Introduction
Sebaceous neoplasms include benign lesions such as sebaceous hyperplasia, sebaceous adenoma and sebaceoma, and malignant lesions such as sebaceous carcinoma1-3. The latter is a rare tumor, with a prevalence between 6.4% and 9.7% among malignant eyelid tumours4,5. Metastasis occurs in 14-28% of cases with a specific mortality rate of 30%6. Thus, this diagnosis has an impact on the patient’s treatment and prognosis.
Histologically speaking, the identification of sebaceous tumors is based on the architectural and cytological features of the lesion. These features, however, may be somewhat insufficient to make a conclusion on the benign or malignant nature of a neoplasm, especially when biopsy samples are small, incomplete, or from difficult areas such as the eyelid7-12. Immunomarkers that could be useful for such differentiation have been investigated13-16. The aim of this study was to determine p53 and Ki-67 expressions in benign sebaceous lesions and sebaceous carcinomas.
Materials and methods
An analytical, observational, cross-sectional and retrospective study was conducted. Sebaceous hyperplasia, sebaceous adenoma, sebaceoma and sebaceous carcinomas diagnosed between January 1975 and March 2019 in the Dermatopathology Service of the General Hospital of Mexico “Dr. Eduardo Liceaga” were included. Age, gender, topography, and time of evolution were determined, as well as asymmetry, circumscribed lesion, ulceration, nuclear atypia, mass necrosis, individual necrosis, pagetoid spread, squamous metaplasia, lymphovascular invasion, perineural invasion, and mitotic index (number of mitoses in ten high-power fields) of the lesions. The percentage of p53 and Ki-67 expressions was recorded and an expression lower or higher than 10% for p53 and lower or higher than 25% for Ki-67 were used as cutoff points. The data analysis used both the MS Excel and SPSS (Statistical Package for the Social Sciences, IBM SPSS Statistics, V.23.0) software. Descriptive statistics (range, mean, median, standard deviation, and percentages) and inferential statistics (Mann–Whitney U-test, Student’s t-test, Chi-square, or Fisher’s exact test) were performed. p < 0.05 was considered a significant difference.
Results
Thirty-six cases were included: thirteen (36.1%) cases of sebaceous hyperplasia, eight (22.2%) cases of sebaceous adenoma, five (13.9%) cases of sebaceoma, and ten (27.8%) cases of sebaceous carcinoma. The average age of the patients was 57.5 ± 15.66 years, being more frequent in males than in females (52.8% vs. 47.2%). The most frequent topography was the face (91.7%), followed by the scalp (5.6%) and the periocular region (2.8%). The median evolution was 36 months (range: 4-480 months). Comparing benign neoplasms with sebaceous carcinoma, a statistically significant difference in gender was observed, with sebaceous carcinoma being more frequent in males (38.5% vs. 90%; p = 0.008). Likewise, a more rapid evolution was found in sebaceous carcinoma (median=36 months, range: 6-36 months vs. median=36 months, and range: 4-480 months, p = 0.025).
In relation to histopathology, sebaceous carcinoma was often found present with the following features: asymmetry (11.5% vs. 60%, p = 0.006), non-circumscribed lesion (0% vs. 60%, p < 0.0001), ulceration (11. 5% vs. 50%, p = 0.024), nuclear atypia (3.8% vs. 100%, p < 0.0001), individual necrosis (3.8% vs. 50%, p = 0.003), squamous metaplasia (0% vs. 30%, p = 0.017), and higher mitotic index (0 vs. 5.5 mitosis in 10 fields, p = 0.005) (Table 1).
Feature | Benign sebaceous lesions n=26*** | Sebaceous carcinoma n = 10 | p |
---|---|---|---|
Asymmetry | n (%) | n (%) | |
Yes | 3 (11.5) | 6 (60) | 0.006 |
No | 23 (88.5) | 4(40) | |
Circumscribed | |||
Yes | 26 (100) | 4(40) | <0.000 |
No | 0 (0) | 6 (60) | 1 |
Ulceration | |||
Yes | 3 (11.5) | 5 (50) | 0.024 |
No | 23 (88.5) | 5 (50) | |
Nuclear atypia | |||
Yes | 1 (3.8) | 10 (100) | <0.000 |
No | 25 (96.2) | 0 (0) | 1 |
Mass necrosis | |||
Yes | 0 (0) | 1 (10) | 0.278 |
No | 26 (100) | 9 (90) | |
Individual necrosis | |||
Yes | 1 (3.8) | 5 (50) | 0.003 |
No | 25 (96.2) | 5 (50) | |
Pagetoid spread | |||
Yes | 0(0) | 0 (0) | - |
No | 26 (100) | 10 (100) | |
Squamous metaplasia | |||
Yes | 0 (0) | 3 (30) | 0.017 |
No | 26 (100) | 7 (70) | |
Lymphovascular invasion | |||
Yes | 0 (0) | 0 (0) | - |
No | 26 (100) | 10 (100) | |
Perineural invasion | |||
Yes | 0 (0) | 0 (0) | - |
No | 26 (100) | 10 (100) | |
Mitotic index * | 0 (0-4) | 5.5 (0-28) | 0.005 |
* median (range) |
***Sebaceous hyperplasia, sebaceous adenoma and sebaceoma
Sebaceous carcinoma showed a higher percentage of p53 expression compared to benign lesions (median = 10%, range 0-80% vs. median = 0%, range: 0-5%, p = 0.001), and the latter more frequently expressed p53 < 10% (100% vs. 30%, p < 0.0001) (Table 2, Figs. 1 and 2). As for Ki-67 expression, sebaceous carcinoma displayed a higher Ki-67 expression percentage compared to benign lesions (median = 35%, range 5-80% vs. median = 5%, range: 0-30%, p < 0.0001). Accordingly, it was observed that these showed a higher frequency of Ki-67 expression < 25% in comparison with sebaceous carcinoma (30% vs. 96.2%, p = 0.009) (Table 2) (Figs. 1 and 3).
Immunomarker | Benign sebaceous lesions n=26*** | Sebaceous carcinoma n=10 | p |
---|---|---|---|
Percentage of p53* expression | 0 (0-5) | 10 (0-80) | 0.001 |
p53 expression < 10% | |||
yes | 26 (100) | 3 (30) | <0.000 |
no | 0 (0) | 7 (70) | 1 |
Percentage of Ki-67* expression | 5 (0-30) | 35 (5-80) | <0.0001 |
Ki-67 expression < 25% | |||
yes | 25 (96.2) | 3 (30) | 0.009 |
no | 1 (3.8) | 7 (70) | |
* median (range) |
***Sebaceous hyperplasia, sebaceous adenoma and sebaceoma
Discussion
Sebaceous carcinoma was prevalent predominantly in males (90% of cases) and more often extraocular (90% of tumors), in contrast to what is reported by other authors1-4,7,17-19. When compared with benign sebaceous lesions, it showed a shorter-term evolution. From a clinical perspective, this is of particular importance when making a presumptive clinical diagnosis. The histological features that made it possible to differentiate sebaceous carcinomas from benign sebaceous lesions were similar to those reported in the literature19-22.
Cabral et al.8 suggested that immunohistochemistry should be performed in cases of sebaceous lesions with histological features that prevent their classification as benign or malignant, or in superficial biopsies of the eyelid where it is recommendable to avoid a second biopsy due to its significant morbidity. Ki-67 and p53 were the two immunocytochemical markers used for such differentiation.
Tumor suppressor gene tumor protein 53 (TP53) is involved in stopping the cell cycle and activating apoptosis, preventing uncontrolled cell growth. Mutations of this gene result in defective DNA repair and have been linked to the pathogenesis of some malignancies23-34. When this gene is mutated, it encodes a mutated TP53 protein that tends to accumulate in the cell nuclei, making its detection possible by immunohistochemistry23,24,26-29. In this article, we found that sebaceous carcinoma presented a higher percentage of p53 expression compared to benign sebaceous lesions. Using the cutoff point suggested by Cabral et al.8 that a p53 expression higher than 10% is associated to a sebaceous carcinoma diagnosis, we found that 70% of sebaceous carcinoma cases exhibited a p53 expression higher than 10%.
Conversely, MKI67 is a protein that is expressed in the nucleus of all cells in the active phases of the cell cycle. The Ki-67 immunomarker is used to detect this protein. It has been observed that the higher the rate of Ki-67 proliferation, the greater the probability of a neoplasm being malignant35-38. This finding was corroborated by this study and allows us to infer that sebaceous carcinoma is a neoplasm, whose proliferative activity is increased with respect to benign sebaceous lesions. Statistical significance was found using a cutoff point for Ki-67 > 25% in the diagnosis of sebaceous carcinoma4. About 70% of sebaceous carcinoma cases were found to have a Ki-67 expression > 25%.