Seattle is known for spawning companies that grow up to define entirely new industries like Amazon, Starbucks, and Costco. One startup with similar aspirations for a new niche of the biotech world, Seattle-based Adaptive TCR, is making huge strides in an industry that didn’t exist a couple years ago, and would have been unfathomable then.
This little company has started to emerge since Xconomy had the scoop back in March that it had spun out from the Fred Hutchinson Cancer Research Center with $4.5 million in angel financing. Almost entirely based on buzz around presentations by its scientific founders, in its first year, Adaptive TCR built a customer roster that includes 50 academic research centers. They are writing “five-figure” checks for a service that provides vividly detailed profiles of how an individual’s immune system adapts on a genetic level to pathogens, and which can monitor exposure to drugs and vaccines, says CEO Chad Robins.
Adaptive TCR, Robins says, is on a trajectory to turn profitable in its second year, 2011. It is in what he calls “serious discussions” to offer its technology to five of the world’s top 10 pharma companies. Further down the road, it has a chance to dominate at least one, and maybe two, new industries that will be worth hundreds of millions of dollars, he says.
“People originally discounted this, saying ‘Oh, this is a service business,’ and then they forget about it,” Robins says. “It’s a several hundred million dollar business on clinical trials alone. It’s a matter for us of scaling.”
These are the early days in the business of immune system profiling. The idea has now become feasible because of the breakneck pace of innovation in both life sciences (faster/cheaper DNA sequencing) and computing (low-cost, large-scale cloud computing data servers).
Those technologies have enabled Adaptive TCR to pursue its own big idea. While the 3 billion letters of DNA that make up a human genome are consistent in almost every cell of the body, the immune system’s B cells and T cells are an exception. In these cells, DNA gets shuffled around in bafflingly complex combinations, allowing T cells to recognize specific invaders, such as flu viruses, and bacteria that people get exposed to over time, and B cells to produce specific antibodies against those intruders. Each person is thought to have 50 to 100 million unique T cell receptors in their immune repertoire, for instance, so that’s a lot of DNA getting reshuffled around in novel combinations that nobody really knows anything about.
Scientists today generally have no way to look at that exquisite complexity in individual immune systems, and tend to look at generalized markers of inflammation like C-reactive protein.
The new immune profiling system is supposed to be far more sensitive. It’s the kind of thing that ought to tell researchers how susceptible a patient is to certain infections, or might show whether a vaccine in early development is doing what it is supposed to do. Scientists have long known that elderly people lose much of their immune repertoire as they age, but it’s now becoming possible to actually see how specific defenses fade over time, laying the groundwork for new vaccine strategies, Robins says. One collaborator at the Hutch is using the system to see how quickly a patient’s immune system regains strength after a bone marrow transplant. That could tell physicians when it’s OK to quit giving powerful antiviral meds that cause side effects, Robins says.
The founding team of Adaptive TCR isn’t exactly full of household names in local biotech. As I noted then in the story back in March, Adaptive’s scientific co-founders are a pair of Fred Hutch researchers who are first-time entrepreneurs in their 30s: Harlan Robins, a particle physicist-turned-computational biologist, and Chris Carlson, a geneticist and molecular biologist. Chad Robins, Harlan’s brother, is the business guy with an MBA from the Wharton School. Before joining the startup with his brother, Chad gained experience with investment banking and hedge funds. He tapped his own Rolodex to raise the company’s seed cash, rather than turn to venture capitalists.
Some big-name scientists have shown interest in what this startup company is doing. Arnold Levine of the Institute for Advanced Study in Princeton, NJ, and Gerald Nepom, the director of the Benaroya Research Institute in Seattle, have joined the scientific advisory board, while Nobel laureate Andrew Fire of Stanford University has been using the technology for experiments supported by the Juvenile Diabetes Research Foundation. The startup’s customer roster includes researchers at Houston, TX-based MD Anderson Cancer Center, Stanford University, Yale University, Harvard University, UCLA, Baylor College of Medicine, and Washington University in St. Louis, Robins says. Craig Weissman, the chief technology officer of Salesforce.com, has joined the Adaptive TCR board of directors.
Researchers today are primarily using the Adaptive TCR service, called immunoSEQ, to learn basic things about how the immune system responds to pathogens. Big Pharma companies might use it to see if experimental vaccines are triggering the specific kind of response they want to see early on, which ought to help separate winners from losers in clinical trials early on. Over time, Adaptive TCR wants to use this immune profiling capability to perform not just services, but to create its own proprietary diagnostic tests, Chad Robins says. It’s clearly a heady time for Adaptive TCR, and a moment to think about making the most of its first-mover advantage.
“They got a head-start, and have a quick and economical approach, to filling a gap in research technology/information,” Nepom says. He adds, though, that it’s still unclear if the data from Adaptive TCR will shed light on revolutionary new biomarkers, “or just a step forward in research knowledge.”
A lot of technical things had to coincide to get Adaptive TCR up and running. There is some intellectual property in the wet-lab assays that the company uses to process biological samples, and in proprietary computer algorithms that make sense of vast datasets on T cell receptors. The actual sequencing of samples is done on the latest DNA sequencing machines from San Diego-based Illumina (which Robins showed me on a recent tour). The sequence data is stored on remote cloud computing servers operated by Amazon Web Services, which researchers can download. Adaptive TCR, from its offices in South Lake Union, has set up an unusually snazzy Web 2.0-style website (at least by biotech standards) which serves as the company’s storefront to its academic and would-be industrial customers.
Behind the scenes, Adaptive TCR has set up work flow logistics so that it can smoothly process more than 400 to 500 samples a week, Robins says. Robins has clearly been studying the habits of some successful consumer-oriented businesses, and trying to incorporate them into what could otherwise look like a service for the tiniest of research niches.
“We think our customers should be able to order like they do on Amazon, track like FedEx, and analyze your data like iTunes,” Robins says.
There are a couple of competitors trying to do similar things. San Francisco-based Sequenta, which I wrote about last week, raised $13 million in a Series B venture round to pursue an immune profiling strategy for diagnostic purposes. Another contender is Huntsville, AL-based iRepertoire.
Adaptive TCR, like any company just a year old, has a lot of work to do on setting up operations. It had about 10 employees when I stopped by for a meeting with Robins in late October, and plans to grow to about 30 or 40 people in the next year, Robins says. It waited until October to start doing marketing of the product, partly because the CEO was afraid he would get inundated with orders before the company had the right processes in place to handle them all.
I made sure to make Robins walk me through the business model in detail so I understood how he arrived at some of the bullish projections. The company offers four different tiers of service, based on how much resolution the researcher wants from the sample. The range runs from $500 per sample on the low end to $7,500 per sample at the top of the line. If, say, a big drugmaker or academic coalition was running a clinical trial of 250 patients and wanted to know about immune responses in those patients, the bills really start adding up fast. That’s because it’s likely researchers would want to take multiple snapshots of samples at three, four, maybe five different points in time to see how the immune system adapts in response to the vaccine or therapy.
If this kind of idea catches on Big Pharma, then there could be a whole lot of packages coming through Adaptive TCR on a regular time schedule. This is why Robins brought up logistics several times during our conversation. “We are working very hard on scaling the business and operational excellence,” he says.
Robins wouldn’t say anything about actual revenues, but I was able to gather that his company has essentially doubled its academic customer base in the last seven weeks. There were VCs sniffing around the company back in October, doing some of their due diligence, along with strategic corporate investors who were showing interest. Robins said he didn’t need the money right now, although he was weighing the pros and cons of whether to take it.
The founders aren’t sure where this business will lead. They drew up a long list of possible applications they imagined, although customers keep asking them whether the service can do something specific that had never been thought of before. One postdoc in particular approached the company’s booth at the American Society of Histocompatibility and Immunogenetics conference in Hollywood, FL in September.
The postdoc marveled at how the Adaptive TCR method was able to generate 1,000 times as much data to characterize T cell diversity in three days as he was able to generate in three years of painstaking work with an older technology. “The level of excitement for this technology is palpable,” Robins says.
High expectations can also be a curse, as many entrepreneurs know. Robins says he’s thinking hard about how not to mess this up. It’s not good enough to just perform the sequence runs and send the data back to the customer. And that has everything to do with old-fashioned business principles, and nothing to do with whiz-bang gene sequencing or cloud computing.
“Customers expect customer relations, and some help, especially the first time through,” Robins says. “There is a human element to this.”