 So, dear friends, let's begin with the prelude session for today and we are going to have three important topics. The first will be by Dr. Raj Kedar and he'll be taking MRI and characterizing renal masses. The second will be by Dr. Ravi Thapar. So we'll be talking about MRI difficulty and then there'll be a lecture on fetal MRI. So, let's begin with the first topic for the day by Dr. Raj Kedar, Sir is professor of radiology at University of South Florida and radiologist-in-chief at Tampa General Hospital. He is director of body imaging and fellowships as well as MSK faculty vice president of Florida Radiological Society and he has several awards to his credit including the Teacher of the Year Award and also Teaching Award, Chair Awards and Golden Apple Awards. So, thank you, sir, for your kind consideration and your time. Let's begin with the talk. Good evening. It's my pleasure and honor to be part of this course and I would like to thank Dr. Patkar and Dr. Verma for inviting me to talk about characterization of renal masses with MRI. Unfortunately, incidence of renal cancer has increased by 30 to 40% over the last 30 years or so and this is predominantly due to incidental lesions due to increased imaging with ultrasound and CT scan. Although most of these lesions are less than four centimeter, increase in diagnosis has not resulted in better clinical outcome. In fact, the reported increased mortality from 1.5 to 6.5 deaths per 100,000 and this is predominantly due to over diagnosis and surgeries and aggressive management even slow-growing tumors which may or may not show growth over the time. The cost-effectives strategies are necessary especially to identify the clinically significant renal masses that could evolve into life-threatening disease versus the over-treatment of benign or indolent malignant lesions that can be observed. As we know, ultrasound is very cost effective. It is very diagnostic for simple cyst, however it has its limitations with smaller lesions. Patients body have it as characterized in complex cystic masses, differentiating from complex cystic masses to solid lesions and cannot stage the disease. CT scan is very sensitive and specific for renal masses for greater than two centimeter in diameter. However, it does have its own limitations in complex cyst, cyst containing hemorrhage or protein, legion size, intraparent camel legions that cannot be measured accurately as far as harmful density is concerned because of volume averaging and misregistration artifacts. It can be limited with renal function. So that's where MR comes in. MR has got excellent soft tissue contrast due to which we can use multiple sequences and parameters to characterize the liver legions to renal lesions. We do not need ironated contrast. Some of the patients who have allergies can be still scanned with this. In cases of renal insufficiency, now with newer microcyclic agents, we can give IV as even in end stage renal failures. My learning objectives for this talk are to discuss typical and variable features of common renal tumors, how MRI can help as a problem solving tool to characterize the renal legion by which the clinical management, planning intervention and avoiding unnecessary procedures. The plan for today would be to talk about different types of tumors, how we can characterize those with MRI. I'm not going to discuss port query of all renal masses. I'm not going to talk about inflammatory or pseudo renal masses or cystic legions. We are going to briefly talk about renal cell carcinomas and its subtypes, some typical and other typical AMLs, oncocytomas and other tumors. First of all, technique. Technique is very important. Our MR protocol is same for all abdomens. It has coronary T2, actual T2 fat set, gradient takeout in and out of phase imaging, DWI and coronal and axial Dixon. Post contrast, we typically do all these sequences in the, with post contrast images in arterial venous, three minute delay, five minute delay in coronal. And for all, we do subtractions as well. The many sequences that can really make difference in diagnosing renal as well as liver legions is subtraction in and out of phase and fat set. Remember that for renal tumors, fat is your friend and diffusion-ready images. As far as subtraction technique is concerned, it is grossly underutilized. It drives me crazy when I see outside scans where subtractions is not done. This is very simple. You just have to push a button and machine does it for you. There is no extra table time or patient time. I think it's crime not to do subtraction sequences when it is so easily available and it makes a lot of difference in characterizing these legions. So essentially in this, you are subtracting pre-contrast images from post-contrast images. So it will subtract out the hemorrhage, hyperintensity within the legion and it will just give you a pure enhancement within the legion, which helps you a lot. So for example, in this patient, this is a legion with slightly hyperdensity. There's peripheral calcification following intravenous IV contras. There is no way of measuring how much this legion is enhancing. Same thing with MRI. You can see on T2-rated images, it is slightly, it is hypointains. So this is cystic legion on pre-contrast even. It is markedly hyperintense indicating hemorrhage or protein-ish material with IV contras. You cannot exactly determine whether this is contrast or the hemorrhage within the legion. But when you do subtraction, you can see this ugly nodule within this cyst, which was a papillary renal cell carcinoma, which would have been difficult to diagnose either on the CT images or non-contrast or non-subtracted images. So subtraction technique is very important. Another legion here on CT scan, without contrast, arterial phase, portal venous phase, or nephrographic corticometallia, nephrographic phase. We don't know whether this legion is enhancing or not. But on MRI, you can see that on T2-rated images, there is some heterogeneity within the legion. On pre-contrast images, there is area of hemorrhage. But with IV contrast, we don't know how much of it is enhancing. On subtraction, you can clearly see that it has got low-level enhancement. This was another papillary renal cell carcinoma. This was a 50-year-old patient, shows an ugly-looking bilabilated legion with hyperintensity on T1 pre-contrast-rated images. Even on T2-rated images, the legion looks extremely ugly. But with IV contrast, you don't know whether this legion is enhancing or not. But on subtraction images, you can clearly see that this is just a hemorrhagic legion. There is no abnormal enhancement within it. We followed this legion for three years and you can see that there is hardly change in this legion. There is still some hemorrhage. It is still looking ugly. But there is no enhancement on subtraction images, which saved nephrectomy for this patient. This was a high-grade papillary renal cell carcinoma. You can see that this legion has got extensive hemorrhage within the legion. This area, we don't know what is happening with IV contrast without subtraction. We don't know whether this area is enhancing or not. And on subtraction images, you can see really ugly-looking papillary renal cell carcinoma, which can be easily diagnosed once we subtract that hemorrhage. So subtraction is very important for diagnosing tumors. How about in an autophase imaging? This is especially diagnostic for angiomyelipomas and some of the clear cell renal cell carcinomas. Let's see how we can diagnose those. Angiomyelipomas come in different flavors. They can be very fatty, like this, where you are going to get this indian artifact at the fat-water interface. And this is diagnostic of angiomyelipoma. You can also see that there is fat here, which is going to suppress on fat-side images. Another legion, intraperinchymal angiomyelipoma, is going to show that indian artifact all around it, as well as chemical shift artifact. Some of the legions could be tiny, which will completely show dropout on autophase imaging. If there are legions containing less fat, then that portion of the legion is going to show drop on autophase imaging. However, the problem comes when there is a lipid-poor AML or clear cell carcinomas in an autophase imaging. However, it has been shown that the lipid-poor angiomyelipoma are stained to follow signal of suos muscle in phase, and they will show areas of drop in signal on autophase, depending on how much fat contained is within the legion. Whereas in renal cell carcinoma, they usually follow the perinchymal intensity on in-phase, and on autophase due to intra-cytoplasmic fat, that is microscopic fat, is going to show uniform drop in signal, which is very characteristic of clear cell renal cell carcinoma. This is a clear cell renal cell carcinoma in-phase, out-of-phase. Although this tumor looks slightly heterogeneous, you can see that on out-of-phase imaging there is uniform drop in signal, and that's because of intra-cytoplasmic fat seen in this legion, which is very typical. In lipid-poor angiomyelipoma, again, we can see that it follows the signal of suos on in-phase, on out-of-phase, there are areas of drop in signal within it due to this small fatty tissue within the legion, but it really depends on how much fat is within the legion. This is lipid-rich angiomyelipoma, which follows the signal of surrounding fat. You can see it is getting saturated on fat-set sequences and out-of-phase imaging. It shows areas of drop in signal, typical for lipid-rich angiomyelipoma. How about diffusion ADC sequences? This is especially helpful in differentiating cystic benign legions from cystic renal cell cancers. Can differentiate benign from malignant renal tumors? Maybe helpful in differentiating histologic subtype or grading the severity of the tumor. DWI is also useful in patients if you cannot give gadolinium, and that might be your only chance to grade the tumor. Here is a low-grade oncocytoma, which shows hardly any restriction, whereas this is a clear cell renal cell carcinoma. You can see how much restriction is there and low ADC. So very helpful. This is a busy chart, but let's tease it out. The RCC subtypes are clear cell, which is by far the most common papillary and chromophobic. It's important to differentiate between these three as clear cell carcinoma tend to be very aggressive, depending on its great and stage and sarcomatode changes. One needs to be very aggressive in treating these tumors with nephrectomy, whereas papillary and chromophobic tend to be relatively slow-growing, especially chromophobic. It has got fire survival of 80 to 93%. So the management could be different here. In papillary and chromophobic, you can do nephron-sparing surgery, which will help in preserving renal function, depending on patient's age. So very important to make sure the type of tumor, which you may be able to do with MR, there are four things that you can look to characterize these different types of renal cell carcinomas. One is T2 signal in and out of phase type of enhancement, and whether there's necrosis or cystic changes within the region or not. So depending on this, we can see that in clear cell type renal cell carcinoma, on T2-wetted images, the region tend to be very high or intermediate in signal intensity compared with low intensity on papillary renal cell carcinoma and intermediate signal intensity on chromophobic type in and out of phase imaging. Since clear cell renal cell carcinoma has got that intracytoplasmic fat, it's going to show drop-in signal, usually uniform on auto-phase imaging, which we saw earlier. Whereas in papillary renal cell carcinoma, there is drop-in signal on in-phase, not auto-phase, but in-phase, that's due to hemocidrin or hemorrhage within the region, whereas chromophobic type is usually homogeneous. As far as enhancement is concerned, clear cell type shows avid enhancement, which is similar to the cortex of the region, whereas papillary shows minimal enhancement and chromophobic shows mild to moderate enhancement. Clear cell carcinomas tend to have necrosis within it. Papillary type may show some small cystic changes and chromophobic tends to be uniformly solid without any necrosis or cystic changes. So on auto-phase, you can see here that this is the clear cell type renal cell carcinoma, which is showing avid enhancement similar to the cortex. There are areas of necrosis within it. This is partially exophytic, more than 3 centimeter diagnosis, whereas papillary is relatively hypovascular. You need subtraction images to see the mild enhancement. It is fairly well-defined and chromophobic. It shows mild to moderate enhancement, well-defined, no necrosis within the region. So how does CT and MR perform in characterizing intermediary renal lesions? Margules and his group performed a trisorpsulpecative study of 143 patients with indeterminate renal lesions detected on ultrasound. 77 had CT, 66 had MRI, and MRI showed definitive diagnosis in 95% of patients versus 78% on CT scan. MRI recommended only additional studies or recommendations such as follow-up only in 1% of patients, whereas CT recommended more than 10% additional evaluation. So MRI was more likely to give definitive diagnosis and less likely to recommend further imaging than CT. So let's go through a couple of scenarios to run through different types of tumors and where MR is going to be helpful in diagnosing this lesion. So scenario one, MR can be helpful to further characterize lesion where initial CT protocol was not adequate. So this patient had CT scan without contrast, and you can see that this patient has got multiple renal lesions, except one probably the rest of the lesions are not characterized on this non-contrast CT scan. You can go ahead and do CT protocol with arterial venous phase. However, still there are going to be some lesions which are not going to be characterized. As you can see on this MRI, that all these lesions look cystic. Some of this lesion, for example, this one is hemorrhagic lesion. There is small hemorrhagic lesion here, which would have any problem with CT scan with and without contrast. But on MRI, we can show that all these lesions are benign, and this patient does not need to follow or any other study. MRI, in this patient at CTA examination, which showed an indeterminate lesion here, we went ahead and did MRI in this patient, and we can see there's this lesion which was not characterized on with contrast is, in fact, an angiomyelipoma. We detected another angiomyelipoma here. These two lesions showed gross fat within it, so there was no additional examination necessary in this patient. So summarizing scenario one, MR can be helpful to further characterize lesions, which are where the initial CT scan protocol was not adequate. So let's see whether MRI can be helpful in indeterminate lesion despite of adequate CT protocol. So here, the protocol is adequate, but still the lesions could be non-diagnostic, because in homogeneous lesion, if there is intermediate change in density, that is, increase in density less than 20-hours per unit, but more than 10-hours per unit, then those lesions are going to be indeterminate, whereas if it is a heterogeneous lesion, then it's going to be difficult to exclude smaller nodular areas within the lesion. So this is a renal mass or renal lesion here. You cannot see it on pre-contrast, but there was 90-hours per unit increase in density from pre-contrast to post-contrast, which is non-diagnostic. So this is indeterminate lesion. We also have to remember that smaller cysts, which are intraparent chymal, in central in location, can give false reading because of partial volumic artifact for those smaller lesions within the renal parent chymal. In this patient, we went ahead and did MRI, and we can see that on T2-rated images, it's a cystic lesion. There's some internal density on pre-contrast T1, but however on post-contrast images, as well as on subtraction, there is no enhancement. So this is a lesion containing mild hemorrhage. So advantage of MRI in such patients is we don't have to worry about pseudo-enhancement. We have subtraction images to subtract out the protein or hemorrhage within the lesion, and we got a very sensitive fluid sequences which can help us diagnosing this cystic indeterminate lesions. There was this study done for potential renal donors, which is very important because we don't want to transplant any kidneys with renal lesions. There were 105 lesions, which were with average size of 1.5 centimeter. 94, all this patient had CT scan with subsequent MRI. On CT scan, 94 of these 105 lesions were TSTC, that is too small to characterize by CT. When they did MRI, in 99% of the lesions, MRI was diagnostic and no further imaging was necessary, out of which 83 were simple cysts, it were protein issues or hemorrhage exist, two were AML. So take home point is MRI can characterize small lesions that are indeterminate with CT due to pseudo-enhancement or partial volume artifact. This was a indeterminate lesion, as this lesion was kind of bi-lobed or trial-lobed, and we can see following administration of IV contras, you can see some separations within it, is there some nodularity within it? So this was at least two F lesions by Bosnay classification. On MRI, we can see this portion of the cyst has got dense hemorrhage. The other portion of the cyst did not have any hemorrhage and is an enhancing nodule, which was diagnostic of early cystic renal cell carcinoma in this patient. So MRI is sensitive for detection of enhancing nodules in cystic renal lesions. It upgraded 73% of patients on this study. It had 95% positive predictive value for malignancy with enhancing mural nodules. How about legions with low enhancement that is difficult to detect on CT scan? The solid mass, which is most likely to result in this conundrum is papillary renal cell carcinoma. As we can see it here, this is the enhancement pattern of clear cell renal cell carcinoma. Chromophobe type and papillary type. And papillary type, you can see it shows very low enhancement with IV contrast compared with other two tumors. So if you suspect that there is very low enhancement or rule of thumb is if you have difficulty in deciding whether it is enhancing or not, they consider it to be a papillary renal cell carcinoma because in up to 30 years, patients with CT, this enhancement is undetectable. This legions tend to be hyperdense on pre-contrast CT. So it's difficult to measure the enhancement in many of these patients. This is a patient with, it looks like hemorrhage exist with IV contrast and on pre-contrast images, the density was around 28 harmful units following IV contrast. On cortico medullary and nephrographic phase, there was enhancement of 16 to 17 harmful units. So this was, although the legion size is pretty large, it was still indeterminate. Whereas on MRI, you can see that this legion has intensity similar to perenchyma with IV contrast. You can clearly see that there's some enhancement to be done with IV contrast. You can clearly see that there's some enhancement which on subtraction images, we can clearly see that this is a solid enhancing legion which was papillary renal cell carcinoma. So over half of the papillary renal cell carcinoma did not show definitive enhancement on cortico medullary phase with CT. On nephrographic phase, it did show some enhancement. So nephrographic phase is probably the best phase, but MR did not have any problems detecting this legions at all. So take home point is when reading CT scan for hyperdense legion, nephrographic phase is key. Although you are still going to call some indeterminate legions with this, MR has better sensitivity and specificity for detecting enhancement of low level enhancing legions versus hemorrhagic cyst in these patients. Scenario number three, how about CT protocol is adequate, legion is enhancing, but the diagnosis is uncertain. So in this, the legion is not fully evaluated with CT scan. For example, this patient which was a fat poor AML, but based on multiple sequences and enhancement pattern and diffusion weighted images as well as in an out of phase images, MR can characterize many of these legions. For example, this is a clear cell carcinoma. This is probably typical. It shows hyperintensity on D2 weighted images. It is usually hyper intense on D1 weighted images. It shows enhancement similar to renal cortex with areas of necrosis within it. There shouldn't be any problem in diagnosing this on CT or MR. We can see that on nephrographic as well as carticum medullary phase, nephrographic as well as carticum medullary phase, it's going to show marked enhancement, especially in carticum medullary phase. It is going to show a drop in signal as I showed you before on out of phase imaging which is going to be uniform due to intraceitoplasmic fat which is also pretty helpful. Whereas in papillary renal cell carcinomas which is second most common, they tend to be very cellular. They are hyper intense on D2 weighted images. These are relatively indolent legions so it's important to separate these out from renal cell carcinomas. These tend to be bilateral and are more common in renal failure patients. Here in this patient, there are multiple legions. You can see one legion on right side which shows typical appearance of papillary renal cell carcinoma which is low enhancement and it is low signal intensity on D2 weighted images. Whereas this legion, we could see another legion in the other kidney which had similar features. Whereas third legion which was there was a simple renal cyst. So with help of MRI, we can easily diagnose papillary renal cell carcinoma which may be indeterminate on CT scan and you don't have to have aggressive treatment in these patients, depending on patient's renal function, age and everything, you can do some other type of treatment. So papillary renal cell carcinoma commonly involves in stretch care kidneys often contain hemorrhage or cystic degeneration usually homogenously hypoenhancing. They can show some restriction on D2 weighted images due to their cellularity which doesn't help. But unlike clear cell RCC, papillary renal cell carcinoma is frequently hypo intense on D2 weighted images whereas the clear cell type, there are hyper intense on D2 weighted images. This is one of the ugly looking papillary renal cell carcinoma. We can see that it is hypo intense on D2 weighted images compared with clear cell which are going to be hyper intense on D1 weighted pre-contrast images. We can see some hemorrhage. We don't know what is happening here, but on subtraction images with IV contrast, you can see multiple lobulations, fronts along the surface of this region indicating aggressive type papillary renal cell carcinoma. How about chromophobes? This is probably the least common and more indolent type of low grade tumor. It shares morphologic features and histologic features of oncocytoma. They tend to be iso to hypo intense on D1 weighted images. They show mild enhancement with IV contrast, no necrosis, their enhancement pattern is less than clear cell carcinoma. So there's something in between. So rule of thumb in this legency, there are usually slightly hypo intense, hypo enhancing compared with the cortex. There can be diagnostic challenge in diagnosing oncocytomas. However, with MRI, there are certain features which can be diagnostic. This tumors tend to have central scar. They tend to be slightly hyper intense which is kind of unique for oncocytoma. They are slightly hyper intense on T2 weighted images and they show kind of slightly spoke field type of enhancement with IV contrast in corticomedulary phase. And on the nephrographic phase, they tend to be slightly hypo intense. The another unique feature of these tumors is the scar doesn't enhance initially, but on delayed imaging, this scar will show enhancement which will help you differentiate between the renal cell carcinoma and oncocytoma. The central scar, hyper intensity on T2 of central scar delayed and complete enhancement. Now intravoxal fat that is on the outer phase it's not going to show low intensity was 97% specific and 55% sensitive in patients with oncocytoma versus renal cell carcinoma based on these two studies. So very powerful feature to differentiate this from renal cell, clear cell type renal cell carcinoma. Another feature that has been described is segmental enhancement inversion in oncocytomas. So on corticomedulary phase, some portion of the tumor may enhance which gets washed out and then the other portion of the tumor will enhance on nephrographic phase, which is kind of unique for oncocytomas. Then we often have diagnostic challenge with fat poor and tumor lipo mass. Fortunately, these tumors tend to be the percentage of these tumors is less. Whenever there is less than 25% of fat per high power field on path, they call it fat poor and tumor lipo mass. The key MR features in this patients is they are dark on T2, unlike renal cell carcinomas which are going to be high on T2. They do show restricted diffusion, which is a problem, but they do not show any necrosis. And they might show some areas of dropping signal on auto phase depending on how much fat is within the region. Sometimes you may not be able to distinguish this from malignancy, and there is no shame in doing biopsy in these patients. Because actually this study was just published last month, which showed that if you see, actually lipid poor and tumor lipo mass on T2D images show this uniform low signal, and they show moderate enhancement on corticomedulary phase. And this is actually, this is diagnosed, actually this was kind of very useful feature in diagnosing lipid poor and tumor lipo mass. This was seen in 83% of patients with, sorry, this was, the sensitivity was 83% and specificity was 90% in this patient. So maybe this feature is going to be very useful in diagnosing lipid poor and tumor lipo mass. Sometimes on CT scan, although with naked eye, you might say that there is fat within this region, but when you actually measure those because of increased density from vessels, you might not get typical fat measurements. In those patients, MR is also useful because here you can see that this lesion is clearly has similar intensity as surrounding fat. It has got the India Inc. artifact that fat fluid interface and shows some enhancement with IV contrast. So MRI confirms AML. Sometimes you might be able to see this pushing of fat into the renal perinchima, which is typical for angiomyeliboma from where it arises. So pretty diagnostic if you see it. So how about MR versus CT characterizing renal masses? This group did a study of 120 patients in legions which were indeterminate on CT followed by MRI. The final diagnosis was based on pathology or followup. 39% of these legions were benign masses such as angiomyeliboma, oncocytoma or hemorrhage cyst. The sensitivity of CT for diagnosis of renal cell carcinomas was 91% versus MR, it was 94%. Specificity for CT was only 28%. Remember these are indeterminate legions, whereas with MR it was 68%. So using MRI in approximately 12.5% of patients with malignant diagnosis on CT could have avoided surgery, whereas 2.5% of benign diagnosis on CT would have been appropriately recommended surgery, which was not done with CT scan. So another study which was published in abdominal imaging which had 63 patients with indeterminate legions which were smaller than two centimeter. And of the 63 patients MR was sensitive 100% of the time that means it diagnosed all those 63 patients. MR was 95% specific indicating benign versus malignant legions of this 63, only three remained indeterminate following MRI. So MR, take home message is MR has greater specificity than CT in distinguishing benign versus malignant diagnosis and fewer legions remain indeterminate after MRI. So wrapping up, MRI is indicated in evaluation of indeterminate renal masses on CT and ultrasound and provide additional diagnostic information which can be helpful in diagnosing and managing these patients. MR imaging is particularly helpful in distinguishing solid from cystic legions when enhancement of renal masses is questionable, especially for those where net enhancement on CT is between 10 and 20 harmful units. So we can either upgrade or downgrade BASNEK 2F legions so that these patients do not have to go for follow-up examination every six months to one year for two and a half to three years. Diffusion weighted imaging and dynamic contrast in NAC imaging can provide specific information regarding tumor histology and staging. MR is especially useful for two small to characterized legions. There are not going to be any legions which are too small to characterize on MRI, legions with low level of any aspect, hemorrhagic legions and complex cyst. And that's all from me. Thank you. Thank you once again, sir, for taking this important topic. In such a-