Applied Radiology

Full-field digital mammography

As Trex Medical Corporation (Danbury, CT) awaits FDA clearance for its full-field digital mammography (FFDM) device, the company predicts that this technology will someday replace film-based mammography as the gold standard for breast cancer detection. Until that day arrives, FFDM will most likely be used side-by-side with conventional mammography when it becomes available sometime in 1999.

Clinical trial evaluated more than 500 cases and 4000 images

Today, film-screen mammography has about 85% sensitivity, and two-thirds of missed cancers are occult, said Melinda J. Staiger, MD, director of breast imaging at Good Samaritan Hospital Medical Center (West lslip, Long Island, NY). Dr. Staiger's institution was one of three centers that conducted comparative studies demonstrating that FFDM is clinically equivalent to film-screen mammography, a requirement for FFDM to be granted 510(k) market clearance by the FDA. The trial collected data from more than 500 cases and more than 4000 images.

"The digital mammograms I've seen preserve beautiful definition of the external tissues, the skin, and the subcutaneous tissues we frequently sacrifice in our [film-screen] techniques in order to penetrate through very dense breast tissue," said Dr. Staiger.

"There's the possibility that digital mammography ultimately will outperform film-screen mammography, but in the near term, it's relatively simple to show they are equivalent," said Laurie L. Fajardo, MD, professor of radiology and vice chair of research at the University of Virginia Health Sciences Center (Charlottesville, VA). "In the limited scope of this study, the imaging of patients who have particularly dense breasts appears to be superior on digital systems because signal-to-noise can be enhanced better than with film-screen images," said Dr. Fajardo. "I believe that if digital mammography does become widespread at some point in the future, we will begin to appreciate earlier diagnosis and less misdiagnosis with mammography, particularly in the radio-dense patient population and the younger patient population," she added.

"Greatest potential impact on management of breast cancer"

The National Cancer Institute (NCI) in 1991 identified digital mammography as "the most fertile territory for major advances in x-ray detection and diagnosis of minimal breast cancers." The NCI also predicted that digital mammography is "the evolving technology with the greatest potential impact on management of breast cancer." The technologic challenge for FFDM is to provide high-resolution, high-contrast images with the lowest possible radiation dose to the patient.

Postprocessing: contrast enhancement, computer-aided diagnosis, tomography, and 3D

Digital mammography can provide more consistency than film, the quality of which depends not only on image acquisition but on chemical processing. In addition, FFDM features the potential of postprocessing techniques to amplify subtle contrast differences in suspicious regions of the breast tissue.

During clinical trials to evaluate FFDM, the ability to manipulate mammographic images on the computer screen was a novelty, noted Dr. Staiger. In the real world of clinical practice, though, "we would fiddle around with contrast only for problem cases," she said. For the average patient, it should only take about 1 minute to read a mammogram, she explained, so it would not make sense to take 5 minutes perfecting the contrast resolution for each case.

Computer-aided diagnosis may also be possible with FFDM, serving almost as a "second reader." The computer, for example, may automatically draw a border around areas of abnormal contrast, calling the radiologist's attention to suspicious regions. This technology might be useful, for instance, in cases where the radiologist is "seduced" by obvious findings. "Sometimes big lesions can be missed when the focus is on looking for very small lesions," noted Dr. Staiger. In the future, image reconstruction techniques with FFDM may eventually offer tomographic or three-dimensional breast images.

There may be up to l000 patterns of a normal breast, noted Dr. Staiger, which makes breast imaging a different type of challenge than imaging the brain, liver, lungs, heart, or other organs. Once FFDM becomes available, mammographers will need to develop algorithms for obtaining good images from dense breasts, fatty breasts, and breasts with microcalcifications, she added.

"My impressions, my intuition, from what I've seen so far, is that it's going to bring a lot of cancers to light in the mammogram that might have been missed before," said Lawrence W. Bassett, MD, FACR, professor of breast imaging at the Iris Cantor Center for Breast Imaging, University of California, Los Angeles (UCLA) School of Medicine. "Although we thought it was going to be difficult to use a larger image receptor, the technologist adapted to it quite readily and, in fact, found it was no more difficult to position than our conventional mammograms were," Dr. Bassett added.

Siemens Opdima® provides digital spot imaging today as company plans for FFDM future

Siemens Medical Systems, Inc. (Iselin, NJ) offers the Opdima®, a mammography digital spot imaging system for use with its MAMMOMAT® 3000 upright mammography unit. Radiologists and mammographers don't have to wait for the future to take advantage of digital imaging, according to Maria Di Palermo, mammography product manager at Siemens. When film mammogram results prompt radiologists to request additional spot views, they can do them with the Opdima system and take advantage of postprocessing techniques for magnification, contrast enhancement (window, level, and filters), and measurements of small calcifications.

Opdima technology is based on a large-area spot charge-coupled device (CCD) incorporated in a slim cassette that fits into the standard object table of the MAMMOMAT 3000. "The digital spot camera is wafer-thin, and fits in the same bucky as a film cassette. Technologists can switch back and forth from film to digital spot imaging with no trouble," said Ms. Di Palermo. "We have the largest field of view available today for digital spot imaging, 49 ¥ 85 mm," she added.

The Opdima represents the first time that Siemens is offering a digital imaging system that can be used in mammography for nonstereotactic applications. "We anticipate that this will become an important tool for evaluation of suspected areas, and for presurgical needle localization," said Ms. Di Palermo. Opdima is available and deliverable in the United States and worldwide.

In addition, the company will probably have its investigational full-field digital mammography (FFDM) system in clinical sites next year for beta testing. There are many technical aspects of FFDM that are still a challenge, noted Ms. Di Palermo. "Each FFDM image is about 40 megabytes, so it takes a very sophisticated computer system to handle the typical four views taken of one patient," she explained. Mammographers will probably want the capability to compare studies with baseline mammograms on the monitor, she added.

Kodak signs 3-year agreement to provide mammography film to Tenet Healthcare

Eastman Kodak Company (Rochester, NY) has a 3-year agreement to be the sole supplier of mammography film products to Tenet Healthcare Corporation, which owns 124 acute-care hospitals in 18 states. Tenet also operates a group purchasing organization, BuyPower, which serves Tenet hospitals as well as 2000 independent facilities.

Kodak agrees in August to acquire Imation

In August, Kodak agreed to acquire most of Imation Corp's worldwide medical imaging business, including its DryView™ laser imaging business. The terms call for Kodak to pay Imation $520 million in cash at closing. This acquisition will enable Kodak to provide a broader portfolio of products worldwide, said the company.

SNM 1998: Advances in cardiovascular imaging and oncology

With new peptide and antibody radiopharmaceuticals on the horizon, The Society of Nuclear Medicine (SNM) held its 45th Annual Meeting June 7-11, 1998, in Toronto, Ontario, Canada. Advances in oncology were focused on the expanding use of positron emission tomography (PET) with fluorine-18 [18F] fluorodeoxyglucose (FDG). Nuclear cardiology is compiling huge databases from thousands of patients-proving the effectiveness of myocardial perfusion studies in determining which patients to discharge from the emergency room and in predicting the risk of future cardiac events.

99mTc-labeled peptide detects deep-vein thrombi

With a new radiopharmaceutical now under FDA review, physicians can quickly differentiate between an active blood clot in the leg, which poses life-threatening risk to the patient, and other types of circulatory problems that are far less dangerous-and which require very different medical treatments. Technetium-99m [99mTc] apcitide (Acutect®, Diatide) is a radiolabeled peptide that targets receptors on the cell surface of platelets that have become activated in the blood-clotting process.

Nycomed Amersham to market/sell apcitide

Apcitide was developed by Diatide, Inc. (Londonderry, NH), a company that specializes in radiolabeled peptides. Nycomed Amersham (Princeton, NJ) will handle the marketing and sales of 99mTc apcitide. In medical journal papers, apcitide is sometimes called "P280."

"P280 has been shown in Phase III multicenter clinical trials to have a relatively high diagnostic accuracy for deep-vein thrombosis. We are planning to use it to distinguish between recurrent deep-vein thrombosis and postphlebitic syndrome, which is inflammation of the vein," said Raymond Taillefer, MD, director of the Nuclear Medicine Department at Hospital Hôtel-Dieu de Montreal (Québéc, Canada). "It is important to make this distinction because deep-vein thrombosis is treated with anticoagulants, whereas postphlebitic syndrome is treated with other types of drugs," said Dr. Taillefer, who participated in clinical trials for apcitide.

About 70% of patients with deep-vein thrombosis (DVT) are asymptomatic. Furthermore, the signs and symptoms of DVT can be nonspecific. Among patients who show typical symptoms-such as leg swelling, warmth, calf pain and tenderness-only 30 to 40% actually have DVT. It is important, therefore, to have an objective test that clearly identifies acute DVT.

Differentiating acute from chronic DVT

If a blood clot is actively forming (acute thrombus), the patient usually requires treatment with an anticoagulant, such as heparin. The typical protocol is to hospitalize the patient for several days to administer intravenous heparin, followed by 3 months of oral anticoagulants. On the other hand, if the blood clot is a chronic thrombus that has already stabilized, the risk of embolism is much lower, and the patient may not require anticoagulants.

Some patients may undergo unnecessary anticoagulant therapy, just to be on the safe side and prevent a possible pulmonary embolism, because their physicians are not sure whether the thrombus is active or chronic. Besides the cost of the hospitalization and medication, anticoagulant therapy poses certain risks, such as gastrointestinal tract bleeding or platelet abnormalities. Therefore, any test that helps patients avoid unnecessary use of anticoagulants would be clinically valuable.

Acute DVT may lead to pulmonary embolism

The most dangerous, potentially fatal consequence of DVT is a possible pulmonary embolism. In the United States, an estimated 5 million patients experience one or more episodes of DVT each year, leading to 500,000 annual cases of pulmonary embolism, resulting in 100,000 deaths per year.

"The advantage of apcitide is that we are actually looking at active blood clotting with physiologic nuclear medicine images. With anatomic images, provided by ultrasound and contrast x-ray venography, we can see an anatomic abnormality that may or may not be an active thrombus," said Robert F. Carretta, MD, director of the Nuclear Medicine Department at Sutter Roseville Medical Center (Roseville, CA).

"Another advantage of apcitide is that we can image the patient shortly after injection and get the report to the referring physician very quickly. It will not replace ultrasound, which is an excellent screening procedure. But ultrasound has its limitations in the calf veins, and often does not detect acute deep-vein thrombosis below the knee," said Dr. Carretta, who also participated in apcitide clinical trials.

Therapeutic antibody for non-Hodgkin's lymphoma

A monoclonal antibody labeled with iodine-131 [131I] is being evaluated by Coulter Pharmaceutical (Palo Alto, CA) as a therapeutic agent for low-grade non-Hodgkin's lymphoma. The anti-B1 antibody targets the CD20 B-lymphocyte antigen found on non-Hodgkin's B-cells, as well as on some normal B-cells. The unlabeled anti-B1 antibody and the 131I each have separate therapeutic effects on non-Hodgkin's lymphoma B-cells.

During the first stage of treatment (dosimetric dose), two doses are administered intravenously: a 60-minute infusion of unlabeled antibody followed by a separate 20-minute infusion of 5 mCi of 131I-labeled antibody. Over the next few days, gamma camera scans determine the 131I residence time, used to calculate the patient-specific therapeutic dose, which is also administered in separate "hot" and "cold" infusions. Patients are given oral iodine supplements to block thyroid uptake of 131I.

Two Phase I/II clinical trials with 131I-anti-B1 antibody have been completed. According to Coulter, in one trial of patients with relapsed or refractory non-Hodgkin's lymphoma, 42 of 59 patients (71%) achieved either a complete or partial response (defined as 350% reduction in all measurable lesions with no new lesions). In the other trial, 14 of 45 patients (31%) achieved complete response and 13 patients (29%) achieved partial response. Side effects included a transient flu-like syndrome and bone-marrow suppression.

A Phase III trial is now underway.

PET predicts success of chemotherapy in non-Hodgkin's lymphoma

In another application of nuclear medicine for patients with non-Hodgkin's lymphoma, early evaluation with FDG-PET, after two or three cycles of chemotherapy, may help to predict which patients will succeed with the treatment (Abstract No. 580). [Abstracts are published in the supplement to the May 1998 issue of J Nucl Med.]

In a study of 20 patients, seven patients had abnormal PET scans early in the course of chemotherapy: four did not achieve complete remission and showed signs of disease recurrence within 4 to 15 months of diagnosis; two achieved complete remission, but relapsed at 12 and 20 months; and one remained in complete remission for 30 months. Five of these patients died. Of the 13 patients who had normal FDG-PET scans, eleven remained in complete remission for up to 36 months after diagnosis; two relapsed (at 6 and 14 months) and died of Iymphoma. "Patients with early response to chemotherapy [indicated by a normal FDG PET scan after 2 or 3 cycles of chemotherapy] have a high probability of progression-free survival," said Pierre M. Rigo, MD, professor and director of the Nuclear Medicine Division at University Hospital (Liège, Belgium).

Octreotide scan improves staging of Hodgkin's disease

A prospective study of 126 patients previously untreated for Hodgkin's disease found that nuclear medicine imaging with indium-111 [111In] octreotide (OctreoScan®, Mallinckrodt) detected disseminated disease (Stage III or IV, spread to organs or in lymph nodes at both sides of the diaphragm) in a significant proportion of patients who had been classified as Stage I or II (disease confined to lymph nodes at one side of the diaphragm) by standard staging techniques (Abstract No. 147). In some cases, these newly detected lesions were outside the subtotal nodal irradiation field, which is the standard treatment for Hodgkin's disease patients with a favorable prognosis. "These patients have a high probability of relapse," said Elly Lugtenburg, MD, of the Department of Hematology, University Hospital Rotterdam (The Netherlands). Instead of radiotherapy aimed at the diseased lymph node, patients who are actually at Stage III or IV should be treated with chemotherapy.

FDG-PET detects colorectal cancer recurrence in patients with normal CEA levels

Results of one study may cause some oncologists to reassess the value of tumor markers as a screening test for recurrent colorectal cancer (Abstract No. 531). In a study of 45 patients, whole-body FDG-PET correctly identified lesions in 16 of 17 (94%) patients with normal CEA and CA 19-9 levels. In addition, FDG-PET correctly diagnosed up to 95% of lesions in patients with elevated CEA and CA 19-9 levels. Typically, a rise in plasma levels of tumor markers is one reason to refer a patient to FDG-PET scanning for identifying recurrent colorectal cancer.

"Our data indicate that normal tumor marker levels do not exclude the presence of malignant tissue," said Hans Bender, MD, associate professor of nuclear medicine at the University of Bonn (Germany). "If there are any other indications of tumor recurrence, such as lesions detected by CT or ultrasound, pain reported by the patient, etc., I would recommend whole-body FDG-PET independent of CFA or CA 19-9 levels," said Dr. Bender. High glucose uptake with FDG is typical of tumor recurrence, he explained, whereas lack of glucose uptake would indicate scar tissue. However, it is too early to recommend that FDG-PET replace tumor markers as a screening method. Dr. Bender said he would like to see this question addressed in a future prospective study with a much larger patient population.

FDG-PET for staging head and neck cancer

For patients with squamous-cell cancer of the head and neck, PDG-PET may be more accurate than other imaging tests, and can help steer patients toward the most appropriate treatment.

One study of 51 patients, conducted at the Northern California PET Imaging Center and several California medical centers, found that PET was more accurate than x-ray computed tomography (CT) and magnetic resonance (MR) for the diagnosis of local and regional recurrence of disease (Abstract No. 479). In addition, PET commonly detected unsuspected distant disease. "Pretreatment imaging with FDG-PET avoids the morbidity and cost of attempted curative surgery or radiotherapy in patients with undiagnosed distant recurrences," said Elma Abella-Columna, MD, a clinical fellow at the Northern California PET Imaging Center (Sacramento). "These patients may be candidates for chemotherapy," she added.

In the second study of 44 patients, presented by researchers from the Saint Louis University Health Sciences Center (Missouri) sequential FDG-PET (obtained at 2 and 10 months after therapy) was evaluated for detection of recurrent head and neck cancer (Abstract No. 478). "More accurate detection of recurrence may provide ways of improving survival," said Val J. Lowe, MD, director of PET at Saint Louis University Health Sciences Center. His group found that "PET can detect head-and-neck tumor recurrence when it may be undetectable by other clinical methods."

Nuclear cardiology: Cost-effective in ER patients with chest pain

A study reported that using nuclear cardiology tests in a dedicated chest-pain center to treat emergency patients reduced the yearly rate of myocardial infarction (MI) from 1.8% to 0.1% in outpatients discharged with a principal diagnosis of nonspecific chest pain. Results of this 2-year study of 6,548 patients were presented by investigators from The Miami Cardiac and Vascular Institute and Baptist Hospital (Miami, FL) and Emory University School of Medicine (Atlanta) (Abstract No. 541).

The cost per diagnosis (of nonspecific chest pain) was $606 prior to adding nuclear cardiology-specifically, myocardial perfusion imaging with single-photon emission computed tomography (SPECT)-to the protocol. With SPECT in the protocol, the cost of this diagnosis increased to $1,676. The total cost of SPECT in this patient population, divided by the number of additional MIs detected-preventing those patients from being sent home with a hidden heart attack-was $59,400. However, when factoring in the cost of missing an MI, estimated at $50,000, the cost of SPECT per additional MI detected was only $9,400.

"With a minimum estimated likelihood of mortality from missed MI of 16%, the maximum cost to save a life is only $59,000," noted Jack A. Ziffer, MD, PhD, director of the Cardiac Imaging Institute. "In this year of constrained resources, nuclear cardiology can play a central role in improving patient outcomes, and do it cost-effectively, in the clinical pathways of acute ischemic heart disease," he added.