- Email: [email protected]
Pilomatricoma (calcifying epithelioma): MDCT and MR imaging findings in 31 patients with radiological-pathological correlation
European Journal of Radiology 106 (2018) 92–99
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Research article
Pilomatricoma (calcifying epithelioma): MDCT and MR imaging findings in 31 patients with radiological-pathological correlation Kun-ming Yia,1, Kang Chenb,1, Lu Wanga, Xiao-juan Denga, Ying Zengc, Yi Wanga,
T
⁎
a
Deptartment of Radiology, Institute of Surgery Research and Third Affilated Hospital, Army Medical University, Chongqing 400042, China Department of Radiology, First Affilated Hospital, Army Medical University, Chongqing 400038, China c Department of Pathology, Institute of Surgery Research and Third Affilated Hospital, Army Medical University, Chongqing 400042, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Pilomatricoma Calcifying epithelioma of Malherbe Multi-detector computed tomography (MDCT) Magnetic resonance imaging (MRI) Pathology
Purpose: To describe the radiological characteristics of pilomatricomas on multi-detector computed tomography (MDCT) and magnetic resonance imaging (MRI), and to correlate the radiological findings and pathological features. Materials and methods: The radiological findings of 41 pilomatricomas in 31 patients were retrospectively reviewed. The images were evaluated with emphasis on calcifications, reticular and ring-like appearances, enhancement patterns, circular target sign and peritumoral fat stranding, and correlating these with pathological features. Results: Of the 31 lesions evaluated by MDCT, 25(80.6%) showed different patterns of calcifications which included single in 12(38.7%) lesions and multiple in 13(41.9%) lesions, but peritumoral fat stranding was observed only in 2(6.5%) lesions. MRI scans were performed in 11 patients with 21 lesions, homogeneous and inhomogeneous hypointensities on T1-weighted (T1W) images were showed respectively in 14(66.7%) and 7(33.3%) lesions. On fat-suppressed (FS) T2-weighted (T2W) images, a ring-like hyperintensity was observed in all 21(100%) lesions, reticular hyperintensity, circular target sign, peritumoral fat stranding and secondary anetoderma were seen in 7(33.3%), 4(19%), 4(19%) and 1(4.8%) lesions, respectively; and a significant difference (P < 0.05) was found when comparing the maximum diameter of tumors with (2.3 ± 1.4 mm) and without (1.1 ± 0.3 mm) reticular hyperintensity. On contrast-enhanced T1W images, all 21(100%) lesions were found to have ring-like enhancement and 7(33.3%) of them showed reticular enhancement. The reticular and ring-like appearances on MR images respectively corresponded to the pathological edematous stroma and connective tissue capsule, and the four-layer structures of the circular target sign on FS T2W images also corresponded to pathological calcifications, shadow cells, epithelial cells and connective tissue capsule, respectively. Conclusions: The characteristic radiological findings associated with pilomatricomas include different patterns of calcifications on MDCT images and ring-like, reticular appearances and circular target sign on MR images. Radiological findings are well correlated with pathological nature.
1. Introduction Pilomatricoma, also known as calcifying epithelioma of Malherbe, is an uncommon benign skin tumor of hair follicle origin with a differentiation toward hair cortex cells that was first described by Melherbe and Chenantois in 1880 [1,2]. This tumor can present at any age, about 60% of pilomatricomas occur in patients under the age of 20 years; however, a second peak of occurrence is seen in older patients. There is
a slight female predominance, with a reported female/male ratio of 3:2 [3,4]. Clinically, most pilomatricomas typically present as slowly-growing, well-demarcated, asymptomatic, solitary, hard, mobile superfcial nodule over the underlying skin, with a size range of 0.5–3.2 cm in diameter [3,5,6], while the overlying skin may demonstrate a reddish-blue discoloration or ulceration. Although all hair follicle-bearing skin surfaces can be involved, pilomatricoma most frequently occurs in the
⁎
Corresponding author. E-mail addresses: [email protected] (K.-m. Yi), [email protected] (K. Chen), [email protected] (L. Wang), [email protected] (X.-j. Deng), fi[email protected] (Y. Zeng), [email protected] (Y. Wang). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.ejrad.2018.07.020 Received 13 April 2018; Received in revised form 8 June 2018; Accepted 19 July 2018 0720-048X/ © 2018 Elsevier B.V. All rights reserved.
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
agent and saline solution were injected at 2.5–4 mL/s through an 18gauge plastic intravenous catheter placed in an antecubital vein. Contrast agent volumes were delivered at 2 mL/kg body weight, and the upper limit of dose was set to 120 mL for every patient. Axial, coronal and sagittal reformations were generated for all MDCT studies. The following scanning parameters were applied to 64–row and 16–slice MDCT systems: detector configurations of 64 × 0.625 mm or 16 × 0.6 mm respectively; section thicknesses of 0.625 or 0.6 mm, respectively; reconstruction intervals of 0.625 or 0.6 mm, respectively; and table speeds of 64 or 16 mm per rotation, respectively. The following parameters were applied to all examinations: pitch 0.984, matrix 512 × 512, field of view 180–240 mm, tube voltage 100–120 kV and tube current 300–400 mA.
head and neck regions, and the remainder occur in the upper extremities, the trunk, and the lower extremities. Most pilomatricomas are solitary, but the pilomatricoma can also be familial and associated with autosomal dominant diseases or syndromes, such as Gardner syndrome, myotonic muscular dystrophy (Steinert disease), sarcoidosis, skull dysostosis, Rubinstein-Taybi syndrome and Turner syndrome [5,7]. The mutation in exon 3 of the β-catenin gene (CTNNB1) is associated with pilomatricoma [8]. The treatment of choice is surgical resection and the prognosis is typically good, after complete resection, recurrence is uncommon [9]. Malignant pilomatricoma is an extremely rare condition with few cases reported and overwhelmingly seen in adults [10]. The diagnosis of pilomatricoma is usually based on history, superficial palpation and confirmed by pathologic examination. However, various manifestations of the tumor have often made the diagnosis based on the clinical examination alone inaccurate [1,4,11–14], resulting in unnecessary extensive surgery for an essentially benign condition [1,4]. In such a complicated clinical setting, knowledge of the radiological findings of this uncommon tumor would provide clinicians with the appropriate information and thus could obviate the extensive surgery. Although the CT and magnetic resonance imaging (MRI) findings of pilomatricomas have been reported in a few previous studies [5,6,15–23], some radiological findings found in our clinical practice have seldomly or never been mentioned in the literature. Therefore, the diagnostic accuracy of preoperative radiological examinations and radiological-pathological correlation about this tumor also need to be further evaluated. The purpose of our study was to describe the multidetector computed tomography (MDCT) and MRI findings of the pilomatricomas with pathological correlation, in order to enhance the understanding of this uncommon benign skin tumor.
2.3. MRI examinations
2. Materials and methods
Eleven patients were examined using 1.5-T MRI scanners (Magnetom Aera, Siemens, Erlangen, Germany; Avanto, Siemens, Erlangen, Germany; Sonata, Siemens, Erlangen, Germany). All conventional MR images were obtained at a section thickness of 3–4 mm with an intersection gap of 1 mm, and a 16 × 16 cm – 30 × 30 cm field of view. T1-weighted (T1W) spin-echo (SE) (TR/TE, 620–778 ms/ 10–16 ms), T2-weighted (T2W) fast SE (TR/TE, 1680–4000 ms/ 40–88 ms), and fat-suppressed (FS) T2W fast SE (TR/TE, 2080–4290 ms/27–105 ms) images were obtained in all tumors. Axial and coronal images were obtained with the above sequences. However, in one patient the T2W fast SE images were not included for image analysis due to poor image quality. In all tumors, contrast-enhanced T1W SE imaging (TR/TE, 640–820 ms/8–16 ms) and FS contrast-enhanced T1W SE (TR/TE, 530–840 ms/9–16 ms) images were obtained after intravenous injection of 0.1 mmol/kg of gadopentetate dimeglumine (Magnevist; Bayer Healthcare, Berlin, Germany). Both plain radiography and MRI were obtained in one patient with 1 tumor.
2.1. Study subjects
2.4. Image analysis
This study was approved by the human research committee of the institutional review board. A waiver of informed consent was obtained for this study. We followed Health Insurance Portability and Accountability Act guidelines. All pilomatricomas were retrieved from multiple centers and were confirmed by pathology during the time period 2010–2017. Eventually, 31 patients with preoperative MDCT or/ and MRI examinations and one of them with plain radiography were collected. One of these patients had 11 lesions, a total of 41 pilomatricomas were included in this retrospective study. The following clinical and pathological data were collected: age, sexes, anatomic locations, size and types of the tumor, clinical symptoms and signs, surgical and pathological findings, treatment outcomes, and relevant adverse events and complications. In all patients, the first preoperative MDCT and/or MR images were reviewed. The MDCT and MRI techniques varied somewhat due to the different imaging equipments and the retrospective nature of the study.
All imaging data were directly interfaced with the picture archiving and communication system (PACS), images were reviewed and consensus was reached by two musculoskeletal radiologists (K.Y., Y.W., with 8 and 23 years of experience, respectively). The radiological findings were analyzed considering the following terms: (1) number of lesions; (2) locations of lesions (included head [periauricular region, tempora, eyelid, eyebrow or malar], neck, trunk, extremities [upper or lower extremities]); (3) margins (well- or ill-defined); (4) size (maximum length and width of the lesion were measured on the axial images and maximum craniocaudal depth was measured on the coronal or sagittal images; and volumes were calculated by using the following ellipsoid formula: length × width × depth × 0.5233); (5) calcifications (included number [single or multiple], shapes [micro, nodular or irregular], distribution [diffuse, local multiple or local single] and degrees [complete or incomplete]); (6) density, signal intensity and enhancement patterns on MDCT/MRI scans (compared with the adjacent skeletal muscle); (8) reticular and ring-like hyperintensities on T2W images or FS T2W images (reticular hyperintensity was defined as hyperintensity with fine, septal structures within the lesion, and ring-like hyperintensity was defined as a high signal rim on the periphery of the lesion); (9) circular target sign on FS T2W images (defined as hypointensity in the center of the lesion, hyperintensity in the second layer, hypointensity in the third layer, and hyperintensity in the outer layer); (10) ring-like and reticular enhancements (defined as a thin hyper-density/intensity ring that surrounded at least half of the tumor, and fine, septal enhanced structures within the lesion on contrast-enhanced MDCT/MR images, respectively); (11) secondary anetoderma (defined as peritumoral dermal edema in the skin overlying the lesion which displayed as patchy areas of high signal on FS T2W images); (12) capsules, fat, haemorrhage, necrosis and cysts, and peritumoral fat
2.2. MDCT examinations Twenty-one patients underwent unenhanced and/or dual-phase enhanced (arterial and venous phases) MDCT scanning by a 64–row MDCT (Lightspeed, GE Healthcare, Milwaukee, WI, USA) or a 16–slice MDCT (Emition, Siemens, Erlangen, Germany) system. Both unenhanced and enhanced MDCT scans were obtained in 20 patients (every patient had a pilomatricoma), only unenhanced MDCT scans were obtained in one patient with 11 pilomatricomas and this patient also underwent a MRI scanning. For enhanced MDCT scanning, a dual-head power injector was used to administer contrast agent (Ultravist; Bayer Schering Pharma, Berlin, Germany) at 370 mg iodine/mL followed by 30 mL saline. The contrast 93
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Table 1 Clinical characteristics of patients. Case
Age (y)
Sex
Location
Duration of symptoms before admission (M)
Symptom and sign
1 2 3 4 5 6 7 8 9 10 11 12
14 43 15 16 5 13 26 4 18 67 37 28
M M M F F F M F M M M F
Neck Temporal Neck Neck Periauricular Neck Forearm Temporal Periauricular Periauricular Forehead Eyelid
12 1 6 12 2 10 12 6 60 6 3 3
13 14 15
58 44 17
F M F
Parietal Neck Elbow
6 7 3
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
58 37 48 46 42 114 38 28 10 70 10 11 35 38 42 63
F M F M M F M F M F F M F F F F
Parietal Parietal Parietal Upper arm Parietal Eyebrow Eyebrow Parietal Eyelid Parietal Parietal Eyelid Neck Neck Malar Eyelid
84 192 10 72 24 24 12 6 12 240 3 6 240 240 2 2
Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue redness Soft-tissue Soft-tissue Soft-tissue bullous Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue
Treatment outcome
Adverse events and complications
Follow-up
mass, R-b mass mass mass mass mass, R-b mass mass mass, R-b mass mass mass,
Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable
excision excision excision excision excision excision excision excision excision excision excision excision
No No No No No No No No No No No No
No No No No No No No No No No No No
mass mass mass,
Curable excision Curable excision Curable excision
No No No
No recurrence No recurrence No recurrence
mass mass mass mass, mass, mass mass mass mass, mass mass, mass, mass mass, mass mass
Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable
No No No No No No No No No No No No No No No No
No No No No No No No No No No No No No No No No
ulcer ulcer
R-b ulcer
R-b
excision excision excision excision excision excision excision excision excision excision excision excision excision excision excision excision
recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence
recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence
3.2. MDCT findings
stranding were also evaluated on MDCT/MR images.
The MDCT findings of 31 pilomatricomas are given in Table 2. In the 31 tumors, 24(77.4%) lesions were hyperdensity, 6(19.4%) lesions were isodensity and 1(3.2%) lesion was hypodensity on the unenhanced MDCT images. Of the 20 lesions with contrast-enhanced images, 18(90%) showed a mild to moderate enhancement, 2(10%) failed to evaluate the enhancement patterns due to complete calcification within the tumor. All 31 tumors were well-defined margins, appeared as round (12, 38.7%) or round-like (19, 61.3%) shapes, and 11(35.5%) of them showed slight lobular and 3(9.7%) showed necrosis and cysts (Fig. 1). Their maximal diameter ranged from 0.8 to 3.6 cm (mean 1.4 ± 0.9 cm) and volumes from 0.5 to 32.4 cm3 (mean 2.4 ± 1.4 cm3). Of the 31 lesions, 25(80.6%) had different patterns of calcifications within the tumors and 2(6.5%) showed peritumoral fat stranding (Fig. 1). The number of calcifications was single in 12(38.7%) lesions and multiple in 13(41.9%) lesions. Their shapes included micro (12, 38.7%), nodular (13, 41.9%) and irregular (3, 9.7%), and the same lesion could have more than one shape of calcifications. The most common distribution of calcifications was local single (12, 38.7%), followed by local multiple (9, 29%) and diffuse (4, 12.9%). Complete calcification was found in 4(12.9%) lesions and the remaining 21(67.7%) lesions were observed with incomplete calcification. In one patient with 11 pilomatricomas, all of these lesions had nodular calcification, 2 of them were found with complete calcification and the remaining 9 lesions showed calcification surrounded by a layer of soft tissue (Fig. 2). No significant difference (P > 0.05) was found when comparing the maximum diameter of tumors with (1.8 ± 1.0 mm) and without (1.0 ± 0.4 mm) calcification, multiple (1.2 ± 0.4 mm) and single (2.1 ± 1.1 mm) calcification, and complete (0.6 ± 0.2 mm) and
2.5. Statistical analysis Statistical analyses were performed using the SPSS analysis program (version 18.0, IBM). Student’s t-test was used to compare the maximum diameter of tumors with and without calcification, multiple and single calcification, and complete and incomplete calcification on MDCT images, and with and without reticular hyperintensity on MR images. Prior to assessment, the data were tested for homoscedasticity by Levene’s test. A P value of less than 0.05 was considered significant.
3. Results 3.1. Clinical characteristics The clinical characteristics are given in Table 1. Our patient cohort included 14 men and 17 women with an average of 35.3 ± 23.8 years (range, 4–114 years). The duration of symptoms in all the patients admitted to the hospital ranged from 1 to 240 months, with an average of 42.5 ± 21.6 months. In total, 41 tumors were identified among these patients, and located in the subcutaneous fat layer of the head (31, 75.6%), neck (7, 17.1%) and extremities (3, 7.3%). Among the 31 patients, 11 of the 41 lesions were found in the same patient (case 17) and located in the same anatomic region. The main complaint was either a painless or tender soft-tissue mass for 38(92.7%) lesions. The other indications included foreign body, cyst, swelling, abscess, ulcer, reddish-blue discoloration (abbreviated as R-b), bullous appearance and local pain.
94
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
hyperintensity was observed in all 21(100%) lesions, and reticular hyperintensity, circular target sign, peritumoral fat stranding and secondary anetoderma were found in 7(33.3%), 4(19%), 4(19%) and 1(4.8%) lesions, respectively (Fig. 3). All of the 4 lesions with circular target sign occurred in same patient and both MRI and CT scans were performed on this patient (Fig. 2). Significant difference (P < 0.05) was found when comparing the maximum diameter of tumors with (2.3 ± 1.4 mm) and without (1.1 ± 0.3 mm) reticular hyperintensity (Table 2). Fast SE T2W images were only obtained in 10 lesions in our study, 4(40%) of them were hypointensity and 6(60%) were heterogeneous hyperintensity, and reticular, ring-like and cystic hyperintensities were observed in 6(60%), 2(20%) and 1(10%) lesions, respectively. On FS enhanced T1W images, all 21(100%) lesions were found to have ringlike enhancement and 7(33.3%) of them showed reticular enhancement (Fig. 3).
Table 2 Radiological findings in the 41 pilomatricomas. Radiological findings
Locations Head Neck Extremities Shapes Round-like Round Irregular Slight lobular Density Isodensity Hypodensity Hyperdensity Necrosis and cysts Peritumoral fat stranding Calcifications Number Single Multiple Shapes Micro Nodular Irregular Distribution Diffuse Local multiple Local single Degrees Complete Incomplete No calcification T1W images (n = 21) Homogeneous hypointensity Heterogeneous hypointensity T2W images (n = 10) Homogeneous hypointensity Reticular hyperintensity Ring-like hyperintensity Cystic hyperintensity FS T2W images (n = 21) Ring-like hyperintensity Reticular hyperintensity Circular target sign Peritumoral fat stranding Secondary anetoderma FS enhanced T1W images (n = 21) Ring-like enhancement Reticular enhancement No enhancement Size Maximal diameter (cm) Volume (cm3)
No. (%) of lesions/mean ± SD MDCT (n = 31)
MRI (n = 21)
25(80.6) 6(19.4)
17(81.0) 1(4.8) 3(14.3)
19(61.3) 12(38.7) 0 11(35.5)
7(33.3) 12(57.1) 2(9.5) 6(28.6)
6(19.4) 1(3.2) 24(77.4) 3(9.7) 2(6.5) 25(80.6)
3.4. Accuracy of radiological diagnosis and pathology of the tumor The accuracy rate of the preoperative radiological diagnosis of pilomatricoma by MDCT and MRI in our study was only 12.9%(4/31) and 14.3%(3/21), respectively. The radiological diagnoses included pilomatricoma (n = 7), dermoid cyst (n = 3), malignancy (n = 3), hemangioma (n = 6), calcification (n = 4), lymph node (n = 4) and benign lesion (n = 14). All tumors were subsequently surgically resected. Specimens were fixed in 4% buffered formalin and routinely processed and embedded in paraffin; 4-μm-thick sections were stained with hematoxylin and eosin (HE). Histological features of these tumors were as follows: (1) irregularly shaped islands of epithelial cells which composed of basaloid cells and shadow cells were observed in all tumors; (2) calcification, ossification, hemorrhage, and adipose cells were seen in 40(97.6%), 1(2.4%), 4(9.6%) and 1(2.4%) tumors, respectively; (3) edematous and fibrous stroma with various degrees of inflammatory cell infiltration and lymphatic vascular dilatation was observed in 36(87.8%) tumors; (4) connective tissue capsule was observed to be complete in 40(97.6%) tumors and incomplete in 1(2.4%) tumor; (5) 31(75.6%) of the 41 tumors were found with various degrees of collagen fibrils, inflammatory cell infiltration and vascular proliferation, and most vascular proliferation occurred in the stroma, but some occurred in the shadow cells; (6) various degrees of peritumoral inflammatory changes were seen in 8(19.5%) tumors; (7) intense inflammatory cell infiltration in peritumoral fatty tissues were observed in tumors accompanied by peritumoral fat stranding on MDCT or FS T2W images (Figs. 1 and 3).
12(38.7) 13(41.9) 12(38.7) 13(41.9) 3(9.7) 4(12.9) 9(29) 12(38.7) 4(12.9) 21(67.7) 6(19.4) 14(66.7) 7(33.3) 4(40.0) 6(60.0) 2(20.0) 1(10.0) 21(100) 7(33.3) 4(19.0) 4(19.0) 1(4.8) 21(100) 7(33.3) 0 1.4 ± 0.9 2.4 ± 1.4
3.5. Treatment outcomes and follow-up
1.4 ± 1.1 3.8 ± 2.3
All patients underwent surgical excision using a direct approach by making the incision over the tumor, and one of them with excision of the overlying skin. At surgery, all lesions did not involve the muscle and they were easily dissected free from the subcutaneous fat. All 31 patients were followed up post-operatively for periods ranging from 3 months to 2 years (mean, 9 months), and did not exhibit local recurrence and no new lesion developed.
incomplete (1.6 ± 0.9 mm) calcification.
3.3. MRI findings The MRI findings of 21 pilomatricomas are given in Table 2. The margins of all tumors were well-defined, their maximal diameter ranged from 0.8 to 4.5 cm (mean 1.4 ± 1.1 cm) and volumes from 0.5 to 40.5 cm3 (mean 3.8 ± 2.3 cm3). One of these tumors showed a high density mass with popcorn calcification on plain radiographs. On T1W images, homogeneous hypointensity was found in 14(66.7%) tumors, the remaining 7(33.3%) tumors showed inhomogeneous hypointensity, 3(14.3%) of them were found to have reticular and small nodular hyperintensities and 4(19%) showed ring-like hyperintensity. On FS T2W images, all 21(100%) lesions were heterogeneous signal intensity with combinations of high and low signals. A ring-like
4. Discussion In the current study, we present the largest series of radiological studies of pilomatricomas to date. Our study demonstrates that pilomatricomas typically present as a well-defined subcutaneous soft tissue mass with different patterns of calcifications and mild to moderate enhancements on MDCT images, and ring-like, reticular appearances and circular target sign on MR images. In previous studies, plain radiography demonstrated that pilomatricoma appeared as a solitary, sharply demarcated subcutaneous tumor with extensive sand-like or dense focal calcification [24]. In our study, 95
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Fig. 1. A 13-year-old girl with a pilomatricoma in the right neck. (a) Non-enhanced axial MDCT image shows a well-defined, subcutaneous and heterogeneous mass oppressing the adjacent sternocleidomastoid muscle (black arrows), with diffuse micro and small nodular calcifications within the lesion and peritumoral fat stranding (white arrows). (b and c) Arterial- and venous-phase contrast-enhanced axial MDCT images reveal a mild to moderate enhancement within the mass. (d) Contrast-enhanced coronal MDCT image shows the mass with a slight lobular appearance (arrows). (e) Microscopic view (HE, ×100) of a pathological specimen shows a connective tissue capsule (black arrows), some foreign body type giant cells (∗) and a large number of enucleated shadow cells (☆). Edema (white arrows) is located between the shadow cells and the connective tissue capsule. (f) Another microscopic view (HE, ×100) reveals calcifications (∗) in the center and shadow cells (☆) on the periphery of the mass. Abundant edematous and fibrous stroma with inflammatory cell infiltration (white arrows) and edema (black arrows) are observed among epithelial cell islands.
distribution and degree, to the best of our knowledge, detailed classifications have not been reported in previous literature. Thus, we found that irregular calcification and multiple shapes of calcification in the same lesion can occur in the pilomatricoma which were observed in 3 and 4 lesions, respectively, in our patients. In addition, we also found that 4 lesions were nearly completely calcified, therefore, it is very difficult to distinguish such pilomatricoma from pure calcification on CT images.
one of the lesions showed popcorn calcification in the tumor on plain radiographs. On CT, the pilomatricoma has been described as well-defined subcutaneous masses with varying amounts of calcification and variable enhancement [17–19,23] and calcifications have been reported in 63.6–81% of pilomatricomas [19,23]. In our study, 80.6% of lesions showed different patterns of calcification on MDCT which is consistent with the previous results [19,23]. Furthermore, our study categorized the calcification patterns in terms of number, shape, 96
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Fig. 2. A 37-year-old man with pilomatricoma in the head. (a) Axial FS T2W MR image shows two well-defined subcutaneous nodules with circular target sign (arrows) which appear as hypointensities in the center and third layer, and hyperintensities in the second and outer layers of the lesions. (b) Axial T1W MR image shows homogeneous hypointensity in the center and ring-like hyperintensity on the periphery of the lesions (arrows). (c and d) Contrast-enhanced axial and sagittal T1W MR images show ring-like enhancement (arrows). (e) Contrast-enhanced coronal FS T1W MR image also shows ring-like enhancement, but no reticular enhancement (arrow). (f–j) Non-enhanced axial MDCT images show a total of 11 well-defined subcutaneous nodules (arrows), two of them are nearly completely calcified (thick arrows), and the remaining 9 lesions appear as central calcifications surrounded by a layer of soft tissue (thin arrows). (k) A local magnified view of the figure 2a and (l) scanning view (HE, ×40) of a pathological specimen clearly reveal the one-to-one correspondence between radiological and pathological fourlayer concentric circular structures from the center to the periphery: the central hypointensity (C) vs. calcification (☆); hyperintensity in the second layer (S) vs. shadow cells (∗); hypointensity in the third layer (T) vs. epithelial cells (white arrows); and hyperintensity in the outer layer (O) vs. connective tissue capsule (black arrows). (m) Pathological microscopic view (HE, ×100) shows the connective tissue capsule with collagen fiber and inflammatory cell infiltration (arrows).
[19,23]. In our series, necrosis and cysts were showed in 9.7% of lesions on MDCT, however, pathological examinations revealed necrosis in 21 of the 41 lesions. Therefore, the micronecrosis within the pilomatricoma is also not easy to be visualized on both CT and MRI images. In our study, 90% of lesions showed as a mild to moderate enhancement, which enabled us to confirm the previous observations [3,17–19]. However, Ichikawa et al. described pilomatricoma as a hypervascular
Although CT has the ability to depict the great majority of calcifications, it can hardly visualize microscopic calcifications due to its resolution limit [18,19]. It is also easy to explain that calcifications were seen in 40(97.6%) lesions on pathology, but MDCT images did not show any calcification in 6 of them in our study. To our knowledge, the cystic degeneration is a rare radiological feature in pilomatricoma, there have only been three reports about pilomatricoma with cystic degeneration 97
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Fig. 3. A 17-year-old man with pilomatricoma in the elbow. (a) Coronal FS T2W MR image shows a well-defined subcutaneous mass with ring-like and reticular hyperintensities, and hypointense nodules are showed in its interior (thin arrow) and patchy peritumoral edema is seen in the skin overlying the mass (thick arrows). (b) Axial FS T2W MR image shows peritumoral fat stranding (arrows). (c) Axial T2W MR image shows reticular and ring-like hyperintensities with peritumoral dermal edema (arrows). (d) Coronal T1W MR image shows a nonhomogeneous hypointense mass with small pieces of high signals in its interior (arrows). (e) Contrast-enhanced coronal T1W MR image shows a ring-like enhancement on the periphery of the mass (arrows). (f) Contrast-enhanced axial FS T1W MR image shows reticular enhancement in the interior of the tumor and a ring-like enhancement on its periphery (arrows). (g) Scanning view (HE, ×40) of a pathological specimen reveals epithelial cell islands composed of basaloid cells (☆) and shadow cells (∗). Abundant edematous and fibrous stroma with inflammatory cell infiltration (white arrows) is observed among epithelial cell islands and vascular proliferation (black arrows) is seen in the distribution zone of the shadow cells. (h) Another scanning view (HE, ×40) reveals peritumoral dermal edema (∗) and connective tissue capsule with collagen fiber and inflammatory cell infiltration (black arrows). Abundant edematous and fibrous stroma with inflammatory cell infiltration (white arrows) is observed among epithelial cell islands (☆). (i) Pathological microscopic view (HE, ×100) reveals lymphatic dilation (☆) in the peritumoral dermis.
hyperintense rim showed enhancement on contrast-enhanced T1W images, but no enhancement in the center of the lesion. Since then, Lim et al. [19] described reticular hyperintensity and reticular enhancement on FS T2W and contrast-enhanced T1W images in all of the five patients they reported. Bulman et al. [23] and Kato et al. [6] reported 16 and 9 pilomatricomas, respectively, with reticular hyperintensity on FS T2W images. Similarly, in our study, the reticular hyperintensity and enhancement were also observed respectively in 7 of 21 lesions on FS T2W and contrast-enhanced T1W images. With regard to the pathological basis of the reticular hyperintensity, Hoffmann et al. [21] suggested that the internal reticulations and septations on MR images correspond to the basaloid cells. In other authors’ opinion, the internal reticulations and septations were reported to represent the intercellular edematous stroma, because the basaloid cells were considered as the sheets of vascular epithelial cells [6,19,22]. Our pathological examinations revealed edematous and fibrous stroma with various degrees of inflammatory cell infiltration and vascular proliferation in 7 tumors. Our results, therefore, suggest that the reticular appearances on MR images correspond to the pathological distribution of intratumoral stroma. A few studies [6,19,26] have reported that the ring-like appearance
mass on angiograms [16]. On MRI, most of the pilomatricomas have been described as showing a homogeneous intermediate signal intensity on T1W images and heterogeneous intermediate signal intensity on T2W and FS T2W images [6,16,19–21,23]. However, some pilomatricomas have been described as showing an inhomogeneous signal intensity on T1W images [15,22], the causes are not very clear at all. Masih et al. [22] described a non-calcified pilomatricoma with a lot of keratin accumulation on pathological examinations. Henkelman et al. [25] indicated that particulate calcium can reduce T1 relaxation times by a surface relaxation mechanism. In our study, 3 lesions were also found to have reticular and small nodular hyperintensities on T1W images, and hemorrhage was observed on pathological examinations. These results, therefore, suggest that the high signal on T1W images corresponded to the distribution of keratin accumulation, hemorrhage and calcifications on pathological specimens. The reticular hyperintensity have been known as a MRI characteristic of pilomatricomas on T2W, FS T2W and contrast-enhanced T1W images [6,19,21–23]. Hoffmann et al. [21] first described a pilomatricoma with high signal bands radiating from the lesion center to high signal periphery on T2W and FS inversion recovery images, and the 98
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Conflicts of interest
of pilomatricomas on ultrasound or MR images corresponded to the connective tissue capsule. Our pathological examinations also showed connective tissue capsules with various degrees of collagen fibrils, inflammatory cell infiltration, and vascular proliferation in 31 of 41 tumors. And similarly, in our 21 pilomatricomas evaluated by MRI, all these tumors showed a ring-like appearance. In our study, the ring-like hyperintensity on FS T2W images was showed to be more frequent (21/ 21) than those (2/10) on conventional T2W images. It may be due to the fact that the ring-like hyperintensity is confounded by the surrounding hyperintense fat tissues on conventional T2W images which can result in a difficulty in identification [6]. In our study, 4 of 21 pilomatricomas showed a circular target sign on FS T2W images. To the best of our knowledge, the circular target sign found in the pilomatricoma has never been reported in previous literature. Furthermore, by comparison of radiographic and pathological findings, we found that the central hypointensity on FS T2W images was confirmed to be calcification on MDCT images and pathological specimens. And then, the hyperintensity in the second layer, hypointensity in the third layer, and hyperintensity in the outer layer of the circular target sign were also corresponding to the pathological shadow cells, epithelial cells and connective tissue capsule, respectively. Pathological examinations showed no lymphatic vascular dilatation and increased numbers of blood vessels within these tumors which may explain all of the 4 pilomatricomas with a ring-like enhancement in the outer layer, but no enhancement in the center and other two layers of the tumors. In addition, the 4 pilomatricomas with circular target sign appeared as a central hypointensity surrounded by a layer of hyperintensity on T1W images and a nodular calcification surrounded by a layer of soft tissue on MDCT, this may be explained by our pathological results that the central hypointensity/nodular calcification on T1W/MDCT images corresponds to the pathological calcification, and the peripheral hyperintensity/soft tissue density on T1W/MDCT images is also consistent with the distribution of an absolute majority of shadow cells on pathological specimens. Therefore, our results may suggest that the shadow cells in the pilomatricomas show as hyperintensity not only on T2W and FS T2W images, but also on T1W images. There are several limitations in our study. The relative smaller sample size is a limitation. Pilomatricoma is an uncommon tumor and, moreover, the patients with pilomatricoma may never come for diagnostic imaging because of the lesion’s small size and superficial location. The retrospective nature of the study is another limitation. It means that there were no perfect imaging protocols, so we did not obtain both MDCT and MR images in the majority of patients. In addition, different MDCT and MRI equipments and techniques were used. However, these problems are simply unavoidable due to the rarity of this type of tumor, and this should not have signifcantly affected the radiological characteristics studied.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. There are no conflicts of interest to report. References [1] S. Kaddu, H.P. Soyer, L. Cerroni, W. Salmhofer, S. Hädl, Clinical and histopathologic spectrum of pilomatricomas in adults, Int. J. Dermatol. 33 (10) (1994) 705–708. [2] A. Malherbe, J. Chenantais, Note surl’epitheliome calcifié des glandes sebacées, Prog. Med. 8 (1880) 826–828. [3] N.J. Wells, G.K. Blair, J.F. Magee, D.M. Whiteman, Pilomatrixoma: a common, benign childhood skin tumour, Can. J. Surg. 37 (6) (1994) 483–486. [4] C.G. Julian, P.W. Bowers, A clinical review of 209 pilomatricomas, J. Am. Acad. Dermatol. 39 (2 Pt. 1) (1998) 191–195. [5] S. Duflo, R. Nicollas, S. Roman, G. Magalon, J.M. Triglia, Pilomatrixoma of the head and neck in children: a study of 38 cases and a review of the literature, Arch. Otolaryngol. Head Neck Surg. 124 (11) (1998) 1239–1242. [6] H. Kato, M. Kanematsu, H. Watanabe, A. Nagano, E. Shu, M. Seishima, T. Miyazaki, MR imaging findings of pilomatricomas: a radiological-pathological correlation, Acta radiol. 57 (6) (2016) 726–732. [7] D. Kwon, K. Grekov, M. Krishnan, R. Dyleski, Characteristics of pilomatrixoma in children: a review of 137 patients, Int. J. Pediatr. Otorhinolaryngol. 78 (8) (2014) 1337–1341. [8] U. Gat, R. DasGupta, L. Degenstein, E. Fuchs, De Novo hair follicle morphogenesis and hair tumors in mice expressing a truncated beta-catenin in skin, Cell 95 (5) (1998) 605–614. [9] S.F. Hassan, E. Stephens, S.C. Fallon, D. Schady, M.J. Hicks, M.E. Lopez, D.A. Lazar, M.A. Rodriguez, M.L. Brandt, Characterizing pilomatricomas in children: a single institution experience, J. Pediatr. Surg. 48 (7) (2013) 1551–1556. [10] J.L. Herrmann, A. Allan, K.M. Trapp, M.B. Morgan, Pilomatrix carcinoma: 13 new cases and review of the literature with emphasis on predictors of metastasis, J. Am. Acad. Dermatol. 71 (1) (2014) 38–43. [11] A. Danielson-Cohen, S.J. Lin, C.A. Hughes, Y.H. An, J. Maddalozzo, Head and neck pilomatrixoma in children, Arch. Otolaryngol. Head Neck Surg. 127 (12) (2001) 1481–1483. [12] N. Kumaran, A. Azmy, R. Carachi, P.A. Raine, J.H. Macfarlane, A.G. Howatson, Pilomatrixoma — accuracy of clinical diagnosis, J. Pediatr. Surg. 41 (10) (2006) 1755–1758. [13] M.Y. Lan, M.C. Lan, C.Y. Ho, W.Y. Li, C.Z. Lin, Pilomatricoma of the head and neck: a retrospective review of 179 cases, Arch. Otolaryngol. Head Neck Surg. 129 (12) (2003) 1327–1330. [14] N.A. Roche, S.J. Monstrey, G.E. Matton, Pilomatricoma in children: common but often misdiagnosed, Acta Chir. Belg. 110 (2) (2010) 250–254. [15] T.J. Hsieh, C.K. Wang, K.B. Tsai, Y.W. Chen, Pilomatricoma: magnetic resonance imaging and pathological evaluation, J. Comput. Assist. Tomogr. 32 (2) (2008) 320–323. [16] T. Ichikawa, Y. Nakajima, H. Fujimoto, A. Koyama, M. Honma, M. Yatsuzuka, K. Ohtomo, G. Uchiyama, S. Ushigome, S. Ohba, Giant calcifying epithelioma of malherbe (pilomatrixoma): imaging features, Skeletal Radiol. 26 (10) (1997) 602–605. [17] P.M. Som, J.M. Shugar, A.R. Silvers, CT of pilomatrixoma in the cheek, AJNR 19 (7) (1998) 1219–1220. [18] K.H. Lee, H.J. Kim, C.H. Suh, Pilomatricoma in the head and neck: CT findings in three patients, J. Comput. Assist. Tomogr. 24 (2) (2000) 332–335. [19] H.W. Lim, S.A. Im, G.Y. Lim, H.J. Park, H. Lee, M.S. Sung, B.J. Kang, J.Y. Kim, Pilomatricomas in children: imaging characteristics with pathologic correlation, Pediatr. Radiol. 37 (6) (2007) 549–555. [20] L.H.De. Beuckeleer, A.M.De. Schepper, I. Neetens, Magnetic resonance imaging of pilomatricoma, Eur. Radiol. 6 (1) (1996) 72–75. [21] V. Hoffmann, T. Roeren, P. Moller, G. Heuschen, MR imaging of a pilomatrixoma, Pediatr. Radiol. 28 (4) (1998) 272. [22] S. Masih, S.M. Sorenson, A. Gentili, L.L. Seeger, Atypical adult non-calcified pilomatricoma, Skeletal Radiol. 29 (1) (2000) 54–56. [23] J.C. Bulman, S.O. Ulualp, V. Rajaram, K. Koral, Pilomatricoma of childhood: a common pathologic diagnosis yet a rare radiologic one, AJR 206 (1) (2016) 182–188. [24] J.O. Haller, G. Kassner, A. Ostrowitz, K. Kottmeler, L.P. Perfschuk, Pilomatrixoma (calcifying epithelioma of malherbe): radiographic features, Radiology 123 (1) (1977) 151–153. [25] R.M. Henkelman, J.F. Watts, W. Kucharczyk, High signal intensity in MR images of calcified brain tissue, Radiology 179 (1) (1991) 199–206. [26] J. Hughes, A. Lam, M. Rogers, Use of ultrasonography in the diagnosis of childhood pilomatrixoma, Pediatr. Dermatol. 16 (5) (1999) 341–344.
5. Conclusion In conclusion, we described the radiological characteristics of pilomatricomas in 31 patients. Although the radiological findings of pilomatricomas are still nonspecific and the defintive diagnosis must be made by pathology, we believe that the results of this study may be helpful in distinguishing the tumor from other subcutaneous soft tissue masses. Consideration should be given to the possible diagnosis of a pilomatricoma when single or multiple subcutaneous soft tissue masses are found with different patterns of calcifications on MDCT images and ring-like, reticular appearances and circular target sign on MR images.
99
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Research article
Pilomatricoma (calcifying epithelioma): MDCT and MR imaging findings in 31 patients with radiological-pathological correlation Kun-ming Yia,1, Kang Chenb,1, Lu Wanga, Xiao-juan Denga, Ying Zengc, Yi Wanga,
T
⁎
a
Deptartment of Radiology, Institute of Surgery Research and Third Affilated Hospital, Army Medical University, Chongqing 400042, China Department of Radiology, First Affilated Hospital, Army Medical University, Chongqing 400038, China c Department of Pathology, Institute of Surgery Research and Third Affilated Hospital, Army Medical University, Chongqing 400042, China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Pilomatricoma Calcifying epithelioma of Malherbe Multi-detector computed tomography (MDCT) Magnetic resonance imaging (MRI) Pathology
Purpose: To describe the radiological characteristics of pilomatricomas on multi-detector computed tomography (MDCT) and magnetic resonance imaging (MRI), and to correlate the radiological findings and pathological features. Materials and methods: The radiological findings of 41 pilomatricomas in 31 patients were retrospectively reviewed. The images were evaluated with emphasis on calcifications, reticular and ring-like appearances, enhancement patterns, circular target sign and peritumoral fat stranding, and correlating these with pathological features. Results: Of the 31 lesions evaluated by MDCT, 25(80.6%) showed different patterns of calcifications which included single in 12(38.7%) lesions and multiple in 13(41.9%) lesions, but peritumoral fat stranding was observed only in 2(6.5%) lesions. MRI scans were performed in 11 patients with 21 lesions, homogeneous and inhomogeneous hypointensities on T1-weighted (T1W) images were showed respectively in 14(66.7%) and 7(33.3%) lesions. On fat-suppressed (FS) T2-weighted (T2W) images, a ring-like hyperintensity was observed in all 21(100%) lesions, reticular hyperintensity, circular target sign, peritumoral fat stranding and secondary anetoderma were seen in 7(33.3%), 4(19%), 4(19%) and 1(4.8%) lesions, respectively; and a significant difference (P < 0.05) was found when comparing the maximum diameter of tumors with (2.3 ± 1.4 mm) and without (1.1 ± 0.3 mm) reticular hyperintensity. On contrast-enhanced T1W images, all 21(100%) lesions were found to have ring-like enhancement and 7(33.3%) of them showed reticular enhancement. The reticular and ring-like appearances on MR images respectively corresponded to the pathological edematous stroma and connective tissue capsule, and the four-layer structures of the circular target sign on FS T2W images also corresponded to pathological calcifications, shadow cells, epithelial cells and connective tissue capsule, respectively. Conclusions: The characteristic radiological findings associated with pilomatricomas include different patterns of calcifications on MDCT images and ring-like, reticular appearances and circular target sign on MR images. Radiological findings are well correlated with pathological nature.
1. Introduction Pilomatricoma, also known as calcifying epithelioma of Malherbe, is an uncommon benign skin tumor of hair follicle origin with a differentiation toward hair cortex cells that was first described by Melherbe and Chenantois in 1880 [1,2]. This tumor can present at any age, about 60% of pilomatricomas occur in patients under the age of 20 years; however, a second peak of occurrence is seen in older patients. There is
a slight female predominance, with a reported female/male ratio of 3:2 [3,4]. Clinically, most pilomatricomas typically present as slowly-growing, well-demarcated, asymptomatic, solitary, hard, mobile superfcial nodule over the underlying skin, with a size range of 0.5–3.2 cm in diameter [3,5,6], while the overlying skin may demonstrate a reddish-blue discoloration or ulceration. Although all hair follicle-bearing skin surfaces can be involved, pilomatricoma most frequently occurs in the
⁎
Corresponding author. E-mail addresses: [email protected] (K.-m. Yi), [email protected] (K. Chen), [email protected] (L. Wang), [email protected] (X.-j. Deng), fi[email protected] (Y. Zeng), [email protected] (Y. Wang). 1 These authors contributed equally to this work. https://doi.org/10.1016/j.ejrad.2018.07.020 Received 13 April 2018; Received in revised form 8 June 2018; Accepted 19 July 2018 0720-048X/ © 2018 Elsevier B.V. All rights reserved.
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
agent and saline solution were injected at 2.5–4 mL/s through an 18gauge plastic intravenous catheter placed in an antecubital vein. Contrast agent volumes were delivered at 2 mL/kg body weight, and the upper limit of dose was set to 120 mL for every patient. Axial, coronal and sagittal reformations were generated for all MDCT studies. The following scanning parameters were applied to 64–row and 16–slice MDCT systems: detector configurations of 64 × 0.625 mm or 16 × 0.6 mm respectively; section thicknesses of 0.625 or 0.6 mm, respectively; reconstruction intervals of 0.625 or 0.6 mm, respectively; and table speeds of 64 or 16 mm per rotation, respectively. The following parameters were applied to all examinations: pitch 0.984, matrix 512 × 512, field of view 180–240 mm, tube voltage 100–120 kV and tube current 300–400 mA.
head and neck regions, and the remainder occur in the upper extremities, the trunk, and the lower extremities. Most pilomatricomas are solitary, but the pilomatricoma can also be familial and associated with autosomal dominant diseases or syndromes, such as Gardner syndrome, myotonic muscular dystrophy (Steinert disease), sarcoidosis, skull dysostosis, Rubinstein-Taybi syndrome and Turner syndrome [5,7]. The mutation in exon 3 of the β-catenin gene (CTNNB1) is associated with pilomatricoma [8]. The treatment of choice is surgical resection and the prognosis is typically good, after complete resection, recurrence is uncommon [9]. Malignant pilomatricoma is an extremely rare condition with few cases reported and overwhelmingly seen in adults [10]. The diagnosis of pilomatricoma is usually based on history, superficial palpation and confirmed by pathologic examination. However, various manifestations of the tumor have often made the diagnosis based on the clinical examination alone inaccurate [1,4,11–14], resulting in unnecessary extensive surgery for an essentially benign condition [1,4]. In such a complicated clinical setting, knowledge of the radiological findings of this uncommon tumor would provide clinicians with the appropriate information and thus could obviate the extensive surgery. Although the CT and magnetic resonance imaging (MRI) findings of pilomatricomas have been reported in a few previous studies [5,6,15–23], some radiological findings found in our clinical practice have seldomly or never been mentioned in the literature. Therefore, the diagnostic accuracy of preoperative radiological examinations and radiological-pathological correlation about this tumor also need to be further evaluated. The purpose of our study was to describe the multidetector computed tomography (MDCT) and MRI findings of the pilomatricomas with pathological correlation, in order to enhance the understanding of this uncommon benign skin tumor.
2.3. MRI examinations
2. Materials and methods
Eleven patients were examined using 1.5-T MRI scanners (Magnetom Aera, Siemens, Erlangen, Germany; Avanto, Siemens, Erlangen, Germany; Sonata, Siemens, Erlangen, Germany). All conventional MR images were obtained at a section thickness of 3–4 mm with an intersection gap of 1 mm, and a 16 × 16 cm – 30 × 30 cm field of view. T1-weighted (T1W) spin-echo (SE) (TR/TE, 620–778 ms/ 10–16 ms), T2-weighted (T2W) fast SE (TR/TE, 1680–4000 ms/ 40–88 ms), and fat-suppressed (FS) T2W fast SE (TR/TE, 2080–4290 ms/27–105 ms) images were obtained in all tumors. Axial and coronal images were obtained with the above sequences. However, in one patient the T2W fast SE images were not included for image analysis due to poor image quality. In all tumors, contrast-enhanced T1W SE imaging (TR/TE, 640–820 ms/8–16 ms) and FS contrast-enhanced T1W SE (TR/TE, 530–840 ms/9–16 ms) images were obtained after intravenous injection of 0.1 mmol/kg of gadopentetate dimeglumine (Magnevist; Bayer Healthcare, Berlin, Germany). Both plain radiography and MRI were obtained in one patient with 1 tumor.
2.1. Study subjects
2.4. Image analysis
This study was approved by the human research committee of the institutional review board. A waiver of informed consent was obtained for this study. We followed Health Insurance Portability and Accountability Act guidelines. All pilomatricomas were retrieved from multiple centers and were confirmed by pathology during the time period 2010–2017. Eventually, 31 patients with preoperative MDCT or/ and MRI examinations and one of them with plain radiography were collected. One of these patients had 11 lesions, a total of 41 pilomatricomas were included in this retrospective study. The following clinical and pathological data were collected: age, sexes, anatomic locations, size and types of the tumor, clinical symptoms and signs, surgical and pathological findings, treatment outcomes, and relevant adverse events and complications. In all patients, the first preoperative MDCT and/or MR images were reviewed. The MDCT and MRI techniques varied somewhat due to the different imaging equipments and the retrospective nature of the study.
All imaging data were directly interfaced with the picture archiving and communication system (PACS), images were reviewed and consensus was reached by two musculoskeletal radiologists (K.Y., Y.W., with 8 and 23 years of experience, respectively). The radiological findings were analyzed considering the following terms: (1) number of lesions; (2) locations of lesions (included head [periauricular region, tempora, eyelid, eyebrow or malar], neck, trunk, extremities [upper or lower extremities]); (3) margins (well- or ill-defined); (4) size (maximum length and width of the lesion were measured on the axial images and maximum craniocaudal depth was measured on the coronal or sagittal images; and volumes were calculated by using the following ellipsoid formula: length × width × depth × 0.5233); (5) calcifications (included number [single or multiple], shapes [micro, nodular or irregular], distribution [diffuse, local multiple or local single] and degrees [complete or incomplete]); (6) density, signal intensity and enhancement patterns on MDCT/MRI scans (compared with the adjacent skeletal muscle); (8) reticular and ring-like hyperintensities on T2W images or FS T2W images (reticular hyperintensity was defined as hyperintensity with fine, septal structures within the lesion, and ring-like hyperintensity was defined as a high signal rim on the periphery of the lesion); (9) circular target sign on FS T2W images (defined as hypointensity in the center of the lesion, hyperintensity in the second layer, hypointensity in the third layer, and hyperintensity in the outer layer); (10) ring-like and reticular enhancements (defined as a thin hyper-density/intensity ring that surrounded at least half of the tumor, and fine, septal enhanced structures within the lesion on contrast-enhanced MDCT/MR images, respectively); (11) secondary anetoderma (defined as peritumoral dermal edema in the skin overlying the lesion which displayed as patchy areas of high signal on FS T2W images); (12) capsules, fat, haemorrhage, necrosis and cysts, and peritumoral fat
2.2. MDCT examinations Twenty-one patients underwent unenhanced and/or dual-phase enhanced (arterial and venous phases) MDCT scanning by a 64–row MDCT (Lightspeed, GE Healthcare, Milwaukee, WI, USA) or a 16–slice MDCT (Emition, Siemens, Erlangen, Germany) system. Both unenhanced and enhanced MDCT scans were obtained in 20 patients (every patient had a pilomatricoma), only unenhanced MDCT scans were obtained in one patient with 11 pilomatricomas and this patient also underwent a MRI scanning. For enhanced MDCT scanning, a dual-head power injector was used to administer contrast agent (Ultravist; Bayer Schering Pharma, Berlin, Germany) at 370 mg iodine/mL followed by 30 mL saline. The contrast 93
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Table 1 Clinical characteristics of patients. Case
Age (y)
Sex
Location
Duration of symptoms before admission (M)
Symptom and sign
1 2 3 4 5 6 7 8 9 10 11 12
14 43 15 16 5 13 26 4 18 67 37 28
M M M F F F M F M M M F
Neck Temporal Neck Neck Periauricular Neck Forearm Temporal Periauricular Periauricular Forehead Eyelid
12 1 6 12 2 10 12 6 60 6 3 3
13 14 15
58 44 17
F M F
Parietal Neck Elbow
6 7 3
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
58 37 48 46 42 114 38 28 10 70 10 11 35 38 42 63
F M F M M F M F M F F M F F F F
Parietal Parietal Parietal Upper arm Parietal Eyebrow Eyebrow Parietal Eyelid Parietal Parietal Eyelid Neck Neck Malar Eyelid
84 192 10 72 24 24 12 6 12 240 3 6 240 240 2 2
Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue redness Soft-tissue Soft-tissue Soft-tissue bullous Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue Soft-tissue
Treatment outcome
Adverse events and complications
Follow-up
mass, R-b mass mass mass mass mass, R-b mass mass mass, R-b mass mass mass,
Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable
excision excision excision excision excision excision excision excision excision excision excision excision
No No No No No No No No No No No No
No No No No No No No No No No No No
mass mass mass,
Curable excision Curable excision Curable excision
No No No
No recurrence No recurrence No recurrence
mass mass mass mass, mass, mass mass mass mass, mass mass, mass, mass mass, mass mass
Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable Curable
No No No No No No No No No No No No No No No No
No No No No No No No No No No No No No No No No
ulcer ulcer
R-b ulcer
R-b
excision excision excision excision excision excision excision excision excision excision excision excision excision excision excision excision
recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence
recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence recurrence
3.2. MDCT findings
stranding were also evaluated on MDCT/MR images.
The MDCT findings of 31 pilomatricomas are given in Table 2. In the 31 tumors, 24(77.4%) lesions were hyperdensity, 6(19.4%) lesions were isodensity and 1(3.2%) lesion was hypodensity on the unenhanced MDCT images. Of the 20 lesions with contrast-enhanced images, 18(90%) showed a mild to moderate enhancement, 2(10%) failed to evaluate the enhancement patterns due to complete calcification within the tumor. All 31 tumors were well-defined margins, appeared as round (12, 38.7%) or round-like (19, 61.3%) shapes, and 11(35.5%) of them showed slight lobular and 3(9.7%) showed necrosis and cysts (Fig. 1). Their maximal diameter ranged from 0.8 to 3.6 cm (mean 1.4 ± 0.9 cm) and volumes from 0.5 to 32.4 cm3 (mean 2.4 ± 1.4 cm3). Of the 31 lesions, 25(80.6%) had different patterns of calcifications within the tumors and 2(6.5%) showed peritumoral fat stranding (Fig. 1). The number of calcifications was single in 12(38.7%) lesions and multiple in 13(41.9%) lesions. Their shapes included micro (12, 38.7%), nodular (13, 41.9%) and irregular (3, 9.7%), and the same lesion could have more than one shape of calcifications. The most common distribution of calcifications was local single (12, 38.7%), followed by local multiple (9, 29%) and diffuse (4, 12.9%). Complete calcification was found in 4(12.9%) lesions and the remaining 21(67.7%) lesions were observed with incomplete calcification. In one patient with 11 pilomatricomas, all of these lesions had nodular calcification, 2 of them were found with complete calcification and the remaining 9 lesions showed calcification surrounded by a layer of soft tissue (Fig. 2). No significant difference (P > 0.05) was found when comparing the maximum diameter of tumors with (1.8 ± 1.0 mm) and without (1.0 ± 0.4 mm) calcification, multiple (1.2 ± 0.4 mm) and single (2.1 ± 1.1 mm) calcification, and complete (0.6 ± 0.2 mm) and
2.5. Statistical analysis Statistical analyses were performed using the SPSS analysis program (version 18.0, IBM). Student’s t-test was used to compare the maximum diameter of tumors with and without calcification, multiple and single calcification, and complete and incomplete calcification on MDCT images, and with and without reticular hyperintensity on MR images. Prior to assessment, the data were tested for homoscedasticity by Levene’s test. A P value of less than 0.05 was considered significant.
3. Results 3.1. Clinical characteristics The clinical characteristics are given in Table 1. Our patient cohort included 14 men and 17 women with an average of 35.3 ± 23.8 years (range, 4–114 years). The duration of symptoms in all the patients admitted to the hospital ranged from 1 to 240 months, with an average of 42.5 ± 21.6 months. In total, 41 tumors were identified among these patients, and located in the subcutaneous fat layer of the head (31, 75.6%), neck (7, 17.1%) and extremities (3, 7.3%). Among the 31 patients, 11 of the 41 lesions were found in the same patient (case 17) and located in the same anatomic region. The main complaint was either a painless or tender soft-tissue mass for 38(92.7%) lesions. The other indications included foreign body, cyst, swelling, abscess, ulcer, reddish-blue discoloration (abbreviated as R-b), bullous appearance and local pain.
94
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
hyperintensity was observed in all 21(100%) lesions, and reticular hyperintensity, circular target sign, peritumoral fat stranding and secondary anetoderma were found in 7(33.3%), 4(19%), 4(19%) and 1(4.8%) lesions, respectively (Fig. 3). All of the 4 lesions with circular target sign occurred in same patient and both MRI and CT scans were performed on this patient (Fig. 2). Significant difference (P < 0.05) was found when comparing the maximum diameter of tumors with (2.3 ± 1.4 mm) and without (1.1 ± 0.3 mm) reticular hyperintensity (Table 2). Fast SE T2W images were only obtained in 10 lesions in our study, 4(40%) of them were hypointensity and 6(60%) were heterogeneous hyperintensity, and reticular, ring-like and cystic hyperintensities were observed in 6(60%), 2(20%) and 1(10%) lesions, respectively. On FS enhanced T1W images, all 21(100%) lesions were found to have ringlike enhancement and 7(33.3%) of them showed reticular enhancement (Fig. 3).
Table 2 Radiological findings in the 41 pilomatricomas. Radiological findings
Locations Head Neck Extremities Shapes Round-like Round Irregular Slight lobular Density Isodensity Hypodensity Hyperdensity Necrosis and cysts Peritumoral fat stranding Calcifications Number Single Multiple Shapes Micro Nodular Irregular Distribution Diffuse Local multiple Local single Degrees Complete Incomplete No calcification T1W images (n = 21) Homogeneous hypointensity Heterogeneous hypointensity T2W images (n = 10) Homogeneous hypointensity Reticular hyperintensity Ring-like hyperintensity Cystic hyperintensity FS T2W images (n = 21) Ring-like hyperintensity Reticular hyperintensity Circular target sign Peritumoral fat stranding Secondary anetoderma FS enhanced T1W images (n = 21) Ring-like enhancement Reticular enhancement No enhancement Size Maximal diameter (cm) Volume (cm3)
No. (%) of lesions/mean ± SD MDCT (n = 31)
MRI (n = 21)
25(80.6) 6(19.4)
17(81.0) 1(4.8) 3(14.3)
19(61.3) 12(38.7) 0 11(35.5)
7(33.3) 12(57.1) 2(9.5) 6(28.6)
6(19.4) 1(3.2) 24(77.4) 3(9.7) 2(6.5) 25(80.6)
3.4. Accuracy of radiological diagnosis and pathology of the tumor The accuracy rate of the preoperative radiological diagnosis of pilomatricoma by MDCT and MRI in our study was only 12.9%(4/31) and 14.3%(3/21), respectively. The radiological diagnoses included pilomatricoma (n = 7), dermoid cyst (n = 3), malignancy (n = 3), hemangioma (n = 6), calcification (n = 4), lymph node (n = 4) and benign lesion (n = 14). All tumors were subsequently surgically resected. Specimens were fixed in 4% buffered formalin and routinely processed and embedded in paraffin; 4-μm-thick sections were stained with hematoxylin and eosin (HE). Histological features of these tumors were as follows: (1) irregularly shaped islands of epithelial cells which composed of basaloid cells and shadow cells were observed in all tumors; (2) calcification, ossification, hemorrhage, and adipose cells were seen in 40(97.6%), 1(2.4%), 4(9.6%) and 1(2.4%) tumors, respectively; (3) edematous and fibrous stroma with various degrees of inflammatory cell infiltration and lymphatic vascular dilatation was observed in 36(87.8%) tumors; (4) connective tissue capsule was observed to be complete in 40(97.6%) tumors and incomplete in 1(2.4%) tumor; (5) 31(75.6%) of the 41 tumors were found with various degrees of collagen fibrils, inflammatory cell infiltration and vascular proliferation, and most vascular proliferation occurred in the stroma, but some occurred in the shadow cells; (6) various degrees of peritumoral inflammatory changes were seen in 8(19.5%) tumors; (7) intense inflammatory cell infiltration in peritumoral fatty tissues were observed in tumors accompanied by peritumoral fat stranding on MDCT or FS T2W images (Figs. 1 and 3).
12(38.7) 13(41.9) 12(38.7) 13(41.9) 3(9.7) 4(12.9) 9(29) 12(38.7) 4(12.9) 21(67.7) 6(19.4) 14(66.7) 7(33.3) 4(40.0) 6(60.0) 2(20.0) 1(10.0) 21(100) 7(33.3) 4(19.0) 4(19.0) 1(4.8) 21(100) 7(33.3) 0 1.4 ± 0.9 2.4 ± 1.4
3.5. Treatment outcomes and follow-up
1.4 ± 1.1 3.8 ± 2.3
All patients underwent surgical excision using a direct approach by making the incision over the tumor, and one of them with excision of the overlying skin. At surgery, all lesions did not involve the muscle and they were easily dissected free from the subcutaneous fat. All 31 patients were followed up post-operatively for periods ranging from 3 months to 2 years (mean, 9 months), and did not exhibit local recurrence and no new lesion developed.
incomplete (1.6 ± 0.9 mm) calcification.
3.3. MRI findings The MRI findings of 21 pilomatricomas are given in Table 2. The margins of all tumors were well-defined, their maximal diameter ranged from 0.8 to 4.5 cm (mean 1.4 ± 1.1 cm) and volumes from 0.5 to 40.5 cm3 (mean 3.8 ± 2.3 cm3). One of these tumors showed a high density mass with popcorn calcification on plain radiographs. On T1W images, homogeneous hypointensity was found in 14(66.7%) tumors, the remaining 7(33.3%) tumors showed inhomogeneous hypointensity, 3(14.3%) of them were found to have reticular and small nodular hyperintensities and 4(19%) showed ring-like hyperintensity. On FS T2W images, all 21(100%) lesions were heterogeneous signal intensity with combinations of high and low signals. A ring-like
4. Discussion In the current study, we present the largest series of radiological studies of pilomatricomas to date. Our study demonstrates that pilomatricomas typically present as a well-defined subcutaneous soft tissue mass with different patterns of calcifications and mild to moderate enhancements on MDCT images, and ring-like, reticular appearances and circular target sign on MR images. In previous studies, plain radiography demonstrated that pilomatricoma appeared as a solitary, sharply demarcated subcutaneous tumor with extensive sand-like or dense focal calcification [24]. In our study, 95
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Fig. 1. A 13-year-old girl with a pilomatricoma in the right neck. (a) Non-enhanced axial MDCT image shows a well-defined, subcutaneous and heterogeneous mass oppressing the adjacent sternocleidomastoid muscle (black arrows), with diffuse micro and small nodular calcifications within the lesion and peritumoral fat stranding (white arrows). (b and c) Arterial- and venous-phase contrast-enhanced axial MDCT images reveal a mild to moderate enhancement within the mass. (d) Contrast-enhanced coronal MDCT image shows the mass with a slight lobular appearance (arrows). (e) Microscopic view (HE, ×100) of a pathological specimen shows a connective tissue capsule (black arrows), some foreign body type giant cells (∗) and a large number of enucleated shadow cells (☆). Edema (white arrows) is located between the shadow cells and the connective tissue capsule. (f) Another microscopic view (HE, ×100) reveals calcifications (∗) in the center and shadow cells (☆) on the periphery of the mass. Abundant edematous and fibrous stroma with inflammatory cell infiltration (white arrows) and edema (black arrows) are observed among epithelial cell islands.
distribution and degree, to the best of our knowledge, detailed classifications have not been reported in previous literature. Thus, we found that irregular calcification and multiple shapes of calcification in the same lesion can occur in the pilomatricoma which were observed in 3 and 4 lesions, respectively, in our patients. In addition, we also found that 4 lesions were nearly completely calcified, therefore, it is very difficult to distinguish such pilomatricoma from pure calcification on CT images.
one of the lesions showed popcorn calcification in the tumor on plain radiographs. On CT, the pilomatricoma has been described as well-defined subcutaneous masses with varying amounts of calcification and variable enhancement [17–19,23] and calcifications have been reported in 63.6–81% of pilomatricomas [19,23]. In our study, 80.6% of lesions showed different patterns of calcification on MDCT which is consistent with the previous results [19,23]. Furthermore, our study categorized the calcification patterns in terms of number, shape, 96
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Fig. 2. A 37-year-old man with pilomatricoma in the head. (a) Axial FS T2W MR image shows two well-defined subcutaneous nodules with circular target sign (arrows) which appear as hypointensities in the center and third layer, and hyperintensities in the second and outer layers of the lesions. (b) Axial T1W MR image shows homogeneous hypointensity in the center and ring-like hyperintensity on the periphery of the lesions (arrows). (c and d) Contrast-enhanced axial and sagittal T1W MR images show ring-like enhancement (arrows). (e) Contrast-enhanced coronal FS T1W MR image also shows ring-like enhancement, but no reticular enhancement (arrow). (f–j) Non-enhanced axial MDCT images show a total of 11 well-defined subcutaneous nodules (arrows), two of them are nearly completely calcified (thick arrows), and the remaining 9 lesions appear as central calcifications surrounded by a layer of soft tissue (thin arrows). (k) A local magnified view of the figure 2a and (l) scanning view (HE, ×40) of a pathological specimen clearly reveal the one-to-one correspondence between radiological and pathological fourlayer concentric circular structures from the center to the periphery: the central hypointensity (C) vs. calcification (☆); hyperintensity in the second layer (S) vs. shadow cells (∗); hypointensity in the third layer (T) vs. epithelial cells (white arrows); and hyperintensity in the outer layer (O) vs. connective tissue capsule (black arrows). (m) Pathological microscopic view (HE, ×100) shows the connective tissue capsule with collagen fiber and inflammatory cell infiltration (arrows).
[19,23]. In our series, necrosis and cysts were showed in 9.7% of lesions on MDCT, however, pathological examinations revealed necrosis in 21 of the 41 lesions. Therefore, the micronecrosis within the pilomatricoma is also not easy to be visualized on both CT and MRI images. In our study, 90% of lesions showed as a mild to moderate enhancement, which enabled us to confirm the previous observations [3,17–19]. However, Ichikawa et al. described pilomatricoma as a hypervascular
Although CT has the ability to depict the great majority of calcifications, it can hardly visualize microscopic calcifications due to its resolution limit [18,19]. It is also easy to explain that calcifications were seen in 40(97.6%) lesions on pathology, but MDCT images did not show any calcification in 6 of them in our study. To our knowledge, the cystic degeneration is a rare radiological feature in pilomatricoma, there have only been three reports about pilomatricoma with cystic degeneration 97
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Fig. 3. A 17-year-old man with pilomatricoma in the elbow. (a) Coronal FS T2W MR image shows a well-defined subcutaneous mass with ring-like and reticular hyperintensities, and hypointense nodules are showed in its interior (thin arrow) and patchy peritumoral edema is seen in the skin overlying the mass (thick arrows). (b) Axial FS T2W MR image shows peritumoral fat stranding (arrows). (c) Axial T2W MR image shows reticular and ring-like hyperintensities with peritumoral dermal edema (arrows). (d) Coronal T1W MR image shows a nonhomogeneous hypointense mass with small pieces of high signals in its interior (arrows). (e) Contrast-enhanced coronal T1W MR image shows a ring-like enhancement on the periphery of the mass (arrows). (f) Contrast-enhanced axial FS T1W MR image shows reticular enhancement in the interior of the tumor and a ring-like enhancement on its periphery (arrows). (g) Scanning view (HE, ×40) of a pathological specimen reveals epithelial cell islands composed of basaloid cells (☆) and shadow cells (∗). Abundant edematous and fibrous stroma with inflammatory cell infiltration (white arrows) is observed among epithelial cell islands and vascular proliferation (black arrows) is seen in the distribution zone of the shadow cells. (h) Another scanning view (HE, ×40) reveals peritumoral dermal edema (∗) and connective tissue capsule with collagen fiber and inflammatory cell infiltration (black arrows). Abundant edematous and fibrous stroma with inflammatory cell infiltration (white arrows) is observed among epithelial cell islands (☆). (i) Pathological microscopic view (HE, ×100) reveals lymphatic dilation (☆) in the peritumoral dermis.
hyperintense rim showed enhancement on contrast-enhanced T1W images, but no enhancement in the center of the lesion. Since then, Lim et al. [19] described reticular hyperintensity and reticular enhancement on FS T2W and contrast-enhanced T1W images in all of the five patients they reported. Bulman et al. [23] and Kato et al. [6] reported 16 and 9 pilomatricomas, respectively, with reticular hyperintensity on FS T2W images. Similarly, in our study, the reticular hyperintensity and enhancement were also observed respectively in 7 of 21 lesions on FS T2W and contrast-enhanced T1W images. With regard to the pathological basis of the reticular hyperintensity, Hoffmann et al. [21] suggested that the internal reticulations and septations on MR images correspond to the basaloid cells. In other authors’ opinion, the internal reticulations and septations were reported to represent the intercellular edematous stroma, because the basaloid cells were considered as the sheets of vascular epithelial cells [6,19,22]. Our pathological examinations revealed edematous and fibrous stroma with various degrees of inflammatory cell infiltration and vascular proliferation in 7 tumors. Our results, therefore, suggest that the reticular appearances on MR images correspond to the pathological distribution of intratumoral stroma. A few studies [6,19,26] have reported that the ring-like appearance
mass on angiograms [16]. On MRI, most of the pilomatricomas have been described as showing a homogeneous intermediate signal intensity on T1W images and heterogeneous intermediate signal intensity on T2W and FS T2W images [6,16,19–21,23]. However, some pilomatricomas have been described as showing an inhomogeneous signal intensity on T1W images [15,22], the causes are not very clear at all. Masih et al. [22] described a non-calcified pilomatricoma with a lot of keratin accumulation on pathological examinations. Henkelman et al. [25] indicated that particulate calcium can reduce T1 relaxation times by a surface relaxation mechanism. In our study, 3 lesions were also found to have reticular and small nodular hyperintensities on T1W images, and hemorrhage was observed on pathological examinations. These results, therefore, suggest that the high signal on T1W images corresponded to the distribution of keratin accumulation, hemorrhage and calcifications on pathological specimens. The reticular hyperintensity have been known as a MRI characteristic of pilomatricomas on T2W, FS T2W and contrast-enhanced T1W images [6,19,21–23]. Hoffmann et al. [21] first described a pilomatricoma with high signal bands radiating from the lesion center to high signal periphery on T2W and FS inversion recovery images, and the 98
European Journal of Radiology 106 (2018) 92–99
K.-m. Yi et al.
Conflicts of interest
of pilomatricomas on ultrasound or MR images corresponded to the connective tissue capsule. Our pathological examinations also showed connective tissue capsules with various degrees of collagen fibrils, inflammatory cell infiltration, and vascular proliferation in 31 of 41 tumors. And similarly, in our 21 pilomatricomas evaluated by MRI, all these tumors showed a ring-like appearance. In our study, the ring-like hyperintensity on FS T2W images was showed to be more frequent (21/ 21) than those (2/10) on conventional T2W images. It may be due to the fact that the ring-like hyperintensity is confounded by the surrounding hyperintense fat tissues on conventional T2W images which can result in a difficulty in identification [6]. In our study, 4 of 21 pilomatricomas showed a circular target sign on FS T2W images. To the best of our knowledge, the circular target sign found in the pilomatricoma has never been reported in previous literature. Furthermore, by comparison of radiographic and pathological findings, we found that the central hypointensity on FS T2W images was confirmed to be calcification on MDCT images and pathological specimens. And then, the hyperintensity in the second layer, hypointensity in the third layer, and hyperintensity in the outer layer of the circular target sign were also corresponding to the pathological shadow cells, epithelial cells and connective tissue capsule, respectively. Pathological examinations showed no lymphatic vascular dilatation and increased numbers of blood vessels within these tumors which may explain all of the 4 pilomatricomas with a ring-like enhancement in the outer layer, but no enhancement in the center and other two layers of the tumors. In addition, the 4 pilomatricomas with circular target sign appeared as a central hypointensity surrounded by a layer of hyperintensity on T1W images and a nodular calcification surrounded by a layer of soft tissue on MDCT, this may be explained by our pathological results that the central hypointensity/nodular calcification on T1W/MDCT images corresponds to the pathological calcification, and the peripheral hyperintensity/soft tissue density on T1W/MDCT images is also consistent with the distribution of an absolute majority of shadow cells on pathological specimens. Therefore, our results may suggest that the shadow cells in the pilomatricomas show as hyperintensity not only on T2W and FS T2W images, but also on T1W images. There are several limitations in our study. The relative smaller sample size is a limitation. Pilomatricoma is an uncommon tumor and, moreover, the patients with pilomatricoma may never come for diagnostic imaging because of the lesion’s small size and superficial location. The retrospective nature of the study is another limitation. It means that there were no perfect imaging protocols, so we did not obtain both MDCT and MR images in the majority of patients. In addition, different MDCT and MRI equipments and techniques were used. However, these problems are simply unavoidable due to the rarity of this type of tumor, and this should not have signifcantly affected the radiological characteristics studied.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. There are no conflicts of interest to report. References [1] S. Kaddu, H.P. Soyer, L. Cerroni, W. Salmhofer, S. Hädl, Clinical and histopathologic spectrum of pilomatricomas in adults, Int. J. Dermatol. 33 (10) (1994) 705–708. [2] A. Malherbe, J. Chenantais, Note surl’epitheliome calcifié des glandes sebacées, Prog. Med. 8 (1880) 826–828. [3] N.J. Wells, G.K. Blair, J.F. Magee, D.M. Whiteman, Pilomatrixoma: a common, benign childhood skin tumour, Can. J. Surg. 37 (6) (1994) 483–486. [4] C.G. Julian, P.W. Bowers, A clinical review of 209 pilomatricomas, J. Am. Acad. Dermatol. 39 (2 Pt. 1) (1998) 191–195. [5] S. Duflo, R. Nicollas, S. Roman, G. Magalon, J.M. Triglia, Pilomatrixoma of the head and neck in children: a study of 38 cases and a review of the literature, Arch. Otolaryngol. Head Neck Surg. 124 (11) (1998) 1239–1242. [6] H. Kato, M. Kanematsu, H. Watanabe, A. Nagano, E. Shu, M. Seishima, T. Miyazaki, MR imaging findings of pilomatricomas: a radiological-pathological correlation, Acta radiol. 57 (6) (2016) 726–732. [7] D. Kwon, K. Grekov, M. Krishnan, R. Dyleski, Characteristics of pilomatrixoma in children: a review of 137 patients, Int. J. Pediatr. Otorhinolaryngol. 78 (8) (2014) 1337–1341. [8] U. Gat, R. DasGupta, L. Degenstein, E. Fuchs, De Novo hair follicle morphogenesis and hair tumors in mice expressing a truncated beta-catenin in skin, Cell 95 (5) (1998) 605–614. [9] S.F. Hassan, E. Stephens, S.C. Fallon, D. Schady, M.J. Hicks, M.E. Lopez, D.A. Lazar, M.A. Rodriguez, M.L. Brandt, Characterizing pilomatricomas in children: a single institution experience, J. Pediatr. Surg. 48 (7) (2013) 1551–1556. [10] J.L. Herrmann, A. Allan, K.M. Trapp, M.B. Morgan, Pilomatrix carcinoma: 13 new cases and review of the literature with emphasis on predictors of metastasis, J. Am. Acad. Dermatol. 71 (1) (2014) 38–43. [11] A. Danielson-Cohen, S.J. Lin, C.A. Hughes, Y.H. An, J. Maddalozzo, Head and neck pilomatrixoma in children, Arch. Otolaryngol. Head Neck Surg. 127 (12) (2001) 1481–1483. [12] N. Kumaran, A. Azmy, R. Carachi, P.A. Raine, J.H. Macfarlane, A.G. Howatson, Pilomatrixoma — accuracy of clinical diagnosis, J. Pediatr. Surg. 41 (10) (2006) 1755–1758. [13] M.Y. Lan, M.C. Lan, C.Y. Ho, W.Y. Li, C.Z. Lin, Pilomatricoma of the head and neck: a retrospective review of 179 cases, Arch. Otolaryngol. Head Neck Surg. 129 (12) (2003) 1327–1330. [14] N.A. Roche, S.J. Monstrey, G.E. Matton, Pilomatricoma in children: common but often misdiagnosed, Acta Chir. Belg. 110 (2) (2010) 250–254. [15] T.J. Hsieh, C.K. Wang, K.B. Tsai, Y.W. Chen, Pilomatricoma: magnetic resonance imaging and pathological evaluation, J. Comput. Assist. Tomogr. 32 (2) (2008) 320–323. [16] T. Ichikawa, Y. Nakajima, H. Fujimoto, A. Koyama, M. Honma, M. Yatsuzuka, K. Ohtomo, G. Uchiyama, S. Ushigome, S. Ohba, Giant calcifying epithelioma of malherbe (pilomatrixoma): imaging features, Skeletal Radiol. 26 (10) (1997) 602–605. [17] P.M. Som, J.M. Shugar, A.R. Silvers, CT of pilomatrixoma in the cheek, AJNR 19 (7) (1998) 1219–1220. [18] K.H. Lee, H.J. Kim, C.H. Suh, Pilomatricoma in the head and neck: CT findings in three patients, J. Comput. Assist. Tomogr. 24 (2) (2000) 332–335. [19] H.W. Lim, S.A. Im, G.Y. Lim, H.J. Park, H. Lee, M.S. Sung, B.J. Kang, J.Y. Kim, Pilomatricomas in children: imaging characteristics with pathologic correlation, Pediatr. Radiol. 37 (6) (2007) 549–555. [20] L.H.De. Beuckeleer, A.M.De. Schepper, I. Neetens, Magnetic resonance imaging of pilomatricoma, Eur. Radiol. 6 (1) (1996) 72–75. [21] V. Hoffmann, T. Roeren, P. Moller, G. Heuschen, MR imaging of a pilomatrixoma, Pediatr. Radiol. 28 (4) (1998) 272. [22] S. Masih, S.M. Sorenson, A. Gentili, L.L. Seeger, Atypical adult non-calcified pilomatricoma, Skeletal Radiol. 29 (1) (2000) 54–56. [23] J.C. Bulman, S.O. Ulualp, V. Rajaram, K. Koral, Pilomatricoma of childhood: a common pathologic diagnosis yet a rare radiologic one, AJR 206 (1) (2016) 182–188. [24] J.O. Haller, G. Kassner, A. Ostrowitz, K. Kottmeler, L.P. Perfschuk, Pilomatrixoma (calcifying epithelioma of malherbe): radiographic features, Radiology 123 (1) (1977) 151–153. [25] R.M. Henkelman, J.F. Watts, W. Kucharczyk, High signal intensity in MR images of calcified brain tissue, Radiology 179 (1) (1991) 199–206. [26] J. Hughes, A. Lam, M. Rogers, Use of ultrasonography in the diagnosis of childhood pilomatrixoma, Pediatr. Dermatol. 16 (5) (1999) 341–344.
5. Conclusion In conclusion, we described the radiological characteristics of pilomatricomas in 31 patients. Although the radiological findings of pilomatricomas are still nonspecific and the defintive diagnosis must be made by pathology, we believe that the results of this study may be helpful in distinguishing the tumor from other subcutaneous soft tissue masses. Consideration should be given to the possible diagnosis of a pilomatricoma when single or multiple subcutaneous soft tissue masses are found with different patterns of calcifications on MDCT images and ring-like, reticular appearances and circular target sign on MR images.
99