October 6, 2008

When Is a Product a Product?
By Kathy Hardy
Radiology Today
Vol. 9 No. 20 P. 14

Not long after the spouse and kids have piled into the family roadster and you’re finally rolling toward your vacation destination, someone almost always impatiently asks, “Are we there yet?”

That same anxiety occurs during the technology development journey from an intriguing idea to a product in the imaging marketplace. Looking at the formalities in the development process, new devices are nominally commercially viable once the FDA grants either new device or 510(k) approval affirming that the device is safe and effective. However, much more than regulatory approval goes into developing and successfully launching a new technology to the imaging marketplace. For Carolyn Meltzer, MD, FACR, a professor of radiology, neurology, and psychiatry at the Emory University School of Medicine in Atlanta and the chair of the radiology department, research teams reach their destination when there is sufficient clinical evidence that the product adds value to the care process.

“Sometimes it’s a judgment call,” she says, “but there also needs to be some published evidence that the data is accepted by the medical community. There must be some medical or scientific acceptability that the new technology is better than existing technology and adds value.”

Meltzer’s experience in the area of new technology development comes first hand. She participated in the clinical evaluation of the first combined PET/CT scanner while serving as codirector of the University of Pittsburgh’s PET facility. Combining PET and CT evolved out of surgeons’ frustrations when attempting to match CT scans and PET scans to determine the location of tumors prior to surgery, according to Meltzer. Researchers found that the sensitivity of CT’s anatomical images combined with the functional specificity of PET revealed the location, size, and shape of malignant tissue.

“With PET/CT, we estimated the added value would be that it would help in 20% of the cases,” she says. “Studies showed that it did add value. The specificity was high with CT, the sensitivity was high with PET, and we found the 20% but didn’t know which ones.”

The FDA approved the marketing of a combined PET/CT machine and Time magazine named it the Medical Invention of the Year 2000.

Jeffrey C. Weinreb, MD, FACR, a professor of diagnostic radiology and chief of MRI at the Yale University School of Medicine in New Haven, Conn., is a witness to significant changes within imaging technology in the past 28 years. In his earliest experiences, CT and real-time ultrasound were new, but MRI and PET did not yet exist, he says.

Weinreb breaks new product development into three stages: discovery, diffusion, and maturation. He also speaks to the source of new ideas and where they best percolate.

“Ideas rarely come from the manufacturer side,” he says. “The manufacturer plays a huge role but rarely is that the place where the ideas start. They usually come from a scientist or clinician who sees a need or opportunity. Sometimes people don’t perceive a need but see an opportunity to do something better or more profitably.”

He uses MRI development as an example of this. “I don’t know that anyone saw a need for MRI,” he says. “A number of people realized the limitations of CT and thought they could do better.”

Across the board, all parties believe teams work best. While Meltzer stresses the importance of clinical involvement in the development and acceptance of new technology, her experience speaks to the benefit of working in a group environment.

“More often than not, the idea comes from an academic department with research supported by internal or external funding,” she says. “At some point, you need to build it and get it out to test. It might be the scientist who has the idea, but he doesn’t know if it’s clinically useful. Maybe it was a brilliant idea but he didn’t have partnerships. He was thinking in a vacuum. Clinicians are busy taking care of patients and don’t have the time to get involved. Creating imaging technology takes large teams. You need to work together and interact. The more people involved, the more ideas that come out of the process. When scientifically oriented folks get together, usually useful things come out of it.”

Weinreb sees the lack of combining efforts as a roadblock to successful product development. “Scientists and people who conduct research often work in a vacuum and separate from or not well informed about clinical applications,” he says. “Likewise, clinicians are not always aware of the science. They often have different perceptions or expectations. Sometimes that causes conflicts.”

Those conflicts can impede progress, Meltzer adds. “Someone has to be a leader to move the project along,” she says. “As long as you have a champion of the research you can be very effective. You rarely see imaging papers with just a few authors.”

Teams also play a vital role from the manufacturer’s perspective. Kulin Hemani, vice president of CT at Siemens Healthcare, agrees that the clinical point of view is key to successful product development. Siemens’ product development begins with input from clinical user groups, he says, adding that everyone needs to work together to succeed.

“You have race car builders and race car drivers,” he says. “Each has their skills, and they all need to work together. You have to admit that you need both. Any good development team will have good technical experts and good clinical experts.”

Clinical input is vital to product development success, Hemani says. “We work with high-profile clinics involved in the healthcare fields for product development,” he says. “From interaction with them, we devise a series of clinical requirements they want to resolve. The priority is cost and deliverability. You must answer a clinical question or need to succeed.”

Product development is process oriented for the German-based Siemens Healthcare, Hemani explains, and based on two criteria: “We want a product that fulfills a clinical need and a product with the highest value,” he says.

New product prototypes are made in Siemens’ labs, Hemani says, where they translate clinical requests into technology concepts. Following a series of feasibility studies, the product prototype is delivered to the same user group that initially provided input on the product. Further enhancements result in new prototypes, Hemani says, improving stability and performance to reach the point where there’s a final product. That version of the product is then submitted to three to six months of clinical research, just prior to submission for FDA approval. Once the FDA approves the product for marketing to the public, high-profile clinical sites receive the product for further clinical evaluation. In all, the process takes two years to complete.

“This is a lengthy process, but we find it necessary to produce a sustainable product,” Hemani says.

He says the product is done, for all intents and purposes, once it is about six months away from being released. At that point, the specifications are built into the product. “There might be some tweaking, but no more specifications would be added at that point,” he says.

If a product won’t make it through to final production, there’s usually evidence of that within the first three months of the process, Hemani says. Key factors that are considered when determining a product’s feasibility early on are whether it’s too expensive to produce or there is insufficient technology to move the product to the next level.

“Some concepts are lost; others are just delayed until some new technology comes along or the cost decreases,” he says.

Software Development
New imaging software development often takes a slightly different route to market than hardware. According to Dan Bickford, executive vice president of sales and marketing with Confirma Inc, located in Bellevue, Wash., new software development historically comes from research. Once a concept is created and licensed, the company launches in “start-up mode.” The process includes raising capital to fund clinical research and papers, while at the same time setting up studies necessary for FDA approval. Once FDA approval is obtained and clinical data is made available, the software is marketed to physicians.

“It would take three years to get new technology from the ‘center of excellence’ stage to the general population,” Bickford says.

Confirma’s computer-enhancement imaging tool, CADstream, which adds a level of specificity to MRI screening with 3D renderings and data calculations, was in development for only nine months before Confirma released the product. Bickford says this shortened time frame was possible because the company followed a different development process. CADstream was introduced at RSNA 2002.

“Rather than start from scratch, we looked for intellectual property that would handle postprocessing,” Bickford says. “We found a need in the market, then looked for technology to license, then found funding to start the company. It’s the bottom up approach to development.”

Bickford says that spending time with members of the medical community enables product developers to find out what tools radiologists need to provide better patient services and ultimately, better patient care.

“Confirma went to radiologists and asked what was needed in breast MRI,” he says. “We learned that workflow was an obstacle for a test as sensitive as MRI. By going to the community first to determine the specifications for the product, you’re bypassing the three years of research to find out what the users already know.”

Clinicians and manufacturers agree that a variety of influences inside and outside the research team can ultimately alter the course of new technology development.

“The goal often deviates,” Meltzer says. “The research defines us. You try not to have a preconceived bias going into the study. That’s why I have never been associated with a company. My hypothesis is based on whether I think it will add value. If you are too heavily influenced financially or emotionally, it can influence the decision-making process. You need to be totally objective.”

Meltzer also notes a common phenomenon in the development process is that initial studies are conducted with small groups of very defined individuals. Those studies present findings that pertain to that particular group. But as the technology goes out and is used with a more complex population of patients, results may “not be as wonderful as had been hoped.”

“The first papers out on a new procedure are good, but once the new technology is in use, the numbers drop,” she says. “It’s harder to show specific ideal applications if the initial tests aren’t more specific. These studies don’t reflect the general patient population.”

Announcements from within the medical community can also alter the course of product development. The release of standards by the American Cancer Society for women at high risk of breast cancer who use MRI screening along with mammography created a heightened awareness of the use of MRI for that purpose. For example, CADstream version 5, currently pending FDA marketing clearance, includes enhancements that could be necessary as more women undergo breast MRI, Bickford says.

Reimbursement is another large factor in the decision-making process of when new technology is ready for use in the mainstream medical community. Weinreb notes that without peer review saying that a new product shows benefits, reimbursement will be increasingly difficult to achieve.

“Economy has a big impact on new technology,” he says. “If the end user doesn’t have money, they won’t use or approve it.”

— Kathy Hardy is a freelance writer based in Phoenixville, Pa., and a frequent contributor to Radiology Today.


Funding’s Role in Technology Development
To say funding plays a role in the decision-making process for the development, completion, and release of new imaging technology is an understatement. From corporate investments to federal grants, how much or how little money is available for research can affect product release outcomes.

“Vendors are under pressure to optimize cost and preference, as well as risk to the patient,” says Kulin Hemani, vice president of CT at Siemens Healthcare. “Cost optimization is a concern. I don’t believe the market can afford $2 million scanners. We have to live within the boundaries of what is affordable.”

Funding for research projects, particularly grants provided by the National Institutes of Health (NIH), is tight right now, says Carolyn Meltzer, MD, FACR, a professor and the chair of the radiology department at Emory University School of Medicine.

“Statistics show that the number of grants is not down, but grants are being funded at drastically lower levels,” she says. “We can’t do what needs to be done.”

Meltzer refers to legislation introduced July 16, 2008, by senators Tom Harkin (D-Iowa) and Arlen Specter (R-Pa.) to provide an additional $5.2 billion to the NIH for fiscal year 2008. The requested supplement is an attempt to reestablish NIH funding to levels consistent with inflation, as well as to provide the institutes with sufficient resources to continue funding research projects. The bill calls for an allocation of $1.2 billion for the National Cancer Institute and $4 billion to other NIH institutes. This legislation follows the Senate Appropriations Committee’s approval of an $875 million increase in NIH funding for fiscal year 2009; President George W. Bush’s request for 2009 is $29.229 billion, a decrease of $150 million from the previous year.

Jeffrey Weinreb, MD, FACR, a professor of diagnostic radiology and chief of MRI at the Yale University School of Medicine, reinforces the importance of clinical involvement in new product research but notes that scientific and medical evidence of a product’s viability isn’t always enough. “Manufacturers are not going to support something unless they see a profit at the end,” he says.

With funding availability an issue, Weinreb says that’s when disruptive technology can play a role. He cites Harvard professor Clayton M. Christensen, PhD, for defining this term as “not just anything new or different, but something that changes the existing business model.”

In contrast, sustaining technology involves products with highly improved new features that enhance product performance and replace existing technology.

“Disruptive technology is not always easy to recognize,” Weinreb says. “It’s simpler and cheaper than existing technology. The vast majority of people say, ‘This will never work.’”

Established companies usually don’t like disruptive technology, he says, due to the smaller market and lower profit margins that are initially available. “They often do not want to deal with it until the technology is further developed, and they prefer to push the existing technology so that its performance stays above what the disruptive technology can achieve,” he says.

Weinreb cites PACS as an example of a disruptive technology. “PACS changes the whole business model,” he says, “and that’s why it’s disruptive.” This doesn’t mean the technology is disruptive for the medical community. It just offers the opportunity to dramatically change the medical imaging business model.

Conversely, he says, MRI’s progression from development to introduction to the marketplace is an example of the more popular sustaining technology.

“It’s still imaging, the business model is still the same, but it’s a technology that didn’t exist 30 years ago and it’s very fast growing,” he says. “The term ‘sustaining technology’ does not imply that the technology is better or worse than a disruptive one. Breast MRI and multidetector CT are often cited by advocates as disruptive technologies. As exciting and valuable as they are, they are in fact excellent examples of sustaining or evolutionary technologies.”

Weinreb adds, “It’s very rare that something new just pops up in a year. When we hear about a new technology it’s usually been in the works for years.”

— KH