PEPconnect

Risk-adjusted Breast Cancer Screening Strategies (ECR Symposium)

This webinar by Prof.Dr. Luis Pina (Pamplona/Spain), Dr. Ritse Mann (Nijmegen/Netherlands), Prof. Dr. Francesco Sardanelli (Milan/Italy) gives an overview about risk-adjusted breast cancer screening strategies (ECR Symposium).

Due to high data volume of clinical image files, restrictions in the bandwidth may compromise image quality depicted in this webinar.

Good morning ladies and gentlemen. Dear colleagues, welcome to this session about risk and breast cancer screening. Let me introduce myself. I'm doctor Luis Pena from the University of Navarra Clinic in Pamplona, Spain. An I am the chair of this session. This session substitute the originally planned 10th Breast Care Day. And it is a joint symposium from Siemens Healthineers. In cooperation with pair. Not all women are at the same risk level to develop breast cancer in the future. There are high risk women with a likelihood between 25 to 30%. There are intermediate risk women with a likelihood between 15 to 25%. And there are women at the bustle risk with a likelihood lower than 15%. Not all breast cancers are equal. We can divide. The breast cancer, according to the morphological features. This year, yes. An invasive cancers or according to the immuno histochemical patterns. Terminal ABB with solar expression of her two pure her two and triple negative cancers. In this graph you can see how different the breast cancers can be, because some breast cancers grow very fast while other cancers are non progressive or grow very very slowly. And this is when overdiagnosis occur. And there are different imaging modalities. We can use. Conventional mammography. Tomosynthesis ultrasound. Contrast mammography and MRI. It is interesting to divide them into 2 main groups. The morphological techniques such as mammography, davitian, ultrasound, and the Morpho functional techniques. Contrast mammography and MRI. Mom Ography is the choice for breast screening. This modality is good for fatty breasts and good for the detection of DCIS. But definitely it is not good for dense breasts. Not good for the majority of invasive level cancers. And take in mind that overdiagnosis can occur and pre Ferrick lesions can be missed. To examples, one of them showing a typical invasive cancer in a fatty breast, no problem to detect this cancer. However, on the other case, it is impossible to detect the breast cancer becauses completely hidden. Due to the density of the breast. What about Thomas Enthesis? Thomas Synthesis is good for the detection of distortions and speculations, and it is better than mammography for dense breasts. But it is not good for the detection of cancers without speculations. And not good for lesions that are completely surrounded by fibro glandular tissue. Look at these examples. In the first one we can detect subtotal, architectural distortion and invasive cancer. However, in the other example it is impossible to detect. On tomosynthesis the breast cancer be cause it is a very very dense breast. Ultrasound ultrasound is good for dense breasts. It is very good for detection of masses and invasive cancers. An it is also very good for the assessment of populations, but it is not good for fatty breasts. Not good for the detection of DCIS. And some of the main problems of ultrasound are that there are two many false positive results. It's a time consuming technique and operator depending. Some examples dense breast, normal mammography. However, we can detect a breast cancer. But we can also detect too many behind missions. Contrast the morphine. This is a more functional technique. It is good for the evaluation of dense breasts. It is good for the detection of invasive cancers. It can be used. Four intermediate risk women. An it is less expensive than memory. But it is not very good for the detection of DCIS. Not for low risk women. An it is more expensive than conventional memorial. One example of contrast mammography showing a breast cancer. And finally MRI. MRI's good for the detection of invasive cancers. It works in dense breasts. An it is good for high risk women, but it is not as good for this year, yes. There are too many false positive results because it is influenced by the hormones. It is expensive. Usually there is little availability an but spatial resolution of this modality. In the first example, we can see normal. Memoriam, however, Emery was capable of showing an extensive this year yes. However, on the second example. And the strong BPE. Makes impossible to detect a breast cancer. This is one of the limitations of MRI. Ladies and gentlemen, it's a great pleasure for me to introduce you to the next speakers. The first one is Doctor Mann from the Netherlands and the title of the presentation is risk based screening and the potential of artificial intelligence for patient stratification. And the second speaker will be Doctor Francisco, certainly from Milano, Italy, and the title of the presentation strategies of women at average risk. The new guidelines on breast cancer screening and diagnosis from the European Commission Initiative on breast cancer. Welcome in this lecture. I want to talk about risk based screening and potential AI for patient stratification and mainly towards a more screening as that is becoming more and more standard. Mr. Start with screening and basically the premise of screening is that you detect cancers but earlier. Basically, all cancers grow and a certain point they reach a threshold where they become either palpable or otherwise symptomatic. At that point patient usually go to the doctor and get treated screening basically detector counselor earlier before they become symptomatic. And therefore also prevents a number of cases where the cancers become modesta sized. You should realize the gain of survival is not here, but in preventing a number of cases where the cancer actually metastasizes. However, many studies looked at mammography for this and actually did several randomized controlled trials in the 80s and 90s, and they all show the same. You get a relative reduction in breast cancer related mortality in the order of 20 to 25% when you screen with mammography. So earlier detection basically leads to an improved breast cancer specific survival. However, screening is not beneficial for everyone. Women here basically wouldn't have had investors here either, and therefore only know little bit earlier that they have breast cancer. Maybe the therapy would be a little less aggressive, but other than that it doesn't change anything for them. For these women, it doesn't change either. I mean they would have been detected here with metastasis, but they already haven't tested this here and one are very likely to die from the disease. So the game is really only for a fraction of people. Obviously the goal is to increase this fraction as much as we can. Basically by finding Kansas earlier and earlier. Now let's have a look at the screening population. This population has never been screened. And therefore. Not there are many cancers there, but none of them have actually reached the symptomatic threshold. Now we screen with mammography and we find a number of the Kansas present in the population. These are treated, but there are obviously still a lot of cancers that are in that population that do not reach the mammography threshold should really be aware of that in the interval. Subsequently, these cancers grow and some new ones occur as well. And if we wait long enough, some become. Once again, symptomatic, and therefore an interval cancer. Basically, in the forecast gets treated. And subsequently we do another screening round and find the ones that have actually grown towards the mammography threshold. Basically these two. And so this is so called instant screen. You see that the yield of an instant screen is definitely lower than the yield of a preference screen we were discussing before. What is the high risk population? Well, most of the time we just talk about a population with more cancers. Much more Kansas. These women basically get screened in the same way you find many more cancers in the first screening evaluation, and subsequently also more interval Kansas in the interval. That's sort of painful, so we want to prevent this, and one of the things that we can do for that is basically just simply shorten screening interval. If we do that, you get to see that one of these cancers still turns up as an interval cancer, but since the other two grew less fast, they did not become symptomatic and are therefore detectives at the next. Incident screen. Obviously that leads to a lower yield per screening around, but since arm anymore Kansas, you still find quite a lot. In women with Drucker one and broker two mutations, cancers also develop a lot earlier than in the average risk population and therefore another remedy to prevent interval cancers or cancers not detected by screening is simply by starting the screen program earlier. If you do that, you can obviously improve the quality of your screening program. Next thing is that you should choose the right modality Mogra Fi in a young population is very often not a very good idea due to the simple fact that these breasts are commonly very very dense. And if you have fairly dense breast, the yields of mammography screening is simply less. This is best shown in this graph where he gets to see the screening yielding a fatty breast. Now we replace that with a dense breast and this is the net effect. You still find Kansas with screening. However, the gain is less to the simple fact that you only detect them when there are much closer to the symptomatic threshold. So the remedy #3 is to use supplemental imaging modalities and MRI is by far the best. You could use ultrasound, but the amount of additional Kansas you find is a lot less. Supplemental cancer detection rate after a negative mammogram in essence is about 15 to 20 per thousand. The yield of mammography itself, somewhere in the order of 6 to 7 per thousand in this population. So this is an increase of about 300%. And in a graph, this looks like this. We basically screen this additional population and therefore eradicates a lot more cancers in women with dense breast biema screening. We tried it also in the average risk population with off women with very dense press. There was quite recently published in this publication that is about screening effort. Women with very dense breasts with press more and this is the game we have. 4780. Three women screamed with them are here. 79 of those had a more detected cancer. That comes, however, at a price. There are obvious false positive 78 per thousand for positive predicted value of 26%. I'm. This was the incident screen, supplemental detection rate of 16.5 per thousand. However, these are treated and in the interval. Some new Kansas might occur, some cancers grow, but you wouldn't expect all these cancers to reappear. Consequently, in subsequent rounds prevalence screens, the amount of cancers detected should really decrease, and Luckily for us, it also did actually, the added cancer detection rate is only 6000 in the second round, or an also the number of false positives tremendously decreases. Consequently, the PP for biopsy stays more or less the same. Does density, however, really matter when we select patients for a more screening? Yes, the risk of breast cancer increases and the risk of missing Kansas increases, however. Mr France cancelled just a lot earlier. In this trial, we evaluated women with a familial risk of 20 to 50% for the development of breast cancer in their lives. And this was once again randomized controlled trial door far smaller than the dense trial. But you get to see that in the MRI arm, many more cancers were detected than in a mammography arm. The supplemental detection rate is a lot less, however, that comes because this is sort of a mix of an incident in prevalence screen. Not all these women were screaming naive to start with. Many of them have already had prior Amoss before they actually entered study. When you get to see, however, is more or less the same. Cancers detected with them are are a lot smaller. And a lot more often not negative. I would say that's a good indication that we actually increase the life expectancy. However, if we look at the different density categories, you get to see the amount of cancers detected over the density categories here, and this basically reflects the amount of patients in each of the density categories. Be cause this is what is most common, and these are somewhat uncommon in us since there is no real effective density here. Amar just improves detection in all density categories. Why is that while simply have a look at this credit craft, this is the dense breast and the screening of mammography. Now we go back to the fatty breast an the effective screening and mammography. And then we ask the Mr. And in essence. The gain is not so much because the gain of mammography is only there. An Mr takes away many of the Kansas much earlier than you can actually find them with mammography. So in essence, density alone is also not a very good selector for. Screening patients with them are. And one of the things that you should ask yourselves if you do that, if you screen women with them are should you still perform a mammogram? These studies all showed an increase in sensitivity with the full protocol including Mr and mammography. However, the sensitivity increases really small, ranging from zero to 9% compared to the 300% that we get with them are and averages about 5%. You also pay penalty by adding mammography, which is a reduction in specificity of .52%. And in screen, that's actually quite a lot. So mammography is the supplemental technique. When women are already screened with Amar and, for example, under 40, you should screen about 2000 women with mammography to find one additional cancer, and I honestly would say that's not worth doing it at all. The types of cancer you detect with mammography only are predominantly low grade cancers. an A lot of DCIS, so honestly again of mammography is not so big. And one further thing there is that breast cancer do not grow all at the same speed. Some cancers are really aggressive, grow really fast, others grow really slow and might actually never cause symptoms at all. If you look at the sensitivity profiles of mammography and Mr. Then basically Amar finds more of the aggressive Kansas, whereas mammography becomes less sensitive. The more aggressive the cancer is. However, a more unfortunately also detects more low grade ECS than mammography. And that's something that we should somehow still prevent. And nevertheless, if you plot that on top of the growth rates, then basically the gain of Mr. Is more. Here was the gain of mammography is more there. And there are obvious downsides of screening, particularly screening with Mr. And one of the things that we find quite a lot of false positives. The other one is cost. Let's start with the false positives. There are definitely more license in the breast and we see on the mammogram, and some of them are actually visible, only Amar, which means that we really need to screen with the risk of the patients in our in the back of our mind. If we treat a patient with a family history of breast cancer the same way as a patient with a BRC A1 rotation. And the positive predicted value for biopsy might decrease substantially, and therefore in this part women you might just close one eye when he fell awaiting the Amar and just hit on the things that are really, really suspicious rather than a broker. One rotation carriers you might hit on every single enhancing focus. However, this basically comes down to shared decision making. You should explain the risks to the patient and let them actually choose which method of screening they would actually prefer. The costs are more on our side. The patients can't really tell anything about costs, and actually if you look at the data from the fungus study the cost in the MRI arm, total costs are much higher here and there, and this supplemental cost is mainly explained by the simple fact that. I'm a screening itself is much more expensive than mammography screening because the treatment costs here are somewhat higher than here, but per cancer detected, there actually lowered the simple fact that cases here are smaller and more often not negative. So we have a sparsity model. We can't afford anymore for every patient, and therefore we need to select patients. Currently we do that based upon these things apart from the genetic tests. These are family history evaluations and you can see that they predict the risk of developing breast cancer to a certain extent, but are not very good, especially not MBR. See a negative patients. I'm still this is what we use and in essence we tried to screen women with a moderate get breast cancer young and it got breast cancer frequent. We can do a bit better by including breast density, and if we do that, you see that the tire kuzyk model really improves and here you get to see the density alone or at least imaging information alone actually already outperforms the direct kuzyk model. But if you combine both you do best. So currently these kind of models basically should incorporate both the image features and the family history and so on to optimally select patients. Then you get to a level where you select patient for a more screen on relative frequent presence of breast cancer and poor performance of other screening tests. The next step would be to actually use AI a bit further, not only to include density, but actually to include the AI reads to actually find those women that have abnormalities that you really would like to assess with them are here. Both screening examinations are rated negative. And honestly, I was likely still the same. However, in 2017 she was diagnosed with bilateral breast cancer. An honestly looking back. In retrospect, you might have picked up this leyshon already earlier. The cat system actually did, but. You cannot really dive into everything the cat system hits. So instead of just taking this core, you might use the classification score that's given via CAT system to a mammogram to actually figure out how likely it is that that patient has into for cancer and one of the things that you could use for that is the transparent score that divides mammograms in 10 percentile score. So 10% of all mammograms are here 10% of all the mammograms are there. And once that are scored, Tim are deemed most likely to harbor cancer. If you would select those women forum or screening, we would find ourselves more than 40% of all the interval Kansas and has a far stronger pre selection method then density alone and so in the future we might actually select patients on the likelihood that they'd develop into fall Kansas. Um, one further thing is obviously. I would select them based on the type of cancer that someone might develop, but we should realize that that is something that we cannot currently do, because that's basically based on the genetic profile of the patients. I really hope, however, that this will be the future and will only find those cancers that really matter to our patients. Thanks, love for your attention and I'm obviously open to questions. Good morning everybody. Thank you. Wish for the introduction. Now we have to. To face the difficult task of evaluate what the. It happened in the field of the monthly ship in Airy. Balance that are have the task to traduce to translate. The scientific work. The evidence in that region in terms of guidelines at the level of the European Union so. This was not an easy task so took a lot of time becausw. Initiative of the European Commission on Breast Cancer started in 2014 with public calls. Looking for expert, able to compose the panels that were too. We are two partners. One was the group who developed. We should develop the guidelines who is developing the guidelines because they are changing over the time and they updated. And the group of the Quality Assurance scheme development. He had the honor to be elected as a vice president of the Quality Assurance Scheme Group and in some way also give a contribution. I gave a contribution to the group which developed the guideline. Who is developing the guidelines? This work provided the 1st paper published on the Indiana's of Internal Medicine in July. 2009 so one year ago. Explaining which was the method. I have to confess that it was quite complicated in terms of not only selection of the panel. A systematic review of the evidence that was updated to sell 2006 and then two 2018 including human human women advocates as full voting members. ABA big work about conflict of interest. The use of grade evidence to determine the frameworks to the decision, and. A bigger fault for defining recommendation on outcomes relevant to women and rating also the certainty of the evidence that you will see is not so high and also a big effort also to to get the feedbacks from the shake orders. The 2nd paper, published in November 2009, have the first only the first publication of the synopsis of the European Guidelines, and as you probably know, these people open. They hope to beat cause the behavior of the Guideline group development was quite conservative. In fact, as you can see on the slide. Um suggestion was against implementing digital mammography screening from 40 to 44 year wage. Was it was? If ever of digital mammography screening on 45 to 49 favor with a strong category of suggestion from 50 to 69. In favor from 17 so 74. However, in terms of interval that was one of the questions that you spinner proposed to ask the introduction, Gigi Group voted against annual screening from 50 to 69 in five or be on your screening over three Inal from 50 to 69 against annual from 70 to 74 and in favor of three annual from 70 to 74. You can see there is also. Additional recommendation in favor. Digital mammography alone over to synthesis. But this was changed subsequently and also a conditional from 50 to 69 in favor of digital mammography alone. Over almost into this blast digital mammography. So not considering the availability of synthetic images. Uh. In addition, we had a conditional recommendation against the automatic Brazil to sound against handheld ultrasound and against MRI screening in the village risk women. But Fortunately, at least we had a conditional. Informal recommendation in favor of digital restroom. A synthesis for diagnosis for us. So it means for assessment for symptomatic women or for women who were recovered after first screening. Sure. Vehicle European Commission is of breast cancer accepted. The use of tomosynthesis in the case of. Systematically, then, and for recalled for women. Record for speeches mission at mammography screening. And this was a good point to make you also be cause open. Also the acquisition of Mammography of Mammographic unit. Updated with to synthesis all over European countries are probably also outside European countries. However. This the news, the good news of this work that believe me, was quite difficult to do, especially cause the logics the logic of the panels was to include a lot of people and in some situation the really expert, especially radiological topics, was very more fraction of the people that voted in the panel. But this was the general rule. Of the European Commission initiative. But the good news is that the. The group in general was open to updating these guidelines. In any case, when new evidence is available and this is the future from this guideline so. This is a work in progress and the radiological community has to work in order to support the adoption and the evaluation of the new evidence for changing the catalyze and the news were recently sent a couple of months ago on the website of the European Commission Initiative of Breast Cancer. Now I will show you what are these needs, so the first is that tomosynthesis was an early acepted. To be used instead of digital mammography of simple digital mammography, you see a very coaches behavior of the Commission of the Panel of the Expert conditional recommendation. Very low certainty of the evidence. But at the end, DT is permitted to be used instead of digital mammography. In general screening program. This is a very good news, and I think we we publish these again, at least with the letter of the. Make sure or or a completely new paper. Better, the suggestion was not to use both DVT and digital mammography. So exploiting the advantage of 2D synthetic views, there are many aspects I have no time to enter it into days. But you can see these aspects in the on the website. Hi, main consideration from the implementation and policymaking of tomosynthesis, evaluating the variability of the quality of DVT machine. Because the machines are not all the same. Adopting quality assurance standards so using synthesized to the imaging and and also monitoring the screening to distinguish the close test done with DBT and don't test done with only digital mammography. Also, the group defined list of research priorities. That means the distribution of tumor grade biology. Prognostic measures the DBT. Additional detected cancer to clarify over the full potential of a bag. You know this, and this is more more. Most important, in my view, to evaluate the impact on interact cancer incidence stage of breast cancer and detection and mortality reduction at also cost effectiveness of the screening program, including synthesis. So. The vice against to combine the double use of digital mammography and DBT without exploiting the advantage of synthesized image was confirmed. So this wasn't changed the updating. So finally we can say that this update. Introduce the news of. Lowinger adoption all over European countries of Thomas Synthesis as first line examination for screening and these think is a signficant, irrelevant advantage. There are also research priorities adopted, but I think you can find these details. In the in the website I want only two highlight that additional research should also set the computer between DBT and generally possible to sound for display screen. So in order to have evidence. For defining what is the best a bit, the best modality. The best approach using DBT or digital mammography plus and also. Other updated interesting where the recommendation in favor of use clip parking after the vacuum assisted biopsy or middle core biopsy for surgical therapy planning patient with breast cancer lesion. So this is. Quite trivial for us, but I think that it's important looking at the entire spectrum of countries in the European, the European level that there is a an official recommendation in favor of the user clip marking. Even though there are consideration for rotation policy making and also model monitoring and evaluation. With research priority also regarding the effect of the of the clip marking and and so on. Another GNU recommendation, to be honest, quite debatable, is that in women with histological confirmed doctor carcinomas item. The guidance Development Group suggest not to use additional market treasonous imaging for proper ative planning, and by also discussed with the group that there are some inconsistency in these recommendation, especially looking the fact that TG recommends research on abbreviated protocol is an important issue, but I think that is not treatable. In particular for the staging of this year. Yes, because we know that. Part of this year, yes. Cain announcer in Adilet time. So to user ability protocol could be something not useful. Also from the viewpoint of results. But this is something that could be. Updating the changed in the future while. New way that was also unexpected, but I think that is a good news at yet. Becausw in some way, giving as defined in Indiana. Indications for MRI in staging. Invasive cancer so. The GDG recommended the use of consciousness and mammography over my to assist in surgical planning of invasive cancer. This is required under find, in my opinion, still undefined recommendation, but is a big opening to contrast a nested images or what with spina defined the more full function and imaging in the staging of already diagnosed breast cancer. So it's relevant that they wrote women with respect that cannot have any. Right, so they implicitly accept that MRI could be the choice in this case, but you know. The May General view of this work is that this group of this big effort of the European Commission is that we have to work among them in order to to convince what is radiological viewpoint and you have to consider that in general the European Commission interval breast cancer are working progress progress, so there is room for improvement room for change of and updating of the guidelines. At a in the moment we are able to provide new evidence and this is a good perspective in my view. So what about the future, already Rich Man is explaining? Does what is part of the future. We may view the future is we have to think about that while currently with the string extremely high risk, high risk and average risk. Thinking about prophylactic mastectomy, MRI or mammography as different tool to face this problem, but probably in the future as already reads told us we have to think more stratified. Situation with the stream Lee. High risk, high risk with mercy to and other genes. Intermediate risk, average risk and low risk. Probably with can think about. Prolonging the interval or or finally also not to screen when the risk is fairly very low. If the Epidemiology will be able to give us models. Clearly able to also using artificial intelligence, telling the human that the risk is very low. And there is increasing new evidence. I only put three important words. One was a despressed that already reads this. One of the authors explained to us. He was already already complete in spring to result of this work. I want only to highlight that one problem is that the aceptance rate of MRI is. Near to 60% so this could be a problem if we want to enlarge these two, the majority of the women. The role of aggregated protocols has already published in the JAMA by Christopher Comstock and coworkers and also the possible role of dual energy contrast analysis. Digital mammography. As these. Important work published by the Memorial Sloan Kettering Group showed us with an important result. Also in terms of specificity of consciousness. As Rich told as the future of radiology, and in particular of breast radiologists that you know, is one of the. Three source of the future as buyers story showed us because no, we see every month knew rats coming to the radiology world following the guide following these periods of the Pirates. The same is for rich and intelligence. And we are now in the years in which the artificial intelligence can change the world of medicine and the world of breast imaging and also the world of rest. Cancer screening. It was due to the fact that the machine were able to go over the performance of humans in 2014 to 2015. Don't people showing that in reading? Mammography screening good AI alone could be better than radiologist. Plus AI not only better than regards, but better than plus AI in the detection. In the terms of area under the covers rock analysist and this could be. This is a perspective you have to take into account, starting to think that there are repetitive tasks that machine can do better than us. But there are clinical task that only humans radiologists. Are able to do. These were published the. At the beginning of this year, in the natural with one of our. Portal person, next president of the user be Fiona Gilbert between the authors. ISM is a very important work with a lot of members not. A lot of results I cannot summarize in this talk because I have to stay in time, but I want to show you that. If you compare he area under the cool of six readers. With an interval between 0.6 and 0.68. The AI showed at 0.74. That means that there is no room from improvement, but also show that there is sure. Difference between human readers and the AI reading, screening. And there are three important impact on screening that we have to embrace as an improvement for our work and for the women. The 1st. Is to omit the second reader with a combination of human and machine, resulting in performance equivalent at least equivalent to that of the traditional double reading process, but saving 88% of the effort of the second you Marie. This track on these immediate feedback for negative cases and for water results from this work. AI is much, much, much, much better than humans and also immediate feedback for positive cases in which again we can peer priority size 3040% of Kansas, Kansas cases with the AI again much better than humans. Obviously this resulted to be confirmed. But I think we have to start to think a different organization, which, for example we will be able to give an immediate feedback to women. So increasing the accept and the quality, the quality perception of women to the screening. And this is my final message allowing radiologists to have more time to be physicians to be clinicians to be people speaking with women and less time spent only for reading. Mobile graphs thank you for your attention. Thank you very much for your attention. It's been a pleasure to be here with all of you. And they will see you. Next year in Vienna by surely surely by.

Nature I V01577 | 2 January 2020 Changes in cancer detection and false-positive recall in EuroR•an Commission EU Sctence Hub > ECIBC European breast cancer guidelines EuroR•an Commilssion EU Sctence Hub > ECIBC European breast cancer guidelines EuroR•an Comrnussion > EU Sctence Hub > ECIBC > European breast cancer g'Me11nes EuroR-an Comnussion EU Science Hub ECIBC European breast cancer g'Mde11nes Comnussion EU Sctence Hub ECIBC > European breast cancer g'Me11nes Ire Commission > Science Hub > ECIBC > European breast eanc•r > Science Hub > ECIBC > European Comnussion > Science Hub > ECIBC > European breast cancer > Science Hub > ECIBC > European breast > Science Hub > ECIBC > European breast cancer guideunas Eu Science Hub > ECIBC > European breast Science Hub > ECIBC > European breast cancer International evaluation ofan Al system for Al (test 1) CLINICAL GUIDELINE Annals RESEARCH AND REPORTING METHODS Annals of Medicine RESEARCH AND REPORTING METHODS of RESEARCH AND REPORTING METHODS RESEARCH AND REPORTING METHODS PP-GAD-ALL-014g-1 PP-GAD-ALL-0149-1 European Congress of Radiology Screening in women with high breast density: Digital breast Towards the treatment of breast cancer: The use of clip- )logv Annals of Internal Medicine Annals of Intårnal Medicine ECIBC recommendations on organized screening of average-risk asymptomatic women NEW ENGIÄND NEW ENGLAND NEW Screening in women with high breast density Should Screening Using DBT (Including Synthesized Should Screening Using D)BT (Including Synthesized Planning surgical treatment: Contrast-enhanced spectral Tomosynthesis use in screening .810 1-0 1. 1.0 A new concept: CE Breast 1.0 .810 WVhat Test Should Be Used for Diagnosis in What Test Should Be Used for Diagnosis in What Test Should Be Used for Diagnosis in Average- s: Planning surgical treatment: Magnetic resonance imaging mammography using artificial intelligence: a retrospective, The GDG concluded that DBT probably confers a The GDG concluded that DB T probably confers a European ÉtR>oiö The resources needed to adopt DBT plus digital JOURNAL MEDICINE SIEMENS SIEMENS .. the future? breast cancer screening Breast cancer Gurdeline Artificial Intelligence (ML/DL) Individual BC risk assessment: the future? Individual BC risk assessment ECR2020 Risk Women Recalled Because ofSuspicious Lesions Risk Women Recalled Because of Suspicious Lesions 2-Dimensional Images) in Addition to Digital 2-l)imensional Images) in Addition to Digital Individual BC risk assessment: 95 Commission Commarission Remedy no. 3: Supplemental imaging Al on prior mammograms of interval Types of cancer detected with Development of breast cancer in different US/ABUS tomosynthesis alone Sensitivity profile of mammography What happens in the screening Remedy no. 1: shortening the Next Speakers multi reader study Adjsk-based screening ENéxtlSpeakers A Deep Learning Mammography-based Model for The ECIBC is committed mammography were considered large because of the BAYER net health benefit, and although the DBT device is CEDE-M Methods for Development of the European Commission Initiative on marking mearking What is the effect of supplemental MRI A Overall Risk-based screening and the Does density really matter? Early cancer detection is not Healthineers European guidelines on breast cancer screening and Including breast density Al impact on screening at Mammography Screening ? at Mammography Screening? mammography Mammography Women with extremely dense breasts Recommendation Mammography Versus Digital Mammography Alone Age Category COE coE ECIBC recommendations on organized screening of average-risk asymptomatic women Breast Cancer Screening and Diagnosis: A Synopsis of the European B BRCA-t*gative B BRCA-f*gative Nature I V01577 | 2 January 2020 Nature I V01577 1 2 January 2020 ECIBC 2nd round DENSE examinations higher costs of the machines and the necessary human 08 AUC = 0.740 AUC = = 0.740 0.740 —a The premise of screening Thank you for your attention! What defines a high risk population? Sparsity model - Patient selection Results of early screening trials Incremental value of mammography The pres of screening Downsides of screening Effect of screening on survival Effect of density Patient selection - future A naive screening population A prevalence screen Therapy Screening interval Screening with Choosing the right modality Breast cancer growth Breast cancer diagnosis Patient selection - new Breast cancer Gurdeline Patient selection - classic MRI + Mammography? breast cancer screening A Overall A Overall population A Overall populat.ion A Overall popuution Improved Breast Cancer Risk Prediction B participants B BRCA•negative participants Screening Imaging - Imaging = MRI or CEM - MRI or CEM Costs Recommendation 15. The EClBC's GDG suggests Recommendation 15. The ECJBC's GDG suggests False positives Remedy no. 2: starting earlier Number of women needed Mammo:MRl randomization Mammo:MRl randomization = Mammo:.MRl randomization much more expensive than the equipment needed for An incidence screen Supplemental detection: Effect of density Be Used for Early Detection of Breast Cancer in Breast Cancer Guidelines Breast cancer screening European Commission Science Hub ECIBC European Commission Science Hub > ECIBC European Commission Hub ECIBC Mammography p value MRI group (.31) Overdiagnosis diagnosis 0-8 (Al) resources (72), For instance, radiologists' reading time to reviewing and updating the recommendations resources (72). For instance, radiologists' reading time Recommendation Recommendations Acceptance of MRI = 59% — Acceptance of MRI 59% Acceptance of MRI 59% — Acceptance of MRI = 59% using DBT over diagnostic mammography projections "selection based on Considerations for implementation and policy making Subgroup considerations In the context Of an organised screening programme: In the context ct an programme: In the context Of an organlsed screening programme: Breast Guidelines Asymptomatic VVomen? Asymptomatic Women? (n.674) group (n•680) magnification mammography, information for other Summary 40-44 • Against implementing DM screening What does "risk" in breast screening mean? "Al" firstly introduced in 1956 in a modalities "relatively frequent" "relatively frequ{ent" "relatively frequerlt" mammography only mammography risk groups and MRI screening interval screening mean? interval? Interval? UNCHANGED 2013 in women at average risk for breast cancer recalled in women at average risk for breast cancer recafted would at least double by using both techniques (77 to Research priorities For asymptomatic women with an average risk of breast cancer, potential of Al for patient screening in other density categories? Recommendations in the Era of Guideline Transparency Considerations for implementation and policy making Mammography for to develop is the fO' •o the for develop the current standard for screening. to develop the standard for breast to develop the for to develop the fol to current standard for develop current standard for breast to develop standard for develop sundard fol to standard for to develop for to develop for to fol breast •o for aimed to develop NEW the type the type ofecancer the type of cancer beneficial for everyone 27 the standard to develop the standard breast to develop the standard •o the to develop the breast to develop the breast to the standard the standard breasto develop the 'o the breast •o develop the •o current standard breast to develop an standard to develop an standard breast develop an for develop for breast develop an cancers cancers" 2-year IC rate 2+fear IC rate oa_z NEW ENGLAND VIENNA MARCH 11-15 0.8 0.6 'R isc 'Risc Recommendation 1. For asymptomatic women Recommendation I. For asymptomatic women In the context of an organised screening programme, for asymptomatic URI • 926 7.6 the likelihood of 500 (46) 2015 Subgroup considerations .2 22 R isc EtD criteria also generally favored usin DBT for dia - EtD criteria also generally favored usin DB T for dia - Mean at 49-4 (7-1) for suspicious lesions at mammography screening for suspicious lesions at mammoqraphy screening Extremely high risk NEW could -4 research project to solve kinds of model Tyrer•Cuzick model with densit Tyrer•Curick with Tyrer•Cuzick model with &nsity model with The and 5 2019 as required when new evidence becomes available 191 seconds) compared with digital mammography '4 about • / Il / Il / J Il / I Il l' Il I I Il Il MRI-group MRI-group 2.5* MRI-group 2.5%. AUC -0.582-0.681 AUC • for asymptomatic women Recommendation to screen with supplemental and utility in very extendied, suspect multifocal. or multiple disease cculd nat utility in very extendied, suspect multifocal, or multiple disease could not utility in very extendied, suspect multifocal. or multiple disease could not utility in very exterxied, suspect multifocal, ar multiple disease could not utility in very extended, suspect multifocal. or multiple disease could nat utility in very exterxied, suspect multifocal, or multiple disease cculd not utility in very exterxied, suspect multifocal. ar multiple disease cculd nat utility in very exterxied, suspect multifocal. or multiple disease could not utility in very exterxied, suspect multifocal, or multiple disease could not utility in very extendied, suspect multifocal. or multiple disease cculd not Bolger J. Schünemann, MD, PhD, MSc: Donata Lerda, PhD; Nadya Dimitrova, PhD; Pablo Alonso-Coello. MD. PhD; Holger J. Schünemann, MD, PhD. MSc: Donat. Lerda, PhD; Nadya Dimitrova, PhD; Pablo MD. PhD; Bolger J. Schünemann, MD, PhD. MSc: Donat. Lerda, PhD; Nadya Dimitrova, PhD: Pablo MD. PhD; Bolger J. Schünemann, MD, PhD, MSc: Donata Lerda, PhD; Nadya Dimitrova, PhD: Pablo MD. PhD; Bolger J. Schünemann, MD, PhD, MSc; Donata Lerda, PhD; Nadya Dimitrova, PhD: Pablo MD. PhD; Bolger J. Schünemann, MD, PhD, MSc: Donata Lerda, PhD; Nadya Dimitrova, PhD: Pablo MD, PhD; Bolger J. Schünemann. MD, PhD, MSc: Donata Lerda, PhD; Nadya Dimitrova, PhD; Pablo MD. PhD; Bolger J. Schünemann, MD, PhD, MSc: Donata Lerda, PhD; Nadya Dimitrova, PhD; Pablo MD. PhD; Bolger J. Schünemann, MD, PhD, MSc: Donat. Lerda, PhD; Nadya Dimitrova, PhD: Pablo MD. PhD; Bolger J. Schünemann. MD, PhD. MSc: Donat* Lerda, PhD; Nadya Dimitrova, PhD; Pablo MD. PhD; J. Schünemann, MD, PhD, MSc: Donata Lerda, PhD; Nadya Dimitrova, PhD: Pablo MD. PhD; with an average risk for breast cancer, the EClBCs GDG with an average risk for breast cancer; the ECIBCs GDG Ill Tang, An, al. "Caznadian Tang, An, al. Tang. An, al. Tang, An, al. • •Canadian Tang, An, al. "Canadian An, al. An, al. "Canadian 'l Intermediate palpable possibly and palpable and palpable therapy and palpable candidate and Figure 2. Real life possibly and possibly candidate and possibly therapy and NEW Holger J. Schünemann, MD, PhD, MSc; Donata Lerda, PhD; Cecily Quinn, MD; Markus Follmann, MD. MPH, MSc; Holger J. Schünemann, MD, PhD, MSG: Donata Lerda. PhD: Cecily Quinn, MD: Markus Follmann, MD. MPH, MSc: Holger J. Schünemann, MD, PhD, MSc; Donata Lerda. PhD: Cecily Quinn, MD; Markus Follmann, MD. MPH, MSc; Holger J. Schünemann, MD, PhD, MSG: Donata Lerda. PhD: Cecily Quinn, MD; Markus Follmann, MD. MPH, MSc; Screening interval cancers Average Il DISCLOSURES the EClBC's Guidelines Development Group (GDG) suggests not The European Commission Initiative on Breast Cancer (ECIBC) Not all women are at the same risk level to In favor of DM screening Cancers diagnosed during s tudy Sensitivity increase: Not all breast cancers are equal Cancers diagnosed during study NEW that someone might Low (conditional recommendation, moderate certainty of Conditional recommendation for the intervention Variability in the quality of DBT machines and their methods of capturing images. 7070 7020 7060 70m Risk-adjusted breast Mammo-group 5.0%. (p 0.001) Mammo-group (p 0.001) dense breasn Is critical. Incaut*s Informaton about limitatZ•ns cc denge breasts Is ctftfcal about llmirathng denge breasts Is Informatk•s• about cc dense breasts Is ctitlcal, Thi* Includes Informatkn about limitatkne cc dense breasts Is ctltlcal. InformatkM about limitate•nscc dense breasts Is ctltlcal. about limitate•nscc dense breasn Is ctltlcal. Informaton about dense breasts Is This Induces ab0L•1 limitatkns cc RESEARCH — Mammo-group 5.0%. (p 0.001) — Mammo-group 5.0%. (p < 0.001) Mammo-group 5.0%. (p < 0.001) 1392 90 High Research priorities SIEMENS Axel MD: MD; Markus MD. MPH. MSG Robert Mansel. MD: Francesco MD; Axel MD: Quinn, MD; Markus MD. MPH. MSG Robert MD: MD; Axel MD: Quinn, MD: Markus MD. MPH. MSG Robert Mansel. MD: Francesco MD; Axel Grswinghott. MD: Quinn, MD; Markus MD. MPH. MSG Robert Mansel. MD: MD; Axel MD: MD; Markus MD. MPH. MSG Robert MD: MD; Axel Grswinghott. MD: Cecily Quinn, MD; Markus MD. MPH. MSG Robert Mansel. MD: MD; Axel MD: Cecily Quinn, MD; Markus MD. MPH. MSG Robert Mans-el. MD: MD; Axel MD: Quinn, MD; Markus MD. MPH. MSG Robert Mans-el. MD: MD; Axel MD: Cecily Quinn, MD: Markus MD. MPH. MSG Robert Mansel. MD; MD: Axel MD: Cecilyt. MD: Quinn, MD; Markus MD. MPH. MSG Robert Mansel. MD: MD; Axel MD: Quinn, MD; Markus MD. MPH. MSG Robert Mans-el. MD: Francesco MD; Axel MD: Quinn, MD: Markus MD. MPH. MSG Robert Mans-el. MD: MD; Axel MD: Quinn, MD; Markus MD. MPH. MSG Robert Mansel. MD: Francesco MD; Axel MD: Quinn, MD: Markus MD. MPH. MSG Robert MD: Francesco MD; Axel MD: MD: Markus MD. MPH. MSG Robert Mansel. MD: MD; L" v Richard LA v Richard LA v L" v u 'v Richæd L" v Richmd LA v Rich•d UI v Richmd L" v Rich•d u 'v Richard LA v Romeo Richard Richard Rich•d alone (33 to 67 seconds) (56, 59, 73). Although the nosis in women reca e or suspicious esions at mam- cf developing interval Screening interval problems now reserved for humans". Association Indiciation aided 2) k, In FUTURE non.palpable COG 50th types non-palpable GOG types non-palpable GOG non-palpable GOG 50th types and non.palpable GOG types non•palpable GOG types non•palpable GOG types and non-palpable GOG 5cth types non.palpable GOG non-palpable GOG 50ths non-palpable COG non-palpable GOG 50th types 'mnvn won 'mnvn 'mnvn •n 'mnvn •on 'mnvn an 'mnvn wen •n Al aided (test 2) Al aided 2) Al 2) won won •n (test 2) 2) suggests screening with digital mammography alone suggests. screening with digital mammography alone women with an average risk of breast cancer, the EClBC's Guidelines 0-6 be investigated, but the rationale better morphological been be investigated. but the rationale better morphological been be investigated, but the rationale better morphological has been Oct 2019 Oct V Oct 2019 •n (6 readers) (9 readers) (6 • mammogtaphl€ bleastdens,'ty With high mammographic breast density Pablo Alonso-Coello, MD, PhD; Paolo Giorgi Rossi, PhD; Annette Lebeau, MD; Lennarth Nyström, PhD; Mireille Broeders, PhD; Pablo Alonso-Coello, MD, PhD; Paolo Giorgi Rossi. PhD; Annette Lebeau, MD; Lennarth Nyström, PhD; Mireille Broeders, PhD; Pablo Alonso-Coello, MD, PhD; Paolo Giorgi Rossi, PhD; Annette Lebeau. MD; Lennarth Nyström. PhD; Mireille Broeders, PhD; Pablo Alonso-Coello, MD, PhD; Paolo Giorgi Rossi. PhD; Annette Lebeau, MD; Lennarth Nyström. PhD; Mireille Broeders, PhD; Pablo Alonso-Coello. MD, PhD; Paolo Giorgi Rossi. PhD; Annette Lebeau. MD; Lennarth Nyström, PhD; Mireille Broeders, PhD; I alone 0.94 0.294 -0.949 (MG-SCR) It Ir In costs during the JULY 15-19 634 (94%) 10 test accuracy data; EtD available at http://bit./y JULY No cancer cancer Node uncertainty about effects ot tanrz•synlhesls. For women dense uncertainty •bout effects ot For women Otth dertse uncertainty about effects ot For women dentse uncertainty about the effects ot •vcmzynlhesls. For women dene about ettectE ot For women derv" "poor performance" Paolo Giorgi Rossi, PhD: Annette Lebeau, MD; Lennarth Nyström. PhD; Miraille Brooders, PhD: Lydia Ioannidou-Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Lebeau, MD; Lonnarth Nyström, PhD; Miroille Brooders. PhD: Lydia Ioannidou-Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Lebeau, MD; Lonnarth Nyström. PhD; Miroille Brooders, PhD: Lydia Ioannidou-Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Leboau, MD; Lennarth Nyström. PhD; Miroille Brooders, PhD: Lydia Ioannidou.Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Lebeau, MD; Lennarth Nyström. PhD; Miraille Brooders, PhD: Lydia Ioannidou.Mouzaka, MD: Paolo Giorgi Rossi, PhD: Annette Lebeau, MD; Lonnarth Nyström, PhD; Miraille Brooders, PhD: Lydia Ioannidou-Mouzaka. MD: Paolo Giorgi Rossi, PhD: Annette Lebeau, MD; Lennarth Nyström, PhD; Miroille Brooders, PhD: Lydia Ioannidou.Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Lobeau, MD; tennarth Nyström, PhD; Miraille Brooders, PhD: Lydia Ioannidou-Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Lebeau, MD; Lennarth Nyström. PhD; Miraille Brooders. PhD: Lydia Ioannidou-Mouzaka, MD: Paolo Giorgi Rossi, PhD: Annette Lebeau, MD; Lonnarth Nyström, PhD; Miroille Brooders, PhD: Lydia Ioannidou.Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Lebeau, MD; Lennarth Nyström. PhD; Miraille Brooders, PhD: Lydia Ioannidou-Mouzaka, MD: Paolo Giorgi Rossi, PhD: Annette Lebeau, MD; Lennarth Nyström, PhD; Mireille Brooders, PhD: Lydia Ioannidou.Mouzaka, MD: Paolo Giorgi Rossi, PhD; Annette Lobeau, MD; Lonnarth Nyström, PhD; Miraille Brooders, PhD: Lydia Ioannidou.Mouzaka, MD: Paolo Giorgi Rossi, PhD: Annette Lobeau, MD; Lennarth Nyström, PhD; Miraille Brooders. PhD: Lydia Ioannidou-Mouzaka, MD: Conditional Indication Recommendation 1. The GDG suggested improved research with higher quality of evidence High risk Category (BRCA2 and cancers CoE 1) Omitting the second reader JOURNAL of MEDICINE There is also a need for research evidence on repeated D3T white paper on artif•iål white paper an artifciål white paper an artifiial white paper on artlf•iål white paper on artificiål white paper an artifriål white paper an artificial white paper an artif•iål white paper on artifciål screening UK stratification r gui- elines GDG could not determine whether using DBT in addi- GDG could not determine whether using DBT in addi. that did that the did rot that did rot that the did thet did that did rd did p < 0.0001 p 0.0001 be Gou: MRI detection rate 16.5%. — MRI detection rate 16.5%. Brest by bc develop" (1 over screening With DDßT in addition to digital mammog- ECISC guidelines Weight using both digital breast tomosvnthesis (DBT) and diaital over screening with DBT in addition to digital mammog- 47 90 considered stronger In these cases. 2nd round using both digital breast tomosynthesis (DBT) and diqital Lydia Ioannidou-Mouzaka, MD; Stephen W. Duffy, BSc, MSc, CStat; Bettina Borisch. MD; Patricia Fitzpatrick, MD; Lydia Ioannidou.Mouzaka. MD; Stephen W. Duffy, esc. MSc, CStat; Bettina Borisch, MD; Patricia Fitzpatrick, MD; Lydia Ioannidou.Mouzaka, MD; Stephen W. Duffy, BSc, MSc, CStat; Bettina Borisch. MD; Patricia Fitzpatrick, MD; Range o - Range O Range o - 1st round mography screening. mography screeninq. be CI be be Ct ee be stout be stout CI be cr be 9t0u: CI Ct MRI+MG MG Stephen W. Duffy. MSc. CStat.; Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD; Xavier Castells, MD, PhD; Stephen W. Duffy. CStat.; Bettina MD; Patricia Fitzpatrick. MD; Solveig PhD; Xavier MD. PhD; Stephen W. Duffy. MSc. CStat,: Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind. PhD: Xavier MD, PhD; Stephen W. Duffy. MSc. CStat,; Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD:. Xavier MD, PhD; Stephen W. Duffy. MSc. CStat,: Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD;. Xavier MD. PhD; Stephen W. Duffy. MSc. CStat,: Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind. PhD; Xavier MD, PhD; Stephen W. Duffy. MSc. CStat.; Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD; Xavier MD, PhD; Stephen W. Duffy. BSc. CStat.: Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD: Xavier Castells, MD, PhD; Stephen W. Duffy. BSc. MSc. CStat,: Bettina MD; Patricia Fitzpatrick. MD: Solveig Hofvind. PhD: Xavier Castells, MD, PhD; Stephen W. Duffy. MSc. CStat,: Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD: Xavier MD. PhD; Stephen W. Duffy. MSc. CStat,; Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD;. Xavier MD, PhD; Stephen W. Duffy. MSc. CStat.: Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind, PhD; Xavier MD, PhD; Stephen W. Duffy. MSc. CStat,: Bettina MD; Patricia Fitzpatrick. MD; Solveig Hofvind. PhD: Xavier MD. PhD; Category Quality assurance standards must be considered in choosing DBT over DM. Overall sensitivity: risk in breast this is is important that rig that arg' For asymptomatic women, with high mammographic breast In women with histologically confirmed ductal carcinoma in situ In women with histologically confirmed invasive breast cancer, the With MRI: PPVI (recall) PPV3 (biopsy) psy) arg' The recommendations included in the guidelines are developed starting /31KVOmD). n 1 KVOmD). n 1 /'31KVOmD). Price to be paid: KVOmD). Indication Iondication Evaluation Condition Sensitiv•ty Sensitiv• Spensitivity What does "risk" in breast FaMRIsc-triaI intelligence in in Cancers k" Invasö.•e breast cancer Invasi.•e breast cancer ppvl PPVI Cancer Cancer" cancers" Ra 10 .810 7/24 24 (4%) 24 In favor of DM screening risk in this the to in this c' the to this c' the to this at c' the to this at the to this the to this the Performance of Dual-Energy Contrast-enhanced % of women MRI PPVI (recall rate) MRI PPVI (recall rate) ; MRI PPVI (recall rate) 17% MRI PPVI (recall rate) , MRI PPVI rate) (SO size time sizn sime ' St St S' gui- elines The ECIBC'S Guidelines Development Group (GDG) suggests: men 'nun Group too tion to digital mammography in screening programs tion to digital mammography in screerung programs cut the cuf cut cut the In cut the 'n ct cut cf cuf the cut In cut the cf •h Giordano, MD; Sue Warman. MEd; and Zuleika Saz•Parkinson. PhD; for the European Commission Initiative on mreast Cancer Giordano, MD; Sue Warman. MEd; and Zuleika Saz•Parkinson. PhD; for the European Commission Initiative on Breast Cancer Giordano, MD; Sue Warman. MEd; and Zuleika Saz•parkinson. PhD; for the European Commission Initiative on Breast Cancer Livia Giordano, MD; Sue Warman. MEd; and Zuleika Saz•Parkinson. PhD; for the European Commission Initiative on mreast Cancer Giordano, MD; Sue Warman. MEd; and Zuleika Saa•Parkinson, PhD; for the European Commission Initiative on Breast Cancer Giordano. MD; Sue Warman. MEd; and Zuleika Saz•Parkinson. PhD; for the European Commission Initiative on Breast Cancer Solveig Hofvind, PhD; Xavier Castells, MD, PhD; Livia Giordano, MD; Carlos Canelo.Aybar, MD. MSc; sue Warman, MEd: Solveig Hofvind, PhD; Xavier Castells, MD, PhD; Livia Giordano, MD; Carlos Canelo.Aybar, MD. MSc; sue Warman, MEd; Solveig Hofvind, PhD; Xavier Castells, MD. PhD: Livia Giordano. MD; Carlos Canelo.Aybar, MD. sue Warman, MEd; Solveig Hofvind, PhD; Xavier Castells, MD, PhD; Livia Giordano, MD; Carlos Canelo.Aybar, MD. sue Warman, MEd: Solveig Hofvind, PhD; Xavier Castells. MD, PhD; Livia Giordano. MD; Carlos Canelo.Aybar, MD, MSc; sue Warman, MEd: Development Group (GDG) suggests: examinations since the current evidence is mainly restricted to a single Interval from observational studies or where clip-marking is not routinely used raphy, in the context of an organized screening pro. raphy. in the context of an organized screening proy mammography develop breast cancer high risk r gui elines .4, 50% screening strategies or 20 be explainee. The opinion an or 20 be The opinion Of 20 mammography be explained. The opinion an or 20 mammography be The opinion an Of 20 be The opinion or an Caradian Association of Caradian of Car-dian Association of be The opinion Of MIX MX We found 10 studies (72, 89-97) reporting the accu- We found 10 studies (72, 89-97) repotting the accu- New York (1963) New York ( 1963) York ( 1963) York (1963) 083 (0-70-100) 083 0-83 (0-70-100) 0-83 169% 128% There is a need for research examining the classification of p.: mnsoaratnns. GOG are GOG are GOG GOG ara 060 Improving quality of care and reducing inequality in Europe Average 40/6 16 (2%) from relevant "healthcare questions" that below are grouped into main MRI PPV3 (biopsy) 26% MRI (biopsy) MRI PPV3 (biopsy) thx tal carcincx•na in situ thx tal carcinoma in situ thx tal carcinc«na in situ carcin«x•na in situ carcina•na in situ in situ MRI PPV3 detection Rate A combination of human and machine results in performance equivalent to US Low, intermediate and high-grade DCIS. Low', intermediate and high-grade DCIS. LOY', intermediate and high-grade DCIS. Low, intermnediate and high-grade DCIS. b_ Contributor Group* Contributor Group' Contributor Group • Iensity, in the context Of an organised screening programme: the density, in the context of an organised screening programme, the (BRCAI) Radiologists + Al radiologists other genes) 0.881 20.21 Interval cancer Against automated breast US screening over DM alone Linear (MG-SCR) (MG-SCR) (MGI-SCR) Very Slow Very low mammoqraphv (DM) in the context of an organised screening Developing (DCIS), the EClBC's Guidelines Development Group (GDG) Robert Mansel. MD; Francesco Sardanelli. MD; Elena Parmelli, PhD; Axel Gråwingholt, MD; and Zuleika Sa2•ParkinSon. PhD: for Robert Mansel. MD; Francesco Sardanelli. MD: Elena Parmelli, PhD; Axel Gråwinghoit. MD; and Zuleika SaZ-ParkinSon, PhD; for Robert Mansel. MD; Francesco Sardanelli. MD; Elena Parmelli. PhD; Axel Gråwingholt, MD; and Zuleika SaZ-ParkinSon. PhD; for Robert Mansel. MD; Francesco Sardanelli. MD; Elena Parmelli. PhD; Axel Gråwingholt, MD; and Zuleika Sa2•ParkinSon. PhD; for Robert Mansel. MD; Francesco Sardanelli. MD; Elena Parmelli, PhD; Axel Gråwingholt. MD; and Zuleika SaZ-Parkinson. PhD; for Average provided a net health benefit, it concluded that, overall BRCA 2 The ECIBC's Guidelines Development Group (GDG) suggests Furti«ipated i pa ted gram (conditional recommendation, very low certainty "ion Digital Mammography for Screening Women at False positives .2 0.94 0.2 Synthesized 2D imaging is necessary, with specific quality standards. depends on and the depends on end on me on and me on and 0t me end and me surveillance episode. Trials in this area are ongoing and their results will cancers 17. l: racy of DBT compared with assessment mammography No. of women may Ann Intern Med. doi:10.732U/M18-3445 Ann Intern Med. doi:10.7326/M10-3445 Ann Intern Med. doi:10.7326/M18-3445 Ann Intern Med. doi:.10.7326/M18-3445 Ann Intern Med. doi:10.732UM18-3445 Ann Intern Med Ann Intern Med doi:10.7326/M18-3445 Ann Intern Med. doi:.10.732U/M10-3445 Ann Intern Med doi:10.732UM18-3445 0-81 (061-1-07) considering the use of randomised studies for high quality evidence. Malmo 1 (1976) Malmo (1976) 674 680 In women with histologically confirmed invasive breast cancer, the - MRI FPrate8% MRI FP rate 8% There are different imaging modalities 9-5% 6-0% - rate — rate Size Of invasi•.e cancer (mm) Size Of invasiw.e cancer (mm) Size Of invasiw cancer (mm) MRI DBT:. YES DBT: YES i. topics. Each topic includes one or more recommendations presented in topics. Each topic, includes one or more recommendations presented in m. Tlø• 24 DBT mammographiC breast density and standardisation of the classification • screening with either digital boast or High breast igh breast MRI 89,696 MRI Dr. Ritse Mann, Nijmegen, Netherlands Normal 72.0% • screening With either digital breast tomosynthesis (DBT) or digital • Screening with either digital breast tomosynthesis (DBT) or digital • Screening With either digital breast tomosynthesis (DBT) or digital • screening with either digital breast tomosynthesis (DBT) or digital the European Commission Initiative On Breast Cancer (ECIBC) Contributor Group* the European Commission Initiative On Breast Cancer (ECIBC) Contributor Group' 167 70-74 breast cancer ECOC gui • MORPHOLOGICAL FEATURES Moderate Considerations for implementation and policy making of evidence: EtD available at http•.//bit.ly/33aQf6V). of evidence; EtD available at http•.//bit.ly/33aQf6V). the undesirable conse uences were reater than the provided to nem regarding me evidence this madazy 'or provided to mem regarding he evidence tehlra madaäty provided to tnem regarding me evidence provided to tnem regarding tne evidence this provided to mem regarding tne evidence this n•odaity 'or to rmdazy the rrndaty Good for fatty breasts 'i The ECIBC'S Guidelines Development Group (GDG) suggests: EClBc•s Guidelines Development Group (GDG) suggests Not good for dense for diagnosis in women recalled because of suspicious suggests not using additional magnetic resonance imaging (MRI) Monitoring and evaluation *ra ting Fractiong era ting *rating *ratking •ra ting Monitoring and evaluation Good for dense breasts Not good for fatty MRI adverse event rate 0.1% MRI rate 0.1% MRI adverseevent rate 0.1% that of the traditional double-reading process, but saves 88% of the effort gr wgowrk gr wHo'Yk iör\ogre s Large datasets from UK and USA. Kopparberg (1977) Kopparberg ( 1977) 0.58 (0-45-0-76) suggests not using additional magnetic resonance imaging (MRI) suggests not using additional magnetic resonance imaqinq (MRI) suggests not using additional magnetic resonance imaging (MRII using clip-marking after NCB/VANCB for surgical therapy planning Increased Risk of Breast Cancer influence the revision of this recommendation in the future. influence the revision of this recommendation In the future. gr wHork s gr wtgork inmogre s For author affiliations, see end of text. For author affiliations. see end of text. NOVEMBER 28, 2019 gr wHork in brogre s 16.59/1000 Grading of 78/1000 Cancer detection 180 (8-1) 1.0 180 (81) (8-1) 2812 Cancer ESTABLISHED IN 1812 11-9 (12•3) 119 (12-3) (12.3) (129%) (12-3) 381 NO, 22 programme. 2120 Patients 12 (S) Supplemental cancer a question and answer format. 1,800,000 gr inmogre s The GDG noted that improved evidence on the effectiveness is very Screening monitoring to distinguish test done with DBT and with DM succr.ng, (he of The GOG satorco scrccn•ng. thc limitations ot tomosyntncsls. Thc GDG taforco scrccntng, neudlr•g thc limitabons of tomosyntncs1E. Thc GDG scrccntng. thc limitatjons ot tornosyntncs1E. Thc Considerations for implementation and policy making (he of thc limitabons of tornosynthcsis. Thc GDG systems used for breast density, including technology for the automation EClBC's Guidelines Development Group (GDG) suggests using (DCIS), the EClBC's Guidelines Development Group (GDG) We found I RCT (64) and 10 observational studies Conditional Against hand-held US screening over DM alone mammography We found 1 RCT (64) and 10 observational studies gr Ja wHOrk in brogre s Low More cancer • not using either DBT ord digital mammography "Risk-based screening and the potential of Al for patient stratification" Screening costs screening mean? screening? screening interval Screening with density and 61 Good for invasive cancers Not as good for DCIS lesions at mammography screening, Digital breast tomo- lesions at mammography screening. Digital breast tomo- Al significantly better than radiologist for eslra e one . Good for distortions and Not good for cancers • Screening with either digital breast tomosynthesis (DBT) or digdital screening with either digital breast tomosynthesis (DBT) Oincludinq Ostergotland (1978) OstergOtland (1978) 0-76 (061-0-95) breasts Benera/ Electrico: Grant/Research, Advisory Board, Speakers' Bureau Beneray Electric: Grant/Research, Advisory Board, Speakers' Bureau Satellite Symposium at the tv 169% cut c' .1 Penalty: No. of screens 300% 30% This article was published at Annals.org 22 July 2019 2812 3075 Strategies for women at average risle Strategies for women at average risk: This article was pub lashed at Annals.org 23 July 2019 This article was pub lashed at Annals.org 22 July 2019 This a rt•cJe was pub Itshed at Annals.orq 23 July 2019 This article was pub lished at Annals.org 22 July 2019 This article was pub Its-hed at Annals.org 23 July 2019 This art'dle was pub hashed at Annals.org 23 July 2019 This art•cJe was pub Itshed at Annals.orq 23 July 2019 This article was pub Itshed at Annals.org 22 July 2019 This article was pubhished at Annals.orq 22 July 2019 This art•cJe was pub lashed at Annals.org 23 July 2019 This a rticle was pub Itshed at Annals.org 23 July 2019 This article was published at Annals.org 23 July 2019 BRCA 1 BRCAI BRCA 81% Machine 22 @ Siemens Healthcare GmbH, 2020 9 (5-14) 17 (13-22) 13-22) FP rate reduction: USA 5.7%; UK 1.2% FN rate reduction: USA 9.4%; UK 2.7% Description: The European Commission Initiative for Breast Description: The European Commission Initiative [or Breast tional, focused on outcomes that matter to women and provided • DCIS DCIS R isc that this regimen estabfshing a that this regimen a this reqirren a regimen a detection rate: in patients with breast cancer lesions. Minimal VA for preoperative planning. of the second human reader. (55-60, 65-71) that were relevant. Screening with DBT (55-60, 65-71) that were relevant, Screening with DBT Conditional In favor of either biennial or triennial DM screening 45-49 40-44 b' (conditional recommendation, moerate certainty of thesis leads to more true, ositives a len corre thesis leads to more true- 05itives a len carre thesis leads to more true, 05itives a len corre nthesis leads to more true, ositives a len s corre -0-979 -o. 979 thesis leads to more true- ositives a len corre s thesis leads to more truer 05itives a ten s carrec Number Of breast Of breast Number Of Breast detected by mammography NSN for to detect breast missed Breast detected by mammography NSN for mamtnography to detect breast missed e- important. Some members suggested that in the context of clinical Age group MRI-group necded to ad:zpt needed to di necded to •dapt di needed to Odzpt necdcd to adapt necdcd necdcd to adzpt needed odzpt needed •.0 adapt needed di needed to ad:zpt di needed •.0 adapt di to ad:zpt •.0 ad:zpt di to di to adopt •.0 adapt adapt ad:zpt adopt di of the determination of breast density. The overuse ot should be monitorec, especially when there Is a The overuse of should be monitorec, especially when the-re is a Toe overuse of should be monitorec, especially when there is a The overuse of should be monitorec, especially when there is a Toe overuse of Should be monitorec, especially when there is a Toe overuse of should be monitored, especially when there is a The overuse of should be monitored, especially when there is a el he to contrast-enhanced spectral mammography (CESM) over magnetic Recommendations Canada I (1980) Canada (1980) Canada Il (1980) Further research should also assess the cost-effectiveness implications 1,200,000 1,800,000 Interventione Intervention 0.97 (074-1-27) 0.97 (0-74-1-27) 097 097 (0-74-1-27) I. GOG I. GOS I _ GOS I ThB _ GOG _ GOS _ ThB GOS ThB GOS The GOS ThB (0-73-0-89) 130-2% 10-2% Increased risk 54 5. Healthineevs Healthineers 23.3% i' negative DM egative DM detecting cancers presenting as masses Six principles Cork in-'brogre European Congress of Radiology "ion in South In a rating of the certainty of evidence. ea2elire. based en breast eensiry•, alter the lifst exarr.nation. and a cut- baselire. based en breast eensity, the lirst examination, and a cut- me lirsr examination. a spiculations 006St without spiculations Cancer Screening and Diagnosis guidelines (European Breast Cancer Screening ane Diagnosis guidelines (European Breast In Morphofunctional ant: a synthesised 2D images) or digital mammography (0M). Tumour Tumour Stage S. • • • • Stage 15-20 / 1000 oa (years) bc gr s 08 Good for dense breasts 04 0.48 0.2 0-8 0.6 0.64 0.28 0.24 0.881 1.0 Too many false FP rate reduction: USA 5.7%; UK 1.2% Breast cancer diagnosis Breast cancer screening • • • • Training of the staff urgent urgcnt Breast cancer screening in addition to digital mammography increased the can- Breast Cancer Guidelines urgc•m urgc„m Screening recall assessment Good for DCIS (Microcalcs) size sign sime time sizn Conditional Against MRI screening over DM alone • Invasive (NST, Lobular, Mucinous, Tubular, Medular, Cribiform, Metaplastic, Not good for ILC reduction in specificity of 0.5-2% 30% 2015 very srnall DCIS mammcgrapntcally, it unlikely MRI will a very srnall DCIS mammograpmtcally, it 13 unlikely MRI will a very srnall DCIS mammograpmcally, It unlikely MRI a very srnall DCIS mammograpmically, it 13 unlikely MRI will a very DCIS it unlikely that MRI nnd a very DCIS mammographically, it unlikely MRI a very srnall DCIS mammograpmtcally, It unlikely MRI a very small DCIS it unlikely MRI a very srnall DCIS mammographically, it unlikely MRI a very DCIS mammcgrapntcally, it unlikely MRI will a very srnall DCIS mammcgrapn•cally, it unlikely MRI a very srnall DCIS mammcgrapmtcally, It IS unlikely MRI will a very srnall DCIS mammcgrapmically, it unlikely MRI a very small DCIS mammcgrapnjcally, It IS unlikely MRI will a very small DCIS mammographically, It IS unlikely MRI will a very smnall DCIS mammcgrapn•cally, it 13 unlikely MRI will a very DCIS mammograpmcally, it unlikely MRI will a • not Implementing tailored screening With both DBT and digital • not Itnptementlng tauorea screening With both and dlgltal • not implementing tailored screening With both OBT and digital • not implementing tailored screening With both DBT and digital DM+DBT:. NO DM+DBT: NO to tha 003 04 0013 03 highcr of the machirx•s ard the highcr of the ar%d thc highcr of ar%d thc higher of the machir.•s ard thc highcr of the ard eh: highcr of the machirc•s ard the higher of the ar%d thc highcr of thc highcr of ard thc highcr of the ard the higher of the ar%d the higher af ard the tha of ard thc of the ard thc i. Mammography of of the Good for masses and the the s Not good for DCIS Comparison of Abbreviated Breast M RI vs Digital Breast Tomosynthesis 1-02 (0-78-1.33) 1-02 078-1.33) Comparison of Abbreviated Breast MRI vs Digital Breast Tomosynthesis BRCA2 BRCA 2 BRCA 1 92% 92% 0.28 (0-78-1.33) 0.28 19.7 10-2% equipoise randomised trials would provide higher quality evidence. gr in rogre_s point • • • • Guidelines) are coordinated by the European Commission's Smaller size Cancers detected DO'S DOS IS ar of tailored DST screening for high mammographic breast density. of tailored DBT screening for high mammographic breast density. Fast 2. 40 15 Adnditional 11 Additional Conditional Additional 7/24 (29%) or distorsions/asymmetries and for TI Slow -u resonance imaging (MRI) to assist in surgical treatment planning. PPV 26.3% PPV 26.3%h PPV 26.30," radiologists cer detection rate and detection of invasive cancer 16.5/1000 27/1090 56.9/1000 16.5/1000 1 — Specificity I — Specificity Specificity NEW False Positives Recommendations: This synopsis of the European Breast Recommendations: lhis synopsis of the European Breast Supplemental MRI Screening for Women with Extremely July 19, 2020 / 10.00 am - 10.45 am on 8 800,000 Research should also aim at establishing the appropriate density Comparison Comparsison Commarission Very low very big very big one. very blg one. Assessment, positive results Overall 40 year'S O.•erall 40 year'S Overall 40 years 40 years 40 40 year'S Strong 1829 very very one. 1829 Sta at for Breast Cancer Detection Among Women With Dense Breasts for Breast Cancer Detection Among Wornen With Dense Breasts for Breast Cancer Detection Amongx Women With Dense Breasts Apocrine differentiation...) at by bzth to at by to at by bath to Good for dense breasts Fast size Research priorities mammography Overdiagnosis Stockholm (1981) 0-73 050-1-06) o.n (050-1-06) Joint Research Centre. The target audience for the guidelines should be considered as a Al only AlI only A I only Invasive cancers Better than DM for dense Not good for lesions Selection of an expert panel (open call; n = 28) • High risk women: 25-30% 2. at is it 2. is it GOG is it COG is it GOG is at is it at i' it at is il is it is il 11 1.0 0-6 The new ECiBC guidelines The new ECIBC guidelines 10% 7124 (29%) SA VA Morph04functional Morpho-functional Morpho•functional Morpho ofunctional Gain in Gain in I Gain provides essential levels of quality care that are equally accessible AS fix investigation costs Screening ages and frequencies compared with digital mammography alone (55-58, • not using either DBT ord digital mammography How to inform women about their results Guidelines provides recommendations regarding organized Guideiines provides recommendations regarding organized a•rt Mammography readers Average risk 191 191 raammo;raphy 191 mammography 191 cornærcd mammography Mammography HOW to inform women about their results HOW to inrorm women about their results 191 maammnography Screening and Screening ages and trequencies 2. The GDG suggested further research on the local effects of clips and The statements by Siemens Healthineers' collaboration partners and customers described herein are based The statemnents by Siemnens Healthineers' collaboration partners and customers described herein are based Screening with There a reed in dala man;ygement and There is a reed inx:rovemen: in data manogement and a feed in a feed in dala and oa more truene atives women without breast cancer and more true•nc atives women without breast cancer and Morphological techniques Missed with ? 11 01 4.7% with DBT ? Undergoing Screening cancers cancer" Other Screening /100% /100% 0.28 0.237 0.28 21.4 R.M. Mann, MD PhD Chair: Dr. Luis Pinal Chair: Dr. Luis Pina DOS Independent study of 6 radiologists: Al system outperformed all readers: Goteborg (1982) 0-75 058-0.98) 0-75 (0-58-0.98) No threshold for changing imaging techniques. 40-50 years 40-50 50-60 32 it is is cigs —e— Breast specialist (indtv'dual) —e- Breast sp«ialist —e— Breast sp«cialist —e— Breast specialist (indivdual) —E— Breast specialist (indtvdual) —w- Breast specialist (indtvidual) 133 '0 55. 133 55. 133 133 to (33 (33 to 133 '0 —e- Breast specialist —e— Breast »pecialist —e— Breast specialist (individual) —e— Breast specialist Breast sp«ialist Breast sp«cialist g" (conditional recommenoaron very of Influenced by ovarian includes women, health professionals, and policymakers- includes women, health professionals, and policymakers. —e— Breast Breast '0 (112/125) survival survival'* 7/24 (295 Additional research should also assess the comparison between DST be Additional research should also assess the comparison between DBT Cancer detection rate 14.2 / 1000 4.9 / 1000 Peripheric lesions can screening programs for women aged 40 to 75 years who are at 7/24 (29%) 20/24 Deep 904 CEMS were performed. By using CEM, 15 cancers were screening 'mmen have to a breast screening organB300ns. Women nave to a breast Development and 64-66, 69). No differences were found in recall rate screening Women nave to a breast Good for invasive ca. Not very good for DCIS 1.0 il 11 Dense Breast Tissue technique when women are Outcome breasts surrounded by fibro could i", could i", ad'S. could date-min' i", date-min' OCT i", i", ad'S. rot i", rat rat i", "young" i", L' Distribution of tumor grade/bio/ogy/prognostic measures in the DBT-additionally Distribution of tumor grade/biology/prognostic measures in the DBT-additionally - Distribution of tumor grade/biology/prognostic measures in the DBT-additionally across Europe. • IHC Patterns 69 may lhs may the 02 20.2 03 0.2 2012 0.8 lhs the tha Lower stage MRI 81 81.4 NA N.A 12 m impacts on the psychological well being for women and whether there is im acts on the s cholo ical well bein for women and whether there is may Good for palpable lesions Too many false positive SS SS 19, KS klO% In favor of biennial DM over triennial DM screening — Breast specialist (average) — Breast (average) — Brezst specialist (average) — Breztsp«ialist (average) 50% 30% 02 Dr Francesco Sardanelli, Milano, Italy — (average) Il i, UK Age Trial (1991) 083 066-1-04) 083 083 (0-66-1-04) 10-2% UK (1991) UK Age (1991) (1991) Chair & Introduction: Treatment costs DJ Treatment costs 'J Treatment costs •u Treatment costs Communication trainina for the staff Communication training tor the staff PHBC Tomosynthesis use in screening Further assessment after the mammogram Further assessment atter the mammogram Communication training for the staff dqit&' dQiteJ on results that were achieved in the customer's unique setting. Because there is no "typical" hospital or on results that were achieved in the customner's unique setting. Because there is no "typical" hospital or 210% k10% not onptemennng toilored screening magnetic resonance ROC-AUC of Al was greater than the average of radiologist by 11.5%. recall 45 Vim y Tim. y In favor of DBT for diagnosis average risk. The recommendations address digital mammogra- average rtsk. The recommendations address digital mammogra- Time. Y Moderate centre tochmiogy cmtre tochreogy availatlo. centre that OBI tochm•logy avail*lo. centre that 08 DB T tochmcl/ogy avail*le. Fraction Systematic review of evidence (up to 3/2016 — 12/2018) (conditional recommendation, very low certainty of the evidence) 20/24 (83%) 2/8 (25%) 2/8 25 MRI: NO Research priorities cycle 11 25 (55, 56, 58, 64-66, 69), but in 4 of the observational (55, 56. 58, 64-66, 69). but in 4 of the observational • Intermediate risk women. 15-25% • Intermediate risk women: 15-25% and DM Plus ultrasound for dense breast screening. and DM Plus ultrasound forqdense breast screening. Recommendation strength O 04 Although the GDG found no evidence re- Although the GDG found no evidence re. • 15-25% diagnosed in 14 of 904 women (cancer detection rate, 15.5 of stress. f a stress. a stress. may ma a ma f a ha ma 2D-mammography + DBT ? University of Navarra Clinic a haatti• it that. a haatti• it r a f a jr a be missed Family history 0.14 0.14k OA4 0.28 7.6 haattt• it mat. Pamplona - SPAIN Luis Pinat Pamplona Of Clinic VA DCts a it UNCHANGED screening? a it that. i" it that. that. Dismiss 35—41% cases as normal including 1 cancer in every 1,000—10,000 Dismiss 35—41% cases as normal including I cancer in every 1,000—10,000 General radiologist (inchvidual) —E General radiologist (inciividual) General radiologist (inchividual) General radiologist (indiividual) General radiologist (inciividual) —E General radiologist (inchvidual) General radiologist (in&ividual) —E General radiologist (individual) General radiologist (in&vidual) Learning mearking Lea rning Very Slow iiiiöil glandular tissue 3 '45678 3 '*5678 345678 3 '15678 45678 1.0 Methods: An international guideline panel Of 28 multidisci- Methods; Ah international guideline panel 01 23 PPV biopsy Overall Overall p.o.164) Average 26.3% 0-80 (0-73-0-89) 0-80 (0-73089) (0-73-089) (0-73-0-89) 26.3% 23.504 23.5% u, Further research is needed to build the evidence on benefits and harms Al With data Al developed With data Good for intermediate risk Not for low risk Al data Al algorithm With data Al algorithm Al algorithm data Evaluation phy screening and the addition of handheld ultrasonography, phy screening and the addition Of hand held ultrasonography. phy screening and the addition of hand held ultrasonography, 240 Luis Pinat Pamplona / Spain Luis Pina, Pamplona / Spain and results 232 SSI S tar, •.har, far, NO. at an impact of clips on breast cancer proqression or recurrence. an impact of clips on breast cancer progression or recurrence. "Strategies for women at average risk: The new guidelines on breast cancer metastasized garding the consequences of these accuracy results on imaging (MRI) Imaging studies the rate of false-positive recalls was increased studies the cate of false •positive recalls was increased T3 0023 detected cancers to clarify overdiagnosis. 841 421 2013 Luminal A • Luminal A DCts 82 — General radiologist (average) Additional tests for dense breast screening Staging of breast cancer develop breast cancer -5 Conditional YES (moderate 1000). PPV3 was 29.4% (15 of 51). assessment, assessment Based on the latest scientific evidence available, ECIBC seeks to 1 10 10 in •ith radiologists. significant in performance in radiologists. in in breast •ith radiologists. in performance in •ieast •ith radiologists. significant in performance in breast radiologists. in performance in "ith radiologists. significant in performance in •ith radiologists. in performance in "ith radiologists. significant improvement in performance in detects:'" •ith radiologists. significant in performance in radiologists. in performance in radiologists. significant in performance in with radiologists. significant in — BOADICEA la boratory and many variables exist (e.g., hospital size, samnples mix, case moix, level of IT and/or automnation TCv8 0.61) 0.61) laboratory and many variables exist (e.g., hospital size, samples mix, case mix, level of IT and/or automation Monitoring and evaluation Randomized controlled screening trial in women with familial risk THIS WHEN THIS IS WHEN plinary members, including patients, developed questions and TÅÉ 1189 294" 464 S 497 TÅÉ h. automated breast ultrasonography. or magnetic resonance inv automated breast ultrasonography. or magnetic resonance im. automated breast ultrasonography, or magnetic resonancc im. M.F. Bakker, S.V. de Lange, R.M. Pijnappel, R.M. Mann, P.H.M. Peeters, E.M. Monninkhof, M J. Emaus, C.E. Loo, M.F. Bakker, S.V. de Lange, R.M. Pijnappel, R.M. Mann, P.H.M. Peeters, E.M. Monninkhof, M.J. Emaus, C.E. Loo, Against annual DM screening Good for high risk women 70-74 40-44 Strong Expensive of DBT vs. DM through comparison of direct outcomes, including ot DBT vs. DM through comparison of direct outcomes, including aided Al Al aided b' Al a aided Al Al a diagnostic aided Al Al a aided A] Al a aided Al Al to diagnostic Al Al a A] Al Al Al when both techniques were combined, although the when both techmiqoes were combined. although the loading... screening and diagnosis from the European Com clinical outcomes, the group discussed the possible con- clinical outcomes. the group discussed the possible con- negative predictions: Al = humans x 28 Research is needed to define the quality parameters that need to be screening and diagnosis from the European Commision Initiative on Breast nitietivæon Breast More often node negative Lower yield Women advocates as full voting members The GOG recommends research on other MRI techniques. especially The GOG recommends researcn on other MRI techniques. especially (conditional recommendation, the evidence) (conditional recommendation, moerate certainty of 2783 Supplemental detection rate 9.3 / 1000 Supplemental detection rate 9.3 / 1000O 9387 9381 RF.LR (area - 0.67. p < .01) ton OVERDIAGNOSIS — BRCAPRO Recommendation strength O 3. The GDG suggested improved cost-effectiveness evidence on the use • Basal risk women: <15% Findings In this crcxs-sectjonal study with longitudinal follow-up Findings In this crcss-æectjmnal study with longitudinal follow-up Findings In this crtxs-æctimnal study with longitudinal follow-up Findings In this crcss-æctjonal study with longitudinal follow-up Findings In this crcss-æectimnal study with longitudinal follow-up Findings In this crcss-sectjmnal study with longitudinal foUlow-up Findings In this crcss-æctimal study with longitudinal follow-up Findings In this study with longitudinal foliose-up Findings In study With longitudinal follow-up Findings In this crcss-æcticmnal study With longitudinal follow-up Findings In this study with longitudinal follow-up Findings In this crcss-æectjcxul study with longitudinal foulow-up Findings In study with longitudinal follow-up Findings In this cross-æectimal study with longitudinal follow-up Findings In this crcss-æectimonaul study with longitudinal follow-up Findings In this crcss-sectimnal study with longitudinal follow-up Findings In this crcss-æctjmnal study with longitudinal follow-up Less expensive than MRI More expensive than DM Simulation in UK double-reading UK: Al maintained non-inferior performance Risk-based screening and the potential of Al Transpara Score ANGELINA "frequent" False Risk of breast cancer •I-year follow up in 858 women —Y 2 interval cancers •I-year follow up in 858 women 2 interval cancers ?l-year follow up in 858 women 2 interval cancers 21-year follow up in 858 women —Y 2 interval cancers 21-year follow up in 858 women 2 interval cancers *I-year follow up in 858 women —Y 2 interval cancers *I-year follow up in 858 women -+ 2 interval cancers •I-year fdllow up in 858 women -+ 2 interval cancers L Total costs z Total costs z0 Total costs corresponding recommendations that were informed by system- corresponding recommendations that were informed by system. 0-83 2012 01 2016 2013 2015 Costs 083 0-8 0.5 04 0.5 0.5 Time consuming 1,800,000 1,200,000 1,000,000 aging compared with mammography alone. The recomnmenda- offer to healthcare providers and women Clear and independent 100 1. 100% aging compared with mammography alone. The recomntmenda- aging compared with mammography alone. The recommenda- Luminal B Diagnostic Francesco Sardanelli 294" R.H.C. Bisschops, M.B.I. Lobbes, M.D.F. de Jong, K.M. Duvivier,J. Veltman, N. Karssemeijer, H.J. de Koning, Evidence on benefits and harms of DBT vs DM through comparison of direct RCT (64) showed no differences. The GDG agreed that Surgical planning OCCURS cern about radiation dose in DBT. Only 1 study reported Image only (area - 0.68. p < .01) Monitoring and evaluation cern about radiation dose in DBT. Only 1 study repotted Since the GDG made a strong recommendation for screening at ages — BCRAT M a I a i. adoption) there can be no guarantee that other customers will achieve the same results. adoption) there can be no guarantee that other customaers will achieve the samne results. Node status abbreviated The ideal design be randornised control abbreviated The ideal be control abbreviated The ideal design be randomnised control abbreviated The ideal design would be control abbreviated protocols. The ideal design be abbreviated The ideal design be control abbreviated The ideal design be randomised control abbreviated The ideal design wouland bnised control protocols. The ideal design be control The ideal design be control cancer Cancer" cancers 0023 per screening impacts of interval cancer incidence, stage of breast cancer at —Y Conditional No (very low -5 Conditional No (very low Conditional recommendation for the intervention fulfilled for DBT-based breast cancer screening programmes to be • not implementing tailored screening with magnematic bltresonance • not implementing tailored screening with mautomated bresonaste (LTR 20-50%) weRK• weRK 42700 Q 700 ns63 24 certainty of test accuracy data) ow certainty of the accuracy evidence n 269 "269 Start of 9676 SOQ soo 100 500 Screening with Ritse mann MD PhD irrluding 1444 womnen who underwent both abbreviated breast irrluding 1444 women who underwent both abbreviated breast irrludirng 1444 wornen who underwent both abbreviated breast irrluding 1444 wornen who underwent both abbreviated breast irrluding 1444 w.vornen who underwent both abbreviated breast irrluding 1444 wevornen who underwent both abbreviated breast irrluding 1444 '.vomnen who underwent both abbreviated breast atic review.•s of the evidence conducted between March 2016 atic review.vs of the evidence conducted between March 2016 auic reviews Of the evidence conducted between March 2016 atic Of the evidence conducted between March 2016 atic reviewos of the evidence conducted between March 2016 Little availability screenedg tions also discuss the frequency of screening and inform deci- tions also discuss the frequency of screening and inform decß- RR (95% CO for patient stratification Sensitivity Sensitivity' and reduced the workload of the second reader by 88%. and reduced the workload of the second reader by 88%, Conditional recommendation for either the intervention or the of clip-marking and comparing the economic impact of clip-marking on ABUS: NO in the Risk of missing ca cers in the (area - p guidance on screening and care. P.J. van Diest, W.P.T.M. Mali, M.A.A.J. van den Bosch, W. B. Veldhuis, and C.H. van Gils, In favor of triennial over biennial DM screening — p - p Operator depending Conditional recommendation against the intervention (conditional recommendation, moderate certainty of the effect would vary dependtng on the baseline rate. radiation dose (a surrogate outcome to assess the risk for 40-44 the effect would vary depending on the baseline rate. 3, How to invite and inform women about •s 793 •s 293 74 'S 743.8 53436 28139 OC6 003 14018 positives trials (ROT) but there are pree•lerns with leasibility because 0T numbers trials (ROT) bat there are prcbl•erns with leasibility because 0T numbers trials (ROT) bat there are probl•emns with leasibility because 0T numbers trials (ROT) but there are prcbberns with leasibility because 0T numbers trials (ROT) but there are prr»•lerns with leasibility because or numbers trials (ROT) bat there are problemns with leasibility because 0T numbers trials (ROT) but there are problems with leasibility because or numbers trials (ROT) but there are problemns with leasibility because 0T numbers trials (ROT) bat there are prCOblerns with leasibility because 0T numbers trials (ROT) bJt there are problemns with leasibility because 0T numbers trials (ROT) bat there are prob•lemns with leasibility because 0T numbers trials (ROT) but there are prrh•lerns with leasibility because or numbers trials (ROT) bJt there are prcOlerns with leasibility because OT numbers 't Luminal B Her 2 • Luminal B Her 2 Non-progressive Non- progressive Positive 4/24 (17%) 4/24 '17%) age screening Feaeåhility ir the at Fea%iility Md acceptability coud asses-see ir the monrorng at Feagiility acceptability coud ir me monrorng af Feasiility acceptability asses-sed ir the monl•oring af Feagiility acceptability cocd ir the af Md in the md acceptability be In the monronng at round Transparency of conflicts of interest and use of GRADE evidence to determine frameworks Transparency of conflicts of-interest and use of GRADE evidence to determine frameworks 1. 2. 1.0 European guidelines on Bilateral Towards the treatment of invasive breast Towards the treatment ot invasive breast Sian making for women at average risk who are recalled for Sion making for women at average risk who are recalled for sicn making for women at average risk who are recalled for •I r;' •I •I r/ ECR2020 detection, and mortality reduction. ultrasound system (ABUS) and December 2018. GRADE (Grading of Recommendations As- Pesapäne, Pesapäne, Coda", Pesa»ne, CodäH, Pesapane, Cod&h, Pesapane, Cod&H, Cod&Hi, Cod•i, CodäH, Cod&H, CodäHi, Coda•i, Coda"i, MRI and DBT. interpreted independently. abbreviated breast MRI MRI and DBT. interpreted independentty. abbreviated breast MRI MRI and DBT. interpreted .independently. abbreviated breast MRI MRI and DBT. interpreted independently'. abbreviated breast MRI MRI and DBT. interpreteeeindependentty. abbreviated breast MRI MRI and DBT. interpreted independentby. abbreviated breast MRI ultrasound system (ASUS) screening implemented. 2D-mammography + DBT ? + DBT ? Towards the treatment or invasive breast outcomes, including impacts of interval cancer incidence, stage of breast cancer outcomes, including@pmpacts of interval cancer incidence, stage of breast cancer COZ whathar arc r, COZ n COZ r, COZ whathar n COZ tct n COZ whathar arc COZ arc r, CO' whathar arc n, COG arc r, COS tct whathar ate rc COS tct whathar are r, COZ tct arc r, COS arc n COS tct whathar arc r, COGe n.:otcr.rg whathar arc r, r.:otcr.rg whathar r, tct whathar arc n tct whathar whathar r, [email protected] Supplemental detection rate 16.5/1000 Low-energy 8/16 (95% 25-75%) Low-energy 8/16 Low-energy 8/16 (95% Cl 25-75%) (95% Cl 25-75%) W Teh and M Morgan, Duty of candour in interval cancer reviews: Does Al add transparency? SO Ritse Mann, Nijmegen / Netherlands Bad spatial resolution radiation-induced breast cancer), and the GDG judged radiationGinduced breast cancer), and the GDG judged radiatiom-induced breast cancer), and the GDG judged Despite about a 2-fold increase in radiation dose with 50-69, these apply specifically to this age group. for the DENSE Trial Study Group* comparison screening and long tcrrn outcomnes; alternatively observational studies could be and long tcrrn outcomnes;, alternatively observational studies could be and long tcrrn outcornes; alternatively observational studies could be and long tcrrn outcornes•, alternatively observational studies could be and long tcrrn outcomes; alternatively observational studics be and long tcrrn outcomnes: alternatively Observational studies cou-lid be and long tcrrn outco•mes; alternatively observational studies could be and long tcrrn outcornes: alternatively observational studies coulid be and long tcrrn outcomnes;, alternatively observational studies co-rid be and long tcrrn outcomnes; altemnatively observational studies could be and long tcrrn outcornes; alternatively observational studies cou-lid be the need for additional procedures such as biopsies and surgeries. EFFECT octs DCIS Increased radiation risk 02 0.28 0.84 0.8 0.83 1.0 20/24 (83%) (498%) 318 (38%) 3.18 (38%) (38%) vroyrammnes. scr.eeni.ngansLAÅiagnQ5is Positive programmes. 20/24 scr.e.en.ingan.d-gjagngsis screening strategies screegni.ngansLA.iagnosis scr-egnj.ingAnsLA.iagngsis screeningansLA.iagnesis s.crgen.ingansLAjiagngs.is screeningansLdg.iagngsis screeningansLAiagngs.is s-gce.eningand-A.iagn.Q5is [email protected] screen.i.ngAnsLA.iagngsis certainty of evidence) 40 44 festons or who have high breast density. Breast MRI suspicious lesions or who have high breast density. in a in a GCG in a z GCG a GCG a z GCG IN PROGRESS 2) Immediate feedback for negative cases (NPV 2 99.90) 3) Immediate feedback for positive cases (PPV 83%9.90) in a z a' in z Conditional recommendation for the intervention (conditional recommendation, moerate certainty of Conditional recommendation forinst the intervention sessment, Development and Evaluation) Evidence to Decision sessrnent, Development and Evaluation) Evidence to Decision a z Presented at the British Society Of Breast Radiology Annual Scientific Meeting 2019 Bakker et al. Bakker et al. RSNA 2019 detected significantly more invasive camers (17 11.8 detected significantly mnore invasive camrers (17 11.8 detected significantly rnore invasive carrers (17 11.8 detected significantly more invasive camrers (17 11.8 detected significantly more invasive carrers (17 11.8 detected significantly more invasive camrers (17 women: 11.8 detected significantly mnore invasive cancers (17 11.8 detected significantly more invasive camcers (17 11.8 detected significantly more invasive cancers (17 11.8 detected significantly rnoce invasive camers (17 wornen: 11.8 detected significantly mnore invasive carxers (17 11.8 RSNA 2019 recommendation, very of evidence) MRI-arm MxI-arm cancer Pure Her2 . Pure Her2 23 k'" interrnedate interrne&ate hiØ interrne&•ate h•Ø interrned•ate hiØ intermnedate hiØ Research priorities intermne&ate Intermnediate interme&ate intermne&iate interme&iate 14/16 14/16 88% (95% 62-98%) 88% (95% Cl 62-98%) [email protected] Arthur L. Samuel. Some studies in machine that side effects of DBT compared with assessment mam- that side effects of DBT compared with assessment mam-n- False positive Rate False Positive Rate positive Rate CEM use of both DBT and digital mammography, the GDG use of both DBT and digital mammography, the GOG Dc carried out. out. Conditional No (very low UPDATED 'C This robust assessment of the Al system paves the way for clinical trials to 50-69 In favor of screening with DM alone over DM+tomosynthesis Very low Very Slow 1. ow certainty of the accuracy evidence Strategies for women at average risk: Recommendations on outcomes relevant to women and rating of the certainty of evidence (COE) at detection and mortality reduction. Very low certainty of the evidence Dr. Luis Pina and Dr. Ritse Mann receive financial support from Siemens for collaborations. 4. The GDG suggested research on the use of clip-marking for palpable 2011 2016 2011 2012 2012 2016 2012 2013 2016 2016 2015 Terry et al. J. Lancet Oncology 2019 Brentnall et al. JAMA Oncology 2018 frameworks were used to structure the process and minimize the frameworks were used to Structure the process and minimize the Saadatmand et al., Lancet Oncology 2019 the Ouaey and quality' he History auaZy and quallry' standards showd he cmt.ral and quality' standards showd he E. per 1000 womnen) than DBT (7 women; 4.8 per 1000 womnen). per 1000 womnen) than DBT (7 women: 4.8 per 1000 wornen). per 1000 wornen) than DBT (7 women; 4.8 per 1000 wornen). per 1000 wornen) than DBT (7 w•omen; 4.8 per 1000 wornen). per 1000 worne,n) than DBT (7 w•omen: 4.8 per 1000 wornen). per 1000 womnen) than DBT (7 women; 4.8 per 1000 wornen). per 1000 wcynen) than DBT (7 women: 4.8 per 1000 worTpn). per 1000 wtynen) than DBT (7 women; 4.8 per 1000 wornen). per 1000 wornen) than DBT (7 womnen; 4.8 per 1000 worrpn). per 1000 womnen) than DBT (7 women: 4.8 per 1000 wornpn). per 1000 than DBT (7 w•ornen; 4.8 per 1000 women). 1000 wornen) than DBT (7 women; 4.8 per 1000 worrpn). NEW How to organise breast cancer screening • not implementing tailored screening with hand-held ultrasound Ann 'nternMed. doi:10.B3WM19-212S Ann 'nternMed. Ann 'nternMed. doi:10.13WM19-212S Ann 'nternM'ed. doi:10.7326/M19-2125 Ann 'ntern Med. doi:10.Æ3W6tM19-212S Ann hternMed. Ann 'nternMed. doi:10.B26/M19-212S Ann*.org Ann '"tern Med. ABNORMAL ABNORMAL CELL mography (including magnification) were likely to be Evidence-based breast cancer guidelines Evidence-based breast cancel guidelines determined that the absolute increase in radiation- TIME learning using the game of checkers. IBM J Public calls launched in 2014 • Triple negative Triple negative in UNIVERSITÅ DEGLI STUDI DI MILANO UNIVERSITÅ DEGLI STUDI DI MllANO UNIVERSITÅ DEGLI STUDI DI MIL4NO certainty of evidence) of N. Karssemeijer ABUS: NO DEGLI STUDI DI MILANO DEGLI STUDI DI DEGLI STUDI DI MIL4NO resea•th co and resea•th cc and resea•th co ant: cc and co and CT and and 2; DEATH FROM OTHER CAUSES FROM OTHER CAUSES OTHER CAUSES developed. Stmdards should ce deveBpec In the improve accuracy and efficiency of breast cancer screening. influence of competing interests by enhancing transparency. developed. Stmndards should ce devek*ed In the Image developed. Stmndards should ee deveB;ed In pr,EuLar tot the Image Prioritize 30—40% of cancer cases: Al = humans x 6 The new ECIBC guidelines — The new ECIBC guidelines The possible utility in specific subgroups (extremely extended. suspect The pos-sible utility in specific subgroups (extremely ex-tended. suspect The pos-sible utility in specific subgroups (extremely extended. suspect The possible utility in specific subgroups (extremely ex-tended. suspect NO invasive cancer was identified by DBT abnre or as interval No invasive cancer was identified by DBT or as interval NO invasive cancer was identified by DBT or as interval NO invasive cancer was identified by OBT as interval NO invasive cancer was identified by DBT alone or as interval NO invasive cancer was identified by OBT alore or as interval NO invasive cancer was identified by DBT abne or as interval No invasive cancer was identified by DBT alone or as interval h ha ha tha ma ror author affiliations Ond of author affiliations SOQ On-yd Of toxt. author soo Ond Of toxt. ror author affiliations soo and of text. author affiliations. soo Of tøxt- author affiliations end Of author Of author affiliations. Ond Of text. author SOQ and Of toxt. author SOQ and Of author affiliations SOQ and Of toxt. author Of toxt. author SOQ Ond Of toxt. ror author affiliations. soo and of toxt. author affiliations. soo Ond Of cancer cancers N. programmes trivial (91). developed by the European Commission. developed by Commission. developed by the European Low certainty of the accuracy evidence Very low certainty of the evidence vs. non-palpable lesions. (91b program mes induced cancer was probably small (58-60, 64). courtesy of N. Karssemeijer (HHUS) b'c by bc Res Dev Prof. Dr. Francesco Sardanelli receives financial support from Bayer for collaborations. CEM Specificity 93.7%, PPVI 21%, PPV3 29%, NPV 99.7% CEM Specificity 93.7%, CEM specificity 93.7%, PPVI 21%, PPV3 29%, NPV 99.7% Cost-effectiveness of a screening programme using DBT courtesy N. Radboudumc Radboud u mc Conditional In favor of screening with DM alone over DM+tomosynthesis quality ct quality ct the quality ct trom the quabty trom the qual,ty trom the Wry-synthesis quality trom the Very low Very Slow 50-69 Questions and recommendations, expressed as strong or condi• Questions and recommendations, expressed as strong or condi- Questions and recommendations. expressed as strong or condt- Mann et al. JMRI 2019 et 2012 Vreemann et al., Vreemann et al.. BCR 2018 etal., Racfology 2016 2012 Racfology 2016 BCR 2018 2016 2017 Stakeholder feedback This article was published at Annals.org 26 November 2019 This article published at Annals.org 26 November 2019 This Annals.org November 9 This W as 0t November 19 0-6 0-8 0-8 2. 1. multifccal or multiple lesions) should be investigated. multitccal or multiple lesions) should be investigated. 02 0-6 04 US/ABUS 1-0 .810 Francesco Sardanelli, Milano / Italy 2019 DM DBT CEDEM MRI carrer during follow-up. carrer during folbw-up. carrer during follow. up. carrer during folbow- up. carrer during folbow. up. carrer during follow. up carrer during folbw- up. carrer during folbw.up. during folbw.up. S, Vreemann et al. Radiology 2018 H. Geuzinge et al. H. Geuzinge et al H. al. ICSN 2019 S. Vreemann et al. Radiology 2018 . ICSN 2019 ICSN 2019 carrer during carrer during up. Courtesy of N. Karssemeijer https://healthcare-qua/ity.jrc.ec.europa.eu/european-breast-cancer-guidelines h ec. europa. eu/european -breast-cancer-guidelines h ec. europa. eu/european-breast-cancer-guidelines https://healthcare-quality.jrc.ec.europa.eu/european-breast-cancer-guidelines unrcstrc:tcd CJ Siemens Healthcare GmEH, unrcstrt:tcd CJ Sirncnt GmEH, unrcstrt:tcd CJ Sirncnt Healthcare GmEH, unrcstrt:tcd CJ Siemens Healthcare GmEH, unrcstrctcd Siemens GmEH, unrcstrrtr •O Siemens Healthcare CJ Sirncnt Aralth:arc GmbH, CJ Sirncnt GmEH, Sirmcnt Healthcare GmEH, (conditional recommendation, very IOW certainty of the evidence) (conditional recommendation, very low certainty of the evidence) (conditional recommendation, moderate certainty of cf or 1 -specificity Specificity 1-speci6ciO• 1-sc«i6ciO• or tr 'i 1- 'e

  • AI
  • CAD
  • MR
  • tomosynthesis
  • breast screening
  • FFDM
  • mammography
  • screening guideline
  • dense breast
  • risk