Daniel C. Sullivan, M.D. has spent his professional life in pictures — in medical images, that is.

Currently associate director for the Biomedical Imaging Program (BIP) in the Division of Cancer Treatment and Diagnosis of the National Cancer Institute (NCI), National Institutes of Health (NIH), Sullivan, a radiologist by training, has a 23-year history in academic radiology with appointments at Yale University School of Medicine (New Haven, Conn.), Duke University Medicine Center (Durham, N.C.) and the Presbyterian Medical Center, University of Pennsylvania Health Systems, (Philadelphia).

Sullivan, who brings his own clinical and research specialties in nuclear medicine and breast imaging to the associate director’s post, oversees the Program’s four branches — diagnostic imaging, molecular imaging, image-guided therapy and imaging technology development. He also administers the Program’s annual budget for grant applications, which has grown from approximately $59 million in FY1997 to $100 million-plus in FY2000. As Sullivan says, “There is a lot of opportunity now” in the biomedical imaging community due to advances in imaging technology together with recent molecular discoveries, such as the mapping of the human genome, that help scientists identify target areas of research.

Medical Imaging spoke with Sullivan about NCI’s Biomedical Imaging Program, its mission and the role of the imaging industry in advancing research and helping set Program priorities.

How long have you been associate director of the National Cancer Institute’s (NCI) Biomedical Imaging Program (BIP) and what does your role entail?
Three years. My role is to organize and administer the Biomedical Imaging Program, which develops programs to facilitate research on new imaging technologies and molecular imaging in particular; to stimulate areas that we think are important; and to administer the grants we think are important to medical imaging with a particular focus on cancer, but it can be more broad than cancer in some cases.

The Biomedical Imaging Program consists of four branches. What is each branch’s area of concentration?
One branch is the Diagnostic Imaging branch, which focuses on improving medical imaging as most people think of it — MRI, CT scans, mammography and ultrasound.

The Molecular Imaging branch focuses on one of the areas that we think is particularly new and important, which is to get imaging down to the molecular level, as science discovers what the abnormalities are that really cause disease at the genetic level and the molecular level. We would like techniques and technologies that can give us information about those abnormalities in the intact living human, so that physicians do not have to biopsy tissue, for example, and do laboratory tests on it. This is a whole new area for medical imaging. It’s a major challenge, and it will not happen next year; it is something that will take 5, 10 or 20 years to develop.

The Image-guided Therapy branch focuses on using imaging to guide small devices, catheters and more to do minimally invasive surgery or to deliver therapy specifically to a very localized area of the body in a minimally invasive way. That could be any type of therapy, whether it is a drug, heat, radiation, laser, surgery or biopsy, for example.

The fourth branch is the Imaging Technology Development branch, which focuses on fostering research on very new technologies that are not yet clinically useful but might have some promise 5, 10 or 20 years into the future.

How many employees are in the program?
The Program started four years ago and we now have about 16 people. Each branch has about three or four people. We anticipate adding a few more people this year, up to a total of about 20.

Does each branch have its own budget or is there one program budget for the four?
There is one overall budget for the program.

In fiscal year 2000, which ends on Sept. 30, our program will support a little more than $100 million of research, but we do not start out the year with a $100 million. The way the National Institutes of Health works, people send in applications on their own initiative or in response to some program we create. Those applications are reviewed by other experts — the peer review system — and the ones that get the highest scores will get funded. The total amount of grants that are funded during the year will depend partly upon the number of applications that are submitted and the relative merit of those applications. For the next fiscal year, starting in October, we will not start out with some predetermined budget; there will be a certain number of grants that will already be obligated, because they are multiple-year awards that have been funded, but the total amount that will be available at the end of the year is unknown. It depends on the above factors.

The amount of funding for imaging has increased significantly over the last three years from approximately $59 million in FY1997 to more than $100 million in FY2000.

To what do you attribute that funding increase?
It is a combination of encouraging investigators to submit applications in the areas we think are important and also creating programs that, in some cases, do have set-aside money for applications that come in. For example, we created a program to ask for applications for molecular-imaging centers, and last year we funded three of those centers at $2 million each for five years. That is $6 million a year for those centers. In the next fiscal year, we will add two more centers at $2 million each, so next year that will be $10 million for those centers. That program will continue to increase in future years.

We have another program for small-animal imaging centers because of the importance of genetically engineered mice in research nowadays; there’s a need to get biological information from those mice while still keeping them alive. There are five of those centers now and they are funded at the rate of approximately $1 million a year each. We have issued another request for applications that is currently available, so people will be sending in applications and next year we will fund another five, so there will be a total of 10.

Those are just two examples. There are many other requests and program announcements that we have issued in the last two or three years that generate more applications and, in some cases, have specific set-aside money. So both of those things increase the total amount of dollars spent for imaging research.

From where does your budget come? Are grants distributed throughout the United States?
It all comes from the NCI budget. In the last few years, Congress has been favorably inclined to increase the overall budget at NIH in general and at NCI in particular, so there has been more money available for special programs. All of the BIP money comes from NCI.

Grants administered can be anywhere in the United States, and actually some are in Canada and in foreign countries, so they could be all over the world, but the majority are within the United States.

Congress decided that in some cases the public health benefit to the population of the United States is well-served if there is expertise in other countries to do particular research; for example, if they do a particular study that is either not being done in the United States or is not available in the United States. So a small number of grants are made to researchers in other parts of the world — not a huge percentage, and there has to be some justification for doing it.

In our case, we are funding a study in Canada looking at the mammography of women with dense breast tissue to see how much of a risk factor that is, and we are funding a couple of studies in the United Kingdom and Germany that are related to looking at MRI of breast cancer or MRI screening of women at high-risk. In all those cases, we also are funding studies in the United States related to the same topics, but in order to get more patients and more data, some of it is being done in other countries as well.

What role do imaging companies play in your research? Are they involved in your decisions to investigate a particular project by being involved in the those initial discussions or by providing you with equipment or technical support?
All of those things. In workshops we generally include representatives from industry. Last year, we started a specific meeting — the NCI-Industry Forum — to exchange ideas about what we at NCI think is important and we can hear from industry what they think is important and what the problems are in solving those important issues. The second national forum was held last in September. So, we have some interaction specifically with industry, and we include them in other workshops.

Some of our program announcements specifically request that academic researchers form partnerships with industry, and sometimes we sponsor trials of new equipment. We are now beginning to sponsor a trial of digital mammography that will test the equipment from four different manufacturers. They were all involved with us in discussions about what the protocol should be, they will be providing the equipment and they discussed with us the subsequent use of the archive of images that will be produced from that clinical trial. So we do have an increasing amount of discussions with industry.

As part of that, do you have to be sensitive to questions about conflict of interest or ethical concerns? If so, how do you deal with that?
We work very hard to maintain fairness and equity so that all researchers have the same access to us and the information that we provide. At the same time, we provide confidentiality about any information that industry provides to us. Sometimes that is just a verbal agreement; industry has good confidence in our ability to maintain confidentiality. Sometimes they ask us to sign nondisclosure statements and agreements of confidentiality and we are always willing to do that, so that we can discuss with them what they are doing and they can tell us what they are doing and how we can help.

How does industry’s participation affect your choice of projects?
We value industry’s opinion about what things are going to be clinically useful and what things have market value. They pay a lot of attention to those issues and companies that are successful have obviously been making successful decisions about that, so we pay attention to the things that their marketing people and research people have decided will have clinical value, market value.

On the other hand, there might be some things that we know, or we think are important to the public health that might not have significant market value.

Can you cite an example of something that industry deemed to have little market value, but the Biomedical Imaging Program pursued nonetheless?
In the last couple of years in our Program there have not been any examples I can give you. However, in past years, NCI, in the Clinical Trials Program, has developed some chemotherapeutic compounds that industry has thought would not have a large enough market. They are called “orphan” drugs because nobody wants to put them on the market; the market is not big enough.

We anticipate as we develop molecular-imaging agents, some very specific molecules that could target a cancer cell, for example, the market for some of those agents may be very small. If an agent is very specific for a particular type of tumor, and it is going to be used only once or twice in a patient for diagnostic purposes as opposed to a drug that gets used over and over again, the market may be small. Therefore, companies may not see it as a profitable venture, and we anticipate that NCI will have to develop those agents for the public health good.

Digital mammography is a bit of an example of that. The companies have developed the machines and they think there is a market for those machines, but the benefit of digital mammography over conventional mammography in detecting more cancers and providing fewer false-positives is probably relatively small. There probably is some advantage, but it is not huge. To carry out a large trial to prove what that benefit is would be prohibitively expensive for any one company. That’s one of the reasons why NCI is working with the companies to carry out a very large trial that would benefit all companies, so that we can identify what the benefit is and how big it is. The public health interest is that women will then know what the benefit is and can decide whether they want to ask for a digital mammogram as opposed to a conventional mammogram.

Do you find that certain projects, such as mammography, are more popular one year than the next. If so, what might some of those be?
There is some variability in what researchers are interested in. They probably respond to what they think the public is interested in and sometimes the funding agencies like NIH or the Department of Defense put emphasis on one area or another. This is partly because Congress requests that the NIH or other agencies do that, and Congress is usually responding to what they perceive to be the public interest or the public will. So, breast cancer advocates sometimes would lobby and generate a lot of support from Congress so that they (Congress) would ask NIH and the Department of Defense to put more emphasis on that area and other disease-type advocacy groups would do the same thing.

What are some other BIP current projects?
One that has been getting some publicity lately is spiral CT for lung cancer. We are funding a similar study at the Mayo Clinic to get more, similar data, and we will probably within the next year start funding a larger, randomized trial at multiple sites across the United States to evaluate whether early detection with spiral CT leads to decreased mortality from lung cancer.

Is the Biomedical Imaging Program in any way involved in the human genome project? In mapping, perhaps?
Not exactly in the mapping, but in trying to develop imaging techniques to identify gene abnormalities. I mentioned that we are funding three molecular imaging centers right now: One is at Memorial Sloan Kettering, one is at the University of California Los Angeles and one is at Massachusetts General Hospital. All are working on developing imaging techniques that can identify the expression products of normal or abnormal genes. This is still very early in development, but the idea is to develop techniques that will give us a way to identify what genes are functionally normally or abnormally in a tumor or another organ in the body without having to take a tissue sample.

The work is mainly related to diagnostic imaging, but in the process of developing an imaging probe, some kind of molecule that can go to a specific target in the body, which is another molecule. It is possible to also attach some kind of treatment substance to that molecule, the target. For example, you could attach a radionuclide molecule that would destroy the cell by radioactivity or you could attach some kind of a drug to that molecule to get the drug specifically into the cell or into the DNA nucleus that you want to kill. First you would give the imaging probe to identify the abnormality and make sure the probe is going where you want it, and then you would switch to another version of that probe that would have the therapeutic modality attached to it and give that to the patient.

How have technological advances in imaging impacted your research?
There is a lot of opportunity now. Part of the reason that NCI developed this imaging program and decided to hire more staff and put more money into it is because we see a big array of opportunities in terms of imaging because of new technology. There are actually three reasons why imaging is so important now in the thinking of people at NCI.

One is because all of these molecular discoveries, the mapping of the genome and understanding genetic abnormalities have identified the targets that we need to go after.

Second has been the developments in imaging technology itself, that is, devices like the MR scanners and CT scanners that have very high spatial resolution — that means you can see very small structures.

And third is the development of new technology in chemistry, what is usually called combinatorial chemistry, the ability to create hundreds or thousands of different chemical compounds very quickly and to screen those different compounds quickly to determine which ones are likely to be useful or active. The combination of those three things, two of which are technology-dependent, are what has opened lots of vistas for medical imaging.

Do those three also help set the trends in biomedical imaging?
Yes, certainly the first one — understanding what the abnormalities are that we want to go after and we want to find. Certainly, that sets the direction for medical research.

What do you see for the future of the Biomedical Imaging Program?
I think we will continue to see increases in the amount of funding over the next few years and particularly an increase in the focus and emphasis on the types of things we describe as molecular imaging, that is, developing molecular probes that are really specific for molecular abnormalities, to identify those in living patients. That is what we think is the growth area in the next 5, 10 to 20 years.