The aim of this study was to develop a highly sensitive two-marker assay fo
r the detection of circulating melanoma cells in patients' blood using a re
verse transcriptase-polymerase chain reaction (RT-PCR). We analysed the use
fulness of two different sets of markers: tyrosinase and MUC-18 (TYR/MUC-18
), and tyrosinase and MART 1 (TYR/ MART 1). Total cellular RNA was isolated
from 337 blood samples from 80 melanoma patients at different stages of th
e disease. All patients had undergone primary surgery. Assay sensitivity an
d specificity were confirmed using three different melanoma cell lines and
two different fibroblast lines. In addition, blood from 47 healthy subjects
and 10 patients with non-melanoma cancer was used as a negative control. W
e found that two-marker analysis is more accurate than the single tyrosinas
e assay. The frequency of melanoma cell detection in patients' blood was ab
out 10% higher when the TYR/MART 1 two-marker assay was used. Using this as
say we did not find any statistical correlation between the molecular marke
rs and the UICC stage of disease or the Breslow thickness or Clark level of
the primary melanoma. The frequency of melanoma cell detection with the TY
R/MUC-18 two-marker assay was even higher than the TYR/MART 1 assay, but un
fortunately the MUC-18 transcript was also present in about 20% of healthy
subjects. Therefore we do not recommend the use of MUC-18 as a standard val
ue marker. (C) 2001 Lippincott Williams & Wilkins.