Supplemental screening using breast MRI in women with mammographically dense breasts

Abstract

Kaiser and colleagues publish a cost-effectiveness analysis in this issue of the European Journal of Radiology, finding that using MR-mammography (mainly referred to as breast MRI) as a supplemental screening test in women with extremely dense breasts is cost effective compared to the standard mammography only approach [[1]Kaiser C.G. Dietzel M. Vag T. Froelich M.F. Cost-effectiveness of MR-mammography vs. conventional mammography in screening patients at intermediate risk of breast cancer - A model-based economic evaluation.Eur. J. Radiol. 2020; : 109355https://doi.org/10.1016/j.ejrad.2020.109355Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar]. The topic is both of broad interest and complex and therefore warrants some structured discussion. Despite controversial discussions about the harms and benefits of breast cancer screening using x-ray mammography, mammography screening is supported by international expert societies [[2]Sardanelli F. Aase H.S. Álvarez M. Azavedo E. Baarslag H.J. Balleyguier C. Baltzer P.A. Beslagic V. Bick U. Bogdanovic-Stojanovic D. Briediene R. Brkljacic B. Camps Herrero J. Colin C. Cornford E. Danes J. de Geer G. Esen G. Evans A. Fuchsjaeger M.H. Gilbert F.J. Graf O. Hargaden G. Helbich T.H. Heywang-Köbrunner S.H. Ivanov V. Jónsson Á. Kuhl C.K. Lisencu E.C. Luczynska E. Mann R.M. Marques J.C. Martincich L. Mortier M. Müller-Schimpfle M. Ormandi K. Panizza P. Pediconi F. Pijnappel R.M. Pinker K. Rissanen T. Rotaru N. Saguatti G. Sella T. Slobodníková J. Talk M. Taourel P. Trimboli R.M. Vejborg I. Vourtsis A. Forrai G. 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: 2737-2743https://doi.org/10.1007/s00330-016-4612-zCrossref PubMed Scopus (89) Google Scholar]. In the last decade, personalised medicine has gained increasing popularity. This approach adapts diagnostic and therapeutic strategies to the individual patient with the aim of improving healthcare. A major step towards a personalised screening approach was the realisation, that breast density is both a predictor of breast cancer risk and mammographic sensitivity. This holds true both for quantitative and quantitative assessment of breast density [3McCormack V.A. dos Santos Silva I. Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis.Cancer Epidemiol. Biomark. Prev. Publ. Am. Assoc. Cancer Res. Cosponsored Am. Soc. Prev. Oncol. 2006; 15: 1159-1169https://doi.org/10.1158/1055-9965.EPI-06-0034Crossref Scopus (1384) Google Scholar, 4Pettersson A. Graff R.E. Ursin G. Santos Silva I.D. McCormack V. Baglietto L. Vachon C. Bakker M.F. Giles G.G. Chia K.S. Czene K. Eriksson L. Hall P. Hartman M. Warren R.M.L. Hislop G. Chiarelli A.M. Hopper J.L. Krishnan K. Li J. Li Q. Pagano I. Rosner B.A. Wong C.S. Scott C. Stone J. Maskarinec G. Boyd N.F. van Gils C.H. Tamimi R.M. Mammographic density phenotypes and risk of breast cancer: a meta-analysis.J. Natl. Cancer Inst. 2014; 106https://doi.org/10.1093/jnci/dju078Crossref PubMed Scopus (181) Google Scholar, 5Pisano E.D. Gatsonis C. Hendrick E. Yaffe M. Baum J.K. Acharyya S. Conant E.F. Fajardo L.L. Bassett L. D’Orsi C. Jong R. Rebner M. Digital mammographic imaging screening trial (dmist) investigators group, diagnostic performance of digital versus film mammography for breast-cancer screening.N. Engl. J. Med. 2005; 353: 1773-1783https://doi.org/10.1056/NEJMoa052911Crossref PubMed Scopus (1476) Google Scholar]. Many countries have therefore adapted a strategy to report mammographic breast density along with the diagnostic reading results. The facts call for an intervention: women, particularly those with dense breasts are underdiagnosed for breast cancer. This is evident by significant interval cancer and advanced (e.g. lymph node positive) cancer rates and breast cancer mortality leading mortality statistics [[6]Kuhl C.K. Underdiagnosis is the main challenge in breast cancer screening.Lancet Oncol. 2019; 20: 1044-1046https://doi.org/10.1016/S1470-2045(19)30314-6Abstract Full Text Full Text PDF PubMed Scopus (3) Google Scholar]. Of all available breast imaging methods, MRI has the highest sensitivity. And as opposed to a commonly repeated saying, there is no evidence for a lower specificity compared to other imaging methods [[7]Baltzer P. Sardanelli F. The mantra about Low specificity of breast MRI.Breast MRI High-Risk Screen. Springer, 2020: 11-21https://doi.org/10.1007/978-3-030-41207-4_2Crossref Google Scholar,[8]Benndorf M. Baltzer P.A.T. Vag T. Gajda M. Runnebaum I.B. Kaiser W.A. Breast MRI as an adjunct to mammography: does it really suffer from low specificity? A retrospective analysis stratified by mammographic BI-RADS classes, Acta Radiol.Acta Radiol. 1987; 51: 715-721https://doi.org/10.3109/02841851.2010.497164Crossref Scopus (55) Google Scholar]. This is why breast MRI has been suggested for breast cancer screening very early in its development [[9]Kaiser W.A. MR Mammography (MRM). Springer-Verlag, Berlin Heidelberg1993https://www.springer.com/de/book/9783642772580Crossref Google Scholar]. Finally, after decades, a high-quality prospective randomized screening study – the DENSE trial - has been conducted in women with extremely dense breasts [[10]Bakker M.F. de Lange S.V. Pijnappel R.M. Mann R.M. Peeters P.H.M. Monninkhof E.M. Emaus M.J. Loo C.E. Bisschops R.H.C. Lobbes M.B.I. de Jong M.D.F. Duvivier K.M. Veltman J. Karssemeijer N. de Koning H.J. van Diest P.J. Mali W.P.T.M. van den Bosch M.A.A.J. Veldhuis W.B. van Gils C.H. DENSE trial study group, supplemental MRI screening for women with extremely dense breast tissue.N. Engl. J. Med. 2019; 381: 2091-2102https://doi.org/10.1056/NEJMoa1903986Crossref PubMed Scopus (133) Google Scholar]. Diagnostically, the results are highly convincing: MRI demonstrated a by far higher sensitivity, and interval cancers in MRI-screened women were reduced to less than 1/1000. Expectedly, the DENSE study was confronted with opposition: an editorial published along with the study questions its relevance, weighing the number of detected cancers against the effort of additional tests and the psychological and physical harm of false positive biopsies that amounted to 74 % of all 300 biopsied women [[11]Longo D.L. Detecting breast Cancer in women with dense breasts.N. Engl. J. Med. 2019; 381: 2169-2170https://doi.org/10.1056/NEJMe1912943Crossref PubMed Scopus (15) Google Scholar]. As known from mammographic screening, decreasing interval cancers (as a surrogate for screening effectiveness) has a prize: increased recalls [[12]Burnside E.S. Vulkan D. Blanks R.G. Duffy S.W. Association between screening mammography recall rate and interval cancers in the uk breast cancer service screening program: a cohort study.Radiology. 2018; 288: 47-54https://doi.org/10.1148/radiol.2018171539Crossref PubMed Scopus (12) Google Scholar]. Such a trade-off has constraints as the rate of detectable cancers is limited by test sensitivity. Notably, the numbers provided by modelling conventional screening results nicely fit with the DENSE trial results. Therefore, MRI provides an opportunity to substantially reduce interval cancers unrivalled by mammography while the price in terms of recalls is exactly as high as it would be for mammography. While the harms of MR-guided biopsy are considered low, high effort and suboptimal availability do indeed play a role in its application [[13]Clauser P. Mann R. Athanasiou A. Prosch H. Pinker K. Dietzel M. Helbich T.H. Fuchsjäger M. Camps-Herrero J. Sardanelli F. Forrai G. Baltzer P.A.T. A survey by the European Society of Breast Imaging on the utilisation of breast MRI in clinical practice.Eur. Radiol. 2018; 28: 1909-1918https://doi.org/10.1007/s00330-017-5121-4Crossref PubMed Scopus (51) Google Scholar]. In general, a quite substantial amount of research has focused on avoiding unnecessary biopsies in the workup of breast abnormalities [[14]Baltzer P.A.T. Kapetas P. Marino M.A. Clauser P. New diagnostic tools for breast cancer.Memo. 2017; 10: 175-180https://doi.org/10.1007/s12254-017-0341-5Crossref PubMed Scopus (23) Google Scholar]. The dense trialists themselves addressed the topic in a follow-up study by developing an A.I. classifier through combining multiparametric texture analysis with machine-learning, demonstrating the potential to downgrade MRI recalls by more than 40 % [[15]Verburg E. van Gils C.H. Bakker M.F. Viergever M.A. Pijnappel R.M. Veldhuis W.B. Gilhuijs K.G.A. Computer-Aided Diagnosis in Multiparametric Magnetic Resonance Imaging Screening of Women With Extremely Dense Breasts to Reduce False-Positive Diagnoses.Invest. Radiol. 2020; 55: 438-444https://doi.org/10.1097/RLI.0000000000000656Crossref PubMed Scopus (10) Google Scholar]. A recent study published by Jajodia applied the Kaiser score, a simple BI-RADS feature based and machine-learning derived clinical decision rule to equivocal MRI lesions and reported a potential to downgrade >60 % of PPV1 recalls [[16]Jajodia A. Sindhwani G. Pasricha S. Prosch H. Puri S. Dewan A. Batra U. Doval D.C. Mehta A. Chaturvedi A.K. Application of the Kaiser score to increase diagnostic accuracy in equivocal lesions on diagnostic mammograms referred for MR mammography.Eur. J. Radiol. 2020; 134: 109413https://doi.org/10.1016/j.ejrad.2020.109413Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar]. These findings are in line with prior reports applying the same score in high-risk patients [[17]Milos R.I. Pipan F. Kalovidouri A. Clauser P. Kapetas P. Bernathova M. Helbich T.H. Baltzer P.A.T. The Kaiser score reliably excludes malignancy in benign contrast-enhancing lesions classified as BI-RADS 4 on breast MRI high-risk screening exams.Eur. Radiol. 2020; https://doi.org/10.1007/s00330-020-06945-zCrossref PubMed Scopus (9) Google Scholar], lesions presenting as mammographic microcalcifications [[18]Wengert G.J. Pipan F. Almohanna J. Bickel H. Polanec S. Kapetas P. Clauser P. Pinker K. Helbich T.H. Baltzer P.A.T. Impact of the Kaiser score on clinical decision-making in BI-RADS 4 mammographic calcifications examined with breast MRI.Eur. Radiol. 2020; 30: 1451-1459https://doi.org/10.1007/s00330-019-06444-wCrossref PubMed Scopus (10) Google Scholar] and suspicious BI-RADS 4 lesions only visible on MRI [[19]Woitek R. Spick C. Schernthaner M. Rudas M. Kapetas P. Bernathova M. Furtner J. Pinker K. Helbich T.H. Baltzer P.A.T. A simple classification system (the Tree flowchart) for breast MRI can reduce the number of unnecessary biopsies in MRI-only lesions.Eur. Radiol. 2017; 27: 3799-3809https://doi.org/10.1007/s00330-017-4755-6Crossref PubMed Scopus (26) Google Scholar]. Notably, according to a multicentric analysis, the Kaiser score seems to be better suited to this task than the most promising quantitative imaging biomarker in breast imaging, the Apparent Diffusion Coefficient derived from Diffusion Weighted Imaging [[20]Dietzel M. Krug B. Clauser P. Burke C. Hellmich M. Maintz D. Uder M. Bickel H. Helbich T. Baltzer P.A.T. A multicentric comparison of apparent diffusion coefficient mapping and the kaiser score in the assessment of breast lesions.Invest. Radiol. 2020; https://doi.org/10.1097/RLI.0000000000000739Crossref PubMed Scopus (5) Google Scholar]. These findings suggest the clinical application of the Kaiser score as an effective tool to avoid unnecessary biopsies at a high level of evidence. A majority of MRI detected lesions can be identified by second-look or targeted ultrasound (SLU), allowing to perform ultrasound follow-ups and biopsies to further resolve MRI recalls [[21]Spick C. Baltzer P.A.T. Diagnostic Utility of Second-Look US for Breast Lesions Identified at MR Imaging: Systematic Review and Meta-Analysis.Radiology. 2014; : 140474https://doi.org/10.1148/radiol.14140474Crossref Scopus (93) Google Scholar]. Ultrasound is by far better available and much less costly compared to MRI. In the journal, two important findings regarding SLU have been recently published: First, criteria to identify which patients are best suited for second-look ultrasound [[22]Bumberger A. Clauser P. Kolta M. Kapetas P. Bernathova M. Helbich T.H. Pinker K. Baltzer P.A. Can we predict lesion detection rates in second-look ultrasound of MRI-detected breast lesions? A systematic analysis.Eur. J. Radiol. 2019; 113: 96-100https://doi.org/10.1016/j.ejrad.2019.02.008Abstract Full Text Full Text PDF PubMed Scopus (8) Google Scholar]. While patients could indeed be stratified into those having MRI-detected lesions identified by ultrasound with high, medium and low probability, even the low probability group showed a success rate that practically suggests to use SLU in all MRI-suspicious lesions before proceeding to MRI-guided interventions or follow-ups. Second, the superior spatial resolution of ultrasound allows to downgrade lesions presenting as benign on targeted ultrasound even if suspicious on breast MRI, thereby avoiding image-guided biopsy [[23]Kolta M. Clauser P. Kapetas P. Bernathova M. Pinker K. Helbich T.H. Baltzer P.A.T. Can second-look ultrasound downgrade MRI-detected lesions? A retrospective study.Eur. J. Radiol. 2020; 127: 108976https://doi.org/10.1016/j.ejrad.2020.108976Abstract Full Text Full Text PDF PubMed Scopus (3) Google Scholar]. Image-guided biopsies are considered safe irrespective of the method used. Proven is a certain psychological harm in terms of increased anxiety and associated symptoms [[24]Mathioudakis A.G. Salakari M. Pylkkanen L. Saz-Parkinson Z. Bramesfeld A. Deandrea S. Lerda D. Neamtiu L. Pardo-Hernandez H. Solà I. Alonso-Coello P. Systematic review on women’s values and preferences concerning breast cancer screening and diagnostic services.Psychooncology. 2019; 28: 939-947https://doi.org/10.1002/pon.5041Crossref PubMed Scopus (25) Google Scholar]. Despite this, psychooncological research data demonstrate that women are actually willing to accept both the direct harms by recalls and unnecessary biopsies and the possibility of overdiagnosis for the prospect of an earlier diagnosis [[24]Mathioudakis A.G. Salakari M. Pylkkanen L. Saz-Parkinson Z. Bramesfeld A. Deandrea S. Lerda D. Neamtiu L. Pardo-Hernandez H. Solà I. Alonso-Coello P. Systematic review on women’s values and preferences concerning breast cancer screening and diagnostic services.Psychooncology. 2019; 28: 939-947https://doi.org/10.1002/pon.5041Crossref PubMed Scopus (25) Google Scholar]. Overdiagnosis in breast cancer is defined as the diagnosis of cancer that will not become clinically manifest [[25]Welch H.G. Black W.C. Overdiagnosis in cancer.J. Natl. Cancer Inst. 2010; 102: 605-613https://doi.org/10.1093/jnci/djq099Crossref PubMed Scopus (1100) Google Scholar]. Overdiagnosis is a controversially discussed phenomenon: it is both very real and can be estimated epidemiologically and on the other hand very abstract as it cannot be predicted on an individual level [[25]Welch H.G. Black W.C. Overdiagnosis in cancer.J. Natl. Cancer Inst. 2010; 102: 605-613https://doi.org/10.1093/jnci/djq099Crossref PubMed Scopus (1100) Google Scholar,[26]Evans A. Vinnicombe S. Overdiagnosis in breast imaging.Breast Edinb. Scotl. 2017; 31: 270-273https://doi.org/10.1016/j.breast.2016.10.011Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar]. In other words: neither a woman diagnosed with breast cancer nor her physicians will be able to say whether her cancer is biologically relevant or not. While increasing recall thresholds may alleviate the problem somewhat, there is actually no data on how much overdiagnosis is a necessary by-product of diagnosing biologically aggressive cancers and how much of this side effect could be alleviated by improved classifications and individualized treatment strategies [[27]Hewitt K. Son J. Glencer A. Borowsky A.D. Cooperberg M.R. Esserman L.J. The evolution of our understanding of the biology of Cancer Is the key to avoiding overdiagnosis and overtreatment, Cancer epidemiol.Prev. Biomark. 2020; 29: 2463-2474https://doi.org/10.1158/1055-9965.EPI-20-0110Crossref PubMed Scopus (4) Google Scholar]. Why is overdiagnosis a problem at all? As we can’t determine which cancers must be treated and which not, all are treated according to their subtype. In case of a biologically indolent breast cancer, this treatment must be considered overtreatment. Both screening and treatment success estimates are diminished by this phenomenon as evidently e.g. mortality will decrease if enough indolent cancers are diagnosed and “successfully” treated though they didn’t require therapy from the beginning. Several letters and a national health council assessment report pointed out the possibility of overdiagnosis in the DENSE trial and use it as an argument to advise against using supplemental MRI in these women that are currently very definitely underdiagnosed by mammography screening alone [28Sharma V. MRI screening in women with dense breasts.N. Engl. J. Med. 2020; 382: 1282https://doi.org/10.1056/NEJMc1917231Crossref PubMed Google Scholar, 29Welch H.G. Zahl P.-H. MRI screening in women with dense breasts.N. Engl. J. Med. 2020; 382: 1283-1284https://doi.org/10.1056/NEJMc1917231Crossref PubMed Google Scholar, 30Autier P. MRI screening in women with dense breasts.N. Engl. J. Med. 2020; 382: 1282https://doi.org/10.1056/NEJMc1917231Crossref PubMed Google Scholar, 31Ministerie van Volksgezondheid Wen S. The Use of MRI Screening in the Population Screening Programme for Breast Cancer - Advisory Report - the Health Council of the Netherlands.2020Google Scholar]. Two main arguments can be brought up in opposition to using overdiagnosis as a criticism against supplemental MRI screening. First, a more sensitive test such as breast MRI will lead to overdiagnosis as an inevitable by-product of supplemental screening if treatment approaches are not adapted to the changed situation [[26]Evans A. Vinnicombe S. Overdiagnosis in breast imaging.Breast Edinb. Scotl. 2017; 31: 270-273https://doi.org/10.1016/j.breast.2016.10.011Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar]. Whether it is justified harming many to safe few is an ethical dispute pre-dating modern medicine and far beyond the scope of this text. In case of doubt, we should consider the opinion of those affected. Eligible women themselves, being those that are subject to both risks and benefits, generally are in favour of screening despite acknowledging the risk [[24]Mathioudakis A.G. Salakari M. Pylkkanen L. Saz-Parkinson Z. Bramesfeld A. Deandrea S. Lerda D. Neamtiu L. Pardo-Hernandez H. Solà I. Alonso-Coello P. Systematic review on women’s values and preferences concerning breast cancer screening and diagnostic services.Psychooncology. 2019; 28: 939-947https://doi.org/10.1002/pon.5041Crossref PubMed Scopus (25) Google Scholar]. While one could argue that the concept of overdiagnosis is difficult to convey on an individual level and that it is limited understanding of the associated risks that brings women to support screening, such arguments are elitist in nature and deeper discussed in the context of democracy in [[32]Walker J.L. A critique of the elitist theory of democracy.Am. Polit. Sci. Rev. 1966; 60: 285-295https://doi.org/10.2307/1953356Crossref Scopus (157) Google Scholar]. Second, while a substantial reduction of biologically active interval cancers has been proven in the DENSE trial (bringing down the interval cancer rates to those of women without dense breasts, thereby proving that MRI can compensate for the shortcomings of mammography) the presence and amount of overdiagnosis cannot be established yet as pointed out by the DENSE authors in their response to these various letters [[33]Bakker M.F. de Lange S.V. van Gils C.H. MRI screening in women with dense breasts.Reply, N. Engl. J. Med. 2020; 382: 1284https://doi.org/10.1056/NEJMc1917231Crossref PubMed Scopus (1) Google Scholar] and premature conclusions may not be warranted [[31]Ministerie van Volksgezondheid Wen S. The Use of MRI Screening in the Population Screening Programme for Breast Cancer - Advisory Report - the Health Council of the Netherlands.2020Google Scholar]. Still, overdiagnosis will be present as in any screening test. This however, should not be an argument to accept the current fact of underdiagnosis as this can be evidently alleviated by using MRI while research to decrease overdiagnosis by improved risk stratification of breast cancer – including MRI-derived data -is ongoing and showing promising results [34Graña-López L. Herranz M. Domínguez-Prado I. Argibay S. Villares Á. Vázquez-Caruncho M. Can dedicated breast PET help to reduce overdiagnosis and overtreatment by differentiating between indolent and potentially aggressive ductal carcinoma in situ?.Eur. Radiol. 2020; 30: 514-522https://doi.org/10.1007/s00330-019-06356-9Crossref PubMed Scopus (3) Google Scholar, 35Dietzel M. Schulz-Wendtland R. Ellmann S. Zoubi R. Wenkel E. Hammon M. Clauser P. Uder M. Runnebaum I.B. Baltzer P.A.T. Automated volumetric radiomic analysis of breast cancer vascularization improves survival prediction in primary breast cancer.Sci. Rep. 2020; 10: 3664https://doi.org/10.1038/s41598-020-60393-9Crossref PubMed Scopus (5) Google Scholar, 36Baltzer P.A.T. Zoubi R. Burmeister H.P. Gajda M. Camara O. Kaiser W.A. Dietzel M. Computer assisted analysis of MR-mammography reveals association between contrast enhancement and occurrence of distant metastasis.Technol. Cancer Res. Treat. 2012; 11: 553-560Crossref PubMed Scopus (18) Google Scholar, 37Cheon H. Kim H.J. Kim T.H. Ryeom H.-K. Lee J. Kim G.C. Yuk J.-S. Kim W.H. Invasive breast Cancer: prognostic value of peritumoral edema identified at preoperative MR imaging.Radiology. 2018; 287: 68-75https://doi.org/10.1148/radiol.2017171157Crossref PubMed Scopus (34) Google Scholar]. Finally, and probably most interesting in terms of the article by Kaiser and colleagues [[1]Kaiser C.G. Dietzel M. Vag T. Froelich M.F. Cost-effectiveness of MR-mammography vs. conventional mammography in screening patients at intermediate risk of breast cancer - A model-based economic evaluation.Eur. J. Radiol. 2020; : 109355https://doi.org/10.1016/j.ejrad.2020.109355Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar], the Health Council of the Netherlands has assumed that investing in MRI screening would not be worth the investment and, to quote “not future proof” [[31]Ministerie van Volksgezondheid Wen S. The Use of MRI Screening in the Population Screening Programme for Breast Cancer - Advisory Report - the Health Council of the Netherlands.2020Google Scholar]. The authors use the argument of high initial investment costs for MRI facilities while pointing out that Contrast-enhanced mammography (CEM), an increasingly popular mammography-derived technique highlighting tissue vasculature in a similar way as contrast-enhanced MRI, is expected to provide a simpler and cheaper alternative. The cost-effectiveness analysis by Kaiser and colleagues published in this EJR issue provides an econometric analysis using the data provided by the DENSE trial [[10]Bakker M.F. de Lange S.V. Pijnappel R.M. Mann R.M. Peeters P.H.M. Monninkhof E.M. Emaus M.J. Loo C.E. Bisschops R.H.C. Lobbes M.B.I. de Jong M.D.F. Duvivier K.M. Veltman J. Karssemeijer N. de Koning H.J. van Diest P.J. Mali W.P.T.M. van den Bosch M.A.A.J. Veldhuis W.B. van Gils C.H. DENSE trial study group, supplemental MRI screening for women with extremely dense breast tissue.N. Engl. J. Med. 2019; 381: 2091-2102https://doi.org/10.1056/NEJMoa1903986Crossref PubMed Scopus (133) Google Scholar] and comes to the conclusion that the DENSE trial screening approach is cost effective and stays much below international willingness-to-pay standards [[1]Kaiser C.G. Dietzel M. Vag T. Froelich M.F. Cost-effectiveness of MR-mammography vs. conventional mammography in screening patients at intermediate risk of breast cancer - A model-based economic evaluation.Eur. J. Radiol. 2020; : 109355https://doi.org/10.1016/j.ejrad.2020.109355Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar]. Upon dedicated sensitivity analysis, the authors found that it is mainly the costs of MRI, followed by those for x-ray mammography that influence the incremental cost-effectiveness ratio (ICER). Notably, test specificity was of only minor importance regarding cost-effectiveness and did only marginally influence the ICER. The investment of new MRI units is not future proof, the Health Council of the Netherlands states, pointing out that capital costs of MRI considering investment costs including building adaptations too expensive. A Belgian study investigating the dependence of capital costs from field strength and productivity of the system found a 26 % decrease in capital costs per examination in case working hours of the system are increased from 55 to 75 h per week (12.5 h daily for six days). The authors reported an average examination speed of 31 min per examination. A field strength of 3 T had higher net capital costs but showed the same potential for reducing costs by increasing productivity [[38]Obyn C. Cleemput I. Special article the capital cost and productivity of mri in a belgian setting.J. Belg. Soc. Radiol. 2010; 93: 92-96https://doi.org/10.5334/jbr-btr.148Crossref Scopus (8) Google Scholar]. This study provides important data for a breast MRI scenario. First, breast MRI does not have high demands regarding scanner technology: standardized examinations performed at 1.5 T are not diagnostically inferior to 3 T [[39]Dietzel M. Wenkel E. Hammon M. Clauser P. Uder M. Schulz-Wendtland R. Baltzer P.A.T. Does higher field strength translate into better diagnostic accuracy? A prospective comparison of breast MRI at 3 and 1.5 Tesla.Eur. J. Radiol. 2019; 114: 51-56https://doi.org/10.1016/j.ejrad.2019.02.033Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar]. In addition, the breast MRI protocols in the reported scenario did not exceed 15 min of magnet time [[39]Dietzel M. Wenkel E. Hammon M. Clauser P. Uder M. Schulz-Wendtland R. Baltzer P.A.T. Does higher field strength translate into better diagnostic accuracy? A prospective comparison of breast MRI at 3 and 1.5 Tesla.Eur. J. Radiol. 2019; 114: 51-56https://doi.org/10.1016/j.ejrad.2019.02.033Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar]. Already 15 years ago, I regularly assisted Dr. Kaiser (senior) examining more than 40 breast MRI patients during a 12 -h day using a single 1.5 T magnet. Even if only three such working days per week are assumed, the net capital costs of a breast MRI could fall to less than 10 Euros. In addition, MRI units can be cost-efficiently upgraded rather than replaced, providing another angle for reducing capital costs [[40]Kaltenbach B. Roman A. Wichmann J.L. Fischer S. Eichler K. Vogl T.J. Zangos S. MRI Upgrade: A Case Study in Germany.Radiol. Manage. 2017; 39: 17-21PubMed Google Scholar]. Notably, these are costs considering a multi-organ equipped MRI system that could cover a broad bandwidth of clinical examinations, not one dedicated to breast imaging only. Ongoing technical developments towards low cost MRI very much point into the direction of affordable large-scale MRI scanning such as in screening applications [[41]Siemens launches its smallest and most lightweight whole-body MRI, (n.d.). https://healthcare-in-europe.com/en/news/siemens-launches-its-smallest-most-lightweight-whole-body-mri.html (accessed December 20, 2020).Google Scholar,[42]Deepspin – Next-gen MRI, (n.d.). https://deepspin.io/ (accessed December 20, 2020).Google Scholar]. These facts and prospects stress that using MRI for screening women with dense breasts is not only cost effective already but will rather increasingly become so. Dual-energy contrast-enhanced mammography (CEM) is a mammography derived technique that, in analogy to MRI, uses iodine-based contrast media to visualize hypervascularized areas in the breast. The technique provides two sets of images, unenhanced low-energy images that resemble digital mammographies and contrast-enhanced images that only show enhancement. Initial research has shown comparable diagnostic performance estimates of both technologies in selected patients. There currently are no large-scale studies published. Three aspects need to be addressed when considering CEM as an alternative diagnostic test for mammography: In this journal, a pilot study comparing CEM and breast MRI in screening women at high risk for breast cancer has been published [[43]Jochelson M.S. Pinker K. Dershaw D.D. Hughes M. Gibbons G.F. Rahbar K. Robson M.E. Mangino D.A. Goldman D. Moskowitz C.S. Morris E.A. Sung J.S. Comparison of screening CEDM and MRI for women at increased risk for breast cancer: a pilot study.Eur. J. Radiol. 2017; 97: 37-43https://doi.org/10.1016/j.ejrad.2017.10.001Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar]. Though at first glance both methods performed similar, one out of three initially diagnosed cancers was missed by CEM, all were missed by mammography. The PPV of biopsies was similar with 15 % and 14 %. The lack of representative single study data calls for a systematic review and meta-analysis. A recent meta-analysis estimated pooled sensitivity and specificity of CEM as 85 % (95 %-CI: 73–93 %) and 77 % (95 %-CI: 60–88 %) and pointed out the presence of small study effect bias [[44]Suter M.B. Pesapane F. Agazzi G.M. Gagliardi T. Nigro O. Bozzini A. Priolo F. Penco S. Cassano E. Chini C. Squizzato A. Diagnostic accuracy of contrast-enhanced spectral mammography for breast lesions: a systematic review and meta-analysis.Breast Edinb. Scotl. 2020; 53: 8-17https://doi.org/10.1016/j.breast.2020.06.005Abstract Full Text Full Text PDF PubMed Scopus (12) Google Scholar]. Breast MRI data compare favourably to this, reporting pooled sensitivity and specificity of 99 % (95 % CI: 93–100 %) and 89 % (95 %-CI: 85–92 %) in patients without mammographic microcalcifications (no publication bias and no heterogeneity regarding sensitivity and NPV) [[45]Bennani-Baiti B. Bennani-Baiti N. Baltzer P.A. Diagnostic performance of breast magnetic resonance imaging in non-calcified equivocal breast findings: results from a systematic review and meta-analysis.PLoS One. 2016; 11: e0160346https://doi.org/10.1371/journal.pone.0160346Crossref PubMed Scopus (63) Google Scholar]. Though less superior in mammographic calcifications (pooled sensitivity 87 %, 95 %-CI: 81–92 % and specificity 81 %, 95 %-CI: 75–86 %), the NPV of a negative breast MRI for invasive cancers still was 99 % (95 %-CI: 97–100 %) [[46]Bennani-Baiti B. Baltzer P.A. MR imaging for diagnosis of malignancy in mammographic microcalcifications: a systematic review and meta-analysis.Radiology. 2017; 283: 692-701https://doi.org/10.1148/radiol.2016161106Crossref PubMed Scopus (64) Google Scholar]. Clinically, while reflected in these numbers, the soft tissue and contrast-enhancement contrast of MRI, a 3D test not requiring ionizing radiation is far higher than that of CEM. The additional risk of inducing breast cancer by ionizing radiation due to mammographic screening is considered a fact [[47]Yaffe M.J. Mainprize J.G. Risk of radiation-induced breast Cancer from mammographic screening.Radiology. 2011; 258: 98-105https://doi.org/10.1148/radiol.10100655Crossref PubMed Scopus (185) Google Scholar], is higher in women with increased breast densitiy and both tomosynthesis and CEM will rather increase this exposure [[48]Fusco R. Raiano N. Raiano C. Maio F. Vallone P. Mattace Raso M. Setola S.V. Granata V. Rubulotta M.R. Barretta M.L. Petrosino T. Petrillo A. Evaluation of average glandular dose and investigation of the relationship with compressed breast thickness in dual energy contrast enhanced digital mammography and digital breast tomosynthesis.Eur. J. Radiol. 2020; 126: 108912https://doi.org/10.1016/j.ejrad.2020.108912Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar]. Both CEM and breast MRI require the intravenous application of contrast media (CM). Both types of CM are considered quite safe: still, reported rates of 1.54 (iodine-based CM) and 0.40 (gadolinium(Gd)-based CM) adverse effects per 1000 applications are not negligible when screening otherwise healthy women [[49]Hunt C.H. Hartman R.P. Hesley G.K. Frequency and severity of adverse effects of iodinated and gadolinium contrast materials: retrospective review of 456,930 doses.AJR Am. J. Roentgenol. 2009; 193: 1124-1127https://doi.org/10.2214/AJR.09.2520Crossref PubMed Scopus (181) Google Scholar]. In comparison, MRI CM have the better safety profile and concerns regarding possible brain deposits of gadolinium that have been reported mainly in multimorbid patients could not be confirmed in a breast MRI screening setting with multiple Gd-based CM applications [[50]Bennani-Baiti B. Krug B. Giese D. Hellmich M. Bartsch S. Helbich T.H. Baltzer P.A.T. Evaluation of 3.0-T MRI brain signal after exposure to Gadoterate Meglumine in women with high breast Cancer risk and screening breast MRI.Radiology. 2019; 293: 523-530https://doi.org/10.1148/radiol.2019190847Crossref PubMed Scopus (9) Google Scholar]. Finally, it is difficult to estimate the specific risk due to the magnetic field of MRI units. A large report lists a rate of 0.29 % for MRI-related incidences in outpatients, none of them due to interaction of the magnetic field and implants or portable devices [[51]Mansouri M. Aran S. Harvey H.B. Shaqdan K.W. Abujudeh H.H. Rates of safety incident reporting in MRI in a large academic medical center.J. Magn. Reson. Imaging. 2016; 43: 998-1007https://doi.org/10.1002/jmri.25055Crossref PubMed Scopus (11) Google Scholar]. Though the numbers given above refute the notion of a non-inferior diagnostic performance of CEM compared to breast MRI [[31]Ministerie van Volksgezondheid Wen S. The Use of MRI Screening in the Population Screening Programme for Breast Cancer - Advisory Report - the Health Council of the Netherlands.2020Google Scholar], only indirect conclusions regarding cost effectiveness can be drawn. Kaiser and colleagues find supplemental screening test sensitivity and specificity as least important factors on cost-effectiveness [[1]Kaiser C.G. Dietzel M. Vag T. Froelich M.F. Cost-effectiveness of MR-mammography vs. conventional mammography in screening patients at intermediate risk of breast cancer - A model-based economic evaluation.Eur. J. Radiol. 2020; : 109355https://doi.org/10.1016/j.ejrad.2020.109355Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar]. However, their paper does neither address sensitivity and specificity levels as low as reported by Suter and colleagues for CEM [[44]Suter M.B. Pesapane F. Agazzi G.M. Gagliardi T. Nigro O. Bozzini A. Priolo F. Penco S. Cassano E. Chini C. Squizzato A. Diagnostic accuracy of contrast-enhanced spectral mammography for breast lesions: a systematic review and meta-analysis.Breast Edinb. Scotl. 2020; 53: 8-17https://doi.org/10.1016/j.breast.2020.06.005Abstract Full Text Full Text PDF PubMed Scopus (12) Google Scholar] nor do they address different costs associated with different image-guided biopsies. CEM-guided biopsy procedures are mainly under development and far from being state of the art, but would be required when employing CEM as a screening test. Though CEM devices will be cheaper regarding unit costs and necessary building adaptions, their use is limited to breast imaging only and they would be bought for screening purposes. Outside the screening application, comparably low numbers of patients requiring further assessment will be examined with this specialized tool. As discussed above, an MRI unit is not only by far more versatile as it can examine the whole body but data derived from standard MRI scans have already been shown to provide prognostic data in case cancer is diagnosed [35Dietzel M. Schulz-Wendtland R. Ellmann S. Zoubi R. Wenkel E. Hammon M. Clauser P. Uder M. Runnebaum I.B. Baltzer P.A.T. Automated volumetric radiomic analysis of breast cancer vascularization improves survival prediction in primary breast cancer.Sci. Rep. 2020; 10: 3664https://doi.org/10.1038/s41598-020-60393-9Crossref PubMed Scopus (5) Google Scholar, 36Baltzer P.A.T. Zoubi R. Burmeister H.P. Gajda M. Camara O. Kaiser W.A. Dietzel M. Computer assisted analysis of MR-mammography reveals association between contrast enhancement and occurrence of distant metastasis.Technol. Cancer Res. Treat. 2012; 11: 553-560Crossref PubMed Scopus (18) Google Scholar, 37Cheon H. Kim H.J. Kim T.H. Ryeom H.-K. Lee J. Kim G.C. Yuk J.-S. Kim W.H. Invasive breast Cancer: prognostic value of peritumoral edema identified at preoperative MR imaging.Radiology. 2018; 287: 68-75https://doi.org/10.1148/radiol.2017171157Crossref PubMed Scopus (34) Google Scholar]. None of these prognostic features can be assessed by CEM, thus breast MRI seems to be better suited to address the problem of overdiagnosis. A high quality prospective randomized trial demonstrated that using breast MRI as a supplemental tool in women with extremely dense breasts is diagnostically superior to mammography alone and can reduce interval cancers considered as a surrogate for biologically significant breast cancer [[33]Bakker M.F. de Lange S.V. van Gils C.H. MRI screening in women with dense breasts.Reply, N. Engl. J. Med. 2020; 382: 1284https://doi.org/10.1056/NEJMc1917231Crossref PubMed Scopus (1) Google Scholar]. In this editorial, a structured overview over arguments against using breast MRI for screening in women with dense breasts also considering alternative supplemental screening methods has been given. In summary, breast MRI is the most accurate imaging test for diagnosis of breast cancer and, as opposed to common misunderstandings, does not have a worse specificity as compared to mammography. According to Kaiser et al., the approach used in the DENSE trial is cost-effective [[1]Kaiser C.G. Dietzel M. Vag T. Froelich M.F. Cost-effectiveness of MR-mammography vs. conventional mammography in screening patients at intermediate risk of breast cancer - A model-based economic evaluation.Eur. J. Radiol. 2020; : 109355https://doi.org/10.1016/j.ejrad.2020.109355Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar] and an analysis of costs suggests that MRI-related costs can be reduced substantially. The potential competitor, CEM, has a worse CM risk profile, inferior diagnostic performance and would be limited to breast imaging applications while MRI units could be used more broadly, allowing flexible and future-proof planning. Overdiagnosis is a given side-effect of screening, but of all imaging tests used for breast cancer screening, only breast MRI has been shown to provide prognostic information on breast cancer biological behaviour, paving the way for individualized (de-)escalation of treatment strategies.