To the Editor: Melanoma arising in the periocular area presents a unique treatment challenge due to anatomic and functional constraints. To date, most information regarding periocular melanoma is gleaned from extrapolated data from other sites of the body, assuming similar biologic behavior. We aimed to identify the overall incidence, pathologic features, and surgical outcomes of periocular melanoma. Following Institutional Review Board (IRB) approval, using the Rochester Epidemiology Project (a multicenter medical records–linkage system encompassing the Olmsted County population), we identified patients with a first lifetime diagnosis of periocular melanoma between January 1, 1970, and December 31, 2020, while a local resident of Olmsted County, Minnesota. Outside (nonresident) referral cases were excluded. Periocular melanoma was defined as a melanoma arising within the orbital rim – on the upper or lower eyelid, lateral canthus, or eyebrow. A total of 10 patients with primary periocular melanoma were identified. The overall incidence rate of periocular melanoma was 0.2686 per 100,000 patient-years. Among all melanomas diagnosed during that timeframe in Olmsted County, 0.43% were periocular in location. No cases extended onto conjunctiva. The histogenic subtype was lentigo maligna in 8 (80.0%). There was 1 superficial spreading melanoma and 1 case without documentation of histogenic subtype. Incidence of in situ (50.0%) and invasive (50.0%) melanomas was equal. Among invasive melanomas, the average Breslow depth was 0.90 mm (range, 0.2-2.6 mm). Seven (70.0%) were treated with Mohs micrographic surgery (MMS). One (10%) was treated with wide local excision. The 2 cases that did not undergo further surgical treatment were both diagnosed prior to 2000 and were removed entirely via initial excisional biopsy. There was 1 regional metastasis to the parotid gland and lymph node, 2.5 years after diagnosis of a superficial spreading melanoma, Breslow 0.7 mm, status-post treatment with MMS. There were no other local or distant recurrences. There were no melanoma-specific deaths. Mean follow-up time was 7.4 years (range 3.1-13.2). This study demonstrates the relative rarity of periocular melanoma, comprising only 0.43% of all melanomas. Comparatively, basal cell carcinoma (BCC) occurs on the eyelid 16% of the time.1Salomon J. Bieniek A. Baran E. Szepietowski J.C. Basal cell carcinoma on the eyelids: own experience.Dermatol Surg. 2004; 30: 257-263PubMed Google Scholar Our study found an overall periocular melanoma incidence rate of 0.2686 per 100,000 patient-years. A previous Olmsted County cohort reported an incidence of eyelid BCC as 14.35 per 10,000 individuals per year, and an incidence of eyelid SCC as 1.37 cases per 100,000 individuals per year.2Cook Jr., B.E. Bartley G.B. Epidemiologic characteristics and clinical course of patients with malignant eyelid tumors in an incidence cohort in Olmsted County, Minnesota.Ophthalmology. 1999; 106: 746-750Abstract Full Text Full Text PDF PubMed Google Scholar The behavior of melanoma seems to parallel primary cutaneous melanoma of head and neck sites. Most tumors were associated with chronic sun exposure; 80% of cases were lentigo maligna subtype. There was 1 regional recurrence in our study, no distant recurrences, and no deaths due to melanoma. The most significant advantage of our study is in its epidemiologic design. The limitations of this study include the retrospective nature of our analysis and the homogenous demographics of Olmsted County, comprised of mostly Caucasian individuals of higher socioeconomic status.3St Sauver J.L. Grossardt B.R. Leibson C.L. Yawn B.P. Melton 3rd, L.J. Rocca W.A. Generalizability of epidemiological findings and public health decisions: an illustration from the Rochester Epidemiology Project.Mayo Clin Proc. 2012; 87: 151-160Abstract Full Text Full Text PDF PubMed Scopus (487) Google Scholar Additionally, some melanomas may have been missed if the tumor was initially labeled as another anatomic site (cheek, nose, temple, etc.) and had extension onto the lid proper. None disclosed. We would like to acknowledge Austin Todd, MS, for his assistance in statistical analysis.
Read full abstract