Skip to main content
Premium Trial:

Request an Annual Quote

Dave Howson of Accelr8 on the Company s Move to Diagnostics Tools


At A Glance

David Howson, President, Accelr8

Experience: 2001-Present — Consultant, president of Accelr8

2000-Present — Chief technology officer, Amidex.

2001-2002 — Consultant, Cochlear Americas.

1989-2000 — President, The Altro Group (medical technology consultancy).

Education: 1966 — BS, psychology, Hobart College.


Denver-based Accelr8 Technology, a software company that also manufactures the OptiArray microarray product line, last week announced plans to sell off its middleware software business for $500,000 by the end of the company’s fiscal year on July 31. In the same announcement, the company named David Howson, who has served as a consultant to the firm since 2001, as its president.

Additionally, the company said it had filed a patent application for its BACcelr8 microbial analysis platform. The technology’s first application would be to quickly provide diagnostic results for treating bacterial infections, the company said.

Publicly traded Accelr8 is an 18-year-old company that started off as a legacy migration software manufacturer and entered the life sciences business in 2001 when it acquired the OpTest technology from DDx, a privately owned Denver corporation developing the chemical technology, in a stock-and-cash transaction valued at $3 million.

That technology, which Accelr8 has licensed to Schott Nexterion (see BAN 10/22/2003), forms the core of the company’s planned migration to a new market, and the formalization of the relationship with Howson, who has 30 years of experience in the medical device and diagnostics industry.

Howson’s experience started with four years in the doctoral program at Cornell University in neurobiology. He served as a graduate lecturer and conducted research in “ultra-micro biochemical analysis applied to single-cell protein turnover quantitation” and representing graduate students on the faculty council before he opted out for business. Today, he holds 18 issued patents. He has also founded four companies, selling two, and has worked as an executive in a variety of medically related technology firms and consultancies.

BioArray News spoke with Howson about Accelr8’s metamorphosis and his role in that change.

This is quite a change for a company that is nearly 20 years old.

[Tom Geimer, Accler8’s chairman] is a visionary guy — not a technologist, but a business visionary. According to economic theory, what you are supposed to do when you sense that your line of business doesn’t have a lot of life to it is you milk it for cash and use your resources to find new opportunities where you can take advantage of what you do know. Most companies don’t do that, they simply ride the curve down and the shareholders get killed. Tom did what you are supposed to do, which is redeploy the assets, and if necessary, bring in entirely new people, which is the case here. It’s a complete redesign of the business, but it was totally conscious. We didn’t want to continue the management brain damage from attending to two businesses.

What are your new duties?

In a company of this size, a president’s job is almost anything. As a consultant, I was doing almost all the business development. I was going out, getting Accelr8 known in the industry, talking one-on-one to major players. We are not selling to core labs, but trying to get our name known and trying to get the thing moving. Since we are heading into the medical and diagnostic area, where I come from, my role is considerably broader. I am actively leading the science team now, and will be building it more toward the engineering side and doing full blown product development for an integrated instrumentation platform. So, for the moment, president means continuing business development with a shift to the medical side of the industry while playing a much more active role of directing the technical programs, the science and the engineering that creates the product. I didn’t have a lot to do with that, as I am not a chemist. We have guys who are really superb at that.

How did you get involved with Accelr8?

A shareholder of Accelr8 introduced [Geimer] to a private company that had a stalled R&D asset. That company’s main product lines were not doing that well and so they had to suspend funding of the R&D. Tom decided that was worth a serious look. That’s when I first got involved with Tom and Accelr8. That was in late 2000. I knew about the other technology back in the days before it was suspended. I was looking at it for another company. It was a diagnostic application that was looking at extremely small quantities of materials and it looked like it would have a ridiculously high sensitivity type of detection. So when Tom started evaluating it about late 2000, he asked the people in the other company if they knew anybody outside the company that knew about it. I was the guy.

I did an initial assessment of the technology, and took a rough stab at what good applications might be and wild ideas about the technology’s business potential. Tom decided to go ahead and concluded the deal. By that time, all of the original scientists had left, so we started from scratch. Tom asked me to consult with him, putting together a scientific team, and advisors and whatever it took to validate the technology.

So, starting in January ‘01, we immediately hired Steve Metzgar, who is a very good assay developer and a good chemist. Immediately afterwards we got lucky and hired a chemical engineer-PhD, Mike Lochhead, who was moving out this way from the University of New Hampshire. So together, initially, Steve and then Mike took what was a raw, untested R&D asset, and very quickly validated the surface chemistry component, and found that they could replicate it, and that it did have some remarkable properties. That was the start of our surface chemistry focus.

How did that lead to microarrays?

In the first six months or less of ‘01, our mission was find out what this stuff was all about and what we might do with it. It was purely about a solution looking for a problem at that stage. So we talked to a lot of scientists. The consensus was microarrays offered some real promise.

Our twist on it was not to try to push Corning out of the [slide] market. But, because we have special properties with the surface chemistry that really stand out if you are exposing them to things like blood, messy sample materials, that are sticky with protein, that is really where we have a huge advantage. So, we, more or less from the start, started focusing on things with diagnostic potential. But the first job was to create a microarray surface product, commercialize it, prove that it is real, get some business moving before we tackle the instrument portfolio, which is much more complicated to develop. It took longer than we thought to productize the surface chemistry and microarray slides and plates and start to get some licensing deals done. At that point, we decided to really get serious about the instrument side, pull it all together into a unified platform. That was late ‘03.

What will be the initial application?

Pneumonia in the ICU. That is a very high urgency market, medically speaking. In our market research, my first reflex is to try to sell the concept. Forget the focus group, let me go talk to a doc, eyeball to eyeball. I’ll tell him what we think we can do and ask him if he’ll buy it. The response has been incredibly powerful. We really hit a very important problem, by focusing on ventilator- associated pneumonias. We think the market dimensions, US only, are approximately 200,000 cases per year in the US, and that is probably a low number. They all need multiple tests; they need the answers fast, because these are high-mortality types of infections.

What are the pieces that you are putting together to create your platform?

The big thing is I start at the bottom with the surface chemistry. If you don’t have a quiet, low-noise, low-interference background to start with, no matter what assay you develop and what great detection you have on top of it, you are limited by the noise level inherent to that substrate. So what we are doing is taking this coating and putting it on a custom substrate. It’s really not microfluidics, its much larger scale than that, but there is a little flow channel. It’s likely to be a linear array of big spots, Let’s say you have 20 of these in a row, and each spot is to capture a different species of bacterium or virus. Phase 1 is capture. We identify by location, exactly as you do with a microarray. In fact, we are using microarray technology to make these things. We have an array-like thing. It is just a really the ultimate in focused arrays, and it is in a straight line instead of a square and we can flow things through it. So first we capture bugs against specific antibodies. Then we can bounce light directly off of the bug, so picture a bacterium sitting there, or even a virus. We do a version of side scattering, which is an ancient technology in that regard, but you get a huge signal. And since we have such a low background inherent with our OptiChem surfaces, what we see is almost purely the signal of the stuff we want. It is very specific capture. So it is a protein array. But instead of doing a sandwich assay with yet another antibody, which kills your signal and adds more non-specific binding or interference, we bounce light directly off the bug. That allows us to count individual bugs. That’s unique. Everybody else in microbiology says, ‘Let’s brew up a few billion,’ literally, and then we will start to work on them. We start right with what is in the specimen itself. We don’t amplify, we don’t enrich, we don’t grow things for a day. Our whole objective is to give the doc an answer in a few hours — because that is literally a life-or-death difference. So we have the surface chemistry that gives us the clear, quiet, dark background.

We use electrophoresis to actively drive the bugs to the capture surface where the antibodies have been printed. Instead of taking hours, we take seconds. This is where our potential to do screaming-fast hyb comes in. We could do DNA this way, it’s just that we think the diagnostic market is a lot bigger. So we put in the specimen at one end, use the electrical field to drive the bacteria, within seconds, across the capture surface, down this channel that has a line of antibody pads. The antibodies capture their species. There is a Pseudomonas on the first pad, and an E. coli on the second pad, and staph on the third pad. The electrophoresis is key.

We bounce light off of those captured organisms and we can literally count the individual bugs that are stuck to the surface. I’m oversimplifying, of course, but that is pretty close. That takes a fraction of a second, no time. We have high-speed image analysis. And it is pretty brutally simple optics. It’s nothing like doing an array scanner. So now we have counted the bugs, and they are captured and held there. We can then grow them where they sit. And this is where we can do antibiotic resistance testing, which is critical. This is really the make-or-break thing on the clinical side. This is unique. Instead of growing them up to enormous numbers and putting them on agar plates and using expensive machines that take at least three days, we can do it in minutes, at the microscopic scale. It’s array thinking — work through a microscope and make computers do it and not try to do it manually. And that is what gives us our diagnostic platform.

So, how do you plan to put this out in a regulated environment?

I think it will be a long time before [the US Food and Drug Administration] accepts fingerprinting-type diagnostics, where you have got patterns of proteins or genes, or just using genes as diagnostics. To do what we are going to do with antibodies and measuring growth and function with a bunch of DNA primers is hideously complicated. You have to amplify, and PCR and its cousins are just not suited to multiplexing. This is a multiplex application from the get-go. There is no way to break this down and do it onesies. You have every specimen that is going to have some very large numbers, and some variation in the organisms. So you have to have a lot of analytical power in the platform. Trying to do that with gene markers I think is incredibly complicated. So we don’t like the biological story in gene analysis for this purpose. For cancer, it makes enormous sense, but for bacterial detection it’s just way too logistically complicated and expensive, and then you have the FDA to deal with. So what we see is making something that is very close to classic microbiology, something that FDA already understands. The differences are small enough that we think we can get it through Class 2.

The key is going into the FDA — we want the middle track, the Class 2 — and not create such novelty that it will take us three years to get approval. It does take quite a large amount of validation before submitting to FDA. So there is an extensive sort of laboratory clinical trial required before you submit. But Class 2 is a friendly pathway by comparison. The other issues that is huge for medical devices is reimbursement. You can get approval, but that doesn’t mean that you can get anybody to buy your product because they can’t get reimbursed by Medicare. In this case the hospital is spending up to $40,000 per case for an extended stay in the ICU that is not reimbursed. What we are doing is a major loss avoidance for the hospital. We will not have difficult reimbursement issues. If you are launching a medical device, very often you have to worry more about getting Medicare to pay for it, than getting it through FDA. And we just don’t have to deal with that issue at all.

What is your time frame for this?

The milestone we are projecting is that by the end of ‘04, we will have proven everything we need to in our lab and we will be starting to get some outside validation. Starting in ‘05, we will be starting to promote it for licensing within the industry, including diagnostics. We don’t expect to close any diagnostic licenses in ‘05, but we could be closing some research-related licenses. Then, ‘06 is the big FDA push. So we think late’ 06, early ‘07, again, a wild guess, not a forecast, but [one that factors in the] six-month approval cycle. That’s a pretty bold statement, but I have been doing this since FDA started regulating devices.

What do you think the industry will think about this?

I think they will see us as irrelevant to the array per se, even though we are using array type technology and that’s what makes this possible. There will be some people for sure who will say: ‘Hey, this is a very different way to think about focused arrays and a different way of using them.’ You know, go right to the target and combine functional classic analysis along with array-type molecular thinking. It is a different type of architecture. Affy, for example, couldn’t care less, but the guys who are more diagnostic-oriented will ask a lot of questions. I think anybody that is dealing with the FDA on genomic diagnostics will say, ‘Hmm, maybe there is a different way to think about these problems, at least for the interim.”

The Scan

Transcriptomic, Epigenetic Study Appears to Explain Anti-Viral Effects of TB Vaccine

Researchers report in Science Advances on an interferon signature and long-term shifts in monocyte cell DNA methylation in Bacille Calmette-Guérin-vaccinated infant samples.

DNA Storage Method Taps Into Gene Editing Technology

With a dual-plasmid system informed by gene editing, researchers re-wrote DNA sequences in E. coli to store Charles Dickens prose over hundreds of generations, as they recount in Science Advances.

Researchers Model Microbiome Dynamics in Effort to Understand Chronic Human Conditions

Investigators demonstrate in PLOS Computational Biology a computational method for following microbiome dynamics in the absence of longitudinally collected samples.

New Study Highlights Role of Genetics in ADHD

Researchers report in Nature Genetics on differences in genetic architecture between ADHD affecting children versus ADHD that persists into adulthood or is diagnosed in adults.