Medical Device Link.

Originally Published IVD Technology January 2002

IN PERSON

The faster, the better

Cepheid trumpets rapid PCR testing as the wave of the future.

For many U.S. healthcare authorities, the anthrax mailings that followed the terrorist attacks of September 11 provided a much-needed wake-up call. While the potential for bioterrorism and biowarfare had long been discussed in public health circles, few authorities had taken sufficient steps to address the issues that all agreed were looming.

The U.S. Department of Defense (DoD), however, has for many years funded research into methods of counteracting bioterrorism and biowarfare agents—beginning with tools for proper identification of those agents. In the thick of this DoD-funded activity are a number of IVD manufacturers with technologies specifically suited to the task.

One such company is Cepheid (Sunnyvale, CA), developer of the Smart Cycler system, a rapid DNA testing device that can deliver results within an hour. More importantly, this device can detect biological agents that might be used in bioterrorism attacks. In late September, after the stock market caught wind of the potential of Cepheid’s technology, the value of the company’s stock skyrocketed by more than 200%. For this issue, IVD Technology editor Steve Halasey spoke with Kurt Petersen, president and chief operating officer of Cepheid, about the company’s technology, including its new GeneXpert device, how Cepheid develops new technologies with other companies, and how the events of September 11 have affected the company. The full text of the interview can be accessed here.

IVD Technology: What is the history of Cepheid?

Kurt Petersen: My previous company, NovaSensor (Fremont, CA), had developed some interesting micromachining and microelectromechanical systems (MEMS) processes that I wanted to apply to biomedical problems, but NovaSensor wasn’t doing anything like that. So I started looking around to find where this MEMS technology could be applied to biomedical problems. During late 1995 and early 1996, I held meetings with Bill McMillan from Syva (San Jose), Greg Kovacs from Stanford University, previous Syva president Tom Gutshall, and Allen Northrup from Lawrence Livermore National Laboratory (Berkeley, CA). Allen was experimenting with a very rapid polymerase chain reaction (PCR) process for use in DNA analysis. All of us realized that if this fast PCR process could be combined with sophisticated microfluidics, then it would be possible to create DNA analysis systems that were very rapid, easy to use, and maybe even portable and field usable.

The basic concept of Cepheid evolved during that time, and in early 1996 I started putting the rest of the team together. Tom became our CEO, and Bill is our vice president for biotechnology.

We soon began talking with Livermore about licensing Allen’s patent, and we began thinking about applying for government contracts. In late 1996 the U.S. Army’s Aberdeen Proving Ground (MD) gave us our first government contract, which was $1.8 million to build an automated rapid DNA analysis system for biowarfare defense.

In early 1997, we got our first round of financing, which was an angel financing mostly from an organization called the Band of Angels. We then got our license from Livermore, so we were off and running.

How did you find the original investors?

That was through Tom Gutshall, who has a vast network in Silicon Valley. We gave a 10-minute presentation at one of the group’s rowdy dinners, and we got several investors and two board members, Dean Morton and Jerry Casilli, out of that. There were some other angels who invested too.

What was it about your proposal that excited them?

At the time, the Band of Angels was composed mostly of people with investment interest in the semiconductor and software industries but not too much interest in biotechnology. Even so, they were very impressed with what we could do.

The example that typically convinces people about the power of this technology is the situation in which somebody has an infection, and they go to a doctor’s office.The doctor takes a throat or nasal swab, then tells the patient that it will take three days to determine if the infection is viral or bacterial and another day or two to know whether the bug might be resistant to antibiotics.

These traditional methods of diagnosis rely on culturing the patient specimen, a 100-year-old technology which is still the state of the art for bacterial identification. Culturing techniques have become a little more sophisticated, but basically it’s the same concept. This extremely slow method has to make you wonder why there isn’t a faster way.

When we tell investors that our technology can do all of that in half an hour, that pretty much convinces them.

How did you decide to go after the government contract? What kind of grant or program was that?

The wonderful thing about the way our interactions with the Department of Defense (DoD) and other government agencies have developed is that what they want for use against bioterrorism or biowarfare is exactly what we want to make commercially available for use in a hospital, a doctor’s office, or a food processing plant. The technology is immediately transferable from one market to the other.

Sometimes small companies get stuck in a situation where they go after government funding, and the technology they’re asked to develop is so customized for government applications that it becomes difficult to spin off onto the commercial side. We’ve been fortunate to be in a technology area where the needs of both government and commercial users are virtually identical. That’s been a big help for us in terms of putting out products that can be bought by both the DoD and life sciences researchers.

Our first contract was from the Aberdeen Proving Ground, which has for a number of years fielded a system called the Biological Integrated Detection System. This system is basically a biological laboratory in a fancy camper shell on the back of a Humvee, and it contains a variety of biological instruments, such as a flow cytometer and immunoassay-based detection instruments. But because of the very high sensitivity of PCR—the commonly accepted figure is that PCR is three or four orders of magnitude more sensitive than immunoassays—some forward-looking people at DoD back in 1996 felt that DNA detection would inevitably become the preferred way of detecting biowarfare organisms. For these reasons, Cepheid has received funding from the Defense Advanced Research Projects Agency (Arlington, VA), the U.S. Army Medical Research Institute of Infectious Diseases (Ft. Detrick, MD), and the U.S. Army Soldier and Biological Chemical Compound (Aberdeen Proving Ground). In fact, the Centers for Disease Control and Prevention recently circulated an e-mail message instructing laboratories not to use immunoassay-based test strips for anthrax because they’re inaccurate and not sensitive enough.

In Search of Additional Financing

How did you manage to turn the funding from Aberdeen into further financing beyond the angel funding that you got in 1997? Did you have bridge funding between them?

The initial funding in 1997 was around $3 million, so it was a pretty big angel round. We went for a second round of funding in April 1998, and the angels came back in on that round, in addition to some venture capital firms. IVP came in on that round, as well as Invemed.

Again, these were people who Tom or I knew, and that B round ended up raising $8.2 million. And it was the same story, describing to people that this DNA analysis can be done very rapidly, and that it has applications in many different markets, including food safety, biowarfare, cancer and other human diagnostics, and even genetic analysis.

When did Cepheid go public?

We started looking for a C round of funding in the middle of 1999, but we had a lot of problems because all the money was being gobbled up by the dot-com companies. Everybody recommended that we call ourselves Cepheid.com. Things eventually loosened up at the end of 1999, and we finally closed our C round in March 2000. We had been looking for $10 million, but ended up closing the round with $19 million. People were begging to give us money, and at the end of the round we actually had to turn people down.

The market for initial public offerings (IPOs) was very hot at the time, so even though our first product wasn’t quite ready, we decided to do an IPO in June of 2000. A month earlier, we had launched our first product, the Smart Cycler, so the launch of our first product and our IPO were almost simultaneous.

Who was the first customer for the product that you sold in May 2000?

It was a lot of government customers. What we normally say is that 20–25% of our customers are in the DoD or other government organizations that are related to biodefense. So those were some of the first customers.

But the Smart Cycler is a powerful instrument for life sciences researchers who are doing all types of DNA research in their laboratories, and that’s where approximately 50% of our sales come from. It’s a unique instrument because it was designed with diagnostics in mind. It wasn’t designed like the usual 96- or 384-well high-throughput roboticized PCR instrument. It was designed for diagnostics in a diagnostic situation, where samples don’t come in 96 at a time, but rather one or two at a time. And they’re usually stat—samples that need quick results—so a 96-well format isn’t optimal.

Working with Partners

How do you change the market for that device, from its initial 20–25% in biodefense and 75% in life sciences, to enter the clinical market, which was what originally got the Band of Angels excited?

A lot of the people involved in life sciences research are developing assays for diagnostics. One of the things we have found is that there are a lot of people who are using the Smart Cycler and are designing diagnostic assays that can be commercialized. So we look at the Smart Cycler as seeding the biomedical community, have them discover how to do PCR better using the Smart Cycler, and then siphoning off the key diagnostic assays that come out of that.

That makes the Smart Cycler a research tool rather than a diagnostic instrument in its own right.

Yes, although we expect that a partner will be introducing some diagnostic assays on the Smart Cycler this year. There are some types of assays where the sample preparation is fairly simple and can be done alongside the Smart Cycler, and where the system can get a rapid, accurate answer.

One of our key partners is Infectio Diagnostics (IDI; Ste.-Foy, QC, Canada), which has an extensive library of DNA sequences including the detection of antibiotic resistance in many infectious organisms. The first test that we’re working on with IDI is for Group B streptococcus, which is an organism that can infect pregnant women without causing any symptoms that would reveal its presence. However, if a woman delivers an infant while she’s infected, there’s a high probability of the infant becoming infected by the organism. In the United States, about 300 infants a year die as a result of infection by Group B streptococcus.

Today, testing for Group B streptococcus can only be done with culturing. But the culturing takes three days, and if the woman is already in the delivery room, there’s no time for that. So we are going to be introducing a test for Group B streptococcus using the Smart Cycler.

Who will be the sponsor of the FDA submission for that test?

The sponsor will be a joint venture that IDI and Cepheid have put together called Aridia (Halifax, NS, Canada).

Is Aridia likely to follow up the strep B test with other assays?

Absolutely. Vancomycin-resistant enterococci (VRE) and methicillin-resistant staph aureus (MRSA) are two other assays we’re looking at. Those are also tests that are performed in a hospital setting and that need to be done in a fairly big hurry.

Once the platform has been approved for certain assays, will you be partnering with other companies besides Aridia?

We’re doing that now. We have a partnership with ETG (Baltimore), which for many years sold chemical and biological detection systems to the DoD. ETG has all the DoD system integration experience that we don’t have, such as global positioning systems, communications systems, and hardened assemblies and keyboards. ETG is going to be incorporating our technology into new instruments that will be used by the DoD for biowarfare defense.

The GeneXpert and FDA


What other products is Cepheid working on now? Are they mostly an elaboration of things that run on the Smart Cycler?

The GeneXpert completely automates everything, from raw sample preparation all the way to detection and result. It incorporates the amplification and detection technologies that are in the Smart Cycler, but adds automated sample preparation.

With the GeneXpert system, the operator simply places a raw sample into a small cartridge that fits in the palm of the hand. The sample doesn’t have to be purified DNA; the system is designed to handle such raw samples as emulsified tissue, food extracts, powders in liquid solution, blood, urine, or swab eluent. When the cartridge is put into the machine, it captures the infectious organisms on cells, lyses them to release their DNA, extracts and purifies the DNA, mixes the DNA with onboard PCR reagents, pushes those reagents into a special PCR reaction tube, and then performs the amplification and detection—all this occurs automatically in 30 minutes.

We delivered the first GeneXpert system to the DoD at the end of 2001, and they will probably use it first to test for anthrax. In 2002, Cepheid will be providing prototype systems to many of our collaborators who will be transferring their manual sample prep protocols to the GeneXpert. That’s what is going to make this technology take off.

What is the regulatory status of such systems? Is FDA likely to grant waived status under the Clinical Laboratory Improvement Amendments (CLIA) for a DNA-based system?

We have a lot of people at Cepheid who have dealt with FDA extensively in the past, and Bill McMillan, our VP of biotechnology has also been communicating and having regular meetings with the agency for about a year. There’s a huge amount of technology involved in the system’s controls to make sure that there isn’t any cross-contamination or erroneous results. So we recognize that FDA has legitimate regulatory concerns.

We are aiming for a CLIA-waived test, and we’re talking with FDA about doing everything we can to get a waived test. However, we don’t expect the first ones will be waived, but we do expect that the system will eventually migrate there as the technology proves itself. There’s a lot that can be done with validated internal controls.

Has FDA raised questions about the examination of discrepant results or what you need to do to compare the sensitivity and specificity of your tests against previous generations of tests?

Dr. Michel Bergeron, the founder of IDI, last year published a paper in the New England Journal of Medicine which was extremely well received. He compared a series of rapid PCR tests with traditional culturing for Group B streptococcus, and basically got 100% correlation of specificity and sensitivity between culturing and the fast PCR.

This paper caused quite a stir. There was a long editorial about it in that particular issue of the New England Journal of Medicine, and a huge number of people asked us about it. The gist of the editorial was that DNA sequence detection, or PCR, is starting to become a real part of diagnostic testing, and this paper is an early example. We think it’s going to happen, and Cepheid is going to make it happen.

The Future of Molecular Diagnostics

Progress in the field of molecular diagnostics has been very slow. A lot of technology is being developed, but it’s very slow to come into the clinical arena. What do you see as the challenges for the field now and in the future?

A large part of the challenge is in making sample preparation simple to perform as well as robust. That’s a very key thing, and that’s where Cepheid has put a huge amount of its energy.

Another issue—especially with regard to meeting FDA approval requirements and making sure that the assay is absolutely right—is to incorporate many internal controls for different aspects of the assay. What this means is that a multiplexed DNA detection system is needed—and by that I mean a system that can detect multiple DNA targets simultaneously in the same reaction tube. Some of those targets would be control targets, and some of them might be multiple targets of the organism being looked for. For instance, with anthrax, there are three DNA targets of interest for a confirmatory test: the anthrax genome and the PXO1 and PXO2 plasmids that give it toxic properties.

This capability of doing multiplexing in the same reaction tube is important, and Cepheid has also put a lot of energy and work into that. Right now, our systems have the capability of simultaneously looking at four colors, or targets. And we have other systems on the bench that are doing many more colors than that.

In terms of clinical diagnostics, what will make DNA testing acceptable is easy, automated, and robust sample preparation that can be accomplished without the possibility of cross-contamination. In our technology, everything is in the tube or cartridge. The sample is inserted, the lid is closed, and the sample never leaves. So there’s never a possibility of amplicon contamination.

In the past, the very largest IVD companies have pretty much stayed out of this market. How does that affect funding for the field in general?

Early last year, Roche Diagnostics (Basel, Switzerland) announced a partnership with Innogenetics (Zwijnaarde, Belgium) to develop DNA probes for infectious diseases, saying that the technology was finally mature enough to perform microbiology using DNA probes. This is an indication that one of the big players in the field believes this technology is coming of age to the point where it is willing to start investing some money.

This is a vindication and verification of Cepheid’s direction in developing DNA tests for many different targets, including infectious-disease targets. And there are other companies like Roche that are also starting to put money into the field.

Does that lead you to the conclusion that more small companies are likely to be purchased, licensed, or partnered? How does that change the shape of the marketplace?

We’ve always been talking to all the major players. But there are always two parts to the story. There’s the reagent part, and there’s the instrument part.

We’re certain that we have the best instrument part and we are close to having the best reagents. So we’ll continue talking to all of those diagnostics companies and see what happens.

Copyright ©2002 IVD Technology