If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
MRI-detected subchondral bone marrow signal alterations of the knee joint: terminology, imaging appearance, relevance and radiological differential diagnosis
Address correspondence and reprint requests to: Frank W. Roemer, Department of Radiology, Boston University Medical Center, FGH Building, 3rd Floor, 820 Harrison Avenue, Boston, MA 02118, United States. Tel: 1-617-414-4954; Fax: 1-617-638-6616.
Quantitative Imaging Center, Department of Radiology, Boston University School of Medicine, Boston, MA, United StatesDepartment of Radiology, Klinikum Augsburg, Germany
To discuss terminology, radiological differential diagnoses and significance of magnetic resonance imaging (MRI)-detected subchondral bone marrow lesions (BMLs) of the knee joint.
Methods
An overview of the published literature is presented. In addition, the radiological appearance and differential diagnosis of subchondral signal alterations of the knee joint are discussed based on expert consensus. A recommendation for terminology is provided and the relevance of these imaging findings for osteoarthritis (OA) research is emphasized.
Results
A multitude of differential diagnoses of subchondral BMLs may present with a similar aspect and signal characteristics. For this reason it is crucial to clearly and specifically define the type of BML that is being assessed and to use terminology that is appropriate to the condition and the pathology. In light of the currently used terminology, supported by histology, it seems appropriate to apply the widely used term “bone marrow lesion” to the different entities of subchondral signal alterations and in addition to specifically and precisely define the analyzed type of BML. Water sensitive sequences such as fat suppressed T2-weighted, proton density-weighted, intermediate-weighted fast spin echo or short tau inversion recovery (STIR) sequences should be applied to assess non-cystic BMLs as only these sequences depict the lesions to their maximum extent. Assessment of subchondral non-cystic ill-defined BMLs on gradient echo-type sequences should be avoided as they will underestimate the size of the lesion. Differential diagnoses of OA related BMLs include traumatic bone contusions and fractures with or without disruption of the articular surface. Osteonecrosis and bone infarcts, inflammation, tumor, transient idiopathic bone marrow edema, red marrow and post-surgical alterations should also be considered.
Conclusion
Different entities of subchondral BMLs that are of relevance in the context of OA research may be distinguished by specific imaging findings, patient characteristics, symptoms, and history and are discussed in this review.
With its ability to visualize bone, cartilage and soft tissues, magnetic resonance imaging (MRI) is the method of choice for the assessment of acute and chronic joint disorders.
Subchondral bone marrow edema-like signal alterations exhibit typical signal characteristics on MRI and are common but non-specific findings. MRI is the ideal tool to assess these pathologies as the subchondral bone marrow cannot be visualized by X-ray or ultrasound technology. Other imaging modalities reflecting bone metabolism will show increased tracer uptake
, but will not help in the differential diagnosis as they are non-specific. Arthroscopy is excellent for visualizing the articular surface but cannot assess subchondral bone.
On MRI, subchondral bone marrow signal alterations are characterized by ill-defined low signal intensity compared with the unaffected bone marrow on T1-weighted (T1w) images. On T2 (T2w)- or proton density-weighted (PDw) fat suppressed (FS) fast spin echo (FSE) or short tau inversion recovery (STIR) images they are characterized by ill-defined subchondral areas of high signal intensity
. After intravenous administration of contrast agents, enhancement of these signal alterations is evident, indicating hypervascularity and repair activity
STIR vs. T1-weighted fat-suppressed gadolinium-enhanced MRI of bone marrow edema of the knee: computer-assisted quantitative comparison and influence of injected contrast media volume and acquisition parameters.
Subchondral bone marrow signal alterations may be observed in conjunction with trauma, chronic cartilage damage and osteoarthritis (OA), as an idiopathic entity or as a concomitant feature of other pathologies such as osteonecrosis, inflammation or tumor. In OA, the importance of subchondral bone marrow signal alterations for structural progression, as well as explaining pain, has been recognized
Change in MRI-detected subchondral bone marrow lesions is associated with cartilage loss – the MOST study A longitudinal multicenter study of knee osteoarthritis.
Association of bone marrow lesions with knee structures and risk factors for bone marrow lesions in the knees of clinically healthy, community-based adults.
Change in MRI-detected subchondral bone marrow lesions is associated with cartilage loss – the MOST study A longitudinal multicenter study of knee osteoarthritis.
STIR vs. T1-weighted fat-suppressed gadolinium-enhanced MRI of bone marrow edema of the knee: computer-assisted quantitative comparison and influence of injected contrast media volume and acquisition parameters.
MRI assessment of knee osteoarthritis: Knee Osteoarthritis Scoring System (KOSS)—inter-observer and intra-observer reproducibility of a compartment-based scoring system.
. Understanding of the pathophysiology of BMLs is limited. Trauma-induced subchondral bone marrow alterations however, are a result of direct or indirect impact and represent a “footprint” of the injury mechanism
The following article is based on a consensus among experts in the fields of radiology, rheumatology and OA imaging research who have profound clinical insight of possible differential diagnoses and have a research background of investigating BMLs. Aim of this article is not a metanalysis of the published literature, but the clinical focus of the radiological appearance and differential diagnosis of subchondral bone marrow signal alterations of the knee joint. A recommendation for terminology is provided and the relevance of these imaging findings for research purposes is discussed.
Terminology
The term “bone marrow edema” was first introduced to the radiological community by Wilson et al.
. In a group of patients with debilitating knee and hip pain they described ill-defined bone marrow hyperintensities on T2w MR images. Standard radiographs showed non-specific osteopenia or were normal. The authors called their findings “bone marrow edema” due to the “lack of a better term and to emphasize the generic character of the condition”
. Histologically, non-characteristic abnormalities are found in areas of edema-like signal changes in conjunction with OA or cartilage damage. These include bone marrow necrosis, bone marrow fibrosis, and trabecular abnormalities but very little bone marrow edema
. Thus, the term bone marrow “edema” is not appropriate. Zanettti et al. suggested replacing the term “bone marrow edema” with “ill-defined signal intensity abnormality” or “edema-like MR imaging abnormality”
Change in MRI-detected subchondral bone marrow lesions is associated with cartilage loss – the MOST study A longitudinal multicenter study of knee osteoarthritis.
. In this review, however, the term BML in conjunction with cartilage damage or OA is applied exclusively to ill-defined non-cystic subchondral lesions that exhibit the typical MRI signal characteristics. A classification of BMLs according to cause into ischemic, mechanical and reactive has been suggested
. However as the pathophysiology is at best poorly understood for many of the lesions we believe such differentiation might be misleading. Especially for osteochondritis dissecans, spontaneous osteonecrosis of the knee (SONK) and idiopathic transient BML syndrome possible combinations of ischemic and microtraumatic etiology are current theories
. For these reasons we will differentiate only between traumatic and non-traumatic BMLs.
In light of the currently used terminology it seems appropriate to apply the widely used term “bone marrow lesion” to the different entities of subchondral signal alterations. Thus, traumatic BMLs should be differentiated from chronic BMLs in association with OA and from BMLs of other origin. This relatively broad differentiation will not always suffice to describe a patient population accurately. We therefore suggest in addition to specifically and precisely define the analyzed type of BML. Examples could be “traumatic BML without associated fracture”, “traumatic BML in conjunction with osteochondral fracture”, “idiopathic non-traumatic BML”, “OA-associated BML”, “chronic BML in conjunction with osteonecrosis” and so on.
Technical aspects
Research groups in the radiological community have been applying water sensitive sequences such as FS T2w, PDw, intermediate-weighted (IW) FS FSE or STIR sequences to assess non-cystic BMLs in conjunction with cartilage damage or OA, as only these sequences depict the lesions in their maximum extent
Change in MRI-detected subchondral bone marrow lesions is associated with cartilage loss – the MOST study A longitudinal multicenter study of knee osteoarthritis.
. It has been suggested that comparable results might be achieved for T1w FS sequences after intravenous gadolinium administration, however the value of those sequences remains to be proven in larger ongoing OA studies
STIR vs. T1-weighted fat-suppressed gadolinium-enhanced MRI of bone marrow edema of the knee: computer-assisted quantitative comparison and influence of injected contrast media volume and acquisition parameters.
Gradient recalled echo (GRE)-type sequences are very sensitive in delineating subchondral cysts. However, even with robust fat suppression or water excitation, these sequences are insensitive to diffuse marrow abnormalities because of trabecular magnetic susceptibility and will not show the full extent of these lesions
Magnetic resonance imaging of articular cartilage of the knee: comparison between fat-suppressed three-dimensional SPGR imaging, fat-suppressed FSE imaging, and fat-suppressed three-dimensional DEFT imaging, and correlation with arthroscopy.
(Fig. 2). Consequently, GRE-type sequences should not be used to assess subchondral non-cystic ill-defined BMLs in conjunction with OA; the sequences will underestimate the size of the lesion. However, two or three orthogonal sequences of similar image characteristic can be applied to exactly locate the lesion and optimize size estimation.
Fig. 2Comparison of FSE and GRE-type sequences for visualization of BMLs and diffuse marrow changes at 3 T MRI. (A) Sagittal IW image shows large BML in the lateral trochlea (arrows). Note small cyst directly subchondrally within large BML. (B) DESS sequence does not depict BML at all but visualizes the cyst clearly (arrow). (C) Sagittal IW sequence shows subchondral BML in the central lateral tibia (arrow). Note also metaphyseal marrow reconversion in the distal femur (arrowheads). (D) DESS sequence does not show non-cystic BML but depicts only the small cyst within the BML (arrow). Marrow reconversion is not visualized.
Correct assessment of sclerotic lesions requires non-FS T1w sequences. Thus, sclerosis, which exhibits low signal intensity on T1w and T2w/PDw images, can be distinguished from the surrounding physiologic fatty bone marrow (Fig. 3). For BMLs in conjunction with disorders like SONK, osteochondritis dissecans or malignant disease, contrast administration may be very helpful in distinguishing vascularized from ischemic and necrotic areas
. Finally, susceptibility artifacts due to orthopedic surgery and metallic implants may mimic subchondral BMLs (Fig. 4).
Fig. 3Sclerotic lesion. (A) Sagittal T1w SE sequence shows small hypointense lesion subchondrally. Sclerotic changes commonly show marked hypointensity on T1w images. (B) On the sagittal PD FS sequence lesion is barely visualized as both lesions and marrow are hypointense. T1w FS sequence without fat suppression is mandatory for characterizing sclerotic lesions.
Fig. 4Susceptibility artifacts due to metallic implant after ACL reconstruction surgery. Coronal PD FS image. Insufficient fat suppression in the medial subchondral tibia makes differentiation of pathologic subchondral BML and artifact impossible.
Several semiquantitative scoring methods to assess the size of BMLs have been published. The Whole Organ Magnetic Resonance Imaging Score (WORMS) defines BML size according to the percentage of subregional involvement but does not differentiate between individual BMLs
. The Boston-Leeds Osteoarthritis Knee Score (BLOKS) assesses BMLs individually according to percentage of subregional involvement, surface area adjacent to subchondral bone and amount of BML that is cystic, respectively non-cystic
MRI assessment of knee osteoarthritis: Knee Osteoarthritis Scoring System (KOSS)—inter-observer and intra-observer reproducibility of a compartment-based scoring system.
. In addition, quantitative volumetric analysis has been introduced. Different strategies can be applied to localize, regionalize and calculate metric proportions using either manual or semi-automated segmentation of BML volume from surrounding normal bone marrow
STIR vs. T1-weighted fat-suppressed gadolinium-enhanced MRI of bone marrow edema of the knee: computer-assisted quantitative comparison and influence of injected contrast media volume and acquisition parameters.
Fig. 5Segmentation of BMLs for quantitative analysis. (A) Sagittal PD FS image. A large traumatic subchondral BML is depicted in the lateral tibia (white arrows). (B) Color-coded segmentation of cartilage (purple, orange) and BML (black arrows). Image courtesy of VirtualScopics Inc., Rochester, NY, USA.
The following paragraphs will focus on the characteristic imaging findings, the pathology, the prognosis and possible complications of the most common types of BML. A schematic overview of the differential diagnoses and their characteristic imaging findings is presented Table I.
Table IDifferential diagnosis of BMLs
Diagnosis
Typical appearance of BML
Location
History
Typical patient characteristics
Associated imaging findings
Contour deformity of articular surface
Prognosis
Technical remarks
Traumatic BMLS
Bone contusion/“bone bruise”
Diffuse epi- or epi-metaphyseal BML without other associated imaging findings
Depending on injury mechanism (direct/indirect)
Adequate direct or indirect trauma. BML seen immediately after trauma
Young active patients. Elderly patients
Often ligamentary, meniscal or other associated soft tissue injury
No
Good. Usually resolves within months to 2 years
Size best assessed on water sensitive FS FSE sequences
Subchondral impaction
In addition to diffuse BML, finding of subchondral thickened black line
Usually tibia or lateral femur
Adequate trauma (often in conjunction with ACL disruption)
Young active patients. Elderly patients
Often ligamentary, meniscal or other associated soft tissue injury
No
Good. See above
Osteochondral/subchondral/chondral fractures
Diffuse BML. In addition disruption of articular surface for chondral/osteochondral fractures
Anywhere epiphyseal directly subchondral
Adequate trauma
Young active patients. Elderly patients
Often ligamentary, meniscal or other associated soft tissue injury
Yes
Depending on involvement of articular surface and other associated injuries such as meniscal damage
T1w sequences helpful to diagnose fracture line
Insufficiency and stress fractures
Diffuse BML. In addition subchondral fracture line
Femur>tibia>patella
Usually no relevant trauma. History of osteoporosis
Elderly patient with osteoporosis
Usually no other pathology
No
Good with non-weightbearing regimen and therapy of underlying pathology
T1w sequences helpful to diagnose fracture line
Repetitive microtrauma/overuse
Diffuse BML without associated other imaging findings
Anywhere subchondral and subspinous region
Overuse/extreme physical activity
Middle-aged active patients
Signs of cartilage damage may or may not be present. Effusion
No
Good with cessation of causing event and if necessary non-weight bearing
Non-traumatic BMLs
AVN/infarcts
Subchondral lesion of centrally preserved fatty marrow with peripheral rim sign
Epi- and metaphyseal
Risk factors usually present for large lesions (steroid treatment, alcoholism etc) Small lesions may be incidental without apparent cause
Middle-aged or elderly patients
Peripheral rim shows enhancement after contrast administration
Subchondral collapse with deformity is a complication
Usually good, depending on size of lesion
Contrast enhancement may help in differentiation of lesion. T1w sequences helpful for depiction of fracture line
SONK
Subchondral thickened area of necrosis with surrounding large diffuse BML
Epiphyseal. Usually medial femur
Usually no triggering event. Sudden onset of pain
Elderly female patient
Effusion
Deformity and subchondral collapse as complication. Fragmentation of osteochondral fragment possible
Bad. May lead to rapid joint destruction
Contrast enhancement may help in differentiation of lesion. T1w sequences helpful for depiction of fracture line
OA-associated BMLs
Diffuse BML directly adjacent to cartilage lesion
Subchondrally anywhere in joint
Risk factors for OA development
Any patient with cartilage lesions
Other signs of OA
Attrition often seen in later stages of disease
Lesions show fluctuation in size
Lesions show enhancement after contrast application
Idiopathic BMLs/transient bone marrow edema syndrome
Large diffuse BMLs without associated other findings
Anywhere in joint
Sudden onset of pain, otherwise none
Middle-aged men, pregnant women
Effusion possible
No
With non-weight bearing regimen good prognosis Lesions may migrate within joint and from one joint to another
Osteochondritis dissecans
Osteochondral lesion with surrounding BML
Medial femur
Family history, repetitive microtrauma or none
Adolescent males>females
Effusion. Depending on stage – fragmentation of osteochondral fragment possible
Possible in later stages
Depending on stage of disease. Early stages good
Contrast enhancement helpful to assess viability of lesion
Inflammatory BML
BML is associated finding of underlying inflammatory joint
Anywhere in joint
Osteomyelitis, diabetes, rheumatoid arthritis
Middle-aged and elderly patients
Effusion
Usually no
If underlying disease is treated adequately BML will regress
Tumor and malignant infiltration
Perifocal diffuse BML for solid tumors. Diffuse marrow infiltration for hematologic disease
Perifocal or diffuse
Systemic hematologic disease or localized tumor
Any age
Effusion
Usually no
Depending on underlying disease
Contrast administration helpful for tumor size assessment
Red marrow
Diffuse BML-like appearance
Metaphyseally, rarely epiphyseally
None in younger patients, smoking, hypoxia of any cause, malignancy
Any age
Usually none
No
Depending on underlying disease
Post-surgical
Perifocal edema depending on procedure
Depending on procedure
Knee surgery
Any age
Effusion, post-surgical findings
Usually no. SONK possible after meniscectomy in elderly patients
Usually good if surgery successful
Susceptibility artifacts due to metallic implants may mimic BML
Traumatic BMLs including contusions, fractures, SONK and overuse
Trauma-induced BMLs can be differentiated into lesions associated with acute trauma, such as bone contusions due to direct or indirect impact and subacute lesions as a result of overload, such as insufficiency fractures and repetitive microtrauma in conjunction with physical activity. Furthermore, the integrity of the overlying cartilaginous surface has to be considered where purely subchondral traumatic BMLs need to be differentiated from concomitant osteochondral and chondral injury.
The histology of bone marrow affected by acute trauma suggests trabecular fracture, edema, osteocyte necrosis and bleeding in the fatty marrow
. Cartilage adjacent to areas of bone marrow signal alterations after acute ACL injury shows degeneration of chondrocytes and loss of proteoglycans within the affected cartilage
The natural history of bone bruises. A prospective study of magnetic resonance imaging-detected trabecular microfractures in patients with isolated medial collateral ligament injuries.
. However, the data is controversial and a recent publication assessing BMLs quantitatively in ACL injured knees reported BMLs persisting in a majority of knees after 1 year of follow-up
Traumatic bone contusions and subchondral impaction
Bone contusions are characterized as poorly defined, reticulated, heterogeneous signal alterations. Bone contusions are usually found epiphyseally and adjacent to the tidemark but can also be located distant from the subchondral plate. They may resemble subchondral BMLs associated with chronic cartilage damage but show a distinct location pattern according to the mechanism of injury
. Most of the time bone contusions are observed in conjunction with additional ligamentary or meniscal injury and are a footprint of the mechanism of injury
Fig. 6Examples of traumatic bone contusions illustrating the injury mechanism. (A) Sagittal PD FS image shows bone contusions in the lateral femur and anterior lateral tibia after a hyperextension injury and associated PCL disruption. (B) Sagittal STIR image. Characteristic bone contusions in the postero-lateral tibia (arrowheads) and central lateral femur (arrows) in association with an ACL tear (arrows).
. These injuries can be differentiated from simple bone contusions because they show variable degrees of depression of the articular cortical osteochondral surface
Purely chondral fractures, also called flake fractures, are especially common in the knee joint. In most cases, the chondral fragment is displaced. These fractures may show accompanying traumatic BMLs or not. Purely chondral lesions can be differentiated from osteochondral lesions by the subchondral black line that separates the cartilage from the bone marrow. If this black line is normal in shape, thickness or signal, it may be concluded that the underlying bone is not involved and that it is a strictly chondral lesion with or without adjacent bone contusion. In cases of disruption, an osteochondral fracture is diagnosed, which always shows adjacent traumatic BMLs or bone contusions (Fig. 7). Depression fractures with or without a disruption of the cortical bone, also known as deep lateral notch sign
, are frequently observed in combination with ACL injuries.
Fig. 7Osteochondral fractures. (A) Osteochondral fracture associated with an ACL disruption. Coronal PD FS image shows depression of the chondral articular surface (white arrow) and fracture line extending subchondrally (arrowhead). (B) T1w image depicts osteochondral fracture in the lateral femur. Chondral disruption is shown (arrowhead) as well as subchondral fracture line (arrow). (C) Sagittal PD FS image. Small osteochondral depression fracture in the lateral femur is depicted (arrow).
Insufficiency fractures are types of stress fractures that occur in bone unable to withstand the stresses of normal activity. By definition, these fractures are not the result of increased physical activity or preceding trauma, as they occur when normal and physiologic muscular strength is applied to bone that is deficient in mineralization or in its elastic resistance. Insufficiency fractures are observed in a variety of diseases in which bone mineral content is reduced. Patients at risk are elderly women who have postmenopausal osteoporosis or patients with metabolic causes of secondary osteoporosis like hyperparathyroidism, Cushing's, diabetes and rheumatoid arthritis
Fig. 8Insufficiency fracture in a 65-year-old woman with generalized osteoporosis. Subchondral fracture line in the postero-lateral femur is shown (arrows).
Stress-related BMLs can be observed without any co-existing degenerative or traumatic changes in the knee. Although there have been relatively few reports of recreational overuse and altered biomechanics causing the BML visualized on MRI, the association is well accepted by most orthopedists and musculoskeletal radiologists (Fig. 9). In an MR examination of a healthy population without arthritis, subchondral BMLs appeared after mechanical stress combined with artificial malalignment of the mechanical axis
. MRI shows abnormalities in the ankle and foot after marathon races as well as in asymptomatic physically active individuals with no preceding extraordinary strain
Fig. 9Diffuse tibial BML due to repetitive microtrauma in a 40-year-old patient woman practicing for a first-time marathon (arrows). Note missing signs of cartilage defects or OA.
Another form of BML is seen at the insertions of ligaments especially the ACL and posterior cruciate ligament (PCL). The lesions are due to traction and may be observed in conjunction with acute trauma or repetitive microtrauma (Fig. 10).
Fig. 10Traction edema at the tibial insertion of the PCL. Diffuse BML is shown in the subspinous region of the tibia adjacent to the insertion of the PCL (arrows).
. SONK most often involves the weight-bearing surface of the medial femoral condyle. A vascular and traumatic theory of its pathogenesis has been discussed, but recent data seems to be more supportive of a mechanical or microtraumatic origin
. The most prominent MRI characteristics of SONK are an ill-defined BML pattern and a lack of peripheral low signal intensity rim as seen in avascular necrosis (AVN) and bone infarcts
Early irreversible osteonecrosis versus transient lesions of the femoral condyles: prognostic value of subchondral bone and marrow changes on MR imaging.
. The BML pattern is not specific for SONK and may be observed in transient epiphyseal conditions. However, a focal subchondral area of low signal intensity adjacent to the subchondral bone plate and representing local ischemia is a specific MRI finding for SONK, and is seen in almost all cases. This area shows no enhancement on post-contrast T1w MRI. Prognosis is associated with extent and depth of this subchondral ischemic area. Another MRI finding suggestive of SONK is a deformity of the subchondral bone plate (flattening or focal depression) in the weight-bearing area of the involved condyle (Fig. 12).
Fig. 11Epiphyseal bone infarct and secondary collapse of the articular surface. (A) Sagittal PD FS image. The infarct displays hypointense central fatty marrow characteristics and is associated with a peripheral rim of reactive tissue in process of gradual restitution (arrows). (B) T1w SE image shows that the subchondral extension of the medial femoral condylar infarct is associated with the complication of an osteochondral fracture (arrows).
Fig. 12SONK following partial meniscectomy. (A) Medial meniscal extrusion and a horizontal tear of the body are depicted (black arrow). Diffuse reactive BME is shown in the medial femur and tibia (white arrows). (B) Six months post-partial meniscectomy (arrow) diagnosis of SONK with dislocation of osteochondral fragment commencing (arrowhead).
Non-traumatic reactive BMLs occur in a group of disorders in which an underlying disease or a prior surgical procedure dominates the history, clinical findings, prognosis, and course of the disease. BMLs in these joints represent only a concomitant component without significant influence on the therapeutic management.
Vascular BMLs
AVN is a synonym for periarticular bone infarcts. The most common MRI appearance of AVN is that of a geographic lesion of more or less preserved fatty marrow signal. On T1w images the lesion is of high signal intensity, surrounded by a low signal intensity rim that is pathognomonic for AVN. Histologically, this low signal intensity rim correlates with the reactive interface between viable and nonviable bone marrow. On T2w images, this rim may present a double-line sign, which is characterized by the juxtaposition of low and high signal intensity rims
. Progression to larger epiphyseal necrosis and ultimately osteochondral collapse is associated with the appearance of localized BMLs around the infarct, signal intensity changes within the weight-bearing area of the infarct and subchondral bone fracture (Fig. 13).
Fig. 13SONK. Sudden onset of medial knee pain in a 52-year-old man. Diffuse BML is shown in the medial femoral condyle (white arrows). Note subchondral area of necrosis (black arrow).
Subchondral BMLs are a hallmark of knee OA on MRI. Radiologically BMLs in OA are understood as non-cystic subchondral areas of ill-defined hyperintensity in T2w, PDw, STIR or IW images and of hypointensity on T1w images
Association of bone marrow lesions with knee structures and risk factors for bone marrow lesions in the knees of clinically healthy, community-based adults.
. As the disease progresses, an increase in BML volume is seen in the same region subchondrally in many patients, which is positively correlated with an increase in cartilage loss and radiographic joint space narrowing
MRI-detected bone marrow edema-like lesions are strongly associated with subchondral cysts in patients with or at risk for knee osteoarthritis: the MOST study.
Fig. 14Classic BML in conjunction with cartilage damage. (A) Sagittal PD FS image shows ill-defined hyperintensity adjacent to the subchondral plate (white arrows). Note superficial cartilage defect adjacent to BML. (B) T1w SE image depicts BML as diffuse hypointensity (black arrows). Note that full size of lesion can only be appreciated on the PD FS image.
However, clinical experience has also shown regression or even complete resolution of subchondral BMLs. The published data on the natural history of BMLs is ambiguous. In the Dutch Genetics, Arthrosis and Progression Study (GARP) study the majority of lesions increased or appeared at 2 year follow-up, but 20% of prevalent BMLs decreased or resolved
. In the Boston Osteoarthritis Knee Study (BOKS) study less than 1% of patients showed a decrease of BMLs in size in the tibiofemoral (TF) joint after 30 months follow-up
. A recent report from the Multicenter Osteoarthritis Knee Study (MOST) study showed that the majority of subchondral BMLs regress or resolve completely
Change in MRI-detected subchondral bone marrow lesions is associated with cartilage loss – the MOST study A longitudinal multicenter study of knee osteoarthritis.
Change in MRI-detected subchondral bone marrow lesions is associated with cartilage loss – the MOST study A longitudinal multicenter study of knee osteoarthritis.
. The clinical relevance of changing BML size remains controversial. A recent study by Phan et al. reported that changes in BMLs did not correlate significantly with changes in Western Ontario and McMaster Osteoarthritis Index (WOMAC) scores
Fig. 15BMLs in conjunction with OA at 3 T MRI. (A) Sagittal IW image shows large subchondral BML in the femoral condyle (arrowheads) and smaller BMLs in the central medial tibia (arrows). (B) Marked regression of BMLs is seen in the tibia and femur 12 months later.
There is sparse data available on the histology of MRI-detected BMLs in osteoarthritic knees. Zanetti et al. showed, in 16 patients referred for total knee replacement, that the lesions consisted of a mixture of different tissue patterns: mainly normal tissue, the lesions also contained necrotic bone marrow, abnormal trabeculae, bone marrow fibrosis, and infrequently bone marrow edema and bleeding
. In a canine animal model MRI-detected BMLs were correlated with histology; hematopoiesis and myxomatous transformation of the bone marrow and/or intertrabecular fibrosis were present, but no edema
. Specific changes in bone mineralization, remodeling and defects within BML features adjacent to the subchondral plate have been shown recently by Hunter et al. using micro-CT
. The mineral density of these BMLs seems to be reduced, and they appear sclerotic compared to unaffected regions from the same individual, based on the increased bone volume fraction, increased trabecular thickness, and decreased structural model index.
Transient bone marrow edema syndrome
The term transient bone marrow edema syndrome has been used to describe the sudden onset of knee pain and MR findings of a typical large BML pattern without a specific cause. Diagnosis is by exclusion. Some authors include transient osteoporosis within this group. However, others prefer this term be used only for patients who do not have osteopenia radiographically. This syndrome is indistinguishable from transient osteoporosis clinically and with MRI. It is thought to be a self-limited process. Many believe idiopathic transient BML syndrome is the same as transient osteoporosis with the disease coming to clinical attention earlier, prior to the development of osteopenia
. These authors reported a painful syndrome affecting the hip which is associated with circumscribed osteopenia on radiography, in a group of women during the third trimester of pregnancy. However, in recent years it has been acknowledged that the condition seems to be most prevalent in middle-aged men
. Patients report progressive joint pain over several weeks to months, exacerbated by weight-bearing activity. Clinical symptoms are generally unilateral and without discernible traumatic episode. In most cases, there are no significant or only mild MRI findings of OA. Plain films may initially be normal but usually demonstrate demineralization in the affected joint with preservation of the joint space. Some patients later develop similar changes in the contralateral side or in other joints, in which case the term transitory migratory bone marrow edema syndrome has been applied
. A typical example of regional migratory bone marrow edema syndrome is shown in Fig. 16. Idiopathic transient bone marrow edema syndrome usually has a good prognosis.
Fig. 16Transient migratory bone marrow edema syndrome over a course of 4 years. Sudden onset of left medial knee pain without trauma in a 56-year-old man. (A) Coronal PD FS image shows diffuse BML in the medial femoral condyle also extending laterally (arrows). After 12 weeks of non-weight bearing the BML vanished and pain resolved completely. (B) New pain 13 months later in the same knee. Diffuse BML is shown in the lateral femoral condyle (arrows). Regression of BML and pain after non-weight bearing for 23 months. (C) New hindfoot pain in the right foot 16 months later. Diffuse BML is shown in the subtalar calcaneus (arrows). (D) Sudden onset of left-sided ankle pain 1 ½ years later. This time a diffuse BML pattern is shown in the talar corpus, neck and head (arrows).
Osteochondritis dissecans differs from SONK in that it primarily affects boys and young men in their second and third decade and typically involves the lateral surface of the medial femoral condyle
. The pathogenesis of OCD is not fully understood but a multifactorial etiology including microtrauma, alterations in subchondral perfusion and genetic predisposition is probable
. In older patients the appearance of OCD may overlap with the spectrum of MR findings seen in association with SONK. Classic OCD exhibits a semi-oval or semi-circular form and a relatively sharp demarcation against the healthy bone marrow (Fig. 17). The central parts of the OCD fragments of the femoral condyle generally have a low signal on T1w images and a heterogeneous mixture of low, intermediate and high signal on T2w images, with dominance of the low signal
. Contrast enhancement in the central parts of OCD indicates that the lesion is completely or almost completely perfused which should be interpreted as evidence of active reparative processes, even in the center of the lesion. The peripheral transitional or demarcation zone between fragment and epiphyseal defect is always of low signal on the T1w images and is frequently, but not always, of high signal intensity in the T2w images
Fig. 17Osteochondritis dissecans in a 12-year-old boy. In situ osteochondral fragment (large arrow) with intact overlying chondral surface. Note also accompanying BML surrounding the lesion (small arrows).
The most common diseases in which reactive inflammatory BMLs are observed are chronic polyarthritis, reactive arthritis, bacterial arthritis, and osteomyelitis. Differential diagnosis is of paramount importance for therapeutic management. MRI is not necessary as the primary modality but certainly helps when the diagnosis is questionable. In the initial stage of chronic polyarthritis, MRI allows evaluation of joint effusion, synovial involvement, bony erosions, periarticular soft tissue involvement, and relatively early alterations of hyaline cartilage. Rheumatic erosions are predicted by subchondral areas of the BML
. Intravenous administration of gadolinium may be helpful for assessing inflammatory activity. In the acute phase of chronic polyarthritis a more or less severe concomitant BML may be observed
. A concomitant BML in bacterial arthritis may be an indicator of bony involvement i.e., osteomyelitis. However, differentiation between acute infectious alterations within the bone and a concomitant reactive BML may be difficult. Intravenous administration of a contrast agent may help in differentiating an intraosseous abscess or necrosis from surrounding BML
Fig. 18Inflammatory edema in rheumatoid arthritis. Sagittal PD FS image shows reactive BME in the epiphyseal femur, tibia and patella (arrows). Note concomitant aseptic effusion in the joint cavity.
Tumors of the knee joint, benign or malignant, are fairly common. After initial plain radiographs, MRI with contrast administration should be performed for exact assessment of bone involvement and involvement of soft tissue structures. MRI morphology of tumors in the knee joint is the same as in other joints (Fig. 19) Concomitant reactive BMLs can be observed in almost all stages of various tumors. In most cases better delineation of concomitant edema from tumor tissue is possible by using intravenous contrast agents. Diffuse infiltration of the bone marrow may be observed in several hematologic and oncologic diseases such as lymphoma, multiple myeloma, and others. On MRI this will be reflected as diffuse marrow alterations with signal characteristics similar to periarticular red marrow (Fig. 20).
Fig. 19Enchondroma. Classic appearance and location. Coronal PD FS image shows a very hyperintense multi-lobulated lesion in the metaphyseal femur (arrow).
Fig. 20Thalassemia. Sagittal IW image at 3 T MRI. Diffuse hyperintense marrow changes are shown in the meta- and epiphyseal femur and tibia. Thalassemia is caused by an inherited abnormality in globin production and is characterized by ineffective erythropoiesis, hemolysis and anemia.
Diffuse BMLs are a common finding after knee surgery. MRI is indicated for follow-up examinations after surgery and when there is continued or recurrent pain after surgery. This reactive BML pattern can be seen up to 6–12 months after surgery and appropriate diagnosis and assignment are not difficult in most cases. In patients with persisting or recurrent pain after arthroscopies with partial meniscectomies
, subchondral signal alterations have been described. In a group of patients with no subchondral signal alterations on MRI before surgery, partial meniscectomy may lead to incident BMLs in the region of meniscectomy in about 1/3 of patients over a period of 2 years
MRI is regularly performed to assess post-surgical status after focal cartilage repair such as microfracturing, matrix associated autologous chondrocyte transplantation or surgery using osteochondral plugs harvested from a non-weight-bearing area of the same joint. Perifocal BMLs may be observed for up to 2 years after surgery depending on the technique and success of procedure (Fig. 21).
Fig. 21Post-surgical BMLs after cartilage repair. (A) Microfracturing. Sagittal PD FS image shows fibro-cartilaginous reparation tissue at the site of the former full-thickness trochlear defect (arrowheads). (B) Osteochondral autologous transplantation. Sagittal FS spoiled GRE-type T1w image. Osteochondral plug is depicted in central part of lateral femur (arrowheads). Note smooth articular surface and BML surrounding the osteochondral plug (arrows).
Subchondral ganglion cysts need to be differentiated from cysts in conjunction with OA and can be observed at all ages. They are rare benign findings, but large cysts may lead to osteochondral collapse in the adjacent region. These cysts are regularly accompanied by a diffuse BML (Fig. 22).
Fig. 22Subchondral cyst. (A) Chronic lateral knee pain in 23-year-old man. Radiograph shows well-defined radiolucency in the lateral tibia representing a cystic lesion (arrow). (B) Sagittal PD FS image shows well-defined subchondral cyst (arrow) with surrounding BML (arrowheads). Cyst not associated with OA.
Other more common findings that may be mistaken for pathologic BMLs are patches of red marrow that are usually observed metaphyseally and are regularly seen in adults. Bone marrow reconversion affecting primarily the metaphysis is observed regularly in anemic patients, smokers and other conditions that stimulate the red marrow (Fig. 23).
Fig. 23Red bone marrow. Coronal PD FS image. Non-pathologic remnant of metaphyseal red marrow in a 20-year-old woman (arrow).
Ali Guermazi is president of Boston Imaging Core Lab, LLC (BICL), Boston, MA, a company providing radiological image assessment services. He is shareholder of Synarc, Inc.
Frank Roemer, Klaus Bohndorf and Michel Crema are shareholders of BICL.
None of the other authors have declared any possible conflict of interest.
References
Boegard T.
Rudling O.
Dahlstrom J.
Dirksen H.
Petersson I.F.
Jonsson K.
Bone scintigraphy in chronic knee pain: comparison with magnetic resonance imaging.
STIR vs. T1-weighted fat-suppressed gadolinium-enhanced MRI of bone marrow edema of the knee: computer-assisted quantitative comparison and influence of injected contrast media volume and acquisition parameters.
Change in MRI-detected subchondral bone marrow lesions is associated with cartilage loss – the MOST study A longitudinal multicenter study of knee osteoarthritis.
Association of bone marrow lesions with knee structures and risk factors for bone marrow lesions in the knees of clinically healthy, community-based adults.
MRI assessment of knee osteoarthritis: Knee Osteoarthritis Scoring System (KOSS)—inter-observer and intra-observer reproducibility of a compartment-based scoring system.
Magnetic resonance imaging of articular cartilage of the knee: comparison between fat-suppressed three-dimensional SPGR imaging, fat-suppressed FSE imaging, and fat-suppressed three-dimensional DEFT imaging, and correlation with arthroscopy.
The natural history of bone bruises. A prospective study of magnetic resonance imaging-detected trabecular microfractures in patients with isolated medial collateral ligament injuries.
Early irreversible osteonecrosis versus transient lesions of the femoral condyles: prognostic value of subchondral bone and marrow changes on MR imaging.
MRI-detected bone marrow edema-like lesions are strongly associated with subchondral cysts in patients with or at risk for knee osteoarthritis: the MOST study.
00077-6&id=gr1.jpg){kind=link}
00077-6&id=gr2.jpg){kind=link}
00077-6&id=gr3.jpg){kind=link}
00077-6&id=gr4.jpg){kind=link}
00077-6&id=gr5.jpg){kind=link}
00077-6&id=gr6.jpg){kind=link}
00077-6&id=gr7.jpg){kind=link}
00077-6&id=gr8.jpg){kind=link}
00077-6&id=gr9.jpg){kind=link}
00077-6&id=gr10.jpg){kind=link}
00077-6&id=gr11.jpg){kind=link}
00077-6&id=gr12.jpg){kind=link}
00077-6&id=gr13.jpg){kind=link}
00077-6&id=gr14.jpg){kind=link}
00077-6&id=gr15.jpg){kind=link}
00077-6&id=gr16.jpg){kind=link}
00077-6&id=gr17.jpg){kind=link}
00077-6&id=gr18.jpg){kind=link}
00077-6&id=gr19.jpg){kind=link}
00077-6&id=gr20.jpg){kind=link}
00077-6&id=gr21.jpg){kind=link}
00077-6&id=gr22.jpg){kind=link}
00077-6&id=gr23.jpg){kind=link}