I Am A Sequencing Absolutist
If an experiment or diagnostic can be performed using DNA or RNA sequencing, I’m all for it. If it’s 10 times as expensive… that’s still what I want. It’s worth it. Sequencing is just axiomatically better.
I’d perform metagenomic sequencing on anyone with any infection, because I think it’s fundamentally better than qPCR.
Any patient with a condition of unknown origin? Sequence them. And not just sequence them, but use the best sequencer available. Then have a team of Bioinformatians pore over the data.
Scratch that, any human? Sequence them! Screen them for potential conditions. Sequence twice, three times, once a year!
A sequencer in every hospital. In every clinic. In every kitchen!
Unfortunately this isn’t how the world works…
Your GP Is Not Gregory House
Unless you are a clinician or have regularly interacted with clinicians I think there’s a tendency to base your understanding of clinicians on what you see on TV or in the media.
For those of us in the sequencing industry “the media” includes scientific articles and presentations. But the reality is that doctors are just people. In general they operate more like IT Technicians than obsessive opioid addicted geniuses.
If your computer has frozen destroying hours of work an IT Technician is just going to tell you to “turn if off and on again”1 . Same goes for a clinician… feeling a bit under the weather? Probably a cold… rest for a few days and see if it gets better. Child seems to be developing slowly? “they’ll probably catch up”…
Only if things get too serious to ignore will a clinician order a diagnostic test. And this test will be for whatever is the most common cause of the symptoms. And only if they think the test is likely to have some actionable outcome.
Let’s look at how the first COVID19 case was identified and the tests performed (full details in the footnote):2
Basic blood workup. White blood cell counts etc.
Day 6: Chest X-Rays. Chest CTs.
Day 11: More Chest X-Rays.
Anti-gen testing for Flu, Chlamydia and Pneumonia.
PCR testing for the same things because they didn’t believe the anti-gun tests.
qPCR testing for common respiratory pathogens.
Do nothing for 3 days… admitted to ICU.
Sent to another hospital… meta-genomic sequencing.
Clinicians slowly accumulate information, and the reality is that in 90% of cases this patient would have “just got better”. Of course I’m a sequencing absolutist… so I would have sequenced on day 1.
For me “it’ll probably get better by itself” isn’t a standard to be aiming for. That we say “I’ve got a bit of a cold” rather than “I have HRV-A2 CA90-0150” is unacceptable. I may be atypical in this regard. But clinicians are in some cases waking up to the fact that patients like to know what’s wrong with them (even if it’s not immediately actionable):
“Frankly, patient experience scores increase when patients leave our care and have a name for the thing that is bothering them even if there isn’t a targeted antibiotic or antimicrobial that has been identified, that would work against the pathogen. It still is comforting to patients and their families to have a name for the illness that they have.” - Emily Volk
And on a societal level if we paid more attention to the causes of seemingly minor ailments we might prevent or at least better contain future pandemics.
The Germline And The Reality
I’m made the point above using infectious disease testing. Because that’s what I’ve spent most of my time investigating. But the same dynamics apply to other genetic conditions. Our issue is not just that “we lack a complete understanding of the genetic basis” of many obviously genetic diseases3. Or that our current tools are limited4.
It’s that even when diagnostic tests are available clinicians are reluctant to perform them.
I’ll take example of intellectual disability. The leading genetic cause of intellectual disability is fragile X. As such you’d imagine that most clinicians would test for this if they encounter a child showing delayed development. And the good ones will… they’ll use karyotyping… an approach we’ve had since the 1960s.
The bad ones… they’ll say “you can’t fix disabled” and run no tests.5
There are (to my knowledge) no treatments for fragile X or most other forms of intellectual disability. Without effective treatments, clinicians may view a diagnostics as unnecessary6.
Why sequence, why find the cause, if there is no treatment?
I’m a sequencing absolutist. More information is better. More information is better for the patient, to give them a better understanding of their condition. Even just a name, a clear cause, can be helpful.
But more information is also better for society. To help us better understand human genetics. To help us better understand, and perhaps one day treat or prevent genetic diseases.
But I’m also a reluctant realist. And have to accept that without clear actionable results in most cases providers won’t pay for tests and clinicians won’t push for them. For more advanced diagnostic tests (like those based on sequencing) to gain widespread adoption we will likely have to show both high diagnostic yield and clear actionable results.
I’ve both worked in IT Support and faced this issue personally. A frozen copy of keynote losing a few hours work. In that case I could attach gdb to the process, jump out of the busy loop it was trapped in and recover my work. Would I have done the same thing for a user? Probably not…
Mild lymphopoenia was observed, but white blood cell and blood platelet counts were normal in a complete blood count test. Elevated levels of C-reactive protein were observed and the levels of aspartate aminotransferase, lactic dehydrogenase and creatine kinase were slightly elevated in blood chemistry tests. The patient had mild hypoxaemia with oxygen levels of 67 mm Hg as determined by an arterial blood gas test.
On the first day of admission (day 6 after the onset of disease), chest radiographs were abnormal with air-space shadowing such as ground-glass opacities, focal consolidation and patchy consolidation in both lungs. Computed-tomography scans of the chest revealed bilateral focal consolidation, lobar consolidation and patchy consolidation, especially in the lower lung.
A chest radiograph revealed a bilateral diffuse patchy and fuzzy shadow on day 5 after admission (day 11 after the onset of disease). Preliminary aetiological investigations excluded the presence of influenza virus, Chlamydia pneumoniae and Mycoplasma pneumoniae using commercial pathogen antigen-detection kits, and this was confirmed by PCR. Other common respiratory pathogens, including human adenoviruses, also tested negative by quantitative PCR (qPCR). Although a combination of antibiotic, antiviral and glucocorticoid therapy was administered, the patient exhibited respiratory failure and was given high-flow non-invasive ventilation.
The condition of the patient did not improve after 3 days of treatment and he was admitted to the intensive care unit.
The patient was transferred to another hospital in Wuhan for further treatment 6 days after admission.
To investigate the possible aetiological agents associated with this disease, we collected bronchoalveolar lavage fluid (BALF) and performed deep meta-transcriptomic sequencing.
Though this is certainly also the case.
Perhaps long read sequencing or methylation or something else will be the magic bullet that will massively increase diagnostic yield.
While this may seem harsh. It is in fact a direct quote from a clinician.
This is of course pretty ridiculous, particularly in this case as the fragile X pre-mutation may be present in other family members if this is the underlying genetic cause. As such other family member may want to be screened.