Arguments Against Clinical Meta-genomics
One for my interests in clinical facing sequencing instrumentation, in particular the idea of using meta-genomic sequencing as a general purpose diagnostic approach.
The idea being that with a cheap enough sequencer and integrated sample prep, you can just sequencing everything in a sample. This would give you a diagnostic panel that would detect all known pathogens and as the need arises, new ones.
I’ve written about all the advantages of this approach before many times. Now here are some of the arguments against meta-genomic diagnostics:
We can’t even get money for surveillance!
To me it seems obvious that we’d be spending somewhere in the region of at least $1B annually on a novel pathogen surveillance platform. That is, going out and sequencing pools of individual or environmental samples.
This would provide an international early warning system. Advance notice on novel circulating pathogens.
All that we have is a few smaller projects on the order of $1M.
If you can’t get money for this, who is going to pay for individual testing? So, let’s get money for surveillance first then do diagnostics.
We can detect new pathogens using existing methods!
There are some regions of similarity between SARS-CoV-1 and SARS-CoV-2. And in fact some projects picked up SARS-CoV-2 using SARS-CoV-1 primers. A new novel pathogen is probably going to be related to something we’ve seen before. So we should just use existing methods and we’ll probably be ok.
It’s too cheap!
If you build a meta-genomic sequencing platform that will compete with qPCR, it’ll be too cheap. Even if you’re margins are high (90% on a $20 test) you’re going to need to be crazy successful to make the kind of money a VC needs to justify an investment.
Making something expensive that sells in lower volume is a better business.
People like qPCR.
People just like qPCR, they understand how it works, that it looks for particular targets invariant of background and have a good understanding of its sensitivity, they like looking at amplification curves and Ct values. You’re going to have a really hard time trying to sell them on something else.
Even if you provide a full panel at low cost people won’t buy it.
Many physicians just don’t want to run more tests than absolutely necessary, so even if you can provide a full respiratory panel (Biofire FilmArray) for the cost of a COVID/RSV/Flu test, nobody well buy it. They’ll much prefer to stick with the low risk validated technologies they have than invest time and money in a new platform.
We like our margins!
As one publication put it:
“Markups on qPCR in the clinical microbiology lab convert <$5 reagents and minutes of technician time into $200–500 reimbursements. Markups on syndromic panels convert $120–200 in reagents and tens of seconds of technician time with <2 h turnaround time into $500–2000 charges.”
If you can perform a full meta-genomic based test for $20 you're still reducing our margins (unless we can charge more). No thanks, we’ll keep pushing qPCR.
Most of these arguments are pretty reasonable. Of course, on balance I think the world would be a better place if we did have meta-genomics and I’m happy discuss! Feel free to ping me (new@sgenomics.org) or over on the Discord!