normal and neoplastic prostatic tissue, which is also positive in misplaced Skene’s glands including cervical ectopic prostatic tissue and vaginal tubulosquamous polyps. The presence of NKX3.1-positive cells in the TB areas suggests a possible association between these benign hair follicle lesions and Skene’s glands. In fact, sebaceous glands and basaloid proliferations resembling hair follicle structures have been reported in association with misplaced Skene’s glands.²⁰ Interestingly, adenoid basal epithelioma/carcinomas have been found to be diffusely positive for NKX3.1^21,22, raising the possibility that these tumors might be associated with or derived from ectopic prostatic tissue. Therefore, these morphologic and immunophenotypic similarities may indicate some intrinsic link between TA/TE/TB, adenoid basal epithelioma/carcinoma, and ectopic prostatic tissue, although this remains largely speculative.

In our case, the TA/TE/TB-like lesion in the cervix was an incidental finding in a patient with a germline MITF gene mutation who underwent prophylactic hysterectomy. It has been demonstrated that a germline variant of the MITF gene is associated with susceptibility to cutaneous melanoma.^23,24 Beyond melanoma, the pathogenic variant may also have a putative risk with renal cell carcinoma and uterine carcinosarcoma.^24,25 However, it remains uncertain whether these ectopic benign hair follicle-derived lesions in the cervix are related to MITF mutation.

In summary, we report the first case of TA/TE/TB-like proliferation with associated pilosebaceous units in the cervix, expanding the observation of extracutaneous ectodermal lesions in this location. The proliferation focally resembled adenoid basal epithelioma/carcinoma; however, ancillary studies, particularly patchy p16 expression and non-detection of HPV, argued against this diagnosis. Similar to adenoid basal tumors, a TB-like lesion focally expressed NKX3.1, suggesting that it might be derived from ectopic prostatic tissue or exhibit prostaticlineage differentiation. While theories of misplaced embryonal/congenital tissue versus acquired metaplastic processes have been discussed, the histopathologic origin of these lesions remains largely unknown.