Clinicopathological Features of 19 Eyelid Pilomatrixomas

Pilomatrixoma, also known as pilomatricoma and “benign calcifying epithelioma of Malherbe,” is an uncommon neoplasm representing approximately 1% of all pathologically diagnosed benign skin tumors [1]. They show microscopic evidence of differentiation similar to the hair follicle matrix. Most sporadic cases possess a mutation in CTNBB1 (β-catenin), resulting in the activation of Wnt signaling; however, this mutation is identified in a range of other neoplasms as well [2, 3]. The tumors can occur at any age, but about 60% of the cases are diagnosed in the first 2 decades of life with a female preponderance [4]. Pilomatrixoma can arise in any hair-bearing skin; however, they are most common in the head and neck region where they have a propensity for the eyebrow, and involve eyelid much less commonly [5, 6].

Pilomatrixoma is often thought clinically to represent other nodular or cystic lesions, and diagnosis is generally first made after excision and histopathological examination [7-9]. Histologically, pilomatrixoma is usually characterized by well-demarcated lesions in the lower dermis composed of peripherally located basophilic cells transitioning into keratinized eosinophilic “shadow cells” in the center of the lesion [4, 10]. The recurrence rate is low (0–3%), but aggressive cases must be differentiated from pilomatrix carcinoma cases [4, 11]. We report a large institutional case series of eyelid pilomatrixoma in order to better define its clinical and pathological features.

Demographic and clinical data for eyelid pilomatrixoma cases diagnosed at the Wilmer Eye Institute, Johns Hopkins Hospital, Baltimore, since 1981 were obtained from electronic medical records after Institutional Review Board approval. These data included race, age, gender, tumor location, symptoms, clinical diagnosis, gross appearance, and dimensions obtained in a HIPAA-compliant fashion with protection of individually identifiable health information. The research adhered to the tenets of the Declaration of Helsinki. Location in the eyelid was predominantly based on chart descriptions and specimen site designations on submission to pathology; however, when possible, clinical photos were also reviewed. All slides were reviewed by 2 pathologists (S.S. and C.G.E.), while clinical notes and images were reviewed by pathologists as well as surgeons (A.A.C. and T.M.).

A total of 19 eyelid pilomatrixomas were diagnosed in the Ophthalmic Pathology Laboratory of Wilmer Eye Institute from 1981 to the end of 2020. The demographic characteristics of the patients are presented in Table 1. The mean age was 24 years (range 2–63 years) with 8 (42%) and 4 (21%) of patients in the first and second decades of life, respectively. There were 12 (63%) females and 7 (37%) males. Most eyelid pilomatrixomas presented as solitary lesions, except for 3 young female patients who had a second tumor. One 9-year-old had a tumor on the right lower lid and a history of pilomatrixoma on the left temple. A 4-year-old had pilomatrixoma on both the left upper lid and the right arm for a year. Finally, a 7-year-old had a pilomatrixoma on the left lower lid (Fig. 1) and a second one on the face at the same time.

Tumors were most common on the upper (n = 14) as compared to the lower (n = 5) eyelid, with 12 tumors on the left and 7 on the right side. One upper eyelid tumor involved the lash line at the lid margin. In 7 cases, skin overlying the lesion showed hypervascularity and dilated vessels, and therefore had red to blueish discoloration. Initial clinical diagnoses included pilomatrixoma (4 cases), nodule/lesion/mass (9 cases), cystic lesion (1 case), sebaceous cyst (3 cases), chalazion suspicious for sebaceous carcinoma (1 case), and dermoid cyst (1 case).

After surgical removal, gross examination revealed that most of the lesions were well circumscribed, although 4 cases were more poorly demarcated. All of the lesions were firm. The lesions varied from 2 mm to 12 mm in maximum dimension. All 19 cases were confirmed as pilomatrixoma on histopathological examination. Microscopically, most of the tumors were centered in the lower dermis with some extension to subcutaneous tissues, except for 1 case centered in the upper dermis. The tumors were relatively typical, with well-defined lobules of basaloid and shadow cells accompanied by calcification and foreign body giant cells (Fig. 2). However, no cases showed ossification, which can sometimes be seen in pilomatrixoma. Mixed acute and chronic inflammation was noted in 1 case, acute inflammation in 3, and hemosiderin laden macrophages in 1. Basaloid cells did not predominate in any of our cases, but in 4 (21%), basaloid cells were absent. All of the pilomatrixomas were completely excised, and none showed signs of malignant transformation. No tumor recurrence was mentioned in the clinical records.

Eyelid pilomatrixoma, a rare benign adnexal tumor, is commonly misdiagnosed clinically. Only a few case series of pilomatrixoma arising at this site have been reported, limiting our understanding of how the tumor manifests in the eyelid [2, 4, 12]. Most of our eyelid cases presented in the first 2 decades of life. This is similar to prior reports across multiple body sites, in which approximately 40% and 20% of cases developed in the first and second decades, respectively [4, 13]. In our study limited to the eyelid, 42% of patients presented in their first decade and 21% in their second. One other study focused on eyelid pilomatrixoma reported that more than 75% of eyelid tumors occurred before 13 years of age, while in our series, this number was somewhat lower (47%) [3]. We found a wide overall age range for eyelid pilomatrixoma (2–63 years old). The youngest patient ever reported was a 10-month-old female baby, in which the lesion was first noticed at 3 months of age, while prior case series reported 74, 59, and 84 years as ages for the oldest patients [4, 12-15]. We had a total of 7 older adult patients in our series, consistent with an earlier study that mentioned a second peak in the fifth to seventh decades of life [14].

Most studies have noted a slight female predominance, and 63% of patients in our study were female [16]. However, in 1 large series including eyelid cases, 8 out of the total 13 (61%) were male [12].

Pilomatrixomas are mostly reported as solitary lesions, although they can be multifocal in rare cases [4]. In our series, there were 3 cases (all female) with 2 pilomatrixomas. The occurrence of multiple pilomatrixomas should prompt a search for possible genetic syndromes. In prior reports, they have most commonly been associated with myotonic dystrophy [17]. Pilomatrixomas have also been associated with Gardner, Turner, and Rubinstein-Taybi syndromes, and approximately half of all syndromic patients have 6 or more of these tumors [3, 17, 18]. In our 3 cases with 2 pilomatrixomas, no associated syndromic disorder was identified. Interestingly, of the genetic syndromes listed above, only Gardner syndrome is associated with mutations activating the Wnt pathway, as does the CTNNB1 change found in sporadic pilomatrixoma [18].

The upper eyelid is a more common site for pilomatrixoma than the lower one, and in our study, 74% of cases developed in upper eyelid [5]. The exact reason for this predilection is not clear, but greater proximity to the eyebrow, a common site for pilomatrixoma, may play a role. Eyelid location in our series was based primarily on the site specified upon specimen submission, as well as chart review, and challenges in precisely defining the anatomical site represent a limitation of a retrospective series such as this. Some cases for which no clinical images were available may have come from the junction between the eyelid and thicker brow skin, although none were from the eyebrow itself.

The diagnosis of pilomatrixoma is often not considered clinically. In our study, pilomatrixoma was included in the recorded clinical differential diagnosis in only 21% of cases, which is similar to other studies [4]. Clinically, the nodular lesions are slow growing and well circumscribed, firm or hard on palpation, and nontender; sizes generally range from 1 to 3 cm [7]. Overlying skin may appear normal but often shows reddish to blue discoloration. Tumors are generally mobile over the underlying tissue, and when excised, they are often larger than they appear clinically [8]. The clinical differential diagnosis in children can include epidermal and dermoid cysts, pyogenic granuloma, chalazion, and capillary hemangiomas [15]. In adults, while benign cysts and vascular lesions remain in the differential, other benign hair-follicle tumors such as tricholemmoma and trichoblastoma, as well as basal cell carcinoma (BCC), and other malignancies can also be considered.

Microscopically, pilomatrixomas are well-demarcated lesions typically composed of 2 cell types: peripherally located basophilic cells with indistinct borders, little cytoplasm, hyperchromatic nuclei, and mitoses, as well as more central regions which keratinize into the eosinophilic “shadow” or ghost cells showing distinct borders but no nucleus. Calcification in these central regions is frequent [4, 10]. As the tumor ages, the number of the basophilic cells can decrease and the shadow cells become more prominent. While the presence of the 2 cell types in a typical spatial orientation is quite characteristic, in some cases, particularly those with limited sampling, basaloid or ghost cells may predominate, making diagnosis difficult. While reports on the use of fine needle aspiration to diagnose these lesions are limited, these sampling issues can make this approach problematic [15].

The microscopic differential diagnosis for biopsies in which basaloid cells predominate includes BCC and trichoblastoma, although even focal regions with abrupt transition to ghost cells would argue against these diagnoses, and they often have other distinctive features such as the peripheral palisading and clefting characteristic of BCC. Trichilemmal and other cysts can sometimes contain cells similar to those seen centrally in pilomatrixoma, but both these and the peripheral basaloid component are less prominent [15, 19]. Finally, as discussed below, pilomatrixoma must be distinguished from pilomatrix carcinoma.

With respect to patient race, all the patients in our study were white except for a 10-year-old Asian boy. We found 1 additional case report of an eyelid pilomatrixoma occurring in a 10-month-old Asian boy [13]. Most case reports and case series have described white patients; however, the correlation between race and pilomatrixoma is still unclear [15, 16, 19].

Pilomatrixoma rarely recur, most commonly when incompletely excised, and malignant transformation is very rare [4, 7]. Diagnosis of malignancy is based on histologic features like poorly circumscription, pleomorphic basaloid cells with prominent nuclei, atypical mitoses, and invasion of blood vessels [2, 15]. None of our cases showed malignancy or recurrence. Nevertheless, attention to surgical margins and follow-up at regular intervals has been recommended [16].

In summary, our findings of pilomatrixoma in the eyelid support the notion that these tumors are most common in the first 2 decades of life with a slight female preponderance, but can appear at any age including older adults. The tumor is commonly misdiagnosed clinically, and the possibility of pilomatrixoma should be considered when firm upper eyelid tumors identified in children or young adults, particularly when blue or red-blue discoloration, are present.

The study was approved by the Johns Hopkins University Institutional Review Board. The study was deemed exempt from written informed consent as cases were retrospective; identifiable patient information was not included, and the outcomes were not directly impacting the care of the patients involved (IRB-X, Protocol NA_00011133). The research described in this manuscript is compliant with the World Medical Association Declaration of Helsinki and HIPAA regulations.

The authors have no conflicts of interests to declare.

This study was supported in part by an unrestricted department grant to the Wilmer Eye Institute from Research to Prevent Blindness.

Drs. Siadati and Eberhart planned the study, reviewed clinical and pathological material, and wrote and edited the manuscript. Drs. Campbell and McCulley reviewed clinical material and edited the manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.