Missed and Nearly Missed Breast Cancer on Mammo, US and MRI

Introduction to Missed and Nearly Missed Cancers

We're gonna spend the next half hour or so talking about mist and merely nearly missed cancers. We're gonna see an awful lot of cases, so we're gonna get right into the heart of the matter.

But why we miss and nearly missed cancers? It obviously affects both us as well as our patients. Sometimes it's pure human error. Sometimes we have technical difficulties, sometimes we can't see a cancer because of the inherent characteristics of the cancer and the surrounding tissues. Sometimes we use our birads classifications and descriptors improperly and sometimes we don't correlate with the clinical history or other exams that the patient might have had. And obviously when we're doing procedures there can be issues with informed consent or communication.

Let's get right into some cases and if you're a little bit of sleep after lunch, hopefully this will serve to, to wake you up a little bit as we go through some of these.

Case 1: Calcifications Misread as BI-RADS 3

Here's year one. This woman went to a very prestigious institution and had this mammogram with these calcifications. Think in your mind about how you would describe these calcifications and what you would do with them. We have a choice of birads two, three, or four. How many would make it a birads two? Anybody? I don't. How do I go back? There we go. Does anybody wanna make it a birads? Two. How about a birads? Three. How about birads? Four. Unfortunately she didn't see any of you. And at this very prestigious institution, this was read as Birads three. Then she went for a reduction. The plastic surgeon said he would take care of the calcifications, but he neglected to have them localized. Then she came two years later. And here are the calcifications. Obviously there are more calcifications, they're heterogeneous, they're pleomorphic, we biopsied them and they are DCIS. Clearly they never should have been made a birads three. And having been made a birads through the plastic surgeon should have be certain to have localized them before surgery.

Case 2: Clear Nipple Discharge

Here's another case, a screen. This is month one, routine screening mammo. Now she complains of clear left nipple discharge and she has these ultrasounds. Both with the 17 and the 12 megahertz probe, we're not seeing anything. Now what do you wanna do? Now this is a test to make sure you've been paying attention for the previous day and a half 'cause you've seen this discussed. Who wants to do number one? Nobody. How about repeat the mammogram? It's been a couple months. Nobody who wants to make it a birads three and do a mammogram and ultrasound in six months? Nobody Who wants to do a ducto gram? A lot of people Who wants to do an MRI? A lot of people. Okay, we're about split between Dr. Graham and MRI. We did the MRI. And this is very similar to the case that Dr. Comstock showed yesterday where you clearly see the non mass enhancement in a segmental distribution here on the subtracted images. This is another case of DCIS.

Case 3: Screening Exam with Positioning Issues

Here's another screening exam. Normal or abnormal. If I could go back, look very carefully at the inframammary fold. Did you see that the first time there it is. Obviously its positioning is critical. If you don't see the area on the image, there's no way you're gonna find that cancer. Luckily for this patient it was very easy to see on the cranial codal view.

Criteria for Good MLO View

What makes a good MLO view? We have to always be sure that we see the pectoralis muscle to the level of the nipple. We want the pectoralis major muscle to be convex. We wanna see as much posterior tissue as possible. Obviously the complete is best. We need adequate compression. The nipple has to be up and out. The inframammary fold needs to be open and obviously motion on sharpness and blur is always an issue that we'd like to avoid.

Criteria for Good Craniocaudal View

And for a cranial cordal view, we want the breast pulled straight out. Not medially, not laterally. We need good compression and we always wanna see less than a centimeter difference of tissue between the cranial cordal and the MLO and obviously again no blur.

This patient obviously was lucky in the cranial caral view, got the lesion clearly on it. Additional views could have been misleading and certainly didn't help even that cleavage view didn't get as far back as the original cranio caral view ultrasound showed a clear mass that's worrisome. It's shaped margins are all worrisome. This was clearly a birads five lesion which underwent biopsy. The tissue marker is in notice that you can't see now the, the tissue marker on the cranial cardi view or the cleavage view but can only clearly see it on the MLL view there. And the axial MRI shows a location of this abnormality, far medially in the posterior breast. Here it is again, far inferiorly in the breast. The positioning is critical. This technologist did an absolutely fantastic job and really had a large role in saving this woman, from a much more invasive cancer.

Case 4: Converse - Missed Additional Cancer on MRI

Here, however, is is the converse. If we look at the index cancer, which was seen in biopsied where the blue arrow is, this was all straightforward, but now we're looking for extent of disease. Is there any other cancer? This patient underwent MRI. We see that it's again, very, very far back and very inferior in the inframammary fold. But here was this patient's MLO view, the nipples not up and out. The posterior tissue is not included on this image, therefore there was no chance for finding that other than having the MRI, which clearly shows the second cancer.

Ultrasound Cases

Okay, let's switch gears a minute. We're gonna do a lot of ultrasound cases.

Case 5: 30-Year-Old with Palpable Mass

Here's a 3-year-old who's had a palpable mass for one year. She comes to you having previously undergone fine needle aspiration where the result was markedly atypical cells. She had refused mammography and had done nothing for the year. Here's her ultrasound to take a very hard look at this ultrasound. Okay, what do you think BIRADS three. BIRADS two. Is it a fibroadenoma? Who wants a BIRADS two? It's a fibroadenoma. Anybody want a BIRADS three? It's parallel oval. Can I sell that to anybody? Okay, I sold it a little bit, okay, I got a couple of birads threes. Anybody wanna make it a birads four? Okay, the majority is going for birads four. The truth was this had been palpable, as I said for over a year. The FNA was highly atypical. We needed to do the core biopsy which showed this was a moderate to poorly differentiated ductal cancer. We'll see that even though it's parallel, it's not homogeneous, its margins are a little bit irregular. Doesn't strictly meet the criteria for birads two or BIRADS three lesion. And it's the only one that's there. It's not one of multiple, therefore it's clearly a suspicious abnormality that should be biopsied.

Case 6: High-Risk Screening in 27-Year-Old

Here's another case. Now we're doing high risk screening. No palpable abnormality. She's 27 years old. Okay, the va, the vascularity is going around. We heard that was maybe a good sign. Maybe. Who wants a BIRADS two for this? How about birads three? She's only 27. I can't sell that to anybody. How about birads four? Okay, the, the overwhelming majority, unfortunately she did not go to any, anybody here and this was read as a by RADS two and she was sent home with the wording of the report compatible with fibroadenoma return in one year. Obviously it was not correct. We saw the shape was irregular, the margins were not circumscribed, it was heterogeneous. She comes back in a year. Now it's palpable. Now it's obviously much larger. It's highly vascular. She had gone this whole year with this growing cancer because the birads were not followed. The criteria were not applied correctly for determining and evaluating an ultrasound detected mass. It was obviously able to be detected by whoever was doing the ultrasound, but they were not able to characterize it appropriately and make the appropriate recommendations. This obviously caused significant issues and here was her MRI. You can see the extensive involvement of this cancer, which had been growing. Now in A-B-R-C-A patient for over a year, the skin is involved. She luckily was able to undergo neoadjuvant chemotherapy and actually at the time had a complete pathologic response. She is doing well despite that delay in care.

Reasons for the Mistake

Why was this mistake made? Well perhaps whoever was reading this ultrasound didn't have much experience. Maybe they were tired, maybe they weren't paying attention, maybe they didn't have prior images, although I'm not sure why you would've needed them. Maybe they looked at some benign features rather than looking at the most worrisome features and maybe they didn't really realize that she was at significantly high risk.

We see a lot of ultrasound issues that come to our attention by having referrals from surgeons that are in our practice. We get patients to come from literally all over the world and they typically will bring images from all over the world with them. And sometimes you see absolutely phenomenal calls and sometimes not.

Case 7: 64-Year-Old Screening Mammogram

Here's one, here's a 64-year-old for a screening mammogram. By the way the crow flies. This was about 16 centimeters from the nipple. Not generally the way I like to measure things, but we see that it's pretty far out laterally there. I think on the mammogram, nobody is too as much difficulty with this mammogram. It's a pretty fatty breast. There's a solitary mass, everything seems straightforward, margins are irregular shape is the round or irregular. We're all worried about it. Here's a birads question. Are you gonna make this a BIRAD zero and get an ultrasound or a BIRADS five because you're gonna do ultrasound for biopsy guidance. Who wants the BIRAD zero? A couple. Who wants birads? Five. Okay, the majority wants to do birads five because you're gonna do the ultrasound just for guidance. If for some reason you don't see it on the ultrasound, you can always biopsys under stereotactic guidance. Here was the ultrasound that was performed on the same day. This is all the images from the ultrasound. I believe there were four. Here they are. And this is what was submitted and they're labeled as such. Left upper quadrant, left upper inner quadrant. No mass was seen. Now what do your next step? I think most of us would say we need to repeat the ultrasound or biopsy under stereotactic guidance. Three days later, somebody else doing the ultrasound can see this. How do you prevent yourself from falling into this trap? Well when did you have to be certain that you're looking in the right spot? If you notice this is 13 centimeters away from the nipple. It's very far lateral. You have to be sure that you've included the relevant portion of the breast in your ultrasound and not stayed in the central portion of the breast. But really absolutely be certain to have correlated correctly to the right spot. Here this is obviously very easy to now biopsy an ultrasound. Having seen this distance from nipple is important. When doing a screening ultrasound, you also need to be sure that you've included the edges of the breast. Media, lateral inframammary fold, axillary tail, all have to be included. Otherwise you can find yourself in this situation. It happens fairly frequently.

Similar Case: Architectural Distortion

This is a very similar case. Here's the screening mammogram and the arrows pointing towards an area of architectural distortion. Here it is. Now you might wanna call this a four or a five but, you're gonna be worried about this. Here's the ultrasound notice they've got the depth here at six centimeters. Most of this screen is useless. They have nice images here of lung. You don't really need to see all this lung here. You've got the muscle here and you've got this amount, you've got half the screen with breast, but you don't see the lesion. They haven't shown you anything. Now here's your options. If the mammogram you thought perhaps was a birads five, but the ultrasound is normal, who wants to get an MRI to break the tie? Dr. Comstack talked yesterday and you could sometimes use it for problem solving. No, I'm getting a nose on that. How about a birads two? You know, maybe it was a soft call. Maybe I made a mistake on the Mammogram. No takers for that one. How about a birads three? No. We all learned that that would be an inappropriate use of birads three. Who wants to go to the stereo? Okay, we've got a lot of people stereo who wants to repeat the ultrasound? Okay, we're about half and half for stereo and repeat the ultrasound. Repeated the ultrasound at a different facility, lo and behold, there it is. What are some of the differences? The depth is a little bit different. The compression is a little bit different. The focal zone is now appropriately placed higher. This the field of depth is correct. Perhaps we're in a different spot. It's not far, it's only two centimeters from the nipple. There really is not much reason why the first individual didn't see it. Scanning technique is absolutely critical to be meticulous with your scanning technique both with depth, both with nearfield, farfield and focal zone to be certain that you're not gonna miss lesions with somebody else can obviously find, shortly thereafter. This was another example of an invasive ductal cancer.

Case 8: Correlation with Palpable Mass

Now how about correlation? This is going along the same theme. This is a 52-year-old who presents with a palpable mass. It's marked with a triangle here. Here's her mammogram. Now for that palpable mass, we think we're gonna do ultrasound. I think everybody would agree that ultrasound would be our next step for 52-year-old with a palpable mass where she's extremely dense and we don't know. She has big things here, big things here, big things here. But here's where the palpable mass is. Here's her ultrasound. The ultrasound that was presented from the outside facility consisted of six images. They're all labeled six o'clock one centimeter from the nipple. If we remember where that palpable marker was, does that make sense for where this abnormality was? Yes, no. Okay, majority are saying no. Okay, but here's a nice cyst and she was told great news, you have a cyst. She wasn't told we need to redo the ultrasound. She wasn't told to come back in six months. She was told everything is great, you have a cyst, come back whenever. Which is what she did except one year later she went to a new physician. That physician called and says, I have this new patient with a hard palpable mass. Here, let's look at the ultrasound that we did. Here's her left breast. Now seven to eight o'clock, two centimeters from the nipple. We have one irregular mass heterogeneous looks very worrisome. Here's the same mass now at three o'clock, two centimeters from the nipple there's a second mass. And then at nine o'clock, three centimeters from the nipple there's a third mass. Knowing we were gonna end up with multiple issues, she had an MRI. You can see there's extensive abnormal background enhancement as well as extensive abnormalities. She underwent multiple biopsies, all were DCIS or invasive ductal and lobular carcinoma and one biopsy on the contralateral side, which yielded atypical lobular hyperplasia.

Why Masses Become Palpable

Let's digress for a minute. Why do masses become palpable? Why would a 52-year-old be complaining of a palpable mass? The obvious one is that there is a new mass or maybe there's a mass that's been there all along, but now it's undergone some interval growth. Perhaps she just started doing self breast exams or maybe a new medical provider felt something that a previous provider had not. Maybe the patient lost a lot of weight or gained some weight and now is feeling something different. Maybe the breast tissue got softer as she got older. If the prior imaging exists and it's a new finding for you, you don't really don't wanna follow it unless you've got a real specific etiology. If you know it's a lymph node, if it's a clear oil cyst, then you can follow it. But otherwise you can get out that biopsy needle. Follow Dr. Ros's rules of how to biopsy something. It's not that difficult for the patient if you do it well.

Case 9: Failure to See Subtle Findings

Let's see another case. Sometimes it's not that we don't know what to do with something, sometimes we just don't see things. And why do we do that? And perhaps we're not following a consistent hanging protocol. Maybe our search patterns are not consistent and we're not seeing something or sometimes we just fail to see it. Maybe it's on the image, we just don't see it. It could be subtle like a small non spiculated mass. Maybe it's subtle architectural distortion that we're not seeing and not marked by CAD subtly. Symmetries can be very difficult to see, or small clusters of faint calcifications could be difficult to see. Here, if we look at time zero, I would, I would say this basically is no finding here. We have these benign calcifications down here. Here's 18 months later and I would draw your attention. Even the lights are on. It may be difficult to see, but I'll draw your attention to this area over here. This was red as normal. Now here's another year later. Again with the lights on, it may be difficult, but I'll draw your attention to this area here another year later. Here's where we are. Also read as normal here. What do you think? Normal or abnormal? Who wants to biopsy? Okay, most people want a biopsy, but the person who was reading this did read this as normal and this actually came to light as part of routine quality assurance. That was biopsied and it was another case of DCIS.

Case 10: Stable Finding Over Three Years

What about this case? Here are, we're going to look at three years of craniocaudal views. Her patient, her mid fifties who was seen for screening. Here's year one, this was read as normal. Here's year two. Same finding here. Here's year three. What do you wanna do? BIRADS two, it's stable. BIRADS three. It's probably benign. It looks like it's been there for a while. Or birads four. Who's birads two. It's been stable. Everybody knew the title of this talk so I can't trick anybody here who wants to put in a birads three. Probably benign. And who wants to biopsy it? Okay, the overwhelming majority of people wanna biopsy it. She went on to ultrasound. This was a great ultrasound, fine, a little hypo coic mass. You can see there's a little bit of echogenic rim around it. A little bit of shadowing here. This was biopsied under ultrasound guidance. Tissue marker was placed in, in accordance with the tissue marker is where the abnormality had been. And this did yield to tubular carcinoma. Tubular carcinomas are relatively, uncommon type of breast cancer. Usually have a great prognosis. Usually they're low grade, they usually slow growing as this one can see. Very little change over three years. Usually the patients are over 50.

Case 11: Palpable Lump with Implant

Here's the opposite end of the spectrum. This patient had a palpable lump for a year. We can't control what our patients are told, but we really need to try to educate our clinicians about what they can order and when they should order. She is an implant that she had placed maybe 20 or 30 years before. And for multiple years she had been complaining that it just didn't feel right on one side. That one side felt different. But what she was repeatedly told, this was normal for her old implant, it was an aging implant. And not to worry about it, you couldn't get imaging with an implant. Just forget it. No imaging was ordered for five years. Here's five years later. We're seeing a lot less of this. But there have been multiple lawsuits over many years for failure to order. Usually it's in young women who gynecologists will dismiss the complaint and say it feels like assisted is assisted. Not order imaging. But this is a case where an older woman was complaining that something just didn't feel right, but yet was basically told her complaints were reveal and not taken seriously. And therefore this cancer was allowed to grow. And obviously we see all the features of a large cancer. Its shape is irregular, it's very hypervascular, its margins are irregular, it's got calcifications within and there's a satellite lesion as well. And here we can look at her lymph nodes, which were abnormal as well. By ultrasound, they're abnormal as well. We are routinely biopsying these abnormal lymph nodes and leaving tissue markers in the lymph nodes when we biopsy them as well.

Case 12: Post-Mastectomy Recurrence

Here's case. This is a 40-year-old, she had a recent diagnosis of left breast cancer and here's her ultrasound images. We've got a lot of areas of concern. Left breast, two o'clock 15 centimeters from the nipple, three o'clock, 15 centimeters from the nipple, three o'clock 20 centimeters from the nipple, four o'clock 14 centimeters from the nipple. Clearly looking at multicentric disease, her aary ultrasound should abnormal lymph nodes and they were biopsied and that demonstrated metastatic adenocarcinoma. She underwent neoadjuvant chemotherapy, bilateral mastectomy and axillary dissection. Eight months later, tissue expanders are in place but are deflated and she complains of a mass in the left chest wall near the tissue expander. Here's one example of the ultrasound image. Here's the skin and this is right over the area that she's complaining about. We have a choice. Is this scar or is this abnormality? Remember she's had a mastectomy eight months prior. Who votes for benign scar by reds? Two. How about it's probably a scar by RADS three. I can't get any takers for that. Who wants to biopsy it? Okay, everybody wants to biopsy it and why? It's irregular. And even though a portion of it looks like it's going to the skin, it doesn't have the appearance of a normal scar. It's irregularly shaped. Majority is not going to the skin. And actually it's invasive ductal carcinoma despite her mastectomy.

Case 13: Palpable Mass with Calcifications

Okay, here's another patient with a palpable mass. She's 58 years old, we don't have the mammogram, but there are extensive linear and punctate calcifications in the upper outer quadrant of the left breast in an area of increased density. And here's the ultrasound that she had to go along with that. Remember she has a palpable mass. This is labeled left breast three o'clock, left breast four o'clock. This was read as negative. Do you think it was negative? Yes. No, maybe, maybe. Here's three weeks later. Same patient, different facility. Here's a big mass. I don't think anybody would've trouble here. There's a big huge mass, very heterogeneous with shadowing. Its shape is irregular, its margins are irregular. It's involving the adjacent tissue. It's going out into the periphery here. It's three o'clock, five, six centimeters from the nipple. Slightly different labeling than the prior ultrasound. Why, why were some of the reasons why this was missed by the first radiology regroup? One reason is that they didn't know that there was a palpable abnormality. The report in fact stayed. This was a screening mammogram and screening ultrasound. Somehow that information was not correctly conveyed. Technologists did not, not appropriate label. The images, the distance from a nipple or anti radial images were not included. The ultrasound technologist didn't mark down that there was a palpable abnormality. She missed the mass and missed the calcifications. And there's no indication the report that the radiologist knew of the history or knew or went in and scanned themselves. Not knowing the history when it's available is a huge problem. How many people have intake forms that patients fill out or patients sign? Most people who doesn't, okay, everybody really does, which is obviously the only way that you can protect yourself. If the patient were to say, I said there was a palpable mass. And if you say, I didn't know that if your patient then had indicated no problems, I'm here for screening, you would be able to defend yourself in that situation. Here the surgeon biopsied the palpable mass, which was an invasive duct carcinoma and ordered the MRI. Theoretically, this large cancer was missed by the screening ultrasound. It actually was not. Theoretically it was missed by the screening ultrasound.

Case 14: Post-Mastectomy Spot Compression

Here's another case of a 43-year-old who is on in vitro fertilization. She had a prior history of inflammatory breast cancer and undergone a mastectomy. Here's two years ago, here's now. We're looking, really looking here. Here's two years ago, here's now, here's spot compression obtained now. What do you think Is this normal abnormal birads? 3, 2, 4. Who wants a two? Who wants a three? Who wants a zero? Okay, most people want a zero. Unfortunately she didn't come to anybody here. It was read as a birads too. The spots were felt to be normal. No ultrasound was done. Six months later she has this ultrasound. It's not clear why she had the ultrasound, but she probably was complaining of something palpable. And this is obvious BIRADS five lesion here, anti-parallel, irregular shaped irregular margins. Tissue marker is placed. Where's our tissue marker? Right here, right where was spotted before. It's a very difficult to defend not having done the ultrasound six months previously.

Lesson on Spots in Dense Breasts

The lesson to be learned here is beware of spots in, dense breasts. Spots can be falsely negative. You really wanna think about either adding tomosynthesis or in this case ultrasound clearly showed the abnormality.

Thank you very much.