| Abstract|| |
Background: Skin adnexal tumors are daunting diagnostic problems. Cytologic atypia does not always imply malignancy and "typia" does not underscore a benign course. Bernard Ackerman first described criteria on silhouettes that enable distinction between the two. Aims: To evaluate the histologic features on silhouettes of benign and malignant skin adnexal tumors. To identify overlaps and confounding features. Materials and Methods: A blinded retrospective review of all skin adnexal neoplasms between 1995 and 2007 was done, with a total of 68 cases. We studied 16 histologic parameters on scanner view and categorized them as benign or malignant. They were compared with the final histologic diagnosis. Statistical analysis was performed using chi-square test. Results: 15 criteria attained statistical significance. Features that proved highly sensitive and specific were: Circumscription, ulceration, uniform size of cell aggregates, discrete arrangement, preserved adnexae and necrosis. Criteria that were sensitive but not very specific include: Symmetry, V-shape, vertical orientation, smooth margins, compressed fibrous tissue, type of clefting, shelling out and geometric shapes. Presence of epithelial cells in singles was not helpful. Conclusion: Malignant skin adnexal tumors are differentiated accurately from benign ones by their contrasting silhouettes. Pathologists should heed the impression formed on scanner view, before evaluating cytologic features.
Keywords: Histopathology, silhouette, skin adnexal neoplasms
|How to cite this article:|
Tirumalae R, Roopa M O. Benign vs. Malignant skin adnexal neoplasms: How useful are silhouettes?. Indian J Dermatol 2013;58:30-3
|How to cite this URL:|
Tirumalae R, Roopa M O. Benign vs. Malignant skin adnexal neoplasms: How useful are silhouettes?. Indian J Dermatol [serial online] 2013 [cited 2019 Jul 17];58:30-3. Available from: http://www.e-ijd.org/text.asp?2013/58/1/30/105282
What was known?
Skin adnexal tumors are challenging to diagnose, with conventional cytologic criteria often not helpful. A single tumor may show a spectrum of differentiation, making precise diagnosis difficult.
| Introduction|| |
Troublesome tumors, this term aptly describes cutaneous adnexal tumors, as used by Cotton D.  They pose major diagnostic problems to both surgeons and pathologists. The bewildering array of differentiation they display and the ever-expanding list of entities add further to the confusion.  Subtyping lesions with accuracy becomes important in some cases to recognize clinical association, as in sebaceous tumors and this requires a good understanding of the embryologic origins and structure of cutaneous adnexae.  Many a time, it is only an academic exercise and even after giving it a label, one is unsure of the diagnosis; particularly so for those who do not encounter these tumors often in their practice. In such a situation, it suffices to answer two important questions i.e., (a) Is the lesion benign or malignant? (b) Has it been completely excised?
Traditionally, cytologic/nuclear atypia has been used as the single most important feature to distinguish benign from malignant tumors. Dr. Bernard Ackerman challenged this concept and proposed that contrasting silhouettes/architectural attributes accurately distinguish benign from malignant tumors. He described a list of criteria and successfully applied them to differentiate Spitz nevi from melanomas.  He has also stated that the same can be applied to skin adnexal neoplasms.  This forms the basis of the current study, where we evaluate the utility of some of those features.
The study was undertaken to evaluate the histologic features on silhouettes of benign and malignant skin adnexal tumors; also to identify overlaps and confounding features.
| Materials and Methods|| |
We performed a blinded, retrospective review of all skin adnexal neoplasms reported between 1995 and 2007. There were 68 cases in total. H and E stained sections were reviewed, additional sections were cut when necessary. 16 parameters were assessed on scanning magnification (×4) and categorized as benign or malignant.
The criteria used were: Symmetry vs. asymmetry, well vs. ill-circumscribed margins, smooth vs. jagged borders, vertical vs. horizontal orientation, V-shape of the lesion, presence of compressed fibrous tissue around the tumor, clefting between tumor cells and stroma, clefting between compressed fibrous tissue and surrounding stroma, complete vs. incomplete shelling out of the lesion, ulceration, necrosis en masse, irregular vs. uniform cells nests, discrete nests, geometric shape of cell nests, epithelial cells in singles and preservation of adnexae within the lesion.
The results were compared with the final histologic diagnosis. Univariate analysis was performed using the Chi-square test. Sensitivity and specificity were calculated for each of the features.
| Results|| |
Of the 68 cases, 53 were classified as benign and 15 as malignant based on the silhouette features. This correlated with the final histologic diagnosis in all the cases. The final histologic diagnosis was as follows: Benign: Syringoma-15, Hidradenoma-8, Trichoblastoma-14, Mixed Tumors-5, Others-11 cases and Malignant: Basal cell carcinoma-10, Sebaceous carcinoma-4, Porocarcinoma-1 cases. The reason we have included basal cell carcinomas is that they are trichoblastic in nature.
15 of the 16 criteria used attained statistical significance in differentiating benign from malignant tumors. The only feature that was not useful was arrangement of epithelial cells in singles, which was seen equally in both benign and malignant tumors. The p-value, sensitivity and specificity of each parameter is shown in [Table 1]. Criteria which were 100% sensitive include symmetry, V-shape, type of clefting and necrosis. The criteria that had both high sensitivity and specificity (>80%) include circumscription, ulceration, uniform size of cell aggregates, discrete arrangement, preserved adnexae and necrosis.
Six criteria were shared by all benign tumors, namely smooth borders, compressed fibrous tissue, absence of ulceration, discrete arrangement, preserved adnexae and absence of necrosis. All malignant tumors showed asymmetry, were not V-shaped, clefts between tumor cells and stroma and did not shell out completely.
| Discussion|| |
Skin adnexal epithelial neoplasms are an assorted group of tumors that show differentiation towards pilosebaceous, eccrine or apocrine structures. Often, they show evidence of more than one lineage within a single tumor. , This can be better understood with a good knowledge of the embryology of these structures. Basal germinative cells are noticed in the epidermis by the 4 th week of development, together with clusters of mesenchymal cells that congregate beneath them, forming a crescentic mass. These cells proliferate to become the infundibular epidermis, hair follicles, sebaceous glands and apocrine glands.  This common embryologic derivation explains the wide array of differentiation displayed by these neoplasms. Eccrine glands, on the other hand, are first noticed in the palms and soles by 12 weeks, distinctly separate from the sebaceous-follicular-apocrine unit. In spite of being armed with volumes of literature, it is an intimidating task for the surgical pathologist to categorise them correctly; specially so since they are uncommonly met with.
Atypia, more nuclear than cellular, is regarded as the most important histologic trait favouring malignancy. This term refers to nuclei that are larger, darker, variably sized, irregularly outlined, have coarse chromatin, nucleoli and abnormal mitoses. All these are subjective, especially when they are present focally or to a mild degree. For example, a Spitz nevus or a dermatofibroma may possess sufficient "atypia" to be labelled as a melanoma or a malignant fibrohistiocytic tumor. The same also hold good for adnexal tumors. This justified the need for the present study, to evaluate the utility of silhouettes in achieving this distinction.
The silhouette or scanning magnification view offers the best impression of the architecture of a lesion. A set of silhouette features was described by Dr. Ackerman to differentiate Spitz nevi from melanoma.  These features were also found to be useful in adnexal tumors. A few older articles cite these but of late, there is not too much emphasis laid on these features. We have attempted to apply these criteria to distinguish benign from malignant tumors.
Symmetry, V-shape, stroma-stroma clefting and absence of necrosis proved 100% sensitive in recognizing benign tumors. Symmetry implies that if one were to draw a line through the centre of the lesion, the two halves look roughly similar [Figure 1]. V-shape refers to the tumor having a broader front towards the top, tapering towards the deeper portions. [Figure 2] Clefting is a feature that is generally emphasized in basal cell carcinomas, but little appreciated in other tumors. We found that the nature of the cleft is a very useful tool in practice. In all benign tumors, there was a rim of condensed fibrous tissue surrounding the tumor cells that was separated from the surrounding dermis by a cleft [Figure 3]a. This was not seen in any of the malignancies. All malignant tumors showed clefts directly between the cells and surrounding stroma [Figure 3]b. Necrosis was absent in benign tumors. A note of caution is warranted here; the ghost/shadow cells in a pilomatricoma may mimic necrosis on scanning magnification.
|Figure 1: Benign tumor showing circumscription, smooth borders and symmetry. (H and E, ×40 Image courtesy: Dr. Almut Boer)|
Click here to view
|Figure 2: V - shape and vertical orientation in a benign tumor, Trichoblastoma (H and E, ×40)|
Click here to view
|Figure 3: (a) Clefts between the rim of compressed fibrous tissue and surrounding stroma in a benign tumor (Pilomatricoma), (b) Clefting directly between the tumor cells and stroma in a malignant tumor, sebaceous carcinoma. Necrosis is also seen (H and E, ×200)|
Click here to view
There are several other features that proved useful. Benign tumors were well-circumscribed, had smooth borders and were oriented vertically to the skin surface [Figure 2]. Uniform discrete cell nests, often in geometric shapes were commonly seen in benign tumors. Benign lesions shelled out completely upon excision, and this is best appreciated by the surgeon sometimes. It would be good to ask this question specifically if it is not mentioned. Preservation of adnexae was seen in all benign tumors, but also in a few malignant ones.
Ulceration, though touted to be common, was rarely seen in malignancies. But, none of the benign tumors were ulcerated. The only criterion that we did not find helpful was the arrangement of neoplastic cells in singles. They were seen in both benign and malignant tumors. On examination at a higher magnification, we found that the benign tumors displaying this feature were all Syringomas. [Figure 4] The small tubules lacking lumina in Syringomas mimicked single epithelial cells on scanner view. This is an important pitfall that one has to be aware of. Another problem we faced in performing this study was the lack of a 2x objective lens in the microscope. This magnification offers the best silhouette outline and most of us in India do not use it routinely. It will be useful to have at least one microscope with a 2x scanner in a setup that handles these lesions.
|Figure 4: (a) Epithelial cells apparently in single files and scattered individually (H and E, ×40). (b) Higher magnification of the same shows small tubules, consistent with Syringoma. (H and E, ×200)|
Click here to view
In summary, it possible to distinguish benign from malignant skin adnexal tumors based on their silhouettes. Before evaluating cytologic details, pathologists should heed the impression formulated on scanner view. There are always some exceptions and one cannot overly rely on any single feature, but the overall architecture is a good guide to predicting their behaviour.
| References|| |
|1.||Cotton D. Troublesome tumors 1: Adnexal tumors of the skin. J Clin Pathol 1991;44:543-8. |
|2.||LeBoit P, Burg G, Weedon D, Sarasin A. Pathology and genetics of skin tumors. Lyon: IARC Press; 2006. |
|3.||Ackerman AB, Boer A. Histopathologic Diagnosis of Adnexal Epithelial Neoplasms Atlas and text. New York: Ardor Scribendi; 2008. |
|4.||Ackerman AB. Differentiation of benign from malignant neoplasms by silhouette. Am J Dermatopathol 1989;11:297-300. |
|5.||Alsaad KO, Obaidat NA, Ghazarian D. Skin adnexal neoplasms-part 1: An approach to tumors of the pilosebaceous unit. J Clin Pathol 2007;60:129-44. |
|6.||Alsaad KO, Obaidat NA, Ghazarian D. Skin adnexal neoplasms-part 2: An approach to tumors of cutaneous sweat glands. J Clin Pathol 2007;60:145-59. |
What is new?
Architectural features on scanning magnification (silhouettes) are important
clues to differentiate benign from malignant skin adnexal tumors.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]