Inter-observer agreement in the diagnosis of breast atypical ductal hyperplasia: A systematic review

Abstract

OBJECTIVE:

To review the concordance between pathologists in the diagnosis of atypical ductal hyperplasia of the breast.

METHODS:

Systematic review of articles registered in the PubMed database with the terms “pathologists” and “atypical ductal hyperplasia”, “atypical ductal hyperplasia” and “interobserver variability”, in Embase with “interobserver agreement in atypical ductal hyperplasia” and Lilacs with the terms in Spanish “patólogos” and “hiperplasia ductal atípica”, without the use of filters, between 03/16/2022 and 03/26/2022 searching for articles that assess inter-observer agreement in the diagnosis of atypical ductal hyperplasia by using the kappa statistic. Review articles, conference proceedings, and commentaries were excluded.

RESULTS:

With the search terms, 507 publications were found. Of these, 491 were excluded from the screening by title and abstract. Of the full reading of the remaining articles, 6 were selected for qualitative analysis. The kappa values ranged from slight agreement, kappa = 0.17, to substantial agreement, 0.69 obtained from pathologists dedicated or experienced in breast pathology and who received a prior tutorial session to review the general criteria for atypia.

CONCLUSIONS:

This study reviews the reproducibility in the diagnosis of atypical ductal hyperplasia. A better concordance was obtained with a previous teaching process offered to pathologists dedicated or experienced in breast pathology.

Keywords

Breast breast diseases risk factors breast neoplasms pathology diagnosis

1. Introduction

Atypical ductal hyperplasia (ADH) is the most common high-risk proliferative lesion in breast cancer [1]. It is defined as a proliferative epithelial lesion that has cytological and architectural characteristics similar to those of low-grade ductal carcinoma in situ (DCIS) but with less compromise in the architecture, degree of involvement of the terminal lobular duct unit, and extension [2]. Page et al. [3] proposed a cut-off of <2 involved spaces to classify a lesion as ADH, but Tavassoli and Norris proposed that lesions ≤2 mm in contiguous extent be classified as ADH considering that using the involvement of at least two spaces was prone to confusion [4].

It is considered a non-obligatory precursor and independent risk factor for developing invasive breast cancer [5–7]. Its diagnosis is associated with a 3 to 5 fold-increased risk of developing invasive breast carcinoma in either breast, although some studies have found a higher risk in the ipsilateral breast [2]. The cumulative incidence of breast cancer with a previous diagnosis of hyperplasia atypical ductal is estimated between 17 and 26% at 10 years and almost 30% at 25 years [8–10].

Fig. 1.

Information flow through the different phases of the systematic review.

Percutaneous core needle biopsy is a widely used procedure for diagnosing breast disease [11–14] due to its diagnostic performance, low rate of complications, minimal incision, acceptance [14], and ease of the technique in an outpatient setting. However, in atypical ductal hyperplasia, because a major difference from low-grade ductal carcinoma in situ is the extent of the lesion, the distinction between these two histologies is not always easy on breast core needle biopsies because the size of the biopsy creates the possibility of undersampling the lesion. To remedy this limitation, if what is identified in the biopsy is atypical ductal hyperplasia, it is recommended to perform an open biopsy [7] which is considered the standard management [15] and with which 10 to 20% of cases are reclassified in recent studies as ductal carcinoma in situ or invasive cases [2]. Even with open biopsy, it is unlikely that a perfect agreement will be achieved, even strictly complying with the current criteria, considering these are arbitrary and more quantitative than qualitative [2].

The pathological diagnosis of breast biopsy is often considered the gold standard for patient management, but it has significant diagnostic variation, being more susceptible to errors than in other areas of surgical pathology [16]. This variation frequently occurs among pathologists in the diagnosis of atypia so that many benign without atypia breast biopsies might be interpreted as having atypia [17], but also in the analysis of other parameters, so the diagnosis can result in a subjective assessment, prone to variability between observers and limited by the amount of tissue obtained [18]. In breast pathology, a particularly problematic area is the distinction between atypical ductal hyperplasia ADH and low-grade ductal carcinoma in situ of limited extention [2]; These are lesions that exist in a biological continuum but have been separated artifactually based on subtle histologic and extensional differences.

Learning the reproducibility of the diagnosis of atypical ductal hyperplasia becomes important to surgeons who rely on the pathologist’s interpretation to guide their management recommendations. This study reviews the studies that have assessed the consistency of the histological diagnosis of atypical ductal hyperplasia of the breast using the kappa statistic [19].

2. Material and methods

A systematic review of articles registered in the PubMed database with the terms “pathologists” and “Atypical ductal hyperplasia”, “Atypical ductal hyperplasia” and “interobserver variability”, in Embase with “interobserver agreement in Atypical ductal hyperplasia” and Lilacs with the Spanish terms “patólogos” and “hiperplasia ductal atípica”, without the use of filters, between 03/16/2022 and 03/26/2022 searching for articles with the evaluation of inter-observer agreement in the diagnosis of atypical ductal hyperplasia by using the kappa statistic. Review articles, conference proceedings, and commentaries were excluded.

Search terms were provided by the author; with another reviewer, the articles on the objective of the research were selected, according to the title, or by additional information in the abstract. Discrepancies were resolved by reviewing the full article and by a mutual agreement. In the articles that continued to be evaluated, the author reviewed the entire article to ensure the relevance of the articles with the objective of the research.

3. Results

With the search terms, 507 publications were found. Of these, 491 were excluded in the screening by title and abstract, 55 of them because they were repeated articles in the search. The rest were excluded because they did not correspond to the objective of the research, were part of the exclusion criteria and in one the reason was the language (Chinese). From the remaining review of 16 full articles, 10 were excluded mostly because they did not assess inter-observer agreement in the diagnosis of atypical ductal hyperplasia by using the kappa statistic, and 6 were finally selected for qualitative analysis (Fig. 1). These are summarized in Table 1. The kappa values were scored according to Table 2. The strength of agreement in the extracted studies ranged from slight agreement, kappa = 0.17, to substantial agreement, 0.69, obtained by pathologists dedicated or experienced in breast pathology and who received a prior tutorial.

Table 1
Characteristics and values of the kappa statistic in the selected studies

Author Cases Evaluated material Patol Num^∗ Characteristics Kappa Methodological aspects

Jain et al. [24] 81 Cases of proliferative lesions 9 Pathologists of the Department of Pathology and Laboratory Medicine from Indiana University 0.43 Concordance with the diagnoses of the majority was evaluated in hematoxylin-eosin slides and after immunohistochemistry with ADH-5.

75 0.58

MacGrogan et al. [25] 13 Excisional or vacuum-assisted biopsies 20 Pathologists from 16 hospitals in the French cancer network and two university hospitals, one Belgian hospital, and one private laboratory. 0.36 Concordance was evaluated with hematoxylin-eosin slides and with immunohistochemistry CK5/6 and E-cadherin.

0.43

Darvishian et al. [20] 20 Atypia in breast core needle biopsies 4 Dedicated pathologists or, with experience in breast pathology. 0.69 Institutional cases

A previous tutorial was made on

Gomes et al. [23] 610 Cases of proliferative lesions ND^∗∗ A pathologist with experience and training in breast pathology vs reports of general pathologists 0.44 Cases received for a second opinion

Elston et al. [21] 32 Cases of proliferative lesions 23 European pathologists with an interest in breast pathology 0.35 Hematoxylin-eosin stained sections

0,29 3 years later, digitized images of small parts of the evaluated sections

Sloane et al. [22] 12 Pathologies associated with screening 17 Regional coordinators^∗∗∗ 0.25 Six pathologies each with two cases were evaluated, 53 sections were made from each block, and the first and last were left as reference.

186–251 Non-coordinators Pathologists 0.17

Author	Cases	Evaluated material	Patol Num^∗	Characteristics	Kappa	Methodological aspects
Jain et al. [24]	81	Cases of proliferative lesions	9	Pathologists of the Department of Pathology and Laboratory Medicine from Indiana University	0.43	Concordance with the diagnoses of the majority was evaluated in hematoxylin-eosin slides and after immunohistochemistry with ADH-5.
	75				0.58
MacGrogan et al. [25]	13	Excisional or vacuum-assisted biopsies	20	Pathologists from 16 hospitals in the French cancer network and two university hospitals, one Belgian hospital, and one private laboratory.	0.36	Concordance was evaluated with hematoxylin-eosin slides and with immunohistochemistry CK5/6 and E-cadherin.
					0.43
Darvishian et al. [20]	20	Atypia in breast core needle biopsies	4	Dedicated pathologists or, with experience in breast pathology.	0.69	Institutional cases
						A previous tutorial was made on
Gomes et al. [23]	610	Cases of proliferative lesions	ND^∗∗	A pathologist with experience and training in breast pathology vs reports of general pathologists	0.44	Cases received for a second opinion

Elston et al. [21]	32	Cases of proliferative lesions	23	European pathologists with an interest in breast pathology	0.35	Hematoxylin-eosin stained sections
					0,29	3 years later, digitized images of small parts of the evaluated sections
Sloane et al. [22]	12	Pathologies associated with screening	17	Regional coordinators^∗∗∗	0.25	Six pathologies each with two cases were evaluated, 53 sections were made from each block, and the first and last were left as reference.
			186–251	Non-coordinators Pathologists	0.17

^∗Number of pathologists. ^∗∗Not defined. ^∗∗∗The U.K. National Cancer Screening Evaluation Unit. Regional Coordinators (one for each of the 14 English health regions and Scotland, Wales, and Northern Ireland).

Table 2

Interpretation of the kappa statistic [19]

Kappa statistic	<0	0.00–0.20	0.21–0.40	0.41–0.60	0.61–0.80	0.81–1.00
Strength of agreement	Poor	Slight	Fair	Moderate	Substantial	Almost perfect

The evaluations of concordance were made among dedicated pathologists, with experience or interest in breast pathology in two publications [20,21], one publication describes many pathologists with interest in breast pathology [22], a pathologist with experience and training compared to general pathologists [23] and in two publications, the affinity of pathologists with breast pathology was not established [24,25].

4. Discussion

Atypical ductal hyperplasia represents a diagnostic dilemma, with wide variation in reproducibility among pathologists, mainly due to problems in differentiating atypical ductal hyperplasia from low-grade ductal carcinoma in situ. The low consistency observed in the diagnosis raises serious concerns regarding the robustness of the diagnostic criteria. Elston et al. [21] raises two possible reasons for this inconsistency, the differences in interpretation of the specific histological characteristics, the most important, and the second in the selection of a field where the morphology is heterogeneous. These authors emphasize that fair consistency in their study is due to differences in morphological interpretation.

When evaluating the concordance for atypical ductal hyperplasia in the selected studies, it was found that the interobserver reproducibility ranges from slight to a substantial agreement, with reproducibility indices (kappa) ranging between 0.17 [22] and 0.69 [20]. A substantial agreement was found among pathologists with dedication or experience in breast pathology when a previous tutorial on the subject was included, as assessed in the study by Darvishian et al. [20], obtaining a concordance that is higher than that one of the other studies reviewed.

However, these strategies defined as adequate to improve concordance between pathologists are far from resolving these limitations in diagnosis that can lead to inadequate treatments [17]. Schnitt et al. [26]. In their study on proliferative lesions, observed a complete agreement of 58% among the participating pathologists. According to their study, interobserver variability can be reduced using standardized criteria for ductal lesions by using Page’s criteria [3] and providing training slides. In contrast, Palazzo and Hyslop [27] documented a fair kappa value (0.36) in the diagnosis of benign and malignant ductal lesions with the same criteria among the pathologists participating in the study (community and academic pathologists).

The inclusion of studies with immunohistochemistry in the diagnosis improved the concordance from fair to moderate in the study by MacGrogan [25], but continued to be moderate in the study by Jain [24], showing only a marginal advantage. The use of digitized images marking a specific field was not useful either (on the contrary, it slightly decreased the concordance) remaining in the range of a fair concordance [21] reflecting inconsistencies due to differences in the morphological interpretation [2].

The diagnosis of ADH is problematic as it represents a spectrum of atypical architectural patterns ranging from flat epithelial atypia to low-grade DCIS. Distinguishing between some atypical ductal hyperplasia lesions and small low-grade ductal cancers in situ remains a problem, which has implications for inter-observer variability, some less common patterns (solid and Pagetoid) can be difficult to distinguish from classic lobular carcinoma in situ [2] and additionally, there is a considerable variation in the diagnosis of atypia so that some benign without atypia breast biopsies can be considered as having atypia [17].

The correct identification of ADH is an essential step for the adequate clinical management of patients [28–31]. The current treatment for ADH is excision, while the options for DCIS are mastectomy, breast-conserving surgery followed by radiation treatment, and breast-conserving surgery alone. Any of these treatments can be followed by hormone therapy [31].

The included studies in this systematic review were defined by their evaluation with the kappa statistic, which is the most widely accepted concordance test when the data is categorical [19]. Its values are usually between 0 (no agreement) and 1 (absolute agreement). A negative value can be obtained in situations where the actual agreement is less than the value expected by chance [32] Table 2.

Strategies are required to reduce the risks of inter-observer variability in management, such as considering reviewing the diagnostic standards of this pathology as well as its differential diagnoses, referral of histological studies in the event of diagnostic suspicion of ADH to pathologists trained in breast pathology with continuous education processes in their practice and considering novel and creative approaches such as artificial intelligence and molecular markers [33].

5. Conclusion

Studies on concordance in the diagnosis of atypical ductal hyperplasia between pathologists show an important variation that is reduced when pathologists are dedicated or have experience in breast pathology. The histologic continuum of the atypical ductal hyperplasia and the low-grade ductal carcinoma in situ represents a major diagnostic problem that also has therapeutic implications. New developments are required to improve the performance in the diagnosis of this entity, such as artificial intelligence and molecular analysis.

Footnotes

Conflicts of interests

The author has no conflicts of interest.

Funding

The author has no funding to declare.

References

Badan

, Roveda Júnior

, Piato

, Fleury

EFC

, Campos

MSD

, Pecci

CAF

, Diagnostic underestimation of atypical ductal hyperplasia and ductal carcinoma in situ at percutaneous core needle and vacuum-assisted biopsies of the breast in a Brazilian reference institution, Radiol Bras , 49: 6–11, 2016. doi:10.1590/0100-3984.2014.0110.

Allison

, Collins

, Moriya

, Sanders

, Visscher

. Atypical ductal hyperplasia. WHO Classification of Tumours Editorial Board. Breast tumours. Lyon (France): International Agency for Research on Cancer . . 5th ed. WHO Classification of Tumours Series, Vol. 2,18–21. 2019.

Page

, Dupont

, Rogers

, Rados

, Atypical hyperplastic lesions of the female breast. A long-term follow-up study, Cancer , 55: 2698–2708, 1985. doi:10.1002/1097-0142(19850601)55:11<2698::aid-cncr2820551127>3.0.co;2-a.

Tavassoli

, Norris

, A comparison of the results of long-term follow-up for atypical intraductal hyperplasia and intraductal hyperplasia of the breast, Cancer , 65: 518–529, 1990. doi:10.1002/1097-0142(19900201)65:3<518::aid-cncr2820650324>3.0.co;2-o.

Simukayi

, Chang

, Nemer

, Van Sant

, Sun

, McIlvride

, Prospective analysis using a novel CNN algorithm to distinguish atypical ductal hyperplasia from ductal carcinoma in situ in breast, Clin Breast Cancer , 20(6): e757–e760, 2020. doi:10.1016/j.clbc.2020.06.001.

Dupont

, Page

, Risk factors for breast cancer in women with proliferative breast disease, N Engl J Med , 312(3): 146–151, 1985. doi:10.1056/NEJM198501173120303.

Schiaffino

, Calabrese

, Melani

, Trimboli

, Cozzi

, Carbonaro

, Upgrade rate of percutaneously diagnosed pure atypical ductal hyperplasia: Systematic review and meta-analysis of 6458 lesions, Radiology , 294: 76–86, 2020. doi:10.1148/radiol.2019190748.

Hartmann

, Degnim

, Santen

, Dupont

, Ghosh

, Atypical hyperplasia of the breast — risk assessment and management options, N Engl J Med , 372: 78–89, 2015. doi:10.1056/NEJMsr1407164.

Hartmann

, Radisky

, Frost

, Santen

, Vierkant

, Benetti

, Understanding the premalignant potential of atypical hyperplasia through its natural history: A longitudinal cohort study, Cancer Prev Res (Phila) , 7: 211–217, 2014. doi:10.1158/1940-6207.CAPR-13-0222.

10.

Coutant

, Canlorbe

, Bendifallah

, Beltjens

, Benign proliferative breast disease with and without atypia (in French), Journal de Gynécologie Obstétrique et Biologie de la Reproduction , 44: 980–995, 2015. doi:10.1016/j.jgyn.2015.09.037.

11.

Tse

, Tan

, Diagnosing breast lesions by fine needle aspiration cytology or core biopsy: Which is better? Breast Cancer Res Treat , 123: 1–8, 2010. doi:10.1007/s10549-010-0962-4.

12.

Pinder

, Shaaban

, Deb

, Desai

, Gandhi

, Lee

AHS

, NHS breast screening multidisciplinary working group guidelines for the diagnosis and management of breast lesions of uncertain malignant potential on core biopsy (B3 lesions), Clin Radiol , 73: 682–692, 2018. doi:10.1016/j.crad.2018.04.004.

13.

Amitai

, Menes

, Golan

, Use of breast magnetic resonance imaging in women diagnosed with atypical ductal hyperplasia at core needle biopsy helps select women for surgical excision, Can Assoc Radiol J , 69: 240–247, 2018. doi:10.1016/j.carj.2018.03.003.

14.

Cassano

, Trentin

. Integrated breast biopsy for best radiological diagnosis of breast cancer. Breast Cancer Innovations in Research and Management , . Veronesi

, Goldhirsch

, Veronesi

, Gentilini

, Leonardi

(eds), (New York), Springer; 2017.

15.

Latronico

, Nicosia

, Faggian

, Abbate

, Penco

, Bozzini

, Atypical ductal hyperplasia: Our experience in the management and long term clinical follow-up in 71 patients, Breast , 37: 1–5, 2018. doi:10.1016/j.breast.2017.10.003.

16.

Miller

, Flanagan

, Lowry

, Kilgore

, Error reduction and diagnostic concordance in breast pathology, Surg Pathol Clin , 15: 1–13, 2022. doi:10.1016/j.path.2021.11.001.

17.

Elmore

, Longton

, Carney

, Geller

, Onega

, Tosteson

, Diagnostic concordance among pathologists interpreting breast biopsy specimens, JAMA , 313: 1122–1132, 2015. doi:10.1001/jama.2015.1405.

18.

Tsuchiya

, Mori

, Schacht

, Sheth

, Karczmar

, Newstead

, Value of breast MRI for patients with a biopsy showing atypical ductal hyperplasia (ADH), J Magn Reson Imaging , 46: 1738–1747, 2017. doi:10.1002/jmri.25694.

19.

Landis

, Koch

, The measurement of observer agreement for categorical data, Biometrics , 33: 159–174, 1977.

20.

Darvishian

, Singh

, Simsir

, Ye

, Cangiarella

, Atypia on breast core needle biopsies: Reproducibility and significance, Ann Clin Lab Sci , 39: 270–276, 2009.

21.

Elston

, Sloane

, Amendoeira

, Apostolikas

, Bellocq

, Bianchi

, Causes of inconsistency in diagnosing and classifying intraductal proliferations of the breast. European commission working group on breast screening pathology, Eur J Cancer , 36: 1769–1772, 2000. doi:10.1016/s0959-8049(00)00181-7.

22.

Sloane

, Ellman

, Anderson

, Brown

, Coyne

, Dallimore

, Consistency of histopathological reporting of breast lesions detected by screening: Findings of the UK National External Quality Assessment (EQA) scheme. UK National Coordinating Group for Breast Screening Pathology, Eur J Cancer , 30A: 1414–1419, 1994. doi:10.1016/0959-8049(94)00261-3.

23.

Gomes

, Porto

, Balabram

, Gobbi

, Inter-observer variability between general pathologists and a specialist in breast pathology in the diagnosis of lobular neoplasia, columnar cell lesions, atypical ductal hyperplasia and ductal carcinoma in situ of the breast, Diagn Pathol , 9: 121, 2014. doi:10.1186/1746-1596-9-121.

24.

Jain

, Mehta

, Dimitrov

, Larsson

, Musto

, Hodges

, Atypical ductal hyperplasia: interobserver and intraobserver variability, Mod Pathol , 24: 917–923, 2011. doi:10.1038/modpathol.2011.66.

25.

MacGrogan

, Arnould

, de Mascarel

, Vincent-Salomon

, Penault-Llorca

, Lacroix-Triki

, Impact of immunohistochemical markers, CK5/6 and E-cadherin on diagnostic agreement in non-invasive proliferative breast lesions, Histopathology , 52: 689–697, 2008. doi:10.1111/j.1365-2559.2008.03016.x.

26.

Schnitt

, Connolly

, Tavassoli

, Fechner

, Kempson

, Gelman

, Interobserver reproducibility in the diagnosis of ductal proliferative breast lesions using standardized criteria, Am J Surg Pathol , 16: 1133–1143, 1992. doi:10.1097/00000478-199212000-00001.

27.

Palazzo

, Hyslop

, Hyperplastic ductal and lobular lesions and carcinomas in situ of the breast: Reproducibility of current diagnostic criteria among community-and academic-based pathologists, Breast J , 4: 230–237, 1998. doi:10.1046/j.1524-4741.1998.440230.x.

28.

Guerini-Rocco

, Fusco

. Premalignant and pre-invasive lesions of the breast. Breast Cancer: Innovations in Research and Management , . Veronesi

, Goldhirsch

, Veronesi

, Gentilini

, Leonardi

(eds), Springer International Publishing AG; 2017.

29.

Sanders

, Podoll

, Atypical ductal hyperplasia-ductal carcinoma in situ spectrum: diagnostic considerations and treatment impact in the era of deescalation, Surg Pathol Clin , 15: 95–103, 2022. doi:10.1016/j.path.2021.11.006.

30.

Gradishar

, Moran

, Abraham

, Aft

, Agnese

, Allison

, NCCN guidelines (R) insights: Breast cancer, version 4.2021, J Natl Compr Canc Netw , 19: 484–493, 2021. doi:10.6004/jnccn.2021.0023.

31.

Solin

, Management of ductal carcinoma in situ (DCIS) of the breast: present approaches and future directions, Curr Oncol Rep , 21: 33, 2019. doi:10.1007/s11912-019-0777-3.

32.

Zito

, Verderio

, Simone

, Angione

, Apicella

, Bianchi

, Reproducibility in the diagnosis of needle core biopsies of non-palpable breast lesions: An international study using virtual slides published on the world-wide web, Histopathology , 56: 720–726, 2010. doi:10.1111/j.1365-2559.2010.03548.x.

33.

Brem

, Management of breast atypical ductal hyperplasia: now and the future, Radiology , 294: 87–88, 2020. doi:10.1148/radiol.2019192192.