Drs. Zoli and Bosello contributed equally to this work.
Atlantoepistrophic magnetic resonance imaging involvement in early rheumatoid arthritis: An aggressive tight control therapy not fully arresting the disease
Article first published online: 27 OCT 2011
Copyright © 2011 by the American College of Rheumatology
Arthritis Care & Research
Volume 63, Issue 11, pages 1629–1633, November 2011
How to Cite
Zoli, A., Bosello, S., Magarelli, N., D'Antona, G., Amelia, R., Fedele, A., Peluso, G., Bonomo, L. and Ferraccioli, G. (2011), Atlantoepistrophic magnetic resonance imaging involvement in early rheumatoid arthritis: An aggressive tight control therapy not fully arresting the disease. Arthritis Care Res, 63: 1629–1633. doi: 10.1002/acr.20573
- Issue published online: 27 OCT 2011
- Article first published online: 27 OCT 2011
- Accepted manuscript online: 27 SEP 2011 10:32AM EST
- Manuscript Accepted: 21 JUL 2011
- Manuscript Received: 15 JAN 2011
- ASRALES foundation
Rheumatoid arthritis (RA) is the most common inflammatory disorder affecting the cervical spine. The purpose of this study was to characterize the atloaxial involvement with magnetic resonance imaging (MRI) in patients with early RA at the moment of diagnosis and after 18 months of a tight control therapy.
Twenty consecutive patients with early RA without cervical symptoms and 20 healthy controls were enrolled. The patients underwent unenhanced and enhanced gadolinium MRI study of the upper cervical spine at diagnosis and after 18 months of therapy. The presence of pannus tissue at MRI was considered active synovitis.
Five (25%) of the 20 patients presented craniocervical involvement with active synovitis at MRI. At onset, patients with cervical involvement presented higher levels of erythrocyte sedimentation rate, a higher swollen joint count, and a higher Disease Activity Score in 44 joints level. All 5 patients (100%) with cervical involvement presented already peripheral erosions. After 18 months, 1 of 5 patients with atloepistrophic synovial involvement at baseline presented complete regression of the enhancement of synovial periodontoid process, and 1 presented a decrease of this enhancement. None of the patients developed erosive process at the odontoid. The only patient with complete regression of the enhancement presented a very early disease (<3 months).
Our study demonstrates involvement of the atloaxial junction in 25% of early RA patients, in particular in patients with active and erosive arthritis. An early diagnosis and aggressive treatment with a combination therapy, aiming for remission, does not always reduce atlantoaxial synovitis.
Upper cervical spine involvement, which occurs in approximately two-thirds of patients with longstanding rheumatoid arthritis (RA), is the most common inflammatory disorder of the cervical spine. Among the joints of the cervical spine, both atlantoaxial and subaxial joints may be involved and the erosive synovitis involving the atlantoaxial, atlanto-odontoid, and atlanto-occipital joints leads to instability, causing symptoms such as neck pain, stiffness, and radicular pain (1, 2).
Plain radiography, traditionally the first step in the radiologic evaluation of patients with RA, can detect only the bone erosions and joint space narrowing. Indeed, magnetic resonance imaging (MRI) demonstrates early synovitis. The extent of synovitis in early RA correlates with disease activity and is the best predictor of the development of erosions (1, 3).
Joint damage, which can result in joint destruction and bone erosions, begins early in the course of RA, and the patients with a longer disease duration do not respond as well to treatment compared with patients with early disease. The existence of a window of therapeutic opportunity in the earliest phase of RA has now received support (4); thus, it is necessary to identify synovitis and joint damage in RA patients as early as possible to allow early treatment and induce long-term remission. No data, to our knowledge, are available on the MRI atloepistrophic involvement during a tight control step-up treatment approach in early RA. The aim of this study was to characterize the atloaxial involvement with MRI of the cervical spine in patients with early RA before the development of symptoms and during a followup period when patients were treated with an aggressive and tight control treatment approach, aiming for remission.
Patients and Methods
This study included 20 consecutive patients (3 men and 17 women) with early RA (disease duration <12 months), fulfilling the 2010 American College of Rheumatology criteria for RA (5), who were attending our early arthritis clinic. All of the patients gave their informed consent to enter into the study, which was approved by our ethics institutional committee. The mean ± SD age was 54.8 ± 12.6 years and the mean ± SD disease duration was 6.5 ± 3.0 months. When patients presented a disease duration of less than 3 months, they were defined as having “very early RA.” In our cohort, 2 patients had very early RA.
None of the patients had a history of previous trauma or surgery of the craniocervical region, and none presented symptoms of cervical involvement at baseline and during the study period. All of the patients showed active RA at onset (Disease Activity Score in 44 joints [DAS44] >3.7); the visits were performed at baseline and every 3 months until 6 months after the second MRI evaluations. At each visit, a detailed assessment was done to evaluate the “core data set” to define European League Against Rheumatism clinical response (6).
Thirteen patients (65%) were rheumatoid factor (RF) positive and 15 (75%) were anti–cyclic citrullinated peptide antibody (anti-CCP) positive.
At the first visit, when the diagnosis was confirmed, patients underwent cervical MRI and began methotrexate (MTX; up to 20 mg weekly) and a low steroid dose. Patients were evaluated every 3 months and a combination therapy with subcutaneous anti–tumor necrosis factor α (anti-TNFα; adalimumab 40 mg/every 2 weeks or etanercept 50 mg/week) was started after 3 months if the patients did not achieve low disease activity (DAS44 ≤2.4) with MTX. At the twelfth month, 12 patients (60.0%) were receiving monotherapy, while 8 patients (40.0%) were receiving an association with anti-TNFα.
Twenty healthy subjects, sex and age matched, were enrolled as controls and underwent cervical spine MRI.
MRI was performed with a 1.5 superconductive magnet system using a phased-array coil (Signa, GE Medical System), with the head and neck in a neutral position. The paramagnetic contrast medium was Multihance (Bracco), administered intravenously at 0.2 mmoles/kg. The protocol included fast spin-echo T1-weighted images on the sagittal plane and fast spin-echo T2-weighted images on axial and sagittal planes. Fast spoiled gradient-echo T1-weighted images with fat suppression on sagittal, axial, and coronal planes after contrast medium were used. The MRI scan allows assessment of the presence of synovitis and erosive joint damage. Thickening of the synovial membrane showing normal to low signal intensity in T1-weighted images, high signal intensity in T2-weighted images, and marked enhancement after contrast medium administration was considered active synovitis (7).
Two independent radiologists (NM and LB) who were blinded to clinical data and radiography findings interpreted all MRIs. Thereafter they solved all disagreements in consensus by joint reinterpretation of the images. All of the patients underwent an MRI study of the upper cervical spine at diagnosis and 17 of 20 patients after 18 months of therapy.
Radiologic and echographic periferical joints evaluation.
All 20 patients had hand and foot radiographs performed at time 0 and after 18 months of therapy. Radiographs were scored according to the Larsen method (8) considering only the erosion score, not the joint space narrowing score. Patients presenting a new erosion or the worsening of 1 or more points of the Larsen erosion score were considered radiologic progressors (9). Eleven patients (55%) presented erosions at baseline.
Nineteen patients underwent ultrasonography (US) with power Doppler (PD) evaluation of the peripheral joints as previously reported (10) within 3 months of the followup MRI.
All analyses were carried out using SPSS, version 15.0. Categorical variables were expressed as numbers, and quantitative variables as the mean ± SD. Non–normally distributed and ordinal data were analyzed using Mann-Whitney nonparametric tests or Wilcoxon's test for paired data. Categorical data were analyzed using chi-square tests with continuity correction, where necessary. A P value less than 0.05 was considered statistically significant.
At the time of diagnosis, 5 (25%) of 20 patients presented craniocervical involvement with active synovitis at MRI characterized by thickening of the synovial membrane with high signal intensity in T2-weighted images and marked enhancement after contrast medium. No patient presented erosions at the atlantoaxial joint or obliteration of the subarachnoid space and spinal cord compression. None of the healthy subjects presented an enhanced signal at the periodontoid process.
RA patients with cervical involvement presented higher levels of erythrocyte sedimentation rate, a higher C-reactive protein level, and a higher DAS44 level at onset compared to patients without cervical synovitis. At baseline, all 5 patients (100%) with cervical involvement presented erosions on plain radiography of the hands and feet with respect to 6 (40%) of 15 RA patients without cervical involvement. Three (60%) were positive for RF and 4 (80%) were positive for anti-CCP (Table 1).
|Early RA patients with upper cervical spine involvement (n = 5)||Early RA patients without upper cervical spine involvement (n = 15)||P|
|Age, mean ± SD years||54.0 ± 15.3||53.9 ± 12.6||NS|
|Disease duration, mean ± SD months||6.2 ± 2.5||6.6 ± 3.3||NS|
|ESR, mean ± SD mm/hour||74.8 ± 16.4||34.9 ± 25.4||0.005|
|44–swollen joint count, mean ± SD||14.8 ± 7.0||9.5 ± 5.2||0.002|
|44–tender joint count, mean ± SD||15.0 ± 6.6||9.0 ± 5.1||0.002|
|DAS44, mean ± SD||5.0 ± 0.8||3.9 ± 0.2||0.003|
|HAQ score, mean ± SD||1.2 ± 0.7||1.0 ± 0.7||NS|
|IgM-RF positive, no. (%)||3 (60)||10 (66.6)||NS|
|IgM-RF, mean ± SD units/ml||71.1 ± 144.4||35.6 ± 68.4||NS|
|IgA-RF positive, no. (%)||3 (60.0)||9 (60.0)||NS|
|IgA-RF, mean ± SD units/ml||46.5 ± 85.9||45.6 ± 100.4||NS|
|Anti-CCP positive, no. (%)||4 (80)||11 (73.3)||NS|
|Anti-CCP, mean ± SD units/ml||61.4 ± 57.8||48.6 ± 42.9||NS|
|Erosion at peripheral joints, no. (%)||5 (100)||6 (40)||NA|
All 5 patients with atloepistrophic synovial involvement at baseline repeated MRI after 18 months; 1 of these 5 patients presented complete regression of enhancement of the synovial periodontoid tissue (Figure 1) and 1 presented a decrease of this enhancement (Figure 2), while the other 3 did not show any modification at MRI. None of the patients developed erosions at the cervical site. Because of a persistent DAS44 >2.4 after 3 months of MTX therapy, 4 (80%) of these 5 patients underwent a combination therapy with subcutaneous anti-TNFα. All of the patients had reached a DAS44 <1.6 at the time of the second MRI evaluation. Of note, the only patient with complete regression of the active atloepistrophic synovitis presented very early disease.
No differences were found in clinical and immunologic features during the followup in patients with and without cervical improvement. All of the patients with no MRI involvement at baseline and followup remained negative, and only 4 (33.3%) of them needed combination therapy with anti-TNFα after 3 months because they had not reached low disease activity, with respect to 4 patients of 5 that needed anti-TNFα therapy in the group of patients with cervical involvement.
The Larsen erosion score remained stable in the 2 patients with decreased enhancement at the cervical spine and in 2 of 3 patients with stable cervical synovitis, while the score increased by 2 points in 1 patient with stable cervical involvement during followup. Two patients (13.3%) without cervical involvement at baseline presented a worsening in the Larsen erosion score, but their MRI at followup remained negative for synovitis.
Considering 3 patients with stable MRI cervical synovitis, during the subsequent 6 months of followup, 1 patient presented a relapse of the disease with PD positivity at US evaluation, 1 despite persistent clinical remission presented PD positivity at US in 1 wrist, and the other one was in clinical and US remission. Furthermore, the 2 patients with a decrease in synovitis at the cervical site maintained a DAS44 <1.6 without any sign of PD at peripheral US.
After the hands and feet, the cervical spine is the most common site of disease involvement in RA. According to the literature, the prevalence of cervical spine lesions of any kind among patients with longstanding RA ranged between 25% and 86%, although only a small percentage (between 7% and 34%) will develop severe neurologic symptoms requiring surgery (3, 11). Cervical involvement initially causes stiffness and later pain by the time it can lead to instability with neurologic symptoms. One-half of the RA patients who had cervical instability cannot have symptoms, and no single factor is entirely predictive of progression of cervical disease, although some serologic and clinical markers were associated with progressive cervical disease (3, 4).
Contrast-enhanced MRI allows the differentiation of hypervascularized, hypovascularized, and fibrous pannus, and this imaging can detect joint effusion and pannus tissue when the radiographic findings are negative. The advantage of MRI lies in the direct visualization of pannus tissue that may loosen or destroy the ligamentous structures such as the transverse ligament of atlas and the alar ligaments (12), leading to atlantoaxial subluxation.
Because joint damage and erosions may already be seen in patients within a few weeks from the onset of symptoms and cervical involvement can begin early in the disease process (13), we have studied with MRI the cervical involvement since the early phases of the disease.
In our study, MRI-detected synovitis in 25% of early RA patients with active disease at the moment of diagnosis, although none experienced pain in the cervical spine or suboccipital area, presented a decreased range of cervical motion or cervical instability. Our data confirm that involvement of the cervical spine begins early in RA and suggest the opportunity to monitor the cervical spine overall in patients with mild cervical symptoms and with an aggressive disease (3, 7).
Current interest focuses on identifying which patients with early RA will develop a progressive disease based on predictive factors. All early RA patients of our cohort with atlantoaxial synovitis presented higher levels of acute-phase proteins and DAS 44 and erosive arthritis, and 4 of 5 patients were anti-CCP positive at baseline. Consequently, elevated disease activity, high acute-phase proteins, and the presence of peripheral erosive arthritis are risk factors and possible predictors of atlantoepistrophic involvement in asymptomatic early RA patients. Recognizing these possible markers of cervical involvement can help to identify those patients with a more aggressive disease process and who need a stricter followup (7, 13–15).
To our knowledge, this is the first study to evaluate the efficacy of a tight control with step-up approach on cervical spine synovitis detected by MRI. Because previous studies suggest that the patients with refractory inflammation are still at risk of development of axial involvement, the rheumatologist should still keep in mind, when dealing with patients with erosive lesions in other joints and in patients with persistent high disease activity, to perform MRI of the cervical junction to evaluate the cervical synovitis that can be persistent despite effective therapy.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Ferraccioli had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Zoli, Bosello, Magarelli, Peluso, Ferraccioli.
Acquisition of data. Zoli, Bosello, Magarelli, D'Antona, Amelia, Fedele, Peluso.
Analysis and interpretation of data. Bosello, Magarelli, D'Antona, Bonomo.
- 11Prevalence and associated factors of anterior atlanto-axial luxation in a nation-wide sample of rheumatoid arthritis patients. Clin Exp Rheumatol 2004; 22: 427–32., , , , , , et al.