Welcome everybody. This is the refresher course on shoulder imaging in the athlete. I'm your moderator. I'm Redo Sutter from Zurich, Switzerland. We have five speakers that are going to cover all aspects on shoulder imaging in the athlete. And it's my pleasure to introduce the first speaker, Diana Afonso from Lisbon, Portugal. She's going to talk about shoulder instability in the athlete. Good afternoon. Thank you to Professor Redo Sutter for the invitation to speak here today. I'll talk about shoulder instability in the athlete. So I'll start with the definition. What is glenohumeral instability? This is defined as the inability to center the humeral head relative to the glenoid during the range of motion leading to pain and decreased function of the shoulder. As we all know, the shoulder is the most unstable large joint of the body. The overwhelming majority of glenohumeral dislocations, about 95%, are anterial dislocations. And after the initial dislocation, up to half of patients will experience a recurrence, which is more common in younger athletic population. So how does this happen? Our position during anterial dislocation is usually an end range abduction and external rotation, quite a zapper position as shown here, in sports that require significant abduction and external rotation of the shoulder, often involved over harm or throwing motions such as in racquet sports, baseball, swimming, volleyball, and basketball. So anterial dislocations involve injury to the bone and soft tissue stabilizers. Bone damage usually includes anteroglenoid fractures or humeral head fractures. Additionally, different types of soft tissue abnormalities can occur as listed here. So moving on to imaging, that's why we are all here today. Which lesions are we familiar with and routinely report on MRI in cases of anterial instability? Of course, the classic soft tissue Bankert and its Bankert variants that I'll briefly mention. Dr. Choi will talk more about the label in her talk. So we all know the classic Bankert when we have a detachment of the anterior labrum along with disruption of the scapular priostin. We are all familiar with this type of lesion. But then we also have the Bankert variants. The Bankert variants are like the classic Bankert where the scapular priostin is torn. In Bankert variants, the priostin may look abnormal, but it's intact. One type of Bankert variant is Purt's lesion, as shown here, where the anterior infralabrum is detached but not displaced and the priostin is intact, even though it might appear irregular or stripped. Another Bankert variant is the ALBSA, that stands for anterior labral priostin sleeve avulsion, that is defined by the detachment of the anterior infralabrum that extends medially with it intact, but lifted and irregular priostin. But again, it's intact. And then we have GLAD, meaning glenolabral articular cartilage disruption, that refers to a non-displaced tier of the anterior infralabrum with associated chondral damage. And since we are dealing with cartilage loss, loose bodies can also be found inside the joint. In addition to labral injuries, what else do we need to look for and report more often on MRI? Bone damage, whether it's unipolar or bipolar bone loss. So, many patients with anterior dislocation sustain a posterolateral humer head impaction fracture, known as Hill-Sachs lesion, or an anteroglenoid rim impaction fracture, also known as bone banker. Its presence and size depend on the force vector at the time of impact, and these fractures can be fragmented and well seen on imaging or the opposite. They can be compressive and more subtle on imaging. A quick reminder, the threshold for critical glenoid bone loss is now considered to be greater than 15 to 20 percent, not 25 percent as used to be considered in the past, according to a recent consensus from International Panel of Surgeons. So, when humeral and glenoid fractures occur simultaneously, we have bipolar bone loss. Bipolar bone loss should be located and quantified using 3D or 2D CT or MRI, including the recent developed zero echo time technique, which generates CT-like images, which may be very helpful in the setting of patients. And why is this important to assess? Because the Hill-Sachs can engage the glenoid and thereby increasing risk of recurrent dislocation. Hill-Sachs interval and the glenoid track are two key parameters in assessing the glenoid tracking status of the Hill-Sachs. So, we need to measure the Hill-Sachs interval. And how do we measure this? We can measure the Hill-Sachs width by drawing this line here across the Hill-Sachs from medial to lateral, and then the bone bridge, which is a line drawn from the lateral most margin of the Hill-Sachs to the cuff attachment. And the bone bridge will help to account for how medial the position of the Hill-Sachs is. And then the sum of both, meaning the sum of the bone bridge and the Hill-Sachs width, will constitute the Hill-Sachs interval. And then we have the glenoid track. What is the glenoid track? Glenoid track refers as the contact zone of the glenoid on the humeral head during the elbow position that is defined as 83% of the glenoid diameter and serves as a tool to predict humeral head engagement of the glenoid rim. This sounds confusing, but it's not. As shown here in this drawing, this red area represents the glenoid track. So, on average, during the mid-elbow position like this, only 83% of the glenoid diameter maintains contact with the humeral head, while the remaining 70% does not. And when there's a glenoid rim defect, as shown here, this will be subtracted from the glenoid track. So we need to measure the glenoid track. How to do it? We place a best-fit circle over the lower two-thirds of the glenoid on the sagittal plane, and then we measure the glenoid diameter and the width of the glenoid defects. And then we apply this formula shown here. When the Hill-Sachs interval is smaller than the glenoid track, okay, we measure the Hill-Sachs off-zone track, so most likely clinical stable. Otherwise, if the Hill-Sachs interval is greater than the glenoid track, or substantially medial, the Hill-Sachs is off-tracked and most likely clinical unstable. So this is an example. We measure the Hill-Sachs interval on the sagittal plane and the glenoid track on the sagittal plane and apply this formula. Please keep in mind that we need to clearly define all these parameters listed here on our reports. And after that, if the Hill-Sachs interval is greater than the glenoid track, as is in this patient, the Hill-Sachs is off-track, which means there's a risk of engagement with the glenoid and recurrent dislocation. And why is this important? Because off-track to on-track conversion is one of the primary goals of surgery for stabilizing antarestability. What else should we be careful not to miss on MRI? Haggle. Haggle is defined as a humeral avulsion of the anterior band of the infralonganormal ligament, and about one-third are missed on MRI, mostly in the non-acute setting due to lack of fluid inside the joint, but also due to scarring. Up to 20% of haggle lesion will have a bone fragment, and do remember that haggle lesions are less likely to be associated with the Hill-Sachs and anterolabral pathology, which can be misleading for the radiologists. Finally, nerve injuries are often overlooked. Axillary nerve is by far the most frequently affected nerve, often presenting with a low-grade reversible injury that typically recovers with no intervention, while more global brachial plexus injuries tend to occur with higher energy mechanisms. So some take-home points. In the setting of anterior shoulder dislocation, we have injuries that we are familiar with and routinely report on imaging such as soft tissue classic banker and banker variants. Injuries we must look for and report more often on MRI, humeral and galenoid fractures, and measure bipolar bone loss. We need to add all these parameters shown here in the report. Don't forget this is important. And remember, when the Hill-Sachs interval is greater than glenotrach or two medial, the Hill-Sachs is off track, so it will be clinically probably unstable. And finally, injuries we must not miss, humeral sides, Hager lesions, and nerve injuries. Thank you. From Lisbon. Many thanks. So the next speaker is Chang-Ah Choi from Korea. She will be speaking about assessing the labrum and cartilage. Thank you for the introduction. Good afternoon. It's my great pleasure and honor to be here. I'd like to thank the organizers for inviting me. So since Dr. Diana Alfonso covered mostly the anterior part of the labrum, now I'll focus mainly on the superior labral lesions. So in my talk, I will briefly describe the biomechanics in the throwing shoulder as it relates to superior labral lesions, discuss some of the imaging techniques, normal variants, and then mostly focus on slab tears, and at the end briefly mention some cartilage lesions. So reminding ourselves of the biomechanics in the throwing shoulder, the shoulder goes through these six phases of motion, and it's in this late cocking phase where you can see the shoulder most externally rotated, putting a strain on the labral bicipital complex with a torsional force and literally peeling off the glenoid labrum off the glenoid. And so this is called the peelback effect, causing labral damage. And also with further torsion, you can see there's this kinking effect between the glenoid and the humerus head imposed on the supraspinatus tendon, mostly posterior portion and the infraspinatus anterior portion, causing rotator cuff tendon damage as well, and this is called the internal impingement. So please remember the mechanisms of peelback and internal impingement as regards the superior labrum. And when these are chronically repeated, they also cause osseous and capsular changes. Regarding imaging techniques, MRI is classically the modality of choice for evaluation of shoulder pathologies. With intra-articular contrast, direct MRI arthrography is more sensitive than MRI for detection of articular cartilage and labral lesions as reported by many authors, more accurate for slab tears, and 3-Tesla obviously has been claimed to be better than 1.5T, even without the presence of intra-articular contrast. CT arthrography is another diagnostic modality which is excellent for slab tears and cartilage lesions and actually one of my favorite modalities because of the shorter imaging times, lower cost in many countries, and also excellent depiction of bony details. So looking at the labrum, we usually describe it in a clockwise fashion, so I will mostly talk about the 10 to 2 o'clock positions, the superior labrum. So looking at the normal anatomy, you can see that with intra-articular contrast on MRI arthrography, this low signal labrum is actually not well delineated from the biceps tendon anchor. Using 3-Tesla MRI, you can see that the low signal labrum is actually visualized better than the MRI arthrography T1 fat-suppressed images on these fluid-sensitive images with and without fat suppression. And with intra-articular iodinated contrast, you can see that the low-density labrum is very well delineated on all three sequences. Looking at normal variants, first the superior sublabral recess or sulcus is actually quite common. It's found in up to 75 percent of patients without tear in a series. It's the partially unattached labrum at the apex of the superior labrum. Usually does not extend past posterior to the LHBT insertion and less than 2 millimeters in width, but oftentimes it's not very clear cut, so a lot of times the recess is difficult to differentiate from slab type 2. Another famous variant called the sublabral foramen found in about 11 percent of individuals. It's at a typical 2 o'clock position anterior to the biceps attachment. Has to be differentiated from a sublabral recess or a labral tear, but its typical position at 2 o'clock and very smooth margin help to differentiate this normal variant from a labral tear. Next, a Buford or Buford complex found in about 1 to 7 percent of individuals. It's characterized by the absence of anterior superior labrum and a cord-like thick MGHL, which can be confused with a displaced labral fragment. But if you follow the serial axial images or the sagittal image, you can see that it's the very thick MGHL that looks like a labrum and just one cut of axial image. And there's an anterior superior labrum missing. And some have claimed it to predispose to slab. There's another sulcus-like normal variant at 7 to 8 o'clock position of the labrum, which we've called the posterior inferior labral cleft. That can be confused with a Kim's lesion, which I will mention later. So now I'll focus on the slab lesions, which is short for superior labrum anterior to posterior tears. It involves mainly the origin of the long head of biceps tendon, where the tendon inserts onto the superior labrum. And because the physical exam is nonspecific, often poorly correlating with arthroscopic findings, imaging diagnosis is very important for pre-op planning and post-op rehab. The lesions are best identified on the coronal oblique and axial images. The following diagnostic criteria may be used to diagnose slab lesions. So their presence of laterally curved high signal or contrast in the labrum, or multiple or branching lines of high signal or contrast, or a full thickness detachment with irregular high signal margin, and a separation of more than 2 millimeters or 3 millimeters, depending on the imaging modality between the labrum and the glenoid on the coronal image. And if you have a paralabral cyst, you may take a hint that there may be labral tear. So there's all this numbering with the slab classifications, out of which type 2 is the most common that you will encounter clinically, which consists of a partial thickness longitudinal tear of the superior labrum with unstable or unstable biceps anchor. This has been subclassified into 2A, B, and C, out of which 2B is the most common, mainly posterior superior labral tear found in the overhead throwing athlete, oftentimes with a spinocleanoid notch cyst. And I will show you and walk you through the remaining number classification of slab. So this is a typical type 2 slab extending from anterior to posterior portion of the superior labrum with a quite noticeable wide gap and adjacent cartilage damage, which was arthroscopically proven. And even though it's difficult to differentiate from a sublarbar recess sometimes, the surgeons say if you encounter cartilage damage adjacent to the labral lesion, they are assured that it's a slab rather than a sublarbar recess. So we may want to find that hint on the imaging as well. You may have a parlabral cyst extending to the prescapular or spinocleanoid notch. In that case, you should start looking for a type 2 slab lesion, which may be concealed, especially if you don't have an intraarticular contrast. You may not really obviously notice it. If there is an obvious connection between the parlabral cyst extending to the prescapular or spinocleanoid notch, you may be more confident that there is a parlabral cyst extending or originating from a type 2 slab. And this may cause prescapular nerve entrapment syndrome, as in this patient with denervation changes in the IST muscle. Parlabral impingement that I talked about before may cause an articular-sided supraspinatus tendon tear, and this may be associated with a type 2 slab tear. And if it's from a chronic mechanism, you may even see osseous changes. So now I'm just breezing through the rest of the slab lesions. You may see a bucket handle tear, which is called a type 3 slab tear, as you can see here, like in the meniscus of the knee. If the tear extends into the substance of the biceps tendon in anchor portion, it's called a type 4 slab, as you can see here. If the superior labral tear extends all along the anterior labrum, contiguous with a Bankart lesion, this may be called a type 5 slab lesion. And you can see the contrast leaking along the defect of the labrum very well on the coronal images as well. You may see a flap tear with a free edge, such as you see in the knee of the meniscus, and this is a type 6 slab. And if you have a slab type 2 tear with extension of the contrast into the substance of the MGHL, this is called a type 7 tear. You may have the slab tear extending along the posterior labrum, and this would be called a type 8 slab. You can also have the labral tear extending along the anterior as well as the posterior labrum extending all the way along the entire circumference of the labrum, then it would be called a type 9 slab. But the good news is you don't have to remember all the numberings, just it's more important to describe the extent and accompanying lesions as the surgeon would depend on the report that you make. And he may choose to debride as for types 1, 3, and 6 lesions, or do a repair surgery for the remainder of the slab lesions, especially if the patient is an active patient. And remember, we're not going to talk about post-op imaging for slab lesions, but the surgeons usually leave the labrum at 12 to 1 o'clock position unrepaired, because if you repair that, they may cause an external limitation of external rotation. And also, the surgeon usually treats the accompanying lesions, especially if he considered that it's symptomatic. And for older patients with less activity, the labrum may be just left alone, and a biceps tenodesis or a tenotomy may be done. Just a brief word on cartilage imaging. GLAD was mentioned in the talk by Dr. Afonso. It's described as a superficial anterior inferior labral tear associated with anterior inferior cartilage damage. And in a recent review of literature, many authors have associated it with shoulder instability and trauma in ABER, in contrast to the original description in stable shoulders following trauma in adduction. And this is what a GLAD lesion would look like with an anterior inferior labral tear and a cartilage fissuring and fragment attached to the labral tear. Lastly, a Kim's lesion is another kind of a marginal chondral lesion that can be combined with concealed posterior inferior labral tear and retroversion of the glenoid that can be found in a throwing athlete. As you can see here, there's cartilage damage at the 7 to 8 o'clock position. A T2 mapping, lastly, can also be used for glenoid cartilage, but it's quite tricky because the glenoid cartilage is very thin, but it's feasible. And this may provide a new insight into cartilage lesions of the glenoid. So in summary, please remember the biomechanism of the superior labral injury, and you can use MR arthrography, CT arthrography, or a 3T MRI as it's more convenient in your institution and for the patient because all are comparable. And remember these different points that help you to differentiate a slab from a normal variant, and you can use the newer cartilage imaging techniques to study the cartilage in labral tear lesions. Thank you for your attention. Many thanks. So it's my pleasure to introduce my associate, Dr. Ben Fritz from Balgus University Hospital in Zurich. He's going to talk about rotator cuff pathology and the muscles. Thank you very much, Reto, for the introduction. And thank you for the invitation. So this talk is about rotator cuff pathology and muscles. So rotator cuff injuries are actually pretty frequent in elderly patients or elderly individuals, and often asymptomatic, and they're usually due to aging and degeneration. In younger individuals or patients, we find them often associated with trauma, or in overhead athletes working with throwing motions and so on, as a chronic overuse syndrome like the baseball pitcher we saw before. So when we look at the rotator cuff injuries, primarily we look at the tendon and tendon pathologies, but they have a direct influence on the muscle and the muscle pathology. So the tendons can be injured and appear like tendinopathy or partial or full thickness tear, and often the fatty degeneration or atrioterior muscles are subsequently a result of the tendon pathology. On every MRI, we usually look at both abnormalities that they have influenced the therapeutic options dramatically, in particular if this patient can undergo surgery or not. So that's the right shoulder of a normal supraspinatus of a patient with a normal supraspinatus tendon. You see there is this striated pattern of this patient. It's a fairly hyper-intense structure. We have a little bit of magic angle effect. We have a coronal PD fat suppressed sequence like here in this case, but it's a total normal tendon. Tendinopathy, in a tendinopathy tendon, you have a loss of the striated pattern with usually some kind of edema or hyper-intense signal in PD and T2-weighted sequences, and you have a thickening of the tendon. So that's a sign of tendinopathy, in particular that loss of striation. Here's another example. This is infraspinatus tendon. Again, a loss of these striations, thickening, and a T2 hyper-intense signal. That's a sure sign of advanced tendinopathy in this patient. When the injury progresses, it can become a partial thickness tear. So a tear is a partial thickness tear when it only is on the articular surface, on the bursa surface, or intrasubstance, but never crosses the entire tendon thickness. In this case here on the left, we see that articular surface defect, some torn fibers at the articular surface, and some interposed fluid in the defect. The middle image, that's a bursa-sided defect here, right at the footprint. And then we have the intrasubstance defects, which are also, in most cases, kind of delamination type of pattern, but the bursa on the articular surface stays intact. Here's a grading of partial tears, which was proposed by Elman. It says basically that when the tear involves about 25%, which is 3 millimeter, it's a grade 1 tear. Up to from 25 to 50%, it's a grade 2 tear. And when it is more than 50% of the thickness of the tendon, it's a grade 3 tear. So keep in mind the tendon is about 12 millimeter thickness, so at least a supraspinatus tendon. And this is also the numbers 3, 3 to 6 millimeter, more than 6 millimeter. This is an easy tool to grade the severity of the partial tear in clinical practice. That's an example of a grade 1 Elman tear. You see this shallow defect on the articular surface in this patient. And this here is a grade 3 Elman defect. It's pretty deep, about, I would say, 80% of the tendon thickness. So that's probably, if you do not want to use this grade 1, grade 2, grade 3, just describe the amount of percentage of how this tear goes into the tendon, like here 80%, you're probably fine in reporting this partial tear in a good manner in your clinical practice. We have primarily two locations on the supraspinatus tendon where we find the defects. One are the footprint lesions, and one are the critical zone lesions. So these footprint lesions is a detachment of tendon fibers right at the tendon bone junction. And these are like avulsion-type patterns. You see a little retraction of the partial tear and the interposed fluid. So that's a common pattern. And the other is the critical zone. This is a region 1 to 2 centimeter medial to the footprint of the tendon. And this is supposed to be the area of the lowest blood perfusion. And they have probably some hypoxic component. In this case, you see a bursa-sided defect and a small articular-sided defect, but not a full-thickness tear. You may have heard of the expression the Rembrandt tears. A Rembrandt tear, in principle, is a particle-thickness tear at the tendon footprint, so like the tear we saw before with the bursa-sided tear. So all these three patterns are basically rim-brain tears. The most famous subgroup is probably the PASTA tear. PASTA is an acronym for partial articular surface tendon avulsion. So you see basically a partial tear on the articular surface with some degree of tendon retraction. That's the PASTA tear. Or here, for instance, this intersubstance tear is a concealed interstitial delamination. So there are all these acronyms all over the place. If you want to use them, you're free to do. I'm not a big fan of it, but you may have now heard about that when you see them in literature. So that's a pain tear. That's a kind of interesting tear. It has an articular side and an intrasubstance continuation. This is a MR arthrography image with a T1 fat suppressed weighted sequence. And you see there's contrast going into this large delaminating intrasubstance defect. So we know for sure there has to be an articular surface partial tear, too, which is not visualized in this image. But this is what MRI can help you for diagnosing these tears more accurately. When we have a tear that crosses from the articular all the way to the bursa side, we call it a full thickness tear. Again, here we have tears at the footprint, or in this case on the critical zone, our main two locations are the same than impartial tears. A full thickness tear can be a really small fissure on the anteroposterior distance, but they can be quite extensive. Here we have a rather small tear around 1 centimeter in the anteroposterior width, which could be graded as a small to medium tear. Here we have an extensive tear with a full width involvement of the supraspinatus tendon, the infraspinatus tendon, and also goes all the way into the teres minor tendon. This is what we usually classify as a massive rotator cuff tear. It's when it involves more than 5 centimeter in the AP distance or involves two or more tendons. Full thickness tears, in particular when they detach from the footprint, can retract. And there is an important classification from Pat, that's a French physician, and it describes the amount of retraction of the supraspinatus tendon in these full thickness tears. A stage 1 tear is basically almost no retraction when the tendon stumps stays right at the footprint. When it goes a little further back to the 10th of the humeral head, it's called a stage 2 tear or retraction in the Pat classification. And when it goes all the way back to the glenoid or the superior part of the glenoid, it's a stage 3 retraction. This is important for the physician since the further this tendon stump retracts, the harder it is to mobilize and bring it back to the tendon footprint and do the reconstruction. And it may have too much tension then on the reconstruction. So that's important to describe. So we switch over to the rotator cuff muscles. And we look primarily for fatty degeneration and atrophy in the patients who have rotator cuff injuries. And fatty degeneration atrophy is, to a certain degree, a normal mechanism of aging. But it is accelerated in patients with rotator cuff injuries. So the Gutai-Yi classification was published now 30 years ago. And that's the most famous and clinically used classification for describing the fatty degeneration or infiltration of the muscles. So it tells you basically what is the amount of fat inside the muscle in comparison to your preserved muscle fibers. And the grade 0 Gutai-Yi classification has basically no fat. This is solid muscle structures. If you have a little fatty streaks inside, you call it a grade 1 in the Gutai-Yi classification. And then it gets more interesting. So a grade 3, a grade 2 fatty infiltration has already quite a lot fat, but still more muscle than fat. A grade 3 has about the equal amount of fat and muscle, like here in this example, a supraspinous muscle. And grade 4 has then really much more fat than muscle. And clinicians use this to determine which patient can undergo surgery or not. So usually, or for a long time, it was considered that a grade 0 to grade 2 is still a patient which is reconstructable or repairable, since the muscle function is probably good enough to have a good outcome after surgery. On the grade 3 to grade 4, it can be considered to be irreparable, since the muscle already has too extensive loss of the function due to the fatty infiltration. But this has changed a little. And I believe Reto may talk about this in the next talk quite shortly. So with a fatty degeneration, we usually find some degree of muscle atrophy. On the left-hand side, you see a pretty normal muscle. You see that bulky experience in this convex structure, concave structure here on the upper side. And once you get atrophy, you see that the substance of the muscle goes down. It sinks more to the bottom of the fossa of the supraspinatus. So there is the so-called tangent sign. That's a tangent from the spine of the scapula to the upper border of the scapula. And if you have the muscle going over this line, you say that's negative and it's still good muscle volume. When you put this line or this tangent and your muscle belly is below this line, that's not a good sign. This is a sign for advanced muscle atrophy. But the tangent sign tells you about advanced muscle atrophy. How about early muscle atrophy? And that's a very recent publication from our hospital. And Reto was the senior author of this. And my colleagues found that in early muscle atrophy, actually, you first lose volume at the supraspinatus muscle, at the posterior superior aspect of the muscle, when you get this indentation with this wavy appearance. And here on the left-hand side, that's a normal muscle. But you see here on this aspect, you lose muscle volume. And this is kind of interesting. They call it the blackbird sign. And we first thought, OK, what is this, a blackbird sign? And then Reto explained it to me. There is the blackbird with the eye. OK, so whenever you see this, think about the blackbird. And this is early muscle atrophy. And a good sign to see that this muscle is undergoing volume loss. So in conclusion, when we evaluate the tendons, we think about tendinopathy, partial thickness tear, full thickness tear. And we can evaluate with great precision on MRI the tear size or tear location and the tendon stump retraction. For the muscles, we primarily look for fatty degeneration and muscle atrophy to determine if this patient can still undergo surgery or not and if this is a good patient for reconstruction. Thank you very much for your attention. Thank you. So now we've heard about the rotator cuff tendons and muscles in the pre-op situation. In the next 10 minutes, I will speak about imaging of the post-operative rotator cuff tendon and muscles. Let's start with these two radiographs of the left shoulder before and after surgery. There are two changes that can be seen on the radiograph. One is obviously the suture anchor. This has been inserted at the humeral head to fix the supraspinatus tendon. And the other change concerns the bone. The surgeon performed a subacromial decompression with acromioplasty. This means that a part of the inferior surface of the acromion has been removed to enlarge the subacromial space. Also on MRI, the subacromial decompression is readily visible, especially if you compare the findings to the pre-op MRI. Let's look at this next case. But now let's focus on the anchors in the humeral head. This patient had a repair of the supraspinatus tendon with two anchors at the greater tubercle. And also a repair of the subscapularis tendon with a single anchor at the lesser tubercle. Both tendons were repaired with a single row technique. So the anchors for each tendon are in one line. Here we see a comparison with the other technique that is commonly used. In the double row technique, we have a medial row of two anchors shown here in red, and a lateral row shown in blue. The double row technique is more robust against mechanical forces, but only works with tendons where there's little scarring during the surgery that is seen. And if you have a PET3 lesion, then this is not possible with a double row technique. Here are other techniques for reattaching the tendons in the bone. In this patient, the suture plate has been positioned at the lateral edge of the greater tubercle. And small suture canals can be seen extending to the footprint of the tendon. On the axial MR image here, we can see the suture canals very nicely, especially going here to the top. When you reattach tendons, things can go wrong. So here we have three different complications. The first is a subcapillary suture anchor that was loose and has migrated to the axillary recess. The second is a suture plate that came loose, and the third a bio-absorbable screw. So all these materials need to be removed as soon as possible with a surgical procedure to prevent severe cartilage damage and osteoarthritis. Another complication that we see, especially by composite anchors, is this. So this is a double row reattachment of the supraspinatus tendon. And there is a severe foreign body reaction in the upper row with a large osteolysis that can be even seen on the radiograph, and subsequent the loosening of the anchors. Now for some procedures that are used less commonly in patients with irreparable massive rotator cuff tears, a tendon transfer can be performed, such as shown here with the latissimus dorsi transfer. This restores part of the function of the supraspinatus tendon that is torn. On the axial MR image, we see this thick latissimus dorsi tendon that has been attached at the lateral and superior edge of the greater tubercle. And on the coronal image, you see these green arrows. So there's a hole in the subacromial space. That is where the supraspinatus tendon used to be, because this is no longer there. The surgeon did this tendon transfer. Another technique for massive rotator cuff tears is the patch graft, where an insufficient tendon is augmented or the tendon defect is bridged by different types of grafts. So what should a normal tendon look like after shoulder surgery? We expect that there is a broad tendon insertion, nicely covering the whole footprint at the bone. And you see some metal artifacts here. So in the post-op situation, it's useful to apply techniques for metal artifact reduction, such as high bandwidth or Dixon sequences. There's actually a temporal evolution of how the tendon looks after surgery. Initially, you will have some increased signal within the tendon, while at a later time point, the tendon usually looks more homogeneously dark. Sometimes you have a small residual defect after the surgery, such as seen here in the yellow circle. But at a later time point, the tendon has healed and looks now more homogeneously dark. What about alternative methods for evaluating the post-op rotator cuff? Ultrasound is a very good method with a high spatial resolution. Here you see a broad tendon attachment, and you can even see the anchor in the bone. If you use Doppler ultrasound, note that the vascularity within the tendon is increased initially and then decreases at a later stage. Which method is better for evaluating the post-op tendon? In this meta-analysis, ultrasound had a slightly better performance for identifying re-tears than MR, but the differences were not statistically significant. So both MRI or ultrasound can be considered good first-line imaging options. So now let's look at the re-tears. You can see in this pre-op image a large, full thickness tear of the supraspinatus tendon. And after the surgery, a re-tear at the exact same location looks very similar. What about this finding in the subacromial bursa? This is a T1FATZ MR orthogram image, and the fluid you see in the bursa, this is not bursitis. So in the post-op situation, the rotator cuff is not watertight. So you will see fluid in the bursa. Don't over-call this imaging finding. This can be present even years after the surgery. Here's an interesting case that was followed up over time. So first you see a thin, partial thickness tear with a small delamination here. And 15 months later, this has developed into a large, full thickness re-tear of the supraspinatus tendon. So when to call something a tear, or when to call something normal? One criterion is you look at the anchors. So in this case, you see that directly at the anchor site, there is no tendon that is visible. So this is a definite re-tear. And posterior to the full thickness re-tear, you see a partial thickness tear of the posterior part of the supraspinatus and also of the infraspinatus tendon. The same criteria can be applied for this case. Next to the anchor at the lesser tubercle, there is a large tendon defect. So this is a definite re-tear of the subscapularis tendon. Here's a case that shows what we want to prevent. So this patient had a large, full thickness re-tear after the surgery. And then six years after surgery, a whole osteoarthritis of the shoulder developed. So you can see the cranial migration of the urinal head with a severe degeneration at the location of the acromion. If you look at the glenohumeral joint, you see deep cartilage defects here and also these osteocytes. So this is why we perform shoulder surgery, not only for regaining the function, but also for the prevention of osteoarthritis. Now one caveat, if you perform imaging in the post-op situation, not all full thickness re-tears are clinically relevant. We will see these small lesions in about 20% of patients with a good clinical outcome. So don't call this a definite re-tear that is painful. This could be actually asymptomatic. And especially if you have small, full thickness defects with a diameter below one centimeter, this has been shown to be often asymptomatic. So now let's look at the muscles in the last part of my talk. So we have already heard before how we can assess the state of the muscles regarding the muscle volume and the fatty degeneration. So these are data that I'm showing from a prospective study where we followed up patients for 12 months after the surgery. And what we looked at here was a single slice where we drew an area. So within this area, the muscle fat fraction was calculated. This analysis included 40 patients, half with intact tendon re-tears and half with failed tendon re-tears. What we found was that patients who went on to have failed re-tears, they started out with a slightly higher fat fraction preoperatively. So the more fat the patient has before the surgery, the higher risk of the re-tear. And the second message is a very good one. If you look at the green line here, after a successful repair, the fatty infiltration of the muscle can almost be stopped. So these were data from a 2D analysis of the supraspinatus muscle belly. The next level of evaluation is to assess the whole muscle volume in 3D. So in red, you see the segmentation of the supraspinatus muscle here. We used this method in over 130 patients before and after the surgery, and we were able to define cut-off values for the different muscles, ranging from 6% for the supraspinatus muscle to 8.3% for the subscapularis muscle. And we were surprised how low the threshold is here. So a surprisingly small amount of increased muscle fat already has an impact on patient outcome after surgery. So the patient and the surgeon might still decide to go on with the surgery, but they know that the risk for a re-tear is higher in patients with more muscle fat. So let's summarize my talk on post-op rotator cuff. Both ultrasound and MR imaging can be seen as first-line imaging options. Know the different surgical techniques, such as the single and double row techniques or the tendon transfers, and know the normal post-operative course. Pre-op fat fraction is a very important biomarker. It's predictive of the outcome after rotator cuff repairs. And imaging, obviously, is very important for identifying complications after a surgery, such as re-tears, dislodged anchors, or a foreign body reaction. Okay. Many thanks for your attention. It is my pleasure to announce the final speaker. Dr. Mohammad Samin from NYU will talk about post-operative shoulder instability. Thank you, Rita, and thank you, RSNA, for the invitation. So for the interest of time, I'll be focusing on the anterior shoulder instability, which is the most common type of shoulder instability. And as you know, appropriate treatment for shoulder instability depends on several factors, including the characterization of the glenoid, bancard, and humeral heel sacs, lesions known as bipolar bone loss. Following surgery, patients may present with recurrent instability, and typically, these are the patients who get MRI. And what surgeons want to know are the status of the repair tissue, such as labrum, capsule, or tendon, the orthopedic fixation harbor, or the grafts, and other possible complications. In respect of imaging, surgical management focuses on the bipolar bone loss quantification to see if they're engaging or non-engaging, or the more popular and recent concept is the bipolar bone loss, and to see if the bipolar bone lesion is on track or off track. It was beautifully discussed in Dr. Afonso's speech, so I'm going to skip a few slides because it was really explained well. So there are various surgical options to treat anterior shoulder instability in isolation or combined, that I'm going over some. But the general goal is to convert the off-track lesions to on-track. This is a proposed surgical approach based on the glenoid track, starting with the amount of glenoid bone loss, and then whether the lesions are off-track or on-track. Just know that these numbers are changing. The 25% is some institution may still use it, but there are different thresholds. But for the purpose of this talk, I'm just going to leave it as that number. Starting with the Bancard repair, which includes repair of the anterior labrum for patients with glenoid bone loss of less than 25%, it can be done with capsular shift plication, which is essentially tightening of the joint capsule. At the bottom, you see the recurrence rate. These are the numbers in the literature after the surgery, what's the chance of recurrent anterior shoulder instability. It is Bancard repairs mostly done arthroscopically. The torn and detached labrum is reattached to the glenoid using anchor screws or sutures. In terms of post-op imaging, MR arthrography is ideal and more sensitive and specific than conventional MRI. On MR, usually the intact repaired labrum should be present and attached to the anterior inferior glenoid. Signal and shape may be variable, but regardless, there should be sufficient repair tissue present and continues with the osseous attachment without any gap. When I have these cases, I always try to locate the anchors first, and I see them best on the sagittal non-fat suppressed images like T1. Then on axial, I looked at the repaired labrum at the level of the anchors, making sure that I can see labral tissue attaching to the glenoid. And notice heterogeneous and irregular shape of the labrum, which is otherwise intact and attached to the glenoid and the anchor. Another example, at the level of the anchors, there is thickened and heterogeneous but otherwise intact repaired labral tissue with no fluid gap or evidence of re-tear. How about this case? You notice there's irregular and insufficient labral tissue at the level of the anchor. You can see a chronic heel-sax lesion without any bone marrow edema. And when you compare the size of the heel-sax with the preoperative MRI, the interval increase in the size of it tells you that this patient has had more episodes of anterior instability after surgery with chronic recurrent re-tears of the repaired labrum, even in the absence of the bone marrow edema, which only tells you that patient did not have any recent episode of anterior instability. There are other signs of recurrent labral tear, including detachment and fragmentation of the repaired labrum, fluid signal in the labrum, fluid gap between the labrum and the glenoid, imbibition of contrast into the labrum, and the contrast field gap between the labrum and the glenoid when an MR arthrography is performed. This case has almost all signs of anterior instability. At the top, there is heel-sax lesion with bone marrow edema. More inferiorly, there is deficient labral tissue, and then fluid undercutting the repaired labrum with detachment, and finally, fragmentation. And most inferiorly, you can notice soft tissue edema and injury involving the inferior glenohumeral ligament or Hagel lesion. Imbibition of contrast, as I mentioned, in the repaired labrum and gap between the labrum and the glenoid at the side of the anchor mean recurrent tear in this MR arthrography with the patient in the ABER position. There are other complications listed here, including postoperative cartilage defect, as seen in this case, with very ill-defined and irregular repaired labrum and granulation tissue formation along the anchor tract. Next is the glenoid bone augmentation, including Bristow and Latter-J that are considered when glenoid bone loss is greater than 25%, although they may also be performed for glenoid bone loss of less than 25%, specifically in more active and high-functional patients or elite athletes. In both Latter-J and Bristow procedure, ipsilateral coracoid bone block attached to the conjoint tendon is harvested and attached to the anterior glenoid. The attached tendon functions as a dynamic stabilizer, pretty much similar to the IGHL. The difference between the two, in Bristow procedure, only the lateral tip of the coracoid process is transformed. The graft is attached to the glenoid through its resected surface, and the graft is usually smaller and, for that reason, attached using just one screw. In Latter-J, the entire horizontal coracoid graft is transferred, and the graft is attached to its inferior surface, and the graft is usually larger and attached with two screws. Radiograph and CTs with metal reduction technique are the main imaging modalities to evaluate the graft status. The graft position is the most important factor in technical success and should be assessed with serial radiographs. The main complications are non-union of the graft and hardware loosening or fracture. And usually quantitative two-third bone bridging is considered successful fusion, but it is important to know that incomplete graft union is not equal to postoperative pain or instability. Therefore, union status of the graft should be interpreted within the clinical context. Again, radiographs can show the two screws here fixing the graft along the anterior glenoid in a very good alignment. CT, including 3D reconstruction with metal reduction, is great in showing the graft position, which should be flushed with the normal curvature of the native glenoid. Sagittal shows the good position of the graft, typically along the anterior glenoid, and both screws are intact with near-complete fusion of the graft of the glenoid. So MRI with metal reduction may have a role, especially when the concern is not necessarily the graft union or hardware issue. Here MRI shows metallic screws and the black arrows show intact conjoint tendon attached to the coracoid graft. And in another example, MRI shows partially dislodged screws and fluid-filled cleft between the graft and the glenoid, indicating non-union. Screws and graft can get dislodged and displaced into the joint and sometimes surrounding soft tissues. And other potential complications are acute detachment of the subscapularis tendon, fatty degeneration of the subscapularis muscle, infection, or secondary oxalated osteoarthrosis, which should be very well scrutinized when you're reading this post-op imaging. So now look at the sub-surgical management of the Hill-Sachs lesions, which are present in almost 100% of the patients with recurrent anterior instability. Starting with remtissage, which in French means feeling, and it is feeling of the Hill-Sachs defect using joint capsule and infraspinatus tendon. It is usually used for off-track lesion in combination with bankart repair, with smaller glenoid bone loss, or laterje, with larger bone loss. A potential disadvantage of remtissage is the anatomical alteration of the rotator cuff muscles, which can result in loss of shoulder range of motion and stiffness. So remtissage involves the joint capsule and infraspinatus tenodesis, and fixing them with screw anchors along the Hill-Sachs defect. Again, MRI is the post-op imaging modality of choice, and normally we see tendinosis of the infraspinatus tendon, and the tendon and the capsule closely filling the Hill-Sachs defect fixed with suture anchors tendon tissue, which are completely filling the defect with no fluid pooling in the defect. In this case, however, you'll notice fluid pooling between the Hill-Sachs defect and the soft tissue remtissage, with a clear gap indicating the adhesions at the side of the suture anchor. And for those who are interested, there is an MRI grading system based on the feeling index score of the remtissage, which refers to the extent of the fluid or contrast agent pooling in the Hill-Sachs defect on a five-point scale. And finally, reconstruction of the humeral head is usually reserved for the large Hill-Sachs lesions, usually more than 40% in more young, active patients. There are different grafts available. Most popular is the osteochondral allograft, and the surgeon resects a cylindrical-shaped osteochondral plug from the cadaver specimen and prepare the recipient humeral head by performing an osteotomy, then subsequently fix the graft using fixation anchors. Post-op, we should check the alignment of the graft, ideally flush with or slightly above the native humeral head. In early post-op, sclerosis and bone marrow edema signal can be expected. Edema can persist for years, but rule of thumb is that the edema signal should not get worsened over time. There are other signs suggested of graft failure, such as motion or malalignment, worsening, step-off, or gradual increase in the bone marrow edema. And here are two examples I'm showing you. CT can show the graft position, which feeling the Hill-Sachs defect really well. Black arrows show the good osseous incorporation of the graft fixed with two intact screws. And here is an example on MR, show the Hill-Sachs defect before and after allograft reconstruction. The defect is completely filled with the graft flushed with the humeral head. Graft is fully incorporated, and there is no bone marrow edema signal. So the take-home points are the surgical management option depend on bipolar bone loss quantification and glandular bone tract subconcept. The goal is to convert the off-tract to on-tract by addressing glandular and humeral lesions. When you're reading these studies, always look for other signs of recurrent instability. Heterogeneous repair tissue can be normal after bank or repair. Incomplete latergic graft osseous union is not equal to postoperative pain or instability, and interpretation always should be in the clinical context. Thank you very much for your attention.

shoulder imaging

rotator cuff pathology

glenohumeral dislocations

MRI

SLAP tears

Bankart lesions

post-operative assessment

tendon tears

advanced imaging techniques