Prediction of outcome in melanoma has proved elusive. Morphologic standards (thickness) and lymph node status remain the best predictors, in spite of extensive efforts to find tissue biomarkers. Recently, the addition of sentinel node biopsy has resulted in less morbid assessment of metastasis. Sentinel node biopsy is negative in 75-90 percent of the cases (thus the procedure was potentially unnecessary) and as many as 25 percent of the negatives will ultimately be upstaged. This data suggests there is a need for a better method of assessing the likelihood of metastasis at the time of primary diagnosis. The efforts to discover molecular methods for assessment of early stage primary disease have been complicated and slowed by the difficulty in obtaining tissue from small tumors. This proposal is a translational effort to directly address these problems by discovery of a small series of biomarkers, that when measured accurately, tightly correlate to metastasis. We propose to discover this marker set using a novel in situ method of quantitative analysis of tissue sections (called AQUA) which has detected relationships between expression and outcome that are not detectable using the traditional pathologist-based method. Specifically, we will use a tissue microarray with over 500 melanomas with long term follow-up as a method to discover the key set of prognostic markers. We have already completed a series of markers and have over 10 that show prognostic value in primary tumors. Once we have constructed and optimized combined set of markers, we will then test them on a newly constructed tissue microarray which represents primary tumors from a recent Yale-based series of cases, all of which have sentinel node biopsy results and recurrence/outcome data in our database. Once we have optimized and validated the marker set on the new array, we will test it on prospectively collected, intervening, whole histologic sections, from cases seen in the Yale Dermatopathology labs. Many of these cases will have sentinel node biopsies within a month or two of the primary diagnosis, providing a near term intermediate endpoint to test the value of our assay. We anticipate this method could provide a future replacement for sentinel node biopsy that is both less morbid and more accurate. ? ? ?
Showing the most recent 10 out of 22 publications