Advertisement

Diagnostic accuracy of axillary staging by ultrasound in early breast cancer patients

  • Author Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Fabian Riedel
    Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Author Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Benedikt Schaefgen
    Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Hans-Peter Sinn
    Affiliations
    Institute of Pathology, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Manuel Feisst
    Affiliations
    Institute of Medical Biometry and Informatics, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • André Hennigs
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Sarah Hug
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Angela Binnig
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Christina Gomez
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Aba Harcos
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Anne Stieber
    Affiliations
    Department of Diagnostic and Interventional Radiology, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Hans-Ulrich Kauczor
    Affiliations
    Department of Diagnostic and Interventional Radiology, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Christof Sohn
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Michael Golatta
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Author Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Antonia Glaeser
    Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Author Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Joerg Heil
    Correspondence
    Corresponding author at: Department of Obstetrics and Gynecology, Heidelberg University Hospital, Im Neuenheimer Feld 440, D-69120 Heidelberg, Germany.
    Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
    Affiliations
    Department of Obstetrics and Gynecology, Breast Unit, University Hospital of Heidelberg, Germany
    Search for articles by this author
  • Author Footnotes
    1 Fabian Riedel and Benedikt Schaefgen share first authorship. Antonia Glaeser and Jörg Heil share last authorship.
Open AccessPublished:December 04, 2020DOI:https://doi.org/10.1016/j.ejrad.2020.109468

      Abstract

      Background

      Axillary ultrasound (AUS) is a standard procedure in the preoperative clinical identification of axillary metastatic lymph node (LN) involvement. It guides decisions about local and systemic therapy for patients with early breast cancer (EBC). But there is only weak evidence on the diagnostic criteria and standard interpretation. The aim of this study was to assess the performance of AUS in the detection and exclusion of LN metastases.

      Methods

      In a retrospective single-center study, 611 consecutive EBC patients with 622 axillae underwent AUS +/- core needle biopsy (CNB) plus axillary surgery, i.e. sentinel lymph node biopsy and/or axillary lymph node dissection. For all patients, AUS image documentation of at least the most suspicious LN was saved during the initial diagnostic work-up. The diagnostic outcome measures were sensitivity, specificity, accuracy, Youden-index (YI), and diagnostic odds ratio (DOR) on the basis of the daily routine interpretation and on the basis of previously recommended diagnostic criteria by two blinded examiners.

      Results

      On the basis of the daily routine interpretation, AUS had a sensitivity (95 % CI) of 53.3 % (46.4–60.1), a specificity (95 % CI) of 93.6 % (90.8–95.8), an accuracy (95 % CI) of 79.7 % (76.4–82.8), a YI (95 % CI) of 0.47 (0.40 – 0.54), and a DOR (95 % CI) of 16.75 (10.37–27.05). Systematic application of previously recommended diagnostic criteria did not improve the diagnostic accuracy of routinely interpreted AUS.

      Conclusion

      AUS performance alone is not sufficient to accurately identify or exclude axillary metastatic disease in unselected patients with EBC.

      Abbreviations:

      ACR (American College of Radiology), ALND (axillary lymph node), AUS (axillary ultrasound), BI-RADS (Breast Imaging Reporting and Data System), CI (confidence interval), CNB (core needle biopsy), DOR (diagnostic odds ratio), EBC (early breast cancer), ER (estrogen receptor), LN (lymph node), NACT (neoadjuvant chemotherapy), NPV (negative predictive value), PPV (positive predictive value), PR (progesterone receptor), SLND (sentinel lymph node dissection), STARD (Standards for Reporting of Diagnostic Accuracy Studies), TAD (targeted axillary dissection), TN (triple negative), YI (Youden-index)

      Keywords

      1. Introduction

      Axillary ultrasound (AUS) is routinely used as part of the pretherapeutic clinical evaluation of patients with early breast cancer (EBC) in most major European breast cancer units (BCU) and is recommended by recent guidelines [
      • Wockel A.
      • Festl J.
      • Stuber T.
      • Brust K.
      • Krockenberger M.
      • Heuschmann P.U.
      • Jiru-Hillmann S.
      • Albert U.S.
      • Budach W.
      • Follmann M.
      • Janni W.
      • Kopp I.
      • Kreienberg R.
      • Kuhn T.
      • Langer T.
      • Nothacker M.
      • Scharl A.
      • Schreer I.
      • Link H.
      • Engel J.
      • Fehm T.
      • Weis J.
      • Welt A.
      • Steckelberg A.
      • Feyer P.
      • Konig K.
      • Hahne A.
      • Baumgartner T.
      • Kreipe H.H.
      • Knoefel W.T.
      • Denkinger M.
      • Brucker S.
      • Luftner D.
      • Kubisch C.
      • Gerlach C.
      • Lebeau A.
      • Siedentopf F.
      • Petersen C.
      • Bartsch H.H.
      • Schulz-Wendtland R.
      • Hahn M.
      • Hanf V.
      • Muller-Schimpfle M.
      • Henscher U.
      • Roncarati R.
      • Katalinic A.
      • Heitmann C.
      • Honegger C.
      • Paradies K.
      • Bjelic-Radisic V.
      • Degenhardt F.
      • Wenz F.
      • Rick O.
      • Holzel D.
      • Zaiss M.
      • Kemper G.
      • Budach V.
      • Denkert C.
      • Gerber B.
      • Tesch H.
      • Hirsmuller S.
      • Sinn H.P.
      • Dunst J.
      • Munstedt K.
      • Bick U.
      • Fallenberg E.
      • Tholen R.
      • Hung R.
      • Baumann F.
      • Beckmann M.W.
      • Blohmer J.
      • Fasching P.
      • Lux M.P.
      • Harbeck N.
      • Hadji P.
      • Hauner H.
      • Heywang-Kobrunner S.
      • Huober J.
      • Hubner J.
      • Jackisch C.
      • Loibl S.
      • Luck H.J.
      • von Minckwitz G.
      • Mobus V.
      • Muller V.
      • Nothlings U.
      • Schmidt M.
      • Schmutzler R.
      • Schneeweiss A.
      • Schutz F.
      • Stickeler E.
      • Thomssen C.
      • Untch M.
      • Wesselmann S.
      • Bucker A.
      • Buck A.
      • Stangl S.
      Interdisciplinary screening, diagnosis, therapy and follow-up of breast cancer. Guideline of the DGGG and the DKG (S3-Level, AWMF Registry Number 032/045OL, December 2017) - part 2 with recommendations for the therapy of primary, recurrent and advanced breast cancer.
      ,
      • Ditsch N.
      • Untch M.
      • Thill M.
      • Muller V.
      • Janni W.
      • Albert U.S.
      • Bauerfeind I.
      • Blohmer J.
      • Budach W.
      • Dall P.
      • Diel I.
      • Fasching P.A.
      • Fehm T.
      • Friedrich M.
      • Gerber B.
      • Hanf V.
      • Harbeck N.
      • Huober J.
      • Jackisch C.
      • Kolberg-Liedtke C.
      • Kreipe H.H.
      • Krug D.
      • Kuhn T.
      • Kummel S.
      • Loibl S.
      • Luftner D.
      • Lux M.P.
      • Maass N.
      • Mobus V.
      • Muller-Schimpfle M.
      • Mundhenke C.
      • Nitz U.
      • Rhiem K.
      • Rody A.
      • Schmidt M.
      • Schneeweiss A.
      • Schutz F.
      • Sinn H.P.
      • Solbach C.
      • Solomayer E.F.
      • Stickeler E.
      • Thomssen C.
      • Wenz F.
      • Witzel I.
      • Wockel A.
      AGO Recommendations for the Diagnosis and Treatment of Patients with Early Breast Cancer: Update 2019.
      ,
      • Charalampoudis P.
      • Markopoulos C.
      • Kovacs T.
      Controversies and recommendations regarding sentinel lymph node biopsy in primary breast cancer: a comprehensive review of current data.
      ]. It plays an important role in the determination of the pretherapeutic axillary lymph node (LN) status, which is an important prognostic factor for disease recurrence and overall survival and which aids in the selection of therapy regimens for invasive breast cancer. Furthermore, new staging approaches in the context of neoadjuvant systemic treatment, such as targeted axillary dissection (TAD) and axillary marking, aim to identify patients not requiring axillary lymph node dissection. These new staging approaches require pretherapeutic identification of suspicious lymph nodes via AUS and their re-evaluation after neoadjuvant treatment [
      • Caudle A.S.
      • Yang W.T.
      • Krishnamurthy S.
      • Mittendorf E.A.
      • Black D.M.
      • Gilcrease M.Z.
      • Bedrosian I.
      • Hobbs B.P.
      • DeSnyder S.M.
      • Hwang R.F.
      • Adrada B.E.
      • Shaitelman S.F.
      • Chavez-MacGregor M.
      • Smith B.D.
      • Candelaria R.P.
      • Babiera G.V.
      • Dogan B.E.
      • Santiago L.
      • Hunt K.K.
      • Kuerer H.M.
      Improved axillary evaluation following neoadjuvant therapy for patients with node-positive breast cancer using selective evaluation of clipped nodes: implementation of targeted axillary dissection.
      ,
      • Donker M.
      • Straver M.E.
      • Wesseling J.
      • Loo C.E.
      • Schot M.
      • Drukker C.A.
      • van Tinteren H.
      • Sonke G.S.
      • Rutgers E.J.
      • Vrancken Peeters M.J.
      Marking axillary lymph nodes with radioactive iodine seeds for axillary staging after neoadjuvant systemic treatment in breast cancer patients: the MARI procedure.
      ]. ACOSOG Z0011, a prospective randomized trial, showed that in cases with limited metastasis in the SLN, ALND can be omitted with equal overall survival for patients receiving breast conservative treatment (including radiotherapy and adjuvant systemic treatment) [
      • Giuliano A.E.
      • Ballman K.
      • McCall L.
      • Beitsch P.
      • Whitworth P.W.
      • Blumencranz P.
      • Leitch A.M.
      • Saha S.
      • Morrow M.
      • Hunt K.K.
      Locoregional recurrence after sentinel lymph node dissection with or without axillary dissection in patients with sentinel lymph node metastases: long-term follow-up from the American College of Surgeons Oncology Group (Alliance) ACOSOG Z0011 randomized trial.
      ].
      Also in the era since the ACOSOG Z0011 trial, AUS also might gain more importance in the pretherapeutic quantification of axillary nodal burden and in the selection of eligible patients that do not benefit from axillary surgery [
      • Farrell T.P.
      • Adams N.C.
      • Stenson M.
      • Carroll P.A.
      • Griffin M.
      • Connolly E.M.
      • O’Keeffe S.A.
      The Z0011 Trial: Is this the end of axillary ultrasound in the pre-operative assessment of breast cancer patients?.
      ,
      • Ahmed M.
      • Jozsa F.
      • Baker R.
      • Rubio I.T.
      • Benson J.
      • Douek M.
      Meta-analysis of tumour burden in pre-operative axillary ultrasound positive and negative breast cancer patients.
      ,
      • Hieken T.J.
      • Trull B.C.
      • Boughey J.C.
      • Jones K.N.
      • Reynolds C.A.
      • Shah S.S.
      • Glazebrook K.N.
      Preoperative axillary imaging with percutaneous lymph node biopsy is valuable in the contemporary management of patients with breast cancer.
      ].
      Despite the widespread use of preoperative AUS in the clinical routine, criteria for LN positivity have been identified but not yet standardized, due to weak evidence [
      • Ecanow J.S.
      • Abe H.
      • Newstead G.M.
      • Ecanow D.B.
      • Jeske J.M.
      Axillary staging of breast cancer: what the radiologist should know.
      ]. Broadly used classification systems, such as the BI-RADS atlas [
      • D’Orsi CJ S.E.
      • Mendelson E.B.
      • Morris E.A.
      • et al.
      ACR BI-RADS® Atlas, Breast Imaging Reporting and Data System.
      ], propose only a small range of criteria for discriminating between benign and metastatic axillary LNs [
      • Spak D.A.
      • Plaxco J.S.
      • Santiago L.
      • Dryden M.J.
      • Dogan B.E.
      BI-RADS(®) fifth edition: a summary of changes.
      ]. Therefore, it is difficult to draw meaningful conclusions from the literature on the diagnostic accuracy of AUS, as different studies have assessed the sensitivity and specificity of this diagnostic procedure under different conditions and as the examinations have been conducted by observers with varying levels of experience. As a result, different studies examining AUS have found sensitivity values varying from 26 % to 94 %, and specificity values ranging between 53 % and 98 % [
      • Alvarez S.
      • Añorbe E.
      • Alcorta P.
      • López F.
      • Alonso I.
      • Cortés J.
      Role of sonography in the diagnosis of axillary lymph node metastases in breast cancer: a systematic review.
      ,
      • Deurloo E.E.
      • Tanis P.J.
      • Gilhuijs K.G.
      • Muller S.H.
      • Kröger R.
      • Peterse J.L.
      • Rutgers E.J.
      • Valdés Olmos R.
      • Schultze Kool L.J.
      Reduction in the number of sentinel lymph node procedures by preoperative ultrasonography of the axilla in breast cancer.
      ,
      • Esen G.
      • Gurses B.
      • Yilmaz M.H.
      • Ilvan S.
      • Ulus S.
      • Celik V.
      • Farahmand M.
      • Calay O.O.
      Gray scale and power Doppler US in the preoperative evaluation of axillary metastases in breast cancer patients with no palpable lymph nodes.
      ,
      • Lee B.
      • Lim A.K.
      • Krell J.
      • Satchithananda K.
      • Coombes R.C.
      • Lewis J.S.
      • Stebbing J.
      The efficacy of axillary ultrasound in the detection of nodal metastasis in breast cancer.
      ]. While benign LN are typically described as kidney-shaped with a thin hypoechogenic cortex and a central fatty, hyperechogenic hilum (Fig. 1) [
      • Rahbar H.
      • Partridge S.C.
      • Javid S.H.
      • Lehman C.D.
      Imaging axillary lymph nodes in patients with newly diagnosed breast cancer.
      ], multiple sonographic criteria for LN positivity have been identified in the literature. Cortical hypertrophy ≥3 mm, an eccentric cortical hypertrophy ≥3 mm, a round shape, and the loss of central hilum have been shown to generate the highest sensitivity and specificity in the detection and exclusion of LN metastasis [
      • Alvarez S.
      • Añorbe E.
      • Alcorta P.
      • López F.
      • Alonso I.
      • Cortés J.
      Role of sonography in the diagnosis of axillary lymph node metastases in breast cancer: a systematic review.
      ,
      • Deurloo E.E.
      • Tanis P.J.
      • Gilhuijs K.G.
      • Muller S.H.
      • Kröger R.
      • Peterse J.L.
      • Rutgers E.J.
      • Valdés Olmos R.
      • Schultze Kool L.J.
      Reduction in the number of sentinel lymph node procedures by preoperative ultrasonography of the axilla in breast cancer.
      ,
      • Esen G.
      • Gurses B.
      • Yilmaz M.H.
      • Ilvan S.
      • Ulus S.
      • Celik V.
      • Farahmand M.
      • Calay O.O.
      Gray scale and power Doppler US in the preoperative evaluation of axillary metastases in breast cancer patients with no palpable lymph nodes.
      ,
      • Lee B.
      • Lim A.K.
      • Krell J.
      • Satchithananda K.
      • Coombes R.C.
      • Lewis J.S.
      • Stebbing J.
      The efficacy of axillary ultrasound in the detection of nodal metastasis in breast cancer.
      ,
      • Mainiero M.B.
      • Cinelli C.M.
      • Koelliker S.L.
      • Graves T.A.
      • Chung M.A.
      Axillary ultrasound and fine-needle aspiration in the preoperative evaluation of the breast cancer patient: an algorithm based on tumor size and lymph node appearance.
      ].
      Fig. 1
      Fig. 1Example of a sonographically benign, kidney-shaped lymph node with a thin cortex and a central hyperechogenic hilum. Longitudinal diameter 10.8 mm; Cortex diameter 1.7 mm. Schematic representation shown on the right.
      The purpose of this study was to assess the performance of AUS in routine clinical practice in a cohort of patients with EBC and to compare it to the performance of the above-mentioned standardized criteria for LN positivity. We hypothesized that the use of systematic criteria would improve the performance of AUS. Our aim was to develop a set of easy-to-use, standardized criteria that would improve the performance of AUS diagnostics in the clinical routine.

      2. Methods

      2.1 Study design and study cohort

      In total, 611 patients diagnosed with invasive EBC between January 2015 and December 2016 at our institution were included in this retrospective study. Bilateral EBC was diagnosed in 11 patients. Fig. 2 shows the flow of patients through the study, along with the primary outcome as well as exclusion criteria.
      Fig. 2
      Fig. 2STARD study diagram presenting study sample selection (AUS: axillary ultrasound).
      Data from the medical, diagnostic imaging, and surgery reports were retrospectively collected and analyzed. This diagnostic accuracy study was conducted and reported according to the STARD (Standards for Reporting of Diagnostic Accuracy Studies) 2015 guidelines [
      • Bossuyt P.M.
      • Reitsma J.B.
      • Bruns D.E.
      • Gatsonis C.A.
      • Glasziou P.P.
      • Irwig L.
      • Lijmer J.G.
      • Moher D.
      • Rennie D.
      • de Vet H.C.
      • Kressel H.Y.
      • Rifai N.
      • Golub R.M.
      • Altman D.G.
      • Hooft L.
      • Korevaar D.A.
      • Cohen J.F.
      STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies.
      ].

      2.2 Imaging and minimal invasive biopsy assessment

      The patients were initial evaluated by clinical examination, mammography, and breast ultrasound. The diagnosis of EBC was then confirmed by core cut or vacuum-assisted biopsy of the primary tumor according to current guidelines [
      • Ditsch N.
      • Untch M.
      • Thill M.
      • Muller V.
      • Janni W.
      • Albert U.S.
      • Bauerfeind I.
      • Blohmer J.
      • Budach W.
      • Dall P.
      • Diel I.
      • Fasching P.A.
      • Fehm T.
      • Friedrich M.
      • Gerber B.
      • Hanf V.
      • Harbeck N.
      • Huober J.
      • Jackisch C.
      • Kolberg-Liedtke C.
      • Kreipe H.H.
      • Krug D.
      • Kuhn T.
      • Kummel S.
      • Loibl S.
      • Luftner D.
      • Lux M.P.
      • Maass N.
      • Mobus V.
      • Muller-Schimpfle M.
      • Mundhenke C.
      • Nitz U.
      • Rhiem K.
      • Rody A.
      • Schmidt M.
      • Schneeweiss A.
      • Schutz F.
      • Sinn H.P.
      • Solbach C.
      • Solomayer E.F.
      • Stickeler E.
      • Thomssen C.
      • Wenz F.
      • Witzel I.
      • Wockel A.
      AGO Recommendations for the Diagnosis and Treatment of Patients with Early Breast Cancer: Update 2019.
      ,
      • Wockel A.
      • Festl J.
      • Stuber T.
      • Brust K.
      • Stangl S.
      • Heuschmann P.U.
      • Albert U.S.
      • Budach W.
      • Follmann M.
      • Janni W.
      • Kopp I.
      • Kreienberg R.
      • Kuhn T.
      • Langer T.
      • Nothacker M.
      • Scharl A.
      • Schreer I.
      • Link H.
      • Engel J.
      • Fehm T.
      • Weis J.
      • Welt A.
      • Steckelberg A.
      • Feyer P.
      • Konig K.
      • Hahne A.
      • Kreipe H.H.
      • Knoefel W.T.
      • Denkinger M.
      • Brucker S.
      • Luftner D.
      • Kubisch C.
      • Gerlach C.
      • Lebeau A.
      • Siedentopf F.
      • Petersen C.
      • Bartsch H.H.
      • Schulz-Wendtland R.
      • Hahn M.
      • Hanf V.
      • Muller-Schimpfle M.
      • Henscher U.
      • Roncarati R.
      • Katalinic A.
      • Heitmann C.
      • Honegger C.
      • Paradies K.
      • Bjelic-Radisic V.
      • Degenhardt F.
      • Wenz F.
      • Rick O.
      • Holzel D.
      • Zaiss M.
      • Kemper G.
      • Budach V.
      • Denkert C.
      • Gerber B.
      • Tesch H.
      • Hirsmuller S.
      • Sinn H.P.
      • Dunst J.
      • Munstedt K.
      • Bick U.
      • Fallenberg E.
      • Tholen R.
      • Hung R.
      • Baumann F.
      • Beckmann M.W.
      • Blohmer J.
      • Fasching P.A.
      • Lux M.P.
      • Harbeck N.
      • Hadji P.
      • Hauner H.
      • Heywang-Kobrunner S.
      • Huober J.
      • Hubner J.
      • Jackisch C.
      • Loibl S.
      • Luck H.J.
      • von Minckwitz G.
      • Mobus V.
      • Muller V.
      • Nothlings U.
      • Schmidt M.
      • Schmutzler R.
      • Schneeweiss A.
      • Schutz F.
      • Stickeler E.
      • Thomssen C.
      • Untch M.
      • Wesselmann S.
      • Bucker A.
      • Krockenberger M.
      Interdisciplinary screening, diagnosis, therapy and follow-up of breast cancer. Guideline of the DGGG and the DKG (S3-Level, AWMF registry number 032/045OL, December 2017) - part 1 with recommendations for the screening, diagnosis and therapy of breast cancer.
      ]. AUS was performed in 622 axillae by an experienced physician (>250 examinations/year) using a Siemens SonoLine Antares or Siemens Acuson S2000 (both Siemens Healthineers, Erlangen, Germany) with 13.5 MHz transducers. At least one image of the most suspicious LN (in the longitudinal direction) was digitally saved. According to institutional guidelines at the time of the study, a biopsy of a suspicious LN was performed in cases of sonographically “unclear” nodal status. At the time of this study, CNB was performed only in patients with an unclear nodal status after AUS (n = 64). Unclear and suspicious findings by ultrasound were labeled “AUS+” for the ultrasound only perspective. For the combined analysis of AUS + and core cut biopsy– patients the examination was labeled “− (negative)”.

      2.3 Reevaluation of axillary ultrasound images using standardized criteria

      All digital images of the initial AUS documentation were reviewed using the following
      criteria for LN positivity (Fig. 3):
      • -
        cortical hypertrophy ≥3 mm,
      • -
        eccentric / focal cortical hypertrophy ≥3 mm,
      • -
        round shape,
      • -
        loss of the central hyperechoic hilum (leading to a homogenous hypoechoic appearance).
      Fig. 3
      Fig. 3Suspicious axillary lymph nodes showing different sonographic criteria for lymph node positivity. Longitudinal diameter indicated in yellow, cortex diameter indicated in red. Schematic representation of the lymph node shown on the right.
      Because cortical hypertrophy could not be measured in some cases (due to full metastatic affection of the LN), a combined variable (“No Hilum”) was set up combining both variables, i.e. corresponding to either a “loss of the central hilum” or an “eccentric cortical hypertrophy ≥3 mm”, in order to be able to analyze the correlation of the various criteria with LN positivity.
      In order to assess interrater reliability, re-evaluation was carried out separately by a second experienced physician (>250 examinations/year) and by a final-year medical student after brief training. Both observers were blinded to the final pathologic results and to each other’s assessments. All cases with inter-observer discrepancy were reviewed by another experienced physician (also >250 examinations/year). All further statistical analysis was based on this definitive assessment.

      2.4 Surgical approaches to the axilla

      All patients included were treated according to current national and international guidelines on the basis of a multidisciplinary in-house tumor conference decisions [
      • Wockel A.
      • Festl J.
      • Stuber T.
      • Brust K.
      • Stangl S.
      • Heuschmann P.U.
      • Albert U.S.
      • Budach W.
      • Follmann M.
      • Janni W.
      • Kopp I.
      • Kreienberg R.
      • Kuhn T.
      • Langer T.
      • Nothacker M.
      • Scharl A.
      • Schreer I.
      • Link H.
      • Engel J.
      • Fehm T.
      • Weis J.
      • Welt A.
      • Steckelberg A.
      • Feyer P.
      • Konig K.
      • Hahne A.
      • Kreipe H.H.
      • Knoefel W.T.
      • Denkinger M.
      • Brucker S.
      • Luftner D.
      • Kubisch C.
      • Gerlach C.
      • Lebeau A.
      • Siedentopf F.
      • Petersen C.
      • Bartsch H.H.
      • Schulz-Wendtland R.
      • Hahn M.
      • Hanf V.
      • Muller-Schimpfle M.
      • Henscher U.
      • Roncarati R.
      • Katalinic A.
      • Heitmann C.
      • Honegger C.
      • Paradies K.
      • Bjelic-Radisic V.
      • Degenhardt F.
      • Wenz F.
      • Rick O.
      • Holzel D.
      • Zaiss M.
      • Kemper G.
      • Budach V.
      • Denkert C.
      • Gerber B.
      • Tesch H.
      • Hirsmuller S.
      • Sinn H.P.
      • Dunst J.
      • Munstedt K.
      • Bick U.
      • Fallenberg E.
      • Tholen R.
      • Hung R.
      • Baumann F.
      • Beckmann M.W.
      • Blohmer J.
      • Fasching P.A.
      • Lux M.P.
      • Harbeck N.
      • Hadji P.
      • Hauner H.
      • Heywang-Kobrunner S.
      • Huober J.
      • Hubner J.
      • Jackisch C.
      • Loibl S.
      • Luck H.J.
      • von Minckwitz G.
      • Mobus V.
      • Muller V.
      • Nothlings U.
      • Schmidt M.
      • Schmutzler R.
      • Schneeweiss A.
      • Schutz F.
      • Stickeler E.
      • Thomssen C.
      • Untch M.
      • Wesselmann S.
      • Bucker A.
      • Krockenberger M.
      Interdisciplinary screening, diagnosis, therapy and follow-up of breast cancer. Guideline of the DGGG and the DKG (S3-Level, AWMF registry number 032/045OL, December 2017) - part 1 with recommendations for the screening, diagnosis and therapy of breast cancer.
      ,
      • Gnant M.
      • Harbeck N.
      • Thomssen C.
      St. Gallen/Vienna 2017: a brief summary of the consensus discussion about escalation and de-escalation of primary breast cancer treatment.
      ]. Patients who had either non-suspicious LN after AUS or proof of benign LN core needle biopsy (CNB) underwent sentinel lymph node dissection (SLND). Axillary lymph node dissection (ALND) was performed after nodal disease was confirmed by minimally invasive biopsy either before or after neoadjuvant systemic treatment (NACT).

      2.5 Pathologic evaluation as reference results

      From all cases (n = 622) pathologic results from SLND (n = 460; 74.0 %) or ALND (n = 41; 6.6 %) before systemic treatment, a positive CNB (n = 10), or the pathologic diagnosis of post-NACT ALND/SLNB specimen for regressive changes (n = 111) served as the reference results to which the AUS diagnosis was correlated.
      In these cases with NACT without prior surgical axillary staging, an additional evaluation of the ALND specimen for the presence of tumor cells and regressive changes indicative of prior metastases was carried out in order to deduce initial nodal status: If an axilla was positive after NACT (ypN+), we extrapolated that it had already been positive at the time of the initial routine AUS before the start of NACT (pN+). This approach was done in n = 37 cases (5.9 %). The current literature furthermore suggests that 97 % of biopsy-proven node positive axillae (pN+) that convert to ypN0 post-NACT show signs of tumor regression such as the presence of intranodular fibrosis, a foamy histiocytic infiltrate and/or the presence of hemosiderin-laden macrophages [
      • Barrio A.V.
      • Mamtani A.
      • Edelweiss M.
      • Eaton A.
      • Stempel M.
      • Murray M.P.
      • Morrow M.
      How often is treatment effect identified in axillary nodes with a pathologic complete response after neoadjuvant chemotherapy?.
      ]. Therefore, we assumed that the patients in our cohort whose axillary lymph nodes contained no tumor cells, but showed regressive changes had been nodal-positive pre-NACT at the time of the initial AUS (pN+), while the lack of tumor cells and regressive changes was interpreted as an already initially negative axilla (pN0). This approach was followed in n = 74 cases (11.9 %).
      All LNs removed by ALND or SLNB were formalin-fixed, embedded in paraffin, sectioned, stained, and processed according to current guidelines. Immunohistochemistry was only performed in selected cases with unclear ypN0 situations. The presence of parenchymal or subcapsular macrometastases (≥2 mm), micrometastases (>0.2 mm but <2 mm), or isolated tumor cells (≤0.2 mm) constituted a positive result.

      2.6 Statistical analysis

      Descriptive statistics were used to assess the sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), diagnostic odds ratio (DOR), Youden-Index (YI), and the accuracy of AUS in routine clinical services and of the blinded, re-interpreted standardized sonographic criteria. DOR measures the effectiveness of a diagnostic test (defined as the ratio of the odds of true positive test result relative to the odds of a false negative test result). YI represents the performance of a dichotomous diagnostic test. It incorporates both sensitivity and specificity ranging from 0 (test without diagnostic statement) to 1 (test without false positives or false negatives). Accuracy is defined as the proportion of correct predictions (true positives and true negatives together) among all cases examined.
      The initial AUS diagnosis (the index test) was correlated to the final pathologic reference results (ALND, SLNB, or positive CNB). Corresponding 95 % confidence intervals (95 % CI) were calculated based on a normal approximation for binomially distributed data.
      The strength of agreement between the two independent observers was determined through an interrater reliability analysis using Cohen’s kappa (“κ”).
      Furthermore, a logistic regression was performed to evaluate the correlation of the above-mentioned standardized morphologic criteria with LN positivity on final histopathologic examination. The performance of the morphological criteria in predicting nodal metastatic involvement on AUS was then determined separately for each criterion and for different combinations of criteria and compared descriptively using the YI and DOR as parameters for their global diagnostic performance. We evaluated both approaches that used one of several possible criteria (so-called “OR” combinations) and approaches that required more than one criterion (so-called “AND” combination).

      3. Results

      3.1 Patient characteristics

      In total, 611 patients with 622 examined axillae were included. The median age was 56 at the time of diagnosis. Most carcinomas were invasive-ductal (86.0 %) and of intermediate histologic tumor grade (55.0 %). Of the 622 axillae examined, 441 belonged to patients with HR+/HER2−, 84 to HER2+, and 80 to TN tumors. Tumor subtype distribution failed for 17 axillae due to incomplete information. Table 1 shows detailed clinical and pathologic characteristics of the study cohort.
      Table 1Patient Characteristics.
      Variablen%
      Patients611
      Age (years)mean56
      SD13
      Axillae622
      Size of the primary tumor (mm)mean20
      SD12
      Invasive tumor typeinvasive-ductal53586.0
      invasive-lobular6810.9
      others193.1
      Tumor subtypeHR+/HER2-44170.9
      HER2+8413.5
      TN8012.9
      missing172.7
      Tumor stagecT129146.8
      cT223537.8
      cT3365.8
      cT4304.8
      cTis162.6
      missing142.3
      Tumor gradeG19114.6
      G234255.0
      G316726.8
      missing233.6
      cN (AUS +/- CNB)cN047876.8
      cN+14423.2
      CNB performedyes6410.3
      CNB resultCNB positive3554.7
      CNB negative2945.3
      Result of gold standardpN040865.6
      pN+21434.3
      Type of gold standardSLND46074.0
      ALND416.6
      CNB (positive)101.6
      ypN+→pN+375.9
      Retrospective reevaluation post-NACT7411.9
      Absolute values with percentages. TN: Triple-negative; HER2+: HER2-positive; HR+/HER2-: Hormone receptor-positive/HER2-negative; SLND: sentinel lymph node dissection; ALND: axillary lymph node dissection; CNB: core needle biopsy.
      On routine AUS, 482 axillae (77.5 %) were classified as sonographically node-negative (AUS−), while 140 axillae (22.5 %) were diagnosed as sonographically node-positive (AUS+). Considering the results of the CNB performed in 64 cases of unclear clinical assessment (10.3 %), 478 axillae (76.8 %) were classified as nodal-negative (cN0) and 144 axillae (23.2 %) were classified as nodal-positive (cN+). The core cut biopsy, SLNB, or ALND (i.e. the reference result) was negative in 408 axillae (65.6 %) and positive for the remaining 214 axillae (34.3 %) (Table 1).

      3.2 Performance of axillary ultrasound in routine clinical services

      Table 2 shows that the initial AUS diagnosis had a sensitivity (95 % CI) of 53.3 % (46.4–60.1) and a specificity (95 % CI) of 93.7 % (90.8–95.8), resulting in a YI (95 % CI) of 0.47 (0.40 – 0.54) and a DOR (95 % CI) of 16.75 (10.37–27.05).
      Table 2Performance characteristics of axillary ultrasound (AUS), core needle biopsy (CNB), and their combination (AUS +/− CNB) in cases with unclear assessment.
      AUS (n = 622)CNB (n = 64)AUS +/- CNB (n = 622)
      Prevalence (%)34.7 (30.9–38.6)79.5 (55.9–79.8)34.7 (30.9–38.6)
      Sensitivity (%)53.3 (46.4–60.1)79.6 (64.7–90.2)53.6 (46.6–60.4)
      Specificity (%)93.6 (90.8–95.8)100.0 (83.2–100.0)95.5 (92.9–97.3)
      PPV (%)81.4 (74.0–87.5)100.0 (90.0–100.0)86.3 (79.2–91.6)
      NPV (%)79.3 (75.4–82.3)69.0 (49.2–84.7)79.5 (75.6–83.0)
      LR+8.54 (5.65–12.38)33.1 (2.1–514.6)11.84 (7.41–18.92)
      LR−0.50 (0.43 – 0.58)20.5 (11.4–36.6)0.48 (0.42 – 0.56)
      DOR16.75 (10.37–27.05)153.20 (8.50–2771.5)24.34 (14.12–41.97)
      Youden-Index0.47 (0.40 – 0.54)0.80 (0.68 – 0.92)0.49 (0.42 – 0.56)
      Accuracy (%)79.7 (76.4–82.8)85.9 (75.0–93.4)81.0 (77.6–84.0)
      Error rate (%)20.3 (17.1–23.6)14.1 (6.6–25.0)19.0 (16.0–22.4)
      Absolute values or percentages with corresponding 95 % confidence intervals indicated in parentheses. PPV: positive predictive value; NPV: negative predictive value; LR+: positive likelihood ratio; LR−: negative likelihood ratio; DOR: diagnostic odds ratio; Youden-Index: sensitivity + specificity – 1.
      CNB alone had a sensitivity (95 % CI) of 79.6 % (64.7–90.2) and a specificity (95 % CI) of 100.0 % (83.2–100.0). Adding CNB to the AUS examination in these patients resulted in a higher specificity (95.5 % vs. 93.6 %), PPV (86.3 % vs. 81.4 %), YI (0.49 vs. 0.47) and DOR (24.34 vs. 16.75) compared to the diagnosis based on AUS only.

      3.3 Performance of pre-defined criteria in predicting lymph node metastasis on axillary ultrasound

      LNs with a cortical thickness of ≥3 mm (62.6 % vs. 28.7 %), round shape (23.4 % vs. 3.4 %), eccentric cortical hypertrophy ≥3 mm (38.7 % vs. 21.2 %), or loss of the central hilum (33.8 % vs. 2.9 %) were found significantly more frequently in malignant than in benign axillae (p < 0.001) (Fig. 4). The criterion “cortical hypertrophy ≥3 mm” was observed most frequently, both in malignant as well as in benign axillae.
      Fig. 4
      Fig. 4Distribution of morphological criteria for lymph node positivity in benign vs. malignant axillae. The examined sonographic criteria are represented on the x-axis, while the percentage of benign and malignant axillae in which the corresponding criterion was found is shown on the y-axis. The 4 comparisons between the benign axillae and the malignant axillae were all statistically significant at p < 0.001.
      The logistic regression model showed that the criteria “round shape” (OR: 4.43, 95 % CI: 2.29–8.55, p < .001) and “No Hilum” (OR: 2.59, 95 % CI: 1.09–6.15), p = 0.03) predicted the presence of LN metastasis on final histopathologic examination. On the other hand, the criterion “cortical hypertrophy ≥3 mm” (OR: 1.38, 95 % CI: (0.60–3.22), p = 0.45) was not significantly correlated with LN positivity (Table 3).
      Table 3Odds ratio of defined sonographic criteria as predictors of lymph node positivity on final pathological examination and their interrater reliability in the assessment of 622 axillary lymph nodes.
      Sonographic criterionOdds Ratio (95 % CI)PNo. (%) of mismatchesCohen’s kappa (95 % CI)
      Cortical hypertrophy ≥3 mm1.38 (0.60–3.22)0.45352 (8.4)0.83 (0.78 – 0.87)
      Eccentric cortical hypertrophy ≥ 3 mmaaa111 (20.7)0.35 (0.26 – 0.44)
      Round shape4.43 (2.29–8.55)<0.00150 (8.0)0.57 (0.47 – 0.68)
      “No Hilum”2.59 (1.09–6.15)0.03012 (1.9)0.92 (0.87 – 0.96)
      Absolute values with percentages (%) or the 95 % confidence intervals (95 % CI) given in parentheses. The p-value is indicated in the rightmost column. “No Hilum”: Loss of central hilum OR eccentric cortical hypertrophy ≥3 mm.
      aNot applicable if” Loss of hilum” was present, therefore not included in the multivariate logistic regression analysis and mismatches and interrater reliability only evaluated in n = 535 axillae.
      There was a high level of agreement between the two observers’ assignment of the criteria “loss of central hyperechoic hilum” (κ=0.92, 95 % CI: (0.87 – 0.96) and “cortical hypertrophy” (κ = 0.83, 95 % CI: (0.78 – 0.87). For the criterion “round shape”, we found moderate agreement between the two observers (κ: 0.57, 95 % CI: 0.47 – 0.68 (Table 3)).
      Upon separate analysis of the performance characteristics of the criteria listed in Table 4, the criterion “cortical hypertrophy ≥3 mm” showed the highest sensitivity in the detection of axillary lymph node metastasis on AUS. At the same time, a cortex diameter <3 mm had the lowest specificity for excluding axillary metastasis. On the other hand, the ultrasound criterion “round shape” had the highest specificity but a very low sensitivity. The combined criterion “No Hilum” had the highest YI of all examined criteria. Table 4 also shows that only the “OR” conjunctions of the various criteria attained sensitivities over 50 %. The combination of the criteria that significantly predicted LN positivity in the logistic regression (“round shape” OR “No Hilum”) showed a higher specificity (75.7 % vs. 70.3 %) and a lower sensitivity (60.7 % vs. 64.0 %) than the “OR” conjunction of all criteria examined (“cortical hypertrophy ≥ 3 mm” OR “round shape” OR “No Hilum”).
      Table 4Performance of various sonographic criteria and their combinations for identifying lymph node positivity.
      Cortical hypertrophy ≥3 mm (n = 622)Round shape (n = 622)“No Hilum” (n = 622)OR conjunction of all criteria: cortical hypertrophy OR round shape OR “No Hilum” (n = 622)AND conjunction of all criteria: cortical hypertrophy OR round shape AND “No Hilum” (n = 622)OR conjunction of significant criteria: round shape OR “No Hilum “(n = 622)AND conjunction of significant criteria: round shape AND “No Hilum” (n = 622)
      Sensitivity (%)62.6 (55.8–69.1)23.4 (17.9–29.6)59.3 (52.4–66.0)64.0 (57.2–70.4)22.0 (16.6–28.1)60.7 (53.9–67.3)22.0 (16.6–28.1)
      Specificity (%)71.3 (66.7–75.7)96.6 (94.3–98.1)76.5 (72.0–80.5)70.3 (65.7–74.7)97.5 (95.5–98.8)75.7 (71.3–79.8)97.3 (95.2–98.6)
      PPV (%)53.4 (47.0–59.7)78.1 (66.0–87.5)56.9 (50.2–63.5)53.1 (46.8–59.3)82.4 (70.1–91.3)56.8 (50.1–63.3)81.0 (68.6–90.1)
      NPV (%)78.4 (73.9–82.5)70.6 (66.6–74.4)78.2 (73.8–82.2)78.8 (74.3–82.9)70.4 (66.5–74.2)78.6 (74.2–82.6)70.4 (66.4–74.1)
      LR+2.18 (1.82–2.63)6.81 (3.85–12.03)2.52 (2.05–31.0)2.16 (1.80–2.58)8.96 (4.62–17.38)2.50 (2.04–3.07)8.15 (4.32–15.38)
      LR−0.52 (0.44 – 0.63)0.79 (0.74 – 0.86)0.53 (44.8 – 0.63)0.51 (0.42 – 0.62)0.80 (0.74 – 0.86)0.52 (0.43 – 0.62)0.80 (0.75 – 0.86)
      DOR4.17 (2.93–5.91)8.58 (4.62–15.95)4.74 (3.32–6.77)4.22 (2.97–5.99)11.20 (5.53–22.69)4.83 (3.38–6.89)10.16 (5.14–20.07)
      Youden-Index0.34 (0.26 – 0.42)0.20 (0.14 – 0.26)0.36 (0.28 – 0.45)0.34 (0.28 – 0.42)0.20 (0.14 – 0.25)0.37 (0.29 – 0.44)0.19 (0.14 – 0.25)
      Accuracy (%)68.3 (64.5–72.0)71.4 (67.7–74.9)70.6 (66.8–74.1)68.2 (64.3–71.8)71.5 (67.8–75.1)70.6 (66.8–74.1)71.4 (67.7–74.9)
      Error rate (%)31.7 (28.0–35.5)28.6 (25.1–32.3)29.4 (25.9–33.2)31.8 (28.2–35.7)28.5 (24.9–32.2)29.4 (25.9–33.2)28.6 (25.1–32.4)
      Absolute values or percentages with corresponding 95 % confidence intervals indicated in parentheses. “No Hilum”: loss of central hilum OR eccentric cortical hypertrophy ≥ 3 mm; PPV: positive predictive value; NPV: negative predictive value; LR+: positive likelihood ratio; LR−: negative likelihood ratio; DOR: diagnostic odds ratio; Youden-Index: sensitivity + specificity – 1.

      4. Discussion

      In this study, we aimed to assess the accuracy of AUS in routine clinical practice and to compare it to the performance of standardized sonographic criteria for LN positivity. The study cohort constituted a representative group of patients with EBC, with the prevalence rate for LN tumor metastasis of 34.7 % corresponding to earlier diagnostic studies on AUS [
      • Mainiero M.B.
      • Cinelli C.M.
      • Koelliker S.L.
      • Graves T.A.
      • Chung M.A.
      Axillary ultrasound and fine-needle aspiration in the preoperative evaluation of the breast cancer patient: an algorithm based on tumor size and lymph node appearance.
      ,
      • Gipponi M.
      • Fregatti P.
      • Garlaschi A.
      • Murelli F.
      • Margarino C.
      • Depaoli F.
      • Baccini P.
      • Gallo M.
      • Friedman D.
      Axillary ultrasound and Fine-Needle Aspiration Cytology in the preoperative staging of axillary node metastasis in breast cancer patients.
      ].
      We found that even at a specialized breast cancer unit, the sensitivity of AUS was only 53.3 % with a specificity of 93.6 %. In the literature, AUS sensitivity and specificity vary from 26 % to 94 % and from 53 % to 98 % respectively [
      • Alvarez S.
      • Añorbe E.
      • Alcorta P.
      • López F.
      • Alonso I.
      • Cortés J.
      Role of sonography in the diagnosis of axillary lymph node metastases in breast cancer: a systematic review.
      ,
      • Deurloo E.E.
      • Tanis P.J.
      • Gilhuijs K.G.
      • Muller S.H.
      • Kröger R.
      • Peterse J.L.
      • Rutgers E.J.
      • Valdés Olmos R.
      • Schultze Kool L.J.
      Reduction in the number of sentinel lymph node procedures by preoperative ultrasonography of the axilla in breast cancer.
      ,
      • Esen G.
      • Gurses B.
      • Yilmaz M.H.
      • Ilvan S.
      • Ulus S.
      • Celik V.
      • Farahmand M.
      • Calay O.O.
      Gray scale and power Doppler US in the preoperative evaluation of axillary metastases in breast cancer patients with no palpable lymph nodes.
      ,
      • Lee B.
      • Lim A.K.
      • Krell J.
      • Satchithananda K.
      • Coombes R.C.
      • Lewis J.S.
      • Stebbing J.
      The efficacy of axillary ultrasound in the detection of nodal metastasis in breast cancer.
      ]. Although the specificity has risen to 100 % in some of these studies when CNB was routinely added to evaluate all sonographically suspicious LN [
      • Gipponi M.
      • Fregatti P.
      • Garlaschi A.
      • Murelli F.
      • Margarino C.
      • Depaoli F.
      • Baccini P.
      • Gallo M.
      • Friedman D.
      Axillary ultrasound and Fine-Needle Aspiration Cytology in the preoperative staging of axillary node metastasis in breast cancer patients.
      ,
      • Mills P.
      • Sever A.
      • Weeks J.
      • Fish D.
      • Jones S.
      • Jones P.
      Axillary ultrasound assessment in primary breast cancer: an audit of 653 cases.
      ,
      • Solon J.G.
      • Power C.
      • Al-Azawi D.
      • Duke D.
      • Hill A.D.
      Ultrasound-guided core biopsy: an effective method of detecting axillary nodal metastases.
      ], this addition of CNB would not improve the sensitivity, as it only concerns AUS + patients.
      Some authors suggest that the experience levels of the examiners and the use of different criteria for the sonographic evaluation of axillary LNs might be responsible for this wide variability of AUS performance [
      • Alvarez S.
      • Añorbe E.
      • Alcorta P.
      • López F.
      • Alonso I.
      • Cortés J.
      Role of sonography in the diagnosis of axillary lymph node metastases in breast cancer: a systematic review.
      ]. We found that the sonographic criteria examined were easy to apply and had high interrater consistency; they might therefore have the potential to standardize the interpretation of AUS examinations. Contrary to earlier studies, we found that the combination of criteria did not have an acceptable overall diagnostic performance. In this regard, the observed opposing values of sensitivity and specificity were striking. The criteria with a high sensitivity were consistently associated with low specificities and vice versa, while the global diagnostic performance of AUS (that was quantified by YI, DOR, and accuracy) could not be substantially improved. Consequently, a consensus on the prioritization of either sensitivity or specificity seems to be a prerequisite for further improvement of AUS in the clinical evaluation of axillar lymph nodes.
      To our knowledge this study is the largest report of consecutive EBC patients that received a pretherapeutic AUS examination and the surgical therapy with pathologic reference results at the same BCU. Nonetheless, this study has a limitation related to the selection bias due to the retrospective design. In total, n = 220 otherwise eligible patients could not be included in the study, because as sonographic images of the most suspicious lymph nodes had not been saved for documentation purpose at the time of the initial AUS examination. One limitation of this study is the retrospective study design. The reevaluation of the lymph nodes could only be performed on existing, two-dimensional image documentation from the primary examination. If the documentation was suboptimal, the retrospective evaluation is problematic and normal lymph node can appear pathologic or vice-versa. Furthermore, the 2D images saved did not allow the dynamic examination of a three-dimensional impression that normally constitutes one of the major advantages of ultrasound; that shortcoming could have led to a worse performance of the standardized criteria.
      One limitation in the study design is that in cases with NACT without prior surgical axillary staging, the initial nodal status was deduced from the pathological evaluation of the post-NACT nodal status for the presence of tumor cells and regressive changes. This approach is based on the observation that if an initially node positive axilla (pN+) converts to ypN0, the post-NACT lymph node specimen shows signs of tumor regression such as fibrosis, necrosis or invasion of inflammatory cells, as described by Sataloff et al [
      • Sataloff D.M.
      • Mason B.A.
      • Prestipino A.J.
      • Seinige U.L.
      • Lieber C.P.
      • Baloch Z.
      Pathologic response to induction chemotherapy in locally advanced carcinoma of the breast: a determinant of outcome.
      ]. The Sataloff criteria are widely used and have shown to be valid indicators of treatment response [
      • Newman L.A.
      • Pernick N.L.
      • Adsay V.
      • Carolin K.A.
      • Philip P.A.
      • Sipierski S.
      • Bouwman D.L.
      • Kosir M.A.
      • White M.
      • Visscher D.W.
      Histopathologic evidence of tumor regression in the axillary lymph nodes of patients treated with preoperative chemotherapy correlates with breast cancer outcome.
      ,
      • Glaeser A.
      • Sinn H.P.
      • Garcia-Etienne C.
      • Riedel F.
      • Hug S.
      • Schaefgen B.
      • Golatta M.
      • Hennigs A.
      • Feisst M.
      • Sohn C.
      • Heil J.
      Heterogeneous responses of axillary lymph node metastases to neoadjuvant chemotherapy are common and depend on breast cancer subtype.
      ]. Although it is possible that single patients with initially positive lymph nodes convert to ypN0 without any signs of regressions, leading to potential misclassification in our study as initially nodal negative, this would be a very rare event which is not described in the literature.
      Therefore, we used the information from the Sataloff criteria, which are routinely applied at our institution, as a pragmatic approach in this analysis to deduce nodal stage in cases without pre-NACT histological confirmation.
      In order to avoid overtreatment (e.g. ALND in pN0 patients), we recommend to confirm every cN + status by core cut biopsy if technically feasible. This will improve specificity to up to 100 %. On the other hand, we argue that even today sensitivity matters although there is strong evidence that ALND even in low volume pN+ (sn) does not improve oncologic outcomes. Patients with sonographically and minimally invasive confirmed lymph node metastasis suffer from poor prognosis and have different breast cancer compared to those with positive sentinel(s) after negative AUS [
      • Tucker N.S.
      • Cyr A.E.
      • Ademuyiwa F.O.
      • Tabchy A.
      • George K.
      • Sharma P.K.
      • Jin L.X.
      • Sanati S.
      • Aft R.
      • Gao F.
      • Margenthaler J.A.
      • Gillanders W.E.
      Axillary ultrasound accurately excludes clinically significant lymph node disease in patients with early stage breast cancer.
      ,
      • Shigematsu H.
      • Nishina M.
      • Yasui D.
      • Hirata T.
      • Ozaki S.
      Minimal prognostic significance of sentinel lymph node metastasis in patients with cT1-2 and cN0 breast cancer.
      ]. Even in those patients a more targeted axillary surgery might be possible [
      • Henke G.
      • Knauer M.
      • Ribi K.
      • Hayoz S.
      • Gérard M.A.
      • Ruhstaller T.
      • Zwahlen D.R.
      • Muenst S.
      • Ackerknecht M.
      • Hawle H.
      • Fitzal F.
      • Gnant M.
      • Mátrai Z.
      • Ballardini B.
      • Gyr A.
      • Kurzeder C.
      • Weber W.P.
      Tailored axillary surgery with or without axillary lymph node dissection followed by radiotherapy in patients with clinically node-positive breast cancer (TAXIS): study protocol for a multicenter, randomized phase-III trial.
      ].
      To increase the diagnostic utility of AUS, a prioritization of either sensitivity or specificity (or a combination with additional diagnostic procedures like elastography and / or core cut biopsy) seems to be crucial. Future studies might evaluate additional techniques to improve the diagnosis of axillary lymph node status.

      5. Conclusion

      The performance of AUS alone is currently not as accurate to identify or exclude axillary metastatic disease as assumed. Core cut biopsy to confirm nodal disease should be routinely performed to at least improve specificity.

      Funding

      None.

      Ethical approval

      The study was approved by the ethics committee in charge under reference number S-241−2017 and was conducted in accordance with the Declaration of Helsinki.

      Informed consent

      Not applicable.

      CRediT authorship contribution statement

      Fabian Riedel: Conceptualization, Writing - original draft. Benedikt Schaefgen: Conceptualization, Writing - original draft. Hans-Peter Sinn: Writing - review & editing, Resources. Manuel Feisst: Data curation, Writing - review & editing. André Hennigs: Resources, Writing - review & editing. Sarah Hug: Resources, Writing - review & editing. Angela Binnig: Resources, Writing - review & editing. Christina Gomez: Resources, Writing - review & editing. Aba Harcos: Resources, Writing - review & editing. Anne Stieber: Resources, Writing - review & editing. Hans-Ulrich Kauczor: Supervision. Christof Sohn: Supervision. Michael Golatta: Resources, Writing - review & editing. Antonia Glaeser: Conceptualization, Project administration, Writing - original draft, Supervision. Joerg Heil: Conceptualization, Project administration, Writing - original draft, Supervision.

      Declaration of Competing Interest

      The authors report no declarations of interest.

      Acknowledgements

      The authors would like to thank Michael Hanna, PhD, (Mercury Medical Research & Writing) for proof-reading the manuscript.

      References

        • Wockel A.
        • Festl J.
        • Stuber T.
        • Brust K.
        • Krockenberger M.
        • Heuschmann P.U.
        • Jiru-Hillmann S.
        • Albert U.S.
        • Budach W.
        • Follmann M.
        • Janni W.
        • Kopp I.
        • Kreienberg R.
        • Kuhn T.
        • Langer T.
        • Nothacker M.
        • Scharl A.
        • Schreer I.
        • Link H.
        • Engel J.
        • Fehm T.
        • Weis J.
        • Welt A.
        • Steckelberg A.
        • Feyer P.
        • Konig K.
        • Hahne A.
        • Baumgartner T.
        • Kreipe H.H.
        • Knoefel W.T.
        • Denkinger M.
        • Brucker S.
        • Luftner D.
        • Kubisch C.
        • Gerlach C.
        • Lebeau A.
        • Siedentopf F.
        • Petersen C.
        • Bartsch H.H.
        • Schulz-Wendtland R.
        • Hahn M.
        • Hanf V.
        • Muller-Schimpfle M.
        • Henscher U.
        • Roncarati R.
        • Katalinic A.
        • Heitmann C.
        • Honegger C.
        • Paradies K.
        • Bjelic-Radisic V.
        • Degenhardt F.
        • Wenz F.
        • Rick O.
        • Holzel D.
        • Zaiss M.
        • Kemper G.
        • Budach V.
        • Denkert C.
        • Gerber B.
        • Tesch H.
        • Hirsmuller S.
        • Sinn H.P.
        • Dunst J.
        • Munstedt K.
        • Bick U.
        • Fallenberg E.
        • Tholen R.
        • Hung R.
        • Baumann F.
        • Beckmann M.W.
        • Blohmer J.
        • Fasching P.
        • Lux M.P.
        • Harbeck N.
        • Hadji P.
        • Hauner H.
        • Heywang-Kobrunner S.
        • Huober J.
        • Hubner J.
        • Jackisch C.
        • Loibl S.
        • Luck H.J.
        • von Minckwitz G.
        • Mobus V.
        • Muller V.
        • Nothlings U.
        • Schmidt M.
        • Schmutzler R.
        • Schneeweiss A.
        • Schutz F.
        • Stickeler E.
        • Thomssen C.
        • Untch M.
        • Wesselmann S.
        • Bucker A.
        • Buck A.
        • Stangl S.
        Interdisciplinary screening, diagnosis, therapy and follow-up of breast cancer. Guideline of the DGGG and the DKG (S3-Level, AWMF Registry Number 032/045OL, December 2017) - part 2 with recommendations for the therapy of primary, recurrent and advanced breast cancer.
        Geburtshilfe Frauenheilkd. 2018; 78: 1056-1088
        • Ditsch N.
        • Untch M.
        • Thill M.
        • Muller V.
        • Janni W.
        • Albert U.S.
        • Bauerfeind I.
        • Blohmer J.
        • Budach W.
        • Dall P.
        • Diel I.
        • Fasching P.A.
        • Fehm T.
        • Friedrich M.
        • Gerber B.
        • Hanf V.
        • Harbeck N.
        • Huober J.
        • Jackisch C.
        • Kolberg-Liedtke C.
        • Kreipe H.H.
        • Krug D.
        • Kuhn T.
        • Kummel S.
        • Loibl S.
        • Luftner D.
        • Lux M.P.
        • Maass N.
        • Mobus V.
        • Muller-Schimpfle M.
        • Mundhenke C.
        • Nitz U.
        • Rhiem K.
        • Rody A.
        • Schmidt M.
        • Schneeweiss A.
        • Schutz F.
        • Sinn H.P.
        • Solbach C.
        • Solomayer E.F.
        • Stickeler E.
        • Thomssen C.
        • Wenz F.
        • Witzel I.
        • Wockel A.
        AGO Recommendations for the Diagnosis and Treatment of Patients with Early Breast Cancer: Update 2019.
        Breast Care (Basel). 2019; 14: 224-245
        • Charalampoudis P.
        • Markopoulos C.
        • Kovacs T.
        Controversies and recommendations regarding sentinel lymph node biopsy in primary breast cancer: a comprehensive review of current data.
        Eur. J. Surg. Oncol. 2018; 44: 5-14
        • Caudle A.S.
        • Yang W.T.
        • Krishnamurthy S.
        • Mittendorf E.A.
        • Black D.M.
        • Gilcrease M.Z.
        • Bedrosian I.
        • Hobbs B.P.
        • DeSnyder S.M.
        • Hwang R.F.
        • Adrada B.E.
        • Shaitelman S.F.
        • Chavez-MacGregor M.
        • Smith B.D.
        • Candelaria R.P.
        • Babiera G.V.
        • Dogan B.E.
        • Santiago L.
        • Hunt K.K.
        • Kuerer H.M.
        Improved axillary evaluation following neoadjuvant therapy for patients with node-positive breast cancer using selective evaluation of clipped nodes: implementation of targeted axillary dissection.
        J. Clin. Oncol. 2016; 34: 1072-1078
        • Donker M.
        • Straver M.E.
        • Wesseling J.
        • Loo C.E.
        • Schot M.
        • Drukker C.A.
        • van Tinteren H.
        • Sonke G.S.
        • Rutgers E.J.
        • Vrancken Peeters M.J.
        Marking axillary lymph nodes with radioactive iodine seeds for axillary staging after neoadjuvant systemic treatment in breast cancer patients: the MARI procedure.
        Ann. Surg. 2015; 261: 378-382
        • Giuliano A.E.
        • Ballman K.
        • McCall L.
        • Beitsch P.
        • Whitworth P.W.
        • Blumencranz P.
        • Leitch A.M.
        • Saha S.
        • Morrow M.
        • Hunt K.K.
        Locoregional recurrence after sentinel lymph node dissection with or without axillary dissection in patients with sentinel lymph node metastases: long-term follow-up from the American College of Surgeons Oncology Group (Alliance) ACOSOG Z0011 randomized trial.
        Ann. Surg. 2016; 264: 413-420
        • Farrell T.P.
        • Adams N.C.
        • Stenson M.
        • Carroll P.A.
        • Griffin M.
        • Connolly E.M.
        • O’Keeffe S.A.
        The Z0011 Trial: Is this the end of axillary ultrasound in the pre-operative assessment of breast cancer patients?.
        Eur. Radiol. 2015; 25: 2682-2687
        • Ahmed M.
        • Jozsa F.
        • Baker R.
        • Rubio I.T.
        • Benson J.
        • Douek M.
        Meta-analysis of tumour burden in pre-operative axillary ultrasound positive and negative breast cancer patients.
        Breast Cancer Res. Treat. 2017; 166: 329-336
        • Hieken T.J.
        • Trull B.C.
        • Boughey J.C.
        • Jones K.N.
        • Reynolds C.A.
        • Shah S.S.
        • Glazebrook K.N.
        Preoperative axillary imaging with percutaneous lymph node biopsy is valuable in the contemporary management of patients with breast cancer.
        Surgery. 2013; 154 (discussion 838-840): 831-838
        • Ecanow J.S.
        • Abe H.
        • Newstead G.M.
        • Ecanow D.B.
        • Jeske J.M.
        Axillary staging of breast cancer: what the radiologist should know.
        Radiographics. 2013; 33: 1589-1612
        • D’Orsi CJ S.E.
        • Mendelson E.B.
        • Morris E.A.
        • et al.
        ACR BI-RADS® Atlas, Breast Imaging Reporting and Data System.
        American College of Radiology, Reston, VA2013
        • Spak D.A.
        • Plaxco J.S.
        • Santiago L.
        • Dryden M.J.
        • Dogan B.E.
        BI-RADS(®) fifth edition: a summary of changes.
        Diagn. Interv. Imaging. 2017; 98: 179-190
        • Alvarez S.
        • Añorbe E.
        • Alcorta P.
        • López F.
        • Alonso I.
        • Cortés J.
        Role of sonography in the diagnosis of axillary lymph node metastases in breast cancer: a systematic review.
        AJR Am. J. Roentgenol. 2006; 186: 1342-1348
        • Deurloo E.E.
        • Tanis P.J.
        • Gilhuijs K.G.
        • Muller S.H.
        • Kröger R.
        • Peterse J.L.
        • Rutgers E.J.
        • Valdés Olmos R.
        • Schultze Kool L.J.
        Reduction in the number of sentinel lymph node procedures by preoperative ultrasonography of the axilla in breast cancer.
        Eur. J. Cancer (Oxford, Engl.: 1990). 2003; 39: 1068-1073
        • Esen G.
        • Gurses B.
        • Yilmaz M.H.
        • Ilvan S.
        • Ulus S.
        • Celik V.
        • Farahmand M.
        • Calay O.O.
        Gray scale and power Doppler US in the preoperative evaluation of axillary metastases in breast cancer patients with no palpable lymph nodes.
        Eur. Radiol. 2005; 15: 1215-1223
        • Lee B.
        • Lim A.K.
        • Krell J.
        • Satchithananda K.
        • Coombes R.C.
        • Lewis J.S.
        • Stebbing J.
        The efficacy of axillary ultrasound in the detection of nodal metastasis in breast cancer.
        AJR Am. J. Roentgenol. 2013; 200: W314-W320
        • Rahbar H.
        • Partridge S.C.
        • Javid S.H.
        • Lehman C.D.
        Imaging axillary lymph nodes in patients with newly diagnosed breast cancer.
        Curr. Probl. Diagn. Radiol. 2012; 41: 149-158
        • Mainiero M.B.
        • Cinelli C.M.
        • Koelliker S.L.
        • Graves T.A.
        • Chung M.A.
        Axillary ultrasound and fine-needle aspiration in the preoperative evaluation of the breast cancer patient: an algorithm based on tumor size and lymph node appearance.
        AJR Am. J. Roentgenol. 2010; 195: 1261-1267
        • Bossuyt P.M.
        • Reitsma J.B.
        • Bruns D.E.
        • Gatsonis C.A.
        • Glasziou P.P.
        • Irwig L.
        • Lijmer J.G.
        • Moher D.
        • Rennie D.
        • de Vet H.C.
        • Kressel H.Y.
        • Rifai N.
        • Golub R.M.
        • Altman D.G.
        • Hooft L.
        • Korevaar D.A.
        • Cohen J.F.
        STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies.
        BMJ (Clin. Res. Ed.). 2015; 351: h5527
        • Wockel A.
        • Festl J.
        • Stuber T.
        • Brust K.
        • Stangl S.
        • Heuschmann P.U.
        • Albert U.S.
        • Budach W.
        • Follmann M.
        • Janni W.
        • Kopp I.
        • Kreienberg R.
        • Kuhn T.
        • Langer T.
        • Nothacker M.
        • Scharl A.
        • Schreer I.
        • Link H.
        • Engel J.
        • Fehm T.
        • Weis J.
        • Welt A.
        • Steckelberg A.
        • Feyer P.
        • Konig K.
        • Hahne A.
        • Kreipe H.H.
        • Knoefel W.T.
        • Denkinger M.
        • Brucker S.
        • Luftner D.
        • Kubisch C.
        • Gerlach C.
        • Lebeau A.
        • Siedentopf F.
        • Petersen C.
        • Bartsch H.H.
        • Schulz-Wendtland R.
        • Hahn M.
        • Hanf V.
        • Muller-Schimpfle M.
        • Henscher U.
        • Roncarati R.
        • Katalinic A.
        • Heitmann C.
        • Honegger C.
        • Paradies K.
        • Bjelic-Radisic V.
        • Degenhardt F.
        • Wenz F.
        • Rick O.
        • Holzel D.
        • Zaiss M.
        • Kemper G.
        • Budach V.
        • Denkert C.
        • Gerber B.
        • Tesch H.
        • Hirsmuller S.
        • Sinn H.P.
        • Dunst J.
        • Munstedt K.
        • Bick U.
        • Fallenberg E.
        • Tholen R.
        • Hung R.
        • Baumann F.
        • Beckmann M.W.
        • Blohmer J.
        • Fasching P.A.
        • Lux M.P.
        • Harbeck N.
        • Hadji P.
        • Hauner H.
        • Heywang-Kobrunner S.
        • Huober J.
        • Hubner J.
        • Jackisch C.
        • Loibl S.
        • Luck H.J.
        • von Minckwitz G.
        • Mobus V.
        • Muller V.
        • Nothlings U.
        • Schmidt M.
        • Schmutzler R.
        • Schneeweiss A.
        • Schutz F.
        • Stickeler E.
        • Thomssen C.
        • Untch M.
        • Wesselmann S.
        • Bucker A.
        • Krockenberger M.
        Interdisciplinary screening, diagnosis, therapy and follow-up of breast cancer. Guideline of the DGGG and the DKG (S3-Level, AWMF registry number 032/045OL, December 2017) - part 1 with recommendations for the screening, diagnosis and therapy of breast cancer.
        Geburtshilfe Frauenheilkd. 2018; 78: 927-948
        • Gnant M.
        • Harbeck N.
        • Thomssen C.
        St. Gallen/Vienna 2017: a brief summary of the consensus discussion about escalation and de-escalation of primary breast cancer treatment.
        Breast Care (Basel). 2017; 12: 102-107
        • Barrio A.V.
        • Mamtani A.
        • Edelweiss M.
        • Eaton A.
        • Stempel M.
        • Murray M.P.
        • Morrow M.
        How often is treatment effect identified in axillary nodes with a pathologic complete response after neoadjuvant chemotherapy?.
        Ann. Surg. Oncol. 2016; 23: 3475-3480
        • Gipponi M.
        • Fregatti P.
        • Garlaschi A.
        • Murelli F.
        • Margarino C.
        • Depaoli F.
        • Baccini P.
        • Gallo M.
        • Friedman D.
        Axillary ultrasound and Fine-Needle Aspiration Cytology in the preoperative staging of axillary node metastasis in breast cancer patients.
        Breast (Edinburgh, Scotland). 2016; 30: 146-150
        • Mills P.
        • Sever A.
        • Weeks J.
        • Fish D.
        • Jones S.
        • Jones P.
        Axillary ultrasound assessment in primary breast cancer: an audit of 653 cases.
        Breast J. 2010; 16: 460-463
        • Solon J.G.
        • Power C.
        • Al-Azawi D.
        • Duke D.
        • Hill A.D.
        Ultrasound-guided core biopsy: an effective method of detecting axillary nodal metastases.
        J. Am. Coll. Surg. 2012; 214: 12-17
        • Sataloff D.M.
        • Mason B.A.
        • Prestipino A.J.
        • Seinige U.L.
        • Lieber C.P.
        • Baloch Z.
        Pathologic response to induction chemotherapy in locally advanced carcinoma of the breast: a determinant of outcome.
        J. Am. Coll. Surg. 1995; 180: 297-306
        • Newman L.A.
        • Pernick N.L.
        • Adsay V.
        • Carolin K.A.
        • Philip P.A.
        • Sipierski S.
        • Bouwman D.L.
        • Kosir M.A.
        • White M.
        • Visscher D.W.
        Histopathologic evidence of tumor regression in the axillary lymph nodes of patients treated with preoperative chemotherapy correlates with breast cancer outcome.
        Ann. Surg. Oncol. 2003; 10: 734-739
        • Glaeser A.
        • Sinn H.P.
        • Garcia-Etienne C.
        • Riedel F.
        • Hug S.
        • Schaefgen B.
        • Golatta M.
        • Hennigs A.
        • Feisst M.
        • Sohn C.
        • Heil J.
        Heterogeneous responses of axillary lymph node metastases to neoadjuvant chemotherapy are common and depend on breast cancer subtype.
        Ann. Surg. Oncol. 2019; 26: 4381-4389
        • Tucker N.S.
        • Cyr A.E.
        • Ademuyiwa F.O.
        • Tabchy A.
        • George K.
        • Sharma P.K.
        • Jin L.X.
        • Sanati S.
        • Aft R.
        • Gao F.
        • Margenthaler J.A.
        • Gillanders W.E.
        Axillary ultrasound accurately excludes clinically significant lymph node disease in patients with early stage breast cancer.
        Ann. Surg. 2016; 264: 1098-1102
        • Shigematsu H.
        • Nishina M.
        • Yasui D.
        • Hirata T.
        • Ozaki S.
        Minimal prognostic significance of sentinel lymph node metastasis in patients with cT1-2 and cN0 breast cancer.
        World J. Surg. Oncol. 2019; 17: 41
        • Henke G.
        • Knauer M.
        • Ribi K.
        • Hayoz S.
        • Gérard M.A.
        • Ruhstaller T.
        • Zwahlen D.R.
        • Muenst S.
        • Ackerknecht M.
        • Hawle H.
        • Fitzal F.
        • Gnant M.
        • Mátrai Z.
        • Ballardini B.
        • Gyr A.
        • Kurzeder C.
        • Weber W.P.
        Tailored axillary surgery with or without axillary lymph node dissection followed by radiotherapy in patients with clinically node-positive breast cancer (TAXIS): study protocol for a multicenter, randomized phase-III trial.
        Trials. 2018; 19: 667