Next-generation sequencing of disseminated tumor cells

Abstract

Disseminated tumor cells (DTCs) detected in the bone marrow have been shown as an independent prognostic factor for women with breast cancer. However, the mechanisms behind the tumor cell dissemination are still unclear and more detailed knowledge is needed to fully understand why some cells remain dormant and others metastasize. Sequencing of single cells has opened for the possibility to dissect the genetic content of subclones of a primary tumor, as well as DTCs. Previous studies of genetic changes in DTCs have employed single-cell array comparative genomic hybridization which provides information about larger aberrations. To date, next-generation sequencing provides the possibility to discover new, smaller, and copy neutral genetic changes. In this study, we performed whole-genome amplification and subsequently next-generation sequencing to analyze DTCs from two breast cancer patients. We compared copy-number profiles of the DTCs and the corresponding primary tumor generated from sequencing and SNP-comparative genomic hybridization (CGH) data, respectively. While one tumor revealed mostly whole-arm gains and losses, the other had more complex alterations, as well as subclonal amplification and deletions. Whole-arm gains or losses in the primary tumor were in general also observed in the corresponding DTC. Both primary tumors showed amplification of chromosome 1q and deletion of parts of chromosome 16q, which was recaptured in the corresponding DTCs. Interestingly, clear differences were also observed, indicating that the DTC underwent further evolution at the copy-number level. This study provides a proof-of-principle for sequencing of DTCs and correlation with primary copy-number profiles. The analyses allow insight into tumor cell dissemination and show ongoing copy-number evolution in DTCs compared to the primary tumors.

Keywords: circulating tumor cells; clonal evolution; disseminating tumor cells; single tumor cell sequencing; tumor heterogeneity.

Figure 1

Flow chart describing the workflow from collection and detection of the DTC, to library preparation and data analyses.

Figure 2

Comparison of copy-number data of DTCs generated from whole-genome sequencing (middle side) and aCGH (right side), with the primary tumor as a control (left side). The piecewise constant fitting algorithm was used to generate segments with gamma = 25 and k_min = 5 for the sequencing and primary tumor data, and gamma = 60 and k_min = 25 for the aCGH data. (A) DTC003: gain of chromosome 1q was observed with both technologies, where a focal deletion on 1p was only detected in the DTC sequencing data, and not recapitulated in either the DTC aCGH or primary tumor profile. (B) DTC083: copy-number alteration on chromosome 1 shows that aCGH data do not recapitulate the smaller deletion on the 1p loci, however both technologies capture the 1q whole-arm amplification. A false aberration is found in the aCGH data (marked with a black arrow).

Figure 3

Copy-number changes were analyzed from sequencing data from the DTC003 and SNP-CGH data from the corresponding primary tumor, MicMa003. (A) Comparison of aberration pattern in the DTC and primary depicted by Circos plots. BAF reveals allelic loss and copy-number alterations of the DTC003 showed ~99% concordance with its primary tumor. (B) SNP-CGH profile of primary tumor shows few subclonal alterations. Clonal amplifications and deletions were observed.

Figure 4

Copy number changes were analyzed from sequencing data from the DTC083 and SNP-CGH data from the corresponding primary tumor, MicMa083. (A) Circos plots reveal aberration pattern differences between the DTC and primary tumor, with ~89% concordance in the copy number changes. BAF show cnLOH on chromosome 13 in the DTC where the primary tumor had a subclonal deletion. Amplification of chromosome 21 seems to be novel in the DTC. (B) ASCAT profile of primary tumor describes several subclonal alterations, such as amplification on chromosome 1q and 16p, and deletion of chromosome 13.

Figure 5

A proposed and animated model of how the tumor evolution has occurred in MicMa083 at the copy-number level. Aberrations change from clonal to subclonal, where the DTC could descend from one of the primary subclones.