Tomosynthesis Is Taking Small Steps to Become the Standard for Breast Cancer Screening
2021; Radiological Society of North America; Volume: 299; Issue: 3 Linguagem: Inglês
10.1148/radiol.2021210153
ISSN1527-1315
Autores Tópico(s)Digital Radiography and Breast Imaging
ResumoHomeRadiologyVol. 299, No. 3 PreviousNext Reviews and CommentaryFree AccessEditorialTomosynthesis Is Taking Small Steps to Become the Standard for Breast Cancer ScreeningRitse M. Mann Ritse M. Mann Author AffiliationsFrom the Department of Medical Imaging, Radboud University Medical Center, Geert Grooteplein 10, 6525 GA Nijmegen, the Netherlands; and Department of Radiology, Netherlands Cancer Institute, Amsterdam, the Netherlands.Address correspondence to the author (e-mail: [email protected]).Ritse M. Mann Published Online:Apr 6 2021https://doi.org/10.1148/radiol.2021210153MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In See also the article by Johnson et al in this issue.Dr Mann is a breast radiologist in the Department of Medical Imaging at the Radboud University Medical Center and the Department of Radiology of the Netherlands Cancer Institute. His laboratory, the breast imaging group, focuses on the improvement and evaluation of radiological techniques for breast cancer detection, evaluation, and treatment. Dr Mann is chair of the scientific committee of the European Society of Breast Imaging.Download as PowerPointOpen in Image Viewer Breast cancer that is detected early does not have an impact on life expectancy. Therefore, the goal of screening is to detect breast cancers early. Because mammographic screening has unequivocally led to a reduction of breast cancer—specific mortality, the screening paradigm works. Earlier untimely deaths are prevented when relevant cancers are detected (1,2).Breast cancer screening does not work to the extent that we would like. Even in women aged 50–70 years (the group of women in whom screening is universally advocated), breast cancer is the most important cause of cancer-related death worldwide. In North America, it is the second cause of cancer-related death only after lung cancer (3). We fail to detect a large number of cancers in a timely fashion. Tackling the topic of underdiagnosis associated with mammographic screening is therefore important.Digital breast tomosynthesis (DBT) recently emerged as the better mammography. Depth information is obtained by moving the x-ray tube over the breast, which reduces masking of lesions by overprojecting fibroglandular tissue. Several prospective trials and large-scale retrospective cohort studies have shown that tomosynthesis depicts roughly 35% more cancer than conventional mammography (4).However, the biologic behavior of breast cancer is variable. Not all cancers progress rapidly to potentially lethal lesions. In fact, some cancerous lesions may never progress into a harmful lesion before the woman who has cancer dies of other causes, which is a situation usually referred to as overdiagnosis. Overdiagnosis is the major reason that detection of more cancers does not automatically translate into better survival, and it is the major general cause of criticism about cancer screening (1).In several studies regarding tomosynthesis, including the Malmö Breast Tomosynthesis Screening Trial (MBTST), neither the tumor size nor the fraction of lymph node—positive cancers changed substantially compared with those observed with screening with mammography alone (5,6). It is therefore essential to investigate whether the detection of more cancers by using DBT is because of overdiagnosis. To evaluate this question, we would require a study that uses mortality as the end point, a design that is common in therapeutic trials.Unfortunately, the use of mortality as an end point is futile in studies about breast cancer screening. Such a study would require huge numbers and a long continuation (≥20 years) because even women with metastasized breast cancer may live for many years. Therefore, evaluating mortality would lead to an unacceptable delay in the introduction of new screening methods that may better serve the women at risk. Moreover, technical evolution will have occurred by the time the study end points would be achieved, which would render the imaging modality out of date. Waiting for unequivocal proof of a new diagnostic technology before implementation would likely mean lack of deployment of new technology because there is always a potentially better technique around the corner. Finally, the end point of mortality in breast cancer screening trials is also not needed. Only when metastatic breast cancer can be systemically cured without serious adverse effects will the screening paradigm no longer hold.Because early detection of cancers that will become clinically relevant is the goal of screening, a common surrogate to assess whether upcoming screening techniques will eventually lead to an increased mortality reduction is the impact on interval cancers. Interval cancers are cancers that manifest between two rounds of screening. Therefore, by definition they are clinically relevant. In this issue of Radiology, Johnson et al (7) reported on the effect of the MBTST on interval cancers. They compared the frequency of interval cancers in trial participants, who were screened with one-view DBT and two-view digital mammography (DM), with that in age-matched contemporary control patients, who were screened with standard two-view DM. Two control patients per woman in the MBTST were retrospectively selected from women screened at the same screening unit.Johnson et al (7) reported that the frequency of interval cancers was 1.6 per 1000 women screened for trial participants, compared with 2.8 per 1000 women screened in the control group, a reduction of 40%. Furthermore, they reported that the interval cancers that occurred in trial participants were on average smaller.The results from the study by Johnson et al (7) were not necessarily expected because earlier studies (see their Table 4) did not find a significant decrease in the rate of interval cancers. Compared with these studies, the unique feature of the MBTST was that they used wide-angle tomosynthesis in one direction only (mediolateral oblique) to overcome the issues of higher dose and longer reading times. In their report regarding the main MBTST end points, the increase in cancer detection rate was 34% (from 6.5 per 1000 women screened to 8.7 per 1000 women screened), which is in line with other studies. Therefore, this approach seems to be a viable alternative to conventional tomosynthesis in two directions (mediolateral oblique and craniocaudal) (8). However, it is difficult to understand how this could lead to a different interval cancer rate.It should be noted that the MBTST was not designed to evaluate changes in the interval cancer rate. Rather, it evaluated diagnostic accuracy of DM and DBT. This implies that in the study arm all women underwent both DM and DBT, but the studies were interpreted independently. Because double reading is performed in Sweden, all examinations in the study arm were therefore read by four radiologists. Those in the control group were read by only two radiologists. The MBTST yielded eight additional cancers that were only detected with DM and could therefore never manifest as interval cancer. When all the cancers detected at DM alone would be added to the interval cancers, the corresponding reduction of the frequency of interval cancers would be 25%, rather than 40%. However, it is unlikely that these cancers would all manifest in the subsequent screening interval. In addition, one could expect a so-called laboratory effect in the trial participants. However, that the cancer detection rate with screening mammography is equal for the study population and the matched control patients argues against this effect, at least for standard mammography. Consequently, we can presume that DBT screening indeed reduces the frequency of interval cancers, albeit likely to a somewhat lesser extent than the authors suggest. Hence, the results do corroborate those of previous investigations.Johnson et al (7) did not report on breast density in the control group. This prevents an evaluation of the value of DBT for the prevention of interval cancers in different density subgroups. Interval cancers are not equally divided over the four categories of breast density but are more than twice as frequent in women with extremely dense breasts compared with women with predominantly fatty breasts. In the publication of the primary trial outcomes, Zackrisson et al (8) showed that the supplemental detection of breast cancers in the highest density category was substantially higher than in the other categories (7.2 per 1000 women screened in category d; 2.8 per 1000 women screened in category c; 2.6 per 1000 women screened in category b; and 1.7 per 1000 women screened in category a). Therefore, when the distribution of density in the women in the control group was different from that in the trial participants, this might have affected the results. Consequently, it would be of great interest to know how the reduction of interval cancers in the breast density categories is distributed. The next question would be whether the results with DBT could potentially rival the excellent results that have recently been published for screening with breast MRI in women with extremely dense breasts. However, by using MRI in women with extremely dense breasts, the reduction of the interval cancer rate was as high as 84% (9), still substantially higher than the overall reduction in this study.To truly assess the effect of DBT on the frequency of interval cancers, randomized studies where women are assigned to undergo either DBT or DM, but not both, are required. These trials should also be sufficiently powered to assess the impact in relevant subgroups, for example, in women with dense breasts. Moreover, performance regarding other (and potentially more important) surrogate markers for improved outcome such as the presence of a stage shift in prevalent and, more importantly, incident screening rounds (ie, does the new screening modality lead to a higher fraction of small node-negative cancers?) should be investigated. Trials that at least partly address these issues are ongoing and are eagerly awaited, most notably the Tomosynthesis Mammographic Imaging Screening Trial (10). Until then, we must acknowledge that technologic developments in breast cancer screening advance more quickly than clinical assessment. During the span of the ongoing technology assessment studies, the spatial resolution of DBT has improved considerably, image reconstruction has become better, and, most notably, the quality of synthetic mammograms obtained from the tomographic images has dramatically increased (also evident from the lack of use of synthetic mammograms in the MBTST). Hence, for the moment, we should likely take tomosynthesis for what it is: better mammography.Disclosures of Conflicts of Interest: R.M.M. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: disclosed money paid to author for consultancies from Bayer Healthcare, Siemens Healthineers, BD, and Transonic Imaging; grants/grants pending from Siemens Healthineers, Medtronic, Bayer Healthcare, BD, Screenpoint Medical, Seno Medical, and Koning. Other relationships: disclosed no relevant relationships.References1. Sardanelli F, Aase HS, Álvarez M, . Position paper on screening for breast cancer by the European Society of Breast Imaging (EUSOBI) and 30 national breast radiology bodies from Austria, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Israel, Lithuania, Moldova, The Netherlands, Norway, Poland, Portugal, Romania, Serbia, Slovakia, Spain, Sweden, Switzerland and Turkey. Eur Radiol 2017;27(7):2737–2743. Crossref, Medline, Google Scholar2. Independent UK Panel on Breast Cancer Screening. The benefits and harms of breast cancer screening: an independent review. Lancet 2012;380(9855):1778–1786. Crossref, Medline, Google Scholar3. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68(6):394–424 [Published correction appears in CA Cancer J Clin 2020;70(4):313.]. Crossref, Medline, Google Scholar4. Alabousi M, Wadera A, Kashif Al-Ghita M, . Performance of Digital Breast Tomosynthesis, Synthetic Mammography and Digital Mammography in Breast Cancer Screening: A Systematic Review and Meta-Analysis. J Natl Cancer Inst 2020.https://doi.org/10.1093/jnci/djaa205. Published online December 29, 2020. Crossref, Medline, Google Scholar5. Bahl M, Gaffney S, McCarthy AM, Lowry KP, Dang PA, Lehman CD. Breast Cancer Characteristics Associated with 2D Digital Mammography versus Digital Breast Tomosynthesis for Screening-detected and Interval Cancers. Radiology 2018;287(1):49–57. Link, Google Scholar6. Johnson K, Zackrisson S, Rosso A, . Tumor Characteristics and Molecular Subtypes in Breast Cancer Screening with Digital Breast Tomosynthesis: The Malmö Breast Tomosynthesis Screening Trial. Radiology 2019;293(2):273–281. Link, Google Scholar7. Johnson K, Lång K, Ikeda DM, Åkesson A, Andersson I, Zackrisson S. Interval Breast Cancer Rates and Tumor Characteristics in the Prospective Population-based Malmö Breast Tomosynthesis Screening Trial. Radiology 2021.https://doi.org/10.1148/radiol.2021204106. Published online April 6, 2021. Link, Google Scholar8. Zackrisson S, Lång K, Rosso A, . One-view breast tomosynthesis versus two-view mammography in the Malmö Breast Tomosynthesis Screening Trial (MBTST): a prospective, population-based, diagnostic accuracy study. Lancet Oncol 2018;19(11):1493–1503. Crossref, Medline, Google Scholar9. Bakker MF, de Lange SV, Pijnappel RM, . Supplemental MRI Screening for Women with Extremely Dense Breast Tissue. N Engl J Med 2019;381(22):2091–2102. Crossref, Medline, Google Scholar10. TMIST/EA1151 Study: Tomosynthesis Mammographic Imaging Screening Trial. American College of Radiology. https://www.acr.org/-/media/ACR/Files/Research/TMIST-Site-Process-Summary.pdf?la=en. Accessed January 18, 2021. Google ScholarArticle HistoryReceived: Jan 18 2021Revision requested: Feb 08 2021Revision received: Feb 08 2021Accepted: Feb 11 2021Published online: Apr 06 2021Published in print: June 2021 FiguresReferencesRelatedDetailsAccompanying This ArticleInterval Breast Cancer Rates and Tumor Characteristics in the Prospective Population-based Malmö Breast Tomosynthesis Screening TrialApr 6 2021RadiologyRecommended Articles Lessons Learned from the Randomized Controlled TOmosynthesis plus SYnthesized MAmmography (TOSYMA) TrialRadiology2022Volume: 306Issue: 2Closing the Chapter on Supplemental Breast Cancer Screening with USRadiology2021Volume: 298Issue: 3pp. 576-577Breast Density and Breast Cancer Screening with Digital Breast Tomosynthesis: A TOSYMA Trial SubanalysisRadiology2022Volume: 306Issue: 2Supplemental MRI in Extremely Dense Breasts: Sharp Reduction in False-Positive Rate in the Second Screening Round of the DENSE TrialRadiology2021Volume: 299Issue: 2pp. 287-289Reducing False-Positive Screening MRI Rates in Women with Extremely Dense BreastsRadiology2021Volume: 301Issue: 2pp. 293-294See More RSNA Education Exhibits Breast Density Included in the Modern Rules of Mammographic ScreeningDigital Posters20192022 New Trends in Breast Density - What Should We Know?Digital Posters2022Contrast-Enhanced Mammography: Current Indications and Future Directions  Digital Posters2019 RSNA Case Collection Deodorant ArtifactRSNA Case Collection2021Locally Advanced Breast CancerRSNA Case Collection2021Invasive Lobular CarcinomaRSNA Case Collection2021 Vol. 299, No. 3 Metrics Altmetric Score PDF download
Referência(s)