PRISMA 2020-compliant systematic review of 73 case reports/series (95 patients) with histopathologically confirmed eyelid metastases, synthesizing clinicopathological features, IHC utility, preoperative misdiagnosis rates, temporal patterns, and oncological impact across all primary tumor sites.
Breast carcinoma was the most common primary (39/95; 41.1%). Two-thirds of evaluable patients were misdiagnosed preoperatively (41/62; 66.1%), most often as pyogenic granuloma. At eyelid biopsy, 68.5% (61/89; 95% CI 58.6–77.0%) had synchronous distant organ metastases. Eyelid biopsy revealed an occult primary in 11.6% and changed oncologic management in 13.7% of patients. BAP1 loss distinguished uveal from cutaneous melanoma in 2 of 3 evaluable cases.
- Publication bias is substantial: rare or unexpected presentations are over-reported, inflating misdiagnosis rates and unusual primary site frequencies. - No true-negative denominators exist, so IHC marker positivity rates cannot be interpreted as sensitivity or specificity. - 28.4% loss-to-follow-up and median follow-up of only 6.5 months make survival estimates unreliable.
Maintain a low threshold for biopsy of any atypical, rapidly enlarging, or treatment-refractory eyelid lesion — especially in patients with a cancer history — because two-thirds are misdiagnosed clinically and over two-thirds harbor synchronous systemic metastases. Use a site-directed IHC panel (ER/PR/GATA3 for breast; TTF-1/NapsinA for lung; CDX2/CK20 for GI; PAX8/CAIX for RCC; BAP1 for melanoma subtyping) to establish the primary site and guide staging.
Explore related topics
