Cytogenetics of uveal melanoma: a 7-year clinical experience.

PURPOSE: Chromosome 3 loss and chromosome 8 gains in uveal melanoma are associated with metastatic death. Since 1999, we have offered cytogenetic analysis to patients treated by local resection or enucleation. This study correlated our cytogenetic results with clinical and histologic predictors and disease-specific mortality. DESIGN: Nonrandomized case series. PARTICIPANTS: Three hundred fifty-six patients with uveal melanoma with data on chromosome 3 and chromosome 8. METHODS: Tumor diameter was measured by echography. Cell type, presence of closed connective tissue loops, and mitotic rate were determined histopathologically. Fluorescence in-situ hybridization was performed using centromeric probes for chromosomes 3 and 8 and for c-myc. Patients were flagged at the National Health Service Cancer Registry, which notified us of any deaths. Statistics included Cox multivariate analysis and Kaplan-Meier analysis. MAIN OUTCOME MEASURES: Disease-specific mortality, according to clinical, histologic and cytogenetic features as well as correlation between cytogenetic variables and other mortality predictors, including a predictive index. RESULTS: The patients had a mean age of 61.9 years. The tumors showed no cytogenetic abnormalities of chromosomes 3 or 8 in 42%, chromosome 8 gains in 11%, monosomy 3 in 21%, and both abnormalities in 27%. These correlated with ciliary body involvement (P<0.001), extraocular spread (P = 0.007), large basal tumor diameter (P<0.001), epithelioid cellularity (P<0.001), closed connective tissue loops (P<0.001), and mitotic rate exceeding 4/40 high power fields (P<0.001). By the study close, 76 patients had died (67 from metastasis). Cox multivariate analysis showed the most significant factors to be basal tumor diameter (P<0.001), monosomy 3 (P<0.001), and epithelioid cellularity (P = 0.004). A predictive index (PI) was derived from these variables. Kaplan-Meier analysis showed that 5-year metastatic death rates ranged from 0% in 84 patients with low-grade melanoma (PI<19) to 66% in 100 patients with high-grade tumor (PI>26; 95% confidence interval, 53%-80%). CONCLUSION: Cytogenetic analysis of chromosomes 3 and 8 enhances prediction of disease-specific mortality after treatment of uveal melanoma but must be interpreted together with tumor diameter and cell type.