Variability of Bad Prognosis in Uveal Melanoma.

Topic: Survival of patients with uveal melanoma classified to have a bad prognosis.

Clinical Relevance: To explore reasons for reported variability in survival of patients with uveal melanoma classified to have a bad prognosis.

Methods: We searched PUBMED, MEDLINE, and EMBASE for studies reporting survival data for uveal melanoma undergoing prognostic testing with chromosome 3 status by fluorescence in situ hybridization (FISH), comparative genomic hybridization (CGH), microsatellite analysis (MSA), multiplex ligation-dependent probe amplification (MLPA), single nucleotide polymorphism (SNP), gene expression profiling (GEP) class, and exon sequencing. Only studies reporting 1-year, 3-year, or 5-year survival were included in the study.

Results: The initial search resulted in 49 studies. Only 12 studies met inclusion criteria. Three studies reported survival data for FISH, 1 study reported survival data for CGH, 1 study reported survival data for MSA, 3 studies reported survival data for MLPA, 3 studies reported survival data for SNP, 3 studies reported survival data for GEP, and 2 studies reported survival data for a combination of tests. No studies reported survival data for exon sequencing. Six studies reported percent free of metastatic death, 2 studies reported metastasis-free survival (MFS), 2 studies reported overall survival (OS), and 2 studies reported probability of metastasis. Metastasis-free survival (5 years) for monosomy 3 by FISH was 40% to 60%, by MLPA was 30% to 40%, by SNP was 72%, and for GEP class 2 was not reported. Overall survival (5 years) for monosomy 3 and disomy 8 tumors by MLPA and GEP class 2 were not comparable (81% and 55%, respectively).

Conclusions: Variability exists in reported survival for uveal melanoma with a bad prognosis. Several factors, including composition of study population (tumor size, exclusion of iris melanoma, duration of median follow-up), method of obtaining tumor sample, type of prognostic test, and use of variable outcome measures, can explain some of the observed differences in survival. Variations in determining the cause of death (metastatic or nonmetastatic) may be the major reason for the observed differences. Standardization of study methods and outcome measures will allow comparison of survival data derived from different prognostic tests.