The authors retrospectively studied 12 patients diagnosed with bullous pilomatricoma at the JBNUH between January 2009 and May 2018. This study was approved by the institutional review board. Informed consent was obtained from all patients.

The following demographic and clinical data were collected: age, sex, site, duration, clinical impression, history of trauma, size, and change in size.

The biopsy samples were fixed in 10% formalin and embedded in paraffin wax. Then, 4 µm-thick sections were stained with Hematoxylin and Eosin (H&E), elastic (Roche, Basel, Switzerland), and Elastica-van Gieson (EVG, Muto, Tokyo, Japan). Two dermatopathologists evaluated the following histopathological data from Hematoxylin and Eosin (H&E) staining of all biopsied specimens: tumor (chronological stage, squamous epithelium, basaloid epithelium, transitional cell, shadow cell, inflammatory infiltrates, giant cell, calcification, and mitosis) and dermal changes (dilatation of lymphatics and blood vessels, increase in the number of lymphatics and blood vessels, edema, increase in the interstitial space, sparsity and disruption of collagen fibers, and sparsity of elastic fibers). The chronological stages were histopathologically categorized into early, fully developed, early regressive, and late regressive.20 The early lesion (Stage 1) was a small cyst lined by squamous and basaloid epithelium containing keratin filaments and shadow cells. The fully developed lesion (Stage 2) was a large cornified eosinophilic mass containing shadow cells with a peripheral lining of the basaloid epithelium. Early regressive lesions (Stage 3) showed inflammatory infiltrates with multinucleated histiocytic giant cells. At the periphery, small foci of basaloid cells were found. The late regressive lesion (Stage 4) was an irregularly shaped, confluent mass of eosinophilic hair matrix material and shadow cells, with varying degrees of calcification and ossification. It was surrounded by hyalinized and sclerotic stroma. This stage had very few, if any, inflammatory infiltrates. When two or more stages were observed simultaneously, the stage with the most dominant features was selected. The determination of the dermal histopathological findings of collagen and elastic fibers was re-examined by Elastic and EVG staining of all the specimens.

Immunohistochemical staining was performed on paraffin-embedded sections, according to the manufacturer’s instructions and established laboratory protocols. The following primary antibodies were used: CD31 (Dako, Glostrup, Denmark; dilution 1:1000), CD34 (Roche, Basel, Switzerland; RTU), D2-40 (Roche, Basel, Switzerland; RTU), MMP-2 (Abcam, Cambridge, UK; ab97779; dilution 1:1000), MMP-9 (Abcam, Cambridge, UK; ab137867; dilution 1:2000), VEGFR-3 (Abcam, Cambridge, UK; ab27278; dilution 1:100), and VEGF-C (Abcam, Cambridge, UK; ab9546; dilution 1:100). The dermal histopathological findings of lymphatic and blood vasculature were re-examined for all specimens using CD31, CD34, and D2-40 staining by two dermatopathologists. Two dermatopathologists also evaluated the immunoreactivity of MMP-2, MMP-9, VEGFR-3, and VEGF-C using the H-score, which assesses both the staining intensity and fraction of stained cells at each intensity.21,22 The staining intensity was scored as “0” for no evidence of staining, “1+” for weak staining, visible with high magnification, “2+” for moderate staining, “3+” for strong staining, and visible at low magnification. The percentage of stained cells (0%–100%) at each staining intensity was determined visually. The H-score was calculated by the formula: 0×(percentages of “0” cells) + 1×(percentages of “1+” cells) + 2×(percentages of “2+” cells) + 3×(percentages of “3+” cells), with a range of 0 to 300.

To comprehensively investigate the differences in clinical features between bullous and ordinary pilomatricoma, the authors analyzed 159 cases from 150 patients with biopsy-proven ordinary pilomatricoma during the same period at JBNUH. Five patients had two lesions; one patient had three pilomatricoma-related lesions, and two patients experienced recurrence. To analyze the histopathological and immunohistopathological features of the dermis and tumor, 143 cases of ordinary pilomatricoma specimens were excluded: 1) Enucleation biopsied tumor specimen without dermal tissue or 2) Torn, disoriented specimen because it was not sufficient to evaluate the features of bullous transformation in the dermis. The histopathological and immunohistopathological features of 16 completely excised ordinary pilomatricoma specimens, including the epidermis, dermis, and tumor, were evaluated.

Statistical analyses were performed using SPSS version 25 (SPSS Inc., Chicago, IL, USA) and R version 4.2.1. All clinical and histopathological data were analyzed using the chi-square test or Fisher’s exact test, except for the chronological stage data. The authors used the chi-square test if the expected cell frequencies < 5 were less than 20%, and Fisher’s exact test if the expected cell frequencies < 5 were more than 20%. The chi-square test was used to analyze the data of the squamoid epithelium, increased lymphatics, dilated blood vessels, edema, increased interstitial space, sparse elastic fibers, and sparse collagen fibers, whereas the other data were analyzed using Fisher’s exact test. Meanwhile, the chronological stage (Stages: 1–4) and immunohistochemical reactivity (H-score: 0–300) were evaluated using R, the Wilcoxon rank sum test because their ordered categories exhibited a significant linear trend. Results with a p-value < 0.05 were considered statistically significant.

Bullous pilomatricoma showed no gender preference, and ordinary pilomatricoma had a female predilection (M:F ratio = 1:1.79). Bullous pilomatricoma occurred predominantly in patients aged 10–19 years (58.3%), followed by those aged 50–59 years (25.0%). Conversely, ordinary pilomatricoma occurred predominantly in patients aged 0–9 years (44.7%), followed by those aged 10–19 years (31.4%). The mean ages of onset were 31.2 and 15.5 years for bullous pilomatricoma and ordinary pilomatricoma, respectively (Table 1).

Only 5 (41.7%) bullous pilomatricoma cases were clinically diagnosed as pilomatricoma, while 66% of ordinary pilomatricoma cases were clinically diagnosed as pilomatricoma. Bullous pilomatricoma was found most frequently in the upper extremities (41.7%), followed by the trunk (25.0%), while ordinary pilomatricoma was found most frequently in the face (54.1%), followed by the upper extremities (22.6%).

Bullous pilomatricoma was detected in 10/11 (90.9%) patients within three months, and ordinary pilomatricoma was detected in only 43/134 (32.1%) patients within the same period. The size of bullous pilomatricoma was greater than 1.0 cm in all 12 patients (100%), whereas that of ordinary pilomatricoma was less than 1.0 cm in 73/159 (52.2%) patients. The mean tumor size was 1.68 cm for bullous pilomatricoma and 0.99 cm for ordinary pilomatricoma. All recorded cases of bullous pilomatricoma (8/8, 100%) complained of an increase in tumor size. Four patients with bullous pilomatricoma had a history of trauma to the lesion, and 18 patients with ordinary pilomatricoma had a history of injury (Table 2, Fig. 1).

Figure 1.

Bullous pilomatricoma. Clinical appearance (A) and histopathological features (B) Chronological stage 2, Hematoxylin & Eosin, scanning view, inset: ×100.

(0.63MB).

In histopathologic data of tumors, the difference in chronological stage, calcification, and mitotic figures between patients with both types of pilomatricoma was statistically significant. Of the 12 patients with bullous pilomatricoma, 7 (58.3%) and 5 (41.7%) were diagnosed with chronological stages 3 and 2, respectively. Of the 16 patients with ordinary pilomatricoma, 12 (75.0%), 3 (18.8%), and 1 (6.3%) were diagnosed with stages 3, 4 and 2, respectively. Calcification and mitotic figures were observed in 5 (41.7%) and 8 (66.7%) patients with bullous pilomatricoma, respectively. Meanwhile, 14 (87.5%) and 3 (18.8%) patients with ordinary pilomatricoma showed calcification and mitotic figures, respectively.

Unlike ordinary pilomatricoma, nearly all patients with bullous pilomatricoma showed dermal changes, including dilated (66.7%) or increased (100%) lymphatics, dilated (83.3%) or increased (100%) blood vessels, dermal edema (100%), increased interstitial space (100%), sparse elastic fibers (100%), sparsity (100%), and disruption (66.7%) of collagen fibers (Table 3, Fig. 1).

MMP-2, MMP-9, VEGFR-3, and VEGF-C were expressed in dermal inflammatory cells and endothelial cells (Fig. 2). In patients with bullous pilomatricoma, MMP-2- and MMP-9-positive cells had higher H-scores than those in patients with ordinary pilomatricoma. Moreover, VEGFR-3 and VEGF-C expression levels in patients with bullous pilomatricoma were higher than those in patients with ordinary pilomatricoma. The differences in immunoreactivity of MMP-2, MMP-9, VEGFR-3, and VEGF-C in the bullous and ordinary groups were statistically significant (Fig. 3).

Figure 2.

Immunohistochemical staining for MMP-2, MMP-9, VEGFR-3, and VEGF-C in bullous pilomatricoma (A–D) and ordinary pilomatricoma (E–H) with magnification (×400).

(0.89MB).

Figure 3.

Box plot showing the result of immunohistochemical analysis using H-score in bullous and ordinary pilomatricoma for MMP-2, MMP-9, VEGFR-3, VEGF-C by Wilcoxon rank sum test.

(0.16MB).