What TriCon 2015 Foreshadows for Clinical Genomics

I spent a very eventful week at the Molecular TriCon in downtown San Francisco, and have been pondering the very clear trends that emerged by attending the clinical and NGS focused talks.

  • Cancer gene panels make sense economically and as “massively parallel” tests to inform therapy, but they are bound to get more complex.

  • Liquid biopsies of circulating tumor DNA (ctDNA) have the potential to impact how we conduct clinical trials, build early detection regiments and monitor for recurrence of cancer.

  • Clinical Exomes have excellent diagnosis yields, but whole genome sequencing provides a better exome and is close to price competitive.

  • The only constant is change: we all expect the FDA to say more about Laboratory Developed Tests (LDTs) and sequencing platform innovation to change the landscape for test developers.

Cancer Gene Panels go Mainstream

I quickly lost track of how many groups have implemented Laboratory Developed Tests (LDTs) based on onco gene panels. They are usually only going after the  30-50 genes with known molecular targets and are being reimbursed at profitable rates.

While not every cancer type has the same level of actionable “targets”, the NCCN guidelines have been recently updated with recommendations like “EGFR and ALK testing should be conducted as part of multiplex/next generation sequencing”.

And the pipeline of 500-700 drugs in development that will require a biomarker test to be properly matched to a patient ensures a strong future for onco gene panel tests.

While the “hotspots” detected with the current off-the-shelf gene panels catch a lot of the recurrent biomarkers, they also ignore the class of common copy number and gene fusion events that are medically actionable.

Currently, only advanced proprietary tests such as FoundationOne by Foundation Medicine and PCDx by Paradigm tackle this full suite of variants. The VP of Diagnostic Development at Illumina also talked about their plans to provide a similar test based on a future FDA validated “NextSeqDx” platform.

While there may currently only be a few examples of these expanded variant profiles changing the overall treatment and clinical trial recommendations, that is bound to change and with it, the demand for more complex cancer Dx tests.

I talked to a Life Tech FAS standing beside her poster where she showed a fairly simple protocol to add in primers for common gene fusions to a standard hotspot panel. With some tweaked bioinformatics, you get an integrated call set of small variants as well as gene fusion events.

The in-development Illumina solution will run a RNA-Seq tumor transcriptome along with the DNA detection of small variations, known and novel gene fusions and copy number assessments.

The roadmap for VarSeq is well tuned to these developments, with plans to integrate other classes of variants and expression level information into the assessment and interpretation of individual samples.

The Year of Circulating Tumor Cells

As our CEO Andreas Scherer has already written, the conference was buzzing with excitement over the potential of liquid biopsies to change the diagnostic landscape and break new ground for early detection and recurrence monitoring.

Circulating tumor DNA can be found in the plasma of cancer patients, and provides a very current snapshot.

These cell-free DNA molecules only have a 2 hour half-life in the blood stream and are expelled in 180-200 bp fragments (chopped up by histones) from dying tumor cells.

The informatics need to be tuned to detect the variants in these low proportion conditions, and 3000X depth was cited as a minimal sequencing level to call variants.

Having VarSeq able to standardize and repeat these highly tuned variant filtering workflows for ctDNA can improve the accuracy and turn-around for tests utilizing this technology.

Whole Genomes are a Better Exome

Whole genome sequencing provides variants in and around the exonic coding regions as well as across the vast landscape between genes.

In the clinical diagnostic setting, intergenic regions are currently of little to no utility.

But I did hear a great argument by Liz Worthey (Medical College of Wisconsin and Genomic Healthcare Innovations) that whole genomes make for a better diagnostic test than exomes.

Here are few of the reasons why:

  • Exomes have inconsistent and uneven coverage. An 80x exome will have the majority of its variants called in the “tails” of coverage (less than 10).

  • Exome capture kits poorly capture some exons and genes that have known clinical associations.

  • Whole genome sequencing has consistent coverage and results in fewer false positive variant calls on average (especially when comparing calls between family members).

  • There are known pathogenic intronic variants that disrupt splicing, as well as UTR variants that are “off-target” for exome capture kits.

  • Finally, the price of whole genome is being driven down by the HiSeq X Ten (and recently announced X Five) that has as its sole application whole genome human sequencing.

As another example of this, Dr. Rong Mao from ARUP provided an exome diagnosis clinical case study where a variant 120bp into an intron of the TH gene was the causal pathogenic variant they placed in their clinical report. The variant introduced a novel splice acceptor site which introduced a novel coding sequence containing a stop codon that, in effect, knocked out the gene.

She mentioned how lucky they were this happened to be captured by their exome sequencing!

As I demonstrated in my analysis of the whole genomes of 17 supercentenarians, VarSeq is able to scale from exomes up to whole genomes as diagnostic labs consider the shift in the coming years.

The Sound of Change is a-Twitter

To wrap up, here are few of my tweets from the conference that captured the changing landscape of clinical genomics:

Economics of cancer LDTs:

 

Targeted therapies found for the majority of patients doing genetic testing:

 

The genes driving most cancers and now well profiled, allow for focused gene panel tests:

 

Gene fusions needed as part of test:

 

Circulator tumor DNA (ctDNA) as an early detection assay:

 

Cost of whole genome sequencing approaching whole exome sequencing:

 

Comparing the various sequencing platforms, with cost and quality centric metrics:

 

A different type of diagnosis: pathogen detection may be disrupted by mobile nanopore sequencers:

 

Finally, the slides for my presentation in the appended Genomics in Medicine Symposia:

3 thoughts on “What TriCon 2015 Foreshadows for Clinical Genomics

  1. Christophe Lambert

    If circulating tumor DNA testing requires rapid turnaround due to the short half-life of cell-free DNA molecules, that has important implications for the supply chain of DNA testing. As the utility of these tests becomes higher, it will drive the need for sequencing equipment at every cancer clinic.

    Reply
    1. Gabe Rudy

      Good point Christophe,

      I think there are a lot of pressures to geographically localize testing labs. Keeping things in-house, having local expertise and quick turn-around being the big ones.

      I’m not sure if the ctDNA half-life is in the life blood stream only, or if they degrade after pulled into a DNA capture solution as well. But if ctDNA detection and analysis could become as routine as other blood work, it certainly does mean prolific sequencing in small labs!

      Reply
  2. Jeff Victor

    Liquid biopsy via cell-free DNA provide what tissue biopsy can’t offer – time- and cost-efficient, real-time, and longitudinal monitoring with a more complete landscape on tumor clonal evolution. However, it requires multiple tubes of blood, extremely low yield, and poor quality. CirculoGene has a recent breakthrough cfDNA enrichment technology which requires only a drop of blood for NGS mutation detection. CirculoGene could be the Oncology equivalent of Theranos!! Very exciting time…

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.