Breast Magnetic Resonance Imaging:
This policy regarding BMRI as a screening and/or diagnostic (detection) tool was primarily based upon several Blue Cross Blue Shield Association (BCBSA) Technology Evaluation Center (TEC) Assessments.
For patients who are at high genetic risk for breast cancer -- based on the 2003 BCBSA TEC Assessment:
- When applied to high genetic-risk patients, the evidence appears to show at least equivalent performance for magnetic resonance imaging (MRI) in terms in sensitivity in detecting breast cancer compared to mammography. In the studies reviewed, MRI detected 100% of cancer cases, while mammography detected 33%.
- When applied to average-risk patients, the direct benefit of MRI screening among this population has not been proven. Because the prevalence of breast cancer is extremely low in average-risk patients, screening with a test such as MRI that has inferior specificity would result in lower positive-predictive values and many more false-positive results. Compared with mammography, there would be greater numbers of increased workups, biopsies, and increased morbidity if MRI screening were to be applied to average-risk women.
For patients who have breast characteristics limiting the sensitivity of mammography -- based on the 2004 BCBSA TEC Assessment:
- When applied to patients with or without a prior history of breast cancer, evidence is insufficient to draw conclusions on the effects of adjunctive BMRI on health outcomes.
- When applied to patients in the average-risk population, the incremental effects of adjunctive BMRI screening are uncertain.
- When the sensitivity of mammography is limited in patients after breast conservation therapy, there may be improvements in sensitivity with BMRI; however, more prospective studies are needed to confirm this.
For patients who have breast cancer and desire screening of the contralateral breast -- based on the 2004 BCBSA TEC Assessment:
- Several BMRI studies were conducted at various times before, during, and after treatment of breast cancer. Most studies reported that contralateral cancers detected on BMRI were not detected by conventional testing, whereas in some cases BMRI was done to evaluate suspicious findings in the contralateral breast. Four of five studies reported a 4% to 9% prevalence of cancer in the contralateral breast using MRI, with one study reporting a higher 24%. The positive predictive variable ranged from 20% to 80% and specificity ranged from 76% to 97%.
- A larger study of 212 patients with negative findings on mammograms on the asymptomatic contralateral breast found 12 cancers (5%) on BMRI. However the positive predictive variable was 20% and specificity of 76%.
- When the positive predictive variables and specificity for BMRI have not been well established, more confirmatory studies are necessary to support the use of BMRI in screening the contralateral breast.
Diagnostic or Detection Uses:
For patients who have a suspected breast implant rupture which had been originally placed for cosmetic reasons or augmentation (unrelated to breast cancer augmentation or reconstruction), whether clinical conditions appear or not, BMRI is not a covered benefit, as the initial breast implant placement was a cosmetic augmentation procedure. All services relating to the post-cosmetic augmentation procedure, including, but not limited to BMRI, are not a covered benefit.
For patients who have a suspected breast implant rupture which had been originally placed for breast reconstruction resulting from a mastectomy due to breast cancer, accidental injury, or trauma, BMRI would be considered a covered benefit as mandated by Federal Legislation.
For patients who have a suspected occult breast primary tumor with axillary nodal adenocarcinoma when there is a negative mammography and physical examination -- based on the 2004 BCBSA TEC Assessment:
- The adjunctive use of BMRI allows patients to avoid the morbidity of mastectomy in a substantial portion of patients (approximately 25% - 61%) compared to the risk of an unnecessary biopsy estimated at 8%.
- The use of positive BMRI findings to guide breast conserving therapy instead of presumptive mastectomy appears to offer substantial benefit of breast conservation in true-positive BMRI.
For patients who have posteriorly located tumors and determination of chest wall involvement are required -- based on the following clinical studies:
- The available prospective study of 19 patients with posteriorly located breast tumors suspected to involve the pectoralis major muscle based on either mammography or clinical examination, the presence of abnormal enhancement within the pectoralis major muscle on BMRI was 100% sensitive and specific for identifying five tumors that actually had invaded the chest wall. In the available studies, the results of the BMRI were compared with surgical and pathological findings. Prior to the BMRI and based on clinical examination, thirteen of the 19 were thought to be fixed on the chest wall. When based on mammography, twelve of the 19 were thought to have pectoral muscle involvement.
- The available evidence from two other studies reported four cases in which BMRI was able to determine involvement of the chest wall with 100% accuracy. Given the high level of diagnostic accuracy for BMRI, as compared with reference standard and conventional alternative techniques, the evidence is considered sufficient to permit conclusions that BMRI improves net health outcomes.
For patients who have suspicious breast lesions and further characterization is desired -- based on the 2000, 2001, and 2004 BCBSA TEC Assessments:
- The available studies addressed a group of patients who have breast lesions of sufficient suspicion to warrant recommendation to undergo biopsy for diagnosis. Therefore, BMRI results are assumed to have an impact on the decision whether or not to undergo definitive biopsy, considered the gold standard.
- The available evidence did not show that this use of BMRI would improve health outcomes. Considering the relative ease of breast biopsy, the sensitivity of BMRI would have to be virtually 100% to confidently avoid biopsy. While BMRI performs well, it is clear that the sensitivity is not 100%. False negative results tend to occur, particularly in certain subcategories, such as ductal carcinoma in situ, but invasive carcinomas may fail to enhance on MR, leading to false negative findings as well. The potential harm to health outcomes of failing to diagnose breast cancer or at least delaying the diagnosis of breast cancer is of significant concern.
For patients who have low-suspicion findings on conventional testing not indicated for immediate biopsy and referred for short-interval follow-up -- based on the following studies:
- The available evidence suggests that adjunctive BMRI may be very sensitive and specific in patients with low-suspicion findings on conventional testing and may provide a useful method to select patients for biopsy or to avoid prolonged short-interval follow-up. However, none of the available studies use prospective methods in the appropriate patient population to directly compare the sensitivity and specificity of short-interval mammographic follow-up with BMRI and to determine the effects of adjunctive BMRI on cancer detection rate and biopsy rate.
- Well-designed prospective confirmatory studies would be necessary to permit conclusions regarding the effect this adjunctive use of BMRI has on health outcomes.
For patients who have clinically localized breast cancer and preoperative mapping to identify multicentric disease is desired -- based on the 2000 and 2004 BCBSA TEC Assessments:
- The available multiple studies confirm BMRI has a better sensitivity for identifying multicentric breast tumors compared to mammography and/or ultrasound. Approximately 2% to 15% of patients otherwise eligible for breast conserving therapy may have multicentric disease seen on BMRI. BMRI is primarily used to identify multicentric breast tumors that have not been detected on conventional testing using mammography, clinical exam, or ultrasound.
- The effect on health outcomes of multicentric disease detected only on BMRI has not been established. If BMRI information is used to guide mastectomy, then the potential benefit of breast conservation will be decreased. The effects of multicentric disease on locoregional occurrence and survival have not been established after either breast-conserving therapy (BCT) with whole-breast radiation or modified radical mastectomy.
For patients who have locally advanced breast cancer and require preoperative tumor mapping before and after completion of neoadjuvant chemotherapy -- based on the 2004 BCBSA TEC Assessment:
- Compared with conventional methods of evaluating tumor size and extent (such as mammography, clinical exam, or ultrasound), BMRI provides an estimation of tumor size and extent that is at least as good as or better than that based on alternatives. One clinical trial found BMRI to be 100% sensitive and specific for defining residual tumor after chemotherapy. Conversely, mammography achieved 90% sensitivity and 57% specificity, and clinical examination was only 50% sensitive and 86% specific.
- Using BMRI instead of conventional methods to guide surgical decision-making regarding the use of breast conserving therapy versus mastectomy would be at least as beneficial and may more frequently lead to the appropriate surgical procedure.
For patients who have locally advanced breast cancer and require evaluation of response during neoadjuvant chemotherapy -- based on the 2004 BCBSA TEC Assessment:
- The available evidence is limited to a few small studies with inconsistencies in outcome measures, reporting, and use of statistical comparisons. A total of six studies, including a total of 206 patients, performed BMRI during the course of chemotherapy. BMRI outcomes for response to chemotherapy were based on either reduction in tumor size or reduction in contrast enhancement. Three studies reported negative-predictive value (ability to predict a nonresponsive tumor) results of 38%, 83%, and 100%. However, the two lower estimates were from prospective studies, while the highest estimate was from a retrospective review.
- Results are not consistent, and there is insufficient evidence to determine whether BMRI can reliably predict lack of response to neoadjuvant chemotherapy. Ongoing clinical trials are pending.
For patients who have positive surgical margins after lumpectomy and require evaluation of residual tumor -- based on the following clinical studies:
- Seven studies evaluated the diagnostic performance of BMRI to determine the presence of residual disease after prior biopsy or lumpectomy. Histopathology on re-excision was used as the reference standard. Most of these studies, including a single prospective study, reported poor sensitivity and specificity of BMRI for detection of residual disease. Two studies that reported more favorable results have methodological concerns that limit the influence of reported results. Three of the studies were conducted in the same institution and accrued patients during similar time periods, so overlap of reported patients may exist.
- The available evidence is not sufficient to permit conclusions whether BMRI improves net health outcomes when used to identify the presence and/or extent of residual disease after lumpectomy and prior to re-excision.
American College of Radiology Practice Guideline:
In 2004, the American College of Radiology (ACR) issued practice guidelines for the performance of BMRI. These guidelines have undergone “a thorough consensus process in which they have been subjected to extensive review, requiring the approval of the Commission on Quality and Safety as well as the ACR Board of Chancellors, the ACR Council Steering Committee, and the ACR Council. The practice guidelines list 12 different indications for BMRI. However, these indications are not referenced, and no discussion is provided regarding what evidence was used to develop the indications. Therefore, it is difficult to critically evaluate these guidelines and to determine to what extent they are evidence based versus consensus based.
The policy was updated based on review of the American Cancer Society (ACS) Guidelines for Breast Screening with MRI. Some of their recommendations were based on evidence and some were based on expert consensus opinion. Those based on evidence relate to having a BRCA mutation, a first-degree relative with a BRCA (breast cancer gene) mutation, or lifetime risk greater than about 20% to 25% based on models such as BRCAPRO™ that are largely dependent on family history. Except for more detail about the model and level of risk, these recommendations are consistent with the current policy and the policy language is modified to reflect use of these models in defining a high-risk patient. No other changes are being made to the policy with this update.
Lehman recently reported that 3% (30 of 969) of women with a recent diagnosis of unilateral breast cancer were found to have contralateral cancer at the time of initial diagnosis using MR imaging. These contralateral lesions were not detected by mammography or physical exam. Eighteen of the 30 were invasive cancer and 12 were ductal carcinoma in situ (DCIS). In this study, 12.5% of the patients (121) had biopsies with a rate of positive biopsy of 24.8%. With one-year follow-up, sensitivity of magnetic resonance (MR) was 91% and specificity was 88%. The results of this study in a diverse group of patients are similar to the findings of others. As noted above, Liberman found contralateral cancer in 5% (12 of 223) of patients who had negative mammograms of the asymptomatic contralateral breast. Six were invasive cancer and six were DCIS. Biopsies were recommended in 72 patients but completed in 61 patients. Lehman found four contralateral cancers in 103 patients; in this study, ten biopsies were done.
While the long-term outcome of these findings is not fully known; important changes in management will occur as a result of these findings, which should lead to improved outcomes. That is, in addition to the presumed benefits of early detection of these tumors, simultaneous treatment of synchronous cancers can occur rather than multiple treatments on separate occasions.
These data concur with the recommendation made by the ACR practice guideline and with the consensus statement from the American Society of Breast Surgeons.
Thus, based on prior studies and the additional data, imaging of the contralateral breast may be considered medically necessary in those with a new diagnosis of breast cancer when conventional imaging has not detected contralateral abnormalities suspicious of cancer.
A previous TEC Assessment concluded that ipsilateral MRI at the time of diagnosis did not meet TEC criteria because there was insufficient evidence to permit conclusions on the effect on health outcomes of adding MRI in the standard staging workup of early stage invasive breast cancer. However, as noted in the Assessment, long-term recurrence rates for the two approaches (modified radical mastectomy compared to breast-conserving surgery combined with whole-breast radiation) did differ, with lower long-term recurrence rates after mastectomy. For example, the National Cancer Institute (NCI) USA trial (n=247) reported 18-year locoregional recurrence rates of 25.6% for BCT versus 9.5% for modified radical mastectomy. The NCI Italy trial (n=701) reported local recurrence rates at 20 years for BCT at 8.8% and for radical mastectomy of 2.3%. These differences were both statistically significant with p-values less than 0.01. Studies have shown that from 2% to 15% of women with a new diagnosis of breast cancer would have multicentric disease detected on MRI. Thus, given the ability of MRI to detect multicentric disease at the time of diagnosis of breast cancer and thus impact treatment decisions that could reduce long-term recurrences, the policy statement is changed to consider ipsilateral imaging with MR also medically necessary at the time of diagnosis of breast cancer. BMRI does not have 100% specificity, it is important that biopsy of identified lesions be obtained before making a decision to perform a mastectomy. The final decision about the surgical approach using the MRI results should be made through informed consent.
National Comprehensive Cancer Network (NCCN) guidelines also indicate that “breast MRI may be considered for a patient with biopsy-proven breast cancer, when dense breast tissue precludes assessment for extent of disease.” These changes relate to those with a new diagnosis of breast cancer. As additional studies are reported for use of MR in screening for disease, it will be critical to evaluate intermediate and long-term outcomes for the additional findings noted only on MR imaging. Determining longer term outcomes is critical because “MR-detected breast cancer,” that is, lesions shown to be breast cancer seen only on MR imaging, may have different clinical characteristics than cancer identified by physical examination and/or other imaging techniques.
The policy was updated with a literature search using MedLine in June 2008. Kuhl reported results for the diagnosis of ductal carcinoma from a prospective series in a single, specialized referral center. Over a five-year period, 7,319 women who were referred to this center received MRI in addition to mammography for diagnostic assessment and screening. One hundred ninety-three (193) women received a final surgical pathology diagnosis of pure DCIS. Of those, 167 had undergone both imaging tests preoperatively; 93 (56%) of these cases were diagnosed by mammography and 153 (92%) by MRI (p<0.0001). Of the 89 high-grade DCIS, 43 (48%) were missed by mammography, and all 43 cases missed by mammography were detected by MRI. By contrast, MRI detected 87 (98%) of these lesions. The authors note that their results are not representative of the typical screening setting. They also indicate that a multi-institutional trial will be needed to further investigate the role of MR imaging for diagnosing DCIS in a screening population and determining its impact on outcomes such as recurrence rates and mortality.
This update also re-addresses several of the recommendations of the ACS guidelines for use of BMRI that were not evidence based. Two of the indications that were based on expert consensus opinion relate to use of MRI in rare genetic syndromes. In these uncommon conditions, the risk of breast cancer, which often occurs in premenopausal women, is as high as 50%. Thus, given the current policy statement regarding use of MRI in those whose lifetime risk is 20% to 25% or greater. Patients with suspected Li-Fraumeni syndrome (mutations of TP53 gene), Cowden syndrome, or Bannayan-Riley-Ruvalcaba syndrome (mutations of phosphatase and tensin-homlog [PTEN] gene) and their first-degree relatives are added indications. Use of MRI in these situations is also included in NCCN guidelines on genetic or familial high-risk assessment. The third recommendation based on expert consensus is for use of MR in those who received radiation (therapy) to the chest between the ages of 10- and 30-years. The risk of breast cancer in these patients can be quite high but depends on the age at treatment, radiation dose, and concomitant use of chemotherapy. Travis estimates that the cumulative absolute risks of breast cancer for a survivor of Hodgkin’s lymphoma who was treated at age 25 years with chest radiation dose of at least 40 Gy without alkylating agents are 1.4% at age 35, 11.1% at age 45, and 29.0% at age 55. The ACS guidelines also note that more recent treatment approaches using lower doses of radiation and limited fields are associated with lower risks. Given the lifetime risk of breast cancer, the indication relating to prior radiation therapy (RT) is added to the policy. Of note, the NCCN guidelines related to breast cancer screening in those with prior thoracic RT recommend clinical examination and then mammography beginning eight to ten years following radiation therapy, or starting at age 40, whichever comes first.
An additional updated search for pivotal publications on the use of BMRI did not find any studies that would provide strong evidence to alter the conclusions reached above. However, several recent studies, which will be described briefly, provide further evidence on some of the issues discussed above.
Regarding the use of BMRI for cancer patients, in a single institution, retrospective study of 756 women undergoing BCT between 1992 and 2001, outcomes were compared between women who had an MRI and those who did not. It should be noted that the use of MRI in these patients occurred at different points in the disease process; 50% occurred before initial surgery. With a mean follow-up of five years, there were no significant differences in overall survival, cause-specific survival, freedom from distant metastases, any local failure, local-only first failure, and contralateral breast cancer.
As for use of BMRI to gauge effective of neoadjuvant chemotherapy, a study of 51 patients compared MRI determination of tumor response following neoadjuvant therapy with pathological results from BCT or mastectomy. Interestingly, BMRI correctly diagnosed 18 of the 19 partial complete response cases among HER2 (human epidermal growth factor receptor 2)-positive patients versus eight of 16 partial complete response cases among HER2-negative patients. In other words, BMRI’s accuracy in determining complete response to neoadjuvant chemotherapy was higher among HER2-positive patients. The authors note that false negatives were more likely when the residual disease was in the form of scattered cells or small foci, which occurred more often in HER2-negative patients. In detecting complete response also, the accuracy of BMRI, varied by chemotherapeutic regimen used. These conclusions are based on small numbers of “suboptimal” spatial resolution and would need to be replicated in larger studies before being applied to clinical practice.
The American College of Radiology Imaging Network (ACRIN) Protocol A6657 on using BMRI to evaluate patients undergoing neoadjuvant therapy has been amended to increase to 377 and include a component that addresses the use of MR spectroscopy in these patients with an optional diffusion-weighted imaging. As of January 27, 2010, 356 patients have been enrolled in this study.
Additional information was published in 2007 on the elevated risk of cancer in women with dense breasts, where mammography is less sensitive. In three nested case-controlled studies
With 1,112 matched case-control pairs, the authors estimated that the adjusted odds ratio of detecting breast cancer among women with density in 75% or more of the mammogram versus those with density in less than 10% of the mammogram was 4.7 (95% CI [confidence interval]: 1.0-7.4). These cancers were detected through screening or during a period of less than 12 months after a negative screening examination. In younger women, 26% of all breast cancers were in patients with density evident in 50% or more of the mammogram.
Lastly, this update includes the utilization of BMRI in patients with breast implants. Kuhl reported in Radiology (September 2007) that BMRI is “performed to investigate the integrity of the implant shell and delineate breast cancer around or behind the implant.” Kuhl further reports that contrast agents are not needed, although agents “may be helpful to depict chronic inflammatory changes around the implant due to what clinically appears as ‘capsulitis’ or implant fibrosis”.
An updated search through April 2011 for pivotal publications on the use of BMRI did not find any studies that would provide strong evidence to alter the conclusions reached on previous updates. No studies were found supporting the use of BMRI to screen for breast cancer and for women at normal risk for breast cancer, other than those with dense breasts, discussed below.
In a retrospective study, the accuracy of BMRI was evaluated among patients with dense breast and suspected breast cancer or inconclusive evaluations who had a BMRI at a single institution in Italy. The reference standard was histology of six and/or 18 month follow-up. MRI was compared to mammography or ultrasound. About half of the women were found to have breast cancer. Of 238 patients, 97 had all three imaging tests. The sensitivity and specificity of BMRI was 98.2% and 95.2%, respectively for mammography, 72.7% and 45.2%; and, for ultrasound, 85.5% and 40.5%. In this study, MRI is used to evaluate patients of having breast cancer or with equivocal results from other modalities, including clinical examination. Although the specificity is relatively high and the negative predictive value in this selected population is 97.6%, this study does not provide sufficient evidence to use BMRI as a substitute for biopsy in these patients, as the authors themselves state. Joint recommendations from the Society of Breast Imaging and the ACR suggest that the addition of ultrasound to screening mammography “may be useful for incremental cancer detection” for women for whom dense breast is their only risk factor. MRI is not mentioned in this context; however, the authors do mention that the recommendations focus on the use of MRI on the highest risk women because of the considerable cost of MRI and the potential for more false-positives with a lower cancer yield as the risk of developing cancer declines.
A study of mastectomy rates at the Mayo Clinic declined from 1997 to 2003 (45% to 31%, p<0.0001), and then increased to 43% in 2006 among patients with TNM stage 0 to 2 breast cancer. (see “NOTE” following this paragraph. The use of preoperative BMRI also rose from 2003 to 2006 (from 10% to 23%, p<0.001). While mastectomy rates rose among all patients, those undergoing BMRI were more likely to have mastectomy than those with no MRI (54% versus 34%, p<0.001). The results of a multivariable model indicated that both BMRI and surgical year were independent predictors of mastectomy. A retrospective study at Cedar-Sinai Medical Center found, to the contrary, that between 2003 and 2007, the number of MRIs per 100 new patients increased from 1.6 to 2.9, while the mastectomy rate was stable.
NOTE: According the American Cancer Society, the TNM staging system classifies cancers based on their T, N, and M stages:
- The letter T followed by a number from 0 to 4 describes the tumor's size and spread to the skin or to the chest wall under the breast. Higher T numbers mean a larger tumor and/or wider spread to tissues near the breast.
- The letter N followed by a number from 0 to 3 indicates whether the cancer has spread to lymph nodes near the breast and, if so, how many lymph nodes are affected.
- The letter M followed by a 0 or 1 indicates whether the cancer has spread to distant organs -- for example, the lungs or bones.
Controversy continues regarding the use of BMRI preoperatively for patients diagnosed with breast cancer. Preoperative BMRI may be used to gauge the size of the tumor or identify additional lesions in the same quadrant, elsewhere in the ipsilateral breast, or in the contralateral breast. Some have suggested that BMRI can reduce re-excision rates by more accurately identifying the extent of the initial tumor. In a review of data on the use of BMRI, one randomized controlled trial and several observational studies were identified that addressed this issue. None reported a statistically significant difference in re-excision rates or proportion of surgeries with positive margins around the tumor (i.e., where cancer remained) between patients who have a preoperative BMRI and those who did not. The evidence from an additional single-arm study was not strong.
The median prevalence of additional ipsilateral cancer foci detected by preoperative cancer by preoperative BMRI is 16% according to one meta-analysis. According to one study of 119 patients in 2005 to 2006 in Germany, the use of BMRI pre-surgery changed clinical management in 40.3%; 17 patients had mastectomies instead of BCT, eight had an extended excision, 21 lesions were examined by using MRI-guided biopsy, and two ultrasound-detection lesions had negative MRI results and were not biopsied. However, it is not clear whether these changes improved patient outcomes. For example, the patient may have decided to have a mastectomy rather than BCT because of the detection of an additional lesion in the same breast on MRI that is proven to be malignant through biopsy. It is possible, however, that the additional cancer would have been eradicated by the radiation-therapy or chemotherapy following BCT, thus producing the same outcomes in terms of recurrence or survival as mastectomy.
The primary technology for conducting MR-guided biopsies appears to be vacuum-assisted biopsy. In a study by Perlet et al., 578 patients with 649 lesions visible on MRI alone, or that could be localized in three dimensions on MRI only, were considered for MRI-guided vacuum-assisted biopsy. A total of 538 lesions underwent biopsy; the rest were excluded because the lesions could not be reproduced during the pre-biopsy planning or because of technical difficulties with the procedure (e.g., breast too small, could not access lesion, motion). Of the remaining 538 lesions, 96% of the biopsies were successful (517 of the original 649 or about 80%). The yield of malignant lesions varied with the type of patient, from 20% among patients with lesions visible only on one mammographic view and not on ultrasound to 27% among women screened because of a family history suggesting a high genetic risk of breast cancer to 50% among 10 women with axillary metastases of suspected breast origin. The group of patients was diverse, and the method of selection was not described. Therefore, it is not possible to generalize these results to other populations. Some of the women were also evaluated for multifocal or multicentric disease, and the issues in this group are different because of the possibility that these lesions will be treated successfully by radiation or chemotherapy, as discussed here. Multifocal disease refers to having additional malignant lesions in the same breast quadrant as the primary tumor; multicentric disease refers to having additional malignant lesions outside of the breast quadrant where the primary tumor is located.
The policy regarding BMRI as a technique for detection of a suspected occult breast primary tumor with axillary adenocarcinoma was reviewed based on a meta-analysis of studies on the use of MRI. Two hundred twenty patients with mammographically occult breast cancer and an axillary metastasis were evaluated in eight retrospective studies. In seven studies, a potential primary lesion was detected in a mean of 72% of cases (range: 36-86%). Pooling individual patient data yielded a sensitivity of 90% (range: 85-100%) in detecting an actual malignant tumor. The specificity, however, was a pooled value of 31% (range: 22-50%).
A 2009 review cautions about the lack of evidence that additional lesions found by MRI only would lead to worse outcomes when they are not excised (i.e., via mastectomy). They also questioned the impact of patient outcomes of finding lesion in the contralateral breast on MRI only, given the positive predictive value of 47.9% (cited a meta-analysis by Brennan et al., 2009),
the predominance of early stage cancer, and other factors. They strongly recommend that randomized clinical trials be undertaken to gauge the impact of presurgical MRI scans of both breasts on patient outcomes, or that MRI not routinely used preoperatively among women with established, early stage breast cancer.
Others have suggested that BMRI might be useful before the use of accelerated partial breast irradiation (APBI), by identifying the patients with multicentric tumors that would not fall within the radiotherapy field. However, neither the equivalence of APBI to breast irradiation nor the utility of MRI in this context have been demonstrated. In a consensus statement on APBI, a Task Group from the American Society for Radiation Oncology, “agreed that there were insufficient data to justify recommendations of routine use of breast MRI for patients selected for APBI.” Diffusion-weighted imaging (DWI) is also being explored as a suggested use of BMRI. In a study from Partridge et al., DWI was used to determine whether lesion type (mass or non-mass-like enhancement) and size (greater or lesser than 1 cm) affect discrimination of benign and malignant breast lesions. Ninety one patients with 116 breast lesions were identified with dynamic contrast-enhanced MRI. Sixteen of 71 masses and 13 of 45 lesions with non-mass-like enhancement were identified as malignant. The authors report the DWI study shows promise.
In a retrospective study of 208 patients undergoing neoadjuvant therapy, 64 indicated complete response on BMRI scans, but 36 of them (56%) had residual disease on pathology. Conversely, 144 indicated residual disease on MRI, but no invasive cancer cells were found on pathology results in 14 of them, of whom five had DCIS. So the sensitivity of BMRI to detect residual invasive cancer was 78% (95% CI: 0.71–0.83), and the specificity was 67% (95% CI: 0.51–0.79). Furthermore, in 22% of all patients, the tumor size on MRI differed by more than 20 mm from the pathology results. This could alter the treatment choice from mastectomy to BCT or more rarely, from BCT to mastectomy. BMRI appeared to be most accurate in patients with triple-negative tumors, then HER-2 positive tumors, and least accurate in patients with ER (estrogen receptor)-positive tumors. The patients in this study may overlap with participants in the study, by Loo et al., of 188 women who underwent MRI scans before and during neoadjuvant chemotherapy compared the ability of MRI to detect response to treatment by breast cancer subtype. They concluded that the change in the largest diameter of enhancement on MRI was associated with tumor response among patients with so-called triple negative and HER2-positive tumors but not among patients with the more commonly found ER-positive/HER-2 negative tumors.
The medical policy review update was based in part upon a review of the published 2011 NCCN Guidelines criteria of breast cancer screening and diagnosis. The NCCN recommends annual BMRI screening for patients with a “BRCA mutation; first-degree relative of a BRCA carrier, but untested; lifetime risk [of breast cancer based on models, such as Breast Cancer Program] BRCAPRO,… that are largely dependent on family history; radiation to chest between the ages of 10 to 30 years; Li-Fraumeni syndrome and first-degree relatives; Cowden and Bannayan-Riley-Ruvalcaba syndromes and first-degree relatives.” It should be noted that the joint recommendations from the Society of Breast Imaging and ACR therefore recommend that high risk women be screened annually with both MRI and mammography.
Computer-Aided Evaluation for Interpretation of Breast Magnetic Resonance Imaging:
This policy regarding computer-aided evaluation (CAE) for magnetic resonance imaging (MRI) of the breast (BMRI) was based on the 2006 TEC Assessment, and updated with a literature search through June 2008. Key aspects of that assessment are reviewed and summarized in this section.
The TEC Assessment summarized four published articles and four published abstracts that met the search criteria. (To meet search criteria, articles had to compare the sensitivity and specificity of BMRI interpreted with and without the use of CAE systems. While the search focused on commercially available CAE systems, some articles on other systems were included. In addition, studies had to report on cancer detection based on histological results.) Three of the articles reported on development and validation of CAE systems aimed at distinguishing between malignant and benign lesions; and they used information on women with known lesions. The fourth article provided information on one of the non-commercial systems used to evaluate women with cancer who were eligible for BCT. Additional findings (other lesions or larger lesions) were found in 48 of the 116 (41%) women; about 80% of these women had further workup; and in 27 of these women the findings were malignant. The area under the receiver operating character curve was 0.91+0.04 for the radiologist reading and 0.98+0.04 for the combined radiologist and computerized reading (p=0.03). However, the ability to generalize these results and the clinical impact of the findings is uncertain.
Four abstracts of studies were included in the TEC Assessment because of the small number of studies identified. However, the need to exercise caution in using results from abstracts must be kept in mind as these results are reviewed. Of the four abstracts, two used CADstream, one did not report the system used, and one was an excerpt from an article that summarized the results of three earlier abstracts on the 3TP system. It is not clear whether the current 3TP system has been modified substantially from the version used in these studies. Once again, these abstracts report on the results of CAE with BMRI among women with known lesions.
Finally, DeMartini reported on the use of CAE with BMRI in 15 patients to assess the impact of chemotherapy. This small study found there were a substantial number of false-negative results for residual malignancy using CAE—a different type of problem than found with most other uses of BMRI, i.e., too many false-positive results.
Further literature research in 2008 revealed two articles were published, apparently based on the retrospective study presented in an abstract mentioned above. The first article, published in 2006, reported on 33 consecutive lesions biopsied under MRI guidance at a single institution. The second article, published in 2007, reported on 155 consecutive lesions that appeared to subsume the 33 lesions included in the 2006 study; the later article is therefore summarized here. The lesions were not palpable or visible on mammography or sonography and were assessed with and without CAE. All of these lesions were rated breast imaging radiology score 4 or 5, i.e., suspicious or highly suggestive of malignancy. Sixty-four percent of the lesions were in recently diagnosed breast cancer patients, 14% were in high-risk patients being screened, and 14% were for problem solving. Three different MRI imaging protocols were used. CADstream was then retrospectively applied for this study. Each pixel in the image was color coded based on, first, whether it reaches a threshold level of enhancement in the first post-contrast image and, second, whether the enhancement increases, plateaus, or decreases in subsequent post-contrast images. The threshold level of enhancement for the first post-contrast image was varied from 50% to 100%. One lesion was excluded for a technical issue with the initial BMRI. As expected, increasing the level of enhancement required (to 100%) lowered the number of false positive results. Thirty-eight of 41 (93%) malignant lesions enhanced at both thresholds; while enhancement was absent in 23% of benign lesions. Two of the three false negative lesions exhibited enhancement when the cursor was manually placed over the lesion. At the 50% enhancement level, there was no statistically significant difference in the positive predictive value between the initial reading and the subsequent application of CAE; at the 100% enhancement level; however, the positive predictive value was significantly higher with CAE than without (30.4% vs. 26.6%, p=0.02). Because the radiologists who read each set of images with and without CAE were not necessarily the same, it is possible that some of this difference might be due to a variation across readers rather than to the addition of CAE. There was no significant difference in subsequent enhancement patterns (i.e., washout, persistent, or plateau) between benign and malignant lesions; and many lesions included diverse enhancement patterns.
In the first report, the authors highlighted the possibility of using the CAE results to identify lesions that do not need to be biopsied, in other words, to identify a subset of the false positive findings that could safely avoid biopsy. However, the sample was highly selective and not representative of the full spectrum of findings likely to be encountered in practice. In the second report, not all of the lesions identified as benign with CAE were in fact benign, i.e., there were false negative findings. The risk of missing cancers and delaying treatment has to be weighed against the opportunity for reducing the number of unnecessary biopsies. The magnitude of this risk cannot be estimated reliably from a single study.
Unfortunately, the literature on the use of CAE with BMRI was sparse overall, and few studies addressed the specific situations in which CAE with BMRI is used in a clinical setting. Some of the limited number of the articles and abstracts calculated test characteristics on the basis of lesions and not the number of women or breasts. In a screening population, many women would not have any lesions; including these women might alter the results. Given MRI’s lower sensitivity in detecting ductal carcinoma in situ (DCIS), the mix of DCIS versus masses would affect the calculations of sensitivity and specificity and might affect the impact of the CAE system.
Prospective, well-designed and executed studies that look specifically at the addition of CAE with BMRI for the specific uses of regions of interest are needed to determine whether or not the use of CAE provides a positive clinical benefit to these patients.
In summary, there are no high-quality, published studies of the impact of commercially available CAE systems on the sensitivity and specificity of BMRI. The few studies and abstracts available focus primarily on the development of a CAE system or they include samples of women that are highly selective and — or -- usually have far more cases of cancer than would be encountered in a screening population.
A search of peer reviewed literature through April 2011 identified no new clinical trial publications or any additional information that would change the coverage position of this medical policy.
The following studies were assessed:
- One study evaluated the use of the CADstream system, using images from a 3.0 T MRI system. In this retrospective analysis of 426 women imaged consecutively, only 36 patients with 42 lesions were used in the final analysis (exclusions included women imaged for research purposes, women for whom no histology was obtained, technical failures of MRI, and patients with a breast imaging radiology score of six 6 for whom the readers were not blinded to patient history). The final sample included women with indeterminate mammographic and/or ultrasound results or high-risk screening. Images were interpreted by two experienced breast radiologists and two residents. Manual reading was followed 6 months later with reading of the CAE results. The sensitivity and specificity of the manual reading for both experienced radiologists combined was 84.6% and 68.8%, respectively (residents only read CAE results; breast imaging radiology rating of three considered negative). The sensitivity and specificity of the CAE reading for the same two experienced radiologists was 90.4% and 81.3%, respectively; the difference in specificity manually versus with CAE was statistically significant at p<0.05. There were no statistically significant differences among the CAE readers, including residents versus experienced breast radiologists. While the results are interesting, this study has several limitations, including its retrospective nature, highly selective sample with a large proportion of cancer cases, and small number of readers. It also is not clear whether the results apply only to a 3.0 T MRI system. Further research is needed with larger, more robust studies to assess these findings.
- A second retrospective study evaluated the sensitivity and specificity of a new MRI CAD software prototype called CAD-Gaea (Sentinelle Medical, Toronto, Canada; commercially released under the name Sentinelle Aegis software). The patients were women at high risk of breast cancer (BRCA1, BRCA2, or calculated lifetime risk of being a mutation carrier of 25% or greater). From an initial sample of 1,548 MRI studies on a 1.5 T system, the study sample consisted of 56 lesions in 53 women. Thirty-nine percent of the lesions were malignant, and of those, 59% were DCIS. The utility of this study for the present policy is limited because it only includes the breast imaging radiology score 3-5 cases that were biopsied (as well as a fat suppression protocol). Readings were performed by two experienced breast radiologists looking at different aspects of the CAE results, e.g., different thresholds for initial enhancement and initial enhanced versus delayed pattern or both. The primary finding was that high levels of sensitivity could be achieved using CAE (both initial and delayed enhancement patterns) for invasive cancers (100%) but when DCIS was included, the sensitivity dropped to 73%. The specificity when including all lesions was 56%. The prospective radiologist interpretation (apparently as part of clinical care) was more sensitive than the CAE-based interpretation (p=0.05), but less specific (p=0.01); specific values were not reported. The authors conclude that “The breast MRI CAD system used could not improve the radiologists’ accuracy for distinguishing all malignant from benign lesions, due to the poor sensitivity for DCIS detection.”
Research continues on efforts to improve the ability of CAE systems to provide information that increases diagnostic accuracy. These include articles that focus on the use of different metrics (e.g., selecting the “most suspect” enhancement pattern of a lesion versus the distribution of different enhancement patterns within a lesion; use of textural analysis of lesions; and expansion of the use of CAE to help distinguish the type of breast cancer and whether it has metastasized). These studies did not necessarily use commercially available CAE systems. Further research is needed to evaluate the utility of all of these approaches.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.