Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review

Osteoarthritis and Cartilage xxx (2015) 1e15

Review

Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review F.W. Roemer y z *, M. Jarraya z x, D.T. Felson k, D. Hayashi z ¶, M.D. Crema z #, D. Loeuille yy, A. Guermazi z y Department of Radiology, University of Erlangen-Nuremberg, Erlangen, Germany z Quantitative Imaging Center (QIC), Department of Radiology, Boston University School of Medicine, Boston, MA, USA x Department of Radiology, Mercy Catholic Medical Center, Drexel University College of Medicine, Darby, PA, USA k Clinical Epidemiology Research and Training Unit, Boston University School of Medicine, Boston, MA, USA ¶ Department of Radiology, Bridgeport Hospital, Yale University School of Medicine, Bridgeport, CT, USA ~o (HCor) and Teleimagem, Sa ~o Paulo-SP, Brazil # Department of Radiology, Hospital do Coraça yy Service de Rhumatologie, Facult e de M edecine, CHU Brabois, UMR-CNRS 7561, 54500 Vandoeuvre les Nancy, France

a r t i c l e i n f o

s u m m a r y

Article history: Received 11 April 2015 Accepted 22 September 2015

Objective: To give an illustrative overview of Hoffa's fat pad pathology with a radiologic emphasis on the anatomy, on technical considerations, and on imaging differential diagnoses in the context of osteoarthritis (OA) imaging research. Design: A PubMed database search including only English literature and covering a 20 year period was performed. The search was based on but no limited to the query terms “Hoffa”, “Hoffa's fat pad” or “infrapatellar fat pad (IPFP)” in combination with “synovitis”, “OA”, and “magnetic resonance imaging (MRI)”. The literature search yielded 289 publications that were screened for relevance; additional references were included when these were considered of importance. Results: Several anatomic variants and pathologic conditions may be encountered when assessing Hoffa's fat pad including tumors and tumor-like lesions such as osteochondroma, tenosynovial giant cell tumor (TGCT) (and pigmented nodular synovitis) and arthrofibrosis, traumatic changes including contusions and anatomic variants such as recesses. The latter may be accountable for differences in crosssectional area or volume changes over time. Signal changes are commonly used in OA research as surrogate markers for synovitis but are non-specific findings. Conclusions: Quantitative approaches to evaluate 3D parameters of Hoffa's fat pad are increasingly applied and their role in regard to structural progression and clinical manifestations of disease needs to be further elucidated. In applying such approaches, knowledge of the detailed anatomy and potential pitfalls that may be a result of anatomical variants, inflammatory disease manifestations and additional diverse pathologies encountered seems to be paramount. © 2015 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.

Keywords: Hoffa's fat pad Infrapatellar fat pad MRI Osteoarthritis Inflammation

Introduction Hoffa's fat pad, also known as the infrapatellar fat pad (IPFP), which are terms that will be used synonymously in this review article, is an intracapsular extra-synovial structure in the anterior

* Address correspondence and reprint requests to: F. W. Roemer, Department of Radiology, University of Erlangen-Nuremberg, Maximiliansplatz 1, 91054 Erlangen, Germany. Tel: 49-9131-8545620; Fax: 49-9131-8536068. E-mail addresses: [email protected], [email protected] (F.W. Roemer).

knee joint. Structurally it is composed of adipose tissue similar to subcutaneous fat1. It is thought to improve the distribution of the lubricant effect of intra-articular joint fluid by increasing the synovial surface2, and to reduce the impact of loading by absorbing forces generated in the knee joint3. Its preservation even under extreme starvation conditions e where subcutaneous fat is eliminated e suggests a central role in knee joint homeostasis4. Hoffa's fat pad has long been seen as a bystander in knee osteoarthritis (OA) with only marginal metabolic relevance. However, Hoffa's fat pad is a major source of cytokines such as interleukin (IL)-6, tumor necrosis factor (TNF)-a and adipokines such as leptin and the recent

http://dx.doi.org/10.1016/j.joca.2015.09.018 1063-4584/© 2015 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

2

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

recognition of the potential importance of these factors in joint metabolism has lead to increased appreciation for the clinical significance of the fat pad. Further, the correlation of the quantity of leptin and other adipokines produced with the quantity of fat mass has provided a rationale for measuring fat pad size1,4,5. Synovial inflammation seems to be present in a majority of persons with knee OA at least some time during the course of the disease and Hoffa's fat pad appears to play a central role in the mediation of joint inflammation6,7. Magnetic resonance imaging (MRI) has become a recognized tool in OA research including assessment of inflammatory disease manifestations, which is commonly performed on either nonenhanced MRI using surrogate signal changes in Hoffa's fat pad or on contrast-enhanced scans by direct assessment of the synovium8e14. Other MRI applications are also being increasingly applied in the field of OA research such as dynamic contrast enhanced (DCE) MRI, which allows the evaluation of inflammatory synovial activity by assessing the degree of perfusion of the synovial tissue12. MR-detected synovitis has been found in all stages of disease including early ones15,16. It also has been associated with pain, severity of disease17 and shown to be predictive of structural changes18,19. Although CE-MRI is the current standard for assessment of synovitis in rheumatoid arthritis, synovitis in large epidemiological OA studies has usually been assessed on non-CEMRI. In a pathologic study by Fernandez-Madrid and coworkers20, hypointense signal alterations in Hoffa's fat pad seen on non-CE sagittal T1-weighted spin-echo images correlated with mild chronic synovitis. This work led to the assumption that synovitis might be assessed on non-CE-MRI, and hyperintense signal changes seen in Hoffa's fat pad on fat suppressed (FS) proton density (PD)or T2-weighted spin-echo-type sequences were suggested as a surrogate marker for whole knee synovitis21,22. In addition to synovial assessment of Hoffa's fat pad, most recently the volume of the IPFP has become a subject of interest. Beneficial associations have been reported between IFPP maximal area and radiographic

OA, structural changes and pain23, and a protective effect of the IPFP in regard to structural progression, as assessed by tibial cartilage volume change, was evidenced in older women3. As the IPFP may have protective role through shock absorption potentially resection of the IPFP may need to be more carefully discussed in a clinical setting23. Whether IPFP quantification offers promise in the clinical assessment of a patient with knee OA remains to be shown; presently these measurements are performed in a research context only. Despite the increasing interest in Hoffa's fat pad within the field of knee OA research, the diversity of normal anatomic variants and the wide spectrum of pathologic findings commonly encountered in routine clinical practice need to be considered when evaluating this important joint structure and making it a focus of research interest. This narrative review aims to give an illustrative overview to clinicians and also to the interdisciplinary research community with a radiologic emphasis on the anatomy, on technical considerations, imaging pathologies encountered and their differential diagnosis in the context of OA imaging research. Methods and design To extract relevant studies, an extensive PubMed database search including only English literature and covering the period from January 1st 1995 to January 1st 2015 was performed. The search was based on but no limited to the query terms “Hoffa”, “Hoffa's fat pad” or “IPFP” in combination with “synovitis”, “OA”, and “MRI”. The literature search yielded 289 publications that were screened for relevance with focus on the interdisciplinary audience of Osteoarthritis Cartilage and clinical application. Included were original research articles and review articles. Excluded were case reports and publications other than in English. Additional references were included when these were considered of importance. The literature was sorted and will be presented according to the topics technical considerations, normal anatomy and abnormal

Fig. 1. Anatomy of Hoffa's fat pad. A. Sagittal T1-weighted spin echo MRI. The IPFP is an intracapsular but extra-synovial structure (colored in green-transparent) limited by the patellar tendon anteriorly (white arrow), and the synovial-lined knee joint posteriorly. Superiorly it attaches to the inferior surface of the patella (black arrow), and projects posteriorly to the intercondylar notch via two alar synovial folds, which fuse forming the infrapatellar plica. Supero-posteriorly Hoffa's fat pad neighbors the cartilage of the femoral trochlea (black arrowheads). B. Identical image without overlay shows low signal foci seen within the fat pad that represent infrapatellar plica (white arrows) and vessels (small black arrows).

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

conditions of the IPFP as well as the potential role of Hoffa fat pad changes in the pathogenesis of knee OA. Results Anatomy and physiology The IPFP is an intracapsular but extra-synovial structure limited by the patellar tendon anteriorly, and the synovial-lined knee joint posteriorly. Fig. 1 gives an outline of the anatomical relation of the IPFP and its neighboring structures such as the patella, the intercondylar notch, the periosteum of the tibia, the intermeniscal ligament and the anterior horns of the menisci, abutting the deep infrapatellar bursa2,24. The latter has the form of an inverted triangle, with the apex situated inferiorly, limited by the distal patellar tendon anteriorly, the tibial margin posteriorly, and the Hoffa fat pad superiorly (Fig. 2)25. A fat apron extending down from the IPFP partially divides the bursa into anterior and posterior compartments, separate from the knee joint26. The vascular supply of Hoffa's fat pad derives from the superior and inferior geniculate arteries, with a central area left paucivascular. The IPFP is richly innervated, receiving branches from the femoral, common peroneal and saphenous nerves. It also contains a high density of type IVa afferent nerve endings, which explains the significant levels of pain that are perceived when the adipose structure is affected by pathologic processes27. Hoffa's fat pad is the most sensitive tissue within knee joint eliciting severe pain on endoscopic palpation and nociceptive fibers have been found in the adipose tissue among patients with anterior knee pain28,29. Hoffa's fat pad is similar in appearance to subcutaneous tissue on MRI as it contains fibrous structures dispersed through adipocytes30. Histologically, other authors found IPFP to resemble more visceral rather than subcutaneous fat4. From a physiological standpoint several studies have linked levels of synovial and inflammatory biomarkers with severity and progression of OA7,17,31e35. The IPFP was shown to be a source of cytokines, adipokines, and lipid mediators in knees with OA36e38. It is also a site of inflammatory and mesenchymal stem cells37,39. The role of Hoffa fat pad does not seem to be limited to liberation of chemical

Fig. 2. Deep infrapatellar bursa. A deep infrapatellar bursa is present in 20e70% of individuals and does not exhibit communication with the knee joint. Sagittal intermediate-weighted fat-suppressed MRI shows a non-inflamed deep infrapatellar bursa (arrows), which has the form of an inverted triangle with the apex situated inferiorly, limited by the distal patellar tendon anteriorly, the tibial margin posteriorly, and the Hoffa fat pad superiorly. Deep infrapatellar bursitis commonly results from overuse of the knee extensor mechanism.

3

mediators but might also be mechanical considering its anatomy. A recent study focused on sex-differences of the healthy IPFP and showed that men displayed significantly greater ratio of IPFP volume/body weight than women, a finding unexplained to date40. Due to its close contact to cartilage and bone surface, the IPFP may contribute to joint homeostasis by reducing the impact of loading and absorbing forces generated in the knee joint3. Technical considerations Hoffa's fat pad as an adipose structure exhibits high signal intensity on all non fat-suppressed sequences. However, to enable optimal assessment of Hoffa-synovitis, i.e., thickening and inflammation of the synovial tissue abutting the fat pad, several parameters have to be considered as the detection of synovitis not only depends on the MRI pulse sequence and possible use of a gadolinium-based contrast agent, but also on coil type, field strength and voxel size. Both contrast-enhanced MRI (CEMRI) and non-contrast enhanced MRI (NEMRI) have been used for the evaluation of synovitis in or surrounding Hoffa's fat pad8. Without contrast administration it is not possible to discriminate between intraarticular synovial joint fluid and synovial thickening as both appear as fluid-equivalent hyperintensity on fluid-sensitive sequences41. Although one study found associations between semiquantitative (SQ) assessment of knee synovitis on CEMRI and NEMRI42, there is increasing evidence establishing the superiority of CEMRI for the purpose of synovitis evaluation8,41. In fact synovitis displayed on CEMRI was found to superiorly correlate with pain and histologic evidence of inflammation including synovial infiltrates and hyperplasia in early and late disease, compared to NEMRI43,44. Nonetheless NEMRI-based scoring systems remain much more widely used and are a valid alternative when CEMRI is not available, provided that investigators are aware of their limitations8. Considering the anatomy of the IPFP, sagittal and axial images are used to assess Hoffa's fat pad and inflammatory disease manifestations in adjacent locations including the suprapatellar and infrapatellar recesses10,11,14,45. Field strength and coil selection

Fig. 3. Intracapsular (osteo-) chondroma. Intracapsular and paraarticular chondromas are rare benign tumors that result from extra-synovial metaplasia in the capsule or adjacent connective tissues. Other names for this entity include extraskeletal ossifying chondroma, intracapsular osteochondroma, and soft-tissue osteochondroma. Intracapsular osteochondroma as shown in this sagittal PD fat-suppressed MRI is a rare form of chondroma and is characteristically located inferior to the patella. Note hypointense peripheral rim of the lesion representing calcification (arrows) and cartilage-equivalent high signal in the center. In addition there is a large traumatic joint effusion (asterisk).

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

4

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

Fig. 4. L-TGCT is a benign proliferative disorder that has also been referred to as giant cell tumor of the tendon sheath and localized PVNS. The extra-articular form is most commonly associated with the tendon sheaths of the hands and the intra-articular form is seen almost exclusively in Hoffa's fat pad. Evidence of hemosiderin deposition is highly variable, but a helpful finding. A. Sagittal T2-weighted MRI shows a well-demarcated infrapatellar nodular lesion of heterogeneous hypointense signal (arrows). B. Corresponding T1-weighted MRI shows the mass as homogeneously hypointense (arrows). Intralesional hypointensity on both sequences is highly suggestive of intralesional hemosiderin deposits that are typical for L-TGCT.

influence synovitis assessment and need to be considered in the context of Hoffa evaluation. As a relatively superficial structure dedicated surface coils may seem promising in the evaluation of Hoffa pathology46. DCE-MRI has been used to evaluate tissue perfusion and has mainly been applied in oncology. In musculoskeletal research, DCEMRI has been used as an alternative method for the quantification of synovitis in order to assess therapeutic response and disease progression in inflammatory diseases such as rheumatoid arthritis47,48. DCE-MRI is based on a sequential acquisition of rapid T1-weighted MR sequences before and during the intravenous

administration of a gadolinium-based contrast agent. A signal intensity graph over time is subsequently generated for any MR voxel. Various perfusion parameters characterizing the signal intensity graphs can be extracted, such as the initial rate of enhancement, the slope, and maximum enhancement12. A recent study has shown that perfusion variables observed using DCE-MRI reflecting the severity of inflammation in the IPFP were associated with pain severity in knee OA12. Further, DCE-MRI seems to be sensitive at detecting post-therapy synovial changes after intra-articular corticosteroid injection that are associated with symptomatic response31.

Fig. 5. Synovial hemangioma is a rare entity that is commonly seen at the anterior compartment of the knee. MRI findings display a non-specific infiltration of the fat pad. A. Sagittal T2-weighted MRI shows a mass-like hyperintense lesion occupying a large part of the IPFP (arrows). B. Corresponding axial intermediate-weighted fat-suppressed MRI proves that the lesion is not only containing lipomatous tissue but also vascular structures that appear hyperintense. Lesions is expanding beyond the fat pad, which is best seen on the axial image. Signal voids may be observed that may mimic hemosiderin deposits of localized PVNS and calcifications of paraarticular chondroma. MRI is essential for the diagnosis as hemangioma arising deep from the IPFP may not be visible arthroscopically.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

Fig. 6. Anterior knee pain in a patient with a long standing history of gout. Sagittal intermediate-weighted fat-suppressed MRI shows intratendinous hyperintensity and thickening of the patellar tendon (large arrow) and diffuse pre-patellar infiltration (small arrows) consistent with soft tissue gouty tophus. In addition, there is concomitant deep infrapatellar bursitis (asterisk) and minimal suprapatellar effusion.

MRI findings Benign tumors and tumor-like lesions Benign tumors and tumor-like conditions affecting Hoffa fat pad are not rare and are usually classified into intrinsic, arising primarily from the fat pad, and extrinsic, resulting from secondary involvement of the fat pad. Researchers need to be aware of these findings as these may or may not be clinically relevant but potentially will impair image analysis. - Paraarticular chondromas arise from the capsule and/or connective tissue adjacent to the capsule due to cartilage metaplasia. Once ossification has occurred the term paraarticular osteochondroma is used49. MR examination displays a central or

5

eccentric lesion within the IPFP, which is round or lobulated (Fig. 3). The lesion is predominantly isointense to muscle on T1weighted images with areas of increased intensity due to marrow fat when ossification is present. On fluid-sensitive sequences such as T2-weighted and STIR images the mass is heterogeneously hyperintense. Gadolinium-enhanced T1-weighted images display peripheral enhancement50. The differential diagnosis of a paraarticular chondroma includes a mineralizing soft tissue sarcoma such as synovial sarcoma. - Synovial chondromatosis is a non-neoplastic, proliferative and metaplastic disorder of synovium. MR findings consist of lobulated homogeneous masses located not only within the Hoffa fat pad, but also in other locations including the suprapatellar pouch and the popliteal fossa. Foci of hypointense signal corresponding to calcification or foci of isointense signal corresponding to cartilage-like tissue may be noted. In case of ossification, peripheral low signal surrounding fat-like signal is noted. The presence of calcification and ossification depends on the degree of maturation of the synovial proliferation50. - Tenosynovial giant cell tumor (TGCT) may occur within the knee as the diffuse type (also known as pigmented villonodular synovitis or PVNS) or as the localized type (also known as giant cell tumor of the tendon sheath and localized PVNS)51e54. This family of synovial proliferative disorders is now thought to be neoplastic in etiology. PVNS was first described as an entity by Jaffe et al., in 194152 and then reclassified in 2002 and 2013 by the World Health Organization according to their pathologic origin53,54. On MR imaging, the localized type of intra-articular tenosynovial giant cell tumor (L-TGCT) typically presents as a midline mass within the fat pad, often abutting, but not invading the patellar tendon (Fig. 4)50,51. The signal intensity of the mass is variable, however areas of low signal intensity on all pulse sequences is suggestive of hemosiderin deposition and is a typical feature. Use of a gradient-echo pulse sequence allows for confirmation of the presence of hemosiderin, which appears as a prominent “blooming” of low signal intensity due to magnetic susceptibility artifact. When diagnosing the localized type it is important to scrutinize the knee for other evidence of intraarticular involvement to rule out the diffuse type of intra-

Fig. 7. Hoffa's ganglion. While meniscal cyst show communication with a meniscal tear, Hoffa's ganglion cysts commonly do not exhibit a communication with the articular surface. Ganglia are non-synovial lined, gelatinous fluid-filled lesions. It has been hypothesized that ganglia of the IPFP may result from degeneration of the transverse ligament, which connects the anterior horns of the lateral and medial menisci. On MRI, ganglia appear as well-defined, uni- or multi-loculated, fluid-like T2 hyperintense lesions. A. Sagittal PDweighted fat-suppressed MRI shows a multi-lobulated Hoffa ganglion adjacent and surrounding the transverse ligament (arrowhead). B. Another example emphasizes the importance of fat suppression on fluid-sensitive sequences to evaluate Hoffa's fat pad abnormalities. Sagittal PD-weighted fat-suppressed MRI depicts a large multi-lobulated Hoffa ganglion (arrows) and adjacent edema along the inferior aspect of the fat pad (arrowhead). C. Corresponding non-FS PD-weighted image inferiorly visualizes ganglion cyst with identical signal intensity than surrounding fat (arrows). Note that Hoffa's edema can not be distinguished from fat signal.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

6

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

Cysts, ganglia, recesses, and bursa

Fig. 8. A parameniscal cyst is a focal joint fluid collection adjacent to the meniscus resulting from extrusion of synovial fluid through horizontal or complex meniscal tears through a one-way valve effect of the tear flap. Hoffa's parameniscal cysts result from anterior horn tears that may exhibit a communication with the cyst may not be readily detected on MRI. Sagittal intermediate-weighted fat-suppressed MRI shows a horizontal tear of the lateral anterior horn (small arrows), which communicates with a parameniscal cyst (large arrows). While parameniscal cysts (resulting from a meniscal tear) may require partial meniscectomy, ganglia may require an extra-articular rather than arthroscopic approach for decompression.

articular tenosynovial giant cell tumor (D-TGCT or D-PVNS). LTGCT is not associated with bone erosion50. - Synovial hemangioma is a rare tumor that commonly involves the anterior compartment of the knee, and arises primarily from the Hoffa fat pad in 7% of the cases55. MR findings display a nonspecific infiltration of the fat pad (Fig. 5). While presence of a fatty component is usually helpful in diagnosing hemangioma in general, this is less readily appreciable for hemangioma of the IPFP, which arises from a primarily fatty structure. The lesion is hyperintense on T2-weighted fat-suppressed images with signal voids corresponding to abnormal vessels55,56. Other benign mass-like lesions involving the IPFP have been reported but are not very common such as fibroma of tendon sheath, glomus tumor, gout (Fig. 6), tumor calcinosis related to chronic renal disease50.

- Ganglia are common non-synovial lined, gelatinous fluid-filled lesions rich in hyaluronic acid and other mucopolysaccharides57. Hoffa's ganglia are most commonly in contact with the anterior horn of lateral meniscus58. Because of this predilection, Saddik et al. hypothesized ganglia of the IPFP may result from degeneration of the transverse ligament, which connects the anterior horns of the lateral and medial menisus2. Other authors pointed to compressive injury of the meniscal periphery with resultant degenerative changes extending into surrounding soft tissues rather than within meniscus59,60 On MR imaging, ganglia appear as well-defined, uni- or multi-loculated, fluid-like T2 hyperintense lesions (Fig. 7). Depending on their protein content ganglia may be hypo- or isointense on T1-weighted sequences57. As opposed to horizontal clefts, which display a large posterior defect in continuity with the intra-articular space, ganglia commonly do not exhibit a communication with the knee joint. - A parameniscal cyst is a focal joint fluid collection adjacent to the meniscus resulting from extrusion of synovial fluid through horizontal or complex meniscal tears through a one-way valve effect of the tear flap61e63. Hoffa's parameniscal cysts result from anterior horn tears that may exhibit a communication with the cyst may not be readily detected on MRI (Fig. 8). Pathologically both meniscal cysts and ganglia contain fluid with high protein content that is similar to synovial fluid64. Therefore MR characteristics are similar for both entities, i.e., hyperintense on T2 and variable signal intensity on T1-weighted images. The confirmation of a parameniscal cyst is based on the depiction of a communication with an adjacent meniscal tear on any pulse sequence. Despite the apparent similarity between ganglia and parameniscal cysts pathologically and radiologically, the distinction between both entities is important, as the treatment is different. While parameniscal cysts (resulting from a meniscal tear) may require partial meniscectomy65, ganglia may require an extra-articular rather than arthroscopic approach for decompression66. - Prevalence of MR-detected deep infrapatellar bursa varies between 19% and 68%67,68. Cadaveric studies have shown no communication of the deep infrapatellar bursa and the knee joint25,26. Aydingoz et al. noted no correlation between the

Fig. 9. Septic bursitis. A. Sagittal PD-weighted fat-suppressed MRI shows diffuse hyperintense signal within the deep infrapatellar bursa, which appears enlarged and fluid-filled (large arrows). In addition, there is marked fluid-equivalent signal within the pre-patellar bursa (small arrows). B. Corresponding contrast-enhanced T1-weighted fat-suppressed MRI shows marked enhancement surrounding the bursa (large and small arrows and confirming bursitis). Fluid appears as hypointensity at the center of lesion (asterisks). Findings are consistent with post-surgical septic bursitis, a diagnosis that was confirmed by aspiration.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

7

Fig. 10. Indentations or clefts along the posterior margin of the IPFP are commonly seen. A. The first type is located superiorly, close to the inferior border of the patella, and oriented vertically. Sagittal intermediate-weighted fat-suppressed MRI depicts superior cleft, which is in communication with articular cavity (arrows). B. The second cleft is located anterior to the anterior horns of the menisci and has a horizontal direction and is also called infra-hoffatic recess. Sagittal intermediate-weighted fat-suppressed MRI shows infra-hoffatic recess with fluid-equivalent signal (small arrows). Note in addition there is a small parameniscal cyst (large arrow). Hoffa's recesses are highly prevalent on routine MRI. The size and shape of the inferior posterior recess are considerably variable, and may be confused with ganglion cysts especially when globular or ovoid. More commonly though, the infrahoffatic recess is linear and represents a common finding on routine MRI of the knee.

Fig. 11. Visualization and characteristics of these recesses may depend on the concomitant amount of joint effusion and synovitis and may be a reason for observed volume changes or fluctuation in CSA measurements, which needs to be taken into account when analyzing cross-sectional or longitudinal associations with other parameters. A. Sagittal intermediate-weighted fat-suppressed MRI shows a normal Hoffa's fat pad without evident signs of fluid-filled recesses or cyst-like lesions. B. 12 months later, the same knee exhibits some joint effusion and now fluid-filled recesses are well seen. C. The transparent overlay demarcates the maximum CSA of the IPFP at baseline (purple). D. The follow-up image clearly shows a decrease in CSA due to the fluid-filled recesses (CSA depicted as transparent green).

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

8

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

Fig. 12. Another example of injury to Hoffa's fat pad. Sagittal intermediate-weighted fat-suppressed MRI shows a traumatic bone marrow lesion or contusion to the inferior pole of the patella (large arrow). No fracture line is observed. In addition there is a traumatic signal change representing edema and hemorrhage at the superior pole of Hoffa's fat pad (small arrows). Example shows incidental finding of an infrapatellar plica or ligamentum mucosum (arrowheads), which originates from the intercondylar notch, inserts into and then courses within Hoffa's fat pad towards the inferior patellar pole. Posteriorly, the plica may be separated from the anterior cruciate ligament as in this example, or completely or partially attached to it. There is associated traumatic mild to moderate joint effusion.

degree of knee effusion and the detection of bursa, which is consistent with the absence of communication between the two spaces67. Awareness of location and dimension, and sometimes correlation with clinical findings are important to distinguish normal from diseased bursa (Fig. 2)57. Deep infrapatellar bursitis results from overuse of the knee extensor mechanism, especially in runners and jumpers69, and can also be caused by infection, gout and ankylosing spondylitis (Fig. 9). Furthermore, deep infrapatellar bursitis is common in OsgoodeSchlatter disease, both during active disease and after resolution of symptoms25. Because of the absence of communication between the bursa

and the joint, the presence of deep infrapatellar bursitis is not related to knee effusion. - On MR examination of the knee, two indentations or clefts along the posterior margin of the IPFP are commonly described. One located superiorly, close to the inferior border of the patella, and oriented vertically. The second cleft is anterior to the anterior horns of the menisci and has a horizontal direction also called infra-hoffatic recess70. The infra-hoffatic recess is often noted around the intermeniscal ligament lying anterior, posterior or inferior to it71. The latter recess is also reported to be in close contact with the infrapatellar plica (ligamentum mucosum, or infrapatellar synovial fold), which was reported to form the roof of the infra-hoffatic recess72. These recesses are highly prevalent on routine MR imaging (Fig. 10). The infra-hoffatic recess is present in 45e90% of MR examinations, while the superior cleft is present in up to 71% of the cases71,72. The size and shape of the inferior posterior recess are considerably variable, and may be confused with ganglion cysts especially when globular or ovoid. More commonly though, the infra-hoffatic recess is linear and represents a common finding on routine MR imaging of the knee71. Depiction and characteristics of these recesses may depend on the concomitant amount of joint effusion and synovitis and may be a reason for observed volume changes or fluctuation in cross-sectional area (CSA) measurements (Fig. 11). A globular-shaped inferior recess is found in about 13% of MR examinations30. Cadaveric studies showed the inferior recess to be located just inferior to the insertion of the infrapatellar plica30,71. These recesses were shown to be highly symmetrical in paired knees27, however no strong correlation between their presence, size or shape and pathologic conditions such as anterior cruciate ligament (ACL) tear has been evidenced71,72. Nonetheless, knowledge of such recesses is important for proper identification and distinguishing pathology from normal variants.

Traumatic and posttraumatic findings Acute Hoffa's fat pad injuries follow a variety of different mechanisms, and can be associated with injuries to the patellar tendon and anterior cruciate ligament with anterior displacement of the stump. Abreu et al. showed in a cadaveric study that edema,

Fig. 13. Edema at the superolateral aspect of the Hoffa fat pad is a common finding in routine MRI examination of the knee and is thought to be a consequence an impingement between the lateral femoral condyle and patellar ligament from patellar maltracking and may be associated with patellofemoral OA. A. Sagittal intermediate-weighted fat-suppressed MRI shows hyperintense signal at the superior lateral aspect of Hoffa's fat pad representing edema secondary to patellar maltracking syndrome. A high riding patella is the most consistently correlated feature of patellar maltracking with superolateral fat pad impingement. While this example also shows marked cartilage loss at the lateral patella superolateral edema may also be observed in young patients. B. Sagittal intermediate-weighted fat-suppressed MRI shows hyperintense signal at the superolateral aspect of Hoffa's fat pad (arrow), which is commonly associated with pain. Note that edema may also be observed in patients without impingement symptoms.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

Fig. 14. Non-enhanced vs CEMRI for visualization of synovitis. Joint effusion cannot be distinguished from synovial thickening using non-enhanced MRI since both are visualized as hyperintense signal. For this reason, signal changes in Hoffa's fat pad are commonly used as a surrogate marker for synovitis. However, because Hoffa's fat pad is an extra-synovial structure these signal changes are non-specific and represent a heterogeneous mixture of edema, vessels, synovitis (especially at the posterior aspect of the IPFP) and other changes such as clefts. A. Image example (sagittal intermediate-weighted fat-suppressed MRI) shows diffuse intra-hoffatic high signal (arrows). Note that joint effusion cannot be distinguished from synovitis (asterisk). B. The corresponding sagittal contrast-enhanced T1-weighted fat-suppressed MRI clearly shows that only the posterior surface of the fat pad corresponds to synovial thickening and enhancement (white arrows) while the intra-hoffatic signal alterations represent non-enhancing edema in this case. Note that true joint effusion is visualized as hypointensity within the joint cavity (asterisk). In addition there is marked synovitis posterior to the posterior cruciate ligament, the most commonly affected anatomical site of synovitis in OA (black arrows).

Fig. 15. SQ assessment of Hoffa signal changes. Several SQ scoring systems assess signal changes in Hoffa's fat pad as a surrogate marker for synovitis. Commonly these are graded from 0 to 3 according to the volume occupied by the hyperintense signal within the fat pad. A. Sagittal intermediate-weighted fat-suppressed MRI shows minor non-specific signal change centrally in Hoffa's fat pad representing grade 1 synovitis using the MRI Osteoarthritis Knee Score (MOAKS14) scoring system (arrow). B. Grade 2 synovitis is shown in this image (arrows). C. Large portions of Hoffa's fat pad appear hyperintense on this sagittal fat-suppressed MRI (arrows), representing a grade 3 lesion in MOAKS.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

10

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

tears, scars and synovial proliferation are more common in knees with torn ACL than knees with intact ACL73. In addition, changes in the fat pad can be noted as an indirect sign of previous patellar dislocation. In this case MR findings consist of T2 hyperintense signal with cleft formation and fragmentation, typically at the superomedial border of the Hoffa fat pad. Fragmented fat may not easily distinguished from loose bodies, which may require X-ray or computed tomography74. Traumatic injury to the infrapatellar plica can also occur and results in signal changes within the fat pad following the course of ligamentum mucosum, i.e., parallel to ACL toward anterior meniscal horn then through Hoffa's fat pad toward the inferior pole of the patella (Fig. 12). In addition to presence of intra-hoffatic edema, MR imaging displays associated traumatic findings such as ACL tear, chondral traumatic injuries, and bone marrow contusions. History of prior sports related trauma is often present in infrapatellar plica injuries75.

Overuse and impingement syndromes - Patellar tendinopathy results from repetitive microtrauma and is usually encountered in athletes engaging in activities requiring quadriceps muscle contraction. The patellar tendon is increased in diameter and demonstrates high signal intensity on FS fluid-sensitive images such as STIR and FS T2-weighted sequence76. Because of the close contact between patellar tendon and Hoffa's fat pad, the injured patellar tendon is often associated with hoffatic juxtapatellar traumatic edema that should be differentiated from synovitis and other pathologic conditions of Hoffa's fat pad. - Inflammatory changes secondary to impingement of Hoffa fat pad results from chronic trauma of the adipose tissue and may take two main forms, Hoffa's and superolateral impingement of the fat pad.

Fig. 16. Post-surgical changes. Fibrous scarring after surgery can involve the IPFP. A. After arthroscopy, fibrotic stranding is accentuated at the portal site and is typically linear and horizontal. Sagittal T2-weighted fat-suppressed MRI shows hypointense linear band of post-arthroscopic scar formation within the fat pad (arrows). B. Corresponding non-fatsuppressed PD-weighted MRI shows scar as hypointense linear signal change (black arrows). C. Open surgery may produce a more extensive fibrotic reaction within the fat pad, and may cause apparent thickening of the patellar tendon, giving the appearance of chronic tendinosis or jumper's knee. Sagittal T1-weighted image pre-surgery shows hypointense joint effusion and traumatic synovitis at the posterior surface of Hoffa's fat pad but normal signal within the fat pad itself. D. 12 months post surgery there is marked hypointense scarring at the surface (small arrows) and also intra-hoffatic signal changes consistent with fibrosis (large arrow).

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

Hoffa's disease is a syndrome of IPFP impingement thought to be the result of acute or repetitive trauma of the knee causing hemorrhage77. Resulting hypertrophy of the inflamed adipose tissue then predisposes it to crushing and impingement between the femur and tibia and therefore to further injuries. Intra-hoffatic areas with increased signal intensity on T2-weighted MR images suggesting edema and hemorrhage are characteristically associated with bowing of the patellar tendon from mass effect. Low signal intensities on both T1- and T2-weighted images may be seen in chronic stages corresponding to fibrosis and hemosiderin deposits24. In more advanced stages fibro-cartilaginous metaplasia may occur rarely leading to ossification78. Infrapatellar impingement is also commonly associated with deep infrapatellar bursitis79. Edema at the superolateral aspect of the Hoffa fat pad is a common finding in routine MR examination of the knee and is thought to be a consequence of Hoffa's fat pad impingement from patellar maltracking (Fig. 13). Although the clinical syndrome has been known for a long time, superolateral fat pad impingement as termed by radiologists, first came into the radiological literature at the beginning of the last decade80. Several publications have since confirmed the relationship between the superior lateral MR abnormalities and patellar maltracking and instability, although abnormal patellofemoral relations are not always found and some patients with this finding were asymptomatic81e86. High riding patella is the most consistently correlated feature of patellar maltracking with superolateral fat pad impingement82e86. This is explained by the reduction of contact area at the patellofemoral joint in this case resulting from long patellar tendon87. Increased tibial tuberosity-trochlear groove distance was also found to be associated with superolateral fat pad impingement83,86. As a result superolateral fat pad impingement is being increasingly recognized as a marker of patellofemoral maltracking. In a recent study, Subhawong et al. showed an association between medial patellar chondromalacia and superolateral fat pad impingement linking this condition to patellofemoral OA88. The infrapatellar plica (also referred to as ligamentum mucosum), thought to be the vestigial remnant of the embryological vertical septum, originates from the intercondylar notch, inserts into and then courses within Hoffa's fat pad towards the inferior patellar pole. Posteriorly the plica may be separated from the anterior cruciate ligament, or completely or partially attached to it89. On MR imaging the plica has a curvilinear appearance as it courses in an anteroposterior orientation. Although presence or absence of a normal infrapatellar plica has no clinical significance, an injured or degenerated infrapatellar plica commonly present as Hoffa fat pad signal abnormality on routine MR examination of the knee (Fig. 12)75. Infrapatellar plica syndrome is often difficult to be differentiated from Hoffas disease both clinically or by imaging criteria. The relationship between both conditions is still unclear2. Hoffa's and peripatellar synovitis As previously mentioned the supra- and infra-hoffatic clefts are synovial-lined. Therefore synovial pathology and thickening may manifest as Hoffa signal changes, or as an extension for the articular synovial pathology. Because joint effusion has the same intensity as synovial thickening on fluid-sensitive images, it is often difficult to distinguish both conditions (Fig. 14). As detailed above CEMRI best correlates with synovitis histologically43. The presence of Hoffa signal changes on NEMRI is commonly considered as a surrogate for synovial proliferation but is a non-specific finding (Fig. 15)8,41. Postoperative changes Postoperative changes affecting Hoffa's fat pad include postsurgical (most often arthroscopic) fibrosis and the “cyclops”

11

lesion commonly seen after ACL reconstruction2. While fibrosis after open surgery usually is ill defined and confluent, postarthroscopic fibrosis commonly presents as subtle bands coursing horizontally through the IPFP (Fig. 16). Post-arthroscopic scarring or fibrotic changes involving the extensor retinaculum and deep Hoffa's fat pad are common and usually thought to be benign. However, arthrofibrosis along the inner margin of Hoffa's fat pad may cause fibrotic contracture and tethering of the patella, limiting knee flexion, which may require anterior interval release surgery. Fibrotic changes affect the medial or lateral patellar retinaculum and deep Hoffa fat pad, depending on the portal used to access the knee90. Maximal fibrous scarring is seen within 6 months of surgery and then reduces markedly or disappears in about half of the patients by 18 months91. In addition to fibrosis metallic debris can also be identified, which are particularly obvious on gradient echo images due to susceptibility artifacts. A cyclops lesion is a nodular soft tissue mass extending through the intercondylar region and the apex of the IPFP (Fig. 17)2. The term anterior metaplasia has also been suggested as its was shown to contain fibrous tissue, fibrocartilage, bone, synovium and fat92. Nonetheless the exact etiology remains unknown. Due to its fibrous nature a cyclops lesion may appear hypointense on T2-weighted images. A cyclops lesion causes loss of full knee extension related to mass effect (classically with development of an audible and palpable “clunk” in terminal extension), which improves after surgical removal. Recognition of these lesions in MRI is important as these patients are poorly responsive to physical therapy and surgical treatment is often needed2.

Fig. 17. Cyclops lesion, also known as localized anterior arthrofibrosis, is usually found as a complication of ACL repair and represents a focal nodular fibrosis in the intercondylar notch, seen anterior to the ACL insertion and presenting with pain and limited extension. It commonly shows low signal on T1-weighted images (indistinguishable from effusion) and inhomogeneous moderate-to-low signal on T2-weighted images (contrasting with surrounding joint fluid). This lesion is amenable to arthroscopic resection, which is followed by immediate relief of symptoms. Sagittal PDweighted fat-suppressed MRI example shows large Cyclops lesion in typical location (small arrows). Note intact hypointense ACL graft (large arrow) and inhomogeneous fat suppression at the femur due to metal hardware (asterisks).

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

12

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

Table I Different pathologies affecting the infrapatellar (Hoffa's) fat pad and corresponding MRI findings Lesion Benign tumors and tumor-like lesions Para-articular chondroma Synovial chondromatosis Localized-type tenosynovial giant cell tumor (also known as Pigmented Villonodular Synovitis e PVNS) Synovial Hemangioma Cysts, ganglia, recesses, and bursa Ganglia Para-meniscal cyst Deep infrapatellar bursa Infrapatellar clefts Traumatic/posttraumatic changes Acute Hoffa's fat pad injury Traumatic injury to infrapatellar plica

Overuse/impingement syndromes Patellar tendinopathy Hoffa's fat pad impingement Superolateral impingement Infrapatellar plica syndrome Peripatellar synovitis Postoperative changes

MRI findings Central eccentric lesion, round or lobulated, isointense to muscle on T1. Peripheral enhancement after Gadolinium administration Lobulated homogeneous mass, hypointensity corresponding to calcifications, ossifications Midline mass abutting patellar tendon, blooming artifacts from hemosiderin deposition

Non-specific, signal voids, calcifications Well defined, fluid-equivalent signal Adjacent to meniscus, meniscal tear Deep, in contact with tibial cortex Vertical: close to inferior border of patella. Horizontal: anterior to anterior horn of menisci T2 hyperintensity þ traumatic associated findings (patellar dislocation, cleft formation, fragmentation) Curvilinear T2 hyperintensity following the course of ligamentum mucosum from intercondylar notch to inferior patellar pole þ traumatic findings (ACL tear, chondral injuries, bone marrow contusion) Patellar tendon T2 hyperintensity. Hoffatic signal changes if present are located in peripatellar region Extensive T2 hyperintensity with bowing of the patellar tendon Superolateral T2 hyperintensity with high riding patella, increased tibial tuberosity-trochlear groove distance Curvilinear T2 hyperintensity following the course of ligamentum mucosum from intercondylar notch to inferior patellar pole Perisynovial T2 hyperintensity, Contrast enhancement (þ) Ill-defined, confluent fibrotic bands (T1 hypointense) Cyclops lesion: nodular mass through the intercondylar region and apex of IPFP (T1 and T2 hypointense)

Summary and outlook Several anatomic variants and pathologic conditions may be encountered when assessing Hoffa's fat pad (Table I). The location of signal changes within the fat pad, its anatomical outline, its signal intensity, and the presence of associated traumatic and other pathologic features in the knee joint are all important and should be considered in the diagnostic approach. Signal intensity changes have been used for a long time as a surrogate for whole knee joint synovitis. These signal alterations are non-specific findings albeit associated with clinical manifestations of inflammation. The ideal method for evaluating synovitis is CEMRI and there seems to be a potential role in DCE-MRI. As detailed in this review several pathologic conditions need to be ruled out when assessing Hoffa's fat pad based on scoring or using quantitative measurements in OA research. Most of these conditions are rare but in particular anatomic characteristics and pitfalls need to be considered in order to avoid misclassification of change over time. Quantitative approaches to evaluate 3D parameters of Hoffa's fat pad including maximal CSA and volume are increasingly applied in a research context focusing on large OA cohorts and their role in regard to structural progression and clinical manifestations of disease needs to be further elucidated. While presently not applied clinically, knowledge of the detailed anatomy and potential pitfalls that may be a result of the anatomical variants including clefts, inflammatory disease manifestations and additional diverse pathologies encountered seems to be paramount when applying such quantification approaches. Authors contributions (1) All authors were involved in the conception and design of the study, or acquisition of data, or analysis and interpretation of data.

(2) All authors contributed to drafting the article or revising it critically for important intellectual content. (3) All authors gave their final approval of the manuscript to be submitted. Responsibility for the integrity of the work as a whole, from inception to finished article, is taken by F Roemer, MD (first author; [email protected]). Competing interests Dr Guermazi has received consultancies, speaking fees, and/or honoraria from Sanofi-Aventis, Merck Serono, and TissuGene and is President and shareholder of Boston Imaging Core Lab (BICL), LLC a company providing image assessment services. Dr Roemer is Chief Medical Officer and shareholder of BICL, LLC. Dr Crema is shareholder of BICL, LLC. Dr Felson has received research grants from the NIH None of the other authors have declared any competing interests. Funding and role of the funding source No funding was received. References 1. Clockaerts S, Bastiaansen-Jenniskens YM, Runhaar J, Van Osch GJ, Van Offel JF, Verhaar JA, et al. The infrapatellar fat pad should be considered as an active osteoarthritic joint tissue: a narrative review. Osteoarthritis Cartilage 2010;18:876e82. 2. Saddik D, McNally EG, Richardson M. MRI of Hoffa's fat pad. Skeletal Radiol 2004;33:433e44. 3. Pan F, Han W, Wang X, Liu Z, Jin X, Antony B, et al. A longitudinal study of the association between infrapatellar fat pad maximal area and changes in knee symptoms and structure in older adults. Ann Rheum Dis 2014, http://

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

4.

5.

6.

7.

8.

9.

10.

11.

12.

13. 14.

15.

16.

17.

18.

19.

dx.doi.org/10.1136/annrheumdis-2013-205108. May 15. pii: annrheumdis-2013e205108 [Epub ahead of print]. Ioan-Facsinay A, Kloppenburg M. An emerging player in knee osteoarthritis: the infrapatellar fat pad. Arthritis Res Ther 2013;15:225. de Lange-Brokaar BJ, Ioan-Facsinay A, van Osch GJ, Zuurmond AM, Schoones J, Toes RE, et al. Synovial inflammation, immune cells and their cytokines in osteoarthritis: a review. Osteoarthritis Cartilage 2012;20:1484e99. Benito MJ, Veale DJ, FitzGerald O, van den Berg WB, Bresnihan B. Synovial tissue inflammation in early and late osteoarthritis. Ann Rheum Dis 2005;64:1263e7. Stannus OP, Jones G, Blizzard L, Cicuttini FM, Ding C. Associations between serum levels of inflammatory markers and change in knee pain over 5 years in older adults: a prospective cohort study. Ann Rheum Dis 2013;72:535e40. Hayashi D, Guermazi A, Kwoh CK. Clinical and translational potential of MRI evaluation in knee osteoarthritis. Curr Rheumatol Rep 2014;16:391. Hayashi D, Roemer FW, Katur A, Felson DT, Yang SO, Alomran F, et al. Imaging of synovitis in osteoarthritis: current status and outlook. Semin Arthritis Rheum 2011;41:116e30. Peterfy CG, Guermazi A, Zaim S, Tirman PF, Miaux Y, White D, et al. Whole-Organ Magnetic Resonance Imaging Score (WORMS) of the knee in osteoarthritis. Osteoarthritis Cartilage 2004;12:177e90. Hunter DJ, Lo GH, Gale D, Grainger AJ, Guermazi A, Conaghan PG. The reliability of a new scoring system for knee osteoarthritis MRI and the validity of bone marrow lesion assessment: BLOKS (Boston Leeds Osteoarthritis Knee Score). Ann Rheum Dis 2008;67:206e11. Ballegaard C, Riis RG, Bliddal H, Christensen R, Henriksen M, Bartels EM, et al. Knee pain and inflammation in the infrapatellar fat pad estimated by conventional and dynamic contrastenhanced magnetic resonance imaging in obese patients with osteoarthritis: a cross-sectional study. Osteoarthritis Cartilage 2014;22:933e40. Braun HJ, Gold GE. Diagnosis of osteoarthritis: imaging. Bone 2012;51:278e88. Hunter DJ, Guermazi A, Lo GH, Grainger AJ, Conaghan PG, Boudreau RM, et al. Evolution of semi-quantitative whole joint assessment of knee OA: MOAKS (MRI Osteoarthritis Knee Score). Osteoarthritis Cartilage 2011;19:990e1002. Baker K, Grainger A, Niu J, Clancy M, Guermazi A, Crema M, et al. Relation of synovitis to knee pain using contrastenhanced MRIs. Ann Rheum Dis 2010;69:1779e83. Loeuille D, Chary-Valckenaere I, Champigneulle J, Rat AC, Toussaint F, Pinzano-Watrin A, et al. Macroscopic and microscopic features of synovial membrane inflammation in the osteoarthritic knee: correlating magnetic resonance imaging findings with disease severity. Arthritis Rheum 2005;52: 3492e501. Attur M, Samuels J, Krasnokutsky S, Abramson SB. Targeting the synovial tissue for treating osteoarthritis (OA): where is the evidence? Best Pract Res Clin Rheumatol 2010;24:71e9. Roemer FW, Guermazi A, Felson DT, Niu J, Nevitt MC, Crema MD, et al. Presence of MRI-detected joint effusion and synovitis increases the risk of cartilage loss in knees without osteoarthritis at 30-month follow-up: the MOST study. Ann Rheum Dis 2011;70:1804e9. Atukorala I, Kwoh CK, Guermazi A, Roemer FW, Boudreau RM, Hannon MJ, et al. Synovitis in knee osteoarthritis: a precursor of disease? Ann Rheum Dis 2014, http://dx.doi.org/10.1136/ annrheumdis-2014-205894. Dec 8. pii: annrheumdis2014e205894 [Epub ahead of print].

13

20. Fernandez-Madrid F, Karvonen RL, Teitge RA, Miller PR, An T, Negendank WG. Synovial thickening detected by MR imaging in osteoarthritis of the knee confirmed by biopsy as synovitis. Magn Reson Imaging 1995;13:177e83. 21. Hill CL, Hunter DJ, Niu J, Clancy M, Guermazi A, Genant H, et al. Synovitis detected on magnetic resonance imaging and its relation to pain and cartilage loss in knee osteoarthritis. Ann Rheum Dis 2007;66:1599e603. 22. Hill CL, Gale DG, Chaisson CE, Skinner K, Kazis L, Gale ME, et al. Knee effusions, popliteal cysts, and synovial thickening: association with knee pain in osteoarthritis. J Rheumatol 2001;28:1330e7. 23. Han W, Cai S, Liu Z, Jin X, Wang X, Antony B, et al. Infrapatellar fat pad in the knee: is local fat good or bad for knee osteoarthritis? Arthritis Res Ther 2014;16:R145. 24. Jacobson JA, Lenchik L, Ruhoy MK, Schweitzer ME, Resnick D. MR imaging of the infrapatellar fat pad of Hoffa. RadioGraphics 1997;17:675e91. 25. Viegas FC, Aguiar RO, Gasparetto E, Marchiori E, Trudell DJ, Haghighi P, et al. Deep and superficial infrapatellar bursae: cadaveric investigation of regional anatomy using magnetic resonance after ultrasound-guided bursography. Skeletal Radiol 2007;36:41e6. 26. LaPrade RF. The anatomy of the deep infrapatellar bursa of the knee. Am J Sports Med 1998;26:129e32. 27. Biedert RM, Sanchis-Alfonso V. Sources of anterior knee pain. Clin Sports Med 2002;21:335e47. vii. 28. Witonski D, Wagrowska-Danielewicz M. Distribution of substance-P nerve fibers in the knee joint in patients with anterior knee pain syndrome. A preliminary report. Knee Surg Sports Traumatol Arthrosc 1999;7:177e83. 29. Dye SF, Vaupel GL, Dye CC. Conscious neurosensory mapping of the internal structures of the human knee without intraarticular anesthesia. Am J Sports Med 1998;26:773e7. 30. Vahlensieck M, Linneborn G, Schild H, Schmidt HM. Hoffa's recess: incidence, morphology and differential diagnosis of the globular-shaped cleft in the infrapatellar fat pad of the knee on MRI and cadaver dissections. Eur Radiol 2002;12:90e3. 31. Wenham CY, Balamoody S, Grainger AJ, Hensor EM, Draycott S, Hodgson R, et al. The responsiveness of novel, dynamic, contrast-enhanced magnetic resonance measures of total knee synovitis after intra-articular corticosteroid for painful osteoarthritis. Osteoarthritis Cartilage 2014;22:1614e8. 32. Sharif M, Shepstone L, Elson CJ, Dieppe PA, Kirwan JR. Increased serum C reactive protein may reflect events that precede radiographic progression in osteoarthritis of the knee. Ann Rheum Dis 2000;59:71e4. 33. Stannus O, Jones G, Cicuttini F, Parameswaran V, Quinn S, Burgess J, et al. Circulating levels of IL-6 and TNF-alpha are associated with knee radiographic osteoarthritis and knee cartilage loss in older adults. Osteoarthritis Cartilage 2010;18: 1441e7. 34. Livshits G, Zhai G, Hart DJ, Kato BS, Wang H, Williams FM, et al. Interleukin-6 is a significant predictor of radiographic knee osteoarthritis: the Chingford Study. Arthritis Rheum 2009;60: 2037e45. 35. Orita S, Koshi T, Mitsuka T, Miyagi M, Inoue G, Arai G, et al. Associations between proinflammatory cytokines in the synovial fluid and radiographic grading and pain-related scores in 47 consecutive patients with osteoarthritis of the knee. BMC Musculoskelet Disord 2011;12:144. 36. Conde J, Scotece M, Lopez V, Abella V, Hermida M, Pino J, et al. Differential expression of adipokines in infrapatellar fat pad (IPFP) and synovium of osteoarthritis patients and healthy individuals. Ann Rheum Dis 2014;73:631e3.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

14

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

37. Klein-Wieringa IR, Kloppenburg M, BastiaansenJenniskens YM, Yusuf E, Kwekkeboom JC, El-Bannoudi H, et al. The infrapatellar fat pad of patients with osteoarthritis has an inflammatory phenotype. Ann Rheum Dis 2011;70:851e7. 38. Eymard F, Pigenet A, Citadelle D, Flouzat-Lachaniette CH, Poignard A, Benelli C, et al. Induction of an inflammatory and prodegradative phenotype in autologous fibroblast-like synoviocytes by the infrapatellar fat pad from patients with knee osteoarthritis. Arthritis Rheumatol 2014;66:2165e74. 39. Wickham MQ, Erickson GR, Gimble JM, Vail TP, Guilak F. Multipotent stromal cells derived from the infrapatellar fat pad of the knee. Clin Orthop Relat Res 2003:196e212. 40. Diepold J, Ruhdorfer A, Dannhauer T, Wirth W, Steidle E, Eckstein F. Sex-differences of the healthy infra-patellar (Hoffa) fat pad in relation to intermuscular and subcutaneous fat content e data from the osteoarthritis initiative. Ann Anat 2015;200C: 30e6, http://dx.doi.org/10.1016/j.aanat.2014.12.004 [Epub ahead of print] Jan 31. 41. Roemer FW, Guermazi A, Zhang Y, Yang M, Hunter DJ, Crema MD, et al. Hoffa's fat pad: evaluation on unenhanced MR images as a measure of patellofemoral synovitis in osteoarthritis. AJR Am J Roentgenol 2009;192:1696e700. 42. Krasnokutsky S, Belitskaya-Levy I, Bencardino J, Samuels J, Attur M, Regatte R, et al. Quantitative magnetic resonance imaging evidence of synovial proliferation is associated with radiographic severity of knee osteoarthritis. Arthritis Rheum 2011;63:2983e91. 43. Loeuille D, Sauliere N, Champigneulle J, Rat AC, Blum A, CharyValckenaere I. Comparing non-enhanced and enhanced sequences in the assessment of effusion and synovitis in knee OA: associations with clinical, macroscopic and microscopic features. Osteoarthritis Cartilage 2011;19:1433e9. 44. Crema MD, Felson DT, Roemer FW, Niu J, Marra MD, Zhang Y, et al. Peripatellar synovitis: comparison between noncontrast-enhanced and contrast-enhanced MRI and association with pain. The MOST study. Osteoarthritis Cartilage 2013;21:413e8. 45. Guermazi A, Roemer FW, Hayashi D, Crema MD, Niu J, Zhang Y, et al. Assessment of synovitis with contrast-enhanced MRI using a whole-joint semiquantitative scoring system in people with, or at high risk of, knee osteoarthritis: the MOST study. Ann Rheum Dis 2011;70:805e11. 46. Eshed I, Krabbe S, Ostergaard M, Boyesen P, Moller JM, Therkildsen F, et al. Influence of field strength, coil type and image resolution on assessment of synovitis by unenhanced MRI e a comparison with contrast-enhanced MRI. Eur Radiol 2015;25:1059e67. 47. MacIsaac KD, Baumgartner R, Kang J, Loboda A, Peterfy C, DiCarlo J, et al. Pre-treatment whole blood gene expression is associated with 14-week response assessed by dynamic contrast enhanced magnetic resonance imaging in infliximab-treated rheumatoid arthritis patients. PLos One 2014;9:e113937. 48. Workie DW, Dardzinski BJ. Quantifying dynamic contrastenhanced MRI of the knee in children with juvenile rheumatoid arthritis using an arterial input function (AIF) extracted from popliteal artery enhancement, and the effect of the choice of the AIF on the kinetic parameters. Magn Reson Med 2005;54:560e8. 49. Sakai H, Tamai K, Iwamoto A, Saotome K. Para-articular chondroma and osteochondroma of the infrapatellar fat pad: a report of three cases. Int Orthop 1999;23:114e7. 50. Helpert C, Davies AM, Evans N, Grimer RJ. Differential diagnosis of tumours and tumour-like lesions of the infrapatellar (Hoffa's) fat pad: pictorial review with an emphasis on MR imaging. Eur Radiol 2004;14:2337e46.

51. Masih S, Antebi A. Imaging of pigmented villonodular synovitis. Semin Musculoskelet Radiol 2003;7:205e16. 52. Jaffe HL, Lichtenstein L, Sutro CJ. Pigmented villonodular synovitis, bursitis, tenosynovitis. Arch Pathol 1941;31:731e65. 53. Fletcher CDM, de St. Aubain Somerhausen N, Dal Cin P. Diffuse-type giant cell tumour. In: Fletcher CDM, Unni KK, Mertens F, Eds. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon: IARC Press; 2002:112e4. 54. Fletcher CDM, Bridge JA, Hogendoorn PCW, Mertens F, Eds. World Health Organization Classification of Tumours of Soft Tissue and Bone. 4th edn. Lyon: IARC Press; 2013:100e3. 55. Aynaci O, Ahmetoglu A, Reis A, Turhan AU. Synovial hemangioma in Hoffa's fat pad (case report). Knee Surg Sports Traumatol Arthrosc 2001;9:355e7. 56. Ryd L, Stenstrom A. Hemangioma mimicking meniscal injury. A report on 10 years of knee pain. Acta Orthop Scand 1989;60: 230e1. 57. Perdikakis E, Skiadas V. MRI characteristics of cysts and “cystlike” lesions in and around the knee: what the radiologist needs to know. Insights Imaging 2013;4:257e72. 58. Kim MG, Kim BH, Choi JA, Lee NJ, Chung KB, Choi YS, et al. Intra-articular ganglion cysts of the knee: clinical and MR imaging features. Eur Radiol 2001;11:834e40. 59. Pedowitz RA, Feagin JA, Rajagopalan S. A surgical algorithm for treatment of cystic degeneration of the meniscus. Arthroscopy 1996;12:209e12. 60. Lantz B, Singer KM. Meniscal cysts. Clin Sports Med 1990;9: 707e25. 61. Howe TS, Koh JS. Arthroscopic internal marsupialization of meniscal cysts. Knee 2007;14:408e10. 62. Anderson JJ, Connor GF, Helms CA. New observations on meniscal cysts. Skeletal Radiol 2010;39:1187e91. 63. De Smet AA, Graf BK, del Rio AM. Association of parameniscal cysts with underlying meniscal tears as identified on MRI and arthroscopy. AJR Am J Roentgenol 2011;196:W180e6. 64. Parisien JS. Arthroscopic treatment of cysts of the menisci. A preliminary report. Clin Orthop Relat Res 1990;257:154e8. 65. Hulet C, Souquet D, Alexandre P, Locker B, Beguin J, Vielpeau C. Arthroscopic treatment of 105 lateral meniscal cysts with 5year average follow-up. Arthroscopy 2004;20:831e6. 66. Bisicchia S, Savarese E. Infra-patellar fat pad cysts: a case report and review of the literature. Muscles Ligaments Tendons J 2013;2:305e8. 67. Aydingoz U, Oguz B, Aydingoz O, Comert RB, Akgun I. The deep infrapatellar bursa: prevalence and morphology on routine magnetic resonance imaging of the knee. J Comput Assist Tomogr 2004;28:557e61. 68. Tschirch FT, Schmid MR, Pfirrmann CW, Romero J, Hodler J, Zanetti M. Prevalence and size of meniscal cysts, ganglionic cysts, synovial cysts of the popliteal space, fluid-filled bursae, and other fluid collections in asymptomatic knees on MR imaging. AJR Am J Roentgenol 2003;180:1431e6. 69. McCarthy CL, McNally EG. The MRI appearance of cystic lesions around the knee. Skeletal Radiol 2004;33:187e209. 70. Schweitzer ME, Falk A, Berthoty D, Mitchell M, Resnick D. Knee effusion: normal distribution of fluid. AJR Am J Roentgenol 1992;159:361e3. 71. Aydingoz U, Oguz B, Aydingoz O, Bayramoglu A, Demiryurek D, Akgun I, et al. Recesses along the posterior margin of the infrapatellar (Hoffa's) fat pad: prevalence and morphology on routine MR imaging of the knee. Eur Radiol 2005;15:988e94. 72. Patel SJ, Kaplan PA, Dussault RG, Kahler DM. Anatomy and clinical significance of the horizontal cleft in the infrapatellar fat pad of the knee: MR imaging. AJR Am J Roentgenol 1998;170:1551e5.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018

F.W. Roemer et al. / Osteoarthritis and Cartilage xxx (2015) 1e15

73. Abreu MR, Chung CB, Trudell D, Resnick D. Hoffa's fat pad injuries and their relationship with anterior cruciate ligament tears: new observations based on MR imaging in patients and MR imaging and anatomic correlation in cadavers. Skeletal Radiol 2008;37:301e6. 74. Apostolaki E, Cassar-Pullicino VN, Tyrrell PN, McCall IW. MRI appearances of the infrapatellar fat pad in occult traumatic patellar dislocation. Clin Radiol 1999;54:743e7. 75. Cothran RL, McGuire PM, Helms CA, Major NM, Attarian DE. MR imaging of infrapatellar plica injury. AJR Am J Roentgenol 2003;180:1443e7. 76. Sonin AH, Fitzgerald SW, Bresler ME, Kirsch MD, Hoff FL, Friedman H. MR imaging appearance of the extensor mechanism of the knee: functional anatomy and injury patterns. RadioGraphics 1995;15:367e82. 77. Hoffa A. Influence of adipose tissue with regard to the pathology of the knee joint. JAMA 1904;43:795e6. 78. Turhan E, Doral MN, Atay AO, Demirel M. A giant extrasynovial osteochondroma in the infrapatellar fat pad: end stage Hoffa's disease. Arch Orthop Trauma Surg 2008;128:515e9. 79. von Engelhardt LV, Tokmakidis E, Lahner M, David A, Haage P, Bouillon B, et al. Hoffa's fat pad impingement treated arthroscopically: related findings on preoperative MRI in a case series of 62 patients. Arch Orthop Trauma Surg 2010;130:1041e51. 80. Chung CB, Skaf A, Roger B, Campos J, Stump X, Resnick D. Patellar tendon-lateral femoral condyle friction syndrome: MR imaging in 42 patients. Skeletal Radiol 2001;30:694e7. 81. Chhabra A, Subhawong TK, Carrino JA. A systematised MRI approach to evaluating the patellofemoral joint. Skeletal Radiol 2011;40:375e87. 82. Subhawong TK, Eng J, Carrino JA, Chhabra A. Superolateral Hoffa's fat pad edema: association with patellofemoral maltracking and impingement. AJR Am J Roentgenol 2010;195: 1367e73.

15

83. Campagna R, Pessis E, Biau DJ, Guerini H, Feydy A, Thevenin FS, et al. Is superolateral Hoffa fat pad edema a consequence of impingement between lateral femoral condyle and patellar ligament? Radiology 2012;263:469e74. 84. De Smet AA, Davis KW, Dahab KS, Blankenbaker DG, del Rio AM, Bernhardt DT. Is there an association between superolateral Hoffa fat pad edema on MRI and clinical evidence of fat pad impingement? AJR Am J Roentgenol 2012;199:1099e104. 85. Jibri Z, Martin D, Mansour R, Kamath S. The association of infrapatellar fat pad oedema with patellar maltracking: a casecontrol study. Skeletal Radiol 2012;41:925e31. 86. Matcuk Jr GR, Cen SY, Keyfes V, Patel DB, Gottsegen CJ, White EA. Superolateral Hoffa fat-pad edema and patellofemoral maltracking: predictive modeling. AJR Am J Roentgenol 2014;203:W207e12. 87. Ward SR, Terk MR, Powers CM. Patella alta: association with patellofemoral alignment and changes in contact area during weight-bearing. J Bone Surg Am 2007;89:1749e55. 88. Subhawong TK, Thakkar RS, Padua A, Flammang A, Chhabra A, Carrino JA. Patellofemoral friction syndrome: magnetic resonance imaging correlation of morphologic and T2 cartilage imaging. J Comput Assist Tomogr 2014;38:308e12. 89. Kim SJ, Min BH, Kim HK. Arthroscopic anatomy of the infrapatellar plica. Arthroscopy 1996;12:561e4. 90. Discepola F, Park JS, Clopton P, Knoll AN, Austin MJ, Le HB, et al. Valid MR imaging predictors of prior knee arthroscopy. Skeletal Radiol 2012;41:67e74. 91. Tang G, Niitsu M, Ikeda K, Endo H, Itai Y. Fibrous scar in the infrapatellar fat pad after arthroscopy: MR imaging. Radiat Med 2000;18:1e5. 92. Bradley DM, Bergman AG, Dillingham MF. MR imaging of cyclops lesions. AJR Am J Roentgenol 2000;174:719e26.

Please cite this article in press as: Roemer FW, et al., Magnetic resonance imaging of Hoffa's fat pad and relevance for osteoarthritis research: a narrative review, Osteoarthritis and Cartilage (2015), http://dx.doi.org/10.1016/j.joca.2015.09.018