DNA copy number alterations correlate with survival of esophageal adenocarcinoma patients

Despite recent advances in surgical and multidisciplinary treatment, prognosis for patients with esophageal adenocarcinoma remains poor, and the low prognostic significance of pTNM staging suggests that additional parameters are needed. To identify genomic abnormalities characteristic of esophageal adenocarcinoma, a panel of 33 samples obtained at surgery from previously untreated patients were analyzed by muliplex ligation-dependent probe amplification technique. We detected frequent gains of 6p, 8q, 13q, 17q, 20q, and losses of 4q, 5q, 15q, and 18q. When DNA copy number changes were correlated to clinicopathological features of patients no association was found between the number of chromosomal aberrations and gender, age, tumor grade or pTNM staging. However, interestingly, a significant correlation between patient survival and total number of chromosomal aberrations was found when esophageal adenocarcinoma cases were stratified according to the median of survival (20 months) (P=0.002) or the median of aberrations (12 aberrations) (P=0.014). Evaluation of the distribution of gains and losses at the level of single chromosomes indicated that gains on chromosomes 5, 6, 8, 11, 20 and losses on chromosomes 1, 3, 5, 11, and 18 were significantly different in the two survival groups. Furthermore, when single gene imbalances were analyzed in further details, we found that besides alterations that involve genes shared by both survival groups, a few genes (KIAA0170, EMS1, ABCC4, F3, and MIF) were altered only in samples from patients with poor survival. Thus, we established a good correlation between the total number of chromosomal alterations and survival, suggesting that the estimation of total imbalances might represent an additional indicator of disease outcome. In addition, the finding of alterations specific for the more aggressive esophageal adenocarcinoma subset might represent promising biomarkers to increase the accuracy of clinical outcome prediction.