Limits of fine-needle aspiration cytology in diagnosing pilomatrixoma: a series of 25 cases with clinico-pathologic correlations.

BACKGROUND: Pilomatrixoma (PMX) is a benign, quite uncommon, skin neoplasm, which is frequently misdiagnosed by clinicians. AIM: We have analyzed 25 PMX to determine the agreement between clinical diagnosis, preoperative FNA characteristics, and corresponding histopathological specimens; moreover, reliable cytologic criteria for PMX and the differential diagnosis to avoid cytological pitfalls have been emphasized.
MATERIALS AND METHODS: By fine-needle aspiration (FNA) cytology a series of consecutive cases of PMX collected during last 5 years were studied. Smears were stained by Papanicolau and May-Grünwald-Giemsa.
RESULTS: Patients affected by PMX were 11 males, 14 females (ratio 1:1.27); the mean age was 32.72 years with age range 3-78 years, being 72% (18/25) of patients 40 years or less. PMX was mainly distributed in the head-neck region (52%), scalp (16%), upper/lower arms (28%), and chest (4%). The observed diagnostic cytological features were represented by clusters of basaloid epithelial cells, shadow or ghost cells, inflammatory background, calcification, and giant cells. Unfortunately, not all these morphological aspects were always disclosed in smears, thus making the cytological preoperative diagnosis questionable and problematic.
CONCLUSIONS: The experience of a well-trained cytopathologist should distinguish the relevant FNA features in terms of smear background, architecture, and cell morphology. The most dangerous mistake in FNA diagnosis of PMX regards a diagnosis of primary malignant or metastatic cutaneous lesions.

Introduction

Pilomatrixoma (PMX), also known as “pilomatricoma” or “calcifying epithelioma of Malherbe,” is an uncommon benign slow-growing neoplasm originating from pluripotential cells of the outer sheath cell of the hair follicle root.[1-3] PMX usually presents as a small, solitary, asymptomatic, firm, subcutaneous nodule, sometimes exhibiting a bluish discoloration of the overlying intact skin;[45] alternatively, patients may experience pain in response to palpation, discomfort, itching, episodes of infection, discharge, or even ulceration.[56] PMX may be located in any part of the body, except the palms and soles, although it shows a predilection for the head and neck region, followed by the upper extremities, the trunk and the lower extremities.[35-9]

The low clinical accuracy for PMX could be attributed to a possible lack of awareness of the tumor by clinicians, which determines often misdiagnosis toward a variety of benign and malignant skin lesions.[13510-12]

Fine-needle aspiration (FNA) has been described as a relevant preoperative approach in the diagnostic investigation of a mass in the head and neck, but sometimes, the cytological features of PMX are incompletely observed, leading to mistaken diagnosis on FNA. In the present study, we have analyzed a series of 25 PMX to determine the agreement between clinical diagnosis, preoperative FNA characteristics, and corresponding histopathological specimens; moreover, reliable cytologic criteria for PMX and the differential diagnosis to avoid cytological pitfalls in skin tumors have been emphasized.

Materials and Methods

The series studied includes 25 cases PMX, which were consecutively collected at the Cytopathological Laboratory of the Division of Anatomic Pathology, “Polyclinic G. Martino” of Messina (Italy) during the last 5 years. The FNA test was requested by the plastic surgeon to exclude a malignant lesion. Smears were preoperatively obtained by the aspiration procedure with a 23 G needle attached to a 10-ml disposable syringe. The smears collected were air dried and stained with May-Grünwald-Giemsa; additional samples were fixed in 95% ethanol and stained with the Papanicolaou method. No cell blocks were made by each aspirate.

After the FNA procedure, all patients were successively surgically treated and subjected to routine histology. For this purpose, surgical specimens were fixed in 10% neutral-buffered formalin for 12-24 hours at room temperature, then included in paraffin blocks at 56°C, cut into 4-5 μm thick sections, and finally stained with the hematoxylin/eosin routine method.

Results

Patients affected by PMX were 11 males, 14 females (male-to-female ratio 1:1.27); the mean age was 32.72 years with age range 3-78 years, being 72% (18/25) of patients 40 years or less. Data concerning patient's age, sex, location, clinical, preoperative, and cytological as well as histopathological diagnoses of cases were analytically reported in Table 1. All patients had a single localization, mainly distributed in the head-neck region (52%), scalp (16%), upper/lower arms (28%), and chest (4%).

Table 1

Clinical characteristics and pre-/postsurgical diagnoses

Generally, all aspirates were cellular containing moderate to high number of cells; the combination and proportion of different cellular components varied from case to case. The key component consisted of clusters of small- and medium-sized basaloid cells [Figure 1a–b]; their nuclei were regular, round to ovoid, vesicular, with dispersed chromatin, and occasional large nucleoli [Figure 1a–b]. The cytoplasm of these basaloid elements was generally scanty and well defined [Figure 1a–b]. Clusters or sporadic single ghost (shadow) cells were found in about 60% of smears [Figure 1b–c], although keratin fragments and inflammatory lymphoid or granulocytic elements made sometimes difficult their detection [Figure 1c–d]; less frequently naked nuclei, cellular debris, and calcium deposits were observed, while evidence of multinucleated giant cells was occasional.

Figure 1

Clusters of basaloid elements were always present in PMX smears, well evident either by Papanicolau (a, 360×) or May-Grünwald-Giemsa (b, 360×); ghost cells were generally isolated and sporadic (b, c - May-Grünwald-Giemsa, 360×). Note the inflammatory background (c, May-Grünwald- Giemsa, 360× - d, Papanicolau, 360×)

In the current series, the clinical diagnosis of PMX corresponds to an equivalent preoperative FNA diagnosis in four cases (16%), even if the hypothesis of a skin nodule/tumor was performed in nine cases (36%) but the cytology confirmed suspicious findings only in five lesions (20%). By FNA, a correct diagnosis of PMX was established in 11/25, while histopathology of resected tumors, always located in the dermis, showed the peculiar diagnostic characteristics of PMX in all cases; therefore, in PMX, the agreement between cytology and histopathology was 44%.

Discussion

Some reports on FNA of PMX have emphasized the possibility of false positive cytological diagnosis[13-17] particularly underlying the risks of the inadequate FNA sampling as well as the incomplete knowledge of relevant clinical data. Moreover, better results have been obtained when FNA on PMX have been carried out and interpreted by the same cytopathologist,[1318] while difficulties merged when age of patients, size and location of PMX were unusual or inconclusive.[413] In our series, the great majority of PMX was located in common locations, such as head/neck, scalp, and arms, with only an isolated case encountered on the chest; in addition, a slight predominance in female patients was observed, similarly to that elsewhere referred.[41219] Our patients showed a mean age of 32.72 years with 72% of patients 40 years or less; these data are in agreement with those reported by some authors,[412] but in contrast with other previous reports in which patient's age was generally less 20 years.[2021] This appreciable discrepancy may be probably due to the relatively low number of patients examined in all studies present in the literature.

The presence of the complete spectrum of PMX cytological characteristics, such as basaloid cells, calcium deposits, naked nuclei, shadow (“ghost”), and giant cells, inflammatory background, allows a significant well-done preoperative diagnosis. Unfortunately, some cytological findings of PMX, like true patognomonic “ghost” cells, calcium deposits, and multinucleated cells, were absent in about 40% of our cases; consequently, in these cases, the accurate preoperative diagnosis becomes problematic, with a potentially erroneous clinical management. On the other hand, the correct identification of PMX by FNA in our patients was limited to 11 cases (44%), similarly to that elsewhere reported in the literature.[1021-24]

The most dangerous mistake in FNA diagnosis of PMX regards a diagnosis of primary malignant or metastatic cutaneous lesions, which may suggest unnecessary and heavy radical surgery, mainly in young adults. In particular, the absence of atypical mitoses, significant nuclear pleomorphism, and coarse clumping of chromatin may allow us to rule out the diagnosis of malignant skin neoplasm such as squamous cell carcinoma. Moreover, basaloid cells present in PMX may orient toward the occurrence of basal cell carcinoma, although the presence of peripheral palisading cells in clusters and as well as the absence of nucleoli are greatly in contrast with the diagnosis of PMX. However, basaloid elements of PMX observed in FNA have been confused with small cells with moderate amount of cytoplasm present in rhabdomyosarcoma[18] or alternatively in Merkel cell neuroendocrine skin carcinoma[12] or in small cell lung metastatic carcinoma.[19] In order to exclude these latter aggressive lesions, the other classical cytological features of PMX should be emphasized, even if the knowledge of precise clinical informations, such as the elderly age in Merkel cell tumor, is mandatory. Finally, mostly basaloid cells in cohesive groups with hyperchromatic nuclei have been found in skin appendage tumor, such as cylindroma, spiradenoma, hidroadenoma, or adenoid cystic carcinoma; in these neoplasms, the absence of ghost cells as well as multinucleated giant cells and acellular calcific deposits should be critical to exclude the PMX diagnosis. By contrast, the appearance of shadow cells has been reported also in other cutaneous and visceral neoplasm, particularly, ovarian endometrioid adenocarcinomas;[25] therefore, the cutaneous metastatic localization of this neoplasm should be misinterpreted as a primary PMX mimicking the cytological characteristics.