Multiple factors foster first-rate radiology research
Multiple factors foster first-rate radiology research
Great radiology schools are associated with extraordinary universities that honor a commitment to basic scientific research. Many medical breakthroughs occur in this environment, but many also arise elsewhere. The catalysts for excellent research vary widely but include geography, strategic alliances, social structures, and rogue visionaries. This report is not intended to be a comprehensive list of great radiology research institutions and people. But it will review how deliberate-and sometimes random-factors influence imaging across the globe.
Arguably, the most important aspect of a quality research center is leadership, and radiology is gifted with many excellent leaders. As a resident in nuclear medicine at the University of Michigan under Dr. William H. Beierwaltes, Dr. James H. Thrall learned two fundamental lessons. First, research grows proportionally with the number and quality of principal investigators. Second, departmental and institutional support for major facilities and equipment is essential.
Thrall has led the radiology department at Massachusetts General Hospital for 15 years. He subscribes to the "virtual cycle of research" concept. It starts with investing in facilities, which then allows leadership to recruit creative individuals. The principal investigator takes advantage of the startup funds and the facility to compete for extramural funding. Successful extramural funding gives the institution confidence in the next round of recruiting.
During Thrall's tenure, MGH radiology's annual budget has grown from $3 million to $46 million, and the number of principal investigators has increased 10-fold to 85. Nine are radiologists who are guaranteed at least 50% of their time for research. Thrall credits those who preceded him, particularly Dr. Juan Taveras, who led the department for 17 years in the 1970s and '80s, for establishing the foundations of a great research institution.
MGH is an institution that bets on itself, Thrall said. When the hospital had no funds for a new cyclotron, the department bought one through its professional practice monies. Within a year, profits from the production of radiopharmaceuticals offset lease payments. Additionally, when constructing the MRI research program at the Charlestown Navy Yard, the hospital had regulatory approval and financial backing for only one MR scanner. Thrall persuaded the administration to build three fully shielded, fully ready imaging bays. Less than a year later, a partnership was forged with GE Medical Systems that led to the installation of the first hospital-based echo-planar MR scanner, which in turn generated the first clinical functional MRI studies. Today, the NMR lab is largest institution at Mass General.
In the world of elite centers with long histories, MGH does not stand alone. The University of Pennsylvania Medical Center's contributions to radiology span 100 years, from Dr. Henry Pancoast's appointment as the nation's first professor of roentgenology to recent developments in optical imaging. The department continues to reflect the influence of Dr. Eugene P. Pendergrass, who as chair from 1933 to 1960, embraced the newest imaging technologies and recruited basic scientists to collaborate with radiology researchers. Dr. Stanley Baum, chair from 1975 to 1999, added to the mix a staunch support of subspecialization and aggressive pursuit of National Institutes of Health grants.
Over the years, the department's scientists helped establish the principles of neuroradiology and interventional radiology. Penn researchers contributed to the invention of SPECT imaging and laid the groundwork for PET. The synthesis of fluorine-18 FDG arose from a collaboration between Penn and Brookhaven National Laboratories.
Today, Penn ranks among the top three recipients of NIH grants in radiology, having received $17 million last year to support 12 laboratories and 15 basic scientists. Ongoing research includes the quantitative characterization of tissue microarchitecture and its relationship to physiology and function, virtual bone biopsy techniques that noninvasively profile bone mineral characteristics, and MRI lung ventilation applications using hyperpolarized helium.
A new five-year research residency program exemplifies the department's scientific emphasis. Two of this year's nine new residents will train to become imaging scientists. They will receive conventional radiology residency training for two years, research and biostatistics training during the third year, and subspecialty training in year four, followed by mandatory fellowship training.
In Korea, Dr. Myung-Chul Lee has worked for 20 years to promote nuclear medicine research and education. He installed the first PET center in Seoul National University Hospital in 1994 and established the Korean Board of Nuclear Medicine in 1995. Largely through efforts such as these, Korea had the fourth largest number of papers presented at the annual Society of Nuclear Medicine meeting in 2000 and 2001. Korean investigators presented 85 papers at last year's SNM meeting in Los Angeles and 94 this year in New Orleans, nearly half involving PET.
Lee became the president of the World Federation of Nuclear Medicine and Biology three years ago (Korea will host the next WFNMB meeting in 2006). Last year, he was named chair of the Asian Regional Cooperative Council for Nuclear Medicine, an organization that promotes nuclear medicine and educates physicians in underserved areas of Asia. Lee credits his time at Johns Hopkins under Dr. Henry Wagner as having fueled his desire to educate.
His own research interests involve PET and neuropsychiatric diseases. At Seoul National University Hospital, FDG-PET preoperative imaging of 500 epilepsy patients has contributed to a cure rate of 87%. Lee and colleagues have also demonstrated the prognostic importance of auditory-to-visual cross-modal plasticity using PET in prelingual deaf patients with cochlear implants. They found that the degree of preoperative hypometabolism was related to the extent of improved hearing capability after cochlear implantation.
In many parts of the world, research arises from indigenous disease states. When the leading cause of death in Japan shifted from communicable diseases to cancer during the 1980s, the Japanese government launched the first of two 10-year research programs aimed at controlling neoplastic disease in 1983. In the first decade, researchers discovered oncogenes and tumor suppressor genes and demonstrated the concept of multistage carcinogenesis, said Dr. Tadao Kakizoe, director of the National Cancer Center Hospital. During the second program, which concludes next year, researchers identified and characterized a novel tumor suppressor gene for non-small cell lung cancer and detected a molecular signature of early hepatocellular carcinoma.
Additionally, the Japanese National Institute of Radiological Sciences (NIRS) installed a heavy ion medical accelerator in Chiba in 1994. It is the world's first heavy ion accelerator complex dedicated to medical use in a hospital environment, according to Dr. Yasuhito Sasaki, president of the NIRS. Heavy ion therapy has advantages over megavoltage x-ray therapy in terms of safely delivering high doses along with increased cell-killing ability. Results of a clinical trial of carbon beam radiotherapy are promising. Two-year local control ranges from 44% to 100%, and three-year survival rates vary from 40% to 97 %, depending on the protocols for particular cancer sites.
"It is noteworthy that good results were obtained in patients with malignant tumors such as melanoma, bone and soft-tissue sarcoma, and various adenomas, which are resistant to conventional radiotherapy," Sasaki said.
Southern Europe has a significantly higher incidence of hepatocellular carcinoma compared with its northern and eastern neighbors. These patients often have hepatitis and/or cirrhosis, making them unlikely candidates for resection. Consequently, Italy has become the hub for percutaneous ablative therapy for HCC, with more than 100 centers performing the procedure. Interventionalists initially used ethanol injection, but a prospective randomized trial of 104 HCC patients with cirrhosis proved the superiority of radio-frequency ablation.
Investigators have further analyzed that 2001 study to assess long-term survival outcomes. The results are encouraging, said lead investigator Dr. Riccardo Lencioni, an interventional radiologist at the University of Pisa. Complete tumor response was achieved in 91% of HCC nodules treated with RFA, with an average of 1.1 treatment sessions, while an average of 5.4 ethanol treatment sessions produced 82% complete tumor response.
The new study appears in the July 2003 issue of Radiology. Although the researchers could not demonstrate a statistically significant difference in survival rates, a trend toward increased survival in the RF group is obvious, Lencioni said.
Today, he and a handful of researchers are recording favorable results using RFA to treat lung nodules. In contrast to liver patients, nonsurgical lung candidates have no other ablative options.
"In this regard, we are giving hope to patients where there is none," he said.
American trypanosomiasis, or Chagas disease, is directly responsible for a huge number of myocardial fibrosis cases, mostly in northern South America, Central America, and Mexico. Most patients carry Trypanosoma cruzi, the blood-borne parasite that causes the disease, for years or even decades before developing heart disease. But Brazilian researchers have made inroads into early detection using myocardial delayed-enhancement MRI.
Dr. Carlos E. Rochitte, a cardiac MRI laboratory researcher at the University of Sao Paulo Medical School, and his colleagues examined 25 patients known to have Chagas disease, but without clinical history of myocardial infarction. The MR technique identified myocardial fibrosis in 22 patients, mostly located in the apical and basal inferolateral left ventricular segments. Injuries were also found in areas not typically associated with the disease, such as midwall and subepicardial fibrosis.
"Most important was the early detection of micromyocardial injuries that otherwise would not have been found with currently available imaging methods," Rochitte said.
ALLIANCES WITH INDUSTRY
Many centers enter into research agreements with major manufacturers. Such relationships generally benefit both parties: Universities get first crack at cutting-edge technology, which they can tweak without government interference, and manufacturers receive instant feedback that helps push the final product to market more quickly.
"It is not only important to have a good relationship with industry, but it is mandatory for good innovative research in general," said Willi Kalender, Ph.D., director of the Institute of Medical Physics at the Friedrich-Alexander University in Erlangen-Nurnberg, Germany.
Kalender is familiar with the intricacies of research. After 15 years conducting research on CT for Siemens Medical Solutions, he then transferred to Friedrich-Alexander where he continued to build on his strong relationship with the manufacturer. Kalender assembled a world-class team of scientists who in the mid-1990s refined CT's ability to perform cardiac imaging. They worked through several generations of scanners, publishing a seminal paper on cardiac spiral CT in Medical Physics in 1998.
Around the same time, Siemens began to focus on reducing CT radiation exposure. With Kalender's guidance, the Erlangen team developed anatomy-dependent tube current modulations that adapt exposure online in real-time, independent of scout images.
"The radiologist or technologist should not have to know anything about optimal parameter settings," Kalender said.
His vision is to create a system for which operators can specify the image quality, such as isotropic spatial resolution of 0.8 mm and a noise level of 20 Hounsfield units, and let the scanner do the rest.
Germany is no stranger to CT radiation concerns. In 1996, Dr. Mathias Prokop and Dr. Michael Galanski published a book that in part detailed how to reduce dose while maintaining image quality. An English version of the work, Spiral and Multislice Computed Tomography of the Body, (Thieme Medical Publishers), is now available and includes updated material on multidetector technology.
With the major manufacturers working to reduce dose, the next challenge is to determine the optimal image quality level needed for particular clinical questions, Prokop said. Imaging pulmonary embolism, for example, does not require high resolution, but checking for pancreatic cancer does.
"These are just two extremes, but for both we don't exactly know the image quality level that we need," said Prokop, who had ties to Siemens until his recent move to the University of Utrecht in the Netherlands. "Determining this image quality level is the goal for the next couple of years."
Long-standing relations between Toshiba Medical Systems and top Japanese medical researchers have led to numerous innovations. The vision behind Toshiba's multidetector CT product development, for example, stems from collaborations between the company's engineers and Dr. Kazurhiro Katada, radiology chair at Fujita Health University in Aichi.
Based on recommendations from Katada and other academic collaborators in Japan, Toshiba engineers canceled plans for an eight-slice scanner and expedited work on its 16-slice machine. Katada then worked with the engineering team to cut the scanner's camera rotation time from 700 to 400 msec. Because of the prevalence of cardiovascular disease in the U.S., Toshiba enlisted Johns Hopkins University cardiology researchers Dr. Joao A.C. Lima and Dr. Edward Shapiro to explore the effect of the higher temporal resolution possible with a 400-msec cycle time on the quality of MDCT coronary artery imaging.
PUSHING THE ENVELOPE
People with vision stretch the capabilities of available resources in order to bring something entirely new into existence. This is the case in Minneapolis, where the University of Minnesota Center for Magnetic Resonance Research joined the NIH and the University of Alabama-Birmingham as the first sites for 4T whole-body MRI research in the early 1990s. Although Ohio State University built an 8T scanner in 1998, the University of Minnesota developed the scientific expertise needed to make the most of the first 7T placed in operation in 2002.
With these ultrahigh-field devices and blood oxygen level-dependent imaging, the center's founder and director Kamil Ugurbil, Ph.D., and his staff have mapped the functioning brain with unprecedented precision and reproducibility. Ugurbil's group has produced maps at 7T showing the organization and behavior of neurons in the human brain at crisp submillimeter resolution. Its work is beginning to reveal the tonotopic organization of the auditory cortex, which shows the differences in the neuronal activity associated with different frequencies of sound. The group's discoveries are also correcting misconceptions about the physiology of face recognition in humans.
The center has literally become a magnet for top researchers like Wei Chen, Ph.D., and Michael Garwood, Ph.D. Chen is quantifying the brain's metabolic response to neuronal activity, while Garwood unearths cancer's functional and molecular properties. It is the center's high-powered instruments that are drawing such talented people to Minnesota, Ugurbil said.
"Most of the time, people don't come to Minneapolis for the winters," he said.
One man's vision in Baltimore 10 years ago has propelled most of the nation's radiology departments and practices toward becoming totally filmless and paperless. Although the benefits of PACS are widely accepted today, that wasn't the case when Dr. Eliot Siegel, now chief of radiology and nuclear medicine at the VA Maryland Health Care System, first persuaded the Baltimore VA Medical Center to fund his vision. Back then, the radiologist shortage was not so acute, and many feared that the efficiency of a filmless environment would threaten their jobs. As it turns out, PACS perfectly complements today's understaffed and overworked radiology departments.
Siegel went beyond installing new computers, however. Along with Dr. Bruce Reiner, director of radiology research at the VA Maryland Health Care System, he has turned his department into a petri dish, documenting productivity, efficiency, and cost-effectiveness. The two men complement each other well: While Siegel focuses on technical aspects, Reiner studies the practicality of such gadgets. They are currently developing an electronic auditing tool that records how a technologist or radiologist views images. Such data might reveal patterns that vendors could use to modify equipment in ways that reduce viewing time.
Another project springs from the demands of the newest generation of radiologists and residents, who are much more computer-savvy than their predecessors. Siegel and Reiner are designing the workstation of the future, which will give radiologists instant access to current and old images and medical records, as well as the capability to navigate through data sets and manipulate images in 2D, 3D, or cine mode.
"That level of sophisticated functionality is what the new generation intuitively expects," Siegel said.
Dr. Gustav von Schulthess, director of nuclear medicine and codirector of the MRI Center at University Hospital Zurich in Switzerland, is well known for his pioneering work in functional imaging in a multimodality environment. Von Schulthess has been pursuing research into combining PET with morphological image data for almost 10 years. In March 2001, University Hospital Zurich took delivery of the first clinical PET/CT scanner to reach the market. Researchers found that the addition of x-ray data to the nuclear medicine exams suddenly revealed much about PET images that previously hadn't been understood.
"Before we started using the PET/CT scanner, nobody knew that fat in the neck area can accumulate FDG," von Schulthess said. "For the past 10 years, we had been saying that all the FDG uptake in the neck was due to muscle, but it's in part fat tissue."
Beyond tumor and infection imaging, the emergence of combined scanners with high-end CT technology could prove particularly useful for applications involving vascular imaging, according to von Schulthess. Surgeons planning a tumor resection, for example, often need to see the anatomy of nearby blood vessels.
The addition of 16-slice CT could also help realize the potential of PET/CT in developing comprehensive cardiac exams. In a study presented at this year's European Congress of Radiology in Vienna, members of von Schulthess's group used a combined scanner to acquire CT images of the coronary arteries and to obtain stress and rest myocardial perfusion data using PET.
"If you pursue this avenue, obviously you want to have the best CT scanner you can have. You could say that even 16-slice CT is not good enough for this application," he said. "I really believe that the one-stop-shop cardiac PET/CT examination could be the future."
In France, Dr. Afshin Gangi is changing the way interventional radiologists treat lesions. Gangi uses a new bipolar RF ablation method incorporating two probes with perfused needle electrodes. The RF energy heats up between the probes, creating larger and more controllable lesions in half the time as monopolar probes.
He got the idea from surgeons who use bipolar knives to close vessels. At first he merely inserted two probes, but the German company Berchtold now markets a bipolar instrument. The technique is promising for kidney, spine, and difficult liver lesions, according to Gangi, a nonvascular interventional radiologist at the University Hospital of Strasbourg.
Gangi was one of the first interventionalists to use RFA for bone metastases. The treatment is mainly palliative, but in certain cases such as bone metastases of thyroid cancer, Gangi performs curative ablation. He convinced his surgeons that RFA is safer for these hypervascular tumors. He destroys more than 90% of the lesion and completes the ablation of residual tumor with iodine-131 therapy. In the vertebral body, he has treated tumor on one side with RFA and then shored up weak bone on the other side with vertebroplasty.
Some countries have an excellent infrastructure that contributes to a particular area of research expertise. Sweden, for instance, is noted for its mammography studies. The entire screening system is computerized. Letters inviting women to participate in particular screening studies are automatically generated from large databases. It helps that screening compliance runs about 75% in large cities and 90% in rural areas, said Dr. Edward Azavedo, director of mammography at Karolinska Hospital in Stockholm. Forms with results are automatically logged. An electronic file containing results of the entire day's screening exams of about 150 women is sent to the postal service, where it is printed and sent-with results-to the women.
Swedish researchers have begun to evaluate many new technologies. Trials are ongoing comparing digital with conventional mammography, and computer-aided detection with conventional mammography. Researchers are also studying CAD to determine how many more cancers and the types of cancer it detects. They hope to use these data as a teaching aid for young radiologists. Studies are also under way to evaluate MRI's role in screening women with family history and/or a genetic high risk for breast cancer. Investigators want to determine which screening method or methods can best detect cancers in this group.
The existence of the European Union holds certain advantages for sharing research efforts among various countries. An example is the Interpret project, which gathered brain tumor spectroscopy data for five years from eight sites in four countries. John R. Griffiths, Ph.D., a researcher in biochemistry and immunology at St. George's Hospital Medical School in London, guided the effort. Although the EU's micromanagement style can be intrusive, he said, it worked particularly well for this project.
"If we did not meet deadlines for reports, we didn't get paid," Griffiths said.
The result is a database of over 500 brain tumor spectra. Radiologists simply feed the spectra of an unknown case into the program and wait for a match. Stored images and clinical records can be used for purposes of comparison. The database is by no means complete, and Griffiths hopes investigators will be able to add cases in an open-source environment.
Griffiths' next project is mapping the body's metabolites, similar to the Human Genome Project's mapping of genes. St. George's just unveiled the Medical Biomics Center, which incorporates genomics, transcriptomics, proteomics, and metabolomics. Griffiths will use MR spectroscopy to find characteristic signatures of diseases that are associated with gene and protein expression. A recent experiment discovered how cancer cells give themselves a turbo-boost to grow faster. Metabolomics will help researchers design new drugs to attack various diseases, Griffiths said.
In many radiology departments, the loss of turf looms constantly. Dr. Gabriel P Krestin, radiology chair at the University Hospital Rotterdam in the Netherlands, has solved one turf battle without firing a shot. Krestin recruited cardiologist Prim J. de Feyter to head noninvasive cardiac research. De Feyter has a joint appointment and splits his time evenly between radiology and cardiology. Among his staff are cardiology and radiology doctoral students working together on projects dealing with MDCT and MR of the heart.
"Hiring Dr. de Feyter was my best political move," Krestin said. "Everybody is happy. We don't have turf battles with cardiology anymore, and we are not threatened that they will install imaging equipment in their department."
As at many universities, the cardiology department at Rotterdam is well funded, and it could easily have purchased a few MR units.
"We have solved that part of the problem, at least for the next couple of years. And the fact that we are successful scientifically proves it was a good move," he said.
Radiological innovation originates with forward-thinking leaders like Krestin and Mass General's Thrall. It also stems from governments that are willing to commit resources for research and from strategic alliances and partnerships. As imaging takes on a greater role and as resources and personnel become even more squeezed, just about everyone agrees that collegiality and collaboration on a global scale are necessary to effectively address imaging in the 21st century.
James Brice, Harold Abella, and Paula Gould contributed to this report.