State-of-the-art imaging techniques for the evaluation of haemophilic arthropathy: present and future
Scientific Meeting: World Federation of Hemophilia 2010, Buenos Aires, Argentina.
Abstract
Summary. In spite of the fact that the diagnosis of haemophilia is essentially clinical and laboratory-based, imaging has become an important tool for the evaluation of complications, diagnostic confirmation and/or complementation and therapeutic follow-up in haemophilic arthropathy. Radiography remains the workforce horse in the diagnosis and follow-up of haemophilic arthropathy. The radiographical findings in arthropathy follow an expected sequence of events and are overall similar in different joints. Magnetic resonance imaging (MRI) has advantages over radiography based on its capability of visualizing soft tissue and cartilage changes in haemophilic joints. The recent development and standardization of MRI scoring systems for measuring soft tissue and cartilage abnormalities may enable the comparison of pathological joint findings in clinical trials conducted at different institutions across the world. The implementation of high-frequency transducers and colour/power Doppler capabilities has provided new insights for clinical applications of ultrasonography (US) in haemophilic arthropathy. In spite of the imaging modality's technical challenges such as operator-dependency, US has advantages over MRI. One of these advantages is its ability of differentiating synovium hypertrophy and hemosiderin deposition, which is not possible with MRI given the presence of susceptibility artefacts from extracellular hemosiderin on gradient-echo MR images. In addition to the aforementioned conventional imaging modalities, novel imaging techniques (blood oxygen level dependent, ultrasmall superparamagnetic iron-oxide contrast-enhanced, and T1 and T2 mapping MRI, ultrasound biomicroscopy, microbubble contrast-enhanced US and positron emission tomography, among others) hold promise for early assessment of haemophilic arthropathy in the future upon completion of their clinical validation.
Introduction
Joint disease affects 90% of severe haemophiliacs and contributes to most of the morbidity of this condition [1]. Ankles, knees and elbows are the joints most frequently involved [2]. Haemophilic arthropathy is caused by recurrent haemorrhagic episodes into the joint, which can be prevented by regular infusions of factor concentrate replacement known as prophylactic regimens [3].
Treatment includes continuous or on-demand clotting factor replacement and radionuclide or open synovectomy. The radiographical findings of haemophilic arthropathy depend on the stage of disease, the age of the patient at onset and the joint involved. These findings include joint effusion, soft tissue swelling, epiphyseal overgrowth, subchondral cysts, osseous erosion and secondary degenerative changes [4].
The pathogenetic mechanisms involved in haemophilic arthropathy are unknown, but most likely multifactorial [5]. Some authors [6–9] suggest that intra-articular blood has a precursor direct effect on cartilage, as a result of the iron-catalysed formation of destructive oxygen metabolites, subsequently affecting the synovium. Other authors [10,11] suggest that there is a hemosiderin-induced synovial triggering process. Nevertheless, most likely both processes occur in parallel, and while they influence each other, they probably do not depend on each other. It is well known, however, that synovial changes precede cartilage changes over the course of haemophilic arthropathy [10].
Conventional imaging modalities
Radiography
Radiography has been used for diagnostic purposes for over a century [12] and remains the standard for diagnosis of haemophilic arthropathy. Nevertheless, this imaging modality is only able to diagnose late arthropathic changes, most notably subchondral and bony abnormalities. While the World Federation of Hemophilia Pettersson X-ray scale [13] does not contain an item to represent soft tissue changes, the Arnold-Hilgartner X-ray scale [14] grades soft tissue and osteochondral changes subjectively on a 0–5 grading system. Pathological changes such as synovial hypertrophy, joint effusion, hemosiderin deposits and periarticular oedema can appear as nonspecific soft tissue swelling on radiography.
A chief limitation of radiographical scales is that articular cartilage can only be assessed indirectly through evaluation of joint space narrowing. Radiography is typically used for therapeutic planning such as arthrodesis and joint replacement, and to follow the progression of arthropathy as a means of monitoring late effects of clinical therapy. Nevertheless, this imaging modality is inadequate for planning modern prevention and for evaluating early treatment efficacy.
Magnetic resonance imaging
Magnetic resonance imaging (MRI) started to be used as an imaging modality for the assessment of haemophilic arthropathy in the 1980s [15], and since then its use and applications in this disease have increased considerably. MRI has been shown to more accurately assess a haemophilic joint than radiography [16]. MRI has obvious advantages over radiography, including the increased level of detail of soft tissue and cartilage changes and lack of ionizing radiation, but it is more costly, less accessible, more time consuming and requires sedation in younger children [17].
Magnetic resonance imaging enables visualization of early arthropathic changes such as hemarthrosis, effusion, synovial hypertrophy, hemosiderin deposition and small focal cartilage defects without joint space narrowing, which cannot be easily delineated by X-ray imaging. Moreover, MRI can provide detailed information about more advanced changes, such as erosions, subchondral cysts and cartilage destruction with joint space narrowing.
In addition to T1- and T2-weighted spin-echo MR images, T2*-weighted gradient-echo imaging can be obtained more quickly than true T2-weighted images and offer better spatial resolution. The magnetic susceptibility artefact from gradient-echo imaging is especially useful in evaluating blood degradation products. On gradient-echo imaging, hemosiderin deposits are intensely black, conversely to the adjacent soft tissues that appear as light grey [18].
Magnetic resonance imaging is a powerful tool in the diagnosis, staging and treatment of patients with haemophilic joint disease. A recent randomized clinical trial [19] showed that primary prophylaxis with recombinant factor VIII can prevent joint damage and decrease the frequency of joint and other haemorrhages in young boys with severe haemophilia A, using visualization of bone or cartilage damage in index joints on MRI as the outcome measure. Another clinical trial (Kraft JK, Bianchette VS, Babyn PS, et al,unpublished data) has shown that MRI is able to detect chronic microhaemorrhages into the joints of haemophilic patients under prophylaxis without any clinical evidence of hemarthroses, and therefore can be a useful adjunct tool in assessing subclinical joint changes in children with severe haemophilia.
To measure arthropathic changes in clinical practice and in clinical research trials, tentative haemophilic arthropathy scales based on MRI findings have been developed in the last decade [20–25]. In 2005, the International Prophylaxis Study Group (IPSG) presented a preliminary comprehensive scoring scheme [22,26,27] that combined the pioneer Denver [20] and European MRI scores [21]. The use of such scales should result in a more consistent assessment of haemophilic joints and should facilitate the development of more targeted treatment to prevent or delay further destructive osteoarticular changes.
Ultrasonography
Although the first sonographic images were obtained in the 1950s [28], the development of ultrasonography (US) for assessing haemophilic joints in clinical practice occurred in subsequent decades. US has advantages over MRI. The former imaging modality is less costly, more accessible, does not require sedation in young children and does not present with interference of susceptibility artefacts, which are commonly seen on gradient-echo MRI sequences. Susceptibility MRI artefacts are represented by low signal intensity (‘blooming’) that covers areas of hemosiderin deposition within the joint [29]. These artefacts may obscure the joint synovium, impairing or making preradiosynoviorthesis imaging evaluations difficult.
Linear high-resolution (7–13 MHz) probes are typically used for assessing haemophilic joints [30], enabling the visualization of superficial musculoskeletal structures such as synovium, tendons, musculature and the cartilage/osteochondral interface at the edge of the joints on grey-scale sonograms. Grey-scale US can also be used to follow the progression or regression of soft tissue hematomas [4,5,31,32] and pseudotumours. The latter entity is a rare complication that occurs in 1–2% of haemophiliacs. Most develop in the muscles of the pelvis and lower extremity, where the large muscles have a rich blood supply, or in bone following intraosseous procedures.
Furthermore, power Doppler sonography has the capability of evaluating synovial vascularity in haemophilic joints [33]. A recent study [34] showed a strong correlation between power Doppler and dynamic contrast-enhanced MRI measurements in haemophilic knees, elbows and ankles.
Ultrasonography does, however, have limitations, such as operator-dependency and partial visualization of the joints (superficial structures can be extremely well seen in detriment of deeper structures, which move beyond the imaging field of view when high-resolution probes are employed). Previous studies evaluated the US findings of haemophilic joints [32,35] and described a systematic protocol for data acquisition of US findings in haemophilic joints [30,36]. Nevertheless, the value of this technique for the assessment of haemophilic arthropathy in comparison with MRI and physical examination has not been fully evaluated so far. As a result, this technique has been underemployed in clinical practice. Another challenge of US relates to the interpretation of images and comparison with other diagnostic tests. Further studies comparing US and MRI are needed to quantitate the diagnostic sensitivity loss of US (for central aspects of the joint which are deeper in nature) with special regard to the interpretation of subchondral abnormalities and cartilage loss.
Computed tomography
Haemophilic pseudotumours are chronic, encapsulated and hemorrhagic fluid collections that contain coagulated blood and are surrounded by a thick wall. They usually destroy adjacent bone and may become quite large. Both MRI and contrast-enhanced computed tomography (CT) are useful in determining the thickness of the wall and the extent of the lesion [4,37]. In the acute stage, the centre of the pseudotumour appears hypodense on CT, but the periphery is isodense and indistinguishable from surrounding muscle [32]. CT shows the thick wall of pseudotumours more consistently than US does. Multiple irregular echoes on US may represent solid material that cannot be documented on CT. Differential diagnosis from abscesses may be difficult.
Nuclear medicine
Radiosynoviorthesis is effective in limiting the frequency of joint haemorrhage, decreasing pain and improving the function in haemophiliacs [38]. Blood-pool indices can be used to evaluate the efficacy of radioisotopic synoviorthesis (90Y and 186Re radiosynovectomy) in patients with haemophilic synovitis, and therefore can be an objective means for monitoring therapy response in these patients.
An in vivo method for erythrocyte labelling with Technetium-99 m generates a dynamic perfusion sequence, which is obtained using a scintillation camera positioned over the area to be examined [39]. This demonstrates the vascularity of the tissue. Subsequently, equilibrium blood-pool images of the area are obtained and analysed with a densitometer to assess relative regional blood volume. In patients who are not bleeding but have chronic arthropathy, vascularity is not increased, and the blood volume of comparable joints may be similar. By contrast, marked increases in vascularity and image density are typically observed in studies of acute bleeding joints. Chronic hemarthroses are typically associated with persistent, but less marked increases in joint perfusion. Transient increases in joint vascularity are demonstrated after insertion of prostheses. In a patient with a thigh haematoma, the dimensions of the haemorrhage can be clearly delineated. As only a tracer dose of radionuclide is infused intravenously, allergic reactions tend to be rare.
Previous studies [40] showed that the correlation between therapeutic outcome (joint bleeding) and the difference of pre and posttherapeutic blood-pool indices were significant (r = 0.594; P < 0.05). A significant increase in the anterior (P < 0.01) and posterior (P < 0.05) views of the blood-pool phase as well as in the anterior view of late phase (P < 0.01) was noted [41]. Moreover, increased Technetium uptake was shown to correlate strongly with the frequency of haemarthrosis, pain, synovitis, range of movement and radiological changes in knees and elbows, but poorly in the ankles [42]. These results support the theory that haemophilic arthritis is amongst the inflammatory arthropathies.
In spite of the potential value of scintigraphy for evaluating posttherapy joint changes, the limited spatial resolution of this imaging modality for the assessment of osteochondral abnormalities and its radiation-bearing potential has limited its use for follow-up of arthropathic changes. Furthermore, long-term safety studies are needed.
Future directions for conventional imaging modalities
A consensus should be reached in MRI and US definitions, and standardized methods for data acquisition and interpretation of these imaging techniques, including a new standardized reference atlas comparing US and MRI findings need to be created/agreed upon. This atlas should be based on a ‘core set’ of MRI sequences and US protocol and should be intended to provide a standardized semi-quantitative assessment tool through which patients’ images can be compared with standard reference images for different degrees of severity of haemophilic arthropathy. For this purpose, it is crucial that the IPSG consensus scales for MRI and US, which are potential research measurement tools for use in future clinical trials of haemophilic arthropathy, are finalized. This work is currently in progress by the Imaging Work Group of the IPSG.
Further consensus should be reached on the imaging modalities to be employed (MRI vs. US) and on the MRI sequences that should be used for measurements according to the aim of the investigation (follow-up of prophylaxis regimens or radiosynoviorthesis, evaluation of complications, failure of treatment, etc.). By this way, a standardized ‘core set’ of MRI sequences can be adjusted to the number of joints to be imaged (single joints vs. all 6 index joints), to the patient’s age and to the degree of detail that is required for the study purpose.
Finally, studies with longer follow-up periods are clearly needed to fully assess the long-term clinical significance of musculoskeletal changes obtained by imaging and physical therapy scores and how these measurements correlate between them. Depending on whether the very early MRI, US or physical therapy changes can reliably predict for clinically significant haemophilic arthropathy in adolescence and adulthood, these techniques may guide individualized therapy approaches for haemophiliacs in the future. This imaging-physical therapy diagnostic/prognostic approach should then determine whether the significant extra cost of full-dose prophylaxis vs. the dose escalation approach is ‘worth it’ in different cohorts of haemophiliacs across the world.
Novel imaging techniques
The use of novel imaging techniques may allow for earlier and more accurate quantification of arthropathic changes both cross-sectionally and over time. Conventional MRI techniques in clinical use do not provide a comprehensive assessment of cartilage and are lacking spatial resolution or specific information about cartilage physiology. The following techniques have been tested experimentally in animal models of arthritis or small cohorts of patients, but hold promise for future translation into clinical trials.
Soft tissues
Magnetic resonance imaging
Blood oxygen level dependent This method relies on MRI contrast resulting from changes in the microvascular ratio of oxyhemoglobin (oxyHb) to deoxyhaemoglobin (deoxyHb). OxyHb is diamagnetic, whereas deoxyHb is paramagnetic, which produces a local bulk magnetic susceptibility effect and subsequent MRI signal change [43]. The changes are typically observed in T2*-weighted functional MRI scans. However, there has recently been interest in BOLD as a way to evaluate microcirculation of any normal or diseased tissue. This technique detects temporal changes in the synovial response of the joint to a stimulus [44] and holds the potential to predict future cartilage changes in an early stage of haemophilic arthropathy.
Ultrasmall superparamagnetic iron-oxide contrast-enhanced MRI It is well known that synovial iron deposition that is easily detectable by conventional gradient-echo MRI techniques is suggested to be indicative of the severity of haemophilic arthropathy [45]. Previous studies showed that iron deposits at localized sites in the synovium are associated with the production of pro-inflammatory cytokines and an ability to inhibit the formation of human cartilage matrix [46]. Proposed mechanisms include the effects of lysosomal enzymes and catabolic cytokines produced by monocytes/macrophages [47–49]. This supports the hypothesis that iron plays a leading role in the induction of synovial changes and the consequent production of catabolic mediators harmful to cartilage.
Newly developed nanoparticle contrast media, known as ‘ultrasmall superparamagnetic iron-oxide (USPIO)’ particles, have been shown [50] to localize to the synovial macrophages in experimental haemophilic arthropathy. Within the joints, the nanoparticles provide significant MRI ‘negative’ contrast, with signal loss on T2-weighted imaging due to T2 shortening caused by their magnetic susceptibility. This negative contrast effect is highly located to the specific areas of macrophage accumulation within affected joints and appears to be quantitatively measurable.
Microbubble contrast-enhanced ultrasonography
Intravenous microbubble echo-contrast agents have the potential to further increase the sensitivity of the colour and power Doppler signal by raising the intensity of weak signals to a detectable level. They hold potential for detecting changes of low-volume, low-velocity blood flow in small vessels such as those of the synovium in arthritic joints [51].
Positron emission tomography
Positron emission tomography [PET] is a technique that uses molecules labelled with isotopes that emit positrons from their nucleus. The most commonly used tracer is 2-deoxy-2-(18F) fluoro-p-deoxyglucose (FDG) [52]. After intravenous injection, FDG is taken up by the cells according to their level of glucose metabolism.
Animal models have demonstrated that FDG uptake by tumours is not only due to the tumour cells themselves, but also due to the inflammatory cells appearing in association with tumour growth or necrosis. Based on this concept, an on-going study (unpublished data) has demonstrated the feasibility of using PET to detect arthritis-related inflammation prior to visualization by morphologic imaging in a rabbit model of blood-induced arthritis. Preliminary results of this study showed that the number of bleeding events would influence the degree of inflammatory changes and consequently, the FDG uptake in affected knees. These data demonstrated that the increased glucose metabolism of many inflammatory cell types and the FDG uptake by inflammatory tissues are the basis for the potential use of FDG-PET in the detection and monitoring of chronic arthropathic processes in haemophilia.
Cartilage
Magnetic resonance imaging
Dual-echo steady-state imaging Dual-echo steady-state (DESS) imaging results in images with higher T2* weighting, which has bright cartilage signal and bright synovial fluid. This technique has proved useful for assessing cartilage morphology in osteoarthritis [53] and holds potential for the assessment of cartilage abnormalities in haemophilic arthropathy.
Driven equilibrium Fourier transform and fluctuating equilibrium Driven equilibrium Fourier transform (DEFT) and fluctuating equilibrium MRI (FEMR) are techniques that depend on the ratio of T1/T2 in a given tissue [54,55]. These techniques produce contrast by enhancing the signal from synovial fluid rather than attenuation of cartilage signal as in T2-weighted sequences. DEFT and FEMR show much greater cartilage to fluid contrast than spoiled gradient-recalled (SPGR), proton-density spin-echo (PD-SE) or T2-weighted fast-spin-echo (FSE) sequences [56].
Sodium MRI This technique is based on the ability of sodium imaging to depict regions of proteoglycan depletion [57]. High sodium concentration is seen throughout the normal cartilage. This method shows promise in being sensitive to early decreases in proteoglycan concentration in arthritis.
T1 mapping Gadolinium-DTPA-enhanced T1 imaging is also a technique sensitive to the cartilage proteoglycan content [58]. In this technique, the negative charge of the paramagnetic contrast agent distributes into the cartilage inversely to the fixed charge density of glycosaminoglycans [58]. In presence of the contrast agent, T1 relaxation time is approximately linearly related to the glycosaminoglycan content. This technique holds potential for detecting early cartilaginous changes prior to macroscopic visualization of cartilaginous damage in haemophilic joints through conventional imaging.
T2 mapping Alteration in the orderly transition in T2 values within cartilage has been shown to correlate to changes in water content and changes in collagen structure and organization associated with hyaline articular cartilage degradation [59,60]. This technique could serve as a proxy of collagen organization in the articular cartilage in haemophilic joints.
Ultrasound biomicroscopy
High-frequency probes (20–50 MHz) allow the evaluation of hyaline cartilage, intra-articular fibrocartilages and ligaments [61] and cartilaginous changes undetectable macroscopically in rheumatoid arthritis [62]. The development of intra-articular high-resolution probes for ultrasound biomicrsocopy may be useful in the intra-operative procedure of synovectomy of haemophilic joints, demonstrating in real-time the microscopic status of the articular cartilage.
Acknowledgement
Dr Doria is a recipient of a Career Development Award from the Canadian Child Health Clinician-Scientist Program.
Disclosures
The authors stated that they had no interests which might be perceived as posing a conflict or bias.
