Wednesday, June 21, 2017

Brief Blog: A Prototype View of the Digital Genomics Industry

Over the past year, there is increasing consideration of a "digital genomics" industry, which I've sketched in a prototype map below.


Full discussion after the break.


Companies that Cross and Link Institutions
  Companies like Syapse and Flatiron are in the news weekly with new alliances, impressive fundraising, and new ways of accelerating the connection of genomic data, institutions, and research.

  See, e.g., Flatiron and FMI's ASCO publication on a real-world "clinico-genomic database"here.

Big Box Entrants
  Constantly in both the public media and specialty media you'll find IBM Watson with widespread announcements of alliances to Mayo, MD Anderson, and others.  But IBM isn't the only big box entrant; track similar efforts from SAP, Verily, Microsoft, Philips, and others.  E.g. this Philips article, here, and their webpage, here.

Outsourced Bioinformatics Layers
  A prototype company in this space would be Translational Software (here) - software as a service to enable genomics laboratories.  But see also examples like MolecularHealth which in 2016 dropped its wet lab as it pivoted to a pure-play focus on oncology software (here).

   In the chart, I've attempted to divide 11 companies in this space as focused on (A) integrated oncology and other testing, on (B) pure-play oncology, or as (C) primarily focused on non-oncology applications.

Virtual Cell Biology
  Although a form of pure-play oncology software, I'd make a separate category for companies that are trying to create in silico cell biology and use it for drug R&D and/or clinical impact.  These include Cellworks, DarwinHealth, and several others.

Novel Genomics Businesses
  A final category is quite different, the flurry of companies that are emerging as Lab Benefit Managers (Concert Genetics, Evicore, and others).

   At one level, the simplest approach is just letting payers outsource their pay-and-deny edits by CPT code.  But that's the tinker-toy level, and this emerging space could be much more powerful than that.  PAMA law in 2014 is best known in the lab industry for a new Medicare lab test payment system (S. 216). But another section of PAMA mandates the use of sophisticated Appropriate Use Criteria (AUC) in the ordering process for advanced imaging (S. 218; read more here here here).  With Medicare payments for genomics topping $500M and rising, just ten words in some new health law could create a Medicare requirement for genomics ordering to be processed through clinician-facing  AUC software a la what was created by PAMA S. 218 for radiology.

And Something New Every Day
And there's something new every day.

One of the concerns of the point of care test industry is that historically these lacked integration with EHRs and LIS.  However, a San Diego startup is building a new POCT platform around inherent, mission-critical internet and app connectivity (Cue Diagnostics, here).  Clearly, this is another form of the digitization of genomics.

Finally, I recently discussed some academic spinout labs like Navican, from Intermountain, that promise wide ranging sophisticated services besides lab testing for oncology genomics (here).   The conncetion with digital genomics is that Navican's complex outreach goals would be heavily dependent on the IT and software and cloud layers.

...And It Hasn't Always Worked

The pioneers in this kind of thinking left some gravestones.  In 2011, Silicon Valley Bio or SV Bio planned to avoid "sequencing, a low margin commodity business" and focus their success on creating more advanced informatics with the genomics.  But it didn't work, and fizzled out after several years of high-flying press releases and large funding rounds (here).  Maybe they were ahead of the curve.  Similarly, an early and substantial run at the clinico-genomic "brass ring" fizzled between IBM and Watson just this year (here).

____
Footnote.

Medicare policy has been back-and-forth with inconsistent examples of making or blocking separate payment for additional software surfaces on top of imaging procedures; for an update, here.