41J Blog

This post previously appeared on the substack.

I’d guess by now everyone in Biotech has heard of Theranos and John Carreyrou’s excellent book on the company. John now has a podcast on the trial currently underway which I’ve been following with interest. I suspect that anyone who has worked in Biotech will recognize the personality types and dynamics at play. In any case, it’s fascinating to see things play out from this unique perspective.

Recent episodes have discussed Theranos’ use of modified Siemens ADVIA instruments. The claim is that Theranos were saying they were using their own instruments, but actually just processing finger stick samples on modified Siemens ADVIA systems. In defense of this position, some seem to argue that the ADVIA modifications were a significant trade secret, which they only wished to divulged when the patents covering their modifications were published.

So, while it’s not really my area of expertise, I decided to go and have a look, and generally have a poke around in Theranos’ patent portfolio.

The first patent I came across that mentioned the ADVIA system, really just seems to say “you could use a smaller sample container”. It has a single independent claim:

So you can put inserts into the sample container to make it smaller and it can have a angled floor. The other claims are mostly around the cavity geometry and coupling the container to a sensor… the patent doesn’t explicitly discuss the ADVIA (there are no examples). But it does say that assays may be performed on commercial equipment including instruments from Siemens and the ADVIA specifically.

Overall, it’s difficult for me to see how the above patent is novel.

Other patents mention that the ADVIA was used for reference measurements. But beyond this I couldn’t see other references to the ADVIA…

But I decided to poke around a little further and found one or two other interesting things including a patent that focuses on the use of their instrumentation in space, leading with:

“Open top test tube and sample containment will encounter challenges to work effectively in the weightless or microgravity environment found on a space station or on a spacecraft”

Other patents are full of hundreds of exciting pictures of the mechanical system, or discuss testing for Zika virus. But nothing jumped out as particularly novel.

One patent that did seem a bit more interesting was this one. Here they describe an approach called TOSCA (two-step chemiluminescent enzyme immunoassay). They present this as an alternative to regular competitive binding assays. This method “eliminates the need to mix a conjugate and a sample before exposing the mixture to an antibody, which may be desirable when very small volumes of sample and conjugate are used”

Essentially this is a minor modification to the competitive binding approach. In the competitive approach, antibodies are attached to a support (surface). You flow in a labeled Antigen along with your sample. If your antigen of interest in present in the sample, it will compete with the labeled antigen. This results in reduced signal at higher antigen concentration.

The Theranos TOSCA approach is exactly the same, but you flow in your sample first and the labels second. This should give antigens in the sample a better chance of binding to antibodies. I guess the idea is that at low concentration you have more antibodies available than antigens in the sample. So you try and bind all sample antigens you can. Only then do you flow in a labeled antigen to “fill in” any antibody sites which are not occupied.

They show data where the TOSCA approach appears to show stronger signal:

Later in the patent they discuss calibration approaches, for example using a known spike-in. These plots show sensitivity down to about 10 pg/ml. Beyond this response looks pretty linear to me. Elsewhere they describe “low concentration” as being in the 10 pg/ml range. So this doesn’t appear to be an approach that in designed to be competitive with some of the fg/ml detection approaches.

There are hundred of patents to look at, but in my brief search I didn’t come across anything I found hugely exciting. Still, I’ll be be keeping all this in mind as I continue to listen to reports on this fascinating case.

This post previously appeared on the substack.

I initially came across Mesoscale when looking at Quanterix. Like Quanterix, Mesoscale make very high sensitivity claims for their immunoassay, in the fg/ml range. Mesoscale Diagnostics has also been around for much longer that Quanterix having been founded in 1995. They are reported to have an annual revenue of ~$100M. 

In particular I came across Mesoscale, because they show very 1 fg/ml sensitivity on HIV p24:

These results are similar to the 1 fg/ml results shown by Quanterix, which suggest that Quanterix is not unique in their sensitivity claims.

The Mesoscale approach uses electrochemiluminescence (ECL) to detect antibodies. Here we’re electrically initiating a chemical reaction which will emit light. We can imagine this has certain advantages over fluorescence. In particular we don’t have any background fluorescence to deal with, or any excitation illumination to filter. ECL has in fact been using for single molecule super-resolution imaging. Though I imagine that with each reaction only emitting a single photon, the advantage of this over other super resolution approaches is limited.

If you’re familiar with Roche’s Cobas e411 and this all sounds somewhat familiar that’s probably because they use the same basic approach. Mesoscale successfully sued Roche for infringing their patents in a legal situation that seems both fascinating and horribly complex. 

Interestingly, there’s a teardown of a Mesoscale instrument on YouTube. The instrument uses a fairly typical 2000s vintage Princeton/Roper cooled CCD camera with 20 micron pixels. Comments on this video were also fascinating, indicating the the plates cost ~$1000. As these plates require embedded electrodes, we would expect them to be a high cost consumable. But this still seems rather expensive…

Comments also suggest that “multiple competitors…can outperform and do assays cheaper” and “Bio-rad’s multiplex Elisa is pretty much as good”. The Bio-rad kits I could find however only report sensitivity down to ~5pg/mL on HIV p24 which is an order of magnitude worse than Mesoscale. But of course for many other assay’s like Quanterix the sensitivity is likely limited by available antibodies in any case.

For the most part I was interested in looking at Mesoscale to see how unique Quanterix/Simoa is. In terms of raw sensitivity, Quanterix does not appear to be unique. Mesoscale platform is showing similar single digit fg/ml sensitivity. The Roche Cobas e411 p24 assay also gets close, with sensitivity in the 60 fg/ml range. 

It seems likely that Quanterix has a cost advantage over Mesoscale. But Cobas e411 tests in the range of $20 appear to be available. It’s therefore possible that Quanterix may still have an advantage here if they can show very high sensitivity at low cost.

This post previously appeared on the substack.

Somalogic are developing a aptamer based protein detection platform with a microarray-like readout. Aptamers have been around since the 1990s, the basic methods being far older still. The basic idea is that you evolve a DNA or RNA sequence that will specifically bind to a protein of interest. Aptamer’s have a troubled history and I’m not really aware of many successful commercial applications based on aptamer approaches.

Somalogic claim to have solved the aptamer problem by adding 4 additional modified nucleotides to the existing natural set. The basic method was published in 2010. This gives them 8 basic monomers to work with (still a lot less than 20+ amino acids, but in the ballpark of the ~10 that have been proposed as a minimally viable set).

The heart of this paper shows that they can find aptamers for various proteins using modified nucleotides where they previously couldn’t:

Kd values here range from 10⁻⁸ to ~10⁻¹⁰ (smaller is better). This in the range of high affinity antibodies (but not very high). These values also compare well to those for regular aptamers, for example these in with 10⁻⁸ to 10⁻⁹ range.

However even antibodies have a relatively high false positive rate (for example, 0.4% for COVID19 tests). A false positive rate of 1 in 250 may not seem like much. But when you’re looking at ~10000 proteins, you’ll get 40 false positives. Of course, this doesn’t exactly map to the Somalogic approach where you’re looking to quantify expression levels rather than obtaining a yes/no answer.

So, particularly coming from a background in genomics, the Somalogic readout is likely to be noisy. This is likely the case for any large-scale aptamer/antibody platform. Olink, by linking two affinity reagents may be able to reduce this. Nautilus, by using multiple affinity reagents sequentially, may also have a route to a less noisy readout.

But what Somalogic may have, is more resolution on their readout. Unlike Olink and Nautilus, the Somalogic platform doesn’t appear to be using a single molecule approach.

Essentially they fish out the aptamers that bound, and then pull these down onto a microarray like platform (hybridizing the aptamers to probes on the surface). Microarray’s (and hybridization) always seems like a process fraught with issues, with environmental issues (like temperature and ozone) effecting results. All issues, you rarely see with sequencing.

So, there’s potential for binding issues between aptamer-protein and with the aptamer-array. The upside is that the array gives you a large dynamic range (I believe they suggest 12 orders of magnitude). So rather than having to readout billions of molecules (as Nautilus or Olink might) you can get this information from the Somalogic platform by imaging each protein (aptamer type) once. I.e. an array with on the order of 10,000 spots. A fairly small (and likely cheap) array by DNA microarray standards.

Final Thoughts

Somalogics secret sauce seems to be in their modified nucleotide aptamers. I would guess at best these are one order of magnitude better than regular aptamers. So you can make the argument (like Nautilus) that you could just use a couple of lower affinity aptamers rather than one slightly higher affinity one.

Aptamers do give Somalogic a natural way to generate a high dynamic range readout (using microarrays). That’s neat, but brings its own issues. And overall, I suspect compensating for artifacts may be harder than in single molecule platforms (like Nautilus and Olink).

All these platforms are likely to have issues with binding affinity, background signal, and other environmental factors.

These are issues that a true protein sequencing platform could potentially overcome.

This post previously appeared on the substack.

Quanterix was formed in 2007 with a focus on building a “single molecule” detection platform with a focus on proteomics applications. They raised a total of $533.33M and exited via IPO in 2017.

They position themselves as being very high sensitivity, for diagnostics applications. So (as they suggest) you have pretty much every other current and next-gen proteometics player working on “Research Proteomics”, then Quanterix coming in with high sensitivity on the Simoa platform which would be applied to diagnostics:

Which of course is where they suggest the big money is:

They draw a slightly awkward parallel to Illumina/genomics. Where Illumina make the discovery platform, but the bigger market is likely diagnostics. The analogy is awkward of course, because the Illumina platform often gets used both for discovery and for diagnostic applications [1]. Whereas in Quanterix’s case they’re suggesting that different platforms will be used for discovery and diagnostics.

Of course none of this really tells us much about the nuts and bolts of the Quanterix Simoa approach. But you can find a video here which discusses it in some detail.

Essentially what Quanterix are doing is single molecule ELISA.

In the traditional ELISA approach you have an antibody which binds to an antigen (some protein of interest). Then you have a second antibody, which binds against this first antibody with a linked enzyme. After washing off any unbound antibodies, this enzyme can be used to process a substrate into a product that shows fluorescence. So, you get a kind of amplification reaction, where a single antigen generates thousands of fluorophores.

Quanterix mix this up slightly, by isolating antigens in wells. This means each well shows activity from only a small number (ideally single) antigen (target protein). But through the ELISA process each antigen will generate a large number of fluorophores.

This is therefore not single fluorophore imaging and this should be possible using standard cameras and optics. In DNA sequencing, this likely wouldn’t be termed “single molecule”. As using the same logic you could call Illumina sequencing “single molecule” in that each observation ultimately is sourced from a single molecule [2].

The Quanterix process is described in their 2010 Nature paper:

It makes sense that this could give increased sensitivity, as you have more sensing regions to work with, fluorescence is confined to femtoliter wells.

The single well occupancy is Poisson limited. Meaning that you’ll have some percentage of wells with single antigens (at most ~37%) and others with zero or more than one. But of course all wells are telling your something about the overall concentration of the analyte in the solution. What’s also neat, is while at the low end they operate in the digital domain (counting the number of wells with a single active enzyme) at the high end when all wells have one or more enzymes they can switch back to an analogue approach and just average over all intensities. This is part of what gives them higher sensitivity and dynamic range.

Using this process they claim they can measure down to 0.01 pg/ml, femtomole concentration:

This is great, but some public presentations suggest that in practice noise is somewhat higher in the 1pg/ml range, and that perhaps this is limited by issues relating to the processing of real samples:

This would be more in line with Somalogic and other platforms. 

One report where the increased sensitivity does seem to add value is a article where they use the platform to look HIV drugs. In this case using qPCR is problematic (because HIVs mutation rate is so high), so looking at the p24 capsid protein appears to give a better estimate of viral load. Here (at low viral load) the Quanterix platform’s sensitivity appears to add value, showing sensitivity in the 10s of fg/ml range, right at Quanterix’s detection limit:

The Simoa platform appears to use ~216000 femtolitre wells (seemingly per analyte). In the diagram above these are shown as ~5 micron wells. Much bigger than Illumina’s current feature size, and mostly likely larger than used by Nautilus. So this is likely a modest fabrication problem, and I assume the fabrication costs looks somewhat similar to the Illumina BeadArray platform. I don’t have exact costs for these, but as they appear to be used by 23andme, which look to have ~50% margins, I suspect these chips cost <<$50.

This is relatively cheap, and I suspect cheaper than some other approaches, but may still be too expensive for some diagnostic applications. For example, would this complete with Olink’s qPCR based platform? Which likely costs <$1 per sample.

Final Thoughts

Quanterix may have a platform that can provide higher sensitivity protein detection for some applications. There was one application where this appeared to be useful in practice, and potentially better than other next-gen approaches, but I’d need to dig further to come to any strong conclusion here. It seems likely that their sensitivity is at least “as good as” other cutting edge approaches.

What’s less clear is if their focus on diagnostic applications is realistic. In particular they suggest that approaches will be developed on other platforms (like Olink’s) and then transferred to Quanterix because it’s cheaper and more sensitive. But it’s not clear to me why you wouldn’t for example develop a test on the Olink PEA-NGS platform, and then transfer it to the Olink qPCR platform…

Finally, as I often do, I dug through the Quanterix glassdoor reviews. As always the negative reviews are more interesting than the positive, with numerous comments suggest that the company is “incredibly top heavy” and that “key positions in the company have been taken over by the CEO’s cronies”. “menu content drove everything here. It never mattered whether the assays worked of not. There was never a commitment to product quality”. This and a few of their public statements give me some concern.

But Quanterix are on the market, with 25.4M revenue in Q2. So to some extent, the academic publications and revenue should speak for themselves. I remain slightly more suspicious of their ability to break into major diagnostic applications. But like the idea that you can have a digital readout at the low end and scale to an analogue readout at the high end. This seems like a pretty neat idea.

1. You could say that this is reasonable viral diagnostics, i.e. COVID19. Where the original sequence (and variants are detected) on the Illumina platform, but you use qPCR for diagnostics based on the reference sequence.
2. In the case of Illumina sequencing, each cluster is grown from one template, in Quanterix thousands of fluorophores are generated by one antigen.