Use of (routine) preoperative MRI in breast cancer: current evidence Joint Hospital Surgical Grand Round 22 Oct 2011
Mar 30, 2015
Use of (routine) preoperative MRI in breast cancer:
current evidence
Joint Hospital Surgical Grand Round 22 Oct 2011
Presentation outline
Introduction Literature review Our own data
Introduction
Traditional triple assessment gives limited data on precise tumor size, location and margin
And whether there are multifocal (=several foci of tumors in the same quadrant) /multicentric (=foci of tumors in different quadrant) /contralateral disease
Breast magnetic resonance imaging (MRI) is emerging as a new clinical adjunct in this respect
Better surgical planning theoretically translates into less local recurrence and improved survival
BI-RADS (= Breast Imaging Reporting and Data System) 5
Radiology (2007) 244, 356-378
BI-RADS 2
Radiology (2007) 244, 356-378
MR Spectroscopy
Total choline (tCho) peak
Radiol Clin N Am (2010) 48, 1013-1042
Diffusion Weighted Imaging (DWI)
Radiol Clin N Am (2010) 48, 1013-1042
Clinical outcomes
Short term Sensitivity and specificity Alteration in management Re-excision rate
Long term Recurrence and survival
Three reviews conducted by Nehmat Houssami
Concluded that “ Evidence consistently shows that MRI changes surgical management, usually from breast conservation to more radical surgery; however there is no evidence it improves surgical care or prognosis”
J Clin Oncol (2008) 26, 3248-3258J Clin Oncol (2009) 27, 5640-5649CA Cancer J Clin (2009) 59, 290-302
(1) Accuracy and Surgical Impact of MRI in Breast Cancer Staging: Systemic Review and Meta-Analysis in Detection of Multifocal and Multicentric Cancer
19 studies with n=2610 MRI detected additional disease in 16%
(interquartile range 11-24%) of women with breast cancer
Summary PPV 66% (95%CI, 52-77%) TP: FP ratio 1.91 (95%CI, 1.09-3.34) Conversion due to MRI
Wide local excision (WLE) to mastectomy 8.1% WLE to more extensive surgery 11.3%
Unnecessary conversion due to MRI (histology negative) WLE to mastectomy 1.1% WLE to more extensive surgery 5.5% J Clin Oncol (2008)
26, 3248-3258
(2) MRI Screening of the Contralateral Breast in Women with Newly Diagnosed Breast Cancer: Systematic Review and Meta-Analysis of Incremental Cancer Detection and Impact on Surgical Management
22 studies with n=3253 Additional contralateral disease detected by
MRI 9.3% (interquartile range 3.8-13.9%) Summary PPV 47.9% (95%CI, 31.8-64.6%) TP:FP ratio 0.92 (95%CI, 0.47-1.82) No data on pooled management alteration
J Clin Oncol (2009) 27, 5640-5649
(3) Review of Preoperative MRI in Breast Cancer. Should MRI be Performed on All Women with Newly Diagnosed, Early Stage Breast Cancer? RCTs showed equivalent survival between breast
conservation therapy (WLE + radiotherapy) and mastectomy for early stage cancer
Vast majority of MRI detected additional disease are within same quadrant as the index tumor, which can be successful treated with post operative radiotherapy
COMICE trial and two additional observational studies did not show reduction in re-excision rate and on contrary higher mastectomy rate
Average of 22.4 days delay in workup
CA Cancer J Clin (2009) 59, 290-302
Only two observational studies provided data on long term outcome Fischer et al study limited by imbalance of
treatment between two groups Solin et al study
Local recurrence in 8 years (MRI+ vs. MRI-ve, 3% vs. 4%, p=0.51)
Overall survival in 8 years (86% vs. 87%, p=0.51)
Significant false positive rate caused additional cost and procedure; potential impact on cosmetic outcome
CA Cancer J Clin (2009) 59, 290-302
Comparative effectiveness of MRI in breast cancer (COMICE) trial: a randomised controlled trial
Multi-center, randomised 1623 women with biopsy proven breast
cancer scheduled for WLE after triple therapy MRI (n=816) vs. no further imaging (n=807)
Lancet (2010) 375, 563-571
19% 19%
Reoperation rate within 6 months 19% MRI group vs. 19% in no MRI group(odds ratio 0.96, 95%CI 0.75-1.24, p=0.77)
Cost: MRI group £5508.4 vs. No MRI group £5213.5 (p=0.075)
Our own data No previous study conducted in Asian population
whom breast density was considered higher Retrospective review Consecutive 712 biopsy proven breast cancer
patients underwent operation by a single surgeon in Hong Kong Sanatorium and Hospital during the period 1 January 2006 till 31 December 2009
Exclusion criteria (1) prior surgery to ipsilateral breast except excisional
biopsy for diagnosis (n=14) (2) neoadjuvant chemo/hormonal therapy (n=37) (3) missing data (n=2)
Total 659 cases for analysis MRI+ 147 vs. MRI- 512
Table 1 Indications for MRI
Percentage (n=147)
Nodular breast on clinical examination 7.5
MMG showed multiple pleomorphic microcalcification 4.1
MMG showed dense tissue 2.0
USG showed ill-defined border 19.0
USG showed multiple indeterminate shadows 53.1
Suspicion for multi-tumor on CNB 2.0
Discordance between clinical, imaging and histological finding 6.1
To locate occult primary focus with positive axillary LN 0.7
To search for residual tumor after excisional biopsy 4.8
Previous injection mammoplasty 0.7
Total 100.0
MMG=mammogram, USG=Ultrasound
Table 2 Characteristics of the patients included in the study
MRI-(n= 512)
MRI+(n = 147)
p value
Age (years)
Mean ± SD 52 ± 12 48 ± 7 <0.0011
Median 50 47
Range 25 - 91 30 - 70
Menopausal state <0.0012
Premenopausal 290 (56.6%) 113 (76.9%)
Postmenopausal 222 (43.4%) 34 (23.1%)
Family history 0.7922
No 413 (80.7%) 120 (81.6%)
Yes 99 (19.3%) 27 (18.4%)
Breast density on MMG <0.0012
<=50% 211 (41.2%) 45 (30.6%)
>50% 234 (45.7%) 93 (63.3%)
Missing data 67 (13.1%) 9 (6.1%) Student's t test for continuous variables1
Chi Squared test for categorical variables2
Table 3 Pathological characteristics of the included malignancies
MRI- (n = 512 )
MRI+ (n = 147)
p value
Size (cm) 0.2131
Mean ± SD 2.3 ± 1.4 2.1 ± 1.7
Median 2.2 1.8
Range 0.01 - 13.00 0.1 - 10.00
Grade 0.4022
I 104 (20.3%) 30 (20.4%)
II 136 (26.6%) 41(27.9%)
III 171 (33.4%) 35 (23.8%)
Missing 101 (19.7%) 41(27.9%)
Invasive/In situ 0.0222
Invasive present 423 (82.6%) 109 (74.1%)
Only in situ tumor 89 (17.4%) 38 (25.9%)
Focality <0.0012
Unifocal 456 (89.1%) 101 (68.7%)
Multifocal 56 (10.9%) 46 (31.3%)
Estrogen receptor score (H score) 0.9411
Mean ± SD 165 ± 104 165 ± 149
Median 190 180
Range 0-300 0-300
Progesterone receptor score (H score) 0.1541
Mean ± SD 111 ± 103 125 ± 107
Median 100 115
Range 0-300 0-300
Ki67 index (%) 0.3061
Mean ± SD 24 ± 51 20 ± 22
Median 11 9
Range 0-9 0-90
CerbB2 Score 0.5142
Negative 247 (48.2%) 80 (54.4%)
Indeterminate 149 (29.1%) 38 (25.9%)
Positive 107 (20.9%) 29 (19.7%)
Missing 9 (1.8%) 0 (0%) Student's t test for continuous variables1
Chi Squared test for categorical variables2
Table 4 Rate of re-excisions and completion mastectomies in patient undergoing BCT
MRI- (n = 349) MRI+ (n = 89) p value
Re-excision of tumor bed 28 (8.0%) 5 (5.6%)
Completion mastectomy 29 (8.3%) 11 (12.4%)
Total 57 (16.3%) 16 (18.0%) 0.71*
Chi Squared test*
Table 5 Rate of final mastectomies
MRI- (n = 512) MRI+ (n = 147) p value
163 (31.8%) 58 (39.5%) 0.085*
Chi Squared test*
Management alteration with MRI
66.0% (97 out of 147) had change in extent of operation From lumpectomy to wider lumpectomy (23 out of
97) to mastectomy (47 out of 97) to bilateral lumpectomy (15 out of 97) to others (12 out of 97)
Within 97 alterations in management, 12 were considered inappropriately extensive due to false positive finding on MRI
MRI detection of multifocal/ multicentric/ contralateral disease
False positive rate = 12.8% False negative rate = 7.5% Sensitivity = 95.3% Specificity = 80.3%
Conclusion
High sensitivity and moderate specificity Neither alter short term outcome e.g. re-
excision rate Nor sufficient evidence to alter long term
recurrence or survival No concrete evidence to support its routine
use