It's my great pleasure to discuss interventional breast radiology with you today. I'll be starting by talking about localizations and then turn the podium over to Dr. Friedwald. So let's talk first about localizations of non-palpable breast lesions and where we are in 2022. So what's my objective here is to give you a case-based presentation. We're going to talk about what to localize. Be sure you know your target. How many things are you targeting? What tissue marker are you targeting? What do you need to bracket or do you need to bracket? And how are you going to perform localization? Is it going to be a regular mammogram, tomosynthesis, stereotactic, ultrasound, MRI, or even CT? And what device are you going to use to localize? And what do we localize? We localize things that we detect by screening. Palpable things that the patient is going to have breast conservation therapy that we probably have biopsied and placed a tissue marker. We are going to localize high-risk or discordant findings that also require excision. We also have those patients that come to us with large cancers that get neoadjuvant chemotherapy and have tumors or axillary nodes that are no longer palpable, and we localize those. And once in a while, we'll localize a case that is not amenable to an image-guided biopsy. And what do our surgeons want? So our surgeons are one of our key stakeholders, right? We have our patients, ourselves, our operating room staff, our administration, our financial people. But what do the surgeons want? They want us to make their job easy so that they can have a successful retrieval of the target. They don't want to miss the target. They don't want to remove the wrong lesion. They want adequate surgical margins. They don't want to resect a lot of healthy tissue. They want good cosmetic outcomes. They want their patients to be satisfied and send them more patients. The ORs want to be efficient. They don't want to be waiting for us. They don't want to hear that we're running an hour behind or two hours behind. And the patient wants to be on time as well. We've learned over the past years that there is no real significant difference between wires and seeds in positive margin rates or cosmesis. So that's not one of the issues that the surgeons are really keen on. So what's important? Number one, know your target. And I'm going to talk a lot about this because I have had the unfortunate experience of spending a lot of hours in analyses and root course analyses of what has gone wrong in a localization. So before you speak to the patient, before you do the localization, know the surgical plan and how you're going to localize it. Know the number of targets. It's really embarrassing if you tell the patient, hi, Mrs. A, I'm going to localize your left breast, and she says, what? It's the right. Very embarrassing. If you say you're going to do one wire, and she says, well, what about the other three things? Very embarrassing. What if the plan is a mastectomy, and you're talking about localizing her breast, and she's like, what are you talking about? That breast is coming off. You're supposed to localize my axilla. So you want to figure that all out before you talk to the patient, before your time out. So hopefully that your patient comes to you with multiple tissue markers if they've had more than one biopsy, and they're different markers, so you know which one that you're localizing. Hopefully, if the patient had neoadjuvant chemotherapy, they've had a tissue marker placed. And if a tissue marker was placed but displaced, you know where it is in relation to your target, so you know exactly what you're targeting. The patient needs to have confidence in you and provide consent. So let's look at some illustrative examples of what can go wrong. So here, these are your preliminary images, and there are two tissue markers. And then she comes, and she gives you these images. So what's the problem? There is only one tissue marker in these images, but it's the wrong clip. So what happened in this case? Two ultrasound core biopsies were done in her surgeon's office. That surgeon happened to only stock one shaped tissue marker, so they both were the same. There were lots of pieces of paper floating around the department with the order. Most said localized left breast. Only one said localized media left breast. The reviewing radiologist wrote localized media left breast, but nobody ever looked at that piece of paper. The technologist didn't have prelims because these biopsies were done in the surgeon's office, so nobody had a mammogram prior. She did one set of images, saw the clip, and presented the case to the performing radiologist for localization. The radiologist saw a clip and said, yay, there's a clip, let's go localize it. She didn't see that there was a tiny one at the edge of the image, nor did she see the radiologist reviewing notes. So the patient came to the OR, was anesthetized, and the surgeon was scratching his head, wait a second, that is not where I want to be. So then another radiologist was called to the OR to localize the first lesion, the correct lesion, with ultrasound guidance in the OR. And that led to a marker being retained in healthy tissue. So now that marker is still sitting in there, and nobody knows what to do with it because she's a high-risk patient, and to get that marker out is going to require another surgery with the attendant cost and cosmetic results. So that sits there until somebody decides what to do with it. So this is another case. These are made-up cases, but they're historical fiction. So here's a patient, has two fibroadenomas, one is palpable, here's a big one, here's a three-millimeter one. So the radiologist looks and says, oh, well, nobody wants to touch a three-millimeter retroareola fibroadenoma, I'm going to localize that big guy. And she puts a nice seed right in there, what do you think? Well, this is what really happened. The surgeon was planning on localizing both fibroadenomas. So the reviewing radiologist, radiologist A, figured that all out the day before and emailed the performing radiologist B, but that radiologist didn't read the email, never saw the email and thought, ah, must be the larger one, that's the one that's going to come out. The patient got to the OR, the surgeon looked at the films and said, wait a second, that's the wrong fibroadenoma that was localized, the patient has to go back to radiology to have the correct site localized. So obviously, the surgeon is not happy, her OR schedule is delayed. Now, we've all seen this case, there's innumerable tissue markers. So you have to figure out in advance what you're going to do. So in this case, what happened? Surgeon says, can you do two sites? Radiologist says, oh, sure, I know this case, I'll localize two sites, no problem. Radiologist localizes two sites. She localized the two sites that she had biopsy done to MRI. One site was LCIS, one site was DCIS. The surgeon reviews the post-localization mammograms and says, whoa, whoa, whoa, whoa, whoa, the patient had the MRI because she had a cancer, you didn't localize the cancer. Now we've got to go back and localize the cancer. Patient has to come back down from the OR to have the cancer localized because she didn't want to take out the LCIS, but we were able to remove that wire. So what are the common threads in all these scenarios, which are historical fiction, but based on true things that have happened? Multiple tissue markers from multiple biopsies can create confusion. There's no clear order from the surgeon by the performing radiologist reviewing it. There's no agreed upon plan. The standard hospital-wide timeout form was appropriately used, but it's not specific enough. It just gives patient name, date of birth, and site. It doesn't drill down to what we really need. So what's our new plan? Our new plan is for the surgeon to submit orders to radiology, including the site, location, site, including clock face and quadrant, pathology, tissue marker shape, and any other details that they want to include, the methodology, if they know it. The radiologist is going to review that order well in advance and agrees and essentially cosigns the order, or if they don't agree, send it back with comments, requests for clarification, and then that process gets repeated until the surgeon and radiologist agree and there's a final order that's always in the exact same place for the technologist and the performing radiologist to see it. So here's an example of what happened recently. Patient has this cancer, this architectural distortion. The surgeon order was to localize the cancer. The radiologist said, sure, but what about that architectural distortion? That hasn't been evaluated. So the plan gets revised. The surgery is postponed well in advance of the scheduled OR date. The surgeon then is really happy with the new process. So what else can go wrong? So I spent a lot of time, and unfortunately the hospital spent a lot of money on this one. Localizing from below, definitely possible. The surgeon's least choice, patient's least choice, because they're standing instead of sitting. Our least choice, the technologist's least choice, because we're bending down from the patient. But what we learned is that if there's bleeding and blood gets into the camera and the camera gets damaged, that's a five-figure replacement that is not covered by your service contract. It's nonstandard, and they say therefore not covered. So we recently had to spend thousands upon thousands of dollars to replace a camera because of a localization from below that had some bleeding. There's been a lot of talk in the online chat room of SBI about whether to use anesthesia or not. Who in the room uses anesthesia for localization? So almost everybody. So we always do. If we don't give enough anesthesia, we find that our patients complain. Other people in the room, like the technologists and the nurses, will also complain. They'll complain to our administration, they'll complain to our coordinators, they'll complain to me that this new radiologist or that radiologist is not giving enough anesthesia. So we always give anesthesia. Okay, so now let's go back to how we're going to localize. So we have to know what we're localizing. Is it a mass? Is it calcifications? Is it a tissue marker? Are there satellites? What's the extent? You have to pick a needle that's long enough. These needles come in multiple sizes, no matter how you're localizing. You have a choice of needle length. So now you have to pick your needle. Then you have to pick your route or your path. Are you coming medially, laterally, superiorly, inferiorly? Can you see it on a mammogram or is it too far posterior and you need to use ultrasound? If it's a mammogram, do you want to use standard technique or do you want to use tomosynthesis? And think about this. If you're doing your biopsy and something is really far back, put in a tissue marker that you can see easily under ultrasound so you can do that ultrasound low because maybe it's a small lesion that you are completely removing. If you're localizing under a standard mammogram approach, the patient is usually seated. You can use a special chair where the patient can be placed in Trendelenburg. You want the patient to be comfortable with mild breast compression and you want all your supplies to be on hand. And then you want to communicate with the surgeon and tell them exactly what you've done so they can triple check and be sure that everything is correct. We never used to localize anything back in the 1960s or 70s and there's a great paper in Radiology in 2013 that talks about the history of localization and why we needed to localize things. So it's a great paper if you like history. So let's talk about wire localization. Again, various lengths, typically very small. Very sterile, single use. The distal ends can be one size or another. Big needle can be removed or left in place. It's been the standard of care for many, many years. The wire is placed through the center of the target and the hook is beyond. Our most commonly used wire is the Copan spring wires made by multiple companies these days. This hooked end of the wire is overbent to form a spring so that it will open when released from the needle. The needle then is positioned with the tip through the target. The needle hub is held in place. The wire is placed so the black mark is at the hub, this black mark here. And then you push it through. Needle is withdrawn and the hook wire reforms and everything is good. And these are some of the different versions that are available. And generally we'll pick whichever the surgeon likes. We use the standard traditional one in our facility. Some surgeons like the Homer needle because this is a stiff part that they can feel and it's totally removable. These things can get transected. This happened 10 years ago. You can see that the hook was not in the specimen. It's still there today, 10 years later. There's a very nice paper that was recently published just earlier this year that says you can just leave these. If it happens, sure, it doesn't look pretty, but you can leave these for years safely. So that's our current common practice is wire localization. But what are the advantages? Why would we love to keep doing this? So we can reposition these. We can even remove them. We can use them under MRI guidance. They're very inexpensive. No oral equipment is needed and there are no regulatory requirements. But why would we ever think of changing? So there's scheduling conflicts. So sometimes we have to get there really early to get these wires placed for the surgeon's first case. The surgeon has to estimate the target based on the images or follow the needle. The wire can be displaced or migrate after insertion. And obviously we've all had that patient who faints during the procedure and the oral calls. Where's my patient? Where's my patient? Excuse me, she's on the floor. So what about seeds? They're the only way to uncouple? No, there are certain wires that are used that can be placed the day before. And actually that's done a lot in the United Kingdom. They use either a hook wire or a looped wire. And this is a wire that's recently been introduced to the U.S. market that is actually designed to be placed the night before, the day before. So even if you place the wire the same day, there can be problems. So you have to be careful that the wire stays in place. So this was done. Everything was perfect at 2.45. At 5 o'clock they call you and they tell you it's fallen out and you're starting all over again. Lots of other things can go wrong. So here's two sites of DCIS for localization. Everything looks perfect. But what went wrong? Who knows? You want to draw a box, make sure the patient hasn't moved. You want to do the posterior site first. You want to place the hub so that you have a clean shadow that's straight. But if you have not chosen a long enough needle, you're going to find yourself in this position where it's short. You push in a little further. It's still not ideal. You try to move it out and it doesn't really work. You can't see the tips of the needle well. Then you start moving things around. Then you've got this mess and the surgeon can't tell what's what. So you really want to straighten out those needles. Localizing under ultrasound is often an easier way to go for the patient. They're much more comfortable. Obviously, they're lying down in a supinoblack position. You want to choose a skin entry site that's near the lesion but allows for a parallel route. Place the needle through the lesion and then you deploy the needle under real-time visualization with orthogonal images to show that you are in the lesion. And many of our surgeons like a little skin marker with a magic marker over the site so they know exactly where to go. So here's just a poorly differentiated invasive ductal carcinoma. Our needle goes through it. Our wire goes right through it. And there it ends up at the end. One thing to remember is that sometimes the tissue marker is not exactly where your lesion is. So your post-images for an ultrasound look may not show the long part of the wire exactly at the tissue marker and don't worry about that. That's okay. We also localize a lot of nodes for people who have had neoadjuvant chemotherapy and have had a tissue marker placed. So here's a marker inside a node and we're going to localize that marker. Here's our wire going in and there we see it in the orthogonal position. So here are some tips for ultrasound. If the wire is hard to push out, you can put a spinal needle through the needle to make a path. You want to choose a needle, again, that's long enough so you can go parallel and not at a steep angle so it's easier to see the needle. Harmonics makes it easier to see that needle so you may want to turn on harmonics. What about MRI localization? So we rarely do those and we only do those if it's the absolute only thing that shows the lesion. All our equipment must be MR compatible. There is gadolinium used. So one thing we've learned is that maybe we don't want to do these wire localizations but we want to actually bring the patient to the magnet, put in a tissue marker, and then localize the tissue marker. So why is that? Because these wires are really flimsy. So this is one that I took out of the package and you can see it's already all bent and it's going to be really difficult to use and this is just the software that would show it. And these needles can really fall out and it's really embarrassing and difficult to deal with when all of a sudden you're pulling out the block and everything comes with it, the wire and everything and you have to start all over again. So this is a series of cases from NYU. So MRI localization is really necessary. 38% are malignant so these MRI only findings that can't be biopsied, you still need to have a plan to how to localize them at your facility. So what else can you do if you don't want to do these wires? So why would we change and look at some of these seeds or non-radioactive markers? It's really to increase our ability to place the localizing device well in advance and uncoupled from the operating room. We also are uncoupling the wire path and the surgical incision and root and there's less movement of the wires. And actually if you look at different fee-for-service things, it may or may not be more expensive depending upon if you're in a global payment system or a fee-for-service. Obviously most of us are in a fee-for-service system and the cost is much lower with wires and the profit is greater. All the different kind of seeds, whether radioactive or not, are very similar to use. It's an easy technique for the radiologist to learn. You can use mammogram or ultrasound guidance. The patient's scheduling is flexible, patient is not NPO. The seed is placed at the lesion. Seed migration is rare. There's nothing for the patient to see once you're done, but you does have to make an extra trip to the facility and you can bracket as long as the devices are about two centimeters apart. But what are the disadvantages? Today you can't reposition those seeds. Once it's deployed, that's where it is. It could migrate when the needle is withdrawn. Some manufacturers' needle tips are relatively blunt and you may need a skin nick. Bracketing can be difficult for some surgeons. It requires two trips to the hospital for the patient. And if the patient is large breasted or the lesion is very deep, the seed may be difficult to detect. The seed can be damaged or the surgeon may have difficulty and obviously there is a cost issue. So you need to know what's important to your stakeholders. What does the OR maybe need to change? Do they need different retractors? Do they need special pads? Do they need different lights? What about the size of the surgical probe? What sound is made? How does the surgeon know exactly where that seed is localized? How sharp is the needle and do we need to make a skin nick? Now again, going back to my theme that what can go wrong will go wrong, let's look at these examples. So here this was a great bracketing case done. We've got these really two nice reflectors appropriately placed. Everything is good, good, good. Patient comes in for her surgery. Well, my insurance has changed. Here it is. Oh, we don't take that insurance. She leaves. She doesn't have the surgery. So now she has to go across town somewhere else and now she has a different kind of marker put in. So now she's got a separate procedure, separate charge, separate bill. Here's a magnetic seed placed at the site of atypical ductal hyperplasia. Everything is good. Patient gets to the OR. Anesthesiologist says, that's not okay. We are not doing this patient in this outpatient OR. She needs to be done in the hospital. Okay, but the hospital doesn't have the ability to take out that kind of seed. They don't have the equipment in the OR. So then she has to go and now she gets another localization. So again, lots of extra costs. Wires could have been pulled out and are much less expensive. There are lots of seed choices, so let me just run through quickly in the last few minutes the different choices. So radioactive seeds have been around for about 20 years. They're small. They're 5 by 1 millimeter. They're I-125 pellets. There's no special precautions for the patient as the radioactivity is low and the surgeons use the same probe in the OR as they use for their technetium for central node injections. And the surgeons really like these. The disadvantages is that the program is hard to start. Once you've started the program, it's relatively easy to maintain, but the NRC regulations are difficult to deal with in the beginning. And this is what the needle looks like. And this tip is relatively dull, but easy to place. And here's just an example of one being placed in a fibroadenoma. And here's one under mammographic guidance. And here's the specimen radiograph, which we always get, showing the removal. So you want to be sure that the seed has deployed before backing out the needle. You want to prevent seed migration. And if a seed is not in an optimal position, you want to discuss that with the surgeon to know if that's OK and they'll compensate, or if you need to do another kind of localization. These can get transected in pathology. This is a nice article from 2016 from MD Anderson talking about that. So then there are these radar reflectors. So these are very commonly used by some places. Other people are concerned that these antennas can get damaged and they cannot be, therefore, supplied by the surgeon in the OR. Then there are magnetic seeds. So these are also small, 5 by 1 millimeter. The probe temporarily magnetizes the seed. An MRI, if you had to do an MRI after these were placed, you'd get a very large bloom. You need to change the retractors that the OR stocks. But some surgeons like these. Here's an example with some abnormal nodes. Here's the biopsy. And there's the magnetic seed, which has then been removed. There are also RFID markers. Each one has a unique number. These are a little bit larger. They're 9 by 2 millimeters. The introducer currently is 12 gauge, and you need a skin neck, although some smaller ones may be coming. And here's just an example of that being used. These can cause bloom in MR, just like the other ones. There are also an electromagnetic tracker. It's a little bit bigger. It's 1.4 by 8 millimeters. The surgeons will see a 3D image like that. There's also a company out of Canada called Mali. That's where the magnet is always on. These are also really tiny, 1.6 by 3.2 millimeters. And the surgeon gets a 3D localization. So you've got many options for localizations. Wires, non-wires, MAMO and ultrasound are the most common modalities. Other modalities can be used. Every team is gonna have their own preferences. These other choices are gonna change over time. So if you're interested in non-wire localization, I would clearly talk to the vendors because things do change. I want to thank a lot of people who have provided slides for me. Here are some selected references. I really would call your attention to this first one. It's AJR this year. Gary Whitman is the senior author. Has a lot of really practical tips for you. And with that, I want to thank you for your attention. I am going to be talking about a slightly different topic, but also interventional procedures. And I'm going to focus on tomosynthesis guided biopsies. And I'm here in Chicago. So if you ever want to come by Northwestern, you're more than welcome to come visit. So we have always been challenged with tomosynthesis only findings where there's an abnormality identified in the breast that we can't see with ultrasound. And prior to tomosynthesis guided biopsy techniques, we've had to try to estimate where the finding is for removal if desired. And how do we do that? Well, this is just an example of the right breast where there is a tomosynthesis only finding. We see there's architectural distortion here blown up on the digital magnification views here. And there was no sonographic correlate. Fortunately, in this patient, it was nice because she had a coarse calcification that we could help use targeting. But not all patients are as nice as this one to provide internal landmarks. So how are we to localize this or biopsy it so that we can find out what that distortion represents? So the best way to do it is really estimate the depth. So we can see here that that architectural distortion is about a third of the way inside the breast. We can see that it's there. And we can correlate that with the tomosynthesis scroll bar to see, yes, it's about a third of the way in. Because on stereotactic biopsy, you can identify the X and Y coordinates, but you don't know what the depth is. So that is a very rough, crude way to biopsy tomosynthesis only findings. But the question really remains, is it really necessary to biopsy these findings? Are we really finding more cancers or are we biopsying unnecessary lesions? So what's the evidence based on this? I'm gonna start talking about tomosynthesis guided needle localizations, which was performed before tomosynthesis guided biopsy devices were available. And this study, published by Phoebe Freer, really reviewed the results of a series of patients that had tomosynthesis guided needle localizations. And this was, again, pre-tomosynthesis guided biopsy technique, and these were in digital mammography occult, tomosynthesis only, and ultrasound negative architectural distortions. Early on, when tomosynthesis was available, there were 36 needle localizations. As I said, all were sonographically negative. 10 of the patients actually underwent MR out of the 36 that was negative as well. So really the only way to identify where the finding was was with a stereotactic biopsy. And then the other 25 lesions did not have a MR prior to localization. So 34 out of the 36 lesions at the time of excision were concordant, but two patients actually had discordant results. The first patient that had discordant results had a second tomo guided localization six weeks later, showing invasive ductal carcinoma. So it was good that they were persistent in relocalizing that area. The second discordant lesion had some questionable discordance. You've all had this experience where, well, it's a small radial scar. Does that actually really represent the architectural distortion? Short-term follow-up showed no, and serial follow-up showed no evidence of malignancy later on. So what were the results of those excisional biopsies? About half of them were malignant, and three-quarters of the malignant lesions were invasive carcinoma, either grade one or grade two. The other half were benign, but most of them, 15 out of 19 of the benign lesions were actual radial scars. And a third of those benign radial scars actually had atypia associated with them. So it really does demonstrate that even if you have MR negative, ultrasound negative, architectural distortion, we really do need to evaluate and biopsy these lesions. And interestingly, about half of these lesions were in fatty breasts. So because of some of this literature, it was very apparent that we needed to have tomosynthesis-guided biopsy technique. The first FDA approval was for the Affirm Upright system back in 2013. This attached to the standard unit. Then a prone system became FDA approved in 2016. And then subsequent vendors also had tomosynthesis-guided biopsy approval, Siemens in 2018, and then most recently, GE in 2020. So there are two different types of tomosynthesis-guided biopsy machines. We have upright versus prone. And what are the advantages and disadvantages of each of these? For the upright machine, it's nice because you can just add it on to equipment that's already imaging for screening and diagnostic patients. So you can utilize that machine when you're not doing biopsies. So if you've got space constraints, you don't want to dedicate an entire room for a biopsy room, you can use the upright. It tends to be faster in terms of positioning and actually doing the biopsy, and there's no weight limit for the patient. So if you've got a larger patient that can't be placed on the table. Also, patients who are unable to lie on the table, so patients who may be wheelchair-bound or have back issues, it really is helpful for those patients. There are advantages to the prone positioning. There is no setup, so it's already set up like a stereotactic table. Patients are less likely to have vasovagal reactions, and I'm gonna talk about that in a little bit. The patient's head is positioned away so they don't watch the procedure. It's a little bit more of a pleasant experience. And if you want to come in from the inferior approach, then they don't have to have a lateral decubitus position, which I'll show you in a second. So here's just a picture of an upright tomo-guided biopsy, patient positioned, and you can see the needle is headed from a superior approach. If you want to come from an inferior approach, you need to have the patient lie down in a lateral position, and you have to have a table that can adjust the height so that you get to optimal positioning with the biopsy machine. It's a little bit more cumbersome than the prone positioning, but still doable. The prone table is very similar to a stereotactic table in terms of the way it looks, and you can have the patient lying on their stomach with their breast hanging through the aperture so that the biopsy can be performed. So let's just go through what a tomo-guided biopsy procedure is like. Here's an area of architectural distortion. We see it in the outer part of the breast with spot compression. We can see that it persists, and we know this is true distortion. And here's just a close-up view of it. There was sonographically negative, so this needed to be biopsied. So the patient would be positioned, typically this was done on a prone stereotactic table. You take a tomosynthesis sweep, and here we have our tomoslice where the area's in focus, and we localize it with the markers. And then you confirm the location with the biopsy table, just as you would with a stereotactic biopsy, and here's just a representation of where the lesion is in space compared to your needle, and you can match up your X, Y, and Z coordinates. You give local anesthesia, advance the needle until the machine reads green or zero on the X, Y, and Z axis, and we can see that the aperture of the needle is in the center of the lesion and everything is lined up. Then you can take a tomosynthesis sweep, that's optional, it's not required. I don't usually do that for distortions or asymmetries, but I will talk about that for calcifications. Then you deploy the needle, and then remove the needle, place the clip, and you take either single or stereo pair to confirm clip placement. So here is pictures of our original lesion, we can see here, and the biopsy clip in appropriate position. Then just as a reminder, when you take your post-procedure mammogram, typically in another room, you wanna make sure that you do it with tomosynthesis so that you make sure that your clip is actually in the center of the tomo-only finding. So this was actually invasive ductal carcinoma grade one, so a good example of a sonographically negative tomo-only finding that ended up being malignant. So tomosynthesis guided biopsy, it turns out, we actually use it for almost every procedure now instead of just a stereo because it allows for more accurate localization of the findings. So even if it's calcifications, we'll use tomo-guided biopsies, and it is actually a faster procedure, and there's a lot less radiation because if you think about it, you take a single tomosweep, that's one dose to the patient instead of a stereo pair, which is two. The only difference I do with calcifications is that I take a tomosweep post-lidocaine because the lidocaine can often move that group, those groups, that group of calcifications, and you wanna make sure that you target them appropriately after lidocaine. And then it's optional to take a tomosweep after you, or pre-fire picture. Again, my dad always says measure twice, cut once. I always feel like measuring it each step of the way allows for less error. And then at the end, you wanna take a stereo pair because the needle, you can't take a tomosweep after you've fired the needle. It will obscure your finding. Also, if you are gonna charge for a stereotactic biopsy currently in the United States, there is no tomo procedure code. So if you're gonna charge for a stereo, then you need to have a stereo pair somewhere in the procedure. So that's a good time to do it post-fire. And then the rest of the procedure is the same. So here's an example of grouped calcifications. We can see it's in the upper outer quadrant of the breast. I'm gonna zoom in here. You probably, in this light, can't see the calcifications, but they're subtle grouped calcifications, better seen in that fatty tissue. You can see them right here that are indeterminate. So again, we take a tomosweep, localize them. You actually know exactly where they are in the breast, and this is very helpful for grouped calcifications in a background of calcifications, and you're not exactly sure if you have the group on the stereo pair as the same group. Then you insert your needle, and you can identify how much of a depth you have. On the console, you can see how much leftover room this is the detector side. You can see the trough. If you need to use a shorter needle, you can use a petite needle. So then you take your tomosweep. You can localize, again, the calcifications and make sure they didn't move post-lidocaine administration. And then you take your stereo pair after firing. I typically like to take this so that I can see if the calcifications are below the needle trough, then you can sample from three to nine or directionally sample based on where those calcifications are in the trough. You can just see a zoom-in image. We can see that these are below the needle, and so that's exactly what I did. Here's a specimen radiograph showing those grouped calcifications, which were indeterminate, and our post-procedure mammogram with the biopsy clip. Here's, sorry, the post-procedure mammogram here, showing that we got the calcifications that we targeted. And this ended up being PASH. The calcifications were benign, so it's not always that the calcifications are gonna be malignant, but it's just a good demonstration of how you can use tomosynthesis-guided biopsy for not just distortions or asymmetries, but also calcifications. So what's the evidence? So now that we have the tomosynthesis-guided biopsy capability, what's the evidence for this procedure? Well, this is an interesting paper that was published that was about patient experience with the upright biopsy, tomo-guided biopsy equipment. This was before the prone was available. And they looked at 10 attributes that were assessed, some that were associated with the test, and some were associated with the procedure just in general. And they had a scoring system. Basically, you don't have to go into all the details of the scoring, but a higher number means basically a worse experience. And we can see that with the tomo-guided biopsy, there were higher scores, meaning they had a worse experience than the stereotactic biopsy. And why is this? Well, the general thought was that it was because it was an upright machine, and they had a lot of vasovagal reactions, and they were watching the experience. And so it's just something to keep in mind, that even if it's a better procedure, or easier for you to have the upright in your facility, that the patients really may not have as good of an experience. This is one study. It's hard to know exactly if that contributes. Most people find that it's a quick procedure, so that even if the patient's watching it, then it goes by quickly, and that can help. So when you look at the total score for the six items related to the procedure, it was 812 for tomo, and 768 for stereotactic biopsy. And then they showed that there wasn't too much difference between the two that were not attributed to the upright or prone positioning. So then this was a retrospective study looking at all cases of tomosynthesis-guided biopsy where that were also DM occult and ultrasound occult. This is very similar to the needle localization procedure paper, but this had 34 patients underwent tomosynthesis-guided biopsy. Two were not visible at the time of biopsy, and it was an upright piece of equipment. Malignancy was less than with the needle localization, so maybe these architectural distortions were not as obvious as the ones that were sent to surgery. There was a malignancy in about 25% or 26%. Eight were invasive, one was DCIS. Of the invasive cancers, four were grade one, two were grade two, sorry, one was grade three, and one was unknown. Most of them had no lymph node involvement, so these are basically early invasive cancers. Of the benign lesions, 64% were radial scars, and about 20% of them had associated high-risk lesions. Thus, 80% of the benign lesions went on to surgery. One patient did have a vasovagal reaction. Other papers showing similar results with tomo-guided biopsies, we see that the positive predictive value is pretty high, 42%, 54%, and 26%, and these were all pretty much architectural distortion cases. So again, just reemphasizing, although not as dramatic as the needle localization results, still tomosynthesis-only sonographically occult findings do need to be sampled. So I'm just gonna review some of the considerations, some things to think about with tomo-guided biopsy that are different than your typical stereotactic biopsy. The needle is angled, and that's because you won't be able to see the finding with a tomosynthesis sweep. You have to angle the needle out of the way. And that can be challenging to determine if you actually sampled the findings properly, because now it's not just the z-depth that the clip can migrate, it actually, the clip can be placed, or displaced, rather, in the x, y, and z-axis, and therefore it may be hard to know whether you actually got the architectural distortion. It's obviously much easier for calcifications, because if you've got the calcs in your sample, you know you got them, but it is definitely challenging for distortion and asymmetries. As I mentioned before, the post-fire image should be a stereo pair if you're gonna charge in the United States, and the needle can obscure the findings. Skin calcifications, typically these should be detected at the time of the diagnostic examination. You can tell if it's on the first couple slices or the last couple slices, you'll know that they're in the skin. But it can be validated, and I'm just gonna go through an example of a case that was a little bit confusing to me the first time we used the tomoguided biopsy system. So this patient has heterogeneously dense breast tissue. On tomosynthesis, there are grouped calcifications here in the central, slightly outer breast, and we can see that it's actually deep into the stack. So it was determined at the time of diagnostic imaging that these needed to be biopsied, they were not in the skin. So we can see that these are indeterminate calcifications. So she presented to me on my procedure day. We can see here are the calcifications, but what was confusing me is that it was actually on the second to last slice. And I thought, oh gosh, maybe these are skin calcifications, and it didn't really make much sense to me as to why they were thought to be deeper on the diagnostic imaging. So here is what I was looking at, and you can see this is my compression paddle and the tip of the needle, and we can see that the calcifications are right where the skin should be. And again, I think, well, I don't obviously wanna biopsy these. If they're skin and it really can tear up the tissue if you actually sample with the trough outside of the breast. But are these skin calcifications or are they deeper into the skin? And so what I realized is the paddle where the needle's going through is fenestrated, and so there's a window. And it turns out that this patient's breast tissue was really pooching way outside of that window. So these actually were deeper into the skin. So even though on the console it looks like it was in the last two slices, it actually is deeper into the breast because the plate is not completely flat. So that rule only works if it's up against a flat plate. So just keep that in mind because I was tempted to cancel that patient because I said, oh, these are in the skin, I don't wanna biopsy these. But actually, it was deeper into the tissue. So here's our needle into the breast. We can see that the calcifications are there. Here's our post-procedure mammogram and our biopsy clip where those calcifications were. And this was invasive lobular carcinoma grade two. So I was about to cancel a procedure because I thought they were skin calcs, and actually it was invasive lobular carcinoma. Not ideal. Okay, so what about the lateral biopsy technique? And people get confused when I say lateral. It doesn't mean that we're coming from the lateral part of the breast. It means literally we are flipping the needle around with the breast in the same compression. So if you're in the CC position, the breast compressed, you swing the needle around and you can come in parallel to the detector. This is actually really, really nice for patients that have less than two centimeter compression breast. But there are some things that you need to think about when you do this procedure. So this patient has implants here. And we can see grouped calcifications here that are indeterminate on the implant displaced views. Here are the magnification views. Again, indeterminate calcifications. So we can see here that the calcifications are very close to the detector. And when you look at, this is the distance between the detector and the compression plate, we can see that this is less than two centimeters. So instead of taking the patient out of position, they remain in this position, we actually swing the needle around here with the lateral arm and come in parallel to the detector. So this is what it looks like. We target those calcifications. Here is the calcifications with the tomoslice. And you can see that our needle is now coming in parallel to the detector, really nice with the calcifications in the center of the trough. Here are calcifications. At the end of our sample, just to point out, you oftentimes need to make a skin nick because there's nothing on the backside of the breast. So sometimes you can snowplow the breast across. And this was a good example. We just barely got the calcifications in the end of our sampling here. Here is just our biopsy clip. Sometimes, you know, if you have a tech or your fellow or resident or you can actually just hold the tissue on the backside as well. Some people use putty to help compress it, but there are definitely ways where you can actually apply back compression to make sure you're not snowplowing. So in summary, it's quite obvious that tomosynthesis-only findings are really essential to be biopsied because we find enough malignancies. Early data showed it was about 50% of the time. Later data is now showing it's about 20 to 25% of the time with sonographically negative findings. There are different types of tomoguided biopsy, pieces of equipment, prone and upright. They both have their advantages and disadvantages. If you have space considerations, then an upright machine may be beneficial. If you're lucky enough to have a room dedicated for a biopsy procedure room, then the prone table may be more advantageous. And it turns out that it is faster for patients despite our initial concern taking longer and about the patient experience. And you can actually, if you use tomosynthesis-guided biopsy technique on all of your patients, not just the asymmetries and distortion, you'll actually be administering a lot less radiation. And that's the end of my talk. Thank you very much.

interventional breast radiology

non-palpable breast lesions

tomosynthesis-guided biopsies

localization techniques

wire and seed placement

magnetic and radar reflectors

radiation exposure

patient management