Joe Hogan leads technology side of newly integrated GE-Amersham

January 24, 2005

No other company has done what GE has done: The merger of its medical systems unit with British pharmaceutical giant Amersham combined a multimodality maker of imaging equipment with the manufacturer of contrast media.

No other company has done what GE has done: The merger of its medical systems unit with British pharmaceutical giant Amersham combined a multimodality maker of imaging equipment with the manufacturer of contrast media.

This merger of two inherently complementary businesses has raised the potential of synergies in R&D, marketing, and distribution. It has also created extraordinary challenges to reaching those potentials. Two different businesses with two very different approaches to medical technology are now jointly charged with ushering in a new era of personalized medicine.

Joe Hogan is handling the medical device side. Before the merger, he served as president and CEO of GE Medical Systems. He holds the same titles for the medical diagnostics and IT arm of the merged company. But he has the added responsibility of ensuring that developments on his side of the Atlantic complement those on the other side of the pond, at the digs of newly named GE Biosciences and the worldwide headquarters of GE Healthcare.

In an interview with DI SCAN, Hogan described how the merger is going, opportunities that have arisen because of it, and what GE has done about them. Of greater significance is how the merger might change the practice of medicine to reflect what Hogan and other GE executives have described as a coming era of personalized medicine.

SCAN: Where does the Amersham merger stand today?

Hogan: We monitor 21 areas, or variables, to keep track of new acquisitions. They go all the way from compliance to revenue capabilities. And right now, we are where we want to be on most of these 21 variables. We feel good where we stand right now, particularly on the contrast part of the business, which relates to our key business, which is diagnostic imaging.

SCAN: What about research?

Hogan: A lot of the discovery business, where they do mass spec and gene sequencers, has been on the research side of pharmaceutical companies. The legacy GE Medical Systems never had a strong link there. But the legacy Amersham business has had a strong sales force in that area. So we have taken our animal imaging business, small CT and MR scanners, and brought them into that distribution channel. We've also grafted our services team into that channel so that we can bring broadband and service capabilities together on the pharmaceutical side. We have top field engineers who know how to service equipment, so we could readily move into that area. And the sales people's ability to sell a broad portfolio of equipment has already been established.

SCAN: Where have you found synergies?

Hogan: We are seeing progress in areas of overlap. A good example is ultrasound contrast agents, which we have merged with our ultrasound business. The ultrasound people are changing their machines, some of the protocols and software, to optimize contrast agents. The sale of ultrasound contrast in the U.S. is up over 100%. Now, this is off a small base. Our sales of contrast have never been big, so 100% is a couple, three million bucks, but it helps validate the synergy model.

SCAN: Let's look at the opportunities you have in ultrasound. Tissue harmonic imaging developed from the idea of using contrast. Someone pushed the button before the contrast arrived and the idea of using second harmonics was born. Are you making any efforts now to evolve the technology whereby something similar might happen?

Hogan: The Amersham people and our imaging people are collaborating now to a great extent. When you look at the ultrasound side, this collaboration is very derivative in the sense that we have an imaging agent and an imaging modality. We're looking at what kind of tweaks and turns can be made.

Having Biosciences as part of GE really helps us on our product road maps. For one, the Biosciences people see what we have coming up in nuclear medicine - PET technology - and it can help them in the development of their agents.

We see PET moving from oncology to neurology. Last year Amersham announced a deal with the University of Pittsburgh concerning compounds that may be able to stage Alzheimer's disease according to the quantity of amyloid plaque. So rather than FDG, which is a broad functional imaging agent, with this new agent you might find out specific information about amyloid plaques, which may give you a much more definitive statement in vivo about a disease's progression. This is personalized medicine.

SCAN: What are the challenges for GE, which has been an equipment company, working with Biosciences?

Hogan: One of the things we will have to come to grips with is that we have lived in a 510(k) environment. The next iteration of a product has always been six to 12 months away. Now we are looking five to six years away. We also have to examine imaging agents and how they interface with the medical equipment that we have in the pipeline.

SCAN: Can you give me an idea of what you see coming over the next few years in the three key areas for PET: neurology, cardiology, and oncology?

Hogan: PET/CT will see tremendous progress in its key areas. In neurology, it will play a major role in early detection and therapeutic response to Alzheimer's disease. In cardiology, PET/VCT (a PET scanner combined with the 64-slice LightSpeed VCT) will drive increased adoption by cardiologists. Cardiac workflow and streamlined processing will be key for growth. Finally, in oncology, PET/CT will be reimbursed and applied routinely for all cancer types. We will see advancements in acquisition and processing capabilities for earlier detection of cancer and improved lesion detectability.

Over the next three to five years we are going to look at certain targeted imaging agents around specific disease mutations in cancer and neurology. Breast cancer for example, is hereditary in 5% to 10% of cases. Mutations in the genes called BRCA1 and BRCA2 are involved in many of these cases, and we hope to tag genetic predisposition earlier with in vivo testing. We hope to help at-risk women and their doctors by having molecular imaging agents that can find breast cancer at the molecular level when it first presents. One of the biggest problems at-risk women and their doctors face now is knowing what to do next once they've determined there is a genetic problem.

SCAN: Where else might we see imaging become more specific?

Hogan: Breast mammography in the year 2010 might be much more specific. Maybe it will be used on women of a particular genotype who should be imaged between the ages of 30 and 35 because they have two gene mutations. When the population can be segregated that way - when you see tests in vivo that used to be in vitro being used in imaging - those are the mile markers along the way toward personalized imaging.

SCAN: In that time, mammography will probably have changed very much. Breast tomosynthesis may be one of these changes. I'm actually surprised we haven't seen anything yet.

Hogan: We have been doing breast tomosynthesis for about four years. I could have commercialized the product two years ago. But we don't just want to make 3D images. We will come out with a package at the end of '05 or '06. But before we do, we will have determined why someone would want to do tomosynthesis. That is what we are doing now.

SCAN: In the meantime, you are continuing to develop the full-field digital Senographe. Will that serve as the platform for tomosynthesis?

Hogan: Our new 2000DS product will serve as a platform to create a large panel volume. We have learned that in order to increase the DQE (dynamic quantum efficiency) from 2x to 3x, we had to change some things on our panel, we had to change the crystal structure. Tomosynthesis will involve changes in the crystal technology and the electronics that take the signal off the back of the panel, because if you change the crystal, you have to change the electronics and you have to make the panel bigger. That means you have to build a bigger fab (fabrication process) for it. Then we need to put it on an ergonomic platform and incorporate the workflow. PACS has become so important in that interface.

We introduced our Senographe 2000DS product line as the platform of the future and invested in large-panel technology with this new crystal. We have initial panels out there. The DQE looks good and the initial images are looking good. Next up was the tomo piece, which Dr. Daniel Kopans has worked hard on at Massachusetts General Hospital. And now we have tomosynthesis out there in preliminary machines. The data are coming in, but it's not just the data on the tomo piece. We still have to learn how do to do the workflow; how to do CAD (computer-aided detection) with tomo; how to make sure this 3D information feeds into a PACS.

SCAN: Digital mammography equipment costs six times what film-based technology costs. I imagine the tomo equipment will cost even more.

Hogan: There's a financial crisis going on in mammography. Nobody is making any money. We can't launch a product that costs x millions of dollars that doesn't have a workflow piece to it, that's not efficient, that has no proven clinical efficacy.

SCAN: Let's talk about CT, not from a personalized medicine point of view but from a practical one. Will the 16-slice CT become the platform for making niche products?

Hogan: It could be the four-slice. That's a great machine for the price. The 16-slice costs more because it has more channels. But it also has more function. It could also be the 32, because 32 slices is the break point. If you want to make a scanner with more than 32 slices, you have to organize your detector technology differently and it is more expensive to do that. I can see all those machines being niched in certain ways. Sixty-four-slice will be our specific cardiology focus.

But I am still amazed at the number of four-slice systems we sell all around the world.

SCAN: But you're not selling them to radiologists.

Hogan: That's true - in the U.S. In China, we are selling them to radiologists. In Europe, we are selling them to radiologists. You have to look at the global marketplace. In the U.S., it's a different ballgame.

SCAN: Philips has come out with a 16-slice CT for private practice cardiologists. Will GE follow this lead?

Hogan: I don't think it is us following. I have a cardiology business unit. They deal strictly with cardiologists in private practice and hospitals. They channel my CTs into those marketplaces. If you want to just put a tag on a device and say it is for them, in the end it really doesn't mean anything. It is how you channel your sales force. I certainly sell to cardiologists. I am not bashful about it. I have been asked for the technology.

As we get into cardiac CT, we are working against time. You want to do the coronary CTA as fast as you can and with a slow heart beat. Some physicians administer beta blockers. With our LightSpeed VCT, you have one of the fastest ways in the world to image the heart - noninvasively. Whether that is private practice or hospital practice or whatever doesn't matter.

SCAN: So take this to the next logical step. We have 64-slice CT. It is a cardiac machine. But it also has radiological applications. The problem is that it is much more expensive than CTs have ever been. It breaks the $1 million dollar barrier. How do you overcome sticker shock?

Hogan: If you think about invasive angiography now in cardiac cath labs costing more than $5000 per study and you recognize that 25% to 50% of all cardiac cath procedures are nontherapeutic, there is a huge potential cost savings from channeling those patients to CT. So can you afford a $1.6 million piece of equipment in order to do noninvasive angio, because it a cost savings to the system? You are paying for technology that allows you to do something at a lower cost. We truly feel that our LightSpeed VCT - because of its speed and clinical capability, because of the size and specificity of the detector - is by far the best machine in the world.

SCAN: Where do you stand with CT flat panels?

Hogan: We have three in the field. We are gathering information. A lot of it is animal information. We have been working on flat panels for years, and we know what the barriers are for that.

SCAN: Is the flat panel your way of getting to volumetric CT?

Hogan: We see our LightSpeed VCT as a true volumetric scanner in regard to how much data it acquires. As you get into this 128-slice or whatever, you evolve into more of a uniform panel kind of thing. That could be the future of CT.

SCAN: How do you define this in terms of slices? Do you use 64 as a benchmark for volumetric scanning?

Hogan: You do unless you are Siemens. They have a 32-slice CT, but they call it a 64. Our 32 will compete with what they call a 64.

SCAN: Toshiba is saying they have a 256-slice prototype.

Hogan: It is a machine that the Japanese government helped them develop. At least they know what the demands will be for information coming off such a detector. But back to your question, how do you measure the future? I can't tell you. I don't know. We have taken panels from our x-ray business, basically cesium iodide technology, and generated some animal pictures. The definition you see is just incredible.

SCAN: Let's talk about word definitions for a moment. You have what you call "imagination breakthrough" products. The LightSpeed VCT is an imagination breakthrough. How do you decide whether a product is a breakthrough or not? Is it clinically based?

Hogan: Our breakthrough definition revolves around technology. It does not have to do with changing the clinical piece. So VCT would be a breakthrough because it changes the way we pull signals off the detector and because it sets a new platform for the future.

But that doesn't mean some of our breakthrough products will not change clinical practice.

SCAN: Do you have specific criteria for deciding what is an "imagination breakthrough" and what is not?

Hogan: I have to decide what has the most clinical efficacy and promotional potential that our sales team can use out there. And it's fun. It is a great way to package your technology. Internally, people get excited about it.

SCAN: Tell me about SPECT/CT from this perspective.

Hogan: My SPECT/CT attenuation correction product, which we launched five years ago, would have been an imagination breakthrough. It established a new platform in technology that didn't exist before.

SCAN: Why haven't you continued to evolve that platform?

Hogan: The SPECT/CT we have right now is a good product. We can evolve it to the 16- or 64-slice, whatever, but we would have diminishing returns. The major thing you want when you put a CT on top of a nuclear camera is to get the attenuation correction. That's where it really helps. Now, can we get a better anatomicl look with a true diagnostic CT? Sure. But I haven't had a lot of customers clamoring for the diagnostics piece. There is a huge tradeoff in cost and we have been listening to the marketplace.

SCAN: But by not evolving this platform, you have ceded leadership to Siemens and Philips.

Hogan: If you are a competitor this late to the game, you will not come out with just an attenuation correction machine. You will come out with a diagnostic CT. So I respect Siemens and Philips for what they have done, but it hasn't been based on an overwhelming customer push to get a diagnostics piece on this thing.

It's not hard to put a diagnostic CT on a nuclear scanner. If customers say they want both attenuation correction and diagnostics, we will give them both. That is part of the whole competitive thing. But I really think anything over a four-slice for bone and oncology is overkill. Nuclear is still a tough price point.