Introduction
Breast cancer is one of the most common causes of mortality and morbidity in women.
The mortality rate has been reduced more so in the developed world, which is largely
due to new therapies and the spread of screening mammography.[1 ] Breast microcalcifications can be the early and only presenting sign of breast cancer.
Mammography is used worldwide to detect microcalcifications.[2 ]
Calcifications are described by their morphology and distribution according to the
BI-RADS lexicon. In 2013, the fifth edition of BI-RADS was released.[3 ] Stereotactic vacuum-assisted breast biopsy (SVAB) is a minimally invasive technique
and involves high-quality specimens, a high calcification retrieval rate, and a low
false-negative rate.[4 ] When SVAB is difficult to perform or a mammography-guided localization puncture
device is unavailable, ultrasound (US)-guided breast biopsy is an alternative option
if the calcifications are detectable by US.[4 ]
In our study, we have assessed the morphology and distribution of suspicious breast
microcalcifications detected on mammogram based on the fifth edition of BI-RADS lexicon,
evaluated the results of image-guided biopsies, and assessed the malignancy risk in
terms of positive predictive value (PPV) and compared with BI-RADS Atlas, fifth edition
and world literature.
Materials and Methods
This is a hospital-based study conducted at our institute over a period of 15 months.
The study population included all symptomatic and asymptomatic females who underwent
mammograms and were detected with suspicious breast microcalcifications, followed
by stereotactic-guided or US-guided breast biopsy and histopathological examination
(HPE) with immunohistochemistry.
A total of 77 patients with suspicious breast microcalcifications detected on mammography
were included. Patients who are younger than 30 years, patients with benign breast
calcifications, benign breast masses such as fibroadenoma, cyst, etc., and patients
who did not undergo biopsy or with unsatisfactory images were excluded from the study.
Procedure
A detailed clinical history was documented in a preset proforma for each patient.
Digital mammography was performed with a Hologic Selenia 6000 digital mammography
unit. Standard mediolateral oblique and craniocaudal views, three-dimensional (3D)
tomosynthesis views, and synthetic two-dimensional mammogram view (C view—better for
assessing calcifications) from 3D views were obtained for all patients.
Two radiologists specialized in breast imaging reviewed the mammograms and categorized
calcifications as typically benign/suspicious based on the BI-RADS Atlas, fifth edition.
Those with suspicious microcalcifications were offered stereotactic-guided vacuum-assisted
biopsy using Hologic Selenia 6000 digital mammography unit with 9G ATEC biopsy needle
or US-guided core needle biopsy with GE LOGIQ P6 and GE LOGIQ F8 ultrasound machines
using Max-Core/MAGNUM disposable core biopsy instruments MC1410 and MC1610 having
14G and 16G needles, followed by HPE. The findings were evaluated according to the
fifth edition of BI-RADS lexicon. Ethics committee approval was granted for the study.
Results
A total of 77 women with suspicious breast microcalcifications detected on mammograms
were included in the study. Age of the population ranged from 34 to 83 years, with
a mean age of 55 years; 35 (45.4%) were identified in the right breast, and 42 (54.5%)
were identified in the left breast.
Among 77 lesions, 56 (72.7%) were not associated with mass, and 21 (27.3%) were associated
with mass. Out of 56 (72.7%) lesions, without any mass on mammography, 47 (83.9%)
underwent stereotactic-guided biopsy and 9 (16%) underwent US-guided biopsy as the
microcalcifications were seen on US. Overall PPV for malignancy of suspicious microcalcifications
without associated mass was 37.5%.
Twenty-one (27.3%) out of 77 lesions were associated with mass and underwent US-guided
biopsy. Overall PPV of these suspicious microcalcifications when associated with mass
was 71.4% and more likely to be invasive ([Table 1 ]).
Table 1
Overall PPV of suspicious microcalcifications without/with mass on mammogram
Total no. of suspicious microcalcifications
Benign
Malignant
PPV
N
Percentage
N
Percentage
Without mass
56
35
62.5
21
37.5
37.5%
With mass
21
6
29
15
71.4
71.4%
Abbreviation: PPV, positive predictive value.
The PPV of suspicious microcalcifications without any associated mass on mammogram
according to the morphology descriptors was as follows: amorphous, 9.5%; coarse heterogeneous,
45.4%; fine pleomorphic, 50%, and fine linear/fine linear branching, 100% (p -value < 0.001). Fine linear/linear branching had the highest PPV of 100% ([Table 2 ]).
Table 2
PPV of suspicious microcalcifications without mass on mammogram according to the morphology
descriptors
Morphology
Benign
Malignant
Total (n = 56)
PPV
p -Value[a ]
N
Percentage
N
Percentage
Amorphous
19
90.5
2
9.5
21
9.5%
< 0.001 (S)
Coarse heterogeneous
6
54.5
5
45.4
11
45.4%
Fine pleomorphic
10
50
10
50
20
50%
Fine linear and fine linear branching
0
0
4
100
4
100%
Abbreviations: PPV, positive predictive value; S, statistically significant.
a Chi-square test.
The PPV of suspicious microcalcifications associated with mass for morphology descriptors
in our study was as follows: fine pleomorphic, 66.6% and fine linear/fine linear branching,100%.
None were malignant among coarse heterogeneous and none of the microcalcifications
associated with mass were amorphous ([Table 3 ]).
Table 3
PPV of suspicious microcalcifications associated with mass according to the morphology
descriptors
Morphology
Benign
Malignant
Total (n = 21)
PPV
N
Percentage
N
Percentage
Amorphous
0
0
0
0
0
0
Coarse heterogeneous
1
100
0
0
1
0
Fine pleomorphic
5
33.3
10
66.6
15
66.6%
Fine linear and fine linear branching
0
0
5
100
5
100%
Abbreviation: PPV, positive predictive value.
The PPV of suspicious microcalcifications for distribution descriptors was as follows:
regional, 0%; grouped, 38.9%; linear, 66.7%; and segmental, 63.2%. Segmental and linear
distribution patterns were found to have the highest PPV for malignancy ([Table 4 ]).
Table 4
PPV of suspicious microcalcifications according to the distribution descriptors
Distribution
Benign
Malignant
Total (n = 77)
PPV
p -Value[a ]
N
Percentage
N
Percentage
Regional
7
100
0
0
7
0
0.017 (S)
Grouped
22
61.1
14
38.9
36
38.9%
0.195
Segmental
7
36.8
12
63.2
19
63.2%
0.099
Linear
5
33.3
10
66.7
15
66.7%
0.085
Abbreviations: PPV, positive predictive value; S, statistically significant.
a Chisquare test.
The PPV of suspicious microcalcifications according to BI-RADS category was as follows:
BI-RADS IVA, 10.5%; BI-RADS IVB, 16%; BI-RADS IVC, 88.2%; and BI-RADS V, 93.8% ([Table 5 ]).
Table 5
PPV of suspicious microcalcifications according to BI-RADS category
BI-RADS category
Benign
Malignant
Total (n = 77)
PPV
N
Percentage
N
Percentage
BI-RADS IVA
17
89.5
2
10.5
19
10.5%
BI-RADS IVB
21
84
4
16
25
16%
BI-RADS IVC
2
11.8
15
88.2
17
88.2%
BI-RADS V
1
6.2
15
93.8
16
93.8%
Abbreviation: PPV, positive predictive value.
Among 77 individuals with suspicious microcalcifications, 41 (53%) were benign, 12
(16%) were in situ carcinoma, and 24 (31%) were invasive carcinoma ([Fig. 1 ]).
Fig. 1 Distribution patterns of suspicious microcalcifications diagnosed as benign, in situ,
and invasive carcinoma on histopathology.
Among 24 individuals with suspicious microcalcifications diagnosed as invasive carcinoma,
1 (4.1%) was invasive ductal carcinoma grade 1, 13 (54.1%) were invasive ductal carcinoma
grade 2, 6 (25%) were invasive ductal carcinoma grade 3, 3 (12.5%) were invasive ductal
carcinoma grades 2 and 3, and 1 (4.1%) was invasive mucinous carcinoma grade 2.
Among 41 individuals with suspicious microcalcifications diagnosed as benign, 27 (65.8%)
were fibrocystic disease, 6 (14.6%) were benign fibroepithelial lesions, 2 (4.8%)
were benign papillomas, and 6 (14.6%) were benign breast changes ([Table 6 ]).
Table 6
Distribution patterns of suspicious microcalcifications diagnosed as benign on histopathology
Histopathological diagnosis
Frequency (n )
N = 41
Percentage
Fibrocystic disease
27
65.8
Benign fibroepithelial lesion
6
14.6
Benign papilloma
2
4.8
Benign breast change
6
14.6
Discussion
Subcategorization of morphology and distribution descriptors for suspicious microcalcifications
based on the fifth edition of BI-RADS lexicon is important because of the different
PPVs for each type of microcalcifications. Image-guided breast biopsy is accurate,
minimally invasive and can be used as a safe approach for diagnosis in patients with
suspicious breast microcalcifications.
In 2013, the fifth edition of the BI-RADS Atlas was published and incorporated changes
in the management, nomenclature, and descriptors of various lesions and of calcifications.[5 ] In the fourth edition, calcifications were separated into three categories: typically
benign, intermediate concern, and higher probability. They are now consolidated into
two categories: typically benign and suspicious morphology.[6 ]
Suspicious morphology calcifications are described as amorphous, coarse heterogeneous,
fine pleomorphic, and fine linear or fine-linear branching in the BI-RADS lexicon.
Amorphous (historically, “indistinct”) : Small and/or hazy in appearance, hence a more specific particle shape cannot be
determined. These calcifications in a grouped, linear, or segmental distribution are
suspicious and generally warrant biopsy ([Fig. 2 ]).
Fig. 2 (A ) MLO and (B ) CC synthetic mammogram (C view) depicting grouped amorphous microcalcification with
focal asymmetry in the central quadrant of right breast (white arrow). (C ) Specimen radiography after stereotactic biopsy showing sampled microcalcifications.
HPE: DCIS. CC, craniocaudal; DCIS, ductal carcinoma in situ; HPE, histopathological
examination; MLO, mediolateral oblique.
Coarse heterogeneous : Irregular, conspicuous calcifications that are generally between 0.5 and 1 mm and
tend to coalesce, but are smaller than dystrophic calcifications ([Fig. 3 ]).
Fig. 3 (A ) MLO and (B ) CC synthetic mammogram (C view) depicting a large segmental area of coarse heterogenous
calcifications in the superolateral and central quadrant of left breast (white arrow)
with mildly enlarged left axillary lymph node. (C ) Specimen radiography after stereotactic biopsy showing sampled microcalcifications.
HPE: DCIS. CC, craniocaudal; DCIS, ductal carcinoma in situ; HPE, histopathological
examination; MLO, mediolateral oblique.
Fine pleomorphic : More conspicuous than amorphous forms and have discrete shapes. These are distinguished
from fine linear and linear branching forms by the absence of fine-linear particles.
Fine pleomorphic calcifications vary in size and shape and are usually < 0.5 mm in
diameter ([Figs. 4 ]
[5 ]
[6 ]).
Fig. 4 (A ) MLO and (B ) CC synthetic mammograms (C view) depicting a large area of fine pleomorphic with
fine linear branching microcalcifications in the upper inner quadrant of the left
breast in segmental distribution with associated asymmetry (white arrow). (C ) Specimen radiography after Stereotactic biopsy showing sampled microcalcifications.
HPE: DCIS. CC, craniocaudal; DCIS, ductal carcinoma in situ; HPE, histopathological
examination; MLO, mediolateral oblique.
Fig. 5 (A ) MLO and (B ) CC synthetic mammograms (C view) depicting segmental area of fine pleomorphic microcalcifications
in left lower inner and central breast (white arrow). (C, D ) Focal masses with internal microcalcifications are seen on ultrasound. (D ) Ultrasound-guided biopsy showed invasive ductal carcinoma. CC, craniocaudal; MLO,
mediolateral oblique.
Fig. 6 (A ) MLO and (B ) CC synthetic mammogram (C view) depicting segmental area of fine linear branching
and pleomorphic microcalcification in the upper outer quadrant of left breast (thin
white arrow) seen as (C ) echogenic foci on ultrasound US (thick white arrow). (D ) US-guided biopsy showed DCIS. CC, craniocaudal; DCIS, ductal carcinoma in situ;
MLO, mediolateral oblique.
Fine linear or fine-linear branching : These are thin, linear, irregular calcifications, which may be discontinuous and
which are < 0.5 mm in caliber. Occasionally, branching forms may be seen. Their appearance
suggests filling of the lumen of a duct or ducts involved irregularly by breast cancer[7 ] ([Figs. 4 ] and [6 ]).
In our study, overall PPV for malignancy of suspicious microcalcifications without
associated mass was 37.5%, which is within the acceptable range set out in the BI-RADS
fifth edition (20–45% for abnormal screening findings)[8 ] and comparable to study conducted by Liu and Huang (overall PPV was 37.0%).[4 ]
According to the fifth edition of BI-RADS Atlas ACR 2013,[7 ] PPV for morphology descriptors are as follows: amorphous ∼ 20%; coarse heterogeneous,
∼15%; fine pleomorphic, ∼ 29%; and fine linear and linear branching ∼70%.
In our study, the PPV of suspicious microcalcifications without any associated mass
on mammogram, according to the morphology descriptors was as follows: amorphous, 9.5%;
coarse heterogeneous, 45.4%; fine pleomorphic, 50%; and fine linear/fine linear branching,
100% (p -value < 0.001). Fine linear/fine linear branching had the highest PPV of 100%.
Our study found higher PPVs for coarse heterogeneous, fine pleomorphic, and fine linear/fine
linear branching microcalcifications, slightly lower PPV for amorphous microcalcifications
as compared with the fifth edition of BI-RADS Atlas ACR 2013,[7 ] but was well correlated with the final BI-RADS assessment categories.
Reasons for mild variations in our study compared with the fifth edition of BI-RADS
Atlas could be due to the small sample size and the study was not conducted in a regular
screening population.
Our study had slightly higher PPV for coarse heterogeneous group and similar results
were obtained for other groups of morphology descriptors as compared with studies
conducted by Park et al[9 ] (PPV of amorphous, 15.9%; coarse heterogeneous, 31.7%; fine pleomorphic, 58.2%;
and fine linear or branching, 90.6%), Kim et al[3 ] (Overall PPV: 22.4%. PPVs of morphology descriptors: amorphous, 7.9%; coarse heterogeneous,
17.8%; fine pleomorphic, 63.2%; and fine linear/fine linear branching, 100%), and
Metaxa et al[10 ] (PPV of amorphous, 7.1%; coarse heterogeneous, 33.3%; fine pleomorphic, 48.1%; and
fine linear/fine linear branching, 85.2%).
Our study had higher PPV for coarse heterogeneous, lower PPV for the amorphous group,
and similar PPV for other calcifications as compared with a study conducted by Burnside
et al[11 ] (PPV was 7% for coarse heterogeneous, 20–26% for amorphous, 25–41% for fine pleomorphic,
and >80% for linear and branched calcifications).
In our study, the PPV of suspicious microcalcifications for distribution descriptors
was as follows: regional 0%, grouped 38.9%, linear 66.7%, and segmental 63.2%.
Our study had no malignant lesion among microcalcifications, which were categorized
as regional distribution. Linear and segmental types of distribution had the highest
PPV of being malignant. This correlated well with BI-RADS Atlas, fifth edition according
to which the probability of malignancy for grouped is ∼31%, linear is ∼60%, and segmental
is ∼62%. The probability of malignancy is described as ∼26% for regional distribution.
The PPV of segmental and linear type of distribution in our study correlated well
with the studies conducted by Kim et al[3 ] (regional, 8.8%; grouped, 14.3%; linear, 87.5%; and segmental, 63.6%) and Liu and
Huang[4 ] (regional, 46%; grouped, 36%; linear, 68%; and segmental, 78%). Variations were
found with other descriptors.
Our study had similar results as compared with the study conducted by Park et al[9 ] except for regional distribution (regional, 31.5%; grouped, 31.3%; linear, 50%;
and segmental, 77.9%) (p < 0.001).
In our study, 21 (27.3%) out of 77 lesions were associated with mass and underwent
US-guided biopsy. Overall PPV of these suspicious microcalcifications when associated
with mass was 71.4%, and more likely to be invasive which was similar to the study
conducted by Soo et al[12 ] (PPV of 69%) and Bae et al[13 ] (66.2% [51/77] vs. 23.2% [46/198]; p -value < 0.001) or of higher BI-RADS category (61.0% [47/77] vs. 22.2% [44/198]; p -value < 0.001).
PPV of suspicious microcalcifications associated with mass for morphology descriptors
in our study were as follows: fine pleomorphic was 66.6% and fine linear/fine linear
branching was 100%. None was malignant among coarse heterogeneous and none of the
microcalcifications associated with mass was amorphous.
Hence, the presence of mass along with microcalcifications increases the likelihood
of malignancy. Our study also showed sonographically visible lesions with microcalcifications
to be of higher BI-RADS category and more likely to be malignant ([Figs. 5 ] and [6 ]).
In our study, the PPV of suspicious microcalcifications according to BI-RADS category
was as follows: BI-RADS IVA, 10.5%; BI-RADS IVB, 16%; BI-RADS IV C, 88.2%; and BI-RADS
V, 93.8%. This correlated well with the fifth edition of BI-RADS Atlas 2013 according
to which the PPV for each category is as follows: BI-RADS IVA, > 2–< 10%; BI-RADS
IVB, >10–< 50%; BI-RADS IVC, >50–< 95%, and BI-RADS V, >95%.
Clinical Implications
Breast microcalcifications can be the early and only presenting sign of breast cancer,
which can be detected on mammograms. Hence, it is very essential to perform a proper
assessment of various calcifications. Assigning a BI-RADS category to each of these
suspected microcalcifications helps in risk assessment, determines the need for image-guided
biopsy, and guides further management strategies.
