| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University of California, Los Angeles, 200 UCLA
Medical Plaza, Rm. 165-47, Box 956952, Los Angeles, CA 90095.
2 Department of Radiology–Breast Imaging, Washington University School of
Medicine, St. Louis, MO.
3 Department of Radiology–Breast Imaging, University of California at San
Francisco, San Francisco, CA.
Received September 25, 2007;
accepted after revision March 4, 2008.
J. Eradat was supported by training grant 5M01 RR000827 from the National
Institutes of Health.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We invited the 1,696 active physician members of the Society of Breast Imaging to participate in a survey addressing whether and how they performed and interpreted breast MRI. Respondents were asked to select one member of their practice to complete the survey. A total of 754 surveys were completed. Every respondent did not reply to every question.
RESULTS. Contrast-enhanced breast MRI was offered at 557 of 754 (73.8%) practices. Of these, 346 of 553 (62.6%) performed at least five breast MRI examinations per week, and only 56 of 553 (10.1%) performed > 20 per week. Radiologists qualified under the Mammography Quality Standards Act supervised the performance of and interpreted breast MRI in the majority of facilities. Of 552 respondents, breast MRI was interpreted as soft copy with computer-aided detection (CAD) in 280 practices (50.7%), as soft copy without CAD in 261 (47.3%), and as hard copy in 11 (2.0%). Of 551 respondents, 256 (46.5%) never and 207 (37.6%) rarely interpreted breast MRI without correlating mammography or sonography findings. The majority of respondents never (269/561, 48.0%) or rarely (165/561, 29.4%) interpreted breast MRI performed at an outside facility. Screening breast MRI was offered at 359 of 561 (64.0%) practices. Of the practices performing contrast-enhanced examinations, 173 of 557 (31.1%) did not perform MRI-guided interventional procedures.
CONCLUSION. Contrast-enhanced breast MRI is now widely used in the United States. The information gained from this survey should provide reasonable approaches for the development of professional practice guidelines.
Keywords: breast imaging • breast MRI • practice of radiology
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There are other problems. First, there is difficulty interpreting examinations performed at other facilities because of the lack of standardized technical methods for performing the examination (e.g., different T1/T2 sequences), different preferences for scanning planes (e.g., sagittal vs axial vs coronal), some examinations are recorded on hard copy whereas others are recorded on soft copy, some are designed to be interpreted using computer-assisted programs, and portable media (e.g., CDs) are not always readable at other facilities. Second, indications for performing breast MRI are inconsistent from one practice to another. Third, some facilities that perform breast MRI do not have the equipment or expertise to perform MRI-guided biopsies of lesions seen only on MRI.
The purpose of our study was to investigate the protocols of radiology practices throughout the United States regarding the performance and interpretation of breast MRI. Using a survey we developed, we obtained information regarding the type and characteristics of radiology practices, availability of breast MRI, and breast MRI techniques and protocols. It will be important to have this information to develop national practice guidelines, as well as reasonable local approaches, for the performance of this relatively new procedure.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The survey was primarily Web-based using SurveyMonkey [9]. Using the e-mail addresses listed in the SBI directory, a request to participate in this survey was sent to the 1,696 active SBI physician members. The participants were directed to the survey Website. Respondents were instructed to assign only one radiologist from their practice to complete the survey.
Along with the invitation to participate in this survey, recipients were informed that the purpose of the study was to better understand the current trends and practices of imaging centers in the United States with respect to breast MRI. In addition, they were informed that the results of this study might be an important step in the development of practice guidelines and eventual standardization of the utilization and technical performance of breast MRI. The message also stated that participation in the study was completely voluntary, any information obtained in connection with the study would remain confidential, and all responses would be kept anonymous.
A total of 632 surveys were completed online. Another 122 surveys were received by mail, for a total of 754 completed surveys. Each survey consisted of 25 questions about a facility's practices regarding breast MRI (Appendix 1). More specifically, we asked whether and how the respondents performed and interpreted breast MRI. Not every respondent answered every question on the survey, and this was noted. The survey was developed by several of the authors who are involved in breast imaging education and who are familiar with current issues regarding breast MRI. The results of the completed surveys were evaluated by all of the authors, including the study statistician. The 25-question survey covered various topics, including the type and characteristics of the radiology practice, availability of breast MRI, and breast MRI techniques and protocols.
|
Type and Characteristics of the Radiology Practice
The purpose of these questions was to determine the types of radiology
practices that took part in this survey and to determine any correlation
between the type of practice and the use of breast MRI. Questions included the
type of radiology facility, the number of overall radiologists in the
practice, the number of facilities covered by the practice, the number of
radiologists who read mammograms, and the number of mammography examinations
performed annually.
Availability of Breast MRI
The purpose of these questions was to determine the extent of breast MRI
use across the country. Each participant was asked whether the facility
offered breast MRI, and if so, for what purpose. Those respondents who did not
offer contrast-enhanced breast MRI in their practice were asked to give
reasons.
Breast MRI Techniques and Protocols
The purpose of these questions was to ascertain the scanning protocols of
imaging centers with respect to breast MRI. Questions included the number of
breast MRI examinations performed weekly on average at a particular center;
the indications for a diagnostic breast MRI; whether screening breast MRI was
performed; and if so, what the indications were. Additional questions
regarding the scanning protocols included the scanner field strength type, the
imaging plane used for dynamic contrast scans, whether unilateral or bilateral
breast MRI was used, and whether compound (parallel) imaging was used. Another
inquiry pertained to who supervised and who interpreted breast MRI
examinations performed at the imaging center.
Whether a facility used a hard-copy or soft-copy workstation with or without computer-aided detection (CAD) to interpret breast MR examinations was asked. How often breast MRIs were interpreted without correlating with mammography or sonography findings, whether technologists who performed mammography also performed breast MRI, and if "second-look" sonography examinations for unexpected suspicious breast MRI findings were performed, were all asked.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Type and Characteristics of the Radiology Practice
The practices in this survey included 334 hospital-based outpatient
facilities, 323 private outpatient facilities, 300 facilities solely within a
hospital, 144 facilities in academic settings, 25 county facilities, and seven
military or veterans' facilities. Numerous practices contained more than one
facility, which explains why the total number of facilities exceeds 754.
Regarding the number of radiologists in these practices, 90 had 1–3
radiologists, 32 had 4–5, 114 had 6–10, 90 had 11–15, 100
had 16–20, and 292 had more than 20. In these practices, the number of
radiologists reading mammograms was 1–3 in 162 practices, 4–5 in
160, 6–10 in 238, 11–15 in 86, 16–20 in 39, and more than 20
in 34.
Availability of Breast MRI
Breast MRI was offered at 569 (75.5%) of 754 facilities responding to this
question. Of these, 12 (1.6%) performed only MRI without contrast enhancement
for implant evaluation. Of those responding facilities that did not offer
contrast-enhanced breast MRI, 85 (43.2%) of the 197 responding to this
question indicated a program is anticipated but not yet in place, 52 (26.4%)
did not have MR equipment, 11 (5.6%) reported lack of radiologist expertise in
MRI, 11 (5.6%) were not interested in incorporating breast MRI into their
current practices, and 38 (19.3%) gave other reasons for not performing breast
MRI.
Breast MRI Technique and Protocols
Of the 553 facilities responding to the question, 207 (37.4%) did < 5
per week, 167 (30.2%) did 5–10, 79 (14.3%) did 11–15, 44 (8.0%)
did 16–20, 22 (4.0%) did 21–25, 10 (1.8%) did 26–30, 8
(1.4%) did 31–35, 11 (2.0%) did 35–50, and 5 (0.9%) did >
50.
Regarding the usual indications for diagnostic breast MRI, 554 practices responded, and Figure 1 shows the responses from those 554 practices. The most common indications were to evaluate the extent of disease (518 responses, 93.8%) and an equivocal diagnostic imaging workup (500 responses, 72.2%).
Regarding the indications for breast MRI screening, Figure 2 shows the results from 375 practices that did perform screening. The most common indication was patients who were positive for the BRCA gene (339, 90.4%), followed by patients with a mother or sister with premenopausal breast cancer (280, 74.7%).
|
|
Regarding the questions of who supervises and who interprets breast MRI examinations in the responding practices, refer to Figure 3, which illustrates the percentages of various radiologists who compose these two groups.
|
3-T magnets, 16 (2.9%) used
equipment designed specifically for breast MRI, and seven (1.3%) used 
1-T
magnets. Many facilities reported the use of more than one type of magnet,
which explains why the absolute numbers exceed 550.
Of 545 respondents, 492 (90.3%) always did bilateral breast MRI. Three hundred ninety-eight (73.0%) performed dynamic contrast-enhanced scans in the axial plane, 321 (58.9%) in the sagittal plane, and 79 (14.5%) in the coronal plane. Compound imaging was used by 48 (8.8%).
Of 552 respondents regarding how breast MRI examinations are interpreted in their practices, 261 (47.3%) indicated at a softcopy workstation without CAD, 239 (43.3%) indicated a soft-copy workstation always using CAD, 41 (7.4%) indicated a soft-copy workstation sometimes using CAD, and 11 (2.0%) indicated using hard copy.
Table 2 addresses specific practice protocols. Of 551 practices responding to the question of whether they interpreted breast MRI without correlating with mammography or sonography, 359 (46.5%) answered "never" and 207 answered "rarely." Concerning whether 512 practices performed MRI-guided biopsy or localization based on an outside MRI without repeating the MRI, 359 (70.1%) reported "never" and 89 (17.4%) "rarely." Of 554 practices, 297 (53.3%) performed "second-look" sonography for unexpected suspicious breast MRI findings, 256 (70.1%) indicated "never," and 89 (17.4%) "rarely." Regarding whether they interpreted outside breast MRI examinations for referring physicians, 269 (48.0%) answered "never" and 165 (29.4%) "rarely."
|
Concerning whether technologists who performed mammography examinations also performed breast MRI, of the 552 responding practices, 501 (90.8%) answered "never," 18 (3.3%) "rarely," 13 (2.4%) "sometimes," 10 (1.8%) "frequently," and 10 (1.8%) "always."
Regarding whether they performed MRI-guided interventional breast procedures, 173 (30.6%) answered "no" out of 566 responding to this question. Of the 393 responding practices that performed breast MR interventional procedures, 292 (74.3%) did both MR-guided localization procedures and MR-guided core needle biopsy, 72 (18.3%) only did MR-guided core needle biopsy, and 29 (7.4%) only did MR-guided localization procedures.
Table 1 shows the responses to the question of whether a practice performed MR-guided core needle biopsy or localization procedures for suspicious findings on breast MRI performed elsewhere without first repeating the breast MRI. Table 2 shows the responses to the question of whether responding practices interpret breast MRI examinations performed at outside practices brought to them by surgeons or other health care providers. And if they did, Figure 4 shows the conditions that must be met before the practice would interpret outside breast MRI examinations.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There was wide variability in the number of breast MRI examinations done per week. Of the 553 respondents to this question, 37.4% did fewer than five per week, 44.5% did 5–15 cases per week, and 18.1% did more than 15 per week. Less than 1% of practices indicated that they performed more than 50 per week.
A recent multiinstitution study conducted by Lehman et al. [10] concluded that women at high risk for breast cancer would benefit from screening MRI. In that study, "high risk" included women 25 years or older who were genetically at high risk, defined as BRCA1 or BRCA2 carriers or with at least a 20% probability of carrying such a mutation. The study found that screening MRI had a higher biopsy rate, with the positive predictive value (PPV) of biopsies performed as a result of MRI being 43%, and helped detect more cancers than either mammography or sonography. The cancer yields for each test were 3.5% for MRI, 1.2% for mammography, and 0.6% for sonography. The use of MRI in addition to mammography for screening women at high risk for breast cancer is becoming more evident in practice and in the literature. The diagnostic and screening indications for breast MRI at practices in our survey are reported in Figures 1 and 2.
Recently, the American Cancer Society (ACS) recommended breast MRI screening for women at high risk for breast cancer (BRCA1 or BRCA2 mutation or first-degree relative with this mutation, 20–25% or greater lifetime risk for breast cancer, radiation to the chest between ages 10 and 30, history of Li-Fraumeni syndrome, Cowden disease, or Bannayan-Zonana syndrome, or first-degree relative with such syndromes) [11]. These recommendations will likely lead to greater use of breast MRI for screening, and an updated survey will be needed in the future.
At the time of our survey, which was conducted before the release of the ACS guidelines in March 2007, 90.4% of practices performing breast MRI were screening for patients positive for the BRCA gene and 74.7% were screening women with a history of premenopausal breast cancer in a mother or sister (Fig. 2). Although the ACS guidelines have not yet recommended screening women with a previous biopsy diagnosis of atypical ductal hyperplasia (ADH), lobular carcinoma in situ (LCIS), or atypical lobular hyperplasia, 61.1% of practices included these in their own indications for screening breast MRI.
The ACS guidelines for indications for screening MRI continue to be revised, and there still appears to be considerable uncertainty regarding whether screening MRI should be performed in women at a higher-than-average risk of breast cancer. This would include women with a 15–20% lifetime risk of breast cancer, history of LCIS, atypical lobular hyperplasia, ADH, very dense or unevenly dense breasts when viewed on a mammogram, or prior breast cancer including ductal carcinoma in situ [11].
As to referrals for breast MRI from clinicians, 83.2% of respondents to these questions considered requests to be frequently or always appropriate, and 86.3% thought that the majority of referring physicians were following the radiologists' recommendations for ordering breast MRI (Table 1). However, 56.5% of respondents thought that referring physicians frequently or sometimes gave in to patients who wanted breast MRI when it was not indicated.
Regarding the supervision and interpretation of breast MRI, an overwhelming majority of practices had MQSA-qualified radiologists supervising the performance of breast MRI examinations (Fig. 3). In addition, the majority of practices reported that interpretation of breast MRI was done by MQSA-qualified radiologists who interpreted mammograms more than 50% of their time.
The technical aspects of performing breast MRI are not currently
standardized. Our survey showed that 96.9% of practices used 1.5-T MR magnets,
9.5% used 3-T magnets, and 2.9% used magnets designed specifically for
breast MRI. The total percentage is > 100% because some practices used more
than one magnet for breast MRI. In terms of protocols, 90.3% always performed
bilateral breast MRI. In terms of which planes were used for contrast-enhanced
dynamic scans, 73.0% used the axial plane, 58.9% used the sagittal plane, and
14.5% used the coronal plane.
In terms of how they interpreted breast MRI, 98.0% of practices indicated that breast MRI was interpreted at soft-copy workstations, and only 2% used hard-copy images. Of those using soft copy, 50.7% (280) had a breast MRI CAD system available, and 43.3% of practices with CAD available used it for all of their breast MRI interpretations. The majority (84.1%) of practices indicated that they never or rarely interpreted breast MRI without comparison with mammography and sonography examinations.
Lack of standardization of breast MRI protocols often makes it difficult to interpret breast MRI performed at another facility. In our survey, 77.4% of respondents "never" or "rarely" interpreted outside breast MRI examinations for their referring physicians. Of those that did interpret outside breast MRI examinations, 67% required at least a review of the mammography and sonography, and 64.2% required a complete clinical history.
In addition, our survey disclosed that 31.1% of practices that performed contrast-enhanced breast MRI did not perform breast MRI interventional procedures. This leaves a question as to the treatment of the women with suspicious findings on breast MRI who cannot have the biopsy performed at the facility that performed and interpreted the examination. In addition, 87.5% of the practices that performed interventional breast MRI procedures reported that they never or rarely performed interventional procedures on the basis of an outside MRI examination without repeating the examination. That repeated examination is unlikely to be reimbursed, creating real problems for the facility that does the biopsy. Perhaps national guidelines can address this serious problem. One solution might be for facilities without biopsy capability to make agreements with facilities with biopsy capability that would include using the same imaging protocols so that studies would not have to be repeated before the biopsy.
Our study revealed some statistically significant differences between academic and nonacademic practices. For example, academic practices were more likely to offer breast MRI than nonacademic practices (88.9% vs 70.3%). Academic practices were more likely to perform screening breast MRI than nonacademic practices (83.2% vs 58.5%). A greater percentage of academic practices than nonacademic practices performed MRI-guided breast interventional procedures (83.3% vs 65.5%). Also, the academic practice radiologists less frequently responded that they "never" or "rarely" interpreted breast MRI performed at outside practices when compared with nonacademic radiologists (52.0% vs 84.6%). The supervising radiologist was more likely to be an MQSA-qualified radiologist who interprets mammograms more than 50% of the time in the academic practices than in the nonacademic practices (75.2% vs 59.3%). The breast MRI interpreting radiologist was more likely to be an MQSA-qualified radiologist who interprets mammograms more than 50% of the time in the academic than in the nonacademic practices (74.4% vs 64.5%).
Our results were subject to several biases. First, there is always the possibility of selection bias with voluntary respondents. In addition, because our participants were selected from the SBI, they may not reflect the average practice.
We anticipate that in the future, breast MRI, like other areas of breast imaging, will be subject to some sort of federal regulation, possibly under an expanded MQSA. We hope the information gained from this survey will provide reasonable approaches for the development of any such regulations. In addition, the findings may be useful in the development of professional practice guidelines. Furthermore, we hope that the results of this study will be helpful to radiologists establishing or updating their breast MRI programs.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  W. A. Berg Reply Am. J. Roentgenol., October 1, 2009; 193(4): 1189 - 1189. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nakano, M. Yoshida, K. Fujii, K. Yorozuya, Y. Mouri, J. Kousaka, T. Fukutomi, J. Kimura, T. Ishiguchi, K. Ohno, et al. Fusion of MRI and Sonography Image for Breast Cancer Evaluation Using Real-time Virtual Sonography with Magnetic Navigation: First Experience Jpn. J. Clin. Oncol., September 1, 2009; 39(9): 552 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. B. DeMartini, P. R. Eby, S. Peacock, and C. D. Lehman Utility of Targeted Sonography for Breast Lesions That Were Suspicious on MRI Am. J. Roentgenol., April 1, 2009; 192(4): 1128 - 1134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Berg Tailored Supplemental Screening for Breast Cancer: What Now and What Next? Am. J. Roentgenol., February 1, 2009; 192(2): 390 - 399. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
