Volume 73, No.11: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
727
Original Article
SMJ
Warapan Numprasit, M.D.*, Norasate Samarnthai, M.D.**, Tichakorn Srianuchat, M.D.***
*Department of Surgery, **Department of Pathology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, ***anyarak Breast Center,
Siriraj Hospital, Bangkok 10700, ailand.
Pure Flat Epithelial Atypia of the Breast on Core
Needle Biopsy: No Need for Surgical Excision
ABSTRACT
Objective: From previous data, pure at epithelial atypia (FEA) of the breast demonstrated on core needle biopsy
(CNB) was related to malignant upgrading. However, FEA itself is not an independent factor for developing breast
cancer; therefore, the necessity of subsequent surgical excision is controversial. is study aimed to evaluate the
upgrade rate of FEA aer surgical excision and to demonstrate the necessity of surgical excision in FEA lesions
identied on CNB.
Materials and Methods: is retrospective study involved a review of the clinical features, mammographic and
ultrasound ndings, and pathological reports of patients with pure FEA found from CNB specimens between
January 2010 to January 2019. FEA accompanied with atypical ductal hyperplasia, atypical lobular hyperplasia,
ductal carcinoma in situ (DCIS), and invasive cancer (IC) in ipsilateral breast were excluded. FEA upgrade is dened
as patients with in situ or invasive cancer presented in surgical excision specimens. e breast imaging results of
pure FEA and FEA upgrade subsets were compared.
Results: In total, 45 pure FEA specimens were revealed from CNB; of which, 6 of the pure FEA (13.33%) did not
undergo further surgical excision, however, they showed no recurrence during follow-up (median follow-up time:
2.68 years). e majority of FEA cases were detected by mammography in 39 patients (86.67%). Of the 45 patients,
32 were classied into BI-RADS 4B (71.11%), 11 as BI-RADS 4A (24.44%), and 2 as BI-RADS 4C (4.44%). One
patient was upgraded to DCIS (2.7%). BI-RADS classication did not dier between upgrade FEA and non-upgrade
FEA groups (p=0.49).
Conclusion: Only a 2.7% upgrade rate, omitting the surgical excision of pure FEA from CNB, was possible. Even
though our study could not demonstrate a correlation between FEA upgrade and radiological ndings, BIRADS
4A was less likely to carry the malignant cells. Furthermore, segmental microcalcication tended to be associated
with upgraded lesions, but not signicantly.
Keywords: Breast cancer; at epithelial atypia (FEA); upgrade rate (Siriraj Med J 2021; 73: 727-731)
Corresponding author: Warapan Numprasit
E-mail: gi_warapan@gmail.com
Received 5 November 2020 Revised 11 May 2021 Accepted 12 May 2021
ORCID ID: https://orcid.org/0000-0001-9027-7406
http://dx.doi.org/10.33192/Smj.2021.93
INTRODUCTION
In recent decades, the diagnosis of at epithelial
atypia (FEA) has been increasing. Previously, FEA was
known as intraepithelial neoplasia, albeit with an unclear
exact pathological description. In 2003, the World Health
Organization (WHO) coined the term “at epithelial
atypia” to describe a lesion where native ductal cells
are replaced by atypical columnar or cuboidal cells.
1
e incidence of upgrading to malignancy, in which
atypical cells occur concomitantly with malignancy (in
Volume 73, No.11: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
728
situ or invasive carcinoma) in larger specimens, aer
FEA has been diagnosed in core needle biopsy (CNB)
specimens as varying from 0 - 42%.
2-4
erefore, the
current standard treatment for FEA aer CNB is surgical
excision. However, FEA itself is not an independent factor
for developing breast cancer. Previous studies found
no risk of breast cancer in FEA lesions aer 13 years
follow-up.
5,6
As a result, the necessity for subsequent
surgical excision is controversial. is study aimed to
evaluate the upgrading rate of FEA aer surgical excision
and to evaluate FEA patients identied by CNB who
have the potential to avoid surgical excision.
MATERIALS AND METHODS
is study involved a retrospective analysis of medical
records and was approved by the Ethics Committee of the
Siriraj Institutional Review Board, Faculty of Medicine
Siriraj Hospital (Si 482/2019). All pathological reports
from CNB specimens diagnosed as FEA from January
2010 to January 2019 were reviewed. Pure FEA from core
needle biopsy specimens, dened as only FEA or FEA
concomitantly occurring with other non-proliferative
or benign proliferative epithelial lesions, were included.
All patients had mammography (MMG) and breast
ultrasound (US) performed followed by core needle
biopsy at Siriraj Hospital. FEA accompanied with atypical
ductal hyperplasia (ADH), atypical lobular hyperplasia
(ALH), ductal carcinoma in situ (DCIS), and invasive
cancer (IC) in ipsilateral breast were excluded. However,
pure FEA from CNB in one breast with contralateral
ADH, ALH, DCIS, and IC were included. All patients
underwent excisional biopsy, otherwise in the case of
patients for whom surgical excision was not performed,
they needed to undergo surveillance at Siriraj Hospital.
Clinical presentation, mammographic and
ultrasonographic ndings, and pathological reports were
collected. For core needle tissue sampling, suspicious
microcalcifications were biopsied by stereotactic-
guided biopsy using a 14-gauge core-biopsy needle
(BARD®MAGNUM®), 14- or 9-gauge vacuum-assisted
device (Eviva®, Hologic), and if a mass was targeted,
ultrasound guidance with a 14-gauge needle (BARD®,
MAGNUM®). e average was obtained from ve cores.
A radiographic clip was placed in some patients for
facilitating follow-up.
Statistical analyses were performed using SPSS
soware version 21 (IBM SSPS, Chicago, IL). Categorical
data was reported as the median with the interquartile
range, mean with SD, or as a percentage. e chi-square
test was used to examine the association between upgrade
to malignancy, the morphology, and the distribution of
the microcalcications. Student’s t-test or Mann–Whitney
U-test was applied to analyze the continuous data. For
the categorical data, the
c
2
-test or Fisher-exact test were
used to analyze for statistical signicance. A p-value
less than 0.05 was considered statistically signicant
throughout this study.
RESULTS
During January 2010 to January 2019, 45 pure
FEA lesions diagnosed from CNB were identied. e
baseline characteristics of the patients are described in
Table 1. Of these 45 lesions, 39 were surgically removed.
Six lesions (13.3%) were observed with no following
surgery; however, they showed no recurrence during
surveillance (median follow-up, 2.7 years; range, 29–2844
days). In this study, only one lesion was found upgraded
to DCIS. e median age at diagnosis was 49 years old. Of
the 45 patients, 5 (11.1%) had a history of benign breast
disease at the index breast, while no-one had a family
history of breast or ovarian cancer. e majority of FEA
were detected by mammography (39 lesions; 86.7%);
whereas, 6 lesions were detected by ultrasonography.
When classifying according to BI-RADS classication
(Breast Imaging Reporting and Data System, established
by the American College of Radiology), 11 lesions were
categorized into BI-RADS 4a (24.4%), 32 into BI-RADS
4b (71.1%), and only 2 into BI-RADS 4c (4.4%). All 45
lesions were biopsied; 73.3% stereotactic-guided, 13.3%
vacuum assisted, and 13.3% by US-guided (Table 1).
In cases in which surgical excision was done, FEA
from CNBs were found as pure FEA in the nal surgical
specimens in 26 of 39 patients (66.7%) and coexisted with
either ADH, IDC, or DCIS in 13 patients (33.3%) (Table 2).
Of these latter 13 patients, 1 patient was upgraded to
DCIS (2.6% of total pure FEA from CNB), while no-one
was associated with invasive cancer, and 1 patient had
FEA with ALH (2.6%); meanwhile, 11 patients had FEA
accompanied with ADH (28.2%): 8 found with ADH
and 3 found with ADH, ALH, or LCIS.
Regarding the radiological ndings of the 26 patients
whose nal surgical specimens had conrmed FEA,
8 of the 26 patients were BIRADS 4a (30.8%), 17 were
BIRADS 4b (65.4%), and 1 was BIRADS 4c (3.9%).
Four lesions were detected by US and 22 by MMG. e
mammographic ndings of pure FEA presented with
microcalcication (MC) were as described. In terms
of the shape, 19 were amorphous, 2 punctate, and 1
round. In terms of the distribution, there were 9 clusters,
8 groups, 4 regional, and 1 linear.
ere were 12 lesions for which the nal pathological
report from the surgical excisional specimen demonstrated
Numprasit et al.
Volume 73, No.11: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
729
Original Article
SMJ
TABLE 1. Baseline characteristics of pure FEA from CNB patients.
Characters CNB pure FEA (n = 45)
Age (years)
Range 37–61
Median 49.0
Detected lesions
Microcalcication 39
Mass 5
Microcalcicationwithmass 1
BIRADS
1–3 0
4a 11
4b 32
4c 2
5 0
Target lesion sampling
Microcalcication 39
Stereotactic 33
Vacuumassisted 6
Mass
USG-guided 6
Residualcalcication
Yes 16
No 18
Unknown 5
Surgery after FEA identied from CNB
Yes 39
No 6
Histology of excision specimens
Pure FEA 26
FEAwithAH 12
FEAwithcancer 1
Abbreviations: AH: atypical hyperplasia, CNB: core needle biopsy, FEA: at epithelial atypia
TABLE 2. Radiographic ndings according to the nal histological results.
Group Total Pure FEA FEA with AH FEA with DCIS P-value
Factor n = 39, (%) n = 26, (%) n = 12, (%) n = 1, (%)
Detectionmethods
MMG 35 22 12 1
US 4 4 0 0
BI-RADS
4A 9 8 (30.8) 1 (8.3) 0 0.49
4B 28 17 (65.4) 10 (83.3) 1 (100)
4C 2 1 (3.9) 1 (8.3) 0
Volume 73, No.11: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
730
FEA combined with high-risk benign breast lesion. Of these
12 lesions, 10 were BIRADS 4b (83.3%), 1 was BIRADS
4a (8.3%), and 1 was BIRADS4c (8.3%). In terms of the
shape of the microcalcications in these 12 lesions, 10
were amorphous, 1 was ne linear, and 1 was punctuate.
Interestingly, 1 patient whose nal surgical excision
specimen showed DCIS was categorized into BIRADS
4b with segmental amorphous microcalcication. When
comparing the BIRADS classications, no dierence
was observed between the pure and combined FEA
subgroups (p = 0.49).
DISCUSSION
Overall, 1 lesion in our study was upgraded to DCIS
(2.6%) and none of the lesions were invasive carcinoma.
In addition, breast cancer did not occur whether omitting
the surgical group or following surgery in the group
without the upgrading. Despite the risk of subsequent
breast cancer of FEA being unclear, just as in previous
studies, this could not be proven conclusively
5,7
, but
FEA itself tended to not increase the risk of subsequent
cancer. Said et al. found no increasing long-term breast
cancer events when FEA was found concomitantly with
either atypical hyperplasia (AH) or proliferative lesions
(PL): [AH + FEA 4.74 vs. AH 4.23; p = 0.59] and [PL +
FEA 2.04 vs. PL 1.90; p = 0.76], respectively.
6
In the case
of pure FEA aer surgical excision, de Mascarel et al.
also found that none of the pure FEA patients in their
study had breast cancer during 10-year follow-ups.
8
e ndings from our study were similar in terms of
the rarity of events following upgrading to malignancy
in FEA lesions diagnosed on CNB.
In practice, pure FEA from CNB should be followed
by surgical removal. is relies on previous evidence,
which demonstrated a 10-40% combination of FEA
with other malignant lesions (either in situ or invasive
carcinoma) from CNB specimens.
2-4,9
However, when
emphasizing cancer upgrading (DCIS or invasive cancer),
the possibility of FEA accompanying these malignant cells
is low (7.5%) compared to its co-incidence with other
high-risk lesions (18.6%).
2
When vacuum-assisted core
needle biopsy (VCNB) was rst introduced, the rate of
upgrading decreased dramatically. Recent investigations
using VCNB reported decreasing histologic upstaging at
0 - 3%.
10,11
However, even our practices did not routinely
use VCNB in all microcalcication-detected cases, and
our upgrading rate was only 2.6%. The necessity of
surgical excision, therefore, is controversial.
Despite the low incidence of malignant upgrading,
the rate of combining ADH was high (30.7%). In addition,
our study could not demonstrate the correlation between
FEA upgrading and radiologic ndings. Even though most
patients presented with suspicious microcalcications
detected by mammography and were classied into
BIRADS 4b, the study population was too small to
specify the signicance of the imaging classication to
discriminate whether the lesion was pure or mixed FEA.
In the case of mammographic abnormalities, amorphous
microcalcications were more commonly identied in
30 of 39 specimens (76.9%) than other shapes and the
majority were distributed as clusters. In 1 lesion with
DCIS upgrading, the patient presented with segmental
amorphous microcalcication, which was the only case
where the microcalcication was distributed as a segmental
shape. Likewise, previous studies failed to identify the
specic radiographic characters for FEA upstaging
3,12
,
which tended to be clustered or segmental amorphous
microcalcications.
3,13
Although, no distinctive breast
imaging was noted, FEA diagnosed as BIRADS 4a was
less likely to be upgraded, and occurred in only one of
9 patients (11.1%), meanwhile, FEA also occurred with
BIRADS 4b (11 of 28; 39.3%) or 4c (1 of 2; 50%); therefore,
BIRADS classication was not a good independent predictor
for selecting patients to observe instead of performing
surgery. Among one of several studies, Alencherry et al.
recently reported that one of the independent risk factors
to upstaging was a history of cancer in individuals or
rst-degree relatives
13,14
, but no patients in our cohort
had a family history of breast cancer.
However, there were some limitations in our study
to note. As the pathological denition of FEA has only
recently been introduced in the past few decades, there may
have been interobserver variabilities of the interpretation
among pathologists and it might have been reported
as columnar cell change in some cases, thus causing a
smaller number to be included in the study population,
especially as we could not review all the slide specimens.
Another shortcoming, because of the higher cost of
vacuum-assisted devices, is that in our hospital, core
needle biopsy with a 14-gauge needle rather than VCNB
is the most practised technique, which might be less
accurate.
CONCLUSION
Flat epithelial atypia is a marker of carrying a high
risk lesion rather than for upgrading to breast cancer.
Even though the histological nding may show atypical
cells, the risk of subsequent breast cancer is very low
compared to ADH. Surgical excision may be omitted
particularly in cases of pure FEA from core needle biopsy.
Numprasit et al.
Volume 73, No.11: 2021 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
731
Original Article
SMJ
REFERENCES
1. Tavassoli FA DP. Pathology and Genetics: Tumours of the
Breast and Female Genital Organs series WCoT, editor. Lyon,
France: IARC Press; 2003.
2. Rudin AV, Hoskin TL, Fahy A, Farrell AM, Nassar A, Ghosh K,
et al. Flat Epithelial Atypia on Core Biopsy and Upgrade to
Cancer: a Systematic Review and Meta-Analysis. Ann Surg
Oncol. 2017;24(12):3549-58.
3. Solorzano S, Mesurolle B, Omeroglu A, El Khoury M, Kao E,
Aldis A, et al. Flat epithelial atypia of the breast: pathological-
radiological correlation. AJR Am J Roentgenol. 2011;197(3):
740-6.
4. Khoumais NA, Scaranelo AM, Moshonov H, Kulkarni SR,
Miller N, McCready DR, et al. Incidence of breast cancer in
patients with pure at epithelial atypia diagnosed at core-needle
biopsy of the breast. Ann Surg Oncol. 2013;20(1):133-8.
5. Boulos FI, Dupont WD, Simpson JF, Schuyler PA, Sanders ME,
Freudenthal ME, et al. Histologic associations and long-term
cancer risk in columnar cell lesions of the breast: a retrospective
cohort and a nested case-control study. Cancer. 2008;113(9):
2415-21.
6. Said SM, Visscher DW, Nassar A, Frank RD, Vierkant RA,
Frost MH, et al. Flat epithelial atypia and risk of breast cancer:
A Mayo cohort study. Cancer. 2015;121(10):1548-55.
7. Neal L, Sandhu NP, Hieken TJ, Glazebrook KN, Mac Bride MB,
Dilaveri CA, et al. Diagnosis and Management of Benign,
Atypical, and Indeterminate Breast Lesions Detected on Core
Needle Biopsy. Mayo Clin Proc. 2014;89(4):536-47.
8. de Mascarel I, MacGrogan G, Mathoulin-Pélissier S, Vincent-
Salomon A, Soubeyran I, Picot V, et al. Epithelial atypia in
biopsies performed for microcalcications. practical considerations
about 2,833 serially sectioned surgical biopsies with a long
follow-up. Virchows Arch. 2007;451(1):1-10.
9. Lavoue V, Roger CM, Poilblanc M, Proust N, Monghal-Verge
C, Sagan C, et al. Pure at epithelial atypia (DIN 1a) on core
needle biopsy: study of 60 biopsies with follow-up surgical
excision. Breast Cancer Res Treat. 2011;125(1):121-6.
10. Chan PMY, Chotai N, Lai ES, Sin PY, Chen J, Lu SQ, et al.
Majority of at epithelial atypia diagnosed on biopsy do not
require surgical excision. Breast. 2018;37:13-17.
11. Hugar SB, Bhargava R, Dabbs DJ, Davis KM, Zuley M, Clark
BZ. Isolated Flat Epithelial Atypia on Core Biopsy Specimens
Is Associated With a Low Risk of Upgrade at Excision. Am J
Clin Pathol. 2019;151(5):511-5.
12. Elif A, Burcu S, Nazan C, Sumru CZ, Kemal AN. Columnar cell
lesions of the breast: radiological features and histological
correlation. Med Ultrason. 2015;17(2):147-54.
13. Alencherry E, Goel R, Gore S, ompson C, Dubchuk C, Bomeisl
P, et al. Clinical, imaging, and intervention factors associated
with the upgrade of isolated at epithelial atypia. Clin Imaging.
2019;54:21-4.
14. Berry JS, Trappey AF, Vreeland TJ, Pattyn AR, Clion GT,
Berry EA, et al. Analysis of Clinical and Pathologic Factors
of Pure, Flat Epithelial Atypia on Core Needle Biopsy to Aid
in the Decision of Excision or Observation. J Cancer. 2016;7(1):
1-6.
