The outcome of ACTs can be determined by gene expression level at the RNA level. However RNA handling is challenging. In contrast tumor DNA is robust and therefore easier to use.
Aim: To characterize the ACTs DNA alterations; to identify markers with diagnostic and prognostic value using tumor DNA.
Methods: The mapping of chromosomal gains and losses of 60 ACTs (39 adenomas (ACAs), 21 carcinomas (ACCs)) was performed with CGH arrays (4452 probes BAC arrays). DNA copy number determination at specific loci was performed with quantitative PCR (qPCR), in the same tumors (training cohort), and 79 independent ACTs (49 ACAs, 30 ACTs, validation cohort). Patient informed consents and local ethic committee approval were obtained. Statistics were performed in R (www.R-project.org).
Results: The number of chromosomal gains and losses measured by CGH differs between ACAs and ACCs (10 vs 44%, P=2−10). Common gains in ACCs include chromosome 5, 7, 12, 16, 19 and 20, and common losses include chromosome 13 and 22. Using qPCR, the best discrimination between ACAs and ACCs was obtained by substracting the DNA copy numbers at 6 loci (5q, 7p, 16q and 11p, 13q, 22q) in the training cohort (sensitivity 100%; specificity 100%). This result was confirmed in the validation cohort (sensitivity 100%; specificity 80%).
In ACCs, interestingly, the number of chromosomal gains and losses was not associated with survival (Cox P=0.84). However comparison of bad and good prognosis ACCs showed specific differences (good prognosis: gain in 5q; bad prognosis: gain in 1q and losses in 10p and 17q). Using qPCR, the substraction of DNA copy numbers at 1q17q showed the strongest association with survival in the training cohort (Cox P=5.10−4), and was confirmed in the validation cohort (P=0.02).
Conclusion: Tumor DNA can be used for the diagnosis of malignancy and the prognosis of ACTs.
Prague, Czech Republic
24 - 28 Apr 2010
European Society of Endocrinology