NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Primary Care (UK). Familial Breast Cancer: The Classification and Care of Women at Risk of Familial Breast Cancer in Primary, Secondary and Tertiary Care: Update [Internet]. London: Royal College of General Practitioners (UK); 2006 Jul. (NICE Clinical Guidelines, No. 41.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Familial Breast Cancer

Familial Breast Cancer: The Classification and Care of Women at Risk of Familial Breast Cancer in Primary, Secondary and Tertiary Care: Update [Internet].

Show details

4Clinical Effectiveness of MRI

4.1. Methods

The question being addressed - is as follows:

What is the effectiveness and cost effectiveness of MRI versus mammography versus MRI and mammography in surveillance for breast cancer in women at increased risk compared to the general population?”

Population: Asymptomatic women at an increased risk of breast cancer

Intervention: MRI or MRI & mammography (digital or film)

Comparator: Mammography (digital or film)

Outcomes: sensitivity, specificity, cases identified, positive predictive values, negative predictive values, mortality, cost effectiveness

Inclusion criteria: RCT’s, cohort or case-control studies evaluating MRI Vs mammography (X-ray mammography is the gold standard) or MRI and mammography Vs mammography in the detection of breast cancer in asymptomatic women with an increased risk of breast cancer.

Exclusion criteria: Insufficient information to allow construction of 2×2 table,. women without an increased risk of breast cancer (unless reported by sub group), computed radiography.

It was not within our remit to consider quality of life issues surrounding the use of MRI. However, this may be a question to be addressed when the full guideline is updated.

The search strategy used in the original guideline was repeated and updated from December 2002 when the last searches were conducted. Foreign language papers were excluded.

The Cochrane Database of Systematic Reviews (CDSR) Database of Abstracts of Reviews of Effects (DARE), Health Technology Assessment (HTA), CENTRAL, Medline, Embase, Cinahl, & PsycInfo databases were searched from 2003 until 30th November 2005. The abstracts were read and 17 papers obtained. 12 papers were rejected because they were not relevant or did not meet the inclusion criteria. 5 papers were included for review that assessed the diagnostic accuracy of MRI vs mammography in women at increased risk of developing breast cancer.

Areas without evidence:

It was anticipated that we would not find studies on the accuracy of digital imaging in an increased risk population group. Therefore we included any studies conducted in a normal population where subgroup analysis had been undertaken in the under 50 age group

An additional search was carried out for diagnostic studies of digital mammography in an average risk breast cancer population. The databases above were searched between 2003 and 22nd December 2005. Foreign language studies were excluded. Seven studies were obtained. Five studies were excluded because they were not relevant or did not meet the inclusion criteria. Two papers were included for review because they had carried out subgroup analysis in the population of interest to this guideline.

Where possible the data from each of the included studies have been reproduced in a 2×2 table to give figures of the true positive, false positive, false negative and true negative results of the diagnostic tests undertaken.

4.2. MRI Evidence

MARIBS study (Leach et al 2005)

Table 1Sensitivity and Specificity Table adapted from MARIBS study [Leach et al., 2005]

MammographyMRIMRI & mammography
All womenBRCA1BRCA2All womenBRCA1BRCA2All womenBRCA1BRCA2
No. women649139866491398664913986
No positive screens
True positive
143627127331211
False positive121301334476414289551
No negative screens
True negative
172533521915022891911418270181
False negative21106815211
Sensitivity %40%23%50%77%92%58%94%92%92%
Specificity %93%92%94%81%79%82%77%74%78%
predictive value %
positive
10%9%32%7%14%15%7%11%18%
negative99%97%97%99%100%97%100%100%99%

The MARIBS study was a multicentre prospective cohort study of 649 women aged 35–49 years with a BRCA1, BRCA2, TP53 mutation or strong family history of breast cancer. Annual screening of MRI and mammography was carried out for 2–7 years (1881 screens). A total of 35 cancers were diagnosed. 19 by MRI only, six by mammography only and eight by both, with two interval cases. 11 invasive cancers were less than 10 mm in greatest dimension. Of these, six were detected by MRI, three by mammography, one by both modalities, and one interval case. Four invasive cancers were between 10–14 mm. Three were detected by MRI and one by both modalities. Five invasive cancers were between 15–19 mm. Four were detected by MRI and one by both modalities. Nine cancers were 20 mm or larger in dimension, six were detected by MRI and three by both modalities. There were six cases of ductal carcinomas in situ (DCIS), of which four were less than 10 mm in diameter. Three were detected by mammography, two were detected by both modalities and one was an interval cancer. Of the 29 invasive cancers, three were grade 1, seven were grade 2, 19 were grade 3. Of cancers detected by screening or in a screening interval, 21 of 26 were node negative.

Sensitivity was significantly higher for MRI than for mammography and was particularly pronounced in BRCA1 carriers (13 cancers). The authors note that MRI is able to detect tumours earlier in their development compared with mammography, although mammography is relatively good at detecting DCIS compared with MRI. In spite of annual screening with two modalities, some large, node positive tumours were identified. This reflects the rapid growth characteristics of cancers in women with germline mutations.

Overall the study shows that the combination of MRI with mammography is the most effective screening examination for BRCA1 and BRCA2 carriers and the full high-risk cohort studied here. The authors also conclude that their results suggest that MRI screening would be of most benefit to BRCA1 carriers.

MRISC study (Kriege et al 2004)

Table 2Sensitivity and Specificity Table adapted from [Kriege et al., 2004]

MammographyMRI
Total screens41694169
No positive screens
True positive
1832
False positive207420
No negative screens
True negative
39173704
False negative2713
Sensitivity %4071
Specificity%9590
predictive value %
True87
Negative99100

Women who had a cumulative lifetime risk of breast cancer of 15 percent or more were included in this cohort study [Kriege et al., 2004]. 1909 asymptomatic women including: 358 carriers of germline mutations, 1052 high risk (30–50% cumulative lifetime risk), 499 moderate risk (15–30% cumulative lifetime risk) were screened, with a mean follow-up of 2.9 years. Among the women examined by both methods at the same screening visit, 45 tumours were detected including 6 ductal carcinomas in situ (DCIS). Five patients were excluded from analysis.

Of the invasive cancers 19 were 1cm or less in diameter, 14 were between 1–2 cm, 11 were more than 2 cm in diameter. Six of the 42 invasive tumours with known axillary status were node positive.

11 grade 1 cancers were found in women at high risk (68.8%), 6 in moderate risk women (75.0%), 2 in mutation carriers (10.5%).

One Grade 2 cancer was found in women at high risk (6.2%), 2 in moderate risk (25.0%), 5 in mutation carriers (26.3%).

Four Grade 3 tumours were found in women at high risk (25.0), 1 in moderate risk (12.5%), 12 in mutation carriers (63.2%).

The authors comment that larger tumours were found in women with BRCA1, BRCA2 and TP53 mutations than the other two risk groups in the study, suggesting that more frequent screening is necessary for this group of women.

The study found that the sensitivity of MRI was higher than mammography, but that the specificity and positive predictive value were lower. MRI detected 20 cancers (including 1 DCIS) that were not found by mammography, and the stage of these cancers was favourable, 11 of the 19 invasive tumours being less than 10 mm.

The study also showed that mammography had a higher sensitivity than MRI for detecting ductal carcinoma in situ (DCIS) 83% (five out of six cancers detected), compared with 17% (one out of six) for MRI. In this study, screening with MRI led to twice as many unneeded additional examinations compared with mammography (420 versus 207) and three times as many unneeded biopsies (24 versus 7). The authors conclude that the MRI can detect breast cancer at an earlier stage in women at risk for breast cancer.

Warner et al study 2004

Table 3Sensitivity and Specificity Table adapted from [Warner et al., 2004]

MammographyMRI
Year 1Year 2Year 3TotalYear 1Year 2Year 3Total
Total screens2361368545723613685457
No positive screens
True Positive
5308115117
False positive1001154120
No negative screens
True negative
2221298343420812582415
False negative842142215
Sensitivity %384303685715077
Specificity %*99.61001009993979995.4
Predictive value %
True
83100N/A8842565046
Negative9797989799989999

Abbreviations: MRI magnetic resonance imaging; N/A not applicable

*

Definition of specificity is based on biopsy rates

A cohort study of 236 asymptomatic women with BRCA1 or BRCA2 mutations underwent 1–3 annual screenings [Warner et al., 2001]. The study found that MRI was more sensitive for detecting breast cancer than mammography or CBE alone. 22 cancers were detected in total; 16 invasive, 6 ductal carcinoma in situ (DCIS). There was one interval cancer.

All six of the DCIS cancers were found in the BRCA2 group. Four in year 1, (two detected by MRI, tumour size between 3.0–4.0 cm, one detected by mammography, tumour size not given because specimen consisted of few small scattered foci, one detected by both modalities, tumour size 1.5 cm). One in year 2, (detected by mammography, no remaining cancer was observed at time of excisional biopsy). One in year 3, (detected by MRI, tumour size 6.0 cm).

Nine invasive cancers in year 1 were detected. Six had tumour sizes between 0.5–1.0 cm (3 detected by MRI, 1 detected by mammography, 2 detected by both modalities), three had tumour sizes of 1.5–2.0 cm (2 detected by MRI, 1 by neither (ultrasound.)) Six invasive cancers were detected in year two. Three had tumour sizes of 0.6–1.0 cm (3 detected by MRI), three had tumour sizes of 1.5–2 cm (2 detected by both modalities, 1 by neither (ultrasound). One cancer was node positive.

The study shows that the addition of annual MRI to mammography improves the sensitivity of surveillance for detecting early breast cancers. The authors suggest that mammography appears to be a useful adjunct to MRI for BRCA2 carriers because of the high incidence of ductal carcinoma insitu (DCIS) in this group.

The authors note that MRI recall rates decreased from 26% on the first round of screening to 13% on the second round and 10% on the third. The authors conclude that the study supports the position that MRI-based screening should be used for breast cancer surveillance for BRCA1 and BRCA2 mutation carriers. Further research is required to demonstrate whether this modality lowers breast cancer mortality before it can be recommended for general use.

Kuhl et al study 2005

Table 4Sensitivity and Specificity Table adapted from [Kuhl et al., 2005]

MammographyMRIMRI & Mammography
AllRisk 20%Risk 20–40%Mutation CarriersAllRisk 20%Risk 20–40%Mutation carrierAllRisk 20%Risk 20–40%Mutation carrier
Total screens170135275116717013527511671701352751167
No positive screens
True positive
14352396208406208
False positive451118539101645514149
No negative screens
True negative
136430267615413702546781551354299673150
False negative29315640003000
Sensitivity %32.650252590.7100.0100.0100.093.0100.0100.0100.0
Specificity %96.896.597.496.997.297.497.797.596.195.597.094.4
Predictive rates %
Positive PV
23.721.421.728.650.042.955.666.742.130.051.247.1
Negative PV97.9199.097.496.899.710010010099.7100100100

This cohort study comprised of 529 asymptomatic women who were suspected or proven to carry a breast cancer susceptibility gene [Kuhl et al., 2005]. A total of 1542 surveillance rounds were completed with a mean follow-up of 5.3 years. This study found that in patients at high familial risk for breast cancer MRI had the highest sensitivity, specificity and positive predictive rates for the detection of cancer. Forty three cancers were identified in the total cohort (34 invasive, 9 ductal carcinoma in situ (DCIS)). Forty of the forty three were diagnosed by imaging studies. Two cancers were palpable at the time of diagnosis (one at the regular screening interval, one was an interval cancer diagnosed in between screening rounds). These two clinically palpable cancers were also visualised by MRI but not mammography.

Nineteen cancers were diagnosed by means of MRI only. These included five intraductal (all high grade) and 14 invasive cancers with a median size of 7.5 mm. All fourteen invasive cancers were staged pT1 and all had negative axillary lymph nodes.

14 cancers diagnosed by mammography. These included three intraductal and ten invasive cancers with a median size of 12.0 mm, four were node positive.

39 cancers were detected by MRI. These included eight intraductal and 31 invasive cancers with a median size of 11.0 mm, five were node positive.

This study found that MRI had the highest sensitivity, specificity and positive predictive value for the detection of invasive as well as intraductal cancer. The addition of mammography to MRI did not improve sensitivity to a statistically significant degree. The authors conclude that compared with mammography, surveillance with MRI allows an earlier diagnosis of familial breast cancer and at an earlier stage. The specificity of MRI was equivalent to that achieved with mammography, which the authors suggest is due to the highly experienced readers used in the study. The number of cancers in the subgroup at moderate risk was too low to make valid recommendations regarding which surveillance modalities to recommend. The findings of this study lead the authors to recommend that MRI should be an integral part of surveillance for women at high familial risk, particularly in documented mutation carriers, but also for women without documented mutation The authors also note that further work is required to assess the risk/benefit ratio of mammography and MRI in young BRCA1 mutation carriers who may exhibit an increased radiosensitivity.

International Breast MRI Consortium Working Group study (Lehman et al 2005)

Table 5Sensitivity and Specificity Table adapted from [Lehman et al., 2005]

MammographyMRI
Total screens367367
No of positive screens
True Positive
14
False positive320
No of negative screens
True negative
360343
False negative30
Sensitivity %25.0100.0
Specificity %99.095.0
Predictive rates %
Positive Predictive value
25.017.0
Negative predictive value99.0100.0

This prospective study compared the performance of screening mammography with MRI on 367 asymptomatic high risk women age 25 or above [Lehman et al., 2005]. The objective of this study was also to ascertain if imaging and biopsy procedures are reliable and do not result in excessive false positive examinations. Imaging results recommended 38 biopsies of which 27 were performed. 4 cancers were detected overall, all were detected by MRI, and mammography detected 1 of these. The biopsy recommendation rates for MRI and mammography were 8.5% (95% CI 5.8–11.8) and 2.2% (95% CI 0.1–4.4), respectively. Twenty four women underwent biopsy based on a positive MRI and four based on a positive mammogram. Of the lesions that were identified as malignant, two were identified in women with scattered fibroglandular density, and two were identified in women with heterogeneously dense breast tissue. Three of the four lesions were identified as infiltrating ductal carcinomas ranging in size from 5 mm to 13 mm, and one lesion was DCIS. All were lymph node negative.

The limitations of this study are that only one screening round was undertaken and no follow-up was carried out to identify potential false negative MRI results or delayed diagnoses of those who declined biopsies. The authors conclude that although the specificity of MRI has been challenged they found only 5% of women underwent benign biopsy and the PPV of biopsies performed was 17%. They recommend that MRI should be considered as a complement to mammography.

4.3. Digital mammography Vs film mammography Evidence

DMIST study (Pisano et al 2005)

There was insufficient data provided to construct a 2 × 2 table.

This prospective study was conducted to assess whether the use of digital mammography had a higher sensitivity than film mammography [Pisano et al., 2005].

The results from 42,760 asymptomatic women entered into the trial were reported. Sub group analysis was undertaken in the following: under 50 age group, pre-menopausal and peri-menopausal n=15803, and those with heterogeneously or extremely dense breasts n=19897. In the entire population the diagnostic accuracy of digital and film mammography was similar. The accuracy was significantly higher for digital mammography in the under 50 age group, women with dense breasts and pre-menopausal & peri-menopausal women.

A total of 335 cancers were diagnosed. Of these 254 (75.8%) were diagnosed within 365 days after study mammography and 81 (24.2%) were diagnosed between 366–455 days after study mammography.

In the pre-menopausal & peri-menopausal subgroup film mammography identified seven (2.1%) invasive cancers, four (1.2%) ductal carcinoma in situ (DCIS), three were node positive. Digital mammography identified 19 (5.7) invasive cancers, 14 (4.2%) DCIS, five were node positive.

In the heterogeneously dense or extremely dense breast subgroup, film mammography identified 12 (3.6%) invasive cancers, seven (2.1%) DCIS, five were node positive. Digital mammography identified 26 (7.8%) invasive cancers, 14 (4.2%) DCIS, five were node positive.

The authors conclude that digital mammography was significantly better than conventional film mammography at detecting breast cancer in these groups. The cancers detected by digital mammography and missed by conventional mammography included many invasive and high-grade in situ cases. The authors conclude that this justifies the use of digital mammography in these groups.

Oslo II study (Skaane et al 2004)

This randomised controlled trial [Skaane and Skjennald, 2004] was conducted to compare cancer detection rates, recall rates and positive predictive value of film mammography (FM) with digital mammography (DM). 25,263 women aged 45–69 years were randomised to either film or digital mammography. Sub group analysis was carried out on the 45–49 age group (n=7607). 17 cancers were detected with FM (10 invasive cancers, seven DCIS). The median size of invasive cancers detected by FM was 11mm. 8 cancers were detected with DM (six invasive cancers, two DCIS). The median size of invasive cancers detected by DM was 10 mm.

Recall rates in both groups were significantly higher with DM than FM (P<0.05), but positive predictive value was not significantly different. In the 45–49 age group the cancer detection rate was nearly equal for the two modalities (P=0.686.) The authors state that a limitation of the study was that comparisons between FM and DM were available only during review of positive mammograms. Low recall rates and no follow-up for probably benign lesions might have caused cancers represented by a positive score on images in either modality to be dismissed at consensus meetings, where decisions about which women should continue in the screening programme and which be recalled for diagnostic workup were taken. Follow-up for two years would be necessary to detect incorrectly dismissed cancers and to evaluate interval cancers in a subsequent screening round.

The number of breast cancers in the group was small and the authors state that the results do not permit any final conclusions regarding the comparison of FM and DM in women younger than 50 years.

Copyright © 2006, National Collaborating Centre for Primary Care.
Bookshelf ID: NBK62046

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (580K)

Original Guideline

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...