Julia Draznin Maltzman, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 3, 2003
October is national breast cancer awareness month and it is estimated 210,000 women will be diagnosed with the disease in 2003. It is the second leading cause of cancer death in North American women. Although significant progress has been made in identifying risk factors, most cases occur in patients without major predictors. This is unfortunate, because the key to curing breast cancer is its early detection and initiation of treatment. To that end, screening protocols have been paramount in breast cancer. A good screening test tries to find the disease before there are symptoms. The conventional approach to screening for breast cancer has been three fold: self-breast exams, physical exams by a health care provider, and mammograms.
Mammograms are an important tool that doctors use to diagnose breast cancer. A mammogram is a special x-ray of the breast. The breast is placed on a flat surface while firm pressure is applied via a compression device. Compressing the breast is necessary as it flattens the breast and enables the doctor to see it in its entirety. However, mammograms are not perfect and normal breast tissue can hide a malignancy. This is called a false negative when a cancer is present but it is not shown on the mammogram. Furthermore, a mammogram can pick up an abnormality that may look like cancer but turn out to be benign. This, by contrast, is called a false positive. False positives can be damaging or even harmful as they lead to unnecessary biopsies and other invasive interventions, let alone the emotional roller coaster ride that it creates. Advances in technology, however, have led some doctors to question the whether other imaging techniques may be better screening tools for breast cancer.
Magnetic Resonance Imaging (MRI) is a technique that uses radio waves and magnetic fields to map out abnormalities in the body. Patients lay on a narrow table while they are advanced into a tube that contains a very high power magnetic field. MRI's are often able to identify lesions that mammograms miss. Similarly, because the MRI is considered more sensitive, it leads to many false positives. Cost is also an issue. An MRI costs about 10 times more than a mammogram.
Traditionally breast cancer screening has been accomplished with mammography. However, recently doctors have wondered whether MRI may be a better and a more powerful tool. This has been in the forefront of many discussions in the past few years and, in fact, was highlighted in the plenary session at the American Society of Clinical Oncology (ASCO) meeting last May covered by OncoLink.
I. The US - Memorial Sloan Kettering Cancer Center Study
This small trial studied only BRCA-1 and BRCA-2 mutation positive women. (This population is considered to be at great risk of developing breast cancer). In a five-year follow-up period, 54 women were asked to have a screening breast MRI in addition to their routine mammograms. Ninety-seven MRI's were completed for screening purposes. In addition, 28 MRI's were done for follow-up reasons, indicating that these women had an abnormal MRI within the preceding six months. Finally, four more MRI's were conducted at the time of breast cancer diagnosis in the opposite breast or immediately before a prophylactic mastectomy. In total, 129 MRI examinations were completed in this study. Twenty-one biopsies were recommended based on these 129 studies. Only three breast cancers were diagnosed. One invasive breast cancer and two high-risk lesions make up the three diagnoses. All three of these cancers were detected by the MRI test, whereas all three were missed by mammography. Based on these data, researches concluded that MRI is 100% sensitive if there is cancer, MRI will find it; and 83% specific - if the MRI scan is negative, however, then there is an 83% that there is no cancer.
II. The Dutch Study
This is the largest, prospective, non-randomized, multi-center trial done to date. Women with either a strong family history or those harboring the BRCA-1 or BRCA-2 mutations were followed for 2 years with annual mammography and MRI scans, as well as bi-annual physical examinations. Of the 1,874 women who participated in this study, 41 breast cancers were diagnosed. Of these, 34 were invasive carcinomas, six high-risk lesions, and one lymphoma. Mammography detected 26% of the invasive cancers; but the MRI was able to find 83%.
Study researchers concluded that MRI specificity was 88%. (Mammogram specificity was calculated to be 95%). This study also found the sensitivity of the MRI to be only 71%. The sensitivity of mammography in this trial, however, was also quite low -- 36%.
Trial investigators were quick to point out that 42% of the tumors detected by MRI were less than 1cm in size. Only 13% of the tumors seen by mammogram were that small. In addition, three fourths (76%) of the tumors found by MRI did not have lymph node involvement. These date suggested that MRI might be useful in the diagnosis of early stage disease, which is considered curable.
III. The German Study
This interesting study went a step beyond the other two, in that it tried to compare and contrast mammograms with MRI's and ultrasounds. There were 462 high-risk women enrolled to receive annual physical exams, mammograms, high-resolution ultrasounds, and MRI's. This study, with five years of follow-up data, yielded 51 diagnoses of breast cancer in 47 different women. All but two cancers were picked up by MRI. By contrast, mammography and ultra sonography were able to detect less than half of the 51 cancers. The two cancerous lesions not seen by MRI were also not visible on ultrasound or mammography, but rather, incidentally found upon prophylactic mastectomy.
In this study, the MRI sensitivity was 96%, significantly higher than mammography or ultrasonography, 43% and 42% respectively. This study also touted 95% specificity for the MRI, noting that it led to the least number of unnecessary breast biopsies. The specificity of mammography was equally high in this trial (94%). Investigators offered their twelve-year experience as the reason for such high specificity rates. These researchers concluded that an MRI has better diagnostic ability than a mammogram and concluded that it may be safer for women with BRCA-1 and BRCA-2 mutations. The BRCA gene defects affect tumor suppression, making it more difficult for women with these mutations to fight the mutagenic effects of ionizing radiation that is emitted with mammography. MRI's have no ionizing radiation and maybe a better tool for these high-risk patients.
Table 1: Sensitivity and Specificity of the MRI as a screening method
Dr. Elizabeth Morris, a radiologist at Memorial Sloan-Kettering Cancer Center, looked at the frequency of biopsies following screening breast MRI's. This study, presented at the American Roentgen Ray annual meeting in San Diego, looked at 367 high-risk women who had normal physical examinations and mammograms who underwent screening MRI's. As a result, 64 patients were recommended a biopsy. For unrelated reasons, five women did not get biopsies. Of the 59 who did, 14 patients (24%) were found to have invasive cancer and 13 women (22%) had high-risk lesions.
Dr. Morris was the invited discussant for the ASCO plenary sessions. She reviewed all currently published studies that looked at MRI's as a breast cancer screening tool. When examined as an aggregate, she explained, 103 cancers were diagnosed in 3,482 women. Dr. Morris cautioned that these studies varied greatly in their inclusion criteria and techniques and thus difficult to compare, but the percent of invasive cancers detected only by MRI ranged from 2-6%. She did note, however, that lymph node negative cancers and small cancers (less than two centimeters) are indeed seen more often on MRI. MRI may be a good screening tool for high-risk women who chose not to have a prophylactic mastectomy, but with the current evidence and data, it does not replace mammography. More studies are needed to change the current standard of care.
The German investigators, however, suggested that the MRI be used as pre-operative screen for high-risk women going for mastectomies, to ensure that the opposite breast does not have other cancerous lesions not detected by a pre-operative mammogram.
There is general agreement that screening MRIs should not be done outside of a center that has much experience with this tool. MRI interpretation requires skill and precision held by radiologists who sub-specialize in breast imaging. Radiologists who do exclusively breast imaging are usually located in a big tertiary care institution and not in smaller hospitals or private clinics. In fact, the German study touted success due to their high experience in breast MRIs. Furthermore, an MRI machine is large, expensive, and not widely available outside of a large medical center. Even in these centers, there are not enough machines to perform screening for the general population. Cost is also an issue. MRI's cost 10 times more than a mammogram. Since there are no convincing data that an MRI is indeed better than a mammogram, then society stands to incur inordinate costs by instituting the MRI as a screening method.
MRI's can lead to many false positives and this in turn leads to many unnecessary biopsies. Surgical biopsies or core needle biopsies can result in significant morbidity and the psychological and emotional anxiety that patients experience cannot be quantified. Because of the high rate of false positivity, screening MRIs were only evaluated in high-risk populations such as women with the BRCA-1 or the BRCA-2 mutation carriers. Outside of a clinical trial, screening MRIs should not be done on the general population.
Technology is constantly improving and both mammography and MRI's are getting better. Each is now able to simulate digital three-dimensional images that can be enlarged for better visualization. Two studies presented at the 2003 Roentgen Ray Society meeting in San Diego showed that digital breast imaging technology may be better than standard plain film mammography. Soon, there may be a plethora of screening techniques. What is clear, however, is that early detection and prompt treatment can save lives. Therefore, research must continue to answer these and other important questions.