INTRODUCTION
We invented Digital Breast Tomosynthesis(DBT) in the 1990’s and patented the technique in 1999 (the patent has run out). Some have called our approach “3D Mammography”. This is inaccurate since the images are not truly 3-Dimensional since the voxels are not isotropic. In the U.S. DBT is replacing 2D Full Field Digital Mammography (FFDM) for screening. The reason is that DBT is a major advance in x-ray mammography of the breast. It eliminates the “structure noise”caused by normal breast tissues that can obscure cancers on standard 2D FFDM while greatly reducing the recall rate from screening. It permits the detection of more small, potentially curable, cancers that are hidden on FFDM. DBT IS SIMPLY A “BETTER MAMMOGRAM”One of the main limitations of FFDM is the fact that cancers can be hidden on the 2-Dimensional images. Although some cancers that develop in dense fibroglandular tissues with no surrounding fat, no associated calcifications,and no architectural distortion can still be hidden on DBT, DBT allows the detection of an additional 30-40% of cancers that will not be evident on 2D mammography, Virtually, every study that has been under taken has shown that DBT is superior to 2D mammography ([1],[2],[3],[4],[5],[6],[7],[8]). Used for screening, DBT detects more cancers at a time when cure is likely. Furthermore, DBT gas the added benefit of reducing recalls from screening. DBT IS A REPLACEMENT FOR 2D MAMMOGRAPHY SCREENINGWhen we were developing DBT in the late1990’s and early 2000’s at the Massachusetts General Hospital it was clear tous that it would replace 2D FFDM for screening. We learned very early in its development that it could reveal cancers that were hidden by normal breast tissues ([9])on 2D FFDM. In addition, since 25% of the women we were recalling from 2D mammography were ultimately shown to have superimposed normal tissues (“summation shadows”) it was clear that DBT, by eliminating superimposed normal tissues, would prove to reduce the need to recall many women for additional evaluation ([10]). This was a surprising, but a highly desirable development that significantly reduces the recall rate from screening. Usually when you improve sensitivity, it comes at the expense of more “false positives” (calling more women back) and reduced specificity, but DBT has the major and unusual advantage of increasing sensitivity and simultaneously increasing specificity.Many of the early observational studies of DBT involved recruiting women who had lesions that were found on 2D screening mammography. In those early years, with limited DBT systems, the only way to have a DBT study was if you had a 2D screening mammogram and DBT was used, “diagnostically”, to analyze the lesions found by 2D screening. Consequently, it was impossible for DBT to detect cancers not seen on 2D mammography. Obviously, studies of patients selected by their 2D screening studies, made it impossible for DBT to be more sensitive than 2D mammography. It also made it impossible for these “diagnostic” DBT studies to reduce the recall rate since DBT was not being used for screening. Infact, DBT is not very effective when only used diagnostically.As the technology proliferated, more accurate comparisons of DBT to 2D mammography involved large numbers of women who had been screened with 2D mammography over a period of time and then compared to large numbers of women who were, subsequently (in later years),screened using DBT. Virtually all of these comparisons showed that DBT detected more cancers per 1000 women than 2D mammography. Given the capability of DBT, this is not surprising, but cohort studies of this type suffer from the possibility that by comparing differing cohorts of women, the groups may haved iffering prior probabilities of having breast cancer making them not actually comparable. An ideal study would be to obtain DBT on all women which should include FFDM MLO and CC projections as well (see below). The FFDM images should be interpreted first with no access to the planar information from the DBT collection. The studies should then bereread using the planar images along with the FFDM images. Since anyone who has used DBT for screening knows that there are cancers that are only revealed on the planar images, it is certain that when the same women are screened with DBT, the cancer detection rate is higher than FFDM alone. Since there are no, scientifically derived,data that show that there are any invasive cancers that do not have lethal potential, it is likely that the earlier detection and treatment of cancers found by DBT alone will lead to additional cures. FULL EXPOSURE FFDM IS PART OF DBT SCREENINGInitially we had hoped that all that would be needed for screening was DBT in the MLO projection. However, we soon learned that there are some cancers that, because of their geometry, can only be seen on CC projections. Consequently, DBT should be obtained in both the MLO and CC projections. We also learned that it would be very time consuming to compare the present year’s DBT to prior studies using just the planar images. Furthermore, despite the fact that once clustered calcifications are detected, their morphology is better seen on DBT planes ([11]),detecting a cluster using just the planar information can be problematic. As you “page through” a cluster with one calcification in the first plane, two in the second, none in the third, two in the fourth, etc., your brain may not register it as a cluster, and you run the risk of missing them. It turns out that the 2D mammographic images allow calcifications to be, efficiently detected. Consequently, it is important to obtain FFDM 2D images with every DBT screening study. CAN SYNTHETIC 2D IMAGES REPLACE FFDM ASPART OF THE DBT SCREENING STUDY?In 2000 we recognized that the planar images could be reassembled to provide a “synthetic” maximum intensity projection that could simulate 2D FFDM and we thought this might replace the need for full exposure 2D FFDM. However,as of now, no one has developed synthetic images that perfectly reproduce the information available on the full exposure studies. The present-day algorithms either create false calcifications, or eliminate some important calcifications. Sincer adiation risk to the breast is so low and probably nonexistent for women ages 40 and over, FFDM images should be acquired as part of screening DBT studies. This being the case, in order to replace full exposure FFDM 2D images using synthetic images we cannot allow a single cancer to be missed by the synthetic images. In other words, given the low priorprobability of cancer in a population, missing even a single cancer by switching to synthetic images instead of obtaining FFDM, when the cancer detection rate is 4-8/1000, is not acceptable. Since radiation risk to the breast is extremely low and possibly even nonexistent for women ages 40 and over, until synthetic images are perfect renditions of 2D FFDM, the full exposure images should be collected with the planar information.Given that the radiologist must review many more images when using DBT, it is not surprising that it takes longer to interpret a DBT study. However, DBT also reduces the recall rate. This means that there will be less time spent in performing diagnostic studies which take up agreat deal of time. No one, to my knowledge, has done an accurate accounting, but I suspect that the time saved in fewer diagnostic evaluations for women recalled from screening, will help offset the longer interpretation times, especially in light of the increased cancer detection rates.
[1] Skaane P, Gullien R, Bjørndal H et al. Digital breast tomosynthesis(DBT): initial experience in a clinical setting. Acta Radiol. 2012 Jun1;53(5):524-9.[2] Skaane P, Bandos AI, Gullien R, et al. Comparison of digital mammography alone and digital mammography plus tomosynthesis in apopulation-based screening program. Radiology 2013; 267:47–56[3] Ciatto S,Houssami N, Bernardi D, et al. Integration of 3D digital mammography with tomosynthesis for population breast-cancer screening (STORM): a prospective comparison study. Lancet Oncol 2013; 14:583–589[4] Lång K, Andersson I, Rosso A, Tingberg A, Timberg P, Zackrisson S.Performance of one-view breast tomosynthesis as a stand-alone breast cancer screening modality: results from the Malmö Breast Tomosynthesis ScreeningTrial, a population-based study. Eur Radiol. 2015 May 1. [Epub ahead of print]PubMed PMID: 25929946[5] Haas BM, Kalra V, Geisel J, Raghu M, Durand M, Philpotts LE.Comparison of tomosynthesis plus digital mammography and digital mammography alone for breast cancer screening. Radiology. 2013 Dec;269(3):694-700[6] Rose SL, Tidwell AL, Bujnoch LJ, Kushwaha AC, Nordmann AS, Sexton RJr. Implementation of breast tomosynthesis in a routine screening practice: an observational study. AJR Am J Roentgenol. 2013 Jun;200(6):1401-8[7] Friedewald SM,Rafferty EA, Rose SL, Durand MA, Plecha DM, Greenberg JS, Hayes MK, Copit DS,Carlson KL, Cink TM, Barke LD, Greer LN, Miller DP, Conant EF. Breast cancer screening using tomosynthesis in combination with digital mammography. JAMA.2014 Jun 25;311(24):2499-507. doi: 10.1001/jama.2014.6095[8] Sharpe RE Jr,Venkataraman S, Phillips J, Dialani V, Fein-Zachary VJ, Prakash S, Slanetz PJ, Mehta TS. Increased CancerDetection Rate and Variations in the Recall Rate Resulting from Implementation of 3D Digital Breast Tomosynthesis into a Population-based Screening Program.Radiology. 2015 Oct 9:142036. [Epub ahead of print][9] E. A. Rafferty, D. Georgian-Smith, D. B. Kopans, R. H. Moore, T.Wu, L. Niklason, D. A. Hall, K. A. McCarthy, E. Yeh, M. M. Staffa, “Comparisonof full-field digital tomosynthesis and conventional two view film screen mammography in lesion detection and assessment of lesion conspicuity,” ARRS annual meeting, 2003[10] E. A. Rafferty, D. Georgian-Smith, D. B. Kopans, R. H. Moore, T.Wu, L. Niklason, D. A. Hall, K. A. McCarthy, E. Yeh, M. M. Staffa, “Evaluation of the call-back rate for screening mammography using full-field digital tomosynthesis versus conventional film-screen mammography,” ARRS annual meeting, 2003.[11] Kopans DB,Gavenonis S, Moore RH, Halpern E. Calcifications In The Breast And Digital Breast Tomosynthesis. The Breast Journal 2011 Nov-Dec;17(6):638-44
