5044

The prognostic value of sentinel lymph node micrometastasis as defined by histology and/or immunohistochemistry is clearly demonstrated for melanoma patients. However, around 20% of patients with tumor negative lymph nodes have recurrent disease within 10 years.The prognostic value of the detection of mRNA coding for differentiation or tumor markers is still debated. Furthermore, to our knowledge, no data is available so far regarding angiogenesis markers.

From July 1999 to July 2002 all patients referred to our department for primary melanoma Breslow>1mm, and/or Clark level IV or V and /or ulcerated underwent the sentinel lymph node procedure. During this period, lymph nodes were bivalved, one half processed for histological examination, the other half was kept frozen for molecular analysis. Using quantitative RT-PCR we analysed the expression of 2 melanocytic differentiation antigens (tyrosinase and MART1) and the following genes involved in angiogenesis and invasion : uPA, PAI1, EMMPRIN; VEGF, VRGFR2 as well as genes involved in lymphangiogenesis, VEGFC, VEGFR3, LYVE and PROX1 and looked for a correlation with the presence of micrometastasis, disease free survival and overall survival.

92 patients were included with a median of 1.3 node per patients (1-3), median Breslow of 2.2mm [0.9 to 19], 29% ulcerated and a median follow-up of 41 months. Micrometastasis were present in 15% of patients. The expression of tyr (p<0.0001) , mart (p<0.0001), VEGF (p 0.0070) and PAI1 (p 0.024) were significantly associated with the presence of a micrometastatic lymph node (Wilcoxon rank-sum test). When all markers and sentinel lymph node histology were joined in a univariate analysis, histology, Tyr and Mart were significantly associated with DFS and OS . Neither Tyr nor mart remained independent factors when they were adjusted to histology.

Most series in the literature reported results using nested PCR for differentiation or tumor antigens. Such techniques failed to demonstrate any benefit as compared to histological assessment because of poor specificity. Our data are in accordance with 2 recent series using quantitative PCR for Tyr Mart or other melanocytic antigens. A significant association with survival was reported in both studies as in ours, independently from histology in one series. However further data are needed to evaluate the benefit of such technique for routine diagnosis. On the other hand we show for the first time a significant association between the presence of a micrometastasis and the expression of VEGF and PAI1. Although our study was probably underpowered and failed to demonstrate a link with tumor progression, the expression of these markers provides a rationale for testing antiangiogenesis agents in stage IIIA melanoma patients.

98th AACR Annual Meeting-- Apr 14-18, 2007; Los Angeles, CA