Biopsy for molecular risk stratification in uveal melanoma: yields and molecular characteristics in 119 patients

Background: Prognostic cytological and molecular features of uveal melanoma have been well researched and are essential in management. Samples can be obtained in vivo through fine needle aspirate biopsy, vitrector cutter, forceps or post-enucleation for off-site testing. This study aims to examine cytological and chromosome microarray yields of these samples. Methods: A retrospective cohort analysis of 119 uveal melanoma biopsies submitted to our laboratory. Samples included those taken in vivo (n = 57) and post-enucleation (n = 62). Patient and tumour features were collected including age, sex, primary tumour location, basal diameter and tumour height. Prognostic outcomes measured include cell morphology, chromosomal status and immunohistochemistry. Results: Post-enucleation biopsies accounted for just over half of our samples (52%). Post-enucleation samples had a more successful genetic yield than in vivo biopsies (77% vs. 50%, p = 0.04) though there was no difference for cytological yields. There was no difference in cytological or microarray yields between instruments. The vitrector biopsy group had the smallest tumour thickness (5 mm vs. 10 mm [fine-needle aspirate biopsy], p = 0.003). There was a strong correlation between monosomy 3, BAP1 aberrancy and epithelioid cell type in post-enucleation samples (T_b = 0.742, p = 0.005). However, epithelioid morphology was not associated with either monosomy 3 (p = 0.07) or BAP1 aberrancy (p = 0.24) for in vivo biopsies. Conclusions: All three biopsy instruments provide similar cytological yields as post-enucleation sampling, although post-enucleation samples had a more successful chromosome microarray yield. Epithelioid cytomorphology alone is insufficient for prognostication in in vivo biopsies, immunohistochemistry would be a useful surrogate test.