Prognostic factors of liver metastases from uveal melanoma.

OBJECTIVES: This study was designed to assess survival and identify prognostic factors for liver metastases diagnosed by systematic screening in uveal melanoma patients.
METHODS: Among 602 consecutive patients treated over 10 years for uveal melanoma and followed by systematic semi-annual hepatic screening (abdominal ultrasonography), 63 (10.5%) developed liver metastases; these patients form the basis of this study. Factors including patient demographics, characteristics of the uveal tumor, metastasis-free interval, severity of liver metastatic involvement, and treatments of metastases were studied retrospectively regarding their prognostic value, using univariate (Kaplan-Meier method) and multivariate (Cox model) analyses.
RESULTS: Thirty-five patients (55.6% of the metastatic population) received systemic chemotherapy or best supportive care only; 14 patients (22.2% of the metastatic population) diagnosed with diffuse liver involvement had cytoreductive surgery and intra-arterial chemotherapy; 14 (22.2% of the metastatic population) had complete surgical removal of liver metastases followed by postoperative intra-arterial chemotherapy. No significant surgical complications were experienced. The median overall survival after diagnosis of liver metastases was 15 months. It reached 25 months for selected patients with complete resection (P=0.0002). In this cohort of 63 patients, ten or fewer preoperatively diagnosed metastases and primary uveal melanoma not involving the ciliary body were independently associated with better prognosis.
CONCLUSIONS: This study suggests that selected patients with screened liver metastases from uveal melanoma may benefit from aggressive treatment, including surgery. The two independent favorable prognostic factors are fewer than ten metastases at screening and the absence of ciliary body involvement.