Cytologic Grading of Cutaneous Sebaceous Neoplasms: Does it Help to Differentiate Benign From Malignant?

Abstract

Both architectural and cytologic characteristics are used to distinguish benign from malignant sebaceous neoplasms; however, specific cytopathologic features of sebocytes have not been well defined. The authors assessed architectural and cytological features of 63 sebaceous neoplasms [15 sebaceous hyperplasias, 12 sebaceomas, 16 sebaceous adenomas (SA), 14 sebaceous carcinomas (SC), and 6 ocular sebaceous carcinoma (OSC)] to investigate whether cytological grading may facilitate classification of lesions. Among other criteria, nuclear pleomorphism (size, nucleolar appearance, membrane irregularity, crowding, mitoses, and chromatin pattern) was assessed and 3 theoretical nuclear grades established. Immunohistochemistry for CK10, p16, adipophilin and ki67 was performed on 7 cases of each type of tumor. Most sebaceous neoplasms, except OSC, showed a bland architectural silhouette. However, SA, SC, and OSC revealed larger nuclei (≥14 μm in ≥50% of cases), evident to multiple nucleoli, membrane irregularity, coarse to clumped chromatin, and nuclear grade ≥2 (latter in ≥56% of cases); by contrast, sebaceous hyperplasia and sebaceomas showed smaller nuclei (≤10 μm in ≥50% of cases), smooth borders, inconspicuous nucleoli, fine chromatin, and grade 1 nuclei (latter in 100% of cases). In the setting of a well-circumscribed architecture, cytologic features gain importance in the classification of sebaceous neoplasms. Interestingly, cytologic similarities found in SA and SC may indicate a close relationship of both neoplasms.

Machine learning for classification of cutaneous sebaceous neoplasms: implementing decision tree model using cytological and architectural features

“…Our suggested decision tree model is highly consistent with a recent study that implemented this method to classify sebaceous neoplasms. Nevertheless, we cross-validated our model on five different folds of the dataset to reduce overfitting and mitigate the effects of chance associated with fitting on a single random data splitting [ 28 ]. Details of our most accurate model, with an accuracy of 83%, are shown in Fig.…”

Tumores con diferenciación sebácea. Comunicación de tres casos y revisión de la literatura

“…2 En el presente se consideran tres grupos de tumores con diferenciación sebácea: adenoma sebáceo, sebaceoma y carcinoma sebáceo. 3,4 La hiperplasia sebácea y el nevo sebáceo no se consideran como verdaderas entidades tumorales sebáceas. La diferenciación sebácea también se observa en forma secundaria como fenómeno poco frecuente durante el desarrollo de otros tumores cutáneos con diferenciación folicular y apocrina.…”