Effect of integrating 3D-mammography (digital breast tomosynthesis) with 2D-mammography on radiologists’ true-positive and false-positive detection in a population breast screening trial

doi:10.1016/j.ejca.2014.02.004

European Journal of Cancer

Volume 50, Issue 7, May 2014, Pages 1232-1238

https://doi.org/10.1016/j.ejca.2014.02.004 Get rights and content

Abstract

Objective

We investigated the effect of integrating three-dimensional (3D)-mammography with 2D-mammography on radiologists’ detection measures in the ‘screening with tomosynthesis or standard mammography’ (STORM) trial.

Methods

STORM, a prospective population-based trial (Trento and Verona breast screening services) compared sequential screen-reading: 2D-mammography alone and integrated 2D/3D-mammography. Radiologist-specific detection measures were calculated for each screen-reading phase for eight radiologists: number of detected cancers, proportion of true-positive (TP) detection, and number and rate of false-positive (FP) recalls (FPR). We estimated the incremental cancer detection rate (CDR).

Results

There were 59 cancers and 395 false recalls amongst 7292 screening participants. At 2D-mammography screening, radiologist-specific TP detection ranged between 38% and 83% (median 63%; mean 60% and sd 15.4%); at integrated 2D/3D-mammography, TP detection ranged between 78% and 93% (median 87%; mean 87% and sd 5.2%). For all but one radiologist, 2D/3D-mammography improved breast cancer detection (relative to 2D-mammography) ranging between 0% and 54% (median 29%; mean 27% and sd 16.2%) increase in the proportion of detected cancers. Incremental CDR attributable to integrating 3D-mammography in screening varied between 0/1000 and 5.3/1000 screens (median 1.8/1000; mean 2.3/1000 and sd 1.6/1000). Radiologist-specific FPR for 2D-mammography ranged between 1.5% and 4.2% (median 3.1%; mean 2.9% and sd 0.87%), and FPR based on the integrated 2D/3D-mammography read ranged between 1.0% and 3.3% (median 2.4%; mean 2.2% and sd 0.72%). Integrated 2D/3D-mammography screening, relative to 2D-mammography, had the effect of reducing FP and increasing TP detection for most radiologists.

Conclusion

There was broad variability in radiologist-specific TP detection at 2D-mammography and hence in the additional TP detection and incremental CDR attributable to integrated 2D/3D-mammography; more consistent (less variable) TP-detection estimates were observed for the integrated screen-read. Integrating 3D-mammography with 2D-mammography improves radiologists’ screen-reading through improved cancer detection and/or reduced FPR, with most readers achieving both using integrated 2D/3D mammography.

Introduction

Reviews of the evidence from the breast screening randomised trials conclude that mammography screening reduces breast cancer mortality [1], [2]. Digital breast tomosynthesis (DBT), a three-dimensional (3D) form of mammographic technology [3], [4], [5], [6] has been investigated in various clinical studies [6], [7], [8], [9], [10], which have been reviewed by Houssami and Skaane [6] highlighting the need for evidence from prospective screening trials. Recently, two large screening trials, the Oslo trial [11] and the ‘screening with tomosynthesis or standard mammography’ (STORM) trial [12], have provided remarkably consistent findings on 3D-mammography: both have shown that interpreting mammography using standard (2D) images in combination with the 3D images increases breast cancer detection and can also reduce false-positive (FP) detection. Both of these contemporary screening trials were based on double-reading involving several participating radiologists who had interpreted subsets of the screens in the respective trials [11], [12]. The implications of the emerging evidence on 3D-mammography (DBT) for breast screening necessitates investigation of its effect on radiologist-specific detection measures – little evidence is available on this aspect of integrated 2D/3D mammography in the context of population breast screening.

We report an evaluation of the effect of integrating 3D-mammography with 2D-mammography for breast screening on individual radiologists’ true-positive (TP) and FP detection, based on radiologists who participated in the STORM trial.

Section snippets

Methods

We investigated reader-specific detection for radiologists who participated in screen-reading in the STORM trial [12], a prospective population-based study comparing sequential mammography screen-reading implemented in two phases: 2D-mammography alone and integrated 2D/3D mammography. The STORM study methods and population have been described in a recent publication from Ciatto et al. [12], and will be outlined only briefly in this paper. The STORM trial recruited asymptomatic women aged 48

Results

Across all radiologists’ reads, 59 breast cancers were detected at screening in 57 subjects from 7292 screening participants (median age 58 years; range 48–71 years). The mammographic findings for these 59 screen-detected cancers are shown in Table 1 by detection method. The distributions of histology type and stage for the 59 cancers have been described in our primary report of STORM [12] – there were 52 invasive cancers and seven ductal carcinoma in situ (DCIS). The primary outcomes of the

Discussion

We report the effect of screen-reading using 3D-mammography with conventional 2D-mammography on detection measures for radiologists who participated in the STORM trial [12]. The primary outcomes from our trial were based on double-reading, and showed a significant increase in breast cancer detection for integrated 2D/3D mammography screening relative to that using 2D-mammography alone. The present evaluation describes radiologist-specific detection measures, to inform of the potential effect of

Role of funding source

The funding sources did not have any role in the design of the study, in the collection, analysis or interpretation of the data, in the writing of the report, or in the decision to submit the paper.

Conflict of interest statement and acknowledegment of funding

Centres and investigators in Italy participating in this trial declare that they received assistance from Hologic (USA) and Technologic (Italy) in the form of tomosynthesis technology and technical support for the duration of the study, and travel support to attend collaborators’ meetings. We thank Mr. Enrico Tedesco for technical support.

Assoc/Prof Houssami receives research support via a National Breast Cancer Foundation (NBCF Australia) Practitioner Fellowship. Professor Macaskill receives

References (17)

P. Glasziou et al.
The evidence base for breast cancer screening
Prev Med
(2011)
J.A. Baker et al.
Breast tomosynthesis: state-of-the-art and review of the literature
Acad Radiol
(2011)
M.A. Helvie
Digital mammography imaging: breast tomosynthesis and advanced applications
Radiol Clin North Am
(2010)
N. Houssami et al.
Overview of the evidence on digital breast tomosynthesis in breast cancer detection
Breast
(2013)
M.J. Michell et al.
A comparison of the accuracy of film-screen mammography, full-field digital mammography, and digital breast tomosynthesis
Clin Radiol
(2012)
S. Ciatto et al.
Integration of 3D digital mammography with tomosynthesis for population breast-cancer screening (STORM): a prospective comparison study
Lancet Oncol
(2013)
Independent UK Panel on Breast Cancer Screening
The benefits and harms of breast cancer screening: an independent review
Lancet
(2012)
N. Houssami et al.
Digital breast tomosynthesis: the future of mammography screening or much ado about nothing?
Expert Rev Med Devices
(2013)

There are more references available in the full text version of this article.

Cited by (45)

ACR Appropriateness Criteria® Imaging After Mastectomy and Breast Reconstruction
2020, Journal of the American College of Radiology
Citation Excerpt :
There is no relevant literature to support the use of FDG-PET breast for screening in this clinical setting. Although insufficient studies have been performed to assess the utility of DBT in this setting, multiple investigations have demonstrated that DBT is helpful in the screening setting of the native breast, thus decreasing recall rates and increasing cancer detection rates compared to a conventional mammographic workup [21-26]. Evidence is limited, but a few retrospective studies suggest a benefit to screening women with autologous reconstruction after mastectomy for cancer in the reconstruction side.
Mastectomy may be performed to treat breast cancer or as a prophylactic approach in women with a high risk of developing breast cancer. In addition, mastectomies may be performed with or without reconstruction. Reconstruction approaches differ and may be autologous, involving a transfer of tissue (skin, subcutaneous fat, and muscle) from other parts of the body to the chest wall. Reconstruction may also involve implants. Implant reconstruction may occur as a single procedure or as multistep procedures with initial use of an adjustable tissue expander allowing the mastectomy tissues to be stretched without compromising blood supply. Ultimately, a full-volume implant will be placed. Reconstructions with a combination of autologous and implant reconstruction may also be performed. Other techniques such as autologous fat grafting may be used to refine both implant and flap-based reconstruction. This review of imaging in the setting of mastectomy with or without reconstruction summarizes the literature and makes recommendations based on available evidence.
The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.
Breast Cancer Screening and Optimizing Recommendations
2018, Seminars in Roentgenology
Citation Excerpt :
DBT may address some of the limitations previously outlined with FFDM. Several studies have shown increased cancer detection rates with the use of DBT as compared with FFDM alone.62-80 The recall rate has also been shown to be reduced with DBT as compared with FFDM alone.63-65,69,70-72,74,76-80
Effect of integrating digital breast tomosynthesis (3D-mammography) with acquired or synthetic 2D-mammography on radiologists’ true-positive and false-positive detection in a population screening trial: A descriptive study
2018, European Journal of Radiology
We previously reported the Screening with tomosynthesis or standard mammography-2 (STORM-2) trial, showing that tomosynthesis (3D-mammography) screening detected more cancers than 2D-mammography in double-reading practice. In this study, we report reader-specific detection measures for radiologists who performed the screen-reading in this trial.
This is a sub-study of the STORM-2 trial which prospectively integrated 3D-mammography with acquired or synthetized 2D-mammograms in parallel double-reading arms. Asymptomatic women ≥49 years who attended population-based screening (Trento, 2013–2015) were recruited. Screening participants were recalled at any positive sequential screen-read in either reading arm of the trial. Radiologist-specific detection measures were calculated for each of seven radiologists who performed screen-reads: number of detected cancers, proportion of true-positive (TP) detection, and number and rate of false-positive (FP) recalls (FPR). We estimated incremental cancer detection rate (CDR) from integrating 3D-mammography in screen-reading.
Across all radiologists, TP detection (relative sensitivity) ranged between: 46% and 100% (median 59.5%) for 2D-mammography; 75% and 100% (median 76%) for integrated 2D/3D-mammography screening; 56% and 76% (median 64%) for 2Dsynthetic; 67% and 88% (median 78%) for 2Dsynthetic/3D-mammography. Integrating 3D-mammography led to incremental CDRs between 0/1000 and 3.5/1000 screens. FPR ranged between: 1.2% and 2.7% (median 2.25%) for 2D-mammography; 1.5% and 3.4% (median 2.75%) for 2D/3D-mammography; 1.6% and 4.6% (median 2.4%) for 2Dsynthetic; and 1.8% and 6.7% (median 3.0%) for 2Dsynthetic/3D-mammography.
There was variability in the magnitude of effect from integrating 3D-mammography (relative to screen-reading with acquired or synthesised 2D-mammography alone) on individual radiologist’s TP and FP detection, although there was an overall pattern of increasing cancer detection and also increasing FP recall for most readers.
Interval breast cancers in the ‘screening with tomosynthesis or standard mammography’ (STORM) population-based trial
2018, Breast
The prospective ‘screening with tomosynthesis or standard mammography’ (STORM) trial recruited women participating in biennial breast screening in Italy (2011–2012), and compared sequential screen-readings based on 2D-mammography alone or based on tomosynthesis (integrated 2D/3D-mammography). The STORM trial showed that tomosynthesis screen-reading significantly increased breast cancer detection compared to 2D-mammography alone. The present study completes reporting of the trial by examining interval breast cancers ascertained at two year follow-up.
9 interval breast cancers were identified; the estimated interval cancer rate was 1.23/1000 screens [9/7292] (95%CI 0.56 to 2.34) or 1.24/1000 negative screens [9/7235] (95%CI 0.57 to 2.36). In concurrently screened women who attended the same screening services and received 2D-mammography, interval cancer rate was 1.60/1000 screens [40/25,058] (95% CI 1.14 to 2.17) or 1.61/1000 negative screens [40/24,922] (95% CI 1.15 to 2.18). Estimated screening sensitivity for the STORM trial was 85.5% [59/69] (95%CI 75.0%–92.8%), and that for 2D-mammography screening was 77.3% [136/176] (95%CI 70.4%–83.2%).
Interval breast cancer rate amongst screening participants in the STORM trial was marginally lower (and screening sensitivity higher) than estimates amongst 2D-screened women; these findings should be interpreted with caution given the small number of interval cases and the sample size of the trial. Much larger screening studies, or pooled analyses, are required to examine interval cancer rates arising after breast tomosynthesis screening versus digital mammography screening.
Breast biopsy with tomosynthesis guidance
2018, Imagerie de la Femme
Les macrobiopsies sous stéréotaxie sont actuellement réalisées en routine dans de nombreux centres de sénologie. Elles sont indiquées en cas de foyer de microcalcifications suspect, de distorsion architecturale, de masse ou d’asymétrie de densité sans traduction échographique ou IRM. Depuis une dizaine d’années, elles sont réalisées sur table dédiée en procubitus ou sur mammographe diagnostique équipé d’un système embarqué adaptable en position assise ou en latérocubitus. Aussi, avec l’émergence de la mammographie tridimensionnelle détectant des images non visibles en imagerie bidimensionnelle, des systèmes de biopsie sous tomosynthèse se sont développés. L’objectif de cet article est de décrire cette nouvelle technique et de partager notre expérience.
Nowadays biopsies under stereotactic guidance are performed in many breast clinics. They are used to assess suspicious cluster of microcalcification, masses, stromal deformity and focal asymmetry of density without ultrasound scan and MRI abnormality. Until few years ago, they were only performed on a specialized table with a prone positioned patient. Now we can practice on a diagnostic digital mammography machine with a patient in a laterocubitus or a seated position thanks to a removable device added on the mammography machine. Although, the use of diagnostic digital breast tomosynthesis is increasing and allowing detection of abnormalities, as stromal deformity, which are not seen with 2D mammography or ultrasound scan. That's why radiologists have to practice biopsies under tomosynthesis guidance. The aim of this article is to describe this new technic of biopsies and to share our breast clinic experience.
Tomosynthesis and breast cancer screening
2018, Imagerie de la Femme
La tomosynthèse est une méthode d’acquisition dans laquelle plusieurs clichés de basse dose sont réalisés sous différents angles. L’ordinateur génère alors une image de la glande mammaire en trois dimensions (3D) qui permet de s’affranchir au moins partiellement des structures parenchymateuses de voisinage siégeant au contact d’un cancer et pouvant gêner potentiellement sa visualisation, et de négativer les phénomènes de convergence engendrés par la superposition de plusieurs structures mammaires normales qui ne sont pas dans le même plan, potentiellement à l’origine de faux positifs. Les premières expériences publiées ont démontré sur des populations enrichies avec une prévalence importante de cancer du sein, ses performances en termes de sensibilité comme de spécificité. Depuis 5 ans, en Europe comme aux États-Unis, un certain nombre d’études menées avec des méthodologies différentes dans le cadre du dépistage organisé ont montré l’apport de la tomosynthèse, avec une augmentation du nombre de cancers variant entre un et trois cas pour 1000 femmes dépistées et une diminution du taux de rappel variant entre dix et trente rappels pour 1000 femmes dépistées. Cette mise au point détaille ces résultats, analyse les causes scientifiques pouvant limiter l’adoption de la tomosynthèse dans les politiques de dépistage, et souligne l’intérêt pratique d’une acquisition en 3D avec une imagerie bidimensionnelle synthétique reconstruite, dont les performances sont égales à une double acquisition bi- et tridimensionnelle et qui pourrait devenir le standard en dépistage.
Digital breast tomosynthesis is an imaging method for which multiple projection views are acquired and reconstructed to provide three dimensions (3D) mammography. It may allow to diagnose a lesion masked by breast density or overcome fibroglandular tissue superposition misinterpreted as a lesion when none is present. First studies published in diagnostic population with a high prevalence of breast cancer have shown the accuracy of tomosynthesis including an increase of both sensitivity and specificity. For the last 5 years, in Europe as well as in the United States with different design, some studies performed in screening populations have shown the value of tomosynthesis in screening. There was an increase of the rate of detected cancers reaching one to three cancers for 1000 screened women and a decrease of the recall rate varying between ten and 30 recalls for 1000 screened women. This topical review describes these results, analyses the scientific causes that could limit the use of tomosynthesis in screening policy and underlines the practical interest of a 3D acquisition with synthesized two dimensions (2D) digital mammograph which allows the same accuracies than 2D plus 3D acquisitions and which could become the reference standard.

View all citing articles on Scopus

¹: Posthumous (deceased May 2012).

View full text

Effect of integrating 3D-mammography (digital breast tomosynthesis) with 2D-mammography on radiologists’ true-positive and false-positive detection in a population breast screening trial

Abstract

Objective

Methods

Results

Conclusion

Introduction

Section snippets

Methods

Results

Discussion

Role of funding source

Conflict of interest statement and acknowledegment of funding

Prev Med

Acad Radiol

Radiol Clin North Am

Breast

Clin Radiol

Lancet Oncol

The benefits and harms of breast cancer screening: an independent review

Lancet

Digital breast tomosynthesis: the future of mammography screening or much ado about nothing?

Expert Rev Med Devices