Copy number alterations in small intestinal neuroendocrine tumors determined by array comparative genomic hybridization

Abstract

Background: Small intestinal neuroendocrine tumors (SI-NETs) are typically slow-growing tumors that have metastasized already at the time of diagnosis. The purpose of the present study was to further refine and define regions of recurrent copy number (CN) alterations (CNA) in SI-NETs.

Methods: Genome-wide CNAs was determined by applying array CGH (a-CGH) on SI-NETs including 18 primary tumors and 12 metastases. Quantitative PCR analysis (qPCR) was used to confirm CNAs detected by a-CGH as well as to detect CNAs in an extended panel of SI-NETs. Unsupervised hierarchical clustering was used to detect tumor groups with similar patterns of chromosomal alterations based on recurrent regions of CN loss or gain. The log rank test was used to calculate overall survival. Mann-Whitney U test or Fisher's exact test were used to evaluate associations between tumor groups and recurrent CNAs or clinical parameters.

Results: The most frequent abnormality was loss of chromosome 18 observed in 70% of the cases. CN losses were also frequently found of chromosomes 11 (23%), 16 (20%), and 9 (20%), with regions of recurrent CN loss identified in 11q23.1-qter, 16q12.2-qter, 9pter-p13.2 and 9p13.1-11.2. Gains were most frequently detected in chromosomes 14 (43%), 20 (37%), 4 (27%), and 5 (23%) with recurrent regions of CN gain located to 14q11.2, 14q32.2-32.31, 20pter-p11.21, 20q11.1-11.21, 20q12-qter, 4 and 5. qPCR analysis confirmed most CNAs detected by a-CGH as well as revealed CNAs in an extended panel of SI-NETs. Unsupervised hierarchical clustering of recurrent regions of CNAs revealed two separate tumor groups and 5 chromosomal clusters. Loss of chromosomes 18, 16 and 11 and gain of chromosome 20 were found in both tumor groups. Tumor group II was enriched for alterations in chromosome cluster-d, including gain of chromosomes 4, 5, 7, 14 and gain of 20 in chromosome cluster-b. Gain in 20pter-p11.21 was associated with short survival. Statistically significant differences were observed between primary tumors and metastases for loss of 16q and gain of 7.

Conclusion: Our results revealed recurrent CNAs in several candidate regions with a potential role in SI-NET development. Distinct genetic alterations and pathways are involved in tumorigenesis of SI-NETs.

Figure 1

Mapping of CN losses detected in chromosome 18 by a-CGH analysis. At the top, the location of the EMILIN2, DCC, BCL2, CDH19 genes analyzed by qPCR are indicated by arrows next to an ideogram of chromosome 18 (UCSC Genome Browser). Alterations in cases with detected CN losses are schematically illustrated by bars with red indicating losses and green marking gains. Fifteen samples exhibited loss of the entire chromosome 18, while samples 7, 11, 16, 17, 27P and 31 showed losses restricted to parts of the chromosome. Below are shown a-CGH profiles for case 4 with entire chromosome 18 loss, case 27P with deletion of 18pter-11.21, and case 16 with a 2 Mb deletion at 18q22.1.

Figure 2

Mapping of CN losses in (A) chromosomes 16 and (B) 11 (B) and (C) localization of CN gains in chromosome 14 by a-CGH. For each chromosome is shown an ideogram (UCSC Genome Browser) together with bars indicating losses in green and gains in green. The genomic location of the CDH1 and SDHD genes analyzed by qPCR are marked by arrows.

Figure 3

Unsupervised hierarchical clustering and overall survival for recurrent regions of CNAs in SI-NETs. (A) Clustering analyses identified two tumor groups (I and II) as well as five chromosomal clusters (a-e). CNAs detected in the primary tumor and metastasis of case 27 were pooled to prevent magnification of alterations in multiple samples from the same case. (B) Kaplan-Meier plots showing shorter overall survival for cases with gain in 20pter-p11.21. The table below indicates the number of subjects at risk at different time points during follow-up.