Current Topics on Knee MRI

Abstract

Knee pathology is one of the most common complaints worldwide. Among the most common complaints is ligamentous and meniscal injuries, for which MRI is the main diagnostic tool. Advances in MRI have improved the accuracy of detecting Anterior Cruciate Ligament (ACL), posterior cruciate ligament (PCL) and meniscal tears, which have helped orthopedic surgeons treat and classify injuries accordingly. Understanding the anatomy, different protocols and the advances will help orthopedic surgeons to deliver better patient care. MRI is especially important in ACL pathology due to its implication in femoral and tibial tunnel positioning; the more anatomically we can reconstruct the ACL, the better the functional outcomes. This is true for most of the ligamentous pathology of the knee. This chapter will review the current indication and further research areas in knee pathologies.

1. Introduction

Magnetic resonance (MR) is the preferred non-invasive imaging method to assess knee musculoskeletal injuries due to its high soft tissue resolution, and it is considered the reference standard with the additional benefit of avoiding exposition to ionizing radiation [1]. New advances in magnetic field and gradient strength allow the development of sequences for ultrastructure imaging and even postoperative ligament reconstructions or meniscal repair, which has represented a challenge to date.

Isotropic three-dimensional MR imaging has a thin slice of less than 1 mm and less partial volume artifacts with the use of thin continuous sections and oblique planes that are helpful for complex structure analysis like static and dynamic knee stabilizers with their tissue-osseous relationships [2]. These advances will be tools for preoperative planning and clinical decisions in patients with ligament and meniscal injuries.

2. Current topics in anterior cruciate ligament on MRI

The three-dimensional configuration of the anterior cruciate ligament is important to determine different conditions in association with the prognostic of the anterior cruciate ligament reconstruction, as parameters related to the most anatomical possible position. Ortiz et al. described the orientation of the anterior cruciate ligament in resonance, proposing a triplane trigonometric method; as a result, they found that the mean angle in sagittal, coronal and axial projection were 76.95, 81.65 and 33.17 degrees, respectively. It is expected that this method may be applicable for planning anterior cruciate ligament reconstruction, using the position of the anterior cruciate ligament of the uninjured knee as a reference [3].

In the evaluation of an acute lesion of the anterior cruciate ligament in MRI, there are different signs that can help with the diagnosis, the most sensitive being the discontinuity of the fibers or irregularity, increased signal on T2, bone bruises and abnormal orientation of the fibers. Secondary findings include bone lesions mainly at the level of the external femoral condyle and external tibial plateau, and it is less common to present lesions at the internal femoral condyle and tibial plateau; other findings are the anterior translation of the tibia and uncovering of the posterior horn of the lateral meniscus greater than 3.5 mm [2].

The normal appearance of the anterior cruciate ligament is characterized by a uniform caliber with a course parallel to Blumensaat’s line, a high or intermediate signal on T1 and T2, and some signs of fluid between the fibers [4].

2.1 Indirect MRI signs related to anterior cruciate ligament injury

The lateral femoral notch sign is a radiographic phenomenon defined as an impaction greater than 2 mm of the lateral femoral condyle, and it may result from a traumatic episode in which the lateral femoral condyle collides with the proximal tibia, causing a defect in the lateral femoral condyle with high signal on T1 (Figure 1) [5, 6].

Figure 1.

Lateral Femoral Condylar Notch, Indirect Sign of Anterior Cruciate Ligament Injury.

Haluk Yaka validated a posterior base measurement of the medial and lateral meniscus, defined as a line passing through the tibial edge of the meniscus and a line passing through the capsular edge on the sagittal side of the posterior horn of the meniscus based on the previous studies performed by Hohman where the relation of the posterior angle of the base of the meniscus and its relationship with the anterior cruciate ligament injury was discussed, it was concluded that the medial and lateral angles above 84.5 and 93.15, respectively are an indirect finding of anterior cruciate ligament injury (Figure 2) [7].

Figure 2.

Posterior base meniscus angle, a line through the base of the meniscus and a line through de posterior aspect of the meniscus in the sagittal plane.

2.2 Location of anterior cruciate ligament injury point in MRI

Within the management protocol for anterior cruciate ligament injuries, there is the option of performing primary repair in cases in which the injury is proximal near the insertion of the femur and a contained injury with good tissue. In order to validate this condition, Sherman classified injuries into five types, been I and II types the proximal tears, in which a primary ACL repair can be considered. Now, before performing an arthroscopy, it is important to know if the patient is a candidate for repair or not; it could change the treatment and rehabilitation approach. For this purpose, it has been described in recent years to resonance as a method that allows us to know the location of the rupture of the anterior cruciate ligament [8, 9].

In a study carried out by Guillien et al. at the University of Rennes, they evaluated the correlation between the anterior cruciate ligament lesion point determined in MRI in comparison with the findings in arthroscopy, and this study was based on the Sherman classification (Figure 3). It was identified that the correlation in relation to the position of the lesion is approximately 70%; it was not the same for the evaluation of the quality of the ligament determined in resonance, in which the correlation was 50%. In another work carried out by Vanderlist, the predictive capacity of preoperative resonance is evaluated in relation to anterior cruciate ligament repair in Sherman type I and II injuries, finding that in 90% of the cases that a lesion was diagnosed pre-surgical type I repair of the ACL was performed and in 88% of type II [4].

Figure 3.

Anterior cruciate ligament tear in the proximal third of the ligament.

As mentioned previously, the complex three-dimensional ACL orientation does not allow for the complete visualization of the ligament in a single image. ACL runs obliquely through the intercondylar notch; this is why various MRI techniques including oblique planes have been investigated (Figure 4) [3].

Figure 4.

A 55-year-old female left knee. Oblique coronal PD FSE MR images, performed using a 1.5 T system with 1.2 mm thickness (A) anterior bundles of ACL (arrow) and it is possible evaluate posterior meniscal roots (red open arrows). (B) Posterior bundles of ACL (arrow).

Kwon et al. from the Department of Radiology and Center for Imaging Science and Department of Orthopedic Surgery of Samsung Medical Center, Seoul, Korea, published a study whose purpose was to evaluate the diagnostic role of additional use of oblique coronal and oblique sagittal imaging for an ACL injury [10]. The study population consisted of 101 patients with a mean age of 35 +/− 12.6 years who required knee arthroscopy for suspected of having a torn ACL on MRI examination with both orthogonal and oblique images using 1.5 MRI system conventional protocol sequences with section thickness 3 mm, TR/TE 2000–3800/20-30 ms and additional oblique coronal/sagittal proton density-weighted imaging. The oblique sagittal image was made in the plane parallel to the medial border of the lateral femoral condyle on an orthogonal coronal image and the oblique coronal image was obtained in plane parallel to course of the femoral intercondylar roof using a sagittal image. Two musculoskeletal radiologists analyzed the knees MRI retrospectively, without knowledge about arthroscopic results, then determined intact, probable tear or definite ACL tear. They evaluated sequences by using four methods (M), MA: orthogonal images only, MB: orthogonal and oblique coronal, MC: orthogonal and oblique sagittal, and MD: orthogonal, oblique coronal and sagittal images.

Diagnostic performance with sensitivities, specificities and accuracies of each method respect arthroscopy (partial or complete ACL tear) as a gold standard was as follows respectively, MA 95%, 83.6%, 88.1%, MB 97.5%, 95.1%, 96%, MC 97.5%, 95.1%, 96% and MD 97.5%, 98.4%, 98%, meaning that specificities and accuracies for methods B, C D were statistically significantly higher than method A. No difference was found between methods B, C and D. This study concludes that some oblique imaging added to standard MRI sequences improves the ability to diagnose ACL tears [10].

2.3 Mucoid degeneration of the anterior cruciate ligament

A topic of interest in recent years has been mucoid degeneration of the anterior cruciate ligament as a rare entity, which can be confused with an ACL lesion, being difficult to diagnose. Bergins et al. described an incidence of 1.8% in an analysis of 4221 patients, in addition to describing different aspects that can help diagnose mucoid degeneration of the anterior cruciate ligament, which includes the presence of a uniform thickening with a bulging ligament, semiologically described as “celery stalk,” increased intermediate intraligamentary signal on T1, and hyperintensity on T2, maintaining orientation and continuity of the anterior cruciate ligament fibers. More recently, Cilengir et al. described MRI findings that can help differentiate mucoid degeneration of the anterior cruciate ligament from injury and found an increased prevalence of intraosseous femoral cysts being part of the mucoid degeneration. Other authors have described anatomical conditions in resonance that can be associated with mucoid degeneration of the anterior cruciate ligament, such as an increase in the angle of the tibial slope, a decrease in the width of the intercondylar groove, male sex [11, 12, 13].

3. MRI accuracy of posterolateral and posteromedial corners injuries

3.1 Normal anatomy of the posterolateral corner

Posterolateral corner (PLC) is currently a diagnostic challenge.

PLC structures are grouped into primary stabilizers, which are statics, and secondary stabilizers, which are static and dynamic [14]. PLC structures are the main mechanism to protect against knee varus stress and posterolateral rotation of the tibia with respect to the femur [15, 16, 17].

Table 1 shows all the PLC structures with their respective origins and insertions, and the graphic scheme of the structures is shown in Figure 5.

Structure	Origin	Insertion
Fibular collateral ligament	Small bony depression proximal and posterior to the lateral epincondyle	Fibular head distal to the tip of the fibular styloid.
Popliteus tendon	From popliteus sulcus of the lateral femoral condyle	Runs posteromedially deep to the fibular collateral ligament, exits the joint capsule through the popliteus hiatus and inserts along the posteromedial aspect of the proximal tibia.
Popliteofibular ligament	From popliteus tendon proximal to the myotendinous junction	Inserts onto the medial downslope of the fibular styloid.
Midthird lateral capsular ligament	Thickening of the lateral capsule from the lateral epicondyle of the femur	Capsular attachments to the lateral meniscus and inserts onto the tibia anterior to popliteal hiatus
Popliteomeniscal fascicles	Popliteus tendon	Posterior horn of the lateral meniscus
Lateral gastrocnemius tendon	Posterior lateral femoral condyle	Achilles tendon onto calcaneus
Fabellofibular ligament	Thickened distal aspect of the biceps femoris that extends from an osseous fabella (sesamoid bone)	Fibular styloid.
Arcuate ligament	Y-shaped thickening of the posterolateral join capsule	Fibular styloid with the fabellofibular ligament
Biceps femoris	Ischiatic tuberosity and femoral diaphysis	Consists of a long and short head and also has numerous insertions arms on fibular styloid and tibial lateral condyle
Iliotibial band	It is a thick band of fascia formed proximally at the hip by the fascia of the gluteus maximus, gluteus medius and tensor fasciae latae muscles.	Gerdy’s tubercle

Table 1.

Components of the posterolateral corner.

Figure 5.

Scheme of the structures of the posterolateral corner.

The three primary stabilizers are the fibular collateral ligament (FCL), popliteus tendon (PLT) and popliteofibular ligament (PFL). FCL is the main stabilizer with varus stress, and PLT acts as a stabilizer regarding tibial external rotation.

On magnetic resonance (MR) imaging FCL is visualized on axial and coronal plane as low signal-intensity band extending from the lateral epicondyle to the lateral aspect of the proximal fibula. The PLT is seen as a low-signal-intensity structure on axial or sagittal sequences [18].

The PFL has an anterior and posterior bundle that embraces the popliteus myotendinous junction, and it is best seen as a low-T2-signal structure on coronal and sagittal planes, deep to the inferior lateral genicular vessels. However, the PFL is not currently described in conventional MR imaging. Coronal oblique sequences and isotropic 3D MR improve visualization of these tissues (Figure 6) [19, 20].

Figure 6.

MR images of a left knee of a 40-years old female patient. (A) Axial fat-saturated T2 weighted sequence demonstrates posterolateral structures. Popliteus tendon (PLT), fibular collateral ligament (FCL), biceps femoris tendon (BF), lateral capsular ligament (ALT), iliotibial band (ITB). (B) Sagittal proton density-weighted image lateral knee with popliteus hiatus formed by popliteomeniscal fascicles (red open arrows) and Popliteofibular ligament (PFL) located anterior to lateral inferior genicular artery (LIGA). (C) Coronal T1 image that shows PFL and its association with respect to LIGA.

Secondary stabilizers include the mid-third lateral capsular ligament (ALT), popliteomeniscal fascicles (PMF), lateral gastrocnemius tendon (LG), fabellofibular ligament (FFL), arcuate ligament, biceps femoris tendon and iliotibial band.

The ALT that is a thickening of the lateral capsule of the knee is seen on isotropic 3D MR axial images [20].

The popliteomeniscal fascicles form the roof and floor of the popliteus hiatus. They are visible in 60–94% of patients and can be seen on sagittal images of isotropic 3D MR.

FFL is best seen on sagittal and coronal MR imaging posterior to the lateral inferior genicular artery; however, this ligament is visible in 33–48% of patients.

Arcuate ligament is also inconsistent, but it may be identified as a thin band overlying the PLT on axial sequences [20].

Biceps femoris tendon (BFT) is composed of multiple distal insertion arms that may not be easily distinguishable but just long and short arms (Figure 6).

3.2 Posterolateral corner injuries and MR imaging

Injuries to the PLC most commonly occur with varus forces, particularly to a hyperextended knee or associated with knee dislocation. Diagnosis may be difficult in the setting of acute trauma because of the patient’s joint effusion. However, prompt diagnosis and management are important, as unrecognized PLC injuries may result in chronic instability and premature osteoarthritis [21]. Therefore, it would be desirable to predict not only by clinical testing but also by imaging. Figure 7 shows PLC injury patterns on MR imaging.

Figure 7.

MR images of a 23-year-old man with left knee PLC injury caused in a motorcycle accident. (A) Coronal proton-density weighted fat-saturated T2 sequence with popliteomeniscal fascicles disruption (red open arrow) and complete tear of fibular collateral ligament (FCL) and popliteus tendon (PLT). Lateral inferior genicular artery (LIGA), fibular styloid (FS). (B) Sagittal proton-density weighted fat-saturated T2 image with PLT tear and disruption of posterolateral ligamento-capsule complex (LCC) and evident disruption of poplitemeniscal fascicles (red open arrow). (C) Axial proton-density weighted fat-saturated T2 sequence with posterolateral capsule injury, midthird lateral capsular ligament (ALT).

A meta-analysis established that 1.5-T or 3.0-T MRI offers high diagnostic accuracy for evaluating injuries involving the meniscus, anterior cruciate and posterior cruciate ligaments. However, in multi-ligament injured knees, MRI had been found to have lower accuracy for the detection of PLC ligament tears [22].

A recent retrospective study from the Ottawa Hospital Research Institute and Department of Radiology determined the diagnostic performance of preoperative MRI for diagnosing PLC injuries of patients with knee dislocations compared to intraoperative findings [21]. They included 39 patients who required repair/reconstruction of the posterolateral corner between May 2005 and April 2020. Preoperative MRI of these patients was on 1.5 T or 3.0 T scanners, and all protocols included sequences in standard imaging planes.

The fibular collateral ligament, bicep femoris and popliteus tendon were categorized as normal, partial tear or complete tear. The posterolateral ligamento-capsule complex (LCC) was evaluated as a single unit that includes popliteofibular and fabellofibular ligaments. This complex and posterolateral capsule was classified as intact or torn; the same classification was used for intraoperative findings.

The diagnostic performance of MRI was a sensitivity (Se) of 95% and specificity (Sp) of 100% for detecting fibular collateral ligament (FCL) tears, Se of 100% and Sp of 77% for BFT tears, Se of 88% and Sp of 71% for PLT injuries and Se of 97% and Sp of 33% for LCC tears. The correlation between surgical findings and magnetic resonance of PLC structures was strongest for the BFT and weakest for the LCC.

This study reports accuracy ranging from 82 to 95% for detecting PLC injuries with MR imaging, even higher than other previous studies [23], probably associated with MRI evolution in the last few years.

Longer time between injury and surgery may allow some injuries to heal and can be found intact at surgery but still presenting with abnormal signals at MRI, leading to higher false positive counts; obtaining MR images closer to the time of injury may make their interpretation more challenging due to inflammatory process also affecting the radiologist reading [24].

Finally, this study concludes that despite the challenges of evaluating knee posterolateral corners, MRI has an acceptable accuracy for detecting their injuries.

Statistically, up to 20% of MRIs made in patients with an Anterior Cruciate Ligament (ACL) tear may reveal PLC injuries. A Swiss retrospective cohort study of The University of Zurich determined the diagnostic performance of different MR imaging findings for helping to predict posterolateral instability in patients with acute complete ACL tears by performing a decision tree analysis [25]. Their sample comprises 162 patients who underwent ACL reconstruction with or without concomitant posterolateral corner reconstruction. Clinical diagnosis of PLC instability requiring reconstruction served as gold standard, and there were obtained conventional MRI of all patients. Results demonstrated a low sensitivity and high specificity for posterior cruciate ligament, biceps femoris, popliteus tendon and lateral collateral ligament. Decision tree analysis results showed that a complete tear or fibular avulsion of the FCL was the most statistically significant finding to help predict posterolateral instability. These results are shared with other studies that affirm it is sufficient to assess the FCL, BFT and PLT to predict PLC instability [26].

With respect to small structures, this study confirms variable visibility of popliteofibular ligament, fabellofibular ligament and popliteomeniscal fascicles, which are not always observed at conventional two-dimensional MRI.

Limitations of this study are sample of patients limited to ACL reconstruction, MRI performed in the acute trauma 10 days or less, and it was different protocol and scanners to take images like in many other studies.

In the next years, the use of isotropic three-dimensional high-resolution sequences could allow for oblique reconstructions and individual examination for each patient. A retrospective study performed at Gachon University of South Korea aimed to document the appearance of PLC structures on 3D isotropic and routine two-dimensional MR images and to determine the significance of pathologic findings in patients with confirmed posterolateral instability. They evaluate conventional 3.0 T MRI of 25 patients with surgery indication as the gold standard and also of 25 control patients with any radiological or clinical finding, but in addition to standard sequences 3D isotropic SPACE (Sampling perfection with Application optimized Contrasts using different flip angle Evolution) images were obtained until adequate visualization of posterolateral corner. Their findings were the following:

• The popliteofibular ligament was best seen with 3D isotropic images. The lateral geniculate artery appeared as a landmark.

• 3D images detected normal and partial tears (grade 1 or 3) of PLC, and 2D images just when there were complete tears (grade 4).

• The “fibular cap sing” that represents no-osseous avulsion of the distal FCL from the tip of the proximal fibula on 3D images was found to be useful for the diagnosis of PLC tears.

A disadvantage of 3D isotropic imaging is the lack of fat suppression that may underestimate slightly altered ligament signals.

Despite these limitations, 3D isotropic SPACE MRI could be an interesting examination method in institutes interested in multiligamentary knee reconstruction [26].

3.3 Normal anatomy of the posteromedial corner

The posteromedial corner (PMC) contains the structures lying between the posterior margin of the superficial medial collateral ligament (MCL) and the medial border of the posterior cruciate ligament (PCL). These structures avoid anteromedial rotational instability and provide restraint to valgus stress. Although some authors do not consider MCL to be part of posteromedial corner, recently, an international expert consensus panel has included it [27, 28]. Figure 8 illustrates the borders of the PMC, Table 2 shows PMC structures with their respective origins and insertions, and Table 3, Figures 9 and 10 describe normal MR imaging of PMC.

Figure 8.

Axial fat-saturated MR image of a left knee illustrated the borders of the PMC (green outline).

Structure	Origin	Insertion
Semimebranosus tendon	Ischiatic tuberosity	5 expansions; Direct arm (principal attachment) Capsular arm Extension to the OPL Anterior arm or reflected arm Inferior or popliteal arm
Oblique Popliteal Ligament (OPL)	Arises from the capsular arm of the POL and lateral expansion of the semimembranosus	Attaches to fabella, the meniscofemoral portion of the posterolateral joint capsule and plantaris muscle.
Posterior oblique ligament (POL)	Origin distal and posterior to adductor tubercle	3 arms in the posteromedial aspect of medial meniscus and the tibia.; Central or tibial arm Superior or capsular arm Distal arm
Posteromedial capsule	Includes deep MCL with its meniscotibial and meniscofemoral components	It is reinforced externally by the POL and expansions from the semimembranosus
Posterior horn of the medial meniscus	Posterior horn has a “Brake stop” function to anterior translation of the tibia	The medial meniscus attaches to the capsule posteromedially and the meniscotibial ligament anchors the meniscus to the tibia

Table 2.

Components of the posteromedial corner.

Structure	MR Imaging
Semimebranosus tendon; Figure 7,; Figure 8	Direct arm is seen in sagittal and axial images along the posteromedial aspect of the tibia distal to the articular margin Anterior arm is seen on coronal plane as an oval structure deep to the MCL Other expansions are difficult to identify [29, 30]
Oblique Popliteal Ligament (OPL); Figure 7,; Figure 8	The OPL is a thin structure and is difficult to distinguish from the posterior joint capsule.; When is thicker, it is seen on sequential sagittal and axial images as a band extending obliquely from the main tendon of semimembranosus laterally toward lateral femoral condyle [27].
Posterior oblique ligament (POL); Figure 7,; Figure 8	POL is best visualized on coronal and axial images at the level of the femoral condyle. The three arms run continuously with each other, are difficult to distinguish from each other [30, 31].
Posteromedial capsule	The attachment of the peripheral surface of the meniscus to the capsule and the tibia is best evaluated on sequential coronal and sagittal images posterior to the superficial MCL [30].
Medial collateral ligament; Figure 7, Figure 8	Best visualized on coronal and axial images.

Table 3.

Normal MR imaging of PMC structures.

Figure 9.

Axial proton density weighted fat-saturated MR images of a normal knee, proximal to distal. (A) At the level of the femoral condyles illustrates medial collateral ligament (MCL), posterior oblique ligament (POL) and semimembranosus tendon (Sm). (B) At the level of joint space, oblique popliteal ligament (OPL). (C) At tibial plateau level, tibial semimembranosus expansion (TSm), inferior arm (ISm) and direct arm (DSm).

Figure 10.

Coronal proton-density weighted fat-saturated images and sagittal proton density images of a normal left knee. (A) Posterior oblique ligament (POL), anterior expansion of semimembranosus (ASm). (B) Inferior semimembranosus arm (ISm). (C) Oblique popliteal ligament (OPL), (D) Tibial and inferior expansions of semimembranosus.

3.4 Posteromedial corner injuries and MR imaging

The semimembranosus is the main dynamic stabilizer of the PMC; without its dynamic support, the remaining PMC structures fail over time and lead to instability that affects the anterior cruciate ligament or posterior cruciate ligament [32].

MR imaging is the modality of choice for PMC injury assessment; however, at present, there are no specific studies that have evaluated the sensitivity and specificity of PMC structure injuries on MR imaging, unlike PLC structures.

A retrospective study of patients with symptomatic anteromedial rotational instability who were treated with ligament reconstruction and based on surgical descriptions found injury of the posterior oblique ligament (POL) in 99% of the cases, injury to the semimembranosus in 70% and peripheral meniscal detachment in 30% [33].

Tears in these three structures are well defined on MRI with an established classification system for each one of the ligament structure injuries.

House et al. have proposed the same classification used for the MCL injuries in the POL injuries, and it is grade I, intact ligament with edema surrounding it (T2 high signal). Grade II thickening of the ligament with partial disruption of fibers and Grade III with complete disruption of the ligament [34]. The coronal plane allowed for visualization of the POL; however, the coronal oblique plane, in combination with the axial plane, improved the analysis of the POL. In case of doubt, the addition of intraarticular contrast material can optimize the visualization of the POL and capsular layers in the axial plane as these structures are displaced away from the femur (Figure 11) [35].

Figure 11.

MRI of a 53-year-old man with right knee PMC injury caused in a twisting knee injury. (A) (B) axial proton density weighted fat-saturated sequence at the level of joint space and at tibial plateau level respectively illustrate OPL partial tear, POL, MCL complete tears and capsular separation (red open arrow). (C) Axial T1 fat-saturated image shows tibial (TSm) and inferior (ISm) semimembranosus arms injuries. (D) Coronal proton-density weighted fat-saturated image with POL complete tear.

With respect to medial meniscocapsular injuries, a “reverse Segond fracture” represents meniscotibial ligament osseous avulsion, also associated with posterior cruciate ligament rupture. Meniscocapsular separation is best visualized in the sagittal sequences. When there is increased signal intensity and thickening of the capsule, it may be associated with capsule sprain, but it could be a ruptured popliteal cyst, too. Hence, it is important to interpret according to clinical history and other imaging features.

4. MR imaging of the postoperative meniscus

Meniscal surgery is a frequent orthopedic procedure [36]. Clinical examination and MR imaging are both the current way to assess patients who complaint about knee pain after meniscectomy or meniscus repair. However, the evaluation after surgery can be difficult and represents a challenge to date.

The normal medial and lateral meniscus are hypointense on T1- and T2-weighted MR images. On axial plane, they are like C-shape; on sagittal images, they appear as a wedge configuration; on coronal plane, they are seen as a right-triangular with a free edge oriented to intercondylar notch [37, 38, 39].

Sagittal and coronal images of intermediate and/or T1-weightened sequences of conventional MRI are the method of choice to assess signal changes in nonoperative meniscus, however after meniscal surgical procedure, it is the T2-to-intermediate-weighted fluid-sensitive images in sagittal and coronal planes to detect synovial fluid signal extending into the substance of the meniscus indicating that the articular surface has been breached due to a new retear [39].

In regard to magnetic resonance with contrast, direct magnetic resonance arthrography (MRA) is useful to evaluate recurrent tears or unhealed repair when there is an extension of contrast into a meniscus substance. Disadvantages of this examination include other invasive procedures, infection, bleeding and allergic reactions. If there is fluoroscopic-guided injection, radiation exposure is another risk and finally entails more costs. In some countries, it is considered an off-label use of gadolinium-based contrast agents, according to FDA [40].

Indirect MR arthrography involves the intravenous administration of gadolinium-based contrast; it allows the identification of sites of hyperemic synovitis associated with vascular tissue enhancement. Nevertheless, the stable healed granulation tissue, as expected after meniscus surgery, may be difficult to differentiate from a residual tear, and it may result in potential false positives. Disadvantages are costs, patient time and adverse reaction to contrast agents [40].

4.1 Imaging findings after meniscectomy

Low signal linear fibrotic tissue in Hoffa fat pad is. A sign of precious arthroscopic knee surgery [39, 40]. With respect to partial or total meniscectomy, it could be found:

Diminution of meniscal tissue, ranging from a large portion of the meniscus being removed to mild blunting of the apical margin [39].

Baker et al. reviewed PubMed published evidence from 1990 to 2017 about recurrent tears after partial meniscectomy imaging. They found nine studies that reported the accuracy of conventional MRI, direct MRA and indirect MR arthrography compared to second-look arthroscopy. Conventional MRI had accuracy ranging from 57 to 80%, direct MRA from 85 to 93% and indirect MRA from 81 to 93% [40]. However, some other studies, specifically a randomized cohort study made by White et al., published in Radiology Journal compare the accuracy of conventional MRI, direct MRA and indirect MRA, with inconclusive results that do not find statistical differences among the three techniques in the setting of a recurrent meniscal tear, although there was a trend toward increased diagnostic performance for both direct and indirect MRA [41].

Intermediate-signal intensity extending to the articular surface of the postoperative meniscus on fluid-sensitive sequences has been the most specific sign of retear [41], meaning that its absence could be a negative predictive sign of an intact meniscus after surgery (Figure 12) [39, 42].

Figure 12.

(A) A 62-year-old man following partial medial meniscectomy, sagittal fat-saturated T2 and sagittal proton density images respectively that show diminutive body and posterior horn of medial meniscus (orange arrows) and (B) fibrotic stranding in Hoffa fat pad (arrowhead). (C) and (B) same patient. Sagittal fat-saturated T2-weighted image illustrates surfacing intermediate signal extending to the apical meniscal articular surface characteristic of a new tear.

4.2 MRI findings after meniscal repair surgery

In this kind of surgery, intrinsic high signal may be seen on intermediate and T2 MRI such as the features of preoperative meniscal injury.

A cohort study performed by the Institute of Sports Medicine of Peking University evaluated the diagnostic performance of MRI compared with second-look arthroscopy as the gold standard in 81 patients to evaluate the healing of the repaired meniscus. They found T2-weighted sagittal and coronal sequences had higher specificity (89.6–98.7%, respectively) and accuracy (85.4–91%), while T1 and proton density had higher sensitivity, 91.7% and 75%-83–3%, respectively. The diagnostic value could be improved by a combined application of different sequences [43].

The same study by Pekin University also demonstrated that approximately 50% of the patients with intact menisci in diagnostic arthroscopy illustrating features of surfacing increased intermediate-weighted linear signal at the healed repair site [43].

On MRA, one of the findings that may represent a partially healed repair or a partial thickness recurrent tear is the extension of intraarticular contrast through the meniscal repair site from one articular surface to another [39].

As a clinical care point, MRI or direct and indirect MR arthrography is the examination of choice for patients with suspected meniscus retear. However, those methods can be challenging because conventional diagnostic criteria of a meniscal tear may be normal findings postoperatively.

5. Three-dimensional isotropic MRI of radial and root tear of meniscus

Almost all meniscal injuries have suggested that the sagittal imaging sequences are the most accurate for detecting them; however, the radial and the root tears could be missed on conventional 2D images due to thicker slices of axial sequences, which is the preferred plane to evaluate both. Several studies describe thin-Section 3D FSE sequences such as Volume isotropic turbo spin echo acquisition (VISTA), CUBE and SPACE to analyze menisci injuries with better quality image of peripheral and radial tears (Figure 13) [44].

Figure 13.

A 45-year-old female left knee. MRI were performed using a 3.0 T system with 2 mm thickness (A) PD FSE FS axial 2D image with suspected tear of posterior root of medial meniscus (arrow). (B) Multiplanar reformatted axial 3D VISTA (volume isotropic turbo spin echo acquisition) image of 0.5 mm slice thickness with better delineation of posterior root tear (arrow).

Daekeon Lim et al. from Yonsei University College of Medicine, Seoul, Korea, assessed the diagnostic value of FS 3D VISTA (Volume isotropic turbo spin echo acquisition) protocol imaging compared to 2D standard imaging in detecting arthroscopy-confirmed (gold standard) radial and root tears. Their results reported sensitivity and specificity of 96% and 96% with VISTA protocol imaging, respectively, versus 87% and 91% with 2D imaging. They found higher sensitivity and specificity with isotropic 3D imaging and excellent interobserver agreement for detecting meniscal radial and root tears. Some limitations and bias were retrospective study design, size of sample, and MRI readers were aware that patients had undergone arthroscopic surgery that could overestimate the lesions [44].

6. Conclusions

The advances in magnetic resonance research imaging have made it possible to achieve greater detail in the diagnosis of knee joint pathologies. Different protocols and MRI sequences have been described, as well as clinical signs for different conditions such as posterolateral corner lesions and mucoid degeneration of the anterior cruciate ligament. These tools should be used as part of the clinical approach to patients with traumatic knee injuries.

Conflict of interest

The authors declare no conflict of interest.