Study Design, Setting, and Ethics Approval

This study was a retrospective chart review of all Japanese patients with SGC who underwent wide excision with paraffin section control between April 3, 2006 and September 1, 2015 at Aichi Medical University Hospital. This study was approved by the Institutional Review Board (IRB) of Aichi Medical University Hospital (No. 2016-016) and followed the tenets of the Declaration of Helsinki. The IRB granted a waiver of informed consent for this study on the basis of the ethical guidelines for medical and health research involving human subjects established by the Japanese Ministry of Education, Culture, Sports, Science, and Technology and the Ministry of Health, Labour, and Welfare. The waiver was granted because the study was a retrospective chart review, not an interventional study, and because it was difficult to obtain consent from patients who had been treated several years prior to this study. Nevertheless, at the request of the IRB, we published an outline of the study, available for public viewing on the Aichi Medical University Hospital website. This public posting also gave patients the opportunity to decline participation, although none of the patients did so. Personal identifiers were removed from the records prior to data analysis.

Patients with metastatic lesions at the first examination or a follow-up time <6 months were excluded from the study. None of the patients had a history of prior irradiation, retinoblastoma, Muir-Torre syndrome, diuretic use, immunosuppressive treatment, or human immunodeficiency virus infection.

Data Collection

The following patient data were collected: patient’s sex, patient’s age, initial diagnosis at other clinics, topical treatment at other clinics, and interval between appearance of symptoms and referral to our institution. Topical treatments included eye drops, ointment, curettage, and cryosurgery for blepharitis, chalazion, and other benign tumors.

The following clinicopathologic data were collected: tumor location, size, origin, and presentation pattern; T category; presence/absence of pagetoid spread; tumor differentiation; mitosis grade; and growth pattern. SGC originates from the meibomian glands, Zeis glands, lacrimal caruncle, pilosebaceous glands in the eyelid, conjunctiva, and/or multicentric regions. The tumor presents in nodular, diffuse, and pedunculated patterns. The T category was defined using the TNM (tumor–node–metastasis) criteria as outlined by the AJCC 7th edition guidelines. SGC was classified into 3 groups according to the degree of differentiation. Well-differentiated tumors had a nodular pattern with cells at the center of the lobules exhibiting sebaceous differentiation. Moderately differentiated tumors were predominantly composed of anaplastic cells, with only a few areas of highly differentiated sebaceous cells. Poorly differentiated tumors had cells with primarily pleomorphic nuclei, prominent nucleoli, and amphophilic-positive cytoplasm. Mitosis was classified into 3 groups: grade 1, 0–1/high-power field (HPF), magnification ×400; grade 2, 2–5/HPF; grade 3, >5/HPF. The tumor growth patterns included lobular, comedocarcinoma, papillary, and mixed.

Management

Two oculoplastic surgeons (Y.T., H.K.) performed the surgical procedures. If a pathologic diagnosis had not been obtained prior to referral to our institution, an incisional biopsy was performed at the first examination. Immediately after the pathologic diagnosis was obtained, the tumor was excised with 5-mm surgical margins ( Figure 1 , Left and Center). All surgical margins were pathologically confirmed with permanent paraffin section analysis using hematoxylin-eosin stain 1 week after the excision. In patients with pathologic findings in a gray zone, the specimens were stained using epithelial membrane antigen (EMA) (3 patients) and/or adipophilin (5 patients). When a residual tumor was found on the paraffin sections, an additional 5-mm-wide excision of the involved lesion was performed. This procedure was repeated until the margin sections were free of tumor. When a negative result was obtained in the margin sections, eyelid reconstruction was performed using a local flap, skin graft, and/or free tarsal graft ( Figure 1 , Right). None of the patients had undergone map biopsy or adjuvant cryotherapy, chemotherapy, or radiotherapy before the development of a local recurrence/metastasis.

Wide excision of sebaceous gland carcinoma with 5-mm surgical margins and reconstruction after pathologic confirmation of a negative result in the margin sections with permanent paraffin section analysis at 1 week post-excision. (Left) A position 5 mm away from the tumor margin is marked. (Center) The lesion is widely excised along the incision line. (Right) The defect was reconstructed using a Hughes flap and skin graft.

All patients underwent a systemic computed tomographic evaluation at the first examination and at every 6-month follow-up visit for 3 years to check for metastatic lesions.

Review of the Margin Sections With Adipophilin

Adipophilin has recently become well recognized in the ophthalmic field as a useful immunohistochemical stain for SGC. For this study, all surgical margins were reviewed immunohistopathologically using adipophilin in patients with local recurrence/metastasis. The paraffin-embedded margin tissues were sliced and stained with primary antibodies against adipophilin (clone: AP125; Acris Antibodies, Inc, Rockville, Maryland, USA) in accordance with the manufacturer’s guidelines. Adipophilin immunostains were examined by light microscopy.

Statistical Analysis

The patient’s age, interval between the appearance of symptoms and referral to our institution, and tumor size were expressed as means ± standard deviations (SD) and were compared between patients with and without local recurrence/metastasis using the Mann-Whitney U test. Comparisons of the other parameters between patients with and without local recurrence/metastasis were carried out using the Pearson χ ² test or Fisher exact probability test. For 2 × 2 tables, when the number of patients was <5, we used Fisher exact probability test. Logistic regression analysis was carried out to determine risk factors for local recurrence/metastasis of SGC. For the logistic regression analysis, all categorized factors were expressed using a binary system (a dummy variable, 0 and 1). The percentage of patients with local recurrence/metastasis was compared between the surgeons using Fisher exact probability test. All statistical analyses were performed using SPSS version 22 software (IBM Japan, Tokyo, Japan). P < .05 was considered to indicate statistical significance.