Festival of Genomics Boston: Tumor heterogeneity and single cell tracks

Festival of Genomics Boston

Among other very interesting tracks on genomics, Festival of Genomics Boston this year included a track dedicated to Tumor Heterogeneity, which is one of the most intriguing aspects of tumor biology. In a separate track, Single Cell Genomics and the technologies and approaches for investigating it were discussed. I followed both tracks, and these are the highlights of some of the talks.


On the Tumor Heterogeneity track, Dr. Kornelia Pollack (Dana Farber Cancer Institute) presented his research applying the concepts of evolutionary biology to assess tumor heterogeneity. Tumor cells are considered unique species within an ecosystem (tumor environment), and they study tumor microenvironment variability and the topological distribution of cells. She presented experimental data from breast cancer in which gene expression levels and copy number variation were used to determine intratumoral cellular variability and to predict the chemotherapy outcome. Furthermore, they created topology maps of the tumors by single cell information and showed that patients with the highest pre-treatment intratumoral diversity have resistance to treatment of breast cancer.

Next, Dr. Charlotte Ng (University Hospital Basel) presented data on tumor heterogeneity during the progression from pre-invasive to invasive disease in breast cancer. By exploring clonal heterogeneity, they showed that the majority of Lobular Carcinoma In Situ (LCIS) were clonally related synchronous Ductal Carcinoma In Situ (DCIS). By building phylogenetic trees of the tumors, they showed that progression from LCIS to Invasive Lobular Carcinoma (ILC), and from DCIS to Invasive Ductal Carcinoma (IDC), may result in subclonal selection as defined by the genomic signatures. Dr. Ng also highlighted that temporal heterogeneity during metastasis formation is higher than spatial heterogeneity. She showed that mutational heterogeneity and copy number heterogeneity are different between primary tumor and metastasis.


On the Single-Cell Genomics Track, Dr. Timour Baslan (Memorial Sloan Kettering Cancer Center) spoke about performing single cell sequencing experiments in formalin-fixed paraffin-embedded (FFPE) tissues. Previous work from his lab was successful in creating single cell genomic profiles from cells extracted from fresh frozen tissues. However the vast majority of archived material is FFPE samples. They developed a method for performing single cell sequencing in FFPE samples. Dr. Baslan discussed the technical improvements of the methods and showed proof of a principle experiment in triple positive breast cancer where they demonstrate that CNV profiles produced by single cells from FFPE tissues using their technology were equivalent to those from fresh tissues.

In an Industry-sponsored presentation, Dr. Kev Livak (Fluidigm) discussed the applications of the C1 device for multi-omics single cell analysis. Currently, single cell sequencing is the new standard for transcriptome analysis, and the C1 device is widely distributed across the research community. Dr Livak highlighted that C1 provides versatility through an Open App program. In addition, the C1 Script Builder enables the creation of user-designed protocols that can be shared at the C1 Script Hub. He described these tools in detail as well as the downstream applications that can be performed after the C1 protocol.

Dr. Mike Schmitt (University of Washington / Twin strand biosciences) talked about the study of subclonal mutations to study how tumors evolve. He highlighted that the current challenge is the high error rate for NGS, either coming from the inaccuracy of the sequencing platform or the amplification error prior to sequencing. To reduce the number of errors, they developed a method called duplex sequencing in which double stranded DNA is tagged with asymmetric primers. Dr. Schmitt showed the application of this method in studying Chronic Myeloid Leukemia, in which they were able to detect low-frequency mutations in refractory patients which could not have been detected using standard NGS approaches.


To close, there was a panel on the future of single-cell genomics. The panelists were Dr. Peter Seems (Columbia University), Dr. Alex Shalek (MIT), Dr. David Kaufman (Merk) Dr. Annett Hann-Windgassen (Stanford University), Dr. Timour Baslan (MSKCC). Through interacting with the audience, the panelists shared their opinion on the outlook of this field. In terms of future challenges, it was agreed that there is a need for smart management and the standardization of a variety of protocols as well as the hundreds of thousands of samples being analyzed. In this regard, robust bioinformatics solutions are needed.

Another point discussed was the need for tools for the macroscopic evaluation of tissue architecture towards the selection of cells from the point of interests with minimally invasive approaches. Moreover, to define cellular trajectories, more precise and sensitive wet lab methods—as well as powerful algorithms to synchronize cells and separate signal from noise—are needed. It was also discussed that the most adequate strategy for single cell molecular profiling (i.e. genomics, transcriptomics, epigenomics, etc.) should be determined according to the experimental question and that methods for integrating information are needed. Given the costs of performing single cell profiling, the panelists favor analyzing a higher number of cells at less coverage and resolution instead of conducting deep profiling in a small number of cells.

Taken altogether, these interesting talks and the panel discussion highlighted the potential and challenges of this burgeoning field. It is clear that tumor heterogeneity can be understood, one cell at the time.

Prepared by Rene Cortese, PhD

Menu Title