Scientists at Silicon Biosystems (Bologna, Italy, and San Diego, CA) demonstrated that the company's DEPArray cell sorting allows downstream analysis of tumor genetic characteristics via next-generation sequencing (NGS) with unprecedented precision. Their study involving the method was able to isolate 100% pure tumor and stromal cell populations from minute formalin-fixed, paraffin-embedded (FFPE) specimens.
Related: Next-generation sequencing helps shed light on cancer metastasis, tumor evolution
The application of precision medicine in oncology requires a rigorous understanding of the genetic characteristics that drive tumorigenesis. Available technologies, like NGS, hold incredible potential, but their accuracy is limited by the fact that biopsied tissue specimens represent a mixture of different cell types present in different proportion.
Moreover, heterogeneity inside the tumor cell population itself limits the possibility to identify tumor drivers, as low represented clones may actually be the one responsible for more malignant traits. Sample heterogeneity can thus affect the performance of powerful downstream molecular analysis by hiding genetic variants because of the dilution of samples.
Current available technology to solve samples heterogeneity—like laser capture microdissection and FACS sorting—still lack the accuracy and purity required for clinical implementation, and their power is often limited by the size and quality of the starting sample materials.
Starting from a wide collections of FFPE clinical samples of different size and with different tumor cellularity, the researchers used the DEPArray cell sorting system to digitally separate precise numbers of pure, homogeneous cells based on their marker phenotype and DNA content. Distinct tumor populations with different DNA content (diploid and hyperdiploid fractions) could be separated in purity from the diploid stromal fraction as demonstrated by results of downstream targeted NGS.
When comparing variant frequencies between pure-DEPArray sorted populations, the researchers could assign—unambiguously and quantitatively—the genetic variants to the different classes of cells (stromal population, diploid, or hyperdiploid tumor populations). On the contrary, targeted NGS results from the unsorted fraction gave a blurred picture of the genetic profile of the tumor, and didn't allow detection of somatic mutations or loss of heterozigosity (LOH).
The researchers also confirmed the results by low-pass whole genome sequencing (WGS). The copy-number profiles of sorted pure populations confirmed the copy number variations (CNVs) and LOH events predicted with targeted NGS and further allowed to interpret dual events on the same locus, like mutation and LOH.
The power of the DEPArray cell sorting was further demonstrated using FFPE specimens with low tumor cellularity (5%), which are usually inaccessible to accurate genetic analysis and are associated with poor patients’ prognosis. The low-pass WGS profile of sorted pure tumor cells from these samples allowed identification of numerous gain and losses along the genome in contrast to the analysis of the unsorted sample, where the signal from tumor cells was so diluted by the contaminating stromal DNA that most of the gains and losses were undetectable.
The researchers' study shows that the DEPArray cell sorting technology, followed by NGS analysis, reveal comprehensive genomic information from any FFPE sample, regardless of tumor cellularity and size of the specimen, and has the potential to revolutionize translational cancer research, including biomarker discovery, and setting a new gold standard for accuracy in oncology precision medicine.
Full details of the work appear in the journal Scientific Reports; for more information, please visit http://dx.doi.org/10.1038/srep20944.