Genetic Analysis of Tumor Cells and Their Protein Expression Emmanuel Petricoin, M.D., Co-Director, FDA-NCI Clinical Proteomics Program, Bethesda, MD

The field of molecular medicine is moving beyond genomics to proteomics. While DNA is the information archive, proteins do all the work of the cell - and ultimately dictate all biological processes and the cellular fate. The challenge and opportunity within proteomics is much more than just developing a list of all the proteins. The true scientific goal of proteomics is to characterize the information flow within the cell and the organism. This information flow is mediated through and by, protein pathways and networks.

The cause of most human disease lies in the functional disregulation of protein-protein interactions. Understanding the role that protein networks play in disease will create enormous clinical opportunities, since these pathways represent the drug targets of the next decade. In the future, entire cellular networks, not just one disregulated protein, will be the targets of therapeutics. The next technologic leap will be the application of proteomic technologies to the bedside. It will soon be possible to analyze the state of protein signal pathways in the disease-altered cells, before, during, and after therapy. This can herald the advent of true patient-tailored therapy.

Our program is focused on the understanding of mechanisms of carcinogenesis, identification of new drug targets, and discovery of new biomarkers for early detection in actual human tissue tumor specimens. Tissue-based proteomics requires technology that can overcome the complex cellular heterogeneity one encounters when studying disease in tissue specimens. To that end, we employ the use of Laser Capture Micro-dissection (LCM) for the proteomic analysis of microdissected subpopulations of human solid tumors (prostate, breast, ovary, and esophageal) as a model for the study of disease progression. These studies encompass and employ:

a. Differential protein profiling and discovery technologies, such as two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) coupled with mass spectrometry for new target and biomarker discovery- over 140 protein have been identified to date.
b. High-throughput proteomic pattern profiling using surface-enhanced laser desorption and ionization (SELDI) to identify disease-related proteins and protein patterns and
c. Focused proteomic approaches through the use of multiplexed phospho-specific antibody arrays, and general antibody and lysate arrays for signal transduction pathway profiling.

Signal Transduction Inhibitors as Anti-Neoplastic Agents Clifford Hudis, M.D., Chief, Medical Oncology of Breast Center, Memorial Sloan-Kettering Cancer Center, New York City, NY

The growth in understanding of the underlying mechanisms and pathways contributing to malignant transformation and tumor progression is fueled in large part by the expectation that this knowledge will lead to improved therapeutics. This hope is based on successes in the past beginning most notably with hormone therapy which was the first targeted therapy in oncology. A variety of chemotherapy agents have been developed with some degree of target specificity but increasingly the exact mechanisms of cytotoxicity are questioned. Very recently targeted therapy in general has been energized by the availability of several active monoclonal antibodies including trastuzumab (Herceptin). The exact mechanism underlying the activity of trastuzumab is still debated and may include cell mediated immunity. However, inhibition of signal transduction through its prevention of HER family heterodimerization appears to be a key activity of this agent. When this signal is prevented a variety of downstream events are seen, cell growth is inhibited, and apoptosis ensues. Because this targeted therapy offers the potential for improved efficacy with minimized toxicity there is now a growing list of drugs under development which may be even more effective. Examples include drugs targeting the tyrosine kinase activity associated with HER2 or other members of the HER family (ie, HER1 or EGF-R), as well as drugs aimed at other non-HER kinases. Ongoing clinical trials are exploring the role of relatively specific EGF-R inhibitors, such as OSI-74 and ZD 1839 (Iressa) as well as other non-HER targeting agents, such as imatinib mesylate (Gleevec). From these trials we should learn not only about the potential efficacy of these agents but also more about the underlying molecular biology.

Targeting the epidermal and HER2/neu growth factor receptor Debu Tripathy, M.D., UCSF Carol Franc Buck Breast Care Center, University of California, San Francisco, CA

Specific genetic, biochemical and physiological events predispostion and lead to the development of cancer. As these biological pathways are further understood, targeted therapies that modulate these pathways are being developed. Several aspects of tumor cell behavior are amenable to different types of pharmacological manipulation that could produce a significant clinical advantage

Growth factor receptors are one of the most relevant and studied targets in breast and certain other cancers. The epidermal growth factor receptor family consists of the epidermal growth factor receptor (HER1/erbB-1), HER2/neu (HER2, erbB-2), HER3 and HER 4. HER2/neu has been extensively characterized and targeted, leading to the approval of trastuzumab (Herceptin) as the first biological therapy for breast cancer. See the syllabus entitled “Herceptin as single agent and/or combination with cytotoxic or hormonal agents” for details.

This diagram illustrates the signal generated by the interaction of ligand and growth factor receptor and the potential elements that can be modulated therapeutically. Numerous downstream effects that can be seen with receptor signaling, primarily mitogenesis, but also changes in motility, metabolism, and even paradoxically, enhancement of programmed cell death can all be seen.

Along with the HER2/neu receptor, the epidermal growth factor receptor (EGFR) has also been a target of interest. EGFR is overexpressed in lung and head/neck cancers as well as a proportion of colon and breast cancer. A monoclonal antibody to EGFR, C225, has activity against EGFR expressing cancer cells in vitro. Human trials that have shown promise include C225 plus radiation for head and neck cancer and C225 plus irinotecan in refractory colon cancer. Small molecule kinase inhibitors that act in a receptor-specific fashion on the cytoplasmic kinase domain can effectively squelch the downstream signaling mediated by the receptor. Two such agents, ZD 1689 (Iressa), and OSI-774 (Tarceva) have shown some activity in lung cancer or head and neck cancer. Since breast cancer cells also express EGFR and since EGFR may heterodimerize with HER2/neu, there are ongoing trials testing these agents in breast cancer. Combinations of these small molecule inhibitors with hormonal therapy are also being tested in breast cancer. Given the potential heterodimerization of HER2/neu and EGFR, a trial of Iressa plus Herceptin is also planned.

The vascular endothelial growth factor receptors (VEGFR-1 and VEGFR-2) mediate angiogenesis and vascular permeability and may have an important role in tumor growth and metastasis. Targeting these growth factor receptors have an anti-angiogenic effect in animal models and are also being tested in the clinic. In a small phase II study of anti-VEGF antibody for advanced breast cancer, a small number of responses were seen with side effects including hypertension and proteinuria. Currently, this agent is being tested in combination with capecitabine or in combination with paclitaxel in two separate randomized trials comparing the combination to chemotherapy alone. Small molecule inhibitors to VEGFR are also being tested in breast and other cancers.

Components of the receptor signaling pathways that may predict responsiveness to such therapies are of great interest since these may optimally choose patients who have a good chance of response. It is unlikely that any single agent will have a dramatic effect on breast cancers universally since a given pathway may not be operative in all cases of cancer. Hence, predictive markers may not only help select the type of drug to be employed in a given patient, but may also identify novel therapeutic targets.

Evolution of Premalignant Lesions from Normal Epithelium to DCIS D. Craig Allred, M.D., Professor of Pathology, Breast Center, Baylor College of Medicine, Houston, TX

Most (probably all) invasive breast cancers (IBCs) arise from certain pre-existing benign lesions over long periods of time. There are many types of benign lesions in the breast and only a handful appears to have significant premalignant potential. The most important premalignant lesions include hyperplastic unfolded lobules (HULs), atypical ductal hyperplasia (ADH), atypical lobular hyperplasia (ALH), ductal carcinoma in situ (DCIS), and lobular carcinoma in situ (LCIS). Several converging lines of pathological, epidemiological, and biological evidence support the premalignant nature of these lesions including: (1) they are on histological continuum, (2) they are much more common in cancerous than non-cancerous breasts, (3) they are risk factors for developing IBC, and (4) they share identical genetic and epigenetic defects with synchronous IBC in the same breast. Currently, premalignant lesions are defined by their histological features. By definition, lesions within specific categories look alike under the microscope but only a relatively small proportion appear to progress to IBC, emphasizing that there must be underlying biological differences causing some to remain stable and others to progress. Understanding this biology has become an important topic in cancer research and there have been a large number of publications in the literature over the past decade. Nearly all have been descriptive or correlative studies of archival (i.e. formalin-fixed and paraffin-embedded) clinical tissue samples because few if any cell lines or animal models exist to support mechanistic studies. Despite these limitations, a great deal has been learned about biological phenomena that appear to be important in the evolution of premalignant disease including growth kinetics (proliferation and apoptosis), the estrogen receptor (ER), oncogenes (e.g. erbB2), tumor suppressor genes (e.g. p53), allelic imbalances, and so on. Alterations in ER expression, structure, and function appear to be particularly important in the early development of premalignant lesions from normal cells and will be mentioned in more detail during the presentation. However, like fully developed IBCs, premalignant lesions as a whole are almost certain to contain a large number of as yet unknown critical biological abnormalities which, hopefully, some of the new high-throughput technologies will help unravel. Understanding this biology is an important goal because it may help identify new prognostic and predictive factors to improve the treatment of patients with premalignant disease and, more importantly, it may identify new therapeutic targets for breast cancer prevention.

Transition from normal-hyperplasia-DCIS-invasion and metastasis using Laser capture microdissection. Clinical implications Emmanuel Petricoin, M.D., Co-Director, FDA-NCI Clinical Proteomics Program, Bethesda, MD

The study of the biological changes that underpin the transition from normal through frank breast cancer has been the focus of intensive genetic studies for the past decade or more. However, since the mechanisms by which the cancer cell arises and survives are ultimately dictated by protein function, proteomics offers a new opportunity to study carcinogenesis. However, because the proteome is in a constant state of flux, and is cell-type and context dependant, simply grinding up a human tumor specimen and analyzing the protein content will not yield the information sought. Cellular heterogeneity confounds these types of analysis as the proteome of stromal cells, infiltrating lymphocytes, normal, DCIS, and cancer epithelial cells cross-contaminate each other. Consequently, we have developed and utilized Laser Capture Microdissection (LCM) to procure pure homogeneous populations of cells directly from breast cancer human tissue specimens to study the protein changes that occur as the transition from normal to frank cancer occurs. Proteomic technologies such as 2D-PAGE separation followed by mass spectroscopy, and implementation of new types of protein arrays are coupled to LCM for discovery of new proteins that may be therapeutic targets or biomarkers for early detection and staging.

Angiogenesis Inhibition; Conceptual Framework and Challenges Kathy D. Miller, M.D., Associate Professor of Medicine, Indiana University, Indianapolis, IN

Angiogenesis, the process of new blood vessel formation, plays a central role in both local tumor growth and distant metastasis. Angiogenesis is a tightly regulated, multiply redundant process required only for wound healing, endometrial proliferation, and pregnancy in healthy adults. Thus the inhibition of angiogenesis offers an attractive therapeutic target with little expected (at least theoretically) toxicity. Rather than a comprehensive review of all agents currently in development, we can conceptually group agents into several categories based on the mechanism of action: protease inhibitors which either directly inhibit or otherwise interfere with the action of proteases critical for invasion, growth factor/receptor antagonists which thwart signaling of pro-angiogenic growth factors, endothelial toxins which specifically target endothelial antigens, and natural inhibitors which stimulate or mimic substances known to naturally inhibit angiogenesis. The clinical experience with each category and potential barriers will be reviewed.

NSABP Chemotherapy Prevention Trials. Richard Margolese, M.D., Herbert Black Professor of Surgical Oncology, McGill University, Montreal, Canada

The traditional model for attacking breast cancer has been to detect the cancer at an earlier stage when it is more curable and to add adjuvant therapies to improve cure rates after surgery. A useful model of breast cancer suggests that DCIS is a pre cursor of invasive cancer and other changes such as atypical hyperplasia are even earlier changes that frequently lead to cancer. Intervening to interrupt the evolution of these changes to invasive cancer with the potential for metastasis is a logical and useful therapy.

In two randomized clinical trials of DCIS treatment, NSABP protocols have shown that lumpectomy with post operative radiation and tamoxifen therapy reduced the rate of invasive breast cancer to approximately 2%

In other reports survival after mastectomy for DCIS varies from 98-99%.

In addition, tamoxifen has shown in the adjuvant therapy setting that it can inhibit the appearance of new cancers in the contralateral breast. This led to the design and completion of a prevention trial involving 13,600 high risk participants. Therapy with tamoxifen reduced breast cancer risks by 50% with acceptable levels of side effects. These two trials illustrate the possibility of controlling the evolution of breast cancer with SERMs which can be considered as growth factor inhibitors. Future research should expand and extend these findings for better control of breast cancer incidence.

hmemon/haldoc/abstract\osman02.abs

Results of the ATAC trial (Arimidex, Tamoxifen and Combination). Implications for chemoprevention. Professor Anthony Howell, CRC Department of Medical Oncology, University of Manchester, UK

The new non-steroidal and steroidal aromatase inhibitors have been shown to be superior to megestrol acetate as second line endocrine therapy and to atamoxifen for first line endocrine therapy for advanced breast cancer. The first adjuvant trial to be reported using an aromatase inhibitor (anastrozole) now indicates that aromatase inhibitors may become the treatment of choice in the adjuvant situation and thus this leads to the suggestion that they may be of value in breast cancer prevention. Anastrozole is a highly selective potent aromatase inhibitor which is non-steroidal and orally active. It is given as a once daily dose (1mg) and there are now over 460,000 patients years experience with this drug. The ATAC trial was initiated approximately 5.5 year’s ago and compares anastrozole 1mg plus tamoxifen placebo versus anastrozole placebo plus tamoxifen 20mg versus a combination of the two drugs. The primary endpoints of the study were disease free survival, safety and tolerability. The secondary endpoints included incidence of new breast (contralateral) primary cancers. The first results of the trial were given at the San Antonio Breast Symposium on 10.12.01 (M Baum). Patients were recruited from 21 countries into the trial between July 1996 and March 2000. In all, 9,366 patients were randomised. Analysis was pre-defined when 1,056 events had occurred. The mean age of this postmenopausal population was 64 years, 64% of patients had T1 tumours and 34% were node positive. Eighty four percent of tumours were oestrogen receptor positive and/or progesterone receptor positive and 9% were receptor unknown.

The remainder were receptor negative. The total number of first events was 1,079, 766 of which were in the receptor positive population. The median duration of therapy was 30.7 months and the median follow up 34.3 months. The Kaplan-Meier curves of disease free survival in the intention to treat the population showed that anastrozole was superior to tamoxifen (HR 0.83, 95.2% CI, 0.71-0.96, p0.0129). The combination was equivalent to tamoxifen (HR 1.02, 95.2% CI, 0.88-1.18, p0.7718). There was a diminution in both locoregional and distant relapses. Analysis of the incidence of new (contralateral) breast primaries showed that anastrozole was markedly superior to tamoxifen (OR 0.52, 95% CI, 0.22-0.79, p0.0068). There was no significant difference between the combination and tamoxifen. There were significantly fewer hot flushes, weight gain, vaginal problems and thromboembolic phenomena in the arimidex arm. However, there were more musculoskeletal disorders and more fractures in the anastrozole alone arm. Thus, the early analysis of the ATAC trial shows a disease free advantage for anastrozole and a marked reduction in the numbers of contralateral tumours compared with tamoxifen or the combination with an improved safety profile. These features, particularly the reduction in contralateral breast cancers, make anastrozole a potentially useful agent for breast cancer chemoprevention. A trial comparing anastrozole with tamoxifen and placebo in women at risk of breast cancer is due to start later this year (IBIS II).

The Intraductal Approach to the Breast Susan M. Love, M.D. MBA, Adjunct Professor of Surgery UCLA, Pacific Palisades, CA

The idea of studying breast cancer by accessing the milk ducts is not a new one. Interest in nipple fluid dates back to the early 1900’s. As early as 1914 Nathan reported a case in which breast cancer was diagnosed on the basis of malignant cells found in nipple discharge. In 1950 George Papanicolaou started a research project with an aim “to ascertain the proportion of asymptomatic women from whom breast secretion could be obtained with a consideration of the possibility of utilizing breast secretion smears in screening for mammary carcinoma.” He applied a suction to the nipple and but was able to obtain fluid from only 18.6% of women. This approach languished until the 1970’s when several groups studied this “breast pap smear” . Sartorius reported on a series of 1503 women from his breast surgical practice and was able to obtain fluid from 55% of subjects. Using his device Buerhing and Petrakis reported similar yields in volunteers. Petrakis and Wrensch contributed greatly to the field over the years by analyzing the fluid for cells as well as biochemical, hormonal and exogenous substances. Their recent twenty year follow up confirmed that women who have atypical cells in their nipple aspirate fluid have a doubling of their risk of subsequent breast cancer. If they have a family history of breast cancer and aypia the relative risk is 4.5 . Most recently Sauter has championed this approach and reported the ability to obtain fluid from 100% of women with multiple attempts.

The first description of ductal lavage and indeed intraductal biopsy was by Raul LeBorgne of Uruguay in the early 1940’s. He describes cannulating the milk ducts and doing ductograms and then collecting the contrast material for analysis. This he called ductal rinse. Sartorius in the 1970’s continued this practice and in 1974 suggested that “ perfusion of the non-productive ducts with isosmotic, sterile saline solution via nylon catheters followed by nipple aspiration, is a ‘cell washing” technique found to be extremely effective in procuring cells from the terminal ramifications of the duct structure.” Of 27 women suspected of having carcinomas on cytology, eighteen were found to have cancer. Seven of these were identified on the basis of cytology alone. All of the lesions were smaller than 8mm.In 49 cases of carcinoma diagnosed by another modality no fluid could be aspirated from the breast. Sartorius postulated that larger tumors had disrupted the ducts and that the fluid was no longer contained.

It was this work of Sartorius which led me to explore the concept of ductal lavage. Both Sartorius and LeBorgne had instilled fluid into the ducts and then removed the catheter and aspirated the fluid they squeezed out of the nipple from its surface. In order to be sure which duct was abnormal, I felt we needed a technique to cannulate a single duct and aspirate the fluid from that duct. The ProDuct catheter was developed on that premise. The current application of ductal lavage has been presented already by Darius Francescatti and will be reviewed in the satellite session this evening.

Parallel to the ductal lavage development has been the exploration of ductoscopy. First described by Teboul in association with ultrasound of the ductal lobular units, this ability to directly visualize the lining of the ducts takes the intraductal approach to another level. Several researchers in Japan, the US, Spain, Germany and Argentina have employed ductoscopy in women with spontaneous nipple discharge. Our early limited work with this technique in non discharging ducts has been followed by a report from Dooley in Oklahoma, who has used ductoscopy to investigate abnormal findings on lavage.

The ability to access the ductal systems has great potential to help in our understanding of the physiology of the non lactating breast. The non lactating breast appears to be metabolically active. It has immunoglobulins, proteins, carbohydrates and significant levels of cholesterol and its metabolites at different levels than those found in the blood. Breast duct fluid has been shown to contain nicotine within a half an hour of a woman smoking a cigarette. Small studies have demonstrated PSA, CEA, LDH, P 53, and Her 2 neu in the fluid. Sukumar’s group has demonstrated the ability to detect methylation of genetic markers in breast duct fluid and the NCI has demonstrated the ability to perform mutagenesis assays, amplify DNA and perform protein electrophoresis on nipple aspirate fluid.

The microenvironment of the ductal lobular system is undoubtedly an important factor in the development of breast cancer. Estrogen, testosterone, prolactin, progesterone, and dehydroepiandrosterone sulfate have been found in the breast fluid. Interestingly hormone levels in the fluid are independent of the blood. In premenopausal women the levels of estrogen do not reflect the menstrual period but are much higher. Petrakis found that they go down with pregnancy and breast feeding and then gradually climb over the next four years. They remain high in postmenopausal women even after serum levels have dropped and can be as much as forty times higher. This may reflect the presence of aromatase in the breast ducts. A report from Harding et al demonstrated that proteins normally stimulated or inhibited by estrogen could be measured in the breast duct fluid and be used to demonstrate the effects of Tamoxifen, HRT and goserelin. We have made significant progress in understanding and treating cancers of the cervix and colon, in part because we have had access to where they start. It is my hope that the intraductal approach to the breast will lead to equivalent advances.

Selected references:

1. Love SM, Barksy Sh Breast-duct Endoscopy to Study Stages of Cancerous Breast Disease Lancet (1996) 348:997-9.
2. Petrakis NL Physiologic, Biochemical, and Cytological Aspects of Nipple Aspirate Fluid. Breast Cancer Res Treat (1986) 8:7-19.
3. Phillips HA, Howard GCW, Miller WR Nipple aspirate fluid in relation to breast cancer: A Review. The Breast (1999) 8, 169-174.
4. Sartorius OW, Smith HS, Morris P, Benedict D, Friesen L Cytologic Evaluation of Breast Fluid in the Detection of Breast Disease J Natl Cancer Inst (1977) 59:1073-1080
5. Wrensch MR, Petrakis NL, Miike R, King EB, Chew K, Neuhaus J, Lee MM, Rhys M Breast Cancer Risk in Women With Abnormal Cytology in Nipple Aspirates of Breast Fluid J Natl Cancer Inst (2001);93:1791-8.

Ultrasound-guided needle biopsy of non-palpable breast masses Bruno D. Fornage, M.D., Professor of Radiology & Surgical Oncolory, M.D. Anderson Cancer Center, University of Texas, Houston, TX

Because of its unique real-time capability, sonography (US) has become the preferred modality for guiding needle biopsy of nonpalpable breast masses. Virtually any nonpalpable breast lesion that is clearly seen on sonograms can be sampled with a needle under US guidance. Both fine-needle aspiration biopsy (FNA) and core-needle biopsy (CNB) are effectively guided by real-time US.

Fine-Needle Aspiration
Standard 20-gauge, 1.5-inch (3.8-cm) hypodermic needles are used for FNA. Rarely, a 2-inch (5-cm) needle is needed because the lesion is deep and/or the breast is large. The standard technique of needle insertion for FNA is oblique insertion, in which the needle is inserted from the end of the transducer along the scan plane with an obliquity that depends on the depth of the target. In experienced hands, the oblique insertion technique is 100% accurate and safe.

Cysts and other fluid collections can be readily aspirated with a fine needle. On occasion, an 18-gauge needle may be needed to drain an inspissated cyst, which typically contains toothpaste-like material. In such a case, it may not be possible to drain the cyst completely. Other collections that can be subjected to diagnostic FNA or percutaneous drainage in the proper clinical setting include postoperative hematomas, lymphoceles, and abscesses.

FNA of infiltrating ductal carcinomas usually yields highly cellular cytologic specimens; with US guidance, the diagnosis is established with a single pass in the majority of cases. Other forms of breast malignancy, such as medullary or mucinous carcinomas, lymphomas, or metas-tases to the breast from extramammary primary cancers, can also be correctly diagnosed cytologically. Hormonal receptors and proliferation markers including Her-2/Neu can be assessed in fine-needle aspirates from breast cancer, although this is usually done on CNB specimens. With strict criteria, the diagnosis of fibroadenoma can be reliably established cytologically. FNA can also readily establish the diagnosis of fat necrosis, acute inflammation, or intramammary lymph nodes.

The major limitation of FNA is its operator dependence. When FNA specimens are insufficient, the operator should switch to CNB.

Core-Needle Biopsy
Numerous commercially available devices provide automatic propulsion of a cutting needle with a throw of about 2 cm. Because of the throw of the needle, the cutting needle must be inserted as horizontally as possible to avoid any injury to the chest wall. When the transfixion of the target has been clearly documented with US and cores appear to be of satisfactory size, no more than three or four cores are needed for diagnosis.

In rare cases, US-guided CNB is done to sample an area of microcalcifications without a discrete mass. In this case, it is imperative to process the cores the same way they would be if the biopsy were done under stereotactic guidance, i.e., the cores obtained must be radiographed using a mammographic unit and magnification technique (or dedicated equipment for specimen radiography) to document the presence of the calcifications in the cores.

Other Biopsy Devices That Can Be Used under US Guidance
A handheld version of the 11-gauge Mammotome device (Ethicon Endosurgery) has been designed for use with US guidance. Advantages of the device include a single insertion of the device, convenient automatic core retrieval, and multiple large contiguous samples with the potential to remove small masses entirely. In the United States, the device has been approved for biopsy, but not for therapeutic purposes. Disadvantages of the Mammotome include the large diameter of the cutting device, the greater associated trauma, and the relatively high cost.

A new breast biopsy device, en-bloc(tm) (Neothermia), has been designed to cut with radiofrequency and retrieve a large (1-cm diameter) intact specimen with a capture snare. While such large-bore biopsy devices and the large volume of the cores they yield are well adapted to stereotactically guided biopsy of microcalcifications, their cost-effectiveness compared with standard Tru-cut needles for the biopsy of breast masses remains to be demonstrated.

Keys to Success of US-Guided Needle Biopsy
US-guided needle biopsy of nonpalpable breast lesions requires teamwork. Progress can be made and experience accumulated only through ongoing communication between the radiologist, the pathologist, and the surgeon. US-guided biopsy procedures require excellent eye-hand coordination and a significant amount of practice before the mandatory 100% accuracy level in hitting the target can be reached. Practicing with easy-to-make phantoms shortens the learning curve of beginners. The golden rule in breast biopsy is the concordance between the results of a biopsy and the imaging and clinical findings. Any discrepancy, e.g., a negative result of the needle biopsy in the face of a single suspicious finding-physical, mammographic, or US-should not delay surgical excision.

References

• Fornage BD, Faroux MJ, Simatos A. Breast masses: US-guided fine-needle aspiration biopsy. Radiology 1987;147:409-414.
• Fornage BD, Sneige N, Faroux MJ, et al. Sonographic appearance and ultrasound-guided fine-needle aspiration biopsy of breast carcinomas smaller than 1 cm3. J Ultrasound Med 1990;9:559-568.
• Fornage BD, Coan JD, David CL. Ultrasound-guided needle biopsy of the breast and other interventional procedures. Radiol Clin North Am 1992;30:167-185.
• Fornage BD. Ultrasound-guided percutaneous needle biopsy on nonpalpable breast masses. In: Harris JR, Lippman ME, Morrow M, Hellman S (eds). Diseases of the breast. Philadelphia, Lippincott-Raven, 1996, pp. 152-158.
• Fornage BD. Sonographically guided needle biopsy of nonpalpable breast lesions. J Clin Ultrasound 1999;27:385-398.
• Fornage BD, Sneige N, Edeiken BS. Interventional breast sonography. Eur J Radiol, in press.
• Parker SH, Jobe WE, Dennis MA, et al. US-guided automated large-core breast biopsy. Radiology 1993;187:507-511.

MRI in Planning DCIS Management Steven E. Harms, MD, FACR, Professor of Radiology, University of Arkansas for Medical Sciences, Little Rock, AR, and Medical Director, Aurora Imaging Technology, North Andover, MA

Although high sensitivity is a common attribute in almost all breast MRI papers, the gold standard for this validation is highly variable. Most breast cancer studies used biopsy as a gold standard. With this methodology, only the lesion actually sampled is confirmed. The ability to detect cancer in the breast that is not sampled with the biopsy is not possible. Therefore, this method inherently underestimates false negative breast MRI findings. We know that routine pathology misses additional cancers in 40% of breasts.(1-3) Therefore, in order to truly measure for the potential of false negative breast MRI, mastectomy specimens must be serially sectioned and rigorously analyzed by pathology. This analysis is limited to only a few MRI studies.(4) Without this validation, the true negative predictive value of breast MRI is not known.

Although it has not been proven with serial section mastectomy correlations, it is likely that dynamic, low-resolution breast MRI misses cancer. Most of the reported previously reported studies show an under-reporting of DCIS. Many of these studies cite less than 10% DCIS in the population studied.(5-7) Recent screening mammography experiences report up to 40% representation by DCIS. These data indicate that dynamic, low-resolution breast MRI is missing DCIS. Many experts now conclude that dynamic, low-resolution breast MRI cannot reliably detect DCIS and only cite specificity statistics for infiltrating carcinoma.(8) A major potential limitation of breast MRI is the inability to detect DCIS. This limitation is solvable with high contrast, high resolution images.

High resolution, fat-suppressed breast MRI has been validated using serial section mastectomy analysis for a proven negative predictive value approaching 100%. We evaluated RODEO MRI for the ability to detect DCIS and found it to be highly effective.(9) Because RODEO uses narrow bandwidth RF, microcalcifications can be seen as tiny foci of magnetic susceptibility effect intermixed with the typical clumped enhancement of DCIS.

Because of the high quality definition of DCIS by RODEO MRI, we are now using this information in combination with stereotaxic localization to mark the margins of DCIS extent prior to surgery. DCIS typically does not involve a spherical volume. DCIS usually extends within the ductal system involving a segment or adjacent segments. The ducts are arranged radially extending from the nipple to the chest wall. Therefore, a surgical excision that removes a spherical volume of tissue may likely result in positive margins. RODEO MRI can identify the margins of the DCIS extent for better resections.

Localization wires are a problem for MRI. Although MRI compatible wires are abundant, the access to the MRI relative to surgical procedures is often difficult. We have many patients from other institutions, even out-of-state who are referred for localization. Leaving a wire in place for a long enough period to travel to distant hospitals for surgery is inadvisable. For this reason we developed another system that is working well..

Instead of wire localizations, we typically perform hematoma localizations. For this procedure needles are placed around the lesion to mark the margins. The correct identification of margins is confirmed on a MR image. Approximately 2 cc of the patients’s own blood is injected into each site. The hematoa shows up as a bright spot on the immediate MR image after hematoma localization. The hematoma is identified at surgery with ultrasound and visually when exposed in the surgical field. No needles are wires are left in place to risk infection. The hematoma lasts for days. The patients are free to travel to other institutions without restriction.

References

1. Rosen PP, Fracchia AA, Urban JA, et al. “Residual” mammary carcinoma following simulated partial mastectomy. Cancer. 1975;35:739-47.
2. Holland R, Veling SHJ, Mravunac M, Hendriks JHCL. Histologic multifocality of Tis, T1-2 breast carcinomas: implication for clinical trials of breast-conserving surgery. Cancer. 1985;56:979-90.
3. Holland R, Connolly JL, Gelman R, et al. The presence of an extensive intraductal component following a limited excision correlates with prominent residual disease in the remainder of the breast. J Clin Oncol. 1990;8(1):113-8.
4. Harms SE, Flamig DP, Hesley KL, et al. Breast MRI: rotating delivery of excitation off-resonance: Clinical experience with pathologic correlations. Radiology 187:493-501, 1993.
5. Kaiser WA, Zeitler E. MR imaging of the breast: fast imaging sequences with and without Gd-DTPA. Radiology. 1989;170:681-6.
6. Heywang SH, Wolf A, Pruss E, Hilbertz T, Eiermann W, Permanetter W. MR imaging of the breast with Gd-DTPA: use and limitations. Radiology. 1989;171:95-103.
7. Kuhl CK, Mielcareck P, Klaschik S, et al. Dynamic breast MR imaging: Are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 211:101-110, 1999.
8. Kuhl CK, Schmutzler RK, Leutner CC, Kempe A, Wardelmann E, Hocke A, Maringa M, Pfeifer U, Krebs D, Schild HH. Breast MR imaging screening in 192 women proved or suspected to be carriers of a breast cancer susceptibility gene: preliminary results. Radiology 2000 215: 267.
9. Soderstrom CE, Harms SE, Copit DS, et al. 3D RODEO breast MRI of lesions containing ductal carcinoma in situ. Radiology 201:427-432, 1996.

Treatment of Ductal Carcinoma in Situ Summary of the Consensus Conference, Philadelphia, April, 1999. Gordon F. Schwartz, MD, MBA, Professor of Surgery, Jefferson Medical College, Philadelphia, PA

The goal of treatment for DCIS is breast conservation, with optimal cosmesis and with a minimum risk of subsequent invasive or in situ recurrence. There are some women for whom mastectomy remains the optimal treatment, but most women with DCIS are candidates for breast conservation. Each patient should be apprised of her own situation and each option discussed with her in detail, including local excision and radiation therapy, local excision alone, or mastectomy. Regardless of the goal of breast conservation, mastectomy is an acceptable treatment options for patients with DCIS, irrespective of their eligibility for breast conservation. A minority of patients with DCIS requires mastectomy, probably less than 25% of patients with this diagnosis, but mastectomy may be performed if it were the patient’s preference.

The majority of women with DCIS, therefore, are candidates for breast conservation. Whether radiation therapy, surveillance alone, or either of these plus tamoxifen, is optimal treatment when breast conservation is employed, is a major topic of current discussion. There are groups of patients with DCIS who fall into each of these categories, but no one has yet defined the selection criteria for each of them precisely enough to make dogmatic recommendations. Clinical trials have shown that local excision and radiation therapy in patients with negative margins provide excellent rates of local control. Patients treated by excision alone have a greater chance of local recurrence. This risk decreases with wider surgical margins around the area of DCIS. Although adding radiation therapy to wide local excision benefits all groups of DCIS patients who are candidates for breast conservation, the magnitude of that benefit may be small enough in some patient sub-groups that radiation can be omitted. However, patients who may avoid radiation therapy have not been reproducibly and reliably identified by any clinical trials, but there are institutional and individual reports of large series of such patients treated by wide excision alone who achieve a 1% or less per year risk of invasive recurrence, which is the reported occurrence of invasive cancer following local excision and radiation.

Clear surgical margins are a major criterion for treatment of DCIS by breast conservation (whether or not radiation therapy is employed); the wider the margin, the lower the rate of local recurrence. Margin status is crucial importance because it is the one variable that the physician can control. A 10 mm margin the best compromise between removal of so much tissue that the cosmetic result would be less than desirable and the likelihood of local recurrence.

An axillary dissection is not required in patients with DCIS. The incidence of occult invasive carcinoma that might have spread to axillary nodes is so infrequent that this possibility should not provoke a therapeutic recommendation to dissect the axilla. For patients undergoing mastectomy, some lymph nodes situated in the axillary tail of the breast may be removed incidentally, but an intentional dissection of the axilla is not warranted.

The role of sentinel node biopsy in DCIS is also a major contemporary question. In general, sentinel node biopsy should be considered only in the context of a peer-reviewed clinical trial. Although sentinel node biopsy, using either radioactive isotope or blue dye to locate the node(s), is less invasive and less morbid than the customary axillary dissection performed for invasive cancer, the ease of sentinel node biopsy should not be an excuse for its use. If, by definition, DCIS does not spread to lymph nodes, a procedure to address lymph node status should not be necessary, however free from complications it might be.

There is no indication for adjuvant cytotoxic chemotherapy in patients with DCIS. In patients undergoing mastectomy, in the unlikely event that a focus of invasive carcinoma is found in the specimen, the patient should be treated as for any other stage I carcinoma of the breast of the same size.

A randomized clinical trial has demonstrated that the addition of tamoxifen, 20 mg daily for five years, to the treatment of DCIS by local excision and radiation therapy decreased the incidence of invasive cancer recurrence. This benefit was seen independent of margin status or the presence of comedo-type necrosis. However, no overall survival benefit was observed over the course of the study. Whether these observations are justification for the routine use of tamoxifen in all DCIS patients remains unclear, since the sub-groups of patients who would benefit the most have not been defined.

Treatment of recurrence after breast conservation is another crucial question, since recurrence as invasive cancer is an endpoint that may affect disease specific survival. If invasive cancer is the recurrence, its treatment should be as for any other invasive cancer of the same stage. When DCIS only is the recurrence, a different question exists. When DCIS only (i.e., without invasive cancer) occurs after treatment by local excision alone, the usual treatment is radiation or mastectomy, based upon the same selection criteria as for DCIS treatment when it occurs for the first time. Therefore, for women who underwent local excision only, the choices include re-excision, re-excision and radiation therapy, or mastectomy. For women who underwent radiation therapy as part of breast conservation initially, the usual treatment for recurrence, even if the recurrence was DCIS only, is considered to be mastectomy. There still may be some patients, depending upon their own preference and understanding of subsequent risk, who prefer to be treated by local excision alone in this situation, recognizing that their chance of future recurrence is higher than they were previously faced with the initial diagnosis and chose radiation.

As the conference concluded, the panelists unanimously agreed that the treatment of DCIS is constantly being refined, and the observations and recommendations made at any time may be influenced by new data reported almost contemporaneously as well as in the future. Because this report was based upon the expert opinions and individual experience of the participants as of the date of the conference, these opinions must not be construed as dogmatic guidelines for treatment. Management of individual patients should be based upon each patient’s unique clinical circumstances. Moreover, the subsequently published consensus report was not intended to establish specific standards of care or to be used as a guideline for diagnostic or treatment management decisions by third-party payers.

Back | Top of Page