Radiologic and Histopathologic Correlation of Different Growth Patterns of Metastatic Uveal Melanoma to the Liver

• Authors: Albert Liao, Emory University; Pardeep Mittal, Emory University; David H Lawson, Emory University; Jenny J. Yang, Georgia State University; Eszter Szalai, Emory University; Hans Grossniklaus, Emory University
• Journal Title: Ophthalmology (Section 12 EMBASE)
• Volume: Volume 125, Number 4
• Publisher: Elsevier | 2018-04-01, Pages 597-605
• Type of Work: Article | Post-print: After Peer Review
• Publisher DOI: http://doi.org/10.1016/j.ophtha.2017.09.029
• Subjects: Health Sciences, Opthamology; Health Sciences, Radiology; Health Sciences, Oncology
• Keywords: Science & Technology; Life Sciences & Biomedicine; Ophthalmology; OCULAR MELANOMA; HEPATIC METASTASES; SURVIVAL; EXPRESSION; CHEMOEMBOLIZATION; PROGRESSION; MANAGEMENT; PROGNOSIS; DISEASE; MODEL

Abstract

Purpose: The purpose of this study was to correlate magnetic resonance imaging (MRI) radiographic results with histopathologic growth patterns of metastatic uveal melanoma (UM) to the liver. Design: Clinicopathologic correlation. Participants: Patients with metastatic UM to the liver. Methods: A retrospective review of MRI images of patients with metastatic UM to the liver at a single institution between 2004 and 2016 was performed. The MRI growth patterns were classified as nodular or diffuse. The histopathologic findings of core liver biopsies of liver metastases identified by needle localization in a subset of these patients were reviewed. The core samples were evaluated by routine light microscopy, including immunohistochemical/immunofluorescent staining for CD31, CD105, and HMB45, and classified as exhibiting an infiltrative or nodular growth pattern. Main Outcome Measures: Magnetic resonance images and core biopsy findings. Results: A total of 32 patients were identified with metastatic UM to the liver that was imaged by MRI, and 127 lesions were identified. A total of 46 lesions were classified by MRI as infiltrative and 81 as nodular. There were 9 needle-localized core biopsies that corresponded to MRI of metastatic lesions. Of these 9 lesions, 3 that were classified as infiltrative on MRI exhibited stage I infiltrative histologic growth patterns; of the remaining 6 that were classified as nodular by MRI, 5 histologically demonstrated stage II or stage III infiltrative growth patterns and 1 histologically demonstrated a nodular growth pattern. Conclusions: Magnetic resonance imaging of hepatic infiltrative growth patterns of metastatic UM corresponded to stage I histologic infiltrative growth in the sinusoidal spaces, whereas MRI nodular growth patterns corresponded to stage II/III histologic infiltrative growth that replaced the hepatic lobule or histologic nodular growth in the portal triad that effaced adjacent hepatic parenchyma.