1. Introduction
Hemophilia A and B are X-linked congenital bleeding disorders caused by the absence or dysfunction of clotting factor VIII (FVIII) or factor IX (FIX), respectively. The degree of clotting factor deficiency influences the phenotype and spontaneous bleedings occurring into joints and muscles represent the most common clinical manifestation, mainly in the severe form of the disease (FVIII/FI < 1 IU/dL) [
1].
The predilection for bleeding into large synovial joints is probably a consequence of the rich vascularization of synovial tissue and its exposure to intensive mechanical forces in combination with a shifted hemostatic balance [
2,
3]. Although the incidence of joint bleeding has been significantly reduced over the last 40 years through the extensive use of prophylaxis and clotting factor replacement, patients with hemophilia are still at risk for joint dysfunction due to bleeding.
Recurrent hemarthroses lead to specific changes in the synovium and cartilage, which finally result in destruction of the joint. This process is called hemophilic arthropathy (HA) [
4]. The HA is associated with chronic pain and limited daily functioning that have a big impact on the quality of life in hemophilic patients [
5].
Pathogenetic mechanism of HA is multifactorial and includes degenerative cartilage-mediated and inflammatory synovium-mediated components. Intra-articular blood leads to the formation of iron-catalyzed destructive oxygen metabolites resulting in chondrocyte apoptosis. This process has a negative effect on cartilage, subsequently the intra-articular blood affects the synovium, in addition to hemosiderin-induced synovial triggering. Intra-articular blood derived hemosiderin deposits are thought to be critical in the early phase of HA through the induction of a proliferative disorder with chronic synovitis, which then extends to the articular cartilage surface and ultimately resulting in destructive arthropathy. Both processes occur simultaneously, and while they influence each other, they probably do not depend on each other [
4,
6]. This model resembles the degenerative joint damage found in osteoarthritis (OA) as well as the inflammatory processes in rheumatoid arthritis (RA) [
7].
The usual diagnostic methods of HA include plain radiography, ultrasonography (US) and magnetic resonance imaging (MRI) [
8]. Developing diagnostic tools to identify patients at high risk of early appearance or progression of joint damage remains an important yet challenging aspect. Therefore, there is a big effort to identify a tool reflecting early dynamical changes in the joint. This tool may be represented by biochemical markers (in serum, urine or synovial fluid). Biomarkers are defined as objective indicators of normal biologic processes, pathogenic processes or pharmacologic responses to therapeutic interventions or to an exposure, and have the potential to decrease the length and cost of trials and enrich our understanding of the pathogenesis of a disease [
9]. An appropriate biomarker should, therefore, document disease progression or disease activity and should assess any effects of therapeutic intervention [
10]. Biomarkers are useful to assess joint tissue turnover, they can reflect the pathological processes due to a joint bleed. Ideally, these biochemical markers should monitor joint status more closely and more accurately than radiography, US and MRI, especially in early phases. In osteoarthritis and rheumatoid arthritis, biomarkers related to the degradation of cartilage, bone and synovial tissue have successfully been studied to assess the severity and progression of joint damage, and to control the effects of treatments [
11,
12]. Although HA resembles characteristics of both OA and RA, few data on biomarkers in this clinical setting are available.
The aim of this review is to summarize and categorize publications about biomarkers in HA focusing on biochemical markers.
The goal of this review is also to clarify if the biomarkers can be used in the clinical practice to monitor bleeding, to assess the progression of the HA, and to evaluate the effectiveness of the treatments.
3. Discussion
Early joint damage in patients with hemophilia often escapes diagnosis because of insufficient investigation of biomechanical changes. Biomechanical reactions of muscles and ligaments around the joint have been less investigated, although similarly important, than the biochemical reactions linked to blood-induced inflammation and damage of joint tissues. Furthermore, the biomechanical reactions are the first to occur immediately after bleeding, as muscles react with changes in contraction pattern and ligaments overloading as demonstrated by the gait analysis [
21] and surface electromyography (sEMG) [
25].
As function impairment usually precedes structural damage [
25], it is of major importance to include more sensitive measurement tools for a more accurate musculoskeletal examination of patients with hemophilia. The early diagnosis may facilitate the initiation of appropriate joint-based therapeutic concepts and should trigger the implementation of training programs in order to preserve hemophilic joints from fast deteriorating. Moreover, the motion analysis by means of sEMG and/or gait analysis can be also useful to evaluate the patient’s response to treatment.
The subtle articular changes of the subclinical disease can be detected by diagnostic imaging techniques. The detection of early signs of osteochondral damage is often difficult. Moreover osteochondral damage can be present in asymptomatic patients in which none or just a few bleeding episodes were previously recognized [
30].
Conventional radiography has been successfully used for decades to objectively evaluate HA. Findings of HA on plain radiographs include osteoporosis, osteonecrosis, epiphyseal overgrowth, widening of the intercondylar notch of the distal femur, bone cysts, joint space irregularity and narrowing, angulation of the knee and ankle and bony fusion. However, all these signs represent late arthropathic changes, most notably subchondral and bony abnormalities, and are not useful for an early diagnosis of arthropathy. All these abnormalities are included into the main classification systems, the Arnold–Hilgartner scale and the Pettersson score [
49,
59].
MRI is the gold standard imaging technique to evaluate the musculoskeletal system because of its excellent spatial and contrast resolution. MRI is the most complete imaging technique and the most sensitive for the diagnosis of musculoskeletal complications of hemophilia; this image technique has shown its interest to detect early signs in joint alteration and it has been shown to be more sensitive in detecting the first signs of HA than both clinical examination and plain radiography [
78].
The signal of blood products on MRI varies according to the sequence used. Susceptibility-sensitive T2n GRE techniques result in enhanced visibility of blood products in the acute stage (deoxyhemoglobin) and the chronic stage (hemosiderin).
MRI imaging is a sensitive technique to visualize hemosiderin deposition in a joint, especially using T2* GRE sequences [
58,
59].
If susceptibility artifact from hemosiderin limits interpretation, GRE sequences replaced with T2-weighted tSE sequences is a suitable alternative [
79].
On proton density sequences or 3D spoiled gradient echo (GRE), chronic synovial proliferation is characterized by intermediate intensity signal on T1 and T2-weighted sequences presenting a level of contrast between cartilage and fluid [
58]. However, in the active synovitis phase, the MRI signal intensity of the proliferating synovium may increase becoming similar to the fluid one so that the distinction with effusion could be difficult [
80].
The use of intravenous contrast media may theoretically help distinguish active synovitis from fibrotic synovium [
43]. However, the presence of hemosiderin and synovium within the proliferative synovium in HA limits the degree of visible enhancement; for these reasons intravenous contrast medium is not routinely recommended in the evaluation of HA [
42].
In terms of clinical relevance, detection of synovial hypertrophy, regardless of its degree of detectable vasculature, represents a sign of undertreatment, possibly related to an insufficient treatment regimen or a limited patient’s compliance [
41].
MRI is also the best imaging modality for the detailed evaluation of osteochondral derangements in HA, particularly those pertaining to the central aspect of the cartilage and subchondral bone evaluation [
40,
49,
54,
69,
70,
71].
Detailed imaging of the articular cartilage obtained with either a proton density fat-suppressed or volumetric GRE sequences may demonstrate focal and diffuse cartilage losses, whereas subchondral edema and cysts may be associated with high-intensity signal on fluid-sensitive sequences [
2].
MRI remains difficult to use in routine clinical practice because of its cost and limited accessibility [
42].
US has a low cost, short examination time and it is widely accessible. With the advent of the last generation equipment, by US it is now possible to depict small, superficial structures of the musculoskeletal system as present in the early stages of hemophilic arthropathy [
30].
US has proved able to detect synovial hyperemia, defined as intrasynovial detection of blood flow signals [
51,
52,
53]. However, intrasynovial hyperemia at Doppler imaging is uncommonly observed in hemophilic patients and, in the rare positive cases, only a few blood flow signals are visualized, suggesting mild hypervascularity that cannot be considered relevant enough to redirect treatment and patient management [
69]. In addition, high variability in the interpretation of Doppler images, the need for high-end machines to get better performance and high interequipment variability is expected [
41].
Regarding osteochondral surfaces, US cannot provide a comprehensive evaluation of the cartilage and subchondral bone like MRI can do.
Indeed, the weight-bearing areas, cannot be clearly assessed due to the problem of the access of the US beam [
48,
49,
54].
In regard to the articular cartilage, US is able to detect the full spectrum of abnormalities, from subtle echo textural changes or partial thickness losses through extensive cartilage disappearance. In children, coexisting damage of the epiphyseal cartilage can be recognized [
2].
Although clinical examination and radiological exams (radiographs, US and MRI) remain the gold standard for the diagnosis of hemophilic arthropathy and hemarthrosis in patients with hemophilia, serological biomarkers of bleeding and joint degeneration can be an attractive complement to the diagnostic imaging techniques.
In our review we pointed out that, at the present time, biomarkers are interesting scientifically intriguing, but they are not standardized for the use in clinical practice. Attempts to investigate the role of the biomarkers in the diagnosis of hemophilic arthropathy has yielded conflicting results, because of different inclusion or exclusion criteria adopted by the different studies and/or classification of HA.
Biomarkers for monitoring articular bleeding and consecutive progressive cartilage destruction in hemophilic patients would be a useful tool for the clinician, but further studies are needed to confirm and develop the present knowledge on existing biomarkers and hopefully identify new, more specific ones.
Even though the use of biomarkers such as miRNA in the screening or follow up of this condition is still limited to the animal model studies, they seem to lay a promising basis for future use in the clinic.
The inflammation and angiogenesis process in HA were investigated in several studies and can be classified as diagnostic/prognostic biomarkers. VEGF is the principal signaling molecule in angiogenesis and can be induced by hypoxia and by certain cytokines through interaction with its receptors. Previously research has demonstrated that the synovial reaction in other joint diseases—that share histologic similarities with HA—enhances oxygen demand and shows evidence of de novo blood vessel formation, including endothelialization of the synovium. Importantly, proliferating synovium can secrete chemocytokines, such as VEGF, that might promote recruitment of endothelial cells to sites of active angiogenesis [
81]. Building on this knowledge the changes in VEGF values have been analyzed in several studies with contradictive results. VEGF was found to be increased in HA patients with early joint disease, but other studies obtained contrasting results; SDF-1a and MMP-9 were also investigated and found to be increased in HA patient with joint disease; a correlation between VCAM-1 and severe arthropathy at X-ray was identified in a study and other ones found an association of CRP and MIF with acute joint bleeding in HA patients. Considering the contrasting evidence regarding the sensitivity and specificity of the biomarkers, the detection of arthropathy in HA patients still relies on imaging techniques: for instance, MRI remains the gold standard for the detection of synovitis in HA patients.
The cartilage turnover markers can be classified as diagnostic biomarkers markers. Type II collagen is the major component of the articular cartilage matrix and is degraded by proteolytic enzymes secreted by chondrocytes and synoviocytes [
81]. Several markers of joint tissue turnover and formation were investigated with contradictive results. Some studies have found a correlation between the increase of collagen turnover markers and the increase of JSN and PS, unfortunately another study did not confirm this correlation [
34,
45,
60,
61,
62].
The different results may be explained by heterogeneous assessment for HA or by different populations; for all these reasons comparison between studies is often difficult.
When it comes to bone turnover, results on the use of CTX-I, serum osteocalcin, serum osteoprotegerin and RANK-L are quite heterogeneous, while there seem to be a correlation between the level of serum TRAP-5 and serum sclerostin with the severity of the HA. On the other hand, serum DKK-1, vitamin D and b-ALP were significantly correlated with physical exam findings, but not radiological ones. However, further studies are needed to support their use in the clinical practice.
4. Materials and Methods
A review of the literature was performed on two medical electronic databases, PubMed/MEDLINE and Embase, from 3 to 7 August 2020. The study selection and the data extraction were achieved independently by two authors, meanwhile the senior investigators revise the work. Inclusion criteria were determined a priori:
all studies were written in English;
all studies have available full text;
all studies were published in peer-reviewed journals.
We excluded studies not reporting data concerning hemophilic disease.
Twenty (20) articles not meeting the inclusion criteria were identified by checking the bibliography of relevant articles and searching for the studies cited in all the articles examined.
Eligible studies obtained at the end of the search and screen process were seventy-three (73).
The search string in PubMed/MEDLINE for biochemical biomarkers was:
The search produced forty-six (46) articles.
Two different Embase searches for biochemical biomarkers were performed using two different strings as follows:
The search produced eighty-three (83) articles.
Meanwhile the second search produced two hundred eleven (211) articles.
The two different searches in Embase had in common fifty-eight (58) articles and all three different searches (PubMed/MEDLINE and Embase) had in common thirteen (13) articles.
Finally, fifteen (15) articles were selected.
Regarding the evaluation of imaging techniques, such as radiographs, ultrasound and MRI, a full text articles research on PubMed/MEDLINE database, considering the last 5 years, was performed using the string as follow:
The research produced one hundred and twenty-four (124) articles.
In Embase a search was performed using the follow string, limiting the research to the studies published in the last 5 years:
The research produced seventy-two (72) articles.
Seventeen (17) articles were shared between Embase and PubMed research.
The articles selected were twenty-two (22).
The research for the motion analysis was performed in PubMed/MEDLINE and in Embase selecting full text articles in the last 5 years. The PubMed string was:
and the Embase string was:
The research in PubMed/MEDLINE produced nine (9) articles and the research in Embase thirty-eight (38) articles. The two databases had in common seven (7) articles and for this review after reading the texts were selected sixteen (17) articles.
Unfortunately, a meta-analysis was not performed because the studies showed variability in test measures.