Screening for metastatic malignant melanoma of the uvea revisited.

Abstract

The purpose of the current study was to assess the value of routine imaging and liver function tests in detecting metastases from malignant melanoma of the uvea. Forty-six consecutive patients diagnosed with metastatic uveal melanoma between 1985 and 1996 who had participated in a screening program that included annual liver function tests (LFT), chest X-ray, and abdominal ultrasonography (US) were eligible for this retrospective cohort study. Main outcome measures were the sensitivity of screening tests, presence of symptoms, recurrence free interval, and metastatic burden. Metastases were diagnosed in 74% of patients (95% confidence interval [95% CI], 59-86) at screening and in 26% (95% CI, 14-41) when the patient consulted a physician because of symptoms that developed before the next scheduled visit. Of all the patients, 59% (95% CI, 43-73) were asymptomatic, and 80% (95% CI, 66-91) had only hepatic metastases. The median recurrence free interval, greatest dimension of the largest metastasis, and metastatic burden of the two groups did not differ. US was diagnostic in 78% (95% CI, 64-89), at least 1 LFT test was abnormal in 70% of patients (95% CI, 54-82), and a chest X-ray was abnormal in 2% of patients (95% CI, 0-12). LFTs and US did not reveal hepatic metastases in 33% and 4% of patients, respectively. The sensitivity of individual LFTs ranged from 0.27 to 0.67, and their specificity from 0.90 to 0.96, with lactate dehydrogenase being the most sensitive LFT used. The authors believe that annual screening with LFTs and abdominal US will identify 59% of patients while they are still asymptomatic and that semiannual screening will detect >95% of such patients. Chest X-ray has a very low yield and is recommended only at baseline to exclude metastatic disease to the eye and if pulmonary symptoms develop.