An incredibly deep coverage of the human genome, increased throughput, and a new analysis pipeline designed for the automatic detection of rare variants in low allele fraction to address tumor heterogeneity. Two years after the release of our Saphyr instrument, Bionano did not come to the AACR annual meeting 2019 empty handed.
And we needed to keep innovating: the rise of precision medicine over the past decades reflects the acknowledgment that cancer is an incredibly complex, adaptive disease that continuously takes advantage of evolutionary mechanisms for its survival. It is now clear that each tumor is unique and evolves in particular ways, so one standard of care does not fit every situation.
The presentation of our 3 posters (link at the end of this post) dissecting our new pipeline for the detection of rare structural variants in the cancer genome and the advantages brought by our 400x coverage were the perfect opportunity for us to educate cancer scientists to how our solutions are applicable to their research and to hand them our new white paper on this topic.
From all the presentations we attended during this exceptional 2019 edition, we chose a few examples illustrating how Bionano can make a real difference:
Dr. David Solit from Memorial Sloan Kettering presented the MSK-IMPACT initiative and examples of the results they recently obtained. MSK-IMPACT stands for integrated mutation profiling of actionable cancer targets and is a shared resource. It aims to identify germline variants that contribute to cancer initiation and to facilitate the development of rational clinical trials based on alterations shared by different types of cancer. Over 30,000 tumors were sequenced for a panel of 468 genes. Gene alterations detected include point mutations, copy number alterations, and structural rearrangements. One example of its application is the retrospective study of over 2000 breast cancer tumors, sampled either before, during or after hormonal therapy. The common mutation in the ESR1 gene was easily identified as a post-treatment event conferring the tumor resistance to hormonal therapy. But interestingly, other mutations in the MAPK signaling (ERBB2, NF1, EGFR) were occurring at an early stage providing an intrinsic resistance toward therapy in some patients, and the results of tumor adaptation explaining the failure of the treatment in others. This assay comes with limitations as only about 25% of tumors could be profiled in this initiative, and the assays only detect alterations in the genes included in the assay design. While gene panel sequencing gives Solit’s team input on the single nucleotide polymorphisms (SNPs) and small structural variations in a fraction of the genome, Bionano optical mapping would provide essential information on all copy number alterations and structural rearrangements > 500bp if it were added to the study as a complementary tool.
Another example comes from Dr. Jessica Zucman-Rossi from Centre de recherche des Cordeliers in Paris. Multistep carcinogenesis of the liver can be initiated by toxins (alcohol, tar, mycotoxins) as well as viruses, involving 40 to 60 genes regulating different pathways such as cell cycle, epigenetic modifiers and oxidative stress. Dr. Zucman-Rossi and her team were able to identify new etiologies from the genes altered in the cancer liver such as HBV frequently inserted in the telomerase gene or Adeno Associated Viruses found in 3% of young patients leading to a specific type of genetic rearrangement. But more interestingly, the prevalence of cancer due to either viruses or toxins varies throughout the world and can be traced through particular mutational signatures (defined in the Catalog of somatic mutations in cancer,COSMIC). While one type of signature is associated with age, alcohol, gender and tobacco, another has to do with mycotoxins found in the food in particular regions of the world. This signature is prone to change over time and space: she reported the case of a patient who developed hepatocellular carcinoma in Africa with clonal mutational signature specific from this region, having this “primary” signature washed out after he moved to Europe to undergo chemotherapy, and replaced with a signature typical from patients treated in Europe. This phenomenon illustrates the constant turnover and renewal of the tumor cell population and its plasticity towards external factors. Identifying virus insertion throughout the genome can be challenging to detect by sequencing as two third of the human genome is repetitive and determining the number of viral genome copies integrated is extremely difficult. As such viral integration can lead to specific types of rearrangements, it would be informative to benefit from the high coverage of Saphyr to identify specific signatures in patients, especially in heterogenous tumors exhibiting such plasticity. A paper from a team at Drexel University previously showed direct labeling of viral integration sites using CRISPR, which could be helpful in tumors as well.
During another session, Alan Ashworth from the University of California San Francisco, discussed about the therapeutic implications of DNA repair defects in cancer. Many tumors harbor mutation in the BRCA genes, impairing their ability to repair DNA through homologous recombination and forcing them to rely on alternative DNA damage response pathways such as base excision repair, mediated by PARP proteins. This deficiency in DNA repair mechanisms is a driver of mutagenesis. He thoroughly described the principle of synthetic lethality where clinicians take advantage of the impediment in the BRCA mediated pathway and inhibit the functional base excision repair pathway through specific PARP inhibitors. Ashworth reported that the follow up of patients relapsing after PARP inhibitor therapy led to the identification of no less than 34 reversion mutations in BRCA2 gene leading to the re-expression of a “zombie form” of BRCA2 gene and the rescuing of the homologous recombination (HR) DNA repair pathway. This illustrates the staggering selection pressure for such event and the ability of cancer cell to select particular clones and adapt. Ashworth added that 10 percent of all have BRCAness(tumors that share molecular features of BRCA-mutant tumors without alteration of the BRCA genes, but of genes involved in the same pathway) potentially extending the utility of PARP inhibitors given these patients are screened properly for their HR deficient phenotype. Since focusing on the mutational status of BRCA genes only is insufficient to identify patients eligible to PARP inhibitors, establishing profiles of structural rearrangements signatures of BRCAness seems the best approach. So Far, array based Comparative Genomic Hybridization (aCGH) and deep sequencing have been used to investigate such signature. These techniques are associated with major drawbacks such as high false positive rates, limited access to low allele fraction and the inability to detect balanced translocations. The structural variation detection solution we developed is the answer to these limitations: our deep coverage gives you access to rare events in heterogenous samples, our direct visualization of long DNA molecules (>250kb) greatly reduces the amount of false positive calls and easily identifies events undetectable by aCGH.
At the intersection of these three examples is the need of a comprehensive characterization of the alterations displayed by heterogenous, constantly evolving tumors cells. “We drink from wells we did not dig, we warm by fires we did not build”. By these words, Norman Sharpless and Douglas Lowy illustrated the painfully slow process by which one generation of cancer scientists build upon the discoveries made by the previous one. But efforts are paying with the development of new drugs along with a deeper understanding of the molecular events driving cancer, leading to exciting developments such as the Pediatric MATCH trial where young patients with solid tumors that are not responding to treatment are assigned to an experimental treatment based on the genetic changes found in their tumors rather than on their type of cancer or cancer site. As we build for the next generation, we like to think that Bionano has something to offer to make this journey faster and safer.
Link to our posters:
PDF: Comprehensive Structural Analysis of Cancer Genomes by Genome Mapping
Ernest Lam, PhD, Bionano Genomics Inc.
Yang Zhang, PhD, Bionano Genomics Inc.
Andy Pang, PhD, Bionano Genomics Inc.