Chief Scientific Officer and Sr. VP of R&D
Chief Scientific Officer and Sr. VP of R&D at Metamark Genetics
This executive has an accomplished background in the discovery and development of oncology biomarkers, companion diagnostics, and assays. He has held senior level positions in various areas of research and development at Ciba-Corning, Chiron Diagnostics, Bayer Diagnostics, Decision Biomarkers, and On-Q-ity, and most recently served as Vice President of Biomarker Development at Biothera, working with their novel cancer immunotherapeutic. He is currently the Chief Scientific Officer and Sr. VP of R&D at Metamark Genetics. He received his B.S. in Biology from MIT and his Ph.D. in Tumor Immunology from Brown University, and subsequently received his postdoctoral training at Dana Farber Cancer Institute in biomarkers for lung and breast cancer. He was also a Research Fellow at Harvard with a focus in assay development and clinical application through work with clinicians and surgeons in the area of lung cancer.Read moreView Profile Page
Disclaimer: This interview is for informational purposes only and should not be relied upon as a basis for investment decisions. In Practise is an independent publisher and all opinions expressed by guests are solely their own opinions and do not reflect the opinion of In Practise.
Maybe to start, I understand that you have used Abcam’s antibodies in two separate roles. If I piece this together correctly, it seems as if for the first, you were working with in-vitro diagnostics and for the second, you were more on the research use only side?
Maybe you could give me a little background on each of those roles and what you did and what you were researching and trying to accomplish?
For the IPD side of things, I led a group that was charged with developing a multiplex immunoassay for a point of care diagnostics device. The instrument that our company developed was capable of running up to 30 different immunoassays in a single reagent cartridge. On that cartridge, we had multiple primary antibodies to isolate the individual markers that we were looking at. We had a nitrocellulose background and the protein antibodies would adhere to that nitrocellulose background. We did it on a geometric pattern so we could identify which specific antibody was where, based on its location on the grid.
30 specific antibodies on that cartridge were initially put down. Then the sample was laid across and all of the 30 individual markers, if present in the sample, would bind to those primary antibodies. We could then come in with either a directly labeled secondary antibody or an unlabeled second antibody, followed by a third labeled detector antibody.
You can see that, with 30 different biomarkers and up to three antibodies per marker, that quickly adds up to not only a huge variety of different types of antibodies, but a large quantity of each one of those. As we started doing our initial development work, we needed quite a few to do our trial and error work. Once we had selected the appropriate antibodies, we bought them in bulk to go into the manufacturing and further validation and, ultimately, into commercialization.
It required multiple types and large volumes of antibodies and we needed a larger company to be able to select them. At the beginning, we had to select not only compatible ones per marker, but when we combined all those antibodies together, they all had to play well in the same sandbox and not have any negative interactions that could cause problems. There was an initial selection, based on specificity for the marker, then there was cross-selection for compatibility and then scale up to get the volumes that we required.
We had a significant budget there for antibodies, because the nature of the beast was such that we needed large quantities and large variety. We had multiple of these projects going on. We had several cancer multiplex projects; we had one for a neurodegenerative disease. We had another one for inflammatory disease. Each of these panels were from between 10 and 30 different markers. Our antibody requirements were tremendously large.
The antibody was, effectively, a consumable within the cartridge?
Absolutely correct. It was a consumable and, for each assay, each cartridge was identical. If we were selling this to a reference laboratory that required many of them, or we were doing a clinical trial, we needed many of them and they all needed to be exactly identical. The cartridge would be put into the instrument, the sample would be added, the instrument would do its thing and we’d dispose of the cartridge and prepare for the next sample. Both the cartridge and the antibodies, technically, were disposable consumables.
When you think about Abcam primarily being in research use only, how do you end up with Abcam as one of your vendors on the diagnostic side?
When that happens, when the company is research use only, we have to then go in and speak with the representatives, initially through customer service, and see if we can get licensing contracts to allow us to use them for other purposes. Usually, the larger companies will have that; some of the smaller companies, obviously, don’t. But it’s pretty straightforward, sometimes, depending on how those antibodies were developed in the laboratory. Sometimes it gets a little more complicated and sometimes it’s impossible. It’s a little bit of a back and forth.
If that is the only antibody that is available that works in our system, then we have to go and do what we need to do to make it work. If there are others available from another company, we’ll go the easier route and do it that way. For the most part, larger companies – not just Abcam – were accommodating in terms of rights to using the antibody for specific purposes.
Of those 30 different ABs that you used here, what percentage were supplied by Abcam versus other vendors?
It was variable, depending on the specific markers we had. If, for a cancer marker, we had TIMP, which is a metalloproteinase inhibitor, we needed a primary antibody to capture it and we needed a paired antibody to detect it, so we had to find two antibodies to that particular marker that worked well in assemblage. They had to have the epitopes; they had to be sufficiently distant from each other that there weren’t any interactions between the two. We would either look on the catalogues or call the company and ask them specifically, do you have paired antibodies. If Abcam had the ones that we needed, we would go to them. If R&D Systems had paired antibodies, we would try them. We would try any company that had the specific pairs that we needed.
There was a lot of up-front research done in silico, on the computer and online, to find the appropriate one. I can’t give you an exact number, in terms of a percent of which ones were Abcam specific and which ones were from R&D Systems, but it was based on the particular marker we were looking for and on the availability of paired antibodies.
The detection antibodies were different. If we used a labeled detection antibody, they were generic. Those we could get from pretty much anybody. For one particular panel of multiplex assays, we got those detection antibodies from Abcam. For another one, we got them from R&D Systems. In that sense, those were exclusively from Abcam or exclusively from R&D, for the two different panels. But for all of the individual assays, it was variable, depending on the availability. The pairing was important so we would talk to their customer service people and they would get the technical information back to us to ensure that the pairing would be sufficient for our purposes.
If Abcam had it, absolutely we would buy it. The issue was that, sometimes, multiple companies had the appropriate pairs so, initially, we would test them from each of the companies. We would rank the pairs on an individual basis, running one assay at a time, and seeing which assay gave us the best signal to noise ratio. We would then put the multiple pairs, for all of the various markers, together in a cocktail and ensure there wasn’t any negative interaction. There was a lot of up-front work before we made the final selection. That was just based on trial and error; not necessarily on the quality of the specific antibody. Although the initial quality had to be there so it provided us a high enough signal and a low enough noise level at the low end.
Percentage wise, once it was selected, it was a no-brainer; we would buy it and we would buy it in bulk from whichever company could provide the best one to us. I would say, obviously, they were in the mix. If I were to divide them all, I would say at least 20% came from them because there would be another 20% from R&D Systems and some of the bigger ones, here in the US. I know Abcam has facilities here, as well, but I believe they are UK based.