 Hi, this is Tom Stavros and today I'd like to talk to you about ultrasound in evaluating nipple discharge and intraductal papillary lesions. I think we can all agree that mammography is not very often helpful in nipple discharge. I feel the galactography is the gold standard, but in some centers MRI galactography is replaced standard mammographic galactography. We're going to talk about ultrasound primarily because of its role in guiding ultrasound and guided directional vacuum assisted biopsy, which is I think the ideal way to biopsy these lesions. Now, even though I'm an advocate of galactography, I reserve it for what we consider high risk secretions. I don't use it in low risk secretions. So what's high risk and what's low risk? Well, high risk is spontaneous, unilateral, single duct orifice, and the nature of the discharge is clear, serosanguonous, or frankly, bloody. They're more likely to be cancer or papillom is associated with this kind of discharge. Low risk secretions are expressible only bilateral for multiple ecto orifices and the nature of the discharge is usually greenish or milky. These are usually associated with ductctasia and or fibrocystic change. Now, even though we reserve galactography for high risk secretions, I am willing to put the ultrasound probe on low risk secretions because there are cases where we find papillary lesions associated with ductctasia or fibrocystic change. And the ultrasound is cheap and there's no risk. So it's easy to apply in both groups. These are nice pictures from Laszlo Tabar. These are examples of the appearance of high risk secretions, clear, serosanguonous, and frankly, bloody. And these are low risk secretions, greenish or milky. And again, you can see this is from multiple duct orifices. And this is comedor cheezy discharge, which is sort of mixed. It can be seen with chronic perioductal mastitis, but it can also be seen with grade 3 DAB or DCIS. Now, the question is, are these secretions coming from the statue in the central Piazza of Bologna high risk or low risk? And the answer is both. They're spontaneous and clear, which is both high risk, but it's bilateral from multiple duct orifices, which is low risk. Now, we first started using ultrasound when galactography was suboptimal or failed. I've been doing galactography for over 20 years. When did this happen? Well, intermittent discharge, you know, the distal end of the duct is lined by squamous sepithelium. And keratin plugs can intermittently obstruct the flow of discharge out of the nipple. Sometimes we were unable to get into the duct. Sometimes we got into the duct and extravasated the contrast. Sometimes the duct was so large that the contrast was so dense that it obscured small papillary lesions. And sometimes there were secretions coming from multiple duct orifices, mostly low risk, but one with high risk and we got into the wrong duct. So there are all sorts of reasons why we could use ultrasound when galactography didn't give us what we wanted. This is an example of a patient who scheduled for galactogram, but the day she came in, she couldn't express secretions. You can see that her duct is dilated, filled with ecogenic fluid. And you can see the reason that her discharge has stopped is she's developed a hypercoic keratin plug within the nipple. And it's obstructing the duct so the secretions can't come out. The papillary lesion is still secreting, but the secretions can't come out of the nipple orifice because the keratin plug is obstructing the duct. These keratin plugs are part of the reasons that we use warm washcloths and ophthalmic pylocarpene to dilate the sphincter of the duct. And sometimes peripheral to central massage to try to express this keratin plug so that we can proceed with galactography. Now here's an example where the galactogram failed. It did show an inter ductal papillary lesion, showed in the upper left arrow, but ultrasound showed five different papillary lesions. And these other four papillary lesions were not visible because the duct was so large, the contrast column so thick that the small papillary lesions were not visible on the galactogram, but were visible on ultrasound. Now today we scheduled the patient for a combined ultrasound and galactogram. I always performed the ultrasound first, but regardless of whether the ultrasound's positive or negative, I always do the galactogram because there are weaknesses of ultrasound that the galactogram can fill in for. Now if galactography shows a lesion that was missed on ultrasound, in other words if the ultrasound was falsely negative, we leave the needle in, take the patient back to the ultrasound room and inject the duct with sterile saline to demonstrate the lesion on ultrasound so that we can proceed with the ultrasound-guided vacuum assisted biopsy. It can be very helpful as a learning tool to look after the galactogram. So here's a case where there's an inter ductal papillary lesion, but there's loose stromal tissue around ducts and do I know that this isn't just loose stromal tissue around a duct? No, I don't, but after we do the galactogram and I repeat the ultrasound, I can clearly see that there's an inter ductal papillary lesion about a centimeter and a half long in the peripheral part of this duct. Here's another example where the ultrasound was falsely negative. We did the galactogram, it showed a tiny inter ductal papillary lesion, so I took the patient back in the room with the needle still in the nipple and injected sterile saline, and then we could clearly show a small inter ductal papillary lesion that we'd missed before the galactogram and then we could proceed with the ultrasound-guided vacuum assisted biopsy. So, you know, anytime an ultrasound is negative and the galactogram shows something, this is very valuable because the best way to biopsy these is with ultrasound-guided biopsy. Now, literature talks about papillomas as being ecogenic. Ecogenic is a garbage bag term, it doesn't mean anything. To me, ecogenic sort of suggests hypercoic, but papillomas are virtually never hypercoic unless they're micropapilli. Now, most papillomas are iso-echoic or heterogeneous. This is a typical appearance of a papilloma. You can see that on the left in the radial view, it's about the same ecogenesis as the surrounding fat and definitely less ecogenic than the periductal fibrous tissue. Notice that I've shown the anti-radial view as well. It's always important you have both views because you can volume average periductal white fibrous tissue with black intraluminal fluid and come up with a pseudo-iso-coic lesion in the radial view. But if you can show an intraductal papillation in the anti-radial view as well as the radial view, then you can be sure it's real and not a false positive. Now, most intraductal papillomas of the size that we see that cause nipple discharge are iso-coic. Larger papillomas that cause palpable lumps from mammographic abnormalities often are heterogeneous in ecotexture and contain calcations. One of the problems with lactography is it often just shows a duct cutoff sign. So we're looking on the left, there's a duct cutoff sign. The duct is obstructed by a papillary lesion, but I don't know if this is a tiny grain of rice papilloma or whether this is something that fills an entire segment. Ultrasound clearly shows that this is only 9mm long. It shows me that it's a short papilloma. It shows me that I can even include the entire papilloma in a half aperture or a 10mm aperture DVAP probe. I don't even need the full aperture. Now, there is a way on lactography to get contrast past this papilloma into the more peripheral part of the duct, and that's just a massage from central to peripheral. And by doing that for a few minutes, many times you can actually get contrast peripherally beyond the papillom and show its extent on the lactography as well. Now, anytime you're going to evaluate the ducts for a cause for nipple discharge, you have to see the entire duct, including the part within the nipple. So, we've developed a combination of three maneuvers that we call peripheral compression, two-ended compression, a rolled nipple technique. It's really all combined into a single maneuver, but it allows us to roll the nipple on its side and see the tissue planes within the nipple and the ducts within the nipple to exclude a so-called nipple adenomeral papilloma that lies within the nipple. Not in all patients, but in many patients when you do a straight anterior approach because of the thickness of the skin, the wrinkling of the skin, trapped air bubbles and whatnot, you can get acoustic shadowing that obscures the intranipple and immediate retrorillar part of the duct. And it's possible in some patients to have as much as a centimeter and a half of the distal duct obscured by shadowing. Here we can see an example of this. We can follow an ectatic duct up toward the nipple, but we can't tell what's going on in the immediate subarilla region or within the nipple from a straight anterior approach. Well by doing the rolled nipple maneuver, we roll the nipple on its side. So this is a nipple that's rolled on its side. The tip of the nipple is over on the left and there's an air bubble at the very far left. And then behind the nipple is my index finger. And we can see two ducts entering the base of the nipple from the right side. But we lose them about halfway through the nipple due to volume averaging, near field volume averaging. This is done with a one-dimensional array probe which is focused at about 1.5 centimeters in depth. So once we get inside 5 millimeters or closer to the skin than 5 millimeters, the beam hasn't yet become focused and so there's a lot of volume averaging. This is the same patient, same duct rolled nipple maneuver using a matrix array probe which focuses much more quickly and more tightly close to the skin. And we can now follow both of these ducts all the way through the nipple to their orifice on the surface of the nipple. So we definitely like to combine the rolled nipple maneuver with a matrix array transducer. I would also add as a technical point that harmonics is, we use that widely. It's our default setting for breast ultrasound, but harmonics does not help in the immediate near field. So we can actually see these nipples better with harmonics turned off. So I turn off tissue harmonics when I'm doing the rolled nipple maneuver and evaluating patients with nipple discharge. So let's just look at a few examples of where the rolled nipple maneuver helped us. Here's a patient with nipple discharge, enter a view of the nipple. Maybe there's something inside the nipple. It probably is, but it's hard to be sure. But when we do the rolled nipple maneuver, we can see that there's a long subarilla papilloma shown by the white arrowheads. But to the right in front of IF, which is my index finger, is a small second papilloma entirely within the nipple. Now I can access, I can get at that subarilla papilloma with a DVAB problem, but I can't get at the intranipple papilloma. So this is really not a great case for DAVB. This would require bivalving in the nipple or galactoscopic biopsy and or removal in order to make a diagnosis or treat the discharge. Now, here's a similar case from an anterior view. I see a subarilla papilloma between the arrowheads that I can easily get to with a directional vacuum-assisted biopsy probe, but there's a question of a second papilloma shown by the arrow within the nipple. But when I do the rolled nipple maneuver, I can see that indeed there is a subarilla papilloma, as I originally thought, but what might have been a second intranipple papilloma is merely nipple tissue between two ectatic ducts. So this is an ideal case for DAVB. There is just a single subarilla papilloma accessible by DVAB, and there's nothing inside the nipple. Now, here's a patient where with two-handed compression, I can see a nipple intranipple papilloma or nipple adenoma, as it's called, but you can see how much better I can demonstrate it with the rolled nipple maneuver on both radial and anti-radial views. Clearly, this is not a DVAB case. There's no way I can get at this with a vacuum probe. This is going to require a surgical bivalving in the nipple or galactoscopic biopsy and removal. Now, when you're evaluating patients with nipple discharge, duct ectasia can be a cause, or there could be blood in the duct caused by papillary lesion or DCIS, so there could be ecogenic fluid within the duct. And that means that in this case, there's three things this could be. This could just be ecogenic fluid within the duct. It could be a long papilloma or it could be DCIS. Now, this is where dynamic maneuvers come in. Here I'm doing a heel and toe-blotment maneuver, and I'm clearly showing that these are all just ecogenic secretions, so what could have been a bioregis for papilloma or in sight to DCIS DAB lesion is shown to be a bioregis to duct ectasia. And here we can see on the Tabar 3D slide this one duct contains this yellow, very thick proteinaceous fluid, but the duct immediately adjacent to it doesn't have that, so ecogeneity can vary from duct to duct within the breast. Another way of showing this, I mean, I just showed you a stored video loop with blotment maneuver. A second way to show this is by using blotment maneuvers with color Doppler. And here in this case, as I'm pushing down on the duct, I'm forcing the secretions posteriorly, the ecogenic secretions posteriorly, they're ecogenic enough to create a Doppler signal, and you can see I'm getting a red signal. When I release compression, the ecogenic secretions come back toward the nipple, so I'm getting a blue signal. So by doing split screens with and without compression on a single frozen image, I can demonstrate the same thing that I can by storing a video loop in grayscale. I prefer the video loop in grayscale, most of us can store videos now, so that's the preferred way to do it, but if you just have hard copy, this is the second way to do it. Now, the fat or the protein, and that's what causes the ecos in the secretions, concentrated fat and or proteanaceous debris within the duct, remember that fluid in the duct can be resorbed through the duct wall, and so if ductotasia becomes chronic, whatever protein or fat and or fat that lies within that fluid cannot get resorbed through the wall, but the fluid can. So over time, the fluid tends to become thicker and more concentrated and more ecogenic. Well, the ecos can be diffusely distributed throughout the duct, or they can layer anteriorly if it's fat or posteriorly if it's debris, and that can make it very difficult to distinguish from an intraductal papillary lesion. So here we have a fat fluid level, the fat is layered anteriorly because the patient is lying on her back, and it's very difficult to distinguish from this centrally located papilloma. Here we have proteanaceous debris that's fallen to the posterior part of the duct, again because the patient is lying on her back, and it's very difficult to distinguish from this more peripherally located papilloma. And you can see again on this beautiful Tabar 3D that proteanaceous fluid can indeed layer out and create a protein level that can be difficult to distinguish on ultrasound from a papillary lesion. So here's another example where dynamic maneuvers help. This is an ectatic duct. The question is, is there an expansal bilob papilloma there, or is this just a proteanaceous debris level in an extremely ectatic duct? Well again, if I do heel and toe allotment, I can clearly show that this is all just ecogenic debris, that there's no true inter ductal papillary lesion. And again, a byred's 4-ult sound becomes a byred's 2-ult sound simply by applying dynamic maneuvers. Now, I will say that there's a situation in which we're handcuffed with all of our imaging techniques, and that's with these thick, cheesy secretions that can be due to either chronic periductal mastitis, or grade 3 DCIS with commido necrosis. The fluid in chronic periductal mastitis can be so ecogenic that it's difficult to tell from this DCIS with commido necrosis. Now, 20% of these people are chronically colonized with bacteria, and the only infections I've ever seen after all sound-guided biopsy occurred in patients where it was exactly this differential diagnosis, chronic periductal mastitis versus commido DCIS. And so I've learned to cover these patients with prophylactic antibiotics I use augmentin two days before the day of and two days after the procedure. Now, this is a difficult problem for all of our imaging tests. In galactography, the secretions are so thick that they create a filling defect difficult to tell from papilloma or DCIS. On ultrasound, the filling defects are so thick ecogenic that, again, we have hard time telling whether this is DCIS or papilloma. And on MMR, there's so much intense periductal enhancement that it's difficult to tell from grade 3 DCIS. So all of the three procedures that we used to evaluate nipple discharge failed to definitively distinguish chronic periductal mastitis from grade 3 DAB with commido necrosis. And so we basically just have to biopsy these, but I really do want to cover these people with antibiotics because they're the exact group of operative infection in some cases. Now, Doppler can be very helpful in making the distinction between inspecated secretions and a papillary lesion. Part of the histopathologic diagnosis of papilloma is a fibrovascular stalk. And even small intraductal papillary lesions have a vascular stalk. And if we scan with a very light scan technique with no compression, we can usually see these small intraductal papilloma vascular stalks. Now, this is a case where on the left with the grayscale image this could be a long, preferably located papilloma or it could just be a debris level. Well, when I put on colored Doppler I can see it clearly had a vascular stalk so this was a long papilloma. So in this case Doppler upgraded the diagnosis. Now, it's possible to get artifactual color. These are small vascular stalks, small lesions. But it's helpful to put pulse Doppler on this and make sure that you have an arterial or venous waveform. If it's artifactual, you get a motorboat artifact where it just goes... Now, only a positive Doppler helps you when you're evaluating patients with nipple discharge. Papilloma is frequently undergo hemorrhagic infarction and hyalinize. And obviously an infarcted hyalinize papilloma will not show vascular stalk. So if you see a vascular stalk, you know for sure you're dealing with an intraductal papillary lesion. If you see a vascular stalk on Doppler, it doesn't mean it's not a papilloma. It could still be a hyalinized infarcted papilloma. So, as is always the case in breast ultrasound, you can believe a positive Doppler more than you can believe a negative Doppler. Now, one thing I will warn you about, most of the literature on papilloma says that large duct papillomas are single most of the time. And the reason the literature says that it's based on galactography, which only looks at one low bar duct at a time. But I can tell you with ultrasound, you see multiple large duct papillomas in different segments or lobes of the duct far more often than the galactographic literature would suggest, because with ultrasound we're looking at all the low bar ducts with galactography, we're only looking at one. Now, pictures on the right are all papillomas that might present with a papilloma or mammographic abnormality. The ones on the left, one through five, are the spectrum of sonographic appearances of papillomas that cause nipple discharge. And they're basically an order of frequency. So, number one would be the most common appearance where you see a small papillom with fluid both central and peripheral. Number two would be the second most common where you see a papillom and you only see fluid centrally toward the nipple, but nothing peripherally away from the nipple. Number three would be a long papilloma that might be longer, even than a two centimeter aperture of a DVA probe. Number four would be a long papillom involving a couple of branches, and number five would be a longer papillary lesion involving multiple branch ducts. Now, obviously papilloma is not the only etiology for nipple discharge, and even though ultrasound can be better or more definitive than galactography in some benign papillomas, you know, we have carcinoma which can be in situ or invasive or a mixture of in situ and invasive. You can have papillary duct hyperplasia, atypical duct hyperplasia, duct dictasia, communicating cysts, hyperprolectinemia, and idiopathic. For many of these things, galactography can be better than ultrasound, and that's why I haven't completely given up on galactography. This is an example of mixed invasive DAB, DCIS, cancer. Between the thin arrows, we can see a taller than wide angular lesion with a thick hecrogenic halo, which is the invasive part of the lesion. Over on the right is N, which is the nipple, and in front of the white arrow head is the DCIS part of the lesion. This patient had frank bloody discharge. Now, you might ask, can you replace mammographic galactography with MR galactography? And I would say maybe. I mean, clearly MR's contrast resolution is better, but its spatial resolution is less than galactography, so we can expect galactography to have resolution of 700 to 100 microns, depending on the mammographic resolution. Whereas MR resolution is going to be somewhere in the 600 to 2,000 micron range depending on whether we use submillimeter isotropic voxels or whether we're using two or three millimeter slices. Now, the important thing about MR is it's not just about spatial resolution. MR may not need equal spatial resolution to galactography because let's just look at micropapillary carcinoma site to one of the more common malignant causes of nipple discharge. MRI's better contrast resolution can show the abnormal proteanase fluid within the duct lumen, and its dynamic contrast enhancement can show abnormal perioductal vascularity even though MR is unable to resolve the individual micropapillary. So, you know, MR can make a definitive diagnosis of micropapillary DAB, even though it doesn't have the resolution to show the individual micropapillary. But I want to show you three examples of why I've never stopped performing galactography. Case one, micropapillary DAB presenting with nipple discharge and this shows simply that the spatial resolution of galactography is better than ultrasound and therefore it can show micropapillary in some cases where ultrasound can't. So, here's a patient with nipple discharge the left shows the ultrasound it looked like a small intraductal papillary lesion sorry about that but when we do the galactogram I can see that there's multiple micropapillary lesions in the region where we showed the papilloma and if we look at the circle that's deeper into the right we can see that there's micropapillary extending way peripherally into this lobe involving essentially an entire segment which I couldn't see with ultrasound since because ultrasound has two or three hundred micron resolution up to a thousand micron resolution depending on the machine whether or not it has spatial compounding the depth whereas mammography, at least the mammography we had had 70 micron resolution so a big difference in spatial resolution galactography can show micropapillary better than ultrasound can and on this tabar 3D picture you can see how tiny these macropapillary are, it's not surprising that ultrasound can't resolve them now in this particular case Doppler was helpful because even though the ultrasound findings were not suspicious the Doppler findings were this has far more vascularity more vascular stalks than I would expect with a benign papilloma and there's also peridectal vascularity so Doppler can sometimes help us here when it's positive again negative doesn't help us this too, micropapillary DAB presenting as nipple discharge now this just shows that ultrasound does very well within a probe length of the nipple but when you get farther peripheral than a probe length into more distal or more peripheral and smaller branch ducts, ultrasound starts to have more difficulty so this patient presented with nipple discharge I was easily able to show an intraductal papillary lesion within a probe width of the nipple 3D showed it was involving branch ducts in more than 2 centimeters in length which I'll show you later as one of our higher risk findings but the galactogram it did show the same papillary lesion so I'm showing that within the box now which is what we saw in ultrasound but it showed innumerable micropapillary involving an entire lobe I mean just crazy now when I went back after the galactogram with the needle still in an injected sterile saline then in retrospect I could see some of the micropapillary on ultrasound and this thick mip 3D reconstruction but I totally miss them so the moral of the story is here if you get farther than a probe width from the nipple into smaller branch ducts or deeper in the breast, ultrasound does less well than it does near the nipple within a probe width and this is what I'm presenting as nipple discharge micropapillary DAB again why I still do galactography so patient presented with nipple discharge I did ultrasound it showed a couple of intraductal papillary lesions now normally I would have done a galactogram but I just moved to a new practice I'd given them a list of my galactograph equipment I needed a couple months before I got there but somehow they hadn't obtained it so I was not able to go to galactogram so my choices were to go straight to DVAB or to do some other follow up so I chose to go to DVAB but I do want to show you one thing here this is the periductal vascularity you know we can see one papilloma on the bottom image with a single red vascular stalk within it and we can see a second papilloma in the left upper image just below into the right of the number 2 again with a single vascular stalk but look at the tremendous amount of periductal vascularity in this patient so both of these images are from this patient now I'm going to show you two other patients that had confirmed benign papillomas and notice that there's just a single vascular stalk and no periductal vascularity so in this particular case the ultrasound images were fairly reassuring but again the Doppler images were suspicious the positive Doppler is always going to be more valuable than negative again I didn't have the galactograph equipment didn't want to send her to surgery we didn't have MR then it was a mobile MR and it just wasn't in town that day so I did galactography I would have done galactography but they didn't have the equipment so I went straight to DVAB here's the clip and the benign diagnosis was intraductal papilloma with single vascular stalks were indeed benign intraductal papillomas so that was certainly concordant with the grayscale image maybe discordant with the Doppler which showed tremendous periductal vascularity I wasn't quite comfortable so I asked the patient to come back in six months rather than the usual one year we wait after a benign diagnosis but two months later she returned with a bb-sized palpable lump at first glance you might think it's a sebaceous cyst because it's so superficial but if you look carefully there's a duct coming out of it and it's tremendously vascular and if we do a video we can see that that duct coming out of it goes posteriorly and then it goes into some more dilated ducts more intraductal papillary lesions then it folds back and goes back the other way so it looks like because the patient is supine we might have a segmental duct that's just to sort of wrinkle because she's lying on her back and this might be involving an entire segment so when I get some other still images from deeper parts of the lesion we can see that it's tremendously vascular there's an expansile intraductal papillary lesion then the second fold is on the left image and the third fold is on the third image so you can see that there's multiple ducts they're beaded, they're tortuous, they're angular this really looks to me like a micropapillary DCIS or DAB so I did a second DAB here you can see where my second clip is it's deeper and again, this points out what I told you and over and over and over we're going to see this if you're within a probe width of the nipple, ultrasound does pretty well but when you get farther away, deeper in the breast, in smaller branch ducts ultrasound starts to have some trouble the galactography can save the day on so the diagnosis from the second biopsy was micropapillary DAB or micropapillary DCIS nuclear grade 3 so the first one benign papilloma, the second was micropapillary DAB grade 3 and the moral of the story is ultrasound does better near the nipple than it does in more peripheral and smaller branch duct lesions now here's the MR and you can see that an entire lobe is indeed involved because she's hanging prone, the duct is straightened out it's not wrinkled back and forth upon itself I like the subtraction view because you can see that all this high signal is actually just proteinaceous fluid that's subtracted out giving us a negative subtraction view there now, I want to put the MR up next to the ultrasound and point out that because the patient is prone with hanging breast on MR, this entire segmental or low bar process straightens out because the patient's lying on her back with ultrasound this duct is folded back and forth upon itself, which was part of the reason it was difficult for us to evaluate it on ultrasound an entire lobe can be involved but it's not all straight and linear like we see on MR it's folded back and forth upon itself on ultrasound ok, so those three cases show why I haven't given up with galactography why I continued to do galactography now what about other causes that nipple discharge other than papillom well, duct dictasia is a cause this particular case it's so severe, we did a prolactin level and found out that she had pituitary adenoma and hyperprolactinemia what I want to point out here is that the ectogenicity of the fluid within the ducts can vary greatly so we're going from hypocoic to modally hypocoic to isocoic to hyperocoic another thing I'd like to point out is everybody's aware that you can get severe duct dictasia with hyperprolactinemia it's my experience that people with hyperprolactinemia also get multiple papillomas and it's very, very difficult the more ecogenic the fluid is and the more dilated ecogenic ducts there are the more difficult it is to exclude a papillary lesion within the ducts and again on this Tabar image you can see severe duct dictasia the ducts with yellow have highly proteanaceous fluid the ducts that are clear have much thinner runnier fluid which correspond to the more ecogenic ducts and the ducts with what appear to be clear fluid on the Tabar 3D image I didn't put Tabar's name on this slide so I have to give Laszlo credit but I forgot to do it on this slide sorry Laszlo now it's very important that you understand that duct dictasia is a diagnosis of exclusion there can be tremendous duct dictasia associated with DCIS micropapillary DAB or with papillomas so here's a case where the sonographer came in and said it's just duct dictasia and I said well, yeah the ducts are certainly dilated but duct dictasia is a diagnosis of exclusion you haven't shown me the whole duct as far peripherally as you can see it so when you went in the room and scanned sure enough there was a large papilloma causing this severe duct dictasia so don't be lazy with duct dictasia it's a diagnosis of exclusion you have to interrogate the entire lumen of the duct before you come to the conclusion that duct dictasia is the cause of nipple discharge now in general duct dictasia is going to cause milky or usually greenish secretions but it can cause bloody secretions if there is acute or chronic periductal mastitis now the inflammation occurs one segmental or low bar duct at a time it doesn't involve all the ducts so if you have multiple low bar or segmental ectatic ducts only one that's inflamed will look different so on the right we can see an ectatic duct with a thin wall and on the outside we can see it's thin wall and then on the left we can see what appears to be isochoric thickening of the duct wall which in fact is periductal plasma cell mastitis so what appears to be a thickened wall is actually periductal inflammation so this is an example of how a duct can look on grayscale ultrasound when there's acute inflammation now you can certainly get hyperemia with inflammation so you can get increased vascularity with periductal mastitis or with papillary lesions the important thing is when you see a vessel with a papilloma it's merely passing through the duct wall to feed the papilloma inside the vessel so it tends to be oriented perpendicular to the duct wall so on the right you can see this red vessel is roughly perpendicular to the duct wall it's simply passing through the duct wall to supply the benign and periductal papilloma on the left we see acute periductal mastitis there's tremendous vascularity it's not inside the duct in the periductal tissues and because it's supplying the periductal inflammation the orientation of the vessel tends to be parallel to the duct lumen now you have to be somewhat careful because the only place that vascular stock that supplies the papilloma is going to be perpendicular to the duct wall is where it passes through the duct wall then as the papilloma grows down the duct toward the nipple or up the duct away from the dipel it can drag its vascular stock with it so the right image is showing where the vascular stock passes through the duct wall the left image is showing the vascular stock as it is being dragged toward the nipple on the left or away from the nipple on the right now one thing I'd point out we're going to talk about DVAB later but this is a good time to talk about what to target with the AVB you can take out a 3 cm fibridinoma completely and not get a drop of blood and you can take out a 5 mm papilloma and get a lot of ecomosis and maybe even hematoma papilloma is being nearer the nipple there's just more background breast vascularity near the nipple but papillomas themselves are very vascular and I specifically target the vascular stock it's important that you understand that this papilloma is not attached to any place else the rest of the papilloma within the duct is completely unattached the only point of attachment is where this vascular stock penetrates through the wall so I specifically target that with my DVAB because I know I'm going to get the whole thing in the first specimen if I get the vascular stock the entire thing is just going to suck right back out of the duct like you suck a piece of spaghetti in now I mentioned that in inflammation the vascularity is parallel to the duct wall but it's outside the duct wall so on the left we can see a long parallel vessel on this radial view that's outside the duct wall posterior to it that's a periductal mastitis I do have a vessel that's parallel in the right image but it's inside the papilloma that's a vascular stock that's been dragged down the duct as the papilloma grew toward the nipple so they're both parallel but one is inside the duct the papillary lesion and one is outside the duct the periductal mastitis now another potential cause for nipple discharge is communicating cysts these cysts invariably have papillary aprican metaplasia so they invariably appear complex on ultrasound they may just have tiny excrescences like the one on the right or they may be almost completely filled with papillary aprican metaplasia as we see on the one on the left these typically have a trigger point history the discharge is often intermittent so during certain phases of the menstrual cycle when there's more activity and more secretions the cyst may enlarge and become tender and then when they start to drain out through the nipple with nipple discharge the lump can get smaller and go away also if you push on that cyst you can get projectile secretions so if you want to push on a cyst to see if it's communicating with the duct or causing nipple discharge put on your safety goggles because I guarantee you can hit the ceiling with these discharge if you push on something with a large diameter and the fluid has to egress through something with a tiny diameter there's tremendous acceleration by the venturi principle that's going to cause this thing to accelerate and they can get very projectile discharge now many times the duct communicating so one thing you can do is put a color Doppler box on part of the duct between the nipple and the cyst and so here when I push down on the cyst I get secretions going one direction and when I release I get secretions going the other direction so again with split screen color images I can document that indeed this torturous duct communicates with the cyst now why do we recommend DVAB rather than core biopsy in almost all cases with DVAB the core biopsy specimens are smaller, they're macerated the pathologist is less likely to be satisfied and more likely to recommend surgical excision to get an adequate specimen we know that ultrasound guided DVAB if you do it right removes the large duct papillom and the first specimen about 90% of cases as long as we get the fibrovascular stock and the specimen is much less macerated then core biopsy rarely removes the whole lesion and rarely stops the discharge now the pathologist should not recommend excisional biopsy if we've already excised the lesion with ultrasound guided directional vacuum assisted biopsy but we have to tell him that we've removed it and specifically request that he serial section he or she serial section the entire specimen because it's really not kosher to recommend excision when it's already been excised the important thing is we found that DVAB can stop secretions in about 90% of patients so we don't just get a diagnosis but we actually stop the secretions and that's not the case with core biopsy so here's an example of an inter-article papillary lesion with a vascular stock here we're putting the probe in deep to the lesion and now I've removed everything all the way up to the front front of the duct and I've deployed a marker what we should always do you have to deploy a marker because if there's a tippy or malignancy you're going to have to get back to the spot for surgical excision now it's important to know what the pathologist does with your specimens and most pathologists take only 3 slices each 4 microns thick separated by 40 microns and if you do a DVAB and you completely remove papilloma they might be looking at 1 tenth of a percent or 1% of the lesion recommend surgical excision well that's unacceptable because you've already excised the lesion so the way we finally adapted to this we actually changed our pathology requisition form and we put a slot on there for a percentage of lesion removed and then a checkbox, please serial section whole specimen and that was only partially successful but at least it's a better form of communication to minimize the chances of the pathologist unnecessarily recommending surgical excision of something you've already excised now what we found with DVAB is we got a definitive diagnosis in all cases, we removed evidence of the lesion and 90% of the lesion and discharged up to the same 90% I'm not sure why that's because we completely removed the intraductal papillary lesion whether we interrupted its blood supply or whether we interrupted the duct but the bottom line is when we reviewed our surgeon's success rate it was the same, 88% and they were doing things like ligating all the subrelar ducts or segmental resections so they were doing much more extensive procedures and getting the same success rate we were getting with the DVAB more than 98% of the lesions were benign large duct papillomas but we did identify a subset of papillary lesions that had 6% chance of incite to carcinoma 7% chance of ADH so what were the group that had a low enough chance that could be considered Byred's 3 less than a 2% chance of malignancy less than or equal to 2% but they were essentially the grain of rice lesions they're typically about 5mm in length the shape of a grain of rice they're non-expansile they don't expand the duct more than the associated grift fluid in the duct they're less than 2cm long in other words they're less in length than the aperture of a normal DVAB probe they don't involve branches they don't involve TDLUs and they don't have any extra ductal findings we found that more than 98% of those were benign now as a practical matter it is a patient care if we call it Byred's 3 not if she presented with nipple discharge she didn't care whether it's Byred's 3 or 4 she wants a diagnosis and she wants it out she wants the discharge to stop I'll talk about it anyway but I will talk about it in the very conclusion cases where it might be helpful now those are the findings that I showed you that correlate with Byred's 3 but these are the findings that correlate with a 13% chance of either DCIS or etypia presence of extra ductal strong suspicious findings like destruction of the duct wall or angles presence of inter ductal weak suspicious findings like micro calcs of the duct more than the associated degree of fluid dilatation a length greater than 2 cm again greater than the length of a DVV aperture involvement of branch ducts or origin in the TDLUs or involvement of TDLUs because by definition these are peripheral papillomas rather than large duct papillomas and then adopter finding a multiple inter ductal vessels or periductal increased vascularity here's an example of destruction of a duct wall and angular margins you can see the posterior duct wall is gone here and posteriorly we see a couple of angles it's more evident when I compress so on the left is without compression it's really hard to appreciate that when I compress I can see the angles and I can see that the posterior wall has been completely chewed through and eroded these are all examples of expansion of the duct greater than the associated degree of fluid one is an invasive papillar carcinoma another is a grade 2 carcinoma a ductal carcinoma in site 2 and another one is ADH but all of these expand the duct in the high-risk category this one has 3 high-risk features it's expand cell it has microcalcifications and it has more periductal vascularity and more vascularity internally than we would expect with a benign papilloma these are longer than 2 cm and they involve branch ducts and these were all grade 3 DCIS and these are a lot of Tabar pictures showing why this finding would tend to suggest a higher chance of DAB for grade 3 DCIS now this is a peripheral papilloma and it's easy to see that because it's in a large TDLU and it's also far from the nipple so that fits our concept of what peripheral should mean we can see that there are 3 other TDLUs that are much smaller closer to the nipple and we can also see that this thing is tremendously vascular much more vascularity than we'd expect with a papilloma the literature is very clear that peripheral papillomas are much more likely to have a tippiora in site 2 carcinoma and therefore a higher risk and should not be viewed as a low risk ever like you might in some cases view a benign large duct papilloma now here's 2 papillomas side by side more superficially I have a peripheral papilloma involving a TDLU and more posteriorly I have a typical large duct papilloma now what I want to point out here the nipple is high on the right on this image which one is closer to the nipple the large duct papilloma or the peripheral papilloma the peripheral papilloma is closer to the nipple so peripheral has nothing to do with distance from the nipple it's entirely related to the fact that it arises in a TDLU not in a large duct so a large duct papilloma by definition arises in a large duct and peripheral papilloma by definition arises in a TDLU now here they are without compression you can see the TDLU shape better but I want to show you is the biologic behavior of these so here's the low risk large duct papilloma single vascular stock as I told you they typically have and here's the peripheral papilloma tremendous vascularity not just in the TDLU but the entire paeductal area this is a combination of a benign papilloma with micropapillary DAB it's been both lasso to ours experience in mind that when you see this combination of multiple benign papillomas with um micropapillary DCIS that there's often an underlying HPV hopefully this goes down with the HPV vaccine so as a practical matter almost all patients who present with a noxious nipple discharge do not care whether the intraductal papillary lesion does or does not appear to be a benign intraductal papilloma i.e. whether it appears by Ritz 3 or not they want it removed and they want the nipple discharge to stop however we're doing counter situations in which assessing the risk of an intraductal papillary lesion might be of value to the patient and or referring physician the first would be an intraductal papillary lesion found as an incidental lesion in a patient who isn't complaining of nipple discharge do you always have to biopsy it or could you maybe follow it well i think if it meets by Ritz 3 criteria maybe you could follow it but obviously you'd have to do that with informed consent the key thing i find is multiple intraductal papillol lesions that are too numerous to biopsy them all so what you do is you look for the papillomas that are largest and or highest risk you biopsy them and you follow the rest and this is not uncommon as i told you with ultrasound we find multiple papillomas far more often than the galactic graphic literature would have ever suggested because we see all the lower ducts at one time where as a galactogram sees only one duct at a time so multiple intraductal papillol lesions are not they're not common but they're not rare so in summary, ultrasound can be effective as a first test for nipple discharge, galactography or MR is useful when ultrasound fails ultrasound of the ducts within the nipple and suboreal area require special maneuvers combine peripheral compression two-handed compression and roll nipple maneuver we schedule a patient for both ultrasound and galactography we always do the ultrasound first but regardless the ultrasound findings we perform the galactogram because we know that ultrasound has weaknesses for micropapillae for lesions more distal from the nipple and in smaller branch ducts and deeper in the breast if the galactogram is positive we repeat the ultrasound with the galactographic guidance in other words leaving the galactograph needle in and injecting sterile saline while we're examining the patient with ultrasound why do we do that? because ultrasound guided DAVB is the best way to biopsy these and we need to know exactly where the lesion is and what it looks like with the DAVB and repeating the galactogram with sterile saline in the ultrasound room gives us exactly that information if the nipple is involved it's a surgical or galactoscopic problem we can't do DAVB on an intranipple lesion but that's part of the reason for doing the roll nipple maneuver ultrasound can detect causes a discharge other than papilloma but for some galactography may be better and it's why we still do galactography or MR ductctasia requires caution it's a diagnosis of exclusion just because you see ductctasia doesn't mean there isn't a papillary lesion there you need to interrogate the entire length of a dilated duct to make sure there's not a papillary lesion causing it and DAVB is definitive in the diagnosis and stops secretions 90% of the time always deploy a marker when you do DAVB because if there is a tibular malignancy you're going to have to go back spring-loaded corbiopsy is much less often definitive and almost never stops nipple discharge now it's important that you think you've removed all imaging evidence of the intranipple papillary lesion with DAVB you informed the pathologist that you've completely excised and specifically requested that they serial section the whole specimen before recommending surgical excision recommending surgical excision of a lesion that we've already excised is a ridiculously unscientific concept thank you