Earlier Detection: Improving the Chances for a Cure

Until breast cancer can be prevented, detection at the earliest time possible remains an essential goal. In its first four cycles, CBCRP has funded 40 grants aimed at the earlier detection of breast cancer either through improving access and detection technologies or the discovery of biomarkers that could be used to signal the presence of the disease.

Conclusions

Developing and improving imaging technologies: better and easier detection

The thrust of research in imaging technology has been toward reconfiguring current imaging devices such at PET and SPECT scanners to a size and configuration more suitable to breast imaging; improving the ability of all techniques to detect abnormal masses at their smallest size; and bringing new technologies such as optical imaging into use.

Edward Hoffman, Ph.D., at the University of California, Los Angeles has developed a promising small scintillation camera configured to more easily image the breast by combining a NaI(Tl) crystal and a Position Sensitive Photomultiplier. This camera could provide a method of identifying malignant lesions identified as suspicious by mammography and has shown in laboratory studies to be able to detect tumors as small as 0.5 centimeter in diameter. A postdoctoral fellow, Yiping Shao, Ph.D., at the University of California, Los Angeles was also funded to design and develop a detector—in this instance a PET scanner—specifically for imaging the breast. He was able to successfully bend optical fibers at large angles with insignificant light loss, thus improving detector performance and exceeding the spatial resolution and count rates (this would lower the radiopharmaceutical dose to the patient) expected in the original proposal. Its predicted performance may result in many significant improvements over conventional clinical PET scanners, and at much lower cost.

Responding to the need to reduce the number of erroneously interpreted breast images, Bruce Tromberg, Ph.D.,of the University of California, Irvine, undertook research to improve the potential of non-invasive optical imaging using near-infrared light. The promise of optical imaging to provide detailed physiological information on breast tissue has been hampered by the light scattering that blurs small, buried, lightabsorbing tumors. His results (from 18 patients) show that pre- and post-menopausal normal breasts exhibit detectable wavelength dependent differences in both absorption and scattering, and suggests that optical methods are quite sensitive to cellular, molecular and structural differences between normal and malignant breast tissues. Further work is necessary to pin down the precise hemodynamic, oxygen consumption, and structural changes that are encoded within optical signatures. With such information, a new generation of safe, non-ionizing diagnostic tools is possible that could complement existing methods and enhance the overall accuracy, specificity and sensitivity of breast cancer detection.

Robert Brasch, M.D., at the University of California, San Francisco, completed a 3-year Research Project for the MRI Detection of Mi-crovascular Status in Breast Cancer. The blood vessels feeding tumors are increasingly being appreciated as a unique target for therapy. This project explored improvements in imaging tumors in rats with specific attention to selecting the best MRI detection methodology and the use of contrast agents to analyze the tumor blood vessels. A correlation was found between the severity of the tumor and the leakiness of the blood vessels, and an angiogenesis inhibitor (anti-VEGF) was found to decrease this leakiness. Interestingly, differences in the tumor accumulation of chemotherapeutic agents was observed when blood vessel leakiness was altered. These are key insights in technology development for the anticipated uses of chemotherapy and angio-genesis inhibitors, and it presents a rationale for pursuing MRI to evaluate therapy and to measure parameters of tumor leakiness prior to treatment.

Michael Buonocore, M.D., Ph.D., at the University of California, Davis, completed a one-year project intended to develop and test new methods of measuring the blood flow through breast tumors, to help determine the presence of abnormal tissue, and to help assess whether the abnormal tissue is malignant or benign (cancerous or non-cancerous). Previous MR studies, using injections of a contrast material (i.e., a drug-like compound that is absorbed by the body's tissues) to improve the ability to see abnormalities, had shown that the blood flow in malignant tumors is higher than that in benign tumors and normal tissue. Using the new methods, he and his team hoped to perform the same type of measurement without the use of an injection of contrast material.

He was able to obtain measurements in eight normal subjects without breast lesions and five subjects with palpable and/or mammographically visible breast lesions. He was able to show that his method is reproducible and extremely sensitive to the amount of blood flowing through breast tissue, and is able to detect small differences that exist between abnormal and normal tissue.

George Leopold, M.D., and Michael André, Ph.D., at the University of California, San Diego, undertook a one-year study to examine the suitability of a new ultrasound system for imaging the breast. This system makes cross-sectional image “slices” by transmitting sound waves into the breast from all sides and reconstructing the images in a computer similar to the operation of a x-ray computed tomography scanner (CT or “CAT”). This technology needed to be evaluated for its ability to distinguish between normal tissues and the many types of “lumps” or masses that may occur in the breast.

During the period of study, progress was made in understanding the appearance of different tissues as well as their unique ultrasound “signature”. Preliminary results in very dense breasts, which are difficult to image well with mammography, were also encouraging. The technical quality of the ultrasound CT images and the potential limitations of the method are now better understood.

Tumay Tumer, Ph.D., of Nova, Inc., undertook a one-year study to try to improve the technology for producing X ray images of the breast that does not use film as the exposure medium. Traditional mammography systems use phosphor screens to generate light from X rays and photographic film to record the resulting images for display on a viewbox. Image quality is limited by the chemical processing procedure and has a narrow contrast range, and altering this image requires re-exposure of the patient to x-rays to obtain another film. In a digitally recorded mammogram, alterations can be made on the computer.

Specifically, in this project they proposed to develop a two-dimensional array of detector elements on a combination of cadmium, zinc and telluride crystals so that an image of the X rays passing through the patient's breast could be produced. Directly converting x-ray energy into electrons would provide an image with smaller visible detail, using less radiation, and showing less interference on the image from the imaging device itself (“noise”) than present x-ray films.

The detectors proved more difficult to assemble than anticipated, and at the project's end work was just beginning on lining a narrow column with these detectors to scan underneath the patient's breast. If this study is eventually successful, Nova expects to obtain funding for the development of pixel detectors from this material for a commercial digital mammography system called DigiMAM™.

Biomarkers and novel screening approaches: unmasking the hidden signs

Nicholas Petrakis, M.D., of the University of California, San Francisco, was funded to expand on his previous research showing that women whose nipple aspirate fluid contained unusual or “atypical” breast epithelial cells had a three-fold risk of breast cancer compared to women with normal findings, and a five-fold increased risk compared to women who did not yield any breast fluid. His CBCRP grant investigated whether combining nipple aspirate fluid findings with measures of mammo-graphic density (shown to be predictive of breast cancer risk in White women) would improve the ability to estimate a woman's risk of developing breast cancer. While this combination did substantially improve the positive predictive value of developing breast cancer, unfortunately the increase was of no clinical value due to the excess number of false positive predictions.

Per Borgstrom, Ph.D., at the La Jolla Institute for Experimental Medicine, completed a 1-year IDEA project to study Platelet Factor 4: a Marker for Malignant Breast Tumors. A visualization technique of intravital microscopy was used to show that platelet factor 4 that is labeled with fluorescence was associated with regions of tumor cell-induced blood vessel growth (angiogenesis). The breast cancer cell line MCF-7 had the highest level of platelet factor 4, which was consistent with a higher tumor growth rate and concurrent blood vessel growth. Dr. Borgstrom was also able to inject labeled platelet factor 4 into the capillaries of mice with tumors and show that it rapidly localized to the tumor sites and represented a new and potentially useful biomarker for breast cancer.

Ashraf Imam, Ph.D., at the University of Southern California, completed an IDEA award investigating the validity of a new prognostic marker for breast cancer called LEA.135. He found that the presence of this marker in human invasive breast cancers strongly correlates with a longer tumor-free and overall survival rate. This finding held up in White and Black women even when other prognostic markers were considered. When Dr. Imam examined the presence of LEA.135 in pre-cancerous cells (ductal carcinoma in situ), he found the presence of LEA.135 indicated a slightly lower risk of developing breast cancer.

In order to determine how to take advantage of a tumor marker, it is always helpful to understand how it functions in a normal cell and how its function has changed in a tumor. Christina Niemeyer, Ph.D., at The Burnham Institute, completed a postdoctoral fellowship exploring the function of a breast cancer marker (Cripto) in breast cancer formation. She examined the levels of Cripto in the breasts of mice at different stages of development and found that the Cripto levels vary with the physiological state of the gland. Dr. Niemeyer also found that when breast cells were grown in the culture dish, the Cripto levels could be increased by exposing the cells to pregnancy hormones such as prolactin. Cultured cells with high levels of Cripto grew at an increased rate, whereas cells with lower levels grew more slowly. When normal cells with low Cripto were reintroduced into mice they were unable to grow; however, when breast tumor cells with low Cripto levels were reintroduced they still formed tumors. These studies show that Cripto may have its greatest impact on the susceptibility of the pre-cancerous breasts rather than on tumor growth.

Improved access to screening: reaching every woman

While increasing numbers of women are receiving mammograms and clinical breast examinations, the number of women who obtain screening on a regular basis remains low. Nicole Howard at the CHG Foundation and Gregory Talavera, M.D., M.P.H, at San Diego State University, formed a team to identify interventions that show potential for increasing annual rescreening within the context of a state-funded program. Among women studied, they found that only 32% were receiving annual rescreening. Of those who had been seen for rescreening, 68.9% had received a reminder from the primary care provider. Of those who had not been seen for rescreening, only 38.7% had received a reminder. Clearly reminders from women's doctors beneficial, but they found that, within the population served by the state-funded program, yearly reminders may miss a significant proportion of women. 27% of women selected for the study could not be contacted (e.g., disconnected telephone, interviewers reached a wrong number). An additional 25.7% of women could not be reached after repeated calls, suggesting that they too may be difficult to reach with standard rescreening reminders. Women who were contacted were more than twice as likely to have been rescreened than women who could not be contacted. Thus, the team suggests that interventions delivered during the 12-month interval between initial and repeat screening may be more effective than standard anniversary date reminders.

Research in Progress

Developing and improving imaging technologies: better and easier detection

Among grants to improve to improve existing technologies is the work at the University of California, Davis by John Boone, Ph.D. and Anthony Seibert, Ph.D. Dr. Boone's research is to improve the spatial resolution (the capacity to image small objects) of digital mammography systems. The current technologies that enable the capture of more X rays also “blurs” the visible light produced from these rays. Dr. Boone hopes to use a “hybrid detector” which uses microplate technology (a honeycomb matrix of leaded glass with tiny square pores aligned with the x-ray source and packed with x-ray phosphor) to channel the visible light toward the light detector and reduce blurring and improve resolution. To date he has manufactured the honeycomb, verified the performance of a necessary x-ray spectrum measuring device, and is conducting computer simulations to study the ability of the system to exclude scattered radiation. Dr. Seibert intends to construct and test a novel digital mammographic imaging detector (a fiber-optic scintillator coupled to a digital camera) that may be useful for more sensitive detection of breast cancer, and which may prove to be a cost-effective solution to the problem of adequately imaging the dense breast. Under study is the alternative phosphor (light producer) cesium oxide. Initial testing of this phosphor in the detector (Complementary Metal-Oxide Semiconductor or CMOS) camera reveals a resolution satisfactory for imaging microcalcifications in the breast. Other functional aspects of the camera are now being tested as a method for correcting image imperfections.

William Moses, Ph.D., of the Lawrence Berkeley National Laboratory and Manbir Singh, Ph.D., of the University of California, Los Angeles are both working to improve existing detection technologies. Dr. Moses is working to optimize PET cameras (an acronym for Positron Emission Technology) so as to better determine whether suspicious structures observed in mammograms have the increased metabolism associated with breast cancers. These cameras would image the metabolic activity in the breast (to detect a cancer) or the axilla (to detect any metastasis to the lymph nodes). The first phase of the work of improving and adapting three of the basic components (an array of scintillator crystals, a photomultiplier tube, and a photo-diode array with one element per scintillation crystal) has been partially completed. The electronics necessary for reading out the photomultiplier tube and the photodiode array have been fabricated and tested, and a prototype detector tested. Work has also begun on the software necessary to operate the camera, acquire data, and reconstruct images from the data.

Dr. Singh will attempt to design a system that will provide, using a newly designed camera configured for the breast, a three-dimensional image from the gamma rays emitted from radiopharmaceuticals. To date, they have successfully conducted computer simulation and experimental studies using the new generation of semiconductor detectors (cadmium zinc-telluride, CZT) arranged either in a cylindrical or hemispherical configuration that does not require breast compression. Results are promising and for the first time suggest the feasibility of using CZT detectors in a SPECT (single photon emission computed tomography) system. Further experiments will be conducted in the next year under a realistic clinical imaging situation.

Jack Sklansky, Eng. Sc.D., of the Charles R. Drew University of Medicine & Science, is attempting to develop a computer system to help radiologists use large collections (“databases”) of digitized mammograms as aids in determining whether or not to recommend biopsies with the goal of reducing the number of unnecessary biopsies. The first test of this system enabled radiologists to cut approximately in half the number of benign (non-cancerous) biopsies and the number of missed cancers. This test also indicated the potential of radiologists electronically “browsing” a large database to find mammograms that are visually and medically similar to a mammogram under analysis.

A key issue in early detection is combining more advanced imaging technology with the ability to analyze the specific characteristics of breast tumors that could indicate a higher potential towards malignancy or invasion. Orhan Nalcioglu, Ph.D., at the University of California, Irvine, is combining Magnetic Resonance Imaging (MRI) with specific probes to detect tumor blood vessels. The critical research hurdle is developing an imaging contrast agent that both works for the MRI detection and is selective for compounds present in blood vessels. He is presently evaluating a protamine-based MRI agent that should recognize heparin, a component of the extracellular matrix, which is present in tumor blood vessels. Prota-mine is a biological compound that is already approved for human use, so this project has the potential for more rapid clinical development.

Two other studies are investigating exciting applications of the new digital mammography technology: one within a medical center, and the other linking the community with a medical center. Daniel Valentino, Ph.D., of the University of California, Los Angeles, intends to develop computer systems that enable cost-effective digital mammography in a hospital or community clinic. To date they have:

  1. implemented an infrastructure to acquire, archive, and distribute full-field digital mammography images, and have integrated it with the hospital's imaging and information system;
  2. established an image database for evaluating systems for processing and display of digital mammography images;
  3. developed software to reduce the size of original breast images and restore the compressed image back to its original size, evaluated the ability of radiologists to use these restored images, and determined the limits of useful compression (a factor of 25);
  4. developed workstation software for the rapid display and screening of digital mammograms which is now available in a software toolkit called the UCLA Digital Viewbox/Tk™.

H.K. Huang, D.Sc., of the University of California, San Francisco, is attempting to demonstrate that: 1) telemammography technologies can be developed for routine clinical operation; and 2) real-time telemanagement (i.e., converting a mammogram to a digital image at the examination site and electronically transmitting it to a breast imaging center for immediate expert consultation) can be established for mammography practice. To date, they have been able to complete the telemammography chain with a full field direct digital mammography system (FFDDM) between Mt. Zion Hospital (the breast imaging expert center) and the Laboratory for Radiological Informatics UCSF. Preliminary results using the FFDDM indicate that image quality is at least as good as, if not better than, conventional film mammograms.

Another study applicable to the use of both digital and conventional mammography is that being conducted by Laura J. Esserman, M.D., at the University of California, San Francisco, who is investigating the factors required to achieve high cost-effectiveness in mammography screening in California, and will then develop a plan to promote the provision of the highest quality screening services for the least possible cost. This extremely ambitious, many-faceted research involves quality control studies of mammography and mammographers, an overview of the Northern California market for mammographic screening, identifying strategies for adoption of a low-cost regionalized screening program, estimating the cost of implementing telemammography and its potential benefits, and examining the possible role of publicly funded programs. To date, the quality control studies have been completed and the results are being analyzed and the telemammography study is underway.

In order to help ensure the continued entrance into breast cancer research of promising students in various fields within the imaging technologies discipline, CBCRP has funded a biomedical physics graduate training program at the University of California, Los Angeles, under the direction of Carolyn Kimme-Smith, Ph.D. To date, five students have been enrolled and are making excellent progress in their studies and research endeavors.

Biomarkers and novel screening approaches: unmasking the hidden signs

William M. Pardridge, M.D., at the University of California, Los Angeles, School of Medicine, extending previous CBCRP work, is testing ‘cationized’ HER2 monoclonal antibody in a suitable strain of mice (a strain of mice which do not reject implanted human tumor tissue) developed specifically as part of this study. Dr. Pardridge previously demonstrated that the migration of this antibody across the blood-tumor barrier into the tumor is facilitated when the antibody has been given an electrical charge (cationized). He is now preparing to conduct a comparison study of native (uncationized) and cationized HER2 antibody, both tagged with a radioisotope, to evaluate the extent to which the cationization improves the tumor uptake of the antibody, thus paving the way for future improved early detection and possible treatment of breast cancer in humans.

David Vera, Ph.D., at the University of California, San Diego, is working toward a similar goal, improved uptake of a radiopharmaceutical, in this case within the lymph nodes. The portion of his work funded by CBCRP is to measure the size distribution of a new sentinel lymph node imaging agent and subsequent versions of this agent—a parameter important to the ability of such an agent to rapidly enter the lymph channel and clear the injection site. The first measurements have been made, and next year the size distribution of the final version will be made.

Malcolm C. Pike, Ph.D., of the University of Southern California School of Medicine, is investigating whether mammographic densities can be used to predict the risk of breast cancer in African-American and Asian-American as has been found for White women, and in younger as well as in older women. They plan to study the mammograms of 1260 women. To date they have obtained and read a portion of the necessary mammograms, but no findings are available to date.

One goal for detecting breast cancer earlier is to take advantage of tumor biology to identify tissues at risk of becoming invasive cancers. Two investigators are using histological approaches to characterize these tissues. Susan P. Hawkes, Ph.D., of the University of California, San Francisco is determining whether a protein called TIMP-3 is a reliable early marker of breast cancer. Dr. Hawkes has developed an assay for TIMP3 that can detect faint traces of TIMP3 in early cancers and is in the process of testing whether it can be used as a non-invasive test for the presence of pre-cancers. Margaret Huflejt, Ph.D., of the La Jolla Institute for Allergy and Immunology, is looking at whether a protein called galectin- 4 can be used as a breast cancer marker. She has confirmed that galectin-4 is expressed in 100% of breast cancers, but is found in only a few pre-cancerous and noncancerous lesions. She has developed nucleic acid based probes and a new antibody to look at blood levels in patients.

Recently Initiated Research

Improved access to screening: reaching every woman

Bruce Allen Jr., Dr. P.H. of the Charles R. Drew University of Medicine & Science is studying the effectiveness of using a culturally-specific telephone intervention designed to increase participants' selfefficacy, increase their knowledge about the importance of regular mammograms, identify and modify their intentions to undergo screening, and provide women with information to counter their reasons for not having a mammogram. Using input from focus groups, an intervention was designed, and initial work on the survey began in December 1996 with the programming of the survey in a Computer Assisted Telephone Interviewing (CATI) system. The testing of this telephone intervention is expected to continue through September 1998.

Sora Park Tanjasiri, Dr. P.H. at the University of California, Irvine, is investigating the breast cancer knowledge, attitudes and behaviors toward screening of two groups of Pacific Islander immigrants, the Chamorros and the Tongans, and is attempting to elucidate the predictors of their screening behavior. Initial focus group work has been done during the past year and the survey instrument is under development.

Two studies are evaluating various aspects of self-care with respect to breast cancer. Jacqueline O'Connor, Ph.D., at the University of California, Davis, is investigating psychological characteristics that motivate breast cancer early detection practices, and variables that predict continuity and discontinuity in health practices following breast biopsy with a benign (non-cancerous) result. Dr. O'Connor's analyses of her data suggest that, while experiencing a threat to breast health (in this case a breast biopsy) can generate heightened perceptions of personal vulnerability and cancer anxiety that may persist for as long as one year following the event, these psychosocial consequences may not necessarily interfere with subsequent early detection practices . Noreen C. Facione, Ph.D., R.N., at the University of California, San Francisco, is studying how English and Spanish-speaking African-American/black, Hispanic/Latina, and Anglo/white women in the San Francisco Bay Area decide whether or when to consult their health care provider about a breast change that may be a signal of breast cancer. Dr. Facione has found that twenty three percent of the women in this communitybased convenience sample described themselves as potentially likely to delay seeking a professional evaluation of a breast symptom that worried them. Younger, less well educated, Black and Latina, and lower income women were more likely to say they would delay. Many of these differences were explained by the women's judgments about their access to health care services, perceptions of prejudicial treatment in the health care setting, habits of using self care remedies rather than seeing a health care provider, beliefs that treatment was futile should the symptoms prove to be cancer, prior habits of health care utilization, and eagerness to engage in making difficult decisions.

Many breast cancers are still diagnosed by clinical and self breast examinations using touch. It is the “viscoelastic” characteristics of the breast tissues that enable one to detect irregularities in tissue firmness and consistency. Michael Buonocore, M.D., Ph.D., at the University of California, Davis Medical Center, proposes that magnetic resonance imaging (MRI), coupled with a device to mechanically vibrate the tissue, could be used to detect these same tissue characteristics. The advantages of such a device would be greater sensitivity and the capability to “see” deeper within the breast.

After suspicious masses have been detected in the breast, they must often be biopsied to determine whether or not they are cancerous. Henry Van Brocklin, Ph.D., at the Lawrence Berkeley National Laboratory, is attempting to develop a pharmaceutical that will carry a radioactive label only to tissue cells that are cancerous. If the proposed pharmaceutical can be developed, it would enable physicians to locate suspicious masses and determine whether they are cancerous in one step.

Most breast cancers that metastasize (spread to other parts of the body) do so through the lymphatic drainage system which radiates out from the breast to nodes in the axilla (armpit). If the node which first receives this drainage (and thus likely the cancer cells) is free of cancer, then it is very likely that the cancer has not spread outside of the breast. Determining whether this first node (called the sentinel lymph node) is free of cancer is important because it most often means that it is not necessary to remove all of the lymph nodes—a procedure that often results in chronic physical problems, discomfort or pain. However, finding this sentinel node can be difficult, even for the most experienced surgeons. Recent techniques using either a blue dye or a radioactive tracer show great promise to dramatically improve the surgeon's ability to easily find the sentinel lymph node. David Vera, Ph.D., at the University of California, San Diego, is proposing to combine a dye and a radioactive tag into a single agent which should allow surgeons to easily and precisely locate the sentinel node. The goal is to develop a sentinel node detection agent that will enable a 100% success rate.